DIVISION S-6—SOIL AND WATER MANAGEMENTAND CONSERVATION

Effect of Phosphorus on Some Physical Properties of Soils: I. Modulus of Rupture1

J. F. LUTZ2 AND RAFAEL A. PiNto3

ABSTRACTApplications of phosphorus, both as phosphoric acid and as

monocalcium phosphate, to three soils were very effective inreducing the hardness of soils as measured by modulus of rup-ture. Phosphoric acid was more effective than monocalciumphosphate on all three soils. Both sources of P were moreeffective on Iredell soil, which consists predominantly of mont-morillonite with some kaolinite, than on White Store soil,which contains montmorillonite and kaolinite but with someillite. Even though both sources of P produced significantdecreases in modulus of rupture on the Georgeville soil, whichcontained predominately 1:1 clay minerals, the results were notas striking as on the Iredell. The greatest effect per unit of Pwas produced by the smaller increments, especially by die50-ppm rate. This is especially important from both agronomicand economic standpoints since 50 to 100 Ib/acre are feasiblerates.

'"pHE RED-YELLOW Podzolic soils of the Southeastern•*- United States are known to have a high phosphate-

fixing capacity. Laboratory studies have shown that theGeorgeville silt loam in North Carolina has the capacityto fix approximately 75 ppm of phosphorus. To test thisunder field conditions several experimental fields wereestablished in 1956 on Georgeville silt loam in ChathamCounty, North Carolina, using 150 Ib of P/acre as thebase rate and divisions or multiples of that amount onother plots. The P was added as triple superphosphate.At the first plowing, about a year following the field appli-cations, the phosphate-treated plots were moist, loose, mel-low, and easy to plow, whereas, the plots which receivedno phosphate were hard, dry, compact, and extremely diffi-cult to plow.

Preliminary invest igat ions the first year, and moredetailed investigations the second year, showed significantdifferences in certain physical properties of the soils onthese plots. The phosphated soils had a lower bulk densityand a higher moisture content throughout the growingseason than the nonphosphated ones (6).

This earlier work indicated that the phosphate effectswere more physical-chemical than biological. There is muchevidence of this in the literature (1, 2, 3, 4, 5, 7) eventhough some of it also indicates a strong biological effect(8, 10).

LUTZ AND PINTO: EFFECT OF P ON PHYSICAL PROPERTIES OF SOILS: I. MODULUS OF RUPTURE 459

Since there was such a pronounced difference in hard-ness of the phosphated and unphosphated soils in thefield, it seemed desirable to study this in the laboratory,and on soils with different dominant clay minerals.

MATERIALS AND METHODSSoil samples were collected from the B horizon of the George-

ville, White Store, and Iredell series of soils. These soils occurin the Piedmont Plateau Region of North Carolina. A brief descrip-tion of them follows.

The Georgeville is a member of the Red-Yellow Podzolic greatsoil group and is formed from weathering products of slightlymetamorphosed volcanic rocks in the Carolina Slate Belt. TheB horizon contains from 35% to 40% clay and about the sameamount of silt. The free iron oxide ranges from 4.5% to 6%.

The White Store is classified as a planosol and is developedfrom parent material derived from shales of the triassic forma-tion. The B horizon has a clay content of 50% to 60% and a freeiron oxide content of 2% to 3%.

The Iredell soil is also classified as a planosol. It is formedfrom the weathering products of basic crystalline rocks with ahigh content of albite. The B horizon usually contains a largeamount of exchangeable sodium and usually has a pH of 7 orabove. The clay content averages about 50% in the B horizonand the free iron oxide content between 3.5% and 4%.

X-ray diffraction pa t t e rns and differential thermal analysis(DTA) studies of these clays showed the following mineralogicalproperties:

Soil

Georgeville

White Store

Iredell

Samples of

Mineral species

Kaolin2:1-2:2GIbbsiteGoethiteMontmorilloniteIllite (clay mica)KaolinMontmorilloniteKaolin

these subsoils werecould be passed through a 2-mm

Amounts(relative)highmediumlowlowhigh

medium to lowmedium to low

highmedium

% freeiron

4.93

2.19

3.72

dried sufficiently so that theysieve without puddling. Dupli-

cate subsamples of 200 g each were treated with 2 sources and7 rates of P as follows: sources, phosphoric acid and monocalciumphosphate; rates, O, 50, 100, 200, 400, 800, and 1,600 ppm P.

To insure uniform distribution the 'P compounds were putin dilute solution and sprayed with an atomizer onto the soilwhich initially was spread in a thin layer but was mixed duringthe spraying. All samples of each soil were moistened to thesame moisture percentage and stored in cotton-stoppered, wide-mouth bottles at room temperature for 30 days.

At the end of the incubation period briquets were prepared bypuddling the soil, placing it in brass molds 1 by 3.5 by 7 cm,and smoothing with a spatula. The molds were placed on smallpieces of blotting paper after the inside had been coated witha thin layer of vaseline to avoid sticking of the soil. The filledmolds were transferred to a screen-bottom tray and dried in anoven for four hours at 70'C. Modulus of rupture determinationswere made according to the method described by Richards (9).

RESULTSThe effects of the phosphate on the modulus of rupture

are shown in Fig. 1 and 2. The data for the Georgevillesoil show that the modulus of rupture was decreased, com-pared to the check, by all rates of P, with the lowest rateof 50 ppm P giving the lowest average value (average ofboth sources). The reduction in modulus of rupture wassignificant at the 1% level with both sources of P. Phos-phoric acid gave a significantly greater reduction than themonocalcium phosphate. The variation in the data as afunction of rate is in general agreement with data reportedpreviously on the Georgeville soil (6).

