Tải bản đầy đủ - 0 (trang)
IV. The Effects of Fertilizers on the Relationship of Evapotranspiration and Yield

IV. The Effects of Fertilizers on the Relationship of Evapotranspiration and Yield

Tải bản đầy đủ - 0trang

2.34



FRiKK G . VIETS, JR.



nutrient availability is reached at which yield increases become infinitesimal. Therefore, the points derived from fertilizer experimentation on a

plot of ET vs. 2’ draw closer together as 2’ increases if fertilizers were

applied in equal increments of a nutrient. Overfertilization may produce

yield decreases that are ignored in the models.

In these models it is presumed that water is available for evapotranspiration and that the water conductivity of the soil and the capacity

of the root system to absorb water are not seriously limiting the ability



FIG.1. Six possible models of the relation between evapotranspiration (E T ) and

yield of dry matter (Y) (top part of each diagram), and water-use efficiency (Y/ET)

reiative to yield (lower part). These models presume that water is nonlimiting for

yield or evapotranspiration and that fertilizer is applied in equal increments

resulting in declining increments of yield.



of the soil-plant system to meet the evaporative demand. Data presented

in this section are chosen with this criterion in mind, but there is always

some question as to the extent to which it is fulfilled as the soil dries

and the moisture tension increases. Increase in moisture tension affects

not only the capillary flow of water to the soil surface for evaporation,

but also the flow to the root surface (see Kelley (1954), Veihmeyer and

Hendrickson (1950), Hagan (1955), and Russell (1959) for recent

reviews on the plant availability of soil water at various tensions, and

contents at equal tensions). Tanner (1960a, b ) drew attention to difficulties of interpreting field experiments relating plant growth to soil



FERTlLIZERS AND THE EFFICIENT USE OF WATER



235



moisture tension because of the confounding effects of soil moisture on

net radiation and the distribution of net radiation between crop and soil.



A. ET AND Y INCREASE

LINEARLY;No ET WHENY Is ZERO

Model A in Fig. 1illustrates these relationships. This case arises only

in containers where evaporation is blocked by a vapor proof seal. Briggs

and Shantz (1913a, 1914), Montgomery and Kiesselbach (1912), and

Miller (1916) used this technique. ET becomes transpiration ( T ) only.

T (or ET) is usually a linear function of Y and the regression line goes

through the origin for both total water use and total yield. De Wit (19%)

discusses these experiments in detail. The data of Ballard (1933) working

with barley and Sudangrass in solution culture in which nitrate was

varied, data of Lawes in 1850 cited by Briggs and Shantz (1913b) for

wheat, barley, and clover, and data of Thom and Holtz (1917) for

wheat in culture solutions fit this concept. Note that in this model there

is no increase in water-use efficiency with increase in yield through

fertilization. Such a situation could exist only in commercial hydroponics

with extreme advection or in the field if plastic or rock mulches were

used and all water was added below the mulch so that no evaporation

could occur.

B. E T AND Y INCREASE

LINEARLY;

E T Is APPRECIABLE

WHENY Is ZERO

See model B in Fig. 1. Figure 2 illustrates these functional relationships of timothy grown in lysimeters 56.42 cm. in. diameter at Gunnison,

Colorado, with rates of nitrogen application of 0, 200, 400, and 800

pounds per acre. The data are the sums of two cuttings made on June

30 and September 1, 1960, respectively. The lysimeters were surrounded

with timothy for a few feet but were subject to both vertical and

horizontal advection from the surrounding dry mountains. The extrapolation of ET to zero yield gives an intercept of about 100 kg. or 40cm. of

water. The actual evaporation from a bare soil in a lysimeter, irrigated

like the grass grown in the others, when the tension reached 500 to 600 cm.

of water at the 8-cm. depth was 56.7 kg. or 22.6 cm. for the same period.

Linear ET vs. Y functions with an intercept give hyperbolic functions of

Y/ET vs. Y in which Y/ET approaches a limit as ( E , T)/ET approaches

1, where E , is evaporation from the soil surface and T is transpiration,

In other words evaporation becomes a smaller proportion of evapotranspiration. Such ET vs. Y functions give increasing water-use efficiency as yield increases simply because water evaporated when plant

growth is zero is wasted.

