EFFECT OF CROP AND SURFACE MULCHES ON RUNOFF, SOIL LOSSES, ANDSOIL AGGREGATION1

C. A. VAN DOREN AND R. S.

IN RECENT YEARS the effect of cropping systems andsurface mulches on infiltration, runoff, soil loss,

and soil structure have received considerable atten-tion. It is generally recognized that different croppingsystems do affect soil structure differently. It also hasbeen definitely established that surface mulches ofplant material are very, effective in reducing soil andsurface water losses. However, many problems relat-ing to the practical application of stubble mulch cul-ture are still unsolved. Further study is also neededon the specific effects of the continued use of surfacemulches on the soil, particularly on soil structure.Although many general principles have been estab-lished, specific results may vary considerable depend-ing upon the kind of mulch, kind of crop grown, till-age practices, soil type, and climatic conditions.

Musgrave (g)3 has called attention to the impor-tance of rainfall penetration to crop production.Browning (2) discussed many changes in erodibilityof soils which are brought about by the application oforganic matter. Hendrickson (6), Borst and Wood-burn (i) , Duley (4), Musgrave (9), and Kidder,et al. (7) emphasized that raindrops falling on ex-posed soil compact the surface and clog the pores ofthe soil at the surface with fine soil particles. Thiscompaction and clogging reduces the permeabilityof the surface layer and may result in relatively lowinfiltration on some otherwise well-aggregated soils.Such a condition is brought about when the surfaceof the soil is unprotected from the impact of fallingraindrops by a mulch or by the canopy furnished bya growing crop. Therefore, a soil that is well aggre-gated and would ordinarily be expected to be quitepermeable to water may be highly susceptible to sur-face runoff and erosion if the surface is unprotected.

This paper reports the results of 3 years of study todetermine the effect of corn and soybeans producedwith and without surface mulches on soil and surfacewater losses and on structure of the surface soil. Aprevious paper (7) reported the infiltration obtainedfrom the various plots during the early part of theinvestigation.

PROCEDURE AND METHODSIn the spring of 1941, and again in 1942 and 1943, 24 plots,

each 24.75 feet square, were planted in four series of six plotseach. The following crops and treatments were located atrandom on the six plots in each series4:

Plots I—Soybeans, clean tilled, no residue.

Plots 2—Soybeans, subsurface tilled, 2 tons wheat straw.Plots 3—Soybeans, clean tilled, residues, returned after

harvest.Plots 4—Corn, clean tilled, no residue.Plots 5—Corn, subsurface tilled, 2 tons wheat straw.Plots 6—Corn, clean tilled, stover returned after harvest.On plots 2 and 5 the wheat straw was applied immediately

after planting the respective crops. Subsurface tillage for con-trol of weeds was practiced on these plots by the use of handtools as the plots were too small to permit the use of fieldtools. On plots 3 the soybeans were cut for seed and thesoybean straw was returned to the plots immediately afterthreshing. On plots 6 after the corn was harvested the stalkswere broken down on the contour. The entire crop was re-moved at harvest time from the plots numbered i, 2, 4, and 5.After completion of artificial rainfall tests in April or Mayof each year, residues were removed from all plots beforeplowing and preparation of seedbeds for the following crop.Soybeans were planted in rows 20 inches apart and corn inrows 40 inches apart.

Artificial rainfall was applied to the plots at the rate of1.75 inches per hour with the type F rainfall simulator. Theequipment used and its operation were described in the previ-ous paper (7). Runoff was determined by weighing all of therunoff in samples collected for intervals of i to 10 minutes,depending on the rate of runoff. Soil losses were determinedin 1941 and 1942 by drying samples of the runoff taken atfrequent intervals. In 1943 the runoff was saved and at theend of i hour, or in some instances at 2-hour intervals, tripli-cate samples of the runoff were taken to determine soil losses.

Soil samples for aggregate analysis were taken in May,1943, before the land was plowed. The samples were takenwith a spade to a depth of S inches from three locations oneach plot and were composited. After becoming air-dry, thesamples were put through a sieve with ^2-inch openings, andaggregate analysis was made by a method very similar tothat described by Peele and Beale (n).

The plots were located at Urbana, 111., on a permeableprairie soil with a 4% slope. The soil is similar to Saybrooksilt loam, but is leached of carbonates to a greater depth thanis usual for Saybrook.. A more complete description of thesoil was given in a previous publication (7).

