Field Crops Research E:LSEVIER Field...
Field
Crops
Research
E:LSEVIER
Field Crops Research 36 (1994) 125-131
d
Consistency of genotypic ranking for carbon isotope
discrimination by cowpea grown in tropical and subtropical zones
A.E. HalP*, $ Thiawb, D.R. Krieg”
“Department of Botany and Plant Sciences, Qniversity of California, Riverside, CA 925;?1-0124, USA
bISRA/Centre National de Recher’ches Agronomiques, BP 53, Banbey, Senegal
‘Department ofAgronomy, Texas $ech Universiry, Box 42122, Lubbock, TX, USA
(Accepte4 16 November 1993)
Abstract
Measurements of plant composition of stable carbon isotopes ( ‘3C/‘2C) cari be used to estimate the extent that plants
discriminate (3) against the heavier carbon isotope. Theoretical and experimental studies have shown that d cari be negatively
correlated with transpiration efficiency ( W = seasonal total biomass production/seasonal transpiration) in C3 plants. Conse-
quently, Select ion for low A ( and high W) may be useful in breeding Ca plants for some water-limitedenvironments. Development
of effective breeding methods depends upon the extent bf genotype X environment interaction (G XE) as it influences the
consistency of genotypic ranking across environments.
Sets of cowpea cultivars and lines were grown under çontrasting water supply regimes in subtropical zones in Riverside,
C’alifomia, and Lubbock, Texas, and in tropical zones in Senegal. Leaf carbon isotope composition was measured and significant
genotypic differences in A were observed in most mals with little G X E for mals within these regions. Genotypic rankings were
quite consistent between wetter and drier environments artd different years within these regions with significant correlation
coefficients for genotypic means in most cases. However, when comparing A values between the twlo subtropical zones (Califomia
versus Texas), correlation coefficients for genotypic means only were moderate and not significant. Comp<arisons of A values
between the tropical zone (Senegal) and the two subtropical zones indicated no consistency in genotypic ranking, and correlation
coefficients for genotypic comparisons were very small.
Apparently.. G X E for A would not necessarily constrain cowpea breeding programs which aim at developing improved
cultivars for specific target production regions, such as semiarid tropical Senegal. However, cowpea performance, with respect
to A and W, uould not be transferable to radically different production zones where attainment o:f high W may require different
s#:ts of genes.
&ey words: Cowpea; Transpiration efficiency; Variety tria]; Vign&
1. Introduction
( W = seasonal total biomass production/seasonal
tran-
spiration) could be increased without associated neg-
Yields of crops for water-limited environments
ative effects. Richards and Condon ( 1993) provide a
would be enhanced if transpiration efficiency
discussion of possible associated negative effects, such
-~
as slow canopy growth and associated increases in soi1
*Corresponding author.
evaporation. Direct selection for W in field conditions
Elsevier Science R.V.
SXDl0378-4290(93)E0068-9
1 2 6
A.E. Hall et al. /Fi& Crops Research 36 (1994) 125-131
would be very difficult due to the need to extract and
tant trop for rainfed semiarid tropical regions of Africa
measure root biomass, and estimate soi1 evaporation
where water limitations cari substantially reduce yield
and total water-use SO that transpirational water-use cari
(Hall and Patel, 1987; Thiaw et al., 1993). Studies
be estimated. However, theoretical and empirical S<ud-
conducted in southem Califomia have discovered cow-
ies indicate that carbon isotope discrimination (A) may
pea accessions with significant differences
in A (Hall
be effective for indirectly selecting for W (i.e., reviews
et al., 1990) that could confer substantial differences
and case studies in Ehleringer et al., 1993), and A cari
in W (Ismail and Hall, 1992)) and genotypic ranking
be estimated in field nurseries from measurements of
for A was consistent over different levels of drought
“C/ ‘.7C in plant tissues. Discrimination against ‘me
and years (Hall et al., 1992). Consistency of genotypic
heavier isotope ( 13C) occurs during the diffusion and
ranking for A over a very wide range of environments
fixation of C02. According to the theory for C, plants,
has not been evaluated for cowpea or most other crops.
A and W both depend upon the concentration of CO,
Studies were conducted to: ( 1) determine whether a
within leaves (Farquhar et al., 1982). Increases in
set of contrasting cowpea genotypes exhibit significant
interna1 photosynthetic capacity or partial stomatal clo-
differences
and any G XE in A measured in leaves of
sure cause decreases in the concentration of CO, within
plants grown in two tropical locations in semiarid Sene-
leaves and concomitant increases in W and decreases
gal; and (2) compare genotypic ranking with values of
in A. Consistent with this theory, plants growing in the
A already obtained for thie same genotypes in subtrop-
same aerial environment have exhibited a negative lin-
ical regions in southem Califomia (Hall et al., 1992)
ear regression between W and A (Farquhar and
and the high plains of Texas.
Richards, 1984). Development of effective breeding
m e t h o d s depends u p o n t h e e x t e n t o f geno-
type X environment interaction (G X E) for A as it
‘2. Materials and methods
influences the consistency of genotypic ranking (Hall
et al., 1994). Selection would not be effective if gen-
Experiments were conducted in a semiarid tropical
otypic ranking for a trait were to vary from year to pear
region of Senegal having a single summer rainfall sea-
for the same genotypes.
son on the Bambey and Louga research stations of the
Some studies have not detected significant G XE.
Institut Senegalais de Recherches Agricoles in 1991
Sixteen peanut genotypes were evaluated in 10 aon-
and 1992. These stations have similar air temperatures
trasting subtropical and tropical environments in
during the growing season with daily maxima averag-
Queensland, and G X E was not significant for A, but
ing 33°C and daily minima averaging 24°C (Hall and
was highly significant for kemel yield (Hubick et al.,
Patel, 1987). Louga is located on the arid boundary of
1988). Studies with 21 or 23 genotypes of barley over
the semiarid zone with an average annual rainfall of
2 years in two environments in the United Kingdbm,
269 mm in recent years compared with 471 mm at
and three locations in Syria, gave positive and sigmifi-
Bambey (Hall and Patel, 1987).
tant correlations among genotypic values of A for 20
During the summer of 199 1,42 land races, cultivars,
out of 25 comparisons
with different pairs of environ-
and breeding lines of cowpea were grown at Bambey
ments (Craufurd et al., 199 1) .
( 14”42’N 16”28’W) on a deep, slightly leached, trop-
In other studies, significant G X E has been detecited.
ical ferruginous soil, called a “Dior” soil, which has a
Studies with about 20 genotypes in a range of field
volumetric moisture content at field capacity of 0.16.
environments across the southem Australian wheat-
Fertilizer was broadcast and incorporated at the rate of
belt ( Condon and Richards, 1992) indicated significant
150 kg/ha of 6:9:8 (N::P:K). Seed was sown at the
1;
G XE in rnany cases, but it was generally a minor
beginning of the rainy season on 12 July in rows 50 cm
component of variation in A compared with the geno-
apart with 50 cm between plants for spreading geno-
typic variation. For common bean (Phaseolus uulgaris
types and 25 cm for erect genotypes, consistent with
L. ) , significant G X E has been observed for genotypes
current recommended practices. The experimental
grown under different watering regimes (White,
design consisted of four :randomized complete blocks.
1993).
