H triml'a
t ~v ..
SWIM
0781 7616
(:29 LIBRARY I
Michigan State
‘ University
~— I
*—
This is to certify that the
dissertation entitled
/K&Lf AJ—ALL-Ifl-A M) IL/V/Pb-f J./' 2;ng j] “Ifxf'Qz/e‘v‘.
( .
. I /( vim. vW-L-‘LW'I "4. 7 r‘nn.
{Cir-pi» ' I*/L Z L {M ‘ 3 C (/L «I..L,.vut
1Z6 l.-.. '~£-{// ,L, {[‘I [(obf:/ (L '- ‘V Z; L". k“ i/th ~f L ~L .. i 1. r i 3
/ ’ presented by
F f‘ « .. ‘ "‘ /i i ' .
f) ' / 7 / ./..’ IL/ . fit
I J l ‘ ' \ r , t 1 ‘ _. ’I' ., VI , t . ,
Vi K. I \_.( I ‘J' 'I ,s . ' .l f I: 14' 'l , u t - 4
has been accepted towards fulï¬llment
of the requirements for
V1? D degreein C :fC‘W'LC/i grwl >x «1"
QM J
Major profeéZor
ZYIZVy
Date
MS U is an Afï¬rmative Action/Equal Opportunity Institution 0-12771
PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
DATE DUE DATE DUE DATE DUE
DECnéz’ I
WW3
MSU Is An Affirmative Action/Equal Opportunity Institution
extrema-1139.1
THE EFFECT OF SEED SIZE, DENSITY AND PROTEIN CONTENT ON FTELD
PERFORMANCE, VIGOR AND STORABILITY OF TWO WINTER WHEAT VARIETIES
By
Riad Zouheir Baalbaki
A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Crop and Soil Sciences
1988
ABSTRACT
THE EFFECT OF SEED DENSITY, SIZE AND PROTEIN CONTENT ON FIELD
PERFORMANCE, VIGOR AND STORABILITY OF TWO WINTER WHEAT VARIETIES
EL’lï¬â€˜lul
3?
Riad Zouheir Baalbaki
A study was conducted to investigate the effects of different seed
characters of wheat (Igigigum aestivum L.) on field performance, vigor
and storability. Two winter wheat varieties, Augusta, a soft white
variety, and Hillsdale, a soft red variety, were divided according to
density, size and protein content and planted in two locations in
Michigan for two consecutive years. After planting, the same seed lots
were subjected to several vigor tests including accelerated aging, speed
of germination, conductivity index, ATP level, glutamic acid decarboxy-
lase activity (GADA), standard germination and cold test. Seed lots from
the first year experiment were stored up to 32 months under room
conditions. At various stages during the storage period, the ATP, GADA
and standard germination tests were performed and their results related
to vigor test results and viability in storage.
The field experiment results showed that differences in emergence
did not reflect differences in seed characters. However, heavy and large
seeds in both varieties consistently resulted in increased yields
compared to light and small seeds. Tiller number per meter had the
greatest effect on yield variation in both years, followed by seed
number per spike and lOOO-seed weight. Location effects significantly
affected yield, and Huron county yields were higher than those in East
Lansing in both years.
4—4‘
Vigor tests that involved stressing the seeds, such as the
accelerated aging test and the cold test were better indicators of yield
potential than non-stress tests like standard germination. Stress tests
were also better able to differentiate performance of the different
density, size and protein classes. Biochemical methods such as the CADA
and ATP tests were also good indicators of yield potential, but were not
sensitive enough to detect differences among all seed classes.
The viability of the different seed categories, as measured by the
standard germination test, did not change significantly up to 18 months
of storage, but declined rapidly thereafter. The CADA and ATP levels
showed a continuous decline throughout the entire storage period. Heavy.
large and high protein seeds stored better than light, small and low
protein seeds, and had higher germination at the end of the experiment.
While the ATP level was among the best tests in predicting storability,
the accelerated aging and speed of germination tests showed very little
correlation with viability during storage.
T0 MARIYA, MAI AND
DR. COPELAND
iv
ACKNOWLEDGMENTS
I would like to express my sincere gratitude and appreciation to
Dr. L. 0. Copeland, my major professor, for his guidance, advice and
constant encouragement throughout this study. He has shown me how to be‘
a better scientist and, perhaps more importantly, how to km: a better
person.
Thanks are also due to my other committee members, Drs. F. Dennis,
E. Foster, and R. Freed for their valuable suggestions and reviews of
this manuscript.
I would also like to thank Dr. E. Everson, D. Glenn, and L.
Fitzpatrick for their help with the field study, and Dr. D. Penner and
F. Roggenbuk for their help with the laboratory study.
I would also like to thank all my friends for their constant
support and their valuable comments on this work.
TABLE OF CONTENTS
Page
LIST OF TABLES ................. 2 ................................... vii
LIST OF FIGURES ..................................................... x
INTRODUCTION ........................................................ 1
CHAPTER I. THE EFFECT OF SEED DENSITY, SIZE AND PROTEIN
CONTENT ON FIELD PERFORMANCE OF WHEAT ............................... 4
ABSTRACT ......................................................... A
REVIEW OF LITERATURE ............................................. 6
MATERIALS AND METHODS ........................................... ll
RESULTS ......................................................... 15
DISCUSSION ...................................................... 30
REFERENCES ...................................................... 37
CHAPTER II. THE EFFECT OF SEED DENSITY, SIZE AND PROTEIN
CONTENT ON VIGOR TESTING OF TWO WHEAT VARIETIES .................... 41
ABSTRACT ........................................................ 41
REVIEW OF LITERATURE ............................................ G2
MATERIALS AND METHODS ........................................... 52
RESULTS ......................................................... 56
DISCUSSION ...................................................... 65
REFERENCES ...................................................... 72
CHAPTER III. THE EFFECT OF SEED DENSITY, SIZE AND PROTEIN
CONTENT ON STORABILITY OF TWO WHEAT VARIETIES ...................... 77
ABSTRACT ........................................................ 77
REVIEW OF LITERATURE ............................................ 78
MATERIALS AND METHODS ........................................... 81
RESULTS ......................................................... 82
DISCUSSION ...................................................... 92
REFERENCES ...................................................... 95
SUMMARY AND CONCLUSIONS ............................................ 97
APPENDICES ........................................................ 100
APPENDIX A ..................................................... 100
APPENDIX B ..................................................... 105
vi
LIST OF TABLES
. Page
Tablel. Emergence number and emergence rate index (E R.I.)
of two winter wheat varieties, Augusta and Hillsdale,
grown in East Lansing in 1985 and 1986 .................... 16
Table 2. Correlation coefficients between emergence number
and emergence rate index for Augusta and Hillsdale, 1985
and 1986 .................................................... 18
Table 3. Biological yield, straw yield, and grain yield of
two winter wheat varieties, Augusta and Hillsdale, 1985 ..... 18
Table 4. Biological yield, straw yield, and grain yield of
two winter' wheat varieties, Augusta and Hillsdale,
1986 ........................................................ 20
Table 5. correlation coefficients among biological yield,
straw yield, grain yield per meter and harvest index of
Augusta and Hillsdale, 1985 ................................ 22
Table (3. Correlation coefficients among biological yield,
straw yield, grain yield per meter and harvest index of
Augusta and Hillsdale, 1986 ................................ 22
Table 7. Yield components of two winter wheat varieties,
Augusta and Hillsdale, 1985 ................................. 23
Table 8. Yield components of two winter wheat varieties,
Augusta and Hillsdale, 1986 ................................. 25
Table EL Correlation coefficients among yield components
and grain yield of Augusta and Hillsdale ................... 26
Table. l0. ‘Yield. (kg/ha) of In“) winter wheat varieties,
Augusta and Hillsdale, grown. in two locations, East
Lansing and Pigeon, Michigan .............................. 28
Table 11. Effect of various seed characters on standard
germination of Augusta and Hillsdale in 1985 and 1986 ..... 57
Table 12. Effects of various seed characters on accelerated
aging test results cu? Augusta and Hillsdale in 1985
and 1986 .................................................... 57
vii
IIIIIIIIIIIIIIIIIIIIIIIIIll-lll---——________f
Table 13. Effects of various seed characters on cold
germination test results for Augusta and Hillsdale in
1985 and 1986. Percent germination .......................... 59
Table 14. Effects of various seed characters on speed of
germination test results for Augusta and Hillsdale in
1985 and 1986 ............................................... 59
Table 15. Effect of various seed characters on
conductivity index test results for Augusta and
Hillsdale in 1985 and 1986 .................................. 61
Table 16. Eff ct of several seed characters on ATP test
results (10- M/Lit) for Augusta and Hillsdale in 1985
and 1986 .................................................... 61
Table 17. Effect of various seed characters on glutamic
acid decarboxylase activity test results (ppm COZ/gm of
seed) for Augusta and Hillsdale in 1985 and 1986 ............ 6A
Table 18. Correlation coefficients between different vigor
tests, field emergence and yield of Augusta, 1985 and
1986 ........................................................ 67
Table 19. Correlation coefficients between different vigor
tests, field emergence and yield of Hillsdale, 1985 and
1986 ........................................................ 70
Table 20. Correlation of standard germination, ATP and GADA
results of the storage experiment for Augusta and
Hillsdale ................................................... 90
Table 21. Correlathmu of the accelerated aging, speed of
germination, conductivity index, cold test, ATP and GADA
tests with the standard germination test results after
one year, two years and 32 months of storage ................ 9O
Table A1. Analysis of variance for emergence (plants per
meter), and emergence rate index (E.R.I.) for Augusta
and Hillsdale, 1985 ........................................ 100
Table A2. Analysis of variance for emergence (plants per
meter), and emergence rate index (E R.I.) for Augusta
and Hillsdale, 1986 ........................................ 100
Table A3. Analysis of variance for biological yield, straw
yield, grain yield and harvest index per meter for
Augusta and Hillsdale, 1985 ................................ 101
Table A4. Analysis of variance for biological yield, straw
yield, grain. yield and. harvest index per meter for
Augusta and Hillsdale, 1986 ................................ 101
viii
—7—
Table A5. Analysis of variance for yield components of
Augusta and Hillsdale, 1985 ................................ 102
Table A6. Analysis of variance for yield components of
Augusta and Hillsdale, 1986 ................................ 102
Table A7. Analysis of variance for yield of two varieties,
Augusta and Hillsdale, grown. in two locations, East
Lansing and Pigeon in 1985 and 1986, and emergence rate
index (E.R.I.) for Augusta and Hillsdale, 1985 ............ 103
Table A8. Analysis of variance results for standard
germination (WC), accelerated aging (AA), cold test
germination (CC), and Speed of germination index (SC),
1985 ....................................................... 103
Table A9. Analysis of variance results for standard
germination (WG), accelerated aging (AA), cold test
germination (CC), and speed of germination index (SC),
1986 ....................................................... 104
Table A10. Analysis of variance results for conductivity
index (CI), ATP, and glutamic acid decarboxylase
activity (CADA), 1985 ...................................... 1014
Table All. Analysis of variance results for conductivity
index (CI), ATP, and glutamic acid decarboxylase
activity (GADA), 1986 ...................................... 104
Table Bl. Mean monthly temperature for two Michigan
locations, East Lansing and Pigeon for the 198A-1986
period ..................................................... 105
Table 82. Mean monthly precipitation for two Michigan
locations, East Lansing and Pigeon for the 1984-1986
period ..................................................... 106
ix
LIST OF FIGURES
Page
Figure l. Emergence number per'meter (E.N.) and emergence rate
index (E.R.I.) of Augusta and Hillsdale, 1985 ................... 31
Figure 2. Emergence number per meter (E.N.) and emergence rate
index (E.R.I.) of Augusta and Hillsdale, 1986 ................... 32
Figure 3. Storage effect on germination, ATP level and GADA of
two seed density classes, Augusta ............................... 83
Figure 4. Storage effect on germination, ATP level and GADA of
three seed size classes, Augusta ................................ 8a
Figure 5. Storage effect on germination, ATP level and GADA of
two seed protein levels, Augusta ................................ 85
Figure 6. Storage effect on germination, ATP level and GADA of
two seed density classes, Hillsdale ............................. 86
Figure 7. Storage effect on germination, ATP level and GADA of
three seed size classes, Hillsdale .............................. 87
Figure 8. Storage effect on germination, ATP level and GADA of
two seed protein levels, Hillsdale .............................. 88
INTRODUCTION
Usually seed quality is measured by germination and purity tests.
However, under a wide variety of environmental and soil conditions, the
standard germination test rarely gives an accurate indication of the-
performance of a seed lot in the field. However, vigor tests can provide
a better indication of field performance. Although vigor tests are now
commonly inuni to determine field emergence, stand establishment, and
sometimes yield potential of species such as corn, soybeans and cotton,
they have not been commonly used for wheat. Nor has the relationship of
vigor tests to overall field performance of wheat been studied
sufficiently. Another area that needs further study ii; seed physical
characters. Physical characters snufli as seed. size, seed. density and
protein content have not always been successfully related to seed vigor
of wheat and no consistent relationship has been established to
correlate such characters with plant performance during the entire
growing season. Furthermore, more work is needed to study the effects of
different seed characters on the results of vigor tests and to determine
the vigor test that is most sensitive to these physical differences.
Another aspect that also needs further study is the relationship of
seed vigor auui physical characters to seed storability. More work is
also needed to try to relate certain biochemical changes during storage
to viability and vigor, and to determine if such changes can be used to
indicate storage potential of seeds of varying quality levels.
1
This study had three broad objectives. (1) The first was to the
relationship between the seed physical characters (i.e., density, size,
protein content) and field performance as measured by stand
establishment, growth rate and yield. (2) The second was to study the
relationship beWeen seed vigor as determined by a series of vigor
tests, and field performance, and to identify the vigor test(s) that
most successfully predict the performance of a seed lot in field tests.
(3) The third objective was to relate seed vigor and physical
characteristics of density, size and protein content to storability.
The research consisted of three related experiments. The first was
a field experiment in which seed of two winter wheat varieties, Augusta,
a soft white variety, and Hillsdale, a soft red variety was divided into
different: density, size, and. protein. categories. The different
categories were then planted in the field for two consecutive years and
data on emergence and yield was obtained and correlated with the
different seed characters.
For the second experiment, the same varieties and seed categories
were used. Several vigor tests were performed and the results correlated
with field emergence and yield of the first experiment. The vigor tests
included the accelerated aging test, the cold germination test, the
electrical conductivity index, the speed of germination test, the
glutamic acid decarboxylase activity test, the ATP test, and the
standard germination test.
The third experiment consisted of storing the different seed
categories for 32 months and periodically testing the seeds using the
standard germination, ATP and glutamic acid decarboxylase activity
tests. Results of the standard germination test after different storage
.1
intervals were then correlated with results of vigor tests performed in
experiment two. In addition, results of the ATP and GADA tests at
different storage intervals were correlated. with the standard
germination results and to the performance of the different seed
characters during storage.
CHAPTER I
THE EFFECT OF SEED DENSITY, SIZE AND PROTEIN CONTENT
ON FIELD PERFORMANCE OF WHEAT
ABSTRACT
A two year field study was conducted to examine the relationship
between seed characters of winter wheat (Triticum aestivum L.) and field
performance. Two winter wheat varieties, Augusta, a soft white variety,
and Hillsdale, a soft red variety were divided according to size,
density and protein content and planted in two locations in Michigan.
Data were collected on emergence rate index, emergence number,
biological yield, straw yield, grain yield, and yield components (number
of tillers, seeds per spike, and lOOO—seed weight). Our results showed
that field emergence was significantly affected by treatment as well as
by replication effects. Small seeds resulted in reduced biological and
straw yield for both varieties in both years, while light and small
seeds resulted in lower grain yields than heavy and large seeds,
respectively. Tillers per meter, seeds per spike and lOOO-seed weight
contributed significantly to variation in grain yield under normal
conditions, but under stress conditions only the number of tillers and
seeds per spike contributed significantly to variation in yield. Tillers
per meter had the greatest influence on yield for both years and for
both 'varieties. Treatments did not significantly influence lOOO-seed
4
5
weight for either year, but had a significant influence on seeds per
spike and tillers per meter. location effects significantly affected
yield, and Huron county yields were higher than those in East Lansing in
both 1985 and 1986.
6
REVIEW OF LITERATURE
1. Seed Size
Successful stand establishment and consistent yields are essential
for efficient field crop production. Seed size has long been recognized
as an important factor affecting field emergence and seedling
establishment and can thus indirectly affect crop yields.
Variation in seed size within and among plants can be due to
genetic differences, interplant competition, effects of disease, and
location within the inflorescence and differences in flowering and
nutrition of the developing seeds (48). Wood et a1. (48) reported that
35 percent of the variation in size of barley (Hordeum vulgare L.) seed
was due to between-plant differences, 13 percent to difference between
ears (spikes), and 52 percent to differences in locations within the
ear. Evans and Bhatt (15) and Boyd et a1. (7) reported a significant
positive correlation between seed size and seedling vigor as measured by
rate of dry weight gain in young wheat (Triticum aestivum L.) and barley
plants. Freyman (16) found that plants from large wheat seeds were more
cold hardy than those from small seeds, which in turn were slightly
hardier than those from seeds with half the endosperm removed. Large
barley seeds produced larger seedlings, more tillers, and higher yield
compared to small, medium or ungraded seeds (13, 26, 27). Furthermore, a
highly significant correlation, between seed size and seedling fresh
weight, shoot length and root length of sorghum (Sorghum bicolor L.) was
observed (41).
Kaufmann and Guttard (26) found that the rate of seedling growth of
two barley varieties was greater for large than small seeds until the 2-
7
leaf stage, but afterward no differences in growth rate were observed.
Lawan et a1. (28) reported a significant increase in seedling emergence
of pearl millet (ï¬gnniggtum amggigangm L.) with increased seed size.
However, not all studies indicate a positive relationship between
seed size and field performance. Demirlicakmak et a1. (13) found no
effect of seed size on emergence of barley, while small amd medium sized
seeds of three corn (Zea may; L.) varieties had significantly higher
germination than large seeds under water stress conditions (36).
Abdullahi and Vanderlip (1) found no relationship between seed size and
yield of sorghum.
B. Seed Density
While seed density is independent of seed size, seed weight is a
measurement that incorporates both seed size and seed density. Ries et
al. (39) reported that seed weight of wheat varied according to its
position in the spike and was higher in the outside florets than the
middle one, and in the bottom 10 spikelets than in the terminal ones.
Yamazaki and Briggle (50) concluded that seed density of soft wheat was
dependent on environment rather than variety. They reported that air
spaces within the kernel largely determined seed density, and that the
extent of such air spaces was dependent on the rate and extent of seed
filling and was influenced by periods of wetting and drying.
Austenson and Walton (3) found a significant positive correlation
between initial seed weight of three wheat cultivars and number of
spikes per plant, seeds per plant, straw yield, grain yield and
biological yield. Ries and coworkers (38, 39) reported that seedling
vigor’ measured. by shoot dry weight was significantly and positively
correlated. with higher initial seed weight of wheat. McDaniel (33)
8
concluded that seedlings from heavy barley seeds had a higher growth
potential than those from light seeds and that higher amount of
mitochondrial respiratory activity of heavy seeds was indicative of a
greater amount of energy production and higher vigor.
