CHARACTERIZATIQN 6F $OIL AERATION DURING SPRINKLER IRRIGATION Thesis for I119 Degree of Ph. D. MICHIGAN SIATE UNIVERSITY Earl Car! SIegman 19.65 THESIS a. 2L 1.113.sz P. v - _ Michigan .5 4e ' University This is to certify that the J thesis entitled CHARACTERIZATION OF SOIL AERATION DURING SPRINKLER IRRIGATION presented by EARL CARL STEGMAN has been accepted towards fulfillment of the requirements for 21'1ng degree in Agricultural ' Engineering LQZI Major professor Date Degember 17; 1965 0-169 ABSTRACT CHARACTERIZATION OF SOIL AERATION DURING SPRINKLER IRRIGATION by Earl Carl Stegman Optimization of crop production requires maintenance of a proper balance between the many factors which affect plant growth. In this study, the feasibility of maintaining a balance between soil aeration and soil moisture content during irrigation periods was investigated. Inferences regarding the feasibility of maintaining adequate aeration during sprinkler irrigation are predicated on a knowledge of the degree of pore space saturation (St) which can be tolerated. Because of undefinable parameters, estimates of St must, at present, be made with indirect methods. The platinum electrode technique was chosen as the method for characterizing aeration with the express purpose of inferring St' For this purpose, adequate aeration 8 was defined as an oxygen diffusion rate (ODR) ; U0 X 10- g/cm2/min. ODR measurements were incorporated’with a soil moisture retention test procedure applied to thin soil samples. ODR distributions as a function of soil moisture content were obtained for eight soil fractions ranging from silica sand exhibiting a primary porosity only to a complex Earl Carl Stegman aggregated natural soil. Aeration was deemed adequate when an ODR distribution possessed both a mean and median in excess of the critical ODR. Tolerable limits of St were defined for each soil fraction. Values of St were unique for a given media. St was consistently less than the saturation in equilibrium with the air entry capillary head. With the exception of the aggregated Brookston sandy clay loam natural soil, St was empirically approximated by a parameter defined as the Brooks and Corey saturation intercept (SI)° This relation- ship suggested that in granular soils SI may be taken as a first approximation of St in the absence of ODR measure- ments. For the granular soils studied, soil moisture content was linearly related to bulk density at critical levels of ODR. Infiltration tests utilizing an irrigation sprinkler were conducted on shallow soil cores to test the feasibility of limiting profile saturations to values less than the appropriate S The theoretical basis for all tests con— t' ducted was that profile saturation approaches a limit de- fined by R=K(SL). The influence of application rate on degree of pro— file saturation was clearly demonstrated in infiltration tests conducted on Brookston 0.5—0.25 mm, Brookston < 2 mm and Hillsdale < 2 mm air dry aggregates. Profile moisture Earl Carl Stegman contents increased with sprinkler application rate. Trans- mission zone moisture contents obeyed the relationship S=CRb. In each of these soils, a feasible application rate Rt existed such that St was not exceeded in the transmission zone. High transition zone moisture contents which occurred in both the Brookston and Hillsdale < 2 mm soils appeared to be the result of structural disturbances produced by impacting water droplets. Exploratory infiltration tests were conducted on Brookston sandy clay loam and Hillsdale sandy loam soil cores obtained from the A—horizons of these soils. In each of these soils, application rates which limit profile saturation to levels less than the appropriate 8 existed. t Concurrent ODR measurements substantiated S that was t determined for each soil in independent soil moisture re- tention tests. A field infiltration test conducted in Hillsdale sandy loam soil indicated that adequate aeration could be main- tained with an appropriate application rate during nearly realistic irrigation periods for this soil Maj r Professor Approved W“ W Department Chairman CHARACTERIZATION OF SOIL AERATION DURING SPRINKLER IRRIGATION By Earl Carl Stegman A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Engineering 1965 ACKNOWLEDGMENTS The author wishes to acknowledge the professional and moral influence of Professor E. H. Kidder, who gave freely of his time and skill to the direction of the author's graduate program. In the course of this study, I have profited greatly from the counsel of Dr. A. E. Erickson (Soil Science). I am indeed grateful for his help and for his sincere interest in my progress and accomplishments. I would like to extend my thanks to Dr. F. H. Buelow (Agricultural Engineering) and to Dr. 0. Andersland (Civil Engineering) for serving on my guidance committee and for their time spent on my behalf. The most heartfelt thanks go to the writer's wife, Dorothy, whose faith that some day the requirements for the degree might be completed was of constant encouragement. ii TABLE OF CONTENTS ACKNOWLEDGMENTS LIST OF TABLES LIST OF FIGURES LIST OF APPENDICES FIGURES SYMBOLS AND ABBREVIATIONS. I. INTRODUCTION 1.1 Objectives. II. REVIEW OF LITERATURE .l .2 3 [\JRJR) Infiltration of Water Into Soils Theory of Infiltration. Measurement of Unsaturated Hydraulic Conductivity. 2.“ 2.5 2. 2 2 Mechanisms of Soil Aeration Characterization of Soil Aeration Sl Composition of Soil Atmosphere and Air Capacity of Soils .52 Diffusion of Gases thru Air Filled Pores. .53 Diffusion of Oxygen thru Moisture Films III. PRELIMINARY CONSIDERATIONS AND METHODOLOGY 3.1 3.2 Preliminary Considerations Methodology . . IV. EXPERIMENTAL . . . . . . . . . . “.1 4.2 Description of Soils Investigated Oxygen Diffusion Rate and Soil Moisture Retention Investigations. A. A. A. 4.3 21 Apparatus . . . . . . . 22 Sample Preparation and Test Procedures for Air Dry Soils 23 Sample Preparation and Test Procedures for Undisturbed Soils Oxygen Diffusion Rate and Infiltration Investigations in the Laboratory iii Page iii vii xi xii JI'UO l2 17 17 17 2O 21 27 32 3A 3A 35 35 38 143 AA A A. A. 31 Apparatus. . . 32 Preparation of Samples and Test Pro— cedures for Air Dry Soils . 33 Preparation of Samples and Test Pro- cedures for Undisturbed Soils . A.A Oxygen Diffusion Rate and Infiltration Investigations in Situ . . . . A .A1 Apparatus and Procedure V. RESULTS AND DISCUSSION 5. 5. l 2 5 5 5. 5 3 D .A Preliminary Tests Soil Aeration and Soil Moisture Relations 21 Electrode Behavior as Related to Capillary Head and Soil Moisture Content 22 Determination of Tolerable Limits of Saturation for Adequate Aeration 23 Critical Oxygen Diffusion Rate and Soil Moisture Relationships 2A Approximate Comparisons with Other Data. Sprinkler Infiltration and Oxygen iffusion Rate in Soils Initially Air Dry .3l Existence of Rt in Brookston 0.50 — 0. 25 mm Aggregates. .32 Existence of Rt in Brookston and Hillsdale ‘ 2 mm Aggregates. Sprinkler Infiltration and Oxygen Dif— fusion Rate in Natural Soils 5 5. 5. Al Existence of R in Brookston Un— disturbed Soil Cores . . A2 Existence of Rt in Hillsdale Un- disturbed Soil Cores A3 Existence of Rt in Hillsdale Sandy Loam Soil in Situ. . . . VI. CONCLUSIONS LIST OF REFERENCES APPENDIX A iv Page AA A6 A8 50 50 511 5A 71 76 80 87 88 88 93 106 106 118 128 135 137 1A5 LIST OF TABLES Table Page A.l Description of soils studied. . . . . . . . 3A A.2 Particle size distributions of soils studied . . 35 5.1 Summary of soil moisture retention measurements at the critical level of oxygen diffusion rate . . 81 5.2 Summary of initial moisture conditions for Brookston sandy clay loam cores . . . . . . 108 5.3 Summary of infiltration and oxygen diffusion data for Brookston sandy clay loam cores receiving 1.02 cm/hr application rate . . . . . . . 109 5.A Summary of infiltration and oxygen diffusion data for Brookston sandy clay loam cores receiving 0.51 cm/hr application rate . . . . . . . 110 5.5 Summary of infiltration and oxygen diffusion data for Brookston sandy clay loam cores receiving 0.25 cm/hr application rate . . . . . . . 111 5.6 End of test oxygen diffusion rates for Brookston sandy clay loam cores grouped into pore space saturation intervals. . . . . . . . . . 112 5.7 Summary of initial moisture conditions for Hillsdale sandy loam cores. . . . . . . . 119 5.8 Summary of infiltration and oxygen diffusion data for Hillsdale sandy loam cores receiving 2.0A cm/hr application rate . . . . . . . . . 120 5.9 Summary of infiltration and oxygen diffusion data for Hillsdale sandy loam cores receiving 1.02 cm/hr application rate . . . . . . . . . 121 5.10 Summary of infiltration and oxygen diffusion data for Hillsdale sandy loam cores receiving 0.51 cm/hr application rate 122 5.11 End of test oxygen diffusion rates for Hillsdale sandy loam cores grouped into pore space saturation intervals. . . . . . . . . . 123 V Table Page 5.12 Summary of infiltration and oxygen diffusion data for Hillsdale Sandy loam cores receiving 1.02 cm/hr application rate for 120 minutes. . . 129 5.13 Summary of infiltration data obtained in Hillsdale sandy loam soil in situ . . . . . 131 vi Figure LIST OF FIGURES A.l Commercially available electrode compared with wire electrode employed in this investigation Soil moisture retention - ODR test samples on a tension table showing the connection of lead wires and the use of plexigl ss plates to minimize evaporation . . . . . , . Enlarged View of a soil moisture retention - ODR test sample showing arrangement of electrodes within the sample. Sprinkler laboratory Example of soil core as prepared for an infil- tration test. Catchment can illustrates method employed to measure infiltration rate. Measurement of initial oxygen diffusion rate prior to an infiltration test on an undis- turbedsufld.core. . . . . . . . . . . Installation of cores and electrodes prior to infiltration test in situ in Hillsdale sandy loam soil . . . . . . . . . Hillsdale sandy loam soil site in situ fully prepared for infiltration test Graphic summary of moisture retention and oxygen diffusion data for silica sand; (a) capillary head vs. saturation (b) effective saturation vs. capillary head, and (c) percentage cummulative frequency distributions of oxygen diffusion rate as functions of capillary head Graphic summary of moisture retention and oxygen diffusion data for Brookston 2 - 1 mm aggregates; (a) capillary head vs. saturation; (b) effective saturation vs. capillary head, and (c) percentage cumulative frequency distri- butions of oxygen diffusion rate as functions of capillary head vii Page 36 39 39 A5 A5 51 53 53 60 Figure 5.3 Graphic summary of moisture retention and oxygen diffusion data for Brookston 1 - 0,5 mm aggregates; (a) capillary head vs. saturation, (b) effective saturation vs, capillary head and (c) percentage cumulative frequency distributions of oxygen diffusion rate as functions of capillary head 5.A Graphic summary of moisture retention and oxygen diffusion data for Brookston 0.5 - 0.25 mm aggregates; (a) capillary head vs. saturation; (b) effective saturation vs. capillary head, and (c) percentage cumulative frequency distri- butions of oxygen diffusion rate as functions of capillary head . . . . . . . . . 5.5 Graphic summary of moisture retention and oxygen diffusion data for Brookston < 2 mm aggregates; (a) capillary head vs. saturation, (b) effec— tive saturation vs. capillary head, (c) percen- tage cumulative frequency distributions of oxygen diffusion rate as functions of capillary head, and (d) percentage cumulative frequency distributions of oxygen diffusion rate as functions of capillary head in absorption 5.6 Graphic summary of moisture retention and oxygen diffusion data for Brookston undisturbed soil; (a) capillary head vs. saturation, (b) effec- tive saturation vs. capillary head, (c) percen- tage cumulative frequency distributions of oxygen diffusion rate as functions of capillary head, and (d) percentage cumulative frequency distributions of oxygen diffusion rate as functions of capillary head in absorption . 5.7 Graphic summary of moisture retention and oxygen diffusion data for Hillsda1e to the liquid content versus capillary head relationship for the media. With this relationship, the Marshall equation can be employed to estimate the saturated conductivity. Brutsaert (1963) modified Marshall‘s equation enabling its use for estimating unsaturated conductivity. Brooks and Corey (loc. cit.) found that the Burdine equation successfully predicted unsaturated and relative permeabilities for a number of materials. They also discovered that when s-s Log 12—3-3 = Log se (2.9) P was plotted as a function of Log PC/y then / p A b Se = F; for PC/Y i'Pb/Y (2.10) where 16 S = saturation S = residual saturation S = effective saturation P = capillary pressure P = air entry capillary pressure y = weight per unit volume of fluid A = pore size distribution index Equation (2.10) is recognized as an equation for a straight line on a log-log plot. The value is the intercept of this line with the abscissa. Brooks and Corey further noted that this intercept corresponded to the air entry capillary head for the materials they investigated. The slope (A) of the line was called a pore size distribution index. Small values of A were characteristic of media having a wide range of pore sizes and conversely large A indicated a media with a relatively uniform pore size. As is the case for the Marshall equation described above, the Burdine equation is considered applicable to isotropic media of stable geometry or structure. The interesting relation— ships that were obtained by Brooks and Corey from a log- 1og plot as described above appear to have some potential for furthering the characterization of porous media in a manner which is not readily discernable from a standard capillary head versus soil moisture plot and thus will no doubt be tested further. 1? 2.A Mechanisms of Soil Aeration Two distinct types of mechanisms are responsible for the interchanges of gases between the soil and the at— mosphere. These two mechanisms are: (1) mass flow resulting from total pressure gradients and (2) gaseous diffusion resulting from partial pressure gradients pro- duced by the consumption of oxygen and the evolution of carbon dioxide by root respiration and other biological processes occurring in the soil. The factors responsible for mass flow are largely meteorological in nature. Rommel (1922) studied these various factors involving mass flow; namely, the effects of temperature difference, barometric pressure, wind action, and rainfall. Rommel concluded that the combined effect of all these factors accounted for approximately 1/10 of the normal soil aeration. Thus, gaseous diffusion accounts for 90 per cent or more of the aeration process. 2.5 Characterization of Soil Aeration 2.51 Composition of Soil Atmosphere and Air Capacity of Soils In the earlier literature, the air requirements of plants were frequently investigated in terms of the air capacity or non capillary porosity of soils or in terms of so called critical oxygen concentration determinations. Wollny (1889) observed that non capillary porosity of soil surfaces protected by plant canopies from impacting 18 rain drOps was 3A to 53 per cent higher than in unprotected soil. KOpecky (1927), Yoder (1937), and Bowen and Farnsworth (19A0) proposed non capillary porosity require- ments ranging from six to ten per cent for wheat and oats to over 30 per cent for cotton. Fulton and Weber (1953) failed to find a significant correlation between soil porosity and growth of maize. Flockner et_al. (1959) in studying the effects of soil compaction found that growth of tomatoes increased with decreasing air volume in the soil until the air volume reached a certain value after which the growth again decreased. Thirty per cent air space pro— duced optimal growth. The oxygen concentration factor has proved to be a rather erratic measure of aeration. Cannon (1925) summarized experiments on the effect of relatively low oxygen concentrations on the growth of plant roots. Interesting relationships of aeration to other environmental factors were indicated. For example, when oxygen concentra- tion was three per cent, good growth was inhibited at all temperatures between 18° and 30° C. Normal growth took place at 18° C when oxygen concentration was increased to ten per cent,however, this concentration was inadequate at 30° C. Russell (1952) noted that this behavior was likely due to increased respirational demands of plants with increased temperatures. 19 Blake e£_gl. (19A9), Epstein et_al. (1957), and Wiegand and Lemon (1958) found that critical levels of oxygen con- centration observed in laboratory studies failed to exist in comparative field studies. Stolzy et_al. (1961) observed the growth of roots in tall plexiglass cylinders. The composition of the air above the cylinders was varied from one to 21 per cent oxygen. Oxygen diffusion measurements made with the platinum electrode technique served to place oxygen concen- tration measurements in their proper perspective. Visual measurements revealed that depth of root growth decreased with decreasing oxygen concentration in the atmosphere maintained above the cylinders. The oxygen diffusion rates (ODR) revealed that, for the particular plant specie studied, the roots stOpped growing at the level within the cylinders where the oxygen diffusion rates decreased to 20 x 10.8 g/cm2/min irrespective of oxygen concentrations maintained in the atmosphere above the cylinders. If the capacity or porosity factor had been included as a variable in this study it seems likely that the critical ODR would not have changed but rather this factor would have manifested itself by further altering the depth of root penetration. The factors of oxygen concentration and air capacity or non capillary porosity might thus be seen as non unique parameters but as indirectly important factors. 