Individual plots consisted of four rows of plants, 4 m
Cowpea (V@a unguiculata L. Walp.) is an impor-
long. Samples consisting of three leaflets were taken
A.E. Hall et al. /Field Çrops Research 36 (1994) 125-131
127
fi om the eighth node of three plants in each plot on the
plains of Texas at the research station of Texas Tech
33th day after sowing. The nine leaflets taken fro@
University at Brownfield in Terry County in a loamy
e,lch plot were pooled and dried, and carbon isotope
fine-Sand soil. Similar experimental methods were used
composition was measured as described in Hall et al.
in both years except that eight cowpea genotypes were
(1992).
evaluated in 198X and six cowpea genotypes were eval-
During the summer of 1992, nine cowpea genotypds
uated in 1989. The experimental design was a split plot
u’ere grown in two experiments at Bambey and Louga
with three replications. The main plots were two levels
( 15”36’N 16’13’W). The soi1 at Louga is sandy with
of trickle irrigation and the sub-plots were the different
a small clay content (3%) and a volumettic moistuqe
genotypes. Individual plots consisted of four rows, with
c$>ntent at field capacity of 0.08. Seeds were sown ht
75 cm between rows, and 12 m long. A fertilizer appli-
the beginning of the rainy season on 17 July in Bambey,
cation of 50 kg N ha-’ and 20 kg P ha-] was made
a Id 10 August in Louga, and the plants were growp
prier to sowing, and additional N was provided in the
uiing the same methods as in 1991, except that indj-
irrigation water at a rate of 1 kg N ha - ’ mm- ‘. A well-
vtdual plots consisted of three rows, 6 m long. Sampljs
watered treatment was established which received
cl,nsisting of three leaflets were taken from the fifth
maximal water needs every 3 days. A drier treatment
nade of four plants on the 28th day after sowing. The
was established where plants were provided with max-
12 leaflets taken from each plot were pooled, dried, and
imal water needs for the first 35 days, after which irri-
a lalyzed for carbon isotope composition as described
gation was terminated. Samples consisting of three
in Hall et al. ( 1992). Forage and grain yield weqe
leaflets were taken from the eighth node of three plants
drtermined on the tenter row of each plot at the end df
in each plot on the 60th and 58th day after sowing in
the growing season.
1988 and 1989, respectively. The nine leaflets taken
Experiments were conducted in the summers of 19S;X
from each plot were pooled and dried, and carbon iso-
a Id 1989 in a semiarid, subtropical region in the high
Tible 1
Clrbon isotope discrimination measured in leaves of cowpea genoqypes grown at Bambey, Senegal, in 1991. Genotypic effects were very highly
.
siqxhcant wlth an LSD,,.,,,
- -
0 4%”
Rank
Genotype
A
Rank
Genotype
A
@‘oo)
(%Cl
i
480
2 2 . 3
2 2
Baye Ngagne
2 1 . 7
,
Tn 88-63
2 2 . 2
2 3
Vita 7
2 1 . 7
3
Kaedi Gris-blanc
2 2 . 2
2 4
Ndouc.
2 1 . 6
.l
4R-0267- 1 F
22.1
2 5
C B 5
2 1 . 6
i
ER- 1
2 2 . 0
2 6
Ndiaga Aw
2 1 . 6
0
Ndiambour
22.0
2 7
ER-7
2 1 . 5
7
473
2 1 . 9
2 8
Diongama
2 1 . 5
‘I
5x-57
2 1 . 9
2 9
Chino Ml
2 1 . 5
0
572
2 1 . 9
3 0
CBS-El
2 1 . 4
10
601
2 1 . 9
31
Bambey 21
21.4
I l
Mougne
2 1 . 9
3 2
8517
2 1 . 4
1.3
482
2 1 . 9
3 3
TVx 1841.OlE
2 1 . 4
1.i
Mouride
2 1 . 9
3 4
Prima
2 1 . 4
I-1
514
2 1 . 9
3 5
WR 237A
21.4
1:;
TVu 374
2 1 . 9
3 6
CB46
2 1 . 3
10
Melakh
2 1 . 8
3 7
TVx 309.1G
2 1 . 3
1”
475
2 1 . 8
3 8
58-111
2 1 . 3
18
TVx 75-4
2 1 . 7
3 9
66-20
2 1 . 3
1’)
590
2 1 . 7
40
Bambey 29
2 1 . 3
20
TVx 1836-015
2 1 . 7
41
Casa 16
2 1 . 3
21
484
2 1 . 7
42
IAR 16-96
21.1
--_~
128
A.E. Hall et 01. /Field Crops Research 36 (1994) 125-131
T a b l e 2
Carbon isotope discrimination measared in leaves of cowpea genotypes grown at Bambey and Louga, Senegal, in 1992
-
Genotype
Bambey
Louga
_-
-
Rank
A
Rank
A
i%)
60)
-
-
58-57
(3)
2 0 . 2
(1)
2 2 . 8
Vita 7
(4)
2 0 . 2
(2)
2 2 . 7
Bambey 2 1
(1)
2 0 . 3
(3)
2 1 . 9
C B 5
(2)
2 0 . 3
(4)
2 1 . 9
4R-0267-IF
(5)
20.1
(5)
2 1 . 8
CB46
(6)
2 0 . 1
(6)
21.4
Prima
(7)
20.1
(8)
2 1 . 0
UCR 237A
(8)
‘ 1 9 . 8
(7)
2 1 . 2
TVx 309- 1 G
(9)
1!9.5
(9)
2 0 . 9
Mean
20.1
2 1 . 7
I-SJ4.m
0 . 7
Mean Squares
Genotype effects
1.53
**
Location effects
50.72
***
G X E effects
0.56
n s
Errer
0.51
LSD tests differences betw&n genotypic
-
.-
means. ** a.nd *** = significant at P<O.Ol and 0.001, respectively. and ns is not signitïcant at
P=O.38.
T a b l e 3
versity of Califomia. These experiments have been
Hay and grain yield of cowpea genotypes grown at Bambey, Senegal,
described previously (Hall et al., 1992). In 1987, cow-
in 1992
pea genotypes were grown under well-watered and
Genotype
Hay
Grain
Total
stored-soil-moisture conditions in a split-block design
f kg/ha)
Wha)
(kdha)
with four replications. The main blocks consisted of
weekly, well-watered furrow irrigation and a treatment
58-57
4433
1 9 7 3
6406
t.hat received no rain or irrigation after plant establish-
TVx 309-1G
3 1 5 1
21:20
5 2 7 1
ment. The sub-plots consisted of the different geno-
4R-0267- 1 F
2278
2197
4475
Vita 7
2033
2 0 8 1
4114
types. In 1988 a third, intermediate irrigation treatment
Bambey 2 1
2 2 9 1
15.35
3826
was included that was irrigated on altemate furrows
LJCR 237A
1503
18:29
3332
every 3 weeks during the growing season. Samples
Prima
1 8 4 7
1 1 0 5
2952
consisting of three leaflets were taken from the eighth
CB46
1 1 3 9
1 4 1 2
2 5 5 1
node of three plants in each plot 5.5 and 62 days after
.
C B 5
1 1 1 6
13.54
2520
sowing in 1987 and 1988, respectively, and processed
Mean
2205
17.34
3939
to determine A as described in Hall et al. ( 1992).
LSQ,.m
937
406
3. Results
tope composition was measured as described in Hall et
al. (1992).
