AN EVALUATION OF SOIL AND WATER MANAGEMENT PRACTICESON A LOWLAND SOIL WITH POOR NATURAL DRAINAGEbyJOHN STANLEY HEINONENB.Sc., The University of Toronto, 1978.A THESIS SUBMITTED IN PARTIAL FULFILLMENT OFTHE REQUIRMENTS FOR THE DEGREE OFMASTER OF SCIENCEinTHE FACULTY OF GRADUATE STUDIES(Department of Soil Science)We accept this thesis as conformingto the required standardTHE UNIVERSITY OF BRITISH COLUMBIADECEMBER, 1992© JOHN STANLEY HEINONENIn presenting this thesis in partial fulfilment of the requirements for an advanceddegree at the University of British Columbia, I agree that the Library shall make itfreely available for reference and study. I further agree that permission for extensivecopying of this thesis for scholarly purposes may be granted by the head of mydepartment or by his or her representatives. It is understood that copying orpublication of this thesis for financial gain shall not be allowed without my writtenpermission.Department of ,S002.).2LL--CS2_ The University of British ColumbiaVancouver, CanadaDate )a,e& /8DE-6 (2/88)iiAbstract Low-lying areas of the Lower Fraser Valley tend to havepoor natural soil drainage. Excess moisture, because of a highwater table and ponded surface water, restricts the range ofcrops that can be grown and influences farm managementdecisions regarding the nature and timing of field operations.When economic considerations lead farmers to commence fieldoperations when the soil is too wet, the result can bestructural degradation of the soil, leading to the formationof surface seals and crusts and a compact ploughpan. The aimof this study was to evaluate the effectiveness of variousmanagement practices in controlling the water table, to gainan understanding of the mechanisms governing excess wateraccumulation, persistence, and removal from lowland soils, andto make management recommendations from these.Monitoring was conducted on two adjacent farms on WesthamIsland. They differed in their management of excess water: onerelied on surface ditching, subsoiling, mole drains, and covercrops, while the other used a system of perforated subsurfacedrainlines emptying into a ditch that was pumped to keep thewater level low. Soil-water pressure potentials were measuredautomatically and manually with piezometers located inselected depressions and surrounding slightly higher groundfrom February to April 1981. Depth of ponded surface water andwater level in the ditches draining each farm were monitored.Rainfall was recorded with four rain gauges. Bulk densitiesand saturated hydraulic conductivities of the surface crust,ploughpan, and subsoil were measured. Soil penetrationresistances and aggregate stabilities (wet sieving) weremeasured as indices of compaction and structural stability.Earthworm numbers and biomasses were measured in both farm.Land drainage through a system of subsurface drainlineswas found to lower the water table more rapidly afterrainfall, and reducing the depth and duration of pondingcompared to undrained land. Drainage resulted in moretrafficable and opportunity days for field operations.Drainage dramatically increased earthworm populations which inturn improved the drainage via the pathways for water movementprovided by burrows. Although there was a more severeploughpan in the undrained farm there was no evidence that itimpeded water movement. There were no differences in aggregatestabilities between farms.The presence of ponded water in depressions was found tobe due mainly to a low permeability surface seal which wasmore severe in the undrained farm. The presence of the sealresulted in surface runoff to low areas and slowed waterinfiltration there. Subsoiling and moling may also aggravateexcess water problems by directing water to low areas. Surfaceditches were found to be ineffective in removing ponded water.Using a cover crop, which reduced the severity of crusting andponding, was a more effective practice than subsoiling whenthe soil was undrained.ivTABLE OF CONTENTS PageABSTRACTTABLE OF CONTENTS^ ivLIST OF TABLES viLIST OF FIGURES^ viiLIST OF APPENDICES ixACKNOWLEDGMENTSCHAPTER 1 - INTRODUCTIONIntroduction and Purpose^ 1Literature Review^ 61. Drainage^ 6a. Drainage Design Criteria^ 7b. Subsurface Drainage and Water-Table Depth^8c. Water-Table Control: Trafficability-Workability and Timeliness 9d. Subsoiling^ 13e. Mole Drainage 142. Aggregate Stability 153. Soil Sealing and Crusting^ 184. Earthworms and Drainage 19CHAPTER 2 - MATERIALS AND METHODS1. Soil and Landscape Descriptions^ 21a. Topography, Elevation, and Drainage^21b. Parent Material and Texture 212. Site Description and Location 223. Piezometers and Ditch Water-Level Recorders^284. Surface Crust Saturated Hyraulic Conductivity 31and Bulk Density5. Soil Profile Saturated Hydraulic Conductivity^33a. Large Core Method^ 34b. Auger-Hole Method 356. Earthworm Numbers and Biomass^ 387. Penetration Resistance Profile 388. Bulk Density Profiles^ 399. Aggregate Stability 3910. Rainfall^ 41VPageCHAPTER 3 - RESULTS AND DISCUSSIONIntroduction^ 421. Piezometer Data^ 422. Surface Ponding in the Lowest Spots in Each^44Site3. Analysis of Recording Piezometer Graphs 474. Comparison of Water-Table Behavior in^52Depressional Areasi. Sites D1 and D2^ 52ii. Sites D2 and U2 56iii. Sites Ul, U2, and U3^ 565. Variations in T with Depth 61i. Depressional Sites 61ii. High Sites^ 686. Water Table - Trafficability Relationships^687. Ditch Water Levels 808. Profile Hydraulic Conductivity^ 809. Surface Crust Ks^ 8610. Surface Crust and Soil Profile Bulk 89Density11. Earthworms^ 9312. Water-Stable Aggregates^ 9613. Penetration Resistance and Compaction^9914. Effects of Moling, Subsoiling, and Subsurface^101Drainage on the Water Regime of DepressionalAreasi. Moling^ 101ii. Subsoiling 102iii. Subsurface Drainage^ 10215. Surface Ditching to Remove Ponded Water^104CHAPTER 4 - SUMMARY AND CONCLUSIONS 107BIBLIOGRAPHY^ 111APPENDICES 116viLIST OF TABLES PageTable 1 Crescent Soil Description 23Table 2 Westham Soil Description 24Table 3 Surface Ponding Data for Depressions 45Table 4 Saturated Hydraulic Conductivity Data: Large 82Core Method (plough layer removed)Table 5 Auger Hole Method Hydraulic Conductivity Data 83Table 6 Surface Crust Saturated Hydraulic Conductivity 87Table 7 Surface Crust Bulk Density 90Table 8 Profile Bulk Density (Core Method) 91Table 9 Earthworm Numbers and Biomass 94Table 10 Water Stable Aggregation Data 97viiLIST OF FIGURES FigureFigure12Map of Low-Lying Portions of the Lower FraserFraser River ValleyClimate Data for Vancouver InternationalAirportPage23Figure 3 Pea Harvesting Equipment and Compacted Soil frompotatoe harvesting under wet conditions17Figure 4 Location of the Research Site on Westham Islandin relation to British Columbia25Figure 5 Location of Research Sites on adjacent drainedand undrained farms26Figure 6 Pumphouse in Farm D with float on right 27Figure 7 Examples of the two piezometer designs used 30Figure 8 Laboratory Aparatus for Measuring Soil Crust K, 32Figure 9 Large Core Technique for measuring K, 36Figure 10 Field measurement of K, using large core method 37Figure 11 Recording piezometer "T data from depressionalportions of sites Dl and Ul48Figure 12 Comparison of manual and recording piezometer 53AFT data for site UlFigure 13 Water table comparison from manually readpiezometers in Site D153Figure 14 Comparison of water table heights from sites Diaand Dld54Figure 15 Comparison of water table heights from sites D2a,D2b, and D2c55Figure 16 Comparison of water table heights from sites U2aand D2a57Figure 17 Comparison of water table heights from sites Ulaand Ule58viiiPageFigure 18 Comparison of water table heights from sites U2a 59and U2bFigure 19 Comparison of water table heights from sites U3a, 60U3b, and U3cFigure 20Figure 21Figure 22Figure 234,7, vs Z (depth) graphs for Sites U2a, D2a,^62and U3a4rT vs Z (depth) graphs for Sites Ula, Dla,U1b,^64and Dlb4%. vs Z (depth) graphs for Sites D2c, Did,^69U3c, and UleRelationships between ponded days, trafficable^72days, opportuity days and days on which rainwas recordedTotal Ponding-Trafficability relationships for^73all SitesDitch water level data for Farm D and Farm U^81Crust properties and Earthworm Behavior during^95Wet PeriodsFigure 24Figure 25Figure 26Figure 27 Penetration Resistance Profiles^ 100Figure 28 Effects of mole draining at a constant depth^103below the soil surfaceFigure 29 Unsuccessful surface ditch installation^105ixLIST OF APPENDICESPageAppendix 1 UBC Evaporation Pan Data^ 116Appendix 2 Westham Island Rainfall Data 117ACKNOWLEDGEMENTS My wife Yiolanne, and children Julia and Keith, have providedme with continual support and encouragement which was needed tocomplete this project.I thank my thesis committee for the direction given, and JandeVries for framing the problem and his many suggestions aboutapproaches to take. Martin Driehuyzen proposed the use ofpiezometers for the study and this proved invaluable.Many people helped with the field work but none more thanTrevor Murrie. Thanks to the Reynolds and Swenson families forallowing me to conduct my field research on their farms. Bernie vonSpindler and Bill Cheang assisted in many ways. The community ofthe Soil Science Department provided enthusiasm and comradeship.Thanks to Jack Dobb, Peter Mills, Charlie Arshad, and AlanStewart for helping to lighten the load in the last phase of myM.Sc. program. I am also grateful for the support my boss JohnPayne has shown.Finally, I deeply appreciate the positive approach of AndyBlack and Mike Novak and their help in hurdling a myriadbureaucratic barriers. Where there's a will there really is a way!Free at last, free at last, .1CHAPTER 1INTRODUCTIONIntroduction and PurposeLow-lying portions of the Lower Fraser River Valley, asshown in Figure 1, generally have poor natural drainage. Thisis due to a number of factors including the deltaic nature ofthe area, the proximity to the Fraser River and sea level,and runoff originating from adjacent forested and urbanizedupland areas. The region's humid maritime climate (see Figure2) is also a major contributing factor.Nearly level to gently undulating topography andmedium- to fine-textured soils combine with the abovefactors to produce a fluctuating, but generally high,regional water table in the lowland areas. Excess watercan accumulate in and on the soils in these areas at any timeof the year as a result of poor natural drainage; this mostcommonly occurs in the off-season from November 1 to March 1.Fraser Valley farmers find that surface water (floodingand ponding) reduces soil trafficability and workability andthereby reduces the timeliness with which land preparation,planting, harvesting and manure spreading can be carriedout. It is a major cause of soil structure degradation2Figure 1: Generalized map of the agricultural lowland areas inthe Lower Fraser Valley of British Columbia. Based oninformation from Winter (1966) and Luttmerding(1980).3Figure 2: Climatic normals based on the period from 1951-1980for Vancouver International Airport (Atmospheric EnvironmentService, 1980,1984).4leading to the formation of a surface seal (Abbaspour,1988). Excess water can also lower crop yields and quality,or result in complete crop failure.Although lowland areas of the Lower Fraser Valleyare protected from flooding by over 500 km of river and seadykes (Victoria, 1968) regional water management is generallyinadequate. It relies heavily on passive methods of waterremoval such as gravity operated flood boxes that open onlyat low tide and close during high tide. Most pumpingstations are located around urbanized lowland areas and asa result the removal of excess water on a regional scale isnot sufficient to meet agricultural requirements.Whenever this regional problem is combined with runofffrom uplands and inadequate on-farm water management, soilwater-logging and surface ponding often occur throughoutthe rainy winter months. Water frequently covers the soilsurface in local depressions, sometimes to depths of 0.25 m ormore. This tends to discourage cultivation of overwintercrops such as cabbage and cauliflower, made possible by thearea's mild climate. If this surface water persists intoMarch, April, or even May, significant delays in tillageoperations may result since the decision to commence landpreparation usually depends on how dry these depressionalareas are. Farmers are generally reluctant to bypass wetdepressions in their fields except in the wettest springsbecause of the difficulties caused for plowing.5When wet conditions continue, economic and otherconditions lead farmers to begin land preparation when thesoil is too wet. This can degrade soil structure in thesurface layer, and, in the case of plowing and rotovating,also results in the formation of a compacted plowpan. Sohne(1958) found that compaction from surface applied pressurescan occur down to depths of up to 0.60 m.Surface structure degradation can increase thesusceptibility to surface sealing and crusting, whichdecreases infiltration rates, causing local surface runoff andponding. If the plowpan remains intact it can also restrictvertical water movement and result in wet surface soilconditions. The problem is therefore somewhat cyclic innature.The difficulties associated with the accumulation ofexcess water in and on lowland soils are ofconsiderable agricultural importance in the Lower Fraservalley. Good soil and water management practices canalleviate many of the negative consequences of poor regionaldrainage, however adequate regional drainage