Front Cover
 Title Page
 Table of Contents
 Types of drought
 Physiological factors
 Breeding for improved drought...
 The CIMMYT experience
 Back Cover

Title: Breeding and selection for drought resistance in tropical maize
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00080064/00001
 Material Information
Title: Breeding and selection for drought resistance in tropical maize
Physical Description: 16, 3 p. : ill. ; 28 cm.
Language: English
Creator: Fischer, K. S
Johnson, E. C
Edmeades, G. O
International Maize and Wheat Improvement Center
Publisher: International Maize and Wheat Improvement Center
Place of Publication: México D.F. México
Publication Date: 1983
Subject: Corn -- Drought tolerance -- Tropics   ( lcsh )
Corn -- Breeding -- Tropics   ( lcsh )
Corn -- Climatic factors -- Tropics   ( lcsh )
Corn -- Drought resistance   ( nal )
Plant-breeding   ( nal )
Plants, Effect of drought on   ( nal )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Bibliography: Includes bibliographical references (p. 17-19).
Statement of Responsibility: K.S. Fischer, E.C. Johnson and G.O. Edmeades.
General Note: "An earlier version of this paper was presented at a Symposium on "Principles and Methods in Crop Improvement for Drought Resistance: with emphasis on rice" at the International Rice Research Institute (IRRI), May 4-8, 1981."
 Record Information
Bibliographic ID: UF00080064
Volume ID: VID00001
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 10833108

Table of Contents
    Front Cover
        Page i
    Title Page
        Page ii
    Table of Contents
        Page iii
        Page iv
        Page 1
    Types of drought
        Page 1
    Physiological factors
        Page 2
        Page 3
    Breeding for improved drought resistance
        Page 4
        Page 5
        Page 6
    The CIMMYT experience
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
    Back Cover
        Page 20
Full Text


breeding and selection

for drought
resistance in tropical maize

K.S. Fischer**, E.C. Johnson*** and G.O. Edmeades****

* An earlier version of this paper was presented at a Symposium on "Principles and
Methods in Crop Improvement for Drought Resistance: With Emphasis on Rice" at
the International Rice Research Institute (IRRI), May 4-8, 1981
** Maize physiologist, CIMMYT, Asian Regional Maize Program, G.P.O. Box 2453,
Bangkok, Thailand
*** Former plant breeder, CIMMYT, present address IPRI, 853 Industrial Road, San
Carlos, California, 94070
**** Maize agronomist/physiologist, CIMMYT, Crops Research Institute, P.O. Box 3785,
Kumasi, Ghana






iv Summary

1 Introduction

1 Types of Drought

2 Physiological Factors

4 Breeding for Improved Drought Resistance

7 The CIMMYT Experience

7 Choice in Germplasm

8 Development of Suitable Selection Criteria

10 Intrapopulation Variation and Improvement

12 Evaluation of Progress

16 Conclusions

17 References


Throughout the lowland wet tropics, periodic
nonprotracted drought caused by irregular rainfall distribu-
tion is responsible for sizable reductions in maize yield. This
is particularly true when reduced water availability coincides
with the critical stage of crop development--flowering.
Such a drought cannot be escaped by genotype maturity or
planting date, nor are other species necessarily better
adapted. Improving resistance of maize to this particular
type of drought, then, could enhance productivity and
minimize farmer risk.
From a comparison of sorghum and maize under
drought, it appears that sorghum maintains photosynthesis
and growth at lower water levels and has more developmental
plasticity than maize. Increasing the dry matter available
for ear development around flowering may help to reduce
the detrimental effects of drought occurring at this critical
stage in maize.

Many morphological and physiological characters
have been suggested for modification so as to enchance
drought resistance in maize. At CIMMYT, one lowland
tropical maize population is being improved for drought
resistance through a recurrent selection program. Progenies
are selected using an index based on grain yield under no
stress and stress, leaf elongation rate, interval between
anthesis and silking, canopy temperature and leaf area loss
during grain filling. Evaluation of the progress after three
cycles of recurrent selection shows a significant increase in
yield under severe drought conditions. Improvement is
approximately 9.5 percent per cycle and is mainly associated
with a decrease in the number of barren plants.
Other morphophysiological traits are being
evaluated for their effectiveness in changing plant response
to drought. Selections for reduced tassel, leaf and height
may improve yield under severe stress conditions. Also,
through multilocation testing and selection in its Maize
International Testing program, CIMMYT has been able to
achieve improved tolerance to drought stress.


This paper first defines a specific type of non-
protracted drought which affects large areas of the lowland
tropics. It compares the drought-resistance mechanisms of
sorghum and maize. The ability of sorghum to continue
photosynthesis at lower water levels, along with its develop-
mental plasticity, may account for some of its advantages
over maize in this type of drought.
It then describes the philosophy of current breeding
strategies for drought resistance and reviews techniques
used to measure differences between maize genotypes for
Finally, evidence is given for the effectiveness of
recurrent selection for drought resistance using a selection
criteria of leaf elongation rate, interval between silking and
anthesis, canopy temperature, leaf area loss and grain yield
under stress and no stress. This method is being practiced
at CIMMYT in one lowland tropical material for the specific
type of drought mentioned above.


Throughout the tropics, periodic drought, caused
by irregular rainfall distribution and accentuated by soils
with low water-holding capacity, causes sizable reductions
in maize yields (Wolf et al. 1974). Estimates reveal that
drought may account for an average loss of 15 percent of
production in tropical areas, even where total rainfall is
reasonably high. Further, the probability of yield loss due
to drought influences the use and utilization of fertilizer

and other inputs. Drought, therefore, is probably responsible
for a much higher economic loss than indicated.
Based on greenhouse experiments, field trials and
historical analysis of on-farm yield data, agronomists and
meteorologists have concluded that drought occurring
around flowering has a major effect on grain yield. Deficits
of water for periods lasting one to two days during tasseling
or pollination may cause as much as 22 percent reduction
in yield (Robins and Domingo, 1953), while stress during
the grain filling stages (McPherson and Boyer, 1977) and
vegetative stages (Denmead and Shaw, 1960) may have
much less effect on yield.
It is clear that the most effective means of reducing
the effects of drought on maize would be to escape periods
of low moisture availability through the manipulation of
genotype maturity and planting date. An example of this is
given by Kasam et al. (1975) for maize grown at Ibadan,
Nigeria (Figure 1). At this site total rainfall is about 1,140
mm and is spread from March to November in a bimodal
pattern. The first season is long enough for a 120-day maize
crop; in the second season, however, crop water requirement
can be met without high soil moisture deficits in maize with
80-90 day maturity. Plant breeding programs, therefore,
should aim at providing high yielding genotypes with a
range of maturities to best fit the season as determined by
moisture availability.
Maize grown in large areas of the tropics, however,
is affected by drought occurring during and usually in the
middle of the main summer growing season. Total rainfall
for the crop season may be adequate, but a reduction in the
number of days with rain, particularly around the critical


This paper first defines a specific type of non-
protracted drought which affects large areas of the lowland
tropics. It compares the drought-resistance mechanisms of
sorghum and maize. The ability of sorghum to continue
photosynthesis at lower water levels, along with its develop-
mental plasticity, may account for some of its advantages
over maize in this type of drought.
It then describes the philosophy of current breeding
strategies for drought resistance and reviews techniques
used to measure differences between maize genotypes for
Finally, evidence is given for the effectiveness of
recurrent selection for drought resistance using a selection
criteria of leaf elongation rate, interval between silking and
anthesis, canopy temperature, leaf area loss and grain yield
under stress and no stress. This method is being practiced
at CIMMYT in one lowland tropical material for the specific
type of drought mentioned above.


Throughout the tropics, periodic drought, caused
by irregular rainfall distribution and accentuated by soils
with low water-holding capacity, causes sizable reductions
in maize yields (Wolf et al. 1974). Estimates reveal that
drought may account for an average loss of 15 percent of
production in tropical areas, even where total rainfall is
reasonably high. Further, the probability of yield loss due
to drought influences the use and utilization of fertilizer

and other inputs. Drought, therefore, is probably responsible
for a much higher economic loss than indicated.
Based on greenhouse experiments, field trials and
historical analysis of on-farm yield data, agronomists and
meteorologists have concluded that drought occurring
around flowering has a major effect on grain yield. Deficits
of water for periods lasting one to two days during tasseling
or pollination may cause as much as 22 percent reduction
in yield (Robins and Domingo, 1953), while stress during
the grain filling stages (McPherson and Boyer, 1977) and
vegetative stages (Denmead and Shaw, 1960) may have
much less effect on yield.
It is clear that the most effective means of reducing
the effects of drought on maize would be to escape periods
of low moisture availability through the manipulation of
genotype maturity and planting date. An example of this is
given by Kasam et al. (1975) for maize grown at Ibadan,
Nigeria (Figure 1). At this site total rainfall is about 1,140
mm and is spread from March to November in a bimodal
pattern. The first season is long enough for a 120-day maize
crop; in the second season, however, crop water requirement
can be met without high soil moisture deficits in maize with
80-90 day maturity. Plant breeding programs, therefore,
should aim at providing high yielding genotypes with a
range of maturities to best fit the season as determined by
moisture availability.
Maize grown in large areas of the tropics, however,
is affected by drought occurring during and usually in the
middle of the main summer growing season. Total rainfall
for the crop season may be adequate, but a reduction in the
number of days with rain, particularly around the critical

flowering stage of the crop, may have a marked effect on
grain yield. The first season at Ibadan illustrates this form of
drought. Even though total moisture is adequate, the margin
between crop water requirement and water availability is
small, particularly in the period just prior to flowering.
When water stress occurs at this time, there may be a
sizable reduction in grain yield. Mosino and Garcia (1968)
suggest that summer drought may affect over 7 million
hectares of rainfed maize in Mexico, including most of the
lowland tropical area where total rainfall is greater than
1,000 mm. In El Salvador, even though monthly rainfall
may average 300 mm, summer drought known as canicula
may be responsible for reducing yields by 20 percent and is
cited as the most frequent source of crop loss as compared
to insects, lodging and excess water (Walker, 1980).
Since the exact timing of this drought during the
growing season is unpredictable, it cannot be avoided by
either genotype maturity or planting date. Furthermore,
maize may be the best adapted cereal for these conditions,
since high humidity and rain at harvest could be harmful to
an alternative crop such as sorghum. To minimize the risk
of yield loss, farmers may stagger their plantings of maize,
plant maizes of different maturities or intercrop different
species. Improved agronomy, such as better weed control
and the maintenance of surface mulch (minimum tillage),
will have a substantial effect on maize yields under drought.
Maize varieties which are better able to resist the effect of
reduced moisture, particularly around flowering, also would
help stabilize grain yields under these conditions.
This paper therefore examines selection criteria
that may be useful in the development of such varieties. It
is suggested that for situations where the duration of
adequate moisture is limited, yields could be stabilized
through the use of genotypes and planting arrangements
which escape stress. In situations where moisture availability
is predictably inadequate to sustain maize, alternative
crop species should be grown or additional moisture


