BNVE’RORKENTAL ANS} STATISTICAL
CGNSIEERKFIMS EN EVALUATING THE
EMERIHESS OF FWWER BEDS FBGM
SELECTED BEGHBUSH
BLUEBERRY CULHVARS

TSIONS for “to Dogma of M. 5..
MCHEGAR STATE UNIVERSITY
Harry Clair Bittenbender
3974

 

THESIS

 

 

 

‘,—-M.--. i “V -,-

ABSTRACT

ENVIRONMENTAL AND STATISTICAL CONSIDERATIONS IN
EVALUATING THE HARDINESS OF FLOWER BUDS FROM
SELECTED HIGHBUSH BLUEBERRY CULTIVARS

By

Harry Clair Bittenbender

The flower bud hardiness of seven commercial highbush

blueberry cultivars (Vaccinium australe Small) was deter-

 

mined from fall to spring via field survival and controlled
freeze data. Good agreement was shown between hardiness
rank of cultivars determined by these methods and the
results of previous researchers. Quantal response data
were adapted to fit the Spearman—Karber equations. The use

of these equations enabled the T temperature at which

50’
50% bud mortality occurs, and the variance around the T50
to be computed. A computer program was written to determine
the T50, T50 mean, '8' statistic, and field survival from
the data collected from field observations and controlled
freezes. The effects of induced changes in bud moisture
content and storage temperature on hardiness were studied.
The results seem to indicate that separate mechanisms exist

for dehardening due to increased moisture content and

dehardening by high temperatures. Multiple regression

Harry Clair Bittenbender

equations using components of hardiness, based on one
year's phenological and environmental data, were created
to predict TSO'S within 1°C of observed TSO's for a given
cultivar and date.

ENVIRONMENTAL AND STATISTICAL CONSIDERATIONS IN
EVALUATING THE HARDINESS OF FLOWER BUDS FROM

SELECTED HIGHBUSH BLUEBERRY CULTIVARS

By

Harry Clair Bittenbender

A THESIS

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

MASTER OF SCIENCE

Department of Horticulture

197A

This thesis is dedicated to my advisor, Stan Howell,

who has always treated me as his equal, both as a
researcher and a man. His door has always been open,

his ear always ready to listen.

ii

and

ACKNOWLEDGMENTS

I wish to thank the following persons: my committee,
Drs. Stan Howell, Bob Andersen, and Don Ramsdell, who have
guided my work and constructively criticized my writing;
Dr. D. R. Dilley, in whose laboratory I worked; Dr. S. K.
Ries, for use of his computer; John Nelson and Ron Bodke,
research director and grower-cooperator, and other members
and staff of the Michigan Blueberry Growers Association; my
parents, Clair and Nina Bittenbender; and finally fellow
graduate students and friends who have aided, guided, and
provided moral support in my research, C. Dogras, D. Cain,
B. Stergios, S. Engstrom, P. Bourland, J. Stuurwold,

S. Stackhouse, and V. Wilson.

iii

TABLE OF CONTENTS

Introduction
Literature Review
Preface to Papers

Adaptation of the Spearman-Karber Method for Estimating the
T50 of Cold Stressed Flower Buds

Interactions of Temperature and Moisture Content on
Deacclimation of Post Rest Flower Buds of Vaccinium
australe Small, Highbush Blueberry

 

Cold Hardiness of Flower Buds from Selected Highbush Blue-
berry Cultivars (Vaccinium australe Small)

 

Predictive Environmental and Phenological Components of
Flower Bud Hardiness in Highbush Blueberry (Vaccinium

australe Small)

 

Appendix

iv

INTRODUCTION

Cold and drought stresses are considered by some to be
the most limiting factors of plant distribution on the
earth's surface. The effects of cold stress, e.g. winter
injury, sun scald, spring frosts, or its interactions with
other environmental components such as diseases, nutrition,
water status, and mechanical injury, limit horticultural
production in the northern half of the United States. In
Michigan, peach, cherry, and highbush blueberry farms pro-
tected by the moderating effects of Lake Michigan are being
forced into the state's interior by land development of the
Lake Michigan shoreline. The challenge for the fruit
industry now and increasingly in the future will be the
solving of the problems of cold hardiness through physio-
logical and genetic research. The extent and perhaps the
existence of a viable fruit industry in Michigan will depend

upon how successful that research effort is.

General Hardiness Theory

 

Recent reviews have dealt with the question; what is
freezing injury? (“0,58). The following may be viewed as a
brief summary of their conclusions. The freezing of inter—

cellular water and subsequent concentration of intercellular

solutes causes a drop in water vapor pressure outside the
cell. This vapor pressure and solute gradient initiates
pure water flow from the cell to maintain an equilibrium.
The alternative to cellular dehydration, i.e. water loss, is
intracellular freezing, which is lethal. The changes that
result within the cell are reduced water content with
resultant concentration of solutes which can cause precipi—
tation of proteins (salting out) if severe enough. Other
changes include pH shift and reduced distance between macro—
molecules. These can result in death of the cell even
though intracellular freezing did not occur.

Though the exact freezing resistance mechanism(s) is
not known, certain components of hardiness that are respon-
sible for induction, maintenance, reduction of hardiness are
known. The effects of environmental stimuli, such as photo-
period and air temperature are very broad. Documentation
of the role of the short day (SD) photoperiod for inducing

hardiness has been done with apple (28), Hedra helix (6“),

 

Acer negundo (65), and Cornus stolonifera (66). One theory

 

 

 

of hardiness suggests that acclimation occurs in two stages
(28,58). The first stage of hardiness is induced by SD,
and the second stage by frost. This second stage might be
better viewed as a zone, the upper limit being the minimum
hardiness level (MHL) of stage two, the lower limit being
set by the inherent cold resistance of the species (A8).

In peach, flower bud hardiness fluctuates below the MHL in

response to air temperature, while the MHL is maintained

until rest (chilling hours) is satisfied, regardless of
naturally occurring air temperatures (A8). Fluctuating air
temperatures have been correlated with peach bud (22,Al,A8,
52) and apple wood (3A) hardiness changes in the winter,
during and after rest. Air temperatures are also effective
hardiness regulators in other species. Air temperatures,
lower than in their natural habitats hardened tropical

willows to —50°C (55). Hardiness of Fucus vesiculous L.,

 

a brown, marine algae was also correlated with air
temperature (A6).

Deacclimation of buds and woody tissues under con-
trolled, artificially high temperatures (20°C—30°C) has been
observed in apple (27), peach and cherry flower buds (52),

Forsythia (26) and raspberry (6). Findings from high tem—

 

perature studies are very important in regards to proper
handling and storage of materials to be used in hardiness
determination work prior to cold stress treatments.

Non-cyclic morphological changes that occur as part of
flower development are not analogous to the cyclic growth
patterns of stems, certain changes appear to render the bud
less hardy. Pollen maturity and appearance of pink in the
bud indicate rapid transition from hardy to the tender
status of peach buds (50).

Water content of peach flower buds (29), Rhododendron

 

florets (25,38) and wheat and barley crowns (A2,70) has been
shown to be directly related to the killing temperature.

Artificially dehydrated Cornus stems increased in hardiness

by as much as 12°C from a 15% moisture loss (16). Buds are
highly permeable to water, during or after rest (2). Bud
scales and stem pith proximal to the bud seem to act as ice
sinks during freezing, water flows from the flower primordia
to ice masses in the scales (20). Intercellular ice has not
been observed in flower primordia except at the killing tem—
perature (20,38,39,67).

Another method to determine the temperature at which
tissue freezes, other than visual observation, is exotherm
analysis. This technique allows the experimenter to plot
tissue temperature as a function of time. Curves are pro-
duced which are deflected by the heat of water crystalliza-
tion at points where water in the plant tissues freezes (28,

61). Exotherm analysis of Rhododendron showed that the

 

scales and peduncle have a much higher exotherm temperature
than the florets. Florets had only one exotherm, and that

was lethal (25). Woody tissues, however, have 2-3 exotherm
when living and only one when dead (58,61). These observa-
tions appear to underline possible differences between wood

and bud hardiness.

Hardiness in the Highbush Blueberry

 

The cultivated highbush blueberry formerly Vaccinium

 

corymbosum L. now predominantly'y. australe Small, is grown

 

on scattered sites through the northern half of the United
States, Washington, Minnesota, Michigan, Indiana, New

Jersey, New Hampshire, and Maine. Environmental factors

limiting production are length of growing season, high soil
pH, low water table, and cold temperatures (21,23). In
Michigan most commercial sites are found on a strip of land
along Lake Michigan that is approximately A0 km. wide in
southwest Michigan and 15 km. wide in the north at Grand
Haven. Cold injury and short growing season limits plant—
ing farther north. Tests conducted in the northern lower
peninsula and in the Upper Peninsula show that 'Rancocas'
is the only highbush cultivar able to survive (32). How-
ever, 'Rancocas' is no longer planted commercially for
other reasons.

Cold injury can occur during any part of the cold
season, the period of most frequent injury varies among
geographic locations. Cold injury occurs in the fall in
Minnesota (5,13), in the winter in New York and Arkansas
(9,AA), and in spring in Michigan, Massachusetts and
Washington (3,30,31,68).

The first reported hardiness experiments with the high-
bush consisted of growing plants in pots without snow cover
to test their ability to survive the winter and to determine
whether a cold requirement existed (13). A cold requirement
of at least 1,000 hrs. at or below 7°C has been shown to
exist for commercial cultivars (1A). The answer to the

question; why does the lowbush V. canadense Lam and'V.

 

angustifolium Aiton. (V. lamarckii Camp.) have greater

 

 

winter survival compared to the highbush seems to be that

a small physiological hardiness edge exists in the lowbush
plus insulation in the snow cover (5,32).

The remaining blueberry hardiness papers fall in three
categories; cultivar comparisons via field observations (3,
A,9,16,l7,19,30,3l), controlled freeze comparisons of
cultivar differences in wood hardiness (12,53,68) and
reports of hardier germplasm for breeding (15,18,33,A3,5A,
59). Few researchers have worked on flower bud hardiness
in the highbush blueberry (3,A,30,68). To date one study
has included controlled freeze comparisons of cultivar
differences based upon flower bud hardiness (68). Some data
are available that correlate wood and flower bud hardiness
from field observations (A). Breeding programs directed at
cold hardiness improvements have utilized interspecific
crosses with V. australe. These sources of germplasm are,

the Ashworth blueberry, V, corymbosum (18,21,A3,59), V.

 

 

lamarckii, a northern lowbush (21,33,A3) and V, uliginosum

 

L., the bog whortheberry (5A). Present breeding efforts are
hampered by the inability to identify hardy seedlings with—
out test winters (71).

Several methods have been reported for breeding and
screening for cold hardiness in tree fruits (37,57,60). A
screening program for either flower buds or wood has not
been developed for blueberries. Such a development would
facilitate the current breeding programs by providing
standardized test winters every year. A screening program

for cold hardiness of flower buds requires informational

inputs from at least three areas. Firstly, background
hardiness information should include seasonal acclimation
patterns, environmental controls, and phenological charac—
ters related to hardiness. Secondly, freezing techniques
and viability criteria need to be developed for specific
tissue(s) in question. Thirdly, improved statistical
methods of handling for hardiness data need to be employed.

The relationship of acclimation patterns to environ-
mental control has been discussed earlier. The establish—
ment of the critical hardiness period for a site reflects
knowledge of micro—climate and the cultivar hardiness cycle.
Phenological characters that closely correlate with hardi-
ness could be useful in selection. Some characters
reported in the literature are moisture (2A,29,38,A2,69)
and anthocyanin contents (A5,56). Others that should be
evaluated include leaf drop, bloom, and harvest dates.
These could provide criteria for selecting hardy seedling
progeny.

The importance of using the proper freezing technique
and viability criteria should not be overlooked (62). Tech—
niques and criteria that have been developed for use with

peach (ll,A9,51) and Rhododendron (2A,29,38) buds may be

 

applicable to the highbush blueberry flower bud.

The choice of statistical methods for analysis is
especially important when working with hardiness data. Data
from controlled freezing tests are normally plotted on a

temperature-survival curve, (T-S), with percent tissue

survival as a function of tissue temperature. The tempera-
ture at which 50% of the cold stressed material is killed

(T50), is analogous to the LD of toxiological studies and

50
is the most statistically accurate estimate of any point on
the T—S curve (25,37,51). However, some workers have felt
that T10 and T90, temperature at 10% and 90% survival
respectively, better describe the hardiness of a tissue as
they provide a range on the T-S curve (3A,35). Several

methods are available that estimate T 's from quantal

50
response data, which is the type data normally collected
in flower bud studies (1,7,2A,25). A determination should
be made concerning the strengths and weaknesses of these

methods for use in a program designed to breed for hardiness

improvement (1,7,8,10).

10.

11.

Literature Cited

Armitage, P. and I. Allen. 1950. Methods of estimating
the LD50 in quantal response data. J. of Hygiene A8:298-
322.

Badanova, K. A. and B. B. Vartapetyan. 1967. Permea—
bility to water of the protoplasm of dormant buds of
woody plants. Doklady Akademi Nauk SSSR 176(2):A76-A77.

Bailey, J. S. l9A9. Frost injury to blueberries.
Fruit Var. and Hort. Dig. A:98.

Bittenbender, H. C. 1973. Varietal hardiness data from
a blueberry test planting, 1970-73. Unpublished data of
the Mich. Blueberry Growers Assoc., Grand Junction, MI

A9056.

Brierly, W. G. and A. C. Hildreth. 1928. Some studies
on the hardiness of certain species of Vaccinium. Plt.
Phys. 3:303-308.

, R. H. Landon, and R. J. Stadtherr.
1952. The effect of daily alterations between 27° and
39°F on retention or loss of cold resistance in the
Lagham raspberry. Proc. Amer. Soc. Hort. Sci. 59:173—
17 .

