the integrated farm system model - usda ars · the integrated farm system model ... models with a...

Title Page

THEINTEGRATEDFARMSYSTEMMODEL

ReferenceManualVersion4.4

C.AlanRotz,MichaelS.Corson,DawnS.Chianese,FelipeMontes,SashaD.Hafner,HenryF.BonifacioandColette

U.Coiner

PastureSystemsandWatershedManagementResearchUnit

AgriculturalResearchServiceUnitedStatesDepartmentofAgriculture

April2018

1 Table of ContentsTitle Page -3

1. Table of Contents -2-1

EXECUTIVE SUMMARY 2-4

INTRODUCTION 5-11

Model Overview 11-16

Figure 1.1 17

CROP AND SOIL INFORMATION 18-20

Alfalfa 20-22

Perennial Grass 22-27

Corn 27-30

Small Grain 30-31

Soybean 31-32

Table 2.1 33

GRAZING INFORMATION 34

Pasture Production 34-36

Pasture Equipment and Operations 36-37

Pasture Use 38-39

MACHINERY INFORMATION 40

Work Performance 40-43

Power Performance 43-45

Energy and Labor 45-47

Table 4.1 47

Table 4.2 48-49

TILLAGE AND PLANTING INFORMATION 50

Suitable Days 50-52

Tillage and Planting Operations 52-54

Table 5.1 55

CROP HARVEST INFORMATION 56

Forage Crops 56-68

Grain Crops 68-71

FEED STORAGE INFORMATION 72

Grain Storage 72

Dry Hay Storage 72-76

Silo Storage 76-83

HERD AND FEEDING INFORMATION 84-88

Dairy Herd 88-95

Beef Herd 95-103

Table 8.1 104

Table 8.2 105

Table 8.3 106

Table 8.4 107

Table 8.5 108

MANURE AND NUTRIENT INFORMATION 109

Manure Handling 109-112

Nutrient Balance 112

Manure Import and Export 112-114

Anaerobic Digestion 114-117

Manure Composting 118-125

Table 9.1- Model equations 126-128

Table 9.2 - Simulation settings 128

Figure 9.1 - Model flow for a compost windrow 129

Figure 9.2 - Compost Simulation Profiles 130

Figure 9.3 - Compost carbon (C) and nitrogen (N) flows 130

ENVIRONMENTAL IMPACT 131

Phosphorus Loss 131

Surface Phosphorus 131-133

Inorganic Soil Phosphorus 133-135

Organic Soil Phosphorus 135-136

Sediment Phosphorus and Erosion 136-138

Figure 10.1 138

Figure 10.2 139

Figure 10.3 139

Ammonia Emission 140

Formation and Emission Processes 140-144

Animal Housing 144-152

Manure Storage 152-154

Field Application 154-156

Grazing Animals 156-157

Table 11.1 - Bedding Material Properties 157

Hydrogen Sulfide Emission 158

Formation and Emission Processes 158-163

Enteric Emissions 163

Animal Housing 163-164

Manure Storage 164-165

Field Application 165

Grazing Animals 166

Figure 12.1 167

Figure 12.2 167

Greenhouse Gas Emission 168

Carbon Dioxide Emission 168-174

Methane Emission 174-182

Nitrous Oxide Emission 182-191

Table 13.1 192

Table 13.2 193

Table 13.3 194

Table 13.4 194

Figure 13.1 195

Figure 13.2 196

Figure 13.3 196

Figure 13.4 196

Volatile Organic Compound Emission 197

Silage Sources 197-203

Manure Sources 203-207

Table 14.1 - VOC Concentrations 208

Table 14.2 - EBIR Values 208

Table 14.3 - Manure VOCs 209

Table 14.4 - Volatility and Reactivity 209-211

Table 14.5 - Emission Parameters 211

Reference Manual | 0

Figure 14.1 - VOC Simulation Diagram 212

Figure 14.2 - VOC Concentration 213

Figure 14.3 - Emission Potential 214

Environmental Footprints 215-216

Water Use 216-218

Reactive Nitrogen Loss 218-220

Energy Use 220-221

Carbon Emission 221-224

Table 15.1-Resource input factors 225

ECONOMIC INFORMATION 226

Production Costs 226-232

Revenue and Net Return 233

Table 14.1 234

REFERENCES 235-250


EXECUTIVE SUMMARYWithtighterprofitmarginsandincreasingenvironmentalconstraints,strategicplanningoffarmproduction systems is becoming both more important and more difficult. This is especially true fordairy andbeef production. Livestock production is complex with a number of interacting processesthat include crop and pasture production, crop harvest, feed storage, grazing, feeding, and manurehandling.Computersimulationprovidesausefulprocedureforintegratingtheseprocessestopredictthelong-termperformance,environmentalimpact,andeconomicsofproductionsystems.Developmentofasimulationmodelofthedairyforagesystembeganintheearly1980s.Thismodel,knownastheDairyForageSystemModelorDAFOSYM,linkedalfalfaandcornproductionmodelswithadairyanimalintakemodeltopredict feedproductionanddisappearanceonthefarm.This model was expanded with additional components for simulating feed storage and animalperformance.Manurehandling,tillage,andplantingoperationswerethenaddedtoextendthemodelto a simulation of the full dairy farm. The dairy farm model was broadened further by addingcomponents for simulating grass, small grain, andsoybeangrowth, harvest, andstorage. Throughamajor revision, a beef animal component was added along with a crop farmoption (no animals) toformtheIntegratedFarmSystemModelorIFSM.Thismodelhascontinuedtogrowascomponentswere added to simulate environmental impacts including gas emissions, nitrate leaching, andphosphorus runoff and a life cycle assessment to determine the carbon footprint of productionsystems.Unlikemostfarmmodels,IFSMsimulatesallmajorfarmcomponentsonaprocesslevel.Thisenables theintegrationandlinkingof components inamanner that adequately represents themajorinteractions amongthe manybiological and physical processes on the farm. This provides a robustresearch and teaching tool for exploring the whole farm impact of changes in management andtechnology. Process level simulation remains an important goal as additional components aredevelopedandadded.InanIFSMsimulation,cropproduction,feeduse,andthereturnofmanurenutrientsbacktothelandaresimulatedovermanyyearsofweather. Growthanddevelopmentofalfalfa, grass,corn,soybean, and small grain crops are predicted on a daily time step based upon soil water and Navailability, ambient temperature, and solar radiation. Performance and resource use in manurehandling,tillage,planting,andharvestoperationsarefunctionsofthesizeandtypeofmachinesusedanddailyweather. Fielddryingrate,harvestlosses,andnutritivechangesincropsarerelatedtotheweather,cropconditions,andmachineryoperationsused.Lossesandnutritivechangesduringstorageare influenced by the characteristics of the harvested crop and the type and size of storage facilityused.Feedallocationandanimalresponsearerelatedtothenutritivevalueofavailablefeedsandthenutrientrequirementsofuptosixanimalgroupsmakingupeitherdairyorbeefherds.Dietsforeachgroup are formulated using a cost-minimizing linear programming approach, which makes the bestuseofhomegrownfeedsandpurchasedsupplements.Proteinandenergyrequirementsaredeterminedforeachanimalgroupbaseduponthecharacteristicsoftheaverageanimalinthegroup.Oneortwoprotein supplements are used to balance rations. These can include both high and low rumendegradableproteinfeeds.Feedcharacteristicscanbedefinedtodescribeessentiallyanysupplementofeachtypeincludingblendedfeeds.SupplementalPandKfed,ifneeded,isthedifferencebetweentherequirementofeachanimalgroupandthesumofthatcontainedinthefeedsconsumed.Nutrientflowsthroughthefarmaremodeledtopredictpotentialnutrientaccumulationinthe


soil and loss to the environment. The quantity and nutrient content of the manure produced is afunctionofthequantityandnutrientcontentofthefeedsconsumed.Ammonia,hydrogensulfideandvolatile organic compound emissions occur in the barn, during manure storage, following fieldapplication,andduringgrazing.Denitrificationandleachinglossesfromthesoilarerelatedtotherateofmoisturemovementanddrainagefromthesoilprofileasinfluencedbysoilproperties,rainfall,andthe amount and timing of manure and fertilizer applications. Erosion of sediment is predicted as afunction of daily runoff depth, peak runoff rate, field area, soil erodibility, slope, and soil cover.Phosphorus transformation and movement is simulated among surface and subsurface soil pools oforganicandinorganicP.Edge-of-fieldrunofflossesofsediment-boundPandsolubleParepredictedasinfluencedbymanureandtillagemanagementaswellasdailysoilandweatherconditions.Thenetemission of greenhouse gases includes the net exchange of carbon dioxide and the loss of nitrousoxide during the production of feed crops, the emission of methane from enteric fermentation inanimals,andthelossesofallthreegasesfrommanureonthebarnfloor,duringstorage,andfollowingland application. Following the prediction of losses, whole-farm balances of N, P, K, and C aredeterminedasthesumofallnutrientimportsinfeed,fertilizer,deposition,andlegumefixationminustheexportsinmilk,excessfeed,animals,manure,andlossesleavingthefarm.Simulatedperformanceisusedtodetermineproductioncosts,incomes,andeconomicreturnforeach year of weather. A whole-farm budget is used, which includes fixed and variable productioncosts.Annualfixedcostsforequipmentandstructuresaretheproductoftheirinitialcostandacapitalrecovery factor where this factor is a function of an assigned economic life and real interest ordiscount rate. The resulting annual fixed costs are summed with predicted annual expenditures forlabor,resources,andproductsusedtoobtainatotalproductioncost.Laborcostaccountsforallfield,feeding, milking, and animal handling operations including charges for unpaid operator labor. Thistotal cost is subtracted from the total income received for milk, animal, and excess feed sales todetermineanetreturntotheherdandmanagement. By comparing simulation results for different production systems, the effects of systemdifferences are determined, including resource use, production efficiency, environmental impact,production costs, and net return. Production systems are simulated over a 25 year sample of recenthistoricalweather.Allfarmparameters,includingprices,areheldconstantthroughoutthesimulationsothat theonlysourceofvariationamongyears is theeffect of weather. Distributionoftheannualvaluesobtaineddescribespossibleperformanceoutcomesasweathervaries.Inter-yeardynamicsarenot considered; initial conditions such as soil nutrient concentrations and feed inventories are reseteach year. Therefore, the simulated data indicate the range of variation in economic andenvironmentalperformancethatcanoccurgiventhevariationinweatheratthefarmlocation,i.e.thedistribution of simulated annual values indicates weather-related risk experienced by the simulatedproductionsystem.Awidedistributioninannualvaluesimpliesagreaterdegreeofrisk.TheIntegratedFarmSystemModelfunctionsonallrecentWindowsoperatingsystems.Inputinformationissuppliedtotheprogramthroughthreeparameterfiles.Thefarmparameterfilecontainsdata describing the farm such as crop areas, soil type, equipment and structures used, numbers ofanimalsatvariousages,harvest,tillage,andmanurehandlingstrategies,andpricesforvariousfarminputs andoutputs. Themachineryfile includesparameters for eachmachineavailable for useonasimulatedfarm.Theseparametersincludemachinesize,initialcost,operatingparameters,andrepairfactors.Mostfarmandmachineryparametersaremodifiedquicklyandconvenientlythroughdialogboxesintheuserinterfaceoftheprogram.Manyofthesefilescanbecreatedtostoreparametersfordifferentfarmsandmachinerysetsforlateruseinothersimulations.Theweatherfilecontainsdailyweather data for many years at a particular location. The daily data include the date, incident solar


radiation,maximumandminimumtemperatures,andprecipitation. Simulation output is available in four files, which contain summary tables, report tables,optional tables, and parameter tables. The summary tables provide average performance,environmentalimpact,costs,andreturnsfortheyearssimulated.Thesevaluesconsistofcropyields,feeds produced, feeds bought and sold, manure produced, nutrient losses to the environment,productioncosts,incomefromproductssold,andthenetreturnorprofitabilityofthefarm.Valuesareprovided for the average and standard deviation of each over all simulated years. The report tablesprovide extensive output information including all the data given in the summary tables. In thesetables,valuesaregivenforeachsimulatedyearofweatheraswellasthemeanandvarianceoverallsimulated years. Optional tables are available for a closer inspection of howthe components of thefull simulation are functioning. These tables include very detailed data, often on a daily basis.Parameter tables summarize the input parameters specified for a given simulation. These tablesprovideaconvenientmethodofdocumentingtheparametersettingsforspecificsimulations.