is requiredbefore the full benefits of proper on-farm managementpractices can be attained.The primary aim of this study was to evaluate theeffectiveness of various management practices in controllingthe behaviour of the water table, since water table depthgreatly affects surface trafficability and workability. A6broader purpose was to gain an understanding of the mechanismsgoverning excess water accumulation, persistence, and eventualremoval from lowland soils. This information is used torecommend soil and water management practices best suitedfor the control of excess water, and to suggest improveddrainage design criteria. Differences in soil physicalproperties resulting from variations in management were alsostudied, as they indicated how to optimize managementpractices.LITERATURE REVIEW1. DrainageDrainage of agricultural land is practiced in humidareas in order to reduce the water content of wet soils andthereby increase their aeration, temperature and strength(Marshall and Holmes, 1979).Most permanent^subsurface drainage^involvesinstallation of perforated plastic drain pipes in conjunctionwith drainage ditch water level control. The depths,spacings, and lay-out of the drainpipes vary according todesign criteria. Short term subsurface drainage improvementscan also be achieved by subsoiling and moling. Surfacedrainage, which is done by land levelling, or shaping toform temporary shallow ditches, can be used to removeexcess surface water.7a. Drainage Design CriteriaThe agricultural function of a drainage system is tohelp increase the economic returns from the farm enterprise(Bouwer, 1974). Economic criteria have to be included intothe design process of farm water management systems. It isobvious that in every circumstance a point is reached beyondwhich equal increments of increased benefit are achieved onlyat the expense of increasing additional costs. The optimumdrainage design is one which maximizes the difference betweenannual benefits and annual costs (Buras, 1974). However, ourpresent state of knowledge is such that an assessment of thefull economic implications of better drainage is not yetpossible (Morris and Calvert, 1976).In practice, drainage criteria are based on acombination of environmental conditions (climate, soil,weather, etc.) and management concerns (eg. croprequirements, trafficability - workability and timeliness).In each case the design choice is made within the generaleconomic constraints as perceived by the farm manager. Givenall the variables, there is considerable potential forsub-optimal drainage designs. It is therefore essentialto assess the effects of various drainage systems in orderto obtain appropriate drainage criteria that will resultin improved design (Oosterbaan, 1980).According to Bouwer (1974), the drainage designcriterion expresses the drainage deficiency of a field to be8corrected by a drainage system. Drainage criteria can beevaluated for a variety of conditions depending on the mainfunction of the drainage system (Bouwer, 1965).The lack of adequate design criteria constitutes theweakest link in rational drainage design (Bouwer, 1974).Dieleman (1979) stated that the cause of failure in drainagedesign is more often a lack of understanding of the broadinter-relations between drainage and other farm or watermanagement matters than the lack of precise data. Drainageis therefore more than just the determination of the correctdimensions of the system. Optimum depths and spacings areprobably strongly dependent on local conditions(Oosterbaan,1980).b. Subsurface Drainage and Water-Table Depth Shkinkis (1979) presented results showing significantlowering of water tables due to drainage, the effectincreasing with drain depth and reduced spacing (sandy loamto sandy clay soils, 0.9-1.8 m drain depths, and 12-45 mspacings). Trafford and Oliphant (1977) concluded,however, that relatively shallow drainage systems alonecaused no decrease in mean water table levels, but whencombined with subsoiling or moling the mean depth of the watertable was increased from 0.23 to 0.40 m ('heavy' clay soil,0.80 m drain depth, 15-60 m drain spacings,subsoiling at0.45 m and 1.5 m spacings, moling at 0.55 m and 2 m spacings).9They also found that the effect of decreased drain spacingswas small especially when drainage was combined with moling orsubsoiling, likely due to the drain depth being shallow.Wind (1976) demonstrated the influence of drainage depthon soil workability in the spring under climate conditions inthe Netherlands. He showed a significant increase in thenumber of work or opportunity days in March and April asdrain depth is increased from 0.9 to 1.1 m. A drain depth of1.5 m appeared to give optimum water-table control forworkability. These varying results serve to confirmthat the effectiveness of a drainage system is partiallydependent on local soil and hydrologic conditions. Thedepth of the regional water-table is an example of a localcondition which could be important in this regard. Theaverage depth of the water-table resulting from on-farmwater management is however, generally an excellentindicator for trafficability, workability, and plant growthconditions and it can serve as a good criterion forevaluating the effectiveness of a drainage system(Oosterbaan,1980).c. Water-Table Control: Trafficability-Workability and Timeliness Timeliness in the performance of farming operations is amajor objective of drainage. Draining excess water from asoil reduces drying time and causes an increase in soil1 0strength and therefore trafficability, thereby makingadditional time available for carrying out farming operationsand lessening the risk of soil compaction and structuraldegradation. A long-term loss in productivity potential mayresult from too much traffic or soil manipulation at timeswhen the soil is not trafficable or workable (Reeve & Fausey,1974).Paul and deVries (1979b) established linear relationshipsbetween water-table depth and cone penetration resistance (asoil strength index) of the plow layer, and used theserelationships to establish critical water-table levels fortrafficability of drained mineral soils. They definedtrafficability as "the ability of a soil to support trafficwithout receiving structural damage beyond the limits forgood crop growth".Paul and deVries (1979b) found that cultivated mineralsoils became trafficable as much as 25 days earlier in thespring where effective subsurface drainage systems loweredthe water-table rapidly in comparison with the same soils intheir natural poorly-drained state. Similar results werereported by Fausey and Schwab (1969) who found that for aclay soil, the moisture content in the upper 0.3m was 4-5%higher throughout the spring with surface drainage thanwith subsurface drainage. This resulted in a 17-day delayin planting date. Subsurface drainage can thereforesignificantly improve the timeliness of farm operations, with11consequent gains in crop yields (due to earlier and moreflexible planting dates) and hence in farm income. Theadditional number of trafficable days can be critical on claysoils, in some years being the difference between normalyields and total crop failure (Armstrong, 1978).The definition of trafficability given by Paul anddeVries (1979b) should, for most practical purposes, benormally extended to include workability. Their definitionis satisfactory for cases where machinery rides on the soilas with manure spreading, but it does not adequately applywhere tillage or planting are involved.Workability, in a general context, refers to the easewith which the soil may be manipulated and the relativesuccess of the manipulation in achieving the desired results.Since the types and purposes of soil manipulation are sovaried, and the methods of characterizing the soilinadequate, it is difficult to generally define workabilityin a quantitative manner. The workablility criteria developedby van Wijk and Feddes (1986) was a soil and crop specificvalue (a workability limit). If the pressure potential ("I') at5 cm depth in the clay soil they studied was below a thresholdvalue of 4rp = -70 cm, then moderate to good field conditionsfor potatoe planting prevailed. On days with ir p higher thanthe threshold value, planting could not be accomplishedwithout damaging the soil structure. Van Lanen et al (1992)used this criteria and simulated pressure potentials to12calculate workable days. A higher probability of workable dayswas found to result in the possibility of a longer growingseason.Trafficability and workability are sometimes usedinterchangeably in an agricultural context. If anagricultural soil is workable it is certainly alsotrafficable, however, the reverse is not always true, eg.a hard dry soil is quite trafficable but marginallyworkable. For tillage and planting the range of moisturecontents where a given soil is both workable and trafficableis of most interest. This range extends from the moisturecontent where the soil becomes friable rather than hard, tothe plastic limit. The optimum part of the range is likelysomewhat removed from its wetter or drier ends.Wind (1975) stated that the most important factorgoverning workability is weather. In humid regions thetiming of farming operations is highly dependent on weather,particularly in the early and late season operations. Sincemoisture inputs to the soil are mainly from precipitation,the most practical means of gaining control over soilmoisture conditions in humid areas is to provide drainage(Reeve & Fausey 1974).Since the time suitable for field work is directlyrelated to the weather, there is an element of risk in allfarming operations. By reducing weather-induced uncertainty,drainage can provide a more predictable environment for farm13management decisions. The number and distribution of workabledays in spring is important for the choice of type and sizeof farm machines and for the organization of farm work(Wind,1976). The drainage investment can help towardsminimizing the considerable 'hidden' costs of untimeliness infarm practices and as a result reduces the necessity toover-invest in farm machinery system capacity as a hedgeagainst the vagaries of weather (Morris & Calvert, 1976).Since soil moisture is directly related to the weather,the concepts of risk and uncertainty associated with weathershould be logically incorporated in the design ofdrainage systems (Reeve & Fausey, 1974). Where drainagesystems are designed for trafficability control, a random-typelayout is used whereby drain lines are located so as todrain as many low areas or wet spots in the field as possible(Bouwer, 1974). In lowland areas a partial random layout maybe used to augment a system of parallel spaced drains. Byincreasing water-table draw-down in depressions a partialrandom layout can affect the desired control withoutrequiring major, and possibly expensive, adjustments to thedrainage system as a whole.d.SubsoilinqSubsoiling is a form of deep tillage which disrupts thesubsoil without inverting it or bringing it to the surface(Wells, 1956). The basic aim of subsoiling is to loosen14compacted soil such as a plowpan, and thus improve verticalwater movement and the growth of plant roots. In order to beeffective subsoiling should only be carried out when the soilis dry and brittle, for maximum shattering. The distancebetween subsoiler blades should also be at least equal to, orless than, the depth of subsoiling to ensure overlap oflifting zones (Swain, 1975). Subsoiling is only beneficial insoils with a compacted layer that restricts root developmentor water movement, and where there is a better structuredor drained horizon below (Wells,1956).Compaction can also be alleviated by using a recentlyintroduced slant-legged soil loosening implement, theParaplow. It works to depths of 0.25 to 0.35 m; soilpassing over the slanting legs is placed under tensioncausing cracking to occur along natural planes of weakness.The degree of soil disturbance and fracturing is varied byadjusting the spacing between the legs and by a flap hingedto the trailing edge of the leg, the amount of soil liftingand cracking being altered by the angle between the flap andthe leg (Chaney and Hodgson, 1984).e.Mole DrainageThe aim of mole drainage is to create a channel, usuallycylindrical, at a depth of 0.5 to 0.7 m, with cracksradiating from it, that will conduct water laterally to adrain line or open ditch. Mole drains will not persist for15more than a few months unless the clay content is greater than40% and the soil is structurally stable under saturatedand near-saturated conditions. The operation is also onlysuccessful in clay if the soil consistence is plastic atmoling depth and friable above (Swain, 1975). Spoor et al(1982b) identified clay mineralogy and soil bulk density atmoling depth as being factors influencing the suitability ofsoils for mole drainage. They identified numerous factorsimportant in the formation, stability, and failure of moledrainage channels. It is evident from their results thatit is difficult to accurately predict the suitability ofsoils to mole drainage.2. Aagreciate StabilitySoil structure can change in response to differentagricultural practices and can, in turn, influence theproductivity arising from agricultural practices. Artificialdrainage diminishes the length of time that the soil has highwater contents and low structural stability and enhances theextent of wetting and drying (Kay, 1990). Differences intrafficability also result in the potential for structuraldegradation being greater on undrained fields, given