Drought resistance in an agricultural sense refers to
the ability of a crop plant to produce its economic product
with limited available water. Drought resistance in an
evolutionary context, however, would normally be the
ability of a plant or species to survive and eventually
reproduce under limited moisture. It is likely that the
mechanisms responsible solely for survival of a species may
in fact differ from those which provide for maximum
economic production. The fact that the survival of maize
has always relied on the intervention of man, therefore,
reduces the probability that this species has evolved strong
mechanisms for survival under moisture stress (Qualset,

Levitt (1972) suggested that the mechanisms for
drought resistance (used in the generic sense) be divided
into mechanisms of drought escape, drought avoidance and
drought tolerance. Drought escape tends to minimize the
interacting of drought with crop growth and yield; tolerance
gives the ability to produce despite loss of plant water
status; avoidance increases the ability to maintain relatively
high plant water status despite a shortage of moisture in the
environment (O'Toole and Chang, 1979; Fischer and
Sanchez, 1979; Fischer and Turner, 1978). However,
O'Toole and Chang (1979) note that too often these
mechanisms are viewed in terms of either/or (exclusive),
implying that a choice is necessary, rather than in terms of
and/or (complementary).
Drought escape is often the most important
and successful form of drought resistance and is usually
imparted through the combination of genotype maturity
and planting date. However, due to the unpredictability of
the drought being discussed here, drought resistance through
escape is generally not feasible and the remainder of this
paper concentrates, therefore, on selection for avoidance
and/or tolerance mechanisms.
Fischer and Turner (1978) have analyzed plant
productivity under arid and semi-arid conditions in terms of
total water transpired (obtained), the efficiency with which
this water is used (water use efficiency as g dry matter
produced per g water transpired), and harvest index (the
ratio of economic yield and total dry matter). They found
little evidence for consistent cultivar differences in water
use efficiency and, thus, yield under moisture-limiting
conditions was determined by total transpiration (root
exploration, etc.) and harvest index (these two parameters
may be antagonistic, i.e. an increase in dry matter
partitioned to the root to allow extra root exploration
could reduce harvest index of grain). With wheat there is
evidence that there is genetic potential to improve both
harvest index and root patterns (Fischer and Turner, 1978;
Passioura, 1981).
Although the body of information needed to
explain the physiological basis for drought resistance
continues to grow, it is difficult to discern a major associa-
tion between a trait (or traits) and drought resistance with
application to a breeding program (Fischer and Wood,
1979). One approach used to study the usefulness of a
factor for crop improvement is the development of isogenic
lines or divergent selections for the character being
considered. This approach is time consuming (Moss et al
1974) but may be necessary to unravel the complexity of
drought-resistant mechanisms. To determine which factors
are more likely to be of consequence, it is useful to compare
two species, such as maize and sorghum, which differ
markedly in drought resistance but are otherwise adapted
to similar environments. In so doing, it must be recognized
that the information provided will be as influenced by the

genotype genotypess) chosen to represent each species as it
is by any trait within the species.
A number of studies exist in which maize and
sorghum have been compared. Under tropical nonstress
conditions in Nigeria, Kassam (1976) measured water use
efficiencies of 3.9 and 3.7 mg dry matter/g water for maize
and sorghum respectively, while Ludlow (1976) reported
values of 2.8 and 3.3 mg/g for the two species respectively.
There is no comparable data for stress conditions.
Insofar as the association between particular
traits and drought resistance is concerned, Table 1 lists those
characters examined by various workers, together with the
importance they place on those characters in explaining
differences in drought resistance between maize and
sorghum. There is not always agreement as to the impor-
tance of any one character and this is noted in the table.
Of the traits shown in Table 1, the two that
probably are of most importance for the nonprotracted
drought are differences in critical water potential (the
potential close to zero turgor when stomates close) and
differences in developmental plasticity. Neumann et al.
(1974) measured critical leaf water potentials of -8.1 to
-9.6 bars for maize and -11.2 to -13.8 bars for sorghum,
while the field data for the two species show values of -16
and -21 bars respectively (Turner, 1974). Thus, sorghum
tolerated higher internal water deficits before closing stomata
and could continue photosynthesis at lower water potentials
(Boyer, 1970a; Beadle et al. 1973).

The species also differ in their developmental
plasticity. As to sorghum's being more able to avoid the
effects of moisture stress at a critical stage of plant
development, Whiteman and Wilson (1965) found that
inflorescence development could be suspended during stress
and resume development after rewatering. Moisture stress,
during various stages of panicle development in sorghum,
causes a reduction in grain number but, even under severe
stress, plants will exert partial panicles (Eastin, 1980). This
is in contrast to maize, where stress initially reduces ear size
but then reduces the number of ear-bearing plants
(increased barrenness). This may be a special featureof the
inflorescences of maize--the staminate flowers are produced
in the terminal inflorescence and the pistillate flowers on
lateral shoots (ears). Further, in sorghum, individual grains
have a greater capacity to compensate for a reduction in
grain number (Fischer and Wilson, 1975; Fischer and
Palmer, 1980). The ability of sorghum to form panicle-
bearing tillers makes possible the recovery and grain
production of these organs upon relief of water stress.
Understanding the factors controlling ear
development and barrenness of plants grown under moisture
stress should prove useful to developing drought resistance
in maize. Shaw (1977) has estimated the sensitivity of
various developmental stages to water stress (Figure 2). The
critical period includes flowering and coincides with the
time of maximum crop transpiration (Downey, 1971;
Andre et al. 1978). The pattern of yield reduction due to

Table 1. A Comparison of Sorghum and Maize Traits Associated with Drought Resistance and Estimates of their
Relative Importance in Explaining Differences in Productivity under Nonprotracted Moisture Stress

Level of
Character Importance References

Root Density High-low Martin, 1930; Sanchez-Diaz etal. 1969

Root Exploration Low Hsiao et al. 1976; O'Toole and Chang, 1979
Root Osmotic Potential High Martin etal. 1931; Sullivan and Blum, 1970

Leaf Cuticular Resistance High-low Martin, 1930; Yoshida and de los Reyes, 1976

Leaf Stomatal Resistance Low Glover, 1959; Boyer, 1970a; Sanchez-Diaz and Kramer,
1971; Beadle et al. 1973

Leaf Size and Rolling Low Martin, 1930

Developmental Plasticity High Whiteman and Wilson, 1965; Sullivan and Blum, 1970;
Eastin, 1980

Critical Potential* High Boyer, 1970a; Sanchez-Diaz and Kramer, 1971; Neumann
et al. 1974; Turner, 1974; Ludlow, 1976

Osmoregulation High Stout and Simpson, 1977; Jones and Turner, 1978; Jurgens
et al. 1978; Turner et al. 1978

Desiccation and Heat Tolerance Low Sullivan and Blum, 1970

Water potential at zero turgor

drought stress is similar to the effects of reduced radiation
as shown by Fischer and Palmer (1980) and Prine (1971).
Tollenaar (1977) recently reviewed the control of grain
yield in maize and concluded that the irradiance per plant
during flowering was a dominant factor determining grain
number. Thus maintenance of photosynthesis during this
stage was critical to yield.
In addition to the influence of total dry matter
accumulation at the critical period, the partitioning of dry
matter to the developing ear and factors affecting spikelet
fertility are important in determining grain yield through
control of grain number. Fischer and Palmer (1980) reviewed
a number of morphophysiological traits that may affect
grain number under nonstress conditions; some of these
may also be relevant under nonprotracted drought stress.
In tropical maize grown under stress levels of
nitrogen, water and density, careful removal of the male
inflorescence prior to flowering increased grain yield by
9.5, 21.0 and 17.9 percent respectively (Poey et al. 1977).
Hybrids with heavier tassels had longer anthesis-to-silking
intervals (Daynard, 1968) and lower grain yield under
density stress (Buren etal. 1974). The drought resistance of
a number of hybrids was also related to less leaf area and a
shorter interval from mid-anthesis to mid-silking. In the
leafier genotypes, leaf development may be at the expense
of growth of the developing ear (Dow, 1981).
The interval between anthesis and silking increases
under most stress conditions, including drought and high
density. This delay in silk development may be related to a
decline in nitrate reductase (Hsiao et al. 1976), to a reduc-
tion in current assimilate supply (Dow, 1981) or to other
factors. A number of authors (Jensen, 1971; Duvick, 1977)
have advocated selection for reduced interval between

anthesis and silking under population density stress for
better performance under moisture stress. Genotypes with a
tendency toward prolificacy (two ears) also have better
population tolerance (Buren et al. 1974). Hallauer and
Troyer (1972) reviewed the performance of prolific types
and concluded that this character contributes to the reduc-
tion of genotype x environment interaction through its
ability to adjust to environmental stresses, including
Dow (1981) concluded that hybrids resistant to
density stress were also more drought resistant. However,
he warned that, while selection for a decrease in the
anthesis-to-silking interval under high density, nonmoisture
stress conditions would improve drought resistance, other
parameters conferring drought resistance could be lost or
selected against.


