 

 

 

 

 

.x 1.. a.
.3 :d5x1|)1.3

.3. Lint-Etna...
J 5.15.1...»

L..- 1....
L... .1 :33: J a:

s .1. 3.71....»
e.i..£....!:. 2
1.....- .9»; ... .

.33... «1.13:5:
F:v!!9 3'71 7!.wlv- Y

1.1?!
rllil‘yllzr

tall
1: I 1:210:
1....

1.1.5.: .
:12...th
3.5.7.3.... . . .
nil) 1.92.1! ,
5.1.... I$:¢3I$0)D\DII1“1
v|\ .Ivan-itll

.1}: r a. L...
.3::....:...:..:.I.
.: {6!}. x...

 

 

 

 

 

7.5::

 

 

 

 

 

 

 

 

 

 

 

 

 

 

\-

5 I
.14).“...
L nit-.5: r
53.3.11...
&.

LL.

.1, a .1. .011.er a... ‘4: I...

35?... A)Iblu\...ltf‘ .2

(51..-va

A: Iirkiltl

‘l

I. li‘v‘;.3.l

it... b)... T... .I.¢\.L.r.’b..¢
iltx 133125,»... i..lul|

.1... 2:... Ll
: i. L 35.5

o“ ILW'ILI itb’
l.“ i!..lhnn.>r.r.

 

THESIS

/)

/
/
0’"

|||||IIIIIIIIIIIUIIIIIIIIIllIIIIIIIIIIIIIIHHI

31293 01563 585

This is to certify that the
dissertation entitled

ANALYSIS, EVALUATION, AND IMPROVEMENT OF FLAVOR
TRAITS IN LONG-DAY ONION GERMPLASM

presented by

Cheng Luo

has been accepted towards fulﬁllment
of the requirements for

 

 

Ph. D. degree in PBG'HRT

WEEWV

Major professor

Date May 29, 1997

 

MSU is an Afﬁrmative Action/Equal Opportunity Institution 0-12771

 

 

LIBRARY

Michigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

mm. we]

ILJ
:_l-
J IE] I
lLJELJ
C] I:
_II IE—E

[__

MSU Is An Affirmative Action/Equal Opportunity Institution
MWMS-M

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

ANALYSIS, EVALUATION, AND IMPROVEMENT OF FLAVOR TRAITS IN
LONG-DAY ONION GERMPLASM

By

Cheng Luo

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Plant Breeding and Genetics Program
Department of Horticulture

1997

ABSTRACT

ANALYSIS, EVALUATION, AND IMPROVEMENT OF FLAVOR TRAITS IN
LONG-DAY ONION GERMPLASM

By

Cheng Luo

Two important ﬂavor traits, sweetness and pungency were evaluated in
long-day, storage onions (Allium cepa. L). Sweetness was determined by
refractive index for soluble solids (SS) and by HPLC for sugars. Pungency was
determined by Schwimmer and Western procedure for pyruvic acid (PA)
concentrations in plant tissues. Initial experiments involved screening of 36
onion lines involving 7 inbreds, 6 parental single crosses, 7 MSU experimental
hybrids, 9 long-day and 7 short-day cultivars. Significant differences in the
percent of SS and sugar, and tissue concentration (umole per gram fresh tissue
weight) (umol.g") of enzymatically developed PA in the bulbs were detected
within the inbred populations and among the hybrids and cultivars. The SS level
increased about 2-3% and the PA concentration decreasing about 3-4 umolg“.
in the ﬁrst two months after storage in response to cold storage. lntraline and
interline crosses were made among the plants selected for various SS and PA
levels. Selection for maximum change for SS or PA was most effective from

high x high or low x low populations. Broad sense heritability and realized

heritability for SS were 84.5 % and 78.6 %, respectively. The levels of heritability
for PA were 67.8% and 69.8%. Heritability levels were estimated from the data
of the parental, F1, F2, and backcross populations over ﬁve years. Frequency
distribution for SS and PA, using the data from an experiment of two-way
crosses followed by backcrossing, conﬁrmed previous reports of additive

dominance of high SS over low SS and low PA over high PA.

ACKNOWLEDGMENTS

I would like to express my gratitude to my major professor, Dr. Lowell
Ewart for his consistent guidance and invaluable assistance in the design and
execution of my research project and for his insight during the preparation of this
thesis. I also sincerely thank my committee members for helping me to learn the
science and art of plant breeding and overcome the academic and cultural
difﬁculties I encountered here in the US. I am very grateful to Dr. Jack Kelly, Dr.
Bernard Zandstra and Dr. Randy Beaudry, for their helpful suggestions and
accurate criticisms in preparation of this manuscript. A special thanks is
extended to Dr. David Dilley and Dr. Randy Beaudry for helping and allowing me
to use the postharvest research lab for analysis of my experimental materials. I
am very thankful to Dr. Royal Heins for his help and generosity which enabled
me to process my research data with his computers. Finally, to my mom, dad
and sister, especially my wife, Yanni, and son, Shawn, without their lifetime love

and encouragement, I would not have come this far. Thank you!

TABLES OF CONTENTS

LIST OF TABLES ........................................................................... . vi
LIST OF FIGURES .......................................................................... viii
INTRODUCTION ............................................................................ 1
LITERATURE REVIEW .................................................................... 4

The Onion ............................................................................ 4

Breeding and Seed Production ................................................. 5

Onion Flavor Quality ............................................................... 7
MATERIALS AND METHODS ............................................................ 24
RESULTS AND DISCUSSION ........................................................... 46
GENERAL DISCUSSION .................................................................. 105
LIST OF REFERENCES ................................................................... 109

 

 

 

 

LIST OF TABLES

Table 1. SS (%) determined by the refractive index
method and PA (pmol.g") measured by
the SW method in onion tissues of MSU
experimental hybrid 2518-1 .............................................................. 48

 

 

Table 2. Soluble solids (%) in seven onion inbred lines ................................. 55 3

Table 3. PA levels in seven MSU onion inbred lines
harvested in 1992 ............................................................................. 57

Table 4. PA levels in nine onion inbred lines
harvested in 1993 ............................................................................. 58

Table 5. Total reducing sugars in MSU onion breeding
ﬁnes .................................................................................................. 63

Table 6. Analysis of sugars in onions from three
different sources (MSU experimental
hybrids, commercial storage onions and
sweet non-storage onions) by HPLC ................................................ 68

Table 7. Sugar content of onion inbreds and
backcross lines following two successive
selections of bulbs with and without regard
to sugar content ................................................................................ 71

Table 8. Pyruvate contents determined with the SW
method in onion inbreds and seed parent
populations derived from low pyruvic acid (PA)
bulb selections ........................................................ _ .......................... 76

VI

 

Table 9. Soluble solid contents (%) in three-way cross
onion hybrids produced from high SS selected
parent bulbs or from regular parental bulbs ...................................... 78

 

Table 10. Pyruvic acid (PA) contents (umol.g") in two
three-way cross onion hybrids produced from
parent bulbs with or without the selections for
low PA levels as determined by the SW
method ............................................................................................. 80

Table 11. Soluble solids (SS) content of onion progenies
derived from parental bulbs selected for
different SS levels determined from refractive
index of onion juice ........................................................................... 82

 

 

Table 12. Pyruvic Acid (PA) levels in onion progenies
derived from parental bulbs of different PA
levels determined with the SW method ............................................ 83

Table 13. Correlation coefﬁcients (r) of pyruvic acid
(PA), soluble solids (SS), and bulb sizes in
ﬁve onion varieties ................................................................ 86

Table 14. Broad-sense heritability (H2) estimated for
soluble solids (SS), pyruvic acid (PA)
concentrations and SS/PA ratio for four onion
populations ....................................................................................... 88

Table 15. Realized heritability (Hz) for soluble solids
(SS) (%) in ﬁve inbred onion fertile (B) lines
derived from selﬁng high SS or low SS
parental bulbs ................................................................................... 90

Table 16. Realized heritability (H?) for pyruvic acid (PA)
content (umol.g") in ﬁve inbred onion fertile

(B) lines derived from selﬁng high PA or low
PA parental bulbs ............................................................................. 91

Table 17. Variation in SS, PA and SS:PA ratio in MSU
experimental hybrid 2518-1 over ﬁve years ...................................... 98

Table 18. Pyruvic acid content (umol.g".) of onion bulbs
from inbred parents by using the MSU color
value procedure .............................................................................. 103

VII

 

Figure 1.

Figure 2.

Figure 3.

Figure 4.

Figure 5.

Figure 6.

Figure 7.

Figure 8.

Figure 9.

Figure 10.

LIST OF FIGURES
Hydrolysis of ﬂavor precursor in onion .............................................. 15

Formation of volatile sulfur compounds in
onion after hydrolysis of ﬂavor precursors.................... ..................... 15

The effect of storage time on SS content (%)in
six different onion germplasms .......................................................... 50

Pyruvic acid (PA) levels over time (5 months)
of storage for six different onion germplasms,
using the SW method ........................................................................ 51

Sugar percentage (g.sugar/g.dry tissue) and
PA (pmoI/g.) concentration in storable onion
hybrids and sweet onions .................................................................. 61

Regression of percent reducing sugars

measured by the Hodge and Hofreiter (1962)

method on percent SS as measured from the

juice refractive index of 9885 B and 5718 B

onions ................................................................................................ 65

Sugar content of two parental onion lines and
hybrid progeny onions derived from high
soluble solids bulb selection .............................................................. 74

Frequency distribution of percent SS content
of the onion parents and progenies of 5718 B
x 826 B .............................................................................................. 94

Frequency distribution of PA concentration of
the parents and progenies of 5718 B x 826 B ................................... 95

Relationship between the MSU color value
procedure and the standard method
(Schiwmmer and Weston, 1961) ..................................................... 100

VIII

 

 

 

INTRODUCTION

Bulb onions (Allium cepa L.) are of major economic and dietary
importance in the US. and many parts of the world. The US. production of bulb
onions during the past decade has increased as much as four times of that in the
19708 (Love, 1994). It is predicted that the production and value of the US.
onions will continue to increase and reach $1 billion by 2000, compared to the
$666 million averages for 1991-1993 (Love, 1994). The quality of onions in
terms of taste, shape, color, storability, and resistance to pathogen and insect
attack has been improved remarkably due to extensive new cultivar development

in the past decade (Pike, 1986).

Sweet onions, one of the fresh market onions, always have been in high
demand on the market, and the price of sweet onions is generally 25-50% higher
than that of storage onions (Smittle, 1988). Most of the production of sweet
onions, however, is found in the southern and western regions of the USA, but
the poor storability of sweet onions generally restricts them to short-term market
sales. They are generally available only during spring and summer. Limited

success has been realized with controlled atmosphere (CA) storage to extend

 

 

 

2

the sweet onion season (Smittle,1988). The major part of the US. storable
onion production is in the northern and northeastern regions of the USA and

these onions are usually available year-round.

In Michigan, onions ranked number one in value for fresh market
vegetable production at $16 million on 7,100 acres in 1994 ( Fedewa and
Pscodna, 1995). This value placed Michigan ninth in the nation for onion
production. Most of Michigan’s production is storage onions with remarkably
pungent ﬂavor. These onions often are referred to as long-day onions. The
Michigan onion producers have a long-standing desire to grow their own onions
with improved ﬂavor qualities to improve competition with onions produced in

other states.

Breeding for onion ﬂavor. quality is difﬁcult for several reasons. First,
ﬂavor quality is a quantitative trait, and therefore, it cannot be evaluated fully until
later generations. Secondly, the onion is a biennial plant which normally requires
two years to complete a life cycle, which adds a measure of inefﬁciency to a
breeding program. Thirdly, to fully evaluate ﬂavor quality, several component
traits need to be measured, which primarily include sugars and pungency levels.
However, the methods available for measuring these ﬂavor components are
generally complicated and inefﬁcient for breeding programs. Finally, storability of
onions appears to be somehow related to onion ﬂavor. Reduced pungency in

onions can lead to a substantial loss of the bulbs after a prolonged period of

 

 

 

3

storage (Owen, 1950). Lin and coworkers (1995) reported a negative correlation

between onion pungency and neck-rot disease index.

A major objective of this study was to develop a system which could be
used efﬁciently in breeding programs for ﬂavor evaluation and improvement of
storage onions, and to determine the change and stability of ﬂavor traits,
primarily sugar concentrations and pungency in onion parent and progeny
populations. Crosses between onions with determined ﬂavor levels were made
to study how selection and parent line combination could affect the ﬂavor traits in
subsequent generations. Another objective was to determine ﬂavor heritability
through genetic analysis based on parent and offspring sugar and pyruvic acid
levels. The last objective was to estimate the effects of yearly environment

variables on ﬂavor traits in the germplasm studied in this research.

 

 

 

LITERATURE REVIEW

The Onion

Bulb onion is classiﬁed taxonomically to the genus of Allium under the
family of Alliaceae. The Allium genus contains a large group of more than 600
species (Hanelt, 1986) including many economically important vegetables like
garlic, leek, and green bunching onions. Allium cepa L. is the botanical name for
bulb onion. This group also includes several species with the basic chromosome
number x=8, 2n=16, and are mostly diploid (Astly, 1989). The occasional
occurrence of individual tetraploid bulbs, however, has been reported (Vosa,

1976).

The bulb onion is one of the oldest cultivated vegetables in the world.
Archaeological discoveries date onion culture to at least 2800 BC. (Vavilov,
1951). Its primary center of origin was central Asia, with secondary centers in
the Near East and the Mediterranean region. The onion was one of the ﬁrst
cultivated plants taken to the Americas from Europe. Onion was introdused ﬁrst
to the Americas by Columbus, who took it to the Caribbean. Later, it was
imported and established in the northern regions of the US. in the early 17th

century (McCollum, 1976).

In the US. onions generally are classiﬁed as ‘short-day’ or ‘long-day’

onions in regard to bulbing. Short-day onions are produced in the southern

 

 

 

 

 

5

regions and long-day onions are produced in the northern regions of the US.
This is an inaccurate classiﬁcation because all onions are considered to be long-

day plants, the various types bulbing differently in response to day length.

The onion is grown widely and is economically important in the US. and
worldwide. The onion is the third most valuable commercial vegetable in the
US. ($589 million in 1991), following only tomato and lettuce (USDA, 1991).
Production of fresh market onions is concentrated in the southern and western
states, whereas storage onions are grown primarily in the northeastern and
midwestern regions. Love (1995) predicted that by' the year 2000, the US. onion
industry will produce a $1 billion crop, 50% more than the $666 million average
for 1991-93. Globally, the leading producers in 1987 were China (3.6 x 107 MT),

India (2.8 x 107 MT) and the USA (2.0 x 107 MT each) (Yamaguchi, 1983).

Breeding and Seed Production

Onion generally is regarded as an outcrossing species, although
estimates of the amount of natural outcrossing vary somewhat. Van der Meer
and Van Bennekom (1968) reported outcrossing rates in the ﬁeld in the range of

73 to 100%, whereas under glass the rate varied from 29 to 82%.

As an outcross-pollinating species, onion exhibits severe inbreeding
depression upon selﬁng. Selﬁng for several generations produces uniform lines,
but much vigor is lost in the early generations (Jones and Davis, 1944).

Because of the severe inbreeding depression, isogenic lines are uncommon.

 

 

 

6

Uniform lines for a speciﬁc trait can be achieved and vigor maintained by
recurrent selection, mass selection, sibbing, or backcross strategies (Barham,

1950; Clark, 1949; Pike, 1986).

Jones and Clark (1943) published a classical paper on the inheritance of
male sterility in onion. They demonstrated that male sterility is determined by the
interaction of cytoplasmic and nuclear factors. Plants with the sterile 'S'
cytoplasm and homozygous for the recessive ms gene were male sterile; plants

with any other cytoplasm and gene combination were fertile.

In a typical onion-breeding system, having obtained a number of selected
male—sterile and potential pollinator lines, a series of experimental F1 single
crosses are made among them and the progenies are evaluated subsequently.
The crosses generally are made from ﬂowering plants raised from mother bulbs
of the parents, and grown in insect-proof isolation cages. Pike (1986) provided a
detailed description of the type of cage and summarized the practical procedures

for using it under US. conditions.

Flies and bees commonly are used to aid in plant pollination (Van der
Meer, 1968). In onion-breeding cages, bees have been found to be better

pollinators than ﬂies.

The use of weak inbred parents may result in low seed yield, especially
when female plants have small seed heads (Campbell, 1968). Erickson and

Gabelman (1958) showed that inbred lines produced on average only about half

 

 

 

4—4

 

7

of the seed yield of hybrid crosses. A partial solution to this problem has been
achieved through development of three-way hybrids, whereby a vigorous male-
sterile F1 is used as the female parent with a third inbred as the ﬁeld pollen
parent. This system usually produces signiﬁcantly higher seed yield than a two-
way hybrid. According to Dowker (1989), it is doubtful that if a three-way hybrid

has the same potential for heterosis or uniformity as a two-way hybrid.
Onion Flavor Quality

Flavor is one of the most important quality components of onions.
People like onions because of their characterisic ﬂavor. Onion ﬂavor generally
can be described as 'sweet’ or ‘bitter’, 'mlld’ or ‘hot’, and 'mildly pungent’ or
‘strongly pungent’ (Schwimmer and Guadagni, 1962; Pal and Singh, 1988; _
Randle et al .,1993). These terms, however, can be misleading and confusing to
the public, and even to the onion bulb producers and breeders. For instance,
’sweet’ usually is perceived by people as a sugar-like taste, but a sweet type
onion is more likely considered to have a less pungent ﬂavor rather than a sugar-
like taste. In fact, in many onion ﬂavor articles, the concern is with pungency
rather than sweetness (Randle, 1992; Wall and Corgan, 1992). Many. onion
researchers have agreed that the characteristic of an onion taste is primarily
determined by sulfur compounds and modiﬁed by sugars (Plateniu and Knott,

1944; Randle et al., 1993).

 

 

 

 

Onion Sweetness

The main components of sweetness of onions are the nonstructural
carbohydrates including glucose, fructose, and sucrose together with a series of
oligosaccharide, the fructans (Suzuki and Cutcliffe, 1989). The nonstructural
carbohydrates also account for the major portion of the dry weight. They range
from 41% to 88% of the total dry weight in various onion cultivars (Bajai et al.,

1980)

Bajai et al ., (1980) tested 12 onion cultivars for an assessment of sugar
contents. These cultivars included red and white onions that were thought to be
different in sugar content. They showed that the reducing sugar content of the 12
cultivars ranged from 12 to 22 g.100 g'1 in dry weight, while nonreducing sugars
varied between 25 to 62 g. 100 g". No differences in reducing sugar were noted

between red and white onions.

Starch, which is found commonly in many crops as one of the storable
carbohydrate forms, usually is not detectable in onion bulbs. Wilson et al.,
(1985) reported, however, that foliage of the plants, during growth, may contain

starch.

