Tải bản đầy đủ - 0 (trang)
II. Nitrogen Balance Sheet for the Cropped Soils of the United States

II. Nitrogen Balance Sheet for the Cropped Soils of the United States

Tải bản đầy đủ - 0trang

215



SOIL N I T R O G E N BALANCES



diately comes to mind. In this study a major attempt was made to account for the income and outgo of soil nitrogen in the United States. A

mass of data was assembled and studied, but they emphasized that in

many cases quantitative information was inadequate for arriving at accurate values for the quantities of nitrogen gained and lost through the

various channels. Their nitrogen balance sheet for cropped soils is

shown in Table I.

TABLE I

Nitrogen Balance Sheet for the Harvested Crop Area of the United States, 1930

Pounds of nitrogen

per acre per year

Additions:

Rain and irrigation

Seeds

Fertilizers

Manures

Symbiotic nitrogen fixation

Nonsymbiotic nitrogen fixatiou



4.7

1.0

1.7

5.2

9.2

6.0

27.8



Losses:

Harvested crops

Erosion

Leaching



25.1

24.2

23.0



Net annual loss



72.3

44.5



It will be observed that no value is given in the table for nitrogen

losses by volatilization. The net annual loss of nitrogen, therefore, may

have been even greater than the value of 44.5 shown.

If Table I were to be revised to correspond to conditions in 1950,

the most striking change in nitrogen additions would be an increase in

the use of fertilizer nitrogen from 1.7 to about 6.7 pounds per acre

annually. The value for addition of nitrogen as manure would be increased only slightly from 5.2 to about 5.7 pounds. The symbiotic nitrogen fixation figure would be higher by approximately 2.5 pounds, owing

chiefly to much larger acreages of alfalfa, lespedeza, and soybeans.

There is no sound basis upon which to estimate nonsymbiotic fixation.

On the other side of the balance sheet, the nitrogen removed in harvested crops has increased by about 6 or 7 pounds per acre, owing

chiefly to larger yields of grain crops, especially corn, and to the growing of larger acreages of legumes. It is impossible to give an accurate



216



F. E. ALLISON



estimate of the changes in erosion that have occurred during the period

since 1930, but undoubtedly there has been a marked decrease as a result of the concentrated attack on this problem by federal and state

agencies. Nitrogen losses by leaching and volatilization may also have

decreased somewhat, but there are few facts upon which to base an

estimate.

The nitrogen balance sheet for the United States, as presented in

Table I, emphasizes the large annual net loss from soils; an up-to-date

revision of the table, based on the most reliable information available,

still shows a large loss but the figure is undoubtedly smaller. The mere

fact that soils are further removed from the virgin state, and that the

soil organic matter level is becoming more nearly stabilized at a lower

level, assures some improvement in the nitrogen balance sheet. Improved farming practices are also undoubtedly exerting a major influence in the same direction.

A balance sheet of the type prepared by Lipman and Conybeare is

interesting but of limited value in the present discussion because it deals

with averages. When a soil nitrogen balance sheet is encountered that

does not balance, usually general information, or average values, are

inadequate for a solution of the enigma.



111. LYSIMETER

EXPERIMENTS

A more accurate idea of what is happening to nitrogen in soils can

be obtained by studies of individual experiments where soils have been

maintained under near-natural conditions, and nitrogen income and

outgo measured over a long period of years. I n general, such data can

be obtained only from lysimeters, where the drainage waters can be

collected and analyzed. Field plot experiments furnish supplemental information, but, of course, accurate balance sheets cannot be constructed.

An occasional pot experiment may also have been performed in such a

manner that it supplies valuable information.

Reference will first be made to several lysimeter experiments reported from state experiment stations. Kohnke et al. (1940) g'ive an

excellent survey and discussion of the types of lysimeters that have been

used for various purposes, and the results obtained. In nitrogen studies

the filled-in type is usually used. Sometimes the soil is placed in layers

corresponding to the natural soil horizons, but a few investigators have

used top soil only. I n either case, the soil is well mixed prior to being

placed in the lysimeters so as to reduce sampling difficulties and to assure that the soils in all lysimeters are exactly alike. Since filled-in

lysimeters are usually buried in soil or gravel in the open to near

ground level, they present conditions for crop growth that vary little



SOIL NITROGEN BALANCES



21 7



from those in undisturbed soil, except for the differences in water penetration, water movement, and drainage conditions.

