Texture and ripening physiology of tomato (Lycopersicon esculentum mill.) fruit


Material Information

Texture and ripening physiology of tomato (Lycopersicon esculentum mill.) fruit
Physical Description:
vi, 126 leaves : ill., photos ; 28 cm.
Ahrens, Milton Joseph
Publication Date:


Subjects / Keywords:
Tomatoes -- Physiology   ( lcsh )
Polygalacturonase   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph. D.)--University of Florida, 1989.
Includes bibliographical references (leaves 112-125).
Statement of Responsibility:
by Milton Joseph Ahrens.
General Note:
General Note:
General Note:
Includes abstract.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001547072
notis - AHG0610
oclc - 22450079
sobekcm - AA00004772_00001
System ID:

Full Text








I will ever be indebted to the postharvest group at the

Citrus Research and Education Center in Lake Alfred for

providing me the opportunity to begin my graduate studies.

Appreciation is further extended to the Vegetable Crops

Department (especially the postharvest group) for allowing me

to continue study toward the Doctor of Philosophy degree. I

would like to thank the members of my supervisory committee

for the strong support and encouragement they provided during

the course of my studies at the University of Florida.

Especially, I am grateful and honored to have been directed

by my supervisory chair, Dr. Donald J. Huber, who, to say the

least, provided an interesting, challenging, and rewarding

environment in which to pursue these studies. Of course, most

of my appreciation and admiration is directed at my wife,

Lynne, to whom I owe the most for my having successfully

completed this graduate program.



ACKNOWLEDGMENTS................................. ii

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



Introduction............................. 1
Cell Wall Structure and Composition............. 5
Fruit Softening.................................. 8
Models of Softening in Fruits..................... 10
Calcium Content of Pericarp and Fruit Firmness... 17
Polygalacturonase Activity and Ripening in Tomato
Fruit ............................................ 20
Fruit Mealiness......... ......................... 26

2 PREAMBLE.......... ........... ...................... 32
Discussion.. ... ................................... 32
Objectives of this Study........................ 35


Introduction..................................... 37
Materials and Methods ........................... 41
Results and Discussion........................... 47


Introduction.................................... 72
Materials and Methods............................ 75
Results and Discussion........................... 79

FRUIT............................................. 91

Introduction ...................................... 91
Materials and Methods............................ 93
Results and Discussion........................... 98

6 SUMMARY AND CONCLUSIONS.. ...................... 109

REFERENCES.............. ......... .......... ...... ... 112

BIOGRAPHICAL SKETCH..................................... 126

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



Milton Joseph Ahrens

May 1989

Chairman: Donald J. Huber
Major Department: Horticultural Science

Methods for determining texture in tomato fruit were

evaluated and an objective test for mealiness was developed.

The relationship between pericarp and whole fruit firmness

was investigated. Tomato genotypes varying in intrinsic

firmness were examined to determine the quantitative relation-

ships between polygalacturonase (PG) activity, autolysis,

firmness, and other ripening parameters including rate and

intensity of ripening. Firmness, respiration, and ethylene

production were monitored in the immature-green throughout

ripening. Polygalacturonase activity was measured by direct

assay of salt-extractable wall protein or by monitoring the

release of pectins from isolated, enzymically active wall.

Pectin content, hemicellulose fractionation, ethanol-insoluble

solids, and calcium were determined in mealy and non-mealy


Whole fruit firmness was not highly correlated with

firmness of inner and outer pericarp (r = -0.139 and -0.582,

respectively). Mealiness was not correlated with firmness.

Calcium and pectin content were significantly higher in mealy

fruit than in less-mealy fruit (25 to 104% and 12 to 44%,

respectively). There were no significant differences in

ethanol-insoluble-solids and % dry weight between the two

fruit types. In all fruit, polygalacturonase activity was

highly correlated with pericarp softening, but only moderately

correlated with softening of whole fruit (r = 0.920 and 0.757,

respectively). Polygalacturonase activity was positively

correlated with cell-wall autolytic activity in pink (r =

0.969) and red (r = 0.900) fruit. Firmer genotypes exhibited

lower rates of respiration and ethylene production during

ripening. Polygalacturonase activity in isolates prepared

from fruit at the climacteric peak was positively correlated

with ethylene production and respiration, and negatively

correlated with days to ripen (r = 0.929, 0.805, and -0.791,


The data indicate 1) whole fruit firmness is not indica-

tive of pericarp firmness; 2) mealiness and firmness are

separate metabolic events; 3) changes in firmness of pericarp

are due to the activity of PG. Furthermore, the data are

consistent with the hypothesis that pectin fragments released

by PG contribute to the production of autocatalytic (System

II) ethylene.



Tomato (Lycopersicon esculentum Mill.) fruit is the

primary vegetable crop produced in Florida. Three hundred

forty million kgs were marketed in the 1986-1987 season with

a wholesale F.O.B. value of over 408 million dollars (Florida

Tomato Committee, 1987). Seventy percent of the crop is sold

to be consumed fresh, with the balance used by the processing

industry. Terminal markets may be roadside, local, distant,

or export. Firm fruit are needed for shipping and most

handling. The highly mechanized bulk fast food market

requires firm fruit for slicing. On the other hand, softer,

fully ripe fruit are desired by the roadside market and for

processing. Due to these diverse marketing requirements, past

and current breeding efforts have been aimed at developing

tomato varieties with handling characteristics suitable to a

wide range of shipping, storage, and marketing conditions,

while attempting to meet consumers' demands for high quality

fruit (Hamson, 1952b; Rushing and Huber, 1983).

Firmness has long been recognized as an important

physical attribute of tomato fruit to be shipped successfully

(Hamson, 1952a, 1952b). Past workers have undertaken studies

to identify mechanical parameters of specific cultivars in

order to characterize and predict handling qualities of

certain breeding lines (Hall and Augustine, 1981; Hamson,

1952b; Rushing and Huber, 1983). These attempts have been

largely successful and are a tribute to past breeding and

selection programs. Today, there are tomato varieties

available that offer increased firmness over cultivars

available in the past (Rushing and Huber, 1983). However,

excessive firmness has often been associated with mealinesss,'

an undesirable texture trait (John Scott, Gulf Coast Research

and Education Center, Bradenton, Fl, personal communication).

The biochemical and physiological bases of firmness have been

studied in the past and are of great current interest (e.g.,

Haller and Harding, 1937; Shafshak and Winsor, 1964; Smith et

al., 1988). On the other hand, investigations into the

chemical basis of mealiness and its relationship to ripening

and firmness have received little attention. Recent inves-

tigations have focused on the relationship of firmness, cell-

wall degradation, and ripening (Baldwin and Pressey, 1988;

Brecht and Huber, 1988; DellaPenna et al., 1987; Smith et al.,


Tomato fruit are classified as climacteric. That is,

during ripening they exhibit a rise and peak in respiration

and autocatalytic ethylene production (Abeles, 1973; Biale,

1960). The onset of ripening and increased carbon dioxide

and ethylene production are temporally correlated with many

ripening phenomena among which are de novo mRNA accumulation

(Lee et al., 1987) and protein synthesis (Bartley et al.,

1982; Grierson and Tucker, 1983; Grierson et al., 1985; Tucker

and Grierson, 1982; Tucker et al., 1980), increases in the

activity of certain enzymes (Bartley, 1974; Bartley et al.,

1982; Brady et al., 1983; Buescher and Tigchelaar, 1975;

Buescher et al., 1976; Crookes and Grierson, 1983; Hobson,

1964, 1965; Knegt et al., 1988; Poovaiah and Nukaya, 1979;

Sawamura et al., 1978; Tucker and Grierson, 1982; Tucker et

al., 1980), fruit softening (Ahmed and Labavitch, 1980;

Bartley, 1974; Bartley et al., 1982; Brady et al., 1983;

Buescher and Tigchelaar, 1975; Buescher et al., 1976; Hobson,

1965; Knee, 1973; Yamaki et al., 1979), changes in acid and/or

sugar composition (Ahmed and Labavitch, 1980; Bartley, 1974;

Bartley et al., 1982; Gross, 1984; Gross and Sams, 1984; Gross

and Wallner, 1979; Huber, 1984; Knee, 1973; Yamaki et al.,

1979), and carotenoid synthesis (Tucker and Grierson, 1982).

Another characteristic of climacteric species is that fruit

harvested after a specific point in development (physiological

maturity), but before the climacteric rise, can be unripe, yet

will continue the normal ripening process even though detached

from the plant (Biale, 1960). The time it takes for fruit to

reach physiological maturity, the nature of the climacteric,

and length and characteristics of the ripening process are

characteristic for individual cultivars (Biale and Young,

1981). As an aid in marketing and to assure uniform prac-

tices, the USDA has classified tomatoes into ripeness categor-

ies according to their external color (U.S. Dept. of Agri.,

1975). Fruit in each category is perceived by wholesalers and

retailers as having certain handling characteristics, with

greener fruit being firmer and able to withstand longer

periods in shipping channels without breakdown. The earliest

stage at which tomato fruit can be harvested and still ripen

satisfactorily is designated the 'mature-green' stage. Fruit

in this classification are physiologically mature but show no

external signs of ripeness. Internally, ripening has been

initiated, but changes in quality attributes such as softening

are minimal at this point (Rick, 1978). Fruit of this class

are used for shipping to the most distant markets, for

specific types of processing, and generally can withstand the

rigors of marketing with the least damage. Another advantage

of harvesting at the mature-green stage is that they can be

allowed to ripen under controlled conditions. With the

manipulation of temperature and ethylene, it is common

practice for entire lots of tomatoes to be ripened uniformly

to meet particular market slots and demands (Ethygen Catalytic

Generators, 1988). As a rule, as tomatoes ripen and move

through the ripeness categories of mature-green, to breaker,

pink, and finally red, they soften (Brady et al., 1985;

Gertman and Fuchs, 1974; Rushing and Huber, 1983). It is

believed that changes in the structure of the plant cell wall

result in softening. The cell wall by its composition and

location is the most important structural element of the plant

cell (Bartnicki-Garcia, 1984). A primary component in this

softening is the enzymic hydrolysis of certain wall polymers

(Huber, 1983b).

Cell Wall Structure and Composition

Albersheim's group (Albersheim, 1978; Keegstra et al.,

1973) has proposed a model of cell-wall structure based on

their work with cell wall from suspension-cultured sycamore

cells. A diagramatic representation of this model is shown

in figure 1-1. The primary cell wall can be divided into two

general structural components (microfibrils and matrix) and

four general chemical components (cellulose, hemicellulose,

pectin, and protein).


Cellulose molecules consist of long unbranched chains of

a 1-4 glycosidic linked D-glucopyranose residues (8000

residues chain1', average). These molecules overlap, have

their long axis' arranged parallel to each other, and are

grouped together into unit structures called microfibrils

(Frey-Wyssling, 1969). A cross section of a microfibril would

cut across roughly 150 cellulose chains. The core of the

microfibril is composed of cellulose molecules organized into

a three-dimensional crystalline lattice held together by

hydrogen bonds. This core is in turn surrounded by other

cellulose molecules running parallel to the core, but not

organized into a crystalline structure. This area is called

the paracrystalline region or cortex (Mark, 1967; Wilson,

1964). Included in the cortex are non-cellulosic polysac-

charides and glycoprotein.


The microfibrils are embedded in a matrix which is

analogous to steel reinforcing rods being embedded in concrete

(Mark, 1967). The primary component of the matrix is polysac-

charide which can be subdivided into pectins (Aspinall, 1980;

Worth, 1967) and hemicelluloses (Whistler and Richards, 1970)

classified on the basis of their solubility in water/chelator

or alkali, respectively.

Hemicelluloses were originally thought to be precursors

of cellulose, but they now are known to be made up of polysac-

charides classified according to their monosaccharide com-

ponents (Bauer et al., 1973). There is considerable diversity

among the hemicelluloses, and entities in this group are

related only by their solubility (Towle and Whistler, 1973).

Albersheim (1978) has found that the only hemicellulose

present in primary cell walls of suspension-cultured sycamore

cells is a xyloglucan, which also appears to be the case for

primary walls of all dicotyledons.

Pectin in higher plants is comprised mostly of polyuronic

acid unbranchedd chains of P 1-4 linked galacturonic acid

residues). The major identified configuration consists of a

zig-zag arrangement of a variable number of galacturonic acid

residues separated from one another by a rhamnogalactose

trisaccharide. The trisaccharide appears to be the site of

a glycosidic link to other pectin components, specifically the

arabinogalactans. These are thought to be the 'connecting

bridge' between the pectins and the hemicelluloses. In

Albersheim's model (1978), the cellulose microfibrils are

coated by the hemicellulose xyloglucans, attached by hydrogen

bonding. The xyloglucan is glycosidically linked to a pectic

arabinogalactan. Further, this arabinogalactan is in turn

glycosidically linked to the pectic rhamnogalacturonan.

