GAS EXCHANGE AND WATER USE EFFICIENCY OF GRAIN SORGHUM

by

SHAOBING PENG, B.S., M.S.

A DISSERTATION

IN

AGRICULTURE

Submitted to the Graduate Faculty of Texas Tech University in

Partial Fulfillment of the Requirements for

the Degree of

DOCTOR OF PHILOSOPHY

Approved j ,

May, 1990

ACKNOWLEDGEMENTS

I would like to express my special thanks to Dr. Daniel R. Krieg, my major

professor, who makes me understand science not only from textbooks but from the

practices throughout my graduate smdy program.

Special thanks are also extended to my advisory members, Drs. Anthony E. Hall,

Ronald E. Sosebee, Henry T. Nguyen, and Robert J. Lascano, for their suggestions

and criticism during the research and preparation of this manuscript.

Sincere thanks to Dr. Fekade S. Girma for the time he spent with me on the

discussion of my research projects. I am also grateful for the assistance provided by

James, Sakis, Saranga, Pankaja, Mike, and Chad during these four years.

It is my duty to thank my family and Mr. Zhang, my high school teacher, for their

kind blessings and support. I must also express my deepest thanks to my fiancee, Jing,

for her encouragement and love.

Finally, I would like to express thanks to my parents, who never went to school

themselves but have supported me as I have pursued my education.

XX

CONTENTS

ACKNOWLEDGEMENTS ii

ABSTRACT v

LIST OF TABLES vii

LIST OF HGURES viii

PREFACE X

CHAPTER

I. GENOTYPIC VARIATION FOR WATER USE EFHCIENCY 1

Introduction 1

Materials and Methods 2

Results and Discussion 5

References 10

II. GENOTYPIC VARIATION FOR GAS EXCHANGE TRAITS AND

THEIR RELATIONSHIP WITH WATER USE EFHCIENCY 18

Introduction 18

Materials and Methods 19

Results and Discussion 21

References 26

m . CANOPY GAS EXCHANGE AND ITS RELATIONSHIP WITH

SINGLE LEAF RESPONSES 33

Introduction 33

Materials and Methods 34

Results and Discussion 36

References 39

i i i

IV. GENERAL SUMMARY AND CONCLUSIONS 46

i v

ABSTRACT

Grain sorghum [Sorghum bicolor (L.) Moench] production is usually confined to

environments which are considered too dry and hot for other cereals to be productive.

Increasing yield and water use efficiency (WUE) of grain sorghum through

identification and utilization of superior germplasm in breeding programs is a major

goal in the improvement of this crop. Breeding for increased WUE has been limited by

the lack of screening criteria and methods that could be used to select desirable

genotypes from large populations under field conditions.

This study was conducted to determine whether genotypic differences in WUE

and gas exchange traits are present and to determine whether the variation in gas

exchange rates and efficiencies reflected the genotypic differences in WUE and thereby,

could be used as a screening criteria for identifying genotypes of grain sorghum with

higher WUE.

Five sorghum genotypes (TX 378, TX 430, SC 35, TX 399, and TX 2741)

which are parental lines used in hybrid seed production were tested in the greenhouse

and field during 1989 under well watered conditions. In the greenhouse study,

individual entries were planted in plastic pots containing 10 kg of sterilized potting mix.

The field experiment was conducted on an Amarillo loamy fine sand. In boUi studies,

shoot biomass production, amount of water used, and leaf area development were

monitored during die growing season. Water use efficiency was calculated as shoot

biomass production per unit water use. Gas exchange measurements were made

throughout the vegetative stage on uppermost fully expanded leaves and whole canopy

using a portable photosynthesis system. Gas exchange efficiency was expressed as the

ratio of photosynthetic rate (A) to transpiration rate (T).

The sorghum genotypes exhibited significant variation for A, A/T, shoot biomass

production, and WUE. TX 378, TX 430, and SC 35 had higher A, A/T, shoot biomass

production, and WUE than TX 399 and TX 2741. No consistent genotypic variation

was observed for T or whole plant water use rates. Single leaf measurements of gas

exchange traits (A and A/T) reflected single plant and whole canopy WUE differences

among the genotypes largely due to the positive correlation between A and shoot

biomass production. In addition, there was a positive correlation among the genotypes

between leaf area and A. The results indicate that measurements of A and leaf area may

be used to select for increased WUE in grain sorghum.

vx

UST OF TABLES

1.1. Average daily shoot biomass production (BP), maximum leaf area QM^A), average daily water use (WU), water use efficiency (WUE), and shoot biomass production per unit leaf area (BP/LA) of five genotypes in the greenhouse (GH) and field (FD) smdies. 11

2.1. Mean carbon assimilation rate (A), transpiration rate (T), and ratio of A to T of five genotypes. F-values based on analyses of variance for the effect of genotype (G) and genotype x environment (G x E) interaction on A, T, and A/T ratio. 28

2.2. Correlation matrix for gas exchange and agronomic traits of five sorghum genotypes in the greenhouse (GH) and field (FD) studies. 29

2.3. Mean transpiration rate (T), vapor pressure difference between the leaf and air (VPD), cuvette relative humidity (RH), and leaf temperatiu^ (LT) of five genotypes measured by a closed system (LI-6200) and an open system (LI-1600). 30

v i i

LIST OF FIGURES

1.1. Leaf area development (a) and cumulative water use (b) of five sorghum genotypes in the greenhouse study. Vertical bars indicate LSD (0.05). 12

1.2. Leaf area index (a), cumulative shoot biomass production (a), and cumulative water use (c) of five sorghum genotypes during the period firom 35 DAP to flowering in the field smdy. Vertical bars indicate LSD (0.05). 13

1.3. Relationship of average daily shoot biomass production (a) and average daily water use (b) during the period from 35 DAP to flowering with average l e ^ area over the same period in the field study. Each point represents individual plot (three replications x five sorghum genotypes. 14

1.4. Water use efficiencies of five sorghum genotypes during the period from 35 DAP to flowering versus their average leaf areas over the same period in the field study. 15

1.5. Relationship of average daily shoot biomass production (a) and average daily water use (b) during the period from emergence to flowering with average leaf area over the same period in the greenhouse study. Each point represents individual plant (four replications x five sorghum genotypes). 16

1.6. Water use efficiencies of five sorghum genotypes during the period from emergence to flowering versus their shoot biomass production per unit average leaf area (BP/LA) calculated during the same period in the greenhouse smdy. 17

2.1. Diumal changes in carbon assimilation rates (A) of TX 378 and TX 399 on 51, 58, and 65 days after planting (DAP) in die greenhouse smdy. Vertical bars indicate LSD (0.05). 31

2.2. Water use efficiencies of five sorghum genotypes versus their ratios of carbon assimilation rate to transpiration rate (A/T). Larger and smaller symbols represent genotypic means in the field and greenhouse smdy, respectively. 32

3.1. Diumal changes in (a) photosynthetic photon flux density (PPFD), (b) air temperature (AT), and (c) saturation vapor pressure dificit (VPD) of ambient air when diumal gas exchange was measured 35 days after planting. 40

v i i i

3.2. Diumal patterns of single leaf (A) and canopy photosynthetic rates per unit ground area (CAP/G) 35 days ^ter planting. Data were pooled across two sorghum genotypes. Vertical bars indicate standard errors of the means. 41

3.3. Diumal pattems of single leaf (T) and canopy transpiration (CT) 35 days after planting. Data were pooled across two sorghum genotypes. Vertical bars indicate standard errors of the means. 42

3.4. Seasonal changes in (a) canopy photosynthetic rates per unit ground area (CAP/G) and (b) leaf area index (LAI). Data were pooled across two sorghum genotypes. Vertical bars indicate standard errors of the means. 43

3.5. Seasonal changes in (a) canopy transpiration rates (CT) and (b) saturation vapor pressure dificit (VPD) of ambient air. Data for CT were pooled across two sorghum genotypes. Vertical bars indicate standard eirors of the means. 44

3.6. Comparison of TX 378 and SC 35 in (a) single leaf (A) and (b) canopy photosynthetic rates per unit leaf area (CAP/L). Vertical bars indicate LSD (0.05). 45

XX

PREFACE

A major concern of grain sorghum [Sorghum bicolor (L.) Moench] research has

been the question "can water use efficiency (WUE) be improved by identification and

development of superior genotypes?" Breeding for increased WUE has been limited

due to limited information concerning genotypic variation in WUE within species and

due to the lack of screening criteria and methods that could be used to select desirable

genotypes from large populations under field conditions.

