Published September, 1963
Caryopsis Developmentin Several Grain SorghumVarieties and Hybrids’
2J. w. Collier
T HE relationships between moisture content,
maturity, seed size, and number of days after
stage of
blooming
are very important to those working with crop seeds.
Unfortunately these relationships are not well defined and
usually vary with both environmental conditions and
genetic differences within the crop. The purpose of this
paper is to present data on caryopsis development in sev-
eral grain sorghum varieties and hybrids during 1958 and
1959.
LITERATURE REVIEW
Aldrich (1), Andrews (2), Dessureaux et al. (4),
and Thom (13), and Hallauer and Russell (6) defined
maturity in corn as the point or time at which the grain
reaches its maximumdry weight. The studies by Brown
and Garrison (3) and Shaw and Thorn (13) indicated
the interval from silking to maturity in corn appears to be
fairly constant; however, Dessureaux et al. (4), reported
that the interval from silking to maturity ranged from 52
to 64 or more days. They also found that an inbred line
of corn high in moisture content generally increased in dry
matter more rapidly or for a longer period of time or both.
Shaw and Thorn (13) found that maximumdry weights
were absolute values and as such, were dependent on ear
size to a greater extent than a percentage or ratio. They
proposed a ratio of dry weight of sample to total or final
dry weight as a measure of maximumdry weight accumu-
lation. Hallauer and Russell (6) estimated maturity as the
time when the maximum dry ,weight of a 100 kernel
sample was obtained.
Kersting et al. (9) found that maximum dry weight
accumulation in male sterile Combine Kafir-60 sorghum
occurred at 45 days after pollination with 23% moisture
in 1958 and at 33 days after pollination with 30% mois-
ture in 1959. Wikner and Atkins (14) reported that max-
imum dry weights in 6 strains of sorghum were reached
at moisture percentage ranges of from 31-35% to 34-
39%. Harlan and Pope (7) reported that maximum dry
weight accumulation occurred in barley at about 40%
moisture and Scott et al. (12) found that wheat also
obtained its maximumdry weight at about 40% moisture.
Grabe (5) reported that maximum dry weight in smooth
bromegrass was reached in about 18 days after anthesis
with a moisture content of about 47%.
Hesketh and Musgrave (8) showed that respiration rates
in corn ears increased as temperatures increased from 10
to 40° C. and decreased as moisture content decreased from
80 to 30%. Kersting et al. (10) suggested that respiration
continued in sorghum for some time after physiological
maturity of the kernel had been reached. Kramer (11)
stated that plant water stress sometimes causes increased
rates of respiration.
EXPERIMENTAL PROCEDURE
The strains of grain sorghumused in this study were grownat
the Blackland Experiment Station near Temple, Texas, in 1958
and 1959on uniformHoustonBlack clay soil. The varieties Martin
and Combine7078 and hybrids RS 610 and E 56a were grown in
both years. CombineShallu (SA 394) was grown in 1958 only
1Published as Technical Contribution No. 4367 of the Texas
Agricultural ExperimentStation. ReceivedApr. 4, 1963.
~Associate Agronomist, Departmentof Soil and Crop Sciences,
Texas A & M College.
419
and the male sterile strain of combinekafir (A3197) was grown
in 1959only. Hereafter the different varieties and hybrids will be
referred to as strains.
Each strain was planted in 2 blocks of 4 rows 30 feet long.
Rowwidth was 40 inches and the blocks were randomized. Plant-
ing dates were March 26 in 1958 and March 4 in 1959. Good
stands were obtained and each row was thinned to a uniform
stand of about 4 plants per foot of row. Althoughthe two grow-
ing seasons were different both growth and developmentof grain
sorghumwere normal for central Texas in both years.
Approximately100 heads of each fertile strain were tagged and
numberedthe secondday after initiation of anthesis. At this time
the florets in the upper 25 to 40%of the head had bloomed.In
1958 all heads were tagged from June 25 to June 30 and in 1959
from June 12 to June 17. The sterile strain A3197was bagged
before anthesis and the bags were removed7 days later at 8 a.m.
on June 17. Amplepollen was present and June 17 was used as
the date of blooming for the sterile strain.
Headsamples were taken first at 10 or 11 days after blooming
and at 1- to 3-day intervals during the sampling period. During
1958 the sampling period was from July 7 to August 6. At each
sampling date, 5 heads of each strain were selected at random,
harvested, placed in bags, taken to the laboratory and 200 kernels
from the top portion of each head were removed. These °samples
were weighed immediately and placed in an oven at 70 C. to
obtain total dry weight. During 1959, the sampling period was
from June 23 to July 24 and the sampling procedure differed
slightly from that used in 1958. At each samplingdate only a few
branches from the top portion of 3 heads of each strain were re-
moved,placed in glassine bags, and taken to the laboratory where
200 kernels were removedfor use in determining dry weight. All
heads from whichsamples were taken remained in the field until
45 days after blooming, at which time each head was harvested
and threshed and final dry weight of 100 kernels was determined.
In addition to the dry weight of 100 kernels, the percentage of
the final dry weight also wascalculated for each sample.
RESULTS AND DISCUSSION
A summary of several climatic factors from 30 days
before blooming through the sampling periods of 1958
and 1959 is shown in Table 1. Open pan evaporation was
included to show the combined effects of temperature, rela-
tive humidity, and wind movement. The combination of
lower rainfall, higher temperatures, and higher evapora-
tion rates during 1958 were less favorable for grain sor-
ghum production than the near ideal conditions during the
1959 season. Mean grain yields of the two hybrids, RS 610
and E 56a, were about 4,200 pounds per acre in 1958 and
5,200 in 1959.
Changes in moisture percentages and dry weights of 100
kernels of the various strains during the sampling period
are shown in Figures 1 and 2. Although the samples were
taken at 2- to 3-day intervals in 1958 and 1- to 2-day inter-
vals in 1959, most points on the curves are smoothed aver-
ages of 3- or 4-day periods. This procedure reduced the
variation due to sampling, particularly during the last half
of the sampling period.
For convenience in reporting and discussing the results
obtained, the sampling periods were divided into two 13-
Table 1--Summaryof 3 climatic factors on 5 dates duringtest
periods in 1958 and 1959.
Period Preclpl- .
Mean max, Open pan
ration, In. temp.,° F. evap.,in./day
1958 1959 1958 1959 1958 1959"
30 days befor~ blooming 3, 81 4.15 93 87 0.24 0.18
Blooming to first sample (12 days) 0.80 0.80 92 93 0.27 0.25
¯ Samplingperiod
12 to 24 days after blooming 0.54 1.56 92.3 91.7 0.26 0.23.
25 to 37 days after blooming
38 to 42 days after blooming
0.00
0.91
0.00
2.71
97. ~ 99.9 0.33
0.34
0.25
0.18
100.0 92.2
420 CROP SCIENCE