 

Bross, I. 1950. Estimates of the LDSO: A critique.
Biometrics 6:Al3-A23.

Brown, B. W. 1966. Planning a quantal assay of potency.
Biometrics 22:322-329.

Cain, J. C. and G. L. Slate. 1953. Blueberries in the
home garden. Cornell Ext. Bull. 900.

Chang, P. C. and E. A. Johnson. 1972. Some distribu—
tion free properties of the asymptotic variance of the
Spearman estimator in bioassays. Biometrics 28:882—889.

Chaplin, C. E. 19A1. Some artificial freezing tests of
peach fruit buds. Proc. Amer. Soc. Hort. Sci. 52:121-
129.

12.

13.

1A.

15.

16.

l7.

18.

19.

20.

21.

22.

23.

2A.

25.

26.

10

Constante, J. F. and B. R. Boyce. 1968. Low tempera-
ture injury of highbush blueberry shoots at various
times of the year. Proc. Amer. Soc. Hort. Sci. 93:267—
272.

Coville, F. V. 1960. Experiments in blueberry culture.
USDA Plt. Ind. Bull. 193. »

Darrow, G. M. 19A2. Rest period requirements of blue-
berries. Proc. Amer. Soc. Hort. Sci. A1:189-19A.

. 1960. Blueberry breeding. Nat. Hort.
Mag. 39(1) 15-29.

 

and J. N. Moore. 1962. Blueberry growing.
USDA Farmers Bull. 1951. Revised.

 

and . 1966. Blueberry growing.
USDA Farmers Bull. 1951. Revised.

 

 

, L. Whitton, and D. H. Scott. 1960. The
Ashworth blueberry as a parent in breeding for hardiness
and earliness. Fruit Var. and Hort. Dig. 1A:A3-A6.

 

, R. B. Wilcox, and C. S. Beckwith. 19AA.
Blueberry growing. USDA Farmers Bull. 1951.

 

Dorsey, M. J. 193A. Ice formation in the fruit bud of
the peach. Proc. Amer. Soc. Hort. Sci. 31:22—27.

Eck, P. and N. F. Childers. 1966. Blueberry culture.
Rutgers Univ. Press, New Brunswick, N.J. 378 pp.

Edgerton, L. J. 195A. Fluctuations in the cold hardi—
ness of peach flower buds during rest period and
dormancy. Proc. Amer. Soc. Hort. Sci. 6A:175-180.

Finney, D. J. 196A. Statistical methods in biological
assay. Second. Ed., Hafner Pub. Co., N.Y., N.Y.

. 1971. Probit analysis. 3rd Ed.,
Cambridge Univ. Press. 333 pp.

 

Graham, R. P. 1971. Cold injury and its determination
of selected Rhododendron species. M.S. thesis, Univ. of
Minn.

 

Hamilton, David F. 1973. Factors influencing deharden-
ing and rehardening of Forsythia x intermedia stems. J.
Amer. Soc. Hort. Sci. 98:221—223.

 

27.

28.

29.

30.

31.

32.

33.

3A.

36.

37.

38.

39.

A0.

11

Howell, G. S. and C. J. Weiser. 1970. Fluctuations in
the cold resistance of apple twigs during spring
dehardening. J. Amer. Soc. Hort. Sci. 95:190—192.

and . 1970. The environ-
mental control of cold acclimation in apple. Plant
Physiol. A5 390- 39”.

 

 

Johnston, E. S. 1923. Moisture relations of peach buds
during winter and spring. Univ. Md. Ag. Exp. Sta. Bull.
255.

Johnston, S. 1939. Resistance of certain blueberry
varieties to frost. Mich. Ag. Exp. Sta. Quart. Bull.
22:10-12.

1939. The frost control problem with
special reference to blueberries. Mich. Ag. Exp. Sta.
Quart. Bull. 22(1):3—10.

 

. 1950. Problems associated with cultivated
blueberry production in northern Michigan. Mich. Ag.
Exp. Sta. Quart. Bull. 33(A):293-298.

 

. 1956. Blueberry breeding in Michigan.
Fruit Var. and Hort. Dig. 11:20.

 

Ketchie, D. 0., C. H. Beeman, and A. L. Ballard. 1972.
Relationships of electrolytic conductance to cold injury
and acclimation in 'Red Delicious' apple trees under
natural conditions during four winters. J. Amer. Soc.
Hort. Sci. 98:257—261.

Lapins, K. 1962. Artificial freezing as a routine
test of cold hardiness of young apple seedlings. Proc.
Amer. Soc. Hort. Sci. 81:26—3A.

Levitt, J. 1956. The hardiness of plants. Academic
Press, N.Y., N.Y. 278 pp.

Lumis, G. P. 1970. Winter hardiness of evergreen
Azalea (Rhododendron cv.) flower buds. Ph.D. thesis,
Mich. State Univ.

 

, R. A. Mecklenburg, and K. C. Sink. 1972.
Factors influencing winter hardiness of flower buds and
stems of evergreen Azaleas. J. Amer. Soc. Hort. Sci.

97(1):l2A-l27.

 

Mazur, P. 1969. Freezing injury in plants. Ann. Rev.
of Plant Physiol. 20:Al9-AA8.

A1.

A2.

A3.

AA.

A5.

A6.

A7.

A8.

A9.

50.

51.

52.

12

Meader, E. M. and M. A. Blake. 19A3. Seasonal trend
of fruit bud hardiness in peaches. Proc. Amer. Soc.
Hort. Sci. A3z91—98.

Metcalf, E. L., C. E. Cress, C. R. 01ein, and E. H.
Everson. 1970. Relationship between crown moisture
content and killing temperature of three wheat and
barley cultivars. Crop Science 10:362—365.

Moore, J. N. 1965. Improving highbush blueberry by
breeding and selection. Euphytica 1A:39-A8.

Moore, J. N. and G. R. Brown. 1971. Susceptibility of
blackberry and blueberry cultivars to winters injury.
Fruit Var. and Hort. Dig. 25:31-32.

 

Parker, J. 1962. Relation among cold hardiness, water
soluble protein, anthocyanins, and free sugar in Hedra
helix L. Plant Physiol. 37:809-813.

1960. Seasonal changes in cold hardiness
of Fucus vesiculosus. Biol. Bull. 119:A7A—A78.

 

 

 

Proebsting, E. L. 1963. The role of air temperatures
and bud development in determining hardiness of dormant
'Elberta' peach fruit buds. Proc. Amer. Soc. Hort. Sci.
83:259—26A.

1970. Relation of fall and winter
temperatures to flower bud behavior and wood hardiness
of deciduous fruit trees. HortScience 5(5):A22-A2A.

 

and H. W. Fogle. 1956. An apparatus
and method of analysis for studying fruit bud hardi-
ness. Proc. Am. Soc. Hort. Sci. 68:6-1A.

 

and H. H. Mills. 1961. Loss of
hardiness by peach fruit buds as related to their
morphological development during the pre-bloom and
bloom period. Proc. Am. Soc. Hort. Sci. 78:10A—110.

 

and T. S. Russell. 1966. A
standardized temperature—survival curve for dormant
'Elberta' peach fruit buds. Proc. Am. Soc. Hort. Sci.
89:85-90.

 

and . 1972. A compari-
son of hardiness responses in fruit buds of 'Bing'
cherry and 'Elberta' peach. J. Amer. Soc. Hort. Sci.
97:802—806.

 

 

53.

5A.

55.

56.

57.

58.

59.

60.

61.

62.

63.

6A.

65.

13

Quamme, H. A., C. Stushnoff, and C. J. Weiser. 1972.
Winter hardiness of several blueberry species and
cultivars in Minnesota. HortSci. 7(5):500-502.

Rousi, A. 1963. Hybridization between Vaccinium
uliginosum and cultivated blueberry. Ann. Ag. Fenniae
2:12-18. -

 

 

Sakai, A. 1970. Freezing resistance in willows from
different climates. Ecology 5A(3):A85—A91.

Steponkus, P. L. and F. O. Lanphear. 1969. The rela-
tionship of anthocyanin content to cold hardiness of
Hedera helix. HortSci. A(1):55-56.

 

Stushnoff, C. 1972. Breeding and selection methods
for cold hardiness in deciduous fruit crOps. HortSci.
7(1):10-13.

Weiser, G. J. 1970. Cold resistance and injury in
woody plants. Science 169:1269—1278.

Whitton, L. 1960. Breeding for winter hardiness. In
Blueberry Research, ed. J. N. Moore and N. F. Childers.
pp. 53-55. Rutgers Univ., New Brunswick, N.J.

Young, R. and C. J. Hearn. 1972. Screening citrus
hybrids for cold hardiness. HortSci. 7(1):1A-18.

McLeester, R. C., C. J. Weiser, and T. C. Hall. 1969.
Multiple freezing points as a test for viability of
plant stems in the determination of frost hardiness.
Plant Physiol. AA:37-AA.

Stergios, B. G. and G. S. Howell, Jr. 1973. Evalua—
tion of viability tests for cold stressed plants. J.
Amer. Soc. Hort. Sci. 98:325—330.

Li, P. and C. J. Weiser. 1969. Increasing cold
resistance of woody stems by artificial dehydration.
Cryobiology 6:270.

Steponkus, P. L. and F. 0. Lanphear. 1967. Factors
influencing artificial cold acclimation and freezing
tests of Hedra helix 'Thorndale'. Proc. Amer. Soc.
Hort. Sci. 91:735—7A1.

 

Irving, R. M. and F. O. Lanphear. 1968. Regulation
of hardiness in Acer negundo. Plant Physiol. A3:9-13.

 

66.

67.

68.

69.

70.

71.

1A

Van Huystee, R. B., C. J. Weiser, and P. H. Li. 1967.
Cold acclimation in Cornus stolonifera under natural

and controlled photoperiod and temperature. Bot. Gaz.
128:200—205.

 

Chen, P. and P. H. Li. 1973. Viability of apple
blossom buds after test freezing. HortScience 8(6):

508-509.

Doughty, C. E. 197A. Cold injury and hardiness of
blueberries in western Washington. Proc. 197A Annual
Meeting North American Blueberry Council.

Levitt, J. 19A1. Frost killing and hardiness of plants.
Burgess Pub. 00., Minneapolis, Minn. 211 pp.

Vasil'yev, I. M. 1961. Wintering in plants. Roger and
Rogers, Washington, D. C. 300 pp.

Draper, A. D. and D. H. Scott. 197A. Current blueberry
breeding work. Proc. 197A Ann. Meeting of North Ameri-
can Blueberry Council. pp. 68-7A.

PREFACE

The following papers are an attempt to answer some of
the questions raised in the literature review. Each paper
is written in the format of the journal to which it will be
sent. The first and third will be submitted to the Journal
of the American Society for Horticultural Science. Paper
two will be submitted to the Canadian Journal of Plant

Science and paper four will be submitted to HortScience.

ADAPTATION OF THE SPEARMAN-KARBER METHOD FOR ESTIMATING

THE T50 OF COLD STRESSED FLOWER BUDSl

H. C. Bittenbender and Gordon S. Howell, Jr.
Michigan State University, East Lansing

Abstract. Survival data for cold stressed
blueberry flower buds are used to demonstrate
application of the Spearman—Karber (S-K) method

for estimating T Quantal responses are used.

50'
The S-K method determines the T for buds stressed

50
over a range of temperatures which produce 0 to
100% bud survival. The general equation can
accommodate data containing unequal temperature
treatments. A single statistic, S, can be calculated
using a second equation which is an unbiased indi-
cator of the variance of the temperature—% survival
curve around the T50 point. The major advantages
of the S—K method are speed, adaptability to
computer programming, and accuracy with small

sample sizes. Agreement is within 0.5°C of

graphically determined T50 values.

The value of T50 (temperature at which 50% of the buds

are killed) in flower bud hardiness research has been

 

1Received for publication . Michigan

Agricultural Experiment Station Journal Article No. 6A75.

 

established (5,10,11,12). To date all the methods used to
determine T50 have been graphic. The use of graphic pro-
cedures requires several points to determine the sigmoid
tolerance distribution called the temperature—survival (T-S)

curve, is estimated from the T—S curve. There are

T50
several disadvantages to using graphic methods. These are:
1) the need to transform raw quantal (live vs. dead)
responses to percent survival or to probits (5,10,1A);
2) the time spent preparing graphs to determine the T50;
3) difficulty of adapting these methods for a computer
analysis program;
A) difficulty in comparing the shapes of T-S curves without
a regression equation.

Screening seedlings for flower bud hardiness demands a
more rapid method. Ideally a T50 determination method should
be rapid if used on a hand calculator, adaptable to computer
programming, and if possible, should provide information
about the shape of the temperature survival curve at the
T50 point in the form of a single statistic. Finney (3)
describes such a computational method, the Spearman-Karber
(S-K) method for finding LD 50 using quantal response data.
Our purpose was to demonstrate the applicability of the S-K

method to flower bud hardiness research.

Materials and Methods

 

Eighteen 3—node terminal stem pieces were collected

from four A-bush replicates (reps) of each of 2 cultivars

of highbush blueberry, (Vaccinium australe, Small), planted

 

at Grand Junction, Michigan. The age of the bushes was 8
years. This procedure provided 3 stem pieces for each of 5
cold stress treatments and 1 field control from each rep or
a total of 12 pieces (36 buds) per treatment per cultivar.

A11 stem pieces in a treatment were placed in stacked
foil weighing pans, and the stack inserted into a wide mouth
vacuum flask. Each pan represented a cultivar-rep unit con—
taining 3 stem pieces. The temperature of a flask was
monitored by a 26 gauge thermocouple taped to a stem piece
in the center of the stack and attached to a potentiometer
(8).