Sung and Delouche (45) observed that germination percent, radicle
length and plumule length increased with increased density of rice
(9111; satixa L.) seed. They also reported that the benefit of high
density seeds was even more apparent under conditions of higher
emergence. Lawan et a1. (28) found that percent germination, seedling
height 24 days after planting, and proportion of vitreous endosperm
starch were positively related to seed density in pearl millet. In
studies with wheat seeds, specific gravity was positively correlated
with field stands, but showed no consistent relationship with yield
(11).
C. Protein Content
Any reserve nutrient that can influence the rate of germination and
seedling development can also influence emergence and crop yield. Meizan
et a1. (35) and CarciaDelMoral et a1. (17) reported that location,
genotype, location by genotype interaction and crop year all had
significant effects on seed protein content of wheat and barley. DeDatta
et a1. (12) showed that the protein content of rice increased at lower
plant densities since more N was available to each plant, and Cochran et
a1. (9) reported that when the supply of N limited yield, deep placement
of N increased both yield and protein content of wheat.
Many studies have shown that an increase in wheat seed protein can
be achieved with spring nitrogen applications in excess of that needed
for maximum yields (19, 23, 24, 46). Simazine application at flowering
9
time was also effective in increasing protein content of rye (37), brown
rice (47) and wheat (38). Wu and McDonald (49) found that increased N
application increased protein content, gluten, soluble protein, non-
protein N, and the nitrate content of wheat. They also found that the
ratio of protein to non-protein N did not change with application rate.
Ries and Everson (38) and Ries et a1. (39) found that seedling
vigor was highly correlated with seed protein content in wheat, and
Evans and Bhatt (15) reported that seedling vigor was positively related
to protein content when seed size was held constant. Lowe and Ries (32)
found that high protein content in either the endosperm or aleurone
layer resulted in more vigorous seedlings, irrespective of the embryo
protein content. Bittenbender and Ries (6) noted that high seed protein
affected vigor of rice by making seeds more resistant to loss of
viability during storage and by producing larger seedlings due to better
endosperm nutrition.
Lowe et al. (30) observed that proline and glutamic acid were the
sources of protein contributing to increased seedling vigor of wheat and
that protein content was positively correlated with seedling vigor and
grain yield. Ching and Rynd (8) found that high protein seeds produced
larger wheat seedlings than did low protein seeds after four days of
germination. They attributed these differences to increased efficiency
of metabolic activity and substrate transfer from the endosperm to the
seedling axis. Ayers et a1. (4) indicated that some catabolic
enzymatic component was more active in high than in low protein wheat
seed, and that rapid use of storage reserves was associated with greater
vigor during early growth, especially if rapid utilization was coupled
with other favorable enzymatic changes.
IIIIIIIIIIIIIIIIIIIIIIIIIll-ll-I:::———————_____i
lO
Oat (Agggg sativa L.) seeds with high protein content have been
reported to yield 21 to 42 percent more the controls, and wheat seed
with increased protein content developed into larger seedlings (40).
Similar results were obtained. by Lowe and Ries (31) where a high
correlation existed between high seed protein content of wheat and
shoot, root and total seedling dry matter 3 weeks after planting. Lopez
and Grabe (29) reported that high protein wheat seeds performed better
under stress conditions and GarciaDelMoral et a1. (17) found that the
protein content of barley seed was positively correlated with grain
yield†They also Showed. that: both factors ‘were correlated. with the
number of spikes per plant and to a lesser extent with seeds per spike
and grain weight.
The primary objective of this study was to examine the
relationship of seed density, seed size and protein content of two
winter wheat varieties to stand establishment, total yield and grain
yield. A second objective was to determine whether some kind of seed
selection during the seed processing operation was possible and
sufficient to improve field performance and crop yield. The final
objective was to determine the association between different seed vigor
indices and field performance.
11
MATERIALS AND METHODS
Experiment 1. 1985
Seed source: Untreated seeds of wheat grown in Michigan were
obtained from the Michigan Crop Improvement Association. Two winter
wheat varieties were used, Augusta, a soft white variety and Hillsdale,
a soft red variety. Seed lots from each variety were cleaned, uniformly
mixed and divided into two 5-kg sublets. One provided an unselected
control and the other was used to obtain different seed size and density
categories. Throughout the study, seed were stored at 5° C and 35
percent relative humidity.
Seed size: Seeds were divided into three size categories. For
Augusta, seeds retained on a 7/64" x 3/4" screen were considered as
large; those passing through a 7/64" x 3/4" screen but retained on a
6/64" x 3/4" screen were considered as medium; and seeds passing through
a 6/64" x 3/4" screen were considered as small. For Hillsdale, seeds
retained on a 6.5/64" x 3/4" screen were considered as large; those
passing through 6.5/64" x 3/4" screen but retained on a 5.5/64" x 3/4"
screen were considered as medium; and seeds passing through a 5.5/64" x
3/4" screen were considered as small.
Seed density: Seeds were divided into light, medium and heavy
density classes by using a Forsberg gravity table, Model 1052. Since no
clear dividing line existed between the light and medium and between the
medium and heavy categories, the medium category was discarded and only
the light and heavy categories were used.
Protein content: To obtain seeds with varying protein contents, the
two varieties had been planted the previOus year in 12 plots each at the
————
12
Crop and Soil Sciences field laboratory at East Lansing. At anthesis,
half the plots from each variety were sprayed with a solution containing
28 percent N at the rate of 20 kg/ha of N. The spraying was repeated 20
days after anthesis. After harvesting, four random samples of 200 grams
from each plot were ground, dried and analyzed for N content in
triplicate using the micro-Kjeldahl procedure (2). The total crude
protein of each sample was obtained by multiplying percent N per gram of
seed by a factor of 5.7. Since the unselected seed lots of Augusta and
Hillsdale had an average protein content of 11.6 and 12.1 percent,
respectivily, seed lots of Augusta and Hillsdale with 12.6 and 13.1
percent protein were selected as high protein and those with 10.6 and
11.1 percent were selected as low protein, respectively.
Field study: Forty eight plots were planted in a factorial
(variety x treatment) experiment in a randomized complete block design
with 4 blocks. The factors were varieties (Augusta and Hillsdale) and
treatments (2 seed densities, 3 seed sizes, 2 protein contents, and an
unselected control) so that each block contained a total of 16
treatments [2 x (3+2+2+l)] or plots. Each plot consisted of 5 rows
established with a seed drill delivering approximately 1730 seeds in a
1.2 x 3.7 meter area giving a seeding rate of about 140 kg/ha. The plots
were fertilized N at the rate of 90 kgs/ha split into a preplanting and
a spring application. The experiment was planted in 2 locations; one at
the field laboratory in East Lansing, and the other near Pigeon in Huron
County, Michigan.
Field data: Prior to seedling emergence, one meter from the middle
row of each plot at the East Lansing location was marked off for data
collection. Emergence number per meter, emergence rate index, biological
IIIIIIIIIIIIIIIIIIIIIIllllllll---—________.
13
yield per meter, straw yield per meter, grain yield per meter and number
of tillers per meter were recorded in this sampling unit. Grain yield
(bu/acre) and lOOO-seed weight were determined using the whole plots
harvested. with a small plot Hege combine. Only yield in kg/ha was
obtained from the Huron county plots.
Emergence rate index (E.R.I) was calculated using the same formula
developed by Maguire (34) to measure the speed of germination and was as
follows:
E.R.I - No. of seedlings emerged/No. of days to first count +...
+ No. of new seedlings emerged/No. of days of final count
Counting was started 3 days after planting and terminated 21 days after
planting, and the final count recorded as emergence number per meter.
After collecting emergence data all plots were thinned back to an
average of 65 plants per meter. For the rest of the data, only three
blocks were used in each location.
Biological yield was determined by weighing the above-ground
portion of the plant after drying. Grain yield per meter was obtained by
threshing the plants and then straw yield calculated as the difference
between biological yield and grain yield. Total grain yield was
calculated by converting the yield of each plot to yield in kg/ha. All
seed weights and yields were reported on a 12.5 percent moisture content
basis. lOOO—seed weight was obtained by counting four lOOO-seed samples
from each plot, weighing them and averaging the results. The number of
tillers per meter was obtained by counting the number of seed-bearing
stems per meter, and seeds per spike was obtained by dividing the total
number of seeds per meter by the number of tillers. The harvest index
was calculated by dividing the seed weight per meter by the biological
IIIIIIIIIIIIIIIIIIIIIIIllllllllI--——________,
14
yield and multiplying by 100.
2. Experiment 2. 1986
The second experiment was conducted to verify the results of the
first experiment and to further our understanding of the relationships
under study. This second experiment was similar to the first with two
exceptions; the number of blocks was increased from three to four and
the number of treatments per variety was increased from 8 to 11. The 3
extra treatments were medium protein, unselected minus light and
unselected minus small.
The unselected minus small category was obtained by using one
screen with a hole size of 5.5/64" x 3/4" and 6/64" x 3/4“ for Hillsdale
and Augusta, respectively. The seeds retained on the screen were
considered as the unselected minus small treatment. The unselected minus
light category was obtained by removal of the light seeds using the
gravity table and combining medium and heavy seeds in one density
category. Medium protein was considered to be 12 percent for both
varieties. All other treatments were the same as those used in the first
experiment.
3. Statistical Analysis
In both experiments, all variables were subjected to analysis of
variance following the procedures outlined by Steel and Torrie (42).
Means were separated using Duncan’s Multiple Range Test (DMRT) at the 5
percent level of probability. Simple correlations and a forward
selection stepwise multiple regression analysis to select the best fit
model were calculated using all data points combined for both varieties.
The SAS personal computer package was used for the analysis.
—7—*
15
RESULTS
In 1985, emergence number of the unselected control of Augusta was
significantly lower than that of all other treatments (Table 1). While
no significant differences existed between the different classes of
density and size categories of Augusta, the low protein treatment had
significantly lower emergence than the high protein treatment. No
treatment effects were significant in Hillsdale. Emergence rate index
(Table l) was lower for the unselected control treatment of Augusta than
for the heavy, medium, large and high protein treatments. As with the
emergence number results, only the protein classes significantly
differed from each other while density and size classes did not. Again,
differences in emergence rate were not significant in Hillsdale.
Analysis of variance (Table A1) indicated that while block and
treatment significantly affected the emergence number and emergence rate
index, varieties did not and only the treatment by variety interaction
was significant for emergence number.
In 1986, blocks and treatments had a significant effect on
emergence number and emergence rate index (Table A2), while varietal
effects and treatment by variety interaction did not. In Augusta (Table
1) the emergence number and emergence rate index did not differ
significantly between any classes of the three seed categories; neither
were any treatments significantly different from the control.
Hillsdale results in 1986 (Table 1) indicated that while no
treatment differed significantly from the control in emergence number
and emergence rate index, the small seed treatment was significantly
lower than the medium, large and unselected minus small treatments. For
l6
Tablel. Emergence number and emergence rate index (E.R.I.) of two
winter’ wheat varieties, Augusta and Hillsdale, grown in East
Lansing in 1985 and 1986.
1985 1986
Variety Treatment E No. E.R.I E No. E R I
AUGUSTA Unselected 69c 7.76d 78ab 8.74abc
Light 81b 9.1lbcd 98a 11.03a
Heavy 93ab 10.38abc 88ab 9.99abc
Unsel-light* - - 86ab 9.27abc
Small 87b 9 27abcd 69b 7 61c
Medium 90b 9.68abc 88ab 9.69abc
Large 93ab 10.63ab 89ab 10.05abc
Unsel-small** - - 77ab 8.37bc
Low protein 89b 8.49cd 92a 10.50ab
Medium protein - - 81ab 9.4Zab
High protein 104a 11.14a 96a 11 25a
Hillsdale Unselected 86 9.36 74ab 8.02ab
Light 84 9.42 94a 10.54a
Heavy 87 10.41 82a 9.423
Unsel-light - - 88a 10.01a
Small 83 9.59 61b 5.99b
Medium 91 10.25 923 10.07a
Large 86 10.74 84a 9.62a
Unsel-small - - 91a 10.00a
Low protein 94 10.48 85a 9.4la
Medium protein - - 89a 9 96a
High protein 91 10.19 79ab 8.81s
n.s n.s
* - unselected minus light, ** - unselected minus small
For each 'variety, means followed. by the same letter in each
column are not significantly different at the S % probability
level according to DMRT.
—
17
both varieties in both years, a highly significant correlation existed
between emergence number and emergence rate index (Table 2).
2. Biological, Straw and Grain Yield
In the 1985 experiment (Table 3), small seeds of both varieties
gave significantly lower biological yield and straw yield than all other
treatments, but TM) other differences were significant. Most variation
occurred in grain yield. per' meter. In .Augusta (Table 3), the light
treatment yielded significantly less than the heavy treatment; the small
treatment yielded significantly less the medium treatment and both in
turn were significantly lower than the large treatment; and the low
protein treatment yielded significantly less than the high protein
treatment. The light, small and low' protein treatments all yielded
significantly less than the unselected control.
No treatments of Hillsdale in 1985 produced grain yields differing
significantly from that of the control (Table 3). While density classes
did not differ, the small seed treatment yielded significantly less than
the large treatment, and the low protein treatment yielded significantly
less than the high protein treatment.
The harvest index in 1985 (Table 3) was similar for both varieties
in that small seeds had ‘very high indices compared to most other
treatments. While significant differences were observed among size and
protein classes of' Augusta, no significant differences occurred in
harvest index among different density classes. In Hillsdale, the small
seed treatment had a significantly higher harvest index than the medium
treatment, while I“) differences existed between different density and
protein classes. Analysis of variance (Table A3) showed that both
treatment and variety significantly affected all variables tested except
Table 2.
18
Correlation coefficients between emergence number
and emergence rate index for Augusta and Hillsdale, 1985 and
1986.
Augusta
Hillsdale
1985 1986
0.87:: 0.97::
0.73 0.98
** - Significant at the 19 probability level.
Table 3. Biological yield,
straw yield,
wheat varieties, Augusta and Hillsdale, 1985.
and grain yield of two winter
Variety
AUGUSTA
HILLSDALE
For each variety, means
DMRT.
Treatment
Unselected
Light
Heavy
Small
Medium
Large
Low protein
High protein
Unselected
Light
Heavy
Small
Medium
Large
Low protein
High protein
Yield (gms/m)
.................................
355a
344a
367a
305b
357a
375a
342a
377a
2283
222a
237a
l86b
226a
235a
220a
237a
l27de
139abc
ll9e
l36bcd
148a
l3lcd
150a
129ab
l23b
130ab
119b
l30ab
139a
l23b
140a
index
.......................................................................
41.
38.
40.
38.
41.
36
35.
35
39
36
37
35.
37.
2a
3bc
lab
2bc
1a
.2b
7b
.5b
.0a
.4b
.2ab
8b
2ab
followed by the same letter in each column are
not significantly different at the 5 % probability level according to
lIIIIIIIIIIIIIIIIIIIIIIIIIlllllll--::r—————____i
19
for‘ biological yield. where only treatment effects were significant.
In 1986 (Table 4), no treatment in either variety differed
significantly from the control in biological or straw yield except the
small seed treatment of Hillsdale. Also for both varieties, no
significant differences in biological yield were observed between the
density and protein classes, while the small seed treatment was
significantly lower than all other size classes. Straw yield results
revealed rm) significant differences between classes of tin; three seed
categories.
As in 1985, most of the variation in 1986 occurred in grain yield
(Table 4). While no treatment differed significantly from the unselected
control of Augusta, light seeds yielded significantly less than heavy
and unselected minus light seeds, and small seeds yielded significantly
lower than medium, large and Luwelected minus small seeds. No
significant differences were observed between protein classes. Grain
yield the light, small and medium seed treatments of Hillsdale in 1986
(Table 4) were all significantly less than that of the control. As with
Augusta, significant differences in grain yield occurred between
different density and size classes but not among protein levels. Only
the harvest index of the light treatment was significantly different
from the control of Augusta in 1986 (Table 4). However, for Hillsdale,
the light, small and medium treatments were significantly different from
the control. Significant differences were recorded among density classes
for both varieties. While the high protein treatment of Augusta had a
Significantly higher harvest index than the low protein treatment, the
large seed treatment had a significantly higher harvest than the small
and medium seed treatments of Hillsdale.
20
Table 4. Biological yield, straw yield, and grain yield of two winter
wheat varieties, Augusta and Hillsdale, 1986.
Yield (gms/m)
--------------------------------- Harvest
Variety Treatment Biological Straw Grain index
AUGUSTA Unselected 222ab l40ab 82abc 37.1ab
Light 233a 160a 73bc 31.4c
Heavy 243a 151ab 92a 37.8ab
Unsel-light* 238a l49ab 89a 37.7ab
Small 198b 126b 72c 36.3abc
Medium 232a lSlab 81abc 35.0abc
Large 234a l42ab 93a 39.6ab
Unsel-small** 241a 155ab 86ab 35.8abc
Low protein 233a 153ab 80abc 34.3bc
Medium protein 239a lSOab 88a 39.8a
High protein 233a l4lab 93a 37.2ab
HILLSDALE Unselected 240a 152ab 88ab 36.8a
Light 242a 177a 65d 26.4c
Heavy 259a 165ab 94a 36.3a
Unsel-light 254a l68ab 86abc 33 7ab
Small 208b l44b 64d 30.7bc
Medium 2443 l70ab 74cd 30 2bc
Large 250a 158ab 92ab 36.7a
Unsel-small 243a 162ab 80bc 33.1ab
Low protein 251a 17lab 80bc 32.Sab
Medium protein 240a 158ab 81abc 36.9a
High protein 250a 158ab 92ab 33.9ab
* - unselected minus light, ** - unselected minus small
For each variety, means followed by the same letter in each column are
not significantly different at the 5 % probability level according to
DMRT.
IIIIIIIIIIIIIIIIIIIIIllllllll---—_______,
21
Both treatment and variety effects in 1986 significantly influenced
all results (Table A4) except grain yield, which was influenced only by
treatment effects. While the biological, straw and grain yield were all
positively and significantly correlated with each other for both
varieties in 1985 (Table 5), the harvest index was significantly but
negatively correlated with straw yield of Augusta and biological and
straw yield of Hillsdale. In 1986 (Table 6), the harvest index of
Augusta was negatively correlated with biological and straw yield but
positively correlated with grain yield. Grain yield was positively
correlated with biological yield, and the straw and biological yield
were positively correlated. Both straw and grain yield of Hillsdale were
negatively correlated with the harvest index and positively correlated
with biological yield; however grain yield and straw yield were not
significantly correlated.
3. Yield Components
The components of grain yield per meter, were tillers per meter,
seeds per spike and 1000-seed weight. In 1985, the small seed treatment
resulted in significantly less tillers than all other treatments for
both varieties while no other treatment differed significantly from the
control (Table 7). While the small seed treatment of Augusta was
significantly lower in seeds per spike than the control (Table 7), the
large and high protein treatments were significantly higher than the
control. For Hillsdale, the heavy, large and high protein treatments
were all significantly higher than the control in seeds per spike. In
1985, there were significant differences in both varieties between
different classes of density, size and and protein content.