day periods, 12 to 24 days after blooming and 25 to 37 was upset somewhat by precipitation from 39 to 42 days
days after blooming. All strains and hybrids, except the 2 after blooming in both years.
male sterile strains grown in 1959, were sampled during Regression analyses of the moisture data show that over-
this period each year. The drying process, in particular, all decreases per day in 1958 ranged from 1.7~ in Shallu

80-

70 - 2.5 70-

60-

z 5O SO-

o
40-

u
~ 30 30-

MARTIN
20-
ES6a- 1958
1958-.-- .... 1959- ~
I0 L 1.0 10-
1959-~ SHALLU
1958............

10 2O 30 10 20 ~0 40

90 - 3.0

8O 80-

7O ’- 2.5 70- 2.5

¯ ~60 60-

_z 50 - 2.0
!
0 40-
i I
I I
i
I
! 1.5~
~ 30 - 1.5 30-

20 // 7078
1958
RS610-

1959- 1959
10 - 1.0 10- 1.0

I I I I
10 20 30 40 10 20 30 40
NUMBER
OF DAYSAFTER
BLOOMING NUMBER
OF DAYSAFTER
BLOOMING
Figure l~Changes in moisture percentages and dry weights of 100 kernels of the various strains during the sampling period
from 10 to 40 days after blooming.
 COLLIER: CARYOPSISDEVELOPMENT
IN GRAIN’SORGHUM 421