20 2.52 Diffusion of Gases thru Air Filled Pores. With diffusion as the generally accepted process governing soil air exchange, numerous investigations of the influence of air voids on the process of diffusion have been conducted. Much of this work deals with apparent diffusion coefficient and its relationship to porosity. Penman (19A0) for a variety of dry materials, found a linear relationship existed between the ratio D/DO and porosity. The relation— ship obtained was D/DO = 0.66 n where D = apparent diffusion coefficient Do= diffusion coefficient when n=l n = porosity VanBavel (1952) obtained a value of 0.60 rather than 0.66. Taylor (19A9), Marshal (1959), and Currie (1960), however, found the ratio D/DO to be a nonlinear function of air filled porosity in wet soils. Currie also demonstrated that the coefficient of diffusion through a porous medium is a function of both particle geometry or shape and porosity thus concluding that any attempt to find a unique D/DO versus n relationship was futile even in dry materials. He further noted that the influence of particle shape is very much modified by the presence of water which not only tends to isolate some of the air filled pores but also rounds off boundaries between particles. 21 Currie (1961) noted that there has been little prac— ‘tical application of presently developed theory for gaseous gflaase diffusion in soils largely because the appropriate bcnindary conditions, diffusion coefficients, and oxygen sink stxrength or root zone respiration rates cannot, at present, bee adequately defined for other than overly simplified root systems . 2?.53 Diffusion of Oxygen thru Moisture Films Russell (1952), in his review of soil aeration litera- tnire, was led to the observation that an evaluation of con- ditions at the interface between the root and soil system would offer the greatest possibility of establishing the influence of soil aeration on plant growth. Recognizing the potential of such an evaluation, Lemon and Erickson (1952) first employed the platinum microelectrode method for charac— terization of soil aeration. It was their contention that Electrolytic reduction of oxygen at the surface of a platinum microelectrode simulated oxygen consumption by a plant root. A principle requirement for reliable operation of the Electrode is the necessity of a complete film of moisture enveloping the electrode. Thus, if the electrode simulates a plant root, it is seen that not only must oxygen diffuse thru the air filled porosity of the root zone, but ultimately it must be diffused thru a very limiting moisture barrier 22 surrounding the rOOt. Because active roots are believed to be surrounded by moisture films (Lemon 1962), particularily when subject to conditions of high soil moisture content as commonly associated with poor aeration, the requirement of a moisture film surrounding the electrode is very likely more of an asset than a liability of the method. Lemon and Erickson's (loc. cit.) application of this method involved applying a suitable potential between the platinum electrode inserted into a soil and a non polarizable anode placed in contact with the soil. The potential applied was chosen so that only oxygen was reduced at the surface of the electrode and the oxygen flux rate was a maximum. This latter condition was achieved when all oxygen diffusing to the electrode surface was reduced. The resulting current flow was measured with a micro ammeter. The magnitude of the current was composed of two components; that arising from the reduction of oxygen and a residual current arising from electrolyte transport. VanDoren (1958), however, demonstra- ted that the residual component was negligible for soil conditions commonly encountered. Because diffusion to the electrode is unsymmetrical and not subject to exact mathematical interpretation, the diffusion current measurement is, at present, converted to oxygen flux rate (usually called oxygen diffusion rate; ODR) by applying the mathematical model for linear diffusion developed by Kolthoff and Lingane (1952), As thus derived from Fick's first law: 1t - n F A fx—o,t - n F A D 8* -1 (2.12) x—o,t where it= electrode current in amperes at time t f,_ = moles of oxygen diffusing to the electrode at x-o,t time t 3 C = oxygen concentration in moles/cm D = diffusion coefficient in cm2/sec n = number of electrons used per molecule of oxygen electrolyzed = A A = area of electrode in cm2 F = Faraday constant = 96,500 coulombs Lemon (1962), as a first approximation to what was recognized as a complex and highly dynamic system both in time and space, developed the equations describing the oxygen flux to a root model assumed to consist of a perfect cylinder surrounded by another perfect coaxial cylinder composed of a homogeneous solid-liquid matrix. The equation describing the oxygen flux rate to the exterior surface of the root also approximates oxygen flux to the electrode. The re- sulting equation for steady state diffusion per unit length of electrode was given by: 2 H De (L -C ) f = a (In re 8 l: a) (2'13) where f = oxygen flux to the electrode De: diffusion coefficient of oxygen in the solidaliquid matrix surrounding the electrode 2A C = oxygen concentration in the liquid at the air liquid interface C = oxygen concentration at the electrode surface; taken = O at steady state conditions a = radius of the electrode r = distance from center of the platinum wire to the point in the soil where de = O dre Although their data exhibited considerable scatter, Kristensen and Lemon (196A) applied equation (2.13) to oxygen flux measurements obtained in a number of specially prepared soil aggregates. A reasonably straight line re— lationship between log (re - a) and moisture content was noted in each soil. Cp was measured and found equal to oxygen concentration at the surface of water in contact with normal air. The value of De was assumed constant. Probably because of considerable scatter in the data, however, Kristensen and Lemon did not conclude that De was unaffected by changes in moisture content even though a straight line relationship was observed. Since its introduction in (1952), platinum micro- electrode measurements have been successfully correlated with plant response by numerous investigators. A partial list includes Archibald (1952), Hanks and Thorp (1956), Cline (1957), VanDoren (1958),Stolzy et_al., (1961),Finn et_al. (1961), and Letey et_al. (1961). These investigators clearly demonstrated that critical oxygen diffusion rates can be established below which optimum plant growth can not 25 persist. It has been further demonstrated, as noted earlier, that short term oxygen deficiencies are damaging to plant growth and ultimate yields. Recently the accumulation of toxic ethanol was clearly related to low oxygen diffusion rates (Fulton, 1963). For the particular crop investigated, it was also shown that ethanol began to accumulate at approximately the same oxygen diffusion rate established as critical in earlier studies (VanDoren, l958). Stolzy and Letey (1964) summarized many of the studies relating ODR to plant response. Based on the accumulated data of various investigators, they noted the following generalities con- cerning ODR and plant growth. If: ODR = 50 to 70 x 10—8 g/cm2/min ———————— required for good germination ODR > NO x lO_8 g/cmC/min ------------- optimal plant growth occurs ODR = 20 to 30 x 10—8 g/cm2/min ———————— root growth retarded ODR < 20 x lO_8 g/cm2/min —————————————— root growth stops This literature review has revealed no specific attempts to relate ODR measurements to the soil moisture status of a root zone with the express purpose of inferring a tolerable limit of saturation. Kristensen (1959) obtained ODR versus soil moisture content relationships for several sieved soil fractions. He separated each of several soil types into three different sieved fractions; namely coarse (2 — 1 mm), medium (1 - 0.5 mm) and fine (below 0.5 mm). 26 If a critical ODR were assumed, a corresponding tolerable degree of saturation might be inferred for these aggre— gates. III PRELIMINARY CONSIDERATIONS AND METHODOLOGY 3.1 Preliminary Considerations To deal with soil aeration and infiltration processes, orie is immediately faced with tremendous natural complexities; riaflmly, soil and plant root systems. Neither soil as a pubrous media or root systems.are at present subject to pre- czise mathematical descriptions. Few models of either beyond Egross simplifications are available. The literature review :repeatedly revealed that even when theoretical models are cirafted, the basic parameters required for solution of the resulting equations are often undefined or subject to in- accurate experimental determination. In the case of soil aeration, models applicable to the diffusion process require that the diffusion coefficients through both the air-solid and liquid-solid phases be defined. In addition, the oxygen requirements of the plant root systems must be quantitatively defined. Among these parameters only the diffusion co- efficient in the air—solid phase has been investigated to any extent. Very little is known about either the diffusion coefficient through the solid—liquid matrix surrounding plant roots or the specific oxygen requirements of plant roots. In View of these unknowns an approach, in terms of first principles applicable to transient state soil aeration 27 28 during the infiltration process, is impossible until more information regarding these basic parameters becomes avail- able. Because of these unknowns, an approach to this problem in the author's VieWpoint is, at present, largely dictated by the platinum electrode method for characterizing soil aeration. The ODR—plant response studies involving the electrode clearly demonstrate that this method best inte- grates the influence of elements in the root environment most likely limiting oxygen availability. These elements are illustrated by assuming as Lemon (1962) that the dif— fusion equation in the cylindrical coordinate system reasonably describes electrode behavior. In this model, ODR= f(De, C re-a). The independent effect of these p’ parameters on ODR can only be estimated at present. The oxygen diffusion coefficient (De) in the solid-liquid phase surrounding the electrode may be influenced by the matrix geometry (oft times referred to as a tortuosity factor). The degree of tortuosity may be in turn influenced by changes in pore space saturation. Cp is defined as the equilibrium oxygen concentration at the gas-liquid inter- face surrounding the electrode. Wiegand and Lemon (1958) stated that gaseous oxygen composition of the soil atmos— phere varies ordinarily within fairly narrow limits near 20 per cent. It seems safe to assume that this would be the case for most soils in need of irrigation. 29 The thickness of the moisture film (re-a) surrounding the electrode is an apparent thickness and is obviously influenced by the matrix geometry and pore space saturation. Data obtained by Kristensen (1959) indicated that the greatest influence on ODR is produced by changes in (re—a) or film thickness. So called critical ODR's have been defined for optimal growth of several plant species. Since the principle factor influencing ODR is moisture content, and in view of the objectives set forth, it is desireable to define a corres- ponding parameter St’ In this dissertation St will be taken to represent the maximum degree of pore Space saturation which can be tolerated for maintenance of adequate aeration in a soil where the gaseous phase oxygen concentration approximates 21 per cent. A critical ODR of NO x 10‘8 g/cm2/min will be assumed as the minimal rate necessary for adequate aeration. This value was selected on the basis of a summerization of ODR-plant response studies presented by Stolzy and Letey (1964). If St can be defined for a soil, the question then arises "is it feasible to limit the degree of profile saturation during the infiltration process so that St is not exceeded?" Recent theoretical and experimental evidence presented by Rubin and Steinhardt (1963 and 1964) for the rain infil- 3O tration process indicated that rain intensity, to a great extent, determined the ultimate degree of profile saturation in homogeneous and isotropic soils. The principle ascertain by Rubin and Steinhardt was that the relationship R=K(SL)l defines the saturation limit (SL) of the wetted profile when ). It seems reasonable to assume that the rate of R : K(Ssat application will also determine the degree of profile satu- ration for the infiltration process produced by sprinkler irrigation even though a sprinkler applies water in a manner which deviates considerably from the boundary condition (see literature review) describing rain infil- tration. This assumption appears justifiable in view of Youngs (1960) observation that the intermittency of the manner in which a sprinkler applies water may influence the wetted profile only during the early stage of infiltration. During this initial stage, it was noted that the profile tended to approach a saturation corresponding more to the instantaneous rate of application. However, as the wetted profile lengthened, this effect was damped out and the de4 gree of profile saturation was determined by the average rate of application (R). 1S=6/porosity; see literature review. 31 Assuming that R=K(SL) is applicable to the infiltration process produced by Sprinklers, then the desirability of defining a tolerable limit of saturation (St) for adequate aeration becomes even more apparent. For if St can be de— fined, then from a K versus S relationship one can infer the maximum sprinkler application rate (Rt) which can be tolerated. To define St’ ODR as a function of pore space satura- tion is required. Because K (S) varies sharply with small changes in S, an accurate value of St is desirable. Kunze and Kirkham (1962), for example, noted that small changes in S of one to two per cent resulted in two to five fold changes in K (S). Hence, a reliable laboratory procedure for obtaining an accurate St is indicated. Such a pro— cedure might well take the form of an appropriate in- corporation of ODR measurement with a laboratory technique commonly employed to obtain soil moisture retention curves. If such a procedure proved feasible, then the parameter St could be added to the list of parameters already being estimated from these curves. In addition, an insight into possible relationships between St and other soil moisture parameters might be obtained. Once S is defined for a given media, the corresponding t K (St) is needed so that one can infer R Unfortunately, t' the literature review revealed that a best procedure for defining K (S) is presently unresolved. In fact, it appears 32 likely that no one procedure will be applicable to all cases of unsaturated flow. Because of these present difficulties, it appeared to the author that a great deal of effort should not be directed at obtaining this specific relationship. It may be argued that a complete K versus S relationship is actually an unnecessary part of this study. Once St is established, the principle requirement then becomes one of determining whether a feasible Rt exists for the media in question. Based on present design and economic limitations an Rt 1 0.25 cm/hr will be assumed a feasible application rate. The most direct approach for evaluating the existance of a feasible Rt would be to employ an irrigation sprinkler directly. Not only is R then evaluated in the truest sense, t but if S for a given application rate R becomes reasonably l l invariant with depth(%:— —-+0) then R : K (31) as predicted l by the Rubin and Steinhardt theory. Also, in view of the limited data on profiles produced by irrigation sprinklers, their effect on the infiltration process can be further evaluated. 3.2 Methodology Based on the preliminary considerations, the following general experimental design was arrived at: l. The platinum electrode was selected as the method for characterizing soil aeration. 33 Measurements of ODR as a function of moisture content needed to establish St for a soil would, if possible, be incorporated with a procedure which could be used to also obtain the soil moisture retention curve for a soil. St would be determined for a range of materials. Sieved homogeneous2m3'1fractions would be investigated initially to arrive at an appro— priate procedure for defining S Finer soil t’ fractions and some natural field soils would be ultimately selected for study regarding the existence of a feasible Rt’ An irrigation sprinkler would be used to inves- tigate the existence of a feasible Rt' In addition ODR measurements would be made in conjunction with the infiltration tests to observe the behavior of the electrode during the in iltration process and if possible to substantiate St in an independent manner . IV EXPERIMENTAL A.l Description of Soils Investigated The soil types together with the pretest condition and fraction of each type which were investigated are listed in Table 4.1. TABLE A.l.