I
!

 

A STUDY OF THE SOLIDS,
SUGAR AND SWEETNESS CONTENT OF
SELECTED INBRED CARROT LINES
AND THEIR HYBRIDS

“tests for ”19 Degree oI M. S.

MICHIGAN STATE UNIVERSITY
Susan Ann Engstrom Bittenbender

I975

JHESI

0]

',_.H__.._.—.__ ,

,_+__--.._- -_.—..

Yo.“ Y..'-’;.;“:ch§.~ . .. -. , . .. - ‘

. ‘ v- ‘MLA&.Q' g" h 'l x, f . -
'.- ,p

. ' .

F 13' ' ‘ ‘ T" ‘a 9 v‘$, (a
a ‘N I

. ~ r k .
' v

' . ?
'l V m -l ‘
‘IV-' -. ‘1 t‘. '. Ii 5'. \

.7
t ,

x:
32.343

9
- .“‘
0‘

I

,,
'J
9"
‘g I! 1 5“,” ,. —,f- ’A
’ U1“: “2'55”"?
.-

3?“ _-
" ' ' "-' w» aura-u

bug 9
A»
0
)
J
u-
{‘ .r‘
“,3?
w"
J
’ -. £1;

 

 

ABSTRACT

A STUDY OF THE SOLIDS, SUGAR AND SWEETNESS
CONTENT OF SELECTED INBRED CARROT LINES
AND THEIR HYBRIDS
By

Susan Ann Engstrom Bittenbender

Inbred carrot lines and their single cross hybrids
developed at Michigan State University were examined for
differences in solids, sugar and sweetness content. Gas-
liquid Chromatography was used to quantify the sucrose,
alpha-glucose, beta—glucose and fructose found to occur in
the carrot roots. A taste panel was used to evaluate
relative sweetness. Total and soluble solids content was
also determined.

Among the inbred and hybrid lines, total and soluble
solids and total sugars were mutually and positively
correlated. Among the inbreds total sugar varied directly
with each individual sugar. Among the hybrids total sugar
was related only to sucrose. Correlations involving sweet-
ness were evident among the hybrids.

The inbred MSU 6000 contained more solids and sugars
than the other inbreds. Hybrid solids and sugars content
suggested heritability of these traits. Other parent lines

also appeared to influence the sugar and solids content.

Susan Ann Engstrom Bittenbender

The data suggested that both nuclear and cytoplasmic control
and interactions of these controls may be responsible for

sugar content in the hybrids.

A STUDY OF THE SOLIDS, SUGAR AND SWEETNESS
CONTENT OF SELECTED INBRED CARROT LINES

AND THEIR HYBRIDS

By

Susan Ann Engstrom Bittenbender

A THESIS

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

MASTER OF SCIENCE

Department of Horticulture

1975

to my husband
Skip
He has provided me with steadfast support as

a fellow horticulturalist and as a friend.

ii

ACKNOWLEDGMENTS

Appreciation is extended to the author's committee
chairman, Dr. Grant Vest, who has provided guidance and
appraisal in data reduction and manuscript preparation.

Thanks is extended to Dr. Larry Baker who provided the
plant material for this study and who served as major advisor
prior to his sabbatic leave.

Dr. C. L. Bedford is acknowledged for his service on the
guidance committee and for the use of his laboratory space.

Dr. Kenneth Sink is also thanked for his service on the
guidance committee.

The author also offers her thanks to her colleagues for
their contributions of advice and labor, H. C. Bittenbender,
E. A. Mielke, N. Jennings, B. Stergios, M. Byrne, M. Uebersax,
K. Uebersax, Y. Lee and B. Vibbert.

A special thanks is extended to Dr. Matthew Zabik for the
use of the chromatograph and computer equipment.

The author wishes to thank the Rackham Foundation for the

provision of funds to support this and related research areas.

iii

TABLE

List of Tables
Introduction . .
Literature Review
Materials and Methods
Results and Discussion
Summary and Conclusions
Recommendations

Literature Cited

OF CONTENTS

iv

Page

CDUOl-J

ll
26
29
32

Table

LIST OF TABLES

Page

Breeding Material Codes . . . . . . . . . 13
Mean Sugar and Solids Content and Sweetness

Values Associated with Roots of Six Inbred

Carrot Lines . . . . . . . . . . . . 1A
Simple Correlations Among Measurements Made upon

Roots of Six Inbred Carrot Lines and Twenty—

five Hybrid Combinations . . . . . . . . 16
Mean Sugar and Solids Content and Sweetness

Values Associated with Roots of Twenty-five F1

Carrot Hybrids . . . . . . . . . . . 18
General Combining Ability Means of Six Inbred

Carrot Lines Based on Total and Reducing Sugar

Content . . . . . . . . . . . . . 22

INTRODUCTION

Michigan ranks third in crop value among the top
eighteen carrot producing states (Whitaker et al. 1970).
This is an improvement over the situation that existed about
twenty-five years ago - at which time Michigan had relatively
few acres in carrot production and was not producing a high
quality root (Carlton 1958). Recent improvements can be
attributed in part to successful trials and releases of
lines and hybrids developed at Michigan State University.

Cultivar improvementrms been characterized by uniform
shape, size and interior and exterior root color. In cross;
section the improved carrot root exhibits higher phloem to
xylem area ratios, uniform orange color and the absence of
yellow or watery cores. A Sweeter taste has also been
associated with the newer genotypes, particularly among the
hybrids.

Compared with other vegetables, the carrot is high in
food production efficiency (MacGillivray 33 al. 1942).
Standard cultivars of carrots in 100 gram portions supply
200% of the adult recommended daily allowance (RDA) of
provitamin A and percentages of other nutrient allowances
slightly below or comparable to many common vegetables (Watt

and Merril 1963). A recent study on advanced breeding lines

showed that 256 to 330% of the adult RDA of provitamin A may
be obtained from a 55 gram fresh portion (Leveille 33 al.
197“). Thus it is reasonable to attempt to widen the accept-
ance of this vegetable in the human diet.

Carrots now rank ninth in annual commercial production
value on a nation-wide basis (Whitaker at _l. 1970). This
may be partially explained by consumer preference, although
it also reflects the relative low cost of carrots to the
consumer.

