E. L.

PINNELL
E. H. RINKE
and
H. K. HAYES
University of Minnesota

Chapter 24

Gamete Selection for

Specific Combining Ability*

Gamete selection as a breeding method was designed for more efficient

sampling of open-pollinated varieties. It was suggested by Stadler in 1944.
The method was outlined in detail by Stadler (1945) and preliminary data
presented. Hayes, Rinke, and Tsiang (1946) proposed that the same technic
could be used to select gametes from such sources as synthetic varieties,
single or more complex crosses, and inbred lines. They discussed gamete
selection in its relation to the improvement of a particular double cross
combination.
MATERIALS AND METHODS
In 1945, three double crosses, Minhybrids 602,607, and 406, were selected
for a method study in gamete selection. Single cross performance data shown
in Tables 24.1 and 24.2 indicate that A344 is low in combining ability in
Minhybrids 602 and 607, and that the same is true for inbreds A25 and A73
in Minhybrid 406.
A344 was crossed to Minnesota #13 (Morris strain) and to 8 inbred lines
namely, Oh51A, A97, 1234, A315, A348, A367, A396, and Ill. 4226 as sources
of gametes. The inbreds were selected because of their diversity of origin
and good general combining ability. In addition, A367 had yielded well in
specific tests with A357, A385, and A392. A315 and A348 had performed well
in crosses with A392. The remaining five inbreds had not been crossed to
A357, A385, or A392 in previous years. A25 was crossed with Golden King
and A73 with Murdock. In 1946, individual F1 plants of these crosses were
* Paper No. 2591 of the Scientific Journal Series, Minnesota Agricultural Experiment
Station.
378
 GAMETE SELECTION FOR SPECIFIC COMBINING ABILITY 379
selfed and crossed to the opposing single cross parent of the double cross.
Table 24.3 gives the number and type of test crosses made.
Approximately 60 inbred X variety F1 plants fromeach of the three sources
were selfed and crossed to the testers. Selection of the better F 1 plants at
harvest reduced the number for further study to 35, 32, and 38 as listed in
Table 24.3. Experiments 1 and 2 were tested in separate randomized blocks
TABLE 24.1
PERFORMANCE OF A344 AND A334*

Av. OF CROSSES

INBREDS

% Moist. Bu.

A344XA357, A392, A385 ..... 28.3 69.8

A334XA357, A392, A385 ..... 30.1 84.3

* As indicated by average moisture and yield when crossed in

non-parental single cross combinations of Minhybrid 602 (A344 X
A334J (A357 XA392) and Minhybrid 607 (A344XA334) (A357 X
A385).

TABLE 24.2
PERFORMANCE OF A25, A334, A73, AND A375*

. Av. OF CROSSES

INBREDS

% Moist. Bu.

A25 XA73, A375 ..... . . . . ... 24.6 76.2

A334XA73, A375 ....... ... 24. 7 79.4
A73 XA25, A334 ........ . . . 24.6 74.8
A375XA25, A334 .......... 24.7 80.8

* As indicated by moisture and yield in non-parental single

crosses of Minhybrid 406 (A25XA334) (A73XA375).

using two replicates at each of three locations in central Minnesota. Data

from the two testers were averaged to give a total of 12 replicates as a test
for each gamete. Experiments 3 and 4 were also grown in randomized blocks
with three replicates at each of four locations in southern Minnesota. One
location of experiment 3 was discarded and one replicate of experiment 4 was
abandoned prior to harvest.
On the basis of yield trial results in 1947, gametes were selected from all
varieties and from two inbreds, for use as parents in the development of
new lines by straight selfing and by backcrossing to the sampler inbred.
Study of the performance indices and agronomic characters of the test
crosses led to the selection of three Golden King gametes and four Murdock
380 E. L. PINNELL, E. H. RINKE, AND H. K. HAYES
gametes as higher in general desirability than the sampler inbred parent. In
addition, gametes of low yield potential but of relatively satisfactory agro-
nomic characters were selected from both varieties, three from Golden King
and two from Murdock.
F 2 populations were grown in 1948 from the selfed ears of the twelve F1
plants selected in the above manner. Visual selection was practiced for desir-
able plant and ear characters, and these individual plants were crossed to
the appropriate single cross tester. Remnant seed of the test crosses of the
selected F 1 plants was used to make a direct yield comparison in 1949 with
the test crosses of the selected F 2 plants. Two randomized block experiments
were made at each of three locations in southern Minnesota with three
TABLE 24.3
GAMETE SOURCES, TESTERS USED, AND TEST CROSSES
MADE IN SELECTING GAMETES FOR IMPROVEMENT
OF INBREDS A344, A25, AND A73