On the White Store soil all rates of phosphorus fromboth sources produced decreases, compared to the check,

in the modulus of rupture. The greatest decrease per incre-ment of P was produced by the lowest rate of application;that is, 50 ppm P. The decreases were significant at the1% probability level. In contrast to the values on theGeorgeville soil, those on the White Store showed a morenearly consistent decrease with increasing rates of P, thequadratic regression being significant at the 5% level. Ason the Georgeville the greatest decreases were caused by thephosphoric acid, the difference between the two sourcesbeing significant at the 1 % level.

The effects of the P were most pronounced on the Ire-dell soil. Except for a 200-ppm rate of phosphatic acidthere was a cont inuous decrease in modulus of rupturewith increasing rate of phosphorus. The greatest decreaseper increment of P was caused by the lower rates. The 100-ppm rate of P from phosphoric acid produced a 35%decrease below the check. Both the linear and quadraticregressions were significant at the 1% probability level.There was a significant difference between sources, withthe phosphoric acid being more effective.

DISCUSSIONFigure 1 shows the effect of phosphoric acid on the mod-

ulus of rupture of the three soils. Several important pointsare of interest. First, the phosphoric acid was more effectivein reducing the modulus of rupture of the Iredell andWhite Store soils, which contain considerable quantities of2:1 type of clay, than of the Georgeville, which containspredominately 1:1 type of clay. Second, the smaller incre-

ppm of P (x 100)

Fig. 1—Effect of P on the modulus of rupture of 3 differentsoils using HsPCX as the source of P.

24

20

D IredellWhite Store

O Georgeville

ppm of P (X 100)

Fig. 2—Effect of P on the modulus of rupture of 3 soils usingCa(H2PO4)3 as the source of P.

460 SOIL SCIENCE SOCIETY PROCEEDINGS 1965

ments of 50 or 100 ppm P were the most effective in reduc-ing the modulus of rupture of all three soils. Third, thevalues for Iredell soil are higher at all increments of treat-ment than the White Store soil, even though both containpredominately 2:1 types of clay and both have approxi-mately the same amount of clay. The principal differencein the two clays is that the White Store contains a mediumto low amount of illite, whereas the Iredell contains noillite.

Figure 2 shows the effect of monocalcium phosphateon the modulus of rupture of the three soils. It is obviousthat the values for the Iredell were decreased considerablymore than those of the other two soils, and that the great-est decrease was caused by the 50- and 100-ppm incrementsof P. On the White Store and Georgeville soils rates above200 ppm were less effective than on the Iredell.

On all three soils the modulus of rupture decreased withincreasing rates of P. Analysis of variance showed that theregression of modulus of rupture on the log-rate of Presulted in highly significant values (1% probability level)indicating a negative relationship between the two factors,with the results being much more pronounced on the Ire-dell than on the Georgeville and White Store soils. Similarresults were obtained with both sources of P, the phos-phoric acid being much more effective than the monocal-cium phosphate on all three soils.

The use of various rates up to 1,600 ppm P makes itpossible to relate modulus of rupture to rates of P asfollows:

Y = a + b log P,

where Y is the modulus of rupture and P the rate of phos-phorus applied.

Taking the derivative the expression becomes:dY/dP = b/P

which states that the rate of change in modulus of rup-ture with the rate of phosphorus becomes smaller as Pincreases. With the White Store and Iredell where thequadratic regression was significant or highly significantthere is implied an important second term proportionalto log P/P, which also becomes smaller as P increases.

This relationship implies that increments of less than50 ppm P would be quite effective in reducing the modulusof rupture of these soils. This was indicated by earlierwork in which 50 Ib of P/acre produced significantchanges in some physical properties of the Georgeville soil(6). This becomes especially important from a practicalstandpoint since 50 and 100 Ib/acre are feasible amountsto use, both economically and agronomically.

The fact that the smaller applications are more effectiveper increment is extremely interesting from a technicalstandpoint. It is difficult to understand why such small

amounts of P per acre should produce such pronouncedchanges in the physical properties of a soil. One-hundredpounds of P per acre is exceedingly small compared to theamount of free iron in all of these soils. Even thoughdeterminations were not made of exchangeable or freealuminum, it is assumed that the aluminum content isconsiderably higher than the free iron. Thus, the amountof P becomes almost negligible as compared with the ironand aluminum. Since the phosphoric acid was much moreeffective than the monocalcium phosphate and since thephosphoric acid would allow the formation of more ironand aluminum phosphates than would the monocalciumphosphate, it is possible that the change is produced by theformation of aluminum and iron phosphates. Preliminarydata indicate a relation between the iron phosphate/alumi-num phosphate ratio and the charge of the disperse phaseof suspensions of these soils, and between charge and mod-ulus of rupture. Also, there is considerable evidence thatthe larger increments increased the negative charge of theparticles to the point that dispersion occurs and this isverified or indicated by the variation in the modulus ofrupture with increas ing increments, particularly on theGeorgeville soil. This variation has been quite pronouncedin both laboratory and field experiments (6). Previous work(6) indicates that small increments of phosphoric acid pro-duces microaggregates, probably as a result of particle link-age by the phosphate anion. This would result in a de-creased modulus of rupture. There was no evidence of anincrease in macroaggregates.