The screening tria1 with 42 diverse cultivars and lines
Experiments were conducted in the summers of 1987
at Bambey, Senegal, in 1991 received 347 mm of rain.
and 1988 in a semiarid, subtropical region of southem
There was little plant-available moisture in the soi1 at
Califomia at the Riverside research station of the Uni-
the beginning of the cropping season. This rainfall
A.E. Hall et al. 1 Field Crops Research 36 (1994) 125-131
1 2 9
Table 4
C>wpea genotypic ranking for carbon isotope discrimination meaaured in leaves in California and Senegal
G motype
Riverside, California
Senegal
1 9 8 7
1 9 8 8
1991
1 9 9 2
-
-
Wet” Dry
Wet
Med. Dry
Mean
Bambeyb
Bambey
Louga
Mean
Rank
Rank
Rank
Rank
Rank
(%o)
Rank
Rank
Rank
C%@l
-
CB46
1
2
2
1
1
19.8
8
7
6
20.9
Rima
2
1
3
2
2
19.7
6
6
8
20.8
CB5
3
3
1
3
'3
19.5
4
2
4
21.2
431-0267-1F
4
7
4
4
4
19.0
1
5
5
21.4
Bunbey 21
5
5
6
5
7
18.8
5
1
3
21.2
511-57
7
4
5
6
5
18.8
2
4
1
2 1 . 6
TVx309-IG
6
6
7
7
6
18.7
9
9
9
20.6
Vita 7
8
8
9
8
8
18.1
3
3
2
21.5
UCR 237A
9
9
8
9
9
17.9
7
8
7
20.8
‘Wet, Med., and Dry denote well-watered, intermediate, and limit&irrigation treatments, data from Hall et al. ( 1992).
‘Itambey and Louga are locations in Senegal with intennediate and limited supplies of rain, respectively.
Table 5
(Hall and Patel, 1987). Genotypic differences in leaf
C~~~clation coefficients for genotypic mean values of carbon isotope discrim-
A were very highly signilïcant (Table 1)) but the range
ination measured in leaves of cowpea
-
of values ( 1 .2%0) was smaller than the range obtained
G%nparisons
r
nP
at Riverside, California (2.1%0) with a set of 50 gen-
otypes including 13 of the same ones (Hall et al.,
W ithin rqions
1990). Maximum differences in W cari be estimated
Sonthem Califomia
using the following equation:
Riverside Wet vs Dry 1987
0 . 9 0 9 9 <O.ool
Riverside Wet vs Dry 1988
0.916 9 <O.Obl
Riverside 1987 vs Riverside 1988
0 . 9 9 4 9 <O.chl
W,/W,=(b-d-A,)/(b,-d-d,)
(1)
Jmegal
Bambey 1992 vs Louga 1992
0 . 7 0 0 9
0.05
where (b - d) are parameters for the inherent discrim-
Bambey 1991 vs Bambey 1992
0.477 9 ns
ination associated with carboxylation (b) and other
Htgh Plains of Texas
metabolic processes (d). The value of (b-d) was
Lubbock Wet Y:; Dry 198X
0.907 8
o.o,
estimated from the empirical studies of Ismail and Hall
Lubbock Wet vs Dry 1989
0.317 6 ns
( 1992) as being 24.3%0. The genotype with the lowest
A (IAR 16-96) in Table 1 is estimated to have 58%
Bntwcen regions
Senegal (1991 -r 1992) vs Riverside
-0.125 9 ns
higher W than the genotype with the highest A (#480).
(1487+ 1988)
This is similar to the range of difference in W estimated
for the study at Riverside, California (Hall et al., 1990)
Senegal(l991+ 1992) vs Lubboçk (1988)
-0.122 6 ns
of 62%.
Senegal ( 1991 -C 1992) vs Lubbock ( 1989)
-0.098 6 ns
At Bambey, Senegal (Table 1) , there was a tendency
for well-adapted local cultivars to have high leaf A
Lubbock (1988) vs Riverside (1987+ 1988)
0.569 8 ns
Lubbock ( 1989) vs Riverside ( 1987+ 1988)
0.518 6 os
(e.g., Tn 88-63, Ndiambour, 58-57, Mougne, Mouride,
and Melakh) and poorly a’dapted exotics to have low
aould have provided about 77% of the maximal water
leaf A (e.g., CB46, Prima, 8517, and CBS-El). For the
rcquirement of a medium-cycle cultivar such as 58-57
13 genotypes included in the trials at both Bambey,
.-_.- -_-----
---.
130
A.E. Hull et (11. / Field Crops Reseurch 36 (1994) 12.5-131
Senegal, and Riverside, California, the correlation
correlation coefficients were very small for compari-
coefficient for genotypic means was small and nonsig-
sons between the tropical and subtropical regions
nificant (r=0.295, IZ = 13), and substantial rank
( Table 5) indicating inconsistent genotypic rankings
changes were present for Chino M 1,4R-0267- 1 F, and
between Senegal and either Riverside, California, or
Vita 7.
Lubbock, Texas. Correlation coefficients for Riverside
The trials with nine cultivars and lines at Bambey
versus Lubbock were moderate and were not signifi-
and Louga, Senegal, in 1992 received 34 1 and 202 mm
tant.
of rain, respectively. There was little plant-available
moisture in the soi1 at the beginning of the cropping
season in either location. This rainfall would have pro-
4. Discussion
vided about 75% and 44% of the maximal water
requirements of a medium-cycle cultivar, such as 58-
57, at Bambey and Louga, respectively (Hall and Patel,
Genotypic differences in leaf A were observed for
1987). Genotypic differences in leaf A were highly
cowpea cultivars and lines grown in Senegal (Table 1)
significant, and there were significant effects of loca-
that were relatively consistent over three trials in con-
tion but very little G X E or change in genotypic ranking
trasting conditions (Table 4). Similar consistency in
(Table 2). Total shoot biomass production at Bambey
ranking for leaf A had been reported for southern Cal-
( Table 3 > was not correlated with leaf A, but the well-
ifornia (Table 4) and may occur in the high plains of
adapted local land race, 58-57, had the greatest shoot
Texas (Table 5). For cowpeaprograms concerned
with
biomass production and relatively high leaf A. Two
breeding cultivars for rainfed production in semiarid
exotic lines, CB46 and Prima, had small shoot biomass
Senegal, it is not clear whether selection should be for
production and relatively low leaf A. Contrasting per-
high, intermediate, or low leaf A. The productive and
formance was exhibited by TVx 3109-1G which had the
broadly adapted local land race, 58-57, had high leaf A
lowest leaf A, but the second largest shoot biomass
( Tables 2 and 3)) as did Tn 88-63 (Table 1) which is
production. Grain yields at Louga were very small due
well-adapted to dry zones of West Africa. In contras&
to the drought (an average value of 13 1 kg/ha com-
TVx 309- 1 G had consistently Iow leaf A (Tables 2 and
pared with 1734 kg/haat Bambey:) andgenotypiccom-
4) ^ but was productive at Bambey (Table 3).
parisons are not reliable.