similartillage or harvesting dates. The problem of soil degradationhas also become more serious generally, because of the trendto more intensive field traffic by heavier machines (Raghvanet al, 1978). These points are illustrated by the photographs16in Figure 3.Aggregate stability, as determined in the laboratory, iscommonly used as an index of soil structure stability in thefield. The type of disrupting force exerted on aggregatesduring laboratory experiments depends on the situation ofinterest in the field (Kemper, 1965). In humid areas, suchas coastal British Columbia, the water-stability is a measureof the relative susceptibility of the soil to compaction,surface sealing and crusting, and is a factor in theinfiltrability and aeration of the rootbed.Comparatively little research has been done on theeffects of drainage treatments on soil structure. Wesselingand van Wyk (1957) concluded that improvements to soilstructure due to drainage are small. Leyton and Yadav (1960)working in Great Britain found a significantly higherproportion of water-stable aggregates after five years ofdrainage. This difference was attributed to lower water-tablelevels. Sieben (1974) reported deterioration of soil structurecaused by persistance of water-table levels near thesoil surface. Fausey and Schwab (1969) noted adverse soilstructure and more severe crusting on undrained plots wherethere were long periods with ponded water. Kay (1990) statedthat there is little information on the rate of change instructural characteristics when artificial drainage isintroduced in agricultural soils.183. Soil Sealing and CrustingSoil seals form where the aggregated structure near thesurface is more or less destroyed and the primary particlesare rearranged. This change can come about under theinfluence of externally applied mechanical pressures such asraindrop impact or vehicle traffic (Bonsu, 1984), or throughspontaneous slaking of soil aggregates in the course ofnatural wetting-drying cycles (Hillel, 1960). Soil crustsform as the seal dries and becomes harder and more compact.Surface seals can form at the soil surface when sedimentfills in surface pores and builds up a layered, dense soil(Abbaspour, 1988). Sediment can originate from erosion andrainsplash in adjacent higher areas surrounding a depressionor from wave action where there is surface water.One of the most significant effects of a surface seal isthe decrease in infiltrability, resulting in ponding. Surfaceseals and crusts also inhibit seedling emergence and aerationof the rootbed. McIntyre (1958) found that surface crustsreduced the hydraulic conductivity of the soil profile by asmuch as a factor of 10 4 .When rainfall rates exceed the infiltrability of a sealexcess water either accumulates or runs off. In individualfields both may occur. The water accumulating in surfacesealed depressional areas, may be augmented by runoffgenerated from surrounding higher areas with surface seals.The depth of water accumulated in the depression may be19several times the amount of actual rainfall depending on theseverity of sealing, and the effective catchment area. Upondrying, higher areas will develop a network of cracks whichbreach the seal. These cracks permit entry of water into thesoil until the infiltrability is exceeded or the cracks areclosed by swelling, infilling due to rainsplash and erosion,and the reformation of a continuous seal. In the pondeddepressions cracks cannot form until the water is removed byinfiltration through the seal and evaporation. The water mayalso be removed through construction of a shallow surfaceditch, or direct pumping from a sump into the main ditchusing a portable pump.4. Earthworms and Drainage Earthworms require reasonably aerated conditions andhence do not thrive in poorly drained soils prone towaterlogging or where the soil surface is covered by pondedwater (Guild, 1948). They can, however, survive for a time inwater provided that it is sufficiently oxygenated (Russell,1973).Two reasons why the presence of earthworms in a soil isdesirable are: (1) they contribute to the formation ofwater-stable aggregates (Hopp and Hopkins, 1946; Dawson,1947; Dutt, 1948, Evans, 1948), and (2) their burrowingactivities improve air and water movement in the soil byproviding non-capillary pore space (Guild, 1955). Bouma et al20(1977) and Bouma (1981) studied the effects of macropores,including earthworm burrows on soil hydrology. The influenceof earthworm burrows increases dramatically with the number,diameter, and vertical continuity of burrows. Bouma'sresearch has demonstrated that earthworms can significantlybenefit agriculture where they are found on finer-texturedsoils with poor natural drainage.Earthworm populations have been shown to be particularlysensitive to crop rotations and their associated tillageoperations (Russell, 1973). Earthworm numbers have been foundto decrease after plowing and discing, but increase inbarley-clover rotations which leave the soil undisturbed forperiods sometimes greater than one year (Carter et al, 1981).Qualitative field observations have indicated dramaticdifferences in the number of earthworms between drained andundrained farmland in the Lower Fraser Valley. In undrainedland significant residual earthworm populations are onlypresent along ditches where the soil is better aerated duringthe wet months. In drained land earthworms are more evenlydistributed throughout the fields. These differences inearthworm populations appear to have significant effects onsoil physical properties, especially hydraulic conductivity.21CHAPTER 2MATERIALS AND METHODS 1.^Soil and Landscape DescriptionThe soils in this study consisted of a Crescent andWestham series complex.(a)Topography. Elevation and Drainage The topography of the complex varies from nearly-level togently-undulating with slopes up to two percent. The relief isgenerally 0.5 m or less. Westham soils are poorly drainedand sometimes have slightly depressional topography. TheWestham soils usually occur at slightly lower elevations thanthe adjacent Crescent soils which are moderately poorly topoorly drained. Elevations are all less than 5 m above sealevel.(b)Parent Material and Texture The parent materials of the Crescent and Westham soilsare medium to moderately-fine textured, stone-free, deltaicdeposits of the Fraser River, usually deeper than 1 m andunderlain by sand. Moderately- to strongly-saline materialsare usual below 1 m in the Crescent series and below 0.5 to 1m in the Westham series. Sulfurous compounds are also usuallypresent at these depths(Luttmerding, 1981a). The Crescent soil22is described in Table 1, the Westham soil in Table 2.2.^Site Description and LocationFive sites were established at Westham Island (located asshown in Figure 4) on two adjoining farms with the same soiltypes but having contrasting soil and water managementpractices (see Figure 5). Two of the sites were on Farm D (Dfor drained) which has perforated plastic drainpipes installedat an average depth of 1.2 m and spaced 18 m apart. Thesedrainpipes empty into a central ditch in which the water levelis controlled by a pump (Figure 6) automatically triggered bya float switch. The water is pumped into a drainage canal.This drainage system was installed nine years prior to thestudy. Farm D had also not been subsoiled for a period of atleast eleven years and had no shallow surface ditches in placeduring the study. Site D2 was in an area beyond the end ofthe central ditch where the drain lines fan out, so thespacing is consequently greater than 18 m.The remaining three sites were located on Farm U (U forundrained) which had no permanent subsurface drainage system.The problem of excess water and poor drainage was dealt within three ways in Farm U: (1) subsoiling to break up the plowpan and speed vertical water movement, (2) mole drains to movewater laterally, and (3) shallow surface ditches to removeexcess surface water. Site Ul had been subsoiled and a covercrop of fall rye was planted to protect the soil surface fromraindrop impact and hence reduce the degree of crusting. Site23Table 1: Crescent Series - Profile DescriptionHorizon^Depth DescriptionAp^0-18cm^Dark greyish brown (10YR4/2,moist)silty clay loam. Weak medium tocoarse subangular blocky structure.Friable hard. Abundant fine roots. pH5.4.Bg^18-30cm Grey (10YR5/1,moist) silty clay loam.Massive to coarse angular blockystructure. Slightly sticky, firm,slightly plastic. Few fine roots. pH4.8.Cgl^30-48cm Grey (10YR5/1,moist) silt loam.Massive structure. Slightly sticky,firm, slightly plastic. Very few fineroots. pH 4.4.Cg2^48-97cm Dark Grey (5Y4/1,moist) silt loam.Massive^structure.^Non-sticky,friable, slightly plastic. Commonrhyzo-concretions. pH 3.9.Cg3^97+ cm Dark Grey (2.5Y4/l,moist). Massivestructure. Non-sticky, non-plastic.Common rhyzo-concretions. pH 3.4.(from Luttmerding, 1981a)24Table 2: Westham Series - Profile DescriptionHorizonAp2Ap2CglCg2CsgDepth^Description0-15cm^Dark^grayish^brown(10YR4/2,moist) silt loam. Fineto medium subangular blockystucture. Friable when moist.Many roots in grassed areas. pH6.1.15-25cm Grayish brown (2.5Y5/2,moist)silt loam . Medium subangularblocky structure. Friable whenmoist. Abundant roots in grassedareas. pH 5.3. Abrupt boundaryat plow pan.25-40cm Gray (5Y5/1,moist) silt loam.Massive structure. Firm whenmoist. Yellow (5Y8/6, moist)mottles. pH 4.140-95cm Dark gray (5Y4.5/l,moist) siltloam. Massive to weak subangularblocky structure. Reddish-brownmottles; rhyzo-concretions. pH3.495+ cm Dark-gray (5Y4/1,moist) sandyloam. Massive. Friable whenmoist. pH 3.4.(from de Vries, J. and C.L. Paul, 1978)25Figure 4: Location of the research site on Westham Island inrelation to British Columbia.26Figure 5: Schematic plan view of the research sites on WesthamIsland. The letter "a" is located in the lowest spot in adepressional area and each subsequent letter indicatesprogressively higher areas.11111111111/61+Lh _„...„„...„27Figure 6: Pumphouse in Farm D with float on right.28U2 was located in a field which was not subsoiled in the fallof 1980 and where the soil surface was severely crusted. SiteU3 had been subsoiled, however, the soil was left bare and asresult had a severe surface crust. The ditches surroundingFarm U are connected to large regional drainage ditch thatflows into the south arm of the Fraser River at low tide viaa sluice gate. These ditches had not been properly cleaned forseveral years. The sediment and vegetation in them preventedoptimum water flow.3. Piezometers and Ditch Water-Level Recorders Piezometer nests were installed at each of the five sitesin late January and early February of 1981. The piezometersare referred to by the abbreviation assigned to them, forexample: Dia, where "a" is a location within Site Dl. Eachsite was located in a depression where ponding had beenobserved. Nests of two to six piezometers, with the testsections located at a various depths in the 0.10 m - 0.80 mrange were installed in the lowest point in each depressionand also on adjacent higher areas where there was no prolongedsurface water accumulation. Another piezometer nest waslocated at an intermediate position in all sites except U2.Piezometer nests and a recording piezometer were located insites Dl and Ul, for more intensive monitoring. Stevenswater-`level recorders were installed on the ditches in bothfarms in order to relate ditch water-levels to water-table29behavior in the field.The recording piezometers consisted of an 0.18 m diameteraluminum tube with a removeable covered platform containing aStevens water-level recorder. The lower (open) end of thetube was covered with fibreglass window screening bound onwith baling wire. The other piezometers consisted of 0.04 minside diameter galvanized steel pipes pressed together at oneend. In the test section slits were sawed into the bottom0.05 m. An alternate design employed fibreglass windowscreening and bailing wire as in the recording piezometer (seeFigure 7).All piezometers were installed in auger holes 0.04 mlarger in diameter, with gravel placed around the test sectionto prevent clogging. The pipes were then sealed in withconcrete. Small tins were placed over the tops of thepiezometers to prevent rainwater entry. Before measurementscommenced each piezometer was emptied once or twice to removemud which had entered through the test section duringinstallation.The water level in each manual piezometer was measuredwith a float tube and a calibration was carried out toaccount for the volume of water it displaced. The recordingpiezometers and ditch water-level recorders were checkedregularly to ensure accurate time and water-level.3 0Figure 7: Examples of the two piezometer designs used.314.