Another container experiment is that of Cassady ( 1957), who grew

PLAINSMAN milo in paraffin-coated, 10-gallon garbage cans filled with



+



236



FRANK G . V E T S , JR.



Dalhart fine sandy loam that was fertilized well with phosphorus and

potash. The cans, 14.16 inches in diameter and 16.5 inches deep, were

buried to a depth of 13.5 inches on a rectangular plan with 45-inch

centers. Bare, dry soil surrounded the cans. The experiment was conducted outdoors at Las Cruces, New Mexico. Seed was planted May 21,

and the total yield of grain, tops, and roots was determined after



I



0



/



/

04

0



I



i

I



I



1



200



400



600



Yield in grams



FIG.2. A. Relationship of evapotranspiration (ET) to yield ( Y ) of timothy grown

in lysimeters. B. Relationship of water-use efficiency to yield for the same lysimeters.

The yields in each figure from left to right were produced with 0, 200, 400, or 800

pounds of nitrogen per acre, respectively. (Unpublished data of H. K. Rouse, F.

Willhite, and A. R. Grable, Soil and Water Conservation Research Division, Agricultural Research Service, U.S.D.A., and the Colorado Agricultural Experiment Station

cooperating. )



harvest 123 days later. Forty-eight amounts of nitrogen were applied in

solution as NH4N03 at planting or at thinning to give a series from 0

to 940 pounds per acre. Plants were thinned to three per can after they

reached a height of 6 inches. Water was added when tensiometers placed

at either 6- or 14-inch depths showed a tension of 750cm. of water.

Record was kept of all irrigation water applied and rainfall received.

Drainage was prevented. Only a portion of the data is presented in



237



FERTILIZERS AND THE EFFICIENT USE OF WATER



Table I. ET is a linear function of the total top weight. Extrapolation

of this line to zero yield gives an evaporation of 28 inches. The actual

evaporation from a bare soil similarly irrigated was 12.54 inches. Wateruse efficiency is a hyberbolic function of yield. So the data conform in

all respects to model €3.

TABLE I

Total Top Yield, Evapotranspiration, and Water-Use Efficiency of PLAINSMAN

Milo

Grown in Containers at Las Cruces, New Mexico, as Affected by

Nitrogen Supplp

Nitrogen

applied

(lb./acre)

0

0



100

200

300

400

500

600

700

800

900

940

a



b

C



Y



ETb

(in./can)

12.540

30.68

36.70

42.10

51.53

60.28

65.68

69.71

74.96

78.49

78.49

78.49



( g./can 1



No crop

22.4

62.0

102.5

157.5

208.9

240.4

310.2

325.9

353.9

355.2

351.9



Y/ET

(g./in./can)



-



0.73

1.69

2.43

3.06

3.46

3.66

4.45

4.35

4.51

4.52

4.48



From Cassady (1957).

Includes 0.88 inch precipitation.

Private communication from C. F. Cassady, Jr.



Other container experiments fit this model: e.g., Scofield (1945)

at Riverside, California, grew alfalfa in garbage cans filled with

87kg. of soil and measured ET and Y for plants on unfertilized soil

and on soil irrigated with nutrient solutions (pertinent data are shown

in Table 11.) The shape of the ET vs. Y function cannot be deduced

from only two points, but fertilization did increase both ET and Y, and

the extrapolated intercept of E T at Y is zero is positive. Fertilization increased water-use efficiency slightly. Allison et al. (19%) showed that the

TABLE I1

Yield and Evapotranspiration of Alfalfa Grown in Containers at Two Nutrient Levelsa

Number of

cuttings

6



7

a



ET

Growth period



Treatment



(kg.)



Jan. 6 to

Oct. 16, 1943

Jan. 8 to

Nov. 22, 1944



None

Fertilized

None

Fertilized



605

682

608

756



From Scofield (1945).



Y

(g.)

754

861

860

1143



Y/ET



1.25

1.26

1.42

1.52



238



FRANK G . V E T S , JR.



evapotranspiration of a broad array of crops variously fertilized and

grown in 63-inch-diameter lysimeters filled with Lakeland sand was

a linear function of crop yield. Extrapolation of E T to zero yield gave

ET of 18.7 inches. They showed that the dry matter produced per inch

of evapotranspiration was a linear function of yield but admit that a

“slightly-curved” line would fit the data better. Since fertilizers increased

yields of some of the crops markedly, use of fertilizers on this infertile

sand increased efficiency with which the precipitation and irrigation

water, applied in drought periods, was used. These lysimeters were

unguarded and surrounded with sand. Hence, advective flow of energy

was undoubtedly high.