Proteins can be either structural or enzymic in nature.

Protein isolates prepared from isolated cell wall have been

shown to contain a variety of distinct enzymes including

exopolygalacturonase and p-galactosidase in apple (Pvrus malus

L.) fruit (Bartley, 1974, 1977, 1978), a- and P-galactosidas-

es and glucosidases, a-mannosidase, a-arabinosidase, 3-

xylosidase, pectin esterase (sic), and polygalacturonase in

pear (P. communis L.) fruit (Ahmed and Labavitch, 1980;

Bartley et al., 1982); 8-galactosidase, 8-1,3- and exo-/-l,4-

glucanase, Cx-cellulase, pectinmethylesterase (PME), and

endopolygalacturonase (endoPG) in tomato fruit (Buescher and

Tigchelaar, 1975; Gross and Wallner, 1979; Huber and Lee,

1988; Kivilaan et al., 1961; Poovaiah and Nukaya, 1979;

Sobotka and Stelzig, 1974; Strand et al., 1976; Themmen et

al., 1982; Tucker et al., 1980; Wallner and Bloom, 1977;

Wallner and Walker, 1975).

Lee et al. (1967) proposed that cell walls of corn (Zea

mays L.). coleoptiles contained hydrolytic as well as syn-

thetic enzymes. A decrease in weight of isolated wall in

buffer was thought to be due to autolysis of a non-cellulosic

glucan component with concurrent release of arabinose and

xylose. Huber and Nevins (1981) determined that this

autolysis was the result of glucanases hydrolyzing hemicell-

ulose glucans. Other cell-wall associated proteins include

the polygalacturonases (PG), which play an important role in

cell-wall degradation and associated fruit softening in

numerous fruit types (Crookes and Grierson, 1983; Hobson,

1964; Wallner and Walker, 1975).

Fruit Softening

The major structural component of the cell is the cell

wall (Bartnicki-Garcia, 1984). It is obvious that the

modification of this structure should play a major role in

changes or differences in fruit texture. Huber (1983b)

reviewed the role of cell-wall hydrolases in fruit softening.

Cellulases (individual enzymes or complexes) have been

identified in avocado (Hatfield and Nevins, 1986; Pesis et

al., 1978), peach (Hinton and Pressey, 1974), pear (Yamaki et

al., 1979), strawberry (Fracaria ananassa Duch.) (Barnes and

Patchett, 1976), and tomato fruit (Buescher and Tigchelaar,

1975; Hall, 1963, 1964; Huber, 1985; Pharr and Dickinson,

1973; Poovaiah and Nukaya, 1979; Sobotka and Stelzig, 1974;

Sobotka and Watada, 1971; Wallner and Walker, 1975). Although

reports of cellulase complexes apparently capable of complete-

ly degrading native cellulose have been characterized in

tomato fruit (Sobotka and Stelzig, 1974), there is no evidence

for extensive degradation of this polymer in vivo in any

fruit. Cellulose degradation has been shown to be minimal in

apple (Nelmes and Preston, 1968), avocado (Hatfield and

Nevins, 1986), pear (Jermyn and Isherwood, 1956), and peaches

(Sterling, 1961). However, Cx-cellulase does possibly play a

role in the terminal stages of the degradation of the locular

gel in tomato fruit (Huber, 1985).

Few studies have been undertaken to identify changes

occurring in the alkali-soluble (hemicellulose) components of

the cell wall during ripening and softening of fruit. Huber

(1983a) reported an increase in smaller-molecular-weight and

a decrease in larger-molecular-weight hemicelluloses in ripe

tomato fruit as compared with mature-green fruit. This shift

in polymer molecular-weight was coincident with pectin

degradation, but was shown to be an independent phenomenon

(Huber, 1983a). The basis for this modification of the

hemicellulose fraction is unknown. As mentioned by Huber

(1983a), there could be a precursor-product relationship

and/or separate synthesis of modified, smaller polymers as

ripening progresses. Similar results were reported in a study

on strawberry fruit (Huber, 1984). On a mole-% basis, sugar

composition of the hemicelluloses showed little change during

ripening from small-green through red. However, as with

tomato fruit, there was a loss of large-molecular-weight

polymers and a gain in smaller-molecular-weight polymers over

the same developmental period.

Models of Softening in Fruits

Pear Fruit

Jermyn and Isherwood (1956) reported on changes in the

cell wall of pear during ripening. There is a loss of

arabinan, galactan, and total hemicellulose. Total polysac-

charides fall, with a slight decrease in cell wall on a per

fruit basis. Yamaki et al. (1979) reported similar data for

the Japanese pear. They found a decrease in arabinose and

galactose content and concluded that these sugars were lost

from the hemicellulose fraction, possibly due to increased

arabanase and f-galactosidase activity (Yamaki and Matsuda,

1979). Cellulose decreased, while total pectin and water

soluble pectin (WSP) increased. It was concluded that the WSP

increase was due to the activity of PG (Yamaki and Matsuda,

1977), and the decrease in cellulose due to the activity of

exocellulase (Yamaki and Kakiuchi, 1979). However,

Labavitch's group (Ahmed and Labavitch, 1980b) failed to

identify cellulase and arabanase activities in 'Bartlett'


Ahmed and Labavitch confirmed the substantial loss of

arabinose and soluble pectin from ripening pear (1980a).

However, the most rapid loss of these polysaccharides does not

occur until after substantial fruit softening. They suggested

that the initial loss of firmness was due to the metabolism

of cell-wall components which did not affect polyuronide

solubility. Treatment of unripe tissue with a purified PG

preparation solubilized a pectic arabinan similar to that

which was solubilized in vivo. The large MW of the arabinan

released indicated that extensive degradation was not neces-

sary to release the polymer. Based on these studies, they

concluded that PG was the only enzyme participating in cell-

wall metabolism in ripening pear fruit. Bartley et al. (1982)

agreed that PG was the major determinant in pear softening and

furthermore reported a coordinated degradation of pectin

polymers with the appearance of endoPG.

Apple Fruit

Although also a pome, apple-fruit cortical softening

appears to be distinctly different than that which occurs in

pear. Knee (1973) demonstrated that apple fruit exhibit an

increase in soluble polyuronide, a decrease in hemicellulose

correlated with loss of wall glucan, and no change in cel-

lulose during postharvest ripening. Bartley (1974) described

a /-galactosidase in ripening apples, the activity of which

preceded the increase in soluble polyuronides. However,

whether this enzyme contributed to polyuronide solubilization

was unclear. Later (1977), Bartley proposed that in the

absence of changes in other components of the cell wall

(Bartley, 1976), hydrolysis of the galactan (probably through

the activity of P-galactosidase) in cell wall is responsible

for the loss of firmness in apple. Lidster et al. (1985) were

able to retain firmness in 'McIntosh' and 'Gravenstein' apples

in store at 20C after vacuum infusion of a partially purified

P-galactosidase inhibitor. These results indicate that p-

galactosidase can contribute markedly to loss of firmness in

apple fruit.

Knee (1974) described a two-stage breakdown of the cell

wall during apple ripening, similar to that suggested by

Bartley (1974). In the first stage, changes in firmness are

minimal while there is a decrease in wall galactan. The

galactose decrease was speculated to originate from a polygal-

acturonan. In the second stage, there is a marked accelera-

tion in softening and an increase in soluble polyuronide,

thought to be bound to a hydroxyproline protein in the middle

lamella. Endopolygalacturonase activity was not evident in

apple, an observation supported by the fact that the MW of the

liberated polyuronide was very high. However, Bartley (1977)

identified an exoPG in apple fruit. Still, there is no

plausable explanation of how this enzyme contributes to wall


Knee (1978) studied properties of polygalacturonate and

cell cohesion in apple fruit and suggested that cohesion

depended on the degree of esterification of the polygalac-

turonate. He suggested that free carboxyl groups maintained

cell cohesion through cooperative binding of calcium ions.

Strawberry Fruit

A different mechanism was proposed for softening in

strawberry fruit (Barnes and Patchett, 1976). In these fruit,

PE increases during fruit development to the early red stage

and decreases thereafter, while PG and polymethylgalacturonase

activities are not observed at any point in development. Cx-

type cellulytic activity increases as fruit senesce. The

authors suggested that loss of firmness in strawberry was not

due to degradation of pectic material, but (in agreement with

Neal, 1965) rather to an increase in pectic methylation,

thereby eliminating calcium cross-links, leading to a scission

of the middle lamella.

Knee et al. (1977) determined that cell expansion in

strawberry receptacles continued through the period that the

fruit turned red and softened. There is no increase in total

polysaccharide during ripening, whereas the middle lamella

swells and 70% of the polyuronide cell wall becomes soluble,

with a concurrent loss of arabinose and galactose. The

authors concluded that polysaccharide synthesis failed to keep

pace with expansion and wall maintenance requirements. This

causes the wall to weaken and allows cellular contents to

hydrate the middle lamella. Huber (1984) agreed that total

polysaccharides may remain constant (or decrease on a fruit

FW basis) but provided evidence which showed that polyuronide

increased on a per fruit basis during ripening. Based on

neutral sugar analysis of polyuronides throughout ripening,

it appears that the increase in polyuronide solubility is due

to the synthesis of new, more-soluble polymers. In support

of previous work, no evidence of PG-mediated cell-wall

degradation was found.

Tomato Fruit

Tomato fruit softening appears to be more related to the

system found in pear than in apple or strawberry fruit.

Hobson (1964, 1965) first suggested that the firmness of

tomato fruit was related predominately to the activity of PG.

Softer genotypes show increased PG activity (Hobson, 1964).

In addition, pericarp from areas exhibiting the physiological

disorder 'blotchy ripening' (Picha, 1987) fail to soften as

ripening precedes. Hobson (1964) reported that this disor-

dered tissue is lower in PG activity than the surrounding

'normal' areas. Hobson (1965) confirmed his earlier work by

measuring firmness of tomato fruit at various stages of

ripeness and correlating the measurements with PG activity.

There was a high positive correlation (r = 0.952) between

softness (whole fruit compression) and PG activity. This

study was substantiated by Buescher and Tigchelaar (1975) and

Buescher et al. (1976) who reported similar results working

with 'Rutgers' and tomato ripening mutants.

Wallner and Walker (1975) reported on the presence of

3-glycosidases in ripening tomato fruit. As it was under-

stood at this time that arabinans and galactans were major

structural components of primary plant cell walls (Talmadge

et al., 1973), Wallner and Walker (1975) speculated that these

glycosidases possibly function to hydrolyze these polymers and

aid PG in degrading the cell wall (Wallner and Walker, 1975).

Wallner and Bloom (1977) described degradation of tomato cell

walls in vitro and in situ. They reported less wall galactose

loss in situ, leading to their suggestion that PG is only

partially responsible for the wall hydrolysis which accom-

panies fruit softening. Sawamura et al. (1978) agreed with

these findings. They exposed tomato fruit to exogenous

ethylene. Polygalacturonase activity increased, but only

after an initial increase in WSP, leading to further specula-

tion that other enzymes were involved in softening. However,

these studies did not follow changes in firmness. Gross and

Wallner (1979) confirmed that galactose as well as arabinose

are lost from the cell wall during ripening. This loss

appeared to be separate from polyuronide solubilization.

However, they suggested that the effect of the decline in

these sugars on fruit firmness was minimal. Themmen et al.

(1982) exposed cell wall from normal and mutant tomato fruit

to protein extracts and purified PG from ripe fruit and agreed

that PG was the major enzyme responsible for cell-wall

degradation. However, they did point out that other enzymes

may contribute to softening in vivo.

Huber (1983a) followed the changes in molecular weights

of polyuronides during ripening in tomato fruit and confirmed

that extensive degradation occurred as evidenced by the

appearance of lower-molecular-weight polymers. This degrada-

tion corresponded with the trend of PG activity. Ultrastruc-

ture studies of ripening tomato fruit (Crookes and Grierson,

1983) showed a dissolution of the middle lamella and eventual

disruption of the primary cell wall. These changes were

correlated with the appearance of PG. Application of PG

isolates to unripe fruit tissue brought about identical

ultrastructural changes. Other workers have continued to

confirm that PG activity and loss of polyuronides from the

cell wall during ripening are the primary agents responsible

for softening in tomato fruit (Brady et al., 1985; Gross,

1984; Huber and Lee, 1986, 1988; Pressey, 1986; Rushing and

Huber, 1984); however, there have been some suggestions that

this relationship is only a general one (Brady et al., 1983,


Most researchers agree that the major cell wall event

occurring during tomato softening is solubilization of

polyuronide (Brady et al., 1985; Buescher and Tigchelaar,

1975; Buescher et al., 1986; Crookes and Grierson, 1983; Gross

and Wallner, 1979; Hobson, 1964, 1965; Huber, 1983a; Huber and

Lee, 1986, 1988; Pressey and Avants, 1971; Sawamura et al.,

1978; Themmen et al., 1982; Wallner and Bloom, 1977; Wallner

and Walker, 1975). It is recognized that there is a general

inverse relationship between PG activity and fruit firmness

within cultivars as fruit ripen (and soften) and between

cultivars varying in intrinsic firmness. Of those studied,

firmer tomato cultivars had less PG (Brady et al., 1983;

Buescher and Tigchelaar, 1975; Buescher et al., 1976; Hobson,

1964; Malis-Arad et al., 1983; Poovaiah and Nukaya, 1979;

Tucker et al., 1980). In addition, Rushing and Huber (1984)

developed an in vitro system for indirectly quantifying the

levels of PG activity by measuring the autolytic activity of

enzymically active cell wall. Wall from firmer fruit release

less soluble polyuronide than does wall from softer fruit.