One detectable storage sugar form in onion is a series of fructosyl
polymers based on sucrose with varying degrees of polymerization (Bacon,

1957; Darbyshire and Henry, 1979). The fructans can range from 20% to about

 

 

 

 

9

50% of the total carbohydrate in the leaf base and later can be hydrolyzed to

give an increased fructose concentration in the bulb tissues (Bacon, 1957):

Heat-processed or boiled onions have a characteristically sweet taste.
The sweetness is not due to the level of sugars in the onions. Yamanishi and
Orioka (1955) found that n-propanethiol increased in concentration during

boiling. This compound is 50 to 70 times as sweet as sucrose.

Birth and Dull (1985) studied the composition of the low-dry-matter onion
‘Granex’ and the high-dry-matter onion ‘Creole’. They reported that more than
80% of the dry matter in these cultivars is made up of ﬁber and sugars. The
same information was provided by Darbyshire and Henry (1979) who also found

an association of soluble oligosaccharides with dry weight in onion.
Suqar Distribution

Sugar concentration per unit of fresh weight of an onion leaf is lowest at
the top and highest in the basal part. Steer and Darbyshire (1979) tested the
distribution of sucrose, fructose, glucose and trisaccharides in onion leaf tissue.
Their study showed that the basal part was about 6 mg/g higher in sugar content

than the top tissues.

Bacon (1957) examined the distribution of the lower molecular weight
fructans in onion bulbs and found these compounds to be absent from outer,

older leaf bases, and present in increasing amounts from the outer to the inner

 

 

 

10

leaf bases. Darbyshire and Henry (1978) reported that free fructose
concentration was highest in the outer leaf bases and lowest in the innermost.
They also observed that fructans and sucrose were high in the inner region, and

low in the outer region. Glucose did not vary much across the bulb.

Sugar Determination

 

1. Soluble solids
Mann and Hoyle (1945) studied the use of a refractometer as a means of

determining soluble solids (SS) and dry matter content of onion bulbs. In their

 

results, the refractive index of the onion juice was expressed as a percentage of
sucrose concentration. They found the sugar concentration and the percentage
dry matter closely correlated (r = 0.91). High-dry-matter onions can, therefore,
be selected using a refractometer, which is much faster than the oven-drying

determination.

2. Individual sugars and total sugar

High pressure liquid chromatography (HPLC) and gas chromatography
(GC) can be used for analysis of individual sugar components in plant tissue.
Darbyshire and Henry (1978) used HPLC to study the distribution of fructans in
onions. They successfully separated fructose, glucose and sucrose from the
extraction. A series of fructans with different degrees of polymerization also were
separated, but the concentrations were much lower than the other three sugars.
Darbyshire and Henry (1979) also later made a comparison among cultivars of

low, medium, and high percentage dry matter. The HPLC results showed that

 

 

 

11

low-dry-matter cultivars contain high levels of glucose, fructose, and sucrose and
only trace amount of fructans; whereas, medium-and high-dry-matter cultivars
contain lower glucose and fructose concentrations and substantial amounts of

fructans.

Genetic Factors Inﬂuencinq Onion Sugars

Because soluble carbohydrates account for most of the soluble or
dissolved solids, most of the published literature related to the genetics of onion
sugars deal with soluble. solids. An early publication about the heritability of
soluble solids in onions was made by Warid in 1952. He used a components of
variance method on parents and F2 generations to study the heritability of
soluble solids in onions. After crossing onions with high solids with onions with
low solids, he concluded that the heritability was 71% from parent-offspring
regression. He also estimated that there are 4-10 gene pairs for soluble solids,
and that partial dominance of low solids is involved. Later, Owen (1961)
examined the segregation of soluble solids genes, and he postulated an additive

gene effect in the inheritance of solids.

Many researchers (McCollum, 1968; Kadams and Nwasike, 1986; Lin et
al., 1995; and Simon, 1995) reported the heritabilities of solids in several onion
populations to be signiﬁcantly high (60-80%). McColIum (1968) also reported a

negative correlation between SS and bulb size.

12

Simon (1995) investigated the genetics of pungency and SS in long-day
onions, using an eight-inbred onion diallel for two years and a four-parent subset
of this diallel for one or two more years. He obtained a very high broad sense
heritability of 83% for SS. The generation means analysis of four crosses
indicated an acceptable ﬁt with a simple additive-dominance model to explain the

inheritance of the SS trait.

Environmental and Storage Factors

Randle (1992) reported that sulfur (S) nutrition affected the concentration
of total and individual sugars in short-day onion germplasm. On average, the
high-S treatment increased sugar content in most of the onions. Bulb dry weight
correlated negatively with bulb S concentration. Sucrose and fructose levels
changed differently in response to S supply, but no such change was found for

glucose.

Steer (1982) examined the effect of day/night temperature on the
accumulation of sugars by an onion cultivar. High temperature during both day
and night increased dry matter but had no effect on sugars. It was stated that
this could be due to the change of water content that simultaneously occurred
with the change of temperature. As a result of water changes, dry matter

concentration would ﬂuctuate accordingly.

Onion sugar concentrations tend to decrease with prolonged storage

(Smittle, 1988). Fractan hydrolysis and fructose increases may occur in onions

 

 

 

13

after a short period of storage (Darbyshire,1978). However, the ratio of sucrose,
glucose and fructose may not be affected by storage methods and duration
(Smittle, 1988). Under controlled atmosphere conditions, the decrease of onion
sugar concentrations tend to be slower than under cool temperature conditions

which commonly are used for onion storage.

Onion Punqencv

Lancaster and Boland (1990) summarized that there are over 80 ﬂavor
volatile compounds that have been identiﬁed from onions. Over 60 of them are
S-containing compounds. It is likely that other compounds also are involved, but

their detection and contribution has yet to be reported (Schwimmer, 1969).

Onion tissue has no pungent odor until the cells are injured (Lancaster et
al., 1989). Onion ﬂavors quickly develop after cell injury. The enzyme alliinase
starts hydrolysis of S-alk(en)yl cysteine sulfoxides which are the primary
precursors for production of S volatiles associated with ﬂavor and odor. Cultivars
that have high levels of the ﬂavor precursors and high alliinase activity have the
potential to produce strong pungent ﬂavor upon injuring of the cells. Soil and
environmental factors such as soil type, S content, temperature, and irrigation

levels also can affect substantially onion ﬂavor development.

1. The precursors
Onion ﬂavor develops from ﬂavor precursor compounds. Four distinct S

precursors exist in the genus Allium (Lancaster and Boland, 1989). They can

 

 

14

change from nonvolatile into volatile S compounds when the cells are injured,
and serve as the base materials that can cause Allium plants to have a strong

ﬂavor. The four S precursors are:
(1) (+) S-Methyl-L-cysteine sulfoxide
(2) (+) S-PropyI-L cysteine sulfoxide
(3) trans-(+)-S-(L-propeny|)-L-cysteine sulfoxide
(4) S-(2-Propenyl)-L-cystein sulfoxide.

An Allium species very often has a ﬂavor distinguishable from other
alliums because of the quantitative and qualitative difference in ﬂavor precursor
content. Freeman and Whenham (1976 a) measured these precursors in some
Allium species with gas liquid chromatography. The proportion of S-alk(en)yl
radicals (e.g., Methyl, Propyl, and Propenyl) in Allium species are different.
Onions usually contain S-methyl, S-propyl, and very high S-1-propenyl cystein
sulfoxides, but no S—(2-Propenyl) cystein sulfoxide, which is the dominant ﬂavor

precursor in garlic.

2. Formation of onion ﬂavor

The enzyme ultimately responsible for the development of onion ﬂavor
compounds after cell injury is alliinase, S-alk(en)yl-L—cysteine sulfoxide lyase.
Allinase catalyzes the reaction in which the S-alk(en)y| sulfoxide group is

eliminated from the substrate during hydrolysis. Two products, the lacrimator

 

 

15

identiﬁed as 1-propenyl sulfuric acid and the cysteine, are formed at the end of

this reaction:

0 (1) (4) NH (3)
T alliinase (2) ||
R-S-CH2-CH(NH2)COOH > R-SO: + CH3-C-COOH

 

 

Figure 1. Hydrolysis of ﬂavor precursor in onion (1) S-alk(en)yl cysteine sufoxide;
(2) S-alk(en)yl sulfenic acid; (3) cysteine; (4) S-alk(en)yl-L-cysteine sulfoxide
lyase.

 

The two products are very unstable. Sulfenic acids have a half-life of
about 90 seconds (Carson et al., 1966), and can undergo nonenzymatic
rearrangements to produce a wide range of volatiles that together give the
characteristic aroma to an onion. Cysteine can be broken down into ammonia
and pyruvic acid in onion homogenates (Lancaster and Boland, 1989). This

reaction is described commonly as:

O (1) (2) (3) (4)
’i alliinase
R-S-CHZ-CH(NH2)COOH + H2O --—-> R-SO +2 CH3COCOOH +2 NH3

Figure 2. Formation of volatile S compounds in onion after hydrolysis of ﬂavor
precursors. (1) S-alk(en)yl cysteine sufoxide; (2) S-alk(en)yl sulfenic acid; (3)
pyruvate; (4) ammonia.

 

16

When the lacrimator factor, thiopropanal s-oxide, rearranges propanal,
carbonyl derivatives of propanal and elemental S are produced. This was
proved by Freeman and Whenham (1976 b), who used a synthetic in vitro
system to demonstrate the importance of thiosuﬁnates as intermediates in the

formation of onion volatiles.

Alliinase is found not only in Allium cepa, but also in most members of
the genus Allium. Puriﬁcation of Allinase has been successful both in garlic and
onion by using standard precipitation and chromatographic separation methods

(Tobkin and Mazelis, 1979; Lancaster and Boland, 1989).

Besides the sweet and pungent taste, onion may develop a bitter,
alkaloid-like taste an hour after homogenization of the tissue. The role of the S-
alk(en)yl cysteine sulfoxides and alliinase in the development of this bitterness is
not clear, but the development of the bitterness requires an enzyme, since heat-
treated or acid-treated onions do not develop bitterness (Schwimmer, 1967).
The bitter taste appears at a slower rate than the onion odor and pungency.
This could be due to secondary reactions following the initial action of alliinase
on S-propenyl cystein sulfoxide (Schwimmer, 1969). The exact compound

responsible for the bitterness remains unknown.

Measurement of Pungencv

Schwimmer and Guadagni (1962) studied the method of estimating onion

ﬂavor by smell. They deﬁned a threshold concentration of onion juice in water

 

 

17

for a smell test. The threshold was the minimum concentration of onion juice
diluted with water that could be detected by 70% of the judges. Pungency
evaluation by smell alone, however, may not reﬂect the pungency that would be
perceived by taste accurately. Aromatic compounds may be perceived more
easily by taste than by smell, because samples are warmed by the mouth, and

aromatic constituents become more volatile (Jellinek, 1985).

The measure of pungency by a taste test is considered by some
researchers to be very inaccurate, because of the accumulative effect of
successive tasting (Platenius and Knott, 1941). Development of chemical
methods for quantiﬁcation and identiﬁcation of ﬂavor-related substances has

made it possible to express the result in numerical values.
Chemical methods of measuring onion ﬂavor

Sampling procedures to estimate ﬂavor potential in onion was studied by
Randle (1992). He suggested that a ﬁve-bulb sample and four replications were
sufﬁcient to detect desired differences for sugars and PA. He also indicated that

analysis of S in onions requires a larger sample size and more replications.

The pioneering work of Platenius (1935) on the determination of volatile
S compounds involved the application of acid hydrolysis of the onion at high

temperatures followed by analysis of S in a distillate.

 

 

 

18

A thin-layer chromatography method for evaluation and separation of the
ﬂavor compounds of onion juice was developed by Luke (1971). This method
can measure the relative concentration of other products involved in the action of

alliinase on the ﬂavor precursors.

Gas-chromatographic (GC) methods and GC-mass spectrometry have
been used successfully in separation and analysis of ﬂavor compounds in
onions. Freeman and Mossadeghi (1971) ﬁrst used GC methods to analyze
ﬂavor and odor compounds. They also indicated that this method may alter the
amounts of ﬂavor components due to the instability of 1-propenyl compound at

the high temperature employed.

By using supercritical carbon dioxide extraction, Sinha et al., (1992)
obtained ﬂavor components from onion tissues. GC analysis results showed the
presence of 28 S-containing compounds, including diallyl thiosulﬁnate, propyl
methanethiosulfonate, dithin derivatives, diallyl sulﬁde, diallyl trisulﬁde, and six

other compounds.

Bennet (1945) ﬁrst qualitatively detected the presence of relatively large
amounts of PA in onions. Morgan (1946) proved its presence by isolation PA
with 2,4-dinitrophenyl hydrazine (DNPH) from an non-heated macerate and

demonstrated that it arises enzymatically from precursors.

Relationship between PA and ﬂavor perception for onion pungency

determination has been studied by many researchers. Schwimmer and

 

 

 

 

 

19

Guadagni (1962) found a highly signiﬁcant correlation (r = 0.97) between the
amount of enzymatically developed PA in the onion juice and an olfactory
threshold concentration of the juice. Wall and Corgan (1992) found that PA
values were signiﬁcantly correlated with mean sensory ratings. Correlation
coefﬁcients (r) were 0.92, 0.84, 0.95, and 0.79 in four separate experiments. The
high correlation again indicates that PA analysis can be used as a reliable

selection technique for pungency in onion-breeding programs.

Schwimmer and Weston (1961) (SW) developed a relatively simple but
still somewhat time consuming method to determine the amount of PA. It has
become a classical procedure for pungency evaluation. They also reported that
onion tissues release PA quickly after the comminution process. Within a period
of six minutes after destruction of the cells, 95% of the maximum level of PA

developed.

Because of the stability and the analyzability of PA in the plant tissues,
many onion researchers use the SW method of PA analysis to measure onion
pungency. Some modiﬁcations on the procedure have been attempted to
reduce the analysis time. Randle and Bussard (1993) reported a new
streamlined pungency analysis that reduced the time of sample preparation from
74 minutes to 19 minutes. The key technical innovation in this system was a
speciﬁcally designed press that extracted onion juice from onion bulbs without

solid particles. Yoo et al., (1995) also reported a simpliﬁed PA procedure using

 

 

20

undiluted homogenate. They established a highly signiﬁcant correlation (r2

0.988) between the new method and the SW method.
Thiosulﬁnate Determination for Estimate of Pungency

An alternative method for the evaluation of pungency in onions involve
the determination of thiosulﬁnates (Carson and Wong, 1959; Nakata et a.
1970). The procedure involves derivatizing the thiosulﬁnates with I
ethylmaleimide and measuring the absorbency at the conjugate of 515 nn
Freeman and Whenham (1976) noted that other nonthiosulﬁnates also may giv
a positive reaction and interfere with the detection by adding values of detectabl

non-thiosulﬁnates to the result.
Factors Affecting Onion Flavor

Many factors are involved in the determination of onion ﬂavor. Generally
there are genetic factors existing in the onion, environmental factors occurrin
during onion production and storage, and soil conditions including soil type, sc

S content and soil moisture conditions.

1. Genetic factors of onion ﬂavor

The genotype of onion cultivars is one of the most inﬂuential factc
affecting onion ﬂavor intensity. Lancaster (1984) found that the proportion <
ﬂavor precursor varies among cultivars. Flavor levels of over 50 cultivars c

onions have been studied and summarized by Lancaster and Boland (1989

21

Some of the more pungent onions are high-dry-matter cultivars. Most of the mild
onions are the Early Grano-type and Japanese cultivars. There was a ten-fold

difference in the range of ﬂavor levels among these onions.

The genetic basis of pungency and sweetness of onions has been
studied by several researchers. Warid (1952) estimated that the heritability for
onion pungency was 71% from parent-offspring regression. McColIum (1968)
concluded the heritability of soluble solids in several onion populations to be
60% to 80%. Owen (1961), based upon several segregation populations,
reported that there are several genes involved in onion pungency. Lin et al.,
(1995) and Simon (1995) concluded that additive gene action was more
important than dominance for both pungency and soluble solids, based upon
their diallel experiment results. Generation means obtained from three crosses
made by Lin et al., (1995) indicated an acceptable ﬁt with a simple additive-
dominance model, and heritability of 40% for pyruvate and 80% for soluble

solids.

2. Soil, nutrition, and environmental factors

Soil and other ecological factors influence the pungency of onion.
Platenius (1941) demonstrated that onions grown on peat soils were more
pungent than those grown on sandy soils whereas those grown on loam or
sandy loam soils were intermediate in pungency. S deﬁciency in the culture

medium of onion plants grown in greenhouse was found to produce very low-

22

pungency bulbs (Freeman and Mossadeghi, 1973). Randle and Bussard (1993)
found that S levels interacted with cultivars in inﬂuencing bulb pungency and
individual sugars, except for fructose. Freeman and Mossadeghi (1973) also
suggested that the water content of the soil in which onions are grown has a
marked inﬂuence on onion ﬂavor. Their study on S nutrition indicated that an
abundant supply of water in the development stage tends to result in large bulbs

with relatively low pungency.

Onion ﬂavor is inﬂuenced strongly by high temperatures and water
supply during the growing season, especially in summer. Freeman and
Mossadeghi (1973) found that under dry conditions, onions tend to have higher

volatile S content and increased pungency levels.

3. Environment and cultivar interaction

Randle (1992) reported that location and cultivar difference exerted a
signiﬁcant effect on SS and PA in short-day onions in a comparison of green-
house grown onions with ﬁeld-grown onions. However, the interaction of location
and cultivar was not signiﬁcant for these two traits in short-day onions. This
interaction was found to be effective only in regard to S content in the onions

they studied.
Relationship between Soluble Solids and Pvruvic Acids

Platenius (1944) ﬁrst reported that total solids content reﬂects pungency

to some extent. Schwimmer and Guadagni (1962) found that there is a modest

23
correlation (r = 0.57) between SS and PA. Lin et al., (1995) conﬁrmed the

existence of a signiﬁcant correlation between the two traits in an 8-parent diallel
experiment. Simon (1995) reported that pungency and soluble solids are

correlated among parental inbreds and hybrids but not within F38.

Flavor and storability

It has been reported repeatedly that onion flavor may determine onion
storability to some extent. Earlier investigators (Owen, 1950; Schwimmer and
Guadagni,1962) reported that pungent cultivars were more resistant to neck - rot
disease, and they indicated that pungency and soluble solids were positively
correlated. Foskett and Peterson (1949) found that mild ﬂavor, poor keeping
quality and low dry matter were associated. Recent studies by Lin et al., (1995)
found a negative correlation between PA and a neck— rot disease index. They
also pointed out that this correlation between pungency and disease was not

high in strength and varied among cultivars.

MATERIALS AND METHODS

Plant Materials
Bulb Production

The onion bulbs were produced each year in the ﬁeld at the Muck Farm of
the Crop and Soil Sciences Department, Michigan State University, Lanigsburg,
Michigan. Bulbs for sugar and pungency analysis was produced each year from
1991 to 1995. Seeds of the onions were planted in May of each year in ﬁeld
blocks. In each block, the seeds were sown in plot rows. Each plot was 10 m
long and 1.3 m wide and had three rows. The plant density was about 225,000
seeds per acre, a typical density used in commercial onion production in
Michigan. Two guard rows at the edges of each block were planted using onion
hybrids. Each row was sown with 2 g of seed by a hand-operated seed planter.
In 1991, 1992 and 1993, Lorsban, a pesticide, was used during the seed planting

to control onion maggot.