INTRODUCTION Dairy and beef production in the United States are facing two major challenges in order toremain viable industries. The first is an economic challenge: inflation-adjusted milk prices haveremainedstableordeclinedformanyyears,whilethecostsofmostproductioninputshaveincreased.As farm profits continue to decrease, production systems must become more efficient. One of themosteffectivewaysofimprovingefficiencyhasbeentoincreasethenumberofanimalsperunit ofcropland(i.e.,intensification).Thistrendhascontributedtothedevelopmentofthesecondchallenge:thefarmsimpactontheenvironment. Livestock farms, particularly dairy farms, have grown more dependent upon the use ofcommercialfertilizersandtheimportofsupplementalfeeds.Theirusehasincreasedcropyieldsandanimal production, which have improved the efficiency and profitability of the dairy and beefindustries. With heavy import of nutrients, however, there is greater opportunity for buildup ofnutrientsinthesoilandthelossofexcessnutrientstogroundandsurfacewaters. Formoresustainabledairyandbeefindustries, improvedproductionsystemsareneededthatincrease the profitability of farmswhile maintainingor reducinglong-termnegative impacts ontheenvironment. Many alternative technologies and management strategies are available to today'sfarmers. These include choices in the number andtypeof animals, landarea, cropmix, equipment,feed-storage facilities, animal facilities, manure-handling options, and much more. Changes in onecomponent ofthefarmoftenaffect other components, andthis interactioncancausechangesintheperformance, environmental impact, and profitability of the farm that are not obvious or easilyunderstood.Quantifyingandcomparingthebenefitsandcostsofalternativetechnologiesandmanagementstrategiesinfarmingisnoteasy. Aproductionsystemthat performswell underoneset ofcropandweather conditions may not perform well under other conditions. Long-term studies are needed toquantifythebenefitsandcostsoverawiderangeofconditions.Fieldstudiesofthistypearecostly,impractical,andperhapsimpossible.Anotherapproachistousecomputersimulation.Process-basedmodels developed and validated with limited field experimental work can be used to study systemperformanceovermanyyearsofweather.TheneedforaresearchtoolthatintegratesthemanyphysicalandbiologicalprocessesonafarmhasledtothedevelopmentoftheIntegratedFarmSystemModel(IFSM).Themodelhasbeenusedtoevaluate a wide variety of technologies and management strategies, and these analyses have beenreported in the scientific and farm-trade literature. Systems research in dairy and beef productionremains as the primary purpose for this tool, but the model also provides an effective teaching aid.With the model, students gain a better appreciation for the complexity of livestock forage systems.Theylearnhowsmall changes affect manyparts of the system,causingunanticipatedresults. Theymayalsousethemodeltodevelopamoreoptimumfoodproductionsystem.Whenusedinextension-typeteaching,producerscanlearnmoreabouttheirfarmsandobtaininformationusefulinstrategicplanning. By testing and comparing different options with the model, those offering the greatesteconomicbenefitwithacceptableenvironmentalimpactcanbefound.

History of Model DevelopmentThecurrentfarmmodelistheproductofover25yearsofsystemsresearchandmodelingwork.TheUSDAsAgriculturalResearchServicehascarriedamajorroleinthiseffort.WiththebeginningoftheU.S.DairyForageResearchCenter(USDFRC)inthelate1970s,aportionoftheCentersfirst


fundingwasprovidedtoMichiganStateUniversityfordevelopmentofasimulationmodelofdairyforageproduction. Anintegrated model of alfalfa growth, harvest, andfeedingwascreatedthroughthecooperativeeffortoftwograduatestudentsandseveraloftheuniversitysfaculty(Savoie et al.,1985). The model, known as DAFOSYM, was written in FORTRAN for use on a mainframecomputer. This version was relatively crude, but it provided a structure for further development.Development and application of the model continued withUSDAsupport after the East LansingClusterprogramoftheUSDFRCwasstaffedin1981.Duringtheearly1980s,mostofthemodelingeffortwasgiventorefiningtherelationshipsusedtodescribefieldcuringandharvestlossesinforageproduction(Rotz, 1985).In1985,themodelwasconvertedtofunctiononpersonalcomputers.Developmentcontinuedtowardmakingthemodelmoreconvenienttouseandmoreadaptabletoothertechnologyandlocations.Inthelate1980s,amajoreffortwasundertakentoupgradethestorageandanimalsubmodelsofDAFOSYM. With the help of others in the USDFRCand cooperators in the NE-132 RegionalResearch Project, the hay and silo storage and the animal component submodels were completed(Buckmaster et al., 1989a, 1989b; Rotz et al., 1989).Forthenextfiveyears,emphasiswasdirectedtoward the application of the model to evaluate alternative forage systems. Benefits and costs ofvarious technologies for hay conditioning, swath manipulation, hay drying, and preservation wereanalyzed with the model. The model was also used for making management decisions such asmachineandsiloselectionandsizing.In1991,theuserinterfacewasupgradedtoallowthemodeltobeusedasateachingaid.ThisDOS version of the model used overlaying menus for editing model parameters and a plottingpackageforhigh-qualitygraphicaloutput.Copiesofthispackageweredistributeduponrequest,withtheprimaryaudiencebeingforageextensionandteachingfacultyintheU.S.andCanada. In theearly 1990s, development of themodel continuedas submodels for manurehandling,tillage, and planting were added (Borton et al., 1995 and Harrigan et al., 1996). This expansionenabledthemodelingofnitrogenlossesandthefarmbalanceofphosphorusandpotassium,providinganewenvironmentalaspecttothemodel.Theexpandedmodelwasusedtocomparevariousmanure-handlingandtillagesystemsondairyfarms. In themid1990s, DAFOSYMwasconvertedtoaWindowsoperatingsystem.Anewuserinterface was developed to provide a more user-oriented model. This conversion allowed furtherexpansionofthemodeltoincludeanimalfacilitiesandessentiallyallcostsincurredontypicaldairyfarms, making it a more complete dairy-farm model. This version of the model was placed on theInternetfornationalandinternationaldistribution. Late in the 1990s, a new corn-growth submodel was added based upon the CERES-maizemodel.Othercrop-productionsubmodelswerealsoaddedforgrass,smallgrain,andsoybeancrops.Theharvest,storage,feeding,andeconomicsubmodelswereexpandedtoincorporatethesenewfeedsonthefarm(Rotz et al., 2001). Grazingofforageandawidevarietyofpossiblefeedsupplementswerealsoadded(Rotz et al., 1999b and Rotz et al., 1999c).Thisexpandedmodelwasusedtostudytheeffectsofcroprotationandfeedsupplementationonfarmperformance,profit,andnutrientlosstotheenvironment.Beefandcroppingoptionswereaddedtothemodel,andthenamewaschangedtotheIntegratedFarmSystemModel.Inthepastseveralyears,animprovedpasturesubmodelwasincorporated,allowingevaluationand comparison of pastures with multiple-plant species (Corson et al., 2007b) or warm-seasongrasses(Corson et al., 2007a).Routineswerealsoaddedtopredictnitrogenvolatilizationoccurringin the barn, during manure storage, following field application, and during grazing (Rotz and


Oenema, 2006).Denitrificationandleachinglossesfromthesoilwererelatedtotherateofmoisturemovementanddrainagefromthesoilprofileasinfluencedbysoilproperties,rainfall,andtheamountandtimingofmanureandfertilizerapplications.Thesoilsubmodelwasextendedtoincludeadetailedsimulationofsoilphosphorusdynamicsandlosses.Erosionofsedimentwaspredictedusingaversionof the Modified Universal Soil Loss Equation (MUSLE), and phosphorus transformation andmovement was simulated among surface and subsurface soil pools of organic and inorganicphosphorus (Sedorovich et al., 2007). Edge-of-field runoff losses of sediment-bound and solublephosphoruswerepredictedasinfluencedbymanureandtillagemanagementaswellasdailysoilandweatherconditions.

Application as a Research Tool The primary goal in the development of the farm model was to create a research tool forcomprehensiveevaluationandcomparisonofdairy-productionsystems.Manydifferenttechnologiesandstrategiesfordairyfarmshavebeencomparedusingthismodelandtheresultsarepublishedinscientificjournalsandconferenceproceedings.Theearliestsimulationstudieswereconductedtoevaluatethefeasibilityandeconomicbenefitsofnewtechnologiesinhaymaking.Chemicalconditioningofalfalfawasintroducedinthelate1970s.Fieldexperimentsconductedtodevelopapracticalsystemforhayproducersprovidedthenecessaryequipment parameters and data to develop and validate the field curing submodels (Rotz, 1985).Simulations on representative farms in the Midwest and Eastern United States indicated that thechemicalconditioningprocessreducedfieldcuringtimeanaverageof12honfirstcuttingand24honlatercuttings.Thisresultedinmorehigh-qualityhay,whichreducedfeedcostsonthedairyfarm.With a treatment cost near $5/t DM of hay, the technique returned the cost of the treatment andprovidedasmalleconomicgainforproducersthroughimprovedhayquality.Matdryingofhaywasanexperimentaltechnologywhereforagewasshreddedandpressedintoamatthatwaslaidbackonthefieldforrapiddrying.Themattedforagedriedtobalingmoistureinaboutonedaywithminimallosseveninhumidclimates.Shreddingalsoimprovedthedigestibilityofthe forage. Experimental work quantified the drying rates, losses, and machinery requirements formodeling the process, and farm level simulations showed that the new technology could be quiteeconomical(Rotz et al., 1990).Theproposedequipmentwascostly,butthemodelpredictedthatinthe Midwest the process could provide a return of up to $4 for each dollar spent on increasedequipmentcoststhroughimprovedhayquality.Chemicalandbiologicalagentsareoftenusedtopreservehigh-moisturehay.Bybalingdamphay, field losses are reduced, but storage losses are increased. Hypothetical treatments with a widerange of effectiveness in preserving high-moisture hay using several strategies were simulated todeterminepotentialbreakeventreatmentcosts.Actualtreatmentcostswereconsiderablygreaterthanthebreakevencosts determinedthroughsimulation, whichindicatedaneconomicloss withcurrenttreatments (Rotz et al., 1992). These simulation results provided preservative manufacturers withguidelinesoneffectivenessversuscostforfutureproductdevelopment.Largeroundhaybalescanbestoredusingavarietyofmethods.Thelong-termperformance,costs,andreturnabovefeedcostsforsixstoragemethods,threebalesizes,twofeedingmethods,andtwomilk-productionlevelswerecomparedon60-and400-cowdairyfarms(Harrigan et al., 1994).The value of bale protection was influenced by bale size, amount of hay in the diet, level of milkproduction, and feeding method. Shed storage was usually, but not always, more profitable thanunprotectedstorage.Thegreatesteconomicreturnfrombaleprotectionoccurredwhensmall-diameter