22
Table 5. Correlation coefficients among biological yield, straw
yield, grain yield per meter and harvest index of Augusta and
Hillsdale, 1985.
Biol. yld. Straw yld. Grain yld.
Augusta Harvest index -0.30 -0 59** O 30
Grain yld. 0.82** 0.59**
Straw yld. O.95**
Hillsdale Harvest index -O.53** -O.67** -0.ll
Grain yld. 0.90** 0.81**
Straw yld. 0.98**
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
Table 6. Correlation coefficients among biological yield, straw
yield, grain yield per meter and harvest index of Augusta and
Hillsdale, 1986.
Biol. yld. Straw yld. Grain yld.
Augusta Harvest index -0.38* -0.7l** 0.53**
Grain yld. 0.58** 0.23
Straw yld. 0.92**
Hillsdale Harvest index 0.03 -0.55** -0.83**
Grain yld. 0.58** 0.01
Straw Yld. 0.82**
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
Table 7.
Yield components of two winter wheat varieties,
23
Augusta and Hillsdale, 1985.
..............................................................
AUGUSTA
HILLSDALE
Treatment
Unselected
Light
Heavy
Small
Medium
Large
Low protein
High protein
Unselected
Light
Heavy
Small
Medium
Large
Low protein
High protein
Tillers
per meter
120a
118a
124a
lOOb
116a
126a
118a
128a
Seeds per
spike
30bc
30bc
32a
29c
31ab
32a
30bc
32a
lOOO-seed
wt. (gms)
41.63bc
43.04ab
40.46c
42.25ab
43.34a
42.97ab
43.6la
40.61
40.34
40.49
40.32 i
40.56
40.43
40.23
40.00
n.s
For each variety, means followed by the same letter in each
column are not significantly
probability level according to
DMRT.
different at the 5 %
IIIIIIIIIIIIIIIIIIIIIIIIIllllllll----—_______i
24
Only the small seed treatment of Augusta had a significantly lower
lOOO-seed weight than the control in 1985 (Table 7), while no
significant differences occurred for Hillsdale. The small seed treatment
of Augusta also had a significantly lower lOOO-seed weight than both the
medium and large treatments.
In 1986 (Table 8), the small seed treatment had significantly fewer
tillers than the control of both varieties, and only the heavy treatment
of Hillsdale had significantly more tillers than the control.
Significant differences in tillering of both varieties occurred among
different density and size classes, while no significant differences
were observed between different protein levels.
No significant differences in seeds per spike were recorded between
the control and any treatment for either variety (Table 8). For Augusta,
the light treatment was significantly lower than the heavy treatment,
the small and medium significantly lower than the large, and the low
protein significantly lower than the high protein treatments. For
Hillsdale, the small and unselected minus small treatments were
significantly lower in seeds per spike than the large seed treatment,
and the low protein treatment significantly lower than the high protein
treatment.
No significant differences in lOOO-seed weight were recorded in
1986 between any of the treatments for either variety (Table 8).
Treatment in 1985 significantly influenced the number of tillers
per meter and seeds per spike but not the seed weight (Table A5).
However, variety significantly influenced the number of tillers and seed
weight but had no effect on seeds per spike. The treatment x variety
interaction was significant for seeds per spike and seed weight. In 1986
25
Table 8. Yield components of two winter wheat varieties,
Augusta and Hillsdale, 1986.
Tillers Seeds per lOOO-seed
Variety Treatment per meter spike wt. (gms)
AUGUSTA Unselected 90ab 32abc 27.87
Light 79bc 3lbc 29.A2
Heavy 99a 35a 29.77
Unsel-light* 91ab 33ab 31.72
Small 75c 3lbc 30.02
Medium 87abc 3lbc 29.97
Large 99a 34a 31.78
Unsel-sma11** 98a 3Zabc 30.72
Low protein 90ab 30c 30.22
Medium protein 90ab 32bc 31.84
High protein 97a 33ab 31.58
n.s
HILLSDALE Unselected 92b 31ab 29.56
Light 89b 30b 29.99
Heavy 106a 32ab 30.51
Unsel-light 91b 31ab 30.45
Small 76c 30b 31.67
Medium 89b 32ab 30.94
Large 98ab 33a 29.21
Unsel-small 98ab 30b 31.57
Low protein 92b 30b 30.78
Medium protein 91b 31b 30.62
High protein 97ab 33a 30.59
n.s
* - unselected minus light, ** - unselected minus small
For each variety, means followed by the same letter in each
column are not significantly different at the S % probability
level according to DMRT.
—"
26
Table 9. Correlation coefficients among yield components and grain yield
of Augusta and Hillsdale.
1000-
seed Wt.
Augusta
Tillers
per meter
Seeds per
spike
1985 ------------------------------
Tillers/m 0.S7*
Seeds/spike 0.7l**
Grain yld. 0.82**
1986
Tillers/m 0.14
Seeds/spike 0.3a
Grain yld. 0.41
0.53*
0.78**
0.68**
1000-
seed wt.
.........
Hillsdale
Tillers Seeds per
per meter spike
0.60*
0.82** 0.61*
0.23
0.45* 0.66**
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
—7—'
27
(Table A6), treatment affected both number of tillers per meter and
seeds per spike, while variety only affected seeds per spike. The
interaction between the two was not significant.
While all yield components were significantly and positively
correlated with each other and with grain yield for Augusta in 1985
(Table 9), in 1986 the lOOO-seed weight was not correlated with either
yield components or grain yield. lOOO-seed weight of Hillsdale (Table
10) was not correlated to either yield or yield components in either
year, while the number of tillers and seeds per spike were significantly
and positively correlated with grain yield for both years.
The resulting regression equations after forward stepwise selection
were:
Y1985 - 452.367 + 3.961X1** + 0.489X2** + 2.091x3* (R2 - 0.72)
rm“ - 45.1.1.7 + 0.59ox2** + 2.36x3** (R2 - 0.56)
where Y - grain yield per meter
X1 - lOOO-seed weight
X2 - number of tillers per meter
X3 - seed number per spike
The yield of whole plots from both East Lansing and Pigeon was
compared in 1985 and 1986 (Table 10). In 1985, only the small seed
treatment of Augusta in East Lansing was significantly different from
the control. While both the light and small seed treatments of Hillsdale
yielded significantly less than the control in the Pigeon location, no
treatment differed significantly from the control at East Lansing. While
only size classes of Augusta differed significantly in yield in both
locations, no significant differences were recorded for Hillsdale in any
seed category at East Lansing. All categories showed significant
28
Table 10. Yield (kg/ha) of two winter wheat varieties, Augusta and
Hillsdale, grown in two locations, East Lansing and Pigeon, Michigan.
1985 1986
Variety Treatment E.Lansing Pigeon E.Lansing Pigeon
AUGUSTA Unselected 5972abc 6127abc 353la-d 3873ab
Light 5539cd 5976abc 3128cd 3408bc
Heavy 6065abc 6270abc 3982ab 4358a
Unsel-light* - - 3934ab 4166ab
Small 519ld 5676c 3039d 3053c
Medium 5669de 5901bc 338lcd 3873ab
Large 618lab 6495ab 3968ab 4440a
Unsel-small** - - 3695a~d 4016ab
Low protein 6024abc 6113abc 3517a-d 3449bc
Medium protein - - 3750abc 3955ab
High protein 6325a 6591a 4146a 4214ab
HILLSDALE Unselected 5628ab 6hl3ab 3750ab 39h8abc
Light 5334b 5457d 29l6d 3388cd
Heavy 5703ab 6878a 3920a 4317a
Unsel—light - - 3620abc 4194ab
Small 5444ab 5539cd 3046cd 3087d
Medium 5532ab 6359ab 3156bcd a064ab
Large 5949ab 6632ab 3893a 4501a
Unsel-small - - 3422a-d 4084ab
Low protein 5375ab 6051bc 3470a-d 3558bcd
Medium protein - - 3&97a-d 407lab
High protein 6004a 6721a 3893a 4603a
* - unselected minus light, ** - unselected minus small
For each variety, means
are not significantly
according to DMRT.
followed by the same letter in each column
different at the S % probability level
—:—
29
differences among classes in Pigeon.
In 1986, the small seed treatment yielded significantly less than
the control for both varieties and locations, with the exception of
Augusta in East Lansing. Though significant differences in yield were
always observed between different density and size classes, no
differences in yield among protein levels were recorded for Augusta at
Pigeon or for Hillsdale at East Lansing.
Location and treatment effects were significant in both 1985 and
1986, while variety effects were not (Table A7). Treatment x variety
interaction and location x variety interaction were also significant in
1985.
30
DISCUSSION
Results showed that the final emergence level was closely
associated with the speed of emergence or emergence rate index.
Correlations between the two were highly significant for both years and
for both varieties. Figures 1 and 2 clearly illustrate that regardless
of the specific treatments, both the total emergence and emergence rate
index followed the same trend, especially in 1986. This indicates that
the emergence capacity of a seed lot is dependent on the speed of
emergence; thus a seed lot that exhibits quick emergence is also
expected to have a high final emergence. Conversely, slow emerging lots
tend to have a low final emergence, even after extended periods of time.
Although our choice of treatments influenced emergence results,
grouping the data by seed characters (Figures 1 and 2) demonstrates that
seed size consistently influenced seedling emergence, while seed density
and protein level did not. With the exception of emergence number for
Hillsdale in 1985, large and medium Seeds always had better emergence
than small seeds. However, the density data show that while the light
seeds gave a lower or equal total emergence and emergence rate index
than the heavy seeds in 1985, the trend was reversed in 1986, and the
same kind of inconsistensies occurred for seeds of different protein
levels. These results indicate that while seed size had a consistent
effect on emergence, seed density and protein content did not.
Seeding depth could also affect emergence. Many other studies (5,
10. 43) have concluded that seeding depth affected rate and final
' and rotein
emergence. Since seed size varied little across denSity p
level, Size effects and their interaction with planting depth probably
31
xenc. 30m cocootoEu
{12
f 1
xenc. 20m mocongm
H2
1
CL
A n
T D
S S
U .L
m m
A. H in
x _. _
a m WHWZXWZW
7 b J I /hW(
L m 1 m _
1mfm I.--
-e\ve .7?
\ICVIC _
Zmï¬m _
«v.5 . m. __
name 1
whim MKHMflWK¢V
.miw 47/
N 0R n
tutu
I _
m"m_ Smfl
q<<<d<< _. 4 i W J ~ 4‘ 1“ n‘V
m m w a
t302\2co_d
tomo2\3co_d
ecvy Smofl Medium Large be Prat HI Poet
H
Ugh!
size and protein
Emergence number per meter (E.N.) and emergence
of several density,
index (ERI)
categories of Augusta and Hillsdale, 1985.
Figure 1.
rate
32
3
1|
_
xoos 30m cocootoEm
xocE 30m oocomtoEm
W H“ m o. .o 7 H mm o. 8 v: 6 5
h L h _ L b— F _ _ h _ _ “
flMH/Mdflrwfl/ _qu./mmw
. _
IIISI..:Tll- F. -!:.-+il!! lx|:!:li
A gï¬W/Mfl/ Na W - .-.-¢-_m--.l. 7
rent g1. JV- w - _- . -
w ._ L _ _
nu _F um . H
A M� H g???
x r _ .11
r a a - -r - %
e3.m s _. :1
2. m z e WWW/Va. -Wï¬. :3-
8 u d. _
. “mm a _ _
mMMWMVm _ _
â€D 9 . W _ _
. u D e .
sham _ _
w and. a .1 but- %
Fr. .u Elf/.3..-
N MRW _ H
Enoch 15
m _ w W - . - a
_ w _ 3
W4 1 1 ~ 1 1“ — 1 1 1 14 d 1 1 1 1 41 m 1 fl 1 —
O 0 O O O 0
1| 0 9 8 7 6
touosimucoi
cou02\wuce_d
Large Lo Pro: HI Pro!
Medium
Small
Ugh!
gence
and protein
size
y.
1986.
of several densit
Emergence number per meter (E.N.) and emer
(ERI)
index
categories of Augusta and Hiilsdale,
Figure 2.
rate
33
overshadowed other differences. Since block effects were significant in
both years despite the uniformity in soil type and fertility of the
experimental area, other types of soil variation (e.g., soil compaction
or seeding depth) may have influenced the emergence results. Many other
studies on emergence have produced similarly contradictory results.
While some indicated that emergence improved with larger and/or higher
density seeds (18, 25, 28), others have associated poor or slow
emergence with the same seed characters (14, 20, 22), or have indicated
no influence of size on emergence (13).
The reduction in all three yield types, i.e., biological, straw and
grain yield, in 1986 compared to 1985 was mainly due to unfavorable
conditions in the 1986 season. While environmental conditions in 1985
were conducive to good growth and yield, in 1986 an extended dry period
occurred through anthesis and seed development, followed by a wet period
during time of harvest (Appendix B). Such environmental conditions could
be the principal cause of the change in relationship between biological,
straw and grain yield, irrespective of variety.
The harvest index is a measure of the proportion of grain yield to
biological yield, or a measure of the plant’s efficiency in producing
seeds. Although the small seed treatments produced the poorest grain
yields, in most cases they had a relatively high harvest index. Plants
from small seeds, with low straw yield and low overall growth, may be
under more pressure to improve their grain production efficiency, while
plants with better vegetative growth and more nutrient reserves would
require less seed production efficiency. This relationship is further
illustrated when results of the first and second year experiments are
compared. Although a substantial reduction in grain yield occurred in
IIIIIIIIIIIIIIIIIIIIIIII::___________________——ï¬
34
1986 relative to 1985, the harvest index was only slightly lower in
1986.
The harvest index can also be used to illustrate some differences
between two low yielding seed lots. Unlike the small seed treatment,
which had a high harvest index, the light seed treatment with a similar
low yield, had a low harvest index in 1986 because it produced a
relatively high amount of straw. Since the small seed treatment had a
slower growth rate than the light seed treatment, it perhaps had a
better chance to adjust to the stress conditions, while the plants
produced by light seeds were unable to shift their emphasis to increased
grain production and decreased vegetative growth. It should be noted
that since all plots were thinned back after emergence data were
collected, differences in treatments were not due to differences in
plant number but to differences in performance of equal numbers of
plants.
Most of the variation due to our treatments occurred in grain yield
rather than biological or straw yield. An exception was the small seed
treatment which yielded least for both years and both varieties. To
examine the sources of variation in grain yield, an examination of the
different yield components was necessary. By comparing the different
partial regression coefficients of the yield components, the number of
tillers per meter was found to be the most important factor contributing
to yield, followed by seed number per spike and then lOOO-seed weight.
While in 1985 all three yield components had a significant contribution
to yield variation, in 1986, and under less favorable conditions, only
the number of tillers and seed number per spike significantly
contributed to yield variation. These results indicated that favorable
35
conditions allow production of more tillers, more seeds per spike, and
larger seeds. Under unfavorable conditions plants produce more seeds per
spike but smaller seeds. Our results were in agreement with those of
other studies that concluded that number of tillers was the prime factor
in determining yield of barley (27), rice (45), and sorghum (44).
While most treatments did not affect tillering in 1985, more
response was observed in 1986. If plant growth is vigorous, as in 1985,
interplant competition should develop sooner, and the advantage of
better quality seed might not be realized. However, under environmental
stress (1986), growth is diminished and less competetion results in
greater differences in tillering. Furthermore, since the number of
tillers is the primary factor contributing to yield, yield differences
under stress will be more pronounced.
Our results support such a premise in that the differences in yield
between the control and the treatments were more pronounced in 1986 than
in 1985. These results agree with reports by Boy and Gamble (21) who
reported that effects of seed size and density of soybeans were greater
under greater field stress. Whole plot yields showed the same response
to treatments as grain yield per meter, and correlations between the two
were highly and positively significant so that we can assume that the
yield per meter was a good estimate of the whole plot yield. When plot
results were combined over locations, treatment and location effects
were significant in both 1985 and 1986. Since only the plots in East
Lansing were thinned, we cannot say whether the location effects were
entirely or partly due to thinning. Since the Pigeon yields were higher
than those at East Lansing, some effect of thinning should be assumed.
However, a further study on thinning effects is neede to verify this
36
point. Since there was no treatment by location interaction, our
treatments are expected to produce the same effects and have the same
trends in either location regardless of yield level.
pr
—:——
N
w
b
VI
0"
\J
on
10.
11.
12.
. Austenson,
37
References
. Abdullahi, A., and R. L. Vanderlip. 1972. Relationships of vigor
tests and seed source and size to sorghum seedling establishment.
Agron. J. 64: 143-144.
. Association of Official Analytical Chemists. 1975. Methods of
analysis, 12th Ed. Assoc. Off. Anal. Chem. Washington, D. C., p.
1094.
H. M., and P. D. Walton. 1970. Relationships between
initial seed weight and mature plant characters in spring wheat.
Can. J. Plant Sci. 50: 53-58.
Ayers, G. S., V. F. Wert, and S. K. Ries. 1976. The relationship of
protein fractions and individual proteins to seedling vigour in
wheat. Ann. Bot. 40: 563-570.
Beveridge, J. L., and C. P. Wilsie. 1959. Influence of depth of
planting, seed size, and variety on emergence and seedling vigor in
alfalfa. Agron. J. 51: 731-734.
. Bittenbender, H. C., and S. K. Ries. 1977. Germination and growth of
rice seedlings from high and low protein seeds after exposure to
various storage conditions. J. Seed Technol. 2: 62-72.
Boyd, W. J. R., A. G. Gordon, and L. J. LaCroix. 1971. Seed size,
germination resistance and seedling vigor in barley. Can. J. Plant
Sci. 51: 93-99.
Ching, T. M., and L. Rynd. 1978. Developmental differences in
embryos of high and low protein wheat seeds during germination.
Plant Physiol. 62: 866-870.
. Cochran, V. L., R. L. Warner, and R. F. Papendick. 1978. Effect of N
depth and application rate on yield, protein content, and quality of
winter wheat. Agron. J. 70: 964-968.
Dalianis, C. D. 1980. Effect of temperature and seed size on speed
of germination, seedling elongation and emergence of berseem and
Persian clovers (Irifglium alexagdrinum and I; resupinatum). Seed
Sci. Technol. 8: 323-331.
DasGupta, P. R., and H. M. Austenson. 1973. Analysis of
interrelationships among seedling vigor, field emergence, and yield
in wheat. Agron. J. 65: 417-422.
DeDatta, S. K., W. N. Obcemea, and R. K. Jana. 1972. Protein content
of rice grain as affected by nitrogen fertilizer and some triazines
and substituted ureas. Agron. J. 64: 785-788.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
38
Demirlicakmak, A., M. L. Kaufman, and L. P. V. Johnson. 1963. The
influence of seed size and seedling rate on yield and yield
components of barley. Can. J. Plant Sci. 43: 330-337.