Improvement in carrot taste properties is one method of
enhancing the acceptability of carrots. Carrot taste is
characterized by bitter, oily and sweet components in a
relatively bland flavor background. Some of the compounds
responsible for these properties have been identified and
assayed (Sondheimer 1957, Otsuka and Take 1969, Heatherbell
gt al. 1971 and Alabran and Mabrouk 1973). Sweetness is
particularly importanttp cultivar improvement, as demon—
strated by the fact that small, sweet carrots command a
higher price on the market.

Accordingly this study was undertaken to determine
differences in total and/or individual sugars between inbred
carrot lines and among selected Fl hybrids of these lines
developed at Michigan State University. Differences in sugar

content could be used as a criterion of further selection

for sweet tasting carrots.

LITERATURE REVIEW

The sweeter naturally occurring sugars are fructose,
sucrose and glucose. When sucrose is given a sweetening
power value of 100, fructose and glucose have the relative
values 173 and 7“, respectively (Guthrie 1971). The sweet-
ness of fructose, however, will vary depending on the
presence of tautomeric isomers (Shallenberger 1971).
Maltose has a sweetening value of 33 when sucrose is set at
100 (Green 1971).

Carrots contain both reducing and nonreducing sugars.
The reducing sugars range from 0.1 to 5% and the nonreducing
sugars range from 2.5 to 8.5% on a fresh weight basis
(Hasselbring 1927, Werner 1941, Carlton and Peterson 1963,
Sistrunk gt a1. 1967, Otsuka and Take 1969, Engel' gt 3;.
1970, Alabran and Mabrouk 1973 and Phan and Hsu 1973).
Therefore total sugars vary from 2.6 to 13.5%.

The ratio of nonreducing sugars to reducing sugars has
been reported as 2:1 for both xylem and phloem of harvested
roots of 'Imperator 11' (Phan and Hsu 1973). Earlier work,
however, showed a higher ratio of Azl in the phloem and
approximately 2:1 in the xylem of mature 'Nantes' and 'Red-

Cored Chantenay' roots (Werner 19A1).

The sugar content of carrot roots has been shown to
depend upon the stage of root maturity (Yamaguchi gt gt.

3 1952 and Phan and Hsu 1973), morphological source of the
tissue (Werner 19U1 and Phan and Hsu 1973L cultivar (Werner
19N1, Carlton and Peterson 1963, Sistrunk gt gt. 1967 and
Kraut 197A) and environmental location (Leveille gt gt.
197“).

Individual mono— and disaccharides which comprise the
reducing and nonreducing sugars have been reported. Alabran
and Mabrouk (1973) determined by gas-liquid chromatography
that an 'Imperator' strain of carrots contained 3.4% sucrose,
1.5% alpha-glucose, 1.9% beta-glucose, 1.1% fructose and
0.A% of one unknown sugar. Otsuka and Take (1969) found that
carrots ofénlunnamed cultivar contained 2.A% sucrose, 2.1%
maltose and 0.4% glucose. Moderate amounts of arabinose
have been reported in addition to sucrose, fructose and
glucose in Egyptian—grown 'Chantenay' carrots (Wali and
Hassan 1966).

A postharvest decline in total sugar content was
reported for 'Imperator 11' (Phan gt gt. 1973) and for
'Nantes' and 'Chantenay' carrots (Werner 19A1). Phan gt gt.
(1973) did note a rise in total sugars at the sixteenth week
due to the appearance of raffinose. The ratio of nonreducing
to reducing sugars has also been shown to decrease following
harvest and during storage (Platenius 1932, Werner, 19A1, and

Phan gt gt. 1973).

It has been suggested that invertase activity is partly
responsible for the postharvest changes in sugar content
(Platenius 1932 and Phan gt gt. 1973). Increased activity
of the tricarboxylic acid cycle has also been indicated
(Phan gt gt. 1973). This suggests metabolic conversion of
the sugar carbon skeleton. Inversion of sucrose alone may
not impair and may even improve sweet taste properties during
storage since the resulting monosaccharides contribute to
sweetness. TCA activity, however, could remove a percentage
of the sugars altogether. Deterioration or preservation of
sweetness would depend on the relative magnitudes of these
two activities.

It is suggested that sweet taste properties may be best
preserved in roots initially containing relatively high
levels of sucrose and/or fructose. Such roots would be more
salable immediately after harvest and following periods of
cold storage.

Hasselbring (1927) found1x>cu1tivar of carrot to
predominate in sugar content. Carlton (1958) reported
greater variability within than among three standard
cultivars; 'Nantes', 'Imperator' (two seed lots) and 'Gold
Pak'. The cultivars did differ, however, in reducing sugar
content. In an inbreeding experiment Carlton and Peterson
(1963) were successful in reducing within cultivar variation.
They reported Significant differences in sugar content among
the S progeny of a 'Nantes' and of three 'Long Chantenay'

2

seed lots. The roots were Chosen individually according to

their differing total sugar content for the S generation

1

and according to percent soluble solids for the S genera-

2
tion. More recently Kraut (1974) determined Significant
differences among advanced carrot breeding lines.

Carlton and Peterson (1963) showed a high positive
correlation between soluble solids and sugars on an indi-
vidual root basis and proposed that refractometer readings
could be used to screen for high sugar roots. This technique
would require destruction of part of the root to be used for
seed. Since stecklings are commonly stored intact and are
'cut' for selection in the spring or after vernalization is
satisfied, such treatment could cause a rise in the suscepti—
bility to fungal and bacterial attacks.

Another screening technique has been demonstrated by
Bassett (197A). Brine solutions were used to determine the
relative specific gravities of individual carrot roots.
Specific gravity has been shown to correlate significantly
with sugar content (Bassett 1973). This approach does not
damage the root and may prove valuable for screening material
on an individual root basis or by pedigree.

The present study did not deal with these screening
methods but was designed to assess the status of some
currently available advanced carrot lines in terms of their
sugar content pg; gg. The sugar assay gives direct informa-
tion about sugar content on a mean and single root basis.

This may be important since the correlative methods, although

proven quite reliable, have r values varying between 0.75 and

0.96 (Carlton and Peterson 1963 and Bassett 1973). Small
differences in breeding material may not be detected easily
by these methods. Caution in analysis of and selection
between genotypes, such as in a diallel study, would be
required since it has been established that the regression
relationship of the above-mentioned traits varied with
genotype (Carlton and Peterson 1963).