to 2.0% in E 56a and in 1959 from 1.4% in RS 610 to Table 2--Moisturepercentagesof developing caryopses at 3
1.8%in 7078. The differences in daily water loss for the dates after bloomingin 1958and 1959.
Strain or ~o moisture in caryopses
two years were very consistent amongthe different strains. hybrid Days after blooming, 1958 Days after blooming, 1959
Simple correlation coefficients for days after bloomingand 23 38
12 25 38 12
moisture percentage were highly significant and ranged 7078 65.2 35.0 14.0 67.0 39.0 23.0
from --.96 to --.98. In both years, the decreases per day RS 610 60.0
-
34.0
-
14.5
-
62.5 36.0 23.0
21.0
A3197 63.0 38.5
were slightly higher and more consistent amongstrains Martin 61.0 30.0 10.2 62.5 34.0 17.5
E 56a 64.5 30.0 10.0 65.0 34.0 18.0
from 12 to 24 days after blooming. During this period, Shallu 55.0 28.2 11.5 - - -
decreases per day ranged from 1.9% in A3197in 1959 to A398 60.0 34.5 15.0
2.6%in E 56a in 1958. From 25 to 37 days after bloom-
ing, decreases in moisture _percentages were smaller and ences appear early in the development of the caryopses
more dependent on strains and environmental conditions. and persist until moisture percentages reach 12 or 14%.
For this period, daily decreases in moisture ranged from Although the decreases in moisture percentages were
1.0% in RS 610 in 1959 to 1.6% in 7078 in 1958. Mois- linear during the sampling period, the developing cary-
ture changes from 39 to 42 days after blooming were opses did not lose moisture during the entire period. It
erratic as rainfall wasreceivedeach year during this period. was possible to calculate, from the data showngraphically
Onevery significant and consistent characteristic of the in Figures 1 and 2, the amountsor weights of water present
moisture percentage data was the higher moisture percent- per 100 kernels of the various samples. These figures, as
ages in the kernels of 7078, A3197and RS 610 than in such, are not included, but they show that maximum
those of Martin and Shallu at comparable dates after amountsof water in the caryopses occurred as early as 11
blooming. These data are summarized in Table 2. The to 13 days after bloomingin Shallu, Mar~inand E 56a in
moisture percentages in kernels of E 56a were almost iden- 1958 to as late as 16 to 18 days after blooming in 7078
tical to those in RS610 from 12 to 24 days after blooming and A3197in 1959. Amountsof water in the caryopses
but during the next 13 days the kernels of E 56a dried remainedfairly constant for another 2 to 6 days and actual
considerably faster than those of RS610.Themoisture per- losses of water did not occur until 16 to 17 days after
centages at any time from 25 to 37 days after blooming blooming in 1958 and 16 to 22 days after blooming in
indicate the 2 distinct groups commonlyreferred to as 1959. Thus, moisture percentage decreases during the first
"slow driers" or greenheadedstrains and the "fast driers" 14 to 16 days after blooming were caused by dry matter
or dryheaded strains. In the case of Martin and Shallu accumulations exceeding water accumulations in the young
compared with 7078 and RS 610, the difference appears caryopses’. Actual water losses from caryopses usually did
to be more than just the rate of drying, since the differ- not occur until moisture percentages were below 50%.
The curves for dry weights of 100 kernels are shownin
90 ,~: Figures 1 and 2. Regression analyses were made also on
3.0 dry weights of 100 kernels and days after blooming. Dur-
1 ing the first sampling period, from 12 to 25 days after
8o / blooming, dry weights of kernels increased in a linear
:/ ." k fashion with very small deviations. Actual weight increases
t~.:~ "%.[ "% J
of 100 kernels per day ranged from .08 to .12 gram with
6 of 10 strain-year combinations averaging .11 gram gain
per day. No reason can be suggested for the smaller dry
’
%..., .!J weights of caryopses during the first 5 to 7 days of the
/ sampling period in 1959 than in 1958.
Correlation coefficients of days after bloomingand dry
weights per 100 kernels for the entire sampling period
z so~1_ "~ ",.~.! ¯ -!- 2.0~ were all highly significant primarily because of the excel-
lent linear relationship during the first half of the sam-
- I piing period. Several types of regression equations were
calculated for the dry weight curves but since the curves
during the last half of the sampling period were of several
different shapes, the equations were of little or no value in
~-30 ~
15 determining the points at which maximumdry weights