Edwards, C. J. Jr., and E. E. Hartwig. 1971. Effect of seed size
upon rate of germination in soybeans. Agron. J. 63: 429-430.
Evans, L. E., and G. M. Bhatt. 1977. Influence of seed size, protien
content and cultivar on early seedling vigor in wheat. Can. J. Plant
Sci. 57: 929-935.
Freyman, S. 1978. Influence of duration of growth, seed size, and
seedling depth on cold hardiness of two hardy winter wheat
cultivars. Can. J. Plant Sci. 58: 917-921.
GarciaDelMoral, L. F., J. M. Ramos, and L. Recalde. 1985.
Relationships between vegatative growth, grain yield and grain
protein content in six winter barley cultivars. Can. J. Plant Sci.
65: 523-532.
Gardener, J. 1980. The effect of seed size and density on field
emergence and yield of pearl millet [Pennisetum americanum (L.) K.
Schum]. M. S. thesis. Agronomy Dept., Kansas State University,
Manhattan, KS.
Hobbs, J. A. 1953. The effect of spring nitrogen fertilization on
plant characteristics of winter wheat. Soil Sci. Soc. Amer. Proc.
17: 39-42.
Hopper, N. W., J. R. Overholt, and J. R. Martin. 1979. Effect of
cultivar, temperature and seed size on the germination and emergence
of soya beans (glycine mg; (L.) Merr.) Ann. Bot. 44: 301-308.
Hoy, D. J., and E. E. Gamble. 1985. The effects of seed size and
density on germination and vigor in soybean (Glycine mg; (L.)
Merr.). Can. J. Plant Sci. 65: 1-8.
Hoy, D. J., and E. E. Gamble. 1987. Field performance in soybean
with seeds of differing size and density. Crop Sci. 27: 121-126.
Hucklesby, D. P., C. M. Brown, S. E. Howell, and R. H. Hageman.
1971. Late spring applications of nitrogen for efficient
utilization and enhanced production of grain and grain protein of
wheat. Agron. J. 64: 274-276.
Hunter, S. A., C. J. Gerard, H. M. Waddoups, W. E. Hall, H. E.
Cushman, and L. A. Alban. 1958. The effect of nitrogen fertilizers
on the realtionship between increases in yield and protein content
of pastry-type wheats. Agron. J. 50: 311-314.
Johnson, J. R., C. C. Baskin, and J. C. Delouche. 1973. Relation of
bulk density of acid delinted cottonseed to field performance. Proc.
Assoc. of Off. Seed Anal. 63: 63-66.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
39
Kaufmann, M. L., and A. A. Cuttard. 1967. The effect of seed size on
early plant development in barley. Can. J. Plant Sci. 47: 73-78.
Kaufmann, M. L., and A. D. McFadden. 1963. The influence of seed
size on results of barley yield trials. Can. J. Plant Sci. 43: 51-
58.
Lawan, M., F. L. Barnett, B. Khaleeq, and R. L. Vanderlip. 1985.
Seed density and seed size of pearl millet as related to field
emergence and several seed and seedling traits. Agron. J. 77: 567-
571
Lopez, A., and D. F. Grabe. 1973. Effect of protein content on seed
performance in wheat (Tritigum aestigum L.). Proc. Assoc. Off. Seed
Anal. 63: 106-116.
Lowe, L. B., G. S. Ayers, and S. K. Ries. 1972. Relationship of seed
protein and amino acid composition to seedling vigor and yield of
wheat. Agron. J. 64: 608-611.
Lowe, L. B., and S. K. Ries. 1972. Effects of environment on the
relation between seed protein and seedling vigor in wheat. Can. J.
Plant Sci. 52: 157-164.
Lowe, L. B., and S. K. Ries. 1973. Endosperm protein of wheat seed
as a determinant of seedling growth. Plant Physiol. 51: 57-60.
McDaniel, R. G. 1969. Relationships of seed weight, seedling vigor
and mitochondrail metabolism in barley. Crop Sci. 9: 823-827.
Maguire, J. D. 1962. Speed of germination -aid in selection and
evaluation for seedling emergence and vigor. Crop Sci. 2: 176-177.
Miezan, K., E. G. Heyne, and K. F. Finney. 1977. Genetic and
environmental effects on the grain protein content in wheat. Crop
Sci. 17: 591-593.
Muchena, S. C., and C. O. Grogan. 1977. Effects of seed size on
germination of corn (Zea mars) under simulated water stress
conditions. Can. J. Plant Sci. 57: 921-923.
Ries, S. K., C. J. Schweizer, and H. Chmiel. 1968. The increase in
protein content and yield of simazine-treated crops in Michigan and
Costa Rica. BioScience. 18: 205-208.
Ries, S. K., and E. H. Everson. 1973. Protein content and seed size
relationships with seedling vigor of wheat cultivars. Agron. J. 65:
884-886.
Ries, S. K., G. Ayers, V. Wert, and E. H. Everson. 1976. Variation
in protein, size and seedling vigor with position of seed in heads
of winter wheat cultivars. Can. J. Plant Sci. 56: 823-827.
Schweizer, C. J., and S. K. Ries. 1969. Protein content of seed:
41.
42.
43.
45.
46.
47.
48.
49.
50.
40
Increase improves growth and yield. Science. 165: 73-75.
Singh, A. R., and V. G. Makne. 1985. Correlation studies on seed
viability and seedling vigour in relation to seed size in sorghum
(Sorghum bigglgr). Seed Sci. Technol. 13: 139-142.
Steel, R. G. D., and Torrie. 1980. Bringiples and procedures of
ggggisgiggg Second Ed. McGraw-Hill Book Co.
Stickler, F. C., and C. E. Wasson. 1963. Emergence and seedling
vigor of birdsfoot trefoil as affected by planting depth, seed size
and variety. Agron. J. 55: 78.
Suh, H. W., A. J. Casady, and R. L. Vanderlip. 1974. Influence of
sorghum seed weight on the performance of the resulting crop. Crop
Sci. 14: 835-836.
Sung, T. Y., and J. C. Delouche. 1962. Relation of specific gravity
to vigor and viability in rice seed. Proc. Assoc. Off. Seed Anal.
52: 162-165.
Terman, G. L., R. E. Ramig, A. F. Drier, and R. A. Olson. 1969.
Yield-protein relationship in wheat grain, as affected by nitrogen
and water. Agron J. 61: 755-759.
Vergara, B. S., M. Miller, and E. Avelino. 1970. Effect of simazine
on protein content of rice grain (erga sagiva L. ). Agron. J. 62:
269-272.
VOOd. D. W., P. C. Longden, and R. K. Scott. 1977. Seed size
variation, its extent, source and significance in field crops. Seed
Sci. Technol. 5: 337-352.
Wu, K. Y., and C. E. McDonald. 1976. Effect of nitrogen fertilizer
on nitrogen fractions of wheat and flour. Cereal Chem. 53: 242-249.
Yamazaki, W. T., and L. W. Briggle. 1969. Components of test weight
in soft wheat. Crop Sci. 9: 457-459.
and
cha
vig
con
dec
tes
agj
abi
am
CHAPTER II
THE EFFECT OF SEED DENSITY, SIZE AND PROTEIN CONTENT ON VIGOR
AND VIGOR TESTING OF TWO WHEAT VARIETIES
ABSTRACT
A two year experiment was conducted to evaluate several vigor tests
and their ability to predict stand establishment and field performance
of winter wheat (Triticum aestivum L.). Different seed physical
characters were also evaluated as to their effect on seed vigor. The
Vigor tests included the accelerated eging test, the cold test, the
conductivity index, the speed of germination test, the glutamic acid
decarboxylase activity test, the ATP test, and the standard germination
test. Tests that involved physiological stress such as the accelerated
aging test anui cold test were better than non-stress tests ixi their
ability to predict yield. They were also better in their reproducibility
and ability to differentiate between performance of different classes of
size, density and protein content. Biochemical tests such as the
glutamic acid decarboxylase activity and the ATP test were also good
predictors of yield but were not sensitive enough to detect differences
betweerz all seed. classes. The conductivity index: and speed of
germination index tests were the least sensitive in measuring seed vigor
and were not correlated with seed yield. No test was found to be
consistently correlated to either emergence rate index or total
emergence.
41
42
REVIEW OF LITERATURE
1. Seed vigor and vigor testing
The inability of the germination test to consistently predict field
emergence has encouraged interest in the development of seed vigor and
vigor testing (61). McDonald (50) stated that environment, inheritance,
mechanical injury and deterioration during storage were the four factors
affecting seed vigor, while Heydecker (36) listed physiological,
cytological, pathological and mechanical factors as causes of vigor
loss. Abdul-Baki (1) reported that in order to maintain seed vigor,
dehydration and subsequent rehydration of the organelle/membrane system
during seed development, maturation and germination must proceed in an
orderly manner. Ching (17) divided the germination process into three
distinct yet overlapping phases, each of which has a direct bearing on
seed vigor: first, reactivation of pre-existing systems; second,
synthesis of enzymes and organelles for metabolism of reserves; and
third, synthesis of new cellular components.
Isely (38) classified vigor tests as direct, based on the
interaction between the seed and the environment, and indirect, based on
some physiological characteristic of the seed. Woodstock (70) remarked
that both approaches were similar in that both measured a physiological
reaponse involving germination or seedling growth and both involved
seed-environment interaction, but differed in the severity of the
environment. McDonald (50) divided vigor tests into three categories;
physical tests that measured seed characters such as size and weight,
physiological tests that utilized some parameter of germination or
growth, and biochemical tests that monitored chemical reactions involved
43
in cellular maintenance. Grabe (32) stated that tests related to
germination are usually best for estimating stand establishment,
biochemical tests are more adapted to measuring subtle differences
affecting storability and yield, and growth tests are better indicators
of uniformity of crop growth. The nature of vigor testing and the many
tests available have prompted many workers to suggest the use of a
combination of vigor tests (4, 32, 36, 23). Abdul-Baki and Anderson (4)
proposed the use of a multiple criteria approach for evaluating soybean
seed ‘vigor by measuring parameters such as 02 uptake, leaching of
metabolites, C02 production, and uptake of labeled glucose or leucine.
2. Glutamic acid decarboxylase activity (GADA)
The abundance of glutamic acid in seed proteins and its important
role in seed metabolism at the initial stages of seed germination have
prompted many investigators to link the activity of glutamic acid
decarboxylase, an enzyme that decarboxylates L-glutamic acid to produce
S-aminobutyric acid and C02, to seed viability and vigor (2).
Galleschi et a1. (28) found that glutamic acid decarboxylase
activity (GADA) was very low in embryos of milky-ripe wheat seed,
increased significantly until the dough-ripe stage, then remained
constant up to the waxy-ripe stage. During the first three hours of
germination GADA remained constant, reached a maximum at six hours and
then quickly declined (24, 29). Linko and Milner (44) found that
moisture levels as low as 18 percent activated enzyme systems in wheat
seeds and triggered the decarboxylation of glutamic acid.
Galleschi et al. (29) found no GAD activity in the endosperm of
durum wheat (Triticum durum L.) and much higher levels in the embryo
axis than in the scutellum. When Lamkin et al. (42) divided barley
th‘
en:
mi,
Sh(
cor
rel]
——7—
44
seeds into germ (embryo) and endosperm segments, almost 80 percent of
the enzyme activity was in the germ, and Inatomi and Slaughter (37)
found activity only in the embryo and growing parts of the barley plant.
Determinations of GADA in the cotyledons and embryonic axes of bean
seeds indicated that the cotyledons contributed 87 percent of the
activity in viable seeds and 91 percent in deteriorated seeds (39).
Cheng et al. (15) showed that the instantaneous evolution of C02
upon wetting of the wheat embryo is due to enzymatic decarboxylation of
free glutamic acid. In another study, it was found that wetting of wheat
embryos resulted. in immediate activation of CAD, releasing CO2 and
increasing the levels of free K-amino butyric acid (45). Lamkin et al.
(42) found that damaged barley kernels had an enzyme activity no higher
than 14 percent of that of healthy barley kernels and germination at
four' and seven. days was highly correlated yï¬xflx GADA†Grabe (33)
concluded that GADA was a more sensitive index of storability of corn
than either the germination or cold tests and recommended the use of the
GADA test to forecast future losses in viability. However, he believed
that use of GADA to predict field emergence would be impractical since
early stages of deterioration might not affect stand establishment in
the field (31). Linko and Sogn (43) observed no correlation between GADA
and viability if! freshly harvested wheat, but suggested that the test
might provide a good indicator of the storability since it detected germ
damaged seeds.
Doubt has been cast on the usefulness of GADA as a vigor test by
showing frequent inconsistencies in test results. One study showed
considerable fluctuations in GADA even though germination of bean seed
remained high (39). Furthermore, in some varieties, GADA remained high
45
in seeds which had lost their viability. In another study, GADA was
better correlated with variety than with age of soybean seeds and showed
little or no correlation with germination at four and eight days or with
seedling growth rate (25).
3. Accelerated aging
The concept behind the accelerated aging test is that seeds loose
viability rapidly when stored under high temperature and relative
humidity. Therefore, differences in viability and vigor are more
apparent. Seeds which deteriorate rapidly tend to perform poorly in
long-term open storage and achieve a less than satisfactory field
performance (63). The accelerated aging test is conducted by exposing
seeds to temperatures in excess of 40° C and 95-100 percent relative
humidity for varying different lengths of time and then assessing their
viability using the standard germination test (8).
Tekrony and Egli (68) concluded that the accelerated aging test was
a good indicator of soybean performance only under adverse conditions,
and added that the highest prediction accuracy occurred when accelerated
aging, 4-day germination and standard germination test results were
combined into a single vigor index. Studying rice storage under
different temperatures and relative humidities, Azizul-Islam et a1. (9)
found that the accelerated aging test was superior to the standard
germination test but not as sensitive as the GADA test in monitoring
seed deterioration. In interpreting the GADA test results, Delouche and
Baskin (22) placed more emphasis on seed survival than on the quality or
condition of the seedling. Since the seeds being tested had deteriorated
considerably, the definitions of normal and abnormal seedlings were
relaxed. Thus, they considered sorghum as germinable if it produced a
46
shoot and root, regardless of their size or presence of surface molds.
Priestley (62) found no reason to assume that the physiology of aging of
dry seeds resembled that occurring at higher moisture levels. Abdul-
Baki and Anderson (3), studying the leaching of sugar from artificially
aged barley seed, also concluded that accelerated aging was not similar
to normal aging, even though both resulted in loss of germination. Tao
(67) demonstrated that location in the aging chamber affected moisture
content, fungal growth and seedling vigor following the aging treatment.
McDonald (51) concluded that germination results derived from a wire-
mesh basket system may be biased by seed position, with the outer seeds
deteriorating more rapidly than the inner ones, and that sample size
also influenced final moisture content and loss of germination.
4. Electrical conductivity test
The electrical conductivity test has been proposed as a vigor test
on the assumption that low vigor seeds generally possess poor membrane
structure and leaky cells, resulting in greater loss of electrolytes
such as sugars and amino and organic acids. Matthews and Bradnock (58)
observed that pea and bean seeds which exuded electrolytes readily had
poor field emergence despite good laboratory germination. In two other
studies on pea seeds, Matthews and coworkers found that the conductivity
of the soak water was negatively correlated with the predisposition of
the seeds to pre-emergence mortality. Since conductivity was shown to be
highly indicative of soluble carbohydrate content of the water, they
concluded that the glucose concentration of the soak water could be used
as an indicator of field emergence (56, 57). In another study, the
conductivity of sunflower seed leachate was better correlated with
seedling emergence under all environments tested than were germination
af1
vi;
ex1
IIIIIIIIIIIIII7______________________________——__4
47
after accelerated aging, seedling growth rates, cold test, seedling
vigor, cool germination or standard germination (6). Yaklich et al. (71)
expressed conductivity' on a per-gram-of—seed basis and improved its
correlation with field emergence of soybeans. Similarly, McKersie et al.
(53) observed that conductivity per 100 seeds of birdsfoot trefoil was
only correlated with laboratory germination, while conductivity per gram
of seed was correlated with percent germination, seedling length and
field emergence.
Abdul-Saki (l) cautioned that in interpreting results of the
conductivity test, types of injuries such as bruises, cracks and tissue
breakage that cause increased leakage should be distinguished from the
increased leaching of solutes due to loss of membrane function. On the
other hand Gill and Delouche (30) concluded that the electrical
conductivity test could not be used as a meaningful index of seed
deterioration because results varied with seedcoat injuries, temperature
changes, and interval and intensity of stirring.
The ASA-610 has been developed to measure conductance in 100
individual cells for concurrent evaluation of 100 individual seeds.
McDonald and Wilson (52) reported that although the instrument measures
soybean seed leachate, seed size and initial seed moisture content both
influenced readings, while seed treatment did not. They also reported
that conductance values varied with fluid level per cell, solution
temperature and soak temperature. The ASA-610 could accurately predict
high or low quality lots but not those in between. In another study
using the same machine, Hepburn et a1. (35) observed considerable
overlap in conductivity readings between 'viable and non-viable seed
lots. Moreover, they doubted the ability of a single partition value to
48
predict laboratory germination, since a larger seeded cultivar of pea
gave higher conductivity values than a smaller seeded one.
5. ATP
Lunn and Mason (48) reported that differences in ATP levels could
predict seed vigor in cauliflower (ï¬gaggiga glegagga L.) despite similar
germination levels. ATP levels of rape and ryegrass (Lolium multiflorum
L.) seeds were positively correlated with seed vigor as measured by seed
weight, haday seedling length, dry weight and fresh weight (17). In
another study, significant correlations were recorded between ATP
content of imbibed seed and seed weight, seedling size and dry weight of
clover, ryegrass and rape (16). Ching and Danielson (18) reported that
ATP contents were highly and significantly correlated with seedling
length in several lettuce cultivars.
Lunn and Madsen (QB) found that while ADP and AMP contents of
germinating rape, cauliflower and sugarbeet (Beta vulgaris L.) seed did
not change during 16 weeks of aging, ATP levels fell long before loss of
viability could be measured by the standard germination test. In another
aging study, ATP content of lettuce seeds, as well as germination
percent and seedling size, were reduced. by aging (18). Many other
studies have confirmed the relationship between deterioration and ATP
levels. Anderson (5) showed that levels of ATP in soybean seeds varied
inversely with degree of deterioration. They related low ATP levels to
lowered rates of RNA and protein synthesis. Studies on other crops have
produced similar results (48, 69). However, some others have cautioned
against using ATP as a vigor test. Yaklich et a1. (71) concluded that
ATP levels in the embryonic axis of soybeans did not always correlate
with field emergence, and therefore cannot be used as a measure of seed
I a.“
IIIIIIIIIIIIIIIIIIIIIIIlIllIll-I::————————————————*
49
vigor. Similarly, ATP content was not well correlated with seedling
growth in several vegetable species (66).
6. Cold test
DasGupta and Austenson (21) found a good correlation between
standard germination, cold germination, 02 uptake and yield of wheat.