The flasks were placed in a pre-cooled freezer (-A0°C)
and allowed to cool at a rate of 5-6°C/hr. The flasks were
removed at 2.5°C intervals in the predetermined stress
range. The wood was allowed to thaw in the flasks at
ambient temperature for 12 hr. After thawing, the stacks
were removed from the flasks, separated into individual
units and placed inAa humid chamber for two days. Tissue
browning was used to evaluate viability.

Evaluation of bud injury was based upon quantal
responses, i.e. buds were dead or alive. Occasionally all
ovaries in a bud were not killed. In this case, the number
of dead ovaries from all live buds in a given treatment-
Ineplicate combination were totaled and divided by 8, the

Ineuan number of ovaries per bud. This quotient was rounded

ta) ‘the nearest whole number and was added to the number of

dead buds counted to give the total number of dead buds for

that treatment-replication combination.

Spearman-

Karber Method

 

Calculation of the T

for a cultivar with n buds per

50

treatment and 1 rep.

(1) T50

b.
1
n
Equation
(1)
(2)

(3)

(A)

(5)

TL + 1/2 d — dei where
n

= temperature at which 50% of the buds are killed

= highest lethal temperature, i.e. temperature
below which there are no live buds

= temperature interval between treatments

= no. of dead buds in the 1th treatment, T1

= no. of buds per treatment

1 assumes that:

d is constant;

n is constant;

the treatment-temperature values are transformed so
that T is the smallest positive (1.0) number and

1

TL the largest positive number;

the field control may be used to calibrate the T1
to 100%;

the treatment—temperature range used includes one
treatment, normally T1’ with 100% survival and the

last treatment T with 0% survival. If T does not

L l
have 100% survival and/or TL is not 0% then tem-
peratures T0 = 100% and/or TL+l = 0% may be
assumed, e.g. T0, T1’ T2 . . . TL’ and TL+1‘

(6) it is possible that a 0% survival will be followed
by some living buds. TL, by definition, is the
highest temperature (warmest treatment) below which

there is no survival.

The equation for T if d and n are not constant is:

 

50
(x. + x.)
+
(2) T50 = z [(P1+1 — Pi) l 12 l l,
bi
where P. = ——
1 Hi

and x1 is the temperature value of the ith treatment.
The variance of the temperature-survival curve through
the T50 point can be approximated and thereby compared to

other curves using:

d2

(3) S = ———————-2 [b.( . - b.)].
n2(n-l) 1 n1 1

In equation 3, S approaches m as the T-S curve become

horizontal, and approaches 0 as the T-S curve becomes

vertical. The same assumptions apply for equations 1 and 2.
A more general S equation, where n is not constant is:

P1(1 - Pi)
(ni-l)

 

(u) s = d2 z [

Results and Discussion

 

Data are presented here from 2 highbush blueberry
cultivars, Jersey and Earliblue, the flower buds were

collected and cold stressed on December 17, 1972. From

Table 1, 'Jersey', replicate 1 using equation 1, the S—K

 

estimate of T50 is:
2.5(1 + 6 + 6 + 9)
= . + . _
T50 13 5 l 25 9
= 8.6A.
This transformed T50 value is converted (by adding 1A and
multiplying by —1) to -22.6A°C. The remaining T 's are

50

calculated in the same manner from the values in Table l.

The T50 estimates for both cultivars calculated using

the S—K method can be compared with the T50 estimates

calculated using the graphic approach (Fig. 1, Table 2).

The 2 T50 values determined (S-K and graphic) for

'Earliblue' rep 1 differ by 1.1°C (Table 2). One possible
explanation for this large difference is found in Table 1

where the b0 value for —l5°C (To) is assumed to be 0. The

S—K method requires that the stress range include 0 and
100% survival responses or that 0 and 100% survival
responses would be observed if the treatment-temperature
range were extended 1 d (in this case 2.5°C). The assump—

tion that T is 100% was made, though —15°C probably would

0

not have given a 100% survival response in rep 1. The T

50

means and T composites for 'Jersey' differed only 0.2°C

50
between methods (Table 2). Looking at 'Earliblue', the

mean T50 values are the same for both methods, and the

composites differ by 0.3° (Table 2). By using the T mean,

50

two or more cultivars can be compared because a variance of

the T50 can be found whereas there is no variance if T50

composites are compared.

Proebsting and Mills (12) discussed the development of
a standard T-S curve for dormant Elberta peach flower buds.
They noted that the curve tended to be steeper in cold
weather and flatter in warm weather. Ketchie and Beeman
(9) used the T10 and T90 statistics to describe the differ-
ences in T—S curves in apple cultivars. The need for a
statistic that describes the slope of the T—S is readily
apparent.

Using the S-K equation 3, a variance, S, can be derived

that is an estimate of the slope of the T-S curve.

The S value for rep 1 of 'Jersey' on December 17, 1972

 

was:
S=(2.5)2[<1-8)+(6-3;+<6-3>+<9-o>3
9 ° 8
= 0.A2A.
The mean (5 composite) T-S curve for 'Jersey' (ES =
0.183) has a steeper slope than 'Earliblue' (x8 = .213)
(Table 2).

The relationship between lepe and S is inverse. The
cultivar with the flattest slope has the largest S value,
and that with the steepest slope has an S value of 0, even
though the line is not vertical (Fig. 2). When the T-S
curve goes from 100% to 0% survival in one temperature
interval, this puts a zero term in equation 3. The S = 0
phenomenon has been observed in rep response data, there-
fore we report S for each cultivar as a mean of the rep S

values (Table 2). The best way to decrease the likelihood

of S = 0 is to decrease the temperature interval, d. Data
(not shown) with d = 1°C and n = A resulted in no S = 0
values.

The S statistic quantifies the slope of the curve

through the T 0 point, thus providing a statistic with

5
which to compare T-S curves of a cultivar grown in different
environments or to compare different cultivars. For

instance, a breeder would prefer a selection with the lowest

possible T to withstand low temperature stresses, and the

50
flattest T—S curve (largest S value) to insure that a per—
centage of the buds would survive to produce a crop, if

temperatures dropped below the T A seedling with the

50°
desired hardiness characteristics could be selected using
the S—K equations.

It would be unfair to finish this discussion of quantal
response analysis without comparing the S—K method with the
more famous computational method advocated by Finney (6,7),
the iterative approach to the maximum likelihood method with
probits. The main disadvantage in using the maximum likeli—
hood is that when using a desk calculator one should start
with a hand drawn regression line. However, if a computer
program is available, the maximum likelihood method is just
as rapid as the S—K method. The greater accuracy and
precision of the maximum likelihood method has been stressed
over the S—K method for large sample size n (6). An

analysis with small n, say 2—5, which most breeders would

use, has shown the S-K to be superior (1,2).

Brown (3) has developed methods to estimate the

interval d and sample size n for a S—K LD analysis.

50
Chang and Johnson (A) have shown that the asymptotic
variance of the S—K equation can be expressed as the
distance between the 5th and 95th percentile of the
tolerance distribution. This would be analogous to the
and T

distance between the T of Ketchie and Beeman (9).

10 90
As the plant breeder working on cold hardiness prob-
lems requires measurements of non-discrete characters, the
S-K method provides two statistics to which normal
statistical methods of analysis can be applied.
The Spearman—Karber method can fit into any quantal
response hardiness evaluation program in which speed,

whether on the desk calculator or computer, and accuracy

are important.

10
Literature Cited

Armitage, D. and 1. Allen. 1950. Methods of estimating
the LD50 in quantal response data. J. Hygiene A8:298-
322.

Bross, I. 1950. Estimates of the LD : A critique.

50
Biometrics 6:A13—A23.

 

Brown, B. W. 1966. Planning a quantal assay of potency.

Biometrics 22:322—329.

 

Chang, P. C. and E. A. Johnson. 1972. Some distribu-
tion free properties of the asymptotic variance of the

Spearman estimator in bioassays. Biometrics 28:882-

889.

 

Chaplin, C. E. 19A1. Some artificial freezing tests of

peach fruit buds. Proc. Amer. Soc. Hort. Sci. 52:121—

 

 

129.

Finney, D. J. 196A. Statistical methods in biological
assay. Second Ed., Hafner Pub. 00., N. Y., N. Y. p.
662.

Finney, D. J. 1971. Probit analysis. Third Ed.,
Cambridge University Press. p. 333.

Howell, G. S. and C. J. Weiser. 1970. Fluctuations in
the resistance of apple twigs during spring deharden-

ing. J. Amer. Soc. Hort. Sci. 95:190-192.

 

 

10.

ll.

12.

ll

Ketchie, D. 0. and C. H. Beeman. 1973. Cold acclima-
tion in 'Red Delicious' apple trees under natural

conditions during four winters. J. Amer. Soc. Hort.

 

_Sc_i_. 98:257-261.

Levitt, J. 1956. The Hardiness of Plants. Academic
Press, N. Y., N. Y. p. 278.

Proebsting, E. L., Jr. and H. W. Fogle. 1956. An
apparatus and method for studying fruit bud hardiness.

Proc. Amer. Soc. Hort. Sci. 68:6-1A.

 

 

Proebsting, E. L., Jr. and H. H. Mills. 1966. A
standardized temperature-survival curve for dormant

Elberta peach fruit buds. Proc. Amer. Soc. Hort. Sci.

 

 

 

 

89:85-90.

12

Table 1. Value of b- (no. of dead buds in a sample of 9)
per cold stress treatment—replicate combination
for 'Jersey' and 'Earliblue', respectively, high—
bush blueberry flower buds collected December 17,

 

 

 

 

1972.
W -15yx -17.5 -20 —22.5 —25 -27.5°c
R°p11°at°5 C 1.0Z 3.5 6.0 8.5 11.0 13.5
Jersey
1 0 0 0 1 6 6 9
2 O O O 0 2 9 9
3 O O O 0 O 9 9
A O O O O O 6 9
Composite 0 0 0 l 8 30 36
Earliblue
1 0 0 5 7 9 9 9
2 O O O 3 9 9
3 0 0 0 3 9 9 9
A O O O O 9 9 9
Composite 0 0 5 13 30 36 36
ZTransformed cold stress temperature = I temperature I -1A.

yCold stress temperatures in °C.
XThe bi values for -15°C are assumed.

WField control.

13

Table 2. Comparison of the replicate, mean, and composite
T5 values, T 0 variance, S value, and slope of
T-g curve thrgugh the T point obtained using
the S—K and graphic methds for 'Jersey' and
'Earliblue'.

 

 

 

Jersey Earliblue
S-K Graph S—K Graph
T50

1 -22.6 —2l.7 -l7.9 -l6.8

2 -23.20 —23.A -22.1 -23.1

3 —23.8 —23.8 -20.A -20.6

A —2A.6 -2A.A -2l.3 7 -21.2

Mean -23.5 —23.3 —20.A -20.A

Composite —23.A -23.6 -20.A -20.7
Variance .68A 1.158 3.232 3.50A

is 0.183 slope

2.5 Is 0.213 slope = 1.

9

 

Fig.

1.

1A

Composite and replicate temperature—survival curve
for 'Earliblue' and 'Jersey', respectively, high-
bush blueberry flower buds cold stressed on

December 17, 1972.

u. mmah<mmazmh o. 2.35.3.2”...

am. on- nu- on. 2-

 

 

 

o. 0.
3.33.3... II on o~
utmoaxou II. on up one»
9 S
no no
A M
on I on A
A V
.. w a 1
on utmoaxou I 2.
am 33.43... I on

owh<40u<mhxml I 00
00

 

Fig.

2.

16

Hypothetical temperature-survival (T—S) curve of
four cultivars demonstrating the relationship of
the '8' statistic and the slope of the curves as

they pass through similar T points. The 'S'

50
statistic is an estimate of the variance around
the T50 point, the slope of the T-S curve provides

an estimate of the survival response at tempera-

tures other than the T50 temperature.

 

S

0 -2|.5

.. — — .0293 - 22.0
o o o .0440 - 22.0
. .0756 - 22.0

 

 

'l0 'IS '20 '25

TEMPERATURE ’C

INTERACTIONS OF TEMPERATURE AND MOISTURE CONTENT
ON DEACCLIMATION OF POST REST FLOWER BUDS OF
VACCINIUM AUSTRALE SMALL, HIGHBUSH BLUEBERRY

 

H. C. Bittenbender and G. S. Howell
Department of Horticulture, Michigan State University
East Lansing, Michigan A882A U.S.A.

Michigan Ag. Exp. Sta. No.

ABSTRACT

Different moisture contents and storage tem-
peratures were examined during the spring of 197A
for their effect on the hardiness of highbush blue—
berry (Vaccinium australe Small) flower buds.
Increasing either storage temperature or water
content decreased bud hardiness. The hardening
effect of reduced water content was dependent upon
storage temperature. Dehardening due to increased
moisture content appears to be via a different

mechanism than that due to high temperature.

RESUME

Au cours de printemps de 197A nous avons
étudé les effets de l'humidité des bourgeons

floraux de bleuet (Vaccinium australe, Small) cv.

 

'Jersey', et de la temperature de conservation,

sur la resistance au froid.

La resistance a été réduite par l'augmentation
d'humidité des bourgeons ou la conservation sous
temperature élevée. L'endurcissement induit par la
décroissance d'humidité a été trouvé dependant sur
la temperature de conservation. 5

Il semblerait que les mécanismes d'action
pour 1'humidité et temperature élevées étaient

différents.