Analysis using gas chromatography is particularly use—
ful since the reducing sugar fructose can be quantified
directly. High fructose concentrations are desirable from
a sweetness standpoint.

Previous to the early 1960's standard methods of
carbohydrate analysis consisted of titrimetric procedures,
paper and thin layer chromatography and polarimetric deter-
minations. The formation of volatile trimethylsilyl sugar
derivatives (Sweeley gt gt. 1963) has permitted a direct and
more accurate look at the individual sugars in tissue systems
via a single analysis. Comprehensive reviews and reports
are available on the gas-liquid chromatography of carbohy-
drates (Sweeley gt gt. 1966, Holligan and Drew 1970,

Holligan 1970, Birch 1973 and Dutton 1973).

MATERIALS AND METHODS

The plant material consisted of Six inbred lines and
twenty-five hybrids developed at Michigan State University.
The carrots were harvested at Grant, Michigan on 16 October,
1974. Six randomly selected, representative roots from each
line were stored at NOC for three weeks before being removed
to 1°C lockers. Over a six day period they were removed
from storage for sample preparation. Roots were sliced
cross—sectionally into three equal lengths. The top was
returned to storage for subsequent seed production, the
center was reserved for subsequent sugar and solids analysis
and the tip portion was tested organoleptically by a six
member taste panel.

The center portion was Sliced into two millimeter
discs. Thirty grams of discs selected randomly from
each portion were weighed and frozen at -10°C. The
samples were lyophilized in an automatic Virtis unit at a
platen temperature of 60°C, a condensor temperature of -60°C
and a vacuum of less than 5.0 pm. The freeze—dried samples
were weighed, ground through a #AO mesh screen in a Wiley
Mill and collected in four ounce jars. These Jars were kept

under desiccation at —10°C. Refreezing was chosen since

preliminary observations with test material indicated that
room temperature storage resulted in loss of sugars.

Total solids were determined by taking the weight of
lyOphilized discs as a percentage of the recorded fresh
weight. Percent soluble solids were determined from a
reconstituted carrot slurry of 6.67% dry matter. One gram
of dry carrot powder was allowed to stand in 1H milliliters
of water for five minutes with periodic agitation to insure
complete hydration. The slurry was filtered through 15
centimeter Whatman #5 filter paper. Three milliliters of
the filtrate were collected and capped in one dram screw
cap vials. Readings were taken from a Valentine Precision
Model 350A refractometer. The following formula was used to
calculate theoretical soluble solids values for fresh weight
of each root:

(percent total solids) x [(measured soluble solids)/
(6.67fl = soluble solids of fresh tissue

Preliminary experiments on carbohydrate content of the
dried, ground tissue indicated that 98°C distilled water
(Carlton 1958), 98°C 80% ethanol (AOAC 1975) and room-
temperature distilled water (Waldron gt gt. 19A8) can be
used interchangably as solvents. A five minute magnetic
stirring extraction in room—temperature water proved as
effective as the longer official procedures for extracting
carbohydrates. The stirring procedure was chosen because of

its convenience and speed.

10

Preliminary use of the gas—liquid chromatograph and the
mass spectrometer indicated that no contaminating or
cochromatographing materials were present in the unpurified
carrot filtrate. The sample preparation procedure used
follows:

One gram of carrot powder was weighed into a 150 milli-
liter beaker. One—half gram of CaCO3 and 100 milliliters of
distilled water were added. The solution was placed over a
magnetic stirrer for five minutes and filtered through a
15 centimeter Whatman #5 filter paper. One milliliter
of the filtrate was pipetted into a one—half dram vial. One
milliliter of a known concentration of internal standard was
added and the vial was placed under 0.07 atmospheres at
50°C for 20 hours. The dried samples were stored in glass
desiccators.

Trimethylsilyl derivatives were prepared using Tri-sil-
'Z' (Pierce Chemical Co.) and allowed to stand for eight hours
prior to injection into the chromatograph. A Beckman GC A
chromatograph equipped with a flame ionization detector was
used to quantify the sugars. A 1.83 meter glass column with
a 6.25 millimeter outer diameter and a 2.00 millimeter inner
diameter was packed with 3% SE-30 on Chromosorb WHP 80/100
mesh (Supelco, Inc.). This gave satisfactory resolution
among sugar and internal standard (hexa + (0 - TMS) -
sorbitol) peaks. A temperature program permitted elution of
the monosaccharide and sorbitol derivatives within ten minutes
at 185°C and elution of the sucrose derivative at 255°C after

five minutes.

RESULTS AND DISCUSSION

Preliminary analysis of the dried carrot tissue
identified the sugars fructose, alpha—glucose, beta-glucose
and sucrose. Their respective retention times were 5.0
minutes, 6.5 minutes, 9.7 minutes and 20.2 minutes. The
inbred lines contained on the average 1.20% fructose, 0.AO%
alpha-glucose, 0.68% beta-glucose and A.5A% sucrose. The
hybrids contained 0.91% fructose, 0.32% alpha-glucose, 0.5A%
beta-glucose and 5.38% sucrose. The alpha and beta anomers
of glucose were combined for the subsequent statistical
analyses Since chemical equilibrium during the extraction
and trimethylsilylation invariably occurred (Sweeley gt gt.
1966).

Previously unidentified peaks eluted intermittently
during the course of analysis. The first occurred on the
trailing solvent peak at 2.0 minutes. It has been tentatively
identified as arabinose.

The second peak appeared as an ascending shoulder on the
fructose peak. Mass spectrometry analysis revealed that only
the TMS-derivative or isomeric forms of this derivative were
present. Two very small peaks followed the main fructose
peak in original and sample fructose preparations. Ellis

(1969) reported that this fructose derivative when dissolved

ll

12

in hexamethyldisiloxane (HMDSO) will deliver four peaks due
to anomeric and ring isomer formations. The relative
retention times reported for these peaks corresponded well
with those observed in this study.

HMDSO is a reaction by-product of sugar derivatization
in the presence of water. It is likely that HMDSO was formed
in the carrot sample Sirups. It's probable formation is
offered as an explanation for these peaks. Individually
these peaks interfered negligibly or not at all with sugar
quantification.