Water use efficiency can be defined in several ways. The basis definition of WUE

is the amount of biomass produced per unit water use. In physiological terms, gas

exchange efficiency defined as the ratio of photosynthesis (A) to transpiration (T) is

suppose to estimate WUE. However, the relationship between the instantaneous

measurements of gas exchange efficiency and long-term WUE has not been well

established.

Our previous efforts have demonstrated the existence of genetic variation in not

only leaf area per plant, biomass production, grain yield, and seed number per unit leaf

area, but also A and gas exchange efficiency in grain sorghum. The ftirther questions

which have to be addressed arc whether the genetic variation in gas exchange rates at

the single leaf level translates into whole canopy differences and whether the genotypic

differences instantaneous gas exchange efficiency reflects the long-term WUE

differences among the genotypes of grain sorghum. If the instantaneous gas exchange

measurements at the single leaf level reflect the long-term plant performance, then they

can be used as evaluation criteria for determination of genetic or management

approaches for crop improvement

This study was undertaken with four major objectives:

1. To determine whether genotypic differences in WUE were present in parental

lines of grain sorghum and the causes for these differences.

2. To determine the genotypic variation in gas exchange traits among the five

genotypes.

3. To determine whether gas exchange measurements accurately reflect genotypic

variation in seasonal WUE.

4. To determine the diumal and seasonal responses of canopy gas exchange rates

as compared to single leaf gas exchange rates of grain sorghum.

X I

CHAPTER I

GENOTYPIC VARIATION FOR WATER USE

EFHCIENCY

Introduction

On a world-wide basis, grain sorghum [Sorghum bicolor (L.) Moench]

production is usually confined to low rainfall areas where water stress represents the

major growth-limiting factor. In the Great Plains where over 85% of U. S. sorghum is

grown, the annual rainfall is less than 500 mm and less than 20% of the sorghum area

receives irrigation. Under irrigated conditions, the cost of the water supply (pumping

and distribution) represents the major production cost and reduces the profit potential of

the crop (Krieg, 1988). Therefore, if water use efficiency (WUE) defined as shoot

biomass production per unit water use can be increased through identification and

utilization of superior germplasm and development of proper management strategies,

yields should increase within the limits of the water supply.

Krieg (1988) has summarized that management approaches offer the greatest

short-term opportunities to increase grain yield within limits of the existing water

supplies. One approach is to optimize the plant population such that the rate of water

use does not exceed the potential for resupply. Another approach is to optimize the

planting date such that the oppormnity for plant water stress during the critical panicle

differentiation period is minimized and seed number per unit leaf area is maximized.

Canity et al. (1982a) demonstrated that WUE was variable depending upon the growth

stage when stress occurred. Stress during the period from the end of bloom to

physiological maturity reduced yield to a greater extent than water use and therefore

reduced WUE. Several studies indicated that narrow-row sorghum produced higher dry

matter and grain yields for a given level of evapotranspiration (ET) under irrigated

conditions (Grimes and Musick, 1960; Porter et al., 1960). Stewart et aL (1983)

reported that the limited irrigation-dryland farming system for the conjunctive use of

rainfall and limited irrigation of graded furrows increased irrigation WUE by 67% over

conventional irrigation practices.

A major concern of sorghum research has been the question "can WUE be

improved by identification and development of superior genotypes?" To address this

question, the existence of genotypic variation in WUE must be demonstrated. Briggs

:ind Shantz's (1913) pioneering work revealed not only differences between species for

biomass produced per unit water consumed, but also differences within species (the

sorghums in particular). Garrity et al. (1982b) evaluated the yield-ET relationships of

three sorghum hybrids as a function of water supply. The slopes of the yield-ET

relationships differed significantiy among the three hybrids ranging from 11.8 to 19.2

Kg ha"l mm'l . However, breeding programs have made no direct attempts to increase

WUE due to limited information conceming genotypic differences in WUE within

species and environmental influences on WUE. Therefore, the objectives of this study

were to determine whether genotypic differences in WUE were present in five parental

lines of grain sorghum and the causes for these differences.

Materials and Methods

Greenhnnsft smHy Five sorghum genotypes (TX378, TX430, SC35, TX399 and

TX2741) which are parental lines used in hybrid seed production were grown in a

complete randomized block design with four replications. These lines offer

considerable genetic variation in plant size, leaf area development, yield potential, and

photosyntiietic capacity (Hutmacher, 1983; Kidambi, 1987; Krieg and Dalton, 1988).

The greenhouse was located on the campus of Texas Tech University, Lubbock,

Texas. All entries were hand-planted in plastic pots containing about 10 kg of sterilized

potting mix on 7 March 1989. The potting medium consisted of a 1:2:1:0.5 (v/v/v/v)

mixmre of peat moss/Ten-egreen clay (Quaker State, Lubbock, TX) /perlite/vermiculite.

Before planting, the pots were over-watered and allowed to drain until they reached a

constant weight hereafter referred to as field capacity. Pot mass was measured with a

mechanical scale (Toledo 1072, Toledo, OH) with increments of 5 g. After seedling

emergence, the most vigorous plant in each pot was kept. Deionized water was applied

every day to bring the pot back to its field capacity weight. The surface of the pot was

covered with styrofoam packing material to reduce evaporation. Nitrogen was applied

weekly with the irrigation water at a rate of 4 g N L'^ water.

Intact plant leaf area was measured weekly using a portable area meter (LI-3000,

LI-COR Inc., Lincoln, NE). The plants were harvested at flowering and separated into

stems and leaves. Dry masses were determined following drying at 70 °C for 72 hoiu-s.

The total amount of water used was calculated as the difference between final and

initial pot mass plus the total amount of water applied, minus the total amount of

evaporation from the pot surface during the growing season. The evaporation from the

covered |X)t surface was determined from four control pots which did not have plants.

Field smdies The same five genotypes used in the greenhouse smdy were grown

in the field during the 1989 growing season at the Plant Stress Laboratory, Brownfield,

Terry County, Texas. The soil was an Amarillo loamy fine sand (fine, loamy, mixed,

thermic Aridic Paluestalfs). All entries were grown in a complete randomized block

design with three replicated plots under well watered conditions. Supplemental

irrigation was provided by a trickle irrigation system. Each entry was planted on 14

June 1989 in plots four rows wide (with 0.75 m between rows) and 12 m long. The

plant density after thinning was 24 plants m"2, A fertilizer mixture consisting of 50 kg

.S and 20 kg P ha"^ was applied preplant. Additional N was provided in the irrigation

water at a rate of 1 kg N ha'l mm"!.

Growth analysis was done from 0.5 m^ ground area biweekly starting 35 days

after planting (DAP) until flowering to determine leaf area and shoot biomass. Shoot

biomass production was expressed as aboveground dry mass per m^ area or on an

individual plant basis. A representative plant was selected from each growth sample to

Jciermine leaf area with an area meter (LI-3100, LI-COR Inc., Lincoln, NE). Total leaf

;irea of the sample was determined from the specific leaf weight of the single plant. The

amount of water used was determined by the water balance method: water use = soil

water at beginning of experiment + rainfall -i- irrigation - soil water remaining at the end

of experiment. Gravimetric analysis of the top 1 m of the soil (at the depth interval of

0.33 m) was used to determine soU water content for each plot every two weeks. Soil

bulk densities of 1.49, 1.51, and 1.48 g cm^ were used to convert gravimetric to

volumetric water content for the top, middle, and bottom depths, respectively. The soil

bulk density was determined by standard laboratory analyses (USDA Textural Scheme)

using pressure plate and pressure membrane apparatus. Drainage and runoff were

assumed to be negligible.

Water use efficiency was defined as the total shoot biomass produced from

emergence to flowering divided by the total water transpired over the same period in the

greenhouse smdy. In the field smdy, WUE was calculated as the shoot biomass

production over the period from 35 DAP to flowering divided by the total water

evapotranspired over die same period.

The statistical sigruficance of genotypic variation in shoot biomass production,

leaf area, water use, and WUE was determined by analysis of variance. Duncan's

::-iultiple range-test was used to separate the genotypic means (SAS Institute, Inc.,

•;0S5), and the consistency of genotypic differences across the greenhouse and field

environments was determined by Spearman's rank correlation analysis (Lindeman et

.il.. 1980).