One might anticipate a high relationship between PG

levels and firmness rather than the general trends previously

reported (Brady et al., 1983, 1985). However, these inves-

tigators determined firmness based on compression of whole

fruit. Hall (1987) has emphasized the importance of measuring

tissue (pericarp) firmness rather than whole-fruit firmness

when investigating the relationship between enzyme levels and

changes in texture. When measuring whole fruit firmness, the

relationship between enzymes and texture may be confounded by

the contribution to fruit texture of parameters which have no

relationship to the enzyme in question. Such parameters could

include the thickness of the pericarp and radial fruit-walls,

the amount and composition of the locular material, the

internal morphology of the fruit, the presence of columnella,

the point of measurement of compression (whether taken over

a locule or at a radial wall intersection), and the overall

water status of the fruit (Shafshak and Winsor, 1964). By

measuring only tissue firmness, the effects of most of these

confounding influences can be either eliminated or greatly


Calcium Content of Pericarp and Fruit Firmness

Calcium is a divalent ion. Due to its double positive

charge, it readily associates with pectin, forming calcium

pectate. The stereochemistry of the pectin chain may allow

for intra- or inter-polymeric binding (Rendleman, 1978). It

has been proposed that this cooperative binding may strengthen

the chains by holding hydrolyzed segments together through the

ionic bonds, or alternately making the polymers less vul-

nerable to attack due to stereochemic interference. The

degree of binding would depend on the amount of calcium

available and the degree to which the polymers were esteri-

fied. In this light, PME may play a role in fruit softening

by regulating the number of binding sites available for


If calcium does regulate PG activity, it would appear that

the concentration or distribution of this cation would change

during fruit softening, or be correlated with firmness among

cultivars of varying firmness. Investigations in this area

have yielded conflicting data. Poovaiah (1979), and Suwwan

and Poovaiah (1978) reported a decrease in total calcium as

fruit ripened and softened. Fruit were harvested at selected

stages of development and assayed. This study was supported

by Rigney and Wills (1981) who showed a shift in calcium

partitioning during ripening of on-vine-developed tomato

fruit. Total calcium increased as fruit developed, to a

maximum level at 80% of full development (incipient color)

then decreased as fruit ripened. Concurrently, bound calcium

(that calcium thought to be associated with pectin) decreased

and soluble calcium increased as a percent of total calcium.

However, Brady et al. (1985), using similar vine-attached

fruit, reported no change in calcium during ripening, whether

measured as total or acid-extractable. Ferguson et al. (1980)

pointed out that calcium extraction techniques are often

unreliable in determining calcium partitioning. Any method

which uses tissue homogenization and water extractions will

be influenced by other ionic components of the cytosolic

fluids. For instance, citric acid, which is often present at

high levels in fruits, readily complexes with calcium.

However, in an artificial system, Ferguson et al. (1980) were

able to recover over 90% of the calcium in calcium pectate

using an 80% acetic acid extraction. In order to overcome the

problems inherent in any extraction procedures, Burns and

Pressey (1987) measured calcium in the cell wall-middle

lamella (CW-ML) of tomato fruit using energy-dispersive X-ray

microanalysis. They reported that as fruit ripen and soften

on the vine, calcium in the CW-ML increases. The hypothesis

offered to explain this phenomenon was that calcium served to

hold the hydrolyzed polymers together, allowing for some

pericarp softening, but preventing cell-wall degradation until

the later stages of ripening and senescence.

Calcium infiltration has proved effective at delaying

flesh softening in apple fruit (Glenn and Poovaiah, 1987; Sams

and Conway, 1984). In addition, a deficiency in calcium in

tomato fruit has been linked to 'blossom end rot' (Geraldson,

1957; Gerard and Hipp, 1968; Spurr, 1959). In this disorder,

tissues in the area of the stylar end soften, become water

soaked, and are subject to invasions by pathogens.

Polvaalacturonase Activity and Ripening in Tomato Fruit

Polvyalacturonase as a Ripening Initiator

The most striking characteristic of tomatoes is that they

undergo a change in color (internal and external) during

ripening (Rick, 1978). As an aid in shipping and handling,

the United States Department of Agriculture has developed a

visual aid based on the external color of tomatoes which is

used in describing the 'ripeness' of the fruit (U. S. Dept.

Agri., 1975). This aid has become useful to researchers as

a reference for stages of development during ripening.

Typical commercial cultivars of tomatoes move through the

color classifications (from mature-green to red) in around 12

days (Rick, 1978). Another feature of developing tomato fruit

is that they soften dramatically as they ripen (Garrett et

al., 1960; Rushing and Huber, 1983; Shafshak and Winsor,

1964). As mentioned previously, PG activity is implicated as

the major cause of softening in tomato fruit.

Fruit with the physiological disorder known as 'blotchy

ripening' ( Picha, 1987) have areas which do not synthesize

lycopene and which fail to soften. Hobson (1964) first

demonstrated that these areas were lacking in PG. This was

the first recorded instance of PG being associated with

abnormal ripening. Buescher and Tigchelaar (1975) did a

comparative investigation of the tomato ripening mutant rin

and the normal cultivar Rutgers. Rin does not soften appreci-

ably, fails to develop lycopene (Robinson and Tomes, 1968) and

lacks the normal climacteric of tomato fruit (Herner and Sink,

1973). Buescher and Tigchelaar (1975) concluded that the

failure of rin to soften was due to the lack of PG activity.

No inhibitors of PG activity were detected. The authors,

along with Hobson (1964), speculated that the lack of PG

activity could be a primary or secondary agent responsible for

the failure of the fruit to ripen.

Attempts have been made to overcome the ripening inhibi-

tion expressed in rin and other mutants. Since ethylene

levels are low in these mutants, application of exogenous

ethylene (or its analogs) has been used in an effort to induce

ripening. Propylene caused an initial increase in respiration

in rin, but endogenous ethylene production was not initiated

(McGlasson et al., 1975). However, development of a slight

yellow color was somewhat advanced over control rin. Mizrahi

et al. (1975) demonstrated that attached rin fruit develop

lycopene if exposed to 130 ppm ethylene, although only at

levels a third of that in normal fruit. This response is the

only report of ripening related changes in rin due to exo-

genously applied ethylene.

Tigchelaar et al. (1978) reviewed genetic regulation of

ripening in tomato fruit, based on studies of the ripening

mutants. They reported that exogenous ethylene does not

stimulate PG activity in rin. Based on this and evidence that

rin exhibits normal levels of preclimacteric ethylene produc-

tion, yet fails to ripen when exposed to higher concentrations

of ethylene, the authors proposed that the 'ripening event'

which was suppressed in rin and nor mutants was PG activity.

In their model, PG hydrolyzes cell-wall pectin, resulting in

the release of cell wall-bound enzymes. They speculated that

these enzymes could contribute to various ripening phenomena

such as ethylene synthesis, carotenoid synthesis, and flavor

and volatile development. Studies by Hobson et al. (1983)

supported this hypothesis by demonstrating that tomato PG

released protein from tomato cell wall preparations. In

addition, an earlier investigation by Strand et al. (1976)

showed that the action of fungal PG released cell wall-bound


Suwwan and Poovaiah (1978) determined that bound calcium

levels in attached rin were high and continued to increase

with age compared to normal attached tomato fruit, which

showed a drop in bound calcium. Poovaiah (1979) agreed that

low PG activity in rin was responsible for the failure of the

fruit to ripen; furthermore, this low activity was due to the

influence of calcium functioning in maintaining cell-wall

structure and membrane integrity. A subsequent study

(Poovaiah and Nakaya, 1979) determined that PG (not cellulase)

was inhibited in rin. Autocatalytic ethylene production and

the respiratory climacteric do not occur in the absence of PG

and in normal fruit PG is detected prior to the respiratory

climacteric. Based on these results, the authors suggested

that the failure of rin to ripen was a consequence of the lack

of active PG.


Other investigators did not support the idea that PG

initiated ripening. Sawamura et al. (1978) reported that the

rise in ethylene production preceded an increase in water-

soluble pectins and PG activity in tomato fruit. They

concluded that ethylene initiated ripening in tomato and that

PG was a secondary ripening response. Salveit and McFeeters

(1980) demonstrated that PG activity increased in immature

cucumber fruit in response to exogenously applied ethylene.

Polygalacturonase activity was not detectable until after the

initial burst in ethylene production.

Hobson (1980) sought to introduce the non-ripening mutant

genes, through backcrossing, into normal tomato lines. Except

for the heterozygote (rin rin), all progeny had lower PG

activity, prolonged shelf life, and were firmer. The mutant

fruit also showed reduced levels of phosphofructokinase and

NADP'-malic enzyme, offering an explanation for the weak

climacteric in the mutants.

Grierson and Tucker (1983) followed ethylene production

and PG synthesis in normal tomato fruit. Enhanced ethylene

production occurred prior to the appearance of PG. Further-

more, fruit held in containers with the ethylene absorber

mercuric perchlorate remained green and failed to produce PG

during the same time period as control fruit. The authors

suggested that ethylene was in part required for the synthesis

of PG. Maunders et al. (1987) supported the role of PG as a

secondary ripening agent by demonstrating that ethylene

stimulates the accumulation of mRNAs in tomato, one of which

yielded in vitro translation products similar in molecular

weight to products obtained from known PG mRNA.

Brecht (1987) demonstrated that the initial increase in

System II (McMurchie et al., 1972) ethylene synthesis during

the early stages of ripening occurs in placental tissues.

This tissue has low, if any, detectable PG (Wallner and

Walker, 1975), yet is temporally advanced over pericarp tissue

in terms of cell-wall changes (Huber and Lee, 1986) and

lycopene accumulation.

Polygalacturonase and System II Ethylene

Recent studies have suggested that PG may contribute to

the production of System II ethylene in pericarp tissue.

VanderMolen et al. (1983) reported on the induction of

vascular-plugging gels in castor bean and banana by extracts

of Fusarium cultures. Polygalacturonase was isolated from the

pathogen-free extract. Data show that PG alone is able to

stimulate the production of the gels, possibly through the

solubilization of wall components. However, PG was shown to

stimulate ethylene production in castor bean leaves. Applica-

tion of exogenous ethylene to leaves produced the occluding

gels. These data indicate that PG of fungal origin elicits

ethylene production which, in turn, causes production of the

gels. Roby et al. (1985, 1986) treated melon leaves with an

elicitor isolated from melon leaf cell walls. This envoked

the synthesis of ethylene and cell-wall hydroxyproline-rich


Baldwin and Pressey (1988) vacuum-infiltrated PG and PME

into unripe normal and mutant tomato fruit and found that PG

had only a transitory effect on ethylene production in normal

fruit, but that in the mutant lines ethylene was maintained

at high levels. Pectinmethylesterase administered alone has

no effect. They suggested that possibly low and as yet

undetectable levels of PG prior to the onset of System II

ethylene in green fruit promote ethylene production to levels

that reach a physiologically active threshold, which in turn

would cause an increase in PG levels. These higher levels

would induce even more ethylene, which would be System II

ethylene. Kim et al. (1987) reported that infiltration of

galactose into pre-ripe tomato fruit, at levels present in

ripening fruit due to degradation of wall galactan polymers,

enhanced ethylene biosynthesis. Thus, it might be that a

variety of products of wall origin may contribute to enhanced

ethylene biosynthesis.

Brecht and Huber (1988) obtained pectin fragments from

autolytically active cell wall and vacuum infiltrated them

into pre-ripe tomato fruit. Ethylene biosynthesis was

enhanced, possibly due to an elicitor effect of the infil-

trated material. The response was specific in that only

certain size classes of fragments were effective. The authors

advanced the hypothesis that PG promotes System II ethylene


Fruit Mealiness

Mealiness is a general term used to describe a dry,

grainy, coarse texture in fruits. It is an important cellular

feature affecting a number of horticultural commodities. In

apples, fruit that become soft concurrently exhibit mealy

symptoms and the terms are often used interchangeably (Fisher,

1942; Haller and Harding, 1937). Peaches which have been

stored continuously at chilling temperatures become mealy upon

rewarming (Ben-Arie et al., 1970; Buescher and Furmanski,

1978). There is currently no literature available on meali-

ness in tomato fruit. However, tomato fruit which are firm

are thought to be mealy, although there are cultivars which

appear to be soft and yet have a grainy, coarse internal

appearance (John W. Scott, Gulf Coast Research and Education

Center, Bradenton, Fl, personal communication).