baleswerefedtohigh-producingcowswithallalfalfafedasdryhay.Comparedtounprotectedhay,annualnetreturnincreasedasmuchas$155/cowwithshedstorageand$143/cowwithtarp-coveredstacks.Thelowestbenefitfrombaleprotectionwasrealizedwhenlarge-diameterbaleswerechoppedand fed as a small amount of a total mixed ration. With this system, annual net return was within$8/cowforallstoragesystemsindicatinglittlebenefitforprotectedstorage. The technique of ensiling direct cut alfalfa has long been of interest in humid climates toeliminate field wilting losses. Simulation was used to compare the long-term performance and theeconomicsofconventionalwiltedsilagesystemstoadirect-cutalfalfaharvestandstoragesystemthatused a treatment such as formic acid to enhance preservation (Rotz et al., 1993). Reducedharvestlosses with direct-cut silage were largely offset by increased effluent losses from the silo, so littledifferencewasfoundinthequantityandqualityofforageavailable totheanimals. Handlingofthewetter material increased machinery, fuel, and labor costs for transport and feeding. The economicvalueofdirect-cutsilagewasfoundtobeverypoor.Producersofhigh-moisturesilageexperiencedaneconomicloss,evenwithnocostforapreservativetreatment.Manydairyfarmershaveconsideredtheuseofgrazingtoreducefeedcostsandimprovefarmprofit.DAFOSYMwas used to model the performance and economics of a 60-cow dairy farm incentral Pennsylvania and a 100-cow operation in southern Michigan with and without the use ofgrazedalfalfa(Rotz and Rodgers, 1994 and Rotz, 1996).Thenetcostoffeedingtheherddecreasedwith grazing through reduced use of conserved forages, corn grain, and soybean meal. Becausegrazinganimalsspentlesstimeinthebarnduringthegrazingseason,lessbeddingwasrequiredwithlessmanurehauledeachyear.Altogether,theseeffectsprovideda12%reductionintheaveragefeedand manure handling cost. Grazing reduced the total feed and manure handling cost by $0.73 to$1.00/cwt of milk produced compared to the confined feeding system where the savings wasdependentuponotherassumptionsonfarmmanagement.Thenetreturnorprofitmarginofthefarmincreasedbyabout$150/cowor$60/acre.DAFOSYMwasusedtoevaluatetheeconomicbenefitsofmeasuringpastureyieldasatoolinmanaginggrazingdairycows(Sanderson et al., 2001). Errorinpasturemeasurementwasfoundtoreduce farm annual net return by $8 to $198/ha depending upon the type of grazing and feedingstrategy used. IFSMwas used to illustrate that using more complex mixtures of forage species inpasturecouldincreaseannualnetreturnofaPennsylvaniadairyfarmbyupto$200/cow(Sandersonet al., 2006). In another application, IFSMwas used to determine the environmental benefits ofconverting a beef farm in Maryland from a corn based system used prior to 1990 to a currentperennial grassland system with intensively managed grazing. The change reduced nitrate leachingloss 56%, denitrification loss 50%, and phosphorus runoff loss by 75% while increasing farm netreturn(Crosson et al., 2007).DAFOSYMwasusedtoevaluatethepotentiallong-termenvironmentalimpactandeconomicbenefit of varying the level of concentrate supplementation on seasonal grazing dairies inPennsylvania(Soder and Rotz, 2001).Farmprofitabilityincreasedassupplementationincreased,butat a decreasing rate with each successive level of supplement. At higher supplementation levels,grazing dairy farms showed greater profitability than a farm with animals fed in confinement.Economic risk or year-to-year variation also decreased as concentrate supplementation increased.Grazing farms showed an environmental benefit compared to the confinement farm by decreasingnitrogenleachingloss. In a relatedstudy, feedinga partial total mixedrationto grazingdairy cowswasfoundtoprovideaviablefeedingstrategyfordecreasingenvironmentalimpactwhilemaintainingprofitability(Soder and Rotz, 2003).Aconfinementfarmshowedthegreatestannualnetreturn,butthis return was only a little greater than that of a grazing farm supplemented with mixed rations.


Economicriskwashighestfortheconfinementfarmcomparedtograzingfarms.Manurehandlinghasbecomeanimportantissueinanimalproduction.DAFOSYMwasusedtoevaluateandcomparemanuresystemsusinglong-termstoragewithspreading,injection,orirrigationtotheless costly daily-haul systemcommonly usedin the upper Midwest (Borton et al., 1995).Incaseswherelong-termstoragesystemswererequiredtoprotecttheenvironment,theannualnetcostofmanurehandling(totalmanurecostminusthevalueofmanurenutrients)wasfoundtoincreasebyupto$65/cowforsmall(60cow)and$45/cowforlarge(250cow)dairyfarms.Comparisonsofthreetillageandfourmanure-handlingsystemsonrepresentativedairyfarmsshowedmulchtillagetobethemosteconomicaltillagesystem(Harrigan et al., 1996).Mulchtillagereturned$15to$25/coweachyearoverconventionaltillagewitha30%reductioninmachinery,fuel,and labor costs. A modified no-till system provided a higher return than conventional tillage, butwhencomparedtomulchtillage,savingsinfuelandlaborwereoffsetbyhighercostsforpesticides.Thehighest net returnamongmanure-handlingsystemswas associatedwith short-termstorageanddaily hauling, but this economic advantage diminished if credit was not given for the value of allmanure nutrients when spread daily. Long-termmanure storage concentrated labor for spreading inthespringandfall.Withlimitedlaborandequipment,thisdelayedtillageandplantingandincreasedannualfeedcostsasmuchas$24/cow.Twoprimaryroughagesfordairyherdsarecornsilageandalfalfa.Wholefarmsimulationwasused to compare the relative merits of these two forages when varying amounts of the foragerequirement(none,one-third,two-thirds,andallonaDMbasis)camefromammoniatedcornsilageandtheremainderfromalfalfa(Borton et al., 1997).Thehighestnetreturnwasfromalfalfaat100%oftheforagerequirement,butdifferencesinnetreturnsacrossforagesystemsweresmallcomparedwiththevariationamongyearscausedbyweather. Changesinfarmsize,soiltype,cropyield,milkproduction, relative prices, and manure handling assumptions did not affect the conclusions of theanalysis. Given the lack of a strongeconomic advantage amongthe forage systems, the practice ofhavingatleastone-thirdoftheforagerequirementprovidedbyeachoftheforagecropswasfavoredtoimprovecropmanagement,feedingmanagement,manuredisposal,andlaboruse. Whole-farmimpacts of using a corn silage processor on the forage harvester were assessedthrough long-term simulations (Rotz et al., 1999a). Processing improved packing in the silos,increased the digestibility of the silage, which reduced supplemental feed requirements and/orimproved milk production. Whenprocessing was used on farms having 100 or 400 high-producingHolsteincowswith40%oftheforagerequirementmetbycornsilage,thetreatmentprovidedabouta2% increase in milk production, a small decrease in supplemental grain feeding, and a $50/cowimprovement intheannual net returnorprofit ofthefarm.Without anincreaseinmilkproduction,theannualeconomicbenefitdroppedto$5/cow.Byincreasingtheamountofcornsilagefedto75%ofthetotalforagerequirement,processingprovideda4%increaseinmilkproductionwithanannualeconomicbenefitnear$100/cow. More efficient use of protein feed supplements on dairy farms can potentially reduce thenitrogen import in feeds, excretion in manure, and losses to the environment. A simulation studyillustrated that more efficient feeding and use of protein supplements increased farm profit andreduced nitrogen loss from the farm (Rotz et al., 1999c). Compared to soybean meal as the soleproteinsupplement,theuseofsoybeanmealalongwithalessrumen-degradableproteinfeedreducedvolatile loss by13to 34kg/ha of cropland with a small reduction in leaching loss (about 1 kg/ha).Using the more expensive protein supplement along with soybean meal improved the annual netreturn by $46 to $69/cow, depending upon other management strategies used on the farm.Environmental and economic benefits were generally greater with more animals per unit of land,


highermilkproductionlevels,moresandysoils,and/oradailymanure-haulingstrategy.Soybeanproductionisrapidlyincreasingondairyfarms.Awholefarmanalysiswasconductedtodeterminethepotentiallong-termeconomicbenefittoproducersandtheenvironmentalimpactofthismanagementchangetogrowingandfeedingsoybeansasaproteinfeed-supplement(Rotz et al.,2001).Theproductionofsoybeansasacashcropincreasedannualfarmnetreturnbyupto$55/cowwhenample cropland was available to produce most of the feed requirement of the herd. Once thesoybeanswerefedineitheraraworroastedform,mostofthiseconomicbenefitwasoffset,reducingtheincreaseinannualnetreturntolessthan$15/cow.Withamorerestrictedlandbase,therewaslesseconomic benefit in shifting land from corn or alfalfa production to soybeans, whether they wereproduced as a cash crop or for feed. Little environmental benefit from reduced N loss or soil Paccumulationwasobtainedbygrowingsoybeansondairyfarms.Useofsmallgraincropsintherotationincreasedfarmnetreturnwhilereducingtheriskoryear-to-year variation in net return (Rotz et al., 2002a). Annual net return was increased by up to$116/cowwhendouble-croppedbarleyorsingle-croppedwheatwasharvestedasgrainandstraw,byabout$30/cowfordouble-croppedbarleysilage,and$50/cowfordouble-croppedryesilage.Nitrogenleachinglossoverthefarmwasreducedby10kg/hawhen40%ofthecornwasdoublecroppedwithsmallgrain,andsoilphosphorusaccumulationwasreducedby2kg/ha.Whole-farmsimulationwithDAFOSYMwasusedtoevaluatethelong-termeffectsofchangesin feeding, cropping, and other production strategies on phosphorus loading and the economics ofactual dairy farms in southeastern New York (Rotz et al., 2002b). Alternative farm managementoptions provideda long-termphosphorus balancefor thefarmaslongas theproduction anduseofforage was maximized and recommended minimum dietary phosphorus amounts were fed.Management changes were demonstrated that eliminated the long-term accumulation of soilphosphoruswhileimprovingfarmprofitability.IFSMwasverifiedtosimulatetheproductionandnutrientflowsoftheDeMarkeexperimentaldairyfarmintheNetherlands(Rotz et al., 2006). Onthisfarm,technologysuchasalowammoniaemissionbarnfloor,enclosedmanurestorage,manureinjectionintothesoil,andtheunderseedingofa grass cover crop on corn land were used to reduce nitrogen loss and improve nutrient recycling.Simulation was then used to evaluate the environmental and economic impacts of using thistechnologyonrepresentativefarmsinPennsylvania.Totalnitrogenlossfromthefarms,primarilyinthe formof ammonia emission, was reducedby25to 55%with an8 to 55%reduction in Prunoffloss.Thecostofthistechnologywasgreaterthanthevalueofthenutrientssavedcausingareductioninannualnetreturnof$65to88/cow. Simulation of farm production systems, supported by case study farm data from fourPennsylvania dairy farms, was used to compare economic benefits and environmental impacts ofdairy production systems using either organic or conventional practices. Four production systemswere compared representing organic grass, organic crop, conventional crop with grazing, andconventionalconfinementproduction.Whole-farmbudgetsusingpricesthatreflectrecentconditionsshowed an economic advantage for organic over conventional production. A sensitivity analysisshowed that this economic advantage was dependent upon a higher milk price for producers oforganic milk as influenced by the difference in milk production maintained by herds using organicandconventionalsystems(Rotz et al., 2007).Environmentalconcernsfororganicproductionwere1)long-term accumulation of soil nutrients due to the importation of poultry manure for cropfertilizationand2)greatersoilerosionandrunofflossofphosphorusduetogreateruseoftillageforweedcontrolinannualcrops.