Lysimeter studies of most value in the present discussion are those

(1) where nitrogenous fertilizers or manures were added in considerable quantity, (2) where no legumes were grown, (3) where all crops

and soils were analyzed, and (4) where the experiment was continued

for five years or longer, thereby reducing the errors due to sampling

and analysis of the soils, plants, and leachates. In all such experiments

erosion is, of course, eliminated.

All of the available data that meet the above requirements cannot

bc discussed here, but the most pertinent and representative results

from the five experiment stations in this country where the most work

of this type has been done will be considered.

1 . Experiments at Ithaca, New York

During the period since 1910 several lysimeter experiments have

been conducted at the Cornell Experiment Station. Concrete tanks were

filled with layers of soil corresponding to the natural soil horizons.

In one experiment, reported by Bizzell (1943), a sandy loam soil

was cropped for 15 years with two vegetable crops each year followed

by a winter cover crop of rye. The treatments were replicated three

times. Nitrogen was added in the forms of ammonia or nitrate to each

vegetable crop at the time of planting. Table I1 shows the nitrogen

balance sheet.

It will be observed that there was a marked decrease in soil nitrogen

during the 15-year period, even though large quantities were added

each year. Furthermore, only a little more than half of the nitrogen

added or made available from the soil was recovered in the crop. About

half of the remainder was found in the leachate, whereas the remaining

portion was unaccounted for. The author states that this unrecovered

nitrogen was presumably lost by volatilization. This loss occurred in a

well-aerated sandy loam that was fertilized with excess nitrogen, intensively cultivated, and limed at intervals to keep the pH near 6.0.

Wilson (1943) believed that such losses are commonly due to loss of

nitrite in the exudate from plants or to loss of nitrogen gas formed as

a result of the reaction of nitrite with amines in plant juices.

Another interesting Cornell lysimeter experiment, using Dunkirk

silty clay loam, was reported by Bizzell (1944). I n this experiment

timothy was grown for 8 years, and sodium nitrate in varying amounts

was added to duplicate lysimeters in three installments each year. The

experimental procedures were similar to those used in the experiment

just discussed. Table I11 shows the nitrogen balance sheet.



F. E. ALLISON



218



TABLE I1



Recovery of Nitrogen from Ammonium Sulfate and Sodium Nitrate

in a 15-Year Lysimeter Experiment' at Ithaca, New York

(Results expressed on a per acre per year basis)

~



~



~



Ammonium

sulfate



Sodium

nitrate



Lb.



N

Nitrogen sources:

Rain

Fertilizer

Soil



Lb.

N



%



6

143

47



6

143

31



Total



%



180



100



196



100



Nitrogen reroveries:



Crops



94



Leachate



43



52

23



110

46



56

23



Totul



136

44



24



156

40



20



Not accounted for

1



Cropping system: two vegetable crops and rye cover annually.



TABLE I11

Recovery of Nitrogen in an 8-year Lysimeter Experiment' a t Ithaca, New York

(Results expressed on a per acre per year basis)

Nos.

4and9

Lb.



N

Nitrogen sources :

Rain

Sodium nitrate



%



6

413



Nos.

5 a n d 10



Nos.

6 a n d 11



Lh.



Lb.



N



6

155



219



100



154

3

17



70

1

8



Total

Not accounted for



174

45



21



Cropping system: continuous timutliy.



N



%



6

144

_



~~



Tutal

Nitrogen recoveries :

Crops

Leachate

Soil



1



%



Nos.

S a n d 12



Lb.

N



%



6



99



_



~



161 100



130



100



99



100



121

2

14



75

1

9



98

1

19



75

1

15



78

4

15



79

%

15



137

24



15



118

14



9



95

4



4



219



SOIL NITROGEN BALANCES



Under continuous timothy the soil showed a gain of 14 to 19 pounds

of nitrogen annually, and the crop recovered 70 to 79 per cent of the

nitrogen available to it. There was almost no loss of nitrogen in the

drainage water, even where 213 pounds were added annually. The

small loss that did occur was in the first year before the timothy was

well established. Regardless of this, there was an unaccounted-for loss

of nitrogen that varied from a negligible amount at the lowest rate of

nitrate addition to 45 pounds per acre per year at the highest rate.

Bizzell considered the losses to result chiefly from volatilization.



2. Experiments at Geneva, New York

Collison et al. (1933) reported the results of a 16-year lysimeter experiment conducted at Geneva, New York, in which soils of relatively

low and high fertility were used without fertilizer additions. A 4-year

rotation of timothy (2 years) , barley, and wheat with two replications

was used. The balance sheet for these data is given in Table IV.