Mealiness Determinations

For some time it has been recognized that mealiness can

reduce consumer acceptance of fruits (Haller and Harding,

1937; Harding and Haller, 1932; Liu and King, 1978). There

exist many pre- and postharvest programs aimed at detecting,

evaluating, and reducing this internal textural parameter;

however, there has to date been no objective means available

to directly quantify this cellular trait. Direct determina-

tion is often made by subjective sensory evalaution (Clark and

Rao, 1977; Finney, 1971; Finney et al., 1978; Liu and King,

1978; Looney, 1975a, 1975b). Although sensory panels are

discriminating and able to distinguish mealy tissue, they are

limited in accuracy and precision, and do not provide a

reproducible numerical distinction.

Based on the observation that soft apple tissue is mealy

(Fisher, 1942; Haller and Harding, 1937), most tests for

mealiness determination in apple have incorporated resistance

to puncture (Liu and King, 1978; Mason et al., 1975; Riley and

Kolattukudy, 1976). There are reports of using sonic reson-

ance, which produces a "stiffness coefficient" (Finney et al.,

1978). In addition to sensory evaluations and puncture

(pressure) tests, bulk compression (Clark and Rao, 1977) has

been used to determine 'wooliness' mealinesss) in peach

tissue. Since mealy fruit appear dry and cakey, Buescher and

Furmanski (1978) have used % expressible juice to quantify

mealiness in peach fruit.

Physiological Aspects of Mealiness

Meheriuk and Porritt (1971) reported on the effect of

high (2%) CO2 on mealiness of apples in storage. Apples under

CO2 were firmer than those in air. Looney (1975b) confirmed

these findings. A sensory panel found that 'McIntosh' apples

which had been stored under CO2 were crisper and firmer than

those which had been stored in air. Meheriuk et al. found

that either prestorage (1977b) or during-storage (1977a) CO2

treatments of apples resulted in less softening in storage as

compared to control fruit; however, the high concentration of

CO2 required for the prestorage treatment damaged 'Golden

Delicious' apples beyond consumer acceptance (Meheriuk et al.,

1977b). Unrath (1972) demonstrated that preharvest applica-

tion of 2-chloroethylphosphonic acid (ethephon) increased

softness of apple flesh at harvest. This was later confirmed

by Looney (1975a). Miller (1975) looked at the effect of

ethephon on apples in storage. Treated 'McIntosh' apples were

softer after 1 week in storage, but effects as compared to

controls were minimal after 3 months. Hammett (1976) reported

similar but improved findings with ethephon treatment of

'Starkrimson Delicious' apples. These results all indicate

that apples become mealy in storage due to the action of

ethylene. Indeed, Lougheed et al. (1973) were able to

demonstrate that ethylene removal from storage atmospheres

allowed apples to remain firm.

Bramlage et al. (1974) demonstrated that calcium has an

effect on firmness. Apples which had received up to 700 ppm

Ca were firmer than control fruit after 7 months storage. A

postharvest dip of calcium chloride solution was shown to be

effective in reducing softness in 'McIntosh' apples after 4

months (Mason et al., 1975). The inclusion of a thickener

(for retention of the dip solution) gave greater results.

This same treatment was repeated by Mason (1976) with the same

results. Riley and Kolattukudy (1976) confirmed the effects

of calcium on apple firmness by demonstrating that prestorage

dips of any of several calcium-containing solutions effective-

ly reduced poststorage softness in 'Golden Delicious' apples.

Glenn and Poovaiah (1987) have recently determined the mode

of action of calcium treatments on delaying softening in apple

fruit. Dipping apples in calcium solutions prior to storage

prevents or delays the mealy condition by preventing the

dissolution of the CW-ML. Fruit which received calcium

treatments were judged to be less mealy and had less breakdown

and dissolution of the CW-ML as evidenced by scanning and

transmission electron microscopy. There were fewer cell-to-

cell contacts and less middle lamellar material in non-treated


Peach fruit are harvested prior to the initiation of

ripening. They are stored in this unripe state at low

temperatures for periods ranging from several weeks up to

several months (Buescher and Furmanski, 1978; Harding and

Haller, 1932). At the end of the storage period, they are

returned to ambient temperature and are allowed to ripen.

However, unripe peaches are chilling sensitive (Harding and

Haller, 1932, 1934). A symptom of this chilling injury is

mealiness of the pericarp (wooliness), which becomes evident

upon rewarming (Haller and Harding, 1939). In order to

overcome expression of this symptom, the fruit are intermit-

tently warmed during the second week of storage (Anderson and

Penney, 1975; Ben-Arie et al., 1970). Buescher and Furmanski

(1978) determined that reduced levels of pectinesterase and

PG were associated with reduced juiciness and mealiness in

peaches. The authors proposed that intermittent warming

prevents the development of mealiness by protecting the

ability of the tissue to produce adequate levels of these

enzymes during post-storage ripening.

While mealiness as a textural trait in tomato warrants

attention in breeding programs (John W. Scott, Gulf Coast

Research and Education Center, Bradenton, Fl, personal

communication), there is no documented evidence of the

relationship between mealiness and development (on the plant

or during storage) or its underlying biochemical causes.

However, it is often thought that firmer fruit tend to be

mealier than softer fruit.


Cellulose Fiber


Figure 1-1. Model of primary cell wall components of suspen-
sion-cultured sycamore cells as proposed by Albersheim. The
model is not intended to be drawn to scale, but only repre-
sents relationships. After Albersheim, 1978.



Changes in texture are a function of PG activity and are

related to the ripening process. In this research I want to

address the very broad area of the relationship between

texture modifications, PG activity, and ripening and to some

extent how calcium influences their interactions. Several

questions come to mind. What is the character of texture in

tomatoes? Is mealiness related to firmness? If not, what is

the biochemical basis of mealiness? What role does PG play

in mealiness and in ripening in general? Do softer fruit

ripen more quickly than firmer fruit? Does PG have more of

a role in ripening than just texture modification? And, how

does calcium influence and/or regulate these processes?

Polyqalacturonase Activity and Texture

Polygalacturonase activity is an accepted major component

of tomato fruit softening; however, correlations between this

enzyme and changes in firmness have ranged from quite high to

only moderate. All previous investigators have extracted PG

from specific tissues, yet have measured firmness by compres-

sion of whole fruit. In this methodology, effects of the

enzyme are dilluted and confounded. Furthermore, most

researchers in this area report firmness as absolute compres-

sion. This introduces another error. Cultivars, especially

non-improved lines, vary in size. A compression of 10 mm on

a 25 mm diameter fruit is not comparable to the same compres-

sion on a 150 mm diameter fruit. Using % compression would

minimize errors of this type and allow for more precise

comparisons between cultivars.

In apple and watermelon, mealy fruit are soft. In the

previous discussion, it was seen that treatments which

diminish the effects of ethylene delay or reduce mealiness in

apple. However, in peach, mealy fruit are excessively dry,

but not soft, and there has been no evidence linking ethylene

activity to mealiness in peach fruit. On the contrary,

mealiness is due to a failure of PG to adequately degrade

pectin after storage upon rewarming. It is completely unknown

at this time which system (if either) most closely represents

the one existing in tomato fruit.

Because a knife blade tends to pass between cell walls

of mealy fruit and through the cell walls of non-mealy tissue,

the cell wall in mealy tissue is probably stronger than that

in normal tissue, and/or the middle lamella in mealy tissue

is weaker than that in non-mealy tissue. As an alternative

hypothesis to lower PG activity causing mealiness, possibly

an increase in the hemicellulose component coupled with a

higher PG activity could result in mealy fruit.

Polvaalacturonase Activity and RiDening

In comparing the amounts of COz evolved by fruits of

different species, it appears that those fruits which have the

highest rates of respiration during ripening require less time

to ripen. In addition, bananas, which evolve copius amounts

of ethylene and COg, have high levels of PG. Avocados also

ripen quickly and have high levels of PG, yet less than

bananas. Tomatoes are intermediate in rate (days to ripen)

and intensity (amount of ethylene and COz evolved at during

ripening), and have less PG than avocado. Recent preliminary

observations have indicated that those tomato cultivars which

evolve the most ethylene, take less time to ripen (once

initiated) yet application of excessive amounts of exogenous

ethylene does not increase this rate. Lastly, apples take

much longer to ripen, and lack endoPG, having only exoPG. It

appears intuitive that the level of PG in a ripening fruit

directly affects the rate and intensity of ripening in that

fruit. While there are cases of fruits which evolve moderate

amounts of ethylene and CO2 yet produce no PG (muskmelon),

there are no known or reported instances of fruits which have

active PG and do not evolve moderate amounts of ethylene.


Bound calcium has been reported to decrease, increase,

or remain the same during tomato fruit ripening and softening.

There have been two major flaws with these investigations.

First, there are recognized problems with fractionating

calcium. Second, these investigators harvested fruit at

selected stages of development and attempted to correlate the

amount of bound and free calcium in the pericarp to the

fruit's point of development. Any changes in calcium could

have been influenced by the dynamic flux of calcium into or

out of the fruit while attached to the plant. Investigations

which attempt to look at changes in the state and amount of

calcium in relationship to changes in development during

ripening need to analyze detached fruit harvested un-ripe.

Objectives of this Study

Three major related hypotheses are proposed.

1. Texture is inherently linked to ripening in tomato fruit,

but in addition, its primary modifying agent (PG activ-

ity) is directly related to ripening regulation through

the biological activity of cell wall-released pectin


2. Mealiness in tomato fruit results from a failure of the

cell wall to rupture under mechanical force. The binding

forces between cells are greater than that within

individual cell walls.

3. Mealiness is related to solubility of hemicellulose in

conjunction with differences in pectin solubilization.

In order to address these theories, this work will

attempt to:

1. Characterize the texture of selected tomato cultivars

from "immature" through "red" ripeness classifications;

2. Evaluate alternative methods of determining firmness in

tomato fruit with a view to contrasting whole fruit and

tissue firmness;

3. Develop an objective method of determining mealiness;

4. Identify those anatomical characteristics that define


5. Determine the relationship between firmness and meali-


6. Follow changes in respiration and autocatalytic ethyl-

ene production through the climacteric and relate these

parameters to PG activity and texture;

7. Characterize changes associated with the cell wall

including changes in hemicellulose and pectin polymers;

8. Relate cell-wall autolysis to texture;

9. Determine the role of calcium in modifying texture.



The Incidence of Mealiness and Firmness

Texture is an important quality parameter in many

horticultural commodities. In fruit, texture can be sub-

divided into two basic components: firmness and mealiness.

Apple fruit. In apple fruit the terms are often used

interchangeably, since soft apple flesh is typically mealy

(Fisher, 1942; Liu and King, 1978). Data from subjective

evaluations which have determined fruit flesh to be mealy have

been highly correlated with objective tests for firmness as

determined using the Magness-Taylor and Effe-gi devices

(Finney, 1971; Finney et al., 1978; Liu and King, 1978). In

one investigation where the author did attempt to separately

analyze mealiness and firmness (Finney, 1971), mealiness was

found to be highly correlated (r = 0.93 to 0.96) with soft-

ness, as determined by sensory evaluation. Scanning electron

micrographs of the cut surface of apple fruit show that mealy

and soft fruit have less middle lamellar material between

cells, resulting in fewer cell to cell contact points (Glenn

and Poovaiah, 1987).

Peach fruit. Mealiness and firmness appear to be

independent aspects of texture in peach fruit. Clark and Rao

(1977) demonstrated that sensory panelists are able to

discriminate between maturity stages of peaches based upon

"elasticity", "hardness", and "wetness" (prob > F = 0.0001 in

each case), but not by "graniness" (prob > F = 0.8699).

Wooliness in peach fruit is described as a "dry and mealy"

condition (Buescher and Furmanski, 1978). It is a chilling

injury symptom which becomes evident upon rewarming. Harding

and Haller (1932) demonstrated that peaches stored at high

temperatures (25 to 30C) soften and are juicy. Those that are

stored at lower temperatures (2 to 4C) and then allowed to

ripen at 25C are less juicy, display a chilling injury symptom

called "breakdown". We can infer from these reports that in

peaches, mealiness, breakdown, and wooliness are terms

describing the same disorder, mealiness is independent of

firmness, and mealy fruit are dry. The only similarities in

mealiness between apple and peach fruit appears to be that the

texture is dry and coarse.