Results and Discussion

Cumulative water use and leaf area of the greenhouse plants are presented in

l-igure 1.1. TX 430 had the slowest rate of leaf area development diuing the early part

of the growing season, but increased rapidly at 45 DAP and surpassed the leaf areas of

the rest of the genotypes at 60 DAP (Fig. 1.1a). TX 378, SC 35, and TX 399 produced

significantiy more leaf area than TX 2741 after 45 DAP. The pattems of cumulative

V. ater use for the five genotypes were basically similar to the pattems of leaf area

development (Fig. Lib). TX 378, SC 35, and TX 399 used almost die same amount

of water during the entire growing season. TX 430 consumed less water during the

early part of the season, but the final water use was not significantiy different from TX

378. SC 35, and TX 399. TX 2741 developed die least amount of leaf area and used

the least amount of total water.

In the field study, the five genotypes showed significant (P<0.05) differences in

leaf area development during the entire measurement period (Fig. 1.2a). TX 430 and

TX 2741 produced die greatest and least leaf area, respectively, which was consistent

^ idi the genotypic ranking of leaf area at the end of die greenhouse smdy. Significant

differences existed among die five genotypes in cumulative shoot biomass production

3s plants developed (Fig. 1.2b). TX 378 and TX 430 consistendy produced more

biomass than TX 399 and TX 2741. However, the genotypes did not exhibit

consistendy significant differences in the cumulative amount of water used (Fig. 1.2c).

Since the flowering dates differed among the five genotypes by 7 days, shoot

biomass production and water use were expressed on a daily basis calculated for the

periods from emergence to flowering for the greenhouse smdy and from 35 DAP to

Howering for the field smdy (Table 1.1). Analysis of variance revealed that the

sorghum genotypes exhibited highly significant differences (P < 0.01) in shoot

biomass production rate in both the greenhouse and field smdies. The variation in shoot

biomass production (calculated as ((maximum - niinimum)/maximum x 100)) among

the five genotypes was about 45 and 32% for the greenhouse and field study,

respectively. Even though the five genotypes exhibited larger differences in the

greenhouse study than in the field smdy, genotypic ranking in shoot biomass

production was similar across the two different environments with a rank correlation

coefficient of 0.90. Genotypic differences in maximum leaf area (leaf area at

flowering) among the five genotypes were also highly significant (P < 0.01) in the two

studies (Table 1.1). Leaf areas of greenhouse grown plants were almost twice that of

Held plants.

The amount of water used represented cumulative transpiration from emergence to

flowering in the greenhouse study and cumulative ET from 35 DAP to flowering in the

field smdy. Although the differences among the genotypes in the amount of water use

were statistically significant (P < 0.05) in each environment, the significant differences

were caused by only one genotype (TX 2741 in the greenhouse smdy and SC 35 in the

field study) with no significant differences among the other four genotypes (Table 1.1).

Water use efficiency was calculated based on the amount of shoot biomass

produced per unit water consumed. Sorghum grain yield is well correlated Avith total

shoot dry matter production (Krieg, 1988). The differences in WUE for shoot biomass

production should reflect the variation in WUE for grain production. Significant

genotypic variation in WUE existed among the five genotypes (Table 1.1). The

variation in WUE among die five genotypes was 17 and 35% for the greenhouse and

the field smdy, respectively. The relative ranking of die five genotypes in WUE was

similar across the two different environments (rank correlation coefficient was 0.70).

TX 378 and SC 35 had higher WUE, while TX 399 and TX 2741 had lower WUE in

both the greenhouse and field studies.

Higher WUE was observed in the field study compared with the greenhouse

study even though water use in the field included soil evaporation. This might be due

to: (i) the higher gas exchange efficiencies resulted from higher carbon assimilation

rates in the field smdy than in the greenhouse study, which is presented in the next

chapter, (ii) the WUE in the field was determined during the most rapid growing period

(35 DAP to flowering) whereas the WUE in the greenhouse represented the period

from emergence to flowering, and (iii) greater proportion of leaves of greenhouse

plants were intercepting solar energy and transpiring water compared to the field-grown

plants.

The regression analysis between average daily shoot biomass production and

average leaf area during the same period indicates that shoot biomass production was

strongly and positively correlated widi leaf area in die field smdy (Fig. 1.3a). The

differences in leaf area explained over 70% of the variation in shoot biomass

production, suggesting that genotypic variation in shoot biomass production was

largely due to die differences in leaf area among die genotypes. However, diere was no

relationship between water use rate and leaf area in die field smdy (Fig. 1.3b). By die

time the water use rate was determined in ±e field study, die leaf area index already

exceeded 3.0 (Fig. 1.2a). At this leaf area index, die evaporative demand dictated die

daily rate of water use. Apparentiy die differences in leaf area (Fig. 1.4) and shoot

biomass production among the genotypes were mainly responsible for the genotypic

\ ariation in WUE at the community level.

In the greenhouse smdy with individual plants, shoot biomass production and

water use were both positively cortelated with leaf area (Fig. 1.5). The differences

.imong the genotypes in WUE were not associated with the genotypic differences in leaf

.:rea. Nearly 30% of the variation in shoot biomass production of the greenhouse

:rown plants was not explained by leaf area differences (Fig. 1.5a), suggesting the

possibility of genotypic differences in shoot biomass production per unit leaf area

(BP/LA). Table 1.1 shows that the differences in BP/LA were highly significant

(P<0.01) among the five genotypes in the greenhouse study. The shoot biomass

production per unit leaf was calcidated from the average daily shoot biomass production

during the period from emergence to flowering in the greenhouse study and from 35

DAP to flowering in the field study and average leaf area over the same period.

Differences in BP/LA could result in differences in WUE since water use was highly

correlated with leaf area (Fig. 1.5b). Figure 1.6 demonstrates that BP/LA of the five

genotypes were strongly cortelated with their WUE. However, the five genotypes did

not exhibit significant differences in BP/LA in the field study (Table 1.1).

Physiologically, differences in BP/LA should result from differences in carbon

assimilation rate per unit leaf area, and the genetic differences in leaf area production

should also be a function of photosyntiietic rate. Therefore, photosynthetic capacity

"light be die ultimate control for die genotypic differences in WUE.

In summary, die sorghum genotypes exhibited significant and consistent variation

in shoot biomass and leaf area production in bodi die greenhouse and field studies.

Genotypic differences in shoot biomass production were largely due to the variation in

leaf area. Water use was not significantiy different for most of the genotypes. The

sorchum genotypes exhibited significant and consistent variation in WUE due to

si enificant genotypic variation in leaf area at the community level and due to significant

cenotypic differences in BP/LA at the single plant level.

10

References

Bricgs. L. J., and H. L. Shantz. 1913. The water requirement of plants: I. Investigations in the Great Plains in 1910 and 1911. USDA Bureau Plant Industry Bull. 284.

Carrity. D.P., D.D. Watts, C.Y. Sullivan, and J.R. Gilley. 1982a. Moisture deficits and grain sorghum performance: Effect of genotype and limited irrigation strategy. Agron. J. 74:808-814.

Garrity, D.P.. D.D. Watts, C.Y. Sullivan, and J.R. Gilley. 1982b. Moisture deficits and grain sorghum performance: Evapotranspiration-yield relationships. Aeron. J. 74:815-820.

(irimes. D.W., and J.T. Musick. 1960. Effects of plant spacing, fertility, and irrigation management on grain sorghum production. Agron. J. 52:647-650.

I lutmacher, R.B. 1983. Stomatal and nonstomatal limitations of photosynthesis in field grown cotton and sorghum. Ph.D. diss. Texas Tech Univ., Lubbock, TX.

Kidambi, S.P. 1987. Genetic control of gas exchange processes affecting water use efficiency in grain sorghum. Ph.D. diss. Texas Tech Univ., Lubbock, TX.

Krieg, D.R. 1988. Water use efficiency of grain sorghum, p. 27-41. In D. Wilkinson (ed.) Proc. 43rd Annual Com and Sorghum Industry Res. Conf., Chicago, IL. 8-9 Dec. 1988. American Seed Trade Association, Washington, D. C.

Krieg, D.R., and G. Dalton. 1988. Physiological basis of yield analyses of genetic and cultural improvements in sorghum production. In Proc. Intemationai Congress of Plant Physiology. New Delhi, India (in press).