The soil type of the Muck Farm is a dark-brown peat. Fertilizer, irrigation
and herbicide sprays were applied as needed. Field-roguing was conducted to
eliminate suspect plants other than the planted genotype based on plant

morphology.

Onions were harvested in late September or early October each year. All

the bulbs were packed and labeled in onion sacks and transported for storage at

24

25
the Horticulture Teaching and Research Center (HTRC), MSU. Freshly

harvested onions were allowed to remain at room temperature for curing for
about ﬁfty days prior to storage at 0-2 C. The onions were stored until planting in

April of the following year.

Seed Production

From 1991 to 1995, selected onion bulbs with known SS and PA were
planted along with non-selected bulbs as the control in a sandy ﬁeld at HTRC in
early April of each year. Three types of isolation cages including small cages,
medium cages and large tent cages were used to provide the environment for
pollination control. Two male and two female bulbs or just four male bulbs were
placed together in each small cages (35.5 x 50 cm) which had cloth tubes
attached on the bottom. For production of somewhat large amounts of seed,
cages of 1 x 2 m (medium sized) were used. Each of the medium sized cages
enclosed a small plot of three to four rows. The pollination row was planted in
the center of the plot with the seed parents on both sides for crosses. Each row
had about ten plants. For a large quantity of seed, onion bulbs were planted in
large cages (2.8 x 6.7 m) with 1 or 2 male rows and 4 or 5 female rows (40

bulbs/row) for crosses or six rows of male plants for increase of line populations.

Irrigation and herbicides were applied as needed after planting. The
insect-proof isolation cages were set above each of the plots before ﬂowering.

Flies were introduced weekly into the small and medium cages and bee hives

26

were placed in the large cages for pollination. The time frame was usually the
last week of June through the third week of July. Roguing was conducted during

plant development and ﬂowering to eliminate suspect off-type plants.

Seed harvesting was performed whenever the seed heads were mature
usually the last week of August. The harvested seed heads were placed in a
drier at about 32-35 C for about 24 hours. After cleaning, the seeds were stored

in a 4 Cl40% RH-seed storage room.

Onion Germplasm for Improvement of SS and PA

The germplasm chosen for this study generally consisted of the inbred
lines which either have been commercially used for hybrid production or have
potentials for new cultivar development. The inbred lines, 5718, 8155 and 826
used in this experiment were the parents of the cultivar, Sweet Sandwich. MSU
inbreds, 9885 B, 9161 and K B were chosen for this study because they have
shown values as possible parents of new hybrids in experimental cases. The
cultivar, Sweet Sandwich, is considered to be a low pungent, storable onion.
Other commercial cultivars were used as comparison material as necessary and

depending on availability at the time of need.

27

Pollination

1. Self pollination
The onion bulbs of fertile plants were planted in pollination control cages.
Seeds from these plants were sown the following year for bulbs of the S1
generation. Selection of the S1 bulbs was made again for SS and PA levels in
the spring of the third year. These bulbs were grown again in isolation. Four bulb
groups were allowed to mass pollinate in the summer. Seeds from the S1 plants

were used to produce 82 bulbs in the fourth year.

2. lntraline crosses between sterile and fertile plants
Crossing between sterile (A line) plants and fertile (B line) plants within an
inbred line was performed by planting the selected bulbs of the A and B line
plants together in one isolation cage using bees or flies for pollination. Seeds
harvested from the A line plants were used to produce the male sterile bulbs of
the backcross (BCO) generation the following year. The following diagram is an

example of how a new inbred 5718 A was produced.

Two plants of 5718 A line X Two plants of 5718 B line
(high in SS or low in PA) (high in SS or low in PA)
I

5718 A-HS-LPA (HS and LPA stand for
high SS and low PA, respectively)

28

3. Interline crosses between fertile plants for development of new B lines
A desired trait from a fertile onion plant can be introduced into another line
by interline crossing followed by backcrossing. The B line development for an
onion inbred usually is carried out by crossing a selected B line with desired
characteristics to another B line from which a new fertile inbred line can be

developed. The general procedures are diagramed as the following.

Parent1 (P1) X Parent 2 (P2)
B line plant with a new trait B line plant with the trait to be
(high SS or low PA) ¢ improved (low SS or high PA)

Backcrossing0(BCo)generation X P2

I Backcrossing with P2 for 5

I generations (BC,... BC.)

I with selection for high SS or Low PA
New B line

In this study, selected bulbs of known SS and PA from two different B lines
Were planted under a controlled pollination environment in a greenhouse at
MSU. The umbels of the plants to be crossed were bagged when the ﬁrst ﬂower
c)lbened. Plants destined to be pollen parents were allowed to ﬂower.
Er\“iasculation of the B line seed parents was performed in the early morning and

early afternoon every day for a period of 2-3 weeks during ﬂowering. The

29

anthers of the seed parent ﬂowers were removed by hand before the flowers
opened. This was performed by using forceps sterilized with 75% alcohol to
avoid possible contamination. Pollen from the designated male plants was
transferred with a brush to the stigmas of the seed parent plants. The brush

was cleaned in a 75% alcohol solution and dried before the next use.

4. Backcrosses for development of new A lines or B lines with high SS and
low PA

Selected bulbs from a two-way crossed sterile line or a three-way crossed
sterile line were grown with one of the respective parent B lines (fertile) in the
small isolation cages for a backcross generation. Seed harvested from the A line
plants were sown the second year to produce the ﬁrst backcross generation
(301). The 801 bulbs were selected again for PA and SS and used to make the
Second backcross seed parents for BC2 generation. The following diagram
Shows an example of producing two new sterile lines, 5718 A and 826 B, from a

two-way crossed sterile line, 5718 A X 826 B.

Backcrossing Backcrossing
$973.8AX826 B) X 57188 (5718AX826 B) X _ 826B
PA $8 or low PA high SS or low PA high SS or low PA high SS or low
I I
5718 A-HS-LPA 826 B-HS-LPA

30

5. Two-way crosses and three-way crosses for hybrid seed parents

The seed for two-way cross parent combinations was obtained from the
sterile plants (A line) of inbred populations that had been selected and grown

with a pollen parent (B line) population of another inbred line in isolation.

The F1 bulbs derived from a two-way cross combination were selected for
SS and PA, and planted with the selected bulbs of another inbred line in
isolation, to make a three-way cross combination. Seeds from the A line (two-
way cross plants) were sown the following year to produce the three-way hybrid
onion bulbs. This three-way crossing system is the same as used for the

production of commercial hybrids (Pike, 1986).

Two-way cross (sterile) x inbred (fertile)
I

three-way hybrid (sterile)

Laboratory Analysis

Onion Bulb Samplinq

For parental bulb selection for sugars and pungency, the bulbs to be
tested were cut into halves. The bottom half of each bulb was saved and labeled
for later planting in the ﬁeld. Only the top half was used for laboratory testing.

After cutting, the bottom half cut surface was put into white sand to form a

31

protective cover. This layer of sand on the onion cut was used to help seal the
cut surface, and to reduce moisture loss and pathogen development. Then, the

bottom half of the bulb was stored in a refrigerator at -2 C before planting.

To determine the mean pungency and sugar levels in a particular
genotype, a random sample of ﬁve to eight bulbs was used for a PA test and of

six to ten bulbs for sugar analysis.

To determine the transverse distribution of sugars and PA in onion scale
layers, scale pieces were cut from two outermost scales, from the two center
scales, and from two middle scales. To determine the longitudinal distribution of
sugars and PA, slices were taken from the top, the middle and the base of the

bulbs.
SS Test

To estimate the total sugar level in a bulb, onion juice from the bulb was
used to measure SS. The sample was taken from the inner and middle scales of
the bulb. The tissues were sliced and placed Into a garlic press, and drops of the

onion juice were collected in a 25-ml glass beaker.

Refractometer readings of percent SS were taken individually on each of
the bulb samples by applying 1-2 drops of the collected juice on the platform of
the refractometer. The SS of each bulb was read under artiﬁcial light. Each

sample was measured twice, and if the two readings exceeded 13%, the sample

32

was checked again. The platform and the glass cover of the refractormeter was
washed with distilled water and wiped-dry after each test in preparation for the

next sample.
Determination of Sugar Concentration

Results from the colorimetric chemical analysis of sugars may reﬂect the
true level of sugars in the plant tissues more accurately than the refractometer
result because it excludes the interference of non-sugar SS which may account
for up to 10% of the total SS. To determine the total sugar content of onion
bulbs, the colorimetric method (Hodge and Hofreiter, 1962) was applied and
modiﬁed in this study. Details of the modiﬁcation and procedure are given later

in this section.

The colorimetric method is based on the reaction in which phenol in the
presence of sulfuric acid will react with the carbohydrates that have a reducing
group (-CHOH) on their molecular structures. This reaction leads to the
formation of color products which make it possible for quantitative colorimetric
determination of sugars. The method is simple, rapid and sensitive. The color

produced is stable.

A wedge of the bulb was cut off longitudinally after the outer, dry scales
were removed. A 20-g sample of tissue was diced into cubes (1 cm"), and
blended with 50 ml 80% ETOH in a test tube. The tube was placed immediately

in a hot water bath at a temperature of 80 C for 10 min. The extraction tube then

33

was placed in ice to cool the sample and then was centrifuged at 1.8 k Kg RCF
for 6 min. The pellets were extracted with another 50 ml 80% ETOH. After well
mixed, the solution in the tube was allowed to stand for 15 min then centrifuged
again. One ml of the supernatant then was diluted with 199 ml of double-distilled
water. Two ml of the dilution was pipetted into a colorimetric tube and 0.05 ml of
80% phenol was added, then 5 ml of concentrated sulfuric acid was added
rapidly. The stream of acid was directed against the liquid surface rather than
against the side of the tube to obtain good mixing. The tube was allowed to
stand for 10 min in a water bath at 30 C. The absorbency of yellow-orange color
of the sample was measured at 490 nm. The sugar content was determined in
reference to a standard curve constructed from a series of glucose

concentrations (0, 0.5, 1.0, 5 mg/ml).

All solutions were prepared in duplicate, except the standard solutions
which were in triplicate to minimize errors. Sugar content was calculated and
recorded in percentage (%) of the bulb fresh tissue weight (9 sugar I100 g fresh

tissue weight).

HPLC Analysis for Individual Suggrs

In 1993 and 1994, onion sugars were extracted and analyzed by HPLC.
Cold-stored onion bulbs were cut into slices and 20 g of samples from each bulb
were placed in a styrofoam container. Liquid nitrogen was poured immediately

into the container to freeze the tissues for about 2 min. A lyophilizing dryer then

34

was used to dry the samples for 92 h. The dried onion slices were ground and
forced through a 25-mesh screen. Then 100 mg from each sample was
extracted with 3.5 ml 80% ETOH. Then each sample was vortexed for about 15
sec and allowed to stand for 15 min, and then, were centrifuged 5 min at a
setting of 3000 rpm (=1879 kg RCF) on a Sorvall RT 6000 B tabletop centrifuge.
The pellets were extracted two more times using the same procedures as above.
The ﬁnal sugar extract (10.5ml) in each tube was diluted with 5 ml of double-
distilled H20. In some samples, if any chlorophyll content that may give green
color in the tissue sample was suspected, 5 ml of chloroform was added to the
tube containing the diluted extract followed by shaking and centrifuging the tube

to remove the chlorophyll.

The supernatant, which contained the extracted sugars, was dried
thoroughly in a Speed Vac AC 200. Then, 1 ml double-distilled water was added
to dissolve the dried sugars. Before being injected into a HPLC analyzer, 1 ml
of dissolved dry sugars was diluted with double distilled water to 1: 200 by

volume, and 20 ul from the dilution was used for the injection.

The HPLC used was a Dionex Series 4000 I, equipped with a carbopak
PA 1 Column (4x250 mm). Each sample was injected with a mixture of 65%
water and 35% 200 mM NaOH at a ﬂow rate of 1.0 ml/min. From 0.1 min to 1.5
min the column was eluted with 75% water, followed by 25% 200mM NaOH, and
then eluted for 7.5 min with 200 mM NaOH. The carbohydrates then were

detected by pulsed amperometry. Standardized 20 PPM of glucose, sucrose,

35

and fructose were injected before and after three injections of test samples.

Each sample was run twice to minimize error.

Pungencv ATnaIvsis

1. Schwimmer and Weston (1962) pyruvic acid analysis

Onion pungency was estimated chemically by using the SW method of PA
analysis. Some modiﬁcations were made in this study to ﬁt the requirement of
our breeding program. A comparison was made between the result of SW
method and the results of the new MSU-developed PA procedure. Onions from
same sources were tested for PA levels separately with the two systems in this

study.

In this study, the dry outer scales of the bulb were removed, and the bulb
was cut into small cubes. Tissue samples of 50 g were chopped into slices. The
50 9 samples were placed in a blender with 100 ml of distilled water and blended
for 40-60 sec, until no large chunky tissues were visible. After blending, the
samples were allowed to stand for 10 minutes to allow development of the PA.
Then 10 ml of the onion juice was placed in a 25-ml centrifuge tube. The
centrifuge was run at 1.8 k kg RCF for 5 min. Then 1 ml of the supernatant was

removed carefully and diluted with 4 ml of distilled water.

Background reactive carbonyls of the samples were checked by adding

about 100 ml of trichloracetic acid (25%) to 50 g bulb tissue before blending or

36

heat application, using a microwave on a maximum heating level (1.5 kw) for 5
min. These treatments were to deactivate alliianase so that enzymatically

developed PA would not appear in these treated samples.

After blending and centrifuging, the samples were taken from the
prepared supernatant dilution. One ml of the dilution from each sample was
placed into a test tube (50 ml capacity). One ml of 0.06% DNPH in 2 N HCL and
1 ml of distilled water then were added to the tube. The tubes were seated in a
water bath with a shaking device at 37 C for 10 min. Five ml of 0.6 N NaOH then
was pipetted into each sample tube and shaken for 3-5 sec. After the purple
color appeared in the tubes, the samples were measured with a
spectrophotometer with absorbency at 420 nm, using a reagent blank (2 ml

water + 1 ml DNPH + 5 ml NaOH) set at 0 absorbency.

The calibration curve was obtained using sodium pyruvate as the
standard. All samples were duplicated. Sodium pyruvate solutions were

triplicated.

This procedure was used to determine the section of the bulb best suited
for PA sampling (Table 1), to measure the average PA level of the onion lines,
and to compare the results obtained with the new MSU-developed pungency

procedure.

37

Develooment of a New Punqencv Procedure for Mass Screening of Onion

Populations

2. Pungency determination procedure

The SW PA analysis (1962) requires homogenization of the tissues and
spectrophotometric measurement for PA in the ﬁnal extracts. This method
involves substantial amounts of labor and equipment, and restricts the analysis
to expensive equipment and well trained personnel. It also destroys the sample
bulbs so that the selection for individuals becomes impossible. A new (MSU)
procedure was developed in the fall of 1995 which makes PA analysis much
simpler. This procedure was designed to be suitable for breeding programs in
which mass screening of populations for pungency is desired by breeders, using

simple equipment.

In this new procedure, the extraction of onion PA is performed by soaking
onion slices with distilled water in petri dishes. A test sample from each petri dish
is added to a depression on a porcelain plate and later treated with 2,4-
dinitrophenyl hydrazine(DNPH). Then after coloration of the PA-DNPH
derivatives, the samples are scored visually according to color intensity. The
color intensity of the samples reflected the PA levels in the tissues. Samples high
in PA result in a dark purple color, and samples low in PA give a yellow color.

lnterrnediate PA samples show color intensities of light to medium purple.

38

The MSU procedure was compared with the SW method by analysis of
the PA of ﬁve onion cultivars, made up of types considered to be sweet or
pungent (Fig. 11). Six bulbs from each of ﬁve hybrids (Vidalia, Texas 1015,
Sweet Sandwich, MSU2518-1 and Duration) were used in this analysis. Vidalia
and Texas 1015 were purchased from a local retail store. The other three
cultivars were out of the storage from MSU trials. The onion slices were cut
horizontally from the middle of the onion bulbs. The size of the slices was about
60 mm in diameter and 5 mm thick. After cutting, the slices were placed in Petri
dishes for 10 min to let PA develop. Twenty-ﬁve ml of distilled water then was
added to each of the dishes to extract PA from the slices. Two drops of the PA
sample from each dish were pipetted into a depression on a porcelain plate.
Two drops of 2,4-DNPH (0.006% in 2 N HCL) then were added to each
depression. The plate was kept at room temperature (about 28 C) and gently
tipped back and forth a few times during a 12-min time span. Finally, ﬁve drops
of NaOH (6 %) were added to each sample. Color intensities of samples were
estimated visually with four values: Yellow (very low PA) = 1, light purple (low

PA) = 2, purple (high PA) = 3, and dark purple (very high PA) = 4.

Samples from the same onion bulbs also were taken and analyzed with
the SW method to compare the two methods. The PA concentrations and color
values obtained by the two methods were analyzed statistically, using the SAS

general linear model (GLM) (SAS Institute, Cary, NC).

39

To check the background PA that was not involved in enzymatic
development of pungency, samples of the onion bulbs were wrapped in a plastic
bag, and then placed in a microwave oven (1.5 kw) for 5 min to stop enzymatic

activity. These samples then were tested for PA by the SW method.

Storage and Onion Flavor Quality

Changes in SS, PA, and Bulb Qualitv Related to Storage

Onion quality may dramatically decrease due to handling and storage
conditions. The bulbs generally can be stored through winter if the temperature
and humidity in storage are satisfactory. Onions may respond to storage
conditions differently according to their genetic make-up. The ﬂavor quality also
changes depending upon genetic and storage environmental factors. To
determine how Michigan onions respond to storage time and condition, one
experiment was designed and conducted to measure total sugar by the
colorimetric method, PA by the SW method, and storability of onion cultivars and

breeding material stored from October 1991 to May 1992.

In 1991, six onion samples, K B (line bred cultivar), 826 B and 5718 A (2
MSU breeding lines), 2518-1(an experimental MSU hybrid), 5718 x 8155 (2-way
experimental cross) and 'Sweet Sandwich' (hybrid cultivar) were collected from

the MSU breeding workshop planting. The onions were harvested in late

40

September and stored at room temperature (24 C) for two months. After that,
the onions were kept in boxes and stored under refrigeration at 0 C from
December, 1991 to May, 1992. Each onion line was sampled, with 10 bulbs for
total sugar testing and six bulbs for PA analysis during 1, 2, 3, 4 and 5 months’

storage. The results are showed in Fig. 3 and 4.

Genetic and Yearly Factors Inﬂuencing Onion Flavor

Crosses Amonq Parental Onions with Different SS and PA

Four cross types involving eight combinations were designed to study how

the parental difference in SS and PA inﬂuence the ﬂavor strengths of the

progeny.