Some experiments with adequate water for evapotranspiration in

the field show that fertilizers increased E T under the experimental

conditions. Stanberry ( 1959) presents graphically the effects of nitrogen

fertilization (none to 240 pounds per acre) on the yield and water use by

barley grown on Superstition fine sand at Yuma, Arizona. The plots were

small, and advected energy in this desert environment could be large.

Fertilization increased yield from 15 to SO bushels per acre in a typical

Mitscherlich-type curve, increased water use from 1s to a maximum of

24 inches (with 160 pounds of nitrogen an acre) and increased Y/ET

from 0.83 to 3.7 bushels per acre-inch of water with a Y/ET vs. Y

function that can be fitted with either a curve the shape of that shown

in Fig. 1, model B, or a straight line with positive slope. Table I11

gives a portion of the data of Jensen and Sletten, who have published

some of their results on the relationships of irrigation regime, yield, and

evapotranspiration of grain sorghum (Jensen and Sletten, 1957; Jensen

and Musick, 1960). Attention here is called to the data for the wet

TABLE 111

Effect of Sitrogen Fertilizer on Yields and Evapotranspiration of Hybrid Grain

Sorghum (RS-610) at Bushland, Texas, in 1958a

Moisture treatment

Nitrogen

applied

1.’

(Ib./acre) (Ib./acre)



0

120



240



2979

2626

2430



M.2



M,b

ET

(in.)

14.8

15.4

14.8



Y/ET

Y

(lb./acre-in.) (Ib./acre)

201

171

164



3442

6964

7232



ET

Y/ET

(in.) (Ib./acre-in.)

21.0

22.3

22.9



164

312

316



a Jensen, h4. E., and Sletten, W. H. Unpublished data, southwestern Great

Plains Field Station, USDA, Bushland, Texas, in cooperation with the Texas

Agricultural Experiment Station.

b Preplanting irrigation only. Rainfall from planting to harvest was 11.30 inches.

c Preplanting irrigation plus 4-inch irrigations on August 12 and 30.



239



FERTILIZERS AND THE EFFICIENT USE OF WATER



treatment M4, in which nitrogen fertilizer almost doubled yields and

increased ET sigdcantly by almost 2 inches. Water-use efficiency was

almost doubled. A plot of ET vs. Y appears to be linear with a low

positive slope, which makes the Y/ET vs. Y curve a very flat hyperbola.

Fertilizer subplots in this experiment were 15 by 50 feet. On August 12

plants were in the late boot stage on the unfertilized plots and were

headed and blooming on the plots getting 120 pounds of nitrogen. The

average plant heights were 3.73, 4.35, and 4.05 feet for the 0-, 120-, and

240-pound rates of nitrogen application, respectively. Mean integrated

moisture tensions to a depth of 4 feet were always higher on the plots

getting 240 pounds of nitrogen than on those getting 60 except soon

after irrigation. This difference in available moisture could have reduced

the differences in ET due to fertilization.

Winter wheat Y vs. ET relationships as affected by nitrogen application with adequate moisture ( M 4 ) for a dry year (1956) and a normal

year (1957) as given by Jensen and Sletten are shown in Table IV.

Some aspects of this work have been published (Jensen, 1956; Jensen and

Musick, 1960). Nitrogen fertilization significantly increased both yields

and evapotranspiration rates. Water-use efficiency was increased by

TABLE IV

Effect of Nitrogen Fertilization on Yields, Consumptive Use, and Water-Use

Efficiencv of Conch0 Winter Wheat at Bushland. Texa9

Moisture treatment

M4c



Year



1956 (dry year)



1957a (normal year)



N

Y

(lb.,/ (bu./

acre) acre)



Y

(bu./

acre 1



ET

Y/ET

(in.) (bu./acrein.



