llllllllllllHl||l|l|l||||||ll||l|Hillllllllllllll||llllll L’XL 5% 0 “b I b 3 1293 00563 5374 -«. . v V .1: _ “Huh-5" ~ Date LIBRARY Michigan State l. University This is to certify that the thesis entitled LINKING GEOGRAPHIC INFORMATION SYSTEMS WITH SIMULATION MODELING FOR RISK-BENEFIT ANALYSIS IN POTATO PRODUCTION presented by Mark Sadler Swartz has been accepted towards fulfillment of the requirements for M . S. degree in Resource Development Qmjtgc/fi/Za Major professor 12/05/88 0-7639 MS U is an Affirmative Action/Equal Opportunity Institution 4V- ”A MSU LIBRARIES RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. LINKING GEOGRAPHIC INFORMATION SYSTEMS WITH SIMULATION MODELING FOR RISK-BENEFIT ANALYSIS IN POTATO PRODUCTION BY Mark Sadler Swartz A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Resource Development 1988 ABS'. DEVI est. fer E roo pro des mov qrc see was in em 8 5 Was LINKING GEOGRAPHIC INFORMATION SYSTEMS WITH SIMULATION MODELING FOR RISK-BENEFIT ANALYSIS IN POTATO PRODUCTION by. Mark Sadler Swartz ABSTRACT Development of an integrated modeling system capable of estimating risks and benefits associated with nitrogen fertilizers and aldicarb in regional potato (Solanum . tuberosum) production was the goal of this project. RlSk was measured as aldicarb and nitrate mass leached below the root zone. Benefit was measured as on-farm agricultural profitability. An analytical model (SUBSTOR) which describes plant growth, nitrogen movement, and aldicarb movement was used to estimate profitability and potential groundwater contamination under alternative management scenarios. The ERDAS geographic information system (GIS) was used to spatially correlate weather, soils, and land use information for model parameterization in heterogeneous environments. Potato production information from Sections 3,9,16,17 of Douglass Township in Montcalm County, Michigan was used for prototype analysis. I w< and Dr. T Thanks t: Russel J completi Movement Special love, ur abilitie the resl have sh; always Abraczi ACKNOWLEDGMENTS I would like to thank Dr. Joe Ritchie, Dr. George Bird, and Dr. Tom Edens for serving on my guidance committee. Thanks to Brian Baer, Farsad Fotouhi, Dr. Brad Johnson, Dr. Russel Jones, and Dr. Joe Hudson for there help in completion of this project. I would also like to thank the Movement Arts community for their energy and comradery. Special thanks to Fred, Rose Ann, and Matt Swartz for their love, unending support, personal, and professional abilities, John Davenport, Becky Mather, Fred Warner, and the rest of the nematology crew for work and good times we have shared. Thanks to Richard Kemp and Jim Smania. I always knew you were there. Thanks to Laura Abraczinkas,"because she makes me laugh". ii Title: Lin simulation production Acknowledgn List of Tat List of Fig Chapter I. IN“. Prl Pr: II. TABLE OF CONTENTS Title: Linking geographic information systems with simulation modeling for risk- -benefit analysis in potato production Page Acknowledgments ....................................... ii List of Tables ..... . .................. . ...... . ........ ix List of Figures ....... ........ ..... ............ . ...... xiv Chapter I. INTRODUCTION Project Introduction ........................... 1 Project Objectives ........ . .................. 1 Project Overview ......... . ................... .. 2 Project Organization ..... ......... . .......... 2 Spatial Data Base .......................... 4 Alternate Management Practices ............. 7 Aldicarb Movement and Degradation .... ...... 7 Aldicarb Root-lesion Nematode Impact ....... 8 Risk-Benefit Analysis .... .......... . ....... 8 Relationship between Objectives .. .......... 9 Justification ..................... ............. 9 Nitrogen ...... .................... ........... 11 Nitrate Risk ....... ..... . .................... 11 Regional Nitrate Concern .... ................. 12 Aldicarb ............. ........................ 12 Aldicarb Risk .... ............................ 13 Regional Aldicarb Concern .................... 14 II. REGIONAL DATA BASE DEVELOPMENT Introduction .. ........ ....... ........... ...... 15 Materials and Methods ... ...... .......... ....... 15 Data Source ................. ....... . ......... 16 Soils ............... ..... ....... ........... 16 Field Boundaries ..... ........ . ............. 16 Land Use ........................... ........ 16 Weather ................. ......... ..... ..... 17 Data Processing ......... . ......... . ........ 17 Air Photo Interpretation ........ ........... 17 Rectification ....... .. .......... . .......... 17 Geocoding .................. ................ 17 Digitization ............. ........ . ....... 18 Mix-«...» HI. ALTERNAC Introdu< Literati Potati Potat Irr Ris Materie Results Irrig Nitr< Aldit Summar ALDICA Introd Litera Soil Evap SYSt Deg: Impl Materj Assn Comput Exis New] I: I] Pf P P if ifi . . me Rasterization .............. . ......... .... 18 GIS creation ....... ......... .. ........... 18 Soils GIS ............... ... ..... . .......... 20 Land Use GIS ........ ...... ................. 20 Weather Data ...... ....................... .. 22 Potato Production Analysis ................... 22 Results ....................... . ................ 22 Summary ............. . .......................... 2 6 III. ALTERNATE MANAGEMENT PRACTICE DETERMINATION Introduction ................................... 27 Literature Review ...... . ....................... 27 Potato Production Scope . ...... . .............. 28 Potato Management ............. ..... ..... ..... 28 Irrigation ................. ........ . ....... 28 Irrigation Use ................. .......... 28 Irrigation Value ...... ................... 29 Irrigation Concern ............ ...... ..... 29 Nitrogen . ......... ... ...... . ............... 3o Nitrogen Use ..... . ......... . ............. 30 Nitrogen Value .. ........ . ...... . ......... 3O Nitrogen Concern .......... ...... . ........ 31 Aldicarb .... .......... . ......... . .......... 31 Aldicarb Use . ...... ......... ............. 31 Aldicarb Value ..... ...................... 32 Risk Management ....... ...... ............... 32 Materials and Methods .... ......... . ............ 33 Results ..... ............ ........ . .............. 33 Irrigation ...... ......... ... ........... ...... 34 Nitrogen ..................................... 34 Aldicarb ....... ..... ....... .................. 34 Summary ..... ..... . ...... .. ........ . ...... . ..... 36 IV. ALDICARB MOVEMENT AND DEGRADATION MODEL Introduction . .................................. 37 Literature Review ................. . ...... . ..... 37 Soil binding .... ..... ................... ..... 37 Evaporation ................ ..... ....... ...... 38 Systemic Uptake ........................ ...... 38 Degradation Rate ................. ..... ....... 39 Implications of the Literature ............... 40 Materials and Methods ............ ...... . ....... 40 Assumptions .................................. 40 Computer Code Development ............. ......... 41 Existing Routines ................... ......... 41 Newly Developed Routines .. ..... .... ..... ..... 45 Include File ............. .... .............. 45 IPPST ....... ........... ...... ........ . ..... 46 PTRANS ......... ..... ............ ........... 47 PFLUX .................. ........ ..... ....... 49 PFLOW .......................... ............ 50 iv SYST PSTE TTOU SOII OUTI Summary . ALDICARI Introduc Integ] Objt Dat; Dat Pre: Meta- Obj Mod Materia Integ Lit Lit Var Dat F I I l 1 Meta- Mo< Ge] Sb In SYSTEMIC . ......... . ............ ..... ...... . PSTDAY ................ .......... ...... ..... TTOUT ...........; ......... . ................ SOILPST ..................... .. ............. OUTPLCH ........... ...... ........ . .......... Summary ........ ......... .... ................... ALDICARB/ROOT-LESION NEMATODE YIELD IMPACT Introduction . ......... ............ ............. Integrative Research Review .. ................ Objective Definition ..... ....... .. ....... .. Data Collection ......... ............ .. ..... Data Evaluation .................... ........ Presentation ...... ...... ....... ............ Meta—analysis .......................... ..... . Objective Definition ....................... Model Hierarchy ...... ....... ....... V........ Materials and Methods .......................... Integrated Research Review ....... . ......... .. Literature Search Procedure ...... .......... Literature Selection Criteria .............. Variable Description .. ..... ..... ...... ..... Data Evaluation ...... .................... .. Research Bias ............. ...... . ........ Data Bias ................ ..... .. ......... Data Availability ...... .................. Missing Values ................. .......... Estimator Test .............. ........ ..... Meta-analysis ........................ ........ Model Hierarchy ............. ...... ......... General Analytical Methods ....... .......... Study Variability ......... ......... . ....... Specific Methodologies .... ...... ....... Variability in Potato Production Measures ..... ....................... Impact of Selected Management Practices on Tuber Yield .......... ..... ..... ..... Results ....................... ...... ..... Variability in Potato Production Measures .............. ................ Impact of Selected Management Practices on Tuber Yield .. ..... ...... ........... Impact of Aldicarb on Tuber Yield .......... Specific Methodologies .. ......... . ..... .. Mean Yield Loss ................. ....... Cumulative Probability Distribution .... Regression Analysis .... ....... . ....... . Pre-season ..... ...................... Post-season .......................... Stepwise ......... ...... ......... ..... Results ......................... ...... ... V Imp; Imp Yie Mean Yield Loss ........................ 82 Cumulative Probability Distribution .... 82 Regression Analysis .......... . ......... 85 Pre-season . ..... .... ......... . ..... .. 85 Post-season .......... ........ ........ 85 Stepwise .............. ............... 85 Impact of aldicarb on P.penetrans Population 88 Specific Methodologies ................... 88 Regression Analysis ............ ..... ... 88 Model Development . ........ . .......... 88 Threats to Model Validity ............ 89 Autocorrelation .................... 89 Heterogeneity of Variance .......... 90 Distributed Delay ...... ...... .......... 91 Results ..... ............. ..... ...... ..... 91 Regression Analysis ....... ............. 91 Autocorrelation . ..................... 96 Heterogeneity of Variance ............ 96 Distributed Delay ..................... 97 Impact of P.penetrans Populations on Tuber Yield . .............. .. ........... ... 98 Specific Methodology ......... ............ 98 Class Correlation .... ....... . .......... 98 Regression ...... ...... ......... . ....... 98 Early and Late Season ................ 99 Four Time Categories ..... ....... ..... 99 Results .............. ...... ..... ......... 100 Class Correlation .. ...... .. ............ 100 Regression ................ ............. 100 Early and Late Season ......... ....... 100 Four Time Categories ...... ....... 100 Impact of P. penetrans Populations on Plant Development ................... ......... .... 102 Specific Methodologies ... ......... .. ..... 102 Data Base ...... . ........... . ........... 102 Correlation ........ .... ............. ... 103 Regression ............................. 103 Changes in Plant Growth .............. 103 Changes in Partitioning ........ ...... 103 Results .............. ...... ... ..... ...... 104 Correlation ....... ........ . ...... . ..... 104 Regression ........ ........... ... ..... 105 Changes in Plant Growth ..... ......... 105 Changes in Partitioning .............. 106 Discussion ................................... 110 Variability in Study Findings ...... ........ 110 Impact of Aldicarb on Tuber Yield ... ....... 111 Impact of Aldicarb on gépgpetgagg Populations ...................... ..... . 114 Impact of P. penetrans Populations on Tuber Yield . ..... ........... ..... .......... ...... 116 l N'" Impa Deve Model PPEl IPPI PPIl OUT! TRT: Summa: VII. NITRATE Introdu Materia Insta Treat Pla Fer Irr Leach San Nit A16 Results Aldic Simul Plant Implica Experh VIII. SIMULAI ANALYS Introd Materi SUBS Weat Soil Mana Bene Risk Result Simu Risk Summer IX‘ DISCUSS] Regiona] Alternai Aldicarl AldiCarl Risk Bel SUmmary Page Impact of P.penetrans Populations on Plant Development .................... . .......... . 117 Model Parameterization ......... . ............. 118 PPENE.INC ......... . ........................ 120 IPPENE ....... ..... . ............... ......... 121 PPIMPACT . .................................. 121 OUTYIELD ....... .... ........................ 122 TRTSUM . .................................. .. 123 Summary .................... ....... ........... 123 VII. NITRATE AND ALDICARB LEACHING EXPERIMENT Introduction ............. ..... . .............. .. 124 Materials and Methods .......... ................ 124 Installation ............ ............ ......... 124 Treatments ............. ........ ... ..... . ..... 126 Planting ...... ........ ....... .......... .... 126 Fertilzier ........... . ..... .... ..... . ...... 127 Irrigation .. ........... .. ........ .......... 127 Leachate Analysis ......... ...... ..... . ....... 128 Sampling ..... ....... ..... .................. 128 Nitrate ... ....... . ......................... 129 Aldicarb ... .................... . ........... 129 Results ........................... ....... ...... 129 Aldicarb Degradation and Movement ....... ..... 129 Simulated and Observed Results ............... 132 Plant Stress Factors ......................... 133 Implication of Nitrate and Aldicarb Leaching Experiment Results .. ........................... 139 VIII. SIMULATION MODELING FOR REGIONAL RISK-BENEFIT ANALYSIS Introduction ............... ............ .. ...... 140 Materials and Methods ..... ........ ......... .... 140 SUBSTOR ......................... ....... . ..... 140 Weather ................................ ...... 142 Soils .................... ..... ...... ......... 142 Management Strategies .................. ...... 142 Benefit ................. ...... ...... ......... 143 Risk ................................. ...... .. 143 Results ........................................ 143 Simulation ......................... .......... 145 Risk-benefit analysis ....... ................. 148 Summary ................... ......... . ........... 151 IX. DISCUSSION AND RECOMMENDATIONS Regional Data Base ......................... ...... 152 Alternate Management practices ............. ..... . 152 Aldicarb Movement and Degradation . ........ .. ..... 153 Aldicarb/Root—lesion Nematode Impact ............. 156 Risk Benefit Analysis ............. ...... . ..... ... 157 Summary .......................................... 159 vii XI. XII. XIII. XIV. XVI. XVII. XVIII. XIX. XXI. Literati Appendi: Appendi: Appendi: Appendi: Appendi Appendi Appendi Appendi Appendi . Appendi Appendi x, Literature Cited ............................... EI%% XI. Appendix A ................... . ................. 167 XII. Appendix B ....... ............... . .............. 170 XIII. Appendix C ... ...... ..... ....................... 178 XIV. Appendix D .................. .............. ..... 181 xv. Appendix E ..... ................................ 183 XVI. Appendix F . ..... . ............ . ................. 185 XVII. Appendix G ..... . ....... ............. ........... 187 XVIII. Appendix H ............ ................... ...... 190 XIX. Appendix I ................. .................... 194 XX. Appendix J ... .................................. 196 XXI. Appendix K ...... ......... . ..................... 197 viii Table Table Table Table Table Table Table Table Table 1. 2. 3. 4. 5. O\ \1 (D \0 Soil : used : Land . Field truth Soils 1986, Irrig appli strat Nitrc conve Aldic coefi Degra aldic Mean aldil soil Table 10 . SUBS: func Table 11. Rese revi Table 12 . Mean and TAble 13 . Pair With “Sir Popt Yie] LIST OF TABLES . Page Table 1. Soil series names and attr1bute numbers used in geocoding ......... ................... 21 Table 2. Land use attribute codes used in geocoding .. 21 Table 3. Field identification numbers and ground- truthed land use for 1986, 1987, and 1988 .... 24 Table 4. Soils on which potatoes were produced in 1986, 1987, and 1988 ............ .. ....... 25 Table 5. Irrigation dates and total number of applications for alternate management strategies . ........... ....... ....... . .. ..... 35 Table 6. Nitrogen application dates and amounts for conventional and standard management strategies 35 Table 7. Aldicarb and metabolite soil adsorption coefficients .............................. 37 Table 8. Degradation rate constants for aldicarb, aldicarb-sulfoxide, and aldicarb-sulfone ...... 39 Table 9. Mean degradation rates for aldicarb, aldicarb-sulfoxide, and aldicarb—sulfone by soil physical and chemical parameters ........ 40 Table 10. SUBSTOR program routines and their primary functions ......................... .. ........ 44 Table 11. Research not included in the literature review ....................................... 62 Table 12. Mean proportion of tuber size classes to total and A size classes ................. .......... 67 Table 13. Paired t-test probabilities associated with estimation of missing size class measures using mean proportion of total tuber yield .... 67 Table 14. Influence of cultivar, initial nematode population density, and aldicarb on B tuber Yield ....IOOOIOOOOCOOUOCID0.0ICOIICOOOIOICODO 76 ix Table 15. Influe popula yield Table 16. Influt popula tuber Table 17. Influi popul. tuber Table 18. Mean appli Table 19. Summa perc Table 20. Summa perce Table 21. Summa perce perCE perce Table 24. Sum; soil. dens: soil dens Table 26. Pear Perc tube nema Clas Table 27. Impa Perc and in-s 01'] I: Table Table Table Table Table Table Table Table Table Table Table Table Table Table 15 16 17. 20. 21. 22. 23. 25. Influence of cultivar, initial nematode population density, and aldicarb on A tuber yield ............. ........................... 77 Influence of cultivar, initial nematode population density, and aldicarb on Jumbo tuber yield ..... . ..... . ............ . ......... 78 Influence of cultivar, initial nematode population density, and aldicarb on Total tuber yield .................. ...... . ......... 78 Mean percentage yield loss without aldicarb application by variety and tuber size class .. 82 Summary of preseason regression results for percentage yield loss by tuber size class ... 85 Summary of postseason regression results for percentage yield loss by tuber size class ..... 85 Summary of stepwise regression results for percentage tuber yield loss on Superior ...... 86 Summary of stepwise regression results for percentage tuber yield loss on Russet Burbank . 86 Summary of stepwise regression results for percentage tuber yield loss on Atlantic ...... 87 Summary of regression analysis results for soil, root, and total nematode population densities on Superior ............ ..... ....... 91 Summary of regression analysis results for soil, root, and total nematode population densities on Russet Burbank .................. 92 Pearson correlation coefficients for percentage change in B, A, Jumbo, and Total tuber size classes with soil, root, and total nematode population densities for two time classed ...................................... 100 Impact of in—season nematode densities on percentage yield loss by cultivar, B, A, Jumbo and Total tuber size classes ................ 101 Summary of stepwise regression results for in-season total nematode population density on percentage tuber yield reductions ......... 101 X Table 29. Pears: after st0101 plant diffe: Table 30. Pears after stolo perce Table 31. Summa impac plant Table 32. Summa impac perce Table 33. Sum: of de part: Table 34. Sum: part: planl Table 35. Comp Yiell Table 35. Sum vari Table 37- Rela root stud POpu Table 38. Sum: of a Table 39. Nit: fert Table 40. Hit} fert Table 41- Nit] irr: Table Table Table Table Table Table Table Table Table Table Table Table Table 29. 30. 31. 32. 33. 34. 35. 36. 