Experi- Number
ment Inbreds Gamete Source Tester of
Number Crosses

1 ........ A344 Morris 13 gametes A357XA392 35

A344 Inbred gametes A357XA392 8
A344 A357XA392 1
2 ........ A344* Morris 13 gametes A357XA385 35
A344 Inbred gametes A357XA385 8
A344 A357X:A385 1
3 ........ A25 Golden King gametes A73XA375 32
A25 A73XA375 1
4 ........ A73 Murdock A25XA334 38
A73 A25XA334 1

* Same plants as in experiment 1.

replications per location. One experiment consisted of the crosses of 6 F 1

plants (remnant seed) from A25 X Golden King, and 34 F 2 plants by the
tester compared with the cross of A25 X tester. The other included test crosses
of 6 F1 and 37 F2 plants from A73 X Murdock gametes in comparison with
A73 X tester.
EXPERIMENT AL RESULTS
The Morris strain of Minnesota 13 has been a very outstanding open-
pollinated variety in Central Minnesota for many years. Yield trial data
from plants of A344 X Morris 13 crossed with the testers show that a large
proportion of Morris 13 gametes have higher yield potential than A344.
Table 24.4 gives the distribution of yield and moisture data obtained from
thirty-five test crosses of A344 X Morris 13. Sixteen of the thirty-five gametes
gave test-cross yields significantly in excess of A344 X tester. Although not
higher in yield, five other gametes may be considered superior to A344 be-
 GAMETE SELECTION FOR SPECIF1C COMBINING ABILITY 381
cause of their significantly lower moisture content at harvest and yields not
significantly different in test crosses from A344 X tester.
The eight inbreds tested as possible sources of germ plasm for the improve-
ment of A344were A97, A315, A348, A367, A396, Oh51A, Ia234, and Ill.4226.
TABLE 24.4
DISTRIBUTION OF PER CENT MOISTURE AND
YIELD OF 35 F1 PLANTS OF A344 X MORRIS 13
CROSSED TO S.C. TESTERS A357 X A392 AND
A357 X A385*

% Ear +l 3 5 1 _,(MeanA344
Moisture _ 1 2 3 6 6 2 X tester)
-2 3 2 2

-2 -1 +1 +2 +3
Yield
(Mean A344 X tester)
* Classes separated by one or more than one LSD (5 per cent) from the
mean of A344 Xtester.

TABLE 24.5
DISTRIBUTION OF PER CENT MOISTURE AND YIELD OF
TEST CROSSES OF 8 INBRED LINES AS SOURCES OF GAM-
ETES. CROSSES ARE OF THE TYPE (A344 X INBRED) X TEST-
ERS*

+2 1 1
+1 2 1 1
% Ear ->(Mean of A344
Moisture _ 1 1 1 X tester)

-2 -1 +1 +2 +3 +4 +5
Yield
(Mean of A344 X tester)
* Classes separated by one or more than one LSD (5 per cent) from the mean of A344Xtester.
The distribution for yield and moisture of test crosses of A344 X Inbred is
given in Table 24.5. Three inbreds, Ia234, A396, and A97, demonstrated
yield potential significantly higher than A344.
Golden King and Murdock are old, well-adapted varieties formerly
grown extensively in southern Minnesota. Test crosses of thirty-two Golden
King gametes X A25 and thirty-eight Murdock gametes X A73 are shown in
Tables 24.6 and 24.7, respectively.
No gamete from Golden King exceeded A25 significantly in test-cross
yields. However, eight not different in yield were significantly earlier, and
382 E. L. PINNELL, E. H. RINKE, AND H. K. HAYES
are considered superior in yield performance on the basis of maturity. Eight
gametes from Murdock demonstrated yield potential greater than A73, as
indicated by significantly higher yields in crosses. In addition, fourteen
gametes not different from A 73 in yield were significantly earlier in maturity.
The proportion of promising gametes extracted from the three varieties
and the eight inbreds is summarized in Table 24.8. About 25 per cent of the
total number tested were superior to the sampler inbred in yield potential.
Another 25 per cent would be considered desirable parents because they had
a yield potential equal to and a maturity potential which was significantly
earlier than the sampler inbreds.