0

80

120



16.9

18.1

17.5



19.4

19.7

20.3



0.87

0.92

0.85



33.6

45.9

52.4



23.6

30.4

30.2



1.42

1.51

1.74



0



27.1

42.3

44.4

20.2

33.9

26.7



17.1

18.1

17.3

18.4

19.4

19.1



1.58

2.34

2.57

1.10

1.75

1.40



28.8

52.5

48.9

26.3

49.8

38.2



24.1

28.4

27.4

26.6

27.0

26.2



1.20

1.85

1.78

0.99

1.84

1.46



120

180

1958 (wet year)



Y/ET

ET

(in.) (bu./acrein. )



0



120

180



Jensen, M. E., and Sletten, W. H. Unpublished data, Southwestern Great

Plains Field Station, USDA, Bushland, Texas, in cooperation with the Texas

Agricultural Experiment Station.

b Preplanting irrigation only.

0 Preplanting irrigation plus 4-inch irrigations in April and May.

Yields were adjusted for hail damage by counting broken culms per unit length

of row on each treatment.

Q



*



2440



FRASK G. VIETS, JR.



fertilization. In the wet year (1958) fertilization increased yields but

had little effect on ET. In both 1957 and 1958, use of 180 pounds of

nitrogen, as compared to the 120-pound rate, decreased yields and wateruse efficiency because of lodging. The fertilized subplots in this experiment were 12 by 65 feet. In the normal and dry years when advection

would be expected to increase the ET of taller wheat in small plots,

yield and ET were closely associated, but this was not the case in the wet

year 1958. In all three years, nitrogen fertilization increased the efficiency

of water use.

C. ET Is INDEPENDENT

OF Y

This situation could exist where three conditions obtain: ( 1 ) there

is no advective heat; ( 2 ) soil surface is continuously moist; and ( 3 ) net

radiation of bare moist surface is the same as that of area with complete

vegetative cover. The ET-Y relationship and the Y/ET-Y relationship are

shown in model C of Fig. 1. This case gives a linear increase in water-use

eEciency with increasing yield. The author is not aware of data with

land plants that fit this model. This situation might occur with submerged

or free-floating aquatics if their presence did not change the net radiation.



D. ET IS



IhmEPENDENT OF



Y AFTER REASONABLY

COMPLETE COVER

Is A-ITAINED



This relationship is shown in Fig. 1 as model D. A portion of the

data of Holmen et al. (1961) is plotted in Fig. 3 for illustration. In Fig.

3A, ET is plotted against bromegrass hay yields for the 1955 and 1956

seasons at Upham, in central North Dakota. The high-moisture plots were

irrigated when 40 per cent of the available water was depleted to a

depth of 4 feet, and the medium-moisture plots when 70 per cent was

depleted. Rainfall was 9.8 inches each year. In 1955, three cuttings were

taken in a 144-day season, the data are for plots getting 80, 160, or 200

pounds of nitrogen per acre. In 1956, two cuttings were taken in a 133day season; data are from plots getting 0,40,80, or 160 pounds of nitrogen

per acre. Yields were lower in 1956 because of severe winterkilling the

previous winter. ET for medium moisture plots was slightly lower in

both years than for high moisture plots. In each year and in each moisture

regime fertilization markedly increased yields, but had on significant

effect on ET. In Fig. 3B water-use efficiency is shown to be a linear

function of yield whether yield is changed by season, irrigation regime,

or fertilizer application. Only the nonirrigated plots to be discussed later

do not conform to this model. As part of the same study, Holmen et aZ.

(1961), using the same treatments but cutting with twice the frequency

to simulate pasturing, found that again ET was unaffected by fertilizer



FERTILIZERS AND THE EFFICIENT USE OF WATER



241



application even though pasture yields were increased two- to threefold.

Water-use efficiency again was linearly related to yield.

Weaver and Pearson (1956) measured ET and Y of Sudangrass from

2-inch height to soft dough stage (July 3-30) at Auburn, Alabama.

Treatments were two moisture levels, two nitrogen levels (0 or 100

pounds per acre) and three stand densities in all combinations. All

subplots were 14 by 4 2 / 3 feet contained in a bin of Lloyd clay loam



30t



400

.



.-



c



r



A



=;



*+



. ++.



i



B



-



I



0



2



4

Tons per acre



I



6



Fig. 3. A. Evapotranspiration ( ET) in relation to yield ( Y ) of smooth bromegrass

in central North Dakota. B. Water-use efficiency (Y/ET) of smooth bromegrass as

a function of yield. Filled circles ( 0 ) denote results for a high moisture level and

crosses ( + ) for a medium moisture level, both levels being attained with irrigation.