In another study, field stand and grain yield of wheat were positively
correlated with standard germination and cold and modified cold
germination percentages (20). Byrd and Delouche (13) observed that the
cold test was better in predicting the storage potential of soybeans
than were the standard germination, first count germination, radicle-
hypocotyl length, respiration rate or dehydrogenase activity. Johnson
and ‘Wax (40) concluded that the cold test was the most effective
germination test for identifying soybean seed lots that performed well
in the field. Cold test results were significantly correlated with
field emergence and were better than standard germination test results
in detecting problem seed lots.
In studying the relationship of vigor tests to cotton seedling
establishment, Bishnoi and. Delouche (10) concluded. that vigor tests
which simulated adverse field conditions, such as the cold test, could
accurately predict field establishment and detect relative deterioration
levels among seed lots. Another study on wheat indicated that direct
stress vigor tests such as the cold test were more closely correlated
With field emergence than standard germination onLy when soil
conditions were unfavorable (34). Mahdi et al. (55) reported that an
index in which the numerator is the cold test results and the
denominator is the standard germination results was a better indicator
of the vigor of cottonseed under adverse field conditions, especially in
50
early sowing, than either test alone.
Many investigators have questioned the necessity of using actual
soil in cold testing (14, 26, A6). Loeffler et a1. (46) compared two
cold test methods, with or without the use of soil, and concluded that
the no-soil test was better because of its sensitivity, reproducibility
and simplicity. Burris and Navratil (12) showed that sterile cold test
results correlated as well with field emergence as non sterile test
results.
7. Speed of germination
Since the rate of seedling growth can reflect the level of
metabolic activity during germination, many attempts to develop tests
and formulas that reflect seed vigor have emphasized speed or rate of
germination.
Various methods and techniques have been used to measure speed of
germination and correlate it with seed vigor. Maguire’s method (SA), in
which speed and percent germination are combined to form a single vigor
index has probably been most commonly used. Using Maguire’s test, Lopez
and Grabe (47) found that speed of germination was positively correlated
with increased protein content in wheat which in turn was positively
correlated with better field performance. Dalianis (19) reported that
cotton seeds which had a higher speed of germination in the laboratory
emerged faster and more completely in soil than slow germinating seeds.
Mckersie et al. (53) found that while differences in speed of
germination of birdsfoot trefoil could reflect differences in age or
physiological condition, physical differences did not produce varying
speeds of germination. Another way of measuring speed of germination
is by recording the number of germinated seeds at different days of the
r4
3%
ge
se
Oll
fi
fi
pr
th
on
SE
in
de
’i—
51
germination test. Mian and Coffey (60) found that the 3-day germination
count for rice was a reliable seed vigor test that gave better results
than the cold test. For corn, the 80-hour-count germination results
agreed well with the results of the cold test, yet were quicker to
obtain (59). Kulik and Yaklich (41) reported that daily germination
counts for four days were significantly correlated with soybean field
emergence, and Tekrony and Egli (68) found similar correlations with a
A-day germination count along with standard germination and accelerated
aging. Finch-Savage (27), using the slope test, found that rapid
germination rates within seed lots were associated with improved
seedling size and uniformity in cauliflower, leek (Allium pprrum L.) and
onion (5, gepg L.). He concluded that the selection and use of faster
germinating seeds would give larger and more uniform seedlings in the
field.
The purpose of this experiment was to relate several vigor tests to
field. emergence and. yield. data, and t3) establish the best test in
predicting field performance of wheat. Another purpose was to examine
the effects of different seed densities, seed sizes and protein contents
on seed vigor enui to correlate vigor differences with differences in
seed physical characters. Another purpose was to examine whether
individual vigor tests were equally capable of detecting differences in
density, size and protein content of different seed lots.
52
MATERIALS AND METHODS
1. First Year Experiment
Two winter wheat varieties, Augusta, a soft white variety, and
Hillsdale, a soft red variety, were divided into an unselected control,
three size classes, two density classes and two protein levels
categories according Ix) the procedures described if} chapter 1†Seeds
were then stored at 5° C and 35 percent relative humidity until use.
A. Standard Germination Test
Four lOO-seed replications from all treatments were germinated
between moistened blotter paper at 200 C for 7 days in the light, then
classified into normal, abnormal and dead seeds according to the "Rules
for Testing Seeds" handbook by the A.O.S.A (7). Only the percent normal
seedlings was recorded for this study.
B. Accelerated Aging Test
The "wire-mesh" tray procedure developed by McDonald (51) was used
for this test. For each treatment an 11 x 11 x 3.5 cm copper wire mesh
tray held 2 cm above the bottom of a plastic box was used. Forty
milliliters of distilled water were added to each box and 250 seeds were
placed in a single layer on the wire tray. The boxes were sealed with
tape and incubated at A10 C and near 100 percent relative humidity for 3
days. Seeds were then removed, allowed to dry at room temperature for
three days, then germinated using the same standard germination test
procedure described above.
r——ï¬
53
C. Cold Test
Soil from the same location as our subsequent field tests was
collected, dried at room temperature, ground to pass through a 20 mesh
screen, and its water holding capacity determined using the procedures
described in the A.O.S.A "Seed Vigor Testing Handbook†(8). Four
replicates of 100 seeds each from all treatments were used for the cold
test. A 2 cm thick layer of soil was placed on the bottom of a 29.5 x 16
x 8.5 cm plastic box and 100 seeds were evenly spread on top of that
layer and covered by another 2 cm of soil. Enough water was added to
bring the medium to 70 percent of its water holding capacity. The boxes
were covered, incubated at 5° C for 4 days, then transferred to 200 C
for 7 days. Seedlings that emerged above the soil level were counted and
results were reported as percent germination.
D. Speed of Germination Test
Four replicates of 100 seeds each were germinated for 7 days
according to the standard germination test procedures listed above. Each
day the number of newly germinated normal seedlings was counted and a
speed of germination index calculated using Maguire’s method (54)
Speed of germination index -
No. of normal seddlings No. of normal seedlings
+...+
No. of days to first count No. of days to last count
E. Electrical Conductivity Index
The ASA-610 conductivity analyzer that was used for this test is an
instrument that provides simultaneous conductivity measurements from 100
individual cells containing one seed each. When conductivity exceeded
130 microamps, the seed was considered to be nonviable. Four replicates
T————’—
of 100 seeds each were used per treatment. Each seed was placed in a
54
cell containing deionized water, incubated at 22° C for 22 hours, and
then the conductance was read. The conductivity index was calculated by
partitioning the microamp reading from 0 In) 130 into 5 ndcroamp
intervals and assigning a number from 1 to 26 to each interval. The
number of seeds in each interval was then divided by the assigned number
and results from all intervals were added to obtain a conductivity
index.
F. Glutamic Acid Decarboxylase Activity (GADA)
GADA was measured using 5.0 gm of freshly ground seeds and a
substrate solution containing 35.0 ml of 0.1 M L-glutamic acid in 0.50 M
sodium phosphate buffer, pH 5.2. The mixture was placed in a st0ppered
250 ml Erlenmeyer flask and shaken in a water bath for 8 minutes at 30°
C. Five cc of air were then removed with a syringe through a septum and
the concentration (ppm) of C02 determined 'by aux ADC model 225-MK3
Nondispersive Infrared Gas Analyzer. Results were reported as ppm C02
per gram of seed. A flask containing all the above chemicals but no
ground seed was added to every run as a blank control, and the results
in ppm C02 of that blank later subtracted from the concentration of C02
from each treatment.
G. ATP Test
ATP level was determined by the luciferen-luciferase method
suggested by St. John (65). Four replications of 20 seeds each were
imbibed at 20° C then dropped in 20 m1 of boiling distilled water and
extracted for 10 minutes. The extract was then cooled in an ice bath and
an aliquot was diluted 2-fold with a HEPES buffer (25mM N-2-
hyc‘
MgE
int
55
hydroxyethylpiperazine-N-2-ethanesulfonic acid, pH 7.5) containing 25mM
MgSOa.7H20. ATP (0.05uM), was added to another diluted aliquot as an
internal standard. The light emission was determined in 0.2 ml of
diluted extract by an Aminco Chem-Glow photometer after adding 0.4 ml of
a reconstituted firefly lantern extract (Sigma Chem.). The ATP
concentration in the extract was then calculated according to the
formula of St. John (65).
2. Second Year Experiment
Treatments and procedures were the same as in the first year except
for the addition of 3 treatments. The extra treatments were medium
protein, unselected minus light and. unselected minus small obtained
using the procedures outlined in chapter 1.
3. Statistical Analysis
Results from all vigor tests were subjected to analysis of variance
following the procedures outlined by Steele and Torrie (64). All the
vigor tests were analyzed as a completly randomized design with four
replications and two factors: seed characters (treatments) and
varieties. Means were separated using Duncan's multiple range test
(DMRT) at the 5 percent probability level. Correlation analysis was
performed on all test results using the mean values from each test.
gen
tree
sip
les
low
13:
(101'
56
RESULTS
In 1985, the only treatments to differ significantly in standard
germination from the unselected control treatment were the light seed
treatment of Augusta and the small seed treatment of Hillsdale (Table
11). The light seed treatment of both Augusta and Hillsdale was
significantly lower tfluni the heavy treatment, the small significantly
less than the large seed treatment and the low protein significantly
lower than the high protein treatment. In 1986 (Table 11) the heavy,
large and high protein treatments were all significantly higher than the
control for Augusta, while no treatment differed from the control for
Hillsdale. Germination. of the light treatment of Augusta was
significantly lower than the other density classes, and the small was
significantly lower than that of other size classes, but protein content
did not affect percent germination. The small and large seed treatments
were the only ones to differ significantly for Hillsdale.
Ir: both :years, only treatments had. a significant influence on
standard germination, while varieties and variety by treatment
interaction did not (Tables A8 and A9).
‘Following accelerated. aging in 1985 (Table 12), the heavy was
significantly' higher than time light. seed treatment, the large
significantly higher than the medium and small seed treatments, and high
protein significantly higher than the low protein seed treatment.
Although the small Augusta seeds were significantly lower, and the large
and high protein treatments significantly higher than the control, no
treatment was significantly' higher than the control for Hillsdale.
However, light, small, medium and low 'protein seeds were all
Tat
gel
Tn
57
Table 11. Effect of various seed characters on standard
germination of Augusta and Hillsdale in 1985 and 1986.
1985 1986
Treatment Augusta Hillsdale Augusta Hillsdale
Unselected 97abc 97ab 93bc 95abc
Light 94d 95b 91c 95abc
Unselected-Light* - - 96ab 95abc
Heavy' 98ab 98a 98a 96abc
Small 95cd 92c 90c 92c
Medium 97abc 96ab 96ab 94abc
Unselected-Small** - - 96ab 96abc
Large 98ab 98a 98a 98a
Low Protein 96de 95b 95ab 93bc
Medium Protein - - 95ab ' 94abc
High Protein 99a 98a 97a 97ab
* - unselected minus light, ** - unselected minus small.
Means followed by the same letter in each column are not
significantly different at the 5 % probability level
according to DMRT.
Table 12. Effects of various seed characters on accelerated
aging test results of Augusta and Hillsdale in 1985 and
1986.
1985 1986
Treatment Augusta Hillsdale Augusta Hillsdale
Unselected 77bc 73a 75bc 67d
Light 70cd 58b 61d 49f
Unselected-Light* - - 77b 76bc
Heavy 84ab 83a 80ab 83a
Small 65d 35c 56d 65f
Medium 68cd 44c 57d 56e
Unselected-Small** ~ - 69c 67d
Large 90a 82a 79ab 83a
Low Protein 75de 61b 70c 59c
Medium Protein - - 75bc 73cd
High Protein 91a 83a 85a 82ab
* - unselected minus light, ** - unselected minus small.
Means followed by the same letter in each column are not
significantly different at the 5 % probability level
according to DMRT.
signif
same I
signii
small
unsele
not 5
heavy
small
treat
varie
pmi
did 1
heavy
seed:
Sign
larg
con
tre
whi
58
significantly lower than the Hillsdale control. In 1986 (Table 12), the
same differences were observed. Only the high protein treatment was
significantly higher than the control for Augusta, while the light,
small and medium treatments were significantly lower. Only the
unselected minus small and medium protein treatments of Hillsdale did
not significantly differ from the control. Unselected minus light,
heavy, large, and high protein were all significantly higher and light,
small, medium, and low protein significantly lower than the control
treatment. In both years (Tables A8 and A9), treatment, variety and
variety by treatment interaction had significant influence on
germination following the accelerated aging treatment.
In the cold germination test in 1985 (Table 13), protein content
did not affect germination in either Augusta or Hillsdale. However,
heavy seeds germinated better than light ones, and large and medium
seeds better than small ones. Light and small seed treatments were
significantly lower than the control of both varieties, while heavy,
large and high protein treatments were significantly higher than the
controls. In 1986 (Table 13), as in 1985, light and small seed
treatments in both varieties were significantly lower than the control
While heavy, large and high protein treatments were significantly
higher. Hoeever, in Hillsdale, the unselected minus small and low
protein seeds were significantly better than the control. For both years
treatment had a significant influence (Tables A8 and A9) on cold
germination while variety and the interaction of variety and treatment
did not.
In 1985, the light and small seed treatments of Augusta had a
significantly higher speed of germination (Table 14) than heavy, and
Tat
gel
ant
Trl
59
Table 13. Effects of various seed characters on cold
germination test results for Augusta and Hillsdale in 1985
and 1986. Percent germination.
1985 1986
Treatment Augusta Hillsdale Augusta Hillsdale
Unselected 63bc 63c éade 62d
Light 52d 52d 48f a3e
Unselected-Light* - - 72bcd 76bc
Heavy 80a 86a 78abc 85a
Small Ale 43d 64f 39e
Medium 53cd 7lbc 60e 67cd
Unselected-Small** - 69cd 70cd
Large 80a 84a 82a 80ab
Low Protein 72ab 73bc 69cd 76abc
Medium Protein - - 73bcd 70cd
High Protein 82a 81ab 80ab 80ab
* - unselected minus light, ** - unselected minus small.
Means followed by the same letter in each column are not
significantly different at the 5 % probability level
according to DMRT.
Table 14. Effects of various seed characters on speed of 1
germination test results for Augusta and Hillsdale in 1985 ‘
and 1986.
1985 1986
Treatment Augusta Hillsdale Augusta Hillsdale
Unselected 23.99ef 29 OOab 26.41c 28 l6abc
Light 30.610 30.05ab 30.69abc 29.72abc
Unselected-Light* - - 28.54abc 27.29bc
Heavy 25.38de 23.07c 26.45c 26.l3c
Small 39.47a 29.23ab 32.47a 31.86a
Medium 35.09b 25.55c 29.8labc 30.26ab
Unselected-Small** - - 28.48abc 27 06bc
Large 21.51f 23.02c , 26.45c 26.97bc
Low Protein 27.52cd 26.54bc 28 27bc 27.85abc
Medium Protein - - 28.88abc 29.02abc
High Protein 35.40b 30.51a 31.51ab 31.83a
* - unselected minus light, ** - unselected minus small.
Means followed by the same letter in each column are not
Significantly different at the 5 % probability level
according to DMRT.
medi
lows
bei‘
tha
Hil
for
gel
trt
60
medium and large seeds, respectively, while low protein seeds had a
lower germination speed than high protein seeds. In 1986 (Table 14),
most classes did not differ in germination speed, the only exception
being small seeds which had a significantly higher germination speed
than large seeds. Differences between the various seed categories of
Hillsdale were the same as those for Augusta for 1985 and 1986, except
for the small seed treatment in 1986 which had a significantly higher
germination speed than the unselected minus small and large seed
treatments. In 1985 treatment, ‘varietyy and. their interaction. had ea
significant effect on our results (Tables A8 and A9), while in 1986 only
the treatments had a significant effect on test results.
Only the size classes affected conductivity index results in 1985
(Table 15) for either variety; the small seeds having significantly
higher values than medium and large seeds. In Augusta, small seed was
the only treatment differing significantly from the control, while the
small treatment of Hillsdale was significantly higher and the large
treatment was significantly lower than the control. In 1986 (Table 15),
size again was the only category affecting conductivity readings. Small
Auguata seeds had a laigher index; than. all other size classes. For
Hillsdale, conductivity of the small and unselected minus small seeds
was significantly higher than for large seeds. Only the large seed
treatment of either variety had a significantly lower index than the
control. Treatment effects were significant in affecting the
conductivity index results in both years (Tables A10 and All).
In 1985 (Table 16), the small and low protein seeds of both
varieties contained significantly less ATP than the control, while the
large and high protein seeds were significantly higher than the control.
Tal
im
61
Table 15. Effect of various seed characters on conductivity
index test results for Augusta and Hillsdale in 1985 and 1986.
1985 1986
Treatment Augusta Hillsdale Augusta Hillsdale
Unselected 58.64bc 61.17b 65.05ab 72.77ab
Light 57.97bc 61.62b 51.28bc 50 62bc
Unselected-Light* - - 48.83bc 57.79abc
Heavy 59.06bc 59.84b 68.77ab 54.56bc
Small 80.39a 82.66a 79.93a 79.403
Medium 50.87bc 56.68bc 51.66bc 58.01abc
Unselected-Small** - - 45.38bc 65.27ab
Large 41.94c 41.16c 39.14c 36.52c
Low Protein 60.36b 61.36b 58.56abc 54.27bc
Medium Protein - - 52.97bc 59 76abc
High Protein 55.01bc 58.58b 59 70abc 57.58abc
* - unselected minus light, ** - unselected minus small.
Means followed by the same letter in each column are not
significantly different at the 5 % probability level
according to DMRT.
Table 16. ffect of several seed characters on ATP test
results (10' M/Lit) for Augusta and Hillsdale in 1985
and 1986.
1985 1986
Treatment Augusta Hillsdale Augusta Hillsdale
Unselected 1.74cd 2 04c l.7lbc 1.83bc
Light 1 78cd 2.04c l.74bc 1.81bc
Unselected-Light* - - l 86bC 2 09b
Heavy 2.12bc 2.24bc 2 02b 2.00b
Small 0.55e 0.57d 0.63d 0.54d
Medium 1.61d 2.13bc 1.59c 1.62c
Unselected-Small** - - l 84bc 2 55a
Large 2.39ab 2.57ab 2.76s 2 78a
Low Protein 0.63e 0.72d 0.51d 0.60d
.Medium Protein - - 1.67bc l.77bc
High Protein 2.78a 2.89a 2.96a 2.78a
* - unselected minus light, ** - unselected minus small.
Means followed by the same letter in each column are not
significantly different at the 5 % probability level
according to DMRT.
While *
variet:
were Vc
small
only t
variet
test I
gluten
prote
highe
seed,
GADA
varie
1986
GADA
Sig;
Sig
mir
62
While the ATP content increased with size and protein content in both
varieties, the density' classes did run: differ’ significantly; Results
were very similar in 1986 (Table 16). In addition, the unselected minus
small seed of Hillsdale contained more ATP than the control. As in 1985,
only the density classes showed no significant differences for either
variety. Both treatment and variety had a significant influence on ATP
test results in both 1985 and 1986 (Tables A10 and All).