——Description of soils studied . Pretesc Soil fraction 3011 type condition investigated Silica sand Air Dry 0.25-0.1 mm Brookston sandy clay loam Air Dry 2.0—1.0 mm H n n n n n 1.0-0.5 mm H n n n n n 0.5—0.25 mm H H H H H I! < 2 . 0 mm H " " " Natural Undisturbed Hillsdale sandy loam Air Dry < 2.0 mm " " " Natural Undisturbed The Brookston sandy clay loam soil and the Hillsdale sandy loam soil had primary particle size distributions as illustrated in Table A.2. 35 TABLE 4.2.-—Particle size distributions of soils studied.a Per cent by weight Soil type Sand Silt Clay Brookston sandy clay loam 53.4 26.2 20.4 Hillsdale sandy loam 72.8 15.4 11.8 abased on hydrometer analysis 4.2 Oxygen Diffusion Rate and Soil Moisture Retention Investigations 4.21 Apparatus Equipment made available by the Soil Science Department, Michigan State UniverSity, was used extensively to obtain the necessary experimental measurements; namely, ODR, soil moisture content, and capillary head. A tension table patterned after Leamer and Shaw (1941) and pressure plate extractors as developed by Richards (1948) were utilized to obtain the retention curves. ODR was measured with the apparatus shown in Figure 4.6. The basic circuitry for this apparatus was first developed by Lemon and Erickson (1952). A complete descrip— tion for the construction of essentially this same apparatus can be found in VanDoren (1958). In the early phases of this study, a commercially available electrode was used (see Figure 4.1). VanDoren (1958) presented a detailed description of the manufacture of these electrodes. 36 Figure 4.1.——Commercially available electrode compared with wire electrode employed in this investigation. 37 The type of electrode which was found most suitable in this study is also shown in Figure 4.1.. It was manu- factured by fusing 22 gauge copper wire to 22 gauge platinum Wire. The platinum wire was clipped so that approximately one cm remained fused to the copper. The COpper wire was then cut so that the desired electrode length was obtained. The entire electrode except for approx— imately six mm of platinum was then insulated by dipping the electrode into a test tube filled with Tygon—Series TP 21-b1ack plastic paint. The electrode was slowly with- drawn from the paint thus allowing a smooth coat of paint to adhere to the surface. The paint coating the electrode was allowed to dry for at least 24 hours. The electrodes were then inspected for faults and carefully measured and clipped so that the uninsulated platinum tip was four mm in length. The electrodes were calibrated by standardizing all electrodes to an electrode considered to have a perfectly fused joint at the c0pper-platinum junction and an exact four mm platinum tip as determined by measurement under a mag? nifying glass. The standard electrode was inserted, with the electrodes to be calibrated, into a three per cent bentonite suspension which had just been stirred for five minutes in a milk shake mixer to bring the solution into equilibrium with atmospheric oxygen. The suspension was allowed to stand for three minutes after stirring before the potential was applied. A potential of —O.65 volt as recommended by VanDoren(l958) was then applied for five minutes before readings were taken. The electrodes were considered calibrated if their readings were within : 0.2 pa of the standard electrode reading. If an electrode reading was too high, it could be reclipped to a shorter length and then recalibrated; conversely, if too low, it was dis— carded. 4.22 Sample Preparation and Test Procedures for Air Dry Soils. All of the soils listed in Table 4.1 were studied in this phase of this investigation. Except for the two un— disturbed soils, all the fractions received like preparatory treatment. The silica sand was a commercial Ottawa sand. The Brookston sandy clay loam and Hillsdale sandy loam soils were collected as disturbed A—horizon material from the Michigan State University Experiment Station Farm and air dryed in the laboratory. The soils were then sieved into aggregate fractions and stored in unsealed plastic bags. Soil moisture retention and ODR measurements were made on samples as shown in Figures 4.2 and 4.3. The soil sample containers were made by gluing a blotter paper base to each cylinder. The blotter papers were cut from the same material which was used on the tension table. The blotter paper base provided a rigid bottom when the soil samples were being made. Once the samples were on the tension Figure 4.2.—-Soil moisture retention — ODR test samples on a tension table showing the connection of lead wires and the use of plexiglass plates to minimize evaporation. I Figure 4.3.—eEnlarged view of a soil moisture retention — ODR test sample showing arrangement of electrodes within the sample. 40 table, the glue loosened readily as the blotter absorbed moisture thus eliminating any warping problems. The cylinders were cut from 8.2 to 9.1 cm diameter plexiglass tubing. The cylinders were cut to a two mm nominal height. Height measurements which were used in bulk volume calculations were obtained by averaging measurements made with an Ames Gage at eight equally spaced points around the perimeter of each cylinder. Three to five electrodes were inserted thru the wall of each cylinder in which ODR measurements were made. These electrodes were fixed in place at one cm above the bot— tom of a container before soil was poured into it. Each soil was packed by tapping the sample container until apparent settlement had ceased. The electrodes remained fixed through- out the duration of each test. To further minimize dis- turbances on the electrode environment during a test, these samples were left in place on the tension table as much as possible. In addition to the samples containing electrodes, moisture check samples without electrodes were also made. These samples were weighed at each equilibrium point following a tension change thus facilitating gravimetric determination of the moisture content. To account for the changing moisture content of the blotter paper beneath these samples with tension changes, two replicate dummy samples were also made on which the blotter weights could be 41 obtained. The accuracy of this technique was checked in the early phases of the study by removing several blotter papers from silica sand samples which could be scrapped free of material and thus compared with the check blotter samples at selected tensions. Each run on a given tension table usually involved four to six replicate samples with elec— trodes, two replicate moisture check samples, and two replicate blotter weight checks. The test procedure involved saturating the samples slowly on the tension table in order to minimize entrapment of air and to maintain a high degree of aggregate sta— bility. Tension increments considered apprOpriate to the fraction tested were utilized. At least two runs were conducted for each fraction, thus if a gap in data existed due to an inappropriate tension increment, adjustment could be made in succeeding runs. This nonuniformity in test procedure was thought necessary in order to minimize the phenomena of electrode "poisoning" sometimes noted to occur when electrodes are left embedded in soils for extended periods of time (VanDoren, 1958). To further check on possible "poisoning," the electrodes were re— calibrated following several of the tests. The procedure common to each run following the Placement of soil samples on a tension table was as follows: L42 Reduction of tension or capillary head in step— wise increments from an initial level considered appropriate to the soil being tested. Equilibrium was assumed at 24 hours following a tension change. ODR measurements and weight determinations for gravimetric measurement of soil moisture contents at each equilibrium point. For conversion to pore space saturation, a particle density of 2.65 was assumed in all soils. Removal of all samples from the tension table at zero cm tension with immediate placement in trays and subsequent flooding. The water level in the trays was slowly raised to a level just below the tops of the samples. The samples were then allowed to soak for 24 hours at which time the samples were replaced on the tension table which was set at zero cm tension. Subjecting the samples to stepwise increasing increments of tension with accompanying ODR measurements and weight determinations for gravimetric measurement of soil moisture content at each equilibrium point. Continuation of the retention test on all samples until adequate aeration was indicated by the elec- trode readings at which time the ODR part of the 43 test was discontinued. The soil moisture check samples were subjected to increasing tension increments until near residual saturation was reached or one atmosphere of capillary head was attained. It should be noted that at the limit of the tension table (40-45 cm), the samples were transferred to the pressure plate extractors. In the pressure plate extractors, equilibrium was assumed to occur in 48 hours after a stepwise increase in capillary head. 4.23 Sample Preparation and Test Procedures for Undisturbed Soils The undisturbed soils required a different sample preparation procedure. Soil cores which were representative of the A-horizon of the Brookston and Hillsdale soils were brought to the laboratory directly from the field. The details with regard to the core extraction procedure will be presented in Section 4.32. The cores were made up of two and three cm segments which were held together with masking tape. The 2-4 cm, 4-6 cm and 6—8 cm segments from two repli— cate cores of each soil type were selected as the test samples. These depth increments corresponded with the depths in which ODR measurements were made in the infil— tration tests described in Section 4.33. Five wire 44 electrodes were inserted into each of these segments before they were sliced from their respective main cores. Blotter papers were again glued to the bottom of each segment; the glue being applied along the perimeter of each sample. During the preparation of these samples, the weights of all component parts of a sample were made so that gravimetric soil moisture determinations could be obtained. The oven dry weight of the soil within each segment, however, could not be determined until the retention test was completed. After the samples were prepared, they were placed on the tension table which was set at 30 cm tension. From this point, the test procedure became identical with that des- cribed in the previous section for the disturbed soils with the exception that moisture check samples could not be used. Due to inhomogeneity of samples, it was necessary to weigh each sample at the equilibrium points for gravimetric moisture determinations. 4.3 Oxygen Diffusion Rate and Infiltration Investigations in the Laboratory 4.31 Apparatus A part circle rotary sprinkler was used throughout this phase. This sprinkler was installed in a sprinkler testing apparatus developed by Mueller (1965). The testing apparatus and laboratory are shown in Figure 4.4. The apparatus consisted principally of a sprinkler mounted in a 55 gallon barrel with a slot in the side. The sprinkler 45 Figure 4.4.-—Sprinkler laboratory. Figure 4.5.—-Example of soil core as prepared for an infiltration test. Catchment can illustrates method employed to measure infiltration rate. 46 was supplied with water from the campus mains by a centrifugal pump coupled with a surge tank and valves so that a nearly constant operating pressure could be maintained. To minimize the effect of factors other than the rate influence, the part circle sprinkler was Operated at 60 psi with a 0.32 cm (1/8 inch) nozzle for all tests. In addition all infiltration tests were made at a distance of six meters from the sprinkler. The desired application rate was obtained by adjusting the traverse arc mechanism on the sprinkler. Thus the higher the rate of application, the smaller the traverse arc and the shorter the time interval between periods in which the sprinkler supplied water to the infiltration test samples. 4.32 Preparation of Samples and Test Procedures for Air Dry Soils Infiltration tests were conducted on Brookston 0.5 — 0.25 mm, Brookston < 2 mm, and Hillsdale < 2 mm air dry fractions. Test sample containers were made up of two cm segments of plexiglass tubing 9.1 cm in diameter. The containers were put together so that planes of five elec- trodes each occurred at l, 7, 11, and 17 cm from the top of the container. The containers were 29 cm in total length. The segments were held together with masking tape. They were also well perforated to facilitate the escape of air during infiltration. Perforated plexiglass plates were bonded to the bottom of each container. 47 Soil cores for infiltration tests were prepared by pouring well mixed lots of the particular soil to be tested into these containers. The cores were packed by tapping the sides of the core until apparent settlement had ceased. The necessary weighings and height measurements were made during this procedure so that bulk density calculations could be made. Following the weighing of the core, electrode lead Wires were connected and a plastic cover was slipped over the core and secured at the top. A core ready for an infil- tration test is shown in Figure 4.5. The infiltration tests were conducted by: 1. Calibrating the sprinkler for the desired rate of application. 2. Placing the core at the six meter location and initiating the infiltration test. 3. Taking ODR readings at time intervals apprOpriate to the manner in which the wet front penetrated the sample. More frequent ODR measurements were taken with higher application rates because of the faster penetration of the wet front. 4. Removal of the soil core from the infiltration test area well before the wet front had penetrated to the bottom of the core. 5. Sectioning of the core as rapidly as possible. Spoon size samples from each cm of depth were placed in soil moisture tins with tight fitting 48 lids for gravimetric moisture determinations. Due to the small sized samples, all weighings were made to 0.1 milligram. 4.33 Preparation of Samples and Test Procedures for Undisturbed Soil Infiltration tests were conducted on the Ap-horizons of the Brookston sandy clay loam and the Hillsdale sandy loam soil. The Brookston had approximately an eight inch or 20 cm Ap—horizon whereas the Hillsdale had a ten inch or 26 cm Ap—horizon. In order that infiltration tests could be conducted on like samples, with regard to initial moisture contents, all replicate cores for each soil were obtained from the field on the same day. The core containers were built by stacking two and three cm lengths of 7.25 cm diameter tubing to 20 cm total height for the Brookston samples and to 26 cm for the Hillsdale samples. The cores were held together with masking tape. Soil cores were obtained by taping on an aluminum cutting ring having an outside diameter equal to that of the core container, and then jacking the cores into the soil. _To extract the cores, a small trench was excavated alongside the core to a depth slightly below the bottom of the cores. The cores were then freed from the bulk soil mass by additional loosening of the soil along the edges of the core. Upon removal, the cutting ring was sliced free from 49 the core and a perforated plate was bonded to the bottom end of the core. The cores were then appropriately marked for identification and sealed into plastic bags. After the cores were extracted, they were taken to the laboratory and stored at 40 °F until infiltration tests were made on them. The cores were prepared for an infiltration test in the following manner: 1. Removal of the test core from storage and sub- sequent removal of the plastic bag thus allowing the sample to be exposed to room temperature for two hours before the test. 2. Insertion of Wire electrodes thru holes in the core container provided for this purpose. Elec- trodes were inserted in planes at 3, 5, 7, and 9 cm levels as measured from the surface of the core. Five electrodes were inserted in each plane. 3. Insertion of duplicate pairs of lead wires at the 2 cm and 5 cm levels above the bottom of each core. Measurements of electrical conductivity between these leads were employed as a means of observing the passage of the wet front at the levels of these wires. 4. Protection of the core with a plastic slip cover and placement of the core at the six meter location for the infiltration test. 50 5. Measurement of initial ODR as shown in Figure 4.6. 6. Commencement of the sprinkler infiltration test. 7. Measurement of ODR at time intervals appropriate to the application rate. 8. Removal of the test core from the infiltration test area before the wet front penetration had reached the bottom of the core and immediate sectioning for gravimetric determination of the moisture profile. 