The inverse relationship of plant moisture content to
hardiness has been upheld for many years (Levitt, 19A1,
1956). Early workers investigated the moisture-hardiness
relationship by increasing the moisture content of germi-
nating seeds or of grains, and found them to be less hardy
(Levitt, 19A1). More recent research dealing with natural
and induced moisture differences in wheat and barley culti-
vars has supported this view (Metcalf et al., 1970). Root
hardiness of apple between years was related to rainfall
and root hydration (Wildung et al., 1973). Moisture con-
tent was related to known hardiness differences of flower
buds among peach cultivars (Johnston, 1923). Johnston noted
that the average exotherm temperature, AET, correlated with
moisture content in the late winter and spring, but AET was
not precise enough to differentiate between cultivars. The

moisture-hardiness relationship in Rhododendron has been

 

studied using both exotherm analysis that correlated

moisture content and AET from fall through spring (Graham,

1971) and a slow cooling method (2°C/hr) that show signifi-
cant moisture hardiness correlation only in the winter
(Lumis, et al., 1972). Freeze drying woody stems was shown
to increase hardiness 12°C (Li and Weiser, 196A). However,
the dehydration effect may have been confounded by low tem—
perature induced acclimation resulting from the 1yophiliza—
tion process (evaporative cooling).

Reports in the literature attest to the dehardening
effect of warm temperatures in the post—rest period (Howell
and Weiser, 1970; Proebsting, 1963). The length of exposure

and the temperature required to deharden Forsythia X

 

intermedia decreases after the rest period has been satis-

 

fied (Hamilton, 1973). Flower buds of cherry and peach will
deharden at least 5°C after exposure to 21°C for 2A hr in
the post-rest period (Proebsting and Mills, 1972). This
dehardening response to temperature was linked to morpho-
logical development of the flower bud in the spring
(Proebsting and Mills, 1961).

To date there is no literature that directly relates
the individual dehardening effects of temperature, moisture
and their interaction, although in nature thawing (deharden—

ing) temperatures almost always result in H O uptake by the

2
plant. An investigation of this relationship requires that
one be able to modify the moisture content independent of
temperature so that the temperature-moisture interaction can

be tested without statistical confounding. The following

studies were designed to do this.

MATERIALS AND METHODS

Plant Material

'Jersey' flower buds were used in all experiments. The
buds on terminal 2 node stem pieces were taken from the
sides of the bush in all cases (Bittenbender and Howell).
Treatments and cold stress were applied with the flower
buds attached to the stem. The plants were located in a
plantation 2 km east of Lake Michigan and 2 km south of

South Haven, Michigan.

Hardiness Determinations

T is the temperature at which 50% of the flower buds are

50
killed in a cold stress situation. This cold stress was
accomplished by stressing material over a temperature range
known to include temperatures at which there will be 100 and
0% survival. Samples were removed throughout this range so
that a gradient of injury resulted. The freezing rate was
2—5°C/hr. Tissue browning of the ovaries in the bud was

used as the basis for viability evaluation. The T value

50
was calculated using the Spearman-Karber method
(Bittenbender and Howell, 197A). This method of hardiness
determination was used in Experiments 1 and 2.

The AET was found using the exotherm analysis tech-
nique which is based on the heat of fusion of freezing water
(McLeester et al., 1969). When ovary tissue of flower buds
and blossoms supercool and subsequently freeze one exotherm

which is lethal results (Chen and Li, 1973; Graham, 1971).

Florets were excised from the treated flower buds, a 3 mil

copper—constantan thermocouple inserted into the ovary, and
the floret cooled at 100-150°C/hr. The temperature of the
floret was recorded with a potentiometer and pen recorder.
The temperature at which the exotherm was initiated was
recorded as the lethal temperature, the AET was the mean of
6 florets, 2/bud with 3 buds/treatment. This method of

hardiness determination was used in Experiment 3.

Experiment I
A preliminary study was designed to determine the effects of

flower bud moisture content, expressed as gms H 0 in buds/

2
dry weight of 8 buds x 100 (HZO/dwt %), on hardiness,

expressed as T Treatments with A replications (reps)

50'
were as follows: a) Buds were stored for A8 hr at 2°C and
subjected to moisture increase (+H2O) or moisture decrease
(-H2O) or no change (0H2O) in moisture content: b) Buds were
stored at 25°C with moisture decrease (-H20).

'Jersey' stem pieces were collected on February 23,
197A; storage temperature—moisture content treatments were
begun the same day, terminated on February 25, and the
material cold stressed the same day.

The —H2O treatment was achieved by holding the stem
pieces over powdered P205 in a desiccator for a pre—
determined time period until the desired moisture content
was achieved. The time period for moisture changes was

determined empirically and varied according to temperature

and morphological development of the bud. The control buds,

0H20, were held in corked 50 ml erlenmeyer flasks with
nearly 100% relative humidity. The small volume of the
flasks enabled a vapor pressure equilibrium to be estab-
lished rapidly between the buds and water vapor so that
there was little change in the moisture content of the bud.
This relationship was also determined empirically. All
moisture changes were done at the assigned temperature.
After a moisture change treatment was completed, the stem
pieces were stored in 50 m1 erlenmeyer flasks like the
0H2O treatment.

Prior to the cold stress, the moisture content of each
treatment—rep combination was determined by randomly

sampling A stem pieces/rep (8 buds) and taking fresh weight

and dry weight measurements of the flower buds.

Experiment 2

The results of Experiment 1 warranted a more in depth study
of the effects of moisture content, storage temperature, and
their interactions on the hardiness. A 3 X 3 factorial
design with A reps in randomized complete blocks was used to

investigate these factors. The treatments included holding

for 2A hrs at: a) 2°C with moisture levels, +H2O, 0H2O,
-H20; b) 13°C with moisture levels, +H20, 0H20, —H20;
0) 25°C with moisture levels, +H20, 0H2O, -H20.

Plant material was collected on March 18, 197A and
temperature—moisture treatments initiated the same day.

The experiment was terminated March 19, and the material

cold stressed the same day. The +H20 treatment technique
differed from Experiment 1 in that vacuum infiltration of
water at +.18 atm was used. This technique enabled the
moisture increase to be greater and more rapid than the
previous technique. Stems were placed in water under +.18
atm for 2 hrs regardless of temperature, after which the
vacuum was released. This technique enabled a moisture
increase of approximately A5 H2O/dwt % units. All other

treatment methods and moisture content measurements were

the same as in Experiment 1.

Experiment 3

In this experiment the role of temperature and moisture
content on the supercooling of excised florets was evalu—
ated via exotherm analysis. A 2 X 2 factorial design with
6 reps was used. The treatments were stored for 2A hr at:
0; b) 25°C with

a) 2°C with moisture levels, +H 0, —H

2 2

moisture levels, +H20 or -H2O.

Material was collected on April 8, 197A, temp-moisture
content treatments were started the same day, and terminated
April 9, AET determinations were made the same day.

Moisture content change was 1 A0 H2O/dwt % units. All
moisture changes and measurements were the same as

Experiment 2.

RESULTS

Experiment 1

Dehydration of the bud at 2°C caused a significant increase
in hardiness and increased moisture resulted in significant
hardiness reduction. Buds dehydrated at 25° to the same
level as the 2° -H2O treatment were significantly less

hardy than the 2° treatment and no different in hardiness

than the 2° 0H 0 (Fig. l).

2

Experiment 2

From the AOV (Table 1) it can be seen that there are sig—
nificant linear and quadratic effects due to change in
moisture content (H20) and storage temperature (temp), also
the interaction of the linear components of H20 and temp

are significant. The effects due to H O and temp are shown

2
in Fig. 2 and 3 so that these effects and their interactions
can be seen. The +H20 treatment resulted in marked dehard—

ening at all temperatures, while the 0H2O and -H20 treat-
ments seem to be parallel in their response to change in
temperature (Fig. 2). The effects of temperature across
moisture content showed significant dehardening as the
moisture content increased for 2° and 13°. At 25°, however,

no significant hardiness change was observed regardless of

moisture level.

Experiment 3

The temp and H O X temp effects on AET were not signifi-

2
cant (Table 2). The effect of H20 was significant,

therefore the temp main effect was collapsed so that the
H20 effects could be compared across temperatures. The
+H20 treatment regardless of temperature had a significantly

higher AET than -H2O.

DISCUSSION

In Experiment 1, the effect of moisture content on bud
hardiness is almost linear whereas the same 2° treatment in
Experiment 2 performed almost a month later does not appear
linear (Fig. l, 2). The difference between the results of
the two experiments is not due to the different methods of

moisture increase since the slope of the T -moisture curve

50

between 0H 0 and +H 0 treatments is same in both experiments.

2 2
Rather the hardening effect of reduced moisture content is
not as great in Experiment 2 as can be seen in Fig. 2. In
Experiment 2 significant dehardening occurs after 2A hrs at
13° and 25°, this dehardening effect of increasing tempera-
ture appears to override the hardening and dehardening
effects of moisture content. This reduced effect can be
seen as the T —moisture curves go from a steep slope at 2°

50
to an almost flat slope at 25° (Fig. 3). This is the H O-

2
temp interaction.
The bud hardiness decrease from Experiment 1 to Experi-
ment 2 is probably due to the continued morphological
development of the buds and warm temperature between the

mid-February and mid—March experiments. The loss of hardi—

ness between 2° and 25° at —H2O in Experiment 2 is a very

\~l.

ka

.PL

10

rapid loss, almost A°C in 2A hrs, while the moisture con-
tent between the two treatments remained the same. The
hardiness loss between 0H2O and +H20 at 2°, a HZO/dwt %
difference of A0%, was 3°, and yet the temperature remained
the same. Based on this it appears likely that high tem-
perature and increased moisture levels cause dehardening
via different mechanisms. The insignificant hardiness
changes at 25° regardless of moisture content also suggest
separate mechanisms. The results from Experiment 3 suggest
that the exotherm analysis technique for flower buds cannot
detect hardiness per se. Rather, it appears to measure
differences in H 0 content.

2

Graham (1971) reports that H 0 content of the Rhododen-

2
dron florets correlates very highly with AET between culti—

 

vars and between dates. Taken in this light, moisture
content may be the major hardiness control mechanism in

Rhododendron or the AET may be a weak tool for this hardiness

 

study since it cannot discern hardiness losses which are
independent of moisture level. For the highbush blueberry,
also a member of the Ericaceae, moisture does not appear to
be the major dehardening mechanism during the period of this
study. Rather, moisture content appears to act as a subset
of the effects of dehardening temperatures. When there has
been little temperature-morphological dehardening, moisture
changes can increase or decrease hardiness. As a flower bud
continues development in the spring and dehardens, the

effective range of moisture influence on hardiness is

ll

reduced. This is apparent in Experiment 2 when severe
temperature dehardening occurs at 25°C after 2A hrs, and
the effect of moisture content is negligible although the
range is 90 H2O/dwt %.

It is hoped that the results of these experiments will
encourage further work to study the various interacting
factors affecting hardiness rather than looking at single
but n93 independent factors. The more immediate implica-
tions, however, are on the need for proper storage and
handling of flower bud material for hardiness determinations
in the post—rest period, also proper choice of method of

reducing H20 when doing moisture—hardiness studies.

ACKNOWLEDGMENTS

The authors wish to thank the staff and growers of the
Michigan Blueberry Growers Association for providing plant

material.

l2

LITERATURE CITED

BITTENBENDER, H. C. and HOWELL, G. S. 197A. Adaptation of
the Spearman—Karber method for estimating the T50 of cold
stressed flower buds. J. Amer. Soc. Hort. Sci. 99: 187-190.

BITTENBENDER, H. C. and HOWELL, G. S. Cold Hardiness of
flower buds from selected highbush blueberry cultivars

(Vaccinium australe Small). J. Amer. Soc. Hort. Sci. (In

 

Press).

CHEN, P. and LI, P. H. 1973. Viability of apple blossom

buds after test freezing. HortScience 8(6): 508—509.

GRAHAM, R. P. 1971. Cold injury and its determination in

selected Rhododendron species. M.S. thesis, Univ. of Minn.

 

HAMILTON, DAVID F. 1973. Factors influencing dehardening

and rehardening of Forsythia X intermedia stems. J. Amer.

 

 

Soc. Hort. Sci. 98: 221-225.

HOWELL, G. S. and WEISER, C. J. 1970. Fluctuations in the
cold resistance of apple twigs during spring dehardening.

J. Amer. Soc. Hort. Sci. 95:190—192.

JOHNSTON, F. S. 1923. Moisture relations of peach buds
during winter and spring. Univ. of Maryland Ag. Exp. Sta.

Bull. 255.

l3

LEVITT, J. 19Al. Frost killing and hardiness of plants.

Burgess Pub. 00., Minneapolis, Minn. 211 p.

LEVITT, J. 1956. The hardiness of plants. Academic Press,

N. Y., N. Y. 278 p.

LI, P. and WEISER, C. J. 1969. Increasing cold resistance
of woody stems by artificial dehydration. Cryobiology 6:

270.

LUMIS, G. P., MECKLENBURG, R. A. and SINK, K. C. 1972.
Factors influencing winter hardiness of flower buds and

stems of evergreen azaleas. J. Amer. Soc. Hort. Sci. 97:

l2A—l27.

MCLEESTER, R. 0., WEISER, C. J. and HALL, T. C. 1969.
Multiple freezing points as a test for viability of plant
stems in the determination of frost hardiness., Plant

Physiol. AA: 37-AA.

METCALF, E. L., CRESS, C. E., OLIEN, C. R. and EVERSON,
E. H. 1970. Relationship between crown moisture content
and killing temperature for three wheat and three barley

cultivars. CrOp Science 10: 362—365.

PROEBSTING, E. L. 1963. The role of air temperature and
bud development in determining hardiness of dormant

'Elberta' peach fruit buds. Proc. Amer. Soc. Hort. Sci.

83:259-269.

1A

PROEBSTING, E. L. and MILLS, H. H. 1961. Loss of hardiness
by peach fruit buds as related to their morphological
development during the pre—bloom and bloom period. Proc.

Amer. Soc. Hort. Sci. 78: 110.

PROEBSTING, E. L. and MILLS, H. H. 1972. A comparison of
hardiness responses in fruit buds of 'Bing' cherry and

'Elberta' peach. J. Amer. Soc. Hort. Sci. 97: 802-806.