In 1985, the only Augusta treatments to differ from the control in
glutamic acid decarboxylase activity (GADA) were the small and high
protein treatments which had significantly lower and significantly
higher activity, respectively (Table 17). For Hillsdale, only the small
seed, with low GADA, differed significantly from the control. While the
GADA appeared tx> increase with size or protein content in both
varieties, no differences were significant among the density classes. In
1986 (Table 17), the small seeds of both varieties again contained lower
GADA than other treatments. Large and high protein Augusta seeds were
significantly higher than the control. Small seeds of Augusta contained
significantly less enzyme activity than that of nmdium or Lumelected
minus small seeds, and all three were significantly lower than the large
seeds. Low protein seeds did not differ from seeds with medium protein
content, but were significantly lower than high protein seeds in GADA
activity. No differences were recorded among the different density
classes. The density and. proteirx categories of Hillsdale paralleled
those of Augusta with respect to enzyme activity, but the size classes
differed in that the small seeds contained significantly less GADA than
all other classes. Both treatment and variety significantly influenced
GADA l!
in 1984
63
GADA in 1985 (Table A10), but only treatment had a significant influence
in 1986 (Table All).
Tat
64
Table 17. Effect of various seed characters on glutamic acid
decarboxylase activity test results (ppm COz/gm of seed) for
Augusta and Hillsdale in 1985 and 1986.
1985 1986
Treatment Augusta Hillsdale Augusta Hillsdale
Unselected 46Gb 463ab hSOcd 462ab
Light 470ab 475a 483abc 470ab
Unselected-Light* - - 482abc 475ab
Heavy SOlab 478a 480abc 474ab
Small 387c 371c 383e 397C
Medium a79ab 416bc 457bcd 46lab
Unselected-Small** - - ASAbcd 457ab
Large 494ab 488a S2Sa 499a
Low Protein AS7b 419bc 429d a30bc
Medium Protein - - ASSbcd 425bc
High Protein 522a 512a SOAab 490a
* - unselected minus light, ** - unselected minus small.
Means followed by the same letter in each column are not
significantly different at the 5 % probability level
according to DMRT.
seed c
which
inter;
of the
other
imbih
deter
which
exhat
resuf
cont
conc
yie
adV
8C!
le
be
65
DISCUSSION
The accelerated aging test detected significant differences among
seed categories for both varieties in both years, and was the only test
which showed a significant influence of treatments, varieties and their
interaction. Aging had greatest effect on the size classes; germination
of the small and medium seed classes was consistently lower than for all
other treatments. Since size must have a significant effect on degree of
imbibition, these results suggest that an1 important factor in
determining the extent of aging is the rate and speed of imbibition
which, when followed by the activation of metabolic processes, lead to
exhaustion of reserves needed for germination.
Tekrony zuui Egli (68) also reported that accelerated aging test
results were a good indicator of soybean performance under adverse
conditions. Our results show that under both normal and adverse
conditions, the accelerated aging test was significantly correlated with
yield (Tables 18 and 19). However the correlatitwi was higher under
adverse conditions in 1986, especially for Hillsdale. The C.V. of the
accelerated aging test was relatively low and therefore the results were
less variable tï¬uui those for other tests. Although test results have
been reported to vary with seed position within individual aging
chambers (51), such effects were eliminated in our study by using most
of the seed in each container for the germination test.
Like the accelerated aging test, the cold test is a stress test,
and was able to detect differences in germination capacity related to
seed size euui density. However, the cxflxi test rarely detected
Significant differences due to protein content. Since the cold test
conditi
Signif
correl
(Table
standa
seeds,
labor;
not b
also
and E
soil
were
be h
were
neg;
sea
One
66
conditions were similar to those in the field, response should have been
significantly correlated with emergence. However, no significant
correlations between total emergence and cold test results were observed
(Tables 18 and 19). On the other hand, both cold germination and
standard germination were always positively correlated. Thus for wheat
seeds, the cold test is more related to seed germination in the
laboratory than to field emergence. Consequently, the use of soil may
not be an1 important factor in determining test results. Loeffler (46)
also concluded that soil and non-soil cold tests gave the same results,
and Burris and Navratil (12) concluded that the presence or absence of
soil pathogens did not affect the cold test results.
The speed of germination index and the emergence rate index results
were calculated using the same procedure; therefore, the results should
be highly correlated. In Augusta however, the correlation coefficients
were very low. In Hillsdale they were higher (significant in 1985), but
negative. These results 'would incorrectly indicate that low' quality
seeds judged by their field performance are the more vigorous seeds.
One explanation of our results could be that smaller and lighter seeds
reach the critical moisture content for germination more quickly than
large or heavy seeds during imbibition and therefore germinate faster.
When differnces in imbibition time are not a factor, as with seeds of
different protein levels, the speed of germination index indicated that
high protein seeds were more vigorous than low protein seeds. By
contrast, even if smaller and lighter seeds germinate faster in the
field, they should not emerge quickly since their growth rate or dry
matter accumulation is lower than that of large or heavy seeds.
If quality factors other than size and density were considered, the
Table 18.
emer genc‘
/
1985 WC
AA
CG
SG
01
ATP
GADA
EN
ERI
1986 WC
AA
CG
SC
CI
ATP
GAD.
EN
ERI
* - Sig:
H . Si}
W = st.
SG - s
GADA .
ERI - E
67
Table 18. Correlation coefficients between different vigor tests, field
emergence and yield of Augusta, 1985 and 1986.
Yld WC AA CC SC CI ATP GADA EN
1985 NC .78**
AA .91** .79**
CC .96** .75** .94**
SC .59 .59 .52 .59
CI .68 .68 .60 .60 .58
ATP .72* .72* .83** .69 -.32 -.69
GADA .83** .83** .78** .80* -.40 -.8l* .87**
EN .39 .39 .51 .51 .29 -.22 .42 .48
ERI .40 .40 .63 .51 .13 -.37 .68 .59 .89**
1986 WC .89**
AA .93** .71*
CC .96** .92** .89**
SC .45 .54 .50 .58
CI .35 .46 .20 .41 .33
ATP .75** .64* .70* .66* -.23 -.45
GADA .70* .66* .66* .66* -.38 -.66* .89**
EN .30 .34 .31 .31 -.03 -.40 .40 .68*
ERI .34 .33 .38 .36 -.01 -.32 .42 .66* .98**
* - significant at the 5 percent probability level.
** - significant at the 1 percent probability level.
WC - standard germination; AA - accelerated aging; CC — cold test;
5K3 - Speed of germination index; CI - Conductivity Index;
GADA - Glutamic acid decarboxylase activity; EN - Emergence number;
ERI - Emergence rate index.
speed of
However,
factors
growth
incorpox
Thw
between
conduct
seeds (
density
doubted
results
conten1
Theref
asar
repro
corre
inde
Shox
COm
68
speed of germination test might provide an accurate indication of vigor.
However, it was not a reliable indicator of emergence when physical
factors were incorporated in the treatments. Perhaps another measure of
growth such an; dry matter accumulation should be substituted or
incorporated in the evaluation procedure.
The conductivity index test detected significant differences
between size classes but not between density or protein classes. If
conductivity is a true measure of vigor, these results indicate that
seeds of different size vary' in 'vigor"while seeds which differ in
density' or protein. content ck) not. Many researchers, however, have
doubted the the ability of the ASA-610 machine to measure vigor, since
results are affected by many variables, such as initial moisture
content of the seeds, fluid level per cell and seed size (35, 52).
Therefore our conductivity index readings should be regarded primarily
as a reflection of size, rather than vigor.
The conductivity index had the highest coefficient of variability
(C. V.) of any test for both years, indicating high variability and low
reproducibility. Moreover, the conductivity index was not significantly
correlated with yield, and only with emergence number and emergence rate
index for Hillsdale in 1986 (Tables 18 and 19). If the ASA-610 machine
is to be used to compare vigor differences between seed lots, samples
should be sized and brought to the same moisture content, and only
comparisons between similar samples be made.
The two tests of biochemical activity, the ATP and GADA tests, gave
basically the same results. Both were capable of detecting differences
in size and protein content but neither was affected by differences in
seed density for either varieties or years. Both ATP and GADA increased
with sf
correlz
Mo
Likewi
the e1
Accord
and p1
diffel
resuli
corre
seed
appro
less
one
(Tat
Sig
0th
69
with size and protein content and therefore were highly and positively
correlated in all cases (Tables 18 and 19).
Most of the GADA activity occurs in the embryo (29, 37, 39, 42).
Likewise, the ATP activity is associated with mitochondrial activity in
the embryo, especially the first hours of germination (ll, 71).
Accordingly, our results indicate that, while differences in seed size
and protein content reflect differences in both embryo and endosperm,
differences in seed density reflect variations in the endosperm. These
results disagree with the findings of McDaniel (49), who reported high
correlations between ATP content in barley and both seed density and
seed vigor.
The coefficient of variability (C.V.) for ATP content was
approximatly twice that of GADA in both years. Therefore results are
less repeatable and consistent for the ATP test than for the GADA test.
This could be considered as an advantage of the GADA test. However, with
one exception, neither test proved a good indicator of field emergence
(Tables 18 and 19), and while results of the two tests for Augusta were
significantly correlated with yield, in Hillsdale they were not.
Otherwise the two tests appeared to yield similar results.
There are four criteria by which the value of any vigor test must
be judged. The first is its ability to accurately predict field
performance either in terms of emergence or yield. The second is its
reproducibility under different conditions. The third is its ability to
differentiate between seed lots differing in quality or vigor. Finally,
to justify its use, the test in question should be more reliable than
the standard germination test. The latter had the lowest C.V., and
therefore the best reproducibility of all tests we performed; results
Table 19
emergence
/
1985 WC
AA
CG
36
Cl
ATE
GAE
EN
ER]
1986 WC
we.s
ERI .
70
Table 19. Correlation coefficients between different vigor tests, field
emergence and yield of Hillsdale, 1985 and 1986.
Yld WC AA CC SC CI ATP GADA EN
1985 NC .72*
AA .74* .89**
CC .71* .86** .79*
SG -.25 -.4O -.29 -.65
CI -.57 -.81* -.69 -.77* .56
ATP .75* .84* .71* .61 -.15 -.71*
GADA .68 .85** .90** .63 -.O6 -.68 .87**
EN .11 .33 .16 .54 -.15 -.3O .03 .05
ERI .48 .45 .41 .83** -.76* -.62 .23 .21 .57
1986 NC .70*
AA .94** .80**
CC .83** .65* .87**
SC -.47 -.38 -.47 -.56
CI -.31 -.61 -.46 -.52 .40
ATP .57 .92** .74** .52 -.28 -.47
GADA .56 .92** .64* .56 -.34 ‘.67* .84**
EN -.11 .28 .13 .31 -.38 -.60* .38 .41
ERI .04 .40 .28 .43 -.47 -.69* .46 .52 .98**
* - significant at the 5 percent probability level.
** - significant at the 1 percent probability level.
WC - standard germination; AA - accelerated aging; CG - cold test;
SC - Speed. of germination index; CI - Conductivity Index;
GADA - Glutamic acid decarboxylase activity; EN - Emergence number;
ERI - Emergence rate index.
were 5
emerge
l
aging
test
index
not w
best
seed:
vari.
indi
two
refl
con
NOS
71
were significantly and positively correlated with yield but not to field
emergence.
Ranking the tests according tx> these criteria, the accelerated
aging test was the best vigor test, followed by the cold test, the GADA
test and the ATP test. The speed of germination index and conductivity
index tests did not give a consistent or true reflection of seed vigor
nor were they correlated with grain yield for either variety. The two
best tests (accelerated aging and cold germination) both stressed the
seeds prior' to germination†The accelerated test, because of lower
variability was the better of the two.
The difference in quality between different seed characters, as
indicated by differences in field performance and vigor results implies
two things. First, seed characters themselves can be used as a
reflection of vigor. This is especially true for seed size which showed
consistent differences between the small, medbun and large seeds for
most of our tests. Second, the effect of these different seed characters
should be considered and if possible eliminated when studying other
factors that contribute to variation in seed quality and vigor.
Ab
15
do
51
10.
11.
12.
13.
14.
72
References
. Abdul-Baki, A. A. 1980. Biochemical aspects of seed vigor. Hort Sci.
15: 765-771.
. Abdul-Baki, A. A., and J. D. Anderson. 1973. Relationship between
decarboxylation of glutamic acid and vigor in soybean seed. Cr0p
Sci. 13: 227-232.
. Abdul-Baki, A. A., and J. D. Anderson. 1970. Viability and leaching
of sugars from germinating barley. Crop Sci. 10: 31-34.
. Abdul-Baki, A. A., and J. D. Anderson. 1973. Vigor determination in
soybean seed by multiple criteria. Crop Sci. 13: 630-633.
. Anderson, J. D. 1977. Adenylate metabolism of embryonic axes from
deteriorated soybean seeds. Plant Physiol. 59: 610-619.
. Anfirund, PL rw., and A. A. Schneiter. 1984. Relationship of
sunflower germination and vigor tests to field performance. Crop
Sci. 24: 341-344.
. Association of official seed analysts. 1978. Rules for testing
seeds. J. Seed Technol. 3: 1-126.
. Associatitni of' Official Seed..Analysts. 1983. Seed ‘vigor testing
handbook. Assoc. Off. Seed Anal. Publ. AOSA Hand. 32.
. Azizul Islam, A†.1. M., J. (3. Delouche, and C. (L Baskin. 1973.
comparison of methods for evaluating deterioration in rice seed.
Proc. Assoc. Off. Seed Anal. 63: 155-160.
Bishnoi, I}. R., and J. (3. Delouche. 1980. Relationship of vigour
tests land. seed. lots tn) cotton. seedling, establishment. Seed Sci.
Technol. 8: 341-346.
Blowers, L. E., D. A. Stormonth, and C. M. Bray. 1980. Nucleic acid
and. protein synthesis and loss of vigour in germinating wheat
embryos. Planta. 150: 19-25.
Burris, ~I. S., and R†.J. Navratil. 1979. Relationship between
laboratory cold-test methods and field emergence in maize inbreds.
Agron. J. 71: 985-988.
Byrd, H. W., and J. C. Delouche. 1971. Deterioration of soybean seed
in storage. Proc. Assoc. Off. Seed Anal. 61: 41-57.
Cal, J. P., and R. L. Obendorf. 1972. Imbibitional chilling injury
in Zea mays L. altered by initial kernel moisture and maternal
parent. Crop Sci. 12: 369-373.
. Cher
aci<
. Chi]
Pla‘
Tec
15.
l6.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
73
Cheng, Y., P. Linko, and M. Milner. 1960. On the nature of glutamic
acid decarboxylase in wheat embryos. Plant Physiol. 35: 68-71.
Ching, TL PL 1973. Adenosine triphosphate content euui seed vigor.
Plant Physiol. 51: 400-402.
Ching, T. M. 1973. Biochemical aspects of seed vigor. Seed Sci.
Technol. 1: 73-88.
Ching, TH 14., and R. Danielson. 1972. Seedling vigor and adenosine
triphosphate level of lettuce seeds. Assoc. Off. Seed Anal. Proc.
62: 116-124.
Dalianis, C. [I 1982. Rate of radicle emergence as :1 measure of
seedling emergence and vigour in cotton (Gossypium hirsutum). Seed
Sci. Technol. 10: 35-45.
DasGupta, P. R., and H. M. Austenson. 1973. Analysis of
interrelationships among seedling vigor, field emergence, and yield
in wheat. Agron. J. 65: 417-422.
DasGupta, P. R., and H. M. Austenson. 1973. Relations between
estimates of seed vigor and field performance in wheat. Can. J.
Plant Sci. 53: 43—46.
Delouche, .I. C., and (L C. Baskin. 1973. Accelerated aging
techniques for predicting the relative storability of seed lots.
Seed Sci. Technol. 1: 427-452.
Don, R., J. R. Rennie, and M. M. Tomlin. 1981. A comparison of
laboratory vigour test procedures for winter wheat seed samples.
Seed Sci. Technol. 9: 641-653.
Duffus, C. M., J. H. Duffus, and J. C. Slaughter. 1972. Glutamate
decarboxylase in barely aleurone and its relationship to -amylase
development during germination. Experientia. 28: 635-633.
Edje, O. T., and J. S. Burris. 1970. Physiological and biochemical
changes in deteriorating soybean seeds. Proc. Assoc. Off. Seed Anal.
60: 158-166.
Fiala, F. 1981. Cold test. Handbook of Vigour Test Methods, (ed. D.
A. Perry), Int. Seed Testing Assoc. Zurich, Switzerland: 28-36.
Finch-Savage, W. E. 1986. A study of the relationship between
seedling characters and rate of germination within a seed lot. Ann
Appl. Biol. 108: 441-444.
Galleschi, L., C. Floris, and I. Cozzani. 1977. Variation of
glutamate decarboxylase activity and -amino butyric acid content of
wheat embryos during ripening of seeds. Experientia. 33: 1575-1576.
Galleschi,L., C. Floris, P. Meletti, and I. Cozzani. 1975. On the
location of glutamate decarboxylase in the caryopsis of hard wheat
(it!
Expei
30. Gill
durii
31. Grab
of s
32. Grab
18-3
33. Gral
lot:
34. Ham
lab
lin
33. Hep
ass
me;
am
36. he
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
74
(Triticum durum) and its activity during early germination.
Experientia. 31: 28-29.
Gill, N. S., and J. <3. Delouche. 1973. Deterioration of seed corn
during storage. Proc. Assoc. Off. Seed Anal. 63: 33-50.
Grabe, D. F. 1964. Glutamic acid decarboxylase activity as a measure
of seedling vigor. Proc. Assoc. Off. Seed Anal. 54: 100-109.
Grabe, D. P. 1976. Measurement of seed vigor. J. Seed Technol. 1(2):
18-32.
Grabe, D. F. 1965. Prediction of relative storability of corn seed
lots. Proc. Assoc. Off. Seed Anal. 55: 92-96.
Hampton, J. G. 1981. The relationship between field emergence,
laboratory germination, and vigour testing of New Zealand seed wheat
lines. N. Z. J. Exp. Agric. 9: 191-197.
Hepburn, H. A., A. A. Powell, and. S. Matthews. 1984. Problems
associated with the routine application of electrical conductivity
measurments of individual seeds in the germination testing of peas
and soybeans. Seed Sci. Technol. 12: 403-413.
Heydecker,W.l969. The ‘vigour' of seeds - a review. Proc. Int. Seed
Test. Assoc. 34: 201-219.
Inatomi, K., and. J. C. Slaughter. 1971. The role of glutamate
decarboxylase and -aminobutyric acid 1J1 germinating 'barley. J.
Exper. Bot. 22: 561-571.
Isely, D. 1957. Vigor testing. Proc. Assoc. Off. Seed Anal. 47: 176-
182.
James, E. 1968. Limitations of glutamic acid decarboxylase activity
for estimating viability in beans (Phaseolus vulgaris L.) Crop Sci.
8: 403-404.