Application as a Teaching Aid Inadditiontoitsprimarypurposeasaresearchtool,theIntegratedFarmSystemModelalsoprovidesaneffectiveteachingaid.StudentsinBio-SystemsEngineering,Agronomy,CropandDairyandAnimalSciencecanusethemodeltolearnmoreaboutthecomplexityofthemanyinteractionsthatoccurwithinacropandlivestock-productionsystem.Studentsmaystudytheeffectsofrelativelysimplechangessuchasthesizeofatractororothermachines.Suchachangeinfluencesthetimingoffield operations, fuel andlabor requirements, thequality of feeds produced, andmilk production aswell as the cost of production and farm profit. More complex problems may be studied, such asmaximizingtheprofitofagiven-sizefarm,optimizingthemachinerysetorstructuresusedonafarm,oramajorchangeinproductionstrategy. The model can also be used in extension-type workshops. Extension field-staff, privateconsultants,andproducersmayusethismodeltostudytheimpactsofvarioustechnologicalchangesonfarmsintheirarea.Withsomeexperience,themodelcanbeusedtoassistwithstrategicplanningand provide useful information on the selection of equipment, structures, and in planning for farmexpansion. Various cropping systems and feeding strategies can also be compared along withnumerous other options in farm management to determine more economical and environmentallyfriendlyproductionsystems.TheWindowsoperatingsystemanduserinterfaceenhancestheusefulnessoftheprogramasateachingaid.AsinmanyWindows-basedprograms,themainprogramwindowopenstodisplayaseries of menu options and icons that are used to direct the user through major model functions.Dialog boxes are used to viewor modify model parameters. Files supplied with the model providedefault values for all parameters of example farms. Parameters are easily changed by modifyingvalues inanentrybox, selectingtheappropriate optionfromalist box, or settingthedesiredvaluethroughascrollbox.EithermetricorEnglishunitsofmeasurementcanbeused.AWindows-typehelpsystemassiststheuserinpreparingasimulationandinterpretingtheresults. Help can be obtained in any part of the program by pressing the F1 key or by using thecontext-sensitive help button. The internal user guide provides a description of the informationrequired or the output received. Major functions and relationships used throughout the model aredocumentedintheprovidedreferencemanual.

Model AvailabilityTheIntegratedFarmSystemModelisavailablefromthewebsiteofthePastureSystemsandWatershed Management Research Unit of the Agricultural Research Service(http://www.ars.usda.gov/main/site_main.htm?modecode=19-02-05-00).Afterenteringthissite,clickon"Software"intheleftcolumn.Itmayalsobeobtainedbyprovidingthesearch-term"IFSM"intheappropriatesearchboxontheARSwebsite(http://www.ars.usda.gov/main/main.htm).Informationonthemodel andcomplete instructions for downloadingandsettinguptheprogramareprovided. Thename and address of those downloading the program are requested for our records. The programoperatesoncomputersthatuseanyversionoftheMicrosoftWindowsoperatingsystem.

Model Overview TheIFSMmodelisawhole-farmsimulationmodelofcrop,dairy, orbeefproduction. Farmsystems are simulated over many years of weather to determine long-term performance,


http://www.ars.usda.gov/main/site_main.htm?modecode=19-02-05-00http://www.ars.usda.gov/main/main.htm

environmentalimpact, andeconomics. Assuch,themodelisalong-termorstrategicplanningtool.All of the major processes of crop production, harvest, storage, feeding, milk or beef production,manurehandling,andcropestablishmentaresimulated,aswellasthereturnofmanurenutrientsbacktotheland.Bysimulatingvariousalternativetechnologiesand/ormanagementstrategiesonthesamerepresentativefarms,themodelassiststheuserindeterminingalternativesthatprovideadesiredleveloffarmproductionorprofit.

Model DesignThefarmmodelisgenericindesign.Systemsthatuseawiderangeofcroprotations,feedingstrategies,equipment,facilities,andothermanagementoptionscanbeevaluated.Themodelislimitedonlybythecropoptions andmanagement strategies definedasavailable intheprogram. Sincethismodelhassomuchflexibility,however,itcreatesmoreresponsibilityforthemodeluser.Describinga given production system requires the use of many model parameters. Determining appropriatevaluesfortheseparametersmayrequiresometimeandeffort.Cross-checkingparametersisnecessarytomakesurethateverythingneededisentered.Forexample,whenanewcropisaddedtothemodel,theappropriateharvestmethodandassociatedequipmentmustalsobeaddedandthestoragefacilitiesandfeedingstrategymayneedtobeadjusted.Applyingthemodeltonewsituationsalwaysrequiressomecalibrationorverificationtoassurethatthefarmsystemisadequatelydescribed. Thefarmmodelisdesignedtorepresent theperformanceandeconomicsofafarmfirm.Assuch,thesimulatedsystemboundariesarethefarmboundaries. Allresourcesbroughtontothefarmare inputs to the systemand those leaving the farmare the systemoutputs. The economic analysisincludesallofthemajorproductioncostsontypicalfarms.Thesecostsareassociatedwithresourcesbroughtontothefarm,whileincomeisreceivedforproductsleavingthefarm. Anassumption in model designis that nointeraction exists betweenthe farmfirmandthesurroundingmarkets.Thus,theresourcespurchasedbythefarmfirmdonotaffectinputprices,andthecropyieldorproductsproduceddonotaffectcommodityprices.Thissimplificationofignoringmarketconsiderationsandpriceriskisnecessarytoallowthemodeltobeusedmorespecificallytoanalyze the technical and economic production efficiency of a farm system for a given regime ofrelativeprices.Theproductionperiodofthemodeledfarmingsystemisoneyear.Overthisyear,thefarmsresourcebaseisassumedtobeatsteady-statewithneitheracquisitionnordisposalofdurableassets(equipment, facilities, animals, etc.). Although the model is designed for multiple-year simulations,this procedure reflects replications of system performance under various single-year weatherconditions,notaviewofthesystemperformanceoverseveralconsecutiveyears.Theaccountingperiodforthemodelisalsooneyear.Alldollarreturnsfrommilk,feed,andanimalsalesarerealizedinthesameyearasthecostsincurredtoproducethosefeedsandmilk.Thisassumption allows the measure of system performance to reflect one years use of resources toproduce that years production. End-of-year crop inventories are sold and feed shortages arepurchasedtomaintainsteadystateaccountingofresources.Thismodelisdesignedforlong-termorstrategicevaluations.Eventhoughthemodelcanbeusedtotrackfarmperformanceoveraspecificyearortwoofweather,therecommendeduseofthemodelisforlong-termsimulationsovermanyyearsofweather.Whenpredictedvaluesarecomparedto actual farm values for specific years, performance measures such as crop yields may showsubstantialerror.Overmanyyears,however,theseperformancemeasuresshouldadequatelyrepresentthevariationencounteredonrealfarms.


ThefarmmodelisdesignedprimarilyforuseinthetemperateregionsofthenorthernUnitedStatesandsouthernCanada.MostofthevalidationandapplicationofthemodelhasbeendonefortheMidwest,Northeast,andPacificNorthwestregionsoftheUnitedStates,alongwithsomeapplicationin Ontario and Quebec, Canada. Recent applications have also included farms in northern Europe,whereclimaticconditionsarenotgreatlydifferentfromthoseinNorthAmerica.Althoughthemodelhasbeenappliedtootherregionsoftheworld,suchasBrazilandNewZealand,caremustbetakeninverifyingand/orcalibratingthemodeltootherclimates.

Model Input Input information is supplied to the programthrough three data files: farm, machinery, andweather parameter files. Thefarmparameter file containsdatathat describethefarm.Thisincludescrop areas; soil characteristics; equipment and structures used; number of animals at various ages;harvest, tillage,andmanurehandlingstrategies; andpricesforvariousfarminputsandoutputs. Themachinery file includes parameters for each machine available for use on a simulated farm. Theseparametersincludemachinesize,initialcost,operatingparameters,andrepairfactors.Mostfarmandmachineryparametersarequicklyandconvenientlymodifiedthroughthemenusanddialogboxesoftheuserinterface.Anynumberoffilescanbecreatedtostoreparametersfordifferentfarmsand/ormachinerysetsforlateruseinothersimulations.Theweatherdatafilecontainsdailyweatherformanyyearsataparticularlocation.WeatherfilesforallstatesoftheU.S.areavailablewiththemodel,andusersmaycreatenewfilesforotherlocations.Allfilesareinatextformatsotheycanbeeasilycreatedoreditedwithmosttexteditorsorspreadsheets. When creating a new weather file, the exact format of the weather data file must befollowed. This format is similar to the standard format for weather data established by theInternationalBenchmarkSitesNetworkforAgrotechnologyTransfer(IBSNAT)project.Thefirstlinecontains a site abbreviation, the longitude and latitude for the location, the atmospheric carbondioxidelevel,aparameterindicatingthehemisphere(Northern=0.0,Southern=1.0),andaparameterfortheaverageNconcentrationinprecipitation(0.110ppm).Theremainderofthefilecontainsoneline of data for each day. The daily data includes the year and day of that year, total daily solarradiation (MJ/m), mean temperature (C), maximum temperature (C), minimum temperature(C),totalprecipitation(mm),andaveragewindspeed(m/s).Only365daysareallowedeachyear,soonedayofdatamustberemovedfromleapyears.Forthedailyvalues,thefirstcolumnmustbefivecharacterswideandeachoftheothersixcolumnsaresixcharacterswide.

Model Algorithm The model is a structured program that uses numerous objects or subroutines to representvariousprocesses onthefarm.Thereareninemajor submodels that represent themajor componentprocesses.Thesemajorcomponentsare:cropandsoil,grazing,machinery,tillageandplanting,cropharvest, crop storage, herd and feeding, manure handling, and economic analysis. The functions,relationships,andparametersusedineachofthesesubmodelsaredescribedindetailinthefollowingsectionsofthisreferencemanual.Theemphasisofthissectionistodescribethelinkageandflowofinformationfortheoverallmodel(Figure 1.1).Themodelbeginsbygatheringinputinformation.Allparametersstoredintherequestedfarmand machinery parameter files are read. The model user can modify most of these parameters byeditingthedisplayedvaluesintheinputmenusanddialogboxes.Ifthefilesaresaved,themodifiedvaluesbecomepermanentlystoredinthefileornewfilescanbecreatedusingdifferentnames.