400 41. Pearson correlation coefficients for days after planting, delta (soil, root, total, and stolon) nematode population densities with plant growth parameters expressed as a difference ................. .................. 105 Pearson correlation coefficients for days after planting, delta (soil, root, total, and stolon) nematode population densities with percentage plant growth parameters ........... 105 Summary of stepwise regression results for the impact of aldicarb and P.penetrans on delta plant growth parameters ............... ....... 106 Summary of stepwise regression results for the impact of aldicarb and P;penetran§ on percentage plant growth parameters .......... 106 Summary of regression results for the impact of delta nematode population parameters on partitioning ratio ..... .. .................... 107 Summary of regression results for partitioning ratio as a function of days after planting .................. ..... .............. 107 Comparison of models for percentage tuber yield loss estimation ...... ..... ............. 113 Summary of beta coefficient signs for variables selected by regression procedures .. 113 Relationship between average number of soil, root, and total nematode samples reported per study and ability to explain nematode population variation .................. ....... 115 Summary of estimation procedures for impact of aldicarb on tuber yield ............. ...... 119 Nitrate/aldicarb leaching experiment at-plant fertilizer treatments ........................ 127 Nitrate/aldicarb leaching experiment nitrogen fertilizer treatments (lbs./acre) ............ 127 Nitrate/aldicarb leaching experiment irrigation treatments in (inches) ............ 128 xi Table 42. Table 43. Table 44. Table 45 Table 46 Table 47 Table 43 Table 49. Table so Table 51 Table 52 Nitrai result concel concel Compa: resul‘ Simul conse in th . Simul conse in th . Compa perce leach assoc strat . Sensi prodt . Marks benel Cumui conse . 1986' tYpe: - Estii in d: Prac . Summ 0f m mana Table 53 . Summ Perc aPpl LYPe (Sta SWit Table Table Table Table Table Table Table Table Table Table Table Table Table 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. Page Nitrate/aldicarb leaching experiment sampling results for leachate volume, nitrate concentration, and aldicarb metabolite concentrations ...... ..... ......... .......... . 130 Comparison of per acre simulated and observed results for risk-benefit parameters .......... 133 Simulated water stress factors for conservation and standard management strategies in the nitrate/aldicarb leaching experiment .. 134 Simulated nitrogen stress factors for conservation and standard management strategies in the nitrate/aldicarb leaching experiment .. 134 Comparison of simulated and observed percentage decrease in yield, nitrate leaching,and aldicarb leaching parameters associated with a switch to the conservation strategy ... ............. .. ........ . .......... 138 Sensitivity of revised SUBSTOR in relation to production system variables ............ . ..... 141 Market prices used in management strategy benefit analysis ...... ...... ....... .......... 143 Cumulative water applied to standard and conservation management strategies ..... ...... 145 1986-1988 Simulation results for five soil types and two management strategies .......... 145 Estimated management strategy profitability in dollars by year, soil type, and management practice ........................... ...... .... 145 Summary of nitrate mass leached and percentage of mass applied (%AP) by year, soil type, and management strategy ........................ . 147 Summary of aldicarb TTR mass leached and percentage of mass applied (%AP) by aldicarb application timing, management practice, soil type, and year ........ ..... ...... ...... . ..... 148 Decrease in profit and leaching measures (standard - conservation) associated with a switch to the conservation management strategy 149 xii Table 55. Table 56. Percei measu: assoc manag Total leach with strat Page Table 55. Percentage decrease in profit and leaching measures (1.0-conservation/standard) associated with a switch to the conservation management strategy .................... ...... 149 Table 56. Total decrease in profits, nitrate mass leached, and aldicarb mass leached associated with a switch to conservation management strategy in the prototype study area ......... 150 Figure 1. Syst Figure 2. GIS prod Figure 3. Locz Micl Figure 4. Ris] Figure 5. The Figure 6. GIS are. Figure 7. Fie Figure 8. Sim Figure 9. Cum imp Figure 10. Cum imp yie Figure 11. C1111 imp Figure 12. SO Figure 13' Roc Figure 15 . Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 10 11. 12. 13. 14. 15. LIST OF FIGURES Page System diagram for project development ...... 3 GIS analysis for uniform potato production regions .. ........................ 5 Location of Douglass Twn. Montcalm Co., Michigan study area ............. ............ 6 Risk benefit analysis information flow . ..... 10 The geocoding process .............. . ........ 19 GIS representation of soil types in study area ..................... .. .......... ....... 23 Field identification number polygon map ..... .24 Simplified SUBSTOR flow diagram .... ......... 43 Cumulative probability distribution for the impact of aldicarb on Superior tuber yield by tuber size class .......... ...... ......... 83 Cumulative probability distribution for the impact of aldicarb on Russet Burbank tuber yield by tuber size class ....... ..... . ...... 83 Cumulative probability distribution for the impact of aldicarb on Atlantic tuber yield by tuber size class ................... ...... 84 Soil nematode population density on Superior - simulated and observed vs. day of year .... 93 Root nematode population density on Superior - simulated and observed vs. day of year .... 93 Total nematode population density on Superior - simulated and observed vs. day of year .... 94 Soil nematode population density on Russet Burbank - simulated and observed vs. day of Year .0...It.........I.....I.O..QCOO00.0.0... 94 Fi re 16. Root gu Burb year Figure 17. Tota Burb year Figure 18. Supe dens POPc Figure 19. Abox rati Figure 20. Belc rati Figure 21. Tube day: Figure 22. Dif: par Figure 23. 150: lys Figure 24. Mas (AS deg Figure 25. Mid pro sta Figul‘e 26. End lay sta oh: yea Figure 23. Nil nil Fiilllre 29. Nit 0b: CU] Page Figure 16. Root nematode population density on Russet Burbank - simulated and observed vs. day of year ......... .. ......... . ................ ... 95 Figure 17. Total nematode population density on Russet Burbank - simulated and observed vs. day of year . ......... .... ....... . .......... ... ..... 95 Figure 18. Superior check soil nematode population density regression residual vs. initial population density ......... ..... ............ 97 Figure 19. Above-ground to total biomass partitioning ratio vs. days after planting . .............. 108 Figure 20. Below-ground to total biomass partitioning ratio vs. days after planting ..... ....... ... 109 Figure 21. Tuber to total biomass partitioning ratio vs. days after planting ..... .................... 109 Figure 22. Difference in aldicarb and check treatment partitioning ratios vs. days after planting . 110 Figure 23. Isometric projection of non-weighing lysimeter construction ......... ............ . 125 Figure 24. Mass of aldicarb (ALD), aldicarb—sulfoxide (ASO), aldicarb—sulfone (ASN) total mass degraded (DEG) vs. day of year ....... ....... 130 Figure 25. Mid-season distribution of TTR in the soil profile for simulated conservation and standard management strategies .............. 131 Figure 26. End of season distribution of TTR in soil layer for simulated conservation and standard management strategies .............. 132 Figure 27. Nitrate/aldicarb leaching experiment observed nitrate mass leached vs. day of year ocoooo-oooooo-o ..... otofioloooooloocoanoo 135 Figure 28. Nitrate/aldicarb leaching experiment nitrate mass leached vs. day of year ........ 135 Figure 29. Nitrate/aldicarb leaching experiment observed cumulative nitrate mass leached vs. cumulative drainage ......... ...... .......... 136 Figure 30. Nitra simul cumul Figure 31. Nitra simul vs. c' Figure 32. Cumul and 2 Figure 30. Figure 31. Figure 32. Page Nitrate/aldicarb leaching experiment simulated cumulative nitrate mass leached vs. cumulative drainage ................ ......... 137 Nitrate/aldicarb leaching experiment simulated total toxic residue mass leached vs. day of year .......... ....... . ........... 138 Cumulative rainfall during 1986, 1987, and 1988 ...................... .............. 144 xvi Agricultur. problem to anal type, farming p which affect po complexities de irrigation, Che relatively unif impact of regic AnalYtical tool Production inpt atool may be I °Ptimization a] contamination. was The goal I Capable 0f est. ec°n°mi° benef fertiliers an' system Was int management pra CHAPTER I INTRODUCTION Agricultural non-point source pollution is a difficult problem to analyze because of spatial variation in soil type, farming practices, precipitation and other factors which affect pollution occurrence and severity. System complexities develop from the relationships between irrigation, chemical movement, and crop development under relatively uniform environmental conditions and from the impact of regional variation in soil types and land use. Analytical tools are needed to study the trade-offs between production input values and the non-point source risks. Such a tool may be used to meet the dual challenge of crop yield Optimization and mitigation of agricultural ground water contamination. Goal and Objectives. The goal of this project was to develop a system Capable of estimating ground water contamination risks and economic benefits associated with the use of nitrogen fertilizers and aldicarb in regional potato production. The SYstem was intended to be sensitive to agricultural management practices, and spatial variation in environmental parameters. This thes interrelated c: 2) alternate m movement and d model developm benefit analys of the study 9 and system dew were accomplis Spatial data 9 integration t4 2 To meet the goal, five thesis objectives were defined: 1. Quantify spatially variable factors important to potato production (i.e. weather, land-use, soils) in a prototype study area. 2. Identify alternate management strategies which would reduce risk of aldicarb and nitrate ground water contamination while sustaining profitability. 3. Expand SUBSTOR's capabilities to include degradation and movement of aldicarb and its oxidative metabolites in the soil environment 4. Expand SUBSTOR's capabilities to include estimation of the impact of aldicarb and Pratylenchus penetrans (Root— lesion nematode) on potato tuber yield 5. Integrate simulation modeling with geographic information systems to facilitate model parameterization for regional potato production risk—benefit analysis. Project Overview Proiect Organization This thesis was divided into five distinct but interrelated categories; 1) spatial data base development, 2) alternate management practice identification, 3) aldicarb movement and degradation model development, 4) yield impact model development, and 5) simulation modeling for risk benefit analysis (Figure 1). Given the comprehensive nature of the study goal, a diverse array of research procedures and system development parameters were needed. Objectives were accomplished using a variety of procedures including: spatial data gathering, literature reviews, literature integration techniques, computer programming, and _ SEOEZD _ m.m>..(2<.r Ems-HI. In; m u).hUualO i 02 Dam—m EU wOOUwK 20:.(Zuam0 Dam; $(E ma: TU 05. uZ°U>J°A m4 OF >m—bUml h¢w>20U .uJU—u ugh—0.0 — NIKU meg-UH (F‘nu C WIF(N)> KWIF(W>> WP¢WKU WM(I° uw(l (ha‘fl dih‘tn 6-. fl>uuu~l° x2: MEGOEAQ v3.1 mhdxtz >:.:m<.:u_0zn_ 4420.03. 1 ZOFdiOth. moOU Z4<2<.<0m_2 mOhmmDm 20.53023 3055... 20.254245. 052552 02:03 ...(thEZOz52m ZOFEZOU 5.10:2: m_w>4_._.Um_.nO pCoEQ0H0>oc poomoum How thmcfic Eoum>m .H ousmflh >>m_>mz qu:.Uw:—O 20.541052. wOOU Zw: mehdimh: 20F°§ m¢_.—Uu30 hzthTpZmSEUdixx—z >>m.>uz szhd‘zwhS mm_Uu.—_._.Uu;O .552; :0. 3: A 3.0% ..odio iJ m._E m5 9—. m200>40d m4 n22 >u>m3m dOm 3.. 20050.. Li 8.30208 All 0.192.003» Um 523200 ...om 3.00.0 0» 1.55. <20 2.3 253 205000.... Ai 025.. Excezo 20:45.0 525.2 ”3023 m0. 20.55552. . 5.? >005 mi 05.... .v 52: 02005 «2 #5232 03025 G o» _ «05.... ¢.< . «20 5.40 a 02 0.3: 20.2.2930 0.5: «.0 00003. 5: 20 9 “200.50.. 2 as... 3: 20030.. 3.2.4.0209. 0.1.3.003» 2:200 ed; 05.0.0 op .1.sz 3.05.3 2.0a 0.... L $53.5 E35 1 <20 5540.5 355 035 <50 255.3 .— 02502.0 multivariate st activity relate W Informatio grower controll land use was cc was performed L System. ERDAS GIS is an intec entry, manipula multiple layer Information on electronically relationship b used to electr features in or to regional po Land use prototype stud of Douglass to 3). A four—es enough to show small enough 5 manageable in The comp: information 01 Potatoes were 4 multivariate statistical analysis. An overview of the activity related to each thesis objective is presented. Thesis Objective 1 - Spatial data base Information pertaining to spatial variability in non- grower controlled variables such as rainfall, soil type, and land use was collected and analyzed. Spatial data analysis was performed using ERDAS Earth Resource Data Analysis System. ERDAS is a geographic information system (GIS). GIS is an integrated software package designed for the entry, manipulation, analysis, and display of single or multiple layers of spatial referenced information. Information on weather, land use, and soils can be electronically overlaid to produce new maps based on the relationship between map features (Figure 2). ERDAS was used to electronically overlay and relate multiple map features in order to show geographic relationships important to regional potato production. Land use and soil type maps were developed for a prototype study area which included sections 8,9,16, and 17 of Douglass township in Montcalm County, Michigan (Figure 3). A four-section area was used so that it would be large enough to show the impact of spatial variation, but yet small enough so that data handling would be sufficiently manageable in a comprehensive modeling system. The completion of Thesis Objective 1 provided information on soil types in the study area on which potatoes were produced in 1986-1988. Files representing Locati C I. 2. GI igure 1“ regions . >4 Weather Uniform Potato Production Area-5 Location X,Y Figure 2. GIS analysis for uniform potato production regions. Figure 3. Location of Douglass Twn, Montcalm Co., Michigan soil physical c years were use Thesis 0b ' ecti A review to grower fiel potato managem create input f standard growe management pra to optimize th profitability Thesis Ob'ect' A model developed and plant growth a the scientific SUBSTOR, ORgans was de‘ Ritchie and It growth and de‘ as: an aid to multi-year ri yield forecas research need SUBSTOR available we: simulates phm 7 soil physical characteristics and weather for each of these years were used as SUBSTOR input parameters. Thesis Objective 2 - Alternate management strategies A review of the scientific literature and sensitivity to grower field practices were used to identify alternate potato management strategies. This information was used to create input files for simulation models representing standard grower practices and a hypothesized improved management practice. The intent of the improved system was to optimize the relationship between agricultural profitability and risk to ground waters. Thesis Objective 3 - Aldicarb movement and degradation A model of aldicarb movement and degradation was developed and integrated with SUBSTOR, an existing potato plant growth and development model. Findings reported in the scientific literature were used for model development. SUBSTOR, simulation of Underground Bulking STorage ORgans was developed at Michigan State University by Dr. Joe Ritchie and Mr. Dale Magnusson. SUBSTOR is a S.tuberosum growth and development model. SUBSTOR was designed to serve as: an aid to within-year crop management decisions, for multi-year risk analysis and strategic planning, large area yield forecasting, and to assist in the definition of research needs. SUBSTOR operates on a daily time step and uses readily available weather, soil, and genetic data inputs. The model simulates phenological development, soil water balance, and nitrogen transf Eight subr SUBSTOR. The 2 programs capabi movement and de estimation of 1 irrigation man Thesis Ob'ecti The impac was determined and meta-analy statistical a multivariate regression pr analysis was 11 aldicarb and g of SUBSTOR cor growers. Thesis Objectf The appem and 4) was usr factors ident nitrate leach environmental determined th 8 nitrogen transformation in the potato production system. Eight subroutines were developed and integrated with SUBSTOR. The addition of new subroutines upgraded the programs capability to include an estimation of aldicarb movement and degradation. SUBSTOR modifications allowed for estimation of risks associated with alternate aldicarb and irrigation management. Thesis Objective 4 — AldicarbzRoot-lesion nematode impact The impact of aldicarb and P.penetrans on potato yield was determined using integrated research review techniques and meta-analysis. Meta-analysis included a variety of statistical analysis methods such as, analysis of variance, multivariate general linear regression, and stepwise regression procedures. Information obtained from the meta— analysis was used in SUBSTOR for estimation of the impact of aldicarb and P.penetrans on tuber yield. The new version of SUBSTOR could then estimate of the value of aldicarb to growers. Thesis Objective 5 - Riskibenefit analysis The appended version of SUBSTOR (Thesis Objectives 3 and 4) was used to estimate impacts of selected management factors identified in Thesis Objective 2, on potato yield, nitrate leaching and aldicarb leaching under each set of environmental conditions of soil type and land use determined through Thesis Objective 1. Upon comp performed unde under Thesis 0 analysis syste transferring i into computer to determine u subregion, dat collected and Simulation ou profitability, potential ass schemes (Figu The poter test the impac soil types in information a‘ impacts of th irrigation, a and environme integrated me The risk bene designed to 1 makers throng for dealing ‘ Risk-benefit analysis system Upon completion of simulation modeling upgrades performed under Thesis Objectives 3 and 4, the procedures under Thesis Objectives 1, 2, and 5 formed a risk-benefit analysis system. Data flow in this system consisted of transferring information on soil type, land-use, and weather into computer readable formats followed by matrix analysis to determine uniform potato production subregions. For each subregion, data necessary for simulation model operation was collected and used for simulation of management scenarios. Simulation output was used to show regional variation in profitability, nitrate leaching and aldicarb leaching potential associated with alternate production management schemes (Figure 4). Project Justification The potential expense required for field research to test the impacts of alternate management strategies on all soil types in a region limits the amount and quality of information available to agricultural decision makers. The impacts of these multiple factors (i.e. soil type, rainfall, irrigation, and pesticide application timing) on production and environmental concerns can best be analyzed through an integrated modeling approach (Wagenet and Hudson, 1986). The risk benefit analysis system developed in this study was designed to provide information for agricultural decision makers through integration of the best available technology for dealing with potato production system complexities. DATA SOURCE MSU Potato Research Farm, CR-Zi Weather Station DNR Photo #3 MDNR 78-48 40 Ground Truth 56 Soils Map Simulation Modeling 10 DAT DATA SOURCE DATA ELEMENT INTERPRETAATION MSU Potato Research Farm, CR-Zl Weather Station VVeather DNR Photo #3 MDNR 78-48 40 Characterization Field NO. # _) of regional I I I I \;' potato Ground Truth : production I Land Use i E I SCS Soils Map i 1 Soil type : I u__._________..__.._____ ____-_ ...—---- Potato production under uniform conditions Simulation Modeling Pot/SoilNVeather _._......