Johnson, R. R. and L. M. Wax. 1978. Relationship of soybean
germination and vigor tests to field performance. Agron. J. 70: 273-
278.
Kulik, M. M., and Yaklich, R. W. 1982. Evaluation of vigor tests in
soybean seeds: relationship of accelerated aging, cold, sandbench,
and speed of germination tests to field performance. Crop Sci. 22:
766-770.
Lamkin, W. M., S. W. Nelson, B. S. Miller, and Y. Pomeranz. 1983.
Glutamic acid decarboxylase activity as a measure of percent
germination for barley. Cereal Chem. 60: 166-171.
Linko, P., and L. Sogn. 1960. Relation of viability and storage
deterioration to glutamic acid decarboxylase in wheat. Cereal Chem.
37: 489-499.
A4. Linko
relat
glute
45. Link
embr
46. Loef
W0
Tech
hi. Lope
per
Ana‘
48. Lun
rel
49. Mel
an(
50. he?
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
75
Linko, P., and M. Milner. 1959. Enzyme activation in wheat grains in
relation tx>‘water content, glutamic acid-alanine transaminase, and
glutamic acid decarboxylase. Plant Physiol. 34: 392-396.
Linko, P., and M. Milner. 1959. Gas exchange induced in dry wheat
embryos by wetting. Cereal Chem. 36: 274-279.
Loeffler, N. L., J. L. Meier, and J. S. Burris. 1985. Comparison of
two cold test procedures for use in maize drying studies. Seed Sci.
Technolo. 13: 653-658.
Lopez, A., and D. F. Grabe. 1973. Effect of protein content on seed
performance in wheat (Triticum aestivum L.). Proc. Assoc. Off. Seed
Anal. 63: 106-116.
Lunn, C., anui E. Madsen. 1981. ATP-levels of germinating seeds in
relation to vigor. Physiol. Plant. 53: 164-169.
McDaniel, R. C. 1969. Relationships of seed weight, seedling vigor
and mitochondrial metabolism in barley. Crop Sci. 9: 823-827.
McDonald, M. B. Jr. 1975. A review and evaluation of seed vigor
tests. Proc. Assoc. Off. Seed Anal. 65:109-139.
McDonald, M. B. Jr., and B. R. Phaneendranath. 1978. A modified
accelerated aging seed vigor test for soybeans. J. Seed Technol. 3:
27-37.
McDonald, M. B. Jr., and D. O. Wilson. 1979. An assessment of the
standardization and ability of the ASA-610 to rapidly predict
potential soybean germination. J. Seed Technol. 2: 1-11.
McKersie, B. D., D. T. Tomes, and S. Yamamoto. 1981. Effect of seed
size on germination, seedling vigor, electrolyte leakage, and
establishment of bird's-foot trefoil (Lotus corniculatus L.). Can.
J. Plant Sci. 61: 337-343.
Maguire, J. D. 1962. Speed of germination-aid in selection and
evaluation for seedling emergence and vigor. Crop Sci. 2: 176-177.
Mahdi, M. T., A. A. Lotfi, E. Shiltawy, and F. F. Farag. 1971. Cold
test of cotton seed. Int. Seed Test. Assoc. Proc. 36: 279-287.
Matthews, 8., and R. ‘Whitbread. 1968. Factors influencing pre-
emergence mortality in peas 1. An association between seed exudates
and the incidence of pre-emergence mortality in wrinkle-seeded peas.
Plant Path. 17: 11-17.
Matthews, 8., and M. F. F. Carver. 1971. Further studies on rapid
seed exudate tests indicative of potential field emergence. Proc.
Int. Seed Test Assoc. 36: 307-312.
Matthews, 3., anKiiJ. T. Bradnock. 1968. Relationship between seed
exudation and field emergence in peas and french beans. Hort.
.Mi
te
Se
.Mi
P1
(
\
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
76
Research. 8:89—93.
Mian, A†I.., and I“ (3. Coffey. 1971. Eighty-hour-count germination
test -A new test method for measuring seed vigor in corn. Proc. Int.
Seed Test. Assoc. 36: 265-271.
Mien, A. I“, enui L. C. Coffey. 1971. Testing seed vigor in rice.
Proc. Int. Seed Test Assoc. 36: 273-278.
Perry. D. A. 1973. Interacting effects of seed vigour and
environment on seedling establishment. In: Seed Ecologv. (ed., W.
Heydecker). Pennsylvania State University Press. pp. 311-323.
Priestley, D. A. 1986. Seed Aging. Comstock Publishing Associates.
304 pps.
Roberts, E. H. 1972. Viability and vigour of seeds. In Viability of
Seeds (ed. E. H. Roberts) Chapman and Hall, London.
Steel, R. C. D., and J. H. Torrie. 1980. Principles and Procedures
of Statistics. 2nd Ed. McCraw-Hill Book Co. 633pps.
St. John†41. B. 1970. Determinathmo of ATP in chlorella with the
luciferin-luciferase enzyme system. Analytical Biochem. 37: 409-416.
Styler, R†C., D. .J. Cantliffe, and (L B. Hall. 1980. The
relationship of ATP concentration to germination and seedling vigor
of vegetable seeds stored under various conditions. J. Am. Soc.
Hort. Sci. 105: 298-303.
Tao, K. J. 1979. An evaluation of alternative methods of accelerated
aging seed vigor test for soybeans. J. Seed Technol. 3: 30-40.
Tekrony, D. M., and D. B. Egli. 1977. Relationship between
laboratory indices of soybean seed vigor and field emergence. Crop
Sci. 17: 573-577.
Van Onckelen, H. A., R. Verbeek, and A. A. Khan. 1974. Relationship
of ribonucleic acid metabolism in embryo and aleurone to -amylase
synthesis in barley. Plant Physiol. 53: 562-568.
Woodstock, L. W. 1973. Physiological and biochemical tests for seed
vigor. Seed Sci. Technol. 1: 127-137.
Yaklich, R. W., M. M. Kulik and J. D. Anderson. 1979. Evaluation of
vigor tests in soybean seeds: relationship of ATP, conductivity, and
radioactive tracer multiple criteria laboratory tests to field
performance. Crop Sci. 19: 806-810.
CHAPTER III
THE EFFECT OF SEED DENSITY, SIZE AND PROTEIN CONTENT ON
STORABILITY OF TWO WHEAT VARIETIES.
ABSTRACT
A long term storage experiment was conducted to study the effect of
different seed characters of winter wheat (Triticum aestivum L.) on
storability and to determine the changes in ATP and GADA associated with
viability during storage. Seed lots of different density, size and
protein content were stored for 32 monthes at room conditions and tested
periodically for germination, ATP content and GADA. Standard germination
results after one year, two years and 32 nmnths were also correlated
with other vigor tests performed before storage. While viability of the
different seed categories did not change significantly during 18 months
of storage, ATP and GADA levels showed a continous decline throughout
the entire storage period. While all seed lots declined in germination
and biochemical vigor indices with increased time of storage, the high
density, large, and especially high protein seeds performed better than
the others and had higher germination at the end of the experiment. ATP
level was among the best vigor tests in predicting storability. However,
the accelerated aging test and the speed of germination test showed very
little correlation with the percent germination and were not good
indicators of storability.
77
78
REVIEW OF LITERATURE
Seeds stored for long periods of time gradually deteriorate and lose
viability due to degradative processes. The rate of ‘viability loss
depends largely upon the storage environment. Of the four factors
affecting storability, time and oxygen level have very little influence
if the other two factors, relative humidity and temperature, are kept at
Optimum levels (16). Roberts (15) found that the longevity of any seed
was influenced not only by environmental factors, but also by species
and initial seed quality, and Anderson (4) cited field and storage fungi
as contributing factors to deterioration. LikhatcheV' et al. (14),
studying seeds representing as number' of crops and varieties, found
considerable genetic differences in the rate of deterioration.
Aspinal and Paleg (5) observed that although viability of wheat
seed was not adversely affected by storage for up to six years, the rate
of germination was reduced after 3 to 4 years of storage. Egli et al.
(10) reported that storing soybean seed at 10.5 percent moisture and
variable temperatures had no effect on germination of any of 12 seed
lots tested, but at 13.5 percent moisture, the germination of 8 of the
seed lots luui declined significantly after’il months of storage.
Bittenbender and Ries (7) showed that both high and low protein rice
seeds declined in viability within one or two months when stored in
sealed vials or over water at 400 C and 20 percent moisture content;
however, high protein seeds remained viable longer than low protein
seeds.
By transplanting old wheat embryos on young endosperm and young
embryos on aged endosperm, Floris (11) showed that the aging process is
l I
aproy
metaboi
sugges
storag
produc
hours
embryc
primal
metab
soybe
that
cell“
note<
comp
prob
esa
agi
79
a progressive phenomenon accompanied by a gradual accumulation of toxic
metabolites and affects both embryo and endosperm. Abdul-Baki (2)
suggested that the earliest and most dramatic change occurring during
storage is the decline in ability of seeds to utilize glucose for C02
production and for polysaccharide and protein synthesis during the early
hours of imbibition. After studying the aging and deterioration of wheat
embryo and aleurone tissue, Aspinal and Paleg (5) concluded that the
primary factor in the aging process is probably related to decreased
metabolism or loss of intracellular integrity.
Chauhan (8) found that the growing points of the embryonic axis of
soybean and barley were very sensitive to accelerated aging, suggesting
that the meristems of the plumule and radicle are the most vital â€key
cell" regions if seeds are to remain viable. Likhatchev et al. (14)
noted an increase in proteolytic enzyme activity shortly before the
complete death of seeds, and remarked that such a phenomenon was
probably due to the loss of cell membrane integrity facilitating the
escape of enzymes. A rise in activity of hydrolytic enzymes during seed
aging was also observed by Abdul-Baki (2).
Harrington (12) listed symptoms of seed senescence as complete lack
of growth, slow or abnormal seedling growth, loss of membrane integrity,
change of color, loss of enzyme activity, and production of toxic end
products such as free fatty acids. Other researchers have reported that
the effects of seed storage on plant growth include delayed germination,
a decrease in rate of root elongation, the slowing of shoot growth in
the early stages of seedling development, and alteration of growth habit
(1, l3). Agrawal (3) observed that in wheat and triticale, the initial
loss in germination was due mainly to increased seedling abnormality,
80
but losses after 22 months of storage were due to death of seeds.
Delouche and Baskin (9) reported that germination capacity was the last
measure of quality to decline as seeds deteriorated during storage. Egli
et al. (10) reported that while the accelerated aging test was sensitive
enough to predict the deterioration rate of 3 soybean varieties during
storage, the rate of germination (4-day count), and especially the
standard germination test, were less sensitive.
Likhatchev et al. (14) remarked that seed vigor of soybeans
declined more rapidly than germination, and Delouche and Baskin (6)
reported that immature (low vigor) peanut seeds did not store as well as
mature seeds. Egli et al. (10) showed that low vigor soybean lots
exhibited a faster decline in viability during storage than high vigor
lots.
The first objective of this experiment was to study the effects of
several seed characters on storability under laboratory conditions, and
to determine whether differences in density, size, and protein content
lead to differences in viability after an extended storage period.
Another objective was to correlate several vigor tests to viability
during storage and to identify the vigor test(s) that most successfully
predict viability. A third objective was to study some biochemical
changes during storage and relate them to viability.
81
MATERIALS AND METHODS
Seeds of Augusta, a soft-white variety, and Hillsdale, a soft-red
variety were divided into two density classes, three size classes and
two protein levels according to the criteria described in Chapter 1.
After the vigor tests discussed in Chapter 2 were completed, seeds
of the same seed lots were stored in closed paper bags at room
temperature and about 35 percent relative humidity. The initial moisture
content of the seeds before storage was 12.3 to 12.5 percent. Storage
began in November, 1984 and terminated in July, 1987.
At selected times throughout the storage period, tests for
germination, glutamic acid decarboxylase activity (GADA) and ATP level
were performed using procedures described earlier in Chapter 2.
Germination tests were conducted every six months for the first year
(Nov. 84, May 85, and Nov. 85), and then every two months until the end
of the experiment. GADA and ATP tests were performed every six months
starting in November 1984 and ending in May 1987. The data were analyzed
as a factorial experiment in a completely randomized design, with
storage time, seed characters and varieties as the three factors. Simple
correlation analysis between standard germination, GADA and ATP results
were calculated using inean values and analyzed utilizing the MSTAT
statistical analysis program. The standard germination results after 12
months (Nov. 1985), 24 months (Nov. 1986), and 32 months (July 1987)
were correlated with results of the vigor tests listed in Chapter 2 and
included the accelerated aging test, speed of germination index,
conductivity index, cold test, ATP test and the GADA test.
82
RESULTS
Seed density in Augusta did not affect germination for up to 18
months of storage (Figure 3). Thereafter, viability in light and control
seeds declined more rapidly than in heavy seeds. ATP contents declined
steadily with storage time for both density classes. Decline in GADA was
greater during early storage than at later storage, especially for the
light aux! control treatments. While ATP content of light and control
treatments were similar, control GADA values were intermediate between
those for light and heavy seeds at all storage intervals.
Effects of seed size on standard germination parallaled the effects
of seed density (Figure 4). All seeds had high germination up ix) 18
months of storage, then germination declined rapidly in) to 28 months
for the control, medium. and large classes. Like density, seed. size
affected ATP content and GADA with the exception of small seeds which
gave much lower values than the other size classes. Low levels of ATP
occured in small seeds and changed little with storage time.
Protein levels, like seed size, affected germination of Augusta
(Figure 5), with little decline in germination of high protein seeds
after 32 months of storage. Similarly, both ATP and GADA in high and low
protein seeds resembled that in large and small seeds. Low protein seed
had significantly lower ATP levels than the high protein or control
treatments and changed little during storage.
Results for Hillsdale (Figures 6-8) were similar to those of
Augusta. Standard germination did not decline significantly during 18
months of storage for any of the seed classes, but declined rapidly
thereafter. Large and the high protein seeds declined relatively little
PEU\ CC (iii
\ </L4(
GADA (ppm CO /gm)
ATP x 10 M/Lit
Percent Germination
83
LSD (571) = 48
2.00-« LSD (57.) = 0.36
. A—a Hcovy .
4 H Light ‘ a
701 0—0 Control
1 T— T T— T l T T l l i 7 1
NOV MAY NOV ’JAN MAR MAY JULY SEP NOV JAN MAR MAY JULY
Figure 3. Storoge..effect on germination, ATP level
and GADA of two seed denSIty classes, Augusto.
flEmu\ 00 EQQV <D<MU
+u_\‘4 (w
GADA (ppm CO /qm)
ATP x 10 M/Lit
Percent Germination
84
LSD (52) = 48
LSD (5%) = 0.36
1.504
1.20-
0.90...
a—e Lorge
H Medium
X-X Small
* o—o Control
504 r 1
1 T I I I T I I I T '1
NOV MAY NOV JAN MAR MAY JULY SEP NOV JAN MAR MAY JULY
Figure 4. Store 'e effect ongerminotion, ATP level
and GADA of hree seed Size closses, Augusto.
85
LSD (52) a 48
GADA (ppm CO /gm)
LSD (571) = 0,36
1.90—
1.50-
ATP x 10 M/Lit
O.7O—J
lOO—w
‘hf LSD (57;) = 11
A L A
LO
Q
l
(I)
O
Ill-ALL
*1
A
'0
'0
Percent Germination
\J
O
A l. L .l
0
O)
O
11
H High Protein
4 X—X Low Protein
j o—oControl
50 I I I I a I I I I r I F I
NOV MAY NOV JAN MAR MAY JULY SEP NOV JAN MAR MAY JULY
Figure 5. Storo e effect on erminotion, ATP level
and GADA of wo seed pro ein levels, Augusto.
420‘ LSD (57;) = 52
GADA (ppm co /gm)
ATP x 10 M/Lit
Percent Germination
2320‘
2JDOA
1.80-
1.50-
1.20d
LSD (52) = 0.37
H Heavy
LSD (5%) = 8
‘ H Light
0—0 Control
70‘ I I I I I I I I I I I
NOrV— MAY NOV JAN MAR MAY JULY SEP NOV JAN MAR MAY JULY
Figure 6. Storoge effe
0nd GADA of two see
ct on germinotion, ATP level
d denSIty classes, Hillsdale.
ATP x 10 M/th
Percent Germination
GADA (PPm CO /gm)
i3
0
l
87
LSD (5%) = 52
LSD (5%) = 0.37
e—a Large
A—a Medium
‘ x—x Sme
o—o Control
60‘ I I‘ I T’ I I TV I I I I I I
NOV MAY NOV JAN MAR MAY JULY SEP NOV JAN MAR MAY JULY
Figure 7. Store e effect on germination, ATP level
and GADA of hree seed size classes, Hillsdale.
ATP x 10 M/Lit
Percent Germination
GADA (ppm co /gm)
290—
24o~
190J
88
440.. LSD (571) = 52
2.70-J
LSD (SZ) = 0.37
2.3oJ
1.90-
1.504
LID-a
O.70~ W
0.30—
100] -
j c’\ LSD (57) = 8
l ¢' '
90~ :
J Q
1 -
80-4 .-
ï¬ Â§
J \
703 °
601
‘ a—a High Protein
J H Low Protein
SO-jt o—o Control
I I I I I —T r I r I I“ I I
NOV MAY NOV JAN MAR MAY JULY SEP NOV JAN MAR MAY JULY
Figure 8. Store é effect on germination, .ATP level
and GADA of we seed protein levels, Hillsdale.
FIIIIIIIIIIIIIIIIIIlII--""E:_______________
89
in germination during the entire storage period. Germination of the low
protein treatment was significantly lower than that of the high protein
treatment, and the small treatment had significantly lower ATP levels
than the medium and large treatments. Neither the low protein nor the
small seed treatments showed a significant change in ATP level
throughout the storage period.
Results of the standard germination test were more closely
correlated with ATP than GADA levels at different times during storage
(Table 20). However, all three were significantly and positively
correlated. Results of the accelerated aging test were not consistently
correlated with standard germination (Table 21). Correlations were
significant after two years of storage in Augusta and one year in
Hillsdale.
Speed of germination was not significantly correlated with
standard germination for any of the periods tested. Conductivity index
was significantLy but negatively correlated with germination for both
varieties, especially after one year, but decreased with storage time
thereafter. Cold test results for Hillsdale were positively correlated
with germination after one and two years, but not after 32 months. After
2 years, only germination was significantly correlated with the cold
test results in Augusta.
Correlation of ATP content with percent germination increased with
storage time for both varieties, and was significantly correlated at all
three storage intervals except at one year for lï¬llsdale. Correlation
between GADA and germination of Hillsdale also increased with storage
time, but was significantly correlated only after 32 nmnths. However,
for Augusta, correlation with germination was always high, especially
FIIIIIIIIIIIIIIIIIIllI"-""E:_______________
90
Table 20. Correlation of standard germination, ATP and
GADA results of the storage experiment for Augusta and
Hillsdale.