Code numbers used for inbred and hybrid lines in subse-
quent tables are presented in Table 1.

Table 2 contains means of the sugar and solids measure—
ments made upon the six inbred lines. The inbreds differed
significantly in total solids, soluble solids, total sugar,
fructose and glucose content. These differences were due to
the line MSU 6000 which contained the largest amounts of
these components. The other five inbreds did not differ
significantly from each other.

Similar results were reported by Kraut (1974) where MSU
6000 contained higher total sugar than did MSU 5986 in a 1972
Texas planting and more than MSU 5931 in a 1973 Texas plant-
ing. Kraut (197“) also showed that MSU 6000 was higher in
soluble solids content than MSU 872 in a 1973 Idaho planting.

In Table 2 the order of magnitude of means prevails
rather consistently for all measurements except sucrose and

sweetness. Correlation data among these measurements are

13

Table 1. Breeding Material Codes.

I
1

Code Identification

 

Parents
6000 . . . . . . . . . . . . . . MSU 6000
5986 . . . . . . . . . . . . . . MSU 5986
872 . . . . . . . . . . . . . . MSU 872
1302 . . . . . . . . . . . . . . MSU 1302
9541 . . . . . . . . . . . . . . MSU 95Nl
5931 . . . . . . . . . . . . . . MSU 5931

Hybridsz
6000 x 872 . . . . . . . . . . . . MSU 6000

pollinated by
MSU 872

 

ZOnly one code example is given for the hybrids. All other
hybrid codes follow this pattern.

.Ho>ma mo.o
onp pm mcme mcoEm mmocDMDQMHU pCHOHMficmHm ocHEpmpmo op cow: was pmmp 0mm m.mmxsen

.oamom oa.op o o co UOCHEMODDU who: mosam> mmOCpmmzmz

.OSmmfip nmmpm w\wE mm popcomopa Ohm mmzam> pawSm Ha<x

 

coapmfipm> mo

14

 

Rmm.mm Rmm.om gam.mm goo.mfi gmm.mfi Rmo.fim Rma.m mpcmHOHoumoo
:m.m mm.oa mo.mfi mz.m: Am.mm ms.m :m.mH cam: camps
mm.m n mm.m n mm.p mm.m: n mm.mm n mm.m n em.HH Hmmm
mm.m n mm.m n .Hg.m om.:: n mm.mm pm m:.m n :A.HH Hzmm
om.m n 00.0H n mo.oa S©.m: p om.mm a mm.m n mo.mH mama
mm.m n m:.oa pm mm.HH mm.m: n mm.m© gm mH.OH p m:.mH mgm
so.m p HS.OH p HO.HH Sw.m: n mm.nm pm Hm.m p wm.mH mmmm
oo.m m aw.wa a mH.mm Hm.m: m mm.~m m S:.NH Na mo.ma oooo

NAmmOCpmmzm mmoosac mmOpozhm mmopozm xmmwww MWMWWWW QWWWWWM ommwwwmm

 

 

.mmcfiq pohmmo UmhncH Nam no

mpoom Qua: ompmfioomm¢ modam> mmoCummzm paw pampcoo moHHom pew pawsm :moz .m canoe

15

presented in Table 3. Total solids, soluble solids and

total sugars all showed significant positive correlations
with each other. This stands in agreement with previous
reports on carrot tissue (Werner 1941, Carlton and Peterson
1963, Bassett 1973 and Bassett 1974). Of the individual
sugars, fructose and glucose correlated more closely than
sucrose with the total and soluble solids. This pattern of
correlation is also evident when the individual sugars are
compared with total sugar. This is of importance since total
and soluble solids and/or total sugars may be used as a basis
for selection without the expense of any one of the sugars
involved. This finding is in contrast to that reported by
Carlton and Peterson (1963) who showed that total sugar, as
well as total and soluble solids and sucrose, levels were
inversely correlated with reducing sugar content in 'Long
Chantenay' carrots.

Among the individual sugars reported here the relation-
ships are apparently random with the exception of fructose
and glucose. These two sugars Show a strong positive
correlation of 0.97 (Table 3).

None of the measurements correlated with the panel
sweetness ratings at the 95% confidence level. However,
total sugars correlated with sweetness at the 90% confidence
level. Multiple regression analysis indicated that fructose
and sucrose expressed as percentages of the total solids were
related to sweetness. The R value was 0.32 and the

regression was significant only at the 84% confidence level.

l6

.Hm>mH

mocmnfimcoo mo.o mzp pm OLDN Soap pCDDOLMHU mHQCDOfichme Ohm pmpoc om mDSHm> an:

 

 

**om.o *xma.o wo.o *xmm.o **HN.Q :H.o mmOCpmmzm
mH.o **mm.o *mmw.ol NH.OI **NN.OI **mm.ol mmoosam
ma.o *awm.o **om.OI 0H.ol **ON.OI **ON.OI mmouospm
om.o om.OI om.OI **mw.o 3*Hw.o **Hw.o omegosm

O O O O O 0 nH-mm 3m
mm o **©N o **>N o **N: o **©> o *#N@ o HmpOB
W O
:H.Q **mm.o **mm.o **m:.o **mm.o **ma.o mamwmom
. . . . . . mUHHom
ma 0 **mm 0 **mm o *wmz 0 **N5 0 **:w o ampOB
Amwsm moHHom mofifiom
S
mmDCpmozm owoosaw omOpo pm mmopozm Hmpoe mansaom proe
mUmLQCH

 

 

.chHpmcHQEOO cannzm

H

m O>HMImDCOBB pew mocfiq pohpmo
ommncH xfim go mpoom com: com: mpcmEOLSmwoz mcoE< mcofipmamhpoo OHQEHm

.m magma

SPIJQ H

17

The predictive equation follows:

sweetness = -4.67 + 0.02 (% fructose) = 0.01 (% sucrose)
The lack of large and significant correlations was unexpected
as was the absence of significant differences in sweetness
content (Table 2). This was due in part to palate fatigue
suffered due to the many samples per panelist.

Means of the sugar and solids measurements taken on the
twenty-five hybrid combinations appear in Table 4. The
hybrids contained slightly more total sugar and more percent
sugar due to sucrose than the inbred lines. The hybrids
differed significantly for all traits except sweetness.