Blum (1979) has described two major breeding
philosophies aimed at improvement of genotypes to stress:
"The first and very common approach accepts that a
superior yielding variety at the potential level will also yield
relatively well under subpotential levels. Drought resistance
may be present in such a variety and expressed as an
unidentified component of stability in performance over
various environments. During the breeding process, yield
and stability in performance are handled as one complex.
Accumulation of environmentally stable yield genes equates
with better performance also under stress situations."
This approach has been successful in sorghum (Blum,

Figure 2. Effect on Grain Yield of Maize from a Single Day of Moisture Stress (A) and Eleven
Days with 54 Percent Crop Shading (B)

(A) (B)
100 110 0

S96 Temperate | 7 Tropical Variety
SGermplasm -Temperate Hybrid
96 -\ :2 =
=S 70"
s ? 60 o LSD(0.05)
o 50
a) o
S "92 Flowering Maturity 0 o
l5 Flowering Maturity
S2 -30-20- -i 0 K 20 30 o s0o -50 -40 -30 -20 -10 10 20 30 40 50
Days from Flowering Days from Flowering

Sources: Shaw, 1977(A) and Fischer and Palmer, 1980 (B)

1979), wheat (Worrall et al. 1980) and maize (Russell,
1974; Duvick, 1977). In wheat, improved yield of
CIMMYT-derived genotypes over a wide range of conditions
is attributed mainly to an increase in yield potential, an
increase in environmental stability and a small change in
the response of the genotype to environmental conditions
(Worrall et al. 1980). It appears that some traits, for example
improved harvest index of short straw wheats, have a
sufficiently strong positive effect on yield under all conditions
to give them obvious superiority over traditional varieties or
collections, even under dry conditions and despite the
specific drought-resistant mechanisms the latter group may
possess (Fischer and Wall, 1976). However, subsequent
work by Fischer and Wood (1979) did not indicate a relation-
ship between harvest index under irrigated conditions and
grain yield under drought.
Gains in maize yields due to breeding in the
US Corn Belt from 1930 to 1970 indicate an increase in
yield potential; in the more recent commercial hybrids
there is considerable improvement at the lower yielding
environments (Table 2) (Russell, 1974). The improved
performance at the higher stress environments (in this case
probably due to moisture availability) might be due to
better stalk and root quality brought about by initial
selection under high plant density and to the extensive
testing of germplasm for yield stability (Russell, 1974).
In view of the lack of sound information on
specific drought characters and the considerable scope for

improvement of yield potential, a breeding strategy based
on selection under well-watered conditions may very likely
be the most efficient (in time) for bringing about rapid
progress. This system has the advantage that, under optimal
growing conditions, heritabilities for yield are higher than
under suboptimal conditions (Johnson and Frey, 1967).
Testing over a large number of sites with varying
moisture availability, although expensive, should enable the
elimination of those genotypes which may have negative
yield traits under moisture stress. Separation of the effects
of drought escape and the identification of traits specifically
favoring performance under these test conditions could also
facilitate selection. For international crop improvement
centers such as CIMMYT, testing over a wide geographic
range also provides a vehicle for the introduction of
improved germplasm into national programs.
"A second approach to breeding for yield
performance in a stress environment maintains that indeed
potential yield is irrelevant," (Blum, 1979). Varieties must
be selected, developed and tested under the relevant
conditions. There are a few examples of population
improvement in maize based on this procedure. One mass
selection study was done in Colombia for the rainy season
(600 mm) and dry season (300 mm) separately and in
combination (Arboleda-Rivera and Compton, 1974). The
selection criterion was grain yield. Three cycles of selection
in the rainy season increased yield by 10.5 percent per
cycle for that season but increased yield during the dry

Table 2. Changes in Grain Yield since 1930 for Representative US Corn Belt Maize
Materials over a Number of Locations

Approximate Grain Yield (t/ha)
Year Mean Across Regression
of Release Maize Type Locations Maximum Slope

Open-Pollinated 5.48 7.05 1.16

1930 Double-cross Hybrid 5.78 8.03 1.26

1940 Double-cross Hybrid 6.58 8.68 1.01

1950 Double-cross Hybrid 6.75 9.08 1.03

1960 Double-cross Hybrid 7.31 10.15 1.27

1970 Single-cross Hybrid 8.37 10.87 1.06
(Public Line)

1970 Single-cross Hybrid 8.07 9.30 0.63
(Commercial Line)

Source: Russell, 1974

season by only 0.8 percent. Three cycles of selection during
the dry season increased yield by 2.5 percent per cycle in
that season and by 7.6 percent in the rainy cycle. Another
study in Mexico involved mass selection for a number of
cycles under irrigated and/or rainfed conditions (Muioz,
1975). Testing of the synthetic derived from these selections
showed similar performance at the high rainfall, high
yielding site, but a greater yield from the selection made
under stress at the low rainfall, low yielding site.
An alternative approach to the two strategies
described is to improve drought resistance in those materials
which already have high yield potential. As improvement in
yield potential becomes relatively more difficult to achieve,
breeding programs might focus more attention on the
identification of specific drought-resistant mechanisms.
Finlay and Wilkinson (1963) suggested that, in barley, both
yield potential and yield stability over environments could
be independently manipulated in a breeding program. In
maize, data provided by Russell (1974) clearly demonstrate
the importance of improved yield potential in improving
yields over a wide range of environments. However, in that
work, an analysis of the performance of some of the more
recently developed hybrids demonstrated that differences
in yield at higher stress environments were due to factors
other than yield potential (Table 2).
It is inferred, then, that selection must be for
increasing, or at least maintaining, potential yield and, in
addition, for improving drought-resistant traits. In maize, it
is likely that such traits are multigenic and at a low gene
frequency in any given population; their frequency needs to
be increased through recurrent selection programs.
Increasing the frequency of genes for one or two drought-
resistant traits while maintaining yield may lead to an
improvement in yield under stress. Recurrent selection for a
morphological trait which has a physiological relationship
with grain yield has been effective in improving grain yield
under nonstress conditions (e.g. Johnson and Fischer,
1979). It is interesting to speculate on the effect on grain
yield under stress conditions if a program of recurrent
selection for a trait associated with drought resistance is
carried out. This would depend on an understanding of
drought-resistance mechanisms relative to the ecology for
which the material is being developed and on the rapid
identification of such mechanisms in large breeding nurseries.
At the same time, materials should be evaluated under
favorable conditions to maintain or improve yield potential.
Under the influence of natural selection, a few
races of maize in various parts of the tropics have developed
drought avoidance and/or tolerance mechanisms. One
collection, Michoacan 21, was described by Palacios de la
Rosa (1959) as having a distinct response to drought and
frost; the mechanism was called latente. This collection
maintained itself under drought without flowering, recovered
remarkably on rewatering, was more resistant to permanent
wilting at the seedling stage, transpired more than other

lines under irrigation, and transpired less under stress due to
stomatal closure (Muioz, 1975). This response may be due
in part to high levels of abscisic acid (Larque-Saavedra and
Wain, 1974). The latente trait has proven difficult to
transfer to higher yielding, agronomically desirable
germplasm, particularly in the lowland tropics. However,
workers elsewhere have successful used this material as a
source of genes for the improvement of drought resistance
in hybrids for the US Corn Belt (Castleberry and Lerette,
1979). In their study, the latente trait did not appear to be
simply inherited and the development of the drought-
resistant hybrids required the selection of inbred lines
under controlled moisture conditions for yield and other
traits associated with drought resistance.
Many morphological and physiological characters
have been suggested for modification so as to enhance
either drought avoidance and/or tolerance (Moss et al.
1974; Parker, 1968). A number of screening methods have
been used to compare the responses of different genotypes
of maize to drought and, while some of these methods
appear useful in a plant breeding program, there is a paucity
of evidence on their use in a population improvement
program (Qualset, 1979). In almost all cases cited, the
screening of lines was the end product of the breeding
program. There are too few reported programs in which
selected materials have been recombined and tested. In
many cases, also, results obtained in laboratory tests are not
further tested under field conditions.
Hurd (1976) has reviewed numerous accounts
where plant water stress decreased with increased depth and
branching of roots. There may be, however, some ecological
conditions where reduced root growth, particularly early in
the crop cycle, is an advantage (Passioura, 1972). In maize,
Nass and Zuber (1971) measured differences between forty
genotypes in terms of total root volume and weight of
nodal roots at two growth stages prior to flowering. These
characters were correlated with the measured resistance to
root pulling of the plants at maturity. Differences in root
volume in maize genotypes have also been recorded by
Musick etal. (1965) and Thompson (1968). Spencer (1940)
noted large differences between inbred lines of maize in the
rate of development of lateral roots and in the ratio of top-
to-root dry weight of seedlings. Muleba (pers. comm.),
using young plants grown in solution culture, selected
families for superior root weight and length and recombined
them to form experimental varieties. Evaluation of these
experimental varieties under water stress conditions in the
field showed that selection for larger root weight was useful
in increasing grain yield under mild water stress while
selection for increased root length was superior under
severe stress.
The rate of leaf elongation has been shown to be
sensitive to changes in leaf water potential (Boyer, 1970b;
Watts, 1974) and soil water supply (Acevedo et al. 1971).
Boyer and McPherson (1975) have suggested that the rate

of cell elongation in seedlings could be used to screen for
drought tolerance in cereals. Fischer and Edmeades (1977)
used leaf elongation rates to screen maize progenies for
drought resistance under field stress conditions.
There has been a considerable breeding effort to
modify stomatal response and reduce water loss by
transpiration. A number of workers in other crops have been
successful in reducing transpiration per unit leaf area
(Jones, 1979). Selection has been for the frequency and
anatomical structure of the stomata and for measured
stomatal conductance (Wilson, 1975). Infrared thermometry
has been used to screen large numbers of genotypes for
canopy temperature; this can be related to stomatal
conductance (Jackson et al. 1977; Kretchmer et al. 1980).
Williams et al. (1967) compared inbreds and
hybrids for drought resistance by a) the percentage of
seedlings which recovered from a 6-hour exposure to 520C
(heat tolerance), b) germination percentage of seeds
exposed to a manitol solution of 15 atmospheres, and c)
percentage recovery of seedlings watered 14 days after they
had reached wilting. The ratings obtained by each of the
three methods were tested by correlation analysis with field
evaluations based on the ratio of grain yield under stress to
yield under full irrigation. The results suggest that the
information from these techniques is correlated with field
data and, therefore, any of them would aid a breeding
Other workers have used similar techniques. Hunter
et al. (1936), Tatum (1954), and Kilen and Andrew (1969)
showed that the relative differences in response between
inbred lines to high temperature coincided with observations
of leaf firing in the field. Muioz (1975) conducted three
cycles of mass selection of seedlings which showed good
recovery upon rewatering after initially being stressed to the
wilting point. Kilen and Andrew (1969) used chlorophyll
stability as an index of heat tolerance for inbred lines of
maize and found it to be correlated with ratings of leaf
firing in the field.
Screening of seeds or seedlings in solutions of
different osmotic potential was used by workers as early as
1930 and has had limited results (Ashton, 1948). Parmer
and Moore (1968) have modified this technique for maize
by the use of polyethylene glycol solutions, and Johnson
and Asay (1978) have demonstrated the effectiveness
of this osmoticum in differentiating between lines of
crested wheatgrass.
Abscisic acid has been shown to be important in
drought resistance. In maize, Larque-Saavedra and Wain
(1974) measured a large difference of in vivo, free abscisic
acid between a drought resistant line (Michoacan 21,
latente) and two European varieties under nonstress and
stress conditions. There are no examples of the screening
of a larger number of maize genotypes under field
conditions for this trait in maize, although such work is
being conducted in other cereals (Austin et al. 1981).