The open circles ( 0 ) denote results for nonirrigated land. (From Holmen et al.,

1961.)



that had a concrete wall extending 4 inches above the soil surface.

Opportunities for advection would appear to be high. On the highmoisture treatment ( irrigation at %-atmosphere tension at 8-inch depth)

mean of the three populations for no nitrogen were: Y = 1791 pounds/

acre, ET = 4.91 in. and Y/ET = 365 pounds/acre-inch. For the fertilized

plots the means were Y = 3089 pounds/acre, ET = 5.10 inches, and Y/

ET = 606 pounds/acre-inch. Fertilization had no significant effect on

ET but water-use efficiency was markedly increased because Y was



242



FRANK G . VIETS,



JR.



increased. On the low-moisture plots (irrigated to prevent wilting)

ET and Y were lower but fertilization still increased Y, had no effect on

ET, and increased water-me efficiency. Highest water-use efficiency was

obtained with high moisture and N application.

Carlson et al. (1959) obtained marked increases in water-use

efficiency with irrigated corn by the use of 120 pounds of nitrogen per

acre in central North Dakota, in both 1956 and 1957 as shown in Table

V. Two stand densities were used, and plots were irrigated so that

available moisture in the 0- to 4-foot depth did not drop below 40 per

TABLE V

Forage Yield, Evapotranspiration, and Water-Use Efficiency of Corn for Two Levels

Each of Nitrogen Fertilizer, Moisture, and Plant Densitya

Nonirrigated



Irrigated



Treatment



Y

(Ib./acre)



ET

(in.)



A

B

C

D



6560

7330

7170

7240



10.82

11.11

10.57

10.97



1956

606

660

678

660



5870

9330

6870

10630



13.91

14.81

15.56

13.77



422

630

442

772



A

B



5060

5950

5950

5860



10.03

10.39

10.14

9.82



1957

504

573

587

597



6780

9110

8300

10900



16.89

18.38

19.40

16.53



401

496

428

659



C

D



Y/ET

(lb./acre-in.)



Y

(Ib./acre)



ET

Y/ET

(in.) (Ib./acre-in.)



Adapted from Carlson et ul. (1959).

A: No nitrogen; 14,000 plants per acre 1956, 10,000 in 1957. B: 120 lb.

nitrogen; 14,000 plants per acre 1956, 10,000 in 1957. C: no nitrogen; 23,000 plants

per acre 1956, 20,000 in 1957. D: 120 Ib. nitrogen; 23,000 plants per acre 1956,

20,000 in 1957.

a

b



cent of storage capacity. Nitrogen fertilization did not affect seasonal

evapotranspiration. Data of Jensen and Sletten for the M 4 moisture

treatment reported in Table IV showed no effect of fertilization on

evapotranspiration of winter wheat in 1958 at Bushland, Texas, even

though yields and water-use efficiency were almost doubled.

Penman (1956a) plotted the cumulative growth of orchardgrass

(Dactylis gZomeruta) from May, 1954, to November, 1955, against the

potential evapotranspiration on plots where water was nonlimiting at

Woburn, England. Grass was cut when 8 to 9 inches in height. Two

approximately straight lines were obtained; one for a plot getting 0.15

cwt. nitrogen per acre after each cutting gave less yield than for a plot

getting 0.30 cwt. nitrogen per acre. The potential transpiration was



FERTILIZERS AND THE EFFICIENT USE OF WATER



243



calculated from meteorological data; evapotranspiration was not actually

measured. If the actual evapotranspiration was identical with potential

evaporation and not influenced by fertilization, then fertilization produced a marked increase in water-use efficiency.Schofield (1952) stated

that evapotranspiration of grass in lysimters at Rothamsted was measurably the same over a threefold range of dry matter production produced

by the application of fertilizer. The water-use efficiency would therefore

be tripled. Bryan and Brown (1961) found that nitrogen fertilization of

cotton, both irrigated and nonirrigated, had no consistent, measurable

influence on the evapotranspiration rate in studies on Sharkey clay and

Grenada silt loam in eastern Arkansas.