4.4 Oxygen Diffusion Rate and Infiltration Investigations in Situ 4.41 Apparatus and Procedure A field infiltration test was conducted in the Hillsdale sandy loam soil in the immediate area from which the cores were extracted for laboratory investigation. The plot was part of an irrigated field on which potatoes were being grown. The sprinkler which was utilized in the laboratory was also used to make the field trial. After installation of the sprinkler in the field, one inch of water was sprinkled on the site in order to bring the soil moisture content above the level of the initial moisture contents observed in the cores which were sprinkled in the laboratory. The site was then allowed to drain for two days at which time four 26 cm cores were driven within a three foot span between two potato rows. A trench was excavated along one 51 Figure 4.6.——Measurement of initial oxygen diffusion rate prior to an infiltration test on an undisturbed soil core. 52 side of the cores facilitating installation of platinum electrodes at the five and nine cm levels, as measured from the surface, in each of the four replicate cores (see Figure 4.7). The necessary lead wires as well as leads for measurement of conductivity were then connected and installed, respectively, after which the trench was covered with boards and pplyethylene. The prepared test site is shown in Figure 4.8. An additional drainage period of two days produced initial moisture profiles very similar to those in the cores previously investigated in the laboratory as described in Section 4.32. The field test cores were sprinkled with an application rate of 0.81 cm/hr for 7.15 hours. ODR measure— ments were taken at hourly intervals throughout the test. At the conclusion of the test, two replicate cores were sectioned for gravimetric determination of the moisture profile. The remaining two cores were left in place and periodic ODR readings were continued for two days following the termination of the infiltration test. After two days, the remaining two cores were sectioned for gravimetric determination of the moisture profile. 53 Figure 4.7.--Insta11ation of cores and electrodes prior to infiltration test in situ in Hillsdale sandy loam soil. Figure 4.8.—-Hillsda1e sandy loam soil site in situ fully prepared for infiltration test. V RESULTS AND DISCUSSION 5.1 Preliminary Tests The apparatus and procedure which was described in Section 4.2 for study of ODR as a function of moisture retention was the result of considerable testing of a pre— liminary nature. A discussion of these preliminary tests is pertinent in view of present day application of the plat— inum electrode technique. The type of soil sample selected was largely arrived at thru preliminary considerations. The thin samples (2 cm deep) were employed in order to minimize the effects of moisture gradients which develop in deep samples on a tension table. The selection of the thin sample was also prompted by the desirability to minimize both the time required for equilibrium to occur between tension increments and the deviations of Cp from equilibrium with 21 per cent oxygen in the air phase. Thus, the principal requirement became that of properly applying the ODR measurement to a procedure for obtaining a soil moisture retention curve. The first tests were conducted by employing commercial electrodes (Figure 4.1) in a manner which has become what might be termed accepted procedure; i.e., by inserting the electrodes manually into the sample with subsequent with— drawal following the particular ODR measurement. Because of 54 55 the thin sample and obvious lack of support for the elec—’ trode, the electrodes were initially fitted into a clamp and then lowered manually until the electrodes penetrated to the mid-depth of the sample. The clamp rested on a set of stOps which controlled the depth of electrode penetration. Results obtained with this procedure were extremely erratic. The inconsistencies of electrode behavior appeared to result from disruption of the matrix geometry by either the large epoxy adhesive bulb just above the platinum wire electrode (Figure 4.1) or the unsteadiness of human hands during manual insertion of the electrode. In attempts to correct this procedure for these de- ficiencies, wire electrodes as illustrated in Figure 4.1 were constructed. Again, attempts were made to utilize the electrodes in a manner whereby they were inserted for each ODR measurement and subsequently withdrawn. The procedure involved fixing a series of wire electrodes into a clamp which in turn was rigidly held in place. In attempts to eliminate the unsteady manual insertion of the electrodes into a soil sample, the samples were placed upon a tray which could be raised hydraulically until the electrodes penetrated the sample to the desired depth. Results were again erratic and inconsistent from tension increment to tension increment. Not only was there again a great variation in readings from electrode 56 to electrode but also the inconsistency noted in the previous tests in which oxygen diffusion rates would many times decrease rather than increase in response to a tension incre- ment which produced a loss of water. The need for following the individual histories of each electrode involved in a particular test run had become apparent. It was reasoned that with a knowledge of individual electrode behavior from tension increment to tension increment, some rational explana- tion of its behavior might be possible. Electrodes were thus fixed into a sample as illus— trated in Figure 4.3. The description of this apparatus and attendent procedures were presented in Section 4.2. Appli- cation of this approach to a number of materials produced some interesting results, the presentation and discussion of which are presented in succeeding sections. 5.2 Soil Aeration and Soil Moisture Relations The preliminary tests produced a method for placing the electrode in a manner which produced consistent patterns of electrode behavior with respect to variation in soil moisture. These tests also made it apparent that, if inferences were to be successfully applied to a macro scale media with micro scale measurement of ODR, a cautious approach would have to be taken. A silica sand and the first three air dry fractions of Brookston sandy clay loam soil (see Table 4.1) were selected initially for study. The silica sand provided a 57 media of stable homogeneous structure whose moisture re— tention characteristics could be reproduced with a very high degree of consistency. The silica sand also offered an opportunity to observe electrode behavior in a media of single grain structure exhibiting only a primary porosity. The first three fractions of Brookston sandy clay loam were selected largely because it was not clear from the litera- ture review as to whether a feasible Rt might exist for any material. Hence, tests on rather coarse fractions were initially thought desirable. In addition, these fractions were so chosen that an approximate comparison of results might be made with results obtained by Kristensen and Lemon (1964) for similar sized aggregates. The Brookston clay loam soil exhibited a high degree of aggregate stability as well as low swelling and shrinkage characteristics thus providing a suitable media for initial investigation. All of the soil fractions listed in Table 4.1 were eventually investigated in the manner described in Section 4.2 but not in a chronological order. However, for greater clarity in the comparative analysis, the results obtained for these remaining soils, appropriate to this section, are also presented. For comparative analysis, the results obtained during desorption were summarized in a set of three to four figures per soil fraction; namely, (a) capillary head as a 58 function of saturation in desorption, (b) log-effective saturation as a function of log—capillary head, and (0) percentage cumulative frequency distributions of ODR expressed as functions of capillary head. For the Brookston and Hillsdale < 2 mm and undisturbed fractions, it was also possible to obtain ODR measurements during the absorption (decreasing capillary head) phase of the respective mOisture retention tests. Thus, for these soils, percentage cumulative frequency distributions depicting electrode behavior during absorption were plotted as Figure (d). These results follow as Figures 5.1 thru 5.8 for each soil fraction in the order of their appearance in Table 4.1. Except when stated otherwise in succeeding sections, the soil moisture retention curves were based on the mean moisture contents observed in four to six replicate moisture check samples which were divided over two to three indee pendent moisture retention runs. Likewise, the ODR dis- tributions were composed of the pooled ODR measurements observed in these runs. The log—log plots of effective saturation (Se) as a function of capillary head (h) were derived from the soil moisture retention curves in the manner recently presented by Brooks and Corey (1964). This procedure is also given in Appendix A. 59 SATURATION S CAPILLARY HEAD h- cm H,O C 20 4O 60 80 IOO . . 5 I? 1- i9 60 I" n - ' I. o I u~.5 E 20- - ‘0 ~ 0 I | 30 ”I“. g . o .c .z”” :3 o / g (40" ' D IN / p. ' I < ~ I >”50- [ a)— P - '-' 275 cm CI 360' ‘ E’ :05 V3 .1 B l \ - _ ' = 5.67 $70 E - / n eoI "’ . Sr‘ 9 ° \ (a) (b) 30' IOOF 3/ y t) 2: III 8 m E u, / 2 I- 28 ELECTRODES § c“/32 2 /° _ 3 h ODR s m 20.0 17.9 94.0 2 25. 49.9 92.5 E 27.5 64.3 91.0 3 30.0 79.7 62.0 I: M] O. l . 0 30 so 90 :20 OXYGEN DIFFUSION RATE-g cm"min" x IO ° Figure 5.l.--Graphic summary of moisture retention and oxygen diffusion data for silica sand; (a) capillary head vs. saturation (b) effective saturation vs. capil- lary head, and (0) percentage cumulative frequency distributions of oxygen diffusion rate as functions cf capillary head. CAPILLARY HEAD h-cm H‘O 60 SATURATION S OAPILLARY HEAD h-cm II,0 0 20 40 60 com . 5 I0 30 80 —1—__l—'F—_l-—7I _—§r\r l 1,1 I n ./"" : Q‘K ¢s,=88.4 I07 ’ - / ~.5 20» - . 0 m . ”I _ g ‘ - 40- ; é - I 50» t' a I- Pb 8 3.7 cm 4: . / 60- U) "05 Y I.“ " 3 70. Z_ /\ I.5I 80 I E 5,: 35.7 . (0) it (b) \ In tfi I00 A >- 90 79 $3 of 3 80 / “I g 70 36 ELECTRODES ‘é’ 80 .- '1 E5150 -_- 5'3 40 II ODR s 3.0 2.3 92.5 '3 3° 4.0 27.9 v 88.4 :5 5.0 47.4 80.5 g 20 c a: “, I I II.’ I 0 30 80 90 I20 OXYGEN DIFFUSION RATE -9 cm" min" x I0 ° Figure 5.2.n—Graphic summary of moisture retention and oxygen diffusion data for Brookston 2 - 1 mm aggregates; (b) effective saturation vs. head vs. saturation, (a) capillary capillary head, and (0) percentage cumulative frequency distributions of oxygen diffusion rate as functions of capillary head. 13 8888 E GAPILLAIY HEAD h-cm N,0 3; 25 C) CD SATURATION S 61 OAPILLARY I-IEAD II-cm H30 20 408080I00 I "W a’ I ' \' s -84 7 0’. p x 0 P o” L / a. ’85 \ r /' ’ . z . . I 9 . ,3 . . ¢ . ' i:- L :3 :J * ,2 -.05 3 ° ’ ?\ -I 52 b u 8 . I: I . u _ 9,-322 (0) (b) "g IOO O .A s” A E80 70 3 h-60 , S 38 ELECTRODES :’ g 50 ‘0 h 0" n 9 3 30 7.5 3.0 89. 5 : I0.0 25.8- 85.9 2 II. 3 40.9 82.0 §2° I2.5 58.9 77.5 an ID ° 30 80 90 I30 OXYGEN DIFFUSION RATE --9 078 min ‘2 I0‘ Figure 5.3.-«Graphic summary of moisture retention and oxygen diffusion data for Brookston l — 0.5 mm aggregates; capillary head vs. (a) saturation, (b) effective saturation vs. capillary head, and (c) percentage cumulative frequency distri- butions of oxygen diffusion rate as functions of capillary head. 62 SATURATION s OAPILLARY HEAD II-cm mo 0 0 I00 I ' 5 IE 39 9.0 I : O \ O“ 20 I- ./ : 2' a: I" ; .,5 st: 79.0 \ \ E 40L - , / g_ a: 60- r: .. 3 ‘ ‘ In 9° 5 . . 1: F- < .J \ ,. I00 «I - P 5 III: b/ ' 20.5 cm 3, '20 fi -.05 3 3 I40 y A - 2.28 ° u. u: 3 "30t3 "°’ IaI ' IbI ' .0I I00I- m {5 ziéfii—E-""'" 5’13 90- . A g ~ A/ o 80 A I! “ 70 g 36 ELEcTRDDEs '2 .l g II oTa'R s .3 3.9 . 4 .9 75 0 E30 .5 82.0 l . ' 80 0xv9EN DIFFUSION RATE -q' cm“ min" xI0 ' 0 . 50 90 I20 Figure 5.4.--Graphic summary of moisture retention and Oxygen diffusion data for Brookston 0.5 - 0.25 mm aggregates; (a) capillary head vs. saturation, (b) effective saturation vs. capillary head, and (c) percentage cumulative frequency distributions of oxygen diffusion rate as functions of capillary head. SATURATION S O 63 CAPILLARY HEAD h-cm H.O 0 4 I00 \ 800 E “xx/43 =87.5 9, 40- I" . \. I z 1 ..5 \ 80- - g 5/ p I .1: l20~ \ o l 3 "5°“ \ I -I E200. 5 P . 32 0 cm 4 " b ‘ 3240- [05 AI 3280~ /\ - 0.89 320 L Sr'60.0 rIa) . (b) .Ol IOO- . O ./ 4 90- / / ’ O o . 70- ”d// (if .0 48 ELECTRODES PERCENTAGE CUMULATIVE FREQUENCY 0- As 5 Ell“ 40- o g? h OH! 9 C) v v 30.0 I8.2 90.8 3°" / 40.0 38.2 89.0 20_ 45.0 39.3 88.0 A/A (0) I0- 0 3O 60 90 120 OXYGEN DIFFUSION RATE -9 cm"min" xIO ‘ Figure 5.5.——Graphic diffusion data for B head vs. saturation, head, (0) percentage oxyoen diffusion summary of moisture retention and oxygen rookston < 2 mm aggrerates: (a) capillary (6) effective saturation vs. capillary cumulative frequency distributions of rate as functions of carillary head. 64 40306 0000 €‘3 48 ELECTRODES PERCENTA GE CUMULATIVE FREQUENCY on C) 4° * 00R 8 14.4 95.4 3° ’ 27.3 92.: 20_ 49 3 89.5 l0 - O 30 80 90 l20 OXYGEN DIFFUSION RATE - g cm“ min" x IO ' Figure 5.5.--Graphic summary of moisture retention and oxygen diffusion data for Brookston < 2 mm aggregates; (d) per— centage cumulative frequency distributions of oxygen diffusion rate as functions of capillary head in absorption. SATURATION 8 65 CAPILLARY HEAD h-cmH.o 5O ”00 c 20 4o 80 80 mo 300 800 0 r4 1 \ «I 74/ a: 40- ‘ " r\ 5 ans \ $973.5 4) 80- ’ ' H g ‘\.‘€.‘SI : 67. 9 ‘ l20- I .: o . < ’ at, l60- ’” 5 \ .— I 3 - l °\ *200' C. b\ tr w . P =240cm \ < > - b j240- E .95 /“ °\\ Ezeo- E /\‘24° ' \\ u s 84.5 \ 320L . ' ° (a) .. (b) .OI \ IOO 5 9O 12 380~ C! E .L 70' $80» '3 30 ELECTRODES .J 50- (o D 0 a 69/ _ 84°“ /"'0 ~ h ODR S N30. ° I5.0 33.5 79.8 g 25.0 49.0 73.3 p. 2:20 ~ g Io (c) X‘.’ l ‘ O 30 60 90 I20 OXYGEN DIFFUSION RATE - g cm" min" x IO ‘ Figure 5.6.——Graphic summary of moisture retention and oxygen diffusion data for Brookston undisturbed soil; (b) effective saturation vs. (C) percentage cumulative frequency distributions of head vs. head, saturation, (a) capillary capillary oxygen diffusion rate as functidps of carillary head. 66 BIOO 2 /° 3 90 / o/0 A S m 80' II. In 70” Z :2 mi 5' 50 3O ELEOTRODES : h I 3 9° w 40’ f0 \O'OX h O'D‘R S 3 30- 73/6 2.0 25.7 80.I .. 5.0 48.0 72.9 f, 20- I00 82.9 85.I o (d) 5 IO- 0. l l l l l O 30 80 90 I20 OXYGEN DIFFUSION RATE -q cm" min" I: I0 ' Figure 5.6.--Graphic summary of moisture retention and oxygen diffusion data for Brookston undisturbed soil; (d) percentage cumulative frequency distributions of oxygen diffusion rate as functions of capillary head in absorption. 67 SATURATION S CAPILLARY HEAD II-CI'I'I H‘O CI 20 4Q 60 20 I00 10 50 I00 300 600 ‘ ' v 1 1 v r . . v—fi—H— O . ; ’N""@"Sx'87.5 " 40 I- ’ ’ ° I / ”b \\ -.5 E 80L m b . 7, ./ z I- \ ‘ I20- 3 o . < ’ 3 Bob 5 I .- 4:.J )- 200 " (D I- ‘5 u E Pb/ - 27.8 cm j 240’ 2 {.08 I E 280» E . /\ 'O.36 0 t v- , . 320 _ In _ Sr'3I.O IO) . . (DI .OI L3 1 )- g . “I (y‘ =80- 0/ s V E10 ' 2‘ 580 a 840 _ II ODR S E” 30.0 23.3 89.3 I520 40.0 34.7 87.8 g 43.0 39.4 88.3 2 IO n l 1 ' 1 1 0 30 CO 90 I20 - - I OXYGEN DIFFUSION RATE -9 cm"mm ' x IO Figure 5.7.--Graphic summary Of moisture retention and oxygen diffusion data for Hillsdale < 2 mm aggregates; (a) capillary head vs. saturation, (b) effective saturation vs. capillary head, (c) percentage Cumulative frequency distributions Of oxygen diffusion rate as functions of capillary head. 68 IOO~ . O > o :90- 63.01395 ‘0 O D (If 380- " f7 0: o “'70» [A In I ' 3.50.. 0,35 48 ELECTRODES '1 E'IAI 350- 00/9 0 _ g o I - h ODR S o40~ 0'0 2.0 9.3 9O.I u, A 5.0 34.9 87.I 030- / 0.0 4I.9 88.0 q 0 E 320’ (d) EEIOL- .A 4 / A1 1 l l 1 O 30 8O 90 I20 OXYGEN DIFFUSION RATE -g cm" min" x IO ' Figure 5.7.--Graphic summary of moisture retention and oxygen diffusion data for Hillsdale < 2 mm aggregates; (d) percentage cumulative frequency distributions of oxygen diffusion rate as functions of capillary head in absorption. '0m ":0 x —_ CAPILLARY HEAD h 69 SATURATION S CAPILLARY HEAD h-cm HgO c O 4 80 ICO I 50 I00 300 IL . 4o. 7;?" _ f” ‘3’ "5 \ so. I--d _ L... g _ 20- / ,: . . 7:" 60- a < _| ‘ ZOOL Lu ‘0 -. : P - 29.5 cm \ W . b °' E -.05 4‘ o_ g * /\ . 0.94 t I 320- “‘ . Sr‘ 35-5 (a) ... (b .Ol IOO >- f /A $90 . A A} 3 3° / I2 u. 70 [LI 2 80 '3 3O ELECTRODES 5' 50 z _ 840 h ODR S 32.5 28.7 84.5 3 30 40.0 35.8 82.I ,5 45.0 53.9 78.5 2 20 U 2 u, IO a. 1 1 0 90 I20 OXYOEN DIFFUSION RATE -9 cm“ mm" x IO ' Figure 5.8.—nGraphic summary of moisture retention and oxygen diffusion data for Hillsdale undisturbed soil; (a) capillary head vs. saturation, (b) effective saturation vs. capillary head, (c) percentage cumulative frequency distributions of oxygen diffusion rate as functions of capillary head. 70 «I a) In 5 o o o o I I I r m 0 3O ELECTRODES PERCENTAGE CUHULATIVE FREQUENCY 0| C) L __ 4° ODR S 30L 2.0 25.0 84.8 o/ 5.0 47.7 77.8 20L 9 QXA I00 56.? 69.5 / d IO'- 0 ( I f I I . I 0 3O 80 90 I20 OXYGEN DIFFUSION RATE - g cm" min" x IO ' Figure 5.8.—-Graphic summary of moisture retention and oxygen diffusion data for Hillsdale undisturbed soil; (d) percentage cumulative frequency distributions of oxygen diffusion rate as functions of capillary head in absorption. 