Recently, screening certain amino acids which
increase dramatically under stress has been used as a means
of evaluating drought resistance. One of these, proline, was
suggested as being useful for drought screening by several
workers (Singh et al. 1972), but this has been questioned
recently (Hanson et al. 1977). Results from work with the
compound betaine suggest that it may be a valid indicator
of the cumulative stress experienced by plants and, if so,
discarding genotypes with high betaine content might be
effective in selecting for drought avoidance. In maize, Pinter
et al. (1978) reported that the free asparagine and proline
content of plant tissue subjected to drought was positively
correlated with drought resistance as estimated from the
difference in grain yield under stress and no-stress conditions.


The objectives of CIMMYT's Maize Program
are to increase the realized yield and yield potential of a
number of adapted maize populations and to improve their
yield stability. The breeding and selection system used is
described elsewhere (Johnson, 1974, Vasal et al. 1978,
Paliwal and Sprague, 1981). In 1976, limited work was
begun to assess the feasibility of selecting more directly
for drought resistance in tropical maize. The objectives
were to demonstrate the improvement, through recurrent
selection, of the performance of one tropical population
exposed to a particular type of drought. In particular, this
work was aimed at improving resistance to drought occurring
at the critical phase of plant development-flowering. Escape
mechanisms would not be utilized.

Choice of Germplasm
Data from the 1973 Experimental Variety Trials (CIMMYT,
1974) were used to identify a population with high and
stable yield. Only data from sites which were rainfed were
used. Mean yield of each site was significantly correlated
with rainfall during the growing season (r=0.74*),
suggesting that yield was to some extent influenced by
moisture availability. The data were analyzed for yield
stability by regression analysis (Finlay and Wilkinson,
1963) and for similarity of response by cluster analysis
(Mungomery et al. 1974); the results are shown in Figure 3.
Within the group of tropical germplasm entries, those with
a preponderance of the race Tuxpeiio had a slope less than
1.0 and a higher than average mean yield across all sites.
The Tuxpeio race has been described by Wellhausen (1956)
as one of the most important modern productive races in
both the USA and Mexico. Since it is found in areas
experiencing limited rainfall during the summer season, it is
not unlikely that it may have some natural adaptation to
moisture stress. It was therefore decided to use this
germplasm as a basis for population improvement for
* Significant at P = 0.05

Figure 3. Analysis of Adaptation by Grain Yield and Regression
Slope and by Cluster Analysis (Genotypes in Each
Shaded Arm Having a More Similar Response Than
Those Excluded) for an International Experimental
Variety Trial Grown only at Rainfed Sites (CIMMYT,

1.5 I

1.4- Tropical x

1.2 Temperate Tropi
S*4a4 Tropical
.2 1.1

. 1.0 f
Po Tuxpeno
M 0.9- esm P

0.8 Op e



3.0 4.0 5.0 6.0

Grain Yield (t/ha)

Development of Suitable Selection Criteria
The emphasis of this work was on field screening. At
Tlaltizapan, Morelos, Mexico, there is no appreciable
rainfall from October through April; plantings in November
are therefore completely dependent on applied water. This
site is at 900 m elevation with mean temperatures for the
growing season of approximately 280C maximum and 150C
minimum. The soil is a calcareous vertisol of approximately
1.8 to 2.0 m depth, and overlies a moist, calcareous parent
The response of eight maize genotypes (including
Tuxpeio-l**) of diverse genetic background to simulated
drought conditions at this site was used to develop relevant
selection criteria. Irrigation was controlled so that treat-
ments of drought stress commenced from floral initiation
and developed through to flowering (to span the critical
preflowering-flowering stage), and prior to flowering and
continuing through to grain maturity. These treatments
reduced grain yield. Fischer and Wood (1979) have defined
an index of drought intensity in wheat as one minus the
ratio of the mean yield under stress to yield under no
stress. Using this index, drought intensities were 0.48 and
0.47 for the stress from floral initiation to flowering and
from ten days before flowering to grain maturity respectively.
However,-although the stress intensity was similar, the yield
components affected by the stress differed. In the early
stress, grain number was reduced by 45 percent but, because
** Tuxpeiio-1 (CO) represents cycle 11 of recurrent selection
for short plants in the population Tuxpeio Crema 1

Table 3. Drought Index Based on Grain Yield of Stress and No-Stress Treatments, and Leaf Water
Potential and Stomatal Resistance, Measured at Flowering at Various Times of Day,
Subjected to Water Stress from Floral Initiation Onward, Tlaltizapan, 1975
Leaf Water Stomatal Resistance
Drought Potential (bars) (sec/cm)
Genotype Index 0600 1200 0900 1300

Tuxperio-1 1.43 -22 -14.7 5.40 3.60

Pioneer 3369 1.17 -2.1 -14.3 7.41 4.07

Pepitilla 1.09 -2.6 -16.0 7.74 5.25

Mezcla Amarilla 0.75 -3.2 -14.7 8.28 4.83

Super Enanos 0.93 -2.5 -14.6 9.03 4.58

Amarillo del Bajio 0.87 -2.8 -14.7 8.64 4.25

(Mix.1 -Col. Gpo.1)
-ETO Blanco 0.90 -3.0 -15.3 8.34 4.57

Early Tropical Composite 0.98 -2.1 -17.0 8.57 3.75

LSD* N.S. -1.7 2.07 0.30

* Significant at 0.05

of the rewatering after flowering, final grain size was
not affected. Stress from ten days before flowering through
maturity reduced grain number by 33 percent and grain size
by 20 percent.
In the work reported here, the response of the
eight genotypes to drought was assessed by a drought index
based on yield under both fully irrigated and stressed
conditions. (The drought index for any one genotype is the
ratio of its yield under stress to nonstress, relative to the
ratio of the mean yield of all genotypes under stress to
nonstress. Thus, a drought index > 1.0 suggests relative
drought resistance, and an index < 1.0, relative drought
susceptibility.) There were differences between genotypes
in both yield potential and drought index; Tuxpefio-1 had
both the highest yield potential and drought index score.
The ranking of genotypes by drought index was independent
of plant height and maturity, measured under nonstress
conditions, suggesting that the observed differences in
drought index were not due to escape mechanisms.
However, yield under nonstress was correlated with drought
index (r= 0.75*, unpublished data).
Measurements of leaf water potential and stomatal
resistance were made at various stages of crop development
and at different times during the day. There were significant
differences between genotypes in leaf water potential
measured at 1200 hours at flowering (Table 3). However,
differences in drought index were correlated (r= 0.76*)
with maximum leaf water potential measured at 0600
hours, not with minimum water potential taken at 1200
hours (Table 4). There were significant differences between
genotypes in stomatal resistance measured at 0900 and
1300 hours at flowering (Table 3). Stomatal resistance,
particularly when measured in the middle of the day, was
negatively correlated with drought index (Table 4).
Leaf water potential and stomatal resistance also were
measured during grain filling in the stress treatment from
ten days before flowering to grain maturity. Drought index
for this treatment was correlated negatively with stomatal
resistance, especially when measured at 1000 hours (Table
The capacity of genotypes to restore maximum
water potential during the night (before sunrise) and the
ability to maintain open stomata during the day appear to
be associated with better performance under the particular
stress at this site. While it is suggested that the difference in
root morphology may explain some of this differential
response, no observations of roots were made.
In this study, two morphological traits--the interval
between pollen shed and silking under stress (flower delay)
and the rate of stem elongation under stress--were also
measured. Stem elongation was positively correlated
(r=0.84*) and the flower delay negatively correlated (r=
-0.66*) with drought index. Both traits would appear useful
for selection. In subsequent work, a measure of the rate of
elongation of a newly exposed leaf was used, rather than

Table 4. Correlation between a Drought Index Based on Grain
Yield Under Stress and No Stress and Leaf Water
Potential and Stomatal Resistance Measured at Various
Times, Taltizapan, 1975

Trait Time of Measurement Correlation (r)
Leaf Water Flowering 0600 hours 0.76*
Potential 1200 hours 0.08

Grain Filling 0600 hours 0.26
1200 hours -0.22

Stomatal Resistance Flowering 0900 hours -0.51
1300 hours -0.89*

Grain Filling 1000 hours -0.72*
1300 hours -0.52

Data for eight genotypes
* Significant at 0.05

that of the stem. This measurement was made when the
plants in the severe water stress treatment were showing
midafternoon leaf rolling; the height from the ground to
the youngest visible leaf in the whorl was measured. A week
later the measurement was repeated on the same leaf. These
measurements were made on six plants per plot in both the
irrigated and stress treatments. The extension (which
includes components of stem and sheath elongation aswell
as leaf elongation) under drought was expressed relative to
the extension under nonstress so as to free it from genetic
differences in elongation rate under no-stress conditions.
The relative leaf elongation (RLE) is:

RLE = (H ) x 100

Where HI7 = leaf tip height under irrigation at day 7
HI0 = leaf tip height under irrigation at day 0
HS7 = leaf tip height under stress at day 7
HSO = leaf tip height under stress at day 0