Lindeman, R.H., P.F. Merenda, and R.Z. Gold. 1980. Introduction to bivariate and multivariate analysis. Scott, Foresman and Company, Glenview, EL.

Poner, K.B., M.E. Jensen, and W.H. Sletten. 1960. The effect of row spacing, fertilizer, and planting rate on the yield and water use of irrigated grain sorghum. Agron. J. 52:431-434.

SAS Instimte, Inc. 1985. SAS user's guide: Statistics, 5di ed. SAS Institute, Inc., Cary, NC

Stewart, B.A., J.T. Musick, and D.A. Dusek. 1983. Yield and water use efficiency of grain sorghum in a limited irrigation-dryland farming system. Agron. J. 75:629-634.

11

^ t> — _c

^ -. ^ c " ^ • • "

vise

pe

s

>- > N

- u

?3 O

•*J ' • " ^

> <

— t i • ^ —

• ^ "_)

•z ' ^

~ u \j

— ^ S c " 3

:< ci. 2 c = o

— o S CL X -y:

Ii " o ri O in

S ^'-a :5-o 2

. C l lJ h*< •1 r_i '*—^ 1 3 - n

1.1.

Ave

rage

dai

ly

use

effi

cien

cy (

W

ouse

(G

H)

and

fiel(

^-t T- 9 bt)

< i P3

O ^

<

0.

Q b

I

a

Q UH

I U

Q uu

K O

Q

X O

C/]

o c

T—t 1

cx 1

E M

1

1

"O ^ H

'o. bO

, — 1 t

P . <N

E

+-1

•a

p .

so so <n — i n vri

« J m CO

so u-i

ca as

so e n

u-5 u-i

73 . O

T t OS

r<i (N

rt ca m CO <N <S d o

S3 ea

o <s

d d

eo eS OO ro i n 00 fS cs d d

U C<3 <n c> oo in en m d d

ca cs o • * CO <s

-H- . C

r-i i n r~ i n I—1 1 - "

oo O

S P

• ^

Tl-i n

X3

m i n i n

CQ

m i n

CQ CO <N rn

X) O cs d

ca <s m d

.E r-o CN d

X)

m Tf

d

X3 OS P

OO so

m f n U C/5

r- r~-a\ f o •<*• «n

u u m o oo 0\

X ) O

S r •<* CO

T^ CO oo —; CN CO

Xi c3 S3

<s -^ «N CN

d d

cS X) Os O T t CO

d d

X3

CO '-^ • * a\ (N '—1 d d

0 0 CO

o\ -^ CO t s

d d

.E o CN ON O oo ^ d

0\ -rt CO OS

r4 d

O N " ^ O N t ^ CO f S

^ S

on Z

•x-

* *

•)(•

*

*

*

* *

* *

* *

* *

a

u c ea o

"E .2? c o c

C en

a. K

C/3

Z

"5 o

Vj

iX c«"

"o >

JL> ^

X I ca

X i

S o d T3 C ea

i n

o d

4-»

c ea o

C 'c 2p-s •c« >

>. ex, o m c o

* *

> (U —* >-. .ts

IS ea X ) 9 CI.

m o d

-S

ca

'•B

>^ c ea O

s "S .2? 'S w

o c= ?J Sa

E 3

o ea

c

C OJ ea ca " ^ VH

5 « E

- "Si 2 o o e —1 • -4^ C M

"3, S2''P

& «5 o bo i < ca

4— -H-

12

< i2

uj £

u (>0_ 3 02

^ < ^

.

4..I c ca

• Q -

bfl .:<:

0.7

0.6-1

0.5

0.4-

0.3-

0.2-

0.1-

0.0

TX378 TX430 SC35 TX399 TX2741

I I

25 35

DAYS AFTER PLANTING

Fig. 1.1. Leaf area development (a) and cumulative water use (b) of five sorghum genotypes in the greenhouse smdy. Vertical bars indicate LSD (0.05).

13

< -3

on on <

5 E

8 on

on,-«.

Qi E W bo

DAYS AFTER FIRST MEASUREMENT

Fig. 1.2. Leaf area index (a), cumulative shoot biomass production (b), and cumulative water use (c) of five sorghimi genotypes during the period from 35 DAP to flowering in die field smdy. Vertical bars indicate LSD (0.05).

14

Z o

O -ai ' _ 0-on oo <

c ca

H. CO

O

5

P — <: 60

0.15 0.20 0.25

LEAF AREA (nJ plant * )

0.30

Fig. 1.3. Relationship of average daily shoot biormss production (a) and average daily water use (b) dunng die period from 35 DAP to flowering with average leaf area over die same period in die field study. E^h point represents individual plot (diree repUcations x tive sorghum genotypes).

15

>-U z u u, w UJ on

W

<

CO

eo

6.0

5.5

5 .0 -

4 .5 -

4 . 0 -

D O

TX378 TX430 SC35 TX399 TX2741

3.5 0.15 0.i7 0.i9 0.21 0.23 0.25

LEAF AREA (ra plant "^

Fig. 1.4. Water use efficiencies of five sorghum genotypes during the period from 35 DAP to flowering versus their average leaf areas over the same period in the field smdy.

16

Z o

Q -S O -oi a, on on <

O S

c 2 to

I—'rt

0.3 0.4

LEAF AREA (ni plant "^

Fig. 1.5. Relationship of average daily shoot biomass production (a) and average daily water use (b) during the period fk)m emergence to flowering widi average leaf area over the same period in the greenhouse smdy. Each point represents individual plant (four replications x five sorghum genotypes).

17

E - - 3 .3 -

z U

M W on

ai

<

3.7

3.5-

to eo 3 .1 -

2.9-

2.7

• - • . •

n o

-

•

•

TX378 TX430 SC35 TX399 TX2741

o •

n 1 — 1 — 1

A

1 1 » —

•

4.4 5.0 5.6 6.2 6.8

BP/LA (gm"2pIant"M

Fig. 1.6. Water use efficiencies of five sorghum genotypes during the period from emergence to flowering versus dieir shoot biomass production per unit average leaf area (BP/LA) calculated during die same period in the greenhouse smdy.

CHAPTER n

GENOTYPIC VARIATION FOR GAS EXCHANGE

TRAITS AND THEIR RELATIONSHIP

WITH WATER USE EFnCIENCY

Introduction

Crop improvement programs do not emphasize increase in shoot biomass

production per unit water used (WUE) even though this could be an important trait in

w atcr-limited environments. The lack of simple, rapid, and reliable screening criteria

and measurement techniques for WUE of crop plants has gready restricted progress in

this critical area of crop improvement (Hall et al., 1990).

Gas exchange should have direct impact on plant productivity and WUE since

carbon assimilation (A) and transpiration (T) rates are two major physiological

processes in plant growth and water use. If differences in gas exchange traits reflect

genotypic variation in seasonal WUE, they can be used as evaluation criteria to identify

germplasm with higher WUE.

There is limited information about the relationship between gas exchange traits

and WUE. It has been reported that carbon isotope discrimination (A) was negatively

related to WUE in wheat ( Farquhar and Richards, 1984), peanut (Hubick et al., 1986),

tomato (Martin et al., 1989), and barley (Hubick and Farquhar, 1989). A strong and

negative correlation between A and die ratio of A to stomatal conductance (g) was

observed in coffee (Meinzer et al., 1990). Theoretically, A is related to die ratio of A to

8 (A/g) through pj/p^, the ratio of intercellular and ambient CO2 partial pressures

18

19

I Farquhar et al., 1982). Therefore, a close relationship between A and WUE suggests

the existence of a relationship between gas exchange traits and WUE.

A close relationship between single leaf and canopy photosynthetic rates and

productivity has been reported in several species (Zelitch, 1982, Wells et al., 1982 and

19S6). Krieg et al. (1989) investigated 22 sorghum Unes for A and biomass production

under field conditions. The results indicated a very strong cortelation between A and

shoot biomass production (r2=0.77). Heitholt (1989) reported that the ratio of A to T

(.X/T) was cortelated to WUE across N and water stress treatments with winter wheat

! r-=0.76). Frank et al. (1987) found no relationship between A/T and WUE in wheat

crass, which was probably because A/T was determined in a greenhouse study, while

WX'E was monitored under field conditions.