71 5.21 Electrode Behavior as Related to Capillagy Head and Soil Moisture Despite wide variations in material investigated, the general electrode behavior was basically similar in all cases. This behavior is well illustrated by the percentage cumulative frequency distributions for each particular soil. In desorption (Figures 5.lc thru 5.8c), the electrode behavior consistently progressed thru several stages as capillary head was increased in a stepwise manner. At high saturations the ODR measurements were uniformly very low. The low ODR measurements persisted as a peaked distribution skewed right until capillary head exceeded the air entry value (Pb/T, see Appendix A) for a given media as estimated by the Brooks and Corey procedure. The values of Pb/Y are shown in Figures 5.lb thru 5.8b for each soil. As capillary head was increased beyond the air entry value, rapid changes in ODR distribution occurred, particularily in the coarser fractions. In the finer fractions, the change in distributions was also quite pro; nounced. In the finer fractions, ODR distributions tended to approach more normal distributions when the mean ODR approached adequate levels of aeration as defined by the 8 g/cm2/min. critical value in this study; namely, “0 x 10— Following this stage, increases in capillary head gradually produced a peaked distribution skewed left with a relatively high mean ODR. However, because ODR measurements were 72 discontinued during this stage of each test Figures 5.lc thru 5.8c do not, in general, clearly illustrate this latter stage. Exceptions to this general electrode behavior occurred in the silica sand (Figure 5.1c) and Brookston 2—1 mm fraction (Figure 5.20). In the case of the silica sand, the inter- mediate near normal distribution clearly did not develOp. As the capillary head was increased beyond the air entry value, individual electrode readings usually Jumped from low to high readings in steplike fashion. This effect can be noted in Figure 5.1c by the radical departure of the ODR distributions from an S shape more characteristic of a normal distribution. This electrode behavior, it appears, might well be expected in materials that exhibit desorption curves similar to Figure 5.1a. The desorption curve-for this media revealed a very narrow pore size distribution which consequently resulted in a majority of the pores draining over a very narrow range of capillary heads. In the Brookston 2-l mm fraction, the aeration study was continued until it was clear that the effect of water film rupture away from electrode surfaces was causing erroneous ODR readings. This effect can be observed in Figure 5.20 by noting, for example, that 86 per cent of the electrodes read less than 60 x 10—8 g/cm2/min at 5 cm of capillary head as compared with 82 per cent at a lesser capillary head of A cm. 73 The electrode behavior in absorption took place largely in reverse order to that for desorption. During initial stages of absorption, saturations were such that erroneous readings were common due to inadequate "wetting" of the electrode surface. As capillary head was decreased, ODR increased to a maximum followed by a gradual decrease with further reductions in capillary head thus approaching finally a peaked distribution with a very low mean ODR. This general behavior is illustrated in Figures 5.5d thru 5.8d. The general agreement with the desorption phase also indicated that the electrodes were not subject to noticeable hysteresis effects. The electrode behavior was reproducible from run to run in the sense that the mean ODR was consistently greater then the critical ODR at the same capillary head. An example of refinement of results can be noted in Figure 5.3c. The ODR distribution (involving 18 electrodes) at 11.3 cm of capillary head was obtained in an independent desorption run in the manner of Section H.22 for the express purpose of acquiring an ODR distribution at a Capillary head half- way between 10.0 and 12.5 cm. The resulting distribution at 11.3 cm is clearly different from either the 10.0 cm or the 12.5 cm distribution. Recalibration of electrodes following several Of the tests revealed that electrode "poisoning" was not a problem. The lack of poisoning can likely be attributed to the fact that electrodes were seldom embedded for more than two weeks. 74 The desorption phase of the soil moisture retention tests as applied to soil samples with embedded electrodes also provided, at least qualitatively, a very enlightening view of a basic cause for variability between electrode readings at a given capillary head. The behavior of electrodes was seen not to be haphazard as it appeared in earlier tests but rather a manifestation of the mechanism of desaturation in the immediate neighborhood of a given electrode. An enlarged view of such neighborhoods for three electrodes embedded in a Brookston 291 mm fraction is illustrated in Figure 5.9. In this figure, the varying geometries surrounding each electrode are apparent. The desaturation mechanism operative within this porous geometry is frequently modeled by employing the familiar capillary tube analogy h%:%%>. By applying this analogy to a soil moisture retention curve, a "so called” pore size distribution can be generated. By assuming in turn that these pore sizes are randomly distributed, it then follows that many different localized combinations of pore sizes can occur. Thus, a media which on a macroscale appears very homogeneous is in microscale an extremely unordered system. As a consequence, stepwise changes in capillary head produce unordered fluctuations in the film thickness (re - a) surrounding any given electrode as desorption takes place. 75 Figure 5.9.-—Enlarged View of 3 electrodes embedded in 2 — 1 mm Brookston aggregates. 76 Because the electrode has considerable size in comparison to pore sizes of soils, the true unordered de- saturation mechanism will, however, be somewhat averaged out in the electrode response. Some aspects of this likeli- hood were evident in this-investigation. In the case of the silica sand (Figure 5.10% a narrow range of pore sizes undoubtedly produced fewer widely different random combi- nations Of pore sizes. Consequently, most electrodes were exposed to drastic fluctuations in (re - a) at the same capillary head. In the complete fractions, such as the Brookston and Hillsdale K 2 mm fractions (Figures 5.50 and 5.70), which exhibited a greater range of pore sizes, individual electrodes responded in a more gradual fashion with increasing capillary heads. 5.22 Determination Of St Figures 5.10 thru 5.80 illustrate the distributions of ODR obtained as a function of capillary head for each soil investigated. For ready reference, the mean oxygen diffusion rates (05?) and pore space saturation (S) associated with each distribution are also tabulated in these figures. Due to the moderate number of observations involved in each distribution, an ODR distribution was considered indicative Of adequate aeration if both its mean and median 8 value exceeded the critical ODR, i.e., NO x 10- g/cm2/min. Thus, S was defined for each soil fraction by noting first t 77 of all the ODR distribution which Just met the above require- ment for adequacy of aeration. Having established this distribution, the associated capillary head was then applied to the appropriate soil moisture retention curve [Figure (a) for each soil] for the corresponding saturation, or St' A more formal statistical analysis of this data was not considered. ‘The nature of the testing procedure was such that all treatments (capillary heads) were applied to a given sample, thus violating an assumption basic to anal- ysis of variance procedures; namely, independance of experi- mental errors. Whether this fact should have been given more consideration in the experimental design may be a matter for some conjecture. In the author's vieWpOint, this design was not worth persuing in view of the prohibitive number of samples which would have been required. The principle concern for applying the mean and median values as suggested above, was rather, that an adequate number of electrodes be employed. In this regard, VanDoren (1958) noted in natural soils that the standard error Of the mean was reduced little as the number of Observations exceeded 40. This criteria was met to within a few electrodes for each soil investigated. Moreover, with the exception Of the two undisturbed soils, the soils investigated in this study were far more homogeneous then natural soils. Also, because there was general agreement in electrode behavior from run to run,it appears reasonable to conclude that an 78 adequate number of electrodes was employed in all tests of this section (see Figures 5.10 thru 5.80). Further analysis of the application of the mean and median criteria showed that the mean ODR was a more conserv— ative criteria than the median in all cases except for the silica sand. This comparative behavior was not unexpected. The ODR distributions for the silica sand (Figure 5.10) exhibited a definite bi-modal character. Hence the median, being less influenced by extreme values, may be a more conservative indicator of adequate aeration in soils of this nature. The accuracy with which St can be defined clearly depends also on other factors even if an adequate number of electrodes are employed. A principal factor to be noted in Figures 5.10 thru 5.80 is the non linear electrode response to changes in capillary head and resultant moisture changes. Thus, to avoid excessive interpolation, a scheme of capillary head incrementation, appropriate to each soil in question was employed in the region where mean and median values of an ODR distribution approached the criteria for adequate aeration. This scheme, in general, was not difficult to achieve. As capillary head was grad- ually increased in steplike increments, some electrodes in a rather random manner, would in each test, begin to exhibit sharp increases in ODR. These electrodes thus signaled the approach of the region of capillary heads in which aeration would become adequate. The nature of the soil moisture retention test and the resulting precision of the moisture retention curve affect the accuracy with which St can be defined. In the air dry soils, soil moisture retention meaSurements were reproducible to within : 2 per cent of the mean values as plotted in Figures 5.1a thru 5.5a and 5.7a. Some swelling in absorption and shrinkage in desorption effects were noticeable particularily in the Brockston 0.5 — 0.25 mm and < 2 mm fractions but were treated as negligible in that sample volumes and bulk densities were based on original measurements throughout each test. The moisture retention measurements on the two natural soils (see ranges indicated with each mean in figures 5.6a and 5.8a) showed more variation as was expected. These measurements were conducted after a considerable number of the sprinkling tests had been completed. Because the ODR measurements were in agreement with the sprinkler infiltra~ tion results, these tests were not further refined with additional replication. Effects of evaporation were minimized by maintaining free water in large pans within the reasonably vapor tight chamber containing the tension tables. In addition, the soil samples were covered with plexiglass plates as shown in Figure 4.2. During the tests of this investigative phase, tempera— ture measurements were recorded at the time of each ODR 80 measurement. The average of all temperature measurements was 74.5 °F. Most fluctuations were within : 2 °F of this value. All extreme values were within 1 4 °F of the mean. Because of the relatively small temperature fluc— tuations and the fact that ODR is influenced by approximately one per cent per °F (VanDoren, 1958) it is doubtful that temperature adjustments, if applied to the ODR readings, would alter the determinations of St' 5.23 Critical ODR and Soil Moisture Relationships Values Of St were determined for each soil as outlined in Section 5.22 by noting the capillary head (ht) and corresponding (St) of the ODR distribution Just satisfying the requirements for adequate aeration. Variations in ODR between successive increments of capillary head were non linear, thus interpolation was avoided with the exception of the Brookston 2 - 1 mm fraction. Because of the erro— neous electrode readings which were observed at h = 5.0 cm and the large change in pore space saturation between 4.0 and 5.0 cm, an interpolated capillary head (h = u.5 cm) was selected in this case. In all other soils investigated, capillary head increments were more nearly appropriate to the soil in question. The observed values of St and associated parameters for each soil investigated are presented in Table 5.1. The t; (2) (3) mean sample bulk associated parameters include (1) capillary head, h gravimetric soil moisture content, Wt; .udoosoucfi cofiumLSuwm honoo new mxoosm o .< xfipcodo< monsoooono hence can wxooaz on» an poumEHumo no new: mamaafidmo spasm Lfiwm .zufimcmo xasn deEmm some mm.m a s p a Aocaospm mafiow Ham Lou—A cm I HV\mm 3 u m ouozv m on ucoucoo oszuwwoe OHLoOEH>mem .um no use: >LmHHHQwom .cofiumsmm wpmzqoom NCHuSaELc; :oflpmsspmm do uHEflH omnwsoHOp EssemeH .mm.m Cofipoom mom ”moot m>wm3~ocooCHm 81 a.mm o.m o.s~ o.Hm n.am som.H ma.ma o.ma m.ms statesmaacs = = = a.sn o.oH o.om m.sm m.sm ems.s mm.em o.ma m.om as m v sacs macaw maaemaaa: 0.3m o.m m.~s a a oam.a ma.mm o.ma m.ms captapaaec: = E = = ©.Nm c.0H ©.mw m.~w o.mm :rm.a mm.om o.m: o.mw EE m v = = : : 0.05 m.Hm mmc.a mm.:: o.mm o.m> ES mm.o I m.o = = : : s.:w m.m .mfim.o em.wm m.HH o.mm ea m.o I H = = = = :.mm >.m :Hm.o mm.om m.: 0.0m EE H I m Eon mmHo hpcmm couwaOLm o.Hm m.sm ma=.H mo.om m.sm o.Hm as .H.o I mm.o scam meadow a So u a 50 mEo\w u 50 a p: u: an Hm »\nm Im a: on am eaaaeaumosca was» aaom m m H o m a m m H cofipomsm coauosomn< :Ofipqsommn .Oums cofimsuufio :szxo «0 HO>OH Hmofipaso on» we mucosossmmoe cofiucmuop waspmfios aaom ho szEEsmII.H.m mqmAQ I 20 30 40 50 60 MOISTURE CONTENT w -- 9/g Figure 5.ll.—~Relationship between bulk density and the aeration limiting gravimetric.moisture content for the soils of Table 5.1. 85 compared with the other soils in that Stwas considerably greater than S The cause for this deviation cannot be I' completely resolved in view of the limited data. However, it should be noted that in order to Obtain a straight line thru the effective saturations occurring below SI in Figure 5.6b, Sr had to be manipulated in a drastic manner (See Appendix A) such that the effective saturation at 340 cm of capillary head was excluded from consideration. The resulting steepness of this branch of the curve, as re— flected in the pore size distribution index (A = 2.40), appeared unrealistic when compared with the Brookston < 2 mm aggregate fraction. Moreover, when Sr was manipulated to Sr = 52.2 per cent so that the effective saturation at 340 cm was included, a straight line fitted all points reasonably well with the exclusion of the effective saturations at 30 and 40 cm. This straight line had a considerably lesser s10pe and intersected the upper branch at SI = 73.2 per cent as compared with S = 73.5 per cent. t It might be speculated that the effective saturations at 30 and 40 cm are not consistent with the remainder of the data for this soil. Whether this inconsistency was due to the nature of the soil, which was in this case considerably structured, or due to experimental error can only be deter- mined by further experimentation. Because of this incon— clusiveness, the Brookston natural soil results are not included in Figure 5.10. 86 Figure 5.10 illustrates an empirical relation. By comparison with a 1:1 relationship, this relation can be interpreted as follows. Soil aeration did not become adequate until pore Space saturation was reduced to a level below SI' For the silica sand investigated, S I' In the other soils possesSing SI's which decreased from this point, the deviation between t equalled S St and SI increased in a linear fashion as SI decreased. Because the lower values of SI tend to indicate a soil possessing a greater range of pore sizes, it appears rea- sonable that S should bear this relationship to S This t I' trend clearly warrants further investigation. Further testing of this relationship between S and SI’ particularily t in homogeneous granular soils which can be expected to have a rapid draindown rate following irrigation, may show that in the absence of a measurement of St one might use SI as a first approximation of St' The parameter S tended to be a more stable point for I a given soil as compared with the air entry head (Pb/Y) and pore size distribution index (A). These latter parameters, which were recently defined by Brooks and Corey (1964) as hydraulic constants for a media, were noted to be somewhat dependent on the manner in which the residual saturation (Sr) was manipulated. 0n the other hand, manipulation of Sr affected the upper branch of the log—effective saturation versus log-capillary head plot very little. The saturation 87 intercept (SI) between the two branches remained nearly fixed and thus was subJect to a lesser Judgement influence as were both Pb/Y and A. This attribute appears to make SI a desirable parameter for correlation with St' A second empiric relation obtained from Table 5.1 is illustrated in Figure 5.11. Mean bulk densities of the moisture check samples involved in each soil moisture res tention curve of Figures 5.1a thru 5.8a were compared with the corresponding aeration limiting gravimetric moisture content (wt). Figure 5.11 indicates that bulk density and gravimetric moisture content may be linearly related at or near the defined critical ODR. Due to the limited data on which the relationship of Figure 5.11 is based, a critical ODR region of i 2 standard errors of estimate (2 SEy = i 0.08) is suggested. In addition, it must be stated that this type of relationship requires further verification over a broader range of soils. In the absence of these tests, it may be conJectured that this relationship is more feasible in granular soils. It is easily seen that bulk densities on a macro scale may not be very representative of the indi- vidual clods comprising much of the soil mass in well structured soils. 