_.._...__..-_.______.__ ---—.... ______ --....— _____.....__ / For each PoUSoilNVeather/ management strategy Trade-offs between -- Estimated Profitability —"" management --_ Mass of N03' leached, strategies out of soil profile -- Mass of aldicarb leached out of soil profile Figure 4. Risk-benefit analysis information flow. If a chem toxic then it manager. Howe then managemen The two source nitrate nitrog mm In soil, to nitrates b soluble are n leaching. Th plant uptake. converted it movement is a Under saturatt microorganism: aconversion 4 Nitrate i abides by the supplies may nitrogen (Moll nitrate conoe methemoglobin lethemoglobin blood stream oxygen. Thii 11 If a chemical moving into ground water supplies is non- toxic then it may be of little concern to an agricultural manager. However, if the compound is toxic or persistent then management of material leaching is of major importance. The two sources of risk considered in this system were nitrate nitrogen and aldicarb metabolites. Nitrogen In soil, non-nitrate forms of fertilizer are converted to nitrates by soil microorganisms. Nitrates are water soluble are not absorbed by soil and thus subject to leaching. The movement of nitrate in soil is impacted by plant uptake. Plants both remove nitrate from soil and converted it to an immobile organic form. Nitrate mass movement is also impacted by the denitrification processes. Under saturated soil conditions some anaerobic microorganisms use nitrate as an oxygen source resulting in a conversion to nitrogen gas. Nitrate Risk. The Michigan Department of Public Health abides by the EPA standard that public drinking water supplies may not contain more then 10 ppm of nitrate nitrogen (McWilliams, 1984). Ingestion of water containing nitrate concentrations greater than 10 ppm may cause methemoglobinemia in infants under the age of six months. Methemoglobinemia occurs when nitrates enter the infant's blood stream decreasing the blood's ability to carry oxygen. This may result in slightly retarded body growth, reflexes or death (Dorsch et al., 1984). Methemoglobinemia known as the ' because of shi Cows and water are also Nitrate can 1) risk of chron' also been cor Montcalm Coun concentration 1986). An an indicated bac be 1-2 ppm ( above this Is seepage, muni (lloWilliams, production is occurrence on fertilizers, Occurre may also sen ‘In Iowa, 67% nitrate level at al., 1986} significantll 12 known as the 'blue baby syndrome' does not affect adults because of shifts which occur in blood pH during childhood. Cows and other cud-chewing animals that drink well water are also at risk from nitrates in ground waters. Nitrate can be reduced to nitrite in the rumen increasing risk of chronic disease. High nitrate in ground waters has also been correlated with spontaneous abortion of litters in swine (McWilliams, 1984). Regional Nitrate Concern. Public health records in Montcalm County, Michigan revealed groundwater nitrate concentrations above the 10 ppm health standard (Erving, 1986). An analysis of public well water quality records indicated background levels of nitrate in ground waters to be 1-2 ppm (Kruska, 1986; Hallberg, 1986). Concentrations above this level may come from agriculture, septic tank seepage, municipal waste sites, or feed lot operations (McWilliams, 1984; Singh and Sekhon, 1979). Potato production is considered to be a probable cause due to its occurrence on sandy and sandy loam soils, use of nitrogen fertilizers, and use of irrigation. Occurrence of elevated nitrate levels in ground waters may also serve as an indicator of pesticide contamination. In Iowa, 67% of well water samples which contained elevated nitrate levels also contained pesticide residues. (Kelley et al., 1986). Pesticide concentrations were not significantly correlated with nitrate concentrations but co- occurrence was significantly correlated. Aldicarb Aldicarb nematicide. I appreciably bi 1980). This ground water Aldicarb when it was f Island, New Y Aldicarb's re in Long Islan water. Aldicarb water in Wise (Rothschild e‘ been concerns ground water (Jones and Ba llaine ground production (ll Aldicarr is such, it 1 ppb is the he as its maxim 31., 1982). most ground l 13 Aldi_cer_b Aldicarb is a systemic insecticide and contact nematicide. It is highly water soluble and does not appreciably bind to the soil matrix (Bromilow and Leistra, 1980). This makes aldicarb susceptible to leaching into ground water supplies. Aldicarb was first discovered in ground water in 1979 when it was found in shallow test wells in eastern Long Island, New York potato fields (Zaki et al., 1982). Aldicarb's registration as a nematicide has been restricted in Long Island, New York due to its presence in ground water. ’ Aldicarb residues were detected in irrigation well water in Wisconsin's Central Sands potato production region (Rothschild et al., 1982). The state of Florida has also been concerned with the possibility of aldicarb moving into ground water as a result of that state's citrus production (Jones and Back, 1984). Aldicarb has also been found in Maine ground waters in regions associated with potato production (McWilliams, 1984). Aldicarb Risk. Aldicarb is a cholinesterase inhibitor. As such, it is highly toxic. In New York a concentration 7 ppb is the health advisory level, while the EPA sets 10 ppb as its maximum recommended limit in ground waters (Zaki et al., 1982). Once under the anaerobic conditions found in most ground waters, aldicarb degrades very slowly (Lemley and Zhong, 1984; Bank and Tyrell, 1984). Benign County, Michig contamination. Back, Romine, water numerica 9.1 for the Ce rating in Cent these regions ground water. numeric index thirteen envi categories of ground water. 14 Regional Aldicarb Concern. The aquifers in Montcalm County, Michigan also appear to be at risk to aldicarb contamination. This area has received a 5.2 rating using Back, Romine, and Hansen's aldicarb appearance in potable water numerical index. This can be compared to a rating of 9.1 for the Central Sands region of Wisconsin and a 5.1 rating in Central Florida (Back et al., 1984). Both of these regions have experienced problems with aldicarb in ground water. The aldicarb appearance in potable water numeric index was developed based on relationships between thirteen environmental factors which fell under the general categories of application, degradation, transport, depth to 'ground water. production (T weather facto quantified fo potato produc The regi Sections 8,9, County, Michi long history State Univers for ground wa Thirty-f area overlay origins (Unit Optimal potat these soil ty Material software used how the maps the soils on the study. CHAPTER II REGIONAL DATA BASE DEVELOPMENT Development of a regional data base was necessary to quantify spatially variable factors important to potato production (Thesis Objective 1). Soil, land use, and weather factors were included. Soil and land use data was quantified for use in ERDAS for determination of uniform potato production areas. The region used for the prototype study area included Sections 8,9,16, and 17 of Douglass Township, Montcalm County, Michigan. This region was chosen because of its long history of potato production, proximity to the Michigan State University Potato Research Farm, and regional concern for ground water quality. Thirty-five different soil types in this four section area overlay unknown depositional materials of glacial origins (United States Department of Agriculture, 1960). Optimal potato production strategies may differ for each of these soil types (Awad, 1984; Pionke and Urban, 1985). W Materials refer to the spatial data sources and software used to meet Thesis Objective 1. .Methods refer to how the maps were handled to produce a matrix output showing the soils on which potatoes were produced in each year of the study. 15 Spatial Data B The first the soil types L a. 'J.’ base developme potato produci in this study Data Source Informat potato produc sources. m. Department 0 Survey of Mo area was cov map scale was section corne TileBs was obtained Resources. 1 September 4t] 40. Color 51 the Montcalm but was not Lamina Ground truth for these op #‘ S atial Data Base The first step in data base development was to quantify the soil types, field boundaries, land uses, and weather characteristics in the study area. The second step in data base development involved using ERDAS to create uniform potato production area maps (Figure 2). Spatial data used in this study was obtained from several different sources. Data Source Information used in the development of the regional potato production data base came from many different sources. Soils. Soils maps were obtained from the United States Department of Agriculture Soil Conservation Service Soil Survey of Montcalm County Series 1949, No.11. The study area was covered by Map Sheets number 23 and number 24. The map scale was 1:20,000. Soils maps were georeferenced using section corner coordinates. Field Boundaries. Aerial photography of the region was obtained from the Michigan Department of Natural Resources. Black and white imagery was obtained from a September 4th flight during 1978, print number 3MDNR 78—48 40. Color slide imagery from 1986 was also available from the Montcalm County Office of the Soil Conservation Service but was not used because of oblique projection. Land use. Land use information was obtained through ground truthing. The field boundary map was used as a base for these operations. Weather . period 1974 to Service for a Michigan. Wee to 1988 from t Because of its Potato Researt Data processil Data pro map and photo translation i Air photo int Land use photo interp field and 1a roads, tree 1 Rectification "1 The spat aerial photog USGS topogra; using a Bausc map and the a the topograpl unnecessary. Geocoding - Soils a: input into E: # 17 Weather. Weather information was available for the period 1974 to 1986 from the Cooperative Crop Monitoring Service for a weather station located in Entrican, Michigan. Weather information was also available from 1985 to 1988 from the Montcalm County Potato Research Farm. Because of its greater accuracy in solar radiation data, the Potato Research Farm data was used for all analysis. Data processing Data processing consisted of air photo interpretation, map and photo spatial rectification, followed by image translation into computer readable formats (geocoding). Air photo interpretation Land use information was obtained through manual air photo interpretation of DNR image 3MDNR 78-48 40. Major field and land use boundaries were delineated by fence rows, roads, tree lines, water bodies, and textural changes. Rectification The spatial consistency of the soils map, and the aerial photograph was checked by projection onto a 7.5 min USGS topographic base map (Six Lakes and Edmore quadrangle) using a Bausch & Lomb Zoom Transfer Scope. Both the soils map and the aerial photograph were spatially consistent with the topographic base map. Map rectification procedures were unnecessary. Geocoding Soils and land use boundaries were then geocoded for input into ERDAS. The geocoding process consisted of three 1 W... ‘F‘ stages: digit' conversion to Di itiza using an elec State plane c and field be segments whic The region in value which r the polygon. Rasteriz the soil and Hit raster rows. Each p yard square Soil Conserva CRIBS (P for the polyg converted the dimensioned s an empty rast POLYFILL map; the appropriz was used to < §;§ Cre; operation wa: which remove- 18 stages: digitization, creation of raster file, and conversion to GIS file (Figure 5). Digitization. The digitation process was completed using an electronic digitizing board and Mapdigz software. State plane coordinates were used for gee-referencing. Soil and field boundaries were represented by a series of line segments which formed complete polygons on each boundary. The region inside of each polygon was assigned an attribute value which represented the soil type or field number inside the polygon. Rasterization. The spatial information contained in the soil and field number polygon files was translated into 8-bit raster files. The raster grid was 355 columns by 355 rows. Each pixil in the grid represented a 10-yard by 10- yard square corresponding to the minimum mapping size in Soil Conservation Service soil maps. CRIES (POLZDIG, CREATE, and POLYFILL) Software was used for the polygon—to-raster conversion process. POL2DIG converted the overlapping polygons of a digitizer file into dimensioned strings of attribute data. CREATE initialized an empty raster structure of appropriate dimensions. POLYFILL mapped dimensioned strings of attribute data into the appropriate location within the raster structure. ERDAS was used to convert rasterized attribute data to a GIS file. GIS Creation. The output file from the POLYFILL operation was input to the ERDAS strip application program which removed CRIES header information. The pixil index A. Uncod C. Appro map by ce Figure 5. Th‘ unooded base Segments lin? attribute in file c, whic? reconstructi A. Uncoded base map B. Areas with assigned numerical codes. Non- features keyed as zero C Approximation of base map by cells. D. Reconstruction of base map. Figure 5 The geocoding process — Map feature boundaries of uncoded base map A, are represented by a series of line segments linked to numeric attribute codes B. Boundary'and attribute information is converted into a grid based raster file c which is subsequentially converted into a 613 reconstruction of the base map D origin for th left of the f ERDAS REVERSE coordinates. ERDAS FIXHE Soils GIS link to the County Soil forms. Land use GIS Field n unique but n on July 28, order to det truthing, fi interpretati (Table 2). ERDAS to inc' analysis. 20 origin for the CRIES GIS system was located at the bottom left of the file while the ERDAS origin is on the top left. ERDAS REVERSE application was used to renumber pixil coordinates. The geocoding process was then completed using ERDAS FIXHEAD to create a new GIS header file. Soils GIS - Soil series attribute values coded during digitization (Table 1) were used to identify soil series, and provide a link to the soil chemical/physical properties table of the County Soil Survey as well as Soil Conservation Service 232 forms. Land use GIS Field number values created during digitization were unique but nominal. Ground truthing observations were made on July 28, 1986; August 23, 1987; and August 17, 1988 in order to determine land use in each year. During ground truthing, field numbers obtained from air photo interpretation were linked with land-use attribute codes (Table 2). The field number map was then re—coded using ERDAS to indicate regional land-use in each year of the analysis. Table 2. Ian used in geoc lTT M 1 HH 2 3 4 s 6 7 8 9 0 1 2 J—‘H ...-J Table 1. Soil series names and attribute numbers used in nonnadina. ATT SCS % M SYMBOL SOIL SERTRS NAME SLOPE 1 AQ Water 2 Aa Alluvial land 3 Ca Carlisle Muck o - 2 4 Eb Ensley loam and Edmore loamy fine sand 0 - 2 5 Ec Epoufette loamy sand and Ronald sandy loam 0 - 2 6 Ga Gladwin loamy and sand and Palo sandy loam 0 - 2 7 Go Grayling sand 0 - 2 8 Gd Grayling sand 2 — 6 9 Ge Grayling sand 6 —10 10 Gg Grayling sand 10 —18 11 Gk Greenwood and Dawson peats 0 - 2 12 Kc Kerston muck 0 - 2 13 Mb Mancelona loamy sand 0 - 2 14 Mc Mancelona loamy sand 2 — 6 15 Md Mancelona loamy sand 6 -1O 16 Mh McBride and Isabella sandy loams 0 — 2 17 Mt Montcalm loamy sand and sandy loam 6 -10 18 Mw Montcalm and McBride loamy sands and sandy loams 0 - 2 19 Mx Montcalm and McBride loamy sands and sandy loans 2 — 6 20 Ra Rifle and Tawas peats 0 - 2 Table 2. Land-use attribute codes used in flanrndina. ATT N_0_i LAND USE l Potatoes 2 Corn 3 Soybeans 4 Forage Crops 5 Small Grains 6 Grass and Open 7 Apple Orchard 8 Christmas Trees 9 Forest Covered 10 Farmstead 11 Urban Residential 12 Cemetery 13 Marsh 14 Water 15 Unknown 16 Cucumbers Weather data Weather only to be co simulation no the file imp manager. Th was re-coded Soil type wa column varia each year of soil types in The geo lproduced a map represen produced (Pi truthing (Ta developed i W respe Weather data Weather data was already in digital format and needed only to be converted to the proper format for SUBSTOR simulation modeling. This conversion was accomplished using the file import and export functions of a LOTUS 123 data manager. The spatial distribution of available weather data forced the assumption that weather characteristics would be uniform across the study area. Potato production analysis The next step was to determine the soil types on which potato production occurred during the land use year. This was accomplished using the ERDAS MATRIX operation. Land use was re-coded dichotomously as (potato=1, non-potato=2). Soil type-was used as the row variable and land-use was the column variable. The MATRIX operation was performed for each year of the study resulting in new GIS files showing soil types under potato production and relative acreage. 133—em The geocoding process conducted under Thesis Objective 1 produced a GIS file representing soil types (Figure 6). map representing field identification numbers was also produced (Figure 7). Using information gained from ground truthing (Table 3), maps of study area land use were developed for 1986, 1987, and 1988 (Appendix A, Figures 1,2,3 respectively). I WW om mm Q 99 ”g °° i as m Nice-'70 w n . .Q QOAOwaQ 2 ii” QCLQOQQZ N. "a zjq ZZqu moo <2; (<30 q'qqz <1 ©3300 ZZZQEEEE _J_J_J zzzzgzgggmqqqz FUUUCfit—hb >>>>mmzzzmzzzm qqqumdddu0004 xofimcfimw u. l inn % IR: due coowaom gonad: :oflumoflMHpcopfi names .5 ounces Table 3. Field identit ID No. 1986 1987 ‘1 Forest Forest ZC-tree C-tree 3 Forest Forest 4 HM C-tree 5Com Open 6 Corn open 7 Wm C-tree 8 Corn Open 9 Potato Forage 10 Forage Open 11 San-grain Open 12 Forest Forest 13 Open open 14 Farmsted Farmste 15 Fanusted farm“ 16 tom Potato 18 torn Open 19 Corn Corn 20 torn Corn 21 Corn Corn 22 mm Forest 23 Forage Forage 24 Forage Forage 25 Unkm“ Unknown 26 Soybean c'tl'ee 27 Smell" C'tree 28 Farnsted Farmer: 29 “”5th Farmso 30 Famed Farmst. 31 Forage Forage 32 c-tree Forage 33 Forage Forage 3!. Corn Forage 35 Forage FOPage 36 Forest F°rest l7 Farmed Farms t “Famed Farm Forest FOrest 40 Famted F 3mm ‘1 “rusted ”mist ‘3 “rage ‘4 POtato ‘5 Potato ‘6 Forage ‘7 Forage ‘8 Corn open 255 Table 3. Field identification numbers and ground~truthed land use for 1986. 1987 and 1988 1987 1988 ID 4*— 1986 1987 1988 ‘Om‘l‘FU‘bWN—h WUMUWNNNNNNNNNN—Id—I—IA—A—h—e—h—b (bum—aoomwomewm—aoomwombuN—no 35 36 37 38 39 40 41 42 43 44 45 46 47 48 Forest Open Farmsted Farmsted Potato Open Open Corn Corn Corn Forest Forage Forage Unknown C-tree C-tree Farmsted Farmsted Farmsted Forage Forage Forage Forage Forage Forest Farmsted Farmsted Forest Farmsted Farmsted Forage Forage Potato Potato Potato Potato Open Forest Open Farmsted Farmsted Corn Potato Forage Forage Forage Corn Forest Forage Forage Unknown C-tree C-tree Farmsted Farmsted Farmsted Corn Forage Forage Corn Forage Forest Farmsted Farmsted Forest Farmsted Farmsted Potato Potato Potato Potato Corn Open Corn 96 Sm-grain Sm-grain Sm-grain Farmsted Farmsted Forest Corn Corn Open Farmsted Open Open Corn Potato Farmsted Forest Open Corn Farmsted Unknown Open Forest Forest Alfalfa Open Forage Forage Forest Hater Farmsted Open Hater Potato Forage Forage Forest Forage Corn Farmsted Open Farmsted Open Open Open Open Farmsted Forest Open Corn Farmsted Unknown Open Forest Forest Open 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 Farmsted Farmsted Farmsted Corn Corn en Farmsted Farmsted Farmsted Potato Corn Corn Open Open Open Urban Urban Urban Open Open Open Open Open Open Corn Potato Corn Farmsted Farmsted Farmsted Corn Open Open Forage Forage Open Corn Open Open Marsh Marsh Marsh Apple Apple Apple Marsh Marsh Marsh Corn Open Open Potato Forage Forage Corn Open Open Forest Forest Forest Forest Forest Forest Farrnsted Farmsted Farmsted Corn Forage Forage Farmsted Farmsted Farmsted Forage Forage Forage Forest Forest Forest Open Open Open Open Open Open Forest Forest Forest Alfalfa Open Open Farmsted Farmsted Farmsted Farmsted Farmsted Farmsted Potato Open Open Open Open Open Farmsted Farmsted Farmsted Open Open Open Forage Forage Forage Corn Corn Open Open Open Open Open Open Open Open Open Open Open Open Open Open Open Open Corn Open Open Farmsted Farmsted Farmsted Corn Open Open Cemetary Cemetary Cemetary ERDAS MA types for whi the study (Te Table 4. Sci] Soil type Epoufette Grayling Mancelona McBride Montcalm The dat Table 4 was necessary to Practices on “W area. needed to be Twanty Simu] fa°t°r5 eact Deelinj because a In: Study area f to inc1u de I table showi] regiOn W0111< handle that 26 ERDAS MATRIX analysis provided information on soil types for which potato production occurred for each year of the study (Table 4). Table 4. Soils on which potatoes were produced in 1986. 