^Surface Crust, Saturated Hydraulic Conductivity & BulkDensitySaturated hydraulic conductivity (KJ measurements weremade on nine field-moist samples of the 0.03-0.04 m thickcrust taken from site U3. Samples were taken after the onsetof cracking so the main sampling criteria was that there beapproximately a 0.15 m X 0.15 m area unbreached by cracks.Each sample was carefully carved into an 0.11 m X 0.11 mblock. Care was also taken to level the bottom of each blockwithout smearing the soil. The crust was placed on a porousplate composed of closely-packed, very-fine sand, and aplexiglass chamber with internal dimensions 0.12 X 0.12 X 0.07m high was put over it (see Figure 8). A water-tight seal wasobtained by pouring molten paraffin around the sample to thetop of the crust. Once cooled the wax-plexiglass interface wasfurther sealed with stop-cock grease. The sample was thenflooded and inflow measured using a marriot device which alsomaintained atmospheric pressure at about the crust surface.A suction of 0.06 - 0.07 m of water was imposed at the base ofthe sample by lowering the outlet. The hydraulic gradientacross the crust sample was measured using manometersconnected to the inflow and outflow ports, and K, wascalculated using Darcy's Law. Measurements were taken atintervals during periods ranging from 42 to 70 hours.The hydraulic conductivity of the surface crust was alsomeasured by performing a water balance calculation ((depth of32Figure 8: Laboratory apparatus for measuring soil crust K 5 .air ventgasketALYanometersQyrryriot deviceLHaxair bubbler intakeq - - ■•• • ‘'• • •", ••••^ow sandscreenoutletq33water on the surface) = rainfall - evaporation - drainage) forthe ponded area in site U3. Changes in the height of pondedwater were accurately measured by counting the number of turnsrequired to bring a sharpened screw with a known thread pitchinto contact with the water. The hydraulic gradient acrossthe crust was determined by measuring the depth of surfacewater and the hydraulic pressure at a depth of 0.14 m using apiezometer. AL in Darcey's Law was taken to be 0.14 m ratherthan the thickness of the crust, because the exact hydraulicgradient across the crust itself was not measured. Theevaporation data was obtained from class A evaporation panmeasurements (Appendix 1) at the University of BritishColumbia climate station located 21 km north of WesthamIsland. Calculations were only made for sunny days in orderto insure more comparable evaporation regimes between site U3and the evaporation pan location and to avoid errors due tosurface runoff into the depressional area during rain.Bulk density of the surface crust was measured using theclod method after Blake (1965). Three samples were taken fromboth the 0-0.015 m and 0.015-0.03 m layers of each of the ninelaboratory Ks crust blocks.5.^Soil Profile Saturated Hydraulic ConductivityProfile saturated hydraulic conductivity (Kd wasmeasured using two techniques:34(a)Large Core Method In July 1980 K, was measured in both farms D and U. Infarm U two sites were established in the general vicinity ofsite Ul. One of these had been subsoiled the previous autumn(blade spacing i m and depth approximately 0.5 m), the otherhad not been subsoiled for more than one year. The site onfarm D (near site D1) had not been subsoiled for more thaneleven years. Sample locations at each site were randomlylocated in a 20 x 50 m area.The plow layer was removed to the surface of the plowpanand a 0.25 m diameter, 0.30 m high column of soil carefullycarved out, after the technique of Bouma (1978), with an uppersurface at the top of the plowpan. The smeared areas of theupper surface were picked clean with a knife to re-open poresand the soil particles were removed with a battery-poweredportable vacuum cleaner.The sides of the column and a 0.05 m deep trough dugaround its base were painted with puddled Haney clay toprevent infilling of pores with plaster or concrete. Thetrough was filled with a slurry of quick setting dentalplaster and a 0.35 m high tube of plastic-covered single-faced corrugated paper was placed so that its base was set inthe plaster around the base of the column. The top of thecorrugated paper tube was even with the top of the soilcolumn. The space between the tube and the soil column wasfilled with concrete mixed at three parts sand to one part3 5Portland cement. Plaster had originally been used to encasethe soil after Bouma (1978) but it was found to be too weakand prone to leaks.A 0.30 m diameter, 0.10 m high, metal collar (a cut-offpail top) was set in the concrete on an angle and a plexiglasstop with water inflow and air outflow ports was clamped ontothe rim of the collar. The soil was then flooded and the airoutflow port clamped off once all the air was displaced.Inflow was syphoned from a small bucket whose water height waskept level with the soil surface using a float valve, i.e. acondition of "just ponding" infiltration was maintained toensure the water entered the soil at atmospheric pressure.Inflow into the bucket was measured for one hour using a onelitre graduated cylinder, and this was assumed to be equal toinflow into the soil column. The measurements were takenafter three to four hours when steady infiltration wasreached (as indicated by nearly constant readings over onehour) K, was calculated assuming a hydraulic gradient ofunity. The experimental procedure is illustrated in Figures9 and 10.(b)Aucter-Hole Method Profile K, measurements were also made in the low areasin each site in the spring of 1981 using the auger-hole methodas described by van Beers (1965). Ten centimetre diameterholes were augered to depths of between 1.2 and 1.6 m and thewater level in them allowed to equilibrate with the wateroutletstopperpail rim—cardboardoncret eclayplastergrgduated cylinderair ventgasket36Figure 9: Large core technique for measuring K.3 7Figure 10: Large core method of K, measurement in the field.Plaster has been used to encase the soil in the example shown.38table for at least 24 hours before measurements commenced.Water was baled out of the hole and then the rate of refillingmeasured. The number of measurements per site were roughlyproportional to their relative areas. Sample locations wererandomly chosen along a line transect.6. Earthworm Numbers and Biomass Earthworm numbers and biomass were measured in adjacentbarley fields in both farms. Sampling was carried out in thehigh unponded areas of both farms. No attempt was made todifferentiate between earthworm populations in higher areasand those in lower areas prone to periods of ponding. A 0.25m2 quadrat frame was used to sample the top 0.15 m of soil.Six samples were randomly taken along a line transect in eachfield in early January 1981. Samples from farm U werehand-sorted in the field, those from farm D were partiallyhand sorted in the field with the remainder hand sorted in thelaboratory. Earthworm biomass was determined on a dry weightbasis for five of the earthworm samples for farm A (oneearthworm escaped during storage). No biomass measurementswere made on Farm U (only one worm was sampled). Qualitativeobservations of earthworm behavior made in the field alsoappeared to be of importance.7. Penetration Resistance Profile Shortly after the disappearance of ponded surface water,cone penetration resistance was determined with a Gouda type39HSA-5, 0.05 m2 base, 60° tip cone penetrometer. Measurementswere made at depths between 0 and 0.55 m at 0.05 m intervals.Three measurements were made at each of ten locations randomlychosen along a line transect through the area of ponding insites Ul, U2, D1, D2, for a total of thirty replicates persite. Water content was determined gravimetrically for 0.10m depth intervals at each of the ten locations. It wasassumed that the effects of differences in water managementpractices on trafficability and compaction would be mostapparent in these low areas.8. Bulk Density Profiles Bulk density samples of the 0 - 0.50 m depth were takenat 0.10 m intervals using the core method after Blake (1965).Five locations randomly selected along a line transect weresampled in sites Ul, U2 and D1 and two in site D2 (this sitewas disturbed by tillage before completion of sampling).9. Aggregate StabilityTen 500 g samples of the plowlayer (0 - 0.15 m) weretaken from the areas subject to surface ponding in sites Uland D1, and an additional ten were taken in both sites fromadjacent higher areas not subject to ponding. The sampleswere randomly chosen along line transects when the soil watercontent was just below the plastic limit.Each sample was passed through an 0.008 m sieve and the40portion greater than 0.00475 m collected on a sieve belowafter shaking by hand. Fifty grams (moist weight) of fieldmoist soil from the 0.00475 m - 0.00800 m fraction was placedon a nest of four 0.15 m diameter sieves with mesh openings of0.00475, 0.00200, 0.00118 and 0.000212 m. The gravimetricwater content was determined for part of the remaining sample.The samples were allowed to moisten slowly beforeimmersion but were not completely protected from slaking. Fournests of sieves were raised and lowered 0.038 m through waterthirty times per minute, for ten minutes, on a Yoder (1936)type wet sieving machine. A sliding mechanism attached to thedrive shaft caused the nest of sieves to rotate through anangle of 20 ° on up and down strokes. The size distribution ofwater-stable aggregates was expressed as the meanweight-diameter:MWD = Exiwiwhere^MWD is the mean weight-diameter of aggregates (m),n is the number of size fractions,xi is the mean diameter (m) of each size fraction ,andw; is the proportion of the total sample (dry weight)occurring in the corresponding size fraction (Kemperand Chepil, 1965).4110. Rainfall Rainfall was measured with four 0.12 m diameter plasticfunnel rain gauges. Two of these rain gauges were located ineach of sites Dl and Ul.42CHAPTER 3RESULTS AND DISCUSSION1.^Piezometric Data Piezometers measure the pressure potential MO ofsoil-water only below the water table. In this region 4 ,1, isalways greater than atmospheric pressure, ie. ir p>0, and thesoil is saturated. The total potential of soil-water, irT , isthe sum of 4,1, and the gravitaional potential NFG , ie. *-1..110-4,0 .The gravitational potential is generally taken to equal zeroat the soil surface.Water movement in saturated soil is described by Darcy'sLaw:q = -KdARTdZwhere q is the discharge through a cross sectional area perunit of time, Ks is the saturated hydraulic conductivity, diri.is the total potential difference over distance Z (dZ) in thesaturated soil (Hillel, 1971).When (d*r/dZ)=0 there is no discharge, in other words thesoil is hydostatic. When (d4ydZ)=-1 the soil is drainingfreely at rate equal to K. For saturated soil, Ks is at amaximum and is fairly constant in time. Saturation alsoensures that large 'T gradients common in drier soil do notnormally occur. In the absence of artificially imposed43gradients, the dominant driving force causing water movementis gravity, so that pore water in saturated soil moves much asit does in bulk free water (Childs, 1969). Where ditches andsubsurface drains are present significant horizontal flow canoccur in response to resultant horizontal pressure gradients(Luthin, 1957).The position of the water table is mainly determined bythe rate of water percolation through the layers of soiloverlying it. A rise or fall of the water-table indicates anet recharge or discharge of groundwater respectively (Hillel,1971).The location of the water table has been calculated fromthe piezometer data. The water table was taken to be thehighest point in the soil where 'I', the pressure potential,equals zero. The depth of ponded water (if any) is irp at thesoil surface. It was assumed that the soil between twopiezometers in a nest, both containing water, was saturated.For practical purposes, the water table was taken to be locatedat the water surface when there was water on the soil surface.On days where no readings were taken the water table wasassumed to be at the mid-point between the previous andsubsequent values. The two exceptions to this were therain-free periods, March 17 to March 24 and March 26 to March28, where the water table was assumed to be equal to thereading on the last day.442. Surface Ponding in the Lowest Spots on Each SiteFigures 11 to 19 clearly show that no surface water wasrecorded at any of the high spots in the four sites, which isan obvious consequence of their topographic position. A filmof water was noted on the surface of each high spot during andshortly after some rainfall events when the surfaceinfiltrability had been exceeded. This was observed to resultin surface runoff at a rate dependent on the slope, theseverity of surface crusting and other antecedent conditions.In the high areas in the drained sites some runoff wasapparently intercepted by cracks and worm burrows connected tothe surface, which were common. In the undrained sites veryfew worm burrows were observed and it therefore appeared thata higher proportion of surface runoff flowed into the low,ponded areas. The severity of the surface crust appearedgreater in the undrained farm which may have resulted in morerunoff. It was also noted that the surface runoff sometimescarried with it sediment in varying amounts. The mechanism ofsediment generation was not studied, but was thought to be acombination of rainsplash and sheet erosion.The duration and average depth of water ponding differedconsiderably between Farms U and D as shown in Table 3. Thelow areas at sites Ul and U2 had a mean ponding duration of46.5 days during the fifty-five day period from February 12 toApril 7th. During the same period sites D1 and D2 had a meanponding duration of 18 days. The absence of adequate land45Table 3: Surface Ponding Data for DepressionsSite^ Ul U2 U3 D1 D2Average Depthof Pondedwhen ponded -■ 12.5 10.3 6.5 2.7 2.7Water (cm) over measurement-*period8.4 10.3 6.5 0.9 0.91-+Period with Ponded Water^1 38 55* 38* 17 19from Feb. 12 to Apr. 7 (days):The average depth of ponding is expressed in two ways: (i) theaverage depth of ponded water over the period February 12 toApril 7th, (ii) the average depth of ponded water for the dayswith ponding.