The major differences in climate between the regions
In five studies over 2 years at F!iverside, California,
are that semiarid Senegal bas higher night temperatures
with different irrigation treatments (Hall et al.. 1992),
than Lubbock or Riverside: and daytime humidity is
leaf A values exhibited consistent genotypic ranking
higher at Lubbock than Riverside. Substantial changes
( Table 4). The same nine genotypes exhibi ted mod-
in genotypic ranking for low leaf A were observed
erately consistent ranking for leaf A values over three
between regions (Tables 4 and 5) ; although ‘~VX 309-
trials in Senegal (Table 4). Four genotypes exhibited
‘IG and UCR 237A had consistently low A in both
substantial changes in ranking across regions: CB46,
southem Califomia and Senegal (Table 4). In addition,
Prima, 58-57, and Vita 7. Cultivar CB46, which is
there was a tendency for well-adapted genotypes to
productive in California, had higb leaf A in Riverside,
bave high leaf A. These data may be explained by the
but low A in Senegal, where it has small yie1d.s of grain
presence of different genes conferring local adaptation
or hay (Table 3). Genotypes that are well-adapted to
in different environments. These genes may cause sto-
Senegal, 58-57 and Vita 7 (Table 3), had high A in
mata1 conductance to be ‘large, which would then
Senegal, but low A in California. Two genotypes had
enhance CC& assimilation and growth, but cause W to
consistently low leaf A in both regions: UCR :237A and
be low and A to be high. Positive correlations between
TVx 309- 1 G; and the latter cari be productive in Sene-
A and shoot biomass production that might be caused
gal (Table 3).
by genotypic differences in stomatal conductance have
Comparisons of genotypic means using correlation
been reported for wheat (Condon et al., 1987) and
coefficients indicates that genotypic rankings were rel-
Phase&s mdgaris L. (Ehleringer, 1990). The consis-
atively consistent in trials over different watering
tently low A of TVx 309-I G and UCR 237A appear to
regimes and years within regions (Table 5). In contrast,
have different physiological bases (Ismail and Hall,
A.E. Hall et al. / Field Graps Research 36 (1994) 125-131
131
1993 ) : TVx 309- 1G has greater phatosynthetic capac-
and dry matter production in field-grown wheat. Crop Sci.. 27:
ity, whereas UCR 237A has smaller stomatal conduc-
996-1001.
tance. Apparently, the genes conferring these traits may
Craufurd, P.Q., Austin, R.B., Accvedo, E. and Hall. M.A., 1991.
be consistently expressed in different environments.
Carbon isotope discrimination and grain yield in barley. Field
Crops Res., 27: 301-313.
The G X E that was observed would not necessarily
Ehleringer, J.R., 1990. Correlations between carbon isotope discrim-
constrain cowpea breeding programs that aim at devel-
ination and leaf conductance 10 water vapor in common beans.
oping improved cultivars for specific target production
Plant Physiol., 93: 1422-1425.
regions, such as semiarid tropical Senegal or subtrop-
Ehleringer, J.R., Hall, A.E. and Farquhar, G.D. (Editors), 1993.
ical Califomia. Selection for A in one of these regions,
Stable Isotopes and Plant Carbon-Water Relations. Academic
Press, San Diego, 555 pp.
however, may not be effective in producing cultivars
Farquhar, G.D. and Richards, R.A., 1984. Isotopic composition of
of cowpea and, possibly, other crops for the other
plant carbon correlates with water-use eftîciency of wheat gen-
region. Also, Richards and Condon (1993) reported
otypes. Aust. J. Plant Physiol.. 11: 539-552.
th’at genotypic ranking of wheat cari be different in off-
Farquhar, G.D., O’Leary, M.H. and Berry, J.A., 1982. On the rela-
scason glasshouse environments compared with com-
tionship between carbon isotope discrimination and the intercel-
lular carbon dioxide concentration in leaves. Aust. J. Plant
rrercial field production environments. Use of linked
Physiol., 9: 121-137.
genetic markers to indirectly Select for A has been pro-
Hall, A.E. and Patel, P.N., 1987. Cowpea improvement for semi-arid
posed (Martin et al., 1984). One advantage of DNA
regions of sub-saharan Africa. In: J.M. Menyonga, T. Bezuneh
marker selection is that it could be effective in any
and A. Youdeowei (Editors), Food Grain Production in Semi-
environmen& including off-season nurseries (B. Mar-
Arid Africa. OAU/STRC-SAFGRAD, Ouagadougou, Burkina
tin, pers. commun., 1992). This approach would net
Faso, pp, 279-290.
Hall, A.E., Mutters, R.G., Hubick, K.T. and Farquhar, G.D., 1990.
a;>pear to be more effective than selecting for A, at this
Genotypic differences in carbon isotope discrimination by cow-
time, because several markers are needed that may vary
pea under wet and dry field conditions. Crop Sci., 30: 30&305.
dcpending upon the target production environment.
Hall, A.E., Mutters, R.G. and Farquhar, G.D., 1992. Genotypic and
Cienotypes with consistently low A (e.g., TVx 309-
drought-induced differences in carbon isotope discrimination
1G) and consistently high A (e.g., Tn 88-63) may be
and gas exchange of cowpea. Crop Sci., 32: l-6.
Hall, A.E., Richards, R.A., Condon, A.G., Wright, G.C. and Far-
useful for selection experiments to determine for
quhar, G.D., 1994. Carbon isotope discrimination and plant
rainfed production in semiarid zones whether low,
breeding. Plant Breed. Rev. (in press).
intermediate, or high A is adaptive.
Hubick, K.T., Shorter, R. and Farquhar, G.D., 1988. Heritability and
genotype X environment interactions of carbon isotope discrim-
ination and transpiration efficiency of single plants of peanut
5. Acknowledgments
(Avachis hypogoea L.). Aust. J. Plant Physiol., 15: 799-813.
Ismail, A.M. and Hall, A.E., 19Y2. Correlation between water-use
efficiency and carbon isotope discrimination in diverse cowpea
This research was partially supported by the Bean/
genotypes and isogenic lines. Crop Sci., 32: 7-12.
Cowpea Collaborative Research Support Program of
Ismail, A.M. and Hall, A.E., 1993. Carbon isotope discrimination
the United States Agency for International Develop-
and gas exchange of cowpea accessions and hybrids. Crop Sci.,
ment Grant No. DAN-1310-SS-6008-00, and The
33: 788-793.
Southwest Consortium, New Mexico State University,
Martin, B., Nienhuis, J., King, G. and Schaefer, A., 1984. Restriction
fragment length polymorphisms associated with water use effi-
IJSDA Subagreement No. 88-34186-3340. The opin-
ciency in tomato. Science, 24.3: 1725-1728.
ions and recommendations
are those of the authors atid
Richards, R.A. and Condon, A.G., 1993. Challenges ahead in using
net necessarily those of USAID.
carbon isotope discrimination in plant breeding programs. In:
J.R. Ehleringer, A.E. Hall and G.D. Farquhar (Editors), Stable
Isotopes and Plant Carbon-~Water Relations. Academic Press,
San Diego, pp. 45 l-462.
61. References
Thiaw, S., Hall, A.E. and Parker, D.R., 1993. Varietal intercropping
and the yields and stability of cowpea production in semiarid
(:ondon, A.G. and Richards, R.A., 1992. Broad sense. heritability
Senegal. Field Crops Res., 35: 217-233.
and genotype X environment interaction for carbon isotope dis-
White, J.W., 1993. Implications of carbon isotope discrimination
crimination in field-grown wheat. Aust. J. Agric. Res., 43: 921-
studies for breeding common bean under water deficits. In: J.R.
9 3 4 .
Ehleringer, A.E. Hall and G.D. Farquhar (Editors), Stable Iso-
Condon, A.G., Richards, R.A. and Farquhar, G.D., 1987. Carbon
topes and Plant Carbon-Water Relations. Academic Press, San
isotope discrimination is positively correlated with grain yield
Diego, pp. 387-398.