1987. and 1988. Acres in production Soil type 1986 1987 1988 Epoufette 3 . . Grayling 9 2 . Mancelona 24 8 . McBride 100 57 91 Montcalm 8 30 . Total 144 97 91 EEEE§£1 The data on potato production by soil type provided in Table 4 was used to determine what simulations were necessary to estimate the impact of alternate management practices on associated risks and benefits in the prototype study area. For example, potato production on Grayling soil needed to be simulated for 1986 and 1987, but not for 1988. Twenty simulations were needed, ten sets of environmental factors each with two alternate management scenarios. Declining acreage over three years is evident, perhaps because a major potato packing company moved out of the Study area in 1986. If the study area had been large enough to include multiple weather data sets, then a three-way table showing potato production by soil type, and weather region would be required. ERDAS has the capability to handle that condition. ALTEi Alternai the desire t< benefits to 1 agricultural alternate ma] scope of sta] intended to I The alternat implemented ' machinery re background n SYstem inter deve10ped ba quantified f TSSOCiated I The lit Information intended to reTional em Production I °f irrigatic assessment : waters Can 1 CHAPTER III ALTERNATE MANAGEMENT STRATEGY DETERMINATION Alternative management strategies were identified with the desire to meet the dual challenge of optimizing economic benefits to the grower, while protecting ground waters from agricultural non-point source contamination. These alternate management strategies were to fall within the scope of standard chemical intensive practices and were intended to be modifications on existing management schemes. The alternate management strategy was one which could be implemented by growers with little change in cropping or machinery requirements. A literature review provided the background necessary for understanding potato production system interactions. Alternate management practices were developed based on this information. These strategies were quantified for input into SUBSTOR which was used to estimate associated risks and benefits. W The literature review is divided into three sections. Information pertaining to potato production scope is intended to place the project study area within a larger regional environment. Information pertaining to selected production management components describes the use and value of irrigation, nitrogen, and aldicarb application. The risk assessment section provides ideas on how risk to ground waters can be minimized while profitability maintained with 27 out radical c miracle Michiga1 the United Si 198?). Betwa potatoes werl (Michigan Ag Michigan, po peninsulas p production e Service, 198 Montcal Michigan for Agriculture, were plantec average yiei (MiChigan At W Three 1 Study, The nitrogen fe aPplication Irrigation As a s one inch of to five day °r°P needs 28 out radical changes in crop production practices. 22£§£Q_ELQQ¥HH£EL§£QE§ Michigan is the tenth largest producer of potatoes in the United States (Michigan Agricultural Statistics Service, 1987). Between 1985 and 1987 an average of 53,000 acres of potatoes were planted with an average yield of 261 cwt/acre (Michigan Agricultural Statistics Service, 1988). In Michigan, potato production occurs in both upper and lower peninsulas providing for great variation in potato production environments (Michigan Agricultural Statistics Service, 1988). Montcalm County is ranked first in the state of Michigan for potato production (Michigan Department of Agriculture, 1986). An average of 12,950 acres of potatoes were planted between 1985 and 1986 in Montcalm County. The average yield of marketable tubers was 332 cwt/acre (Michigan Agricultural Statistics Service, 1988). Selected Management Components Three potato management factors were considered by this study. They are irrigation application amount and timing, nitrogen fertilizer amount and timing, and aldicarb application timing. Irrigation As a standard practice in Michigan potato production one inch of irrigation water is usually applied every three to five days when natural rainfall is not sufficient of meet crop needs (Vitosh, 1987)~ migefi water stress. water stress to meet crop region indic; yields from Comparison 0 University M 1988 (a very tuber yields irrigation t m the movement Presence of Of water. I natural rair Plant uptak, irrigation ; i“creases t] also the am. °r nitrates AS W011 relationshi c°ntaminati nitrate Con correlated 29 Irrigation Value. The potato plant is susceptible to water stress. Irrigation is used in Michigan to reduce crop water stress when natural precipitation is not great enough to meet crop needs. Research in Wisconsin's Central Sands region indicates that in some years irrigation may increase yields from 100-200 cwt/acre to 500 cwt/acre (Butler, 1978). Comparison of two experimental plots at the Michigan State University Montcalm Potato Research Farm indicates that in ' 1988 (a very dry year) irrigation may have increased total tuber yields as much as 331 cwt/acre. As such, the value of irrigation to potato production is considerable. Irrigation Concern. Two conditions are necessary for the movement of chemicals out of the root zone. They are: presence of the compound in the soil, and downward movement of water. Downward movement of water is a function of natural rainfall, irrigation, soil type, evaporation, and plant uptake (McWilliams, 1984). Of these factors, irrigation is the most easily controlled. Irrigation increases the amount of water available for the plant and also the amount of water available for movement of aldicarb or nitrates. As would be expected, there appears to be a relationship between application of irrigation to crops and contamination of ground waters. In Holt County, Nebraska, nitrate concentration in ground water was significantly correlated (r2 0.66) with the age of irrigation wells. The analysis revealed nitrate levels increasing in shallow ground water: irrigation we and Spalding Using CI applied fert zone (Hubbart rainfall dis When ir with the one application frequent has application 1976). Nitrogen Mitroge acreage to i Nitroge ' Virtually a2 acreage. A lbs/acre spl allplication: urea! potas; ammonia. “Tm tubfirs. I 30 ground waters on an average of 4.92 ppm for each year irrigation was applied to nitrogen fertilized corn (Exner and Spalding, 1979). Using conventional irrigation practices 17% to 53% of applied fertilizer is expected to leach below the rooting zone (Hubbard, 1984; Hallberg, 1986) depending on natural rainfall distribution and irrigation management scheme. When irrigation application is analyzed in conjunction with the uncertain nature of precipitation events, then application of smaller amounts of irrigation water on a more frequent basis will reduce compound leaching more than the application of fewer, heavier irrigations (Singh and Sekhon, 1976). Nitrogen Nitrogen is applied to commercial potato production acreage to increase yield and quality of tubers. Nitrogen Use. Nitrogen fertilizer is applied to virtually all of Michigan‘s commercial potato production acreage. A standard application would consist of 200 lbs/acre split between planting and two side-dress applications. Nitrogen may be applied as animal manure, urea, potassium nitrate, sodium nitrate, or anhydrous ammonia. Nitrogen Value. Nitrogen fertilizers are used in potato production to increase the yield and quality of tubers. In research conducted on Superior potatoes the application of nitrogen fertilizer at 300 and 150 lbs/acre fl increased tu‘ over an appl 1980). Incr decreases th yield. Mm amount, timi Well as the irrigation e (McWilliams, applications mass of hit: with less if: Aldicarb Aldical insecticide primarily f< llflgay at a rate 0: the $011 prt Since 1975, °r°P rotati, been treate, 31 increased tuber yields by 39 and 27 cwt/acre respectively over an application of 75 lbs/acre nitrogen (Vitosh et al., 1980). Increased levels of nitrogen fertilization also decreases the percentage mass of B grade potatoes to total yield. Nitrate Concern. Nitrate movement is affected by amount, timing, and formulation of applied fertilizer, as well as the frequency and magnitude of precipitation or irrigation events, and the growth status of the potato plant (McWilliams, 1984). As with irrigation, frequent applications of small amounts of nitrogen should reduce the mass of nitrogen leaching out of the root zone when compared with less frequent and large fertilizer applications. Aldicarb Aldicarb is a water soluble systemic and contact insecticide and nematicide used in Michigan as Temik 15G, primarily for the control of P.penetrans and Leptinotarsa decemlineata (Colorado Potato Beetle). Aldicarb Use. Aldicarb is usually applied at planting at a rate of 3.0 lbs.a.i./acre and is distributed throughout the soil profile via water movement (Rhone Poulenc, 1988). Since 1975, approximately 25,000 acres have been treated annually with aldicarb in Michigan (Bird, 1987). Although crop rotation is a common control practice, some sites have been treated continuously for as many as eight years. fl Algiaar‘ a yield redu decemlineata reductions o with current impact of L.- considered a in potato p1 Aid—ism are similar only a singl growing sea: root zone m. plant emerg uptake of t Mileage The us irrigation Profitabili These potat threats to it al., 19: "at“ am: ground Watt (William; impacts cm “\- 32 Aldicarb Value. Typical P.penetrans infestations cause a yield reduction of approximately 16 percent. L; decemlineata at high population densities can cause yield reductions of up to 66% but losses of 5% are more common with current control practices (Noling et al., 1984). The impact of L.decemlineata on potato production was not considered as a part of this project. The value of aldicarb in potato production may be underestimated. Aldicarb Concern. Factors affecting aldicarb movement are similar to those affecting nitrogen movement. However, only a single application of aldicarb occurs during the growing season. The risk of aldicarb movement out of the root zone may be lessened with application of aldicarb at plant emergence (Jones et al., 1986) to increase plant uptake of the compound. Risk Management The use of nitrogen fertilizers, aldicarb and irrigation can have a significant impact on the profitability of potato production (Vitosh et al., 1980). These potato production inputs can also pose significant threats to ground water quality (Zaki et al., 1982; Bunyan et al., 1981; Back, at al., 1984). Method of irrigation water application may have an impact on the potential for ground water contamination by impact soil water relations (McWilliams, 1984). Frequency of nitrogen applications also impacts contamination risk to ground waters. Timing of aldicarb application may also be important in-the mitigation fl of risk to g The pur background i necessary fo potential pr while mainta practices is The impacts estimated tr. Two alt identified. literature 1 aldicarb. < represent 51 Potato mom was designe< application directly met Irriga °htained uni timing Was The ch Plunges f m... . 33 of risk to ground waters (Jones et al., 1986). The purpose of this literature review was to provide background information regarding potato production methods necessary for determination of standard grower practices and potential practices which may reduce risk to ground waters while maintaining tuber yields. Quantification of these practices is required for a comprehensive modeling system. The impacts of these alternate management systems are estimated through the use of simulation modeling. M飧£i§l§_§2§_M§££2Q§ Two alternate potato production strategies were identified. They based on information provided by the literature for the management of irrigation, nitrogen, and aldicarb. One management strategy was developed to represent standard grower practices in the Montcalm County potato production region. The second management strategy was designed to improve nitrogen, aldicarb, and irrigation application efficiency through timing of applications to directly meet plant needs. ‘ Irrigation applications were based on weather data obtained under Thesis Objective 1. Nitrogen application methods were subjectively determined. Aldicarb application timing was also studied. amiss The chapter objective was to determine management practices for comparison of associated risks and benefits. Results provided in this section show the treatments used fl for risk-bar treatments :‘ Irhrisaa1 decreased f1 grower prac‘ conservatim of irrigati: a lower tot [Jim (particular application standard gr practice wa excess mate The st followed by lbs./acre 1 treatment 1' 25 lbs. /ac1 days after (Table 5). t° the pot: nitroTen. a. L/icre 1 us" be de AA‘ w,‘ w-- 34 for risk-benefit analysis simulation. The impact of these treatments is provided in Chapter VI. Irrigation. Irrigation application amounts were decreased from one inch per application in the standard grower practice to one-half inch per application in the conservation practice. This results in a greater frequency of irrigation application in the conservation treatment but a lower total volume (Table 5). Nitrogen. Application of smaller amounts of nitrogen (particularly at planting) and making more frequent applications is what distinguished the conservation from the standard grower practice. The intent of the conservation practice was to directly meet plant needs without providing excess materials which would be available for leaching. The standard treatment was 75 lbs./acre at planting followed by 70 lbs./acre 50 days after planting, and 55 lbs./acre 70 days after planting. The conservation treatment involved application of 25 lbs./acre at planting, 25 lbs./acre 25 days after planting followed 50 lbs./acre 50 days after planting, and 25 lbs./acre 80 days after planting (Table 6). This treatment was intended to provide nitrogen to the potato plant just ahead of the growth demand for nitrogen. Aldicarb. The standard aldicarb application was 3 lbs. a.i./acre applied at-planting. Aldicarb application may also be delayed until plant emergence. The intent of delayed application was to make aldicarb unavailable for Table 5. Ir] application: 1986 Standard 9 06/13 06/17 06/23 06/30 07/05 07/10 07/19 07/23 07/29 08/03 08/08 08/13 08/18 1 ' Consep 35 Table 5. Irrigation dates and total number of applications for alternate management strategies. 1986 1987 1988 Standard Conser1 Standard Conser Standard Conser 06/13 06/10 06/12 06/08 06/10 06/10 06/17 06/14 06/16 06/13 06/15 06/13 06/23 06/20 06/20 06/16 06/20 06/16 06/30 06/23 06/24 06/19 06/25 06/19 07/05 06/29 06/29 06/22 06/30 06/22 07/10 07/02 07/03 06/25 07/05 06/25 07/19 07/05 07/07 O6/28 07/12 06/28 07/23 07/08 07/13 o7/01 07/22 07/01 o7/29 07/19 07/17 o7/o4 o7/27 07/04 08/03 07/22 07/22 07/07 08/01 07/07 08/08 07/27 07/26 07/13 08/06 .07/11 08/13 07/30 07/30 07/16 08/12 07/14 5 08/18 08/02 08/30 07/19 08/22 07/20 08/05 09/04 07/22 08/27 07/23 08/09 09/09 07/25 08/31 07/26 08/12 09/14 07/28 09/09 o7/29 08/15 08/29 09/16 08/01 09/01 08/04 09/04 08/07 09/07 08/11 o9/1o 08/14 09/13 08/25 08/28 08/31 09/07 09/10 09/16 No. 13 17 16 22 16 27 1 - Conservation managment strategy Table 6. Nitrogen management strategies for standard and conservation treatments. ‘ standard Conservation Days after Planting N applied Days after Planting N applied 0 75 0 25 . . 25 25 50 70 50 50 70 55 80 50 Total 200 150 leaching by both at-plar estimated . The in] was tested 1 procedures . 36 leaching by increasing plant uptake. Risk associated with both at—plant and at-emergence applications of aldicarb was estimated. aummarx The impact of the two alternate management strategies was tested using SUBSTOR simulation modeling. The procedures and formats used are provided in Chapter VII. A As part simulating a metabolites integrated 1 considered . volatilizat aldicarb an The li of informat and metabol is categorj V01atilizat $011 I or Soil 015 and degrade Organic 11131 (1980) 13.3. Table 7 . A; W $011 w, W CHAPTER IV ALDICARB MOVEMENT AND DEGRADATION MODEL As part of Thesis Objective 3, computer routines simulating aldicarb movement and degradation to oxidative metabolites in the soil environment were developed and integrated with SUBSTOR water movement routines. Factors considered in model development were: binding to soils, volatilization from the soil surface, systemic uptake, aldicarb and oxidation products degradation rates. Literature Review The literature review was developed to provide the base of information necessary for the development of an aldicarb and metabolite movement and degradation model. Information is categorized based on modeling concerns of soil binding, volatilization, systemic uptake, degradation. Soil Binding. Compound binding with soil organic matter or soil clays may retard movement with soil water. Aldicarb and degradation products only weakly partition into soil organic matter as demonstrated by Bromilow and Leistra, (1980) p.372 (Table 7). Table 7. Aldicarb and metabolite soil adsorption coefficients. Adsorption Coefficients (xloégc ..Jmskq'1) Soil OM% Aldicarb A—sfilfoxide A-sulfone Sandy Loam 1.35 64 O 8 §éndy,Loam 5.92 550 160 185 37 Smelt 4 to aldicarb lysimeters, presence of are availab Aldica bind with c clays aldic adsorbed on Volati translocate losses can surface (Ma 0f aldicark m SYStemic a1 I‘Etrieved r uPtake by 1 Mg With its 0: is then ox ProduCts. hydrolllSis are actin relativfly Deg ra 0“ et a1. 38 Smelt et al., (1983), summarized materials pertaining to aldicarb and metabolite binding to soil in soil columns, lysimeters, and arable fields. They concluded that in the presence of water flux aldicarb and its degradation products are available for movement between soil layers. Aldicarb and degradation products do not significantly bind with clay minerals in the soil. In montmorillonite clays aldicarb is excluded from the first layers of water adsorbed on external surfaces (Supak et al., 1978). Volatilization. Aldicarb and its metabolites are also translocated upward by capillary action. Significant mass losses can be expected through volatilization from the soil surface (Maitlen and Powell, 1982). In-furrow application of aldicarb reduces volatilization. Systemic Uptake. In the soil aldicarb exhibits both systemic and contact pesticidal activity. No articles were retrieved which dealt with aldicarb exclusion or active uptake by plant roots. Degradation Rate. The degradation of aldicarb begins with its oxidation to aldicarb-sulfoxide. Aldicarb sulfoxide is then oxidized to aldicarb sulfone and hydrolysis products. Aldicarb sulfone is then degraded to other hydrolysis products. Aldicarb and its oxidation products are active pesticides whereas the hydrolysis products are relatively non-toxic (Leistra et al., 1984). Degradation rates follow first—order conditions (Li-Tse Du et al., 1985) and are highly variable (Table 8). Table 8 . Deg Surrender—i Soil Texturc SAND LDANY SAND LOAN SANDY LOAN SANDY LOAN SANDY LOAN SANDY LOAN SANDY LOAN PEATY SANDY PEATY SANDY PEATY SANDY PEATY SANDY PEATY SANDY SAND SAND SAND SAND SAND SAND LOAMY FINE FINE SAND Rum 1- ALDICAI 2 - ALDICAI 3 - ALDICAI 4 - CITATI< 39 Table 8. Degradation constants for aldicarb, aldicarb- sulfoxide. and aldicarb-sulfone. Temp Deg. Const. k lgdays Soil Texture C° pH 0 __M Aldic A-sox A-son Cit.