Some of these differences were found in total and soluble
solids, total sugars and sucrose content. They were due in
general to the high and low hybrid combinations MSU 6000 x
MSU 872, MSU 5986 x MSU 872, MSU 1302 x MSU 6000, MSU 5931 x
MSU 6000, MSU 6000 x MSU 1302, MSU 6000 x MSU 9541, MSU 1302
x MSU 5931, MSU 5986 x MSU 9541 and MSU 5931 x MSU 5986.
Some of the hybrid combinations responsible for significant
differences in fructose and glucose content (Table 4) are not
among those listed above. This indicates that the reducing
sugars do not vary directly with the solids, total sugar and
sucrose content.

The differences among the hybrid combinations remain
rather parallel for total and soluble solids, total sugars
and sucrose. Fructose and glucose levels also varied
together. Correlation coefficients for these relationships

are shown in Table 3. Among the hybrids total and soluble

18

 

 

 

mm.m pama.m onwmm.m won 0:.mz commom.mm namm.m moon mm.mH Hamm x NOMH
oo.m meH.OH onmaa.oa ponmmm.mm Uonmow.m> nmao.oa opon m:.mH moma x Hamm
No.3 nam:.m unwom.m ooommm.mm onmno.:> ammo.m Opon mm.mH momH x mpm
50.: mmo.mH nwmm.mH pop no.3: poommm.ow Qwam.m moon Hm.ma mum x Hmmm
NH.m o wm.m oomnm.w Uoommm.:m Uonmmm.wm ammo.OH moon ow.ma mwmm x ooom
om.m ommw.> osmom.m Boom:>.mm ponww:.mn nmmm.m mooanm.mH mum x momH
oo.m p mm.m on :w.m nmnw.mm onmmn.m~ nwma.oa opoanm.mH Hmmm x mum
mm.m nmmm.s onwom.m conmms.mm wonmmm.ms nmmm.m moonmma.ma Hmmm x mmmm
NH.: swam.m onmmz.oH commow.mm oommH.:w nmmm.m opopmma.ma Hzmm x ooom
mm.m n ow.m o w:.m onmwm.aw onmmm.:> nmm>.oa moonmmm.MH mama x ooow
Nw.m nwwm.m unam:.HH wonmma.mm onmom.nw owom.OH moonmmm.ma 0000 x Hmmm
NH.: 9 mo.© onmoo.h mom.:o onmmm.mn ommH.OH wonmom.ma mmmm x Hzmm
mw.m ommz.m oanH.OH pmwa.mm pmm>.mw ommm.HH onwam.ma ooow x NOMH
NH.m owmm.m opd:w.m pmmo.:m www.mm nmao.aa QmH:.:H mum x wwmm
nm.: a mm.o on mw.m mum.ow mmm.mm mam.HH NmmH.mH mmm x ooom
hwwsm moflaom mowaom ompmfipom
mmmCDOOBm mmoosaw mmouozpm mmoposm OHQSHom Hmpoe
m xampoe unmopmm pcmopmm owhnmm
.mofihnmm Hm poppmo w>flmlzpcO39

mo mpoom spas UmpmHOOmm¢ mosaw> mmmcpmozm pcm pcmpcoo moflaom ocm howsm com:

.: Danae

.Hm>®H mo.o
on» um mcmoE mcoEm moocohmmmfio pCMOHchme OCHEpmpmo op com: mm; anon 0mm m.mmxsem

.oamom OH on o D co pocfiEhwpmo mum: modam> mmmcpmm3mm

.OSmme 3mmpm w\wE mm oommmgaxo Chm modam> ammSm HH<x

 

COHDMHAH> mo

19

 

gmm mm Rmm mm gaa mm gmw SH gmm OH Nmo SH goo m mpcmHOngmoo
wp.m mm.w HH.m wn.mm m:.aw Hm.m mm.ma mummz camps
mm.m a sm.m onwom.m eon mm.:: e Hm.wm nano.m m AO.HH mmmm x Hmmm
Sm.m nmwo.m onmmm.m eon hm.m: So Am.mm nmzm.m ms mm.HH Hzmm x mmmm
AH.: nmom.g spaso.m eonmfio.mm nonmam.fis nmsm.m mes ms.HH Hmmm x moma
oo.m paws.OH onmwm.HH es mo.H: co mo.mm. n mm.a mes ms.HH momH x Smmm
mw.m nmom.m onwm:.m eon mm.m: eon Hm.m© p mm.» was mm.HH NOMH x Hmmm
SH.: nwmm.0H sawmm.HH eon m=.s: conmmm.mw nasa.m mes mo.ma 0000 x mmmm
00.: nmmm.HH mm:.mfi o mm.o: so Sm.:w pmmm.w was mo.ma Hemm x mam
AH.: cams.w onmpm.m eon mo.m: con oa.gw nwmm.m Deon NH.NH ommm x mam
SH.: nmmm.o opmzo.a eonwmm.:m uopmzm.mm nmms.m Egon AH.NH Hemm x Hmmm
sm.m nmom.m onmmo.w coomsm.om won mm.mo Emma.m Deon mm.mH mmmm x NOMH
.A.s.pcoov : magma

20

solids and total sugars were mutually and positively corre-
lated. Of the individual sugars sucrose was positively
correlated with these traits and was responsible for the
agreement between the solids data and total sugar. Data in
Table 3 also indicated that as total sugars increase in the
hybrids the percentage due to sucrose also increased.
Although the reducing sugars were again positively
correlated, negative correlations existed between the
reducing sugars and sucrose. This suggests antagonism
between their respective formations. The reducing sugars
are also inversely correlated with total solids, and soluble
solids, while a random relationship exists between the
reducing and total sugars. The pattern of inverse correla-
tions among these hybrids is similar to that reported by
Carlton and Peterson (1963) for inbred carrots. These
researchers discussed the association of high reducing
sugars with low total solids and fiber content. Such an
association brings both sweet and tender components together
which is valuable in the fresh market situation. Hybridiza—
tion techniques considering this association may provide an
efficient means to create new cultivars with these traits.
Among the hybrids sweetness appeared to be related to
soluble solids, total sugars, fructose and glucose. Fructose
was the sweetest sugar present in the carrots and glucose,
although about one-third as sweet, was directly related to
fructose concentration. Since sucrose was inversely

correlated with the reducing sugars, its contribution to

21

sweetness was probably reflected in the total sugar correla-
tion value. Multiple regression analysis failed to explain
sweet taste on the basis of all four of these traits. How-
ever sucrose and alpha-glucose expressed on a fresh weight
basis may be used to predict sweetness in the following
equation:

sweetness = —l.98 + 0.06 (sucrose) + 0.76 (alpha—glucose)
The R value for this equation was 0.33 and the regression was
significant at the 95% confidence level.