SELECTION OF GAMETES AS PARENTS AND TESTS

OF F2 PROGENIES
Years of testing at Minnesota have led to the conclusion that, in general,
there is a direct relation between yield and moisture content at husking
among hybrids of equal genetic desirability. On this basis the combining
ability of inbred X gamete plants was determined by considering both yield
and moisture percentage at husking. They were effectively placed on a com-
parable basis by calculation of a performance index using the test cross of
the sampler inbred as 100 for both ear moisture and yield. For example, if
the moisture percentage of an A25 X Golden King plant (in test cross) was
93.5 as compared with A25 X tester, and its yield was 106.5 per cent, its per-
+
formance index would be 13. Where the comparative moisture percentage
is higher than the yield percentage the index becomes a negative value.
The performance indices for the selected gametes for both 1947 and 1949
trials and similar data for the F 2 plant progeny tests appear in Tables 24.9
and 24.10. The tests of F2 plants from gametes of both high and low yield
potential were made as explained in "Materials and Methods". This wa scar-
ried out by selfing selected F 2plants and also crossing them with the appro-
priate tester, and again comparing the results with the crosses of the appro-
priate inbred with the same tester. Agreement between the two tests of the
gametes was very good except for Murdock gametes numbered 12 and 49.
On the average there was good agreement between F1 and F2 progeny
performance. Tables 24.11 and 24.12 show that there is evidence of segrega-
tion for yield factors within almost all of the F 2 families tested.
Mean performance of the F 2 progeny from the high testing gametes was
little different from the F 1 for either ear moisture or yield (Table 24.13).
However, the F 2 progeny from the gametes of low yield performance ex-
ceeded the F 1 parent plant in yield performance on the average. This indi-
cates that visual selection of plants within the F 2 populations was more
effective among the progenies arising from the gametes of low yield per-
formance than for those F 2plants that were selected from high performing F 1
crosses (gamete X inbred).
 TABLE 24.6
DISTRIBUTION OF PER CENT MOIS-
TURE AND YIELD OF 32 F1 PLANTS
OF A25 X GOLDEN KING CROSSED
TO A73 X A375*

+2 1
+1 3 2
% Ear
Moisture -1 2 8 5
-2 3 5 3

-2 -1 +1 +2
Yield
* Classes separated by one or more LSD (5 per
cent) from the mean of A25 Xtester.

TABLE 24.7
DISTRIBUTION OF PER CENT MOIS-
TURE AND YIELD OF 38 F1 PLANTS
OF A73 X MURDOCK CROSSED TO
A25 X A334*

+3 1
+2
+1 1 1
% Ear
Moisture -1 5 7 3
-2 1 2 9 3
-3 2 1 2

-2 -1 +1 +2
Yield
* Classes separated by one or more LSD (5 per
cent) from the mean of A73Xtester.

TABLE 24.8
NUMBER AND SOURCE OF GAMETES SUPERIOR
IN PERFORMANCE TO SAMPLER INBREDS

Higher in
Not Dif-
Total Yield and
Source of Sampler ferent in
Gametes as Early or
Gametes Inbred Yield but
Tested Earlier in
Earlier
Maturity

Morris 13 ...... A344 35 16 5

8 inbreds ...... A344 8 2* 0
Golden King ... A25 32 0 8
Murdock ...... A73 38 8 14
Total. ..... .......... 113 26 27

* An additional gamete higher in yield was also later in maturity.

383
 TABLE 24.9
PERFORMANCE INDICES OF TEST CROSSES OF
SELECTED F, PLANTS AND F2 PROGENY
FROM A25 X GOLDEN KING

PERFORMANCE INDEX

NUMBER
F1 PLANT
1947 1949 OFF,
NUMBER
PLANTS

F, F, F,

19 H* .... +11 +19 +2s 5

20H ..... +14 +9 +14 7
36H ..... +9 +16 +11 7
SL ..... -11 - 3 +s 7
29L ..... -11 - 1 -0 1
46L ..... - 5 + 1 +2 7

*H = high- and L = low-performing gametes.