Tomato fruit. There have been numerous studies involv-

ing firmness measurements of tomato. Hamson extensively

reviewed literature on firmness and the underlying causes in

tomato (1952a) and contrasted firmness in several varieties

(1952b). Hall and Augustine (1981) evaluated three tomato

cultivars during extended storage. Rushing and Huber (1983)

reported on efforts by breeders to improve firmness of tomato,

detailing progress in this area. However, mealiness, as a


separate texture trait, has received little attention in

breeding programs (John W. Scott, Gulf Coast Research and

Education Center, Bradenton, Fl, personal communication). A

preliminary study using sensory evaluation indicated that

mealiness does not appear to be related to firmness in tomato

fruit. The firmest and softest cultivars studied were the

first and second mealiest, respectively. Furthermore, within

each cultivar, it appeared that mealiness does not increase

during ripening and softening of the fruit.

Texture Methodology

Although there exist no objective tests for measuring

mealiness, there are numerous devices and techniques for

determining firmness in fruit. Hamson (1952b) first described

a device which objectively measured firmness in tomato fruit

by simulating squeezing by hand. This device has become known

as the Cornell Pressure Tester (Garrett et al., 1960; Kattan,

1957). Kattan (1957) attempted to improve on this methodology

by constructing a multi-point compression device named the

Firm-o-meter. This device was later commercially manufactured

by Agricultural Specialty Company (Hyattsville, Maryland) and

is known as the Asco Firmness Meter (Garrett et al., 1960).

Correlations between these two devices (Cornell Pressure

Tester and the Firm-o-meter) in one study were r = 0.923 to

0.947 (Kattan, 1957). Other devices based on the single point

compression test have been constructed (Shafshak and Winsor,

1964) and have offered some improvement in ease of use. Gull

(1987) has modified the Cornell device to greatly facilitate

its operation. Compression devices have also been constructed

to measure firmness in cherry and prune (Verner, 1930),

strawberry (Bouyoucos and Marshall, 1951), and peach (Clark

and Rao, 1977).

The Magness-Taylor pressure tester (1925) has been widely

adopted in determining the firmness of apples, peaches, and

pears (Bouyoucos and Marshall, 1951). This spring-driven

device measures the amount of force required to drive a

plunger a given distance into a fruit. In addition to the

Magness-Taylor device and its derivatives, which are destruc-

tive, sonic resonance has been employed as a nondestructive

test for determining texture in apples (Finney, 1971; Finney

et al., 1978).

Universal testing devices for measuring theological

properties, such as the Instron Testing Device, measure the

force required to shear, compress, or pull (tension) samples

prepared from a variety of sources (Corey and Schlimme, 1988;

Hall, 1987; Mohsenin and Gohlich, 1962). These methods are

precise and accurate and they lend themselves to testing a

number of horticultural products; however, they are costly,

destructive, and may not correlate well with consumer evalua-



Mealiness and firmness may or may not be related in

tomato fruit, although a preliminary study indicated they are

not. There are examples in other fruits where mealiness is

associated with softening (apple) and where it is not (peach).

Much of the public and many breeders feel excessively firm

tomato fruit tend to be mealy. Any study investigating

texture must first address the relationship between firmness

and mealiness. In order to do so, objective techniques for

determining each component must be evaluated. The purpose of

this study is to characterize the texture of several tomato

cultivars. In order to do this, a variety of devices to

measure firmness will be evaluated and the collected data

correlated with the results from sensory panel evaluations.

Second, mealiness in tomato fruit will be determined by

sensory panel and an attempt be made to develop an objective

test for determining mealiness. Last, the relationship

between firmness and mealiness will be investigated.

Materials and Methods

Subjective and Objective Measurement of Tomato Fruit Firmness

Plant material. Initially, seven tomato cultivars were

grown in the Spring of 1986 at the Gulf Coast Research and

Education Center, University of Florida, Bradenton, Fl.

Transplants were set in a completely randomized block design

in the field March of 1986, 6 plants per block, 3 blocks per

cultivar. Harvests were initiated in June.

Fruit (50) from each block were harvested green at a size

characteristic of fully ripe fruit of each cultivar. The

pedicel and calyx were removed to prevent injury during

transport. Fruit were shaded and transported in an air-

conditioned vehicle directly to the laboratories at

Gainesville, arriving within 3.5 hours.

Upon arrival, fruit were lightly rinsed in tap water to

remove field debris, dipped in 1% hypochlorite, rerinsed, and

allowed to air-dry. Decayed, abnormal, or damaged fruit were

discarded. Fruit were regraded for uniformity, within

cultivars, of shape and size. Fruit were placed on trays,

stem-end down, in storage rooms at 20C and allowed to ripen

to selected stages of development. These stages were deter-

mined visually by comparison of external fruit color to a

United States Dept. Agri. visual aid (U.S. Dept. Agri. Visual

Aid TM-L-1, The John Henry Company, P.O. Box 17099, Lansing,

Michigan). To insure the exclusion of immature fruit, those

fruit that had not begun to ripen (develop incipient color)

within 2 weeks of harvest were discarded.

Additional fruit were harvested green, rinsed, and graded

as above, and evaluated for firmness with 12 hours of harvest.

Immature-green fruit were separated from mature-green fruit

after firmness determination by cutting through the equatorial

plane and observing the cut surface. Seed coats and locular

gel have not fully developed in immature-green fruit (Brecht,

1987; Kader and Morris, 1976).

Eleven, 9, 14, and 18 cultivars were transplanted into

the field in the Fall of 1986,. Spring and Fall of 1987, and

the Spring of 1988. Planting and harvesting schedules are

shown in table 3-1. Fruit handling procedures in all cases

were as outlined above.

Subjective texture determinations. Sensory evaluations

were conducted by panelists in the Fall of 1986 and the Spring

of 1987. The panels were managed under the guidelines of

Heintz and Kader (1983). Each panel was comprised of 11

individuals, 9 of whom were on both panels. In the fall

evaluation, panelists were not trained other than to be given

instructions on filling out the evaluation form, which is

reproduced in figure 3-1. In the spring test, panelists were

familiar with the range of texture to be expected (having

participated in the previous evaluation). Furthermore, they

received instruction by being provided examples of what a non-

mealy and severely mealy fruit were prior to beginning the


Fifty uniform red-ripe fruit were placed on trays, stem-

end down, in front of the panelists. Each tray contained one

cultivar. Three fruit per cultivar were selected by the

panelists at random. Prior to cutting the tomatoes, firmness

of whole fruit was judged by hand. With a utility knife,

panelists halved the fruit through the equatorial plane and

observed the cut surface. Afterwards, firmness, mealiness,

and mouthfeel were ascertained upon mastication. Each

panelist scored 3 separate fruit of each of the 9 (spring) and

11 (fall) cultivars. Individual scores of each fruit were

recorded as determined using a 5-point scale for mealiness and

firmness (figure 3-1). In another test, members were asked

to rank cultivars by mealiness without regard to individual


Objective firmness determinations. Initially, in the

Spring of 1986, 10 red-ripe fruit of each cultivar from each

block were examined for firmness objectively as follows. A

Cornell Pressure Tester (Hamson, 1952b) as modified by Gull

(1987), was used. This device will hereafter be referred to

as the IFAS (Institute of Food and Agricultural Science)

Firmness Device (figure 3-2). Measurements were taken on

whole fruit at the blossom end and at 3 equidistant points

along the equatorial plane. No effort was made to identify

locular areas or radial walls, since in some cultivars these

areas are not apparent from the exterior. Initial height was

determined and a 1 kg weight was applied to the fruit at these

points. Millimeters of deformation (deflection or compres-

sion) were noted at 5 seconds. Firmness values were expressed

as % deformation of whole fruit. Lower values reflect firmer


Following the compression test, the same fruit were

peeled by hand with a paring knife, taking care to remove the

least amount of tissue as possible. Two areas, 180 degrees

apart along the equatorial plane, were selected at random and

tested for firmness using a Magness-Taylor puncture device

(Magness and Taylor, 1925). Data were the force required (in

grams) to drive a 7.9 mm diameter convex tip into the flesh.

Higher values reflect firmer tissue.

The locular, placental, and columnella material, includ-

ing the blossom and stem scar tissues, of the above fruit were

removed and discarded. The remaining pericarp (fruit wall)

tissue of each individual fruit was further divided into outer

and inner (radial) pericarp by carefully excising the radial

walls intact. This left the pericarp of each fruit in 4

pieces: 2 shells of outer pericarp and 2 hemispheres of inner

radial wall. The pericarp types of each fruit were weighed

separately and placed in a Kramer shear cell attached to an

Instron Universal Testing Device (model #1132, Instron Corp.,

100 Royall St., Canton, Mass.) (figure 3-3). The Instron was

equipped with a 500 kg mechanical load cell. Crosshead speed

was 20 cm min"1. The device was set to drive the shear plates

completely through the cell. Peak heights were recorded and

data were expressed as Newtons (GFW pericarp)" Higher values

reflect firmer tissue.

In subsequent tests, firmness was determined at the

immature-green (IM), mature-green (MG), breaker (B), pink (P),

and red (R) stages of development, using the IFAS and Instron


Objective Mealiness Determination

Plant material. Tomato fruit were harvested and handled

as previously described. Initial evaluations were carried out

on 5 cultivars (Rutgers, Sunny, FL-7136, Suncoast, and Flora-

Dade) and later expanded to include most of the cultivars

grown each season. For subjective determinations, of

mealiness, all cultivars were evaluated in the Fall of 1986

and the Spring of 1987.


Sensory evaluations. The sensory evaluations were

conducted as previously described.

Determination of tissue tonicity. Discs were cut from

the outer pericarp of ripe fruit using a 5-mm cork borer. The

peel was excised from each disc and enough endodermal material

removed to bring the disc thickness to 3 mm. The discs were

incubated for 12 hours at room temperature in 50 ml of a range

of Sorbitol concentrations from 0.2 through 1.0 M to determine

which one concentration would most closely approximate an

isotonic solution for a number of tomato cultivars. Using 5

cultivars, 10 discs were excised from 3 fruit per cultivar per

Sorbitol concentration. Isotonic strength was determined by

measuring change in resistance of the bathing solution using

a conductance bridge (Radiometer CDM, The London Co., 811

Sharon Dr., Cleveland, Ohio), and comparing this against

visual observation of plasmolysis and rupture via light


Cell separation and count determination. Ten discs

(total) were recovered from 3 to 4 fruit as outlined above and

placed in 50 ml of isotonic Sorbitol (0.6 M, as determined

above) in 125-ml erlenmeyer flasks (3 flasks per cultivar).

Each flask was then sealed with parafilm and shaken (120 rpm)

at room temperature using a gyrotory shaker. At the end of

1 hour, 10 ml aliquots of the bathing solution were drawn from

each flask and pipetted into a 25-mm petri plate. Total cell

counts of each sample were determined using a light microscope

(Nikon Diaphot Type 108).

Scanning electron microscopy of mealy tissue. Ripe,

mealy and non-mealy tomato fruit were cut through the equa-

torial plane and a slice of tissue 3 mm thick was excised from

the surface using a utility knife. These sections were freeze

dried for 1 week at -62C and 80 mtorr (Virtis Freezemobile II,

model 10-MR-TR, The Virtis Co., Gardiner, NY), then immediate-

ly transferred to a desiccator. After 2 days, samples were

excised from the outer equatorial pericarp region of the dried

sections with a scapel, gold coated using an IB-2 Ion Coater

(Eiko Engineering, Perkin Elmer, Rockville, MD) and viewed at

50X with an Hitachi S-450 Scanning Electron Microscope

(Hitachi Inc., Tokyo, Japan).

Results and Discussion

Objective and Subjective Firmness of Tomato Fruit

Objective measurements. Firmness of ripe fruit in the

Spring of 1986 as determined by the IFAS device, Instron

device, and Magnus-Taylor is shown in table 3-2. Each

technique was able to discriminate between cultivars, and with

precision. However, ranking of the cultivars by firmness

differed according to technique employed. Correlations among

objective measurement techniques are shown in table 3-3.

Whole fruit firmness as determined by the IFAS device was

poorly correlated with the Magness-Taylor device (r = -0.572).

This is not surprising since the Magness-Taylor device, in

measuring resistance of the outer pericarp to puncture, is

influenced by the contribution of the underlying tissue.


Unlike apples, peaches, and pears, tomatoes are not uniform

throughout their interior. Locular areas containing a jelly

and water-like consistency are surrounded by firmer ovary

wall. In addition to the IFAS device, the Magness-Taylor was

not correlated highly with the Instron inner wall (r = 0.301)

or outer wall (r = 0.435) determinations. The Instron wall

measurements are of a single tissue type (ovary wall), whereas

the Magness-Taylor device measures a composite parameter, as

pointed out. These data indicate that the Magness-Taylor

device is unsuitable for use in determining firmness of fruits

that are not uniform throughout, such as the tomato.

Like the puncture test, the IFAS device measures a

composite type of firmness. Resistance to compression is

likely influenced by inner wall morphology and texture of the

locular material. Correlations between the Instron and IFAS

devices were low. Correlations coefficients were -0.139 and

-0.582 for inner and outer wall, respectively.