AS Y INCREASES

E. ET DECREASES



This model is shown as E of Fig. 1. Since ET is so dependent on

total net radiation and Y so dependent on the shortwave (visible)

portion of the total net radiation, it is quite unlikely that ET could

decrease as Y increases. The author is not aware of data fitting this model.

If means are found to increase the efficiency of photosynthesis, thus

converting more radiant energy into chemical energy and leaving less

energy for heat, then ET could conceivably decrease as L increases.



F. ET INCREASES

FASTER

THANY INCREASES

No data are known to the author which fit this model ( F in Fig. 1 ) .

ET is strictly a physical phenomenon dependent on the energy available

for evaporation, provided that water is available for evaporation and

that transpiration is not limited by the capacity of the plant to transmit

water through root damage or plant senesence. Y on the other hand is a

complex phenomenon depending to a large extent on soil and crop

management practices, such as fertilization, plant spacing, and temperature, that affect the production of leaf area and its assimilation rate.

Therefore, it is reasonable to assume that ET is always at or near the

radiation potential when water is adequate, whereas Y is not. So ET cannot increase faster than dry matter production. Instances could probably

be found in which overfertilization decreases the salable product or

cuts the production of grain while total dry matter production and ET

are increasing. Such instances would fit this model.



THATCANNOTBE CLASSZFIED

G. EXPERIMENTS

The reports cited below contain worth-while information that cannot

be classified in the six models because of insufficient points to establish

an ET vs. Y function with certainty, too little information, or ET was

not measured separately for each fertility treatment.



244



FRANK G. VIETS, JFt.



Stanberry ct al. (1955) conducted a 3-year experiment with alfalfa,

under probably the greatest conditions of advection possible in the field,

on small plots surrounded by desert on the Yuma Mesa, Arizona. The

main plots were three irrigation regimes as follows: (1) “dry”-irrigate

( 5 or 6 inches of water) when plants show darkening of foliage and

top-droop that precedes wilting; ( 2 ) “medium”-apply 4 inches of

water whenever tensions at 12 or 18 inches approximate 600cm. H2O

tension; and ( 3 ) “wet”-add a 2-inch irrigation when tension approximates 200cm. H,O at 12-inch depth. The quantities of water used on

these main plots were 215, 223, and 255 inches for the 3-year period

and include the meager rainfall. Each irrigation plot contained four

subplots (15 by 15 feet) getting 1, 2,4, and 6 cwt. PzOj per acre banded

before seeding. Stanberry et al. stated that there was no deep percolation,

and they assumed that water applied plus rainfall represented the evapotranspiration. No measurements were made of evapotranspiration on

individually fertilized subplots. This means that if fertilizers did increase

ET after irrigations, the effects were compensated by decreased ET

due to decreased moisture availability before the next irrigation. The

significant data are shown in Table VI. The authors concluded that

the irrigation regimes had no significant effect on water-use efficiency,

TABLE VI

Total Yields, Evapotranspiration, and Water-Use Efficiency for Alfalfa for 1950-1952

at Yuma, Arizona, as Affected by Irrigation Treatment and

Phosphate Fertilizationa

Irrigation treatment



P20,

Applied

(lb./acre)

100

200

400

600

ET (in.)

a

b



Medium



Dry



Y



Y/ET



Y



Y/ET



Wet



Y



Y/ET



( T.b/acre)(T./acre-in. ) ( T./acre) ( T./acre-in.) (T./acre) ( T./acre-in.)



25.77

28.62

34.52

34.94



0.120

0.133

0.160

0.162

215.3



28.74

30.73

37.65

40.57



0.129

0.138

0.169

0.182

223.3



33.29

37.72

44.79

47.20



0.130

0.148

0.175

0.185

255.3



After Stanberry et al. (1955).

Ton (2000 pounds).



but that the use of superphosphate increased it. However, Stanberry

(1959) showed for all fertility treatments in the experiment that wateruse efficiency on the medium treatment was about 90 per cent of the

value for the wet treatment and about 80 per cent of the value for the

dry treatment.

Keller (1954), in a study of greenhouse techniques for measuring



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

IV. The Effects of Fertilizers on the Relationship of Evapotranspiration and Yield

Tải bản đầy đủ ngay(0 tr)

×