“ SAND 20 6.4 3.7 0.300' 0.010 0.230 a LOAMY SAND 20 6.9 3.8 0.460 0.010 0.230 a LOAM 20 7.1 9.7 0.240 0.007 0.100 a SANDY LOAM 5 7.0 1.4 0.300 0.015 0.012 b SANDY LOAM 10 7.0 1.4 0.440 0.033 0.020 b SANDY LOAM 15 7.0 1.4 0.210 0.034 0.013 b SANDY LOAM 15 7.0 1.4 0.800 0.035 0.021 b SANDY LOAM 15 7.0 1.4 0.800 0.025 0.016 b PEATY SANDY LOAM 5 6.3 5.9 0.200 0.011 0.005 b PEATY SANDY LOAM 10 6.3 5.9 0.270 0.030 0.010 b PEATY SANDY LOAM 15 6.3 5.9 0.140 0.013 0.005 b PEATY SANDY LOAM 15 6.3 5.9 0.460 0.031 0.012 b PEATY SANDY LOAM 15 6.3 5.9 0.550 0.031 0.015 b SAND 23 7.2 0.2 . . 0.020 C SAND 23 7.2 0.2 . . 0.017 c SAND 23 6.7 1.0 . . 0.011 c SAND 23 6.7 1.0. . . 0.013 C SAND 23 6.7 1.0 . . 0.016 c SAND 10 7.9 0.8 . 0.008 0.008 d LOAMY FINE SAND 10 8.0 1.2 . 0.004 0.006 d FINE SAND 10 5.0 0.4 . 0.002 0.001 d MEAN 0.419 0.019 0.037 1 - ALDICARB (KP) a = Leistra et al., 1984 2 - ALDICARB SULFOXIDE (KA) b = Bromilow et al.,1980 3 - ALDICARB SULFONE (KB) C = Li-Tse Du et al.,1985b 4 - CITATION CODE d = Smelt et al.,l983 Aldica by typical does not se between deg content or Table 9. M aldicarb-st physical ar Parameter Texture Sand Loamy Sand Sandy Loam Loam W The c ”Panic m can be eat] a great d« of aldica 40 Aldicarb degradation rate is not significantly affected by typical soil pH ranges (Chapman and Cole, 1982). There does not seem to be any clearly discernable relationship between degradation rates and soil type, pH, organic matter content or soil temperature (Table 9). Table 9. Mean degradation rate for aldicarb, aldicarb-sulfoxide and aldicarb-sulfone by soil physical and chemical Daramptprs_ Degradation Constant k lgdays _—Parameter Alchar—b mm waffle. Texture Sand 0.300 0.007 0.040 Loamy Sand 0.460 0.007 0.118 Sandy Loam 0.442 0.026 0.013 Loam 0.240 0.007 0.100 pH 5.0 e 5.9 . 0.002 0.001 6.0 - 6.9 0.320 0.021 0.035 7.0 - 7.9 0.460 0.022 0.025 8.0 - 8.9 . 0.004 0.006 % Om 0.0 - 2.9 0.510 0.019 0.013 3.0 - 5.9 0.340 0.019 0.072 6.0 - 9.0 0.240 0.007 0.100 Temp C° 5 0.250 0.013 0.008 10 0.350 0.015 0.009 15 0.490 0.028 0.013 20 0.330 0.009 0.186 23 . . 0.015 Implications of the Literature The degree of aldicarb and metabOlite binding to soil organic matter and clay minerals is small. Pesticide mass can be expected to be lost through volatilization. There is a great deal of variability associated with reported values of aldicarb and metabolite degradation rates. Variability (H in degradat soil textui The e: aldicarb b: small in 07 degradatio: Aldic developed approach w files were data trans contains a common blc Assun based on 1 literature interactic Was C0nsi< assumed t< Coefficie1 t0 actin infOrmati. Volatiliz aldicarb aldiCarb with no a 41 in degradation rates reported is not easily explained by soil texture, pH, organic matter, or temperature. The expected impact on movement and degradation of aldicarb binding to soil organic matter or clay minerals is small in comparison to the uncertainty associated with degradation rates. Materials and Methods Aldicarb degradation and movement routines were developed using Microsoft FORTRAN v.4.0. A structured approach was used to maintain program readability. INCLUDE files were used in place of subroutine common blocks for data transfer between subroutines. Each INCLUDE file contains a data dictionary, variable initialization, and common blocks. Assumptions. Several operational assumptions were made based on the information provided in the scientific literature. The attenuation of aldicarb movement due to its interaction with the soil organic matter or clay materials was considered to be negligible. Pesticide movement was assumed to be a function of soil water movement. A mixing coefficient was used to represent differential mass flow due to active and non-active soil pores. Quantitative information on the loss of aldicarb and metabolites through volatilization was not available. The volatilization of aldicarb mass was assumed to be zero. Plant uptake of aldicarb is assumed to be proportional to root water uptake With no active uptake and no exclusion. Aldicarb degradation aldicarb, a variables c assumed nor rates are r matter cont SUBSTC available A subroutine: were develv The m. CERES corn routine's informatio research 1 mm Mm 42 degradation follows first-order kinetics. The mass of aldicarb, aldicarb-sulfoxide, and aldicarb-sulfone are the variables of concern with all other degradation products assumed non-toxic. Aldicarb and metabolite degradation rates are not affected by soil organic matter, pH, organic matter content, or temperature. Computer Code Development SUBSTOR operates on a daily time step and uses readily available weather, soil, and potato variety inputs. Nine subroutines simulating aldicarb movement and degradation were developed and linked with SUBSTOR. Existing Routines The majority of SUBSTOR routines were adapted from the CERES corn model. A brief statement regarding each routine's function is provided (Table 10). Additional information on routines not developed as part of this research is available in CERES - Maize: A simulation model of maize growth and development (Jones and Kiniry, 1986) or SUBSTOR Model Documentation (Swartz, 1987). 43 MAIN -IPEXP -IPPEN# -IPALD* -IPTRT -IPSOIL -IPVAR -IPNIT -IDWTH -OPECO -IPWTH -PROGRI -OPSEAS _ -SOILRI -SOILNI ~SOILT -CALDAT -PTRANS* -NTRANS -WATBAL -NFLUX -PFLUX* -PFLOW* -NFLUX -SYSTEMIC* -PHENOL -CALDAT -PHASEI -GROSUB ~NUPTAKE -NWRITE -XWRITE -PSTDAY* -TTOUT* -SOILPST* -OUTPLCH* :-PPIMPACT# -OUTYLD# -SUMOUT# * sub-routine developed under Thesis Objective 3 # sub-routine developed under Thesis Objective 4 Figure 8. Simplified SUBSTOR Flow Diagram. New Subroutine IPEXP IPPEN IPALD IPTRT IPSOIL IPVAR IPNIT IDWTH OPECO IPWTH PROGRI OPSEAS SOILNI SOILT CALDAT PTRANS NTRANS WATBAL NFLUX PFLUX PFLow SYSTEMIC PluNOL PHASEI GROSUB “UPTAKE “WRITE XWRITN PSTDAy TTOUT S UMOU\T 44 Table 10. SUBSTOR program routines and primapy functions. Sumo—utine IPEXP Primary Functions Initialization of experiment to be simulated IPPEN Initialize aldicarb/P.penetrans yield impact routines IPALD Initialize aldicarb movement and degradation routines IPTRT Called if run time option to modify experiment variables is selected IPSOIL Modify soils IPVAR Modify potato variety IPNIT Modify fertilizer applications IDWTH Modify weather data used OPECO Writes new experimental parameters to screen IPWTH Initialize weather data PROGRI Starts simulation loop OPSEAS Generates output headings and initialize counters SOILNI Determine nitrogen contribution of stem and roots SOILT Calculates soil temperature CALDAT Converts day of the year to calendar date PTRANS 'Applies aldicarb to appropriate soil layer on application date. Calculates aldicarb and metabolite degradation NTRANS Distributes fertilizer on appropriate days. Calculates nitrification and denitrification WATBAL Determines runoff and infiltration of rainfall Determines movement of water with saturated flow Determines water movement with unsaturated flux Determines evapotranspiration Determines root growth, depth, and water uptake NFLUX Move nitrogen with soil water PFLUX Move aldicarb and degradation products with unsaturated flux as determined by WATBAL PFLOW Move aldicarb and degradation products with saturated flow as determined by WATBAL SYSTEMIC Determine plant uptake of aldicarb and degradation products as determined by WATBAL PHENOL Calculates thermal time PHASEI Determines plant growth stages GROSUB Partitions Photosynthates NUPTAKE Determines nitrogen available and nitrogen desired NWRITE Determines if nitrogen output files are to be written XWRITE Calculates cumulative environmental parameters PSTDAY Calls nitrate and aldicarb daily output routines Resets aldicarb mass matrix for next days degradation . TTOUT Writes output files for aldicarb total tox1c metabolites _ SUMOUT Writes summarv output file for yield and Teaching Newly Devei The f: Objective 2 they are c: Include Fi. ALDIC and metabo Variabl PSTMASS (P, 45 Newly Developed Routines The following routines were developed under Thesis Objective 3. Routines are presented in the order in which they are called by the SUBSTOR program. Include File ALDIC.INC contains variables used in aldicarb movement and metabolism routines. Variable PSTMASS(P,T,L) L JDATE ALDRATE KP,KA,KB CTP,CTA,CTB APDEPTH PSTCOST APDATE TRTVAL CUMLEACH(3) CUMPUP(3) TLEACH TPUP PSTDOWN(3,10) PSTUP(3,10) APLAYR PSTLCH(4) PLANTUP(4,10) DATA DICTIONARY Description IS A THREE DIMENSIONAL ARRAY HOLDING INFORMATION ON PESTICIDE MASS BY SOIL LAYER. RANGES FROM 1 TO 4 STANDING FOR ALDICARB, ALDICARB SULFOXIDE, ALDICARB SULFONE, AND PESTICIDE DEGRADED TO NON-TOXIC METABOLITES MASS RESPECTIVELY (kg/ha). . RANGES FROM 0 TO 1 WITH 0 STANDING FOR PRESENT DAY, AND 1 STANDING FOR PREVIOUS DAY. RANGES FROM 1 TO NLAYR AND REPRESENTS INDIVIDUAL SOIL LAYERS. DAY OF THE YEAR RATE OF ALDICARB APPLICATION:(kg/ha) ACTIVE INGREDIENT DEGRADATION CONSTANTS OF ALDICARB, A-SULFOXIDE, A-SULFONE COEFFICIENT OF TRANSFORMATION FOR OXIDATIVE DEGRADATION 1=COMPLETE 0=NONE DEPTH OF ALDICARB APPLICATION IN CENTIMETERS COST OF ALDICARB APPLICATION $/AC DATE OF ALDICARB APPLICATION PESTICIDE VALUES BASED ON AT PLANTING APPLICATION CUMULATIVE LEACHING OF PESTICIDE CUMULATIVE PLANT UPTAKE OF PESTICIDE TOTAL MASS LEACHED FROM BOTTOM SOIL LAYER TOTAL MASS TAKEN UP BY THE PLANT MASS OF PESTICIDE IN GRAMS MOVED TO LOWER SOIL LAYER MASS OF PESTICIDE IN GRAMS WICKED TO UPPER SOIL LAYER DEPTH INDICATOR USED FOR PLACEMENT OF ALDICARB IN PROPER SOIL LAYER DAILY LEACHING OF PESTICIDE OUT OF PROFILE DAILY PLANT UPTAKE OF PESTICIDE FROM EACH SOIL LAYER. 31 32 33 34 39 ALDFILE SUMOUT ALDFLAG TOXOUT SPSTOUT LCHOUT REAL ALDR +CPSTLCH(4 +psmp (4 , 1 +CUMLEACH( INTEGER o CHARACTER +LCHOUT*11 COMMON / A +PSTCOST , C +PSTMASS , F +0UT3 0 , OU'I +SUMOUT ’ AI IPPST IPPE Aldicarb l format. 1 inClude f: ALDRATE, 1 CTB. ALDFILE SUMOUT ALDFLAG TOXOUT SPSTOUT LCHOUT 46 UNIT NUMBER FOR TOTAL TOXIC OUTPUT FILE UNIT NUMBER FOR SOIL PESTICIDE OUTPUT FILE UNIT NUMBER FOR LEACHING OUTPUT FILE UNIT NUMBER FOR ALDICARB PARAMETER FILE UNIT NUMBER FOR SUMMARY OUTPUT FILE NAME OF ALDICARB INPUT PARAMETER FILE OUTPUT FILE NAME FOR SUMMARY DATA FLAG INDICATING IF ALDICARB DEGRADATION ROUTINES ARE TO BE USED 1-= YES OUTPUT FILE NAME FOR TOTAL TOXIC MASS OUTPUT FILE NAME FOR SOIL PESTICIDE RESIDUE OUTPUT FILE NAME FOR LEACHATE SUMMARY REAL ALDRATE,KP,KA,KB,CTP,CTA,CTB,APDEPTH,PSTCOST, +CPSTLCH(4),CUMPUP(4),PSTDOWN(4,10), +PSTUP(4,10), APLAYR, PSTMASS(4,2,10), PSTLCH(4), +CUMLEACH(4), PLANTUP(4,10), TLEACH,TPUP INTEGER OUT31,0UT32,0UT33,0UT37,0UT39,INAL34,APDATE y CHARACTER SUMOUT*11,ALDFLAG*1,TOXOUT*11,SPSTOUT*11, +LCHOUT*11 COMMON /ALDIC/ALDRATE,KP,KA,KB,CTP,CTA,CTB,APDEPTH, +PSTCOST,CPSTLCH,CUMPUP,PSTDOWN,PSTUP,APLAYR, +PSTMASS,PSTLCH,PLANTUP,TLEACH,TPUP, +OUT30,0UT31,0UT32,0UT33,0UT37,0UT39,INAL34,APDATE, +SUMOUT,ALDFLAG,TOXOUT,SPSTOUT,LCHOUT IPPST IPPST is called from the MAIN program and reads the aldicarb parameter file 34. File 34 "ALDFILE.PAR" is free format. Parameter variable units are provided in the include file, ALDIC.INC. Parameter variables include ALDRATE, APDEPTH, PSTCOST, APDATE, KP, KA, KB, CTP, CTA, CTB. OPEN(34,FILE='ALDFILE.PAR',STATUS='OLD') READ(34,*)ALDRATE,APDEPTH,PSTCOST,APDATE, KP,KA,KB,CTA,CTB Next the summary output file is named and opened. WRITE(*,320) 320 FORMAT(5X,'ENTER NAME OF SUMMARY OUTPUT FILE') READ(*,'(A)')SUMOUT OPEN(OUT39,FILE=SUMOUT,STATUS='NEW') Program execution returns to MAIN. PTRANS SubrOI degradatim date is pr: then the s equals APD. soil layer SUBSTOR. applicatio degradatio IF ELS ELE EN! The a aPPlicatic cm of sci: amount ex: 100 110 A101 dimension from 1 to aldica‘rb The index 47 PTRANS Subroutine PTRANS is used to simulate application and degradation of aldicarb and metabolites. If the simulation date is previous to the pesticide application date APDATE, then the subroutine returns to MAIN. If the simulation date equals APDATE, then aldicarb is applied to the appropriate soil layer as defined by APDEPTH and soil layer depths from SUBSTOR. If the simulation date is after the aldicarb application date then the program executes aldicarb degradation routines. IF (JDATE.LT.APDATE) THEN RE ELSEIF (JDATE .EQ.APDATE) THEN GOTO 100 ELSE GOTO 200 ENDIF The amount of water required to dissolve a standard application of aldicarb is 227.27 Kg corresponding to 0.0056 cm of soil water. The assumption was made that a greater amount exists in the application layer. 100 DO 110 L=1,NLAYR APLAYR = APLAYR + DLAYR(L) IF (APDEPTH .GT. APLAYR) GOTO 110 PSTMAss (1,1,L) = ALDRATE RETURN 110 CONTINUE Aldicarb and metabolite masses are held in a three dimensional array called PSTMASS(P,T,L). The index P ranges from 1 to 4 representing aldicarb mass aldicarb sulfoxide, aldicarb sulfone and mass degraded to non-toxic metabolites. The index T ranges from 1 to 2 with 1 representing today's mass and 2 representing yesterday's mass. Values of L range 7’! from 1 to are in ki] tracked us kinetics. PSTMl PSTMJ PSTMJ PSTM. Thee sOil beir aldicarb is not at the next firstwr, (aldicarl the Seco] COefficil MOSS of ‘ degraded rate con 48 from 1 to 10 representing up to 10 soil layers. Array units are in kilograms per hectare. Daily mass changes are tracked using the following algorithm based on first-order kinetics. PSTMASS(1,1,L) = PSTMASS(1,2,L)*EXP(-KP) PSTMASS(2,1, L) = PSTMASS(l, 2 ,L)*(1- EXP(- KP))2 +PSTMASS(2, 2 ,L)*EXP(-KA)3 PSTMASS(3,1, L) = CTA*PSTMASS(2, 2 ,L)*(1. O-EXP(— KA)) +PSTMASS(3, 2 ,L)*EXP(--KB)5 PSTMASS(4,1,L) = PSTMASS(4, 2 ,L)6 +(1. O-CTA)*PSTMASS(2, 2, L)*(1. 0-EXP(- KA))7 +PSTMASS(3, 2 ,L)*(1. 0-EXP(- -KB))8 First-order degradation of aldicarb mass Add mass of aldicarb degraded to A-sulfoxide mass First-order degradation of A-sulfoxide mass Add mass of A-sulfoxide degraded by oxidation to A-sulfone First-order degradation of A-sulfone mass Yesterdays' cumulative mass degraded to non-toxic products Add mass of A- -sulfoxide degraded by hydrolysis to non- toxic products Add mass of A-sulfone. degraded by hydrolysis to non-toxic products men p19u1H \) 03 These calculations are performed for each layer in the soil being simulated. The one-day time lag is used so that aldicarb maSs degraded to aldicarb sulfoxide on a given day is not available for metabolism to aldicarb sulfone until the next day etc.. The values of KP, KA, and KB are the first-order degradation coefficients for the parent compound (aldicarb), the first metabolite (aldicarb sulfoxide) and the second metabolite (aldicarb sulfone) respectively. A coefficient of transformation (CTA) is used to separate the mass of aldicarb sulfoxide degraded by oxidation from mass degraded by hydrolysis. The mean values for degradation rate constants reported in Table 9 were used. PFLUX PFLUX movement c with satur millimeter water f 101 Pesticide proportior the total constant c movement < is due to EL 49 PFLUX PFLUX is called from WATBAL and is used to simulate the movement of aldicarb and metabolites between soil layers with saturated flux. The value of DRAIN is converted from millimeters to centimeters. DRAIN indicates the volume of water flowing out of the lowest layer of the soil profile. Pesticide mass moved to a lower layer is assumed to be proportional to the water flow out of that layer divided by the total water content of that layer. A proportionality constant of 0.65 was used to represent differential mass movement due to in-layer water mixing. This in-layer mixing is due to soil pore size variability. ' DO 40 P=1,4 (for each pesticide mass) DO 30 L=1, NLAYR (for each layer) IF (L. LT. NLAYR) THEN PSTDOWN(P, L)-=0. 65*PSTMASS(P, 1, L)2*FLUX(L)3 / (SW(L)*DLAYR(L)+FLUX(L)) PSTMASS(P, 1, L) =PSTMASS(P, 1 ,L)-PSTDOWN(P, L)5 PSTMASS(P, 1, L+1) =PSTMASS(P, 1, L+1) +PSTDOWN(P, L)6 ELSE (bottom layer) PSTDOWN(P, L)= O.65*PSTMASS(P,1,L)*DRAIN7 /(SW(L)*DLAYR(L)+DRAIN) PSTMASS(P, L L)— =PSTMASS(P, L L) -PSTDOWN(P, L) CPSTLCH(P) 8=CPSTLCH(P)+PSTDOWN(P, L) ENDIF 30 CONTINUE 40 CONTINUE Pesticide mass moving out of soil layer L Pesticide mass in soil layer L before movement Soil water moving out of soil layer L Total water previously in soil layer Subtract mass moved out of layer L from layer L Add mass moved out of layer L to layer below Soil water leaching out of profile Update cumulative pesticide leaching After execution of this routine the program returns to ”\lmtflhUNH WATBAL where movement with unsaturated flow is determined. 3U PFLOW water from water betv The value between 12 actiOn frc If FLOW it to the lot DO 61 DO 51 IF (I ELSE 50 PFLOW PFLOW is called from WATBAL after the evaporation of water from the surface soil layer and redistribution of water between unsaturated soil layers has been determined. The value of FLOW represents the direction of water movement between layers. If flow is positive then flow by capillary action from a lower level to the next higher levels occurs. If FLOW is negative then water moves from the higher level to the lower level. DO 60 P=1,4 (for each pesticide mass) DO 50 L=1, K (for soil layers 1 - (nlayr-1)) IF (FLOW(L). GT. 0. 0) THEN (upward movement) PSTUP(P, L)— :0. 65*PSTMASS(P, 1, L+1) 1*FLQW(L)/ (SW(L+1)*DLAYR(L+1)+FLOW(L)) PSTMASS(P, 1 ,L) =PSTMASS(P, 1, L)+PSTUP(P, L)3 PSTMASS(P, L L+1) =PSTMASS(P, 1, L+1) -PSTUP(P, L) ELSE (downward movement) PSTDOWN(P, L)=-o. 65*PSTMASS(P, 1, L) 5*(FLOW(L)/ (SW(L)*DLAYR(L)+FLOW(L)6) PSTMASS(P, 1, L)— =PSTMASS(P, l, L)-PSTDOWN(P, L)7 PSTMASS(P, 1, L+1) =PSTMASS(P, 1, L+1)+ PSTDOWN(P, L)8 ENDIF 50 CONTINUE 60 CONTINUE Pesticide mass in lower layer (movement up) Proportion of water movement out of layer to higher layer modified by 0.65 assumed mixing factor 3 Add pesticide mass moved from lower layer to higher layer Subtract pesticide mass moved from lower layer from the mass in the lower layer Pesticide mass in layer (movement down) Proportion of water moved out of layer to lower layer Subtract pesticide mass from upper layer Add pesticide mass to lower layer e NIH ooqoxm The value for FLOW in layer one is always 0.0. This subroutine returns to WATBAL where plant uptake of soil water is determined. SYSTEMIC Subrc uptake of mass is as roots . 51 §X§I§Ml§ Subroutine SYSTEMIC is called from WATBAL after root uptake of water has been estimated. Movement of pesticide mass is assumed to be proportional to water taken in by the roots. DO 80 P=1, 4 (for each pesticide mass) DO 70 L=1, NLAYR (for each soil layer) PLANTUP(P, L)'=PSTMASS(P, 1, L)*(RWU(L) /DLAYR(L)) 2/(SW(L)*DLAYR(L))3 CUMPUP(P) =CUMPUP(P)+PLANTUP(P, L) PSTMASS(L L L) =PSTMASS(P, 1, L)-PLANTUP(L L)5 70 CONTINUE 80 CONTINUE Plant pesticide uptake frOm layer Root water uptake from layer Total soil water in layer Update cumulative plant uptake Subtract pesticide mass taken up by roots from soil layer ' m-hwwb-I Program execution returns to WATBAL. PSTDAY PSTDAY is called by the MAIN program at the end of the simulation day. PSTDAY calls daily pesticide output files (TTOUT, SOILPST, and OUTPLCH) prior to updating the pesticide mass matrix. Today's mass value T = 1 is shifted to the T = 2 position. DO 100, T=2,1,-1 (for time index 2 to_1) DO 75, P=1,4 (for each pesticide) DO 50, L=1,NLAYR (for each soil layer) C=T IF(C. GT. l)THEN PSTMASS(P, T, L) =PSTMASS(P, T- 1 ,L)1 ELSE PSTMASS(P, T ,L)=-99. 