Type 1. high x high: onions high in PA crossed with onions high in PA, or
onions high in SS crossed with onions high in SS;

Type 2. low x low: onions low in PA crossed with onions low in PA, or onions
low in SS crossed with onions low in SS;

Type 3. high x low: onions high in PA crossed with onions low in PA, or
onions high in SS crossed with onions low in SS;

Type 4. low x high: onions low in PA crossed with onions high in PA, or

onions low in SS crossed with onions high in SS.

Correlation Between SS. PA and Bulb Sizes
The correlation coefﬁcients between SS and PA, PA and bulb size, and
SS and bulb size were calculated from the data from ﬁve onion hybrids including

2518-1 a MSU experimental hybrid with a high yielding potential, single bulb

41

center and globe shape, 'Sweet Sandwich', which generally is considered to be a
storable and less pungent hybrid, 'Spartan Banner 80' and 'Norstar' both of which
are two cultivars with good storability and high pungency, and 3506 which is
another MSU experimental hybrid with a high onion oil content. To obtain bulb
size, the diameter of each bulb was measured in cm. Ten bulbs from each entry
were used to measure bulb size and test for SS and PA. The SAS general linear
model (GLM) (SAS Institute, Cary, NC.) in Microsoft Window was used to

process the data.

Heritabilitv Study

1. Broad-sense heritability

Interline crosses of fertile plants by fertile plants were made in 1991 and
1992. The F1 plants were selfed in 1993 and 1994. The F2 bulbs were tested
along with the parent onions for PA and SS and SS/PA ratios in 1994 and 1995.
Broad-sense heritability ( h2) was calculated by using the variances of the F2 8
and parents. The method was proposed ﬁrst by Mahmud and Krammer (1951)
to estimate broad-sense heritability. SS and PA measurement on F2 plants of
four single-cross populations and on ﬁve inbred lines, 5718 B, 8155 B, 826 B,
9161 B, and K B were conducted by taking four bulbs for PA and eight bulbs for
SS from each line.

The formula for estimating h2 was:

h2 = [(02F2 '525)/ 52F2 I X 100

42

where h2 = broad-sense heritability

<52F2 = total variance

0'25: environmental variance (non-genetic variance)
Since the environmental variance can be calculated from the variances of the
non-segregating populations (P,, P2, and F, ):
0'25: (0'2 P1 XO'ZPzX 021:1 )1/3:
GP,2 and on2 = variance of parents of single-cross populations;

therefore, h2 = [02,52 - (0%, x czpzx 021:1 )“3 ]/c52,=2 x 100

62,:1: variance of F,

2. Realized heritability
The realized heritability (hf) of SS or PA was determined by the amount of
genetic improvement that was realized by selection for high SS bulbs or low PA

bulbs within a population.

The formula used was hf = R/S (Falconer, 1981), where R is the
response realized by selection, which is calculated by subtracting the mean of
the S1 generation by the mean of parent generation, and S is the selection
differential, which is calculated by subtracting the mean of the S1 by the mean of
non-selection. The selection differential is the difference between the means of

the population from which they were selected.

Five new breeding lines (Table 15) were developed for low or high SS

through selﬁng. Three of the new selfed lines (8,) derived from two bulbs of

43
8155 B, 5718 B and 826 B were selected for high SS. Two selfed lines from

8155 B and 9161 B were selected for low SS.

Five breeding lines (Table 16) were developed for low PA or high PA
through selﬁng. Three of them (5718 B, 9161* B and 826 B) were selected for

low PA, the other two were selected for high PA.

Two to ﬁve selected bulbs from each breeding line were planted in one
small isolation cage to allow massed pollination. Bulbs produced from the seed
of each group were tested again for SS or PA. Realized heritability was

calculated using the formula above. The results are shown in Tables 15 and 16.

Frequency Distribution for PA and SS Contents in Parent and Proqenv

Populations

Crosses between fertile parent plants (P, and P2) of two different families
(5718 B x 9885 B and 5718 B x 826 B) provided F, seed to produce F2
populations and B, and 82 backcross populations. Bulbs of these six generations
(P,,P2, F,, F2, B, and 82) were harvested and tested for SS in ‘5718 B x 9885’ B
related populations and for PA in ‘5718 B x 826 8’ related populations. The
frequency distribution analyses for SS were made based on the data of ‘5718 B
x 9885 8’ related populations (Fig. 8). The distribution for PA was made from the

data of ‘5718 B x 826 B’ related populations (Fig.9).

 

44
The F1 plants of ‘5718 B x 9885 B’ from the seed harvested in 1991 were

selfed or backcrossed to the parents in 1993, and bulbs of F2 , the B1 and the
82 were analyzed for SS in 1994. The F1 plants of ‘5718 B x 826 B’ from the
seed harvested in 1992 were selfed or backcrossed to the respective parents in
1994. The bulbs harvested in 1995 were analyzed for PA. Relative frequency
was calculated by dividing the number of the bulbs with a determined SS value

or PA value by the total number of the sampled bulbs, and multiplying by 100.

Tests for SS and PA of Onions for Estimating Yearlv Variation

MSU experimental hybrid 2518-1 was used as a standard in SS and PA
analyses over ﬁve years from 1991 to 1995. Each year, ten and six bulbs from
this hybrid were taken for SS tests using a refractometer and PA analysis using
the SW method, respectively. Bulbs were tested for SS and PA in March and
April each year after winter storage, except the bulbs harvested in 1995 were
tested in December 1995. SS:PA ratios were calculated by dividing the mean of
SS (%) by the mean of PA (u.m|/g.f.w.). Least signiﬁcant differences between
different years for SS, PA and SS:PA ratio were calculated using t test at P 5

0.05.

Data Analysis

Analysis of variance was conducted to test the progenies after crosses.
The crossing combinations without selection for SS and PA were used as the

control for the respective progenies. A completely randomized experimental

 

45

design was used with two replications in the trials for evaluation of SS, sugar and
PA levels in developed onion germplasm. The number of replications for
laboratory tests, however, varied depending upon availability of seed and
survival of plants in the ﬁeld. Analysis of variance was performed to test the null
hypothesis of no difference in SS percentage, PA concentration and SS/ PA ratio
among parent and progeny families. The General Linear Model procedure of
SAS in Microsoft for Windows Program was used in regression analysis on
pooled data from test results of PA, SS, and bulb size to study the relationship

among these traits.

In lab tests for sugar content, SS concentration and PA level of an onion
population, 10 bulbs from each population were sampled randomly as ten
replications for SS evaluation, six bulbs were used as six replications for sugar
and PA analysis. Standard deviations of the mean were calculated to estimate
the variability in sugar production and pungency development of each

population.

 

RESULTS AND DISCUSSIONS

Distribution of Soluble Solids and Pyruvic Acid in Onions

The way onion bulbs are sampled affect the ﬁnal results of ﬂavor
analysis. It has been reported that different sections and scale layers of onion
bulbs have different levels of ﬂavor components (Bacon, 1957; Darbyshire et al.,
1978). Screening of an onion population for SS and PA requires that a part of
the onion tissues is homogenized, and the lower base 1/3 of the selected bulbs
be saved for seed production. A preliminary study on the distribution of SS and
PA in onion bulb tissues was performed in 1991. lnforrnation obtained through
this study provided the guidance on how to sample onions for ﬂavor tests and
helped to explain how the concentration of ﬂavor substances varied among

onion tissues.

Changes of SS Contents in Onion Bulbs

To determine the variation in SS among tissues in an onion bulb, ﬁVe
sections of the bulb were sampled, including the outer section (two outer scales),
the inner section (from the third to the ﬁfth scale layers), the innermost section
(the two center scales of the bulbs), the base section, (the base of the bulbs) and

the upper section.

46

47

The percentage of SS content of MSU experimental hybrid 2518-1
increased across the scales from the outer older scales to the inner, younger
scales. Table 1 shows the means of ﬁve bulbs in which the outer tissues
contained only 7.64% SS and the innermost had 9.9% SS. Inner tissues were
intermediate in SS concentration, with an average of 9.14% SS. Substantial
differences in SS contents also were found between the top and bottom sections
of the bulbs. The top tissues had a lower SS of 7.9% and the base tissues had a
SS mean of 10.12%. Similar results were reported by Darbyshire (1978). He
pointed out that older tissues tend to store a smaller amount of SS than younger
tissues. Other researchers (Bacon, 1957; Smittle, 1988 and Lin, 1995) reported
that the older tissues, including the outer ring and the top of the bulb have large
cells. It has been found (Darbyshire, 1978) that large cells are closely associated
with low dry weight, sugar and SS, and high ﬁber content. It is likely that stored

sugars are translocated from older tissues to younger tissues of onion bulbs.

Distribution of PA in Onion Bulbs

To obtain information on distribution for PA in onion bulbs, the ﬁve
different sections of the bulbs were sampled in the same way as for SS tests.
The tissues from each sample were blended and tested for PA levels using the

SW method.

The data in Table 2 show that PA in onion tissues varies within the bulb.

The highest PA concentration was found in the top tissues, which reached 10.92

48

Table 1. SS (%) determined by the refractive index method and PA (umol.g“)
measured by the SW method in onion tissues of MSU experimental hybrid
2518-1 .

 

 

 

 

Tissues ’ PA SS
(umol9") (%)
outer 7.64 10.92
inner 9.14 10
innermost 9.9 9.34
regression
liner 66** 127**
quadratic 0.48"s 0.3"s
F value between groups 8.6** 1286“
base 10.12 9.12
top 7.9 11.38
F value between groups 725* 10.21**

 

z The outer tissues were sampled from the ﬁrst and second scale layers; the
inner tissues were taken from the third to the ﬁfth scales; the innermost tissues
were sampled from the center two scale layers of the bulb. The SS and PA
contents were the means of ﬁve onions.

”s,*,** Nonsigniﬁcant or signiﬁcant at P s 0.05 or 0.1, respectively.

 

49
umol.g". The older and outer onion scale layers also had very high PA at 10
umolg". The innermost and the base portions of the scales contained low levels
of PA of the tissues sampled. In this study, the PA level increased from the
innermost scale tissue to the outer scale tissue and from the base section of the
scale to the upper scale tissue. This variation of PA is opposite to that of SS.
Lin (1995) reported different results for short-day onions, PA being lower in the
second and third scale tissues and higher the outermost (scale 2) and the
innermost layers. These SS and PA studies provided the necessary information
to design the procedure for selecting bulbs for SS and PA levels, in which, the
innermost scales from the center of the bulbs may be used for mean values of
SS and PA. There is a signiﬁcant variation of SS and PA among onion scale
layers. The innermost scales and the base tissues of the bulb are statistically
different (P5 0.05) from the outer scales and the top tissues of the bulb in SS

and in PA contents.

Variation of SS and PA Contents Due to Storage Conditions

Onion ﬂavor changes in response to postharvest handling and storage
conditions. Peterson (1986) reported that the onion cultivar Sweet Sandwich
became mild after three months or more in storage. A study was conducted to
determine the variation of total sugar and PA content in Michigan onion
germplasm in response to storage. These tests also were designed to
investigate if there was any particular time that the bulbs could be saved in order

to select for desired SS and PA concentrations. Changes in total SS and PA

50

 

 

 

 

 

 

   

 

 

 

 

   

 

 

 

 

K B 826 B
12 _- Inbred Line 12 F Inbred Line
25 83
a) co
w a)
5 I I l I I
1 i 2 I 3 i 4 I 5 i
Storage Time (month) Storage Time (month)
5718 A 2518
12 _- Inbred Line 12 _— MSU Experimental Hybrid
a) a)
a) m
s l i 1 l l 6 ' I J I l l
1 I 2 i 3 i 4 i 5 i 1 i 2 i 3 I 4 i 5 i
Storage Time (month) Storage Time (month)
5718 x 8155 Sweet Sandwich (ASG)
12 T Two-way Cross 12 _- Commercial Hybrid
8? Z\°‘
a) m
a) w
6 I l l I i 5 ' i l I I u
1 I 2 I 3 I 4 i 5 i 1 I 2 i 3 i 4 i 5 I
Storage Time (month) Storage Time (month)

Figure 3. The effect of storage time on SS content (%) in six different onion
germplasms. Data points are means of ﬁve measurements.

PA (u,ml/g.f.w.) PA (u,mllg.f.w.)

PA (u,mllg.f.w.)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

51

 

 

 

 

 

 

 

 

 

 

 

K B 826 B
25 Inbred Line 25 Inbred Line
’3? 20 ————————————————————— I ______
23 15 ____________________________
E /a—————-
5 1° 5 ‘‘‘‘ ‘/—_—_;,; — — — -----------
E 5 _ - __7/. _____________________
o o
1 2 ' 3 5 ' 1 3 5 '
Time (month) Time (month)
5718 A 2518 MSU
25 Inbred Line 25 MSU Experimental Hybrid
20 ———————————————— p — _ _ _ —————— E
15 ———————————————————————————— 6:
/”- =
,0 _ _ E = ,K/ ______ g.
5 ————————————————————— — — E
o o
1 3 5 1 1 2 ' 3 ' 5 '
Time (month) Time (month)
5718 x8155 Sweet Sandwich (ASG)
25 Two-way Cross A 25 Commercial Hybrid
20 —————— _ _ — ~ — —— , g:
15 ———————————— — _ g}
10 r- _ _=_—_ . _ _? j S-
5 — — — — — ————— g
0 I I

 

 

 

 

 

 

3
Time (month)

 

 

 

 

 

3
Time (month)

Figure 4. Pyruvic acid (PA) levels over time (5 months) of storage for six
different onion germplasms, using the Schwimmer and Weston method. Data
points are means of four measurements.

52
levels in inbreds, K B, 826 B, 5718 A, a two-way cross line 5718 Ax 8155 B,
MSU experimental hybrid, 2518-1, and a commercial hybrid, Sweet Sandwich
stored at 0 C for 6 months from December of 1991 to May of 1992 are presented
in Figures 3 and 4 respectively. Bulbs of onion populations were collected at
random each month from January to May. Five bulbs of each onion population
were used for SS tests using the refractive index method. Four bulbs from each

onion population were collected for PA analysis using the SW method.

Total SS content in the bulbs showed an early increase over the ﬁrst and
sometimes the second month in storage. K B, 826 B, 5718 A increased in SS
contents during the ﬁrst three months and then there was a sharp reduction of
SS after four months. After an increase for two months in storage, total SS
content in 2518-1, 5718 A x 8155 B and Sweet Sandwich gradually declined.
The ﬂuctuation of the SS content in these three hybrids was less than that in the
three line-bred onions. Dehydration of the bulbs was observed during storage.
This probably is attributable to the sugar increase in the bulbs. Bulbs with higher
water content tend to dehydrate more rapidly in storage which would increase
sugar concentration. Hybrids with larger bulbs tend to dehydrate slower than the
line-bred onions. Therefore, changes of SS content in hybrids should be slower
than that in line-bred onions. The sharp drop of SS content at the end of the
storage period may be attributable to increased metabolism and translocation of
the sugars for energy use in response to shoot development which was

observed in most of the line-bred onions.

53

PA levels in K B and 826 B showed continual increases over ﬁve month
in storage and 5718 A, 5718 A X 8155 B, and 2518-1 exhibited changes mostly
in the last two months of storage. The changes in K B, 826 B, and 5718 X 8155
were signiﬁcant during the ﬁrst three months in storage. Again, bulb dehydration
in storage probably caused the PA concentration to increase, whereas,
increased bulb metabolism after three month of storage (shoot development)
would inﬂuence PA concentration. Freeman and Whenham (1976) previously
reported a similar observation and suggested that during sprouting more
precursors are produced and more available to alliinase than before. As a

consequence, more PA is produced.

Although SS and PA levels in the onions changed signiﬁcantly during the
storage, the data suggests that after two or three months of storage (February
and March), it would be the appropriate time to conduct the selection and
evaluation of onion ﬂavor qualities. This is the period of time that the SS and PA
levels are the most stabilized in regard to being able to save selected bulbs for
planting. This study found that onion bases (lower half of the bulb), if stored
more than three months after cutting, became very dehydrated with a high risk
for diseases. It is, therefore, recommended that mass screening of onion bulbs
from populations for SS and PA be carried out after three months of storage,

especially if bulbs are to be saved for planting.

54

Variation of Sugars and Pyruvates in Onion Germplasm

Variation of SS and PA amonwne-bred Populations

Because of severe inbreeding depression, line-bred onions generally are
developed by the self-pollination of a single bulb for the development of a fertile
B line (or intraline cross pollination for the development of the sterile A line) for
two or three generations followed by mass-pollination among a few selected
bulbs. A sterile complementary A line is produced by crossing the developed B
line to a sterile source and then backcrossing for four or ﬁve generations (Pike,
1986). The inbred population remains somewhat heterozygous, which can lead
to changes in genetic constitution over time by drift or selection. This level of
heterozygosity provides the possibility of genetic variance for the selection of

individuals from inbred populations for SS and pungency.

Table 2 shows the SS levels obtained from the test of onion juice refractive
index of seven MSU inbred populations. Ten bulbs were sampled for each
population. The 88 concentration in the bulbs of the inbred lines varied between
and within populations. K B had the greatest variation within a population with a
standard deviation (STD) of 2.06 and a SS range from 5.1% to 12.3%. The lines
9161 B and 5718 B also had high SS variation within the populations with a STD
of 1.56 and 1.49 respectively. The bulbs of 9885 B exhibited relatively less

variation in SS, having the lowest STD (0.91) among the populations tested. This

55

Table 2. Soluble solids’ (%) in seven onion inbred lines

 

 

Inbred Mean” STD C. V. (%)
(%)
5718 A 8.97 1.33 14.83
5718 B 9.98 1.49 14.93
9161 A 9.08 1.08 11.89
9161 B 9.73 1.56 16.03
826 B 9.45 1.14 12.06
K B 9.34 2.06 22.06
9885 B 7.17 0.91 12.69

 

’ Determined with refractometer.
Y Mean of analyses of ten bulbs

56
breeding line also had the lowest SS of all the populations with a mean of only
7.17%. Among the seven inbred lines tested, ﬁve (5718 B, 9161 A, 9161 B, 826

B and K B) had mean values above 9%.

Although environmental effects on SS of these populations were not
estimated, genetic factors still would have to be considered to have a major role
in sugar production in the bulbs. The large SS variation in 5718, 9161 and K B
may be due to the resident heterozygosity left over during the development of
these inbred lines. For instance, 5718 and 826, ofﬁcially released early in 1982
(Peterson et al., 1986), were originated from self-pollinated single-plant
selections in breeding plots at MSU. The selections were mass-pollinated every
2-3 generations to avoid inbreeding depression. A change of genetic
constitution over the time would be expected. Such heterozygosity would allow

the possibility to select for high or low SS individuals from an inbred population.