The Instron measurements of inner and outer pericarp wall

were highly correlated (r = 0.834). This is not surprising

since these measurements were taken on the same tissue type.

The inner wall was softer than the outer wall (table 3-2),

reflecting the observed pattern of ripening in tomato fruit.

That is, interior radial walls ripen before outer walls and

are developmentally advanced over outer walls (Brecht, 1987).

Hall (1987) found identical patterns when he examined outer

and radial walls of 'Walter', 'Flora-Dade', and 'MH-1' tomato

fruit using an Instron device equipped with a penetrating


Ripening related changes in firmness of selected culti-

vars are graphically illustrated in figures 3-4 and 3-5. For

3 of the cultivars, firmness of whole fruit (as determined by

IFAS device) was not significantly different until the breaker

stage (figure 3-4). Thereafter, the cultivars segregated into

very soft ('Rutgers') and midrange firmness ('Flora-Dade' and

'Sunny') groupings. 'Sunny' is considered 'ideal firm' by the

tomato industry (John W. Scott, Gulf Coast Research and

Education Center, Bradenton, Fl, personal communication). A

fourth cultivar (Fl-7136) consistently increased in softness

from the immature-green through the red stage. However, this

cultivar was the firmest of those tested, and the overall rate

of change was the least. 'Flora-Dade' and 'Fl-7136' were not

significantly different at the pink stage, yet varied drama-

tically at the red stage. These data (figure 3-4) illustrate

the point that firmness of whole fruit up to and including the

breaker stage, and even in one instance through the pink

stage, is not indicative of firmness at the red-ripe stage of

development. Figure 3-5 shows firmness of fruit pericarp (as

determined by Instron device) at selected stages of develop-

ment as fruit ripen. Data are expressed in this figure as the

reciprocal of Newtons (GFW pericarp)-' for easy comparison with

data in figure 3-4. For all 4 cultivars, differences and

changes in firmness were not significant until the pink stage.

At this point, rate of change in firmness of 'Rutgers'

exceeded the rate of softening of the other 3 cultivars which,

although had begun to change, did not differ significantly

from one another. However, at the red stage, the 4 cultivars

had segregated into 3 categories similar to the whole fruit

firmness (fig 3-4). Unlike whole fruit firmness, tissue

firmness was not indicative of final firmness until after the

pink stage.

Subjective texture evaluation. Table 3-4 shows firmness

of 9 cultivars in the Spring of 1987 as determined by the IFAS

device and sensory panel. There was a very high correlation

between blossom-end compression and equatorial compression (r

= 0.946). It appears that the blossom-end measurement can

serve as an indication of whole fruit firmness without sacri-

ficing accuracy. In this test, blossom-end firmness was

higher than equatorial firmness, reflecting the absence of

softer underlying locular material through the polar axis.

Sensory panel determinations were equally correlated with the

IFAS blossom-end measurement (r = -0.877) and equatorial (r

= -0.857) determinations (table 3-4), confirming either

objective technique From the Duncan's Multiple Range Test,

it appears that sensory panelists were not as discriminating

as the objective measurements. Trends in the fall sensory

evaluations were similar to the spring tests (data not shown).

Mealiness Determinations

Preliminary studies have shown that in mealy tomato

fruit, cells appear to easily separate one from another under

slight pressure. Glenn and Poovaiah (1987) have shown that

in mealy apple tisse, cell to cell contacts are reduced.

Scanning electron micrographs of the cut surface of mealy

cortical tissue showed little rupture of cells as compared to

non-mealy tissue. Overripe watermelon, whose pericarp appears

mealy (Donald J. Huber, Vegetable Crops Dept., Univ. of

Florida, Gainesville, FL, personal communication), show a

similar breakdown of the middle lamella (Elkashif and Huber,

1988). It was surmised that the tensile strength between

adjacent cells in mealy tissue is less than that within the

cell walls. This results in cells separating rather than

rupturing when sheared; the contents are not released, thus

the characteristic dry, coarse, grainy appearance of the cut

surface. Figure 3-6 shows the cut surface of non-mealy,

mealy, and extremely mealy tomato fruit. With these assump-

tions in mind, the possibility of determining mealiness by

direct measurement was investigated, based on the propensity

of the cells to separate under slight mechanical force.

Tissue tonicity. Six hundred millimolar (0.60 M)

Sorbitol was the concentration which most closely matched an

isotonic solution for the cultivars tested (table 3-5). Light

microscopy confirmed that there was only slight plasmolysis

at this concentration and little evidence of cell rupture.


Cell separation and count determination. Figure 3-7

shows examples of randomly chosen viewing areas of the

aliquots of the disc bathing solution as seen through the

light microscope at 4X. The fields correspond to aliquots

from slightly mealy (figure 3-7A), mealy (figure 3-7B), and

very mealy (figure 3-7C) fruit. Table 3-6 lists subjective

mealiness scores of 11 cultivars in the Fall of 1986. Also

included are objective determinations (cell counts) of 5

cultivars. Fruit judged not mealy to slightly mealy had

average cell counts of 76 to 101, whereas fruit judged to be

mealy and extremely mealy had average cell counts of 374 and

512. Table 3-7 lists and compares subjective and objective

mealiness scores of 8 cultivars in the Spring of 1987. In

every instance, mealy fruit (determined by sensory evaluation)

had greater average cell counts than less mealy fruit.

Average cell counts of 57 to 116, 160, 244 to 371, and 480

corresponded to slightly mealy, mealy, very mealy, and

extremely mealy, respectively. When sensory panelists were

asked to rank tomato fruit according to mealiness, there was

a 1 to 1 correlation between the sensory panel determinations

and cell count method (table 3-8).

Scanning electron microscopy. Scanning electron micro-

graphs of the surfaces of tomato tissue which had been sliced

with a utility knife are shown in figure 3-8. Cells of very

mealy tissue (figure 3-8A) had less points of cohesion than

did cells of mealy tissue (figure 3-8B), which in turn had

less than cells from non-mealy tissue (figure 3-8C). The

blade appeared to have gone between cells in mealy tissue

(figure 3-8A) rather than through them as in non-mealy tissue

(figure 3-8C). Cells of moderately mealy tissue had a

combination of ruptured and whole cells (figure 3-8B). This

phenomenon in tomato is similar to that reported to occur in

apples (Glenn and Poovaiah, 1987), and substantiates the

hypothesis regarding the physical manisfestation of the mealy


Mealiness as a function of firmness. Results of the

objective mealiness determinations for the Fall of 1986 are

contrasted with the objective IFAS firmness determinations in

table 3-9. The overall correlation across all cultivars

between mealiness and firmness was r = 0.089. The highest

correlation between these textural traits occurred within the

cultivar FL-7155 in the Spring of 1987 (r = 0.451). A

scattergram of mealiness as a function of firmness is depicted

in figure 3-9. These data overwhelmingly indicate that there

is no relationship between firmness and mealiness in tomato


Table 3-1. Planting and harvest dates of tomato cultivars,
conducted at the Gulf Coast Research and Education Center,
Bradenton, Fl.


SPRING 1986 JAN 26 FEB 09 MAR 07 JUN 06

FALL 1986 JUL 28 AUG 05 AUG 28 NOV 13

SPRING 1987 JAN 31 FEB 17 MAR 16 JUN 18

FALL 1987 JUL 16 JUL 26 AUG 26 NOV 23

SPRING 1988 JAN 08 JAN 18 FEB 22 MAY 26

Table 3-2. Objective firmness determinations of ripe tomato
fruit harvested in the Spring of 1986. Methodology is as
outlined in the text. Each number is the mean of 30 fruit,
followed by the standard error.


Flora-Dade 110.083 10.1+0.065 2.99+0.096 4.63+0.112

MR-1 12+0.334 7.5+0.118 1.97+0.082 3.40+0.156

Fl-7136 5+0.042 8.8+0.118 2.29+0.063 4.76+0.132

Suncoast 9+0.196 8.0+0.063 2.03+0.036 3.35+0.072

Sunny 10+0.170 7.7+0.056 3.02+0.051 4.68+0.125

Walter 16+0.171 7.0+0.043 1.85+0.031 2.70+0.055

Marglobe 140.290 6.1+0.078 2.71+0.134 4.30+0.130

ZNewtons (GFW pericarp)"'. Larger values reflect firmer fruit.
Y% compression. Larger values reflect softer fruit.
Grams X 100. Larger values reflect firmer fruit.

Table 3-3. Correlations among objective firmness measurement
techniques, between means of firmness of ripe tomato fruit
harvested in the Spring of 1986.


IFAS -- -0.5722 -0.139 -0.582

M-T -0.572 -- 0.301 0.435

INNER -0.139 0.301 -- 0.835

OUTER -0.582 0.435 0.834 --

ZAll values are significant at the 0.0001 level.

Table 3-4. Firmness of ripe tomato fruit harvested in the
spring of 1987, as determined by sensory panel and IFAS
device. Smaller IFAS values are firmer. Larger sensory panel
values are firmer.


Angora-Hairy 14aU 25a 2.2a

Rutgers lib 16b 3.0b

Angora 10b 16b 2.3a

Fl MH-1 8c 14c 2.8b

Sunny 6d 14c 3.0b

Flora-Dade 5de 12d 3.7d

Suncoast 4e 9e 3.3c

FL-7136 3f 7f 3.9d

FL-7155 3f 8e 4.2d

z% compression, means of 10 fruit
y3 measurements per fruit
means of 33 fruit
Means within columns followed by the same letter are not
significantly different at the 0.05 level

Table 3-5. Effect of molarity of Sorbitol bathing solution
on electrolyte leakage from tomato pericarp discs. Data are
means (to nearest %) of three determinations of % change in
conductivity of solutions after 12 hours of incubation at room

CULTIVAR 0.20 0.40 0.60 0.80 1.00

ANGORA 124 74 30 52 69

RUTGERS 71 42 28 22 94

FL-7136 112 71 37 21 23

SUNNY 80 49 33 39 51

Table 3-6. Mealiness of ripe tomato cultivars as determined
by sensory panel and objective cell count, Fall 1986.


233-A 2.2b

RUTGERS 1.9a 76a


SUNNY 1.4a 101b.

SUNCOAST 2.8c 374c

FL-7136 3.7d 512d

FL-MH1 1.9a

FL-7155 3.2d

FLORA-DADE 1.9a 97a


7136 X (NC-8276 X 7065) 2.7c

means in columns followed by the same letter are not signifi-

cantly different at the 0.05 level.
range of 1 (not mealy) to 5 (extremely mealy)
Means of 3 determinations

Table 3-7. Mealiness of ripe tomato cultivars as determined
by sensory panel and cell count, Spring 1987.


SUNNY 2.3a 57a


RUTGERS 2.1a 67a


ANGORA 4.6ef 371e

SUNCOAST 4.3d 253d

FL-7136 4.5de 244d

FL-7136 X SUNCOAST 4.8f 480f

Means within columns followed by the same letter are not sig-

nificantly different at the 0.05 level
range of 1 (not mealy) to 5 (extremely mealy)
means of 3 determinations

Table 3-8. Ranking of ripe tomato cultivars by mealiness as
determined by sensory evaluation and cell count.



SUNNY 2 132


FL-7136 4 262

ANGORA 5 344

Least (1) to most (5) mealy
means of 3 determinations

Table 3-9. Objective firmness (IFAS device) and objective
mealiness (cell count) of ripe tomato fruit harvested in the
Fall of 1986.


SUNNY 6.5 57


RUTGERS 10.5 67


ANGORA 12.9 371

SUNCOAST 5.1 253

FL-7136 4.9 244

FL-7136 X SUNCOAST 4.7 480

z% deformation. Larger numbers reflect softer fruit.
means of 3 determinations. Larger numbers reflect mealier





Please rate the following texture parameters on a 1 to 5
scale. Use any criteria or technique to determine the
ratings. Choose three (3) fruit PER CULTIVAR. Evaluate one
(1) fruit per form. You will have completed a total of


Very Soft

Slightly Soft


Very Firm


Very Soft


Slightly Soft


Very Firm

Slightly Mealy


Very Mealy

Figure 3-1. Form used to record sensory evaluation of tomato





Figure 3-2. Cornell Firmness Tester as modified by Gull. A
fruit is placed under the 1 kg weight and the handle is
lowered until contact is made with the fruit. The sliding
inclined plane under the depth gauge is used to zero the
instrument. The handle is then released and compression is
noted at 5 seconds.

Figure 3-3. Instron Universal Testing Device fitted with a
Kramer Shear Cell. The machine is set to drive the plates
completely through the sample box. A 500 kg mechanical load
cell transfers the force into electrical energy which is sent
to a strip recorder.


12.0 -

0 9.0l FL-7136
O 9.0 -

..0 -6.

C- 3.0

0.0 I

Figure 3-4. Whole fruit firmness, as determined by IFAS
device, of tomato fruit at selected stages of development.
IM, MG, B, P, and R are immature-green, mature-green, breaker,
pink, and red fruit, respectively. Bars are SE of 30 fruit.
Higher values reflect softer fruit.