TABLE IV

Nitrogen Balance Sheet for 16-Year Lysimeter Experiment' at Geneva, New York

(Results expressed on a per acre per year basis)

Low-fertility soil



Lb. N

Nitrogcn sources:

Rain

Seed

Soil



Total



1



%



More fertile soil

Lb. N



%



9

2

118



9



2

52

63



100



129



100



Nitrogen recoveries:

Crops

Leachate



44



70

13



69



8



14



53

11



Total

Not accounted for



52

11



17



83

46



36



Cropping system: timothy ( 2 years), barley, and wheat.



Both soils lost large amounts of nitrogen during the experiment, and

the loss increased with the fertility level. It is a little surprising that

even though no fertilizer or manure was added, and the soils were in

timothy for half of the time, there was an unaccounted-for loss of 17 to

36 per cent of the nitrogen available to the crop. The author concluded



220



F. E. ALLISON



that the only explanation for the large nitrogen deficits is that nitrogen

escapes to the atmosphere.



3 . Experiments at Windsor, Connecticut

Rather extensive lysimeter experiments (Jacobson et al., 1948; Morgan and Jacobson, 1942; Morgan et al., 1942a; and Morgan et al.,

1942b), using Merrimac sandy loam, have been reported from the tobacco station at Windsor, Connecticut. In one experiment various nitrogen sources were applied without replication to tobacco grown continuously for 10 years. The nitrogen recovery data are given in Table V.

TABLE V

Nitrogen Balance Sheet for 10-Year Lysimeter Experiment' at Windsor, Connecticut

(Results expressed on a per acre per year basis)



Three

nitrates

Lb.

N

Nitrogen sources:

Rain

Fertilizers

Soil



Total

Nitrogen recoveries:

Crops

Leachate

Soil



Total

Not accounted for

1



Two

ureas and

cyannmitl

Lb.



Lb.



N



%



4



%



4

200



200

19



~

223



Two

NH4+

salts



-



_

100



_

100



204



42

55



73

108

6

- _ _ _

93

123



216

7



1



7



%



Lb.

N



No

nitrogen



Lb.

%



N



%



4

200



4

200



-



-



_

204 100



204



100



40



100



39

40



20

31



50

78



9



-



12



51

+11



36

53

3



88

88

10



43

43

5



79

82

19



8



186

18



9



180

24



187

3



N



Six

organic

fertilizers



4

0

36



0



Continuous tobacco.



In this experiment with a very sandy soil that was heavily fertilized

with nitrogen, it will be observed that usually more of the added nitrogen was recovered in the drainage water than in the crop. Nearly all of

the added nitrate was accounted for but there was an average unaccounted-for loss of about 10 per cent of the nitrogen added in the other

forms.

In a similar experiment, calurea was added to cylinders at the rate

of 200 pounds of nitrogen per acre annually for 10 years and various



22 1



SOIL NITROGEN BALANCES



cropping systems compared, using duplicate treatments. Table VI gives

the nitrogen recovery data.

TABLE VI

Nitrogen Balance for a 10-Year Lysirneter Experiment at Windsor,

Connecticut, Where the Soils Were Fertilized with Calurea

(Results expressed on a per acre per year basis)



No nitrogen



Fertilized with calurea



Tobacco



Tobacco

oat

cover



Lb.

N



Lb.

N



+



Fallow



___--___-Lb.



N

Nitrogen sources:

Rain

Fertilizer

Soil

Total

Nitrogen recoveries:

Crops



Leachate

Soil

Total

Not accounted for



%



4

200

67



%



%



Grass

sod



Lb.

N



Tobacco



%



4

a00



100 POG



100



194



94

48



46



77

28

8



%



Lb.

N



%



4

0

51



4

0

40



100 55



100 44



100



16

6



36



4

190



4

no0

28



Lb.