Difficulty was encountered in looking for parental
influence upon hybrid traits partly because the relationships
among solids and sugars differed in the inbred and hybrid
groups. Another difficulty arose due to lack of sufficient
hybrid combinations to complete a 6 x 6 diallel study. Seed
was unavailable for five combinations, three of which,
unfortunately,involved MSU 6000. However, as shown in
Table 5, the general combining ability total and reducing
sugar means did not change appreciably between the complete
4 x 4 diallel and the larger incomplete diallel. It was
evident that all but one of the seven progeny of MSU 6000
fell above the average for total sugars and those traits,
positively correlated with total sugars (Table 4). For
purposes of discussion, total sugars shall be considered
representative of these traits.

As shown in Table 5, MSU 6000 behaved equally well for
total sugars whether it was used as a pollinator or as a

seed parent. In both capacities MSU 6000 ranked high among

22

Table 5. General Combining Ability Means of Six Parent
Carrot Lines Based on Total and Reducing Sugar

 

 

 

 

Content.
4 x 4 6 x 6
Complete Incomplete
Diallel Diallel
Parent z Reducing Reducing

Line Total Sugar Sugar Total Sugar Sugar
__ -_ 75.10 14.97

6000
__ __ 76.60 20.96
73.20 19.23 70.30 19.71

5986
63.60 15.41 67.40 14.64
72.30 16.57 70.40 18.35

872
75.30 20.05 77.40 18.29
70.00 15.55 72.20 16.81

1302
67.70 19.16 70.10 18.06
-- -- 75.40 16.66

9541
__ —- 67.70 19.13
64.80 18.41 68.20 18.25

5931

73.60y 15.14X 73.60y 15.14X

 

X’yMeans followed by the same letter are based on the same

values.

ZAll sugar means are presented as mg/g fresh tissue.

23

the general combining ability means for total sugars. MSU
6000 as a pollinator was characterized by high reducing
sugar content as well. This is of importance because
reducing sugars were not related to the total sugars and
will therefore not necessarily appear in higher amounts as
total sugars increased in the hybrid combinations. MSU 6000
as a female resulted in very low reducing sugar content.

Combinations involving MSU 1302 also showed little
differences in total or reducing sugar when this parent was
used as a female or as a male. In both cases the sugar
contents were average.

The largest difference in total sugars as influenced by
male and female parentage were exhibited by lines MSU 872 and
MSU 9541. MSU 872 had a strong positive influence upon total
sugar when used as a pollinator. This may be of greater
general importance than the influence of MSU 6000 as a
pollinator since the former was based upon four combinations
whereas the latter involved only three. The strength of
MSU 872 as a female parent, also based on four combinations,
was slightly below average for total sugars and about average
for reducing sugars.

MSU 9541 occurred as a pollinator in five combinations
and resulted in below average total sugar content. These
five combinations were quite variable as Shown in Table 4.
However, among the female general combining ability means,
MSU 9541 ranked highest in total sugars. Here it occurred

in only two combinations both of which were above the grand

24

mean. Reducing sugar content for this line as a seed parent
and as a pollinator was near the average.

Five combinations involving MSU 5986 as a pollinator
showed that it was consistently weak in total sugars in both
the complete and incomplete diallel Situations. In only one
of the five combinations did the total sugar content fall
above the grand mean. The influence of MSU 5986 as a female
on the total sugar content was average. MSU 5986 was also
responsible for a large sex-related difference in reducing
sugar content. The female influence was positive whereas
the male influence was negative.

MSU 5931 had the poorest female influence on total sugar
content. Its performance as a pollinator was slightly above
average, although the reducing sugar content was lower than
average.

Reciprocal differences among the hybrids are evident in
Table 4. These may be used as specific examples to clarify
what has been stated previously about parental influences.
MSU 5986 x MSU 872 was significantly higher in total sugar
content than MSU 872 x MSU 5986. Since both of these parent
lines served as average females it is suggested that this
reciprocal difference was due to the exchange of pollinators.

A difference between MSU 9541 x MSU 5986 and the
reciprocal cross falls Slightly short of significance at the
95% confidence level. This difference was, however, signifi-

cant at the 90% confidence level. This exchange of weaker

25

pollinators suggests that MSU 9541 was a strong female
contributor to total sugar content.

The total sugar content of MSU 5986 x MSU 5931 was
significantly different from that of its reciprocal at the
90% level. The former combination involved two average
parents but the latter involved both a poorer pollinator
(MSU 5986) and a poorer female (MSU 5931). Accordingly the
latter combination had the lowest total sugar value among

the hybrid group.

SUMMARY AND CONCLUSIONS

In conclusion the following phenomena are evident from
this parent—hybrid study on the solids, sugar and sweetness
content in carrots.

Differences in total and soluble solids and total and
reducing sugars existed. MSU 6000 was significantly higher
in quantities of solids and total and reducing sugars than
were the other lines. The reducing sugars were not correlated
with sucrose content, but all other solids and sugar measure-
ments were mutually and positively correlated.

Significant differences in solids and sugar content were
apparent among the hybrids. Some of these differences were
due to superior quality of the combinations in which MSU
6000 was one of the parents. MSU 6000 was the only parent
line that served as a positive influence as both a male and
female parent. This behavior suggests that these outstanding
qualities were transmitted genetically and that the MSU 6000
cytoplasmic factors were not strongly influential.

Although the other inbreds did not Show significant
differences in the inbred analysis, their general combining
ability means did suggest genetic influence upon the hybrids.
Another strong pollinating line, MSU 872, and a strong

female parent, MSU 9541, were found. However, in these cases

26

27

the positive influence was associated with the sex of the
parent. This suggests that more than one mechanism of
inheritance most likely involves cytoplasmic influences.
Interactions between cytoplasm and the genome are also
likely since significant reciprocal differences occurred
among the hybrids.