Aftertheinputparametersareproperlyset,asimulationcanbeperformed.Thefirststepofthesimulation execution is the initialization of numerous arrays of information in the model. Thisinitializationsetsallsimulationvariablestothesamestartingcondition.Next,themachinerysystemused on the farm is set up. This procedure links all the appropriate machinery into operations fortillage,planting,harvest,feeding,etc.Theperformanceandresourcerequirementratesaredeterminedforeachoperation(SeeMachinerysection).Theremainderofthesimulationisperformedonadailytime-stepforeachyearofweatherdata.Weatherdataisreadforthe365daysofthefirstyearfromtheweatherfile.Eachofthemajorfarmprocessesissimulateddailythroughthoseweatherconditions,andthenthenextyearofweatherdataisread.Thiscontinuesuntiltherequestednumberofsimulatedyearsiscomplete. In a given year, the simulation begins with spring manure-handling, tillage, and plantingoperations.Asequenceoftheseoperationsissimulatedthroughtimeonadailytime-stepuntilallarecompletedoravailabletimefortheseoperationsisused(SeeTillage and Plantingsection).Uptosixoperations can be usedfor the tillage andplanting of each crop. Onanygiven parcel of land, fieldoperations must occur in a sequence, but more than one operation can occur simultaneously. Soilmoisture on the field surface is tracked through time to predict days suitable for fieldwork. Themoistureisincreasedbyrainfallanddecreasedthroughevapotranspirationandmoistureflowtolowersoillayers.Fieldoperationsareallowedonlyonsuitabledayswhenmoistureisbelowacriticallevel.Tillagefollowsmanurehandlinginthesequenceofoperations. Adelayinplantingduetountimelyoperationscreatesadelayincropgrowth,whicheffectscropyieldandquality.Theaverageplantingdatedeterminedforeachcropisusedastheseedingdateforthesimulationofcropgrowth.Followingspringoperations,growthandharvestofeachcropissimulatedonadailytime-stepover the full year (SeeCrop and Soilsection). Only the crops used on the farm are simulated. Ifgrazingisused,thefirstcropsimulatedispasture.Pastureproductionissimulatedeachdayandthequantityofforageproducedistotaledforeachmonthofthegrowingseason.Thismonthlyproductionprovides a forage source for balancing the rations of animals on pasture (SeeHerd and Feedingsection).Alfalfaandgrassforageforharvestaresimulatednext.Thealfalfaandgrassgrowthroutinespredictdailyyieldandnutrientcontentthroughoutthegrowingseason.Atharvesttime,asubroutinesimulatesfieldmachineryoperations, drying,andrewettinginthree-hourincrementsthroughouttheday (SeeCrop Harvestsection). Losses and nutritive changes due to machine operations, plantrespiration, and rain damage are accounted for in predicting the quantity and quality of forageharvested. Each grain crop is then simulated with the order being small grain, corn, and finallysoybeans. Grain-crop models predict grain and silage yields, and the harvest routines account forlossesandresourcerequirementsduringharvest(SeeCrop Harvestsection).Atthecompletionofthedailysimulationofthegrowthandharvestofeachcrop,thestorageofthatfeedissimulated.Thestorageprocessesaresimulatedonanannualtimestep,wherethedatesoffilling,refilling,andemptyingofstructuresinfluencethelossesandchangesinnutrientcontentthatoccur(SeeFeed Storagesection).Foroutsidestorageofhay,dailyweatherconditionsareconsideredinpredictinglossesandnutrientchanges. Thenextstepinthesimulationisfeedutilizationandherdproduction. Feedallocation, feedintake,milkoranimalproduction,andmanureproductionarepredictedforeachanimalgroupmakinguptheherd.Mostoftentheseprocessesaresimulatedonanannualtimestep,wherefeedrationsforall animals are formulated for the year based upon the feeds produced that year (SeeHerd andFeedingsection).Ifpastureoraseasonalcalvingherdisused,feedingandherdproductionprocesses


are simulated ona monthly timestep. The pasture available ona given month andthe stored feedsproducedthatyearareusedtofeedtheanimalgroupseachmonth.Supplementalfeedsarepurchasedtomeetproteinandenergyrequirementsoftheherd,andexcessfeedsaresold.Followingtheherdsimulation,themanureproducedistrackedthroughthescraping,storage,andapplicationprocessestopredictammonianitrogenlossesandthewhole-farmbalanceofnutrients(SeeManure and Nutrientssection).Manureproductionispredictedfromthefeeddrymatter(DM)consumedandthedigestibilityofthosefeeds.Ammoniavolatilizationissimulatedondailytimestepas influenced by ambient temperature and rainfall. Following the prediction of losses, whole-farmmass balances of nitrogen, phosphorus, and potassium are determined as the sum of all nutrientimports in feed, fertilizer, deposition, and legume fixation minus the exports in milk, excess feed,animals,manure,andlossesleavingthefarm.Falloperationsarethensimulatedonadailytime-stepbeginningwithmanureapplication.Eachfalloperation,includinganymanurehandling,tillage,andplanting,aresimulatedinsequencethroughtimeuntilthelastdayoftheyear(SeeTillage and Plantingsection).Operationsareperformedonlyondayssuitableforfieldwork. Erosionofsedimentispredictedasafunctionofdailyrunoffdepth,peak runoff rate, field area, soil erodibility, slope, and soil cover. Phosphorus transformation andmovementissimulatedamongsurfaceandsubsurfacesoilpoolsoforganicandinorganicphosphorus.Edge-of-fieldrunofflossesofsediment-boundandsolublephosphorusarepredictedasinfluencedbymanureandtillagemanagementaswellasdailysoilandweatherconditions.Attheendofeachyear,aneconomicanalysisisperformedbasedupontheperformanceofthefarmduring that year. All costs associated with growing, harvesting, storing, and feeding of crops,milking and care of the animals, and the collection, storage, and application of manure back to thecroplandareincluded(SeeEconomicssection). Awhole-farmbudgetisused,whichincludesfixedandvariableproductioncosts.Annualfixedcostsforequipmentandstructuresaretheproductoftheirinitialcostandacapitalrecoveryfactorwherethisfactorisafunctionofanassignedeconomiclifeandrealinterestordiscountrate.Theresultingannualfixedcostsaresummedwithpredictedannualexpendituresforlabor,resources,andproductsusedtoobtainatotalproductioncost.Thistotalcostissubtractedfromthetotalincomereceivedformilk,animal,andexcessfeedsalestodetermineanetreturn to the herd and management. No carryover of inventories is considered; so, the economicanalysisofeachyearcanbeconsideredanindependentmeasureoffarmperformanceandeconomicsforthatspecificweatheryear. Following the economic analysis, the simulation proceeds to the next weather year, andtheprocess is repeated. This annual loop continues until the requested number of simulated years iscomplete. After the simulation is complete, all performance andeconomic information is organizedandwrittentooutputfiles. Measures of farm performance, production costs, and the net return over those costs aredetermined for each simulated year. All input parameters, including prices, are held constantthroughout the simulation so that the only source of variation is the exogenous input of weather.Distributionoftheannualvaluesobtainedcanthenbeusedtoassesstheriskinvolvedinalternativetechnologies or strategies as weather conditions vary. Using statistical terminology, each systemalternativecanbeconsideredatreatment,andeachsimulatedyearisareplicateoffarmperformanceforthespecificweatherconditionsoftheyear.Thusamultipleyearsimulationprovidesanestimateofthefrequencyorprobabilityofattainingacertainlevelofsystemperformance.Awidedistributionin annual values implies a greater degree of risk for a particular alternative. The selection amongalternativescanbemadebasedupontheaverageannualmeasureofperformanceortheprobabilityofattainingadesiredlevel.


Model Output The model creates output in four separate files. Following a simulation, the files requestedappear in overlaying windows within the primary IFSM window where they can be selected andviewed.Thefouroutputfilesarethesummaryoutput,thefullreport,optionaloutput,andparametertables.Thesummaryoutputprovidesseveraltablesthatcontaintheaverageperformance,costs,andreturnsoverthenumberofyearssimulated.Thesevaluesincludecropyields,feedsproduced,feedsboughtandsold,manureproduced,abreakdownoffeedproduction,manurehandlingandotherfarmcosts,andthenetreturnorprofitabilityofthefarm.Valuesareprovidedforthemeanandstandarddeviationofeachoverall simulatedyears. Themoreextensivefull report includesthesevaluesandmore. In the full report, values are given for eachsimulated year as well as the meanandvarianceoverthesimulatedyears.Optionaloutputtablesareavailableforacloserinspectionofhowthecomponentsofthefullsimulation are functioning. These tables include daily values of crop growth and development; asummaryofthesuitabledaysforfieldworkeachmonth;dailysummariesofforageharvestoperations;annualsummariesofmachine,fuel,andlaboruse;andabreakdownofhowanimalsarefed.Optionaloutputisbestusedtoverifyorobservesomeofthemoreintricatedetailsofasimulation.Thisoutputcanbecomeverylengthyandassuchisonlyavailablewhenrequested.Toobtainafileofmanageablesize,simulationofonlyafewyearsisrecommendedwhenobtainingdailyormonthlydataoptions.Parametertablesalsocanberequested.Thesetablessummarizetheinputparametersspecifiedforagivensimulation.Anynumberoftablescanberequested,andthesetablesaregroupedbymajorsections of model input. These sections include: crop, soil, tillage and planting parameters; grazingparameters; machine parameters; harvest parameters; storage and preservation parameters; herd,feeding,andmanureparameters;andeconomicparameters.Thesetablesprovideaconvenientmethodfordocumentingtheparametersettingsforspecificsimulations. Several aspects of the model output can be plotted. These include the pre-harvest andpost-harvestcropyields,totalfeedandmanurecosts,netreturnforthefarm,andthewhole-farmbalanceofthethreemajorcropnutrients(nitrogen,phosphorus,andpotassium).Annualvaluesoftheseoutputnumbers are ranked from smallest to largest and plotted as a cumulative probability distribution.Theseplotscanbeviewedonthemonitorandprintedonacompatibleprinter.


Figure 1.1 - Model AlgorithmOverallalgorithmoftheIntegratedFarmSystemModel


CROP AND SOIL INFORMATION A general soil model is used to predict the tractability of soil for field operations and themoistureandnitrogenavailableforthegrowthanddevelopmentofeachcrop.Precipitation, runoff,evapotranspiration,moisturemigration,anddrainagearetrackedthroughtimetopredictthemoisturecontentinmultiplelayersofthesoilprofile.Soilsaregenerallydescribedasclayloam,loam,sandyloam, and loamy sand with deep, moderate, or shallow depths. Parameters used to describe soilsincludeavailablewaterholdingcapacity,surfacealbedo,evaporationanddrainagecoefficients,moistbulk density, runoff curvenumber, andthe organic matter, silt, clay, andsandcontents. With thesecharacteristics, the lower limit of extractable water (permanent wilting point), drained upper limit(field capacity), and saturated moisture contents are determined using relationships describedbySaxtonetal.(1986).Thesoilismodeledinfivelayersforgrassandfourlayersforothercropswithalllayershavingthesamesoil characteristics. Inall crops, thetopthreelayers arerelativelythinsurfacelayers withthicknesses of 30, 45, and 75 mm. In grass, the fourth layer is 200 mm thick and the fifth layerextendsfromthe350mmdepthtothebottomofthesoilprofileorthecroprootingdepth,whicheverisfirstlimiting.Inothercrops,thefourthlayerextendsfromthe150mmdepthtothebottomofthesoil profile or the crop rooting depth. The maximumdepth or bottomof the profile is the assignedavailablewaterholdingcapacitydividedbythedifferencebetweenthedrainedupperandlowerlimitsofthesoil(mmmoisture/mmsoil).Typicalrootingdepthsof1.5mareusedforcornandsoybeans,1.2mforsmallgrains,1.8mforalfalfa,and0.8mforgrass.