WILDUNG, D. K., WEISER, C. J. and PELLET, H. M. 1973.
Temperature and moisture effects on hardening of apple

roots. HortScience 8(1):53-55.

15

 

 

 

 

 

Table 1. Analysis of 3 X 3 factorial, storage temp for 2A
hrs and moisture content (HZO/dwt %) 'Jersey'
flower buds collected March 18, 197A and cold
stressed March 19.

Source df MS F
Block 3 l.A5 2.90
Treatment
Temp

T Linear l A3.7A 87***+
T Quadratic 1 6.12 12**
H2O
H Linear l 22.0A AA***
H Quadratic 1 A.8 9.6**
Temp X H2O
***
TL x HL 1 9.15 18.3
TL X HQ 1 1.57 N.S.$
TQ X HL 1 .38 N.S.
T X H . N.S.
Q Q 1 39
Error 2A 0.50
+Asterisks denote significance, ** < P .01, *** < P =

.001.

iNot significant at P

.05.

16

Table 2. AOV of 2 X 2 factorial with 6 reps, storage
temp and moisture content of 'Jersey' flower

buds collected April 8, 197A, AET determination

 

 

 

 

April 9.
Source Analysis of variance AET ,
df MS F +H2O -H20

Treatment

H20 1 58.59 9.765** -l5.3°C -18.A°C

Temp 1 10.01 1.668 N.S.+

H2O X Temp 1 .36 N.S. HSD .05 = 2.20°C
Error 20 6.00

 

TF test not significant at P = .05.

.223 on...

0.?— Om. ON. 0: 00. 0m 00

NI

m- on

NI

VN...

 

wmahdeQEMF m0 (102%
com .m_ .N

   

 

$23u\0
OMN CNN 0_N CON Om.

 

m:
00—

on.

em.

on.

0.3

 

 

 

0....

Fig. 1.
Fig. 2.
Fig. 3.

17

Effect of induced moisture level changes on T50°C
of 'Jersey' flower buds stored at 2° or 25°C for

A8 hrs and cold stressed on February 23, 197A.

Effect of moisture level +, 0, —, increase, no
change, decrease respectively, on the T50°C of
'Jersey' flower bud stored at 2°, 13°, or 25°C

for 2A hrs and cold stressed March 19, 197A.

Effect of storage temperature and changing mois-

ture content on T °C of 'Jersey' flower buds,

50
treated March 18, 197A and cold stressed March 19.

COLD HARDINESS OF FLOWER BUDS FROM SELECTED
HIGHBUSH BLUEBERRY CULTIVARS
(VACCINIUM AUSTRALE SMALL)1

 

H. C. Bittenbender and Gordon S. Howell, Jr.2

Department of Horticulture, Michigan State University
East Lansing, Michigan A882A

Abstract. The flower bud hardiness of seven

commercial highbush blueberry cultivars (Vaccinium

 

australe Small) was determined from fall to spring
for two consecutive years. Hardiness was expressed
as T50, estimated by the Spearman-Kérber equation.

The T values for cultivars showed good agreement

50
with spring field survival and with cultivar hardi-
ness based on wood hardiness from controlled and
natural freezes reported by other workers. A
comparison of T50 and AET (average exotherm tempera-
ture) methods for determining flower bud hardiness

was made, the AET method had greater variability

than the T50 method.

The cold hardiness of highbush blueberry (Vaccinium

 

australe Small and V, corymbosum L.) has been of interest

 

 

lMichigan Agricultural Experiment Station Journal Article
No. .

2The authors wish to acknowledge the help and cooperation
of the growers and staff of the Michigan Blueberry
Growers Association, Grand Junction, Michigan.

to workers since Coville's studies in 1910 (7). In many
areas where the highbush is grown commercially, the bushes
suffer damage from low temperature stress during various
parts of the dormant season. To date no definitive con-
trolled freeze—stress research has been done on blueberries
in Michigan, the leading producer of highbush blueberries in
the U.S. (25). One report in the literature indicatefthat
cold injury in Michigan occurs in the spring (19). The
season of injury varies with the location. Injury in
Minnesota occurs in the fall (3,32); in New York and Wash-
ington, in the winter (A,1A). The results of controlled
freeze studies to differentiate cultivar hardiness have been
varied and in some cases the conclusions rather suspect.
Work in Vermont on wood hardiness indicated that 'Collins',
a cultivar not known for exceptional hardiness (9,10),was
hardier than Jersey and Bluecrop (6). This conclusion was
based on electrolyte leakage which was expressed as 'index
of injury'. The relationship of 'index of injury' to tissue
browning or even growth of the wood subsequent to stress

was not reported. This relationship has been shown to be
important (26) in determining the value of such a quantita-
tive viability test. Minnesota workers evaluated
differential thermal analysis as a means of determining wood
hardiness among cultivars and found it unreliable. Good
discrimination among cultivars was made using tissue
browning (23). In Washington electrical impedance was used

to relate hardiness of unfrozen wood with the T of wood

25

and buds in controlled freeze studies (1A). Good agreement
between impedance and hardiness (T25) was found in the
period from December to February but not in the fall or
spring.

Reviews indicate that breeding for blueberry hardiness
is being pursued via interspecific crosses (8,28). Pro-
grams are utilizing hardy germ plasm from other species

such as the Ashworth blueberry (V, corymbosum L.) (10), the

 

bog whortleberry (V. uliginosum L.) (2A), and the northern

 

lowbush (V. lamarckii Camp) (20). A report on the present

 

status of breeding for hardiness in blueberry lists the

main obstacle to progress as the inability to identify hardy
germ plasm without test winters (15). In order to success-
fully screen for hardy material several questions need to

be answered: (1) what is the critical economic tissue(s),
(2) what is the best way to evaluate the hardiness of that
tissue, and (3) when is the critical season of hardiness

for that tissue in the field (27)? We have attempted to
answer these questions by studying the acclimation and
deacclimation patterns of highbush blueberry flower buds of

several commercial cultivars grown in Michigan.

Methods and Materials

Sampling and Freezinngechniques

 

In the 1972-73 season, 3 3—node terminal stem pieces

were used as the experimental unit. Enough material was

collected to provide a field control and 8—10 cold stress
treatments/rep/cultivar. Stem pieces were placed on tape
and rolled up in aluminum foil, placed in a vacuum flask
with a 26 gauge thermocouple attached to a stem in the
center of the foil roll and connected to a potentiometer
and pen recorder to record stem temperature. Eight to 10
flasks were prepared in this manner (1 flask with all reps/
stress temperature), placed in a freezer, and removed at
2.5°C intervals within the predetermined temperature stress
range. The cooling rate was 2—5°C/hr. The sampling and
freezing preparation was modified in 1973—7A, in that 2
2-node terminal stem pieces were used as the experimental
unit. The stem pieces were placed in stacked aluminum foil
weighing pans instead of rolled up in aluminum foil, and
the flasks were removed at 1°C intervals. Stacking the pans
made it necessary to use a randomized complete block design
to remove variation due to pan position within a flask.

After removing the flasks from the freezer, the mate-
rial was allowed to thaw for 12 hr at ambient temperature
while still in the flasks. The foil rolls or pans were
transferred to a chamber near 100% relative humidity for 2
days before buds were evaluated visually for browning of
the ovaries.

The hardiness of the buds stressed in this manner was
reported as T50 using the Spearman—Karber (S—K) equations.

The '8' statistic, which is the slope of the line through

the T50 point, is also determined from the S—K equations

for each cultivar (2).

Cultivar Study

 

Cultivars selected for study included clones which
would provide a range of hardiness based on known field
performance in Michigan. These included Jersey, Rubel,
Coville, Berkley, Earliblue, and three recently released
cultivars, Elliot, Northland, and Bluehaven. Two locations
were used in the study. The first planting, a warm site
containing Jersey and Bluehaven, was approximately 3 km
east of Lake Michigan and just south of South Haven (SH),
Michigan. The second planting, a cold site, was a varietal
test planting containing all cultivars except Bluehaven.

It was located at the Michigan Blueberry Growers Association
in Grand Junction (GJ), Michigan, 16 km east of Lake
Michigan and SH. Since Jersey was planted at both locations,
Bluehaven and Jersey could be compared, and thus the hardi-
ness of Bluehaven relative to the other cultivars could be
seen by comparing the Jerseys at both locations. The
comparison of locations using Jersey, however, has no
statistical basis as locations cannot be replicated except

by using 'years' as blocks.

Hardiness determinations (T ) were begun in September

50
of 1972 and 1973, and continued until mid-April 1973 and
early May 197A. The experimental design used was com-

pletely randomized with A bushes/rep and A reps/cultivar.

Data were analyzed in 1972-73 in completely randomized
design, in 1973—7A in a randomized complete block design
because of freezing technique modifications. Cultivars
were compared at each sampling date using Tukey's w test
(HSD) to indicate significance among T50 values. Tempera—
ture data in the form of weekly max and min, and daily air
temperature variations from a 7 day thermograph were taken

for both locations from September-May each year.

Site on Bush Study

 

The cultivars, Jersey, Rancocas, and Berkley, were
studied in 1972-73, and Jersey and Berkley in 1973-7A to
see if bud hardiness varied as a function of bud site on
the bush. These cultivars were chosen because they repre-
sented a known range of hardiness. The positions studied
were the top portion of the bush above 1.9 m from the
ground and the north and south sides of the bush between
1.9 and 1 m from the ground. The bushes averaged 2.5 m in
height. The planting was located approximately 1A km east
of Lake Michigan and SH.

Hardiness determinations (T50) for this study were the
same as for the cultivar study. Treatments i.e. site on
bush were analyzed as randomized complete blocks with mean
TSO's compared within a cultivar and date using the HSD

value.

Exotherm Studies

 

Average exotherm temperature (AET) for individual
florets was determined in January and April, 197A. A 3 mil
copper—constantan thermocouple was inserted into the ovary
of an excised floret, the floret placed in a small Vial to
prevent dehydration, and cooled at 100-150°C/hr. Graham
(16) showed that for flower buds AET was independent of
cooling rate. The temperature of the ovary was recorded
with a potentiometer and pen recorder. The temperature at
which the heat of fusion was released was recorded as the
exotherm temperature and the mean of several florets tested
as the AET (16). The bud material was taken from GJ and SH,
and only 2-node terminal stem pieces were used. The experi-
mental design used was randomized complete blocks with the
temperature at which the florets were prepared for cooling
and the time after sampling as the blocking effects. AET's

were compared using the HSD value.

Results and Discussion

In 1972—73 (Table la), there was no significant differ-

ence among cultivar T 's at a given date until November 25,

50
at which time all cultivars except Earliblue were at maxi-
mum hardiness. Dehardening started some time between
January 13th and March 19th, 1973. The last hardiness
determination of 1972-73, April A was made 3 days after a

killing freeze. On that date Jersey, at GJ and SH, Rubel,

Northland, and Elliot were the most hardy and the cultivars
of GJ did not differ significantly. The least hardy in
descending order were Earliblue, Coville, Berkley, and Blue-
haven. There appeared to be no difference between locations
for Jersey on any sampling date. A

The fall of 1973 showed Earliblue, Northland, and Blue—
haven acclimating earlier than other cultivars (Table 1b).
Maximum hardiness was attained by all cultivars by the Ath
week of November as in the previous year. Spring deharden—
ing began in early March, which was also similar to the
previous year. The mid-April hardiness determination showed

Berkley to be more hardy in relation to the other cultivars

 

than in the spring of 1973 (Tables 1a, lb). In late April,

Berkley and Bluehaven were least hardy and there was no sig—

 

nificant difference among the other cultivars. There
appeared to be a location difference on October 13th, the
SH Jerseys were hardier than the GJ Jerseys.

Correlating cultivar hardiness rank between years
showed significant positive correlation for winter and
spring, r = 0.7A* and 0.80* respectively. The correlation
for fall between years was not significant, this may be
explained by the early and significant increases in hardi-
ness in the fall, 1973 of Northland and Earliblue. These
hardiness differences may be due to the weather differences
between the fall of 1973 when the first severe frost was on
November 11 and the fall of 1972 when the first hard frost

was October 16. Perhaps these cultivars responded to some

environmental cue other than temperature, such as photo-
period (17), or they may be more sensitive to the higher
acclimating temperatures in the fall. The difference
between Jerseys at the two locations during the fall of
1973 may also be explained by weather differences, in the
early fall at SH, Jersey was hardier on October 13 (Table
1b).

A comparison of T50 and field survival is made in
Table 2. The T50 values follow closely with field survival
after the April 11th frost in 1973. However, in 197A the
field survival resulting from the May 6th frost did not

correlate with the T50 values. The T of cultivars on May

50
Ath did not differ significantly.

The discrepancy between T and field survival after

50
May 6 may be due to several factors. The buds were in

pink stage and therefore quite large and spread apart.

This size factor and the freezing technique in which the
thermocouples for recording temperature are attached to the
stem piece and not inserted into the florets may induce an
error in measuring the actual temperature of the flowers at
this stage of development. Earliblue, which had 78% field
survival, was the most advanced, i.e. largest florets, and
consequently had the greatest potential for error in hardi-
ness determination if this assessment is correct. Another

possibility is that the other cultivars were able to

reharden in the two days in which frosts did occur, between

10

the freezer run and the -6°C frost on May 6, but Earliblue
since it was so advanced could not harden.
The above factors are probably important only in a

late spring situation, since very good T -field survival

50
agreement was seen after the mid-April frost in 1973. The
TSO's on this date were still higher than might be expected
from the field survival, this no doubt is a systematic tech-
nique error, which should have no effect on the relative
hardiness differences among cultivars.