Positive correlations existed among the traits measured
in the inbred lines, but negative correlations were found in
the hybrids. Although no explanation is readily available
for the inverse relationship between the reducing sugars and
the solids and sucrose contents in the hybrids it may be of
use in the breeding of sweet and tender fresh market carrots.

The sweetness values were generated in hopes of relating
them to specific sugars or combinations of sugars or
synergisms among sugars in the fresh carrot. Such relation-
ships would provide useful breeding bases. However, the
sweetness data could not be readily interpreted for the
inbred group. Simple and multiple regressions showed sig-
nificance among the hybrids where soluble solids and total
and reducing sugars and sucrose were related to sweetness.
The correlation coefficients were quite small, however.

This parent-hybrid study has served as an extension of
the nutrition analyses of advanced carrot breeding material
recently begun at Michigan State University. This study has
also provided a basis for the comparison of several more

years work on this subject. Repetition over location will

28

also provide information on the type of variation involved in
sugar make-up of this breeding material.

Thorough assessment of the cOmbining ability of the
inbred lines rests upon completion of the full diallel
design. In this manner statistical Significance may be
attached to the general combining ability information. The
five combinations required for this completion are now

available.

Recommendations

Recommendations are offered in order that future
experiments involving this plant material may be more suc-
cessful and meaningful.

The results of this study are not as tenable as they
might be were one to complete the 6 x 6 diallel. Five
missing combinations made analysis of‘heritability by
standard methods impossible. Statistical significance
could not be attached to the combining ability means.

In preparation for analysis the center third of each
root was used. Riddle and MacGillivray'O966)noted that total
solids decreased from the top to the bottom third of the
carrot root. It has already been documented that carrot
sugar content is positively correlated with total solids.
Since all carrot roots were sampled in the same manner, the
differences observed should not have been affected by the
total solids phenomena. However, future workers may wish
to consider alternative sampling procedures.

The statistical analyses reported here were charac—
terized by relatively high coefficients of variation.
Preliminary analyses of sugar content showed a 1.0 to 2.0%
error between injections into the gas chromatograph with the
use of an internal standard. This reflects the within-
machine variation.

Error between subsamples of the same material rose to

4% indicating that improvement of technique is in order. The

29

30
employment of an automatic pipette is strongly suggested for
the placement of sample and standard volumes in the vials.

The statistical design itself contributed most greatly
to random error. Blocks, though employed during analysis,
were absent in the field. Relationships of sweetness values
with the sugar analyses will be more realistic if block
effects can be removed during simple and multiple regression.
It is likely that multiple range tests will also Show a
greater number of differences among means. One may also
wish to consider increasing the number of replicates (blocks)
from six to a larger number if labor is available.

Optimum conditions during the taste panel analysis were
not met. Two violations are of concern - the volume of
samples per panelist and the absence of a reference carrot.
Spencer (1971) offered a discussion of taste panels and
sweetness measurements. Since future workers will presumably
plant a complete diallel, this will increase, not decrease
the number of samples per panelist (block). The use of
incomplete block designs would solve this problem.

Analysis of dried carrot product is preferable to fresh
because of the storage and handling advantages. Sugars may
be selectively removed by respiration during hot air drying
(Smith 1969). It has been reported that sugar losses may
also occur in such tissues due to non-enzymatic browning
(McWeeny 1973). Substantial sugar losses seen in hot air-
dried preliminary sample tissue that was stored at room

temperature are presumable due to one or both of these

31

activities. Sugar loss was greatly circumvented by the use

of the freeze dryer, but enzymes which hydrolyze sucrose are
not inactivated during this process (Smith 1969). Preliminary
experiments indicated that enzyme activity was not apparent
during the extraction of sugars, but this does not preclude
the generation of invert sugar during other handling of the
lyophilized sample. The freeze drying itself employed a
commercial procedure of heating at 60°C at the end of the
operation. Enzyme activity here is not impossible.

The appearance of invert sugar may be of special
interest since fructose has been entirely absent in some
reports on the sugar content on freshly analyzed carrot
roots (Platenius 1934 and Otsuka and Take 1969). Although
fructose-lacking freeze-dried carrots have been recently
analyzed by gas chromatography (Bittenbender 1975), future
workers may wish to investigate the possibility of

artificially induced fructose content in this type of sample.

LITERATURE CITED

Alabran, D. M. and A. F. Mabrouk. 1973. Carrot flavor.
Sugars and free nitrogenous compounds in fresh carrots.
J. Agr. Food Chem. 21:205-208.

Anonymous. 1973. Handbook gt Silylation. Pierce

 

Chemical Co., Rockford, Ill.
Association of Official Analytical Chemists. 1975.

Official Methods gt Analysis, 12th ed. Washington,

 

D. C.

Bassett, M. J. 1973. Relationship between core size,
specific gravity and soluble solids in carrots. Hort—
Science 8:139.

Bassett, M. J. 1974. Screening of carrot roots for high
soluble solids by specific gravity. HortScience 9:232-
233.

Birch, G. G. 1973. Gas-liquid chromatography of food
carbohydrates with Special reference to trimethylsilyl
derivatives. A review. Food Technology 8:229—246.

Bittenbender, S. A. E. 1975. Unpublished data.

Carlton, B. C. 1958. Evaluation and selection of carrot
breeding material for sugar content, dry matter and

soluble solids. M. S. Thesis. Michigan State Univ.

32

10.

ll.

12.

13.

14.

15.

16.

33

Carlton, B. C. and C. E. Peterson. 1963. Breeding
carrots for sugar and dry matter content. Proc. Amer.
Soc. Hort. Sci. 82:332-340.

Dutton, G. G. S. 1973. Applications of gas-liquid
chromatography to carbohydrates: Part 1. Advances tg

Carbohydrate Chemistry and Biochemistry. Ed. R. S.

 

Tipson and D. Horton. Vol. 30, 11—160. Academic
Press, New York.

Ellis, W. C. 1969.. Liquid phases and solid supports for
gas-liquid chromatography of trimethylsilyl derivatives
of monosaccharides. J. Chromatography 41:335—349.

Engel', O. S., A. A. Prokof'ev and T. V. Mikheeva. 1970.
Some physiological differences in carrot root tissues
performing different functions. Fiziol. Rast. 17:
1076—1082.