PA in MSU Inbred Lines

PA analysis was performed every year from 1991 to 1995 to examine the
variation among inbred populations. Tables 4 and 5 show the variation of PA
among several MSU inbred lines in 1992 and 1993. PA in onions varied
signiﬁcantly on a yearly basis. The yearly environmental effects on ﬂavor
changes are discussed in a later section. In 1993, nine inbred lines were
examined for PA, including four populations not analyzed in 1992. The new

populations were K2 B, K3 B, 8155 A and 8155 B. The inbred lines K2 B and

 

57

Table 3. PA Ievels’in seven MSU onion inbred lines harvested in 1992.

 

 

 

Pedigree Mean W STD C. V. (%)
(umol-g")
K1 B 10.89 a 1.29 11.85
5718 A 9.68 bc 1.04 10.74
5718 B 9.02 cd 0.94 10.42
9161 A 9.43 c 2.12 22.48
9161 B 10.67 b 1.69 15.84
826 B 11.84 a 1.36 11.49
9885 B 10.53 b 2.36 22.41

 

2 PA levels were determined with the SW method
W Means with the same letter are not signiﬁcantly different by T test.( P s 0.05).

58

Table 4. PA levels’ in nine onion inbred lines harvested in 1993.

 

 

 

Pedigree MeanW STD C. V. (%)
(um0|.9" )

Kl B V 11.65 a 1.02 8.76%

K2 B 9.21 c 1.98 21.50%
K3 B 7.85 e 2.07 26.37%
5718 B 9.38 bc 1.87 19.94%
5718 A 8.65 cd 0.98 11.33%
9161 A 8.2 d 1.89 23.05%
9161 B 11.45 a 0.89 7.77%

8155 A 8.49 cd 1.23 14.49%
8155 B 7.86 e 1.36 17.30%

 

’ PA levels were determined with the SW method.

y K1B, K2B and K38 are three different selections within the K B line.
WMeans with the same letter are not signiﬁcantly different (T test; P_<_0.05).

59
K3 B are reﬁnement selections out of K1B for single centers and globel shaped
bulbs. The inbred line 8155 B is one of the parents for the hybrid Sweet

Sandwich.

Tables 3 and 4 show that 826 B, K18 and 9161 B had a very high PA
levels more than 10 jumolg'1 suggesting high pungency potentials of these line-
bred onions. The bulbs of the 5718 B and 9161 A had an average PA value
lower than 9.45 pmolg'1 as tested in 1992 (Table 3). The bulbs cf the 8155 B
and K3 B populations were found to be low in PA, less that 7.9 umolg’1 (Table
4), as tested in 1993. The large STD values of the 9885 B and 9161 A
populations (Table 3) indicate that these two lines have a large variation for PA
within their populations. The highest PA levels were about 40-50% higher than
the lowest PA levels in the bulbs within these populations, which should allow
selection for very high or low PA individual bulbs. The average means of PA for
the inbred lines tested in 1992 and in 1993 (Tables 3 and 4) was different for the
two years. A difference in PA was found in the bulbs of K1B which had an
average mean of 10.89 umolg'1 in 1992 and 11.65 umolg’1 in 1993. The
selection for single center bulbs resulted in populations with reduced PA for K2 B
and K3 B, both of which have lower PA than the parent K1 B inbred line. This
needs further study to see if there is any relationship between single centers and

onion PA levels.

 

60

PA and Sugars levels in Storable Onions and Sweet Onions

This analysis was performed in order to understand the difference of total
sugar and PA levels between storable onions and sweet onions. A comparison
of total sugars and PA for 11 storable onion hybrids and four sweet type onions
was made in the summer of 1993 (Figure 5). The storable onion hybrids
analyzed were MSU experimental hybrids 3506, 9490, 5534, and four sister
hybrids of 2518-1 and cultivars Spartan Banner 80, Sweet Sandwich and
Duration. The sweet onions were the cultivars, Vidalia, Sweet Spanish, Walla
Walla and Texas Super Sweet. The sugar content in percentage of dry weight
was estimated from the total sugars by HPLC analysis. PA was measured with
the SW method with modiﬁcations. Figure 5 shows the combined sugar and PA
levels for these onions. Sugar levels in some of the storable onions are as high
as or higher than those were found in the sweet type onions. The four storable
onions 9490, 2518-1, 2518-2 and 2518-3) had sugar levels very close to the
sweet onion cultivars, Vidalia and TX Super Sweet. The cultivar Sweet Sandwich
had a sugar content even higher than Vidalia and TX Super Sweet. Vidalia,
however, generally is considered to be a very mild onion, and in this study it had
the lowest PA of all. The sweet onions all had a PA concentrations (4-7 pmolg")
lower than the 11 storable onions. The major difference between these storable
onions and the sweet onions in regard to ﬂavor is pungency levels. The hybrids
Sweet Sandwich and 2518-2 both had a relatively low PA concentration,

suggesting the

 

61

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

45
A 1' IDsugarII
E,” 40 ,, 7-. .7 7 lpyruvate..-
a 2
:1
V35 I ,1- z a, If
E
I“ I
.2 I I
2325
1:3 I
i”
520
on
:i
m. .-
on
15 7
°\° -
v
I; -
81110 - _ ..
:i — -
m -
75
c5
Pt
08—3118: 2858- 25,23
23°33"? 88583 Era“;
3335
mm g 889,
g8 15 33w
2 a w a
a.” u: §
(I) I—

Figure 5. Sugar percentage (g.sugar/g.dry tissue) and PA (pm0l/g.)
concentration in storable onion hybrids and sweet onions. T bars show the SE
for each mean.

 

62

possibility for selecting within their parent germplasm populations for storable

material with low pungency.

Variation of Reducing sugars in Onion inbred lines and Two-wav Cross

Pooulations

Reducing sugars account for a major portion of the total sugars and SS in
onion bulbs. Analysis of reducing sugars in onion bulbs would provide useful
information for evaluation of the sweetness of onion germplasm or cultivars.
This analysis would be more accurate than SS test because the result of
reducing sugar analysis excludes non—sugar components which make up a part

of the total SS.

Variation in reducing sugar content was found in nine onion populations
tested (Table 5). The inbred line 5718 B and two-way cross 5718 x 8155 had
the highest reducing sugar levels within the materials analyzed. The three-way
cross 5718 x 8155 is the seed parents for 'Sweet Sandwich', a long-day, storable
onion bred for low pungency. Evidently, the selection in these two lines during
the last ﬁve years has made progress toward production for a high sugar onion.
Sugar differences between the A and the B line within a breeding line were
found, indicating different potential for sugar production in A and B lines. Both
breeding lines 5718 B and 9161 B had a higher sugar content than their
respective A lines. Continued backcrosses of these B lines with there respect A

lines should increase the sugar content in the A lines. The 9885 B breeding line

 

63

Table 5. Total reducing sugars in MSU onion breeding lines.

 

 

 

Onion line Reducing sugars ’ STD C. V.
(%)

5718 A 5.95 cy 0.91 15
5718 B 7.8 a 1.08 14
9161 A 5.76 c 0.55 10
9161b 7.1 ab 0.78 11
9885 B 4.68 d 1.13 24
826 B 6.53 be 0.78 12
K B 7.05 ab 1.28 18
9161 A x 826 B 6.55 bc 0.9 14
5718 A x 8155 B 7.77 a 0.53 7

 

’ The Hodge and Hofreiter (1962) method for reducing sugar analysis was used
with modiﬁcations. A standard curve was established using glucose in a series of
concentrations.

Y Separation of means in column by least signiﬁcant difference (I test) at P < 0.
5%.

64
had the lowest reducing sugar. This breeding line is one of the parents for an
MSU experimental hybrid with high onion oil. Probably, some of the energy
products in these plants are converted to oil resulting in less sugar accumulation

in the tissues.
Relationship Between Reducing sugars in Onion Tissue and SS in Onion Juice

In order to understand the relationship between the SS and reducing
sugars in long-day, storable onions, two inbred lines were used for analysis of
SS and reducing sugars. Onion juice was prepared from four bulbs of inbred

line 9885 B and 5718 B.

The SS, as measured by a refractive index was related linearly to total
reducing sugars In onion juice (Fig. 6) as determined by the Hodge and Hofreiter
(1962) method for analysis of reducing sugars in plant tissues with modiﬁcations.
The data support previous reports by other researchers (Mann and Hoyle, 1945;
Bacon, 1957) that the SS and reducing sugar content in onion juice are highly
correlated. This study, however, found that the coefﬁcient of determination
expressed as r2, and the Y intercept value that are involved in the description of
the relationship between SS and total reducing sugars in onion juice may vary in
different onion populations. For the inbred line 9885 B, the equation describing
the relationship between SS content and reducing sugars content in the juice

accounted for 87 percent (r2 = 0.87) of the variability. For the inbred 5718 B, the

 

65

 

 

 

 

 

 

 

12

10 —
3? 8 -
o
“O
.E 6 _
(D
.2.
*9 4 —
“'07-, Y = 3.04 + 0.96 X
t' 2- 8:087
a)
a)
°\° 0 l l i i

0 2 4 6 8 10
% reducing sugar
12

   

5718 B

 
   

.3
CD 0
I I

% SS (refractive index)
0)
I

 

 

 

4— Y=1.25+1.09X
8:088
2—
0 I l l l
0 2 4 6 8 10

% reducing sugar

Figure 6. Regression of percent reducing sugars measured by the Hodge and
Hofreiter (1962) method on percent SS as measured from the juice refractive
index of 9885 B and 5718 B onions.

66

equation describing the relationship between SS content and reducing sugars
concentration In the juice accounted for 66 percent (r2 = 0.66) of the variability.

Although, both populations showed a positive correlation between SS and
reducing sugars, the inbred line 9885 B had a higher Y intercept value of 3.04 for
the regression equation for SS and sugar than the inbred line 5718 B with a Y
intercept value of 1.25. These values are useful indicators of the amount of
nonsugar solutes present in the juice. For instance, the Y value (3.04) for 9885
B can mean that if the juice is tested at a SS level of 3.04, it, theoretically,
contains no sugars but non-sugar soluble solids of 3.04%. In general, 5718 B

had a higher sugar content and less nonsugar solutes in the juice than 9885 B.

Results from this study indicate that onion sugar content in the bulb tissue
can be estimated by using the SS content as measured by the refractive index of
the juice. Although the nonsugar solute present in the juice may interfere
somewhat with the estimate of sugar content in different onion populations when
a refractormeter is used, it can be ignored since the amount of the difference is
small. When a mass screening of individual onion bulbs from a population is
necessary, measurement of the SS would be less expensive and less time-

consuming than the reducing sugars analysis.

HPLC Apalvsis of Onion Sugars

Some onions with low total sugar content may taste sweeter than onions

with high sugar content. Randle (1993) suggested that it is not only because of

 

67

low pungency but also of the different levels of individual sugars. For instance,
an onion with one unit of fructose will be tasted sweeter than the onion with 2

units of glucose.

Many techniques can be used to partition onion sugars. One of the most
efﬁcient and accurate methods is the application of high pressure liquid
chromatography (HPLC). With this method, onion sugars can be separated
when the extraction runs through a column equipped in a HPLC instrument and

analyzed quantitatively according to the elution time and volume.

Twenty-three onion populations were tested in 1993 with HPLC for
individual sugar concentration (% dry weight) in the bulb tissues. The results in
Table 6 are from the ‘Vidalia’ onion, 2518-1, 3506 and ‘Sweet Sandwich’
populations. Three major sugar components, sucrose, glucose, and fructose

were found in the extract injected into the HPLC.

There actually were four peaks being observed on the chromatograms.
Before the elution time of the three major peaks, there was a small amount of a
sugar molecule appearing on the graph, which had an elution time around 0.5
min. This small peak represents one of the oligosaccharides or fructans in the
samples. Fructans are considered to be fructose polymers containing a single
glucose residue and commonly found in onion related species (Darbyshire and
Hanry, 1978). During storage, fructans can be hydrolyzed to fructose and

glucose.

 

68

Table 6. Analysis of sugars in onions from three different sources (MSU

experimental hybrids, commercial storage onions and sweet non-storage onions)
by HPLC.

 

 

 

 

 

 

 

 

 

Sucrose Glucose Fructose Total
Onions

%drywt. %drywt. %drywt. %drywt.
MSU hybrid onions
3506 2.6 2.3 1.9 6.8
2518-12 10.7 11.6 4.9 27.2
2518_2z 11.1 12.9 6.0 30.0
2518-32 11.0 14.8 5.2 31.0
9490 12.4 14.8 7.7 34.9
5534 3.1 3.0 4.1 10.2
LSD ’ 2.07 1.89 2.2 4.01
Commercial hybrid
onions
25134" 10.1 11.5 5.1 26.7
8.8. 80 8.6 9.6 4.6 22.6
Sweet Sandwich 18.3 14.5 8.3 41.1
Duration 7.6 14.5 5.8 27.9
Spartan Banner 10.8 8.5 9.4 30.8
LSD 2.33 1.46 3.86 4.61
Sweet onions
Vidalia 13.3 10.3 10.3 33.8
Sweet Spanish 19.0 20.0 3.5 42.5
Walla Walla 17.1 16.2 7.0 40.3
Texas Super Sweet 12.1 12.4 5.8 30.4
LSD 1.84 2.66 2 2.9

 

2 Different experimental seed production lots of 2518-1.

y Data are means of three samples, with lest signiﬁcant difference (t test) at
P 5 0.05.

x The bulbs of 2518-1 were from a commercial onion trail plot.

69

Previous researchers (Bacon, 1957; Darbyshire and Hanry, 1978; and
Darbyshire, 1978) reported that onions usually had a high content of fructans. In
this research the fructan part showed a very small quantity. Possibly, after winter
storage of these onions, any oligosaccharides could have been reduced

substantially and hydrolyzed in the tissues before the analysis.

Table 6 lists the observed means of sucrose, glucose and fructose of
ﬁfteen onions including six MSU experimental line onions, ﬁve commercial hybrid
onions and four sweet onion cultivars. Total sugar values were calculated by
adding the results of the three major sugars. The oligosaccharides or fructans,

as noted earlier, are not included because of only trace detection.

The four sweet onions ‘Vidalia’, ‘Sweet Spanish’, ‘Walla Walla’ and ‘Texas
Super Sweet’ all had a high total sugar content, above 30%. The ‘Vidalia’ onion
had the highest fructose level (10.3%), which probably accounts for its sweet
taste. The ‘Sweet Spanish’ onion had the highest sucrose and glucose, but it is
not considered as sweet as ‘Vidalia’. This may be due to the low level of
fructose. Large variation in individual sugars and total sugars were found in the
storable, Michigan-produced cultivars. The hybrids ‘Sweet Sandwich’, 9490 and
2518-1 were high in sucrose, glucose and total sugar, whereas 3506 and 5534

were very low in all three sugar components.

Comparing Michigan-produced onions with the sweet type onion, Vidalia,

the major difference is in fructose level. The ‘Vidalia’ onion generally had a

 

70

much higher level of fructose. Other onions in this group all had a relatively low
fructose, except for ‘Sweet Sandwich’ and ‘Spartan Banner’. Considering the
higher level of fructose in combination with sucrose and glucose and low
pungency (Fig. 5), it may explain why the ‘Vidalia’ onion has much sweeter taste
than any other onions. Possibly future breeding work for sweetness improvement

of onions should be directed toward selection for high fructose and low PA

levels.

Improvement of Sweetness and Pungency in Michigan Onion Germplasm
lntraline Crossing and Backcrossing for Development of High-sugar A Lines

Seed parent line development is essential for new hybrids. Onion inbred A line
seeds usually are produced through intraline crossing (A x B). New traits can be
introduced into an inbred line through interline crossing followed by
backcrossing. To increase the sugar level of seed parental lines, intraline
backcrosses were used in this study (Table 7). Total sugar levels of the inbred
lines and progenies were determined by the Hodge and Hofreiter (1962) method.
lntraline crosses were made between high sugar A lines and high sugar B lines
of inbred lines 5718 and 8155. These lines are two of the three parental lines
used to produce the cultivar Sweet Sandwich. An interline single-cross also was
made between the inbred line of 9161 A and 8155 B, and a backcross of the F1
with 81 55 B followed in two years later, using the same selection procedures.

The Sugar levels in the bulbs of the populations from the crosses were compared

 

71

Table 7. Sugar content of onion inbreds and backcross lines following two
successive selections of bulbs with and without regard to sugar contentz.

 

 

 

 

 

 

 

 

Line Cross type Sugar (% fresh weight) in population
generationsy
1992 1994
BC1 BCZ
Backcross (lntraline cross) mean A " F mean A F
5718 A x5718 B no sugar 5.40 4.76
determination"
high X highw 7.15 1.75 2.04” 7.51 2.75 9.78**
9161Ax9161B no sugar 5.55 5.23
determination
high x high 6.87 1.32 1.76“ 7.22 1.99 4.23*
Single-cross (interline cross) F1
9161A x8155 B no sugar 5.66
determination
high x high 6.79 1.23 10.22**
Backcrossing BC1
(9 161 A x 8155 no sugar 6.22
B) x 8155 B determination
high x high 7.54 1.32 5.87*

 

 

 

 

 

z Hodge and Hofreiter method (1962) for determination of sugars in plant tissues
was used.

y First selection was done in 1992. Plants derived from ﬁrst selection are BC1.

Second selection was done in 1994, plants derived from second selection are
B02,

x Regular bulb selection without regard to sugar level.
w Bulb selection with highest sugar level.

V A is the differential value between the ‘high x high’ and ‘no sugar
determination’.

"8, *,** Not signiﬁcant, signiﬁcant (P < 0.05)and highly Signiﬁcant (P E 001)-

72

with the sugar levels in the bulbs of the populations without selection.

The test for sugar content in the progenies was ﬁrst made in 1992. Then
reselected bulbs with high sugar levels were mated again. The second test for

sugar content in the reselected generation was implemented in 1994. Table 8

shows the results from this selection, comparing bulbs selected without regard to

sugar content with bulbs selected for the highest sugar within each population.

The intraline crosses between two high-sugar plants in the families of 5718
and 9161 produced bulbs with sugars higher than their respective selected bulbs
without regard to sugar level in 1992. After reselection, the second backcross
generation (BC2) of 'high x high' had a higher sugar concentration than the ﬁrst
backcross generation (BC1). If comparing, however, the results between 1992
and 1994, the reselection did not signiﬁcantly increase the sugars in the BC2
generation, because the yearly effect was high; the bulbs selected without
regard to sugar level also increased in sugar at about the same percentage over

the 1992 analysis.

The single-cross line, 9161 A x 8155 B also showed an increase in sugar
level of 1.23% (the A value in Table 8). The ‘high x high’ progeny of this cross
was backcrossed with the high sugar parent line 8155 B. This backcross

progeny had a sugar level,1.34% higher than the no sugar determination cross.

Populations of 5718 B and 9161 B inbreds after reselection and selﬁng of

individual bulbs for two generations, exhibited poor vigor, low seed yield, small

 

73

bulb size, and poor storability. Further inbreeding (selﬁng with four bulb masses)
probably would reduce the plant vigor seriously with possible elimination of the

entire population.