TABLE 24.10
PERFORMANCE INDICES OF TEST CROSSES OF
SELECTED F1 PLANTS AND F2 PROGENY
FROM A73 X MURDOCK

PERFORMANCE INDEX

NUMBER
F, PLANT
1947 1949 OFF,
NUMBER
PLANTS

F, F, F,

12H ........ +18 - 3 - 6 8

14H ........ +2s +33 +27 5
49H ........ +20 + 1 + 5 6
S0H ........ +29 +19 +18 7
6L ........ -10 -23 -10 5
35L ........ - 4 -24 - 7 6

384
 TABLE 24.11
FREQUENCY DISTRIBUTION OF PERFORMANCE IN-
DICES OF F 2 PROGENY PLANTS FROM A25 X
GOLDEN KING AROUND MEAN PERFORMANCE
OF A25 X TESTER

INDICES FOR F2 PLANTS

F1 PLANT
NUMBER
-15 -5 +s +1s +2s +JS
---------------
19 H ..... 1 2 2
20H ..... 2 3 2
36H ..... 3 4
5L .... 3 2 1 1
29 L ..... 1
46L ..... 1 1 5

TABLE 24.12
FREQUENCY DISTRIBUTION OF PERFORMANCE
INDICES OF F 2 PROGENY PLANTS FROM A73 X
MURDOCK AROUND MEAN PERFORMANCE OF
A73 X TESTER

INDICES FOR F, PLANTS

F1 PLANT
NUMBER
-15 -5 +s +1s +2s +JS
---------------
12 H ..... 5 1 1 1
14H ..... 1 2 2
49H ..... 2 2 2
50H ..... 2 2 2 1
6L ..... 2 2 1
35L ..... 2 3 1

TABLE 24.13
COMPARISON OF F 1 PLANTS AND THEIR F2 PROGENY IN 1949 TEST
CROSS PERFORMANCE FOR EAR MOISTURE, YIELD,
AND PERFORMANCE INDEX

No.
PERFORMANCE
PLANTS EAR MOISTURE % YIELD IN Bu.
INDEX
TESTED
PARENTS

F1 F, F1 F, Diff. F1 F, Diff. F1 F, Diff.

- - -- --
A25XG. King H .... 3 19 20.3 20.2 -0.1 54.0 54.5 +o.5 +14.4 +15.8 + 1.4
A25XG. KingL ..... 3 15 23.3 23.8 +o.5 52.9 56.3 +3.4** - 2.1 + 3.3 + 5.4
A73XMurdock H .... 4 26 21.8 21.1 -0. 7* 56.0 55.0 -1.0 +12.5 + 9.6 - 2.9
A73XMurdock L .... 2 11 22.8 22.4 -0.4 41. 7 48.7 +7.0** -23.6 - 8.6 +15.0

* Exceeds 5% point of significance. """Exceeds 1% point of significance.