Differences in leaf area duration were not measured
in this study. However, if the supply of assimilates during
grain filling is important to performance under drought (as
suggested by correlation of drought index and stomatal
resistance), then duration of active leaf area also may be an
important criterion in explaining genetic differences. In
subsequent work, plants were scored visually for leaf tissue
death using a scale of 1 to 5 (1--minimum loss, 5--maximum
loss). Ratings were made weekly, commencing three weeks
after flowering and continuing on a weekly schedule until
This initial work resulted in the development of a
selection index to be used to screen a large number of
segregating families. It is based on grain yield under irrigation
(yield potential) and drought, flower delay, leaf area loss

during grain fill and relative rate of leaf elongation (RLE).
The selection index considers these characters in a multi-
spatial arrangement and assigns to them relative distances
from a selection target. The distance for each character
relative to another can be varied by defining the selected
target in terms of standard errors from the mean. Each
character is further given a weighting in the overall selection
index (Schwarzbach, 1976). Correlations among characters
are not taken into account.
An example of the use of this index in selecting
the best 10 families (for formation of an experimental
variety) and 80 families (for recombination of the next
generation) in a progeny trial with 256 entries is given in
Table 5. Two additional characters, plant height and
maturity under irrigation, are included in the selection.
Because of the small plot size used, tall progenies had a
competitive advantage for light and therefore had higher
yields, particularly under irrigation. Similarly, yield under
the stress treatment tended to be positively associated with
earlier maturity. The object of the study was to select for
drought resistance through mechanisms other than escape.
Through the selection index, plant height and maturity are
kept constant (relative to the mean of the population) and
gains are made for the other characteristics.
The selection target expressed in absolute values
and in standard errors from the mean, and the weighting for
each of the characters included in the selection index are
shown in Table 5. Selection intensity is highest for grain
yield under stress and days to flower; it is relatively lower
for all of the other characters. For the eighty families

selected for recombination, the selection differentials (mean
of selected families minus the population mean) for grain
yield under stress and leaf tissue death were approximately
one standard error, while those for grain yield under
irrigation, flower interval and relative leaf elongation were
around 0.5 standard error. The correlations of these
characters with grain yield under stress and nonstress
conditions are shown in Table 6.

Intrapopulation Variation and Improvement
Using these criteria, eighty-five full-sib families of the
population Tuxpeio-1 were screened under moisture
regimes similar to those described earlier. A profile of the
soil moisture available at flowering and at maturity in the
severe stress treatment has been reported elsewhere
(CIMMYT, 1981). Analysis of yield indicated a significant
genotype x water stress interaction. There was, however, a
large increase in the coefficient of variation of the trial
under the stress treatment.
Experimental varieties, based on families selected
for yield under irrigation and yield under drought, and the
divergent selection for resistance and susceptibility based
on the selection index, were formed and again grown under
similar moisture regimes.
The grain yield of the various experimental varieties
under stress and nonstress treatments is shown in Figure 4.
There was no significant interaction of variety by water
stress level. However, F values for preplanned comparisons
among varieties indicate significant varietal differences.
When comparing those experimental varieties selected

Table 5. Statistics for Population of 250 Families of Tuxpefio-1, Selection Criteria, and Selections for each of the Characters
Used in the Selection for Drought Resistance, Tlaltizapan, 1980

Plant Flower Relative Leaf Canopy Temp-
Grain Yield (kg/ha) Height (I) Days to Intervals (S) Elongation Leaf perature (S) *
Irrigated (I) Stress (S) (cm) Flower (I) (days) (o/o) Scores (S) (oC)
Population Statistics:
Mean 5177 1324 175 91.0 5.8 64.6 3.1 28.0
SE 757 428 10.5 2.4 2.3 8.4 0.7 0.8
CV 14.6 32.3 6.0 2.6 40.0 18.0 23.0 2.7
Max. 6865 2918 209 97.6 14.1 93.5 5.0 30.6
Min. 2638 93 141 82.6 .1 44.5 0.9 26.2

Selection Index:
Target (Absolute) 6691 2608 170 92.2 1.2 77.3 1.7 26.5
Target (Standard Error
from Mean) + 2.0 +3.0 -0.5 +0.5 -2.0 2.0 -2.0 -2.0
Weighting 2.0 3.0 2.0 3.0 2.0 2.0 2.0 2.0

10 Families 5796 2171 175 91.4 3.2 70.3 2.4 26.8
80 Families 5529 1732 177 91.3 4.4 68.5 2.6 27.1
(0/o) +6.8 +30.8 +1.1 +0.3 -23.4 +11.0 -26.0 -4.0
SE Units +0.46 + 0.95 +0.19 +0.13 -0.61 + 0.45 -1.00 -1.13

* Additional criteria for 1980
* Selection differential for 80 families

Table. Coefficient of Linear Correlations (r) between Grain
Yields and Other Characters Used for Selecting Drought-
Resistant Families in Tuxpenlo-1, Grown under Non-stress
and Stress Conditions, Tlaltizapan, 1979

Grain Yield
Variable No Stress Stress

Grain Yield (Stress) 0.17* 1.00**

Relative Leaf Elongation -0.15 0.39

Interval between Anthesis and Silking 0.04 -0.36**

Leaf Tissue Death (Stress) -0.15 -0.48**

Canopy Temperature at:
7 Days before Flowering -0.56**
Flowering -0.73**
Grain Filling -0.65**

Total Dry Matter (No Stress) 0.64** 0.25

Harvest Index (No Stress) .07 -0.01

* Significant at 0.05
"* Significant at 0.01

mainly for grain yield, a significant increase was shown by
the variety selected for better grain yield under irrigation
when it was grown under no stress. However, under stress
conditions there was no significant difference between those
varieties, although the experimental variety selected for
better grain yield under stress tended toward better grain
A comparison of the experimental varieties based
on the selection index for resistance and susceptibility
showed a significant difference under stress; the yield of the
resistant and susceptible varieties being 2.3 and 1.5 t/ha
respectively. There was, however, no difference in yield
under irrigation. Under the stress treatment, the grain yield
of the resistant selection was also higher than that of the
experimental varieties selected for yiMl alone.
These studies suggested that a) there is genetic
variation within this tropical maize population for
performance under these specific drought situations, and b)
the inclusion of plant characters in addition to yield
enhances the identification of the drought-resistant families.
Based on these findings, a modified recurrent
population improvement program was initiated using the
Tuxpefo-1 population; it is now in the fourth cycle of
Two hundred fifty-six full-sib families are evaluated
at Tlaltizapan, Mexico, during the dry season under two
water regimes-normal irrigation and severe stress (no
irrigation after planting). Family entries are arranged in a
simple lattice (16 x 16) with two replications. For the stress

Figure 4. Influence of Moisture Regime on Performance of Four Experimental Varieties of Tuxpefto-1 when Selected for High
Grain Yield under No Stress and Stress (A) and for Physiological Characters for Resistance and Susceptibility to Drought
(B), Tlaltizapan, 1976

(A) (B)

7 0


a 1 0
4 LSD P.0.5

> 3T LSD P. 0.5
I Grain Yield Characters
2 Stress
o Grain Yield Susceptible
1 No Stress Characters

0 0 0
0 0.5 1 0 0.5 1
Relative Drought Intensity Relative Drought Intensity

treatment, there are two such trials (i.e. 4 replications). Plot
size is 2 1/2 m in length and 0.75 m wide; plant density is
52,000 plants/ha (2 plants at 50 cm spacing).Thesefamilies
are screened for characters previously discussed and, in
addition, to canopy temperature measured with an infrared
thermometer (Barnes Instatherm Model 14-220 D-4).
Measurements are made in the stress treatment prior to
tasseling and between 1100 and 1300 hours. The measure-
ment allows for an approximately 1 meter by 0.40 cm
section of canopy of each progeny to be evaluated for mean
temperature. The usefulness of this measurement can be
seen from the data shown in Table 6. The canopy
temperature for 256 families, measured before and at
flowering, was negatively correlated with their grain yield
(r = 0.56* and 0.73* respectively). When used in the
selection index, the mean canopy temperature of the 80
families selected for recombination was 1.13 standard error
units lower than that of the population mean (Table 5).
Of the 256 families evaluated at the Tlaltizapan
site, approximately 80 families are selected, with remnant

seed planted in summer in a crossing block at CIMMYT's
lowland tropical station, Poza Rica. A large number of
reciprocal full-sib crosses are made at random among the
families and, at harvest, 256 ears are saved to constitute the
new selection cycle. These families are again evaluated
in Tlaltizapan in the winter (dry) cycle.

Evaluation of Progress
An evaluation of progress was conducted after
three cycles of recurrent selection for drought resistance in
Tuxpefio-1. In addition, there was an evaluation of various
cycles of selection for reduced plant height in the population
Tuxpefio Crema I, and for reduced tassel (cycle 6), leaf size
(cycle 5), and yield and yield stability through the inter-
national progeny testing system (Pop. 21, cycle 3), in
Tuxpefio-1. The selections for reduced plant height had
already been shown to affect maturity and optimum plant
density for maximum grain yield. In the evaluation, planting
dates for the various cycles of selection for reduced plant
height were arranged so that all genotypes in the study

Table 7. Effect on Grain Yield after Selection for Various Characters under Irrigation and Stress
Conditions, Tuxperio Grown at Optimum Density, Tlaltizapan, 1981

Grain Yield Yield Change/Cycle
(kg/ha) (0/o of original)
Character Cycle Irrigation Stress Irrigation Stress
Reduced Height1 6 5276 1129