In Table 3, the hardiness rank of cultivars based on
the T50 flower bud hardiness on April 1A, 1972 is compared
with the average hardiness ranking of the same cultivars by
other workers based on field survival or controlled freeze
evaluation of the wood. Flower bud hardiness as a means of
describing overall hardiness compares very well with results
from other studies (Table 3).

The results of the site on bush study (Table A) show
that in the late fall of 1972 the buds on the north and
south side of the bush were hardier than the buds at the
top. No significant difference was seen during the late
winter or spring of that year or any season during 1973-7A.
Since a significant difference between top and sides of the
bush was seen in all cultivars tested in late fall 1972,
but not in the following year, site on bush effect may be a
function of environment-site interaction rather than simply

difference due to site alone. Even though these differences

in hardiness did not occur both years, it should be noted

11

that it is a potential source of sampling error which can
be very significant.

In a preceding paper (2) we suggested that the '8'
statistic of the Spearman-Karber equations, a variance of
the temperature-survival (T—S) curve, could be used to
describe the slope of the T—S curve of a cultivar. It was
suggested that cultivars having different S values should
be chosen in favor of those with the largest S value since
it would have greater percent survival at temperatures below

the T than another cultivar with the same T but a

50 50
smaller value. No significant difference in the value of S
was found among cultivars during any season. In this study
the precision of the freezing technique and accuracy of
temperature measurement play an important role when using
the '8' statistic which seems to be very sensitive to slight
differences in cooling rates. A significant difference was
seen among seasons in 1972-73 (Table 5) but not in 1973-7A.
The change in the T—S curve from steep in fall (small '8')
to flattest in spring (large '8') is in agreement with the
change in the T-S curve among seasons reported for 'Elberta'
peach flower buds (22).

The bud hardiness of the cultivars used was based on
the 3 most distal nodes in 1972-73 and on the 2 most distal
in 1973—7A. The field survival of buds based on bud posi-
tion on the lateral for four cultivars was observed on

bushes growing at the varietal planting at East Lansing,

approximately 190 km east of Lake Michigan, a poor

12

blueberry site. Distal buds are significantly less hardy
than proximal buds in the same cultivar (Table 6).

Whether the difference in hardiness between distal and
proximal buds is based on delayed morphological development
or some physiological component is not clear. From these
data it appears that cultivars setting many flower buds per
lateral, such as Jersey, sometimes as many as 8—10, have an
added hardiness advantage over those that set few. In this
study, the hardiness of distal buds was used because some

cultivars did not have more than 3 buds per lateral.

AET vs. T50

The formation of ice in flowers within buds or in

 

blossoms is lethal (5,13,16,18). The possible application
of exotherm analysis for determining hardiness was investi—
gated. The results of controlled freezes on January 2A,
197A (Table 7), indicated that for a given number of reps

(N) the T analysis provides the smallest HSD value, even

50
though there seems little difference between cultivars.

The relative ranking of cultivars between the two AET
analyses, N = A, 8, is approximately the same except for
Jersey and Rubel. Using N = 8, a smaller HSD is obtained,
and AET values are lower. This may be due to slight dehy-
dration of A reps during the pre-freezing preparation which
was done at 20°C. Water content (H2O/dwt %) for the buds

is listed for each cultivar on that date. It is possible

that the AET values reflected the difference in water

13

content between cultivars rather than hardiness per se. On

April 2A, 197A, there was no correlation between T and

50
AET and moisture content. AET values which indicate the
exotherm temperature of an isolated floret do not parallel
field hardiness on this date, the reason for this and the
low AET values is not known.

The number of reps for each method and resulting HSD
value can be misleading. The AET analysis, N = 8, each
cultivar required A buds with 2 florets per bud. The T50
analysis using N = A represents A reps X A buds X 6 tempera—
tures = 96 buds/cultivar. Much more material is required
for T50 analysis using the S—K equations, but the resultant
HSD is smaller. The time required for using each method is
about the same, 8-10 hrs, since extra time is needed for
preparation of each floret using AET and time is lost if
only one potentiometer is used. However, only 15-20
minutes is required for the exotherm freeze run at 100-

150°C/hr; AET gives immediate results while the freezer

T50
run may take 6—8 hrs at 2-5°C/hr and another 2A hrs before
buds can be evaluated for browning, which for A cultivars
would take about an hour.

There appears to be much work yet to be done before

AET can be applied for determining flower bud hardiness of

blueberry as it was used by Graham (16) on Rhododendron

 

flower buds.

Flower bud hardiness based on T50 successfully

discriminates between hardy and tender cultivars. Flower

1A

bud hardiness is a good indicator of overall hardiness; when
the hardiness relationship between the flower buds of young
plants and mature plants is known, a screening method using
the T50 of flower buds can be developed. Ideally, seedlings
from crosses between 'hardy' parents based on Stushnoff's
hardiness criteria (27) would be frozen at two temperatures
above and below the T50 of some hardy standard such as
'Jersey' (21). The most hardy 5-10% would be saved for
further selection based on other criteria. The acclimation-
deacclimation pattern would be studied using T50 analysis

for the best selections.

15

Literature Cited

Bailey, J. S. l9A9. Frost injury to blueberries.

Fruit Var. and Hort. Dig. A:98.

 

Bittenbender, H. c. and Gordon E. Howell. 197A.
Adaptation of the Spearman-Karber method for esti-
mating the T50 of cold stressed flower buds. J. Amer,
Soc. Hort. S91. 99(2):l87-190.

 

Brierly, W. G. and A. C. Hildreth. 1928. Some studies

on the hardiness of certain species of Vaccinium.

 

Plant Physiol. 3:303-308.

 

Cain, J. C. and G. L. Slate. 1953. Blueberries in the

home garden. Cornell Ext. Bull. 900.

 

Chen, P. and P. H. Li. 1973. Viability of apple

blossom buds after test freezing. HortScience 8(6):

 

508-509.
Constante, J. F. and B. R. Boyce. 1968. Low tempera—
ture injury of highbush blueberry shoots at various

times of the year. Proc. Amer. Soc. Hort. Sci. 93:

 

 

267—272.
Coville, F. V. 1910. Experiments in blueberry culture.

USDA Plant Ind. Bull. 193.

 

Darrow, G. M. 1960. Blueberry breeding. Nat. Hort.

 

Mag. 39(1):15-29.

and J. N. Moore. 1962. Blueberry growing.

 

USDA Farmers Bull. 1951 (revised).

 

10.

110

l2.

13.

1A.

l5.

16.

17.

18.

19.

16

and . 1966. Blueberry growing.

 

 

USDA Farmers Bull. 1951 (revised).

 

 

, L. Whitton and D. H. Scott. 1960. The

 

Ashworth blueberry as a parent for hardiness and

earliness. Fruit Var. and Hort. Dig. lA:A3-A6.

 

, R. B. Wilcox and C. S. Beckwith. 19AA.

 

Blueberry growing. USDA Farmers Bull. 1951.

 

 

Dorsey, M. J. 193A. Ice formation in the fruit bud of

the peach. Proc. Amer. Soc. Hort. Sci. 31:22—27.

 

 

Doughty, C. E. 197A. Cold injury and hardiness of blue—

berries in western Washington. Proc. g; 197A Annual

 

Mtg. g£_North American Blueberry Council.

 

Draper, A. D. and D. H. Scott. 197A. Current blueberry

breeding work. Proc. 9: 197A Annual Mtg. 93 North

 

 

American Blueberrijouncil.

 

Graham, R. P. 1971. Cold injury and its determination

on selected Rhododendron species. M.S. thesis, Univ.

 

of Minn.
Howell, Gordon S. and C. J. Weiser. 1970. Similarities
between the control of flower initiation and cold

acclimation in plants. HortScience 5(1):18-20.

 

Johnston, E. S. 1923. Moisture relations of peach buds
during winter and spring. SEEK, gflMaryland Ag, EAR.
§E§° 9211' 255.

Johnston, S. 1939. Resistance of certain blueberry

varieties to frost. Mich. Ag, Exp. Sta. Quart. Bull.

 

 

 

22:10-12.

20.

21.

22.

23.

2A.

25.

26.

27.

17

1956. Blueberry breeding in Michigan.

 

Fruit Var. and Hort. Dig. 11:20.

 

 

Lapins, K. 1962. Artificial freezing as a routine
test of cold hardiness of young apple seedlings.

Proc. Amer. Soc. Hort. Sci. 81:26-3A.

 

 

Proebsting, E. L. and H. H. Mills. 1961. Loss of
hardiness by peach fruit buds as related to their
morphological development during the pre-bloom and

bloom period. Proc. Amer. Soc. Hort. Sci. 78:10A—110.

 

 

 

 

Quamme, H. A., C. Stushnoff and C. J. Weiser. 1972.
Winter hardiness of several blueberry species and

cultivars in Minnesota. HortScience 7(5):500-502.

 

Rousi, Arne. 1963. Hybridization between Vaccinium

 

uliginosum and cultivated blueberry. Annales

 

Agriculturae Fenniae 2:12-18.

 

Sheridan, P. and R. A. Seelig. 1973. Blueberries -
Fruit and Vegetable Facts and Pointers. United Fresh
Fruit and Vegetable Assoc., 777 lAth Street, Washing-
ton, D. C.

Stergios, B. G. and G. S. Howell, Jr. 1973. Evaluation
of viability tests on cold stressed plants. J. Amgg.

S99. Hort. Sci. 98(A):325—330.

 

Stushnoff, C. 1972. Breeding and selection methods for

cold hardiness in deciduous fruit crops. HortScience

 

7(1):10-13.

18

28. Whitton, L. 1960. Breeding for winter hardiness. In
Blueberry Research. J. N. Moore and N. F. Childers,

Ed. N. J. Ag. Exp. Sta. at Rutgers. pp. 51-52.

19

.mocmpmMMHU undefimficmfim ozm

 

 

 

 

 

 

 

.sesaoo sashes n. 9 tom omms
mm.su mm.mn Am.emu am.:mu o:.mm- mm.HHI mm.ou .sonsoe
om.su mm.HH- sm.:mu se.:mu ms.em- mm.mH: om.su so .sonnos

mQOHpmooq Spom pm mmmhmh

om.m .m.z mo.H .m.z mm.H .m.z .m.z on:
m:.sn mm.mu sm.:mu em.:m- o:.mm- mm.HH- mm.m- senses
mw.mu Ho.mu :m.Hmu Hw.amu mm.omu oo.oeu Ao.on so>ssosam
Cm>mm cpsom pm mmm>fipazo

:m.m mo.m :m.m om.m we.e .m.z .m.z s.m.z nmo. omm
mH.s- mo.oa- mo.smu :m.mmu oa.mmu om.mfiu mm.ou ooHHHm
me.mu mm.mu Ho.mmu mm.mmn oa.mfiu m:.mfiu em.ou osaoaasam
Hm.mu mw.ml so.Hmu mm.amu wH.Hmu sm.mfia Ho.sl oo.mu oaaa>oo
Hm.:- mo.mu mm.mH- me.om- mm.Hm: mm.HH- mm.o- :m.mn moaxsom
mm.sn mo.mn oa.mmu mm.mmu mfl.mmu mm.HH- Em.o- wm.eu osoarosoz
mm.su mm.mu Hm.sm- Hm.:mu mm.:mn mm.mfiu o:.o- Hm.:- Hoosm
om.su mm.HH- sm.emu se.:m- mp.emn mm.mau mm.hu m~.mu senses
COfiPOCSW UCGLU mpmxfinpflso
msusflun msumfium ms-maua mpusaumfi mslmmuaa mwummuoa msumuoa msumaum Lo>aoaso

.mEummmH

wcflpso mmpmo Uopooamm pm mpm>HpH50 appopmsan Endpcwfls w pom mosam> oo B .ma manna

2O

.oocoLoMMHo pcmoflmficmflm ozNA

 

 

 

 

 

 

 

.QESHoo cfinpfiz m.ome pom mmmn
mw.3| mH.m| oo.HHI oo.mm| mo.mm| Hm.mm| mm.mm| um.mal om.oal m3.~| mm .mompmh
3m.3| Hm.3l mn.oH| mo.mml 3m.3m| mm.>m| mm.mm| oa.om| ms.>| mm.m| ho .mmmpoh

mEOHpmooA zoom pm zmmpoh

.m.z .m.z .m.z mo.a mH.o 3w3.o mm.o .m.z .m.z .m.z mo. 9mm
m©.3l ma.m| .oo.HHI oo.mm| wo.mm| Hm.nml mm.mml sw.mal om.oal m3.>| mompmw
om.3| mm.3| 33.HH| oo.mH| 33.mm| Hw.mml wm.mm| 33.mHI mw.m| ma.m| zo>mcosam
co>wm Epsom pm npw>fipaso

.m.z mm.H mm.m mm.o mm.H 03m.o m.m.z mm.a mm.a mw.H mmo. 9mm
Hw.m| mn.m| om.HHI mm.am| Hm.mm| 3m.mm| mm.mm| 33.0w: mo.m| >3.m| poaaam
Hm.m| mH.m| m>.w| wm.mH| oo.mm| Hm.3m| mm.3m| mm.ma| Hm.mH| m3.oan osanfiammm
oo.m| 33.3: om.HHI mo.omu wo.mml mm.mmu oa.3m| oo.man mo.m| ww.ml moaxpom
mo.m| mm.3l mm.mH| mo.mmu mm.mm| om.mml mm.3m| mw.om| 33.0: mm.m| ocwasupoz
om.m| mw.m| mm.oH| Hw.om| mm.mm| 3m.mml mm.3m| Hm.mal mm.n: 3m.m| Honsm
3m.3| Hm.3| ms.oH| wo.mm| 3m.3m| mm.wml mm.mm| oa.om| mw.u| mm.ml mompmh
coapocSh ocwpw pm mpm>HpH30

3sn3um 3summu3 3snmn3 3sumum 3sn3mna msumanma msnmmuaa msnmuafi msumanoa msummum Lo>fipazo
.3wlmnma wcflpso mopmo Umpooamm pm mpm>HpHSo mmpmnosap cmsncwfic w you modam> ome .nH ofint

21

Table 2. T °C at nearest date and field survival after
sgging frosts for years 1973 and 197A, for

various cultivars.