*ponding during entire period of measurement46drainage resulted in an average of 2.6 times more days withsurface water in the low areas of Farm U. No primary seedbedpreparation such as ploughing can be initiated when there areany ponded areas in a farmer's field, so that times of pondingcan be viewed as "no go" periods. The longer ponding periodin site U2 (when compared to site Ul) can be attributed to itsmore severe crust which resulted from poor surface cover. Thedata suggest that cover cropping is more effective thansubsoiling, if undrained.The fact that there were 17 to 19 days of ponding in thedrained farm suggests that the drainage system was notoperating properly, or that weather conditions exceeded thesystem's capacity. It was observed that the pump was oftenswitched off for up to fifteen minutes due to lowering of thefloat, even though the water level in the ditch feeding it wasabove the level of the drain outlets. This was due to theculverts under the machinery crossing being too small and theditch being too filled with sediment and weeds to efficientlyconduct water to the pump. It should be noted that the averagedepth of ponded water is in part a function of the geometry ofthe low spot and how much water it can hold beforeoverflowing. The average depth of ponded water, when present,ranged from a mean of 11.4 cm for sites Ul and U2 to a mean of2.7 cm for sites D1 and D2. The average depth of ponded waterover the entire 55 day period was 9.4 cm for sites Ul and U2and 0.9 cm for sites D1 and D2. The latter is an order of47magnitude less than the former due to the combined effects ofsmaller depth of ponding and the presence of worm burrowsconnected to the surface in low areas which allowed surfacewater to enter the soil, bypassing the crust. This processhas been referred to as "short-circuiting" by Bouma and Dekker(1978). The capacity of each ponded area did not appear tohave been exceeded during the period of data collection. Itcan therefore be concluded that the data were likely notinfluenced by overflow.The data for sites U2 and U3 are quite similar in thatthere was ponded water on the surface throughout therespective periods of measurement. The average depth of waterat site U3 between March 1st and April 17 was 6.5 cm.3. Analysis of Recording Piezometer DataThe difference in the behaviour of soil water betweenFarm U and Farm D is clearly represented in the recordingpiezometer plots shown in Figure 11. The 4%. in the recordingpiezometers at Sites Ul and D1 commence rising quickly afterthe beginning of rain following prolonged dry periods, such ason February 13 and March 29. There are greater fluctuationsat Site D1 after the initial rise in irT , which indicates thatit is more responsive to rainfall inputs, even with surfacewater present. It also indicates that water is being removedmore effectively through the system of subsurface drainsbetween rainfalls.W 0. hiZ 0 H-O, (-0h00- IIi-I^(D• trO Fart1-•0" ••II(D (DWi-,riL31 OnH. II< am H-Zrt. "1Orort- N.• M(D N0I-, 0O MO ft^.It•^pi M^03I-, liDIo -iI--i-, gvZ(fl (I0li rnM0ID piO ti,M gu' 00M11°0aIVrtcn M1-- '1ItO aul mtoo rt-I-.49At both sites 4%., as measured by the recordingpiezometers, increased to greater than zero during mid- tolate February, though it rose higher and longer in Site Ul.The 4,1. at Site D1 increased by about 0.20 m with the rainfallon March 3. The rate of rise was significantly greater thanthe 0.04 m increase at Site Ul. This relatively sluggish andsmaller rise at Site Ul can be attributed to poorer hydrauliccontact between the soil and the water overlying it due to thepresence of a surface crust. The additional rain input wentprimarily into storage on the surface along with the existingsurface water. The fact that the surface water-level at SiteUl rose 0.05 m in response to an 18 mm rainfall was probablya result of surface runoff inputs from surrounding higherareas. The presence of a surface crust in higher catchmentareas can therefore result in a greater depth of water pondingon the surface of low areas.The increments at site Ul, shown in Figure 11, weregreater on March 15, 16, and 25, despite the relatively lowand falling water table and absence of surface water, due tothe wetter conditions than in Site D1 in the soil above thewater table. At Site D1, for the same rainfall, water thatmight otherwise have contributed to a rise in AFT at depth wentinto storage in the unsaturated soil above the water table.Ponding also resulted from the 29.4 mm rainfall on March29, but only at site Ul. Ponded water was last observed atsite D1 on March 3, and then it was a relatively minor amount.50The period of ponding at site Ul following March 29 differedfrom the earlier one; the depth of ponded water in this caseclosely reflects the rainfall inputs. This was apparently dueto the extensive cracks which were observed to breach thesurface crust during the preceding rain-free period. Theconsequent increase in the surface infiltrability resulted inlittle runoff into the ponded area. The general pattern of 'Tbehaviour during this time was similar to the earlier and moresevere period of ponding.The drop in skr at site Ul after the April 5 rainfalltrails the recession in site D1, however, the rate at whichthe 'YT dropped was very similar. The lag time for site Ul wasdirectly related to the difference in the length of time waterwas present on the surface. At the end of the two periods ofponding'YT was,-0.20 m at 1.21 m depth at site Ul on the lastday water was recorded on the surface. The same was true ofSite D1 on February 27. When there was water on the surfaceAFT generally did not drop to lower than -0.20 m. In view ofthis it is advisable that all possible measures be taken toeliminate, or at least reduce the depth and duration ofponding. Drainage as practiced in Farm D was more successfulin eliminating surface water and in reducing in low areasafter the cessation of rain.Figure 12 shows the manually collected data frompiezometer Ula plotted with the recording piezometer chartfrom site Ul. There is close agreement between the two whichi-s- DI IV 04M 0 0 P..1 P, th1jM 1-t)< Al 0) Pi1-,- a 0 ID0+ M PiM ■ MNO II tvct 11) 0 ..• 0 MIIi) al 01(I—.• ct °ct M 15M Pi Pi11a 1-"~b ril. ori 0(-I- 00I—'^0o a 1-1)co IIIM rt 0ru w0, ... 0LEI 01 tr. piM i...,M ctEl 0' guMO 1 art MI—,^ 0) lM tt Mtt I-a- 0• 0M M PiM ao rt l'4 *0ft 'Q• rtM 0'1sM 1.-■-rt M1-1 NO 0 0O ElEl pi mM 1-, rert m0" co iio oaw l,co 1—• .-152indicates that the recording piezometer data can be used tointerpolate 'PT and water table behaviour between readings frommanually recorded piezometers in comparable locations.Similar close agreement was found in the piezometric data fromsite Dl. This agreement occurred despite the differences intest-section depth and the greater diameter of the recordingpiezometers, which required that twenty times more water bedisplaced from or into the soil for each unit rise or fall inthe water table. That the greater diameter of the recordingpiezometers did not influence the comparison is likely due tothe test-sections being located in the relatively highlypermeable sandy subsoil. The similarity in *1. beneath about0.70 m depth seems to indicate that it does not vary muchvertically below this depth under saturated conditions.Figure 13 shows the close agreement which existed betweenpiezometers Dlb and Dlc which were located about 8 m apart.The data are similar in spite of minor differences in wellpoint depth and surface elevation, which indicates that thewater table is nearly level, at least over short distances.4. Comparison of Water-Table Behaviour in Depressional Areas(i) Sites Dl and D2 Figures 14 and 15 show that the respective water-tablesat Dla and D2a behaved similarly throughout the period ofmeasurement.E Nei▪ Fa-rt ctiqm mII CI) 1C7 Nct• Htr tr.CD gu •o• 1.4 pciCr), 0 pi• tiI-A-tr 1-3 (aa' PI 00p. 0I 0 frt)• cuo oH• m• oo 0^L..)ort(Drt 0,0• 0,rtM"a rta s• trz m0N mE 1-1-"11a rtO 01rf0) i• tiiiiE56ii) Sites D2 and U2 The water tables for sites U2a and D2a are plottedtogether in Figure 16. At the lower piezometer the 4%. at U2awas higher than D2a with the exception of the periodsassociated with the onset of prolonged rainfall in mid-February and late March. On both these occasions the '"T atD2a increased more quickly to rainfall inputs due to betterhydraulic contact between the soil surface and lower portionsof the profile. This is especially due to the absence of acontinuous slowly permeable surface crust at D2a. On severaloccasions there is a rise of 0.1 m or more at D2a whilethere a continued decline or slight rise at U2a for similarreasons; for example refer to March 3 and 16, and April 5.The 4,1, decline after March 3 is initially much faster atD2a, with its rate of fall slowing somewhat after March 5. Thefall in 4%. at U2a generally trails D2a by about ten days.There was ponded water on the surface crust at U2a throughoutthe measurement period, and this dominated the hydrologicbehaviour of the soil.(iii) Sites Ul, U2, and U3 The piezometric data for sites Ul, U2, and U3 are shownin Figures 17, 18 and 19 respectively. Sites U2a and U3a werevery similar in that they both had ponded water on the surfacethroughout the entire period of measurement, and the depth ofthat water followed a similar pattern over time. In both casesdropped to -0.50 m in mid-March despite the presence of thew •)I--, lai ...tO 1•)• 14)H.I 0 l'0tui p)0 t• oO 1-4 aH.'P M • 0^vi^O 0^--.10 1--1artH' (Dit 0.,H-OZ 0M rt.1:D MPi "asm tr• M1E1M 0.as• /""1.0mz 61-rt• glia) ctrt Na gir- 1.-. .;I-1 • •1:1101 Za 0e' 0O H p0.0 m wIn • 111--Cr 1-3 coM ti 0i-a 00 0i-r,E rb1_,.I 0 11)Ill ill0 tr 0O I-I gliig ID I-'•^0^U10 0 03O i-,o 1:1113, fl-1-,- (1)ft 1:1,1-, •0Z mU1 rtgo MII ".- ft13) 0U1 trU1 I-,0 (7)(D o•asE C G• I-'-kgIDtilrt1.4t:r^03•M^• aHa 0• tEi▪ trN•• ti)-•-tr HM 11 0O M B.p I 0 it• 1))0 0' 0W0 • M• 0O 00• 1:11• rt• aP-O 4O pirt• MPi Pi(1-• W(11 erMA. (DJ.0"• 4Mtnrt0'aM rt61ponded water, although this occurred four days earlier at siteU2a. The rate of decrease in AFT was similar at both sitesprior to this time. U3a is more responsive to rainfall inputsthan U2a, eg. March 16 and 29. This may be due to site U3being a smaller, more pronounced depression. The latter mayalso account for the rise of AF T above zero at U3a on March30th and thereafter. The water ponded on the surface crust atU3a dominates the hydrologic behaviour of the soil as it doesat U2a.Site Ula, in contrast to U2a and U3a, was free of pondedwater in early February, mid- to late March and in mid-April.The presence of a cover crop prevented formation of a surfacecrust severe enough to reduce infiltration to the point whereponded water persisted throughout. Yet despite this, and thefact that it had been subsoiled, the water table did not fallbelow 0.21 m in the dry period in mid-March. Even withsubsoiling and a cover crop the absence of adequate drainageresulted in excess soil moisture in this depressional area.Some of these differences remains unaccounted for; they may bedue to the inherent landscape variations between the sites.5.^Variations in AFT with Depth (i) Depressional Sites The "T vs. Z graphs for depressional sites are shown inFigures 20 and 21.62Figure 20: iri. vs. Z (depth) graphs for Sites U2a, D2a, andU3a. Missing data points occur when the piezometer tubecontains no water ie.*.<0.• Site U2a^■ Site U3a^• Site D2a■ Site U3a^• Site D2a• Site U2aMa r18^I^Mar 25Mar.29^ Mar. 31l eII50^ 50I^I'U 00Apr.5^ Apr.9^ Apr.16063Figure 20 continued: 4,1. vs. Z (depth) graphs for Sites U2a,D2a, and U3a. Missing data points occur when the piezometertube contains no water ie.ir<0.4'1' (cm of water)0 -4^-10 0 .1^0501020? 30- 40:9- 50° 60708064Figure 21: i./. vs. Z (depth) for Sites Ula, Dla, Ulb, and Dib.Missing data points occur when the piezometer tube contains nowater ie.ir<0.• Site Dia ■ Site Ulb • Site Ula 0 Site Dlb50- 50-50^t oMar.16?-E. 50a0Mar18•1I0•Mar.29^ Mar.31 Apr.3? •of50-•//60Apr.5^ Apr.9••In50- ••5011AprI6AO65Figure 21 continued: 'T vs. Z (depth) for Sites Ula, Dla, Ulb,and Dib. Missing data points occur when the piezometer tubecontains no water ie.irp <0.