The existence of genetic variation in seasonal WUE for grain sorghum has been

discussed in the previous chapter. Kidambi et al. (1990) have evaluated 30 sorghum

h\ brids under field conditions and reported significant genotypic variation for A and g

both across the season and in any given sampling time-irrigation level combination.

The objectives of this study were: (i) to investigate the genotypic variation in gas

exchange traits among the five sorghum genotypes which have exhibited significant

genotypic differences in shoot biomass production and WUE, and (ii) to determine

^ hether gas exchange measurements accurately reflect genotypic variation in seasonal

\\X'E.

Materials and Methods

The research sites, cultural practices, treatments, experimental design, and

"^asurements of shoot biomass production, leaf area development, and water use have

^ n described in die first chapter. Gas exchange measurements were made between

20

:: ') and 1300 h (CDT) and when photosyntiietic photon flux density (PPFD exceeded

!tH" ^mol m'2 s"^ in the greenhouse and 1800 |J.mol m-2 s"! in the field.

Measurements were made five times during the period from 35 days after planting

D.-\P) until flowering (75 DA' i. An uppermost fully expanded leaf was selected for

:hc measurements of gas exchange rates. Three leaves of each plot were selected as

subsamples for gas exchange measurements in the field smdy. The net exchange of

C0-> and H2O between the leaf and atmosphere was measured with a portable

photosynthesis system (LI-6200, LI-COR Inc., Lincoln, NE) by enclosing the leaf in a

.-h;imber and monitoring the rates at which the CO2 and H2O concentrations in the air

. iunged over fairly short time intervals (20 s). Stomatal conductance to H2O vapor was

calculated based on measured T, vapor pressure difference between the leaf and air

I\PD), and a predetermined boundary layer conductance (Anonymous, 1987). The

boundary layer conductance estimated with wet filter paper was 0.8 mol m"2 s'^. The

infrared gas analyzer of LI-6200 was calibrated periodically against standard CO2 gas

\^iih the concentration of 347 |iPa Pa'^ (Big Three Industries Inc., Houston, TX). A

four-liter chamber was used to minimize CO2 depletion effects on photosynthesis due

•o the high CO2 uptake rate of sorghum leaves. The maximum flow rate (800 |imol s"l

for this particular instrument) was used to circulate dry air into the leaf chamber in order

to slow down the humidity increase during the course of each measuremenL The 20 s

"^asurement duration resulted in an air temperature increase of less than 2 *C inside die

"Chamber.

Diurnal measurements of gas exchange was made on TX 378 and TX 399 in the

Pecnhouse smdy from 900 to 1700 h (CDT) on 50, 58, and 65 DAP to determine

21

^hcdier differences for A among die genotypes extended across die entire photoperiod.

Comparisons of gas exchange measurements using a closed (LI-6200) and open

i 1,1-1600) system were conducted in die greenhouse study on 60 DAP.

The data were transferted from the LI-6200 to a microcomputer via a

communication program (Procomm®, PIL software systems, Columbia, MO). The

significance of genotypic variation in gas exchange traits was determined by analysis of

\ ariance, and Duncan's multiple range-test was used to separate the genotypic means

I.SAS Institute, Inc., 1985). The consistency of genotypic differences across the

greenhouse and field environments was determined by Spearman's rank cortelation

analysis (Lindeman et al., 1980).

Results and Discussion

The mean gas exchange rates and component ratios of the five sorghum

genotypes listed in Table 2.1 were values averaged across all five measurements during

the period from panicle initiation to flowering since most of genotype x environment (G

^ E) interactions for gas exchange traits were nonsignificant or small compared with

genotypic effects. The sorghum genotypes exhibited significant (P<0.01) differences in

A under both die greenhouse and field conditions. Even tiiough G x E interaction for A

•as statistically significant (P < 0.01) in die field study, it was small compared vridi die

genotypic varibility of A, according to die F-values (Table 2.1). Genotypic ranking in

A of the five genotypes was similar across die two different environments widi a rank

'••orrclation coefficient of 0.80. TX 378, TX 430, and SC 35 demonstrated higher A

*anTX 399 and TX 2741.

In die previous chapter we reported diat shoot biomass production was strongly

*"d positively conflated widi leaf area (r^ = 0.71 and 0.77 in die field and greenhouse

22

study, respectively). The 23-29% of variation in shoot biomass production which

remained unexplained by leaf area was mainly due to die differences in A, since die

differences in shoot biomass production among die genotypes were cortelated widi

cenotypic differences in A (Table 2.2), especially in die field smdy ( r^ = 0.79).

Identifying the major contributor to shoot biomass production between leaf area and A

based on the magnitude of r^ vvas difficult because higher A may cause faster leaf

expansion. A significant positive correlation between genotypic differences of leaf area

and A was observed in the field smdy with a r^ of 0.83 (Table 2.2). Therefore, die

jcnotypic differences in shoot biomass production of grain sorghum was caused by

genotypic variability in bodi leaf area and photosynthetic rate.

Diumal gas exchange measurements were made in the greenhouse study on 50,

.*>8, and 65 DAP to determine whether differences for A among the genotypes extended

across the entire photoperiod. Results indicate that differences in the diumal pattern

between TX 378 and TX 399 were consistent across all three measurement periods

(Fig. 2.1). TX 378 always had higher A dian TX 399 during the entire daylight period,

\^hich also indicates that G x E interaction for A was small compared with the

genotypic effect The genotypic difference in A became statistically significant when die

photosynthetic capacity reached the maximum around mid-day and the differences

became even larger in die afternoon. The larger genotypic differences occurring in die

afternoon compared widi mid-day might be due to the different response of A among

ihe genotypes to temporary water deficits or due to different temperature responses.

Transpiration rates among die five genotypes did not show significant (P<0.05)

differences in the field smdy (Table 2.1). In die greenhouse smdy, die differences in T

among the genotypes were statistically significant. The ranking in T among the

genotypes was not consistent across die greenhouse and field studies, which suggests

23

:hai real genotypic variation in T might not exist. The results reported in previous

chapter also indicated that die five genotypes did not exhibit as consistent differences in

:i>ng-term water use as in shoot biomass production. There was no strong relationship

sciween T and whole plant water use rates among die five genotypes in bodi die field

and greenhouse studies (Table 2.2), which suggests that instantaneous T measured at

sincle leaf level did not control the long-term water use.

The comparisons of gas exchange measurements using a closed (LI-6200) and

.-.pen (LI-1600) system were conducted in the greenhouse smdy. The data shown in

lable 2.3 indicate that no differences in leaf temperature existed among the five

genotypes. The cuvette relative humidities measured by the open system were similar

.itToss the five genotypes as expected. However, the cuvette relative humidities

monitored by the closed system were significantiy higher for TX 378, TX 430, and SC

35 than for TX 2741 and TX 399. The differences were attributed to the differences in

T among the genotypes, since the cortelation between the cuvette relative humidities

and T of the five genotypes wassignificant (r2=0.88). Differences in relative humidity

among the genotypes residted in a significant genotypic differences in VPD, while VPD

detected by the open system was not significantiy different among the genotypes.

Differences in VPD among the genotypes resulted from the chamber effects in the

closed system will influence the value of stomatal conductance (g) which is calculated

^«rd on T and VPD. Therefore, g, the ratio of A to g, and internal CO2 concentration

*C,) are not presented in this paper because diey are artifacts generated fi-om chamber

effects. The value of T may not be affected by the differences in VPD since die gas

exchange measurements were taken over short time intervals (20 s). The correlation for

24

1 o( the five genotypes measured by die open and closed system was highly significant

,r-=0.93).

Significant variation (P<0.05) in A/T was observed among die five genotypes in

hoth the greenhouse and field studies (Table 2.1). Genotypic ranking in A/T was

consistent across two locations with a rank cortelation coefficient of 0.90. The

correlations between A and A/T were significant in bodi studies (Table 2.2), indicating

that A was the major component responsible for genotypic differences in A/T. This also

supports our previous conclusion that differences in shoot biomass production among

;hc genotypes determined the genotypic variation for WUE.

The ratio of A to T was positively cortelated with WUE with a r^ of 0.77 in the

greenhouse study and 0.96 in the field study (Table 2.2). Figure 2.2 demonstrates that

the correlation between A/T and WUE of five genotypes across the two different

environments was highly significant (r2=0.97). This close relationship was largely due

:o a positive cortelation between A and shoot biomass production and the insignificant

differences in T and whole plant water use rates among most of the genotypes.