RESULTS AND DISCUSSIONRUNOFF

Effect of crop.—It is generally recognized that acrop of soybeans leaves the soil loose, and it is com-monly assumed that the soil in this loosened conditionis particularly susceptible to runoff and erosion.Browning, et al. (3) suggested three factors whichone may logically assume to be primarily responsiblefor the looseness of the soil following a crop of soy-beans. However, the looseness of the soil following asoybean crop may result in increased infiltration andreduced runoff from soybean as compared with cornplots providing the surface of the soil is protected by

Contribution by the Soil Conservation Service, U. S. Dept, of Agriculture, and the Illinois Agricultural Experiment Station,Urbana, 111., cooperating. Published with the approval of the Director:

2Soil Conservationist, U. .S. Dept, of Agriculture, and Cooperative Agent and Assistant Professor of Soil Physics, IllinoisAgricultural Experiment Station, respectively. The authors wish to express appreciation to E. H. Kidder, formerly AssistantHydraulic Engineer, and E. G. Reimann, Assistant Hydraulic Engineer, Soil Conservation Service, U. S. Dept, of Agriculture,for valuable assistance during the conduct of this study.

3Figures in parenthesis refer to "Literature Cited", p. 101.4The same crop and treatment were applied to the above plots during each of the 3 years included in this study.

97

SOIL SCIENCE SOCIETY PROCEEDINGS IQ43

FIG. i.—Effect of mulch on compactness of soil surface onsoybean plot in November, 1941. Left, sample of soil frommulched plot; right, sample from unmulched plot.

a mulch to prevent sealing of the soil at the surface.The mulch prevents the surface of the soil from be-coming compacted and prevents the raindrops fromdispersing the fine particles which would eventuallyclog the soil pores (Fig. i).

During the summer of 1941 tillage conditions werecomparable on the soybean and corn plots as culti-vation was performed with sweep cultivators without

, throwing soil to the rows. In October 1941 and againin April 1942 the average runoff from the corn plotswith straw mulch was greater than from comparablesoybean plots (Table I, Fig. 2). The difference inrunoff between corn and soybean plots became greaterafter continued application of water. For example, in

April, 1942, the average per cent runoff from themulched plots for the first 60" minutes of the initialstorm was 3% from the soybean plots and 4% fromthe corn plots, which is a difference of only i%,while the difference at the end of 2 hours was 14% 4at 3 hours, 22% ; and at the end of 4 hours, 28%. Thedifferences in runoff during the wet runs which wereapplied on the day following the initial runs weregreater than during the initial runs. These differencesproved significant when analyzed by the chi-squaretest.

Effect of residues.—Corn stover was more effectivein reducing runoff than soybean residues. The averagepercentage runoff for 120 minutes from soybean plotswith residues was 62 in October, 1941, and 47'inApril, 1942, while comparable losses from corn plotswith stover mulch were only 26% in October, 1941,and 18% in April, 1942 (Table i). As explained byKidder, et al. (7), corn stalks and leaves provided agreater volume of mulching material than soybeanresidues. Also, the stalks were broken down on thecontour encouraging formation of large ponds whichprotected the surface from the impact of the raindrops(Fig. 5).

Effect of duration oj runs.—Runoff was practicallyconstant from the bare plots after the second hour,but it continued to increase gradually on all mulchedplots through both a 6-hour initial run and a 4-hour.wet run (Figs. 3 and 4). Neal (10) found runoff tobe practically constant on exposed soil after i to 2hours. Musgrave (8) called attention to the fact thatspecific values for the infiltration rate of a soil type

TABLE i.—Cumulative soil and water losses from artificial rains from initial and wet runs under various cover conditions *

T,yp,

mulch

Dateof

run

Av.lengthof ini-

tial run,min.

Initial run

60 minutes

Soybeans Corn

1 20 minutes

Soybeans Corn

Wet run .

60 minutes

Soybeans Corn

1 20 minutes

Soybeans Corn

Average Runoff in Per Cent of RainfallNone . . .None. . .None . . .None . . .Straw. . .Straw. . .Straw. . .Straw. . .Residue .Residue .

June, 1941Oct., 1941Apr., 1942June, 1942June, 1941Oct., 1941Apr., 1942June, 1942Oct., 1941Apr., 1942

H790

163. H3

234279369496214308

487374855432

4725

508678

3H4

189

898632

6247

620i?2618

73

21292O157769

235352

4546

363120

6it63

57Average Cumulative Soil Losses in Pounds Per acre

None. . .None . . .None. . .None . . .Straw. . .Straw. . .Straw. . .Straw. . .Residue .Residue.

June, 1941Oct., 1941Apr., 1942June, 1942June, 1941Oct., 1941Apr., 1942June, 1942Oct., 1941Apr,, 1942

i,7561,8973,3624,881

82972635

605722

3,0952,4314,118

2794 •36

251159

10,3832222306591

1,168 .2,394

74253266

546725

2,281

153254150H5461

i,378

138365347276519

505425380

720

i,i73*The initial runs were made under existing soil moisture conditions. The wet runs were made on the plots on the day following the initial runs.tEstimated.

70

I

I

VAN DOREN AND STAUFFER: EFFECT OF CROP AND SURFACE MULCHES

TOO

99

- - Corn

W/T1AL W£T

0-1 I-Z Z-3 3-4. 0-1 I-ZTIME - HOURS

FIG. 2.—Average runoff losses in percentage of rainfall fromstraw-mulched plots receiving an application of artificialrainfall at a rate of 1.75 inches per hour. Initial and wetruns, October 1941 and April 1942.

are virtually nonexistent, and measured rates are ofvalue largely from a comparative and relative view-point. He states further that relative infiltration rateson a given soil type are frequently masked by suchfactors as degree of slope, length of slope, vegetativecover, tillage operations, variations in soil moisture,and temperature. Results recorded in Figs. 3 and 4indicate that mulch has materially decreased the rela-tive rates of runoff. Also, the continued increasingtrends in pounds 'of runoff by hourly periods on the

\

^200

//7///C7/

Wheat * >frew

TIME - HOURSFIG. 3.—Hourly runoff in pounds from soybean plots 6 by 12

ft. in size as a result of application of artificial rainfall ata rate of 1.75 inches (656 pounds)per hour. April, 1943.

600

C 3to

I0; 200

Sfov.t,

• Inifial- Wet

I-Z Z-3 3-4 4-5

TIME - HOURSS-&

FIG. 4.—Hourly runoff in pounds from corn plots 6 by 12 feetin size as a result of application of artificial rainfall at arate of 1.75 inches (656 pounds) per hour. April, 1943.

mulched plots indicate that a constant infiltration r.atehad not been reached even after an initial applicationof 6 hours plus an additional wet run of 4 hours onthe following day.

Effect oj season.—Different results were obtainedin runoff values during June as compared to runoffduring the fall and spring months in 1941 and 1942.The crops were seeded on the same, dates in bothyears. Natural rainfall was well above normal andwell distributed during June of both seasons. Theaverage runoff from the bare plots in October, 1941,and the following April was significantly greater thanin June, 1941. The corn plots were not run in June,1942. However, the bare soybean plots which weretested in June and during the following April yieldedsignificant differences showing greater runoff duringJune instead of during April as in the previous season.This apparent discrepancy 'can be explained by thefact that the dates of the runs on the various replicatedplots averaged 13 days later in June, 1941 than inJune, 1942. The increased growth of the plants in1941 provided an additional canopy effect which wasnot present in 1942. Actual effect of season on infil-tration on crop land is difficult to obtain as cropconditions are seldom identical in. different seasonsof one year or in different years.

SOIL LOSSESEffect of mulch and residues.—Average accumu-

lated soil losses in pounds per acre ranged from 14 to149 times greater frorn unmulched plots than fromplots protected by straw mulch (Table i). For ex-ample, the soil lost from soybean plots, unmulched,for the first 60 minutes in April, 1942, was 3,362pounds per acre, which is 129 times the 26 poundsper acre lost from the soybean plots which were

IOO SOIL SCIENCE SOCIETY PROCEEDINGS 1943

mulched with straw. Significant differences betweenthe bare and mulched plots existed during June, 1941,October, 1941, April, 1942, June, 1942, and April,1943, both on corn and soybean plots. Point quadratreadings on the straw-mulched plots during thewinter and spring months indicated that 91% of theground surface was protected by the mulch from thedirect impact of raindrops.

Significantly greater soil losses occurred in all testsfrom the plots mulched with soybean residues thanfrom the plots mulched with corn stover. Soybeanresidues consisting in part of leaves which decomposerapidly did not cover as much of the ground surfaceand did not provide as much surface storage of waterin pounds as corn-stover mulch. Readings made witha point quadrat during the winter months showed thatan average of 27% of the ground surface was exposedto the direct impact of rain on the plots mulched withsoybean residues, while only 13% of the ground sur-face was exposed on the plots mulched with cornstover (Fig. 5).

Effect of duration of runs.—Soil losses by hourlyincrements for the April, .1943, runs were analyzedto show the effect of duration of rainfall. Soil lossesfrom the umnulched plots increased to the end of thethird hour and then decreased for the 6o-minuteperiod ending with the fourth hour. Rainfall was notapplied to the umnulched plots beyond a 4-hourperiod. Neal (10) found that soil losses were ap-proximately constant from an unmulched surface aftertwo hours. In general soil losses remained practicallyconstant on the mulched plots after the end of thesecond hour.

Soil losses were greater from the plots during thewet runs than during the initial runs both with andwithout mulches (Table i). High soil moisture on theplots at the beginning of the wet runs contributed tothe increased soil losses from these runs.

SOIL AGGREGATIONSince the results of the aggregate analyses reported

here are based on one set of soil samples, they should

be considered as preliminary in nature. However, theeffects of both crop and treatment on soil aggregationseemed to be so definite it was considered worth-while to include the results in this paper. It is theintention to pursue these studies further.

The results are given in graphic form in Fig. 6.Each point on the graph represents an average ofduplicate determinations on samples from one plotin each of the four series, or eight single determi-nations. The agreement between duplicate results wasusually satisfactory. While the results secured oncomparable plots in the different series were notalways in close agreement, .the trends were always inthe same direction.

Other investigators have found that soil aggre-gation is affected by systems of cropping. An exami-nation pf the graphs in Fig. 6 show this to be thecase with the results reported in this paper. The soilfrom the bare corn plots contained a higher percent-age of total aggregates larger than o.io mm than thatfrom the bare soybean plots, due chiefly to the higherpercentage of aggregates larger than 0.50 mm in thecorn plots. Using the chi-square test, the differencein aggregation was significant. On,the mulched plotsthere was no significant difference in total aggregationbetween the soybeans and corn, probably because thedifferences due to the crops were masked by theeffects of the straw mulch. When residues were re-turned the corn plots showed significantly higheraggregation than the soybean plots.

The straw mulch very definitely favored aggre-gation of the soil. This was true for both corn andsoybeans and for all sizes of aggregates except theo.io to 0.25 mm aggregates on the mulched corn plot.The straw mulch seemed to be particularly effectivein promoting the formation of the large (0.50 mm)aggregates on the corn plot. This helps to explain thevery loose and' friable condition of the soil understraw mulches, but the results do not help to explainthe more friable conditions of the soybean plots ascompared to the corn plots.

HID O./O - 0.25 Mm.CD 0.25-0.50 Mm.

FIG. 5.—Protection to soil surface provided by crop residues.Left, soybean residue, 27% of surface exposed; right, cornstover, 13% of surface exposed. November, 1941.

FIG. 6.—Effect on soil aggregation of different crops anddifferent surface treatments.

VAN DOREN AND STAUFFER'. EFFECT OF CROP AND SURFACE MULCHES 101

YIELDS

The yields of corn and soybeans when grown withand without mulches are shown in Table 2. The aver-age yield of corn from the plots which were mulchedwith 2 tons per acre of wheat straw for the 3-yearperiod was 16.2 bushels per acre less than the averageyield from the unmulched plots. Soybean seed yieldswere not secured in 1941 and hay yields were notsecured in 1943. The average yield of soybean seedfrom the mulched plots for a 2-year period was 2.3bushels per acre more than from the unmulched plots.The mulched soybean plots yielded more hay in 1941and less in 1942 than the unmulched plots.

TABLE 2.—Effect of mulch on yields of corn and soybeans *

Year

194119421943

Ave.

Difference

.Corn, grain peracre, bu.

Mulch

81.155-193-2

76.5

No.mulch

90.592.595-1

92.7

-16.2

Soybeans

Seed per acre,bu.

Mulch

36.226.8

3L5

Nomulch

37421.0

29.2

2.3

Hay per acre,tons

Mulch

3.283.05

3-17

Nomulch

2.643-73

3-!9

-O.O2*Mulch consisted of 2 tons per acre of wheat straw applied to the plots

immediately after planting.

The direct cause of reduction of yield on themulched corn plots is not known. Since the wheatstraw was applied after spring plowing and aftersimilar preparation of seed beds on all plots, reduc-tion in yields was not caused by failure to plow themulched plots. The fact that the mulched plots yieldedless than the unmulched plots with corn but not withsoybeans indicates that availability of nitrates undermulch may be the most important factor retardingyields of mulched crops,

SUMMARYThe paper reports the results of a study to deter-

mine the effect of mulches of wheat straw, soybeanresidue, and corn stover on runoff, soil losses, soilstructure, and yields. The experiment was conductedon a permeable prairie soil with 4% slope. Runoff andsoil losses were determined by the use of the type Frainfall simulator.

On plots mulched with wheat straw under compara-ble tillage conditions runoff and soil losses weregreater from corn plots than from soybean plots.

All of the mulches which were used were effectivein reducing runoff and soil losses. Wheat straw wasparticularly effective as a mulch. Corn stover whichis greater in bulk and. provides a more complete cover-age of the surface than soybean residue was also moreeffective in reducing runoff and erosion.

Corn yields were reduced by the application afterseeding of a wheat straw mulch. Soybean seed yieldswere reduced slightly in 1942 but were increased bythe application of mulch in 1943. Soybean hay yield?were increased by mulching in 1941 but were reducedin 1942.

Preliminary results of aggregate analysis are re-ported. Straw mulch very definitely favored aggre-gation of the soil. Soil from the unmulched corn plotscontained a higher percentage of total aggregates thanthat from unmulched soybean plots.