* FL-7136



Figure 3-5. Firmness of tomato pericarp, as determined by
Instron device, from tomato fruit at selected stages of
development. IM, MG, B, P, and R are immature-green, mature-
green, breaker, pink, and red, respectively. Data is ex-
pressed as the reciprocal of Newtons GFW pericarp". Bars are
SE of 10 measurements.





Figure 3-6. Examples of, from top to bottom, non-mealy,
mealy, and extremely mealy ripe tomato fruit.


if flC

I -





U S,


Figure 3-7. Typical field views through the light microscope
at 4X of aliquots taken from the bathing solution containing
discs from A) non-mealy, B) mealy, and C) very mealy tomato


vr, -




Figure 3-8. Scanning electron micrographs of the cut surfaces
of A) very mealy, B) mealy, and C) non-mealy tomato fruit.
Bars are 500 microns.




200 300 400

Figure 3-9. Mealiness (cell count) as a function of pericarp
firmness as determined by Instron device. Overall correlation
between mealiness and firmness was r = 0.089.7


0 300







The levels of PG activity in tomato fruit show a general

inverse correlation with whole fruit firmness (Brady et al.,

1985; Hobson, 1965). Firmer genotypes tend to have less PG

activity. Within a cultivar, the softening which occurs

during ripening is accompanied by increases in the levels of

PG. The ripening mutant rin, which fails to ripen and soften

normally, has been reported to lack this enzyme (Buescher et

al., 1976; Hobson, 1980; Tucker and Grierson, 1982), although

this claim has been recently disputed by DellaPenna et al.

(1987). Some tomato genotypes are prone to the disorder

'blotchy ripening', in which affected areas of the pericarp

fail to ripen; these regions have been shown to have a greatly

reduced level of PG (Hobson, 1965).

On the basis of studies with rin and other tomato

ripening mutants, Tigchelaar et al. (1978) proposed that PG

functioned to initiate a number of ripening parameters,

including an increase in ethylene production, by releasing

specific but unidentified enzymes from the cell wall. The

idea that PG functioned to initiate ripening did not receive

support from subsequent studies showing that the initial

increase in ethylene synthesis occurred prior to the ap-

pearance of PG transcripts (Grierson, 1985; Grierson and

Tucker, 1983; Maunders et al., 1987). However, these studies

were based on ethylene production in whole fruit, whereas PG

levels were assessed in isolates of pericarp tissue. Brecht

(1987) demonstrated that the initial increase in System II

(McMurchie et al., 1972) ethylene synthesis during the early

stages of ripening occurs in the locular gel which, while not

containing PG, is temporally advanced over pericarp tissue in

terms of cell-wall changes (Huber and Lee, 1986) and lycopene

accumulation. Therefore, the absence of PG in the locular gel

argues against an initiative role of the enzyme in ripening.

Recent studies have provided evidence that PG may

contribute to the production of System II ethylene in pericarp

tissue. Brecht and Huber (1988) demonstrated that vacuum

infiltration of pectin fragments generated from autolytically

active cell wall from ripe tomato fruit into pre-ripe tomato

fruit enhanced ethylene biosynthesis, possibly due to an

elicitorr' effect. The response was specific in that only

certain size classes were effective. That PG is not asso-

ciated with ethylene production in locular gel does not weaken

the tenability of the idea of PG-enhanced System II ethylene.

Kim et al. (1987) demonstrated that infiltration of galactose

into pre-ripe tomato fruit, at levels present in ripening

fruit due to degradation of wall galactan polymers, enhanced

ethylene biosynthesis. Thus it is possible that a variety of

products of wall origin may contribute to enhanced ethylene


biosynthesis. Numerous studies, primarily those directed at

the biochemistry of host-pathogen interactions (Chappell et

al., 1984; Roby et al., 1985, 1986; VanderMolen et al., 1983),

have demonstrated the capacity of pectin and other carbohy-

drate fragments to elicit a variety of responses, including

ethylene synthesis.

If PG serves in a regulatory manner in fruit ripening,

there would presumably be a high correlation between PG

activity and ethylene production during ripening. Similarly,

one might anticipate a high relationship between PG levels and

firmness rather than the general trends previously reported

(Brady et al., 1985; Hobson, 1965). However, these invest-

igators measured firmness of whole fruit firmness when

investigating the relationship between enzyme (PG) levels and

changes in texture. When measuring whole fruit firmness, the

relationship between enzymes and texture may be confounded by

the contribution to fruit texture of the locular material and

internal morphology of the fruit.

In this study, the relationship between whole fruit and

tissue (pericarp) texture and how these parameters are

correlated with PG levels were examined. Additionally, four

genotypes differing in firmness were examined to ascertain

whether a relationship exists between PG levels and autocata-

lytic (System II) ethylene production.

Materials and Methods

Plant Material

Eleven tomato genotypes were grown at the University of

Florida, Gulf Coast Research Center, Bradenton, FL. Fruit

were harvested at selected developmental stages, determined

by comparing external fruit color with a United States Dept.

Agri. visual aid (United States Dept. Agri. Visual Aid TM-L-

1, The John Henry Company, P.O. Box 17099, Lansing, MI). On

the basis of initial firmness evalautions of all genotypes,

fruit from 4 selected genotypes (Rutgers, Sunny, Flora-Dade,

and FL-7136) were harvested green and at their typical full-

red size. Fruit were rinsed with tap water to remove field

debris, dipped in 1% hypochlorite, and rinsed with deionized

water. After air drying, the fruit were held in storage rooms

and ripened at 23C, or used immediately. Fruit that had not

begun to ripen within 1 week were eliminated. At desired

stages of development, fruit were sectioned through the

equatorial plane and each half quartered by cutting through

the polar (stem-blossom) axis. The blossom and stem scars and

locular and placental material were removed and discarded.

Fruit pieces were randomized within genotypes and 1 kg of

material was stored in high-density polyethylene bags at -30C.

Additional fruit were utilized for firmness measurements as

described below.

Firmness Measurements

Fifty fruit per cultivar were harvested at the red-ripe

stage and evaluated for whole fruit firmness using a Cornell

Firmness Device (Hamson, 1952) as modified by Gull (1987).

Measurements were taken at 3 equidistant points along the

equatorial plane. A 1 kg weight was applied and deflection

in millimeters noted at 5 seconds. Data were expressed using

the equation

Percent Deformation = (d X h"') X 100

where d = deflection after 5 seconds and h = original height

(diameter at point of applied force). For pericarp firmness

determinations, discs were cut from the outer pericarp

equatorial wall using a 15-mm cork borer. The discs were

trimmed to a thickness of 5 mm by excising the peel and

exocarp, then compressed to 3 mm through the short axis.

Compression was accomplished with a 5-cm diameter plate

attached to an Instron Universal Testing Instrument (Model

#1132, Instron Corporation, 100 Royall Street, Canton, MA)

equipped with a 50-kg mechanical load cell. Crosshead speed

was 20 cm min'. Peak heights over the course of travel were

recorded and data were expressed as the reciprocal of Newtons

disc''. Out of the 11 genotypes screened, Rutgers, Sunny,

Flora-Dade, and a non-released line, FL-7136 were selected as

representative of soft, 2 mid-range, and very firm fruit,


Respiration and Ethylene Determinations

Respiration was determined by placing individual fruit

(10 per cultivar) in 500-ml glass containers with lids fitted

with rubber septa. Samples of 0.5 ml were drawn from the

container atmosphere after 30 min and analyzed for CO2 using

a Fisher gas-partitioner (model #1200, Fisher Scientific

Company, 711 Forbes Avenue, Pittsburg, Penn.). Column and

injection port temperatures were 60 and 87C, respectively.

Data were expressed as ml CO2 [(kg FW) hour] Ethylene

production was determined by sampling as above, analyzed with

a Photovac model 10A10 (Photovac Inc., Unit 2, 134 Doncaster

Avenue, Thornhill, Ontario), and expressed as ul ethylene [(kg

FW) h] Both ethylene and CO2 production were monitored

throughout ripening at 24-hour intervals.

Preparation of Enzymicallv-active Cell Walls

Enzymically-active cell walls were isolated from pink and

red fruit pericarp tissue and release of uronic acids deter-

mined as described by Rushing and Huber (1984), with the

exception that the incubation temperature was 30C.

Extraction and Assay of Polvyalacturonase

Polygalacturonase was extracted by homogenizing 40 g of

partially thawed pericarp in 40 ml of cold (5C) 100 mM Na-

acetate, 2.4 M NaCl, pH 6.5, in a Sorvall Omni-mixer in ice

for 2 min at full speed. The homgenate was held on ice for

1 hour, filtered through 2 layers of cheesecloth, and the


filtrate centrifuged at 39200 g (JA-20 rotor) for 20 min at

5C. The pellet was discarded and the supernatant brought to

75% saturation with solid ammonium sulphate and held on ice

for 2 hours. The suspension was centrifuged at 39200 g for

25 min at 5C and the supernatant discarded. The pellet was

resuspended in 2 ml of the extraction buffer and centrifuged

at 490 g for 15 min. A 1 ml aliquot of the supernatant was

desalted on a Sephadex G-25 column (1.6 x 15 cm) equilibrated

with 50 mM Na-acetate, 150 mM NaCl, pH 4.5. Fractions of 2

ml were collected at a flow rate of 50 ml hour"'. Presence of

protein in each fraction was determined visually after

addition of 100 ul of Coomasie reagent (Bradford, 1976) to an

equal volume of aliquots of each fraction. Fractions which

contained protein elutionn volume 20 through 40 ml) were

combined. Polygalacturonase activity was determined using a

reaction mixture containing 100 ul of the desalted protein

preparation and 400 ul of ethanol-purified polygalacturonic

acid (2 mg ml'1) in 30 mM Na-acetate, 150 mM NaCl, 0.01%

Thimerosal, pH 4.5. Enzyme activity after 1 hour at 34C was

determined by the reductometric method of Milner and Avigad

(1967). Protein was determined by the method of Bradford

(1976) using bovine serum albumin as a standard. Activity of

PG was expressed as umoles galacturonic acid equivalents [(GFW

pericarp) hour]'.

Results and Discussion

Over 3 growing seasons, 'Rutgers' fruit were the softest,

'Sunny' and 'Flora-Dade' were intermediate in firmness and

'Fl-7136' were the firmest of the 4 genotypes tested (figures

4-1 and 4-2). All fruit softened as they ripened; however,

differences in firmness among all genotypes were not consis-

tent or continuously significant over the course of ripening

when determined by whole fruit firmness (figure 4-1).

Initiation of softening varied according to genotype. In

whole fruit, softening was not readily apparent until the pink

stage of development in 'FL-7136', the breaker stage in

'Sunny' and 'Flora-Dade', and the mature-green stage in

'Rutgers'. In addition, the rates of change in firmness

varied within and among genotypes. The change in firmness of

1 midrange ('Flora-Dade') and the firm genotype ('FL-7136')

were nearly linear once softening had been initiated. In the

other midrange ('Sunny') and the softest genotype ('Rutgers'),

the pattern appeared sigmoidal. In contrast to whole fruit

firmness, differences in pericarp firmness were discrete past

the breaker stage and rates of change nearly consistent

throughout ripening (figure 4-2). Softening initiation

occurred past the breaker stage of development in all geno-

types. Firmness at the early developmental stages (mature-

green and breaker) was not indicative of firmness at the later

stages of development as measured by whole fruit (figure 4-1)

or tissue firmness (figure 4-2). Absolute differences between

genotypes and the rate of change in firmness within genotypes

varied according to measurement technique. Since outer

pericarp firmness remained unchanged up through the breaker

stage (figure 4-2), the firmness changes occurring in whole

fruit prior to the breaker stage (figure 4-1) may be inter-

preted as the contribution of changes occurring in internal

tissues. The locular gel, for example, undergoes significant

textural changes during the transition from the immature-green

to the mature-green stage (Huber and Lee, 1986). Other

factors possibly contributing to whole fruit firmness include

carpel morphology, thickness of outer fruit wall, and overall

water status of all fruit tissues. In this study, overall

correlation between whole fruit firmness and tissue firmness

was 0.668. Although the general order of firmness was

conserved between techniques, at the pink and red-ripe stage

differences in firmness as depicted by the two measurement

techniques was dramatic (figures 4-1 and 4-2).

Rushing and Huber (1984) reported on the use of enzymic-

ally active cell wall as a system for studying tomato fruit

softening. They demonstrated that cell wall autolysis

increases with ripening, reflecting progressively higher

levels of endogenous PG. Consistent with this relationship

between autolysis and tissue PG levels, autolysis (table 4-1)

was also related to tissue firmness (figure 4-2) among the 4

genotypes examined here. Genotypes with greater autolytic

activity had softer pericarp tissue. Autolysis in wall

prepared from the 4 genotypes was highly correlated with

levels of tissue PG in pink (r = 0.969) and red (r = 0.900)

pericarp tissue (table 4-1). Similarly, PG activity was

highly correlated with pericarp tissue firmness (r = 0.920).