92 ENDIF 50 CONTINUE 75 CONTINUE 100 CONTINUE 1 Set the matrix value at T=2 equal to the matrix value at T=1 2 Set fac Execution 1m TTOU'] informatic taken. up 1 layer of 1 mass is ca and A-sul 52 2 Set the matrix value at T=1 equal to -99.9 to facilitate error checking Execution returns to the MAIN program. TTOUT TTOUT is called from PSTDAY and is used to write information on total toxic residues remaining in the soil, taken up by the plant plant, and leached out of the lowest layer of the soil profile to output file 31. Total toxic mass is calculated as the mass sum of aldicarb, A—sulfoxide, and A-sulfone. It is calculated using: DO 20 P=1,3 (for aldicarb, A—sulfoxide, A-sulfone) DO 10 L=1,NLAYR (for each soil layer) TTSOIL(L) =TTSOIL(L) +PSTMASS (P, 1 , L) 10 CONTINUE 20 CONTINUE Total pesticide mass degraded in the soil is determined: DO 25 L=1,NLAYR (for each soil layer) DEGSOIL=DEGSOIL+PSTMASS (4 , 1 , L) 25 CONTINUE Total toxic mass taken up by the plant, and leached is calculated using: DO 30 P=1, 3 (Sum mass for total toxic residue) TTPUP=TTPUP+CUMPUP(P) TTLCH=TTLCH+CPSTLCH(P) 30 CONTINUE Calculate total degraded mass in plant, leached, and remaining in soil: ’ TDEGMASS=CUMPUP(4)+CPSTLCH(4)+DEGSOIL Check for mass balance: DO 40 L=1, NLAYR CHECK=CHECK+TTSOIL(L) 40 CONTINUE 1 Sum total toxic mass in each soil layer CHECI Add mass :' If 5: total tox: written. date then precipita‘ toxic in ‘ uptake, d routine r SOILPST SOII. Pesticide simulatic is Opened If the sj then the with the °Peratiol 0UTPLCH OUT] Pesticid, equals a] heider i: Sim“lati. date, cu are Writ 5 3 CHECK=CHECK+TTPUP+TTLCH+TDEGMAS S Add mass in other pools. CHECK equals the application rate. If simulation date equals application date then the total toxic output file is opened and header information is written. If simulation date is greater than application date then date, cumulative (rainfall, irrigation, precipitation, total toxic leached, nitrate leached), total toxic in up to five soil layers, and cumulative (plant uptake, degraded mass) are written to the output file. The routine returns to PSTDAY. p k SOILPST SOILPST is called from PSTDAY and writes daily soil pesticide mass information for aldicarb and metabolites. If simulation date equals application date then output file 32 is opened and header information is written to that file. If the simulation date is greater than the application date then the pesticide mass in each soil layer is written along with the soil water content in that layer. Program operation returns to PSTDAY. OUTPLCH OUTPLCH is called from PSTDAY and writes cumulative pesticide parameters to output file 33. If simulation date equals application date then output file 33 is opened and header information is written to that file. If the simulation date is greater than the application date then date, cumulative precipitation and water drainage variables are written to file 33 as well as cumulative nitrate, aldicarb, non-toxic. New 1 existing : existing : routines, routine. code was The expansion aldicarb allowed f aldicarb and metal: 54 aldicarb, A-sulfoxide, A-sulfone, and total mass degraded to non-toxic. Return to PSTDAY. Summary New FORTRAN routines were developed and linked with existing SUBSTOR routines. If variables obtained from pre- existing SUBSTOR routines were modified within these routines, they were reinitialized before exiting the routine. This insured that the execution of the original code was unchanged. The modifications described here resulted in a expansion of SUBSTOR's capabilities to include estimation of aldicarb and metabolite movement and degradation. This allowed for the estimation of how irrigation scheduling and aldicarb application timing affect the movement of aldicarb and metabolites through the soil profile. ALDJ The (Root-1e: yield was research Dur. (MSU) Ne: Montcalm been pub reports contribu we: Use constraj research restrict a limits Pressure This ma) results. Procedu] results. reviews CHAPTER V ALDICARB / ROOT-LESION NEMATODE YIELD IMPACT MODEL The impact of aldicarb and Pratylenchus penetrans (Root-lesion nematode) on Solanum tuberosum (potato) tuber yield was studied from the perspective of an integrative research review and meta—analysis (Thesis Objective 4). During the past 15 years, the Michigan State University (MSU) Nematology Program has conducted research at the MSU Montcalm Potato Research Station. The research results have been published in graduate student theses, MSU research reports and professional journals. The current research contribution uses previously and un—published published research findings as a.base for extended analysis. Use of previous research findings is frequently- constrained by the isolation of each study to its particular research objectives. Most agricultural research results are restricted in that they provide information only for one or a limited number of crop growing seasons with specific pest pressures and distributions of temperature and rainfall. This may result in limited generalizability of research results. Integrated research review and meta—analysis procedures, however, can be used to generalize research results. Integrated Research Review A distinct difference exists between classic research reviews and integrated research reviews. In a classic 55 research : from avai review, t' quantitat literatur have occu time of G from writ inference statistic methodolc to those The research Collecti¢ Presentai The the proj f°r revi therefor M The for info Specific research An integ 56 research review, the reviewer makes cognitive inferences from available literature. In an integrative research review, the researcher uses the rigor and power of I quantitative methodologies to describe the available literature. This alternate approach resembles changes which have occurred in primary information collection since the time of Galileo (Drake, 1981). Researchers have progressed from writing about observable phenomenon (cognitive inference) to using replicated experimental units for statistical testing of hypothesis (quantitative methodology). Integrative reviews use similar methodologies to those of today's primary researchers. The following are the five stages of an integrative research review: 1) objective definition, 2) data collection, 3) data evaluation, 4) meta—analysis, and 5) presentation of results (Cooper, 1984).. Objective Definition The objective definition stage determines the scope of the project by defining research boundaries. Methodologies for reviewing and analyzing data are objective dependant; therefore, it is imperative for objectives to be precise. Data Collection The data collection stage describes the methods used for information retrieval from the scientific literature or specific data bases. It also serves as an indicator of research bias and describes where data how were obtained. An integrated research review summarizing information from one speci an integr of scient Data Eval Each review ma research bias.' Tl collected results i to handle m Pre: an integ: research it summa future r The “59d for framewo: integrat Mei require: analYSi: 57 one specific scientific journal has a value different from an integrated research review which summarizes many sources of scientific information. Data Evaluation Each specific study used in an integrated research review may not contain all the information required to meet research objectives and is a potential source of research bias.' The data evaluation section identifies biases in collected variables and potential problems with using study results in meta-analysis It also describes procedures used j to handle missing values. Presentation Presentation of research results is the final stage of an integrated research review. The value of an integrated research review can be measured in the amount of past work it summarizes and the degree to which the study clarifies future research needs (Cooper, 1984). The five stages of an integrated research review were used for the analysis of thirty-four studies, and provided a framework for synthesis of results. Data collected in the integrated research review provide a measure of variability in available published research results. Meta-analysis Meta-analysis is defined as research on research. This requires an integrative research review and subjects research results to further quantitative analysis. Meta- analysis uses information from the integrative research review to because i global pi The form, alt transcend means a c data col] higher dc position, referenc: position Met; research research Studies analysis Which go term met Proceduz that put Fix 0f the 1 literatl °f mOde 58 review to meet specific research objectives. Meta-analysis, because it integrates findings across studies, can provide a global picture of research results. i The prefix "meta" is defined as a change in position or form, altered, transposed; or going beyond, higher, transcending (Webster, 1979). The term "meta—analysis" means a change in research position to a level above primary data collection; or research on research. Going beyond or higher does not indicate better, it indicates a change of position, a stepping out of a discipline's plane of reference in order to objectively analyze the goals, current position and objectives of the subject. Meta-analysis may or may not be part of an integrated research review. If the objectives of the integrated research review can be met with in the scope of the original studies then this stage is more appropriately termed analysis. If the integrated research review has objectives which go beyond the original research objectives, then the term meta-analysis can be used to describe the analysis procedure. The term meta-analysis is also used to indicate that published study results are being analyzed, not the phenomenon for which the original studies were designed. Objective Definition Five, meta-analysis objectives were developed because of the need to define the current state of the scientific literature, and where possible, develop a hierarchial series of models which could be used to simulate the impact of aldicarb ; this inte« covers re aldicarb objective categoriz Objective Objective Objectiv 0bjectiv Oblectiv aldicarb and P.penetrans on potato production. 59 The scope of this integrative research review (IRR) and meta-analysis covers research pertaining to potato production with aldicarb used to control P.penetrans. The following five objectives were used to describe the information base and categorize 14 meta-analysis methods. objective Objective Objective objective Objective I) Describe the variability in research results showing the impact of aldicarb application and P.penetrans on potato production Analysis 1) Descriptive Statistics Analysis 2) ANOVA II) Determine the impact of aldicarb on potato yield Analysis Analysis 4 Analysis 5 Analysis 6 Average Yield loss 1 Cumulative Probability Distribution Preseason Model Postseason Model (.0 VVVV III) Determine the impact of aldicarb on P.penetrans population dynamics Analysis 7) Regression Model Analysis 8) Distributed Delay Model IV) Determine the impact P.penetrans population dynamics on potato variety tuber yield Analysis 9) Class Correlation Analysis 10) Class Regression Model V) Determine the impact of aldicarb on potato plant development Analysis 11) Correlation Analysis 12) Regression Delta Plant Growth Analysis 13) Regression Percentage Plant Growth Analysis 14) Regression Plant Partitioning 1 . ‘ Aldicarb application is the current normal practice the Michigan potato production. as potential tuber yield loss associated with not applying aldicarb In this thesis,"Yield Loss" is defined The 2 five leve distribut yield 105 parameter and the 6 evaluated also inc1 The parts. ' integrat procedur analyses Provided Se\ reView' Variable availab: % 5 Th. RESeuro? informa informa PUblish 6O . Model Hierarchy The hierarchy of models used in the research contains five levels: average yield loss, cumulative probability distribution, regression prediction, population linked with yield loss, and population linked with plant development parameters. Each level represents an increase in complexity and the degree to which the dynamics of the system are evaluated. The quality of data needed to support each model also increases. Materials and Methods The Materials and Methods section is divided into two parts. The first describes procedures used in the integrated research review. The second describes general procedures used in the meta-analysis. Methods for specific analyses are organized by meta—analysis objective and provided with analysis results. Integrated Research Review Seven procedures were used for the integrated research review, literature search procedure, selection criteria, variable description, research bias, data bias, data availability, missing value, and estimator test.Data Literature Search Procedure The Michigan State University (MSU) Montcalm Potato Research Farm Annual Report provided the majority of the information presented in this study. To augment this information source, a computer—aided search of information published in scientific journals was conducted. CAIN, CAB, BIOZ, and facilities conditions nematode, abstracts aldicarb ‘ potato pr library w and copie base (PPD Literatur The that a pa Conjuncti measureme POpulati< inclusiol these or; TwentY-tl eXclucled PM The inclu'sio as a Sp: f°110win agmnomi 61 BIOZ, and CABA data bases where searched using off-line facilities at the MSU Library. The key word search conditions were: (Pratylenchus penetrans, or root-lesion nematode, and Solanum tuberosum, or Potato).l Citations and abstracts were retrieved and searched for studies in which aldicarb was used as a control measure for P.penetrans in potato production. Theses and disSertations at the MSU library were also searched. These papers were then located and copied for potential inclusion in the P.penetrans data base (PPDB) developed as a part of the research review. Literature Selection Criteria . The criteria for inclusion of a paper in the PPDB, were that a paper had to be a field study, use aldicarb in conjunction with a non-treated control, report tuber yield measurements, and have information on P.penetrans populations. A minimum of three studies were needed for inclusion of a specific potato variety. Papers not meeting these criteria were not included in the integrated review. Twenty-three out of fifty-seven papers retrieved were excluded from the analysis (Table 11). Variable Description The information contained in studies which met the inclusion criteria was coded into the PPDB using LOTUS 123 as a spreadsheet. The spreadsheet was divided into the following four: sections pre-plant measures, yield measures, agronomic measures, and growing season measures. REFER Bernard Biehn et Bird,198 Brown et Burpee a Dickersc Dunn, ls Francel Hawkins Hawkins Kable a1 Kimpins] Kimpins] Kotcon 1 Kotcon a Martin l 01thof, 01thof, 01thof, Oostenb' Patters Riedel Rowe et Mm Table 11. Research not included in the literature review. REFERENCE Bernard and laughlin,1976 Biehn et al., 1971 Bird,1986 Brown et al., 1980 Burpee and Bloom,1978 Dickerson et al., 1964 Dunn, 1972 Francel et al.,1987 Hawkins and Miller,1971a Hawkins and Miller,197lb Kable and Mai, 1968 Kimpinski,1979 Kimpinski,1982 Kotcon et al.,1985 Kotcon and Loria, 1986 Martin et al.,1982 Olthof,1983 Olthof,1985 Olthof,1986 Oostenbrink,1958 Patterson and Bergeson, 1967 Riedel et al., 1985 Rowe et al., 1985 Wong and Ferris.l968 Reason for Non-Inclusion MICRO GREEN NO TILE HOUSE ALE OTHER NO NEMA DATA . . x . . . . NO CHECK x . . . . x . , . . SURVEY . . x. . . x . . . . SINGLE STUDY . . x ' . '. x . . . . . NO YIELD . . . AVERAGED x . . . . x . x . . . x . . . x . . . x . . . . x . . x . . x . . . x . . . Pie included conducte (base 10 nematode treated gig variable category treatmel Burbanki (Append. Ac; informa cultiva B,Table Q; include the deg SamPlir. nematoc' nematoc' Presen( cultiv: A SiZe J Size KliObby '63 Pre;plapp_flea§ppp§. The pre-plant measures section included four variables: 1) the year the study was conducted, 2) date of pre-plant sampling, 3) degree days (base 10) accumulated by planting date, and 4) initial nematode population density /100 on3 soil for aldicarb treated and non-treated plots (Appendix B, Table 1). Yield Measures. The yield section included eleven variables: 1) harvest date, and 2-9) tuber yield by size category (B, A, J, total yield) for both check and aldicarb treatment, 10-11) knobby yield is reported for Russet Burbankz. Yields are reported in hundred-weight per acre (Appendix B,Table 2). Agronomic Measures. The agronomic section included information on N,P,K fertilizer use, rotation crops, potato cultivar, study location, and a code for citations (Appendix B,Table 3). Growing Season Measures. The growing season section included five variables: 1) date of nematode samples, 2) the degree days (base 10) accumulated at the day of sampling, 3) the number of nematodes in 100 cm3 soil, 4) the nematode population in'1.Q gram of root tissue, and 5) total nematode population in soil plus roots. For ease of presentation growing season measures are arranged by cultivar, (Appendix B, Tables 4,5,6 for Superior, Russet 2 B size class tubers, less then 5 cm in diameter A size class tubers, 5-8 cm in diameter J size class tubers, greater then 8 cm in diameter Knobby class tubers, are mis-shaped russet burbank Burbank, midlife Dat research and esti Researcl All Montcall created across : conduct providi researc contrib minimiz Data Bi It be bias aldica: its use PrOduc1 Substal G1 the dai 64 Burbank, and Atlantic respectively). Deta_Eysluatien Data evaluation is divided into five sections: research bias, data bias, data availability, missing values, and estimator test. Research Bias All of the studies selected were conducted at the MSU Montcalm Potato Research Farm on a McBride sandy loam. This created a spatial bias and limitation for generalization across soil types and climate. The studies, however, were conducted over a ten-year period from 1977 to 1987; providing for ample weather variability. Many different researchers, working both independently and in teams contributed the selected studies. This variation should minimize researcher bias. Data Bias Information on aldicarb impact in potato production may be bias in that experiments were conducted by removing aldicarb from a production system which has developed around its use. Aldicarbs value to potato production with in a production system not dominated by pesticide use may be substantially different. Growing season nematode population density measures of the database created unique problems. Relatively few nematode population samples were reported. Variation in the frequency of sampling (multiple measures) and unevenly spaced sampling complicated the statistical analysis. Data - I bias in study tl The bias in a mei Data Av; Pr! density accumul estimat were ma densiti Da regard yield I study 1 report. StUdie: data a Burban B, A, tuhers knobby knobs Classj total 65 bias in this situation relates more to the results of this study than the data reported in the original field research. The bias developed when an attempt was made to use the data in a meta-analysis format. Data Availability Pre-plant measures study year and nematode population density data were complete. Information on degree day accumulation, if not reported in the original study, was estimated using historic weather information. No estimates were made for missing in-season nematode population densities. Data reported for Atlantic potatoes was complete in regard to yield information. However, for Superior tuber yield measures, 13 studies had complete data reported. One study lacked a measure for the B size category. One study reported only the yield of A size category potatoes. Three studies reported only total yield. For Russet Burbank, five studies had complete yield data and three studies reported only total yield. Russet Burbank potatoes were a special case in that in addition to B, A, and J size classes, data regarding deformed or knobby tubers was reported. The decision was made to ignore the knobby class in the size categories, but to include the knobs in the total weight category. B, A, and J classifications represent economic differences, whereas total tuber weight represents biomass production. Missing Si] given t4 points 1 potato j Mi 1984). missing having relativ against The prc accurac conside Tl using 1 Yield < size c1 then u; Partia 17 the CWt/ac th/ac eStima 66 Missing Values Since the original data were not collected with thought given to potential for meta-analysis, some desired data points were unavailable. Of the optimal 132 points in the potato yield section, 22 were missing. Missing values create a dilemma (Tabachnik and Fidell, 1984). Many statistical procedures do not accommodate missing values and disregard all data associated with cases having missing values. It is necessary to weigh the relative worth of the existing measures for an observation against the uncertainty added by estimating missing values. The proportion of missing values to existing values and the accuracy of the estimation procedure are important considerations. . The decision was made to estimate missing yield values using mean proportions of the existing data. Where complete yield data existed, the ratio of each size class to the A size class and total yield was determined. This ratio was then used to estimate missing size class values based on the partial information available (Table 12). In study number 17 the reported total yield of Superior tubers was 196 cwt/acre. The A tuber yield was estimated to be 173 cwt/acre (196 X 0.887). Twenty—two missing data points were estimated using this procedure. Table 1; and A s: Estimatc Aldicarl Check Aldicar] sneer... T - Bas« A - Bas: Estimat Th yield m evaluat paired differs being r Table 1 estimat m Tested Estimgj Aldicai Check Aldica: Leer T ‘ Ba: A ‘ Ba: °f eac adequa Burban °f the Yield 67 Table 12. Mean proportion of tuber size classes to Total and A size Classesp Superior Russet Burbank Atlantic Estimator _B_ A _J_ B A .pl_ _N_ _g_ A J Aldicarb T 4.3 88.7 7.0 17.5 69.7 5.5 7.3 7.1 81.3 11.6 Check T 4.8 90.8 4.4 23.9 66.1 4.5 5.4 7.0 86 7 6.3 Aldicarb A 5.8 7.5 Check A 5.2 4. 