Crops
Research
E:LSEVIER
Field Crops Research 36 (1994) 125-131
d
Consistency of genotypic ranking for carbon isotope
discrimination by cowpea grown in tropical and subtropical zones
A.E. HalP*, $ Thiawb, D.R. Krieg”
“Department of Botany and Plant Sciences, Qniversity of California, Riverside, CA 925;?1-0124, USA
bISRA/Centre National de Recher’ches Agronomiques, BP 53, Banbey, Senegal
‘Department ofAgronomy, Texas $ech Universiry, Box 42122, Lubbock, TX, USA
(Accepte4 16 November 1993)
Abstract
Measurements of plant composition of stable carbon isotopes ( ‘3C/‘2C) cari be used to estimate the extent that plants
discriminate (3) against the heavier carbon isotope. Theoretical and experimental studies have shown that d cari be negatively
correlated with transpiration efficiency ( W = seasonal total biomass production/seasonal transpiration) in C3 plants. Conse-
quently, Select ion for low A ( and high W) may be useful in breeding Ca plants for some water-limitedenvironments. Development
of effective breeding methods depends upon the extent bf genotype X environment interaction (G XE) as it influences the
consistency of genotypic ranking across environments.
Sets of cowpea cultivars and lines were grown under çontrasting water supply regimes in subtropical zones in Riverside,
C’alifomia, and Lubbock, Texas, and in tropical zones in Senegal. Leaf carbon isotope composition was measured and significant
genotypic differences in A were observed in most mals with little G X E for mals within these regions. Genotypic rankings were
quite consistent between wetter and drier environments artd different years within these regions with significant correlation
coefficients for genotypic means in most cases. However, when comparing A values between the twlo subtropical zones (Califomia
versus Texas), correlation coefficients for genotypic means only were moderate and not significant. Comp<arisons of A values
between the tropical zone (Senegal) and the two subtropical zones indicated no consistency in genotypic ranking, and correlation
coefficients for genotypic comparisons were very small.
Apparently.. G X E for A would not necessarily constrain cowpea breeding programs which aim at developing improved
cultivars for specific target production regions, such as semiarid tropical Senegal. However, cowpea performance, with respect
to A and W, uould not be transferable to radically different production zones where attainment o:f high W may require different
s#:ts of genes.
&ey words: Cowpea; Transpiration efficiency; Variety tria]; Vign&
1. Introduction
( W = seasonal total biomass production/seasonal
tran-
spiration) could be increased without associated neg-
Yields of crops for water-limited environments
ative effects. Richards and Condon ( 1993) provide a
would be enhanced if transpiration efficiency
discussion of possible associated negative effects, such
-~
as slow canopy growth and associated increases in soi1
*Corresponding author.
evaporation. Direct selection for W in field conditions
Elsevier Science R.V.
SXDl0378-4290(93)E0068-9
1 2 6
A.E. Hall et al. /Fi& Crops Research 36 (1994) 125-131
would be very difficult due to the need to extract and
tant trop for rainfed semiarid tropical regions of Africa
measure root biomass, and estimate soi1 evaporation
where water limitations cari substantially reduce yield
and total water-use SO that transpirational water-use cari
(Hall and Patel, 1987; Thiaw et al., 1993). Studies
be estimated. However, theoretical and empirical S<ud-
conducted in southem Califomia have discovered cow-
ies indicate that carbon isotope discrimination (A) may
pea accessions with significant differences
in A (Hall
be effective for indirectly selecting for W (i.e., reviews
et al., 1990) that could confer substantial differences
and case studies in Ehleringer et al., 1993), and A cari
in W (Ismail and Hall, 1992)) and genotypic ranking
be estimated in field nurseries from measurements of
for A was consistent over different levels of drought
“C/ ‘.7C in plant tissues. Discrimination against ‘me
and years (Hall et al., 1992). Consistency of genotypic
heavier isotope ( 13C) occurs during the diffusion and
ranking for A over a very wide range of environments
fixation of C02. According to the theory for C, plants,
has not been evaluated for cowpea or most other crops.
A and W both depend upon the concentration of CO,
Studies were conducted to: ( 1) determine whether a
within leaves (Farquhar et al., 1982). Increases in
set of contrasting cowpea genotypes exhibit significant
interna1 photosynthetic capacity or partial stomatal clo-
differences
and any G XE in A measured in leaves of
sure cause decreases in the concentration of CO, within
plants grown in two tropical locations in semiarid Sene-
leaves and concomitant increases in W and decreases
gal; and (2) compare genotypic ranking with values of
in A. Consistent with this theory, plants growing in the
A already obtained for thie same genotypes in subtrop-
same aerial environment have exhibited a negative lin-
ical regions in southem Califomia (Hall et al., 1992)
ear regression between W and A (Farquhar and
and the high plains of Texas.
Richards, 1984). Development of effective breeding
m e t h o d s depends u p o n t h e e x t e n t o f geno-
type X environment interaction (G X E) for A as it
‘2. Materials and methods
influences the consistency of genotypic ranking (Hall
et al., 1994). Selection would not be effective if gen-
Experiments were conducted in a semiarid tropical
otypic ranking for a trait were to vary from year to pear
region of Senegal having a single summer rainfall sea-
for the same genotypes.
son on the Bambey and Louga research stations of the
Some studies have not detected significant G XE.
Institut Senegalais de Recherches Agricoles in 1991
Sixteen peanut genotypes were evaluated in 10 aon-
and 1992. These stations have similar air temperatures
trasting subtropical and tropical environments in
during the growing season with daily maxima averag-
Queensland, and G X E was not significant for A, but
ing 33°C and daily minima averaging 24°C (Hall and
was highly significant for kemel yield (Hubick et al.,
Patel, 1987). Louga is located on the arid boundary of
1988). Studies with 21 or 23 genotypes of barley over
the semiarid zone with an average annual rainfall of
2 years in two environments in the United Kingdbm,
269 mm in recent years compared with 471 mm at
and three locations in Syria, gave positive and sigmifi-
Bambey (Hall and Patel, 1987).
tant correlations among genotypic values of A for 20
During the summer of 199 1,42 land races, cultivars,
out of 25 comparisons
with different pairs of environ-
and breeding lines of cowpea were grown at Bambey
ments (Craufurd et al., 199 1) .
( 14”42’N 16”28’W) on a deep, slightly leached, trop-
In other studies, significant G X E has been detecited.
ical ferruginous soil, called a “Dior” soil, which has a
Studies with about 20 genotypes in a range of field
volumetric moisture content at field capacity of 0.16.
environments across the southem Australian wheat-
Fertilizer was broadcast and incorporated at the rate of
belt ( Condon and Richards, 1992) indicated significant
150 kg/ha of 6:9:8 (N::P:K). Seed was sown at the
1;
G XE in rnany cases, but it was generally a minor
beginning of the rainy season on 12 July in rows 50 cm
component of variation in A compared with the geno-
apart with 50 cm between plants for spreading geno-
typic variation. For common bean (Phaseolus uulgaris
types and 25 cm for erect genotypes, consistent with
L. ) , significant G X E has been observed for genotypes
current recommended practices. The experimental
grown under different watering regimes (White,
design consisted of four :randomized complete blocks.
1993).