5.24 Approximate Comparisons with Other Data An added measure of confidence in the methods employed in this study for obtaining S was attained by a comparison t 88 of the St values for the Brookston 2 — 1 mm, 1 — 0.5 mm, and 0.5 — 0.25 mm fractions with data published by Kristensen and Lemon (1964). For similar sizes of aggregates in a Honyoye loam, it was noted that at 40 x 10.8 g/cm2/min the range Of S for these aggregates was approximately 75 - 91 t per cent pore Space saturation as compared with a 75 - 86 per cent range for the Brookston aggregates. It was very interesting to note a favorable comparison between a criteria for adequate aeration based on diffusion of oxygen thru the air phase of moist porous media as developed by Taylor (1949) and the relationship of Figure 5.11. Taylor noted for a quartz sand (0.5 — 0.25 mm) with ED = 1.45 and an Ontario loam with ED = 1.30 that aeration became limiting at 22.5 and 30 per cent moisture content, respectively. These values, when plotted in Figure 5.11, fall within the critical region of aeration suggested by the data of this study. 5.3 Sprinkler Infiltration and ODR in Soils Initially_gir Dry ‘5.31 Existence Of Rt in Brookston 0.5 - 0.25 mm Aggregates Infiltration tests in the manner of Section 5.32 were initially conducted on the Brookston 0.5 - 0.25 mm air dry aggregates. The selection of-this aggregate fraction was prompted largely by the desirability of beginning this experimental phase in a very stable homogeneous media of relatively coarse texture in which a feasible Rt would be expected to exist. 89 Water content profile data obtained for application rates of 0.36, 1.24 and 2.87 cm/hr are presented in Figure 5.12. The results were obtained from single trials on replicate cores of 1.043 mean bulk density. The wet fronts terminated in a distinct plane well above the bottom of each core but were not sampled. Thus the profiles under consideration were terminated at the last wetted depth at which measurements were made. The deviations within each profile were very likely due to either small inhomogeneities within each core or drainage effects during sectioning of each core for gravimetric determination of the moisture profile. The resulting profiles produced by these three appli- cation rates clearly demonstrate that the intensity of sprinkler application rate affected the degree of profile saturation. The development of near vertical moisture profiles indicated that the flux rates within these portions may be defined by R = K(SL) as postulated by Rubin and Steinhardt (1963). Effects which might be attributed to the intermittency of sprinkler application rate were not evident in the profiles for this soil. By averaging the moisture contents in the near vertical portions Of each profile (3 cm to 10 cm) it appears reasonable to assume that the application rates of 0.36, 1.24, and 2.87 cm/hr correspond to unsaturated conductivities at 50.4, 55.6, and 60.1 per cent pore space saturation, 90 SATURATION S OlOZOSO‘IOSOSOTOSO Q 2' .7 I E \. \ 2"- < I :6‘ I IL \ E \ ! 3‘” . II .x‘ I! l I I2- 1 I II 0.30 cm/I'Ir / I 12.87 CM/I‘Ir I.25 cm/ hr 3 .Figure 5.12.-—Influence Of rate Of water application upon soil water content profiles in Brookston 0.5 - 0.25 mm aggregates initially air dry at various sprinkler application rates. 91 reSpectively. When these corresponding values were plotted on a log—log basis, a straight line relationship (S = CRb) was indicated (Figure 5.13). Substantiation for such a relationship can be found in the literature. The author noted that when the rain infiltration data for Rehovot sand, as presented by Rubin and Steinhardt(l964), was plotted in the above manner, a straight line relationship was un— questionably obtained for application rates producing pore space saturations ranging from 28 to 70 per cent. Gardner (1956) obtained this same type of relationship for three different soils when capillary conductivities obtained by the pressure plate outflow method were plotted as a function of volumetric moisture content. Recalling that an S of 75 per cent (Table 5.1) was t obtained from the ODR—soil moisture retention tests, it was seen from Figure 5.13 that none of the application rates produced profile saturations which came close to exceeding this value. By extrapolation, a rate in excess of 25 cm/hr would likely be required to produce a profile in which pore space saturation exceeded 75 per cent. This high application rate far exceeded the capacity of the experimental apparatus thus preventing a test of this extrapolation. The obJectives, however, were met by the rates employed in that the exist- ence of a feasible Rt was demonstrated. ODR measurements were attempted in the manner of Section 4.32 at 1, 7, and 11 cm levels within each core. In View of the experience gained in observing electrode 92 I00 50 SATUATION S lo I 1 1111111 I [Lllllll 1 1 11 1__ OJ I.O l0.0 I00. APPLICATION RATE R - cm NR" Figure 5.13.~—Relationship between sprinkler application rate and transmission zone water content in Brookston 0.5 — 0.25 mm aggregates. 93 responses during the ODR-soil moisture retention tests, the electrode behavior in these tests was Judged as charac- teristic of behavior produced by incomplete moisture films on the electrode surfaces. The resulting ODR measurements were thus typical of readings obtained in the initial stages of the absorption phase during soil moisture retention tests. As an example, the end of test mean ODR at the 7 cm depth was 22.9, 32.9 and 43.2 x 10—8 g/cm2/min in response to saturations produced by 0.36, 1.24 and 2.87 cm/hr appli- cation rates, respectively. These electrode readings, while quantitatively erroneous, are usually incurred only in soil environments where aeration is obviously adequate (Kristensen and Lemon 1964). Hence, these ODR measurements substantiate the existence of a feasible Rt in this soil. 5.32 Existence of Rt in Brookston and Hillsdale < 2 mm Aggregates Following the discovery that adequate aeration could be maintained rather easily in coarse textured media such as the Brookston 0.5 — 0.25 mm fraction, the complete air dry fractions for the Brookston sandy clay loam and Hillsdale sandy loam soils passing through a 2 mm sieve were selected for study. The selection of these two media was based on the desireability of continuing investigations in a homo- geneous material of a more realistic yet reproducible 94 nature so that infiltration profiles might be subject to explanation on the basis of the rather restrictive Rubin and Steinhardt theory (1963). Water content profile data obtained for application rates of 0.23, 0.89, and 2.03 cm/hr into cores of Brookston < 2 mm aggregates are presented in Figure 5.14. Likewise, profile data for rates of 0.23, 0.91, and 2.29 cm/hr into cores of Hillsdale < 2 mm aggregates are presented in Figure 5.15. The Brookston and Hillsdale cores possessed mean bulk densities of 1.296 and 1.449, reSpectively. The profile data in Figures 5.14 and 5.15 were obtained from single trials at rates which were duplicates of rates applied to an initial set of shorter cores. Since there was little deviation in transmission zones between replicate cores receiving identical application rates, it was thought more descriptive to plot the longer profiles such that the wetting zones could be exhibited as part of each profile. The moisture contents of the wet front, however, could not be accurately sampled and thus are not indicated in Figures 5.14 and 5.15. The irregularities within each profile were very likely due to combinations of sampling error, slight inhomogeneities within samples, and drainage effects during the three to four minutes required for sectioning of each core, thus smooth curves were drawn through the data for each application rate. 95 SATURATION S 30 4O 50 60 7O 80 90 I00 I 1 .5 I04 2" CHI IZI SOIL DEPTH I> 0.23 cm/hr 20 T 22*- 0.89~ cm/hr/O/ 24 L , 2.03 cm/hr Figure 5.14.-—Inf1uence of rate of water application upon soil water content profiles in Brookston < 2 mm aggregates initially air dry at various sprinkler application rates. 96 SATURATION S 30 4O 50 60 70 80 90 I00 / 2. .. f ._ 7‘ . BL o I2 I4- A/ if] 20 O / ' 22 _ A (2122.29 CID/hf 0.23 cm/hr Z-— CHI 6 D\ D 0-0 SOIL DEPTH I 24- ' 0 0.9! cm/hr Figure 5.15.--Inf1uence of rate of water application upon soil water content profiles in Hillsdale < 2 mm aggregates initially air dry at various sprinkler application rates. 97 The general characteristics of these profiles with regard to application rate are clearly similar to those obtained for the Brookston 0.5 — 0.25 mm aggregates. In particular, profile moisture contents increased with increases in application rate. The near vertical portions of each profile indicate that the saturation limit <33== > associated with each application rate was being approached within each transmission zone. Consistently observed deviations which were not clearly observed in the coarser Brookston 0.5 — 0.25 mm fraction took the form of transition zones similar to results of Bodman and Colman (see literature review). These uppermost zones of higher moisture content exhibiting a steep moisture gradient to the transmission zone were also noted by Rubin and Steinhardt (1964) as experimental realities but not predicted by their theoretical consid— erations. While there may be some other contributing factors, the observed transition zones in this set of results were believed to be principally due to effects of droplet kinetic energy incident upon the surface of the soil. This behavior was somewhat expected because protective layers were not utilized to completely remove the kinetic energy from the impacting water drOplets, but measures were taken to insure that the soils were sprinkled with a realistically ideal sprinkler system. Cores were located at a point in the sprinkler traJectory where drop sizes were small. In 98 addition, a high operating pressure (60 psi) typical of that which might be used under realistic conditions was utilized. Based on work by Mueller (1965), a 0.23 cm (1/8 inch) sprinkler nozzle operating at 60 psi produces mean drop diameters of 1.50 mm and a mean drop energy of 160.8 ergs at 6 meters from the sprinkler. For further breakup of drOplets a window screen (Figure 4.6) was placed at 0.5 cm above each core. Nevertheless, impact of water droplets upon the surface of these two soils undoubtedly caused slaking and reorientation of soil particles. Cursory observations of the soil surfaces following sprinkling tests indicated a subsidence of 2 to 3 mm had usually taken place. As a consequence, it might be theorized that the intrinsic per— meability near the surface of these soils became a transient function in response to matrix changes produced by impacting water droplets. Hence, hydraulic conductivity in the transition zone may be not only a function of moisture content but also of time in a manner dictated by the effects of droplet energy on the soil aggregates near the surface. As already noted, the transmission zones for the Brookston 0.5 — 0.25 mm aggregates extended to the surface (Figure 5.13). Since this soil fraction was devoid of the finer particles contained by the Brookston < 2 mm fraction, it appears that this behavior may be taken as a qualitative substantiation of the above argument for reduction in 99 intrinsic permeability near the soil surface. In the case of Brookston and Hillsdale < 2 mm fractions this reduction was sufficient to cause ponding near termination of the infiltration tests when application rates of 2.03 and 2.29 cm/hr, respectively, were applied to the 28 cm cores (Figures 5.14 and 5.15). The trends of profiles produced by 1.52 and 1.78 cm/hr application rates (not shown in Figures 5.14 and 5.15 but included in Figures 5.16 and 5.17) for the Brookston and Hillsdale soils, respectively, also indicated that ponding may have eventually occurred in deeper cores. Thus, with St equal to 88.0 and 86.5 per cent for the Brookston and Hillsdale < 2 mm fractions, respectively, application rates producing runoff conditions could be employed before the corresponding transmission zone profile water contents exceeded these values. The value of St may be diminished as an indicator of adequate aeration in soils where this phenomena takes place, particularily if Rt is also small. Prolonged irrigation at rates which nearly saturate the surface may produce inadequate aeration resulting from insufficient diffusion of oxygen from the free atmosphere to the root zone. This possibility prompts an awareness of a completely different approach towards achieving periods of minimum oxygen difficiency. It is conceivable, in soils where Rt is very low, that the total period of oxygen deficiency during irrigation and the moisture redistribution period following irrigation to levels below St may be greater than periods 100 of deficiency produced by a high application rate and its associated period of moisture redistribution. This approach represents a drastic departure from the obJectives of this study and was not further pursued. The utility of the sprinkling technique for determining Rt can again be demonstrated by plotting the pore space saturation from the region within each transmission zone most nearly satisfying the relationship (g; = 0) as a function of corresponding application rate. Figures 5.16 and 5.17 for the Brookston and Hillsdale soils, respectively demonstrate that straight line log-log functional relation- ships :m.mm mam.a : m mm.:m mm.mm mao.H m o OH x CHE\ Eo\w & & Eo\m 80 -w m m nmmo mm m3 mam N< comp QOfimSEMHo COHpQLSpmm pompcoo zpfimcoo pCmEOpocfl cmmzxo momdm whom oncomfioz gazm comma y comp COfimeHHQdm pn\Eo mo.a wcfi>fioomm mmpoo smoa mmao mocmm Cowmxoomm Lou mumo COHmSMMHc cmwmxo new coapmanHMCfi mo anoEESmII.m.m mqmfimomp momoo.Emoa zwao mocwm zoomxoopm Log mpmo COfimSMMHU cowmxo ocm COHmepHHmcH mo mnwEESmII.:.m mqmHmomp momoo Emoa zmfio H mocmm coumxoopm mom memo scamsmufio cowzxo com COprppHHmcfi mo mmmEEsmII.m.m mqmLcmoo mcoz mocmg consamao crmhxo mop gown: cfl pcosmpocfi Spamw map 00 Scopcoo obshmfloe may he 0ocHELo0m© mos mcoco coco mo :oflpwmoa one v m . 0: m..C.®®E q.mm m.~m m.mm n.0m H.0m 0.0m 0.mm 1.: m.mm \.Nm m.0m 0.0m mnm: m.mm m.m: :.Hm m.:m 0.0m :.Hm 0.:H m.0m 0.mm 0.mm 0.0m 0.H: m.m: m.am 0.0a 0.0m 0.0m H.0m H.Hm 0.mm 3.0: m.0: :.Hm 0.Hm 0.mH m.mm H.0m 0.0m 0.2: 0.Hm m.mm m.0m m.0m 5.0m 0.H: 4.0m 0.mm H.mz 0.Hm m.m: H.0m H.0m m.mm 0.0m m.m: m.mm 0.m: m.ma 3.0: H.3m m.0m H.:: 0.00 00H x HIGHS mIEo m ommpmh scamsmwac cmw0xo 0.00 A 0.00 I H.mm 0.m> I H.05 0.05 I H.m0 0.m0 I H.00 0.00 v name you CH mam>mcpr coflmeSPmm mowqm whom mmam>pmpcfi :oHpMLSpwm mowdm whoa coca consonw mmmoo EMOH zmao 002mm COpmxoopm mom mmpmp coamSMMHo cmwmxo pmmu no cchI.0.m mqm¢e 113 end of test mean ODR's, for each set of 5 electrodes embedded at 3, 5, 7, and 9 cm, were associated with the moisture contents of the 2—4, 4-6, 6-8, and 8—11 cm core increments, respectively. These means were sorted into pore space saturation categories irrespective of application rate as illustrated in Table 5.6. Class intervals of five per cent saturation were chosen so that the VanDoren criteria (Section 5.22) for an adequate number of electrode readings could be met in most of the intervals. Comparison of mean ODR's between class intervals reveals a definite decreasing trend with increasing saturation. It should be noted that the Sprinkler infiltration tests were conducted with water from the campus mains which possessed a temperature near 55°F during these tests. Temperature measurements within the cores were such that an approximate 15 per cent increase should be applied to the means of Table 5.5 if comparisons with the ODR results obtained from the soil moisture retention tests are to be made at comparible temperatures. Recalling that St of 72.5 per cent was obtained from the soil moisture retention tests, it was seen that this value falls within the category (70.1—75.0) of Table 5.6 which possesses a mean of 36.2 x 8 10’ g/cm2/min. If a 15 per cent increase for temperature 8 g/cm2/min to difference is applied, thus raising 36.2 x 10— 42.5 x 10.8 g/cmZ/min, a very favorable comparison with the soil moisture retention results of Section 5.2 is indicated. This agreement between two entirely different techniques added 114 a measure of confidence to the value of St and was taken as evidence that St may be predicted by incorporating ODR measurements with soil moisture retention procedures. Comparing S = 72.5 per cent with the profiles t (Tables 5.2 thru 5.4) produced by 1.02, 0.51, and 0.25 cm/hr application rates, respectively, it can be noted that only the profile produced by the 1.02 cm/hn rate exceeded this saturation to a severe degree. The corresponding mean ODR's in the region of measurement also are indicative of inadequate aeration in this profile. The true effect of application rate on pore space saturation was, however, somewhat masked by the tendency to allow the wet fronts to reach the bottom of the cores before termination of the infiltration tests (compare the moisture profiles of Tables 5.3 thru 5.5 with Table 5.2). This tendency was in part due to irregular behavior of the elec- trical conductivity measurements which were employed to measure the wet front progress within each core and in part due to the desireability for observing the electrode behavior as long as possible at each application rate. Since St was not exceeded and aeration was very nearly adequate at the conclusion of the infiltration tests at 0.51 and 0.25 cm/hr application rates, one might conclude that either rate could be employed to irrigate this soil to a depth of 20 cm without severely restricting aeration. The tendency to allow the wet front to reach the bottom of the core can only affect this conclusion in a conservative manner. 