12 5358 1203 1.0 1.1

15 5893 1718* 1.3 5.8

18 6129 1570* 1.3 3.3

Reduced Tassel 0*** 5608 1213

6 6172 1673* 1.7 6.3

Reduced Leaf 0*** 5608 1213

5 6196 1468* 2.1 4.1

Drought Resistance 0 5859 1224
1.8 9.5
3 6179 1572*

EV** 6311 1647

Pop. 21 (Cycle 3)2 0*** 5608 1213

3 6458 1315 5.0 2.8

LSD (P=0.05) 899 433
CV 0/o 11.7 23.8

Difference from original cycle significant preplannedd F test) at P=0.05
** Experimental variety (40/o selection intensity)
*** Best estimates of original cycle
1 Planting dates arranged so all cycles flowered at or near (i 1-2 days) same time
2 Selected for yield, stability and wider adaptation through international progeny testing system

flowered at the same date; this served to reduce effects
resulting from drought escape mechanisms. The planting
density of the cycles of selection for short plants was also
varied, based on previous experience, in order to provide an
optimum density for each cycle. All of the other selections
in the trial were grown at 52,000 plants/ha. There were four
replications of each entry in the no-stress treatment and
eight replications under stress. Plot size was eight rows of 5 m
with a distance of 0.75 m between rows. All harvests
were from a well-bordered area of each plot.
The analysis of grain yield under stress and
no-stress treatment of various selections grown at their
optimum densities is shown in Table 7. The F values for
preplanned comparisons show that selection for drought
resistance improved grain yield under drought. However,
under no stress, there was only a small, nonsignificant
increase in grain yield. With 30 percent family selection
pressure, the rate of yield increase for the stress conditions
was approximately 9.5 percent per cycle. For the experi-
mental variety, representing a 4 percent family selection
pressure in the latest improvement cycle (cycle 3), yield
under stress was further improved (4.7 percent higher than
for cycle 3).
Selection for reduced plant height, tassel size or
leaf area resulted in an increase in yield under both water
regimes (Table 7). The percentage increase per cycle under
moisture stress was 6.3 and 4.1 for tassel and leaf selection,

respectively. Maximum grain yield under stress conditions
for the height selections was achieved at cycle 15 with an
average gain per cycle of 5.8 percent (from cycles 6 to 15).
Further selection for reduced plant height (cycle 18)
resulted in a reduction in grain yield under stress but not
under the no-stress treatment. The population 21 entry had
the highest (5 percent per cycle) yield increase under no
stress and a 2.8 percent per cycle increase under stress
conditions. For the leaf selection, grain yield per unit leaf
area increased from 44.4 to 67.8 g/m2 of leaf surface
(unpublished data).
Yield was examined in terms of its components--
total dry matter and harvest index (Table 8). The selections
for drought resistance had a nonsignificant increase (3.0
and 4.5 percent per cycle for cycle 3 and experimental
variety respectively) in total dry matter produced under
stress. Under irrigated conditions the total dry matter
increase was 1.5 percent per cycle for the experimental
variety. For the morphological selection for reduced height,
tassel and leaf size, there were no significant changes in
total dry matter either under irrigation or stress.
Total dry matter can be considered in terms of
total water transpired and water use efficiency (g dry
matter produced per g water transpired) of the crop. The
experiment was not designed to measure these components,
but there was an attempt to note differences in rooting
density between entries by visually scoring the amount of

Table 8. Total Dry Matter, Harvest Index and Ears Per Plant for Various Selections in Tuxpefio, Grown under Irrigated and
Stress Conditions, Tlaltizapan, 1981A

Total Dry Matter
(kg/ha) Harvest Index Ears/100 Plants
Character Cycle Irrigated Stress Irrigated Stress Stress

Reduced Heightl 6 13653 5408 34.0 15.1 56
12 13637 5157 39.1 15.0 65
15 12682 5719 42.9 25.6 85
18 13043 5327 44.7 25.0 72

Reduced Tassel 0** 13765 6112 39.5 15.7 58
6 14409 6778 41.4 22.2 65

Reduced Leaf 0** 13765 6112 39.5 15.7 58
5 13200 5510 44.3 24.1 73

Drought Resistance 0 13894 6246 39.6 16.4 57
3 14147 6807 38.3 20.9 65
EV* 14544 7099 38.2 20.8 74

Pop. 21 (Cycle 3)2 0 13765 6112 39.5 15.7 58
3 14995 6147 40.3 15.7 56

LSD (P=0.05) N.S. 996 6.1 2.0 12.8
CV 0/o 19.1 15.1 11.6 32.8 16.5

Experimental variety (40/o family selection)
** Best estimates of original cycle
1 Planting dates arranged so all cycles flowered at or near (+ 2 days) same time
2 Selected for yield, stability and wider adaptation through international progeny testing system

root found in soil probe samples taken at 30 cm intervals
and a depth of 150 cm (1--low density, 3--high density). This
was done at the same time (physiological plant maturity)
that volumetric soil moisture measurements were made.
There were significant varietal effects in both the root
scores and the volumetric moisture content of the soil
profile at 120-150 cm (Table 9). Although there were no
large differences between entries, there is an indication that
both selection for drought resistance and international
progeny testing has increased root activity at this depth.
The early generation of selection for reduced plant height
may have reduced perceived root density at this depth.
However, soil moisture was less for the fifteenth cycle for
reduced height.
The amount of water used by the crop from the
profile 0-150 cm was calculated from measurements of
gravimetric moisture and bulk density taken for each entry
at germination, flowering and maturity (Table 9). Again
there were significant differences, with a tendency for the
drought selection, reduced plant height (cycle 15) and the
international progeny testing to have increased the amount
of water taken up by the crop. Based on these data and
those for total dry matter at black layer, an estimate of the

water use efficiency for these materials was made. There
does not appear to be a consistent trend with the various
selections. The mean value for all selections under stress
was 2.5 mg dry matter/g H20 which is considerably lower
than the value of 3.9 mg/g for nonstressed conditions
reported by Kassam (1976).
Many of the changes in grain yield are associated
with changes in harvest index at optimum plant density
(Table 8). For the drought selection, harvest index under
drought conditions increased by 9.0 percent per cycle; this
appears to be associated with an increase in the number of
ear-bearing plants (selection for drought resistance reduced
the number of barren plants from 33 to 15 percent). Harvest
index increased by 7.7, 6.9 and 10.7 percent for the
selection for reduced height (cycles 6 to 15), tassel size and
leaf size. For these selections there was also an increase in
harvest index under the nonstress conditions (Table 8).
However, there may be a limit to the amount of improve-
ment in yield asa result of improved harvest index. Although
the harvest index for cycle 18 of the short plant selection
was higher under no stress and similar under stress than
cycle 15, grain yield and ears per plants under stress were
lower. In the drought selections, the higher yield of the

Table 9. Visual Estimate of Root Activity, Soil Moisture Content, Water Use and Estimated Water Use
Efficiency for Various Selections of Tuxpeiio, Grown under Stress Conditions

Soil Moisture Estimated Water
Root Density Content Water Use Use Efficiency
Score (120-150 cm) to Black Layer (mg dry matter/
Character Cycle (120-150 cm) (O/o) (mm) g H20)

Reduced Height1 6 2.02 41.7 206 2.62
12 0.96 41.6 210 2.45
15 1.21 31.0 286 1.99
18 0.84 41.0 245 2.17

Reduced Tassel 0** 0.96 42.1 217 2.81
6 1.12 41.6 231 2.93

Reduced Leaf 0 ** 0.96 42.1 217 2.81
5 0.87 41.2 235 2.34

Drought Resistance 0 0.84 42.7 228 2.74
3 1.28 35.2 275 2.47
EV*** 1.27 38.6 257 2.76

Pop. 21 (Cycle 3)2 0 0.96 42.1 217 2.81
3 1.37 39.5 261 2.35

LSD (P=0.05) 0.5 5.4 6 -
CV o/o 30 26 23

Score (1--low, 3--high)
** Best estimate of original cycle
*** Experimental variety (40/o family selection)
1 Planting dates arranged so all cycles flowered at or near ( 1-2 days) same time
2 Selected for yield, stability and wide adaptation through international progeny testing system

experimental variety under stress was associated not with a
higher harvest index but with a tendency for greater total
dry matter and more ears per 100 plants.
Data in Table 10 show the relationships between a
number of characters which may influence the number of
ear-bearing plants (and hence harvest index) and grain yield
under stress. The variation in each character is that which
exists between the various selections of Tuxpeio included
in the trial. Although not all of the correlations are

significant (n=10), the trends in the relationship are similar
to those found for the variation between families indicated
earlier. Thus, grain yield under stress is correlated with
flower interval (-0.71*), relative leaf elongation (0.65*)
and canopy temperature (-0.35*) measured under stress.
Tassel and stem dry weight and leaf area index, measured
under irrigated conditions, also were correlated with grain
yield under stress. Correlation values (r) were -0.58, -0.51
and -0.64* respectively.

Table 10. Statistics for Characters Associated with Grain Yield Production under Stress (Drought Resistance) for Selections
within the Population Tuxpefio-1, Tialtizapan, 1981A

Dry Matter at Flowering Relative Leaf Flower Canopy
(kg/ha)** Leaf Area Elongation Delay Temperature
Parameter Total Stem Tassel Index ** (o/o) (days) (oC)

Mean 13845 4105 504 3.44 59.5 5.3 33.3

Maximum 14995 5339 650 4.15 71.8 9.2 35.0

Minimum 12682 3682 386 2.82 50.8 3.7 31.6

F Ratio 11.7** 4.3** 8.3** 22.1** 7.82** 10.8** 5.5**

CV o/o 19.0 15.3 15.2 5.8 10.4 21.7 5.3

Correlation with Grain
Yield under Stress
(n-= 10) 0.29 -0.51 -0.58 -0.64** 0.65** -0.71* -0.35

* Significant at 0.05
** No-stress conditions


This study shows that there is an opportunity to
select within the population Tuxpeio-1 for improved yield
under stress (drought resistance), while maintaining its
relatively high yield potential. While the findings are for
only one population exposed experimentally to a specific
type of drought, the procedures used are applicable to
other maize populations and localities.
Improvement in drought resistance is much
more rapid when the selection procedure uses more
characters than just grain yield per se; many factors are
involved in conferring drought resistance. The use of a
selection index based on relative leaf elongation, the
inverval between pollen shed and silking, canopy
temperature, leaf area loss, and grain yield under stress and
no stress resulted in maximum gain per cycle in grain yield
under stress (9.5 percent).
Selection for drought resistance requires a field
site with uniform and controlled moisture and sufficient
area for adequate replication of progenies. All of the
criteria used can be measured rapidly and are therefore
suitable for screening a large number of progeny and, with
the exception of canopy temperature, require no
sophisticated instrumentation. Relative leaf elongation,
canopy temperature and the interval between pollen shed
and silking can all be measured prior to pollination and,
when combined, account for 54 percent (R2 = 0.54*) of
the variation in grain yield under stress conditions.
Morphological selection for reduced height,
tassel size and leaf size, made under nonstress conditions,

also improved grain yield under stress. These criteria are
easily incorporated into a breeding program and can be used
for individual plant selection. Tuxpeio-1 (Pop. 21),
improved for general agronomic characters and wide
adaptability through the International Progreny Testing
System, showed high gain per cycle in grain yield under
nonstress conditions and also improvement in yield under
stress conditions (although gain per cycle was lower).
In all of the selections studied, most of the yield
improvement was a result of those processes which reduced
barrenness and increased harvest index. There was, however,
some indication that the selection index also changed the
rooting pattern and enhanced total dry matter production.
Since all of the selection criteria may alter components
of drought resistance, there appears to be a need to
incorporate all characters into the selection process. An
ideal breeding program would simultaneously evaluate
progeny for criteria used in the selection index and make
use of within-family (individual plant selection) variation
for the desirable morphological traits during the recombina-
tion cycle.
Significant gains in yield under stress conditions
were achieved after only three cycles of recurrent selection.
Continuing the selection and accumulation of genes for
drought resistance traits, while at the same time maintain-
ing yield potential, should lead to further improvement
in yield under stress. There should be ample variation for
these traits within the already available productive and
well-adapted tropical germplasm.