Green, L. F. 1971. The balance of natural and synthetic

sweeteners in food. Sweetness and Sweeteners. Ed.

 

G. G. Birch, L. F. Green and C. B. Coulson. pp. 7-18.
Applied Science Ltd., London.

Guthrie, H. A. 1971. Introductory Nutrition. 2nd ed.

 

C. V. Mosby Co., St. Louis.
Hasselbring, H. 1927. Carbohydrate transformations in
carrots during storage. Plant Physiol. 2:225-243.
Heatherbell, D. A., R. E. Wrolstad and L. M. Libbey.
1971. Carrot volatiles 1. Characterization and
effects of canning and freeze drying. J. Food Sci.

36:219—224.

17.

18.

19.

20.

21.

22.

34

Holligan, P. M. 1970. Routine analysis by gas-liquid
chromatography of soluble carbohydrates in extracts of
plant tissues. 1. A review of techniques used for the
separation, identification and estimation of carbo-
hydrates by gas—liquid chromatography. New Phytology
70:239—269.

Holligan, P. M. and E. A. Drew. 1970. Routine analysis
by gas-liquid chromatography of soluble carbohydrates
in extracts of plant tissues. 11. Quantitative
analysis of standard carbohydrates, and the separation
and estimation of soluble sugars and polyols from a
variety of plant tissues. New Phytology 70:271—279.

Kraut, C. W. 1974. A study of the nutritional composi-
tion of selected carrot varieties. Ph.D. Thesis.
Michigan State Univ.

Leveille, G. A., C. L. Bedford, C. W. Kraut and Y. C. Lee.
1974. Nutrient composition of carrots, tomatoes and
red tart cherries. Federation Proc. 33:2264—2266.

MacGillivray, J. H., G. C. Hanna and P. A. Minges.

1942. Vitamin, protein, calcium, iron and caloric
yield of vegetables per acre and per acre man—hour.
Proc. Amer. Soc. Hort. Sci. 41:293-297.

McWeeny, D. J. 1973. The role of carbohydrate in non-

enzymatic browning. Molecular structure and function

 

g£_food carbohydrates. Ed. G. G. Birch and L. F. Green.

 

pp. 21-31. Applied Science Ltd., London.

23.

24.

25.

26.

27.

28.

29.

30.

31.

35

Otsuka, H. and T. Take. 1969. Sapid components in
carrots. J. Food Sci. 34:392—394.

Phan, C. T. and H. Hsu. 1973. Physical and chemical
changes occurring in the carrot root during growth.
Can. J. Plant Sci. 53:629-634.

Phan, C. T., H. Hsu and S. K. Sarkar. 1973. Physical
and chemical changes occurring in the carrot root
during storages. Can. J. Plant Sci. 53:635—641.

Platenius, H. 1932. Chemical changes in carrots during
storage. Proc. Amer. Soc. Hort. Sci. 29:450.

Platenius, H. 1934. Physiological and chemical changes
in carrots during growth and storage. Cornell Agric.
Exp. Sta. Memoir 161:1-18.

Riddle, P. J. and J. H. MacGillivray. 1966. Relation
of dry matter to soluble solids in carrots and peppers.
Proc. Amer. Soc. Hort. Sci. 89:381—385.

Shallenberger, R. S. 1971. Theory of sweetness. In:

Sweetness and Sweeteners. Ed. Birch, G. G., L. F.

 

Green and C. B. Coulson. pp. 42—48. Applied Science
Ltd., London.

Sistrunk, W. A., G. A. Bradley and D. Smittle. 1967.
Influence of preharvest factors on carbohydrates in
carrots. Proc. Amer. Soc. Hort. Sci. 90:239—251.

Smith, D. 1969. Removing and analyzing total non-
structural carbohydrates from plant tissue. Wis. Agr.

Exp. Sta. Res. Rep. 41:1—11.

32.

33.

34.

35.

36.

37.

38.

36

Sondheimer, E. 1957. The isolation and identification
of 3-methy1-6-methoxy-8—hydroxy-3,4-dihydroisocoumarin
from carrots. J. Amer. Chem. Soc. 79:5036-5039.

Spencer, H. W. 1971. Taste panels and the measurement

of sweetness. In: Sweetness and Sweeteners.’ Ed. G. G.

 

Birch, L. F. Green and C. B. Coulson. pp. 112-127.
Applied Science Ltd., London.

Sweeley, C. C., R. Bentley, M. Makita and W. W. Wells.
1963. Gas—liquid chromatography of trimethylsilyl
derivatives of sugars and related substances. J. Amer.
Chem. Soc. 85:2497-2507.

Sweeley, C. C., W. W. Wells and R. Bentley. 1966. Gas
chromatography of carbohydrates. Methods tg

Enzymology. Vol. 8:95-108. Ed. S. P. Colowick and

 

N. 0. Kaplan. Academic Press, New York.

Waldron, D. R., C. D. Ball, E. J. Miller and E. J. Benne.
1948. A study of methods for the determination of
sugar in drop plants. J. Assoc. Off. Agric. Chem. 31:
708—714.

Wali, Y. A. and Y. M. Hassan. 1966. Qualitative
chromatographic survey of the sugars prevailing in
some horticultural crops. Proc. Amer. Soc. Hort. Sci.
87:264-269.

Watt, B. K. and A. L. Merrill. 1963. Composition gt

 

Foods - Raw, Processed, Prepared. Handbook No. 8, U.S.

 

 

Dept. of Agric., Washington, D. C.

37

39. Werner, H. 0. 1941. Dry matter, sugar and carotene
content of morphological portions of carrots through the
growing season and storage. Proc. Amer. Soc. Hort. Sci.
38:267-272.

40. Whitaker, T. W., A. F. Sherf, W. H. Lange, C. W. Nicklow

and J. D. Radewald. 1970. Carrot Production tg the

 

United States. Handbook No. 375, U.S. Dept. of Agric.,

 

Washington, D. C.
41. Yamaguchie, M., B. Robinson and J. H. MacGillivray.
1952. Some horticultural aspects of the food value of

carrots. Proc. Amer. Soc. Hort. Sci. 60:351-358.

"'11111311711777!fllfliujffl’IIIIfliIsIIIES