Previous workers (Brady et al., 1984; 1985; Hobson, 1965,

1981), measuring whole fruit, found only general trends in

relating PG activity to firmness. Consistent with these

reports, PG was only moderately correlated with firmness of

whole red-ripe fruit (r = 0.757). In a recent study, Smith

et al. (1988) were able to reduce active PG levels in tomato

fruit to 10% of that in control fruit using antisense RNA

inhibition. Yet, they reported no concurrent decrease in

softness as measured by compression of whole fruit. As

previously mentioned, changes in whole fruit firmness are not

a good indication of changes in tissue firmness. When

relating these changes to metabolic activity in pericarp

alone, they are indirect at best. Smith et al. (1988) point

out underlying causes of changes in texture are complex. In

this study, the lowest levels of PG were found in the very

firm genotype 'Fl-7136' (table 4-1). These levels (1/4 of

that found in the softest genotype 'Rutgers') were apparently

insufficient to soften the tissue of 'Fl-7136 over the course

of normal ripening (figure 4-2). However, on a whole fruit

basis (compression), softening of this genotype did indeed

occur (figure 4-1). Those genotypes which characteristically

have a large locule to pericarp ratio or vary in size within

the genotype, would show even greater disparity between PG

levels and whole fruit firmness, making comparisons between

genotypes even more difficult.

Ethylene production showed a pattern similar to firmness

in that differences between genotypes were not significant

until after the breaker stage (figure 4-3). Maximum ethylene

production occurred at the pink stage, at which time the

genotypes were concurrently segregated into firm, inter-

mediate, and soft categories. In agreement with Poovaiah and

Nukaya (1979), firmer fruit had lower levels of respiration

at the climacteric peak (figure 4-4). Polygalacturonase

activity was highly correlated (r = 0.929) with peak ethylene

production (table 4-2). This is supportive evidence that

these 2 processes are related. Brecht and Huber (1988)

reported that the onset of ripening in green tomato fruit was

advanced by infiltration of pectin fragments generated by

autolytically-active cell wall from ripe tomato pericarp.

Initiation of ripening in relation to controls was indicated

by advanced autocatalytic ethylene production and lycopene

synthesis. This work was supported by Baldwin and Pressey

(1988) who reported that tomato fruit vacuum infiltrated with

purified tomato PG displayed elevated levels of ethylene

production over control fruit. In the present study, in

addition to greater autocatalytic ethylene production,

genotypes with higher PG had greater levels of respiration and

required less days to ripen (table 4-2, r = 0.805 and -0.791,

respectively). Kagan-Zur and Mizrahi (1987) found that

tetraploid tomato fruit had greater levels of PG, climacteric

ethylene production, and required less days to reach the

climacteric maximum than did diploid fruit of the same

isogenic line with similar pre-climacteric levels of ethylene

production. They pointed out that if ethylene is the trigger

and coordinator of ripening processes, then fruits of the same

genetic background with identical pre-climacteric rates of

ethylene production should have similar ripening responses.

However, DellaPenna et al. (1987) reported finding low

levels of PG in the non-ripening mutant rin, which has been

reported to show no enhanced ethylene production during

development. This appears to indicate that PG activity and

ethylene production are not coupled through the activity of

PG-generated elicitors. However, the presence of PG mRNA or

its product may not, in rin, be coordinated with production

or activity of 1-aminocyclopropane-l-carboxylic acid (ACC)

synthase or ethylene-forming enzyme (EFE), thereby rendering

the tissue incapable of enhanced ethylene production in

response to PG.

Numerous studies have provided evidence for a relation-

ship between stress and ethylene (Lee et al., 1987; Roby et

al., 1985, 1986; Romani et al., 1968; VanderMolen et al.,

1983). Romani (1984) suggests that ethylene produced during

fruit ripening and senescence is in response to stress imposed

by the genetically programmed demise of various cellular

entities, including membranes and cell walls. It is his view

that ethylene serves as a ripening augmentor or accelerator,

but is not the primary causative agent. Other factors, such

as tissue sensitivity to ethylene, determine the capacity of

a fruit to ripen. In climacteric fruit, one of the ripening

parameters is the production of System II ethylene. In

addition to autocatalytic ethylene production, climacteric

fruit show a burst of respiration during senescence. Romani

(1984) proposed that this increase in respiration occurs as

a response to ripening stresses and is a homeostatic response

of the mitochondrion.

These studies demonstrate that there is a highly sig-

nificant relationship between PG levels and pericarp tissue

firmness and, more interestingly, between PG levels and System

II ethylene. Whether the latter relationship is a direct one

is yet unknown. The extensive evidence of pectin-fragment-

mediated increases in ethylene production in host-pathogen

interactions (Roby et al., 1985, 1986; VanderMolen et al.,

1983; West et al., 1984) argues in favor of a direct relation-

ship. The data are consistent with this theory and indicate

a strong dose-response regulatory relationship between PG

activity and rate and intensity of ripening.

Table 4-1. Polygalacturonase activity [umoles galacturonic
acid equivalents (GFW pericarp) hour''] and autolysis [pectin
released as (ug galacturonic acid equivalents) (mg cell DW) 1,
after 4 hours at 30C] in pink and red tomato fruit. Data in
columns followed by the same letter are not significantly
different at the 0.01 level.




RUTGERS 3.57a 76.Oa 4.46a 105.6a

FLORA-DADE 1.86b 62.9b 2.43b 96.8b

SUNNY 1.59b 55.3c 2.02b 73.7c

FL-7136 0.95c 46.4d 1.22c 62.7d

r = 0.967 r = 0.900

Table 4-2. Polygalacturonase activity [umole galacturonic
acid equivalents (GFW pericarp) hour'1] of pink tomato fruit
and rate (days to ripen) and intensity [uL ethylene and ml CO2
(kg FW) hour" at climacteric peak] of ripening in tomato
fruit. Data in columns followed by the same letter are not
significantly different at the 0.01 level.


RUTGERS 3.56a 5.0a 6.82a 31.Oa

FLORA-DADE 1.86b 6.4b 4.31b 21.5b

SUNNY 1.59b 6.6b 4.03b 28.Oc

FL-7136 0.95c 12.4c 0.65c 17.5d





10.0 -

7.5 -

5.0 -

2.5 -

0.0 -

**** FL-7136


Figure 4-1. Whole fruit firmness, as determined by modified
Cornell device, of tomato fruit at selected stages of develop-
ment. IM, MG, B, P, and R are immature-green, mature-green,
breaker, pink, and red, respectively. Bars are SE of 30
fruit. Higher values reflect softer fruit.


0.020 -





0.005 -

j*t r\/~\r\


***** FL-71 36


Figure 4-2. Firmness (resistance to compression) of pericarp
discs, as determined by Instron device, from tomato fruit as
selected stages of development. MG, B, P, and R are mature-
green, breaker, pink, and red, respectively. Bars are SE of
30 discs. Data are expressed as the reciprocal of Newtons
disc"1. Higher values reflect softer fruit.




***** FL-7136

I I I i I

Figure 4-3. Ethylene production of tomato fruit at selected
stages of development at 23C. MG, B, T, P, LR, and R are
mature-green, breaker, turning, pink, light red, and red,
respectively. Bars are SE of 24 fruit.




r?' 25-



**** FL-7136


Figure 4-4. Respiration of tomato fruit at selected stages
of development as measured by carbon dioxide evolution at 23C.
MG, B, T, P, LR, and R are mature-green, breaker, turning,
pink, light red, and red, respectively. Bars are SE of 24

I 1 I 1



Little is known regarding the biochemical basis of

mealiness in horticultural crops. In apple (Malus domestic

Borkh.) fruit, exogenously applied calcium has been shown to

modify mealiness in cortical tissue by reducing the loss of

the cell wall-middle (CW-ML) (Glenn and Poovaiah, 1987). It

remains unknown, however, if changes in endogenous calcium

concentrations in apple lead to mealiness in vivo. Ben-Arie

and Laves (1971) demonstrated that mealiness ('wooliness') in

peach (Prunus persica L.) was due to a decreased solubiliza-

tion of pectic substances. Buescher and Furmanski (1978)

proposed that the basis for this was the failure of the tissue

to develop adequate pectinesterase and polygalacturonase (PG)

activity. From these limited studies, two very different

forms of mealiness can be described. In apple, mealiness is

associated with softening and loss of the CW-ML, and in peach,

a failure to solubilize the pectins of the CW-ML.

Scanning electron micrographs of the cut surface of mealy

apple tissue (Glenn and Poovaiah, 1987) reveal discreet,

intact cells which are not ruptured. Non-mealy tissue cells

are cut open and apparently release their contents. Because

of this, the mealy tissue appears 'dry' and the non-mealy

tissue appears 'wet'. Preliminary investigations (Ahrens,

unpublished) involving scanning electron micrographs of mealy

and non-mealy tomato (Lycopersicon esculentum Mill.) tissue

showed similar results. However, initial studies have

revealed that, unlike apple fruit, in tomato fruit mealiness

and firmness (and concommitant PG activity) are not associated

(Ahrens, unpublished).

Changes in cell-wall components other than polyuronides

in relation to the development of mealiness have not been

investigated. Modifications in the cellulose fraction of the

cell wall could conceivably lead to the mealiness trait.

Although Cx-cellulase activity has been demonstrated in tomato

(Buescher and Tigchelaar, 1075' Hall, 1963, 1964; Pharr and

Dickinson, 1973; Poovaiah and Nukaya, 1979; Sobotka and

Stelzig, 1974; Sobotka and Watada, 1971) and avocado (Persea

americana Mill.) (Hatfield and Nevins, 1986; Pesis et al.,

1978) fruit, no evidence exists for the in vivo degradation

of native crystalline cellulose. It appears unlikely that

modifications of this cell-wall fraction are responsible for

the development of mealiness.

Hemicelluloses have been reported to be modified during

ripening in tomato (Huber, 1983a) and strawberry (Fraqaria X

ananassa Duch.) (Huber, 1984). Since these changes were

monitored throughout ripening in single cultivars, no infer-

ence can be drawn regarding mealiness and these modifications.

It is conceivable that relative differences in modification

of this cell-wall fraction are responsible for the mealiness

trait in tomato fruit.

The purpose of this study was to examine several tomato

cultivars known to vary in mealiness and relate differences

in their cell wall composition to differences in the expres-

sion of this textural trait.

Materials and Methods

Plant Material

Four tomato genotypes varying in intrinsic mealiness were

grown at the Gulf Coast Research and Education Center,

Bradenton, Florida, in the Fall of 1987. Fruit were harvested

green, the calyxes and pedicels removed, and transported in

an air-conditioned vehicle to Gainesville, arriving within 3.5

hours. Tomatoes were then lightly rinsed in tap water to

remove field debris, dipped in 1% hypochlorite, and rerinsed

in tap water before air drying at ambient temperature. Fruit

were graded by removing misshapen, damaged, and under- and

oversized fruit, placed on trays stem-end down, and allowed

to ripen at 20C. When ripe, fruit were halved through the

equatorial plane and each half was quartered through the polar

(stem-blossom) axis. All locular material was removed and

discarded, as well as the stem and blossom scar material, and

the columnella. The remaining pericarp was stored at -30C.

To guard against inclusion of immature-green fruit in the

experiments, fruit in which ripening had failed to commence

by 1 week were discarded.

Mealiness Determination

An objective test for determining mealiness of tomato was

developed. Discs were prepared from tomato pericarp using a

5-mm cork borer and trimmed to 3 mm by removing the epidermis

and outer locular membrane. Ten discs per cultivar (taken

from the equatorial region of 3 to 4 fruit) were agitated in

50 ml of 0.60 M Sorbitol in a 125-mi erlenmeyer flask with a

gyrotory shaker (125 rpm) at room temperature. Aliquots of

10 ml were removed from the bathing solution after 1 hour ,

pipetted into a 25-ml petri dish, and total counts of cells

in the aliquots were taken using a light microscope (Nikon

Diaphot type 108). These numbers were then used to quantify


Preparation of Cell Wall

Twenty-five grams of partly thawed pericarp (with peel

removed) from tomatoes at selected stages of development were

homogenized in 100 ml of 95% ETOH using a Sorvall homogenizer

(full speed for 2 min) set in an ice bath. The homogenate was

rinsed through Miracloth (Biochemical Corp., La Jolla, CA)

with 500 ml 5C deionized water and 5 X 100 ml 5C acetone, then

resuspended in 100 ml 5C acetone. The suspension was vacuum-

filtered through Whatman glass-fiber (GFC) filter paper and

washed with an additional 1 liter of 5C acetone. The cell

wall material was transferred to a petri dish and allowed to

air dry for 2 hours, followed by oven drying (30C) for 12