Soil Water Balance Soil moisture is predicted inthelayers considering thewater entering, movingthrough, andleavingthesoilprofile(Jones and Kiniry, 1986).Moistureenteringthetopsoillayerisprecipitationplusirrigationwaterminusrunoff.Dailyprecipitationisobtainedfromtheweatherdataprovidedasmodel input. If irrigationis used, additional water is addedin20mmincrements ondayswhenthesoilmoisturedropsbelow60%ofthatatfieldcapacity. WaterrunoffiscalculatedusingtheUSDANaturalResourcesConservationServicerunoffcurvenumber,wheretheamountofrunoffisrelatedtotheamountofprecipitationandthemoisturecontentinthetop45cmofthesoilprofile(Jones andKiniry, 1986). The incoming moisture fills the top layer until its drained upper limit is met.Remainingmoisturemovesthroughthefirstlayertofillthesecondlayer.Thisfillingeffectoccursforeachof thelayers until thesoil profile (all layers) is filledtothedrainedupper limit. At this point,moisturedrainstotheunderlyinggroundwaterandisunavailabletothecrop. Moistureisextractedfromthesoilbyevapotranspiration,i.e.waterlossthroughevaporationfrom both soil and plant surfaces. Soil evaporation is determined using the two-stage methoddeveloped by Ritchie (1972). In stage 1, soil evaporation is limited by energy. In stage 2, soilevaporation declines as a function of time fromthe beginning of this stage. Plant transpiration is afunction of the solar radiation level, ambient temperature, crop albedo, leaf area index, and soil-moistureavailability(Jones and Kiniry, 1986).Moisturefromsoilevaporationissubtractedfromtheupper layer of the soil profile and plant transpiration is taken from the lower layers. Transpirationmoisture is divided among layers depending on crop type. In grass, 15%is taken from the secondlayer,25%fromthethird,35%fromthefourth,andtheremaindertakenfromthelargerlowestlayer.Inothercrops,15%istakenfromthesecondlayer,25%fromthethird,andtheremaindertakenfromthelargerlowerlayer. Moistureremovalfromeachlayerislimitedtothelowerlimit ofextractable


moistureforthatlayer.Unsaturatedmoistureflowamongthesoillayersallowsmoisturetomigratetowardequilibrium.Moisturemovesupordownthroughthesoilprofilewhenthemoisturelevelinalayerisgreaterthanthat in an adjacent layer. Moisture flow rate is a function of the soil water diffusivity and thedifferenceinsoil-moisturelevelbetweenlayers(Jones and Kiniry, 1986).Thelinkbetweensoilmoistureandthegrowthanddevelopmentofthecropismodeledusingawaterstressfactor(Jones and Kiniry, 1986). Thisfactorvariesfrom0to1,where1representsnostress on the crop. Values are less than 1 belowthe critical soil moisture where stress begins. Thiscritical soil moisture is normally set at half the available water-holding capacity in the root zone.Belowthislevel,thewaterstressfactordeclinesinproportiontoavailablesoilmoisturetowardzeroat the lower limit of available moisture. Plant transpiration and the associated moisture uptakedeclines in proportion to the decrease in the water stress factor. In grass, species-specific rootingdepthsinfluencehowmuchsoilwaterisavailabletoeachspecies. Theinitial soil moisture content in thespringis set onaspringthawdate. Thethawdate isdeterminedfromanaccumulationofdegree-daysinwhichthedegree-dayvalueforagivendayistheaveragedailytemperatureabovefreezing(C).Untilamaximumaveragedailytemperatureof7Cisreached,theaccumulationofdegree-daysisdividedby6.Ifanaveragedailytemperatureoflessthan0C occurs, the accumulation is reinitialized. The soil is considered thawed when the degree-dayaccumulationreaches14.Theinitialsoilmoisturefollowingthespringthawisnormallysetatfieldcapacity(thedrainedupperlimitmoisturecontent).Inadryclimateorfollowingarelativelydrywinterseason,thisinitialmoisture is reduced. Total precipitation for the first 90 days of the year is divided by the availablewater-holding capacity of the soil. If this ratio is less than one, the initial soil moisture content isreducedinproportiontowardaminimumlevelat30%offieldcapacity.

Soil Nitrogen BalanceSoilNistrackedintwosoillayers.Theupperlayeristhesumofthethreeuppersoillayersdefinedforsoilmoisture,andthelowerlayeristhesameasthatdefinedforsoilmoisture.Nitrogenmovement andtransformationwithin andamongsoil layers is modeled with functions mostly fromthe DAYCENT model (DAYCENT, 2007) with some from the Nitrate Leaching and EconomicAnalysis Package (NLEAP) model (Shaffer et al., 1991). Total soilNin each soil layer includesnitrate, ammonia, crop residueN, manure organicN,and other soil organic matter. Transformationamongthesenitrogenpoolsandflowamonglayersispredictedonadailytimestep.Initiallevelsforthesepoolsaresettorepresentthesoilfollowingagrowingseason.Fertilizerinanitrateorammoniaform and manure organic and inorganicNare added to the upper soil layer on the correspondingapplicationdate.Asmallamountofnitrateisalsoaddedtotheupperlayerfromprecipitationusingauser-assignedNcontent in rainfall. Nitrates flow down through the soil profile with soil-moisturemovement.RotationfromalegumecropalsoprovidesadditionalcropresidueNforusebythesucceedinggraincrop. AddedresidueNis 200kg/ha fromrotated alfalfa and63kg/ha fromrotated soybeans.Consideringthatabout70%ofthecropresidueNisrecycledintothesucceedingcrop,thisprovidesNcreditsofabout140and45kg/haforrotatedalfalfaandsoybeans,respectively.NitrogenuptakebythecropislimitedbyavailablesoilNortheNdemandofthecrop.NitrogenstressfactorsareusedtolinkcropgrowthanddevelopmenttosoilNlevel.Thesestressfactorsvary


between 0 and 1 as defined by Jones and Kiniry (1986). The stress factor on any given day isdeterminedfromtheratioofNuptakeoverNdemandbythecrop.Ingrass,species-specificrootingdepthsinfluencehowmuchsoilNisavailabletoeachspecies. Nitrogenlossesfromthesoilduetovolatilization,leaching,anddenitrificationarepredictedeach day (DAYCENT, 2007). Volatilization is a function of the amount of ammonia in the upperlayer,temperature,andavolatilizationrate.Leachinglossonagivendayisrelatedtotheamountofnitrateinandtheamountofmoisturethatdrainsfromthelowestsublayerofthesoilprofilesimulated(seeNitrous Oxidesection).TheconcentrationofNinmoistureleavingthesoilprofileistheratiooftheNleachedtothetotal amountofmoisturethatdrains. Denitrificationisafunctionofthewater-filled pore space and is limited by either the nitrate or carbon dioxide available in the soil (seeNitrous Oxidesection).

AlfalfaGrowth ProcessesAlfalfagrowthissimulatedusingALSIM1Level2,amodeldevelopedbyFick(1977).Thisdeterministic model simulates the physiological processes of alfalfa growth, incorporating bothbiological and environmental elements. Rather than predicting production on an individual plantbasis,cropproductionismeasuredinunitsofDMmassperunitareaofthefield.AfewmodificationsweremadetotheoriginalALSIMmodelto(a)performmultiple-dayharvestperiodsalongthedailyyield-qualitytimepath,(b)resetregrowthasafunctionofthelengthofthepriorharvestperiod,and(c)usethesoilmodeldescribedabove. Daily growth of alfalfa is predicted for leaves, stems, basal buds, and total non-structuralcarbohydrate reserves. The primary unit for crop growth is material available for top growth andstorage(MATS).MATSrepresents the pool of photosynthates created each day after respiration hasbeendeducted(Fick, 1977).Thismaterialaccumulationonagivendayisafunctionofsolarradiationlevel, crop leaf area, atmospheric CO2 level, day length, ambient temperature, and soil moistureavailability.MATSisusedprimarilyinthegrowthofleavesandstemswiththeremainderstoredastotal nonstructural carbohydrates in the crown and taproots (TNC). A portion of MATSwouldnormally be used for root growth, which is not included in this model. To compensate for thisassumption,aportionofMATSisallocatedtootherplantpartsthatarenottrackedbythemodel. A portion of the TNCis used for the development of basal buds, which then controls thedevelopmentofnewstemsandleaves.Lightmustbesuppliedandmaterialsmustbepresentineithertheleaf orbasal budpoolsforphotosynthesis tooccur. Basal budyields(BUDS)arepredictedasafunctionofthegrowthrateofbuds,thegrowthrateofleavescomingfrombudelongation,thegrowthrateofstemscomingfrombudelongation,TNC,andtherelativegrowthrateofplantmaterial(Fick,1977).Leafandstemgrowthsaremodeledusingsimilarfunctions(Fick, 1977).Leafgrowthisthesumofleafgrowthrateandthegrowthratefrombudelongationminussenescentloss.Leafgrowthrateonagivendayisafunctionofdaylength, current leaf mass,MATS, andawaterstressfactor. Growthfrombudelongationis relatedtosolar radiationlevel, ambient temperature, daylength,BUDS,andwaterstress.ShadingofthecropisthemaincauseoflossduetosenescenceasdescribedbyHuntetal.(1970).Senescenceisafunctionofasenescencerate,thedecaytimeofsenescingleaves,anddaylength. Stem growth is modeled like leaf growth except that stem growth from bud elongation isdefined as 10% of that for leaf-bud elongation. Crop growth continues into the fall until the crop


freezes(i.e.,averagedailytemperaturedropsbelow-3C).Totalnon-structuralcarbohydrateisafunctionofMATS,thegrowthrateofbuds,thegrowthrateof leavesandstems, andTNCrespirationrate(Fick, 1977).TNCrespirationis calculatedfromthemaintenance respiration loss ofTNCand the fraction loss of TNCto respiration whenbudsareformed or regrowth occurs. Plant life depends upon the supply of either photosynthates oraccumulatedTNC. If there is no photosynthesis or the TNClevel drops below 5 g/m the modelsimulatescropdeath.TheTNCrespirationportionofthismodelincludesamaintenancecomponentforoverwinteruseofTNC.ThewaterstressfactorismodifiedfromtheALSIMmodeltoaccommodateIFSMsmultiple-layer soil model. This factor is a function of soil moisture level weighted across soil layers. Tenpercent of the water stress factor depends upon the soil moisture in the upper three layers, and theother90%dependsuponthesoilmoistureinthelargerlowerlayer.Thisdistributionisusedtoreflectadeep-rootedcropthatdrawsmoisturefromdeepinthesoilprofile. AlfalfayieldonanygivendayisthesumofstemandleafDM.Thisyieldrepresentsapurestand of alfalfa in its first production year. To better represent yields found on farms, a yieldadjustmentfactorisusedtoincreaseordecreasethepredictedyieldbyasetamount.Thisamountisthe product of a yield persistence factor and an adjustment factor supplied by the model user. Thepersistence factor represents the yield decline that occurs each year over the life of the stand. Thisfactorisrelatedtothedesignatedlifeofthestandandtheintensityoftheharvestschedule.Increasingthenumberofharvestsduringtheseasonand/orreducingcropmaturityatthetimeofharvestreducespersistence and thus reduces the average yield of alfalfa over the life of the stand. Typical yieldreductions are 0, 0-7, 0-13, and 0-30% for the first, second, third and fourth year of stand life,respectively,wherethehighendoftherangerepresentsmorefrequentcuttingschedules. Thetimeofcropestablishmentalsoeffectsalfalfayield. Whenthecropisestablishedinthespring, a first cutting does not occur on that portion of the crop. So, if a four-year stand life isspecified, 25%ofthecropisestablishedeachyear, andthatportiondoesnotprovideafirst-cuttingharvest. When the crop is established in late summer or fall, a full growth and harvest schedule isassumedthefollowingyear.Harvestresultsinremovaloftopgrowth,andthustheresettingofleafandstempoolstozero.During harvest the updated values of the supply of photosynthates (MATS), the accumulatednonstructuralcarbohydratesintherootreserves(TNC),andthewateravailabletotheplantinthesoilprofilearetemporarilystoredinthemodel. Oncethetotal areaofthealfalfacrophasbeenmowed,thestartingdateforregrowthofthesubsequentcuttingissettoadateonethirdofthetimebetweenthefirstandlastdayofthecurrentharvest.TNC,MATS,andavailablesoilwaterarethenreinitializedatthestoredvaluescorrespondingtotheappropriate regrowthstart date. Regrowthcontinuesforaslong as environmental conditions are appropriate, or until a subsequent harvest is initiated. Thisprocedureallowsthealfalfacroptocontinuetogrowandbeharvestedfollowingapredictedgrowthquality curve through an extended multiple-day harvest period. It also delays regrowth of thesubsequent cutting, reflecting the impact of slow or delayed harvests on yields and quality ofsubsequentcuttings. This model wasdesignedtosimulate alfalfa productionintheGreat Lakesarea, sotheusershouldbeawareofsomeassumptionsandlimitationstothemodel.Thefirstassumptionsarethatthecrop is pure alfalfa and the soil is well drained with no significant fertility problems. With theseassumptions, the model may tend to overestimate average yields. Also, the basic growth-ratecalculationsdependuponleaf-areaandlight-absorptionrelationshipsmeasuredinOntario,Canadaat