Individual plots consisted of four rows of plants, 4 m
Cowpea (V@a unguiculata L. Walp.) is an impor-
long. Samples consisting of three leaflets were taken
A.E. Hall et al. /Field Çrops Research 36 (1994) 125-131
127
fi om the eighth node of three plants in each plot on the
plains of Texas at the research station of Texas Tech
33th day after sowing. The nine leaflets taken fro@
University at Brownfield in Terry County in a loamy
e,lch plot were pooled and dried, and carbon isotope
fine-Sand soil. Similar experimental methods were used
composition was measured as described in Hall et al.
in both years except that eight cowpea genotypes were
(1992).
evaluated in 198X and six cowpea genotypes were eval-
During the summer of 1992, nine cowpea genotypds
uated in 1989. The experimental design was a split plot
u’ere grown in two experiments at Bambey and Louga
with three replications. The main plots were two levels
( 15”36’N 16’13’W). The soi1 at Louga is sandy with
of trickle irrigation and the sub-plots were the different
a small clay content (3%) and a volumettic moistuqe
genotypes. Individual plots consisted of four rows, with
c$>ntent at field capacity of 0.08. Seeds were sown ht
75 cm between rows, and 12 m long. A fertilizer appli-
the beginning of the rainy season on 17 July in Bambey,
cation of 50 kg N ha-’ and 20 kg P ha-] was made
a Id 10 August in Louga, and the plants were growp
prier to sowing, and additional N was provided in the
uiing the same methods as in 1991, except that indj-
irrigation water at a rate of 1 kg N ha - ’ mm- ‘. A well-
vtdual plots consisted of three rows, 6 m long. Sampljs
watered treatment was established which received
cl,nsisting of three leaflets were taken from the fifth
maximal water needs every 3 days. A drier treatment
nade of four plants on the 28th day after sowing. The
was established where plants were provided with max-
12 leaflets taken from each plot were pooled, dried, and
imal water needs for the first 35 days, after which irri-
a lalyzed for carbon isotope composition as described
gation was terminated. Samples consisting of three
in Hall et al. ( 1992). Forage and grain yield weqe
leaflets were taken from the eighth node of three plants
drtermined on the tenter row of each plot at the end df
in each plot on the 60th and 58th day after sowing in
the growing season.
1988 and 1989, respectively. The nine leaflets taken
Experiments were conducted in the summers of 19S;X
from each plot were pooled and dried, and carbon iso-
a Id 1989 in a semiarid, subtropical region in the high
Tible 1
Clrbon isotope discrimination measured in leaves of cowpea genoqypes grown at Bambey, Senegal, in 1991. Genotypic effects were very highly
.
siqxhcant wlth an LSD,,.,,,
- -
0 4%”
Rank
Genotype
A
Rank
Genotype
A
@‘oo)
(%Cl
i
480
2 2 . 3
2 2
Baye Ngagne
2 1 . 7
,
Tn 88-63
2 2 . 2
2 3
Vita 7
2 1 . 7
3
Kaedi Gris-blanc
2 2 . 2
2 4
Ndouc.
2 1 . 6
.l
4R-0267- 1 F
22.1
2 5
C B 5
2 1 . 6
i
ER- 1
2 2 . 0
2 6
Ndiaga Aw
2 1 . 6
0
Ndiambour
22.0
2 7
ER-7
2 1 . 5
7
473
2 1 . 9
2 8
Diongama
2 1 . 5
‘I
5x-57
2 1 . 9
2 9
Chino Ml
2 1 . 5
0
572
2 1 . 9
3 0
CBS-El
2 1 . 4
10
601
2 1 . 9
31
Bambey 21
21.4
I l
Mougne
2 1 . 9
3 2
8517
2 1 . 4
1.3
482
2 1 . 9
3 3
TVx 1841.OlE
2 1 . 4
1.i
Mouride
2 1 . 9
3 4
Prima
2 1 . 4
I-1
514
2 1 . 9
3 5
WR 237A
21.4
1:;
TVu 374
2 1 . 9
3 6
CB46
2 1 . 3
10
Melakh
2 1 . 8
3 7
TVx 309.1G
2 1 . 3
1”
475
2 1 . 8
3 8
58-111
2 1 . 3
18
TVx 75-4
2 1 . 7
3 9
66-20
2 1 . 3
1’)
590
2 1 . 7
40
Bambey 29
2 1 . 3
20
TVx 1836-015
2 1 . 7
41
Casa 16
2 1 . 3
21
484
2 1 . 7
42
IAR 16-96
21.1
--_~
128
A.E. Hall et 01. /Field Crops Research 36 (1994) 125-131
T a b l e 2
Carbon isotope discrimination measared in leaves of cowpea genotypes grown at Bambey and Louga, Senegal, in 1992
-
Genotype
Bambey
Louga
_-
-
Rank
A
Rank
A
i%)
60)
-
-
58-57
(3)
2 0 . 2
(1)
2 2 . 8
Vita 7
(4)
2 0 . 2
(2)
2 2 . 7
Bambey 2 1
(1)
2 0 . 3
(3)
2 1 . 9
C B 5
(2)
2 0 . 3
(4)
2 1 . 9
4R-0267-IF
(5)
20.1
(5)
2 1 . 8
CB46
(6)
2 0 . 1
(6)
21.4
Prima
(7)
20.1
(8)
2 1 . 0
UCR 237A
(8)
‘ 1 9 . 8
(7)
2 1 . 2
TVx 309- 1 G
(9)
1!9.5
(9)
2 0 . 9
Mean
20.1
2 1 . 7
I-SJ4.m
0 . 7
Mean Squares
Genotype effects
1.53
**
Location effects
50.72
***
G X E effects
0.56
n s
Errer
0.51
LSD tests differences betw&n genotypic
-
.-
means. ** a.nd *** = significant at P<O.Ol and 0.001, respectively. and ns is not signitïcant at
P=O.38.
T a b l e 3
versity of Califomia. These experiments have been
Hay and grain yield of cowpea genotypes grown at Bambey, Senegal,
described previously (Hall et al., 1992). In 1987, cow-
in 1992
pea genotypes were grown under well-watered and
Genotype
Hay
Grain
Total
stored-soil-moisture conditions in a split-block design
f kg/ha)
Wha)
(kdha)
with four replications. The main blocks consisted of
weekly, well-watered furrow irrigation and a treatment
58-57
4433
1 9 7 3
6406
t.hat received no rain or irrigation after plant establish-
TVx 309-1G
3 1 5 1
21:20
5 2 7 1
ment. The sub-plots consisted of the different geno-
4R-0267- 1 F
2278
2197
4475
Vita 7
2033
2 0 8 1
4114
types. In 1988 a third, intermediate irrigation treatment
Bambey 2 1
2 2 9 1
15.35
3826
was included that was irrigated on altemate furrows
LJCR 237A
1503
18:29
3332
every 3 weeks during the growing season. Samples
Prima
1 8 4 7
1 1 0 5
2952
consisting of three leaflets were taken from the eighth
CB46
1 1 3 9
1 4 1 2
2 5 5 1
node of three plants in each plot 5.5 and 62 days after
.
C B 5
1 1 1 6
13.54
2520
sowing in 1987 and 1988, respectively, and processed
Mean
2205
17.34
3939
to determine A as described in Hall et al. ( 1992).
LSQ,.m
937
406
3. Results
tope composition was measured as described in Hall et
al. (1992).