Augusta Hillsdale
Warm Ger. ATP Warm Ger. ATP
GADA 0.53** 0.70** 0.57** 0.79**
ATP 0.84** 0,85**
** - significant at the 1 percent probability level.
Table 21. Correlation of the accelerated aging, speed of germination,
conductivity index, cold test, ATP and GADA tests with the standard
germination test results after one year, two years enui 32 months of
storage.
Acc. Speed Conduc. Cold ATP GADA
Storage Period aging germ. index test test test
AUGUSTA 1 Year 0.53 -0.22 -0.86** 0.56 0.71* 0.88**
2 Years O.89** -0.35 -0.75* 0.85** 0.90** 0.91**
32 Months 0.61 -0 28 -0.71* 0.48 0.93** 0.80*
Hillsdale 1 Year 0 78* -0 60 -0.82* 0.96** 0.53 0.59
2 Years 0.68 -O.33 -0.64 0.81* 0.85** 0.69
32 Months 0.67 -0.10 -0.57 0.57 0.97** 0.84**
? - significant at the 5 percent probability level.
** - significant at the 1 percent probability level.
91
after one and two years of storage.
92
DISCUSSION
Regardless of seed size, density or protein content, germination
declined only after some biochemical parameter such as ATP or GADA had
already declined appreciably (Figures 3-8). These results agree with
observations by Delouche and Baskin (9) who noted that germination of
peanuts was the last measure of quality to decline during storage.
Although seed characters did not significantly affect initial
germination, differences in ATP and GADA content were apparent prior to
storage and were generally maintained throughout storage. Unlike the
standard germination, ATP and GADA did run: change appreciably during
storage. However, the differences in germination among different
density, size and protein classes increased during storage. These
results show the inadequacy of the standard germination test to predict
storability of seeds, especially for short periods. They also show the
inability of the standard test to reveal differences in seed vigor.
Our results also show that physiological and biochemical changes
take place prior to an actual decline in viability. Thus, regardless of
initial quality, seeds can be stored for at least a year without
significant reduction in viability even though a continous reduction in
vigor may occur.
It is apparent from Table 20 that though both biochemical indices
were significantly and positively correlated with standard germination
over the entire storage period, the ATP test was a better indicator of
seed viability. Table 21 also shows that initial ATP test results were
superior to GADA or any other test in predicting intermediate and final
viability after storage.
’i—
93
The accelerated aging test, which is thought to simulate the
natural aging process, showed poor correlations with germination results
for both Augusta and Hillsdale, especially after 32 months of storage.
Thus, although the accelerated aging test is a good measure of vigor, it
does not neccessarily simulate natural deterioration during storage, and
therefore cannot predict seed storability. These results are in
disagreement with findings by Egli et al. (10) who reported that the
accelerated aging test was a good indicator of deterioration of stored
soybeans. Other results suggest that tests sudh as the cold test and
accelerated aging test can predict viability during short storage
periods. However, when the storage period is relatively long,
correlations with germination decrease.
Heavy, large and high protein seeds stored significantly better
than light, small or low protein seeds. While ATP content and GADA
declined at the same rate for different categories, their initial levels
were high enough in the good quality seeds to prevent large germination
losses. A second possibility might be the tendency of light, small and
low protein seeds t1) utilize their energy reserves more rapidly than
other seeds, leading to a decline in germination. This was especially
apparent in high protein seeds which had better resources and capacity
to withstand prolonged storage than low protein seeds. By contrast,
seeds with low protein levels had the lowest ATP, GADA and germination
results, comparable with those of' small seeds, which ‘would. also be
expected to have limited energy resources. These results agree with
those of Bittenbender and Ries (7) who also recognized the advantages of
high vs. low protien content for rice storage. They also agree with the
conclusions by Roberts (15), who noted that initial quality was one of
94
the factors influencing seed longevity.
i
10.
11.
12.
. Abdul-Baki, A. A. 1969.
. Agrawal, P. K. 1978. Changes
. Anderson, J. D.
. Aspinal, D., and L. G. Paleg. 1971.
. Baskin, C. C
. Chauhan, K. P. S. 1985.
. Delouche, .1. C., and C. C. Baskin. 1973.
95
REFERENCES
. Abdulla, F. H., and E. H. Roberts. 1969. The effect of seed storage
conditions on the growth and yield of barley, broad bean, and peas.
Ann. Bot. 33: 169-184.
Relationship of glucose metabolism to
germination and vigor in barley and wheat seeds. Crop Sci. 9: 732-
737.
in germination, moisture and
carbohydrate of hexaploid triticale and wheat (Triticum aestivum)
seeds stored under ambient conditions. Seed Sci.
Technol. 6: 711-
716.
1970. Metabolic changes in partially dormant wheat
seeds during storage. Plant Physiol. 46: 605-608.
The deterioration of wheat
embryo and endosperm function with age. J. Exp. Bot. 22: 925-935.
., and J. Delouche. 1971. Differences in metabolic
activity in peanut seed of different size classes. Proc. Ass. Off.
Seed Anal. 61: 73-77.
. Bittenbender, H. C., and S. K. Ries. 1977. Germination and growth of
rice seedlings from high and low protein seeds after exposure to
various storage conditions. J. Seed Technol. 2: 62-72.
The incidence of deterioration and its
localisation in aged seeds of soybean and barley. Seed Sci. Technol.
13: 769-773.
Accelerated aging
techniques for predicting the relative storability of seed lots.
Seed Sci. Technol. 1: 427-452.
Egli, D. B., G. M. White, and D. M. Tekrony. 1979. Relationship
between seed vigor and the storability of soybean seed. J Seed
Technol. 3(2): l-ll.
Floris, C. 1970. Ageing in Triticum durum seeds: behaviour of
embryos and endosperms from aged seeds as revealed by the embryo-
transplanting technique. J. Exp. Bot. 21: 462-468.
Harrington, J. F. 1973.
Biochemical basis of seed longevity. Seed
Sci. Technol. 1: 453-461.
. Harrison, J. G. 1977. The effect of seed deterioration on the growth
of barley. Ann. Appl. Biol. 87: 485-494.
— 7
14.
15.
16.
96
Likhatchev, B. 8., G. V.
Shevchenko. 1984. Modelling
385-393.
Zelensky, Y. C. Kiashko, and Z. N.
of seed ageing. Seed Sci. Technol. 12:
Roberts, E. H. 1973. Predicting the storage life of seeds. Seed Sci.
Technol. 1: 499-514.
Roos, E. E. 1980. Physiological, biochemical and genetic changes in
seed quality during storage. HortSci. 15(6): 781-784.
SUMMARY AND CONCLUSIONS
While vigor tests are commonly used to determine seed quality of
many crops such as corn and soybeans, no single test is accepted as a
good measure of wheat seed quality. Nor has the effect of different seed-
physical characters on field performance, vigor or storability been
studied sufficiently.
Two winter wheat varieties, Augusta, a soft white variety, and
Hillsdale, a soft red vriety were divided into several density, size and
protein classes. The effects of these seed characters on field
performance, vigor and storability were then investigated.
Results of the field study showed that light, small and low protein
seeds yielded significantly less than heavy, large and high protein
seeds, respectively, and that greater differences in yield occured under
unfavorable environmental conditions. Tillers per meter, seeds per spike
and 1000-seed weight all contributed significantly to variation in grain
yield. However, under adverse field conditions, 1000-seed weight made no
significant contribution to variation in grain yield. Yields were also
influenced by location, with the higher yields in the Pigeon location
relative to the East Lansing one. A strong association existed between
final emergence and rate of emergence, indicating that seeds which
emerge rapidly also tend to have a higher final emergence. Field
performance of' the different treatments onus correlated 'with several
vigor tests. The accelerated aging test and the cold test, both of which
97
98
apply stress to seeds, were the best predictors of yield for both
varieties over all treatments. Although glutamic acid decarboxylase
activity and the ATP level were also significantly correlated with grain
yield, they 'were not always able to detect differences among seed
classes. None of the tests correlated well with total emergence or
emergence rate index, thus other factors apparently influence
establishment in the field. Among the least successful tests in
predicting either emergence or yield were the conductivity index and
germination rate index. Since smaller and lighter seeds reach the
critical moisture content for germination more quickly during imbibition
than large or heavy seeds, any measure of germination speed that does
not involve stress or account for factors such as dry matter
accumulation will be biased in their favor.
Viability of the different seed categories, as measured by percent
germination, changed little during the first. 18 months of storage.
Unlike the standard germination test results, the ATP and GADA declined
steadily throughout the entire storage period. The ATP and GADA tests
were therefore more sensitive to changes leading to decreased viability
that were not detected by the standard germination test.
All seed categories showed a decline with increased storage, but
the heavy, large and especially the high protein seeds did not lose
viability rapidly, and therefore stored better than other seed classes.
The ATP test was the best predictor of storability among the different
vigor tests, and high correlations were obtained between initial ATP
level and viability after one year, two years and 32 months of storage.
In contrast, the accelerated aging test and the speed of germinathni
index were poorly correlated with germination after one year, two years
99
and 32 months of storage.
The following conclusions can be made from our study:
- Seeds with a high emergence rate index are also expected to have a
high total emergence.
- Biological yield and straw yield were not significantly affected by
the different seed characters studied.
- Grain. :yield was significantly affected. by the different seed
characters.
- Number of tillers was the most important factor contributing to
variation in grain yield, followed by seeds per spike and 1000-
seed weight.
- Stress tests, such as the accelerated aging test and cold test,
were the best indicators of yield, followed by tests that measure
biochemical parameters, such as ATP and GADA levels.
- No vigor test was consistently correlated with either emergence rate
index or total emergence.
- A decrease in viability during storage is preceeded by a decrease in
GADA and ATP content.
- ATP and GADA tests were the best predictors of long term
storability.
- The accelerated aging test and speed of germination test were not
good predictors of seed storability.
APPENDICES
100
APPENDIX A
Table A1. Analysis of variance for emergence (plants per meter),
and emergence rate index (E.R.I.) for Augusta and Hillsdale, 1985.
Mean Square
--------¢-------------ca--- ............
Source of Degrees of Emergence E.R.I
Variation Freedom Number
Replication 3 299.0** 5.33*
Treatment 7 297.0** 4.54**
Variety 1 3.5 3.93
T X V 7 172.7* 1.71
Error 45 67.6 1.44
C.V. 9.4% 12.2%
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
Table A2. Analysis of variance for emergence (plants per meter),
and emergence rate index (E.R.I.) for Augusta and Hillsdale, 1986.
Mean Square
Source of Degrees of Emergence E.R.I
Variation Freedom Number
Replication 3 968.4** 16.66**
Treatment 10 536.9** 8.89**
Variety 1 84.1 3.06
T X V 10 142.7 2.63
Error 63 165.5 2.58
C.V. 15.2% 17.0%
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
101
Table A3. Analysis of variance for biological yield, straw yield,
grain yield and harvest index per meter for Augusta and Hillsdale,
1985.
Mean Square
Source of Degrees of Biol. Straw Grain Harvest
Variation Freedom Yld. Yld. Yld. Index
Treatment 7 3306.6** 1589.7** 451.2** 9.62**
Variety 1 724.3 2759.6** 671.2** 106.78**
T X V 7 203.9 97.2 69.7 2.17
Error 30 400.2 238.6 39.2 1.60
C.V. 5.7% 7.1% 4.7% 3.3%
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
Table A4. Analysis of variance for biological yield, straw yield,
grain yield and harvest index per meter for Augusta and Hillsdale,
1986.
Mean Square
Source of Degrees of Biol. Straw Grain Harvest
Variation Freedom Yld. Yld. Yld. Index
Treatment 10 1251.5** 669.0* 611.2** 62.23**
Variety 1 3256.2** 507l.5** 200.3 220.44**
T X V 10 78.6 36.7 43.9 5.01
Error 63 429.6 318.7 64.3 10.23
C.V. 8.7% 11.5% 9.7% 9.1%
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
102
Table A5. Analysis of variance for yield components of Augusta and
Hillsdale, 1985.
Mean Square
Source of Degrees of Tillers Seeds 1000
Variation Freedom /meter per Spike Seed Wt.
Treatment 7 338.2** 6.6** 1.5
Variety 1 652.7** 1.0 53.5**
T X V 7 25.8 2.9* 1.9*
Error 30 40.0 1.0 0.7
C.V. 5.5% 3.3 1.9%
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
Table A6. Analysis of variance for yield components of Augusta and
Hillsdale, 1986.
Mean Square
Source of Degrees of Tillers Seeds 1000
Variation Freedom /meter per Spike Seed Wt.
Treatment 10 44l.6** 10.6** 4 6
Variety 1 104.7 17.6* 0.2
T X V 10 19.9 1.0 3.8
Error 63 75.2 1.6 6.3
C.V. 9.5% 4.0% 8 3%
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
103
Table .A7. Analysis of variance for yield of two varieties,
Augusta and Hillsdale, grown in two locations, East Lansing and
Pigeon in 1985 and 1986, and emergence rate index (E.R.I.) for
Augusta and Hillsdale, 1985.
1985 1986
Source of Degrees of Mean Degrees of Mean
Variation Freedom Square Freedom Square
Location 1 4924084** 1 5758451**
Treatment 7 1081989** 10 2378094**
Variety 1 117482 1 21453
T X V 7 432490 10 30335
L X T 7 106008 10 164525
L X V 1 785108 1 563358
L X V X T 7 117763 10 49419
Error 60 119730 126 179997
C.V. 5.8% 11.3%
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
Table A8. Analysis of variance results for standard
germination (WC), accelerated aging (AA), cold test
germination (CG), and speed of germination index (SC), 1985.
Mean Square
Source of ................................
Variation D.F WG AA CG SG
Treatment 7 26.0** 1584** 1837** 141.7**
Variety 1 6.9 2463** 182 120.9**
T x V 7 2.9 193** 78 55.2**
Error 48 2.6 49 51 7.3
C.V. 1 7% 9.9% 10.6% 8.1%
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
Table A9.
germination
104
Analysis of ‘variance results for .standard
(WC), accelerated aging (AA), cold test
germination (CG), and speed of germination index (SC), 1986.
Source of
Variation
Treatment
Variety
T x V
Error
C.V.
Mean Square
D F WG AA CG SG
10 31.5** 1026** 1461** 30.5**
1 0.3 482** 22 0.5
10 5.5 59** 44 1.7 l
66 6.1 20 34 6.2
2.6% 6.4% 8.7% 8.6%
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
Table A10. Analysis of variance results for conductivity index
(CI), ATP,
1985.
and. glutamic acid. decarboxylase activity (GADA),
Source of
Variation
Treatment
Variety
T x V
Error
C.V.
Mean Square
--------------- .............
D.F CI ATP GADA
7 970** 5.2** 14196**
1 89 O.9** 5659**
7 8 0.1 1030
48 128 0.1 1241
19.1% 16.9% 7.6%
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
Table All. Analysis of variance results for conductivity index
(CI), ATP,
1986.
and. glutamic acid. decarboxylase activity (GADA),
Source of
Variation
Treatment
Variety
T x V
Error
C.V.
Mean Square
D.F CI ATP GADA
10 881** 4.3** 8945**
1 116 0.3% 647
10 159 0.1 411
66 226 0.1 989
26.1% 13.1% 6.8%
* - Significant at the 5% level of probability.
** - Significant at the 1% level of probability.
105
APPENDIX B
Table B1. Mean monthly temperature for two Michigan locations,
East Lansing and Pigeon for the 1984-1986 period.
Pigeon East Lansing
Monthly 30 Year Monthly 30 Year
Year Month Average Average Average Average
1984 Jan 15.7 20.7 14.6 21.9
Feb 30.5 22.1 31.5 23.9
Mar 25.3 31.1 25.5 33.4
Apr 44.7 44.4 45.6 46.5
May 53.7 55.2 51.8 57.6
June 69.1 64.9 68.8 67.1
July 68.7 69.4 68.6 70.7
Aug 67.4 67.9 70.5 69.0
Sep 57.7 61.0 58.7 62.1
Oct 53.8 50.7 52.4 51.2
Nov 38.6 38.5 37.3 38.7
Dec 30.8 26.7 31.5 27.0
1985 Jan 18.2 20.7 18.3 21.9
Feb 21.1 22.1 20.4 23.9
Mar 33.3 31.1 35.9 33.4
Apr 48.6 44.4 50.9 46.5
May 56.2 55.2 58.9 57.6
June 61.3 64.9 62.2 67.1
July 69.0 69.4 69 2 70.7
Aug 66.1 67.9 69.3 69.0
Sep 58.2 61.0 62.4 62.1
Oct 50.3 50.7 50.6 51.2
Nov 38.0 38.5 38.0 38.7
Dec 20.4 26.7 20.3 27.0
1986 Jan 20.9 20.7 21.4 21.9
Feb 21.6 22.1 21.2 23.9
Mar 33.4 31.1 34.8 33.4
Apr 46.3 44.4 49.1 46.5
May 55.9 55.2 58.5 57.6
June 62.5 64.9 65.2 67.1
July 70.7 69.4 70.1 70.7
Aug 64.9 67.9 66.6 69.0
Sep 59.6 61.0 62.1 62.1
Oct 48.4 50.7 49.7 51.2
Nov 34.7 38.5 34.6 38.7
Dec 28.0 26.7 28.8 27.0
106
Table 82. Mean monthly precipitation for two Michigan locations,
East Lansing and Pigeon for the 1984-1986 period.
Pigeon East Lansing
Monthly 30 Year Monthly 30 Year
Year Month Average Average Average Average
1984 Jan 0.95 1.79 0.46 1.40
Feb 0.97 1.56 0.79 1.21
Mar 3.26 2.20 2.76 2.09
Apr 4.33 2.66 2.43 2.82
May 4.25 2.58 4.30 2.73
June 4.13 2.88 0.18 3.54
July 1.68 2.93 2.04 3.02
Aug 2.97 3.01 2.64 3.12
Sep 3.99 2.67 3.03 2.58
Oct 3.80 2.49 2.93 2.20
Nov 3.27 2.38 3.14 2.22
Dec 3.26 2.18 3.29 1.84
1985 Jan 4.04 1.79 3.03 1.40
Feb 4.13 1.56 3.58 1.21
Mar 5.81 2.20 4.03 2.09
Apr 3.45 2.66 3.74 2.82
May 2.58 2.58 2.73 2.73
June 1.72 2.88 2.19 3.54
July 2.98 2.93 2.05 3.02
Aug 6.39 3.01 4.04 3.12
Sep 4.52 2.67 3.43 2.58
Oct 3.01 2.49 4.98 2.20
Nov 6.74 . 2.38 3.81 2.22
Dec 1.52 2.18 0.93 1.84
1986 Jan 1.18 1.79 0.72 1.40
Feb 2.33 1.56 3.20 1.21
Mar 1.81 2.20 1.64 2.09
Apr 1.81 2.66 2.76 2.82
May 2.93 2.58 3.51 2.73
June 3.26 2.88 6.67 3.54
July 3.20 2.93 2.76 3.02
Aug 4.63 3.01 3.89 3.12
Sep 13.39 2.67 7.99 2.58
Oct 2.60 2.49 2.70 2.20
Nov 0.41 2.38 1.25 2.22
Dec 1.88 2.18 1.22 1.84
"willWillow