• Site D1a ■ Site Ulb • Site Ula 0 Site Dlb66These graphs show that there is a general trend,especially in the upper portion of the profile, towards alower 1T in the drained low sites (Dia and D2a), and a shorterperiod when free water is present, when compared with thedepressional sites in Farm U. The differences in the depthand duration of ponding certainly accounts for part of this.As long as ponded water persists there is a source of watercontinually supplying the surface soil, keeping it saturated.This is aggravated by the fact that the supply rate isgoverned by the K, of the surface layer which is low, to adegree dependent on the severity of surface crusting. Thesurface crust in the drained sites was less severe, and wasalso extensively perforated by worm holes, and in someinstances breached by cracks, which meant that the supply rateof water into the soil was higher. Water was also moreeffectively removed by the system of subsurface drains in FarmD and this also accounted for the drier conditions.Figures 20 and 21 clearly show that decreases in AFTwithin the profile of the drained soil after rain are muchmore rapid than in the undrained. Figure 22 illustrates thatthis also applies to the high sites, and in both casesdecreases in 4%. are also faster in the drained, ie. there aregreater fluctuations in 4r.r. This may be partially due to thehydraulic contact between the surface and the upper profile inthe undrained being poor due to the presence of a surfacecrust. The direct addition of rain waters, as well as surface67runoff, added to the reservoir of ponded water in depressions.This ponded water receded slowly through evaporation andinfiltration through the crust. The differences between thedrained and undrained farms appears to be due mainly to themuch shorter periods of ponding on the drained farm.There was a general trend to hydraulic gradients of nearzero in the low areas. In the drained farm at site D2a thewater levels in the piezometer nest rose and fell with onlyvery slight vertical gradients indicating good hydrauliccontact within the profile. In the undrained sites verticalgradients of up to 0.6 developed in the lower profile duringdry periods when water was ponded on the soil surface. This isboth an indication of the effect of the surface water on theenergy state of the soil water in the upper profile and of theinfluence of the regional water table as transmitted throughthe underlying sandy. strata. In late March the 4%. below 0.20m depth at site U2a is larger negative than at D2a. This is areflection of the poor hydraulic contact between the surfaceand the soil profile and the consequent lack of response torainfall on March 29, 30, and 31st at U2a.While the data clearly indicate the influence of thesurface crust on soil hydrologic behaviour in Farm U, thereis no consistent evidence of there being other soil layers,for example a ploughpan, restricting vertical water movement.Any influence the ploughpan may have is masked by the effectsof the layer at the soil surface which most restricts water68flow. It is the surface crust on which attention must befocused.(ii) High Sites Figure 22 contains the irT vs Z graphs for the high sites.It is clear that the decline of 4,1, in the profile after wetperiods is faster in the drained sites. This was likely due tothe more effective removal of water from the soil bysubsurface drainage. Hence there was a longer period whenfree water is present in the upper 0.50 m of the profile inundrained high areas and therefore fewer opportunity days.There was a trend towards higher 4 ,1, in the profile at D2cthan at Dld. This may have been a result of D2c beingsituated where the drain lines fan out and the drain spacingis consequently greater. This was most evident during wetterperiods.6. Water table - Trafficability Relationships Paul and deVries (1979) used traction efficiency as acriterion for trafficability, and cited general agreementamong researchers that maximum traction efficiency is achievedat about 20% wheelslip. They used this relationship toestablish critical water table levels for trafficability.It was found that critical water table levels were 0.45m and 0.60 m for Hallart SiCL for grassland and cultivatedconditions respectively. Steinhardt and Trafford (1974) alsorecommended water table depth of 0.50-0.60 m as being required69Figure 22: 4,./. vs. Z (depth) graphs for Sites D2c, Dld, U3c,and Ule. Missing data points occur when the peizometer tubecontains no water ie.irp <0.70Figure 22 continued: iri. vs. Z (depth) graphs for Sites D2c,Dld, U3c, and Ule. Missing data points occur when thepeizometer tube contains no water ie.ir p <0.71to minimize structural damage of clay soils in England. Inanalyzing the water table data collected in this study it willbe assumed that the critical water-table level fortrafficability is 0.50 m. It will be further assumed that thewater table is located at the highest point where 'I' equalszero in a piezometer, or on the soil surface. The water tableis defined as that level in the soil at which the hydrostaticpressure of soil water is zero Childs, 1969). It should benoted that Paul and deVries measured the water table usingopen wells and not with piezometers.Using the trafficability-water table assumption discussedabove, Figure 23 shows when each site had ponded water on thesoil surface, was not trafficable, and those days on which thesite was trafficable. Figure 24 focuses on some of the pointsillustrated in Figure 23 for the fifty-five day period fromFebruary 12 to April 7 when the majority of sites were in"operation". The concept of "opportunity day" is introducedand used to illustrate differences in management optionsresulting from soil-water phenomena. An opportunity day isone on which a site is trafficable based on the critical watertable depth, has no water ponded on the soil surface, and hasnot recorded any rainfall. The latter is included because itis assumed that the farmer is a good manager who would notattempt tillage or traffic on a day that it has rained, due towet conditions in the surface layer and the consequent risk ofsoil structure degradation, ie. soil is untrafficable. ThisH) rt 0 itO 0"1:1 ,-'M •110ttN ti ^'!• rt rt MM 0 0• "L.)co^ft '•"CI `'‹la,^mp. mO h1 f D 01:5 fu 1 11PI fil 1-" rtI-1. 0 01-, 1E1 cl. 0o 14*,i o tr•^,rt).- m --•ft rtoi.i m tim m OP0. 0 1.<74is assumed to be the case even when there is no ponded waterand the water table is below 0.50 m.Figure 23 shows that there were consistently moretrafficable days at drained sites than at comparableundrained sites between February 12 and April 7. There werefewer days where ponded water was present on the surface ofdepressional drained sites and there also was a higherproportion of non-trafficable days, where ponded water was notpresent, in the drained farm. This points to the significanceof ponded water resulting from surface crusting in prolongingperiods of untrafficable conditions on the undrained sites.The number of trafficable days, and also of opportunitydays, is maximum in the highest sites and minimum in thedepressional sites. The difference in the number ofopportunity days between high and depressional locations meansthat a portion of the opportunity days in higher locations are"lost" to primary tillage operations such as ploughing. Theseopportunity days would not be lost to tillage operations likediscing (which commonly precedes ploughing in the Lower FraserValley) or to manure spreading, both of which could moreeasily be carried out while bypassing untrafficabledepressions.It is interesting to note that while site Dld had twomore opportunity days available than at Ule, all opportunitydays at Ule were "lost" for an operation such as ploughing dueto untrafficable conditions in the depression while there75were six "full" or "unrestricted" opportunity days at Dld.It is obviously desirable to maximize the number of"full" opportunity days since this leaves open the widestrange of management options. The soil and water managementpractices, particularly the system of subsurface drains andthe associated water removal system, on the drained farm havebeen successful in providing six and twelve "full" opportunitydays at sites D1 and D2 respectively. All "full" opportunitydays (FODs) occurred between March 11 and March 24. FODsduring this period are generally more important than ones inFebruary for the farms studied since primary tillageoperations and planting of early potatoes (an important cropon both farms) could reasonably begin in March. It should benoted, however, that fertilization of the overwintercauliflower and cabbage crops on the drained farm was carriedout in February to correct nutrient deficiencies soopportunity days can be significant throughout the year.There were no full opportunity days at any of theundrained sites from February 12 to April 7 either due toponding, untrafficable conditions based on the criticalwater-table levels, or both. This does not mean the drainedsites physically could not be ploughed, it simply suggeststhat there is a greater potential for compaction and structuredegradation if they were ploughed on non-opportunity days,especially in the depressional areas which compriseapproximately 10 to 15 percent of the field.76It is significant not only that full opportunity dayswere obtained only in Farm D but also that the lag timebetween the high and depressional areas becoming trafficablewas short. It is highly desirable that this time beminimized, for example, to avoid a situation where ploughingtakes place when the depressions are trafficable, but the highareas have become too dry. While it is apparent that Farm Uwould benefit from installation of a drainage system in thisregard, drained land would also benefit from any changes inmanagement which would expand the period of full opportunitydays. In the case of the soil being studied any managementpractices in addition to drainage that would further reducecrusting, runoff, and ponding would be desirable.The experimental results clearly illustrate thehydrologic results of topographic variations at the WesthamIsland sites. There is an uneven distribution of infiltrationwater due to runoff, followed by its accumulation indepressions. As a result there is great spatial variabilityof trafficable conditions, ponding and opportunity days onboth farms. The effect of a subsurface drainage system is, ona relative scale, to either insufficiently drain depressionalareas or excessively drain high areas to a degree largelydependent on weather conditions. It seems that most of theseproblems could be lessened, or eliminated, if systematicland-levelling were undertaken. This would involve strippingoff the existing Ap horizon and adjusting the levels of the77subsoil materials followed by respreading of the Ap soil. Atno point would it be acceptable to bring saline and acidicsubsoil closer to the final levelled land surface, than it wasprior to levelling. Although low areas comprise at most 10%to 15% of the land area, their elimination through landlevelling would appear to be worthwhile.If anything was surprising in the water-table data it wasnot that there were differences between the drained and theundrained farms, rather it was that the differences,especially after cessation of ponding at site U1 as comparedto site D1, were less extreme than anticipated. This may beaccounted for by one or more of the following reasons:(i) There were indications that the drainage system inFarm D was not functioning at desired efficiency levels duringthe study period. It was noted that the pump would shut offfor over ten minutes during the wettest periods while at thesame time ditch water-levels farther out in the field remainedabove the drain outlets as illustrated in Figure 25. The pumpswitched on for about seven minutes so in effect the capacityof the pump wasn't even tested since it was shut off more thanhalf of the time. As a result, during peak rainfall periodswhen it was needed most the drainlines were not flowingfreely, although they were always observed to be flowing tosome extent.This appeared to be caused by a restriction of flow inthe ditch due to the presence of too much sediment and weeds,78and culverts being too small and possibly high. The level ofthe pump intake may not have been low enough as well. Thedamlocated at the pump also seemed to be leaking since waterwas observed welling up into the ditch when the water level inthe adjacent drainage canal was high at high tide. Remedyingthese problems is recommended.(ii) The Westham soil is underlain by sand at a depthaveraging about 1 m. The thickness of this sandy layer isnot known but it is thought to be several metres. The effectof having this permeable sand underlying the entire area maybe to counteract or retard some of the beneficial effects ofthe drainage system in Farm D. Water may flow upwards intothe "drained" surface soil in the drained farm from this sandylayer. The driving force would result from the elevationdifference between the regional water table and theartificially lowered water table in Farm D. This may be oneof the problems in attempting to improve drainage on a spotbasis in an area where the regional drainage is inadequate. Ineffect a portion of the water being pumped from the ditch inFarm D may have originated from below the drainlines ratherthan the soil above. The influence of the regional watertable appeared when the recession of the water table seemed toslow down when the latter reached approximately 0.50 m indepressions.^Further research into the nature of theseinteractions is required.(iii) Large scale "macro-cracks", 0.02 m or more wide79extending over tens of metres, and up to 1 m into the soilhave been observed in the Westham soil. These macro-cracksform as a result of soil shrinkage during drying and theirlocation and orientation seem to be influenced by tillage andcropping patterns eg. some cracks were located between, andparallel to potato rows. To the extent that they remain open,these macro-cracks provide an important lateral pathway fortransporting ground water to drainlines. They may accountfor part of the differences between drained and undrained landsince, by keeping moisture levels down, a drainage systemimproves the chances of the macro-cracks remaining at leastpartially open while at the same time providing a pathway (thedrain-line) for removal of groundwater from the cracks. Thereis a "positive feedback loop" in that, as long as they areopen, the macro-cracks can transfer water to the drain-line.It was suggested by the farmer (Bob Reynolds) that themacro-cracks had closed up in Farm D the winter of the studyas a result of the persistent wet weather that began inmid-October 1980, thereby removing their benefits. The wetweather likely combined with problems with the functioning ofthe drainage system discussed above to bring this situationabout. The issue of macro-cracks and their effects on soildrainage is highly speculative at this point, and furtherresearch into the question seems worthwhile.807. Ditch Water Levels The changes in ditch water levels with rainfall areillustrated in Figure 25. The generally faster and greaterrise and fall in the ditch water level in Farm D is areflection of the more efficient delivery of water through thesubsurface drainage system. Ideally, however, the water levelin the Farm D drainage ditch should remain fairly constant andcertainly not rise above the level of the drain outlets as itdid after March 29. This points to problems with thedrainage system in Farm D, and these are discussed later.The smaller rise in the water level in the Farm D ditchis likely due to the drier conditions there which resulted ina greater portion of the rain going into storage. It mayalso be the case that water was delivered to the ditch andpump at a slower rate which permitted more efficient pumping.8. Profile Hydraulic ConductivityThe data from the core method are shown in Table 4, andthe auger-hole K, data are presented in Table 5.The core data indicate higher mean K, values on thedrained site, 3.4 times greater than the undrained subsoiledand 5.0 times greater than the undrained unsubsoiled. Thisdifference is attributable to the more abundant macropores,especially worm burrows, which provided preferred pathways forvertical water movement on the drained site. Differences inworm populations are discussed in Section 11.hiH.trl01MNJtn••0I-.