N



Grass

sod



a

~~~~-~

971



100



239



0

995



83



90

97



39

4%



-



-



-



~

225

46



187

45



149

64



23



~

17



19



31



113

81



40

14



19

99



4 -



43



48

7



34

53



-



IS



~

13



9%

29



50



The average percentage of added nitrogen, plus that released from

the soil that was unaccounted for, varied between 17 and 42 per cent

for the fertilized soils, and was either 13 or 50 per cent for the unfertilized soils. It is especially surprising that the lowest recovery was with

grass sod. Most investigators have observed little or n o losses with sod

crops, and some marked gains, supposedly due to biological nitrogen

fixation, have been reported. The recoveries of nitrogen were the same

where tobacco was grown as where the soil was kept fallow. The introduction of an oat cover crop following tobacco made the nitrogen

balance less satisfactory, even though this additional crop reduced the

nitrogen content of the leachate.

Other experiments similar to those reported in Table VI, where

mixed sources of nitrogen were used instead of calurea, gave recoveries

that agreed closely with the calurea data. These results are at considerable variance with most other lysimeter data.



F. E. ALLISON



222



4 . Experiments at Knoxville, Tennessee



Lysimeter experiments on the recovery of fertilizer nitrogen from

soils have been reported from the Tennessee Experiment Station. In the

two experiments of most interest in the present discussion the soils, unfortunately, were not analyzed at the end of the experiments. The recovery of added nitrogen must, therefore, be based on the difference

between the nitrogen in the drainage waters from fertilized and unfertilized soils.

I n these experiments 989 pounds of nitrogen per acre in the form of

nitrate (Mooers et al., 1927), or of ammonium salts (MacIntire et al.,

1952), were added in a single application to an uncropped silt loam

in lysimeters 6 feet deep (Table VII) .

TABLE VII

Recovery of Nitrogen from Uncropped Cumberland Silt Loam in

Lysimeters at the Tennessee Experiment Station

Nitrogen in leachates



Forms of nitrogen added

First expt.'-5 years

No nitrogen

Calcium nitrate

Magnesium nitrate

Sodium nitrate

Second expt.'-12 years

No nitrogen

Ammonium chloride

Ammonium phosphate

Ammonium sulfate



Total,

lb. per

arre



From

fertilizer,

Ib. per acre



Recovery,



Average

recovery,



%



%



68.3

889.8

871.1

987 4



821.5

80%.8

919.1



83.1

81 .a

93.0



85.8



372. 8

1160.1

1111.3

1294.7



787.3

738.3

851.9



79.6

74.7

86.2



80.4



Single additions of 989 Ib. N per acre. Three replications.



The recovery of added nitrate nitrogen averaged 86 per cent, and

that of ammonia nitrogen 80 per cent. The ammonia nitrogen was of

course removed as nitrates following nitrification. Loss of a fair percentage of the nitrogen by volatilization is indicated.

All of the experiments discussed above have dealt with soils in the

humid regions. Similar data from the semiarid and irrigated regions,

where there is little or no leaching, are sparse. Only one such lysimeter

experiment, namely, that reported by the California Experiment Station, will be discussed here in detail.



SOIL NITROGEN BALANCES



223



5 . Experiments at Riverside, California

The experiment at California, described by Chapman et al. (1949)

and by Broadbent and Chapman (1950), is being conducted in lysimeters that are 10 feet in diameter and 4 feet deep. These were filled

with a Sierra loam top soil having a nitrogen content of 0.04 per cent.

The data for the six lysimeters that were cropped to Sudan grass for

the 15-year period are given in Table VIII.

TABLE VTII

Nitrogen Balance Sheet for 15-Year Lysimeter Experiment’ at Riverside, California

(Results expressed on a per acre per year basis)

No cover crop

2.5 tons straw



g



100Ib. N 20Olb. N



NoN



Mustard cover crop turned under

no straw



No N



100lb. N 2OOlb. N



- L _ _ - - - -



Lb.

N

Nitrogen sources:

Rain and irrigation 8

Fertilizer2and

straw

32

Soil

96



Lb.

%



Lh.



N



%



N



%



Lb.

N



Lb.

%



N



%



Ih.



N



8



8



11



13



13



131

96



237

71



0

120



100

75



a00



- _



~~



~-~



%



48



Total

136 100 235 100 316 100 131 100 188 100 261 100

Nitrogen recoveries:

Crop

84

82 140 GO 171

66 139 74 192 74

54

87

15

Leachate

51

31 113 36

16 40

18 14 31

37 73

Total

Not accounted for

1



2



135

1



1



213

22



9



284

3%



10



105

26



20



170

18



10



432

a9



11



Cropping system: audan grass with and without muatard cover crop.

Nitrogen supplied as calcium nitrate; no replication.



In all cases the soil lost much nitrogen, and there was failure to

account for all of the available nitrogen added or released from the

soil. This unaccounted-for nitrogen varied between 1 and 20 per cent

for the 15-year period. For the first 10 years of the experiment (Chapman et al., 1949) there were indications that the losses by volatilization

from some lysimeters were greater than these values. The earlier results also suggested that nonsymbiotic nitrogen fixation was of considerable importance where carbonaceous plant materials were present,

but the values for the 15-year period do not substantiate this view.



224