 

 

Spring, 1973

 

Date A—11-73 A-lA-73
temperature min —10°C T50°C
Cultivar Field

 

survival %

Spring,fl97A

 

5-6-7A 5-A-7A
min —6°C T50°C
Field

survival %

 

Jersey, GJZ 86 —7.
Rubel 100 -7
Northland 8A -7
Berkley ‘AA —A
Coville 53 ‘5
Earliblue 53 -5
Elliot 8O -7
Bluehaven, SH 11 -3
Jersey, SH 86 —7

HSD .05 5O 2

50

.85
.22
.51
.21
.A2
.15
.75
.92

.80

99 -A.9A
98 -5.50
9A —5.06
92 -5.00
78 -5.31
98 —5.81
98 min -A°C -A.50
99 min -A°C -A.63
11 N.S.y

 

ZLocations, Grand Junction (GJ),

yNo significant difference.

South Haven (SH).

22

.mmocflopmn p59 Lozoam wcfipmmp .oNooLm UoHHoppcoox

.H u onofiossms

.mmocflohmn poo: wcflpmop .ommohm Umaaompcoom

.mhsncfi mo condemn

 

mxozmam

m.m

mmm

r-immr-‘I

m.m

wcfipam
.Looooosoooem

mnma

x

3&cmm mwmpo><
Amm.

Haom mung mAmH m.oessso

Aoav mmmfi .zoppmm

Add.
weapon mmmfi .sopnssos

ANHV 33mH .sossom

Amv zwoma .opcwpmcoo
Adv sores: momfi .sossno
sz Leona: mmma .saeo

Adv nwsesom mama .sofiaom

 

co>m£osam mwooocmm

nsfiHHoo

mmathm

mHHH>oo

osfioeashm
sa>eofiso

osnasosoz

poaaam

hompoh

Hoosm

Cowmom Ucm Lozps<

 

 

.moummmm Umaaoppzoo so woman mmocflopmc 059 Lozoam ho coo: mommmpm

Hmpzpmc popmm .Hm>H>LSm UHpHE co woman mpm>fipazo QmSQsmHQ mo mwcfixzwp mmocflwpmm .m manme

23

 

 

 

Table A. T 0°C of flower buds taken from 3 sites on the
bash, from the 3 cultivars.
Cultivars and Sampling Dates
Site on Bush 10—1A—72 11-11-72 12-11-72 A-l2—73
Berkley, north side -6.87 —17.78 —23.5A —6.25
top -6.7A -17.36 —20.69 -6.0A
south side -7.36 -l7.08 -21.81 —5.97
HSD .05 N.S.Z N.S. 2.50 N.S.
Rancocas, north side -7.22 -18.33 -26.0A -6.A6
top -6.60 -16.94 —2u.58 -7.99
south side —6.53 -18.19 -25.28 -6.7A
HSD .05 N.S. 1.18 N.S. N.S.
Jersey, north side -7.92 —19.A6 —26.7A -9.58
top —7.6A —l7.01 -2A.93 -9.86
south side —8.A7 -l8.75 —26.A6 -9.79
HSD .05 N.S. 2.35 1.03 N.S.

 

ZNo significant difference.

2A

Table 5. Change in S statistic among seasons between
years using the average of 8 cultivars.

 

 

 

Year
seas°n 1972—1973 1973—197A
Fall 0.156 0.2u1
Winter 0.238 0.191
Spring 0.258 0.228
HSD .05 0.062 N.S.Z

 

ZNo significant difference.

25

Table 6. Relationship of bud position to winter survival.
Values are mean number of live blossoms in late
spring (May 25, 1973) after late winter freeze.

 

 

 

 

BUd Positionz Jersey RubelCUltigziiley Earliblue
l 2.6 0.0 A.6 l
2 A .7 A.7 2.7
3 6 6.7 6.0 A.l
A 7.A 7.2 7.1 5.7
5 9.1 9.6 8.0, 7.A

HSD .05 (within a column) = 2.9A

 

Z1 = most distal, 5 = most proximal on five node stem.

26

 

 

 

 

Table 7. T 's, AET's, and moisture content for several
cégtivars on two dates in 197A.
Method of Hardiness Determination
Moisture con-
Cultivar T50°C AET °C AET °C tent of bud
. (HZO/dwt %)
l—2A—7A
Jersey -2A.92 —21.68 —23.36 125.5
Rubel —22.62 -22.l5 -23.29 122.A
Berkley —23.05 —17.70 —18.78 135.0
Earliblue —22.98 -19.82 -20.31 132.7
HSD .05 0.89 3.32 2.06 6.A
Preparation temp 0OZ 2 2 2 & 20
Ny u u 8 1:
Cooling rate °C/hr 2-5 100—150 100-150
A—2A—7A
Jersey —A.8l —1A.92 25A.6
Rubel -5.63 -13.67 232.6
Northland —A.88 —13.22 265.8
Berkley -A.AA -1A.85 298.8
Earliblue —5.13 —10.58 29A.5
Elliot -5.75 -1A.02 239.A
Bluehaven —A.63 -l3.52 326.2
HSD .05 °C 1.29 N.S. 17.3
N A A A

 

ZTemperature at which buds were prepared for freezing.

yNumber of replicates.

PREDICTIVE ENVIRONMENTAL AND PHENOLOGICAL COMPONENTS
OF FLOWER BUD HARDINESS IN HIGHBUSH BLUEBERRY
(VACCINIUM AUSTRALE SMALL)1

 

H. C. Bittenbender and Gordon S. Howell, Jr.2
Department of Horticulture, Michigan State University
East Lansing, Michigan A882A

Abstract. Environmental and phenological com—

ponents of highbush blueberry (Vaccinium australe

 

Small) flower bud hardiness were used to predict

hardiness (T ). Multiple regression equations

50
were derived from 1 years data 1973-7A for fall,
winter and spring. Factors considered for the
equations were air temperature, photoperiod, bud
dry weight and moisture content, bark color, dates
of leaf drop and pollen formation in the field and
time interval of flower forcing. The phenological
characters and hardiness were measured for seven
commercial cultivars. Estimated T values were

50

within 1°C of measured T50 values.

Though the exact freezing resistance mechanism(s) is
not known, certain components of hardiness that are respon—
sible for induction, maintenance, and reduction of hardiness

are known. The effects of environmental stimuli such as

 

lMichigan Agricultural Experiment Station Journal Article
No.

2The authors wish to acknowledge the help and cooperation
of the growers and staff of the Michigan Blueberry Growers
Association, Grand Junction, Michigan.

photOperiod and air temperature are very broad. Documenta-
tion of the role of the short days (SD) photoperiod for

inducing hardiness has been done with apple (9), Hedra helix

 

(2A), Acer negundo (11), and Cornus stolonifera (28). One

 

 

theory of hardiness suggests that acclimation occurs in two
stages (10,27); the first stage is induced by SD, the
second by frost. Fluctuating air temperatures have been
correlated with peach flower bud (6,16,19,20,22) and apple
wood (13) hardiness changes in the winter, during and after
rest.

Pollen maturity and appearance of pink in the bud are
phenological characteristics which indicate rapid transi-
tion from a hardy to tender status in peach buds (20).
Other factors that may play a role in or correlate with
hardiness are moisture content (7,12,15), anthocyanin con—
tent (18), date of leaf drop, and date of flowering (1A).

An attempt is made to consider all of these factors as
they affect acclimation in the fall, hardiness maintenance
in winter, and loss of hardiness in the spring in the blue-
berry flower bud. These environmental factors and pheno-
logical characteristics will be entered into multiple
regression analysis to see if they relate to the hardiness
status of highbush blueberry flower buds in a predictive
manner (17). The cultivars used were part of a two year
cultivar and location study (A) including Jersey, Rubel,
Northland, Berkley, Earliblue, Elliot, and Bluehaven. Some

were located at South Haven, and others at Grand Junction,

Michigan. The bushes were at least 10 yrs old and were
growing in A A—bush plots. A controlled freeze method was
used to cold stress the flower buds which were evaluated

for tissue browning of the ovaries. Hardiness was expressed
as T50 using the Spearman-Karber equations (2). The values
for the components of hardiness were determined at every
date from September—April that a freezer run was made.

Means for a cultivar were used except for environmental

data such as photoperiod or air temperature which were the

same for all cultivars at the same location.

Components of Hardiness

 

Bud Dry Weight. Flower bud dry weight (bud dwt) is a

 

measure of size and indicator of growth and was found by
drying buds for 2A hr at 65°C. Four 2-node terminal stem
pieces (8 buds) were taken per plot. The bud dwt is a mean
of four plots and measured to the nearest tenth mg.

Moisture Content. The bud moisture content was

 

determined using the data from the bud dwt material, fresh
weight minus dry weight/dry weight times 100. Moisture
content was expressed as the mean percent H2O g/dwt g of
four plots (12).

S9999. The four stem pieces per plot per cultivar
from which the flower buds were taken for bud dwt and
moisture content were used to determine anthocyanin content
of the bark. The bark was removed with a potato peeler and

the anthocyanin extracted with cold acidified methanol for

18 hr at 5°C using 100 ml methanol per gram fresh weight
of bark. The extract was filtered through Whatman No. A
filter paper and the absorbance read at 520 nm (25). The
mean absorbance of four plots was used to compare relative
anthocyanin content among cultivars at a given date.

Leaf Drop. Date at which a cultivar had drOpped 50%

 

of its leaves was determined by observation. Leaf drop was
expressed as the number of days after November 3, 1973 that
50% defoliation occurred.

Tetrad Formation. An attempt was made to determine

 

mean date of microsporogenesis, this was not successful.
Therefore the date of pollen grain formation was used. The
assumption was made that the interval from microsporogenesis
to pollen grain formation was the same among cultivars (25).
Terminal 2—node stem pieces were collected from April-May,
197A; killed and fixed in FAA (5:90:5), and stained with
acetocarmine (A5% acetic acid) using the squash method (23).

Interval to Flowering. This was expressed as the mean

 

number of days for 50% of the buds on a branch to open at
least one flower. Branches with 7-8 laterals were collected
on three dates in the spring of 197A for all cultivars.

Buds on the laterals were removed so that there were only 2
distal buds/lateral on each branch. The branches were
placed with their butts in water at room temperature. The
50% flowering data is recorded as the time from the start

of forcing. The mean flowering data for a cultivar is the

mean of three dates tested, since only one branch was used
per date per cultivar.

Air Temperature. The mean air temperature i.e. max +

 

min divided by 2, of the seventh day prior to sampling for

a hardiness determination correlated significantly with
hardiness in 1973-7A. Correlations between hardiness and
air temperature were made using max, min and mean tempera-
tures for the days of the sampling dates to 1A days prior to
sampling, and 1 to 1A days running averages for the same
time intervals. Use of the seventh day mean resulted in the
highest r values, regardless of season. In 1972—73, the
best correlation with hardiness was the eighth day prior to
sampling. Air temperature was expressed as air temp °C +
100, to make all temperatures positive. Air temperature
measurements were taken with a 7 day thermograph, 1.6 m from
the ground, at the Grand Junction plots. In South Haven,
the thermograph was 8 km from the plots, here the weekly
temperature charts were adjusted using the max and min
readings from the location of the plots.

Photoperiod. The daylength of the eighth day prior to

 

sampling had the highest correlation coefficient when
correlated with hardiness in 1973—7A; the ninth day corre-
lated the best in 1972—73. Correlations were made between
hardiness and daylength using the days from sampling date
to 1A days prior to sampling. Daylength was defined as the

period from sunrise to sunset. These values were for

Lansing, Michigan, approximately 130 km east and 35 km
north of Grand Junction (1).

Photoperiod X Air Temperature. This product was found

 

simply by multiplying the two values given above for a

given date and location.

Hardiness Equations

 

The multiple regression equation in which the inde-
pendent variables correlate linearly with the dependent
variable takes this form:

Y = BO + Ble + B2X2 + B3X3 . . . BNXN + E

where in our model Y is estimated T50, BO . . . BN are
parameters of the model and are specific for each equation,
Xl . . . XN are independent variables i.e. bud dwt,
moisture content, and E is the residual error. The data
were entered by cultivar for each season with the measured
T50 and associated components for one cultivar at one date
as the unit for regression. The various terms were
sequentially deleted from the equation if their signifi-
cance was greater than the .10 level (5). A stepwise
deletion multiple regression program in CDC 6500 computer
at Michigan State University, was used to calculate the
best fit equation. The fall model contained bud dwt,
moisture content, bark color, mean air temperature, photo-
period, temp X photoperiod, and leaf drop date as possible

components of the hardiness equation. The winter model

included as possible components bud dwt, moisture content,

bark color, mean air temperature, photoperiod, and temp X
photoperiod. The spring model considered bud dwt, moisture
content, bark color, mean air temperature, photoperiod,
temp X photoperiod, date of pollen formation, and mean
flowering interval.

The multiple regression prediction equations are
highly significant for each season. All equations have
standard errors less than i 1°C (Table l). The equation
for fall contained the following significant components:
photoperiod, moisture content, air temperature, and temp
X photoperiod. The equation for winter was composed of bud
dwt, moisture content, and bark color. The spring equation
contained bud dwt, moisture content, bark color, and photo-
period. The components of each equation give the best fit
of the factors measured during each season for estimating
T50. One should not view the highly significant F or R2
values to imply cause and effect; these are merely predic-
tive equations and cannot prove a causal relationship. The
and estimated T

close agreement between actual T values

50 50

can be seen for 'Jersey' (Fig. 1), no estimate differs more

than 1°C from the measured T Comparison of actual and

50'
estimated T50 values for other cultivars could be made by
substituting the specific phenological and environmental
values for each cultivar into the hardiness equations as
was done for Jersey.