115 Electrode behavior in terms of mean ODR associated with each application rate is illustrated in Figure 5.20 as a function of time at the four sampling depths after initiation of infiltration. Because this soil was moist, initial readings could be obtained. The general electrode behavior was most interesting and of the type which was originally also expected in the air dry soils. The behavior in these natural soil cores appeared to have a logical sequence. Initially, the electrodes were inadequately wetted and thus exhibited an initial phase of increasing ODR until the moisture films (re - a) apparently began to limit diffusion of oxygen. Following this initial phase, the electrode readings appeared characteristic of a response to the degree to which moisture films were building up around the electrodes. As evidenced by the sharp decline in ODR after 75 minutes, (Figure 5.20),the 1.02 cm/hr application rate produced a degree of profile saturation which severely limited ODR as compared with the lesser rates of 0.51 and 0.25 cm/hr, respectively. A review of the electrical conductivity data has led the author to conclude that the wet front had not reached the bottom in any of the cores receiving 1.02 cm/hr appli— cation rates in less than 90 minutes. Hence, the ODR measurements at 75 minutes (Figure 5.20) were representative of this rate at a time when the moisture profiles were not affected by the barrier imposed by the core bottom. Because OXYGEN DIFFUSION RATE ODR - g/cm‘lmin x IO‘ 116 4a 42 C \°\A APPLIGATION RATE cm/hr \A - - -l.02 as 3 cm DEPTH 0‘0!“ fl A—O.25 48" . /-‘"'6 :— ———- 76/ \ N\A 42- 365/ \A 30- \. 5 CNIDEPTH 4m / 36« *0 A 30- 24 ° 9 cm DEPTH \ _|_§ L 1 L 1 .j 0 60 IZO ISO 240 300 TUNE -'Ifl”l FiEUre 5.90.~—0xygen diffusion rate vs. time at four samplinw depths after initiation of infiltration tests in Brookston sandy clay loam undisturbed soil cores. 117 the oxygen diffusion rates at 75 minutes were already below critical levels, this rate was indeed producing aeration limitinglsaturations. A further analysis of the results in Tables 5.2 thru 5.4 reveals that the effect of the 1.02 cm/hr application rate on profile saturation and consequent electrode behavior (Figure 5.20) was magnified by the fact that a higher mean bulk density existed in the cores receiving this rate. This inhomogeneity is reflected by a mean B0 of 1.443 in the 2—11 cm depth range of the 5 cores which received 1.02 om/hr as compared with mean bulk densities of 1.370 and 1.379 for the corresponding increments of cores receiving 0.51 and 0.25 cm/hr. This inhomogeneity between experimental units occurred despite a random selection procedure which was employed to select the five cores for this treatment from the original 20 cores. It thus appears reasonable to assume that this five per cent difference was within expected limits of variation for a natural scil. The probable influence of this higher BD on moisture profiles associated with the 0.51 om/hr application rate can be noted in Table 5.3 at the 8—11 cm depth increment. For a mean 80 of 1.449, the mean ODR was 36.4 x 10"8 g/cm2/min at 72.67 per cent saturation. Consequently, there remains good reason to state that an application rate of 0.51 cm/hr could be tolerated for infiltration to a depth of 20 cm. 118 5.42 Existence of Rt in Hillsdale Undisturbed Soil Cores ‘ A series of tests completely analagous to those of Section 5.41 were conducted on Hillsdale sandy loam undisturbed soil cores. This soil had a deeper A—horizon (26 cm) and was definitely less structured than the Brookston sandy clay loam soil. In fact, only a granular structure was exhibited. This soil had been spring plowed approximately one month prior to the time the cores were extracted. At the time of sampling, a potato crOp was being grown on the sampling site. Cores were extracted from between rows. No potato roots were evident within any cores. The profiles of mean BD, W, S, and ODR obtained from five replicate trial cores per application rate are presented in Table 5.8 thru 5.10. The mean profiles obtained from three cores sampled for initial moisture conditions are presented in Table 5.7. This profile indicates a reasonably uniform initial moisture profile existed below 4 cm. Mean ODR's for each depth increment sampled, irre— spective of application rate, were again sorted into five per cent pore space saturation intervals and are presented in Table 5.11. This table also contains the results of four additional cores to be described later. Comparison of the means for each category reveals a decreasing trend in ODR with increasing pore space saturation. The value of St = 78.5 per cent obtained from the soil 119 TABLE 5.7.——Summary of initial moisture conditions for Hillsdale sandy loam cores f’V Depth Bulk Moisture Pore space increment density content saturation AZ BDa wa s61 cm g/cm3 % % 0 e 2 1.064 9.78 17.40 2 a 4 1.444 9.92 31.49 4 — 6 1.565 10.54 40.33 6 — 8 1.538 10.53 38.56 8 — 11 1.538 10.69 39.15 11 - 14 1.493 11.12 37.99 14 — 17 1.531 10.62 38.53 17 — 20 1.552 11.25 42.18 20 F 23 1.523 11.07 39.67 23 — 26 1.508 11.05 38.66 means of 3 replicate samples 120 mmcfiomop moomoooao mm 00 mommE mmadsmm momoaflaoh m mo mommE M NH.:0 Hm.wa 0mm.a cm I mm mm.mm >H.0fi m0m.a mm I 0m 00.05 00.? gm; em I S 00.3. 3.0m wmmé Z“ I i 00.0N 00.Hm mwm.a :H I Ha mém 30> 0m.0m mmmé j” I w m.:m 00.0w m0.am 30m.H 0 I 0 0.? :2: mmém mmmé 0 I : 0.mm m:.00 m~.mm mam.a : I m 0:.Hm mH.0m 0HH.H m I 0 -00H x CHE\mEo\m & R mEo\w Eo ammo mm 03 mam N< comp scamsgmfio COHpmnopmm poopcoo zpfimcoo pcmEopocH commxo woman whom mospmfioz xasm npdoa comp coaomOfiaddm n£\Eo :0.m wcfi>fimoon monoo Emoa mocwm mamomaafim now momo coawsmufio Comxxo 0cm cofiuwpofifihcfi mo anmessmII.m.m mqmfimoop monoo Emoa 002mm mamomaafim pom mono coawzgmfio cowmxo 0cm coapwnoafiucfi mo zom€53mII.0.m mqm¢e mwcfiwmmn moompowao mm mo momma moHQEmw omeHHQoh m 00 mcmoE M 3:.0m wa.afi m0m.a mm mm R0.Rm m0.za m0m.a mm 0m moém 07ml” 034 cm 2 0m.m0 R:.0H 0mm.H RH :H mm.a0 mo.RH mmm.a an a H.m: 00.mR 4:.RH Hm0.H HH 0 m 0.3 2.8 2.: m8; a c 0.m: 03.00 mR.mH 0mm.fi 0 : :.0m 00.? 02.9” Hmmé : m 3R.mm 00.Hm m00.H m 0 @0H x cfiE\mEo\m R R mEo\w Eo ammo mm 03 mom N< mung ooamsmmao coapmmoumm pcoocoo mpfimooo pcosmgocfi cowxxo woman whom magpmfioz xaom comma all comp COfiomofiHddm nc\so Hm.0 mafi>fimomo monoo Ewoa mocmm oflmomafiflm pom mono coamsmmfio oommxo com coapmnpafimofi ho mnmEESmII.0H.m mqmpmmoo mam: mmump :oamstHc cowmxo can scan: CH pcmEopocfi o zpdmv on» mo pompcoo omdpmfloe mop an Umcfispmpoc m0: mhpcm nomm Mo COHpHmOQ one m 0.00 R.Rm R.:: H.0: 0.00 0.00 memo: 0.0: 0.0m 0.00 :.Rm R.0m 0.mm 0.0: 0.Rm H.mm 0.0m 0.H: H.0m 0.0m R.mm 0.0m :.0m 0.:0 0.0: 0.:: 0.m: :.0m R.mm 4.00 0.00 R.m: :.0m 0.0m R.mm m.R0 0.0: m.Hm 0.0: 0.00 :.mm 0.0: 0.0: 0.00 0.0: R.:: 0.0: 0.00 0.0m 0.0m 0.0: 0.00 0.:: 0.0m 0.0m 0:: m.:: T? 100 0.0: 0.:0 0.00 0.0: 0.0: 0.00 0.0m 0.00 0.»: H.0: 0.R: 0.0: 0.Hm 0.0m 0.0: H.0: 0.0: 0.0: 0.:0 m.0: m.0m 0.Rm 0.0m H.0R 00H I80 0 I 00009 2000:00H0 cowmxo 0.00 A 0.00 I H.0R 0.0m I H.0R 0.0» I H.00 0.00 I H.00 0.00 v Mmoo Log CH mam>hopcfi soapmnopmm woman opom mmfim>nopofi COHpmpopmm momma whoa oucfi amazonw momoo Emoa 00c00 onUmHHHm pom moods COHmSMMHU cowmxo ammo mo UCMII.HH.0 mqm<9 124 moisture retention tests falls within the pore space satura— tion category (75.1-80) possessing a mean ODR of 37.7 x 10-8 g/cmZ/min (increased to 44.3 x 10'-8 g/cm2/min if 15 per cent temperature adjustment is applied). The temperature adjust- ment raises the mean of this category to an adequate level while the mean of the next higher saturation category remains indicative of aeration deficiency despite the temperature correction. This agreement between the two methods employed was again demonstrative evidence of the fact that St can be reasonably predicted by incorporating ODR measurements with soil moisture retention measurements. Comparison of S = 78.5 per cent with the profiles t produced by 2.04, 1.02, and 0.51 cm/hr application rates (Tables 5.8 thru 5.10) respectively, shows that this value was exceeded to a slight degree in the mean profile pro- duced by the 2.04 cm/hr rate. The moisture profiles pro- duced by the lesser rates did not exceed S = 78.5 per cent. t The mean saturations in the region of ODR measurement (2-11 cm) were 76.01, 70.56, and 67.86 per cent for the 2.04, 1.02, and 0.51 cm/hr rates, respectively. The corresponding mean bulk densities were indicative of good homogeneity between experimental units. Consequently, it appears likely that profile saturation was influenced by application rate. The specific degree of influence was again somewhat masked by the tendency to allow the wet fronts to reach the core bottom be- fore termination of the infiltration tests. However, based on St as a criteria and the adequacy of aeration indicated in 125 the region of ODR measurement in Table 5.8, one may conserv- atively conclude that a rate of 1.02 cm/hr could be tolerated in this soil to a depth of 26 cm. The results of electrode behavior at each depth sampled are presented in FigureESJILwhere ODR was plotted as a function of time after initiation of infiltration. The general behavior was similar to that observed in the Brookston sandy clay loam cores of Section 5.41. Electrodes at every depth sampled exhibited an initial rise phase characteristic of inadequate wetting until moisture films began to limit oxygen diffusion. The total length of time that a rise phase was exhibited was obviously dictated by the application rate and the depth at which the electrodes were inserted. The true maximum ODR'S could not be established with the intermittent scheme of measurements which was employed. At 3 cm, the peak of the rise phase was not clearly attained by any of the three application rates. A review of the 2—4 cm increments (Tables 5.8 thru 5.10) revealed that these increments were decidedly less dense than the remainder of the profile. Consequently, pore space saturations were much lower in these depth increments and thus the probable reason for this electrode behavior. At the 5, 7, and 9 cm sampling depths, clear reversals in ODR trends took place for all three application rates. The gradual reduction in ODR exhibited by the 1.02 cm/hr curves at 5 and 7 cm between 60 and 120 minutes was the type Figure 5.2l.--Oxygen diffusion rate vs. 126 OXYGEN DIFFUSION RATE ODR -g/cm‘/min x IOa .A .A 42. 0/0 /0 APPLIGATION RATE cm/hr 35‘ / .—2.24 3045 O — L02 A—O.5l 24- 3 cm DEPTH ID 54‘ A\ _ o A 48‘ /O74 \0 \ \9 42% —_—‘—. ‘A 36- m-‘D 5 cm DEPTH __g4 ' O__i '/€r\\\‘\\\\\~ 48 1 ‘\\V<:;(//’/,/”_— ‘\\\\\\0 IA\\\\\\\\\\\ 424 . \\\\\ .A 361 I A 30 7 cm DEPTH __24 544 / 48‘ A 42+ \A 361 9 PT CHI 30 DE H O 60 l20 IOO 240 THE-MIN time at four sampling depths after initiation of infiltration tests in Hillsdale sandy loam undisturbed soil cores. 127 of trend expected in response to a lengthening infiltration profile as produced by an application rate which limits pore space saturation to safe levels. The trends of the 0.51 cm/hr curves at 5, 7, and 9 cm between 120 and 180 minutes were also evidence of expected behavior. The steepening down- trends near the end of infiltration tests at these two rates however, were likely due to the blocking effect of core bottoms which resulted in a rise in saturation not character- istic of the given application rate. A review of the electrical conductivity data indicated that the wet fronts were definitely not at the bottom of any core in less than 60, 120 and 180 minutes for 2.04, 1.02 and 0.51 cm/hr application rates,respectively. This fact can also be reasonably confirmed for both the 2.04 and 1.02 cm/hr rates by noting that the oxygen diffusion rates were still indicating a rise phase at 9 cm at 60 and 120 minutes, respectively (Figure 5.21). Because a severe decrease in ODR was already occurring at 7 cm in 60 minutes following the initiation of infiltration at 2.04 cm/hr, it appears that this rate could be excessive for penetrations much beyond 26 cm. Following the above series of infiltration tests on the Hillsdale sandy loam soil, an additional four cores were irrigated at 1.02 cm/hr for 120 minutes in an effort to obtain the effect of this rate on pore space saturation where the wet fronts definitely had not reached the bottom 128 of the cores. The resulting profiles of mean BD, W, S, and ODR are presented in Table 5.12. The mean pore space saturation profile revealed that the wet front had penetrated to approximately 23 cm in these tests. It can be noted that S was not exceeded at any t depth, and that mean oxygen diffusion rates at 3, 5, 7, and 9 cm were indicative of adequate aeration. The pattern of electrode behavior was very similar to that exhibited in Figure 5.21 for 1.02 cm/hr up to 120 minutes at all depths. Hence, a feasible Rt certainly exists in this soil for irrigations of at least 26 cm of penetration. 5.43 Existence of R in Situ in Hillsdale Sandy Loam Soil t Following the series of laboratory investigations, a limited field infiltration test was conducted in accordance with the apparatus and procedure presented in Section 4.41. An application rate of 0.81 cm/hr was employed to apply a total of 5.7 cm of water to four replicate cores. The basic intent of this test was to explore, briefly in terms of a more extended infiltration test, the likelihood of limiting profile saturation to levels below St = 78.5 per cent in the region of this soil for which this value was found appropriate. This possibility was recognized as most likely in soils possessing subhorizons of favorable permeability. Visual examination of the A horizon in this 2 soil led the author to believe such conditions did exist. 129 wmofiwmmh ocoppooam 0m 00 momma moaasmm opmofiaoop : mo mcon.M :m.m: 00.: 000A 00. I mm 00.:: 0:.HH R00.H mm I om 00.00 00.2 0004 0m I 5 00.00 00.0H 0m0.a RH I :H 0:.00 00.0H 0:0.H :H I Ha 0.0: 0:.0R :0.0H 000.H HH.I 0 H0: 00.: 00.0H 000A 0 I 0 0.H0 0H.0R 0R.0H 0H0.H 0 I : 0.0: 00.00 :0.Hm 0:0.H : I m 0:.0: 0:.mm RHH.H m I 0 W0H x CHE\mEo\w R R mEo\w So 2000 mm 03 000 N0 moms coamsmmfic coameSpmm poopooo mpfimoov pooEmpocH cowmxo oomam onom opopmfioz xasm nuoom monocHE 00H pom mums coapmofiaomm h£\Eo m0.H wofi>fiooon mopoo EmOH 00:00 mamcmflaam 0H ompomaaoo moan coamsoofla cowsxo 00m coaownpaaocfl oo RomsssmII.mH.0 000:9 130 The observed moisture profiles prior to the infil- tration test, at the conclusion of the test, and 48 hours after the test are presented in Table 5.13. The initial moisture profile when compared with Table 5.7 shows that the test site had been preconditioned to a moisture content very similar to the initial profiles of the cores investigated in the laboratory. The observed moisture profile at the conclusion of the infiltration test shows that St for this soil was not exceeded to a depth of 23 cm. Comparison with the initial profile indicated that approximately 2.2 cm of water infil- trated beyond 26 cm. The higher saturation in the last core increment was not readily explained in the absence of moisture measurements below 26 cm. This increase may have been due to slight redistribution of moisture in the time period (10 min) which elapsed between the end of the infil- tration test and core sectioning or to the textural dis- continuity between the Ap and A2-horizons at 26 cm. An approximate ten per cent difference in bulk density existed between the soil'cores of this test and those of the earlier laboratory tests. This difference was likely within limits to be expected as part of natural soil variation. The moisture profiles obtained at 48 hours after the conclusion of this infiltration test are included in Table 5.13 merely as a measure of "field capacity." 1Q31 .wmaoeom monogfiooL m 00 C005 0 wucmmmCooC znpco Comm m 50.00 00.0: mm0.~ :m.00 00.0w 5:0.H 00.0m .m.m 550.: on I m 0C.H0 m0.0~ 0::.H m0.m5 00.0w :00.H m0.:m 00.0: :::.H mm I on 00.0: H5.0H 00:.H «H.00 00.0H 05:.H 00.50 ma.fifi 50:.H on I 5: 05.:0 00.0: 00:.H 05.50 H0.0m 00:.: 05.00 0H.HH m::.H 5: I :~ 05.0: 50.:H 05:.“ mm.00 0n.0m 00:.H :0.0m 0.:H m0:.~ :H I Ha 05.0: 00.:H 00:.fi H:.00 05.0a 00:.H 0:.0: :m.HH H20.H a: I 0 00.00 No.0: 00:.H 00.:0 Ho.Hz 000.: HH.¢: m0.HH H5:.H 0 I 0 0H.m: 00.:H 00m.~ 0H.r0 00.0m :mm.~ 00.00 H5.HH 0H:.~ _ 0 I : 00.H: Hm.:H 50n.