385
386 E. L. PINNELL, E. H. RINKE, AND H. K. HAYES

DISCUSSION
Almost 50 per cent of the gametes studied showed evidence of having
combining ability in excess of the sampler lines. The gametes chosen as par-
ents appear to furnish a desirable source of germ plasm for a selection pro-
gram designed to improve the yield potential of A344, A25, and A73 in
combining ability in specific crosses.
Where a high combining varietal gamete is chosen for an inbred selection
program, the F1 plant of which it is a parent represents a highXlow type of
cross so far as combining ability is concerned. To the extent that such F 1
plants are comparable to crosses of inbreds, the breeding results should be
similar to those from crosses of inbreds differing in combining ability. At
Minnesota (Hayes and Johnson, 1939), crosses of highXlow combiners have
given F 5 lines ranging from high to low, but with a good proportion of high
combiners.
Whether selection of gametes should be followed by test controlled selec-
tion in the F 2 is an important question. In these studies more than 50 per
cent of the F 2 plants from high combining gametes tested at least ten per-
formance index units higher than the sampler inbred. Thus without further
test crosses, the chances of choosing high combining F2 plants would still
have been very good. The number of F2 plants that could be handled in test
crosses was quite limited. This may make for greater difficulty in recovering
or improving the agronomic type of the sampler lines. It was very evident
from field observations that the proportion of agronomically desirable Fa
lines appeared lower than usually found from crosses of highly selected
inbreds.
The greater effectiveness of visual selection among the F 2 progenies of the
low testing gametes is at this stage only an interesting development. Only a
small proportion of the plants arising from the low testing gamete parents
exceeded the sampler inbred in performance by a significant amount.
It was not possible to determine by visual examination which F2 popula-
tions were derived from high gametes and which from low gametes, although
there were wide differences in plant type between populations.
Gametes from eight inbred lines compared fairly well with varietal
gametes from Morris 13, in offering promising sources of germ plasm for the
improvement of specific combining ability of A344. Where a breeder has
available large numbers of inbred lines of diverse origin the use of test
selected inbred parents rather than varietal gamete parents may be the more
feasible approach. Selection for characters other than yield would presum-
ably be done more economically. The same advantage can be claimed for the
use of complex crosses of inbreds. On the other hand, utilization of varietal
gametes in improvement work does not "use up" inbreds so far as their
combination in hybrids for commercial use is concerned. Lines recovered
 GAMETE SELECTION FOR SPECIFIC COMBINING ABILITY 387
from crosses of inbreds may be more restricted than their inbred parents in
commercial use because of relationship. It seems probable to the writers that
the method of gamete selection is worthy of considerable use for further
selection of material from open-pollinated, desirable commercial varieties.
Studies of lines recovered from selected varietal gametes will have to be
carried to F 5 or later generations to determine if the large amount of out-
cross testing is justified economically. The writers would like to emphasize
the importance they attach to method studies of the type presented here.
New ideas in breeding must be explored constantly if continued progress is
to be made in corn improvement.

SUMMARY
Since 1945, a program has been underway at Minnesota to attempt im-
provement of Minhybrids 602,607, and 406 by the method of gamete selec-
tion. The hybrid pedigrees are respectively: (A344XA334) (A357XA392),
(A344XA334) (A357XA385), and (A25XA334) (A73XA375). Detailed
studies of the non-parental single crosses among the inbred parents of each
hybrid led to the conclusion that A344 in Minhybrids 602 and 607, and A25
and A73 in Minhybrid 406 were low in combining ability.
A344 was crossed to the Morris strain of Minnesota 13 and to eight in-
breds of diverse origin. A25 was crossed to the Golden King variety and A73
to Murdock. (InbredXgamete) (tester) crosses were made using the oppos-
ing single cross parents as testers. These were compared with the appro-
priate cross of inbredXtester. Yield trial performance was obtained from a
total of 113 gametes, 35 from Morris 13, 8 inbreds, 32 from Golden King,
and 38 from Murdock.
Sixteen gametes from Morris 13, three from the inbreds, and eight from
Murdock gave significant increases in yield over the test crosses of the checks
A344 and A73. Five gametes from Morris 13, eight from Golden King, and
fourteen from Murdock were not significantly different in yield but were
significantly earlier so that yield performance could be considered better
than the checks on the basis of ear moisture at harvest. These varieties and
the three high testing inbreds thus appear to be good sources of gametes for
improving the relatively low performing inbreds in specific combining ability
for yield.
Both high and low testing varietal gametes were selected for use in a study
of the development of new inbreds. From the crosses, A25 X Golden King and
A73XMurdock, selected F2 plantsXthe appropriate tester were compared
with the progeny of their F1 parental plants when crossed on the same tester.
While there was excellent agreement, on the average, for combining ability
in the F1and F 2, there was evidence of segregation for combining ability from
almost all of the twelve F 2families which were studied. Visual plant selection
within the F 2 generations appeared to be effective in increasing yield per-
388 E. L. PINNELL, E. H. RINKE, AND H. K. HAYES
formance of the plants from the low testing gametes, but appeared to have no
effect in further increasing the yield performance of the F 2 plants from the
high testing gametes.
The economic feasibility of F 2 plant testing in a gamete selection program
is questioned.