The absolute value of A/T was related to seasonal WUE in the field study. The

average A/T across five genotypes was about 5.35 mmol CO2 mol H2O-I or 3.6 g C

'• g H2O-1. Only about 60% of die carbon taken up by crop photosyndiesis contributes

to biomass (McCree, 1988) so diat only 2.16 g C in biomass results fi-om 3.6 g C fixed

""y photosynthesis. This amount of C converts to 5.27 g biomass since the

concentration of C in a sorghum plant averages about of 41% (Stahl and McCree,

^988). Thus, die WUE based on die A/T was about 5.27 g biomass kg H2O-I which

^ as very close to acmal WUE in die field study (average WUE across five genotypes

*as about 5.0 g biomass kg H2O-I).

25

A close relationship between A and WUE was also observed (Table 2.2). In die

previous chapter we concluded diat die genotypic differences in WUE was due to die

differences in leaf area and shoot biomass production per unit leaf area among the

genotypes. This may be explained by the possibility that the variation in A among die

ccnotvpes is the physiological cause for the genotypic differences in both leaf area and

shoot biomass production per unit leaf area. Therefore, genetic variation in A might be

::x- ultimate control for the genotypic differences in WUE.

In conclusion, the sorghum genotypes exhibited significant variation in A, A/T,

r-.omass production, and WUE. No consistent genotypic variation was observed for T

or whole plant water use rates. Single leaf measurements of gas exchange traits

reflected single plant and whole canopy WUE differences among die genotypes largely

due to the positive relationships between A and shoot biomass production.

26

References

Anonymous. 1987. LI-6200 technical reference. LI-COR Inc., Lincoln, NE.

{arquhar. G.D., M.H. O'Leary, and J.A. Beny. 1982. On die relationship between carbon isotope discrimination and intercellular carbon dioxide concentration in leaves. Aust. J. Plant Physiol. 9:121-137.

i arquhar, G. D., and R. A. Richards. 1984. Isotopic composition of plant carbon correlates with WUE of wheat genotypes. Aust. J. Plant Physiol. 11:539-552.

irank, A.B., R.E. Barker, and J.D. Berdahl. 1987. Water-use efficiency of grasses grown under controlled and field conditions. Agron. L 79:541-544.

iiall. A. E., R. G. Mutters, K. T. Hubick, and G.D. Farquhar. 1990. Genotypic differences in carbon isotope discrimination by cowpea under wet and dry field conditions. Crop Sci. 30:300-305.

Heitholt. J.J. 1989. Water use efficiency and dry matter distribution in nitrogen-and water-stressed winter wheat. Agron. J. 81:464-469.

Hubick, K.T., and G.D. Farquhar. 1989. Carbon isotope discrimination and the ratio of carbon gained to water lost in barley cultivars. Plant, Cell and Environment 12:795-804.

Hubick, K.T., G.D. Farquhar, and R. Shorter. 1986. Cortelation between water-use efficiency and carbon isotope discrimination in diverse peanut (Arachis ) germplasm. Aust. J. Plant Physiol. 13:803-816.

Kidambi, S.P., D.R. Krieg, and D.T. Rosenow. 1990. Genetic variation for gas exchange rates in grain sorghum. Plant Physiol, (in press).

Krieg. D. R., S.P. Kidambi, F.S. Girma, and D.T. Rosenow. 1989. Water use efficiency of grain sorghum: genetic and environmental components, p. 169. In Proc. 16th Biennial Grain Sorghum Res. and Utilization Conf., Lubbock, TX. 19-22 Feb. 1989. Grain Sorghum Producers Association, Abemadiy, TX.

lindeman, R.H., P.F. Merenda, and R.Z. Gold. 1980. Introduction to bivariate and multivariate analysis. Scott, Foresman and Company, Glenview, IL.

^'anin. B., J. Nienhuis, G. King, and A. Schaefer. 1989. Restriction fragment lengdi polvmorphisms associated with water use efficiency in tomato. Science 243:1725-1728.

^'-•Cree, K.J. 1988. Sensitivity of sorghum grain yield to ontogenetic changes in respiration coefficients. Crop Sci. 28:114-120.

27

Meinzer, F.C., G. Goldstem, and D.A. Grantz. 1990. Carbon isotope discrimination in coffee genotypes grown under limited water supply. Plant Physiol 92:130-135.

S\S Institute, Inc. 1985. SAS user's guide: Statistics, 5di ed. SAS Instimte Inc • Cary, NC. ' "'

S:ahl. R. S., and K. J. McCree. 1988. Ontogenetic changes in the respiration coefficients of grain sorghum. Crop Sci. 28:111-113.

Wells, R., W.R. Meredidi, Jr., and J.R. Williford. 1986. Canopy photosyndiesis and its relationship to plant productivity in near-isogenic cotton lines differing in leaf morphology. Plant Physiol. 82:635-640.

Wells, R., L. L. Schulze, D.A. Ashley, H.R. Boerma, and R.H. Brown. 1982. Cultivar differences in canopy apparent photosynthesis and their relationship to seed yield in soybeans. Crop Sci. 22:886-890

Zelitch, I. 1982. The close relationship between net photosynthesis and crop yield. BioScience 32:796-802.

28

I able 2.1. Mean carbon assimilation rate (A), transpiration rate (T), and ratio of A to T o! five genotypes. F-values based on analyses of variance for the effect of genotvpe '0) and genotype x environment (G x E) interaction on A, T, and A/T ratio.

ricnot>'pe

1^378

TX 430

.SC 35

TX399

TX2741

G

GxE

A

GH

jimolm

40.04 a"

38.82 ab

37.27 be

33.96 d

35.79 c

14.2**

NS

FD

2s-l

46.04 b

47.77 a

45.99 b

43.54 c

40.79 d

22.5**

7.3**

T

GH FD

mmolm2s'^

8.33 a

8.25 a

7.88 b

7.46 c

7.71 be

9 g**

NS

8.32

8.77

8.33

8.35

8.07

NS

NS

A/T

GH FD

mmol mol"

4.81a

4.71 ab

4.73 ab

4.55 c

4.64 be

5.5*

NS

5.53 a

5.45 a

5.52 a

5.21b

5.05 c

11.8**

8.5**

*.** denotes significance at the 0.05 and 0.01 probability levels, respectively. NS represents nonsignificance at the 0.05 probability level.

* Means followed by the same letter (widiin a column) are not significantiy different at die 0.05 probability level according to Duncan's multiple range-test.

29

lible2.2. Cortelation matrix for gas exchange and agronomic traits of five sorghum •enotypes in die greenhouse (GH) and field (FD) smdies.

A/T LA BP WU WUE

0.95* 0.91*

0.89* 0.54

0.40 0.91*

0.61 0.89*

0.39 0.11

(GH) 0.95* 0.40 0.61 0.71

(FD) 0.91* 0.91* 0.89* 0.94*

r (GH)

(FD)

VT (GH) 0.88*

(FD) 0.98**

'. ** denotes significance at the 0.05 and 0.01 probability levels, respectively.

t A=carbon assimilation rate, T=transpiration rate, A/T=ratio of A to T, LA=Ieaf area, BP=shoot biomass production, WU=water use, and WUE=water use efficiency.

30

r 5t

•J «

:^ X^

•' -^

V

'. • *— "5 u k .