It is of interest that no environmental parameter

enters into the winter equation. This is probably due to

lack of change in mean air temperature, photoperiod, and
hardiness in the winter. In the spring equation, photo-
period is included but not temperature. This should not be
taken to imply that temperature is not important in blue-
berry dehardening. The dehardening effects of temperature
have been noted for many species (8,9,22) including blue-
berry (3). Rather, the temperature correlation may be
confounded with the increasing daylength so as not to have
a significant effect. Another explanation might be to view
temperature as the trigger for dehardening, a chain
reaction; once the temperature rises to a certain level,
dehardening starts and the dehardening rate remains inde-
pendent of any temperature changes within a certain range,
in such a case temperature would not correlate with
hardiness on a per date basis (9).

The potential uses for hardiness prediction equations
are many. The hardiness for a given cultivar at a given
date can be estimated, this is of importance when frost
protection systems are used so as to maximize their
efficiency. Site selection is another area of possible
implementation, the environmental measurements coupled
with phenological characters may be able to predict whether
a cultivar can survive at particular site. The screening
procedure for hardiness can benefit by use of a predictive
equation. After several years of data have been gathered
or by using 10-20 year average weather cycles for tempera—

ture and rainfall for a site (28), it may be possible to

have parameters for air temperature, photoperiod, temp X
photoperiod that do not change from year to year. At this
point a breeder could screen measuring only phenological
characters and freeze only to determine small differences

among cultivars.

10

Literature Cited

Anonymous. 1965. Sunrise and sunset at Lansing,
Michigan. NO. 11A8, Nautical Almanac Office, USN
Observatory, Washington, D. C.

Bittenbender, H. C. and Gordon S. Howell, Jr. 197A.
Adaptation of the Spearman-Karber method for esti-

mating the T of cold stressed flower buds. J.

50
Amer. Soc. Hort. Sci. 99(2):187-190.

and . Inter-

 

 

actions of temperature and moisture content of

deacclimation of post rest flower buds of Vaccinium

 

australe Small, highbush blueberry. Can. J. Plant
Sci. (In Press).

and . Cold

 

 

hardiness of flower buds from selected highbush blue-

berry cultivars (Vaccinium australe Small). J. Amer.

 

Soc. Hort. Sci. (In Press).
Draper, N. R. and H. Smith. 1966. Applied regression
analysis. Wiley & Sons, New York, New York. A07 pp.
Edgerton, L. J. 195A. Fluctuations in the cold hardi-
ness Of peach flower buds during rest period and
dormancy. Proc. Amer. Soc. Hort. Sci. 6A:175-180.
Graham, R. P. 1971. Cold injury and its determination

on selected Rhododendron species. M.S. thesis, Univ.

 

of Minn.

10.

11.

12.

13.

1A.

15.

16.

11

Hamilton, David F. 1973. Factors influencing deharden-

ing and rehardening of Forsythia X intermedia stems.

 

 

J. Amer. Soc. Hort. Sci. 98:221-223.

Howell, G. S. and C. J. Weiser.. 1970. Fluctuations in
the cold resistance of apple twigs during spring
dehardening. J. Amer. Soc. Hort. Sci. 95:190-192.

and . 1970. The environ—

 

 

mental control of cold acclimation in apple. Plant
Physiol. A5:390—39A.
Irving, R. M. and F. O. Lanphear. 1968. Regulation of

hardiness in Acer negundo. Plant Physiol. A3:9—13.

 

Johnston, E. S. 1923. Moisture relations of peach buds
during winter and Spring. Univ. Md. Ag. Exp. Sta.
Bull. 255.

Ketchie, D. 0., C. H. Beeman, and A. L. Bullard. 1972.
Relationship of electrolytic conductance to cold
injury and acclimation in fruit trees. J. Amer. Soc.
Hort. Sci. 97:A03—A06.

Layne, R. E. C. 1967. Relation Of bloom date and
blossom temperature to frost injury and fruit set in
apricot. Fruit Varieties and Hort. Digest 21:2.

Lumis, G. P. 1970. Winter hardiness of evergreen

Azalea (Rhododendron cv.) flower buds. Ph.D. thesis,

 

Mich. State Univ.
Meader, E. M. and M. A. Blake. 19A3. Seasonal trend
of fruit bud hardiness in peaches. Proc. Amer. Soc.

Hort. Sci. A8:91—98.

17.

18.

19.

20.

21.

22.

23.

12

Metcalf, E. L., C. E. Cress, C. R. Olein and E. H.
Everson. 1970. Relationship between crown moisture
content and killing temperature of three wheat and
barley cultivars. Crop Sci. 10:362-365.

Parker, J. 1962. Relation among cold hardiness, water,
soluble protein, anthocyanins and free sugars in

Hedra helix L. Plant Physiol. 37:809-813.

 

Proebsting, E. L. 1963. The role of air temperatures
and bud development in determining hardiness of dor—
mant 'Elberta' peach fruit buds. Proc. Amer. Soc.
Hort. Sci. 83:259—269.

1970. Relation of fall and winter

 

temperatures to flower bud behavior and wood hardi-
ness of deciduous fruit trees. HortScience 5(5):A22-
A2A.

and H. H. Mills. 1961. Loss of hardi—

 

ness by peach fruit buds as related to their morpho-
logical development during the pre—bloom and bloom
period. Proc. Amer. Soc. Hort. Sci. 78:10A—110.

and . 1972. A comparison

 

 

of hardiness responses in fruit buds of 'Bing' cherry
and 'Elberta' peach. J. Amer. Soc. Hort. Sci. 97:802-
806.

Sass, John E. 1958. Botanical Microtechnique. 3rd

Ed. Iowa State Univ. Press, Ames, Iowa. 228 pp.

2A.

25.

26.

27.

28.

29.

13

Steponkus, P. L. and F. O. Lanphear. 1967. Factors
influencing artificial cold acclimation and freezing

tests of Hedra helix 'Thorndale'. Proc. Amer. Soc.

 

Hort. Sci. 91:735-7A1.

and . 1969. The rela—

 

 

tionship of anthocyanin content to cold hardiness of

Hedra helix L. HortScience A(1):55-56.

 

Stushnoff, C. and B. F. Palser. 1969. Embryology of

five Vaccinium taxa including diploid, tetraploid,

 

and hexaploid species or cultivars. Phytomorphology
19(A):312—331.
Van Den Brink, C. 1973. A close look at the weather
for the Horticulture industry. HortScience 8(3):27l.
Van Huystee, R. B., C. J. Weiser and P. H. Li. 1967.

Cold acclimation in Cornus stolonifera under natural

 

and controlled photoperiod and temperature. Bot. Gaz.
128:200-205.
Weiser, C. J. 1970. Cold resistance and injury in

woody plants. Science 169:1269-1278.

1A

Table 1. Hardiness (T O) prediction equations square of
multiple corgelation coefficients (Ré), standard
error of the estimate (SE) and F test for regres-
sion for selected highbush blueberry cultivars,

 

 

1973-7A.
FALL
T50Z = —A5.233 + 12.358 (photoperiod) + 0.0323 (HZO/dwt %)
+ 0.A695 (x air temp) = 0.0528 (air temp X photo-
period)
R2 = 0.9939
SE = 0.61000

Fy = 778***

WINTER

T50 = 60.551 + 19.900 (bud dwt) + 0.0892 (HZO/dwt %) +
8.961 (color)

R2 = 0.9029

SE = 0.5580C

F = 3u***

SPRING

T50 = -2.530 + 9.556 (bud dwt) + 0.0097 (H2O/dwt %) +
10.232 (color) + 6.850 (photoperiod)

R2 = 0.98A0

SE = 0.929°C

F = 291***

 

Z is expressed as T50 + 100, a coding used to make all

T 0
t8mperatures positive.

yAsterisks denote level of significance at the P = 0.001
level.

15

Fig. 1. Fall, winter, spring hardiness (T 0°C) of 'Jersey',

5
1973-197A and estimated TSO'S from hardiness

equations.

O ou.~(!;mu

l 41.53

083‘»

 

APPENDIX

Fortran A program written for CD0 6500 to compute T50,

mean T50, variance of T50,

error of the T50 estimate for a cultivar or any treatment

using the Spearman-Kérber equations. The specific parame-

'S' statistic, and standard

ters of the input format statement are:

6 cols. date code

2 cols. treatment or cultivar code

1 col. replicate number

2 cols. number of observations (N) per rep

A cols. transformed lethal temperature, TL

3 cols. temperature interval value (d)

2 cols. field control value, number of dead buds in the
control in that rep

2 cols. number of b. values i.e. number of cold stress

treatments rom T to T

15 3-cols. b values for eacA cold stress treatment start-
ing with T (or T if necessary) to T

A cols. transformation fa8tor, absolute value of the
warmest cold treatment minus 1

Sample data card for Jersey rep 1, Table 1 S—K paper

Data starts in column 1 Data ends in column 79

121772 01 1 09 13.5 2.5 0 06 000 000 001 006 006 009 1A.0

The potential of this program is:

1-9 reps

99 labels for cultivars, treatments

0-9.9°F or C temperature interval, d

1-15 temperature treatments
1-999 observations per rep, N

A card is placed before each cultivar or treatment set of
data cards, which indicates the number of reps in that

cultivar or treatment.

A blank card follows the data deck, this terminates

the program.

Program SK (Input, Output, Tape 60 = Input, Tape 61 =
Output)
Dimension A (2A,A), Sum (15), Sum 2 (A), T50 (A),
(99)
(1) 9H Jersey

9H Rubel
9H Northland
9H Berkley
9H Coville
9H Earliblue
9H

9H Bluehaven
9H Jersey SH
9H Ran South
9H Ran Top

9H Ran North
9H Berk South
9H Berk Top
9H Berk North
9H Jer South
9H Jer Top

9H Jer North
9H Jersey C
9H Jer 2/Dry
Jer 2/Wet
9H Jer 20/Dry
9H Ear C

9H Ear 2/Dry
9H Ear 2/Wet
9H Ear 20/Dry
9H 2/C

9H 2/H2O
9H 2/Dry

9H 13/C

9H l3/H2O

9H l3/Dry

9H 25/C

9H 25/H2O

(A9 - 9H 25/Dry
Print A0

A0 Format (*1*,* Date Cultivar T50 1 T50 2 T50 3

T50 A Con

1Trol Mean T50 T50 Variance S T508em*)

C Read Number of Reps
10 Read 1, K
1 Format (11)
If (K.EQ.1H) Go to 51
Read 2, ((A(I,J), I=1,2A), J=1,K)

OONChU‘I-EUUNHCDNO‘xU'Itwml—‘HOKOCDNmUl-Il‘wNP-‘Ovvvvv

AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA
vvvvvvvvvvvvvvvvvvvvvvvvvvvvv

ttbtttttwwwwwwwwNMHF—‘HF—‘F—‘l—‘l—‘i—‘f—‘l—‘KDWKW“)

I II II II II II II II II II II II II II II II II II II II II II II II II II II II II
\0
TI:

ZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZZ

2 Format (F6.0,1X,F2.0, 1X, F1.0, 1X, F2.0, 1X, FA.0,
1X, F3.0, 1X,
1F2.0, 1X,F2.0,15F3.0,1X, FA.1)
0 Calculate for S
TS=0.0
DO 12 J-1,K
TN =A(A,J)
SBI =0.0
DO A I=9,23
L=I-8
Sum(L) = A(I,J)
SBI=SBI +Sum(L) * (TN-Sum(L))
A Continue
XK=K
S = TS / XK
C Calculations for Per cent Field Survival
TTN=0.0
l7 TFC=0.0
DO 18 J=1,K
TFC= TFC + A(7,J)
TTN=TTN+A(A,J)
18 Continue
FC=(TFC/TTN)*100.0
C Calculations for T50
DO 5 I=9,NBI
Sum2(J)=0.0
NBI=A(8,J) +8.0
DO 5 I=9,NBI
Sum2(J)= Sum2(J) +A(I,J)
5 Continue
DO 7 J=1,K
T50(J)=0.0
T50(J)=A(5,J) + A(6,J) /2.0-(A(6,J) *Sum2(J))/ A(A,J)
C T50 Transformed to Real Temperatures
T50(J)=(T50(J) +A(2A,J))*(-l.0)
7 Continue
If (K.EQ.A) Go to 25
DO 2A J=K,A
T50(J+1)=0.0
2A Continue
25 SQSum=0.0
SumT50=0.0
DO 27 J=1,K
SumT50= SumT50 + T50(J)
SQSum = SQSum + (T50(J)**2)
27 Continue
XK= K
VarT50 =(SQSum ~(SumT50**2) /XK)/(XK —1.0)
EMT50=SQRT(VART50 / XK)
XT50 =SumT50 / XK
M= A(2,l)
Print 50 , A(1,1), N(M), T50, FC, XT50, VART50, S,
EMT50

50 Format(*0*,F8.0, 1X, A10, F6.1, 1X, F6.1, 1X, F6.1,
1X, F6.1, 6X,
1F5.1, 5X,F6.2, AX, F8.3, 5X, F6.3,3X,F7.3)
Go to 10
51 Continue
End

A sample output, Jersey Table 1, SK paper

DATE CULTIVAR T501 T502 T503 T50A CONTROL MEAN T50
121772 JERSEY -22.6 —23.0 —23.8 —2A.6 100 —23.5

T50 VARIANCE S T5OSEM
.68A 0.183 .Al3

1..

A
A
I'll
A
A
III"
A
|
A
I
II
III
A
A
|
A
II
ll"
5|
5‘
ll
A
III"
III
II
A

3058 0

Li

I“

3

HUMAN