I .,’ ......
/
were reached.
During 1958 maximumdry weights were reached in 24
to 28 days after blooming in all strains except RS 610
which continued to increase in dry weight until 35 days
after blooming. Apparently the more favorable environ-
mental conditions during July of 1959 permitted a longer
period of dry matter accumulation than during the 1958
sampling period. Maximumdry weight of A3197 was
reached in about 26 days after blooming during 1959 and
in 31 to 35 days after bloomingfor all the other strains.
20 :30 40 Maximum dry weights were reached at slightly higher
NUMBEROFDAYS AFTERBLOOMING moisture percentages in 1958 than 1959 in all strains ex-
~Figm:e
2---Changes
in moisture
percentagesanddryweighcs of cept RS 610. Althoughdry matter deposition occurred over
100kernelsof RS610andits parentalstrainsfromI0 to 40
d~ys~f~e~bloomin~du~i~195~. a longer period in 1959 than in 1958, rates of moisture
 422 CROP SCIENCE

loss were only slightly lower in 1959 than in 1958. Maxi-

mum dry weights were reached at moisture percentages
from 22% in Martin in 1959 to 32% in 7078 in 1958.
These data emphasize the difficulty of trying to determine
maturity by the use of moisture percentages if different
strains and years are involved.
A comparison of rates of moisture loss and dry matter
accumulation in RS 610 and its parental strains, A3197
and Combine 7078, are shown in Figure 2. In general,
moisture changes in the hybrid were similar to those of its
parents. On the other hand, dry weights of 100 kernels of
RS 610 were usually higher than those of its parents dur-
ing the period from 12 to 35 days after blooming. During
this particular year, dry matter was accumulated at a slightly
faster rate in caryopses of the hybrid than in either of the
two parents. Comparisons of 7078 and RS 610 in 1958
show that dry weights were not greater in RS 610 than in
7078 but that dry matter was accumulated during a longer
period in RS 610 than in 7078.
In 1959 both the dry weights of 100 kernels and per-
centages of the final weight of each sample were available.
Curves comparing these two methods of determining matu-
rity are shown in Figure 3. The use of the percentage of
the final dry weight reduced the variation due to different
seed sizes among plants and may be useful if only a few
plants are sampled. However, there appear to be definite
limitations or dangers in using this method to determine
the time at which maximum dry weights were reached.
These data, and others (10) indicate that dry weights
decreased after reaching maximum values. Respiration may
have caused all or a portion of these dry weight decreases.
Since the final dry weights were less than maximum dry
weights, percentages of final dry weights usually reached
peaks of 102 to 108%. Thus, the time at which 100% of
the final dry weight was first reached in a particular strain,
was from 2 to 5 days earlier than the time maximum dry
weights were reached.
SUMMARY AND CONCLUSIONS
Moisture percentages and dry weights of developing
caryopses of several strains of grain sorghum were deter-
mined in 1958 and 1959. Final dry weights of kernels
were determined in 1959.
Decreases in moisture content from 12 to 25 days were
similar in both years. From 25 to 37 days after blooming
decreases in moisture percentages were greater in 1958 than
in 1959.
Percentages of moisture in 7078 and A3197 generally
were higher than in Martin and Shallu at comparable days
after blooming. RS 610 and E 5<$a had similar moisture
percentages until 25 days after blooming but E 56a dried
faster than RS 610 from 25 to 40 days after blooming.
The amount of water in the caryopses either accumulated
or remained fairly constant until 16 to 20 days after bloom-
ing after which actual moisture losses occurred.
Dry matter accumulation generally continued over a
longer period in 1959 than in 1958. Maximum dry weights
were reached in 24 to 27 days in 1958 in all strains except
RS 610 which continued to accumulate dry matter until
about 35 days. In 1959 maximum dry weights were reached
in all entries in 31 to 35 days.
Dry weights of kernels appeared to be the most desirable
method to use in determining maximum dry weight accu-
mulation. The percentages of final dry weights were slightly
biased and usually indicated maturity 2 to 5 days earlier
than maximum dry weight curves.