The ‘high’ selection populations were high in sugars, suggesting a good
potential for developing new parent lines with higher sugar levels by selection
within the lines. Variation from bulb to bulb in the high x high selection was

found for all the sugars measured.

Variation in Sugar Components Determined by HPLC after Selection for High

Total Sugar Onions

Sugar content of progeny onions derived from a cross between high SS
bulbs of 5718 A and 8155 B was compared to the original parents in 1994. The
levels of sucrose, glucose, fructose and total sugars in the bulbs of these three
populations were determined by HPLC and are shown in Figure 7. The selection
for higher SS not only increased total sugar concentration in the selected
progeny population, but the sucrose and glucose levels were higher than either
parent. Fructose in the single-cross hybrid population was lower than that of the
pollen parent, 8155 B, but the same as in the female parent, 5718 A. These
results indicated that although the total sugar content of onions can be improved
with a 'high SS x high 88' crossing combination, sugar increase is mainly an
increase of sucrose. This also implies that breeding onions for high fructose,

which is the sweetest component among the three major onion sugars, may

 

74

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

0
sucrose (% dry wt.) glucose (/° dry Wt')

50 50

40 -— 40 -
‘5 30 —— 5:3: 30 ‘
d) m
8 8
‘6 20 —— 1- 2 20 -
3 T U)
(D T

10 —— 1° ‘ T T

0 0

5718A 81558 5718Ax81558 5718A 31553 5713AX81553
fructose(% dry wt.) total sugar (% dry wt.)
50 50
40 — 40 ~ I
’3 T

$3 at
I; 30 — a 30 —
(I) U)
2 53
g 204 a 20 a
u. T §

10 n T L 10 —

0 0
5718A 81553 5713AX3155B 5718A 81558 5718Ax81553

Figure 7. Sugar content of two parental onion lines and hybrid progeny onions
derived from high soluble solids bulb selection. Data are means of four bulbs.
The mean value of the onion lines as measured with HPLC is represented by a
bar, and the error for each bar was calculated at the 95% conﬁdence level.

 

75
require a different system, speciﬁcally for selection of germplasm high in
fructose. Individual bulbs might have to be screened with HPLC which would be
time-consuming and costly. Possibly using the quicker, less costly refractometer
analysis for SS would indicate total sugars, sucrose and glucose in the selected

bulbs, but it would not be an indicator of fructose levels.

Even though the Table 7 results are from reducing sugar analysis and
Figure 7 results are from HPLC analysis, the resulting information from both
shows that selection for high sugar levels can help increase total sugar in
resulting populations. Breeding lines or experimental hybrids might be perceived
to be sweeter if the level of reducing sugars could be maintained at 6.5% or
higher, or total sugar at 35% or higher, or SS at 11% or higher, and the PA level

could be maintained at around 6-7 pmolg".

Reducing Onion Pungency in Inbreds and Seed Parent Populations

To reduce PA in onion germplasm, intraline crosses, interline crosses,
and selfs were used along with low-PA parent bulb selections. Table 8 contains
data of PA tests by the SW method in 1992 and 1994 of the progeny populations
of four fertile inbreds (B lines) developed by selfing, two sterile inbreds (A line)
developed by intraline backrossing, and two single-cross seed parents
developed by interline crossing. The selfed populations derived from the low PA
selection all showed a lower PA concentration than the controls (regular

selection without regard to PA). The difference of PA levels between the two

 

76

Table 8. Pyruvate contents determined with the SW method (1961) in onion
inbreds and seed parent populations derived from low pyruvic acid (PA) bulb

 

 

 

 

 

 

selections.
Line Cross type PA (pmol.g“)

1992 1994
Fertile lines low PA regular low PA regular

selection 2 selection

5718 B selfed 7.92* 8.7 8.46** 10.96
826 B selfed 9.46 "s 9.78 9.25ns 9.9
8155 B selfed n/a y n/a 8.02* 9.15
9161 B selfed 8.33“ 10.22 7.22* 10.64
lntraline cross
9161 A x 9161 B backcross 6.6** 9.2 8.2** 10.2
5718 A x 5718 B backcross 7.2* 8.2 7.4* 8.6
Interline cross
(two-way cross)
5718 A x 8155 B single-cross 5.6** 8.6 7.2* 8.1
9161 A x 826 B single-cross 7.1** 9.7 6.9** 8.7

 

 

 

 

 

"3,*,**, Not signiﬁcant or signiﬁcant by t test at P = 0.05 or 0.01, respectively.
’ Regular bulbs selection without regard to PA content.

y Data are not available due to missing plots and poor bulb harvests.

77

types of selection was signiﬁcant except for the 826 B line.

After a generation of selﬁng of the inbred line 9161 B for lower PA, this
population was more than 30% lower in PA than the regularly selected
population (the control). Severe inbreeding depression, however, was observed
in the 9161 B population as evidenced by small bulb size and poor vigor. Any

further selﬁng after the second generation, thereforee, would not be advisable.

In the intraline crosses between A and B plants of 5718 and 9161, low PA
selection (low x low) produced a progeny of A lines with lower PA in both 1992
and 1994. The difference over regular selection (the control) was signiﬁcant both
years. Signiﬁcant differences in PA also were found in the interline cross
populations indicating that selecting for lower PA was effective, with a potential

for the development of less pungent onion hybrids.

Three-wav Cross Hvbrid Onions with Selection for High SS and Low PA

1. Selection for higher SS

Onion commercial hybrids usually are produced from three-way cross
which involved three parental lines. To understand how parental bulb selections
for high SS and if any particular three-way cross combinations of germplasm
would affect the SS levels in the hybrid populations, two three-way hybrid onions,
(9161 A x 826 B) x K B and (5718 A x 8155 B) x 826 B were produced from high-

SS parental bulbs. Eight bulbs from each hybrid were produced from the same

 

78

Table 9. Soluble solid contents (%) in three-way cross onion hybrids produced
from high SS selected parent bulbs or from regular parental bulbs.

 

 

Hybrid Cross type 1992 1993 1994 Mean Mean gain
(%) (%) (%) (%) (%)

(9161 A x 826 without SS 8.23 8.09 9.46 8.59
B) x K B selection

high x high 9.67 9.88 10.33 9.96 1.37
(5718 A x 8155 without SS 9.75 8.41 8.96 9.04
B) x 826 B selection

highxhigh 11.02 10.68 11.14 11.03 1.91

 

 

 

 

 

Analysis of Variance for (9161 A x 826 B) x K B

 

 

Source Sum of Degree of Mean F Value
Square Freedom Square

between 2.8 1 2.8 822*

groups2

within groups 1.36 4 0.34

total 4.16 5

 

 

 

Analysis of Variance for (5718 A x 8155 B) x 826 B

 

 

Source Sum of Degree of Mean F Value
Square Freedom Square

between 5.96 1 5.96 20.43**

groups

within groups 1.16 4 0.29

total 7.12 5

 

 

 

*, ** Signiﬁcant at P < 0.05 or highly signiﬁcant at P = 0.01, respectively.
7' Groups are the ‘without SS selection’ and ‘high x high’ combinations.

79
seed lots in 1992, 1993 and 1994 were tested for SS using the refractive index
method. The same number of bulbs also were taken from the regular hybrids
(no selection for SS in the parent bulbs) of the same seed lots and tested for SS.
The results are shown in Table 9. The two hybrids developed from high-SS
parent bulb selection are listed as ' high x high'. The regular hybrids are listed as

‘without SS selection’.

Three years of testing for SS of the two three-way hybrids, (9161 A x826
B) x K B and (5718 A x 8155 B) x 826 B, showed consistent higher SS levels in
the 'high x high’ selections. Analysis of variance (ANOVA) results indicate that
the selection in (9161 A x826 B) x K B and (5718 A x 8155 B) x 826 B hybrid
populations signiﬁcantly changed the levels of SS based on the three years of
data. The hybrid (5718 A x 8155 B) x 826 B had an even higher change in sugar
content. The two hybrids from high-SS selected parent materials increased SS
level by 1.37% and 1.91%, respectively. This information, again, suggests that
selection for high SS parent bulbs can be used to produce three-way hybrids

with higher SS levels.

2. Selection for lower pyruvic acid

Selection for lower pyruvic acid parent bulbs for production of the seed of
two three-way hybrids, (9161 A x826 B) x K B and (5718 A x 8155 B) x 826 B,
was performed for three years (1991,1992 and 1993). Five bulbs from each of

these hybrid populations, including the ones from the same seed lots without PA

 

80

Table 10. Pyruvic acid (PA) contents (pmol.g") in two three-way cross onion
hybrids produced from parent bulbs with or without the selections for low PA
levels as determined by the SW method.

 

 

 

 

 

 

 

 

 

 

 

Hybrid Cross type 1992 1993 1994 Mean PA
pmolg'1 umolg". pmolg“. pmolg' reduction
’ (%)
(9161 A x without PA 9.2 8.4 10.6 9.4
826 B) x K B selection
Iowx low 8.2 5.6 6.8 6.9 -2.5
(5718 A x without PA 8.2 9.1 7.4 8.2
8155 B) x selection
826 B
Iowx low 6.2 7.8 5.5 6.5 -2.7
Analysis of Variance for (9161 A x 826 B) x K B
Source Sum of Degree of Mean Square F Value
Square Freedom
between 9.6 1 9.6 6.56 **
groups2
within 5.9 4 1.47
groups
total 15.49 5
Analysis of Variance for (5718 A x 8155 B) x 826 B
Source Sum of Degree of Mean Square F Value
Square Freedom
between 4.5 1 4.51 4.26"“‘
groups
within 4.2 4 1.06
groups
total 8.7 5

 

 

"3,“ Not signiﬁcant or highly signiﬁcant at P = 0.01, respectively.

2 Groups are the ‘without SS selection’ and ‘high x high’ combinations.

 

 

 

 

81

selection, were tested in 1992, 1993, and 1994 for PA contents with the SW
method. Progeny from the two 3—way cross hybrids using parental material

selected for lower PA levels showed lower PA levels over three years of analysis

(Table 10).

The cross of low by low of the hybrid (9161 A x 826 B) x K B , produced
a lower mean PA level of 6.9 umolg", whereas the non-selected PA control
had a higher mean PA of 9.4 pmol.g". The ANOVA showed that the change in
the selections (Table 10) was not found to be statistically different from the
control. The 2.5 and 2.7 percent reduction, respectively in the PA level in these

two 3-way hybrids, however, is within the desired 6-7 range of umolg'1 of PA.

Genetic Analysis of Pungency and SS in Storable Onions

Effects of Selected Parents for SS and PA Levels on Progeny Populations

To obtain an estimate of response to different selection pressures for SS
and PA, progeny population following four different crosses (high x high, low x
low, high x low and low x high for SS and PA) involving six cross combinations
(Tables 11 and 12) were compared to the onions of the same combinations
without selection as the control in 1992 and 1994. The two years of data
showed that the high x high SS crosses and the low x low PA crosses generally
were the most effective for increasing SS and reducing PA in the progeny

populations. Signiﬁcant changes in SS and PA were found in the onions mostly

 

82

Table 11. Soluble solids (SS) content of onion progenies derived from parental
bulbs selected for different SS levels determined from refractive index of onion

 

 

 

 

juice.
Combinations SS (%) z
1992 1994
high low high low control high low high low control
x x x x x x x x
high low low high high low low high
Backcrosses

 

5718Ax5718 B 11.8** 7.2* 7.9Ins 8.2Ins 8.2 13.4** 7.4* 8.2ns 8.0ns 8.6
9161 Ax 9161 B 12.8** 6.6** 8.6ns 8.8"8 9.2 14.5** 8.2** 11.2*8.0** 10.2

 

2-way crosses
5718Ax 8155 B 13.8** 8.0* 8.4"3 7.9* 8.6 13.6** 7.6** 9.0"3 8.8ns 9.4

9161 Ax 826 B 10.6* 7.2** 6.8** 9.5”8 9.2 14.0** 9.2ns 8.8“8 9.2"8 8.6

 

3-way crosses
(5718Ax 8155 14.8** 8.0** 8.8"8 9.4"8 10.6 13.5“ 7.2** 7.8* 10.1”8 9.8
B) x 826 B
(9161 Ax 826 B) 12.5** 7.9"8 8.5ns 8.2"5 8.4 11.8** 8.2** 8.1** 9.0* 10.2
x K B

 

 

 

 

 

"3, *,**, Not signiﬁcantly or signiﬁcantly different from the value of control at P _<_
0.05 or 0.01, respectively.

2 Means of 6 progeny bulbs.

Table 12. Pyruvic Acid (PA) levels in onion progenies derived from parental

83

bulbs of different PA levels determined with the SW method.

 

Combinations

PA levels (pmol.g") z

 

 

 

 

 

 

 

 

1992 1994
high low high low control high low high low control
x x x x x x x x
high low low high high low low high
Inbreding
5718Ax 5718 B 12.3"5 7.8** 10.4** 9.6** 12.1 13.5** 6.8** 10.0* 12.4* 11.3
9161 Ax 9161 B 14.6** 8.4"5 9.8* 8.7"5 8.8 11.0** 6.4** 7.9* 10.5* 9.2
1-way crossing
5718Ax 8155 B 10.9** 5.6** 6.8** 7.3* 8.6 9.6** 7.2* 10.2** 8.2"s 8.1
9161Ax 826 B 16.4** 7.1** 8.6* 8.4* 9.7 14.0** 6.9** 6.8** 9.4"5 8.7
2-way crossing
(5718Ax8155 8.6"3 6.2** 7.9"3 8.0"3 8.2 13.4** 5.5** 8.3* 7.2"8 7.4
B) x 826 B
(9161Ax826 B)18.7** 8.2* 9.9"8 10.3"8 9.2 16.4** 6.8** 8.7** 8.4** 10.6
x K-1

 

ns***
9 i a

2 Means of 5 progency bulbs.

Not signiﬁcant or signiﬁcant at P = 0.05 or 0.01, respectively.

 

 

84
from the high x high and the low x low crosses. The high x high SS crosses
gained signiﬁcantly higher SS (mostly at P5001) than the control and the low x
low gave signiﬁcantly lower SS than the control for ten out of the twelve crosses.
The low x low SS cross in the combination of (9161 A x 826 B) x K B, however,
was not signiﬁcant but still lower than the control. The low x low 9161 A x 826 B
cross of 1994 for SS was completely different to the 1992 cross. The 1992 cross
was highly signiﬁcant for a lower SS level, whereas, the 1994 level was not
signiﬁcantly different. The progenies of the high x low and the low x high crosses
responded differently to the parental selection. Most of these combinations (17

out of 24) for SS levels did not result in a signiﬁcant change in SS.

These same high x low and low x high combinations for PA levels gave
rather varied results. There were eleven values signiﬁcantly lower than the
control and 5 values signiﬁcantly higher than the control. It is difﬁcult to draw any
general conclusions to suggest if any one of these two crosses (high x low ;|ow
x high crosses) would lead to respective increase or a decrease in SS and PA
contents in the progeny of such crosses. The results indicate that if higher SS
and lower PA levels are desired it should be possible by selection of bulbs that
have been analyzed accordingly. This supports suggestions of other researchers
(Bacon, 1954; Lin et al., 1995 and Simon, 1995) that SS and PA in onions are
traits controlled by additive genes and that additive genes have small individual

effects and are inﬂuenced markedly by the environment. In the case of low x

85

high and high x low crosses, possibly any additive inﬂuence would be working

against itself causing varied results such as observed in this analysis.

Correlation A_mong PA. SS and Bulb Sizes

When comparing PA to SS, some researchers assumed that there was a linear
relationship between SS and PA (McCollum, 1968; Lin et al., 1995; Simon,
1995). In short-day onions, PA and SS also were related linearly to bulb size
(Lin et al., 1995). In this study with We storable onion hybrids (MSU
experimental hybrids 2518-1 and 3506; commercial onions ‘Sweet Sandwich’,
‘Spartan Banner 80’, and ‘Norstar’), the correlation coefﬁcients of PA, SS and
bulb sizes calculated from the data of these ﬁve onion lines were signiﬁcant,
except for PA:SS for MSU experimental 3506 and SS:Size for MSU experimental
2518-1 (Table 13). PA content in four of the ﬁve hybrids had a moderately
positive correlation with SS, which ranged from 0.52 to 0.61, suggesting that the
onions with higher PA tend to contain more SS in the bulb tissues. PA content in
3506 had a low correlation coefﬁcient with SS, which was not signiﬁcant,
indicating another possibility that the relationship between SS and PA may vary
according to onion hybrid. Bulb size of all ﬁve hybrids had a negative correlation
with SS and PA, indicating large bulbs tend to have lower concentrations of SS
and PA in their tissues. Because of the positive relationship between SS and PA
in some of the onion populations, it may be difﬁcult to breed for an onion with

high SS and low PA. Due to this relationship, breeding, therefore, for high SS

and low PA storage onions may requiring more extensive analysis over a longer

time frame.

Table 13. Correlation coefﬁcients (r) of pyruvic acid (PA), soluble solids (SS),

and bulb sizes in ﬁve onion varieties.

86

 

 

 

 

Onions PA:SS SSzSize PA:Size
r

2513-1 ’ 0.59 ** -0.49 -0.69 **

Sweet Sandwich 0.53 * -0.62 ** -0.53 *

Spartan Banner 80 0.61 ** -0.54 * -0.64 **

Norstar 0.52 * -0.66 ** -0.51 *

35052 0.22 -0.56 * -0.63 **

 

MSU experimental hybrids.
* Signiﬁcant at 5% level.

** Signiﬁcant at 1% level. Each r was calculated in a regression
analysis based on the data of 10 onion bulbs from one population.

 

87
The negative correlation among bulb size and PA, and SS suggests that

large bulbs of storage onions tend to have lower concentrations of SS and PA. It

is natural to favor large bulbs for breeding germplasm, because of their yield
potential and mild ﬂavor. This relationship, however, may be based upon the
moisture content in the bulb. Large bulbs usually contain a larger amount of
water that dilutes all the substances in the cells. As large bulbs become drier at
the end of a long store period large, mild onions can become more pungent too

(Platenius, 1944).

Broad-sense Heritabilitv for SS, PA and Ratio of SS to PA in Onions

To study inheritance of SS and PA, interline crosses were made between
fertile onions with high SS and low PA in 1991 and 1992. The F1 onions were
selfed in 1993 and 1994. A variance components procedure (Mahmud and
Kramer, 1951) was used to calculate the broad sense heritability for SS, PA and

SS/PA ratio (Table 14).

Broad-sense heritability for SS, PA and SS/PA ratio for four onion
populations was estimated, and the results are presented in Table 14. The
values obtained from different lines vary to some extent. Heritability of SS, PA
and SS/PA ratio ranged from 87% for 5718 B x 826 B, to 73% for 9161 B x 826
B for SS, 68% for 9161 B x 8155 B to 47% for 5718 B x 826 B for PA, and 77%
for 9161 B x 826 B to 59% for 5718 B x 826 B for SS/PA ratio. These values are

useful as they give an estimate of maximum heritability for SS, PA and SS/PS

88

Table 14. Broad-sense heritability (H2)z estimated for soluble solids (SS), pyruvic
acid (PA) concentrations and SS/PA ratio for four onion populations.