•:

c y)

^

O c \> ao u > C

v» o

bJL

^ E c

^" C S3

^

Q OH

>

•o o U

c O H

o T3 1/3

U

t3 V3

o 0

c

o

•a

O

u

o c O

(N CS m r~ m u-i oo m

U en m CO f o en

^

cs en

un so u-5

Ou

• * r~ ' - r~ s c sq en 00 c s c s en c s en en en en

eo en cs >n lo sq so ^ iri >n in "n

.O JO c s oo

o o d d

en CO CO

un oo r- ^ 2 ' 0\ 0\ 0\ O 0\ en en en •^' en

r - m OS t^ ^ c s c s ^ . "^ •^. erj en en en en

ea

g en en

C3

CS

cs en

ea

en

J3 ea o

ON C S >n en

ca .o

o so

Cjs \ 0 op O oo i5\

oo 00 r-

00 C5 S^ S^ r - en m 2 i J^ ^ r t en en c s X X u >< >< H H <zi H H

C/2 Z

on Z

on Z

*

GO

z

*

* *

* *

c u

o. ea

2 O H

o CO

Z

>

ea

O

O d

ea

o u O H C/3

u I i

vT

leve

(U VH

^ _ > i

fica

i

c 6fl

o c 9

c ^ ea 0)

>

• £ |

ta ^

(L> O

o J= r?i! *-» c ^

.1-1 f \ " c ea

C/3 U

o-a

* ?

o ^ ea ^

a i» • r ea ^ -

^ OJ

U O4

•*- .» . . —

i> E S _c« ea "r--S c ^ 3

•^ o •s O "O

o e« ea C _ H ea u <o >

31

cs 'S "o E

800 1000 1200 1400 1600 1800

HOUR

Fig. 2.1. Diumal changes in carbon assimilation rates (A) of TX 378 and TX 399 on 51,58, and 65 days after planting (DAP) in die greenhouse study. Vertical bars indicate LSD (0.05).

32

>-U z y a on

E-<

to

6.0

5.1-

4.2-

3.3-

2.4

• TX378 • TX430 A SC35 n TX399 O TX2741

r 2= 0.97

4.3 —r-4.8

— I — 5.3 5.8

A/T (mmol moF )

Fig. 2.2. Water use efficiencies of five sorghum genotypes versus their ratios of carbon assimilation rate to transpiration rate (A/T). Larger and smaller symbols represent genotypic means in the field and greenhouse smdy, respectively.

CHAPTER m

CANOPY GAS EXCHANGE AND ITS

RELATIONSHIP WITH SINGLE

LEAF RESPONSES

Introduction

Canopy photosynthesis (CAP) has been studied mainly in field-grown wheat,

. otion, and soybean during the past 25 years (Jeffers and Shibles, 1969; Pegelow et

J1.. 1977; Johnson et al., 1981; Larson et al., 1981; Morgan and Willis, 1983;

Pucch-Suanzes et al., 1989). These smdies indicate that CAP is a complex process,

affected by various environmental and plant factors such as irradiance, temperature,

plant water and nutritional status, leaf area index (LAI), genotype, developmental stage,

.md time of day.

Morgan and Willis (1983) measiuied diumal CAP and evapotranspiration (ET) on

spnng wheat and reported that maximum CAP rates always occvured in the morning

and earlier than maximum rates of ET. Ludwig et al. (1965) reported that CAP and ET

•n artificial communities of cotton plants increased with LAI until a LAI of about 3 and

-. respectively.

In soybean, genotypic variation in CAP has been well defined (Harrison et al.,

1981; Larson et al., 1981; Wells et al., 1982). Wells et al. (1982) showed diat

*-'Ortelation coefficients for the association between grain yield and CAP measured

during seed development were significant and ranged from 0.45 to 0.65. Harrison et al.

' J981) observed diat heritabilities based on variance component estimates were 41-65%

'«• CAP. The significant cortelation and higher heritability indicate die feasibility

33

34

.ssclecring for CAP to increase seed yield in soybean. Wells et al. (1986) smdied CAP

and its relationship to plant productivity in near-isogenic cotton lines differing in leaf

rv.orphology and found that lint production was positively related to seasonally

;r,tegrated CAP (r2=0.53).

Zelitch (1982) has stated that plant productivity is more closely related to

measurements of CAP than to measurements of single leaf photosynthetic rates. Leaf

.ipparent photosynthesis measurements usually estimate the maximum potential of a

cirnotype because the uppermost fully expanded leaves (optimum physiological

conditions and plant position) are examined (Elmore, 1980). Canopy photosynthesis

measurements, on the other hand, measure the carbon uptake of the whole stand. This

measurement more accurately describes the photosynthetic activity per unit ground area

and combines genotype efficiency, leaf morphology, and canopy architecture (Wells et

.il.. 1986).

Relatively littie work has been reported on the sorghum canopy gas exchange

traits. The purpose of this smdy was to determine the diumal and seasonal responses of

canopy gas exchange rates as compared to single leaf gas exchange rates of grain

sorghum.

Materials and Methods

The research sites, cultural practices, treatments, experimental design, and

measurements of single leaf gas exchange have been described in previous chapters.

Diumal measurements of canopy simultaneously widi single leaf gas exchange were

"wde from 800 to 1800 h (CDT) on 35 DAP in each plot of two sorghum genotypes

'"TX 378 and SC 35) in die field study. Since dien, only midder-day measurements of

35

• P.opy simultaneously with single leaf gas exchange were made in each plot of the

.•me cenotypes until flowering.

The size of chambers for canopy gas exchange measurements was 1 m^ (1.50 m

^ i).(i7 m X 1.00 m). The four vertical sides of the chamber were covered with

r.i.Tsparent polyvinyl chloride (PVC) film and the top was covered with transparent

I i.non® (polytetrafluoroediylene) film (DuPont Chemical Co., Wilmington, DE) which

•r.insmitted infrared radiation, thereby avoiding heat build-up inside the chamber. Two

cvtric fans (Dayton 4C004, W. W. Grainger, Inc., Lubbock, TX) with 4 m^ s"!

. .sp.icity were bounded to two opposite sides of each chamber for air circulation.

A portable photosynthesis system (LI-6200) was connected to the chamber as a

- ii scd system to monitor the changes in CO2 concentration, relative humidity (RH),

.md air temperature (AT) inside the chamber over short time intervals (50 s). Absolute

•umidity (AH) was calculated based on relative humidity (RH) and air temperature

I AT) inside the chamber according to the equation: AH = 0.01RH{exp[52.57633 -

''^90.4985/(AT+273.15) - 5.028281n(AT-(-273.15)]}. Canopy photosynthetic and

•r.inspiration rates (CT) were determined from the slopes of the CO2 depletion and the

•ijisolute humidity increase during the measurement period, respectively. Samration

^ jpor pressure dificit (VPD) of die air were calculated based on the difference of vapor

pre sure from saturation vapor pressure inside the chamber. The 50 s measurement

^-ration resulted in an air temperature increase of less than 2 °C inside the chamber.

^•c a-lative humidity increased by about 11%, which resulted in a decrease in VPD of

^ l Pa inside the chamber during the measurement period. However, die increase in RH

ind decrease in VPD did not influence die canopy gas exchange rates significantiy

36

sccause the slopes of the CO2 depletion and the absolute humidity increase did not

. hange significantiy over the 50 s period.

A wooden floor consisting of two pieces of plywood was used to cover the

-nind when canopy gas exchange measurements were made in order to minimize the

.:mount of CO2 and H2O vapor evolved from the soil surface. Foam rubber was used

:,. seal the floor around the stem of die plants. The chamber was then placed on the

A ixxlen floor with 24 plants enclosed. Rates of CO2 and H2O vapor evolved from the

>.>il surface were reduced from 4.0 to 0.8 |imol m^ s'l and from 2.3 to 0.7 mmol m^

s '. respectively, after using the floor. Therefore, CAP and CT were not cortected by

Mill respiration and evaporation components.

Saturation vapor pressure dificit (VPD) of the ambient air were calculated based

on the difference of vapor presure from saturation vapor pressure. The data of ambient

j:r temperature and relative humidity were collected at the site with an automated

\^cadicr station (CR-10, Campbell Scientific Inc., Logan, UT).

Results and Discussion

Figure 3.1 shows the diumal pattems of photosynthetic photon flux density

PPFD), air temperature, and vapor pressure differences (VPD) of the ambient air when

^umal gas exchange measurements were made. The diumal pattems of A and CAP per

-nit ground area (CAP/G) were similar and both followed changes in PPFD fairly

-ioscly (Fig. 3.2). Single leaf photosynthetic rate reached its maximum at 1200 h and

'•hen declined gradually to 22 ^imol m'^ s'^ at 1800 h. The canopy apparent

•^tosyndietic rate approached die maximum at die same time as A, but stayed at its

37

mxximum rate until 1600 h. This suggests diat the maximum value of canopy light

ir.terception occurs 2 h before solar noon (at 1400 h) and lasts for 2 h after solar noon.