•rt0E03rt(1)1I--,(D4.76mm<1.00mm 7.90 64.11 63.11 68.89 66.21 16.82 44.84 59.02 71.95 42.50 60.64 22.33 460.70>2.00mm<4.76mrn 17.30 22.31 19.21 17.66 18.68 29.91 27.53 17.86 17.49 23.38 21.13 4.63 18.411DRAINED >1.16mm<2.00mm 8.40 2.44 1.20 2.63 2.67 6.27 4.47 10.77 2.23 3.32 4.44 3.06 8.61>.219mm<1.13mm 27.70 4.05 1.79 4.04 3.19 3.43 8.43 0.035 4.69 7.42 6.48 7.14 66.37>.219mm (dliforono0 38.70 7.09 14.69 6.87 9.25 43.57 76.92 12.31 3.64 23.38 23.63 23.13 433.66Site DI Weproesions9 >4.76mm<11.00mm 37.20 55.02 30.01 70.31 35.13 37.44 50.30 64.14 72.83 63.84 61.71 16.39 227.33>2.00mm<4.75innt 22.70 20.91 22.14 17.34 12.23 24.95 25.05 12.06 18.96 27.42 20.40 6.26 24.86DRAINED >1.111mnto:200nn 7.70 2.22 4.91 2.51 1.96 5.78 5.32 4.32 2.91 2.53 4.02 1.00 3.24>.2111mm<1.111mm 15.10 6.70 10.01 4.40 0.42 14.13 10.39 6.75 5.43 4.70 7.30 4.63 11.86>.219mm (d1ffrorenco) 17.30 14.25 32.73 5.40 50.26 17.70 8.94 0.73 0.00 1.51 14.31 16.02 231.01Site UI 91..) >4.76fiont<1.00mm 43.34 40.59 42.02 14.20 27.17 9.67 19.37 27.03 64.95 41.13 3E96 20.39 301.96>200mm<4.76nwe 15.40 22.44 25.96 14.72 3286 15.46 23.56 11.14 16.70 28.14 20.64 7.00 44.09UNDRAINED >1.111mnt000mm 0.60 3.40 4.51 6.24 6.56 2.80 7.90 7.11 4.08 4.69 4.79 222 4.43>.219mm<1.lSmm 0.80 5.55 6.55 16.87 11.22 12.09 12.61 12.62 6.61 7.28 9.22 4.67 19.63>.219mm (d11111woneo) 9.84 23.02 25.57 47.97 22.19 59.98 36.54 42.10 7.66 18.76 29.36 16.39 261.30Site UI WoOrmiena0 >4.76mm<1.00mm 53.77 4520 38.71 59.45 46.40 27.06 56.23 79.10 58.15 .66.52 63.06 14.62 192.46>2.00mm<4.76mm 21.75 24.60 21.74 24.87 13.80 23.61 18.03 20.20 26.86 19.10 21.41 3.34 13.30UNDRAINED >1.111mm000mm 3.29 %3.20 4.19 2.52 1.11 9.20 5.23 1.51 2.22 2.76 3.31 2.36 4.95>.219rnm<1.111mm 6.47 6.50 8.77 4.40 1.8.4 22.69 3.41 3.67 4.09 5.67 6.76 6.93 31.63>.219mm (difforonera 11.72 20.50 26.59 8.86 37.95 17.44 17.01 0.00 8.68 5.05 16.76 10.06 106.14DI riso fravimoirle 0297 0.314 0.287 0.332 0.329 0.322 0.333 0.317 0.328 0.300 0.316 0.0162 0.0020 CorrelationDi dopressional molaturo coniont 0.343 0.361 0.311 0.380 -- 0.361 0.368 0.351 0.332 0.337 0.364 0.246 0.0006 C" 111141•ntUI Hs* of oomph, 0.268 0.296 0.305 0.298 0.327 0.263 0.251 0.273 0.290 0.309 0.285 0.0237 0.0006 MWDU1 dowssolonal kg vist.d kg soil 0.327 0.372 0.326 0.316 0.271 0.294 0.321 0.345 0.338 0.327 0.324 0.0274 0.0007 vsMoistureContontDi rise Moan . 1.46 4.91 4.70 5.06 5.59 221 3.92 4.54 5.25 3.60 4.124 1.364 1.649 0.374DI dopresslonal Wolght 3.37 4.35 2.82 5.14 2.68 3.42 4.21 4.61 5.37 5.07 4.104 0.979 0.862 0.196U1 Moo Diameter 5.21 3.44 3.67 1.62 3.02 1.27 2.25 2.30 4.82 3.70 3.130 1.296 1.611 0-206U1 &pre ssiona I (mm) 4.25 3.81 3.33 4.70 3.45 2.83 4.30 5.77 4.68 4.97 4.210 0.873 0.690 0.46598agreement with the differences in mean water content, since alower Om would result in more complete aggregate breakdown viaslaking. The soil O m, however, shows only a weak negativecorrelation (r2=0.073) with the proportion of sample less than212 Am when all forty samples are pooled. This indicates thatfactors other than Om account for most of the variations inthe MWD results. It may be that some of the aggregates foundin the low ponded areas are denser and more stable becausethey have been formed by the mechanical breakdown of largercompact clods. The likelihood that tillage or other fieldoperations are carried out when the soil is not trafficable isgreatest in the wetter depressions, and therefore so is theformation of compact soil clods. If present, these denser andmore stable aggregates would tend to counter-balance theeffects of unstable aggregates in calculating MWD values.The wet-sieving measurements show that land drainage andthe higher earthworm populations on Farm D have produced nosignificant increase in the water-stability of soilaggregates. The early laboratory research which indicated thatearthworms contribute to water-stable aggregate formation hasnot been born out in the fields studied on Westham Island.It is not clear what effects, if any, the technique usedhad on the water-stable aggregate results. Further researchinto the wet-sieving procedure as a method of evaluation thestructural stability of field soils seems necessary. The aimshould be to increase the meaningfulness and reproducibility99of results.13. Penetration Resistance and CompactionThe cone penetrometer is a widely used tool in soilcompaction research, despite the fact that the theory behindit is not well understood. Frietag (1971) presents datashowing near linear relationships between cone index and soilwater content in the 0.25-0.35 kg H 20/kg soil range.The hypothesis being tested was that better drained andhence drier soil undergoes less compaction during normaltillage. The penetrometer measurements shown in Figure 27(raw data are in Appendix 3) were made in the low areas atsites Ul, U2, Dl, and D2. It was felt that any differences inthe degree of compaction would be most apparent there sincewater management practices contrasted most sharply in lowareas. It should be noted, however, that ill-timed vehicletraffic on soil can cause severe compaction even under thebest water management regime.The mean penetration resistance (30 replicates) for siteUl is significantly higher (5% level) than the drained sitenearby, site D1, at the 0.05, 0.10, 0.15, 0.20, 0.25, and 0.30m depths. Site U2 had mean penetration resistance values thatwere significantly higher (5% level) from those at near bysite D2 only at the 0.20 and 0.25 m depths. It is interestingthat below 0.35 m the penetration resistance profiles aresimilar, and that they lie below the depth range strongly100Figure 27: Penetration resistance profiles. Each data pointis the mean of 30 replicates.101influenced^by^ordinary^tillage. The^highest meanpenetration resistance at each site was measured atdepth of 0.25 m. This corresponds to the ploughpan depthqualitatively identified in the field, although, as theprofiles in Figure 27 illustrate, the ploughpan is actually azone of more compact soil generally between 0.15 and 0.30 mbelow the surface. The profiles for Farm D, though notidentical, closely follow each other and the same can be saidfor those from Farm U. The penetration resistance profilesagree with the hypothesis that the better drained sites ofFarm D have less compacted ploughpan zones. A preciseinterpretation of these results is hampered however, bydifferences in water content at the time of measurement andlack of knowledge about the relationship between penetrationresistance and water content.14. Effects of Moling, Subsoiling, and Subsurface Drainage onthe Water Regime of Depressional Areas MolingIf mole drainage channels are properly created and thenpersist for some time they may aggravate excess waterconditions even on gently undulating land by directingsubsurface water into depressions from surrounding higherareas. Since mole drains are installed at a fixed depthrelative to the surface on which the tractor rides, water isalways directed into and never out of low areas as shown in102Figure 28. One strategically located low spot maysignificantly reduce or halt flow in a "properly" installedmole drain. This would be much less of a problem in drainedland where subsurface pipes would remove any water sodelivered to a low spot or where mole drain collapse preventswater from flowing.No mole drains were encountered in all the digging thatwas done in farm U, nor were any seen transferring water intothe ditch. The soil texture, along with these observations,indicates that mole drains are quite shortlived in this soiland are therefore an ineffective means of dealing with excesswater. Where they do persist, mole drains would cause theproblems mentioned above.SubsoilinaWithout subsoiling there might be lateral flow of waterinto low areas on top of the ploughpan. With subsoiling theremay be flow along the bottom of subsoiler slits into low areassimilar to the case with mole drains. Subsoiling may alsoresult in increased infiltration and vertical water movementsand hence less flow from high to low areas.Subsurface Drainage Plastic subsurface drainpipes are laid with laser-guidedmachines to give a slight grade from the most distant part ofthe field to the ditch. The drain-pipe depth is not constantbut is inevitably at its shallowest relative to the surface inlow areas. As a consequence of this, low areas receive the103Figure 28: Effect of mole draining at a constant depth belowthe soil surface. Arrows indicate the direction of water flowin the mole channel.104least direct benefit from subsurface drainage. Indirectbenefits may derive from improvements in higher areas due todrainage which lead to decreases in runoff and through-flow.Low areas tend to be more poorly drained than surroundinghigher areas partially as a result of the above. Even if otherfactors are not operating, where there is a "flat" watertable it will be closest to the surface in low areas.15. Surface Ditching to Remove Ponded Water Observations in the field indicated that surface ditchesare generally not an efficient means of channeling away pondedwater to deep ditches. The gently undulating topography inthe sites, and in Delta area in general, make it difficult toconstruct a surface ditch with an adequate grade for waterremoval as shown in Figure 29. The surface ditch illustratedin Figure 29 was constructed using a mouldboard plough; otherimplements designed specifically for the purpose producesimilar results. Trenching equipment is sometimes used and ismore successful because of better depth control.Surface ditches suffer from two main problems even whenan adequate grade can be obtained: (1) the soil thrown up oneither side of the ditch often prevents ponded water fromentering laterally; depending on where the surface ditchbegins only part of the ponded water may be removed, and (2)inadequate water removal on a regional scale results in highwater levels in deep ditches often to the point that water105Figure 29: A surface ditch installed in an attempt to removeponded water was unsuccessful due to an inability to establishan adequate gradient.106flows into the fields rather than outward.Surface ditches, as a remedial measure to deal with theccumulation of ponded water, would be most successful if theyemptied into deep ditches whose water level is adequatelycontrolled for agricultural requirements. A more effectivealternative may be to construct a sump and use a portablegasoline pump to remove ponded water. It is recommended thatpumping be tested in the field.107Chapter 4SUMMARY AND CONCLUSIONSThe effectiveness of contrasting soil and water managementpracticed on adjacent farms with the same soil type was evaluated.The study involved measurement of soil physical and biologicalproperties in the field and laboratory.Land drainage, through a system of subsurface perforated tubesthat empty into a ditch whose water level was controlled bypumping, was found to be superior to the use of mole-drains,subsoiling and surface drainage used in combination. Land drainagewas also more efficient in lowering the water table after rainfall,and thereby providing aerated conditions in the surface layer whichbenefits survival of crops and soil fauna, especially earthworms.Drainage reduced the depth and duration of ponding and resulted inmore trafficable and opportunity days. There were no fullopportunity days during the period of measurements in the undrainedfarm due to persistant high water table levels in depressions whilethere were six to twelve full opportunity days in the drained farmduring the same period.There was a more severe ploughpan in the undrained farm andthis was thought to reflect wetter conditions there for comparableplowing dates. There was no evidence that the ploughpan impededvertical water movement in either farm. Measurements of thehydraulic conductivity of the ploughpan and the lower profile108indicated that their K, was too high to significantly retard watermovement under saturated conditions.There was no measured difference in water-stable aggregatesbetween the two farms. Greater water-stable aggregation indepressions was thought to be due to the physical break-up ofcompact