F. E. ALLISON



These data emphasize some of the limitations of the lysimeter technique, especially when the treatments are not replicated. Broadbent and

Chapman (1950) point out that the errors of soil sampling were large.

This was due in part to the absence of leaching during the last 5 years

of the experiment, which permitted some accumulation and uneven distribution of nitrates. In addition, there was, unfortunately, an accumulation of 610 to 855 pounds per acre (4-foot depth) of nitrate-nitrogen

in the soils at the time of starting the treatments. This nitrate was

formed from the oxidation of the soil organic matter during the 7-year

period that the soil was allowed to settle following the filling of the

lysimeters.

Results obtained in lysimeter experiments at the Arizona Experiment Station (Smith, 1944) are in some respects comparable to those

obtained at California. In the Arizona experiment a Gila loam and a

Mohave clay, cropped to nonlegumes for 12 years (unreplicated),

showed a large loss and a large gain in nitrogen, respectively. However,

after 18 years, when additional soil analyses were made, the average

yearly gaseous loss in the one soil, and gain through bacterial fixation

in the other, were greatly reduced. Construction of accurate nitrogen

balance sheets for the 18-year period is not possible from the limited

data published to date (Smith, 1953).



6. General Remarks

I n considering lysimeter experiments it is well to bear in mind that

a high degree of accuracy in such experiments cannot be expected.

Chief among the sources of error are the inaccuracies of soil sampling

and the errors involved in the storage and analysis of the leachates. In

addition, few experiments have been adequately replicated. Fortunately, the errors of analysis of the soils and crops need not be great,

since the large numbers of analyses required assures that the individual

plus and minus errors will largely cancel out. Regardless of their limitations, lysimeters do furnish valuable information that cannot be obtained by other methods.

The experiments discussed above bring out the following facts:

1. Crops commonly recovered only 40 to 75 per cent of the nitrogen

that was added or made available from the soil. Low recoveries were

usually obtained where large additions of nitrogen were made, where

the soils were very sandy, and where the crop was not adequate to keep

the nitrogen low.

2. The nitrogen content of most soils decreased regardless of how

much was added as fertilizer unless the soil was kept in uncultivated

crops.

3. A large proportion of the nitrogen not recovered in the crop was



SOIL NITROGEN BALANCES



225



found in the leachate, but substantial unaccounted-for losses occurred

in most lysimeters. Nitrogen gains were few.

4. The magnitude of the unaccounted-for nitrogen was largely independent of the form in which the nitrogen was supplied, whether as

nitrate, ammonia, or organic nitrogen.

5. Unaccounted-for nitrogen was commonly slightly higher in

cropped soils than in fallow soils. Fifty-one lysimeters that received

nitrogen and were planted to nonlegume crops showed an average unaccounted-for loss of 20 per cent of the total available nitrogen; the corresponding figure for 106 uncropped soils was 12 per cent. The average

value is near 15 per cent. A small portion of the loss from the cropped

lysimeters can be accounted for as due to insects and birds, loss of leaves

and pollen, and leaching of nitrogen from the leaves onto soil outside

the lysimeters. Any nonsymbiotic nitrogen fixation that may have occurred would of course increase the loss figures for both cropped and

uncropped soils. These data constitute strong evidence that nitrogen

losses from normal, well-aerated soils, by volatilization, are not

negligible.

The statement is frequently made that lysimeter experiments are

too artificial to show what happens under normal conditions. Drainage

losses seem to be much too high, especially if filled lysimeters are used.

Although this criticism is doubtless justified in many cases, this fact

does not seriously affect the value of the data for use in the present

discussion of the nitrogen balance enigma.



IV. FIELDEXPERIMENTS

Accurate soil nitrogen balance sheets cannot be constructed from

experimental data obtained in ordinary field experiments. This is due

chiefly to lack of knowledge of the quantity of nitrogen removed annually by leaching and erosion, frequent failure to analyze the crops

for total nitrogen, and the uncertainty as to the contribution that the

subsoil has made to the feeding of the crop. Nevertheless, field experiments do supply information that is sufficiently quantitative to show the

magnitude of the main soil losses or gains. Two such experiments will

be discussed below in some detail, and reference will be made to several

others that are of general interest.

1 , Experiments at Rothamsted, England

The experiments conducted on the Broadbalk wheat fields at Rothamsted furnish considerable information on nitrogen losses from soils

because rain gauges are located near the plots. Table IX, constructed

from unreplicated data given by Russell (1950), gives the pertinent

information.



Tài liệu bạn tìm kiếm đã sẵn sàng tải về

II. Nitrogen Balance Sheet for the Cropped Soils of the United States

Tải bản đầy đủ ngay(0 tr)

×