43.5N latitude; so, the model may not function as well at more southerly locations with differentlightconditions.Topartiallycompensateforconditionswherethemodelmaynotfunctionasdesired,a user-specified yield adjustment factor canbe usedto adjust the predicted yield while maintainingyear-to-yearyieldvariationsduetoweather.

Nutritive Characteristics The primary nutritive characteristics used in the model to describe forage quality are crudeprotein and neutral detergent fiber contents. Whole crop quality is determined from the individualcharacteristics of leaf and stem components and the portions of the crop that are leaf and stemmaterial. The amount of leaf and stem DM available on a given day is obtained from the growthrelationships described above. Quality contents of leaves and stems are determined by separaterelationshipsusingempiricalmodelsobtainedfromFickandOnstad(1988):

CPL=72.906.96ln(GDD+1.0)[2.1] CPS=26.2-0.039(GDD)+0.000022(GDD)[2.2] NDFL=20.8[2.3] NDFS=24.7+0.083(GDD)+0.0000448(GDD)[2.4]

whereCPLandCPS=Crudeproteincontentofleavesandstems,respectively,%NDFLandNDFS=Neutraldetergentfibercontentofleavesandstems,respectively,%GDD=Growingdegree-daysabove5C,C-d

Perennial GrassMultiple-Species CharacteristicsThemodelallowssimulationofuptofourforagespeciesgrowntogetherinagrasspasture.Onespeciesfromeachofthefollowingplantfunctionalgroupscanbesimulated:(1)cool-seasongrasses,(2)cool-seasonlegumes,(3)cool-seasonforbs,and(4)warm-seasongrasses.Thefollowingspeciesfromeachfunctionalgroupareavailable:

Cool-seasongrasses:

Kentuckybluegrass(Poapratensis)Orchardgrass(Dactylisglomerata)Perennialryegrass(Loliumperenne)Tallfescue(Festuca arundinacea)Userdefined

Cool-seasonforbs:

Chicory(Cichorium intybus)Userdefined


Cool-seasonlegumes:

Redclover(Trifolium pratense)Whiteclover(Trifolium repens)Userdefined

Warm-seasongrasses:

Bermudagrass(Cynodon dactylon)Switchgrass(Panicum virgatum)Userdefined

Usersmayadjustthephysiologicalparametersofaspeciestorepresentlocalvarieties,ortheymaydefinetheirownspeciesbyspecifyingthenecessaryphysiologicalparameters.Tosimulatetwocool-season grasses growing in a mixture, users can modify the parameters of another functionalgroupsspeciestomakeitbehavelikeacool-seasongrass.

Growth ProcessesGrowthofeachspeciesintheswardispredictedfromemergencetotheenddateofvegetativegrowthusingfunctionsfromtheGRASIMmodeldevelopedbyMohtaretal.(1997).Thismodelwasoriginallydesignedtosimulatetheeffectofintensivegrazingmanagementpracticesondailybiomassproduction.Inourmodel,thegrasscomponentisusedtopredictpastureproductionaswellasplantgrowthforhayandsilageproduction.Thismodelincludesphotosynthetictransformationandgeneralgrowthfunctions,wherelightenergyistransformedintocarbohydrates(Johnson et al., 1983).Grossphotosyntheticrateonagivendayisprimarilyafunctionofthesolarradiationlevel,daylength,ambienttemperature,atmosphericCO2level, and the crop leaf area. Photosynthetically fixed carbonis then the product of this grossrate,aCO2-to-carbonconversionfactor,andthemostlimitingoffourpotentialstressfactors.Thesefactorsrepresentstressesoradjustmentsduetoambienttemperature,soil-moistureavailability,soilNavailability,andstoredcarbohydratelevelsintheplant.Thecarbohydratesproducedarepartitionedintorootandshootgrowthandmaintenanceusingpartitioningcoefficients.Thephotosynthateinabove-groundgrowthisallocatedbetweentwopools:storageandstructure(Mohtar et al. 1997).Thedailychangeinthestoragepooliscomputedasthephotosyntheticinputminusstorageandmaintenancerespiration.Thechangeinthestructurepoolisshootgrowthminusthesenescentloss.SenescenceincreaseswiththeamountofstructuralDMinthecrop,ambienttemperature,andthecropphysiologicalstageofdevelopment. Themodel simulates nitrogenfixationperformedbythelegume(e.g., clover) portionof thepasture, if present, using functions adapted fromWuand McGechan (1999). First, using data fromHarrisandClark(1996),themodelassumesaconstant1.11g/mdryweightoflegumerootsforeverypercentageofthepastureoccupiedbylegumes.Forexample,with20%ofthepastureinlegumes,dryweight of legume roots equals 22.2 g/m. The model multiplies this value times a constantnodule:legumerootratioof0.16(Wu and McGechan, 1999)andthemaximumamountofNfixedper gram of nodule mass (110.6 mg N/g nodule dry weight/day) (Wu and McGechan, 1999).MultiplicationofthesevaluesdeterminesmaximumdailyNfixationinthepasture.Forexample,with20% legumes, maximum Nfixation equals 3.9 kg N/ha/day. Subsequently, the model multipliesmaximumNfixationtimesvariables(withvaluesfrom0-1)thatrepresent(1)mineralN(ammoniumandnitrate)intheuppersoillayer,(2)soiltemperature,and(3)legumewaterstresstodeterminetheamountofNfixeddaily.ThesoilmineralNmultiplierdecaysexponentiallyfrom1to0.15asmineralNincreases from 0 to 180 kg N/ha (Wu and McGechan, 1999). The soil temperature multiplier


equalsalinearinterpolationofatrapezoidalfunctioninwhichmaximumNfixationoccursbetween13and26CwithnoNfixationbelow0Corabove30C(Wu and McGechan, 1999).Thelegumewater stress multiplier uses the equation of Jones and Kiniry (1986) described earlier. During asimulation,themodelcalculatestheamountofnitrogenfixedeachdayandthen,usingavaluefromHgh-JensenandSchjoerring(1997),transfers22%ofittothesoilammonium(NH4)pool,whereitbecomesavailableforuptakebygrassandforbcomponentsofthesward.CropDMyieldisdeterminedassumingthatcarbohydratesconstitute40%ofplantDM.Thus,totalDMyieldonanygivendayis2.5timesthesumofthestorageandstructurecarbohydratepoolsofallspeciesinthesward.LeafandstemDMaccumulationisthedifferencebetweenthataddedeachday for each plant component through growth and that removed through senescence. Dry matteradded through leaf growth is a function of the total cropDMaccumulation and the crop stage ofdevelopment(describednext).Theremainingnewgrowthisallocatedtostemgrowth.Stemsenescentlossissetat30%ofthetotalcropsenescentloss,withtheremainderbeingleafDM.

Phenology If the user chooses to simulate a cool-season grass, the model simulates its phenologicaldevelopment through six physiological stages, based loosely on a scale developed by Moore andMoser(1991):

Stage Description Index Value

vegetativeV1 earlyvegetative(germination) cropstage=1.0

vegetativeV2 latevegetative(primordiainitiation) 1.0

In vegetative stage V1, the potential stage-development rate equals the leaf-emergence rate(describedbelow)dividedby3.0.InvegetativestageV2,thepotentialstage-developmentrateequalsamaximumrate(0.28)thatcanbelimitedbythreemultipliersrepresentingtheeffectsofphotoperiod,temperature, and soil moisture. The photoperiod multiplier increases linearly from 0.3 to 0.95 asphotoperiod increases from 8 to 16 hours, then increases linearly to 1.0 as photoperiod increasesabove16hours.Thetemperaturemultipliermonotonicallyincreasesfrom0at0Cto1.0at20C.Thesoil moisturemultiplier rangesbetween1.0and1.4, movingat40%ofthechangeinIFSMswaterstressfactorforplantgrowth.Thepotentialdevelopment-rateduringvegetativestageV2ismultipliedby the photoperiod multiplier times the minimum of the temperature and soil-moisture multipliers.Thenumberofprimordiaequalsthetotalnumberofleavestimes0.6. InreproductivestageR1,thepotential stage-developmentrateequalstheleaf-emergenceratedivided by the number of primordia (plus 1) times the soil-moisture multiplier calculated forvegetativestageV2.InreproductivestagesR2andR3,thepotentialstage-developmentrateequalsamaximum rate (0.05) times the temperature and soil-moisture multipliers calculated for vegetativestageV2.InreproductivestageR4,thepotentialstage-developmentrateequalsamaximumrate(0.1)timesthetemperaturemultipliercalculatedforvegetativestageV2. Themodel represents cool-seasongrass morphologybysimulating thenumber of leaves andtillerspersquaremeterofpasture.Theinitialnumberoftillersissetat8000,allofthemvegetative.The model calculates leaf-emergence rate as a maximum rate (0.15 leaves/tiller/day) that can belimited by three multipliers representing the effects of photoperiod, temperature, and soil moisture.Thephotoperiodmultiplier increaseslinearlyfrom0.7to0.9asphotoperiodincreasesfrom8to16hours, then increases linearly to 1.0 as photoperiod increases above 16 hours. The temperaturemultiplierisaparabolathatreachesitsmaximumvalue(1.0)between20and25C.Below0Candabove 45C, no leaf emergence occurs. The soil moisture multiplier ranges between 0.9 and 1.0,movingatone-tenththechangeinIFSMswaterstressfactorforplantgrowth,whichisafunctionofsoil watercontent, soil water-holdingcapacity,andspecies-specificsensitivitytoaratioofthetwo.The leaf-emergence rate is multiplied by the photoperiod multiplier times the minimum of thetemperatureandsoil-moisturemultipliers. Themodeladdsnewlydevelopedleavestothesimulatedplantaslongastheperiodofleafgrowthhasnotended(i.e.,cropstage=3.0).Themodelcalculatesatillerappearancerateequaltotheleaf-emergenceratetimes0.481.Themaximumnumberoftillersislimitedto10640uptothesummersolstice;afterthesummersolstice,themaximumnumberoftillersdecreasesto8000.Themodelthencalculates"tillerdays",equalto3.0dividedbytheleaf-emergencerate.Thenumberofnewtillersequalstheminimumof(1)tiller-growthrate times the number of tillers or (2) the number of tillers required to reach maximumdivided bytillerdaystimesamultiplierdescribingtheinfluenceofLAI.Thenumberofsenescingtillersequalsthe minimum of (1) 20% of the vegetative tillers or (2) the number of tillers above the maximumtimes the leaf-emergence rate. If the grass is in vegetative state V2 (1.0 < crop stage = 2.0), thenumber of new reproductive tillers equals the number of vegetative tillers times the stage-developmentrate,butonlyuntilthesummersolstice.Afterthesummersolstice,nonewreproductivetillersareproduced.