The screening tria1 with 42 diverse cultivars and lines
Experiments were conducted in the summers of 1987
at Bambey, Senegal, in 1991 received 347 mm of rain.
and 1988 in a semiarid, subtropical region of southem
There was little plant-available moisture in the soi1 at
Califomia at the Riverside research station of the Uni-
the beginning of the cropping season. This rainfall
A.E. Hall et al. 1 Field Crops Research 36 (1994) 125-131
1 2 9
Table 4
C>wpea genotypic ranking for carbon isotope discrimination meaaured in leaves in California and Senegal
G motype
Riverside, California
Senegal
1 9 8 7
1 9 8 8
1991
1 9 9 2
-
-
Wet” Dry
Wet
Med. Dry
Mean
Bambeyb
Bambey
Louga
Mean
Rank
Rank
Rank
Rank
Rank
(%o)
Rank
Rank
Rank
C%@l
-
CB46
1
2
2
1
1
19.8
8
7
6
20.9
Rima
2
1
3
2
2
19.7
6
6
8
20.8
CB5
3
3
1
3
'3
19.5
4
2
4
21.2
431-0267-1F
4
7
4
4
4
19.0
1
5
5
21.4
Bunbey 21
5
5
6
5
7
18.8
5
1
3
21.2
511-57
7
4
5
6
5
18.8
2
4
1
2 1 . 6
TVx309-IG
6
6
7
7
6
18.7
9
9
9
20.6
Vita 7
8
8
9
8
8
18.1
3
3
2
21.5
UCR 237A
9
9
8
9
9
17.9
7
8
7
20.8
‘Wet, Med., and Dry denote well-watered, intermediate, and limit&irrigation treatments, data from Hall et al. ( 1992).
‘Itambey and Louga are locations in Senegal with intennediate and limited supplies of rain, respectively.
Table 5
(Hall and Patel, 1987). Genotypic differences in leaf
C~~~clation coefficients for genotypic mean values of carbon isotope discrim-
A were very highly signilïcant (Table 1)) but the range
ination measured in leaves of cowpea
-
of values ( 1 .2%0) was smaller than the range obtained
G%nparisons
r
nP
at Riverside, California (2.1%0) with a set of 50 gen-
otypes including 13 of the same ones (Hall et al.,
W ithin rqions
1990). Maximum differences in W cari be estimated
Sonthem Califomia
using the following equation:
Riverside Wet vs Dry 1987
0 . 9 0 9 9 <O.ool
Riverside Wet vs Dry 1988
0.916 9 <O.Obl
Riverside 1987 vs Riverside 1988
0 . 9 9 4 9 <O.chl
W,/W,=(b-d-A,)/(b,-d-d,)
(1)
Jmegal
Bambey 1992 vs Louga 1992
0 . 7 0 0 9
0.05
where (b - d) are parameters for the inherent discrim-
Bambey 1991 vs Bambey 1992
0.477 9 ns
ination associated with carboxylation (b) and other
Htgh Plains of Texas
metabolic processes (d). The value of (b-d) was
Lubbock Wet Y:; Dry 198X
0.907 8
o.o,
estimated from the empirical studies of Ismail and Hall
Lubbock Wet vs Dry 1989
0.317 6 ns
( 1992) as being 24.3%0. The genotype with the lowest
A (IAR 16-96) in Table 1 is estimated to have 58%
Bntwcen regions
Senegal (1991 -r 1992) vs Riverside
-0.125 9 ns
higher W than the genotype with the highest A (#480).
(1487+ 1988)
This is similar to the range of difference in W estimated
for the study at Riverside, California (Hall et al., 1990)
Senegal(l991+ 1992) vs Lubboçk (1988)
-0.122 6 ns
of 62%.
Senegal ( 1991 -C 1992) vs Lubbock ( 1989)
-0.098 6 ns
At Bambey, Senegal (Table 1) , there was a tendency
for well-adapted local cultivars to have high leaf A
Lubbock (1988) vs Riverside (1987+ 1988)
0.569 8 ns
Lubbock ( 1989) vs Riverside ( 1987+ 1988)
0.518 6 os
(e.g., Tn 88-63, Ndiambour, 58-57, Mougne, Mouride,
and Melakh) and poorly a’dapted exotics to have low
aould have provided about 77% of the maximal water
leaf A (e.g., CB46, Prima, 8517, and CBS-El). For the
rcquirement of a medium-cycle cultivar such as 58-57
13 genotypes included in the trials at both Bambey,
.-_.- -_-----
---.
130
A.E. Hull et (11. / Field Crops Reseurch 36 (1994) 12.5-131
Senegal, and Riverside, California, the correlation
correlation coefficients were very small for compari-
coefficient for genotypic means was small and nonsig-
sons between the tropical and subtropical regions
nificant (r=0.295, IZ = 13), and substantial rank
( Table 5) indicating inconsistent genotypic rankings
changes were present for Chino M 1,4R-0267- 1 F, and
between Senegal and either Riverside, California, or
Vita 7.
Lubbock, Texas. Correlation coefficients for Riverside
The trials with nine cultivars and lines at Bambey
versus Lubbock were moderate and were not signifi-
and Louga, Senegal, in 1992 received 34 1 and 202 mm
tant.
of rain, respectively. There was little plant-available
moisture in the soi1 at the beginning of the cropping
season in either location. This rainfall would have pro-
4. Discussion
vided about 75% and 44% of the maximal water
requirements of a medium-cycle cultivar, such as 58-
57, at Bambey and Louga, respectively (Hall and Patel,
Genotypic differences in leaf A were observed for
1987). Genotypic differences in leaf A were highly
cowpea cultivars and lines grown in Senegal (Table 1)
significant, and there were significant effects of loca-
that were relatively consistent over three trials in con-
tion but very little G X E or change in genotypic ranking
trasting conditions (Table 4). Similar consistency in
(Table 2). Total shoot biomass production at Bambey
ranking for leaf A had been reported for southern Cal-
( Table 3 > was not correlated with leaf A, but the well-
ifornia (Table 4) and may occur in the high plains of
adapted local land race, 58-57, had the greatest shoot
Texas (Table 5). For cowpeaprograms concerned
with
biomass production and relatively high leaf A. Two
breeding cultivars for rainfed production in semiarid
exotic lines, CB46 and Prima, had small shoot biomass
Senegal, it is not clear whether selection should be for
production and relatively low leaf A. Contrasting per-
high, intermediate, or low leaf A. The productive and
formance was exhibited by TVx 3109-1G which had the
broadly adapted local land race, 58-57, had high leaf A
lowest leaf A, but the second largest shoot biomass
( Tables 2 and 3)) as did Tn 88-63 (Table 1) which is
production. Grain yields at Louga were very small due
well-adapted to dry zones of West Africa. In contras&
to the drought (an average value of 13 1 kg/ha com-
TVx 309- 1 G had consistently Iow leaf A (Tables 2 and
pared with 1734 kg/haat Bambey:) andgenotypiccom-
4) ^ but was productive at Bambey (Table 3).
parisons are not reliable.