The diumal pattems of T and CT were also similar and both followed changes in

evaporative demand (i.e., VPD) fairly closely in the moming (Fig. 3.3). The maximum

rate of T occurred from 1200 h until 1600 h, while CT reached its maximum at solar

noon. Even though the air temperature and VPD were very high during the period from

: (>(X) h to 1800 h, both T and CT declined rapidly over the same period. This rapid

decline was probably due to increased stomatal resistance resulting from lower PPFD

ir.d temporary plant water stress. At the single leaf level, A and T peaked at the same

;:me diumally, and maximum photosynthesis rates occurred earlier than maximum CT

at die canopy level.

Seasonally, CAP/G increased with LAI until LAI reached its maximum (Fig.

v4). After this stage, LAI remained fairly constant but canopy photosynthesis per unit

leaf area (CAP/L) declined rapidly largely due to leaf aging and less sink demand.

However, CT did not followed the seasonal changes in LAI, which acmally decreased

by 15% from 35 to 50 DAP when LAI increased by 2 over die same period (Fig. 3.5a).

The decrease in CT was largely attributed to die differences in VPD between these two

days when gas exchange measurements were taken (Fig. 3.5b). Therefore, the control

of transpiration at tiie community level is the evaporative demand and not the plant

itself, which supports die relationships in single leaf transpiration and long-term water

use reported in previous chapters.

Figure 3.6a demonstrates that TX 378 and SC 35 had similar A. The two

genotypes did not exhibit significant differences in CAP per unit leaf area (CAP/L),

neither (Fig. 3.6b). The season pattems of A and CAP were not as similar as dieir

38

*;umal patterns. This is because LAI and plant age affect more direcdy on CAP dian on

In conclusion, diumal responses of gas exchange rates measured at the single leaf

.c\el were fairly similar to these at the canopy level. Seasonal changes in CAP were

..irccly a function of leaf area development and plant age, while CT was mainly

controlled by evaporative demand. TX 378 and SC 35 did not exhibited significant

d::"t'crences in CAP/L or A.

39

References

l-lmorc, CD. 1980. The paradox of no cortelation between leaf photosyntiietic rates and crop yields, p. 155-167. In J.D. Heskedi and J.W. Jones (eds.) Predicting photosynthesis for ecosystem models. Vol. II. CRC Press, Boca Raton, FL.

iLirrison, S. A., H. R. Boerma, and D.A. Ashley. 1981. Heritability of canopy apparent photosynthesis and its relationship to seed yield in soybeans. Crop Sci 21:222-226.

Jeffers. D.L., and R.M. Shibles. 1969. Some effects of leaf area, solar radiation, air temperature, and variety on net photosynthesis in field grown soybeans. Crop Sci. 9:762-764.

Johnson, R. C , R. E. Witters, and A. J. Ciha. 1981. Daily pattems of apparent photosvnthesis and evapotranspiration in a developing winter wheat crop. Agron. J. 73:414-418.

l^son, E. M., J. D. Hesketh, J. T. WooUey, and D. B. Peters. 1981. Seasonal variations in apparent photosynthesis among plant stands of different soybean cultivars. Photosynthesis Res. 2:3-20.

Ludwig, L. J., T. Saeki, and L. T. Evans. 1965. Photosynthesis in artificial communities of cotton plants in relation to leaf area. I. Experiments with progressive defoliation of mature plants. Aust. J. Biol. Sci. 18:1103-1118.

Morgan, J.A., and W.O. Willis. 1983. Gas exchange and water relations of'Olaf spring wheat. Crop Sci. 23:541-546.

Pegelow, E.J., Jr., D.R. Buxton, R.E. Briggs, H. Miu-amoto, and W.G. Gensler. 1977. Canopy photosyndiesis and transpiration of cotton as affected by leaf type. Crop Sci. 17:1-4.

f^Jech-Suanzes, I., T.C. Hsiao, E. Fereres, and D.W. Henderson. 1989. Water-stress effects on the carbon exchange rates of diree upland cotton (Gossypium hirsutum) cultivars in the field. Field Crops Res. 21:239-255.

^Vells. R., W.R. Meredidi, Jr., and J.R. Williford. 1986. Canopy photosyndiesis and its relationship to plant productivity in near-isogenic cotton lines differing in leaf morphology. Plant Physiol. 82:635-640.

^^'clls, R., L. L. Schulze, D.A. Ashley, H.R. Boerma, and R.H. Bro>yn. 1982. Cultivar differences in canopy apparent photosyndiesis and their relationship to seed yield in soybeans. Crop Sci. 22:886-890

flitch, I. 1982. The close relationship between net photosyndiesis and crop yield. BioScience 32:796-802.

40

y—N

• a j

CS

•E o p

3, Q OH OH

2500

2000

1500

1000

500

0

u

<

ca OH

Q OH >

30-J

25

20

15

600 800 "lObo 1200 1400 1600 1800 2000

HOUR

Fig. 3.1. Diumal changes in (a) photosyntiietic photon flux density (PPFD), (b) air temperature (AT), and (c) saturation vapor pressure dificit (VPD) of ambient air when diumal gas exchange was measured 35 days after planting.

41

o

p a, <

T — ' — I — ' — r — ' — r

600 800 1000 1200 1400 1600 1800 2000

e

"o E

HOUR

Fig. 3.2. Diumal pattems of single leaf (A) and canopy photosynthetic rates per unit ground area (CAP/G) 35 days after planting. Data were pooled across two sorghum genotypes. Vertical bars indicate standard ertors of the means.

42

-C

"o E E

600 800 1000 1200 1400 1600 1800 2000

HOUR

Fig. 3.3. Diumal pattems of single leaf (T) and canopy transpiration (CT) 35 days after planting. Data were pooled across two sorghum genotypes. Vertical bars indicate standard ertors of the means.

43

o (—• c

O OH

<

<

DAP

Fig. 3.4. Seasonal changes in (a) canopy photosynthetic rates per unit ground area (CAP/G) and (b) leaf area index (LAI). Data were pooled across two sorghum genotypes. Vertical bars indicate standard errors of means.

44

E

ea OH

Q cu >

DAP

Fig. 3.5. Seasonal changes in (a) canopy transpiration rates (CT) and (b) satmation vapor pressure dificit (VPD) of ambient air. Data for CT were pooled across two sorghum genotypes. Vertical bars indicate standard errors of means.

45

I

r j

O

E

<

DAP

Fig. 3.6. Comparison of TX 378 and SC 35 in (a) single leaf (A) and (b) canopy photosynthetic rates per unit leaf area (CAP/L). Vertical bars indicate LSD (0.05).

CHAPTER IV

GENERAL SUMMARY AND CONCLUSIONS

Results from diis smdy demonstrated die existence of genotypic variation in WUE

defined as the amount of biomass produced per unit water use. The genotypic variation

in WUE was apparentiy due to the significant genotypic differences in shoot biomass

production without significant differences in water use for most of the genotypes. The

genotypic differences in shoot biomass production were largely due to the variation in

leaf area among the genotypes. At die single plant level (in the greenhouse sttidy), leaf

area was also cortelated with water use. Therefore, the cause of genotypic variation in

WUE was not the genotypic differences in leaf area but die genotypic differences in

shoot biomass production per unit leaf area. At die community level (in the field study),

the daily rates of water use was dictated by die evaporative demand but not plant itself.

Therefore, genotypic differences in leaf area were not correlated with water use. The

cause of the genotypic differences in WUE was the differences in leaf area among the

genotypes.

The sorghum genotypes exhibited significant variation in A. The genotypic

differences in A were positively correlated witii both shoot biomass production and leaf

areas of die five genotypes, indicating that the ultimate cause of genotypic differences in

shoot biomass production was genotypic variation in A. No consistent genotypic

variation was observed for T and T was not cortelated widi long-term water use. Gas

exchange efficiencies defined as die ratio of A to T were significantiy different for die

genotypes, which was largely due to die significant genotypic differences m A. The

genotypic differences in instantaneous gas exchange efficiencies measured at die single

46

47

icaf level reflected single plant and whole canopy WUE differences among the

izenotypes largely due to the positive relationships between A and shoot biomass

production. Therefore, the ultimate cause of genotypic differences in WUE was the

differences in A among the genotypes.

The comparisons of gas exchange measurements at the single leaf and canopy

levels indicated that diumal responses of A and CAP, T and CT were in agreement,

respectively. This supports that single leaf measurements of gas exchange traits

TL'tlected single plant and whole canopy differences in shoot biomass production and

long-term WUE.

The results indicate that measurements of A and leaf area may be used to select for

increased WUE in grain sorghum.