~ «0.00 00.00 50x.H 00.:0 00.HH 050.: : I m 05.:H . Cm.0m 0m.HC m I 0 R R mEO\w R a mEo\5 a R mEo\: Eu 0 I 00 m 3 00 m 3 a: N4 :oHomCzumw acouCoD huMmCoc Cofiuwazumm acoucoo zuflncov cofipaLzowm poouCoo hawmcmv 0:.soLoCH woman Doom opsumfios xasx woman whom Counfinfi :Hom woman 0Com oosuwfic: xasm Comma ammo COfiumnuHHCCH ammo COaumeHHQCw oHHCOCa C0000 mp: 0: am GaRCOCC opsumfioz ho UCm um oaauopa mononHCE oCsomHoe HmeHCH I ".anfl "."Hflfl"l'!h III I-” maufiw CH Hfiom EmoH hUCwm maaomCHH: :: cocamuoo moms coHAmApCfiocw no spasesmII.mC.0 000:9 132 Comparison of this profile with the initial moisture profiles for this test and the laboratory tests reveals that this soil was irrigated at a relatively high antecedent moisture content. The electrode behavior associated with the two cores which remained in place following the conclusion of the infiltration test is shown in Figure 5.22. The electrode behavior in the two cores which were extracted at the conclusion of the test was similar but not included in Figure 5.22 because of incongruency at the end of the test. The mean ODR's at 5 and 9 cm in the two cores which were extracted at the conclusion of the infiltration test were 70.1 and 47.9 x 10_8 g/cm2/min, respectively. Figure 5.22 once again indicates the characteristic wetting patterns of earlier tests. At 5 cm, a clear reversal of the rise phase did not occur while at 9 cm a gradual decline in ODR was taking place at the end of the infiltration test. The decline in ODR in the period following the infiltration test was not readily explainable in the absence of continuous precise moisture measurement in the region of these electrodes. Rather than advance uncertain hypothesis for this behavior it may better suffice to restate that the basic objective of this field test was to further demonstrate that an R existed for this media t such that St = 78.5 per cent was not exceeded during a nearly realistic irrigation. The end of test profile 133 .opflm Cfl HHoo Emofl known ifixwmfiaflm Ca omop coflproflflgcfi Cm mo cowwofloflCH C0000 mCCQoU mCHHoEmm cap 00 oEflp .m> ome COHmommHU commeII.mm.0 oComHm .z:: z. wszh. cow. 00¢. com. 000. com com 00¢ CON 0 o i E\eo 5.0 2 5“,: 22:05:; .8 02m oi Eu 0 lo 0 /O/ %. 0/0 (3 9‘20 V. N 833328 ‘9 G 8 cm x UIUI/,Iu0/b - HOD Elva NOIsnsJIO NBSAXO 134 (Table 5.13) did substantiate this. Also, despite the unexplained decline in ODR, adequate aeration was evidently maintained within the region of ODR measurement. Thus, in this respect the results of this test add to the overall aims of this investigation. The electrode behavior in the re— distribution period following infiltration is here regarded as a subject in need of future investigation. It should be noted that small amounts of compression (1-2 cm) were common to the insertion of cores into this soil as well as the Brookston sandy clay loam soil. This phenemenon was an experimental reality not readily overcome with present day sampling equipment. However, because the existence of feasible application rates were demonstrated in these slightly compressed soils, it is even more likely that feasible rates would exist in the true natural soil. CONCLUSIONS As a result of this study the following conclusions are presented: 1. The platinum electrode can be employed to define the pore space saturation of a soil which limits aeration. Soil aeration, as defined by the platinum elec— trode technique, is inadequate (ODR << 40 x 10"8 g/cm2/min) at pore space saturations in equilibrium with capillary heads less than the air entry head of a soil. Oxygen diffusion rates increase rapidly in response to drainage produced by capillary heads in excess of the air entry capillary head. In soils of the nature investigated in this study, aeration may be expected to approach adequacy 8 (ODR 1 40 x 10_ g/cmZ/min) at pore space saturations 1 81‘ In soils of the nature investigated, the aeration limiting soil moisture Content Wt (dry weight basis) may be expected to decrease linearly as bulk density increases. Moisture contents of soil profiles during sprin- kler irrigation are influenced by application 135 136 rate. In near homogeneous, isotropic soils initially air dry, pore space saturations within the transmission zone can be related to appli- cation rate by S = CRb Experimental results demonstrated clearly that reduced application rates will result in reduced saturation and increased aeration during irri— gation periods. Laboratory evaluation of S coupled with field t testing is a realistic approach in evaluating the aeration adequacy of soils during and immediately following irrigations. LIST OF REFERENCES 137 LIST OF REFERENCES Archibald, J. A. (1952) Effect of soil aeration on germi— nation and development of sugar beets and oats. Thesis for the degree of M. S., Michigan State University, East Lansing. (unpublished). Ashcroft, G., Marsh, D. D., Evans, D. D., and Boersma, L. (1962) Numerical method for solving the diffusion equation: I. Horizontal flow in semi—infinite media. Soil Science Society Of America Proceedings,26:522—525. Blake, 0. R., and Page, J. B. (1948) Direct measurement of gaseous diffusion in soils. Soil Science Society of America Proceedings,l3:37—42. Bodman, G. B. and Colman, E. A. (1943) Moisture and energy conditions during downward entry of water into soils. Soil Science Society of America Proceedings, 8:116—122. Baver, L. D., and Farnsworth, R. B. (1940) Soil structure effects in the growth of sugar beets. Soil Science Society of America Proceedings,5:45-48. Brooks, R. H., and Corey, A. T. (1964) Hydraulic prop- erties of porous media. Hydrology Paper No. 3, Colorado State University, Fort Collins. 25 pp. Bruce, R. R., and Klute, A. (1963) Measurements of soil moisture diffusivity from tension plate outflow data. Soil Science Society of America Proceedings, 27:18—21. Brutsaert, W. (1963) On pore size distribution and relative permeabilities of porous mediums. Journal of Geophysical Research, 68:2233-2235. Butijn, J., and Wesseling, J. (1959) Determination of capillary conductivity of soils at low moisture tensions. Netherlands Journal of Agricultural Science, 2:155-163. Cannon, W. A. (1952) Physical features of roots with especial reference to the relation of roots to the aeration of soil. Carnegie Institute of Washington Publication, 368:1-168. 138 139 Childs, E. C., and Collis—George, N. (1950) Permeability of porous materials. Proceedings of Royal Society of London, Series A, 201:392—405. Cline, R. A. (1957) The effect of applied fertilizer and oxygen diffusion rate on the growth, yield, and chemical composition of peas. Thesis for the degree of M. S., Michigan State University, East Lansing. (unpublished). Currie, J. A. (1960) Gaseous diffusion in porous media. British Journal of Applied Physics, 11:318—323. Currie, J. A. (1961) Gaseous diffusion in the aeration of aggregated soils. Soil ScienceI92z40—45. Ebstein, E., and Konke, H. (1957) Soil aeration as affected by organic matter application. Soil Science Society of America Proceedings, 21:585—588. Erickson, A. E., and VanDoren, D. N. Jr. (1960) The relation of plant growth and yield to soil oxygen availability. Seventh International Congress of Soil Science, Madison, Wisconsin 111:428—434. Finkel, H. J., and Nir, D. (1959) Gravity versus sprinkling methods of irrigation: A comparative study. Soil Science,88:l6—24. Finn, B. J., Bourget, S. J., Nielson, K. F., and Dow, B. K. (1961) Effects of different soil moisture tensions on grass and legume species. Canadian Journal of Soil Science, 41:16—23. Flockner, W. J., Vomicil, J. A., and Howand, F. D. (1959) Some growth response of tomatoes to soil compaction. Soil Science Society of America Proceedings, 23:188—191. Free, G. R., Browning, G. M., and Musgrave, G. W.' (1940) Relative infiltration and related physical charac- teristics of certain soils. United States Department of Agriculture Technical Bulletin 729. Fulton, J. M., and Erickson, A. E. (1964) Relation between soil aeration and ethyl alcohol accumulation in xylem exudate of tomatoes. Soil Science Society of America Proceedings, 28:610—614. Fulton, J. M., and Webber, L. R. (1953) Soil porosity and corn yields under nine three year rotations. Canadian Journal of Agricultural Science, 33:494—499. 140 Gardner, W. R. (1956) Calculation of capillary conductivity from pressure plate outflow data. Soil Science Society of America Proceedings, 20:317—320. Gardner, W. R., and Mayhugh, M. S. (1958) Solutions and tests of the diffusion equation for the movement of water in soil. Soil Science Society of America Proceedings, 22:197—201. Gardner, W. R., and Miklich, F. J. (1962) Unsaturated conductivity and diffusivity measurements by a constant flux method. Soil Science,95:271-274. Gray, A. S. (1950) The very slow sprinkler application rate. Irrigation Engineering and Maintenance, 9:7-17. Hanks, R. J., and Bowers, S. A. (1962) Numerical Solution of the moisture flow equation for infiltration into layered soil. Soil Science Society of America Proceedings, 26:530-534. Hanks, R. J., and Thorp, F. C. (1956) Seedling emergence of wheat as related to soil moisture content, bulk density, oxygen diffusion rate, and crust strength. Soil Science Society of America Proceedings, 20:307—310. Hansen, V. E. (1955) Infiltration and soil water movement during irrigation. Soil Science, 79:93-105. Jackson, R. D., VanBavel, C. H. M., and Reginato, R. J. (1963) Examination of the pressure plate outflow method for measuring capillary conductivity. Soil Science, 96:249-256. Kirkham, D., and Feng, C. L. (1949) Some tests of the diffusion theory and laws of capillary flow in soils. Soil Science, 67:29-40. Klute, A. (1952) A numerical method for solving the flow equation for water in unsaturated materials. Soil Science, 73:105—116. Kolthoff, I. M., and Lingrave, J. J. (1952) Polarography. Volume I and II New York: Interscience Publisers. Kopecky, J. (1927) Investigations of the relations of water to soil. Proceedings 1st. International Congress of Soil Science, 1:495-503. 141 (1954) Estimating and Beenhauwer, 0. 35:719-725. Krimgold, D. B., Agricultural Engineering, infiltration. Kristensen, K. J. (1959) Investigations of some soil factors influencing oxygen diffusion in the liquid Cornell Thesis for the degree of M. S., Ithaca. (unpublished). (1964) Soil aeration Kristensen, K. J., and Lemon, E. R. III. Physical aspects and plant root relations. of oxygen diffusion in the liquid phase of the soil. Agronomy Journal, 56:259—301. Kunze, R. J., and Kirkham, D. (1962) Simplified accounting for membrane impedance in capillary conductivity determinations. Soil Science Society of America Proceedings, 26:421—426. Leamer, R. W., and Shaw, B. (1941) A simple apparatus for measuring noncapillary porosity on an extensive scale. Journal of American Society of Agronomy, 33:1003-1008. Soil aeration and plant root relations: phase. University, Lemon, E. R. (1962) I. Theory, Agronomy Journal, 54:167—170. Lemon, E. R., and Wiegand, C. L. (1962) Soil aeration and II. Root respiration. Agronomy plant root relations: Journal, 54:171-175. (1952) The measurement Lemon, E. R., and Erickson, A. E. of oxygen diffusion in the soil with a platinum microelectrode. Soil Science Society of America Proceedings, 16:160-163. Letey, J. L., Stolzy, H., Blank, G. B., and Lunt, 0. R. (1961) Effect of temperature on oxygen-diffusion rates and subsequent shoot growth, root growth, and mineral content of two plant species. Soil Science, 92:314-321. M. R., and Powers, W. L. (1938) A study of factors Soil Science Society of affecting infiltration. America Proceedings, 3:334-339. Marshall, I. J. (1958) A relation between permeability and size distribution of pores. Journal of Soil Science, 9:1—8. Lewis, Marshall, T. J. (1959) The diffusion of gases through porous media. Journal of Soil Science, 10:79-82. 142 Miller, E. E., and Elrick, D. E. (1958) Dynamic determi— nation of capillary conductivity extended for non— negligible membrane impedence. Soil Science Society of America Proceedings, 22:483—486. Miller, R. D., and Richard, F. (1952) Hydraulic gradients during infiltration in soils. Soil Science Society of America Proceedings, 16:32—38. Mueller, C. C. (1965) Determination of water drop characteristics from a medium pressure rotary irrigation sprinkler. Thesis for the degree of M.S., Michigan State University, East Lansing. (unpublished). Nielson, D. R., Biggar, J. W., and Davidson, J. M. (1962) Experimental consideration of diffusion analysis in unsaturated flow problems. Soil Science Society of America Proceedings, 26:107—111. Nielson, D. R., Davidson, J. M., Bigger, J. W., and Miller, R. J. (1964) Water movement through Panoche clay loam soil. Hilgardia, 35:491—505. Nielson, D. R., Kirkham, D., and Van Wijk, W. R. (1961) Diffusion equation calculations of field soil water infiltration profiles. Soil Science Society of America Proceedings, 25:165—168. Penman, H. L. (1940) Gas and vapor movement in the soil. II. The diffusion of carbon dioxide through porous solids. Journal of Agricultural Science, 30:570—581. Phillip, J. R. (1955) Numerical solution of equations of the diffusion type with diffusivity concentration dependent. Transactions of the Faraday Society, 51:885—892. Rawlins, S. L., and Gardner, W. H. (1963) A test of the validity of the diffusion equation for unsaturated flow of soil water. Soil Science Society of America Proceedings, 27:507-511. Richards, L. A. (1948) Porous plate apparatus for meas— uring moisture retention and transmission by soil. Soil Science, 66:105-110. Richards, L. A. (1930) Capillary conduction of liquids through porous materials. Physics,l:3l8—333. Rijtema, P. E. (1959) Calculation of capillary conduc- tivity from pressure plate outflow data with non negligible membrane impedence. Netherlands Journal of Agricultural Science, 7:209—215. 143 Rommel, L. G. (1922) The aeration of soil as an ecological factor. International Institute of Agri— culture Bulletin 1957. Rubin, J., and Steinhardt, R. (1963) Soil water relations during rain infiltration: 1. Theory. Soil Science Society of America Proceedings, 27:246-251. Rubin, J., and Steinhardt, R. (1964) Soil water relations during rain infiltration: II. Moisture content profiles during rain of low intensities. Soil Science Society of America Proceedings, 28:1-5. Russel, M. D. (1952) Soil aeration and plant growth. Agronomy Monograph II, B. T. Shaw ed. New York: Academic Press. 253—301. Sasso, G. (1958) Tests on sprinkling and furrow irri— gation of maize. Attic cent. naz. Mecc. Agric. 2:247—258. Steward, F. C. (1960) Plant physiology. Vol. I A. Cellular Organization and Respiration. New York: Academic Press. Stolzy, L. H., and Letey, J. (1964) III. Correlation of plant response to soil oxygen diffusion rates. Hilgardia, 35:567—575. Stolzy, L. H., Letey, J., Szuszkiewioz, T. E., and Lunt, O. R. (1961) Root growth and diffusion rates as functions of oxygen concentration. Soil Science SOCiety of America Proceedings, 25:463—467. Swartzendruber, D. (1963) Non—Darcy behavior and the flow of water in unsaturated soils. Soil Science Society of America Proceedings, 27:491—495. Taylor, S. A. (1949) Oxygen diffusion in porous media as affected by compaction and moisture content. Soil Science Society of America Proceedings, 14:55-61. Van Bavel, C. H. M. (1952) Gaseous diffusion in porous media. Soil Science, 73291-104. VanDoren, D. M. Jr. (1958) Relationship between oxygen diffusion rates as measured with the platinum microelectrode and plant growth. Thesis for the degree of Ph. D., Michigan State University, East Lansing. (unpublished). - 144 Wesserling, J., Van Vijk D., Fireman, M., Van't Wouldt, B. D., and Hagen, M. (1957) Land drainage in relation to soils and crops, Monograph VII. America Society of Agronomy, 461—513. Wiegand, C. L., and Lemon, E. R. (1958) A field study of some plant-soil relations in aeration. Soil Science Society of America Proceedings, 22:216—221. Wollny, E. (1890) Untersuchungen Uber die Beeinflussing der Fruchtbarkeit der Ackerkrume durch die thatiheit der Regenwurner. Rorsh. Gebiete Agr. Physik, 13:381—395. Yoder, R. E. (1937) The significance of soil structure in relation to the tilth problem. Soil Science Society of America Proceedings, 2:21—33. Youngs, E. G. (1964) An infiltration method of measuring the hydraulic conductivity of unsaturated porous materials. Soil Science, 97:307-311. APPENDIX A 145 Determination of Log-Effective Saturation as a Function of Log-Capillary Head The following procedure for obtaining a plot of effective saturation (Se) as a function of capillary head (h) was utilized in this investigation (see also Brooks and Corey, 1. 1964). Experimentally obtain the moisture retention curve for a given soil (Figure A.l). Approximate the residual saturation (Sr) by selecting a value of S at which the retention curve appears to approach a vertical asymptote (Figure A.l). With this estimate of Sr’ compute tentative values of S from the relationship S = S _ Sr . e e 15:75- r Plot tentative values of log Se as a function of log h (Figure A.2). Usually, this first plot will not result in a straight line at higher values of h. Re—estimate Sr as indicated in Figure A.2 and continue steps 3, 4, and 5 until a straight line is obtained. Obtain associated parameters; i.e., pore size distribution index (A), estimated air entry capillary head (Pb/Y), and the Brooks and Corey saturation intercept (SI). 146 147 S O 20 4O 60 80 IOO 20. 0I ‘E'I 40» ?| IA'I 60" ml °| 80- w E '3' 2| .C l20~ mI "l (0 I40» El IL l60 Figure A.1.-—Capillary head as a function of saturation. IO ‘0 p ‘Ofir b/Y 0.5bCOMPUTED \ 5: VALUES OF 88 FOR 3r -- 020 ‘VA . W? P \ ‘0 2 nd ESTIMATEQ‘I/S: OF 5r OBTAINED \ I ’ BY FITTING \é GONPUTED POINT ON STRAIGHT LINE s-s, 5" I-s. no. I l J l 1L1}; l l J 1 l I 11 IO so IOO 200 h - cm Figure A22.--Effective saturation as a function of capillary head.