clods. Further research into the wet-sieving method wasrecommended.The presence of ponded water in depressions was found to be amajor problem. It was shown that flooding can occur due to thewater table rising above the soil surface. In most cases, pondingwas due to the presence of a slowly permeable surface crust. Inhigh areas the crust reduced infiltration rates and resulted inrunoff into depressions where ponded water accumulated. In theworst instance the crust K, was measured to be 0.002 m/day. Thismeant that on sunny days removal of ponded water was roughlyequally partitioned between drainage and evaporation. The presenceof high water tables and wet conditions in depressions createdconditions favorable to crust formation.In the undrained farm growth of a cover crop reduced theseverity of crusting and hence of ponding. In the drained farmponding was less severe even without a cover crop. This was aresult of the extensive perforation of the surface crust byearthworms which were abundant due to the more favorable aeratedconditions there. Earthworms were scarce in the undrained farm.Earthworm burrows provided preferred pathways for water to bypassthe crust thereby reducing runoff and ponding.109Surface runoff was also reduced in both farms by crackingwhich breached the crust as the soil dried. Surface cracking wasinfluenced by surface geometry, as was earthworm survival duringponding. Further research into the nature of these influences wasrecommended.It was suggested that subsoiling and mole drainage mayaggravate wet conditions in depressions when their channels areinstalled at a constant depth relative to the soil surface.Subsoiling to break up the ploughpan had no benefit with respect toponding where a crust subsequently formed on the soil surface. Thespacing of subsoiler blades in Farm U was found to be too great.Subsoiling in the absence of measures to control crusting and watertable levels should be discouraged as it is relatively ineffectivewhen applied in isolation.Surface ditches were found to be ineffective in removal ofponded water due to poor installation techniques and topographicconstraints. Even where they are well installed they may channelwater into fields where the adjacent regional water management ispoor.Land drainage is recommended as the most effective system formanaging soil water in poorly drained soils. However, even a welldesigned drainage system will not be fully successful if there isnot an ongoing program to manage the system, including the soil, tomaintain and promote optimum efficiency levels. Land leveling wassuggested as a means of improving the efficiency of a drainagesystem. Good regional water management is required before a110drainage system can provide all its potential benefits.111BIBLIOGRAPHYAbbaspour, K. 1988. Hydrologic Responsiveness of a LowerFraser Valley Lowland Soil. Unpublished M.Sc. Thesis.Department of Soil Science, University of BritishColumbia. Vancouver.Armstrong, A.C. 1987. The effect of drainage treatments oncereal yields: results from experiments on clay lands. J.Agricultural Science, Cambridge. 91:229-235.Atmospheric Environment Service. 1981. UBC Climate Station,Monthly Meteorological Summary. Environment Canada,Vancouver.Atmospheric Environment Service. 1984. Canadian ClimaticNormals, 1951-1980. Environment Canada, Canadian ClimateProgram, Ottawa.Blake, G.R. 1965. Bulk Density. In Black, C.A. (ed.) Methodsof Soil Analysis. American Society of Agronomy, Inc.Madison, Wisconsin.Boersma, L.^1965.^Field measurement of hydraulicconductivity below a water table. In Black, C.A. (ed.)Methods of Soil Analysis. American Society of Agronomy,Inc. Madison, Wisconsin.Bonsu, M. 1984. Hydrologic Behavior of Some StructurallyDegraded Organic Soils. Unpublished Ph.D. Thesis.Department of Soil Science, University of BritishColumbia. Vancouver.Bouma, J. 1978. The drainability of some Dutch knick soils.Agr. Water Mngt. 1:1-9.Bouma, J. 1981. Soil morphology and preferential flow alongmacropores. Agr. Water Manage. 3:235-250.Bouma, J., Jongerius, A.,Boersma, 0., Jager, A. andSchoonderbeek, D. 1977. The function of different typesof macropores during saturated flow through four swellingsoil horizons. Soil Sci. Soc. Am. J., 41:945- 950.Bouwer, H. 1965. Drainage for Efficient Crop ProductionConf., Amer. Soc. Agr. Eng., Pro. p.62-65.Bouwer, H. 1974. Developing drainage design criteria. InVan Schilfgaarde, J. (ed.) Drainage for Agriculture.American Society of Agronomy, Inc. Madison, Wisconsin.112Braun, M.A., Chaney, K., and Hodgson, D.R. 1984. Preliminaryinvestigation on the response of spring barley to soilcultivation with the Paraplow. Soil and Tillage Res.4:277-293.Buras, N.^1974.^Water management systems.^In VanSchilfgaarde, J. (ed.) Drainage for Agriculture.American Society of Agronomy, Inc. Madison, Wisconsin.Carter, A., Heinonen, J. and deVries, J. 1981. How to makeearthworms happy. Country Life in B.C.Childs, E.C. 1969. An Introduction to the Physical Basis ofSoil Water Phenomena. Wiley (Interscience), New York.Dawson, R.C. 1947. Earthworm micro-biology and the formationof water-stable soil aggregates. Soil Sci. Soc. Proc.512-516.deVries, J., and Paul,C.L. 1978. A Preliminary Report on theEffects of Drainage on Soil Physical Properties ofLowland Farms in Delta. Unpublished.Dielman, P.J. 1979. Drainage knowledge and research needs indeveloping countries. In Wesseling, J. (ed.) Proceedingsof the International Drainage Workshop,ILRI Publ. 25,Wageningen, The Netherlands.Dutt, A.K. 1948. Earthworms and soil aggregation. J. Amer.Soc. Agron. 40:407-410.Edwards, C.A., and Lofty, J.R. 1972. Biology of Earthworms.Chapman and Hall, London.Evans, A.C. 1948. Studies on the relationships betweenearthworms and soil fertility II: Some effects ofearthworms on soil structure. App. Biol. 35,1-13.Fausey, N.R. and Schwab, G.O. 1969. Soil moisture content,tilth, and soybean response with surface and subsurfacedrainage. Agron. J. 61:554-557.Freitag, D.R. 1971. Methods of measuring soil compaction. InCompaction of Agricultural Soils. American Society ofAgricultural Engineers, St. Joseph, Michigan.Guild, W.J. McL. 1948. Studies on the relationships betweenearthworms and soil fertility III: The effect of soiltype on the structure of earthworm populations. Ann.Appl. Biol. 35, 181-192.113Guild, W.J. McL. 1955. Earthworms and soil structure. InMcEwan (ed.)^Soil Zoology: Proceedings of theUniversity of Nottingham 2nd Easter School Agric. Sci.,1955. Butterworths, London.Hillel, D. 1960. Crust formation on loessal soils. Trans.Int. Soil Sci. Congr. Madison 7th 1:330-339.Hillel, D. 1971. Soil and Water: Physical Principles andProcesses. Academic Press, New York.Hopp, H. and Hopkins, H.J. 1946. Earthworms as a factor inthe formation of water stable-aggregates. J. Soil, WaterCons., 1:11-13.Kay, B.D. 1990. Rates of change of soil structure underdifferent cropping systems. in Advances in Soil ScienceVolume 12 B.A. Stewart (ed.) pp. 1-52, Springer-Verlag,New York.Kemper, W.D. 1965. Aggregate stability. In Black, C.A.(ed.) Methods of Soil Analysis. American Society ofAgronomy, Inc. Madison, Wisconsin.Kemper, W.D. and W.S. Chepil. 1965. Size distribution ofaggregates. In Black, C.A. (ed.) Methods of SoilAnalysis. American Society of Agronomy, Inc. Madison,Wisconsin.Leyton, L. and Yadav, J.S.P. 1960. Effect of drainage oncertain physical properties of a heavy clay soil. J.Soil Sci. 11(2):305-312.Luthin, J.N. (ed.)^1957.^Drainage of AgriculturalLands.Amer. Soc. Agron. Monograph 7.Luttmerding, H.A. 1980. Soils of the Langley-Vancouver MapArea. Report No. 15 Volume 1: Soil Map Mosaics andLegend, Lowere Fraser Valley (Scale 1:25,000) BritishColumbia Soil Survey. B.C. Ministry of Environment.Victoria.Luttmerding, H.A. 1981a. Soils of the Langley-Vancouver MapArea.Report No.15 Volume 3:Description of the Soils.British Columbia Soil Survey.B.C. Ministry ofEnvironment. Victoria.Luttmerding, H.A. 1981b. Soils of the Langley-Vancouver MapArea.Report No.15 Volume 6: Technical Data - SoilProfile Descriptions and Analytical Data. BritishColumbia Soil Survey.B.C. Ministry of Environment.Victoria.114Marshall, T.J. and Holmes, J.W. 1980. Soil Physics. 345pp.Cambridge University Press, New York.McIntyre, D.S. 1958. Soil splash and , the formation ofsurface crusts by raindrop impact. Soil Sci. 85,261-266.Morris, J. and Calvert, J.T. 1976. Drainage Investment:Measuring the Return. ADAS Q. Rev. 20:151-161.Paul, C.L.^1978. Prediction of Trafficability of TileDrained Farmland. Ph.D. Thesis. University of BritishColumbia, Vancouver.Paul, C.L. and deVries, J. 1979(a). Prediction of soilstrength from hydrologic and mechanical properties. Can.J. Soil Sci. 59:301-311.Paul, C.L. and deVries, J. 1979(b). Effect of soil waterstatus and strength on trafficablity. Can. J. Soil Sci.59:313-324.Raghavan, G.S.V., McKyes, E. and Beaulieu, B. 1978. Claysoil compaction due to wheel slip. Trans. ASAE.21:646-649.Reeve, R.C. and N.R. Fausey. 1974. Drainage and timelinessof farming operations.^In Van Schilfgaarde, J.(ed.)Drainage for Agriculture. American Society ofAgronomy, Inc. Madison, Wisconsin.Russell, E.W.^1973.^Soil Conditions and Plant Growth.Longman. London.Shkinkis, C.N.^1979.^Subsurface drainage results forover-wetted soils of the Latvian S.S.R. Agr. Water Mngt.2:55-65.Sieben, W.H. 1974. On the influence of drainage on nitrogendelivery and yields on reclaimed loamy soils in theIjssel Lake Polders. In Feddes and van Wijk. 1976.Influence of drainage on crop yield: An integrated modelapproach. ICW, Wageningen, the Netherlands.Soehne, W.H. 1958. Fundamentals of pressure distribution andcompaction under tractor tires. Agr. Eng. 39:276.Spoor, G., Leeds-Harrison, P.B., and Godwin, R.J.^1982.Potential role of soil density and clay mineralogy inassessing the suitability of soils for mole drainage. J.Soil Sci. 33:427-441.115Steinhardt, R. and Trafford, B.D. 1974. Some aspects ofsubsurface drainage and ploughing on the structure andcompactability of a clay soil. J. Soil Sci. 25(2):138-152.Swain, R.W. 1975. Subsoiling. In Soil Physical Conditonsand Crop Production, Tech. Bull. 29,00:189-204. Min.Agr. Fish. Food. HMSO, London.Trafford, B.D. and Oliphant, J.M.^1977. The effect ofdifferent drainage systems on soil conditions and cropyield of a heavy clay soil. Expl. Husb. 32,75-85.van Beers, W.F.J. 1963. The Auger-Hole Method. IILRIBulletin 1, H. Veenman & Zonen. Wageningen, TheNetherlands.van Lanen, H.A.J., Reinds, G.J., Boersma, 0.H., and Bouma, J.1992. Impact of soil management systems on soil structureand physical properties in a clay loam soil, and thesimulated effects on water deficits, soil aeration andworkability. Soil and Tillage Research, 23:203-220.van Wijk, A.L.M., and Feddes, R.A. 1986. Simulating effects ofsoil type and drainage on arable crop yield. in A.L.M.van Wijk and J. Wesseling (Editors), Proceedings of theEC-Workshop on Agricultural Water Management. Balkema,Rotterdam, pp. 96-112.Victoria, British Columbia. 1968. Information guide - FraserRiver flood control program. 27pp.Wells, A.A. 1956. Subsoiling. ADAS Report. Great Britain.Wesseling, J. and van Wijk, W.R.^1957.^Soil physicalconditions in relation to drain depth. In Luthin, N.J.(ed.) American Society of Agronomy, Inc. Madison,Wisconsin.Wind, G.P.^1975. Workability and drainage. NOTA 890.Instituit voor Cultuurtechniek en Waterhuishouding,Wageningen, The Netherlands.Winter, G.R. 1968. Agricultural development in the lowerFraser Valley. In Seimens, A.H. (ed.), Lower FraserValley: Evolution of a Cultural Landscape. TantalusResearch Ltd. Vancouver, British Columbia.Yoder, R.E. 1936. A direct method of aggregate analysis anda study of the physical nature of erosion losses. Am.Soc. Agron. 28:337-351.116APPENDIX 1 U.B.C. Climate Station Evaporation (EV)Dateand Bright SunshineFebruary to April,EV^BSS^EVmm^hrs^Date^mm(BSS)Data1981BSShrs^DateEVmmBSShrsFeb.1 0.4 0.3 Mar.1 0.7 7.1 Apr.1 3.4 7.12 0.6 5.3 2 1.2 5.8 2 2.0 1.03 0.4 7.5 3 0.8 0.0 3 2.9 10.34 0.4 7.5 4 1.8 8.5 4 0.3 0.05 0.4 8.4 5 1.4 7.6 5 2.8 5.86 0.6 6.5 6 2.4 6.4 6 2.8 7.97 0.0 0.0 7 0.2 0.0 7 2.8 4.98 0.2 7.7 8 1.2 8.3 8 1.0 3.39 2.4 6.7 9 2.6 9.5 9 2.2 4.610 1.3 8.7 10 2.0 8.8 10 1.9 0.011 1.2 0.1 11 1.8 7.7 11 0.8 3.712 0.1 0.0 12 1.6 9.0 12 0.8 5.813 1.8 0.0 13 3.6 8.1 13 2.4 7.214 0.0 1.1 14 0.9 2.1 14 2.8 12.215 0.2 0.0 15 0.4 0.0 15 2.4 0.416 1.0 3.0 16 2.4 9.0 16 4.3 10.717 0.4 0.0 17 2.4 9.7 17 8.2 11.718 1.2 0.0 18 5.0 10.0 18 4.3 12.519 1.2 2.1 19 3.6 9.8 19 4.7 12.420 1.2 0.0 20 4.0 8.9 20 0.7 0.021 0.1 0.1 21 3.0 6.4 21 0.4 0.022 1.0 3.8 22 3.2 7.6 22 0.4 0.023 0.8 0.8 23 2.2 4.9 23 1.6 0.825 0.2 0.0 25 2.8 8.5 25 3.9 11.326 0.6 0.7 26 2.4 10.3 26 3.0 12.227 2.5 9.3 27 2.2 0.3 27 1.2 0.028 1.9 8.7 28 0.8 0.0 28 0.8 0.029 0.2 0.0 29 2.0 0.430 1.0 0.0 30 2.0 4.131 1.8 3.8TOTAL 22.9 88.3 63.0 179.4 67.8 161.2MEAN 0.8 3.2 2.0 5.8 2.3 5.4117APPENDIX 2WESTHAM ISLAND RAINFALL DATARainfall (mm)Date (1981)^RG1^RG2^RG3^RG4^Mean^Feb. 13^13.3^13.3^----^----^13.314 10.3 9.9^11.2^10.1^10.415^11.7^11.9 8.0^8.4 10.016 14.6^----^18.6^22.1^18.417^5.7 4.4^2.2 2.7 3.718 19.8^18.8^16.4^19.9^18.719^19.6^14.5^16.4^20.4^17.721 7.4 6.6 7.1 7.1^11.525^16.8^17.9^15.5^15.9^16.527 2.8 3.0 4.0 3.8 3.4Mar. 3^18.1^17.9^18.1^17.3^17.97 0.9^---- 1.8 1.8 1.515^4.2 4.0^4.0^4.2^4.116 6.6^6.2 7.1 6.8 6.725^14.2^14.2^15.0^14.6^14.529 28.8^32.9^28.3^27.4^29.430^9.4 9.7 8.9 9.3 9.331 8.9^9.0^9.3^8.9^9.0Apr. 3^4.3^5.0^4.4^5.3^4.74 6.0 5.8 6.6 6.2 6.25^11.5^11.9^15.0^13.3^12.98 4.9^---- 5.3 5.0 5.110 0.9^0.9^0.913^----^----^8.8 8.0 8.416 ----^---- 0.9^0.9^0.9