6 T - Based on reported total tuber yields A tuber yields A - Based on reported Estimator Test The mean proportions of available B, A, J, and Knobby yield measurements to A and Total yield measurements were evaluated as an estimator for missing yield measurements. A paired t-test was performed. The probability of the difference between an estimated value and a measured value being not different from zero was calculated (Table 13). Table 13. Paired t- test probabilities associated with estimation of missing size class measures using mean proportion. Tested Superior Russet Burbank Estimator B A J B A J K Aldicarb T 0.85 0.98 0.93 0.86 0.91 0.86 0.04 Check T 0.82 0.99 0.89 0.92 0.79 0.95 0.00 Aldicarb A 0.06 0.92 ' Check A 0.93 0.83 T - Based on reported total tuber yields A — Based on reported A tuber yields The t-test results indicated that the mean proportion of each size class to available total yield data was adequate for estimation of all size classes except Russet Burbank knobby. Because of the low significance (PR=0.06) of the estimate, the mean proportion of A yield to total yield was not used as an estimator. Mean proportions for each tuu estimati Gel two catc used to of anal Medel_fl A tendenc of aldi fails t only p: informa A additic the deg loss. but st: aPplica M‘ PrEGic initia determ reBree avails This t ’ 68 each tuber size class to Total yield were used for all yield estimations. Meta-analysis General meta-analysis methodologies are divided into two categories. The first is a discussion to the hierarchy used to organize the research. The second is a description of analytical methodologies used in meta-analysis Model Hierarchy I A mean yield loss model provides a measure of central tendency and is the simplest method of estimating the value of aldicarb-to potato production. Its weakness is that it fails to account fOr variation in study results and uses only presence or absence of aldicarb application as the information source. A cumulative probability distribution model, in addition to providing Central tendency information, shows the degree of uncertainty associated with average yield loss. This type of model improves decision making ability but still is based only on presence or absence of aldicarb application. Multivariate regression models increase the degree of predictability by using additional information suCh as initial nematode population and/or planting date to determine aldicarb application value. Variables in regression models can be chosen based on the information available during different portions of the growing season. This type of model may improve grower pesticide use decision making al the syst' A b season n estimate correlat does not A 1 dynamic: stolon . partiti represe because factors Th decisic the deg model. Structl inform. the li resear used t aldica limita I desCr: 69 making ability, but does little to explain the biology of the system. I A biologically based model of the system links in- season nematode population dynamics with functions that estimate yield loss. This type of model shows an implied correlation between nematode populations and yield loss but does not show how the nematode causes tuber yield loss. A fifth type of model would relate nematode population dynamics to plant growth parameters such as root growth, stolon initiation, root uptake of soil water, or plant partitioning of photosynthates. This type of model represents implied causation and should be the most accurate because of its potential sensitivity to potato management factors such as nitrogen application or irrigation. The choice of which type of model to be implemented for decision making should be based on decision objectives and the degree to which available data supports the chosen model. The advantage of using a hierarchical modeling structure is that it can be used to organize available information into a usable formats, while clearly indicating the limitations of the information base. The integrated research review and meta-analysis design of this study was used to make optimal use of information available for aldicarb use decision making, and to document the limitations in currently available information. Research methods which were used for all analyses are described in the general analytical methods section of this if“‘“‘ chapter. results primary fiene£e1_ Th1 aldicarl non-tre: pestici. pests. plots w Al softwar Since t Signifi signifi associa aPPr0p1 Si accomm. analys determ reBree TEgreé 0n Var Signi: the Inc 70 chapter. Because of the number of methods used, methods and results for each analysis are organized based on the five primary research objectives. General Analytical Methods Through out the study treated refers to plots where aldicarb was applied to control P. enetrans, check refers to non-treated controls. In both treated and check plots pesticides may have been applied to control non-nematode pests. The singular difference between treated and check plots was the application of aldicarb. All statistical analyses were performed using SAS software on a VAX 1170 in the MSU Entomology Department. Since this is a descriptive study, 0.15 was used as the significance level for discussion and for the minimum significant difference calculations of ANOVA. The associated probability of each mean is provided where appropriate. SAS General Linear Methods (GLM) procedures accommodates unbalanced data design and was used for analysis of variance and yield impact work for researcher determined models. SAS STEPWISE procedures were also used. In stepwise regression, variables are entered one at a time into the regression equation and then retained or set aside depending on variable statistical significance criteria. A significance level of 0.40 was used for variable entry into the model. A significance level of 0.20 was required for variable that the statisti disadvar accommoc analyze< availabl for des: small t 71 variable retention. The advantage of stepwise procedures is that the variables in regression equations are based on the statistical significance of those variables. The disadvantage in using this procedure is that it does not accommodate unbalanced design.~ Each variety must be analyzed separately, effectively lowering the sample size available for regression analysis. This procedure was used for descriptive purposes and when sample sizes were too small to accommodate a researcher designed GLM model. Me describ De parts. the rar met lit analysf determ: suppori 72 Meta-Analysis Objective 1 Study Variability Methods used under Meta-Analysis Objective 1 are described. Associated results are provided. Specific Methodology Description of original study variability has two parts. The first uses simple descriptive statistics to show the range of results reported in scientific literature which met literature selection criteria. The second is an analysis of variance conducted on selected parameters to determine if results reported in the individual studies were supported aCross studies. Variability in Potato Production Measures For each variable in the PPDB the number of observations, number of missing observations, mean value, standard deviation“ minimum value, and maximum value was calculated. Results reported for this section are for original data. No estimated values were used. Impact of Selected Management Practices on Tuber Yield The hypothesis that potato cultivar, P.penetrans and aldicarb impact potato yield was statistically tested (PR=0.15) to see if results reported in single studies were globally supported. A three way analysis of variance was performed using SAS GLM for unbalanced ANOVA. Unbalanced ANOVA accommodates unequal n-counts for main effects. Yields of B, A, J, and the Total were analyzed separately. Presence or absence of aldicarb, potato cultivar and pre-plant nematode count were independent variables. Aldicar dichotc a conti variab] soil at mm:3 of ini1 repres« to div. (high,: Cultiv Superi In seaso] Soil i Nemat, 0 to . treat. The n 58 wi 73 Aldicarb or no aldicarb (1 or 0, respectively) is a nominal dichotomous variable. P.penetrans population at planting is a continuous variable converted to dichotomous ordinal variable with 0 indicating less than 23 nematodes / 100cm3 soil and 1 indicating greater than or equal to 23 nematodes /1000m3 soil. 23 nematodes /. 100cm3 soil was the midpoint of initial nematode count distribution. It was used to represent high vs. low initial nematode count. An attempt to divide initial nematode count into three categories (high,medium,low) resulted in empty analysis cells. Cultivar was a three level nominal variable with 1 = Superior, 2 = Russet Burbank, and 3 = Atlantic. Meta-Analysis Objective 1 Results Variability in Potato Production Measures Superior. The typical planting date was May 15 with dates ranging from May 1 to May 29. Pre-plant nematode 3 population density ranged from 0 to 54/100 cm: soil with an 3soil. The typical growing season average of 22.3/100 cm length was 114 days with values ranging from 95 to 161 days. Soil nematode population density during the growing season ranged from 0.0 to 54.0/100 cm? for aldicarb treated soil and from 1.2 to 120.8/1OOCm3 for non-treated soils. Nematodes in the roots of aldicarb treated plots ranged from 0 to 48/1.0 gram fresh root. Nematodes in the roots of non— treated plots ranged from 14.8 to 213.8/1.0 gram fresh root. The number of days between nematode samples ranged from 2 to 58 with an mean of 24 days. T1 and 11. respeci were 25 respeci were 21 treate« 281.65 treate S deviat are p: E but da counts 33.5. Value: $6380] Soil ; Nemat. range roots fresh range and e 74 The mean yields of B size category tubers were 12.40 and 11.35 cwt/acre for aldicarb treated and non—treated, respectively. The mean yields of A size category tubers were 280.89 and 236.20 for aldicarb treated and non-treated, respectively. The mean yields of J size category tubers were 21.39 and 11.33 cwt/acre for aldicarb treated and non- treated, respectively. Mean total yields of tubers were 281.65 and 224.54 cwt/acre for aldicarb treated and non- treated, respectively. Sample size, number of missing points, means, standard deviations, minimum, and maximum values for all variables are provided (Appendix C, Table 1). Russet Burbank. The typical planting date was May 7 but dates ranged from May 2 to May 21. Pre-plant nematode counts ranged from 3.2 to 67/100cn9 soil with an mean of 33.5. The typical growing season length was 145 days with values ranging from 136 to 161 days. Soil nematode population density during the growing season ranged from 0 to 31/100 cm? soil for aldicarb treated soil and from 2.5 to 286/100cn? soil for non—treated soils. Nematode density in the roots of aldicarb treated plots ranged from 0 to 5.8/1.0 gram fresh root. Nematodes in the roots of non—treated plots ranged from 9.7 to 269/1.0 gram fresh root. The number of days between nematode samples ranged from 6 to 77 with a mean of 28 days. . The mean yields of B size category tubers'were 54.27 and 62.53 cwt/acre for aldicarb treated and non-treated, .‘ respeci were 2‘ non-tr- catego treate tubers and no S deviat are pi i dates Popula mean 1 days i seaso treat soils range roots fresh range 27.3 respg Were 75 respectively. The mean yields of A size category tubers were 277.55 and 211.67 cwt/acre for aldicarb treated and non-treated, respectively. The mean yields of Jumbo size category tubers were 25.45 and 13.58 cwt/acre for aldicarb treated and non-treated, respectively. Mean total yields of tubers were 346.69 and 274.04 cwt/acre for aldicarb treated and non-treated, respectively. Sample size, number of missing points, means, standard deviations, minimum, and maximum values for all variables are provided (Appendix C, Table 2). Atlantic. The typical planting date was May 7 but dates ranged from April 26 to May 16. Presplant nematode population density ranged from 2.5 to 57/100cm3 soil with a mean of 22.17. The typical growing season length was 136 days with values ranging from 118 to 140 days. Soil nematode population density during the growing season ranged from 0 to 25.3/100 cm? soil for aldicarb treated soil and from 5 to 237/100an soil for non-treated soils. Nematodes in the roots of aldicarb treated plots ranged from O to 19.4/1.0 gram fresh root. Nematodes in the roots of non-treated plots ranged from 0.1 to 37.0/1.0 gram fresh root. The number of days between nematode samples ranged from 2 to 58 with a-mean of 24 days. The mean yields of B size category tubers were 29.3 and 27.3 cwt/acre for aldicarb treated and non-treated, respectively. The mean yields of A size category tubers were 328.73 and 235.75 cwt/acre for aldicarb treated and non-tre categ01 treate< tubers and no: $- deviat are pr I_mpa_ct ’E (PR>F= accour ] impact a mini (+/-) was s: SuPer. and A- (Tabl Table nemat M Lee CHECK ALDIC Lee cHECI" m 76 non-treated, respectively. The mean yields of J size category tubers were 47.17 and 24.03 cwt/acre for aldicarb treated and non—treated, respectively. Mean total yield of tubers were 404.93 and 386.63 cwt/acre for aldicarb treated and non-treated, respectively. Sample size, number of missing points, means, standard deviations, minimum, and maximum values for all variables are provided (Appendix C, Table 3). Impact of Selected Management Practices on Tuber Yield B Tuber Yield. A significant result was obtained (PR>F=0.0001) for B yield. The variables in the analysis accounted for 66 percent of the variance. Initial nematode count significantly (PR>F=0.15)’ impacted B category potato yield (Appendix D, Table l). with a minimum significant difference (MSD) of 6.56. Treatment (+/-) aldicarb was not significant (PR>F=0.73). Cultivar was significant (PR>F=0.0001). The MSD for comparison of Superior and Russet Burbank was 10.91, for Russet Burbank and Atlantic was 12.44, for Superior and Atlantic was 10.35 (Table 14). Table 14. Influence of cultivar, initial nematode population density and aldicarb on B tuber vield. Russet WWW P.p > 23 mean n mean p mean p CHECK 12.34 9 52.30 4 25.64 5 ALDICARB 13.39 9 40.87 3 25.80 4 P.p >=23 CHECK ' 12.34 5 83.00 2 30.50 2 ALDICARB 13.36 5 67.67 3 33.33 3 A. yield explai F count There and Ru was 42 46.66 Table nematc tuber P.p < CHECK ALDIC P.p > CHECK m for ; anal} DITal nema' PPCO sepa sign with 42,7 77 A Tuber Yield. A significant result was obtained for A yield (PR>F=0.0042). The variables in the analysis explained 36 percent of the variance (Appendix D, Table 2). For the yield of A category potatoes initial nematode count and treatment were significant with a MSD of 26.47. There was no significant mean separation between Superior and Russet Burbank. The MSD between Superior and Atlantic was 42.76. The MSD between Russet Burbank and Atlantic was 46.66 (Table 15). ‘ . Table 15. Influence of cultivar, initial nematode population density and aldicarb on A tuberavield. Russet Superior Burbank Atlantic P.p < 23. mean p mean p mean p CHECK 230.73 9 205.75 4 335.00 5 ALDICARB 272.88 9 281.10 3 313.78 4 P.p >=23 CHECK 194.21 9 169.64 5 337.00 (AN ALDICARB 248.68 9 226.57 6 348.67 Jumbo Tuber Yield. Significant results were obtained for Jumbo yield (PR>F=0.0005). The variables in the analysis explained 42 percent of the variance (Appendix D,Table 3). I For the yield of Jumbo category potatoes initial nematode count and treatment were significant. The MSD for PPCODE and TRE was 4.94. There was not a significant mean separation between Superior and Russet Burbank. There was significant difference between Russet Burbank and Atlantic with MSD of 46.6. The MSD between Superior and Atlantic was 42.76 (Table 16). Table 1 nematoc Jumbo t P. p < 1 CHECK ALDICAJ P.p >=: CHECK ALDICA‘ IFS provid in the (Apper treat: Signi: Burba1 MSD f. Table nemat Leta; P.p < CHECF ALDIC .P-p : CHECI M Peso; insL 78 Table 16. Influence of cultivar, initial nematode population density and aldicarb on Jumbo tuber vield. Russet Superior Burbank Atlantic P.p < 23 mean p mean p mean p CHECK 12.38 9 17.13 4 24.44 5 ALDICARB 22.08 9 29.57 3 50.55 4 P.p >=23 CHECK _ 9.04 9 8.06 5 23.00 2 ALDICARB 19.50 9 17.75 6 42.67 3 Total Tuber Yield. Total tuber yield analysis also provided significant results (PR>F=0.0016). The variables in the analysis accounted for 39 percent of the variance (Appendix D, Table 4). e For Total potato yield, initial nematode count and treatment were significant with a MSD of 30.40. There was no signifiCant mean separation between Superior and Russet Burbank. The MSD for Superior and Atlantic was 53.59. The MSD for Russet Burbank and Atlantic was 60.64 (Table 17). Table 17. Influence of cultivar, initial nematode population density and aldicarb on Total tuber vield. Russet Superior Burbank Atlantic P.p < 23 mean p mean p mean p CHECK 255.46 9 304.85 4 385.08 5 ALDICARB 308.34 9 399.40 3 390.13 4 VP.p >=23 CHECK 212.44 9 249.40 5 390.50 2 ALDTCARR 279.58 9 320.33 ‘6 424.67 3 In this analysis of variance with the exception of PPCODE*CUL for B size category, all interaction terms were insignificant with (PR>T=>0.3). The results of this study indica1 signif: Russet but B 79 indicate that aldicarb and initial nematode count do have significant impacts on potato tuber yield. Superior and Russet Burbank yield differences are insignificant for all but B size category tubers. are pr F of ald obseri size < calcui treate ((tre: estim. to an use. Megp_ avera Categ 92mg; Calct meas1 relai freq1 Prob. expe Ont 80 Meta-Analysis Objective 2 Impact of Aldicarb on Tuber Yield Specific methods used under Meta-Analysis Objective 2 are provided as well as meta-analysis objective results. Specific Methodology Potential yield loss associated without the application of aldicarb was calculated for each pair of tuber yield observations in the PPDB, and categorized based on tuber size class and potato cultivar. Potential yield loss was calculated by dividing the difference between aldicarb treated and non-treated yields by the aldicarb treated yield ((treated — non-treated) / treated). This yield loss estimate may be bias. It is an estimate of aldicarb value to an agricultural managment practice dependant on pesticide use. Mean Yield Loss Mean potential percentage yield loss was determined by averaging yield loss values for each cultivar and tuber size category. Cumulative Probability Cumulative probability distributions were developed by calculating the relative frequency of yield reduction measurements by size category and cultivar. The relationship between yield reduction and associated relative frequency was then plotted. The abscissa of the cumulative probability distribution, gives the probability of experiencing equal to, or less than the yield loss indicated on the ordinate. pegres 1' yield coeffi detern varia} type c indica the v; decre varia a gro GLM w of of used not 5 Atlal had 1 Seas. imprl and incl grow 81 W Three regression analysis procedures for percentage yield reduction were developed. The sign of regression coefficient (Beta estimate) is included in an attempt to determine if regression coefficients are a function of that variable's impact on percent yield loss or a function of the type of regression analysis. Positive Beta estimate signs indicate an increase in yield loss with increasing values of the variable. Negative Beta estimate signs indicate decreasing yield loss with increasing values of the variable. Preseason Model. Change in yield based on information a grower has at time of planting was analyzed using GLM. GLM was used because it accommodates the unbalanced number of observations available for each variety. For the GLM procedure, DV1 and DV2 are dummy variables used to indicate variety. DV1 indicates Superior (1), or not Superior (0). DV2 indicates at Atlantic (1), or not Atlantic (0). Russet Burbank was chosen as 0,0 because it had the least complete original data. Postseason Model. A sub-objective was to use post season information available in_a retroactive mode to improve predictive ability. Growing season length (GSL), and total tuber yield in aldicarb treated plots (TWT) were included along with the variable in the preseason model. TWT was used as an indicator of the general quality of the growing season. under proce selec for a and I Mean tube: for I Table e221; Maris Supe] Russ: um Cumu‘ from Yiel from The 80% Tota bein inc: beir 82 Stepwise variable selection. The third sub-objective under regression analysis-involves using stepwise procedures. The postseason variables were available for selection by the stepwise procedure. Results are provided for average yield loss, cumulative probability distribution, and regression analysis sub-objectives. Meta-Analysis Objective 2 Results Mean Yield Loss Mean yield loss results varied from a 17% increase in tuber yield for Russet Burbank to a 52% tuber yield decrease for Russet Burbank Jumbo yield (Table 18). Table 18. Mean percentage yield loss without aldicarb application by variety and tuber size class. Variety _p B A Jumbo Total Superior 18 10 16 47. 