The major differences in climate between the regions
In five studies over 2 years at F!iverside, California,
are that semiarid Senegal bas higher night temperatures
with different irrigation treatments (Hall et al.. 1992),
than Lubbock or Riverside: and daytime humidity is
leaf A values exhibited consistent genotypic ranking
higher at Lubbock than Riverside. Substantial changes
( Table 4). The same nine genotypes exhibi ted mod-
in genotypic ranking for low leaf A were observed
erately consistent ranking for leaf A values over three
between regions (Tables 4 and 5) ; although ‘~VX 309-
trials in Senegal (Table 4). Four genotypes exhibited
‘IG and UCR 237A had consistently low A in both
substantial changes in ranking across regions: CB46,
southem Califomia and Senegal (Table 4). In addition,
Prima, 58-57, and Vita 7. Cultivar CB46, which is
there was a tendency for well-adapted genotypes to
productive in California, had higb leaf A in Riverside,
bave high leaf A. These data may be explained by the
but low A in Senegal, where it has small yie1d.s of grain
presence of different genes conferring local adaptation
or hay (Table 3). Genotypes that are well-adapted to
in different environments. These genes may cause sto-
Senegal, 58-57 and Vita 7 (Table 3), had high A in
mata1 conductance to be ‘large, which would then
Senegal, but low A in California. Two genotypes had
enhance CC& assimilation and growth, but cause W to
consistently low leaf A in both regions: UCR :237A and
be low and A to be high. Positive correlations between
TVx 309- 1 G; and the latter cari be productive in Sene-
A and shoot biomass production that might be caused
gal (Table 3).
by genotypic differences in stomatal conductance have
Comparisons of genotypic means using correlation
been reported for wheat (Condon et al., 1987) and
coefficients indicates that genotypic rankings were rel-
Phase&s mdgaris L. (Ehleringer, 1990). The consis-
atively consistent in trials over different watering
tently low A of TVx 309-I G and UCR 237A appear to
regimes and years within regions (Table 5). In contrast,
have different physiological bases (Ismail and Hall,
A.E. Hall et al. / Field Graps Research 36 (1994) 125-131
131
1993 ) : TVx 309- 1G has greater phatosynthetic capac-
and dry matter production in field-grown wheat. Crop Sci.. 27:
ity, whereas UCR 237A has smaller stomatal conduc-
996-1001.
tance. Apparently, the genes conferring these traits may
Craufurd, P.Q., Austin, R.B., Accvedo, E. and Hall. M.A., 1991.
be consistently expressed in different environments.
Carbon isotope discrimination and grain yield in barley. Field
Crops Res., 27: 301-313.
The G X E that was observed would not necessarily
Ehleringer, J.R., 1990. Correlations between carbon isotope discrim-
constrain cowpea breeding programs that aim at devel-
ination and leaf conductance 10 water vapor in common beans.
oping improved cultivars for specific target production
Plant Physiol., 93: 1422-1425.
regions, such as semiarid tropical Senegal or subtrop-
Ehleringer, J.R., Hall, A.E. and Farquhar, G.D. (Editors), 1993.
ical Califomia. Selection for A in one of these regions,
Stable Isotopes and Plant Carbon-Water Relations. Academic
Press, San Diego, 555 pp.
however, may not be effective in producing cultivars
Farquhar, G.D. and Richards, R.A., 1984. Isotopic composition of
of cowpea and, possibly, other crops for the other
plant carbon correlates with water-use eftîciency of wheat gen-
region. Also, Richards and Condon (1993) reported
otypes. Aust. J. Plant Physiol.. 11: 539-552.
th’at genotypic ranking of wheat cari be different in off-
Farquhar, G.D., O’Leary, M.H. and Berry, J.A., 1982. On the rela-
scason glasshouse environments compared with com-
tionship between carbon isotope discrimination and the intercel-
lular carbon dioxide concentration in leaves. Aust. J. Plant
rrercial field production environments. Use of linked
Physiol., 9: 121-137.
genetic markers to indirectly Select for A has been pro-
Hall, A.E. and Patel, P.N., 1987. Cowpea improvement for semi-arid
posed (Martin et al., 1984). One advantage of DNA
regions of sub-saharan Africa. In: J.M. Menyonga, T. Bezuneh
marker selection is that it could be effective in any
and A. Youdeowei (Editors), Food Grain Production in Semi-
environmen& including off-season nurseries (B. Mar-
Arid Africa. OAU/STRC-SAFGRAD, Ouagadougou, Burkina
tin, pers. commun., 1992). This approach would net
Faso, pp, 279-290.
Hall, A.E., Mutters, R.G., Hubick, K.T. and Farquhar, G.D., 1990.
a;>pear to be more effective than selecting for A, at this
Genotypic differences in carbon isotope discrimination by cow-
time, because several markers are needed that may vary
pea under wet and dry field conditions. Crop Sci., 30: 30&305.
dcpending upon the target production environment.
Hall, A.E., Mutters, R.G. and Farquhar, G.D., 1992. Genotypic and
Cienotypes with consistently low A (e.g., TVx 309-
drought-induced differences in carbon isotope discrimination
1G) and consistently high A (e.g., Tn 88-63) may be
and gas exchange of cowpea. Crop Sci., 32: l-6.
Hall, A.E., Richards, R.A., Condon, A.G., Wright, G.C. and Far-
useful for selection experiments to determine for
quhar, G.D., 1994. Carbon isotope discrimination and plant
rainfed production in semiarid zones whether low,
breeding. Plant Breed. Rev. (in press).
intermediate, or high A is adaptive.
Hubick, K.T., Shorter, R. and Farquhar, G.D., 1988. Heritability and
genotype X environment interactions of carbon isotope discrim-
ination and transpiration efficiency of single plants of peanut
5. Acknowledgments
(Avachis hypogoea L.). Aust. J. Plant Physiol., 15: 799-813.
Ismail, A.M. and Hall, A.E., 19Y2. Correlation between water-use
efficiency and carbon isotope discrimination in diverse cowpea
This research was partially supported by the Bean/
genotypes and isogenic lines. Crop Sci., 32: 7-12.
Cowpea Collaborative Research Support Program of
Ismail, A.M. and Hall, A.E., 1993. Carbon isotope discrimination
the United States Agency for International Develop-
and gas exchange of cowpea accessions and hybrids. Crop Sci.,
ment Grant No. DAN-1310-SS-6008-00, and The
33: 788-793.
Southwest Consortium, New Mexico State University,
Martin, B., Nienhuis, J., King, G. and Schaefer, A., 1984. Restriction
fragment length polymorphisms associated with water use effi-
IJSDA Subagreement No. 88-34186-3340. The opin-
ciency in tomato. Science, 24.3: 1725-1728.
ions and recommendations
are those of the authors atid
Richards, R.A. and Condon, A.G., 1993. Challenges ahead in using
net necessarily those of USAID.
carbon isotope discrimination in plant breeding programs. In:
J.R. Ehleringer, A.E. Hall and G.D. Farquhar (Editors), Stable
Isotopes and Plant Carbon-~Water Relations. Academic Press,
San Diego, pp. 45 l-462.
61. References
Thiaw, S., Hall, A.E. and Parker, D.R., 1993. Varietal intercropping
and the yields and stability of cowpea production in semiarid
(:ondon, A.G. and Richards, R.A., 1992. Broad sense. heritability
Senegal. Field Crops Res., 35: 217-233.
and genotype X environment interaction for carbon isotope dis-
White, J.W., 1993. Implications of carbon isotope discrimination
crimination in field-grown wheat. Aust. J. Agric. Res., 43: 921-
studies for breeding common bean under water deficits. In: J.R.
9 3 4 .
Ehleringer, A.E. Hall and G.D. Farquhar (Editors), Stable Iso-
Condon, A.G., Richards, R.A. and Farquhar, G.D., 1987. Carbon
topes and Plant Carbon-Water Relations. Academic Press, San
isotope discrimination is positively correlated with grain yield
Diego, pp. 387-398.