 

 

 

Crosses H2
88 (%) PA(%) SS/PA(%)
. 9181 Bx8155 8’ 74 68 68
57188xK8’ 84 59 67
5718 B x 826 8" 87 47 59
9181 B x 828 B " 73 60 77
Mean 84.5 58.5 67.8

 

 

 

 

1 H2 value was calculated using a variance components procedure (Mahmud
and Kramer, 1951) based on the data of SS tests by using onion juice refractive
index and PA analysis (Schwimmer and Weston, 1961). Eight and four bulbs
from each population were used for SS test and for PA analysis, respectively.

’ Data of 1994.
" Data of 1995.

89
ratio in these populations. For an example, the two interline crosses, 5718 B x
826 B and 5718 B x K B which have the H2 value more than 84% would have a
higher genetic potential to produce high SS progeny than the two other interline
crosses, 9161 B x 826 B and 9161 B x 8155 B with a H2 value less than 75%. In
general, these four interline crosses have a higher H2 value for SS (73—87%) than
for PA (47-68%), indicating that selection for SS is generally more effective than

for PA.

Reali_zed Heritabilitvin Selfed Inbred Populations

Five reselected inbred fertile lines were developed from selecting progeny
from selfed bulbs for either low SS or low PA and for either high SS or high PA.
The reselections within the inbred populations of 5718 B, 9161 B and 826 B
resulted in changes for SS and PA levels in the S, progenies as is shown in
Tables 15 and 16. The mean values of S, populations from high-SS selection
were all higher than their parents and the mean values of S, populations from
low-SS selections were all lower than their parents. The same trend also was
observed in the PA selection, but the changes in PA were smaller than that in
SS. The amount of genetic improvement realized by this type of selection was
estimated by calculating the realized heritability H3. The highest Hf value for SS
was found in the population of 826 B, indicating that reselections in this
population is highly effective. The second highest H} value for SS was found in
8155 B. These two inbreds are the major parental lines of the three-way cross

hybrid, ‘Sweet Sandwich’. These results indicates that selections for high SS in

90

Table 15. Realized heritability (H?) for soluble solids (SS) (%) in ﬁve inbred
onion fertile (B) lines derived from selﬁng high SS or low SS parental bulbs.

 

 

 

 

Selfed Selection Mean of Mean of S 2 RX H,2 Y
inbreds for SS parent S,
(%) (%) (%)

8155 B high 10.4 14.6 4.8 4.2 87.5
5718 B high 10.2 13.6 4.6 3.4 74
826 B high 8.8 11.2 2.6 2.4 92
8155 B low 11.4 8.0 -4.9 -3.4 69.4
9161 B low 8.2 6.5 -2.4 -1.7 70
Mean 3.86 3.02 78.6

 

2 Selection differential = mean of S1—mean of control (without selection).
y Realized heritability, H2 = R/S (Falconer, 1981).
X Response realized by selection = mean of S1- mean of parent.

91

Table 16. Realized heritability (H?) for pyruvic acid (PA) content (pmol.g") in
ﬁve inbred onion fertile (B) lines derived from selﬁng high PA or Low PA parental

 

 

 

 

bulbs.

Selfed Selection Mean of Mean of S 2 R " H,2 Y
inbreds for PA parent S,

(umol.g") (pmol.g") %
5718 B low 10.96 8.46 -2.91 -2.5 85
9161 B low 10.64 7.82 -3.6 -2.82 78
826 B low 9.9 9.25 -1.89 -0.65 34
5718 B high 10.96 13.6 3.7 2.64 71
9161 B high 10.64 12.5 2.3 1.86 81
Mean 2.88 2.1 69.8

 

2 Selection differential = mean of S1 - mean of control (without selection).

x Response realized by the selection =mean of S1- mean of parent.
YRealized heritability, H2 = R/S (Falconer, 1981).

92
these two line population possibly lead to a higher SS level in this hybrid. In
Table 16, among the ﬁve inbred lines, the response (R) of 9161 B to the
selection for PA was the greatest (-2.82 ), meaning that through the selection the
progeny PA mean value has been reduced and is 2.82 (umol.g") lower than the
mean value for the parent. The largest Hf value, however, was found for the
selection of 5718 B. H} values for PA for both 5718 B and 9161 B were large
and suggested this type of selection is possible for PA improvement of these two
lines. The inbred 826 B line was the only line that failed to show Hf large
enough to justify the genetic improvement by selﬁng and selection for low PA. It
could be due to the homozygosity of the parental population, as low PA was
already high at the time of selection, which would make the selection among the
population less effective. The average H,2 value for SS (78.6%) is higher than
that for PA (69.8%), indicating selection in these selfed inbred line populations

for high or low SS is more efﬁcient than for high or low PA.

Relative Frequency Distributions of Individuals for PA and SS in Two Parent B

Lines and the Progenv Populations

Crosses were made by pollinating 5718 B plants containing high SS or
low PA levels with 826 B plants containing low SS or high PA levels. Knowing
the levels of SS and PA in the parent and progeny populations made it possible
to determine the frequency distribution for SS and PA within the F1, F2, and

backcrossing Bp1 and Bp2 generations. The frequency distributions for SS and

 

 

93

PA are shown in Figures 8 and 9. The relative frequency (%) was calculated by
dividing the number of the bulbs at a SS or a PA level with the total bulb number
sampled from the population. In Figures 8 and 9, the parent and progeny
populations are represented on the horizontal axis, and mean values of SS and

PA are represented on the vertical axis.

These two ﬁgures show the typical mode of inheritance for additive gene
action controlling the two traits, SS and PA levels. Each of them have a
continuous distribution ﬁgure on the graphs with the F18 and F2s intermediate
between the P1s and P2s, and the Bp1s intermediate between the P1s and F25,

and the Bp2s intermediate between the P2s and F1s.

In Figure 8, there were two separate means of SS for the parental
frequency distributions (P1 and P2). P1 had a distribution graph at a mean near
9%, and P2 appeared to have a distribution at a mean near 8%. The F1
distribution was relatively narrow, with a mean positioning from 7 to 11%, similar
to the high SS parent, P1. Bp1 was strongly skewed toward P1 while Bp2 was
skewed toward P2. The F2 frequency distribution was relatively wide and less
symmetrical with a mean positioning from 7 to 10%. This indicates again that
there is additive inheritance for SS. The F1 population had more individuals with
the gene effect for these two traits. The F2 distributions for both SS and PA were
typical and suggestive of additive inheritance. The means of the progeny high
SS than the F2 population. There is also the possibility that heterosis in the F1s

may have resulted in increased sugar production in the bulbs. In Figure 9,

 

94

   

 

 

 

 

 

 

 

 

 

 

 

 

      

 

 

 

 

S" 33
O 0
3 3
2 2
LL u.
3 ,3
E E
I) 0
tr 1:

e 7 a 9 . 10 11 12 e 7 3 9 . 1o 11 12

Soluble Solids % Soluble Solids %
F1 F2

a\° «- S
O 1). ID
3 I a
2 . E
u. ”I u.
3 1&— .3
7‘: 10 E
a , '6
n: n, n:

e 7 a 9 ~ 10 11 12 a 7 a 9 . 1o 11 12

Soluble Solids % Soluble Solids %
Bp1 Bp2
32 m _,h o\°
>. I >
8 m ‘c’
O 30 I)
8 I e
' 7 7 7 7 I.)
E 201 It
I) L z o
2 2
5 1°) 5
0 R7 2 L a 0
tr AA_ 1.:
a‘ __
e 11 12 e 7 11 12

a o 10 a 9 10
Soluble Solids % Soluble Solids %

Figure 8. Frequency distribution of percent SS content of the onion parents and
progenies of 5718 B x 826 B. The P1 and P2 represent 5718 B and 826 B
respectively; the F1 is the progeny of 5718 B x 826 B; the F2 is the progeny from
selﬁng the F1 (5718 B x 826 B); Bp1 is from the backcrossing of the F1 (5718 B
x 826 B) with 5718 B; Bp2 is from the backcrossing of the F1(5718 B x 826 B) x
826 B.

95

   

 

 

         

 

 

 

 

 

    

 

 

 

 

 

    

 

 

 

P1 P2
°\0 60 °\° 60
6 I 6
r: - r: . 2 . .
9 40' g ,0
8 l 8
2', mi '1", 20
.2 1 .2
%o ‘2 A %o
m 6 8 10 12 14 16 I! 6 8 1o 12 14 18
PA content (um/g) PA content (um/g)
F1 F2
°\° 60 °\0 60
6 6
C E
g 40‘, g 40 _
U‘ ‘ 0'
e 9
“I; 20). E 20
.2 . .2
E . , E
a) O . ' (D . .1
'1 6 8 1o 12 14 16 0‘ 6 8 1o 12 14 16 18
PA content (um/g) PA content (um/g)
Bp1 Bp2
c\° o\°
>.‘50 7 7 >5“
8 l 8
g ml 3 40
O' O'
93 1 9
E 201 2;, 20
g o, ,, ,m g “_l
m 6 8 10 12 14 16 0‘ 6 8 1o 12 14 16
PA content (um/g) PA content (um/g)

Figure 9. Frequency distribution of PA concentration of the parents and
progenies of 5718 B x 826 B. The P1 and P2 represent 5718 B and 826 B,
respectively; the F1 is the progeny of 5718 B x 826 B; the F2 is the progeny from
selﬁng the F1 (5718 B x 826 B); Bp1 is from the backcrossing of the F1 (5718 B
x 826 B ) with 5718 B; Bp2 is from backcrossing of the F1 (5718 B x 826 B ) with
826 B.

96
although the frequency distributions of P1 and P2 covered a range of PA from 8
to 10 j.tmol.g'1 while the P28 had two major portions of bulbs (about 55% of the
total) with a PA level in a range from 10 to pmolg". The F1s had more low PA
bulbs than the P1s. The distribution of the F1, however, had a conﬁguration
similar to the P1, except that high PA (16 umolg") bulbs were not found in the
F1s but in the P1s. The F2 distribution showed again a relatively wider and less
symmetrical distribution than the F1 and exceeded the range of that found in the
P1 and P2 graphs. The two backcrosses had distribution conﬁgurations very
close to their respective parent lines in regard to total spread and peak
frequencies. Bp1 had a low frequency of less than 10% in the 14 jtmolg'1 range
and an absence in the 16 umol.g'1 range. This could be due to additive
dominance for lower PA, or possibly the larger sized F1 bulbs had more

moistures which could dilute the PA concentration.

Lower level of PAs and higher level of SS in the F1s than that found in the
F2s suggest that herterosis for vigor in the F1 bulb may have increased SS, and
the large bulb size of the F13 may have improved the PA trait . In general, the
frequency distribution graphs for SS and PA provided useful information about
populations (F1s, F2s and 806) also indicate additive—dominance of high SS
over low SS and low PA over high PA. These results of frequency distribution
for SS are similar to the early reports by Warid (1952) and Owen (1961). Warid
estimated that there are 4-10 gene pairs involved in determination of the SS

level in onions after crossing high SS onions with low SS onions. He also

 

97

concluded that partial dominance of low SS is involved. Owen, who examined
the segregation of SS genes in short-day onions, also suggested an additive-
dominance mode in the inheritance of SS. They both used Spanish type onions
in their studies. The information from this study indicates that the additive-
dominance mode for SS is also suitable for long-day onions. Additive gene
effects on PA also were suggested by other researchers (McCollum,1968;
Simon, 1995; and Lin et al.,1995). They all reported heritability of pungency
measured by PA analysis to be fairly high and suggested additive-dominance

gene action for low PA over high PA in onions.

Yearlv Variation in SS and PA of Onions

Year-to-year variation in SS and PA were evaluated for onion population
2518-1. PA, and SS:PA ratios (Table 17) differed signiﬁcantly among the ﬁve
harvest years. The only year in which the onions had an SS content signiﬁcantly
different from other years was 1994. The percentage of SS showed less
signiﬁcant yearly variation, suggesting that SS content is inﬂuenced less by
environmental variables than PA. Data of 1995 showed lower concentrations for
both SS and PA than any other year. This may be due to the fact that the bulbs
had higher water content; as they were analyzed after two month in storage
which is 2-3 month earlier than when the onions were tested in the other four
years. Year-to-year variation in the SS:PA ratio may be explained by the large
variation of PA among the test results of these ﬁve harvest years. Because SS

levels of onion bulbs are inﬂuenced less by environmental factors, the yearly

 

98

Table 17. Variation in SS, PA and SS:PA ratio in MSU experimental hybrid
2518-1 over ﬁve years.

 

 

 

Year SS PA SS:PA
(%) (umol-g")

1991 35 a2 9.2 b 0.93 b

1992 9.23 ab 11.25 c 0.82 a

1993 8.2 a 8.27 ab 0.99 c

1994 9.7 b 10.93 c 0.89 b

1995’ 7.9 a 7.86 a 1.01 c

 

2 Mean separation within columns by ttest (P s 0.05). Data are means of 10
bulbs for SS and 6 bulbs for PA, respectively.

y Bulbs harvested in 1995 were analyzed 2-3 months earlier than the previous 4
years.

99
change of SS:PA ratio is really due to the large change in PA. A change for PA
in a year, therefore, would have a very strong effect on onion ﬂavor. Randle
(1992) suggested that the strong pungency of an onion can mask the sweet
ﬂavor produced by sugars. The results of this multiple year analysis indicates
that environmental variables have a strong effect on the ﬁnal ﬂavor perception of

onions.

A Simpliﬁed PA Procedure for Determination of PA in a Breeding Program

 

The SW PA method generally has been used for evaluation of onion
pungency by many researchers (Randle, 1992; Lin et al., 1995; and Simmon,
1995). This method, however, is very time-consuming, requiring as long as 100
min for a single test. A simpliﬁed procedure was developed which uses much
simpler equipment and requires less than 30 min for a test. Two studies were
conducted to test the reliability and workability of the MSU PA procedure. A
comparison was made in which PA concentrations of onions were tested using
the simpliﬁed and SW procedures. The PA of ﬁve onion cultivars, made up of
those considered sweet or pungent, were determined. The second study
involved screening two inbred populations 9161 A and 9161 B for high- and low-
PA bulbs. Crosses of plants with high PA or of those with low PA were made
between 9161 A and 9161 B. If the PA levels of the parent bulbs tested with
this procedure are true, the high- and low-PA crosses should then produce the

same type of offspring.

100

Comparison of a Simpliﬁed MSU PA Procedure with the Classical Schwimmer

and Weston PA Method

PA content measured by the SW method and color intensities measured by the
new MSU procedure were highly correlated with a coefﬁcient of determination (r2)
of 0.83. This indicates that 83 percent variation in these tests is accounted for
by the linear regression. These results also show that the visual detection with a
color value of the ﬁnal PA-DNPH derivatives (Materials and Methods, p 35) in a
sample can be used for estimating PA concentrations. Each color value
represents a range of PA concentrations. Figure 11 shows that the color value 1
(yellow) has a PA > 5.4 pmolg", color value 2 (light purple) covers PA from 5. 4
jltmolg'1 to 8.22 umolg", color value 3 (purple) represents PA from 8.6 umolg'1
to 12.8 umolg", and color value 4 (dark purple) covers samples with the highest
PA of more than 13.2 umolg". For pungency evaluation, the yellow color
indicates mildness while the dark color suggests high pungency. This would be
more useful for breeders who are more interested in knowing if an onion
selection is high or low in PA rather than knowing the exact amount. In addition,
it only takes about 25 min to run a group of 25 samples, which makes this

method much more time-efﬁcient.

The result of test for background PA, which is unrelated to pungency
development in the bulb tissue, shows a very low level for background PA. The

same result was previously reported by Yoo et al., ( 1993 ). The background

 

 

 

101

 

 

 

 

 

Color value and PA content
6
5 _
4 - <>
(1)
3
.9
8
2 ‘ Y=0.506+0.21X
(:2 = 0.83")
1 _
0 I I i I
0 5 10 15 20 25
PA (pmol.g")

Figure 11. Relationship between the MSU color value procedure and the
standard method (Schiwmmer and Weston, 1961 ). Color value was visually
estimated with 4 scales: 1 = yellow, 2 = light purple, 3 = purple, and 4 = dark
purple.

 

 

102
PA does not seem to signiﬁcantly affect the result, and probably can be ignored

in breeding programs.
Selection of Two Onion Populations for PA Usingﬁthe MSU PA Procedure

Progeny from the selections for high- and low-PA parental bulbs from the
two inbred populations 9161 A and 9161 B showed a changed PA content
compared to the parents (Table 18). High-PA crosses produced progeny, 9506,
with a high PA content (14.2 umolg") and low-PA crosses produced progeny,
9502, with a low PA content (8.4 uumol.g"). The two new inbred populations
differ signiﬁcantly from each other and the regular selection 9511. These results
suggests that the color value procedure could be a reliable tool for a breeding

program.

There are some limitations to this procedure. (1) Uniformity of bulb size must
be highly considered when comparing different onion varieties since the surface
area of the onion slices determine how much PA will develop and be released;
(2) A color value does not represent an exact PA level. It only refers to a range
of PA in the tissue; (3) Development of PA with DNPH requires a warm
temperature. A hot water bath is involved in the SW system to maintain a warm
condition for the reaction. Although, this method does not require heating of the
samples when the room temperature is around 28 C, low temperature conditions

would slow the reaction and would not be suitable for this procedure. It would ,

 

 

103

Table 18. Pyruvic acid content (pmol.g".) of onion bulbs from inbred parents by
using the MSU color value procedure.

 

 

 

Lines Crosses PA (pmol.g")

P1 y P2 offspringx
9502 9161 A x 9161 B low low 8.4 az
9506 9161 A x 91618 high high 14.2 c
9511 9161 A x 9161 B regular w regular 9.62 b

 

’ PA values followed by different letters are statistically different (P5005). Data
are means of 6 bulbs.

Y Parental PA, estimated by the new procedure.

X Offspring PA, estimated by Schwimmer and Weston method (Schwimmer and
Weston, 1961).

W Regular selection without regard to PA levels.

104
therefore, be suggested not to use this method if the temperature during the

testing would drop below 25 C.

This method is an alternative to other more complicated systems for onion
pungency measurement. Although, it does not result in an exact quantiﬁcation of
PA in onions, it does lead to a quick estimate of onion pungency levels with
simple techniques and conventional tools, which often is desired by onion

breeders and handlers.

 

 

GENERAL DISCUSSIONS

Improvement of onion ﬂavor is a great challenge to onion researchers
because there are so many determining factors involved. From the breeding
point of view, lower pungency and higher sugar levels in onions may be possible
through single-bulb selection and crosses between parent lines high in sugar
and low in PA. Environmental factors, however, may not be controlled easily and
can offset breeding efforts. Most existing sweet type onions are grown in a
particular area where soil, weather, and fertilization factors together provide an
optimum environment for the onion to fully develop its unique ‘mild taste’
potential. Unfortunately, little research work has been directed toward the

determination of how to select and breed storable onions with milder ﬂavor.