Acevedo, E., T.C. Hsiao, and D.W. Henderson. 1971. Im-
mediate and subsequent growth responses of maize
leaves to changes in water status. Plant Physiol.
Andre, M., D. Massimono, and A. Daguenet. 1978. Daily
patterns under the life cycle of a maize crop. 1.
Photosynthesis, transpiration, respiration. Physiol.
Plant. 43:397-403.
Arboleda-Rivera, F., and W.A. Compton. 1974. Differential
response of maize (Zea mays. L.) to mass selection
in diverse selection environments. Theor. Appl.
Genetics 44:77-81.
Ashton, T. 1948. Technique for breeding for drought resis-
tance in crops. Comm. Bureaux of Plant Breeding.
Tech. Comm. 14. Cambridge, England.
Austin, R.B., I.E. Henson, and S.A. Quarrie. 1981. Abscisic
acid and drought resistance in wheat, millet, and
rice. Paper presented: Symposium on Principles and
Methods of Crop Improvement for Drought Resis-
tance: with emphasis on rice. IRRI.
Beadle, C.L., K.R. Stevenson, H.H. Neumann, G.W. Thurtell
and K.W. King. 1973. Diffuse resistance, transpira-
tion, and photosynthesis in single leaves of corn and
sorghum in relation to leaf water potential. Can. J.
Plant Sci. 53:537-544.
Blum, A. 1979. Genetic improvement of drought resistance
in crop plants: a case for sorghum. Pp. 429-445.
Mussell, H., and R.C. Staples, eds. Stress Physiology
in Crop Plants. Wiley, Interscience. New York.
Boyer, J.S. 1970a. Differing sensitivity of photosynthesis to
low leaf water potentials in corn and soybeans.
Plant Physiol.46:816-820.
Boyer, J.S. 1970b. Leaf elongation and metabolic rates in
corn, soybean, and sunflower at various leaf water
potentials. Plant Physiol. 46:233-235.
Boyer, J.S., and H.G. McPherson. 1975. Physiology of water
deficits in cereal crops. Adv. Agron. 27:1-23.
Buren, L.L., J.J. Mock, and I.C. Anderson, 1974. Morpho-
logical and physiological traits in maize associated
with tolerance to high plant density. Crop Sci.
Castleberry, R.M., and R.J. Lerette. 1979. Latente, a new
type of drought tolerance. 35th Ann. Corn and
Sorghum Res. Conf. 35:46-56.
CIMMYT (International Maize and Wheat Improvement
Center). 1974. Results of International Variety Trial
13. CIMMYT, Mexico.
CIMMYT (International Maize and Wheat Improvement
Center). 1981. CIMMYT Report on Maize Improve-
ment 1978-1979. CIMMYT, Mexico.
Daynard, T.B. 1968. Relationships of certain phenotypic
characteristics to population-tolerance in corn. PhD
Thesis. Univ. of Guelph, Ontario, Canada.

Denmead, O.T., and R.H. Shaw, 1960. The effects of soil
moisture stress at different stages of growth on the
development and yield of corn. Agron. J. 52:272-
Dow, E.W. 1981. Resistance to drought and density stress in
Canadian and European maize (Zea mays. L.)
hybrids. M.Sc. Thesis. Univ. of Guelph, Ontario,
Downey, L.A. 1971. Water use by maize at three plant
densities. Exp. Agric. 7:161-169.
Duvick, A.N. 1977. Genetic rates of gain in hybrid maize
yields during the past 40 years. Maydica XXII:
Eastin, J.D. 1980. Sorghum development and yield. Sympo-
slum on "Potential Productivity of Field Crops
under Different Environments," IRRI, Philippines.
Finlay, K.W., and G.N. Wilkinson. 1963. The analysis of
adaptation in a plant breeding programme. Aust. J.
Agric. Res. 14:742-754.
Fischer K.S., and G. Edmeades. 1977. Performance of
tropical maize synthetics selected under simulated
drought. Agron. Abstr. p. 42.
Fischer, K.S., and G.L. Wilson. 1975. Studies of grain produc-
tion in sorghum bicolor (L. Moench). III. The
relative importance of assimilate supply, grain
growth capacity and transport system. Aust. J.
Agric. Res. 26:11-23.
Fischer, K.S., and A.F.E. Palmer. 1980. Yield efficiency in
tropical maize. Symposium on "Potential Produc-
tivity of Field Crops under Different Environments,"
IRRI, Philippines.
Fischer, R.A., and N.C. Turner. 1978. Plant productivity in
the arid and semiarid zones. Ann. Rev. Plant
Physiol. 29: 277-317.
Fischer, R.A., and P.C. Wall. 1976. Wheat breeding in Mexico
and yield increases. J. Aust. Inst. Agric. Sci. 42:139-
Fischer, R.A., and M. Sanchez. 1979. Drought resistance in
spring wheat cultivars. II. Effects on plant water
relations. Aust. J. Agric. Res. 30:801-814.
Fischer, R.A., and J.T. Wood. 1979. Drought resistance in
spring wheat cultivars. III. Yield associations with
morpho-physiological traits. Aust. J. Agric. Res.
Glover, J. 1959. The apparent behaviors of maize and sorghum
stomata during and after drought. J. Agric. Sci.
(London) 153:412-416.
Hallauer, A.R., and A.F. Troyer. 1972. Prolific corn hybrids
and minimizing risk of stress. 27th Ann. Corn and
Sorghum Res. Conf. 27:140-157.
Hanson, A.D., C.E. Nelson, and E.H. Everson. 1977. Evalua-
tion of free proline accumulation as an index of
drought resistance using two contrasting barley
cultivars. Crop Sci. 17:720-726.

Hsiao, T.C., E. Fereres, E. Acevedo, and D.W. Henderson.
1976. Water stress and dynamics of growth and
yield of crop plants. Ecological Studies, Analysis
and Synthesis 19:281-305.
Hunter, J.W., H.H. Laude, and A.M. Brunson. 1936. A
method for studying resistance to drought injury
in inbred lines of maize. J. Am. Soc. Agron. 28:694-
Hurd, E.A. 1976. Plant Breeding for drought resistance.
Pages 317-353 in Water Deficits and Plant Growth.
T.T. Kozlowski, ed. New York: Academic Press.
Jackson, R.D., R.J. Reginato, and S.B. Idso. 1977. Wheat
canopy temperature; a practical tool for evaluating
water requirements. Water Res. 13:651-656.
Jensen, S.D. 1971. Breeding for drought and heat tolerance
in corn. 26th Ann. Corn and Sorghum Res. Conf.
Johnson, D.A., and K.H. Asay. 1978. A technique for
assessing seedling emergence under drought stress.
Crop Sci. 18:520-522.
Johnson, E.C., and K.S. Fischer. 1979. Ideas for the improve-
ment of efficiency of maize varieties. Proc. XXV
Ann. Meeting of the Central Amer. Coop. Prog. for
the Improvement of Food Crops (PCCMCA).
Tegucigalpa, Honduras.
Johnson, G.R., and KJ. Frey. 1967. Heritabilities of quanti-
tative attributes of oats (Avena sp.) at varying levels
of environmental stress. Crop Sci. 7:43-46.
Johnson, G.R., and K.J. Frey. 1967. Heritabilities of quanti-
tative attributes of oats (Avena sp.) at varying levels
of environmental stress. Crop Sci. 7:43-46.
Jones, H.G. 1979. Stomatal behaviour and breeding for
drought resistance. Pages 408-427 in Mussell, H. and
R.C. Staples, ed. Stress Physiology in Crop Plants.
Wiley, Interscience. New York.
Jones, M.M., and N.C. Turner. 1978. Osmotic adjustment in
leaves of sorghum in response to water deficits.
Plant Physiol. 61:122-126.
Jurgens, S.K., R.R. Johnson, and J.S. Boyer. 1978. Dry
matter production and translocation in maize
subjected to drought during grain fill. Agron. J.
Kassam, A.H., J. Kowal, M. Dagg, and M.N. Harrison 1975.
Maize in West Africa and its potential in savanna
areas. World Crop 27:73-78.
Kassam, A.H. 1976. Crops of the West African semi-arid
tropics. International Crops Research Institute for
the Semi-Arid Tropics. India.
Kilen, T.C., and R.H. Andrew. 1969. Measurement of drought
resistance in corn. Agron. J. 61:669-672.
Kretchmer, P.J., D.R. Laing, and S. Zuluaga. 1980. Uso del
termometro infrarrojo para la selecci6n por tole-
rancia a sequfa en Phaseoulus vulgaris L. Proc. XXIII.
PCCMCA. Guatemala, Guatemala.
Larque-Saavedra, A., and R.L. Wain. 1974. Abscisic acid
levels in relation to drought tolerance in varieties of
Zea mays L. Nature 251:716-717.