Nutritive CharacteristicsNutritivecharacteristicscalculatedforplantsharvestedfromtheswardincludewholeplantcrudeprotein andNDFcontents (Buxton et al., 1995). For grass, crude protein equals 6.25 times theNconcentrationintheplantmaterial,whereNconcentrationisthattakenupbytheplantdividedbythe


plant biomassDM. Nitrogen uptake is related to soilNavailability andthe nitrogendemandof thecrop.NitrogendemandonagivendayisthedifferencebetweenthecriticalNconcentrationdesiredby the plant and the actual N concentration in the plant. For cool-season grasses, the critical Nconcentration is predicted as an exponential function of the predicted crop stage of development(Jones and Kiniry, 1986). For cool-season legumes and forbs the critical Nconcentration is afunctionofplantDM(kg/ha) andthemaximumNconcentration(MAXNC)(Gastal and Lemaire,2002):criticalN=MIN(MAXNC,MAXNC(DM/1000)-0.5)[2.5]

Forwarm-seasongrasses,thesamebaseequationisused,butwithadifferentexponent:

criticalN=MIN(MAXNC,MAXNC(DM/1000)-0.37)[2.6]

Theaveragecrudeproteincontentoftheswardequalsmeancrudeproteincontentofallspeciespresent,weightedbytheDMofeach. Neutral detergent fiber (NDF) concentrations (both digestible andindigestible) are predictedseparatelyfortheleafandstemcomponentsofeachspecies.TheNDFconcentrationonagivendateis the totalNDFaccumulated in the leaves or stems divided by the accumulated leaf or stem DM.SimilarrelationshipsareusedtopredicttheNDFaccumulationinleavesandstems.Thataccumulatedeachdayisthedifferencebetweenthataddedthroughgrowthandthatlostthroughsenescence.Thataddedthroughgrowthis a functionof theDMaddedthroughgrowth,ambienttemperature,andforcool-season grasses, crop stage of development. The base rate of NDFaccumulation differs byfunctionalgroup,withwarm-seasongrassesaccumulatingdigestibleandindigestibleNDFatahigherratethancool-seasonspecies(Fritschi et al., 1999;Mandebvu et al., 1999). The model calculates daily total NDFaccumulation in leaves and stems as a maximumrate(1.18) times the daily structural biomass growth times multipliers representing the effects oftemperature and relative total NDFaccumulation rate by crop stage. The temperature multiplierequals 0.87 for mean daily temperatures up to 10C and increases by 0.02 (0.03 for warm-seasongrasses)foreverydegreeover10C.Forcool-seasongrasses,therelativetotalNDFaccumulationratefor leaves varies little, increasing from 0.35 to 0.37 as crop stage increases from 0 to 5, while therelativetotalNDFaccumulationrateforstemsislargerandvariesmore,increasingfrom0.55to0.75ascropstageincreasesfrom0to5.Forcool-seasonlegumes,therelativetotalNDFaccumulationrateforleavesandstemsisfixedat0.30and0.45,respectively.Forcool-seasonlegumes,therelativetotalNDFaccumulationrateforleavesandstemsisfixedat0.40and0.60,respectively.Forwarm-seasongrasses, the relative totalNDFaccumulation rate for leaves and stems is fixed at 0.66 and 0.76,respectively.ThemodelcalculatesdailyindigestibleNDFaccumulationinleavesandstemsasthedailytotalNDFaccumulationtimesmultipliersrepresentingtheeffectsoftemperatureandrelativeindigestibleNDFaccumulation rate by crop stage. The temperature multiplier equals 0.74 for mean dailytemperaturesupto10Candincreasesby0.04(0.03forwarm-seasongrasses)foreverydegreeover10C. For cool-season grasses, the relative indigestibleNDFaccumulation rate for leaves increasesfrom0.5to0.75ascropstageincreasesfrom0to5,whiletherelativeindigestibleNDFaccumulationrateforstemsincreasesfrom0.45to0.82ascropstageincreasesfrom0to5.Forcool-seasonspecies,


the relative indigestible NDFaccumulation rate for leaves and stems is fixed at 0.40 and 0.45,respectively.Forwarm-seasongrasses,therelativeindigestibleNDFaccumulationrateforleavesandstems is fixed at 0.65 and 0.60, respectively. An additional small increase in indigestibleNDFispossibleinleavesorstems,equaltoabaserate(0.003forcool-seasongrasses,0.002forwarm-seasongrasses)timestheamountofdigestibleNDF(totalNDFminusindigestibleNDF)inleavesorstems,respectively,timestheindigestibleNDFtemperaturemultiplier. Senescent loss of NDFis a function of the senescent loss of DMpredicted in the growthcomponentabove,thefractionofthecropthatisleavesorstems,andtheNDFconcentrationinthelost material. In-vitro true digestibility (IVTD) of leaves and stems of each species is calculated bydividing digestible DM (DMminus indigestible NDF) of leaves or stems by leaf or stem DM,respectively.TheaverageNDFandIVTDconcentrationsoftheswardequalsmeanNDFandIVTDofallspeciespresent,weightedbytheDMofeach.

CornGrowth ProcessesCornbiomass(silage)andgrainyieldsarepredictedfromseedingthroughmaturity.FunctionsforpredictingabovegroundgrowthandphenologicalstagearetakenfromtheCERES-maizemodel(Jones and Kiniry, 1986). As implemented in the Decision Support System for AgriculturalTechnology (DSSAT) version 3.0 (Tsuji et al., 1994). The model simulates the growth anddevelopmentofasingleplantthatisrepresentativeofafullcrop.Phenologicaldevelopmentofleaf,stem,ear,andgrainmassispredicteddailybaseduponsoilandweatherconditions.Thisdevelopmentoccurs in six physiological stages (emergence through harvest maturity) using information on theaccumulationofthermaltimeorphotoperiod(Jones and Kiniry, 1986).Geneticparametersareusedinsettingthelimitsforsteppingfromonedevelopmentalstagetothe next. To simplify our model, two genetic parameters are assigned as functions of a relativematurityindexdefinedasdaysuntilmaturity.Thegeneticparameters,P1andP5,asdefinedbyJonesandKiniry(1986)areestimatedwiththefollowingrelationships:

P1=4.0(RMI)220[2.7]

P5=6.0(RMI)+70,butnogreaterthan685[2.8]

whereRMIistherelativematurityindexindays.OthergeneticparametersaresetatP2=0.5,G2=750,andG3=9. Our model differs from the DSSATmodel in that root growth is not modeled. Instead ofpredicting theroot uptakeof moisture to predict themoisture stress effect onplant growth, a waterstressfactorissimplycalculatedfromtheavailablesoilmoisture.Thisfactorvarieslinearlyfrom1atthe soil moisture level where plant stress begins (normally about 50%of field capacity) to 0 at thelower limit of extractable soil moisture. The water stress factor is weighted across soil layers. Forcorn,30%ofthewaterstressfactorisdependentuponthesoilmoistureintheupperthreelayers,andtheother70%isdependentuponthesoilmoistureinthelargerlowerlayer.Thisfactorwasusedtocontrol the growth rates of various plant parts as implemented in theDSSATmodel (Ritchie and


Otter, 1985).Growthisdrivenbycarbonfixedthroughphotosynthesis.Drymatterproductiononagivendayisafunctionofthesolarradiationlevel,ambienttemperature,plantleafarea,andthemoisturestressimposed on the plant (Jones and Kiniry, 1986). Partitioning of theDMproducedamongthe plantcomponents varies with the developmental stage of the crop. In stages 1 and 2, the above groundgrowthisrestrictedtoleafgrowth.Dailygrowthofleafareaperplant,totalplantleafarea,andleafmassaredetermineduntiltasselinitiation.LeafgrowthisrelatedtotheamountofDMproducedandambienttemperatureasinfluencedbyanystressimposedbyinadequateavailabilityofsoilmoistureandnitrogen.Tasselinitiationthroughtheendofleafgrowthandsilkingismodeledinthethirdstage.Inthisstage,dailygrowthofleafmassandareacontinuetobecalculatedinadditiontodailystemgrowth(Jones and Kiniry, 1986).Stemgrowthisafunctionofdailyleafmass,leafnumber,andthenumberof leaves at tassel initiation. The partitioning ofDMbetween leaf and stemgrowth varies with thenumberofleavesontheplant.Instage4,growthispredictedfromsilkingtothebeginningofeffectivegrainfilling.Itisinthisstagethateargrowthbegins,leafgrowthstops,andstemgrowthcontinues.Eargrowthisproportionaltotheaccumulationofgrowingdegree-daystimesthewaterstressfactor(Jones and Kiniry, 1986).Stem growth is then proportional to ear growth. The average accumulation of plantDMover thedurationofthisdevelopmentalstageisusedtosetthenumberofgrainkernelsontheear.Effectivegrainfillingoccursinstage5.Duringthisstage,dailytotalgraingrowthiscalculatedwith daily biomass production divided among grain, stem, and root growth. Plant DMis alsotranslocated fromthe stemsandleaves to assist grain filling. Grain filling is influenced byambienttemperatureandanystressimposedbylowsoilmoistureorsoilN. Atphysiologicalmaturity(stage6),allcropgrowthfunctionsceasebutthesenescenceofcropmaterialcontinues.Totalleafsenescenceiscalculatedthroughoutallsixdevelopmentalstages.Leafsenescencedueto drought stress, competition for light, and lowtemperature are determined based upon total plantleafarea,thesumofdailythermaltime,andsoilmoisturestress(Jones and Kiniry, 1986).Grainandsilageyieldsaretrackedthroughouteachdayofasimulation.Grainyieldisthesingleplantgrainmasstimestheplantpopulation.Silageyieldisthetotalbiomassyield,whichincludesthesumof the plant leaf, stem, cob, and grain masses multiplied by plant population. For control overpredictedgrainandsilageyields,yieldadjustmentfactorsareusedtoincreaseordecreasepredictedyieldsasetamounteachdayoverallsimulatedyears.Thisgivesthemodelusertheabilitytoadjustorsetthelong-termaverageyieldwhilemaintainingyear-to-yearvariationasinfluencedbyweather.Arotationeffectisalsoaddedtoadjustcornyieldaccordingtotheprecedingcrop.Forcornthatfollowscorn, the grain andsilage yields are reduced10%.This reduction represents a typical yielddifferencebetweencontinuouscornandcornfollowingalegumecrop(Rotz

the integrated farm system model - usda ars · the integrated farm system model ... models with a...

Documents