A few studies including ours have shown that production of onion sugars
and PA are genetically determined by dominant-additive genes. To develop
milder ﬂavored inbred lines, selection can be conducted effectively within inbred
populations for high sugar and low PA bulbs for seed parent use. The means for
higher sugar content and lower PA level would be improved, within limits,
generation by generation. Signiﬁcantly increased levels of sugars or soluble
solids and reduced PA would be seen after one or two breeding cycles,

especially if there is considerable variation within the population.

105

 

106

Considering onion ﬂavor traits of sweetness and pungency to be additive
gene-controlled traits, using the crosses of 'high x low’ or ‘low x high’ parent bulb
combinations would not be a suggested approach for higher SS and lower PA.
The main use of such crosses would be to bring in other desirable characteristics

that are lacking in a particular germplasm.

Measurements of onion sugars and pungency must be efﬁcient and reliable.
Any successful improvement on the analysis procedures and new tools for bulb
screening would be very beneﬁcial to the breeding process. The new method of
rapid PA determination developed through this study adds a new option for onion
breeders and would make the screening and ﬂavor evaluation processes more

efﬁcient.

These studies show that Michigan onions have potential for ﬂavor quality
improvement. Compared to the sweet onions, ‘Vidalia’, ‘Texas Super Sweet’ and
‘Walla Walla’, produced in southern and western states, Michigan onions are
more pungent, with higher PA levels and are storable. They are actually not low
in sugar content. This study suggests that future breeding work to develop
sweeter tasting Michigan onions needs more research efforts on redusing PA

level rather than on raising the sugar level.

In this study the experimental MSU hybrid 2518-1 and the commercial
cultivar Sweet Sandwich produced new encouraging results. New Michigan

sweet onions, therefore, may be produced in the future from the parent lines that

 

107

have been selected for low PA. More tests are needed to see how stable such
hybrids can be in terms of SS and PA levels, and how tolerable they will be to

storage effects after bulb pungency is reduced.

The development of high-sugar and low-PA inbred lines might result in
higher disease susceptibility and low storability. Low pungency may be
accompanied by low tolerance to pathogen attacks (Lin et al., 1995). In this
study, some of the newly reselected material from the developed inbreds,
including 5718 B and 826 B showed a large number of rotted bulbs during winter
storage and suffered considerable ﬁeld disease attacks. It is not known if these
problems are primarily due to inbreeding depression, causing susceptibility, or
low pungency and high sugar. Further studies on the relationship between
onion ﬂavor and disease tolerance will be necessary to answer these questions.
Currently, from the present data, a PA level of 6.0-7.0 umolg'1 possibly would be

a level at which severe storage losses could be avoided.

The inheritance of onion pungency and soluble solids can be explained in
most cases with a simple additive-dominance model (Lin et al., 1995; Simon,
1995). In the development of a breeding line, selection for a dominant allele in a
homozygous condition can be hampered by masking of any recessive alleles.
The effect of selection on the genes and genotypic frequency in a population
would not be remarkably high over a short breeding time span. The development

of gene markers with the use of DNA ﬁnger-printing techniques, however, could

108

improve effectiveness for selection and reduce the time of development of

desirable inbreds.

LIST OF REFERENCES

Astly, K. 1989. Conservation of genetic resources, in Onions and Allied Crops.
ed Rabinowitch, H. D., & Brewster, J. L., V.1, p.181, CRC Press, Florida.

Bacon, J.S.D. 1957. The water-soluble carbohydrates of onion, Allium cepa L..
Biochem. J. 67, p5.

Bajai, K. L., G. Kaur, J . Singh and S.P.S. Gill. 1980. Chemical evaluation of
some important varieties of onion (Allium cepa L.). Qual. Plant. Plant
Foods Hum. Nutr. 30: 117-122.

Barham, W. S. and H. M. Munger. 1950. The stability of male sterility in onions.
Proc. Am. Soc. Hortic. Sci. 56: 401.

Bedford, L. V. 1984. Dry matter and pungency tests on British grown onions. J.
Natl. Inst. Agric. Bot. 16: 58.

Bennet, E. 1945. A note on the presence of pyruvic acid in Ebenezer onions.
Plant Physiol. 20:461-463.

Birth, G. S. and G. G. Dull. 1985. Nondestructive spectrophotometric
determination of dry matter in onion. J. Amer. Soc. Hort. Sci. 1102297 -
303.

Block, E., R. E. Penn and L. K. Revelle. 1979. Structure and origin of the onion
lacrimatory, A Microwave study. J. Am. Che. Soc. 101: 2200.

Boelens, M., P. de Valois, H. J. Wobber and AV. Gen. 1971. Volatile ﬂavor
compounds from onion. J. Agric.. Food Chem. 19, 984-990.

Bose, S. and A. N. Shrivastava. 1961. Soluble carbohydrates from onion (Allium
cepa L.). Sci. Cult. 27:253.

Brodnitz, M. H. and J. V. Pascale. 1971. Thiopropanal S-oxide: a lacrimatory in
onions. J. Agric. Food Chem. 19. p. 269.

Campbell, W. F,, S. K .Cotner and B. M. Polock. 1968. Preliminary analysis of
the onion seed production problem in 1966 growing season.
HortSciences. 3, 40-43.

Carson, JR. and FF. Wong. 1959. A color reaction for thiosulﬁnates. Nature.
183 : 1673.

109

110

Carson, J.R., R.E. Lundin and TM. Lukes. 1966. The conﬁguration of (+) S-(1-
propenyI)-L-cysteine-S-oxide from Allium cepa. J. Org. Chem. 31 :1634.

Clarke, A. E. and L. H. Poland. 1949. The amount of self-pollination in male
sterile onion lines. Proc. Am. Hort. Sci. 53, 299-301.

Cooke, R. J., T. M., Smith and A. G Norgan. 1986. The use of isoelectric
focusing of seed seterases to classify cultivars of Allium cepa L.(onion).
J. Natl. Inst. Agric. Bot. 17, 209-221.

Craig, R. 1968. Implications of the new genetics in horticultural plant breeding.
Hort. Sci. V.Ill, 242-246.

Darbyshire, B. 1978. Changes in carbohydrate content of onion bulbs stored for
various times at different temperatures. J. Hort. Sci. 53: 195—201.

Darbyshire, B. and R. J. Henry. 1978. The distribution of fructans in onions. New
Phytol. 81. 29-33.

Darbyshire, B. and R. J. Henry. 1979. The association of fructans with high
percentage dry weight in onion cultivars suitable for dehydrating. J. Sci.
Food Agric. 30, p.1035.

Darbyshire, B. and B. T. Steer. 1989. Carbohydrate biochemistry. in Onions and
Allied Crops. ed. Rabinowitch, H. D., and J. L., Brewster. V.3, p2, 1989.
CRC Press, Florida.

De Miniac, M. 1970. Application of gas chromatography to the study of Allium
cepa bulb carbohydrates. C. R. Acad. Sci. Paris. 270,1583.

Fedewa, DJ. and S. J. Pscdna. 1995. Michigan agricultural statistics 1994.
Michigan Agricultural Statistics Service.

Dowker, B. D. 1989. Onion breeding. in the Onions and Allied Crops. ed
Rabinowitch, H. D., J. L. Brewster. V. 1, p.215, 1989. CRC Press,
Florida.

Erickson, H. T. and W. H. Gabelman. 1958. Potential value of inbred and F1
hybrid onion for seed production. Proc. Am. Soc. Horic. Sci. 71, 435-439.

Falconer, CS. 1981. Introduction to Quantitative genetic. 2d ed. Longman, New
York.

Foskett, R.L. and CE. Peterson. 1949. Relation of dry matter content to storage
quality in some onion varieties and hybrids. Proc. Amer. Soc. Hort. Sci.
55:314-318.

 

111

Freeman, G. G. And N. Mossadeghi. 1973. Studies on the relationship between
water regime and ﬂavor strength in watercress (Ron'ppa nasturtium-
aquaticum L. Hayek), cabbage (Crassica oIeracea capitata) and onion
(Allium cepa). J. Hortic. Sci. 48, 365-369.

Freeman, G. G. And N. Mossadeghi. 1970. Effect of sulphate nutrition on ﬂavor
components of onion. J. Sci. Food Agric. 21:610-615.

Freeman, G. G. and R. J. Whenham. 1976. Synthetic S-alk(en)yl-L-cysteine
sulfoxide Alliinase ﬁssion products: simulation of ﬂavor components of
Allium species. Phytochem. 15: 187-191.

Freeman, G.G. and R. J. Whenham. 1976. Effect of over winter storage at three
temperatures on the ﬂavor intensity of dry bulb onions. J. Sci. Food Agric.
27,p37.

Freeman, G.G. and R. Whenham. 1976. Onion storage at ambient temperature.
J. Sci. Food. Agri. 25, p517.

Grieg, WS. and CL. Marine. 1965. Onions and their processing potentials.
Michigan Agr. Expt. Sta. Res. Rpt. 14.

Hanelt, P. 1989. Taxonomy, Evolution, and History, in the Onions and Allied
Crops. ed. Rabinowitch, H. D.,and J. L., Brewster. V. 1, p.13. 1989. CRC
Press. Florida.

Hayman, BL 1958. The separation of epistatic from additive and dominance
variation in generation means. Heredity 12:371-390.

Hodge, J.E. and B. T. Hofreiter. 1962. Determination of reducing sugars and
carbohydrates. p. 380-394. In: R.L. Whistler and ML. Wolfram (eds.)
Methods in carbohydrate chemistry. Vol. 1. Academic. New York.

Jellinek, G. 1985. Aroma. In: Sensory evaluation of food. Ellis Honrvood Ltd.,
Chichester, England.

Jones, H .A. and A. E. Clarke. 1943. Inheritance of male sterility in the onion
and the production of hybrid seed. Proc. Am. Soc. Hort. Sci. 43:189-194.

Jones, H. and G. Davis. 1944. Inbreeding and heterosis and their relation to the
development of new varieties of onion. USDA Technical Bulletin. No.874.

Jones, HA. and CS. Bisson. 1934. Moisture content of different varieties of
onions. Proc. Amer. Soc. Hort. Sci. 31:165-168.

 

112

Jones, R. N. 1989. Cytogenetics. in the Onions and Allied Crops. ed.
Rabinowitch, H. D. & J. L .Brewster. V1 p. 200. 1989. CRC Press,
Florida.

Kadams, A.M. & C.C. Nwasike. 1986. Heritability and correlation studies of some
vegetative traits in Nigerian loca white onion. Plant Breeding. 97: 232-
236.

Kappes, E. M. 1981. Sugar-Pak I column. Product bulletin, Waters, p7.

Lancaster, J. E. and M.J. Boland. 1989. Flavor Biochemistry. in Onions and
Allied Crops, ed., Rabinowitch, H. D.and J. L. Brewster. v.3, p.36, 1989,
CRC Press, Florida.

Lancaster, J. E., B J. McCallion and M. L. Shaw. 1984. The levels of S-
alk(en)yl-L-cysteine sulphoxide during the growth of the onion (Allium
cepa L.). J. Agric. Food Sci. 35: 415.

Lederhouse, R. C., D.m. Caron and RA. Morse. 1972. Distribution and behavior
of honeybees on onion. Environ. Entomol. 1, 127.

Lin, M.W., J. F. Waston and J. R. Baggett. 1995. Inheritance of soluble solids
and pyruvic acid content of bulb onions. J. Amer. Soc. Hort. Sci.
120(1):119-122.

Love, J. M. 1994. The US onion industry, 1995-2000. Onion World. 09: 15-18.

Mahmud, l. and H. H. Kramer. 1951. Segregation for yield, height, and maturity
following a soybean cross. Agron J. 43:605-609.

Mann, L.K. and B.J. Hoyle. 1945. Use of the refractometer for selection of onion
bulbs high in dry matter for breeding. Proc. Amer. Soc. Hort. Sci. 46:285-
292.

McColIum, GO. 1968. Heritability and genetic correlation of soluble solids, bulb
size and shape in white Sweet Spanish onion. Can. J. Genet. Cyto.
10:508-514.

McColIum, G. D., 1973. Chromosome behavior and sterility of hybrid between
the common onion, Allium cepa, and the related wild A. oschaninii.
Euphytica. 23, p699.

McColIum, G. D. 1976. Onion and allies, Allium (Liliaoeae). ln Evolution of Crop
Plants. Simmonds, N. W. Ed. Longman. London. Chap. 53.

McColIum, G. D. 1980. Development of the amphidiploid of Allium galanthum x
A. cepa. J. Hered. 71, p445.

113

McGregor, S. E. 1976. Insect Pollination of Cultivated Crop Plants. Agriculture
Handbook. No. 496, US. Department of Agriculture, Wadhington, DC.
268.

MoCollum, G. 1966. Heritability and genetic correlation of some onion bulb
traits. J. Hered. 57:105-110.

Morgan, E.J. 1946. Pyruvic acid in the juice of onion (Allium cepa). Nature
157:512.

Ohsumi, C., T. Hayashi, K. Kubota and A. Kobayashi, 1993. Volatile ﬂavor
compound formed in an interspeciﬁc hybrid between onion and garlic. J.
Agric. and Food Chem. 41(10): 1808-1810.

Owen, E. W. 1961. The inheritance of dry matter in onion bulbs. MS. Thesis.
Univ. Idaho, Moscow, Idaho. 35pp.

Owen, J.H., J.C. Walker and MA. Stahamann. 1950. Pungency, color and
moisture supply in relation to disease resistance in the onion.
Phytopathology 40:292-297.

Pal, N. and N. Singh. 1988. Analysis of genetic architecture of pungency
(pyruvic acid) in onion. Plant Breeding Abstr, 19882379.

Peters, R. 1989. Seed production in onions and other Allium species. in Onions
and Allied Crops, ed Rabinowitch, H. D., & , J. L. Brewster. V.1. p.167,
1989, CRC Press, Florida.

Peterson, C. E., P. W. Simon and L. A. Ellerbrock. 1986. 'Sweet Sandwich'
onion. HortSci. 21 (6): 1466-1468.

Pike, L. M. 1986. Onion Breeding, In Breeding Vegetable crops. Bassett,
M.j.,Ed., AVl Publishing, Westport, CT, 357.

Platenius, H. 1935. A method for estimating the volatile sulphur content and
pungency of onions. J. Agric. Res. 51: 847.

Platenius, H. 1944. Factors affecting onion pungency. J. Agr. Res. 62:371-380.

Rabinowitch, H. D.and J. L. Brewster. 1989. Onions and Allied Crops. V.1, p.13.
CRC Press, Florida.

Randle, W. M., 1992. Sulfur nutrition affects nonstructural water-soluble
carbohydrates in onion germplasm. HortScience. 27:52-55.

Randle, WM. 1992. Sampling procedures to estimate ﬂavor potential in onion.
HortSceince 27:1116-1117.

 

114

Randle, W. M., and ML. Bussard. 1993. Pungency and sugars of short-day
onions as affected by sulfur nutrition. J. Amer. Soc. Hort. Sci. 118(5):
766-770.

Randle, W., M. BUssard and D. Warnock. 1993. Ontogeny and sulfur fertility
affect leaf sulfur in short-day onions. J. Amer. Soc. Hort. Sci. 118: 762-
765.

Rutherford, R. P. and R. Whittle. 1984. Methods of predicting long term storage
of onions. J. Hortic. Sci. 59: 537-541.

Saguy, M., C. H. Mannheim, and Y. Peleg. 1970. Estimation of volatile onion
aroma and ﬂavor compounds. J. Food Technol. 5, p165.

Schwimmer, S. and W.J. Weston. 1961. Enzymatic development of pyruvic acid
in onion as a measure of pungency. J. Agric. Food Chem. 9:301-304.

Schwimmer, S. and D.G. Guadagni. 1962. Relation between olfactory threshold
concentration and pyruvic acid content of onion juice. J. Food Sci. 27, 94.

Schwimmer, S. 1967. Development of a bitter substance in onion juice. Food
Tech., 21:292.

Schwimmer, S. 1969. Characterization of s-propenyl-cysteine sulfoxide as the
principal endogenous substrate of L-cysteine sulfoxide lyase in onion.
Arch. Biochem. Biophys. 130:312.

Simon, PW. 1995. Genetic analysis of pungency and soluble soids in long-
storage onions. Euphytica 82: 1-8.

Sinha, N. K., D.E. Guyer, D. A. Gage and C. T. Lira. 1992. Supercritical carbon
dioxide extraction of onion ﬂavors and their analysis by gas
chromatography-mass spectrometry. J.Agric. Food Chem. 40 (5): 842-
845.

Smittle, D. A. 1988. Evaluation of storage methods for "Gannex" onions. Hort,
Sci. 113 (6): 877-880.

Steer, B. T. and B. Darbyshire. 1979. Some aspects of carbon metabolism and
translocation in onions. New Phytol. 82, 59.

Steer, B. T. 1982. The effect of growth temperature on dry weight and
carbohydrate content of Onin (Allium cepa L. Cv. Creamfold) bulbs.
Aust. J. Agric. Res. 33:559.

Suzuki, M and J. A. Cutcliffe. 1989. Fructans in onion bulbs in relation to storage
life. Can. J. Plant Sci. 69: 1327-1333.

115

Tobkin, H. E., and M. Mazelis. 1979. Alliinase: puriﬁcation and characterization
of the homogeneous enzyme from onion bulbs. Arch. Biochem. Biophys.
193: 150.

USDA. 1991. Agricultural Statistics. Washington, DC. P. 156 -157.

Van der Meer, Q. P. and J. L. Van Bennekom. 1968. Research on pollen
distribution in onion seed ﬁelds. Euphytica. 17, p216.

Vavilov, N. I. 1951. The origin, variation, immunity and breeding of cultivated
plants. Chron. Bot., 13, p1.

Vosa, C. G. 1976. Heterochromatic patterns in Allium. the Cepa group and its
allies. Heredity. 36. p.383.

Wall, MM. and J.N. Corgan. 1992. Relationship between pyruvate analysis and
ﬂavor perception for onion pungency determination. HortScience 27:1029
- 1030.

Warid, WA 1952. Inheritance studies in the onion. Ph.D Thesis, Luoisiana
State University. Louisiana. p.221.

Whitaker, J. 1976. Development of ﬂavor, odor and pungency in onion and
garlic. Adv. Food Res. 22: 73 -133.

Wilson, 0., J. W . Cross, and W. J. Lucas. 1985. Sugar uptake into Allium cepa
leaf tissue: an integrated approach. Planta. 164, 227.

Yamaguchi, M. 1983. World Vegetables. AVI Publishing. Westport. CT. Pp. 184 -
206.

Yamanishi, T. and K. Orioka. 1955. Chemical studies on the change in ﬂavor
and taste of onions by boiling. J. Home Econ., 6, p45.

Yoo, K.S., L.M. Pike and BK. Hamilton. 1995. A simpliﬁed pyruvic acid analysis
suitable for onion breeding programs. HortScience 30(6):1306.

 

 

   

"Iillllllllllil‘lllllllli