Levitt, J. 1972. Responses of plants to environmental stresses.
New York: Academic Press.
Ludlow, M.M. 1976. Ecophysiology of C4 grasses. Ecological
Studies, Analysis and Synthesis 19:364-386.
Martin, J.H. 1930. The comparative drought resistance of
sorghum and corn. Agron. J. 22:993-1003.
Martin, J.H., J.A. Harris, and I.D. Jones. 1931. Freezing
point depression and specific conductivity of sorghum
tissue fluids. J. Agr. Res. 42:57-69.
McPherson, H.G., and J.S. Boyer. 1977. Regulation of grain
yield by photosynthesis in maize subjected to a
water deficiency. Agron. J. 69:714-718.
Mosino, A.P., and E. Garcia. 1968. Evaluaci6n de la sequfa
intraestival en la Repiblica Mexicana. C.P.E.N.A.
Chapingo, M6xico.
Moss, D.N., J.T. Woolley, and J.F. Stone. 1974. Plant modifi-
cation for more efficient water use: the challenge.
Agric. Meteor. 14:311-320.
Muleba, N. Visiting scientist, International Maize and Wheat
Improvement Center (CIMMYT), Mexico. Pers.
Mungomery, V.E., R. Shorter, and D.E. Byth. 1974. Genotype
x environmental adaptation. I. Pattern analysis-
application to soya bean populations. Aust. J. Agric.
Res. 25:59-72.
Mufoz, O.A. 1975. Relaciones agua-planta bajo sequfa, en
various sintdticos de maiz resistentes a sequoia y
heladas. Tesis de doctor en Ciencias, Colegio de
Postgraduados, ENA. Chapingo, M6xico.
Musick, G.J., M.L. Fairchild, V.L. Ferguson, and M.S. Zuber.
1965. A method for measuring root volume in corn
(Zea mays L.). Crop Sci. 5:601-602.
Nass, H.G., and M.S. Zuber. 1971. Correlation of corn (Zea
mays L.) roots early in development to mature root
development. Crop Sci. 11:655-658.
Neumann, H.H., G.W. Thurtell, and K.R. Stevenson. 1974.
Leaf water content and potential in corn, sorghum,
soybean, and sunflower. Can. J. Plant Sci. 54:185-
OToole, J.C., and T.T. Chang. 1979. Drought resistance in
cereals: rice a case study. Mussell, H., and R.C.
Staples, eds. Stress Physiology in Crop Plants. Pp.
373-405. Wiley, Interscience. New York.
Palacios de la R., G. 1959. Variedades e hfbridos de mafz
latente y tolerantes a la sequfa y a las heladas.
Mexico. Agr. 107:38-39.
Paliwal, R.L., and E.W. Sprague. 1981. Improving adaptation
and yield dependability in maize in the developing
world. CIMMYT, Mexico. Technical Conference on
Improved Seed Production, Nairobi, Kenya, June
2-6, 1981.
Parker, J. 1968. Drought resistance mechanisms. Pages
195-234 in Water Deficits and Plant Growth. T.T.
Kozlowski, ed. New York: Academic Press.
Parmer, M.T., and R.P. Moore. 1968. Carbowax 6000,
manitol and sodium chloride for stimulating drought
conditions in germination studies of corn (Zea mays
L.) of strong and weak vigor. Agron. J. 60:192-195.

Passioura, J.B. 1972. The effects of root geometry on the
yield of wheat growing on stored moisture. Aust. J.
Agric. Res. 23:745-752.
Passioura, J.B. 1981. The interaction between the physiology
and the breeding of wheat. Evans, L.T. and W.J.
Peacock, eds. Wheat Science-Today and Tomorrow.
Pp. 191-200. Univ. Press, Cambridge.
Pinter, L., L. Kalman, and G. Palfi. 1978. Determination of
drought resistance in different genotype hybrids of
maize (Zea Mays L.) by field trials and biochemical
test. Maydica XXIII:121-127.
Poey, F.R., J.E. Grajeda, O.J. Fernandez, and F. Soto. 1977.
Effect of detasseling on maize grain yield compo-
nents. Agron. Abstr. p. 44.
Prine, G.M. 1971. A critical period of ear development in
maize. Crop Sci. 11:782-786.
Qualset, C.O. 1979. Breeding for drought resistance in maize.
Proc. SAFGRAD/International Institute for Tropical
Maize Workshop. Ougadougou, Upper Volta.
Robins, J.S., and L.E. Domingo. 1953. Some effects on severe
soil moisture deficits at specific stages in corn.
Agron. J. 45:618-621.
Russell, W.A. 1974. Comparative performance for maize
hybrids representing different areas of maize
breeding. 29th Ann. Corn and Sorghum Res. Conf.
Sanchez-Diaz, M., M. Moreyard, F. Gonzalez-Fernandez.
1969. Physiological responses of sorghum and
maize (corn) leaves to water stress I. Transpiration
and leaf elongation responses (in Spanish, English
summary). Anales de Edafologia y Agrobiologia,
Sanchez-Diaz, M.F., and P.J. Kramer. 1971. Behaviour of
corn and sorghum under water stress and during
recovery. Plant Physiol. 48:613-616.
Schwarzbach, E.N. 1976. Seminar presented on screening of
top entries from across locations: data and evaluation
of selections for many characters simultaneously,
March 15, 1976, CIMMYT, Mexico.
Shaw, R.H. 1977. Climatic requirement. Corn and Corn
Improvement. Pp. 591-623. G.F. Sprague, ed. Am.
Soc. Agron. 18.
Singh, T.N., D. Aspinall, and L. Paleg. 1972. Proline ac-
cumulation and varietal adaptability to drought in
barley: a potential metabolic measure of drought
resistance. Nature New Biology 236:188-189.
Spencer, J.T. 1940. A comparative study of a seasonal root
development of some inbred lines and hybrids of
maize. J. Agric. Res. 61:521-538.
Stout, D.G., and G.M. Simpson. 1978. Drought resistance of
sorghum bicolor. 1. Drought avoidance mechanisms
related to leaf water status. Can. J. Plant Sci.
Sullivan, C.Y., and A. Blum. 1970. Drought and heat toler-
ance of corn and sorghum. 25th Ann. Corn and
Sorghum Res. Conf. 25:55-66.

Tatum, L.A. 1954. Breeding for drought and heat tolerance.
9th Ann. Corn and Sorghum Res. Conf. 9:22-28.
Chicago, Illinois.
Thompson, D.L. 1968. Field evaluation of corn root clumps.
Agron. J. 60:170-172.
Tollenaar, M. 1977. Sink-source relationships during repro-
ductive development in maize. A review. Maydica
XXII: 49-75.
Turner, N.C. 1974. Stomatal response to light and water
under field conditions. R.L. Bieleski, A.R. Ferguson,
and M.M. Creswell, eds. Mechanisms of regulation of
plant growth. Bull. 12, pp. 423-432. Royal Society
of New Zealand, Wellington.
Turner, N.C., J.E. Begg, and M.L. Tonnet. 1978. Osmotic
adjustment of sorghum and sunflower crops in
response to water deficits and its influence on the
water potential at which stomata close. Aust. J.
Plant Physiol. 5:197-608.
Vasal, S.K., A. Ortega, and S. Pandey. 1978. Population
improvement and varietal development in CIMMYT's
maize program. In Proc. of XVth Annual Meeting of
Caribbean Food Crops Society.
Walker, S.T. 1980. Decision making by farmers and by the
national agricultural research program on the
adoption and development of maize varieties in El
Salvador. PhD Thesis. Stanford Univ., California.
Watts, W.R. 1974. Leaf extension in Zea mays III. Field
measurements of leaf extension in response to
temperature and leaf water potential. J. Expt. Bot.
Wellhausen, E.J. 1956. Improving American corn with exotic
germplasm. 11th Ann. Corn and Sorghum Res,
Conf. 11:85-96. Chicago, Illinois.
Whiteman, P.C., and G.L. Wilson. 1965. Effects of water
stress on the reproductive development of Sorghum
vulgare. Pers. Univ. Queensland Papers, Dept. Bot.
Williams, T.V., R.S. Snell, and J.F. Ellis. 1967. Methods of
measuring drought tolerance in corn. Crop Sci.
Wilson, D. 1975. Leaf growth, stomatal diffusion resistances
and photosynthesis during droughting of Lolium
perenne populations selected for contrasting stomatal
length and frequency. Ann. Appl. Biol. 79:67-82.
Wolf, J.M., G. Levine, G.C. Naderman, and E. Gonzalez E.
1974. Adverse soil-water conditions and corn
production in Central Brazil. Abstracts of the 66th
Annual Meeting of the American Society of
Agronomy. P. 47. Chicago, Illinois.
Worrall, W.D., N.H. Scott, A.R. Klatt, and S. Rajaram. 1980.
Performance of CIMMYT wheat germplasm in
optimum and sub-optimum production environ-
ments. International Wheat Conference, Madrid,
Yoshida, S., and E. de los Reyes. 1976. Leaf cuticular
resistance of rice varieties. Soil Sci. Plant Nutr.

Correct Citation: K. S. Fischer, E. C. Johnson, and G. O. Edmeades,
1983. Breeding and Selection for Drought Resistance in Tropical
Maize. Centro Internacional de Mejoramiento de Ma(z y Trigo
CIMMYT. El Batin, Mdxico.

The International Maize and Wheat Improvement Center (CIMMYT)
receives support from government agencies of Australia, Canada,
Denmark, European Economic Commission, Federal Republic of
Germany, France, India, Ireland, Japan, Mexico, The Netherlands,
Norway, The Philippines, Saudi Arabia, Spain, Switzerland, United
Kingdom, and USA; and from the Ford Foundation, Inter-American
Development Bank, OPEC Fund for International Development,
Rockefeller Foundation, United Nations Development Programme,
and the World Bank. Responsibility for this publication rests solely
with CIMMYT.


CIMMYT International Maize and Wheat Improvement Center

The CIMMYT International Maize and Wheat Improvement
Center holds all rights to the source text and shall be
considered the copyright holder for the text and images of
these publications.

The CIMMYT International Maize and Wheat Improvement
Center has made this publication available to the University of
Florida, for purposes of digitization and Internet distribution.

The CIMMYT International Maize and Wheat Improvement
Center reserves all rights to this publication. All uses,
excluding those made under "fair use" provisions of U.S.
Code, Title 17, Section 107 are restricted.

Contact the CIMMYT International Maize and Wheat
Improvement Center for additional information and

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs