182 TRAFFIC COMPACTION OF SOIL AND TILLAGE REQUIREMENTS

IV. The Effect of Traffic Compaction on a Number of Soil Properties

W. ARNDT

Large differences in bulk density, soil strength, clod size before and after heavy rain, infiltrationrate and water loss from the tillage layer were measured in adjacent crop, traffic and blank bandsafter row cropping on Tippera clay loam at Katherine, Northern Territory, Australia. These indicatelarge differences in the tillage requirements and responses in adjacent bands usually 18 in wide. Thetillage requirements may eventually be reduced to the control of soil compaction by previous traffic,which could be minimized by choosing the best traffic modules.

1. IntroductionSeveral of the traffic systems described in Part

I of this series are in use on Tippera clay loam, alateritic red earth at Katherine Research Station,Northern Territory, Australia. On severaloccasions a number of soil properties weremeasured in adjacent narrow bands, described astraffic, crop, and blank bands. The technologicaleffects of tillage in the various bands are dis­cussed.

2. Methods2.1. Bulk density

Soil cores of 3 in dia and 2 in long were taken24 h after wetting and covering the sites. Thecrop and traffic bands of bulrush millet stubbleland which had been planted and inter-tilledtwice by the row crop: 2-track: 2-row: fixedtraffic system were sampled to a depth of 6 in in3 X 2 in stages. Similar samples were takenfrom a grazed native pasture and a seedbed ingood planting condition. Samples were alsotaken at 0-2 in in peanut crop bands, in a barefallow after rain, in a traffic band after l2-tractorwheel passages without tillage, and in a fieldplot of loosened soil that was protected fromtraffic during 18 months of open exposure.

2.2. Soil strengthMeasurements were made in the crop and

traffic bands of the stubble of bulrush millet, 7 dafter the first rain storm for the season (1,24 in)when the initial land breaking operations werebeing done.

The calibrated tillage tool was a vertical bar,40 in long, pivotted 10 in from the lower end, towhich a horizontal cutting edge (5 X t in) was

fitted. This was lowered into an excavation andbrought to bear on a soil face 5 in below the sur­face. A spring balance was attached to the upperend of the bar and a manual load was appliedthrough the balance. The load on the cuttingedge at the time of soil failure was calculatedfrom the spring balance reading and the leverratio.

The penetrometer was a simple spring-loadedblade. Readings were taken at the same time asthose with the calibrated tillage tool at horizon­tal intervals of 6 in and vertical intervals of 3 inon a pit face extending across the crop trafficand blank bands of the bulrush millet. The soilsurface level in the crop band was taken as thedatum for the verticle scale.

2.3. Tilth after initial tillageIn the bulrush millet land, sample areas of

crop and traffic bands were dug over by hand toa depth of 6 in with a 6 in width of cut. The dugsoil from 3 samples (18 X 18 in) in each case wasbulked, air dried and separated into clod sizeclasses with sieves of t, t, 1,2,4 and 6 in squareopenings. Adjacent sorghum land planted with arow crop: 2-track: 2-row system and inter-tilledthree times with a row crop: 3-track: 4-row:fixed traffic system was chisel ploughed. Samples,each 16 in square to the full depth of ploughing,were taken from the crop and traffic bands shownin Fig. 1. After air drying they were separatedinto clod size classes with the above sieves.

2.4. Tilth after effective inter-tillageBlocks of land that had been cropped con­

tinuously to peanuts for 4,9 and 14 a respectivelywere available. The sowing and first inter-tillagefollowed the row crop: 2-track: 2-row fixed

W. ARNDT

Fig. 1. Banded differences in soil tilth after chisel plough­ing land cropped to sorghum

traffic system in which alternate inter-rowspaces were subjected to a total ot four passagesof the front and back wheels (T1) . The remaininginter-row spaces were not compacted by traffic(T2) . The second inter-tillage with the sameequipment was staggered so that T1 receivedno further compaction, whilst T2 receivedtwo passages of both wheels.

Soil moisture conditions were excellent for theinter-tillage that followed and the tilth of adjoin­ing inter-row spaces was as uniform as could everbe expected. Five samples to the full depth oftilling and 12 in square were bulked for alternatetraffic bands in the 3 blocks that differed in theirprevious cropping history. After air drying theclod size classes were determined by sieves withL 1. t and I in square openings.

2.5. Soil surface condition after heavy rainIn February 1961 adjacent plots of cotton and

cowpeas in a crop rotation trial and plots ofpeanuts in a monoculture trial were inter-tilledfor the second time following a row crop: 2 track:2 row: fixed traffic system. The traffic bandswere effectively ripped up by extended tines.Elsewhere a bare fallow plot maintained bymeans of a row crop: 2 track: 4 row: fixed trafficof harrowing was ripped up with tines having thesame traffic module as the harrowing operations.

Following heavy rain on all three sites therewere banded differences in the proportions ofsurface soil which remained as stable clodsprotruding above the general level of the sealed

183

surfaces. These stable clods (> 1 in dia) werecollected from 15 ft sections in adjacent trafficand blank bands on the cotton, cowpea andpeanut lands and were weighed after air drying.The differences on the bare fallow plot wererecorded photographically.

2.6. Infiltration rates

A double ring infiltrometer with a 12 in innerring, a 22 in outer ring and a head of 0·7 in wasused. The infiltration into the inner ring wasrecorded after 5, 10, 15,20,40,60 and 90 min forreplicated samples of the crop traffic and blankbands of the bulrush millet area.

2.7. Soil water storage

After artificially wetting the crop and trafficbands of the bulrush millet land during the ringinfiltrometer tests, the top inch of soil wasloosened by hand tilling. After 3 d of hot dryweather the loose top inch of soil was removedand a core of soil 4·5 in in dia was removed fromthe 1-7 in level on each site. The cores were ovendried for the determination of the gravimetricand volumetric soil water contents.

After 7 d of hot dry weather following the firstrain storm for the season (1,24 in) the gravi­metric soil water contents in 2 in stages to 12 inwere determined for the crop and traffic bands.The volumetric water contents were determinedusing associated bulk density determinations.

2.8. The pattern of water loss by evaporationfrom top soil

The water loss pattern for top soil wasexamined in cultivated field plots that weresaturated and firmed by light rolling during thedry season. Soil water contents were determinedat depths of 0-1, 1-2,2-3, 3-6 and 6-9 in after10, 25 and 65 d of open exposure.

3. Results3.1. Bulk density

In the traffic bands the mean bulk density for0-2 in was recorded as 1·3 g/cm3 (Fig. 2). Thismean does not describe the samples adequatelysince they consisted of 1 in of very loose soil over1 in of soil as compact as that in the 2-4 in layer.This occurs because the top inch of soil in atraffic band is pulverized rather than compacted

184 TRAFFIC COMPACTION OF SOIL AND TILLAGE REQUIREMENTS

Fig. 3. Size distribution of clods after the first breakingof the crop bands ofsorghum (1) and bulrush millet (2), andthe adjacent traffic bands of the same sorghum (3) and

millet (4)

3.2. Soil strengthThe loads required to shatter the soil in the

crop and traffic bands were 117 (± 6) and 459(± 123) lb respectively. A similar order of differ­ence was also found by the penetrometer readingswhich were 29 (± 10) lb for the crop band , 28(± 15) lb for the blank band , and 105 (± 5) Ibfor the traffic band.

6

Clod s ize, in dia

o

-g 10

U

3.3. Tilth after initial tillageThe clod size distribution was remarkably

similar for hand-dug bulrush millet land andchisel ploughed sorghum land in both crop andtraffic zones (Fig. 3). Error terms cannot begiven for the bulked samples of bulrush milletland. On the sorghum land the mean contents ofclods « t in) in the crop and traffic bands were74·6 (± 7,0) and 34·5 (± 5·0) % respectively, the

0> 100

.§ 90o" 80E" 70ooo

~

different bulk densities which subsequently needto be reduced and increased respectively bytillage to restore the seedbed conditions.

Fig. 2 suggests that the bulk density decreasesrather than increases when a seedbed is croppedwithout traffic. Elsewhere the bulk density for0-2 in was 1·20 ( ± 0,05) g/cm3 for a bare fallowthat had been levelled by heavy rain prior toplanting and 1·22 (± 0,05) g/cm3 for the cropbands of a peanut crop. After 18 months ofexposure, bare soil without traffic had a bulkdensity of 1-14 (± 0'04) g/cm3• The combinedevidence suggests that this particular soil doesnot have the strong natural tendency for self­compaction that casual observation suggests.

Since there was no significant change in thebulk density at 2-4 in when the number of wheelpassages was increased from 6 to 12, and sincethese bulk densities are similar to that underheavily grazed pasture, tractor wheels do notappear to be any more serious than animalhooves from the point of viewof soil compaction.

The bulk density at 6 in on arable land (seed­bed, crop band and traffic band) was generallysimilar to or lower than that in grazed nativepasture land. In the seedbed, bulk density wasconstant to 6 in except for the loose surfacemulch. As the surface was approached, the bulkdensity decreased in the crop band and increasedin the traffic band except for the loose layer in thetop inch. The bulk density of the seedbed is inter­mediate between those of the traffic and cropbands. As the surface is approached, the effectsof traffic on one hand, and presumably the rootsystem of the crop on the other, produce very

6

5

s: 3

a.<lI

o 4

c

Fig. 2. Bulk density profiles in crop bands (1) and trafficbands (2) ofbulrush millet stubble land, a seedbed (3) and a

grazed native pasture (4)

2

by traffic and shallow tillage owing to the lowwater content at the time of operating. Thislayer is also supplemented by loose alluvialdeposition. The profile in Fig. 2 was drawn toallow for this observation. At 2-4 in the bulkdensity of the traffic band after 6 wheel passages(1'51 ± 0·2 g/cm3) was similar to that in grazednative pasture (1,5 ± 0·1 g/cm3) and after 12wheel passages (1,49 ± 0·2 g/cm3) .

W. ARNDT

crop band containing twice as much fine materialas the traffic band. In the same bands the meancontents of clods (2-6 in) were 7·2 (± 5,5) and48·0 (± 4,2) %respectively. The content of clods0-2in) was 17·6 %in both cases. The outstandingdifferences are the 7-fold increase in the pro­portion of large clods (2-6 in) and the 16-foldincrease in the 50% mean clod size both due totraffic effects.

3.4. Tilth after effective infer-tillage

The number of years of previou s cropping hadno effect on the weight of soil disturbed on eachof two classes of traffic band (Table I). When

TABLE 1The weight of air dry soil (kg) from equal areas (5 samples)of the tilled layer for two types of traffic bands on peanutland, previously cropped for differing numbers of years

Years ofpeanut Type of traffic bandmonoculture

T1 T2 Meall

4 30·3 18·4 24·49 27·0 17·9 22-4

14 26·4 19·7 23·1

Mean 27·9 18·7

the various years of previou s cropping areregarded as replicates, the effect of traffic differ­ences on the weight of soil disturbed is verymarked. This is due to the combined effects ofdifferences in the soil bulk density, surface leveland the rigidity of the respective gangs of tines.Since the test was made under ideal soil moistureconditions, the difference of 50% probablyrepresents the minimum difference expectedfrom similar causes under general field condi­tions. The proportions of various clod sizes areshown in Table II. Once again the age of theland had no effect, and when years of previouscropping are regarded as replicates, traffic effectstended to change the proportions of fine« ! in) and coarse (> I in) clods. The physicalsignificance of this measurement is not known.If differences in tilth have some agronomic signi­ficance the previous traffic history can beregarded as an important determinant of suchdifferences.

185

TABLE IIClod size distribution ( %) in the tilth for two differenttypes of traffic band on peanut land previously cropped for

differing numbers of years

Type Years of Clod size class, illof mono-

traffic culture ! 1 1 t-t t-I 1-2"8"-4

---- - - --4 42·8 7·1 10·6 16·7 22·8

T, 9 48·3 5-6 7-9 13·5 24·714 52-6 7·1 8·0 13-2 19·0

-- - - -- - -Mean 47-9 6·6 8·8 14·5 22·2

- - - - -- - -4 56·7 7-3 10·2 17·1 8·6

T2 9 56·2 5-9 7·9 13·5 16·514 62·2 6·5 9·3 13'1 8·9

--- --- - - -- - - --- -Mean 58·4 6-6 9·1 14-6 11·3

3.5. So il surface conditions after heavy rain

In Fig. 4, the smooth and completely sealedsurfaces on the ex-blank bands and the roughpartly sealed surfaces on the ex-traffic bandsshould be noted. On the cowpea and cottonlands which were similar, the mean weights ofsoil in the weather resistant clods on the trafficand blank band s were 41·4 (± 8,9) and 17·2( ± 3'9) kg per 15 ft section respectively. On thepeanut land the respective amounts were 10·3( ± 3'3) and 1·6 ( ± 0,5) kg. In all cases the morestable tilth was associated with soil compactionpr ior to tillage.

Fig. 4. Banded differences ill the resistance of the tilth toslaking by high-intensity rain after deep tine tillage of a

bare fa llow

186 TRAFFIC COMPACTION OF SOIL AND TILLAGE REQUIREMENTS

Fig. 5. Gravimetric and volumetric soil water contents inloose crop bands (1) and compacted bands (2) 7 dafter

1·24 in rain

3.8. The pattern of water loss from top soilThe soil water profiles found after 10, 25 and

65 d of open exposure (Fig. 6) show that water islost by evaporation at a decreasing rate to adepth approaching 6 in over the first 10 d andthe field capacity water content is maintained atgreater depth. After 10 d water tends to be lostat a more constant rate with increasing depth

3.6. Infiltration ratesThe infiltration rates on bulrush millet land

after a 90 min test commencing with dry soilwere similar for the crop and blank bands andabout 6 times greater than on the compactedtraffic band, as follows:

Crop band 16·5 (± 3'5) injhBlank band 17·6 (± 2,5) injhTraffic band 2·9 (± 0,6) injh

Sail water content, 0/0

0 5 10 IS 20

·~~.3III

",I'0\\ .00.0o

.s 3 "0Ii

s: Li:Q.~ 4

'0 •• •VI 5

II

8

• •7

4. DiscussionThe large differences in the magnitude of soil

physical properties found in highly organizedbands can largely be attributed to differences inthe bulk density of the soil due to differences insoil compaction by traffic.

The bulk density measurements indicate thatcropping alone does not induce undesirablelevels of soil compaction over one croppingseason and therefore should not be responsiblefor large annual expenditures on tillage. Thetillage requirements are largely determined bythe residual effectof trafficwhich in turn is deter­mined by the traffic system and the efficiency oftillage described in Part I. Since the theory forthe overall relative magnitude of soil propertiesin a field (Part II) further shows that adverseeffects of traffic may be ignored while the areaaffected (Xl) remains relatively small, the largeannual expenditure commonly devoted to tillagecan be questioned. The data in Fig. 1 suggestthat the soil in the crop band was loosened by

Fig. 6. Gravimetric soil water contents 10 (1), 25 (2),and 65 (3) d after saturation and open exposure during the

dry season

and from greater depths. Tentatively it is sug­gested that during the first 10 d or more afterrain liquid water movement predominates. Thenet water loss decreases with depth and isrestricted to the tillage layer, as suggested inPart III.

Soil water content

Gravimetric, % Volumetric, cm3/ 100 crrr'

2 4 6 8 10 0 5 10 15 20

In~\ r\III,M

\

4

2

8

12

o(f) 10

s:a. 6.,"0

3.7. Soil water storage- The~gravimetric soil water contents of thecores taken from traffic and crop bands 3 daftersaturation were 14·8 (± 0,9) and 15·2 (± 0·3)%respectively. These values are remarkablysimilar and consequently the volumetric waterstorage in the cores was almost entirely depen­dent upon differences in soil bulk density. Waterstorage in the cores, was 387 (± 23) and 288(± 29) em" respectively.

After 7 d exposure (Fig. 5) the gravimetricwater content of the traffic bands was againslightly lower than that in the crop band and thevolumetric water content was greater, but thedifferences throughout were not significant.

W. ARNDT

the growing crop and should be firmed ratherthan loosened in the next tillage cycle. If the soilcompaction in the traffic bands cannot beignored, then according to the theory for theoverall relative magnitude of soil properties(Part II) it needs to be loosened by tillage. In awell-designed traffic system this should onlyrepresent a small and well-defined fraction of thefield. Since the bands which have very differenttillage requirements are clearly de1iniated by thestubble of the previous crop it is reasonable tosuggest that the tillage treatment could vary asthe soil conditions vary from band to band.

For the better systems described in Part I thefraction of the field that requires loosening bytillage may be as small as one-eighth.

The 4-fold differences in the strength of thesoil in the crop and traffic bands are related to thedifferences in bulk density produced by traffic.In the theory for the overall relative magnitudeof the soil property in a field (R) (Part II) therelative value (r l ) for soil strength is 4. Assumingthat the work, power, or time required for tillageis directly related to soil strength, R = 1·3 whenXl = 0·1 and increases to 3·0 when Xl = 0·66.The general level of R is largely determined bysoil type and conditions of working but the 2- to3-fold difference in,the effort required to till thefield is entirely due to the choice of trafficsystem. Since the traffic systems in common useare frequently faulty, and since they are mucheasier to alter than the physical nature of thesoil, useful and rapid advances are expected fromstudies of tillage traffic systems quite indepen­dent of further advances in the understanding ofthe physical nature of soil compaction.

The increased soil stability due to soil compac­tion by traffic was remarkable, indicating thatthe majority of stable clods in similar fields arederived from the shattering of previous trafficbands. The effect of mechanical pressure on rain­fall resistance of clods was similar to thatpostulated by Vilensky's principleb? and reportsfrom other agricultural regions. In special cases,such as the problem of clods in the mechanical

187

harvesting of potatoes in the U.K., efforts arebeing made to reduce stable clod formation bywheel traffic." In Kansas, U.S.A., the mean sizeof the clods is increased to combat wind erosionby deliberate pre-cultivation compaction. 5 Inregions where high intensity rainfall demands aparticularly stable tilth, as at Katherine, pre­cultivation compaction offers a means ofachieving such stability, but it will involve agreater expenditure on tillage effort. It is becom­ing increasingly clear that variations ofVilensky'sprinciple have a world-wide application that isnot generally recognized as a beneficial side effectof soil compaction by traffic. Where traffic isreduced because of its many adverse effects thisbeneficial side effect may need to be simulated byspecial operations such as rolling.

It appears that for at least 10 d after rain,water loss by evaporation is confined to a fairlyconsistent depth of 6 in in a clay loam, and theamount of water stored and lost in this layerdepends partly upon the bulk density of thelayer. These data and the large differences ininfiltration rates associated with differences inbulk density have been considered in the theorygiven in Part III.

AcknowledgementsThanks are due to R. Wetselaar for providing

the data for Fig. 6 and to E. F. N. Murray forassistance in the design and construction of themeasuring equipment.

REFERENCES

, Arndt, W. Continuous cropping ofpeanuts at Katherine,N.T. Tech. Pap. 16, Div. Ld Res. reg. Surv. CSIRO,Aust., 1961

2 Russell, E. W. Soil structure. Tech. Commun. 37, imp.Bur. Soil Sci., 1938

3 Vilensky, D. G. (Aggregation of soil). USSR Acad.ScL, Moscow, 1945 (Transl. CSIRO, Melbourne)

• Hawkins, J. C.; Brown, N. J. Tillage practices andmechanization. Neth. J. agric, ScL, 1963, 11, 140

5 Lyles, L.; Woodruff, N. P. Surface soil cloddiness inrelation to soil density at time of tillage. Soil Sci.,1961,91 (3) 178