22 Russet Burbank 8 -17 23' 52 20 Atlantic 7 5 -2 50 4 Cumulative Probability Distribution In Superior B grade potatoes (Figure 9), results ranged from an increase of 6% to a decrease of 27% with 50% of the yield losses being less than 8%. A grade yield loss ranged from 8 to 45% with 50% of the losses being less than 13%. The impact on Jumbo tubers ranged from a 5% increase to an 80% decrease with 50% of the losses being less than 48%. Total yield loss ranged from 10 to 51% with 50% of losses being less than 20%. For Russet Burbank (Figure 10), B grade potato yield increase ranged from 33 to 5% with 50% of the increases being less than 20%. A grade yield loss ranged from 8 to 83 CW PROS < GIVEN YIELD REDUCTION - ‘10 1D 30 SCI 70 90 X DECREASE FEW ALDICARB TREATED POTATO TOT Figure 9. Cumulative probability distribution for the impact of aldicarb on Superior tuber yield by tuber size category. CW PROS < GIVEN YIELD REDUCTION I fi‘“ I I I I fl f I I fi -30 — 10 «o 30 so 70 X DECREASE FRW ALDICARB TREATED POTATO B TOT Figure 10. Cumulative probability distribution for the impact of aldicarb on Russet Burbank tuber yield by tuber size categoryi than incre less 50% o tuber 26% w The i 11% I than 42 t< less an i resu 84 43% with 50% of the results indicating a yield loss of less than 13%. The impact on Jumbo tubers ranged from a 5% increase to an 80% decrease with 50% of the losses being less than 48%. Total yield loss ranged from 10% to 51% with 50% of losses being less than 20%. For Atlantic potatoes (Figure 11), impact on B grade tuber yield ranged from an increase\of 7% to a decrease of 26% with 50% of the results showing less than a 2% increase. The impact on A grade tuber yield ranged from an increase of 11% to a decrease of 8% with 50% of the results showing less than a.4% increase. Jumbo tuber yield decreases ranged from 42 to 57% with 50% of the results indicating a yield loss of less than 48%. The impact on total tuber yield ranged from an increase of 4% to a decrease of 13% with 50% of the results indicating a less than 1% decrease. 1 n 0.9 0.8 0.7 0,6 0.5 BA 0.3 0.2 0,1 I l I *1 4D 50 0‘! I T I f T —20 0 20 CW PROS < GIVEN YIELD REDUCTION X DECREASE FRCM ALDICARB TREATED POTATO B + 0 A TOT Figure 11. Cumulative probability distribution for the impact of aldicarb on Atlantic tuber yield by tuber size category. , Regre obtai (APPe signi Regre 19). Table resul Depel Variz Delta Deltz DElti 1.3% obta (APP perc rang Tam em B t1 Var 85 Regression Analysis Preseason. Significant regression results were obtained for percentage B, A, and total tuber yield loss (Appendix E, Tables 1,2, and 3 respectively). No significant regression was obtained for Jumbo yield. Regression r-square values ranged from 0.49 to 0.58 (Table 19). Table 19. Summary of preseason regression analysis results for percentage yield loss by tuber size class. Dependant Model Beta coefficient sign Variable PR>F RA 2 DV1 DV2 APO PJD Delta B wt. 0.0007 0.5 + + + — Delta A wt. 0.0030 0 - - - - Delta J wt. . . . . . Delta T wt. 0.0008 0.49 + - - Postseason Model. Significant regression results were obtained for percentage B, A,and Total tuber yield reduction (Appendix F, Tables 1,2,and 3 respectively), but not for percentage Jumbo yield reduction. Model r—square values ranged from 0.61 to 0.66 (Table 20). Table 20. Summary of postseason regression results for percentage yield loss by tuber size class. Tuber Size Model Beta coefficient sign __§1§§$ _£EZE_ 31.; DE; DE; AEQ BED §§L TEE B 0.0035 0.61 + + + - + + A 0.0018 0.64 + r - + + + Jumbo . . . . . . . . Total 0.0001 0.66 + — + + + + Stepwise Variable Selection. For Superior percentage B tuber yield reduction, PJD was the only variable that met variable selection criteria (Appendix G, Table 1). For Super model signi yield 3). 21). Table for 1 crit usin was redu Mode Was For 86 Superior A yield, PJD was the only variable which met the model criterion (Appendix G, Table 2). No variables met the significance criterion for predicting Jumbo yield. Total yield was estimated using APO GSL and TTW (Appendix G, Table 3). Model r-square values ranged from 0.39 to 0.86 (Table 21). Table 21. Summary of stepwise regression results for percentage tuber yield loss on Superior. Tuber Size Model Beta coefficient sign Class R" 2 PR > F m M & T_Tw_v B 0.39 0.0840 . - - . A 0.70 0.0003 . - . . Jumbo . . . . . . Total 0.86 0.0001 + . + + For Russet Burbank B yield no variables met the entry criterion. Percentage A tuber yield reduction was predicted using PJD (Appendix G, Table 4). No significant regression was obtained for Jumbo yield. Percentage total tuber yield reduction was predicted using PJD (Appendix G, Table 5). Model r-square values ranged from 0.74 to 0.76 (Table 22). Table 22. Summary of stepwise regression results for percentage tuber yield loss on Russet Burbank. Tuber Size Model Beta coefficient sign Class RA 2 PR > F APO PJD GSL TTW B I O O O O t A 0.74 0.0266 . + . . Jumbo . . . . . . Total 0.76 0.0133 . + . . For Atlantic percentage B tuber yield reduction, APO was the only significant variable (Appendix C, Table 6). For percentage A tuber yield reduction PJD and GSL were selec yield Table predi Regre 23). 87 selected (Appendix G, Table 7). Percentage Jumbo tuber yield reduction was predicted using PJD and TTW (Appendix G, Table 8). Total tuber percentage yield reduction was predicted using APO PJD and GSL (Appendix G, Table 9). Regression r-square values ranged from 0.37 to 0.97 (Table 23). Table 23. Summary of stepwise regression results for percentage tuber yield loss on Atlantic. Tuber Size Model Beta coefficient sign Class RA 2 PR > F APO 2gp QSL TEE B 0.38 0.1412 + . V. . A 0.82 0.0307 . - - . Jumbo 0.73 0.0715 . - . - Total 0.97 0.0088 + - - . Objec deve autc Mede acct inde the the: exP Pep' 88 Meta-Analysis Objective 3 Impact of Aldicarb on P.penetrans Populations Specific methods and results for Meta-Analysis Objective 3 are provided. Specific Methodology The determination of the impact of aldicarb on P.penetrans population dynamics has two sections. The first involves using researcher selected regression models to describe population changes. Nematode population density in soils, root, and Total (soil+root) were analyzed for aldicarb treatments and non-treated checks. Threats to model validity such as autoregression and heterogeneous variance structures are discussed. The second involves an attempt to use distributive delay modeling to identify the impact of aldicarb on P.penetrans population dynamics. Regression Analysis Regression analysis methods include those for model development and threats to model validity such as autocorrelation and heterogeneity of variance. Model Development Regression procedures were developed using degree day accumulation and time measured as day of the year as independent variables. Stepwise procedures indicated that the model equation was third order with respect to either of these variables. Degree day accumulation can be used to explain 48 and 64 percent of the variation of nematode population dynamics in non—treated soils on Superior and Russe expla soil: were dynaI day . limi inde is l The day ave] AVR res: acce Thr. are mea reg The is Pre 89 Russet Burbank respectively. Day of the year can be used to explain 47 and 60 of population variation in non-treated soils on Superior and Russet Burbank respectively. There were insufficient measures to model P.penetrans population dynamics on Atlantic potatoes. Day of the year and degree day are correlated. Their combined explanatory ability is limited. Day of the year was chosen as the predominant independent variable. Error associated with day of the year is less than that for calculating degree day accumulation. The gain in explanatory power experienced by using degree day for Russet Burbank is small. For each population dynamics regression analysis the average residual (AVR) was determined using: AVR = Summation l=1,n {predicted-measuredj/n. The average residual is used to provide a measure of the predictive accuracy of the model. Threats to Model Validity Two threats to model validity which must be analyzed are autocorrelation and heterogeneity of variance. Autocorrelation. Autocorrelation occurs when measurements within a data set are not independent. The regression analysis assumes that variables are independent. The number of nematodes at one time period for a given study is assumed to be a function of the number of nematodes in previous time periods. was c non-t Durbf in t: orig: cont; pair with pair divi betw sang 1 ther ‘ whie sam] Stm ext: fou val Var reg mar tin 90 To test the degree of this association, a new variable was created which represented the nematode population in non-treated soils at the previous sampling within a study Durbin-Watson statistic was not valid due to unequal spacing in time series. Each sample was then a pair of samples, the original and its first-order autoregressor. If a study contained five sampling dates the second sample would be paired with the first, the third with the second, the fourth with the third, and so on. The average spacing between samples was 23 days. Data pairs (sampling value and autoregressor) were sorted and divided into classes dependent upon the length of time between a sampling and its first order autoregressive sample. Class 1 contained observation pairs which were less then 11.5 days apart, Class 2 contained observation pairs which were less than 23 days apart, and class 3 contained samples which where more than 23 days apart. To explore higher order degrees of autocorrelation, studies with 4 or more samples per growing season were extracted from the database, and analyzed for first to fourth order autocorrelation. Heterogeneity of Variance. The second threat to model validity is heterogeneous variance. For regression analysis variance associated with nematode population density is required to be uniform throughout the growing season. For many biological systems population variation increases with time. This potential threat to model validity was studied by g] VBISI U ...; m E! ‘ - a re lite seco info Dist temp mort wore deve sea: 91 by graphical analysis of regression analysis residuals versus the independent variables in the regression equation. Distributed Delay The Meta-Analysis Objective 4 requires the existence of a reliable nematode model. For this reason a second literature search was conducted. The objective of this second literature search was to determine the extent of information available for distributed delay modeling. Distributed delay modeling is based on the impact of temperature on population life stage developmental rates, mortality, and natality. BIOL and BIOB data bases were searched using the key words (Pratylenchus penetrans or Root lesion nematode, and development or temperature or degree day). The literature search revealed eight studies meeting these criteria. Meta-Analysis Objective 3 Results Regression Analysis For P.penetrans population density on Superior potatoes model r—square values ranged from 0.17 to 0.45 (Table 24). Complete regression results for soil, root, and total population densities in aldicarb treatments and non-treated controls are provided (Appendix H, Tables l,2,3,4,5, and 6 respectively). Table 24. Summary of regression analysis results for soil, root, and total nematode population densities on Superior. SOIL ROOT TOTAL Check 0.31 0.0153 16.4 0.47 0.0018 33.5 0.45 0.0001 36.9 Aldicarb 0.32 0.0133 7.4 0.17 0.3376 5.1 0.20 0.0478 10.2 * AVR - average residual = Summation i=1,n {measured- predictedI/n dens. prov pota (Tab tota cont den pro 92 The results of regression analysis for population densities in the soil root and total on Superior are provided (Figures 12,13,and 14 respectively). For P.penetrans population density on Russet Burbank potatoes model R-square values ranged from 0.45 to 0.86 (Table 25). Complete regression results for soil, root, and total population densities in treated and non-treated controls are provided (Appendix H, Tables 7,8,9,10,11, and 12 respectively). Table 25. Summary of regression analysis results for soil, root, and total nematode population densities on Russet Burbank. SOIL . ROOT TOTAL Treatment RA 2 PR > F AVR RA 2 PR > F AVR RA 2 PR > F AVR Check 0.66 0.0002 27.9 0.86 0.0001 20.9 0.53 0.0007 57.4 Aldicarb 0.47 0.0124 4.0 0.45 0.1540 0.8 0.83 0.0004 4.4 AVR - average residual = Summation i=1,n {measured- predicted}/n The results of regression analysis for population densities in the soil root and total on Russet Burbank are provided (Figures 12,13,and 14 respectively). WIT 93 130 fi 120 — o n 110 — :‘ 8 100 — 8 so — O 8 ° 0 “ so — a 8 D. U 70 — o >— +; so — lg 0 Lu C) Z 9 e. 4 .J D O. O D. I 200 DAY G YFAR D ALDCARB WSERVED + ALDICARB ESTIMATED 0 CHECK CQSERVED A G-IECK ESTIMATED Figure 12. Soil nematode population density on Superior - simulated and observed vs. day of year I" g 180 8 ° . 150 O O D .; 14o \ C! u 120 U )- ‘: ma 0‘) Z N 0 so § I e— 60 < .1 3 P ‘° DAY a: mm D ALBUM OBSERVED + ALDICARB ESTIMATED 0 CHECK oqsenveo A CHECK ESTIMATED Figure 13. Root nematode population density on Superior - root simulated and observed vs. day of year 94 320 — O 300 — 280 — 250 — ° 0 240 —— o 220 — 200 — POPULATION DENSITY (P.p/(1.0 9‘100 ccjj 40 — 8 o 0 B u 20 — a 0 E 0 WWW o ' "' I H I “'1 U I Y I T 1 ' 160 190 200 220 240 2150 DAY 0" YEAR D ALDXZARB cassava: + ALDICARB ESTIMATED 0 CHECK CQSERVED A CHECK ESTIMATED Figure 14. Total nematode population density on Superior - simulated and observed vs. day of year. POPULATION DENSTIY [P.p/100 CC SOIL) 0 I q P fin r W ”F P’H 9 q I . 130 150 170 190 210 230 250 270 DAY G YW U ALDICARB COSERVED + ALDICARB ESTIMATED 0 CHECK QSERVED A CHECK ESTIMATED Figure 15. Soil nematode population density on Russet Burbank - simulated and observed vs. day of year. 95 250 — O 240 - C o 8 22C] — CI: :5» 200 — o ‘ ‘IBO — ,_ \ 9 150 — u U )_ 140 — I. 33 120 —— Ci 0 100 — § ’_ BU — 3 a 60 — o E 40 _ o o 20 — ° 0 0 w: a '- -' wI a I H I I —1 I 140 150 180 EDD 220 240 250 DAY cs Yam [3 ALDICARB 0055mm + ALDICARB ESTIMATED 0 CHECK oqsaavgp A CHECK ssrwmzo Figure 16. Root nematode population density on Russet Burbank - simulated and observed vs. day of year 400 0 n 350 — o I“ 0‘ o O. 3' 300 — V B 250 - S o \ d o E; 200 - o C ,. ‘2 UJ 150 — D o O 3 o ,2 100 _ ‘1 5 o O. 0 8 5:: ° 0 9 o L? D -r -3 D 3 '11 P: r-' I | 140 160 190 200 220 240 250 DAY 0: W D ALDICARB GISERVED + ALDICARB ESTIMATED 0 CHECK msEHVED A CHECK ESTIMATED Figure 17. Total nematode population density on Russet Burbank - simulated and observed vs. day of year Thre hete non- beta betv were 3 re aut< is ; ind: Cor sam sam and (PE is 1161 SO: 18; 96 Threats to Model Validity Threats to model validity included autocorrelation and heterogeneity of variance. Autocorrelation. For nematode population density in non-treated soil samples the Pearson Correlation Coefficient between values and their autoregressor decreased as time between samples increased. Pearson Correlation Coefficients were 0.61, 0.39, 0.34 for sample spacing categories 1,2, and 3 respectively indicating the degree of first order autocorrelation and its sensitivity to sample spacing. This is important because it shows first order violation of the independence of measures statistical assumption. Second samples were statistically correlated (Pearson Correlation Coefficient) with first samples, 0.81. Fourth samples where correlated with second samples, 0.73. Fifth samples where correlated with fourth and third samples 0.66 and 0.53 respectively. There were no other significant (PR>IRI=.15) correlations. This again supports that there is an independence threat to model validity. Heterogeneity of Variance. Heterogeneity of variance was most clearly associated with the impact of initial nematode population densities on regression analysis for soil population density in Superior check treatments (Figure 18). IO H- rn on] dex f0] 19: To] do te' 0n Cu de 97 n U 3 40 g u \ 9 30 O. \J 2’ 20 D Cl [1 8 CI D D m 10 a CI o u n :’ 0 G '3 ‘3 " “ o u U ‘5' U) D U V -10 D D a D B a D 0 LI) -20 c] E u g —30 a o I j I I 20 40 EU 01 INITIAL POPULATION DENSTIY (P.p/1UOCC) Figure 18. Superior check soil nematode population density regression residual vs. initial nematode population density. Distributed Delay of the eight studies retrieved by the literature search only one contained information regarding P.penetrans development on potatoes. Developmental rates were reported for P.penetrans on alfalfa (Townshend, 1984; Kimpinski, 1981), timothy (Kimpinski, 1981), soybean (Acosta, 1979), Tobacco (Townshend, 1977), and onion (Ferris, 1970). One document was retrieved which reported potato production and temperature (Burpee, 1978). Burpee's analysis concentrated on plant development and not nematode development. Currently, insufficient information exists for distributed delay modeling of P.penetrans population density on potato. unde net and usj the see tre em on Va de tr 98 Meta—Analysis Objective 4 Impact of In-season P.penetrans Populations on Tuber Yield Specific methodology and results of research performed under Meta—Analysis Objective 4 are provided Specific Methodology An attempt was made to determine if there is a window of time in which the nematode populations most affectively impact yield. The impact of in-season P.penetrans population dynamics on yield was analyzed by division of population density measures into time categories. The small number of samples taken during the growing season limits the thoroughness with which this question can be analyzed. Correlation The degree of linear relationship between delta nematode population values for two class (early, and late) and percentage tuber yield reduction measures was determined using Pearson product-moment correlation. For each study the growing season was divided into two (early and late season) segments. The difference between treated and non- treated soil, root, and total nematodes was calculated for early and late season time classes. If there was more than one sampling date reported in a study time class, reported values were averaged. Regression Two regression analyses were used in an attempt to determine the impact of in-season nematode populations on tuber yield. The analysis for two time classes was used to dete root dete port dens pree r00‘ in lat var det P0P P0P nen deI sax be' de Pr an re 0P it PI St 99 determine which portion of the nematode population (soil, root, or soil+root) was most important to yield impact determination. Four time classes were used to determine the portion of the season in which total nematode population density has the greatest impact on tuber yield. Two Time Classes. Five regression models were used for prediction of percentage tuber yield loss based on soil, root, and total nematode populations. Impact of P.penetrans in soil was determined using pre-plant, early season and late season soil population density measures as independent variables. Impact of P.penetrans populations in roots was determined using early and late season root nematode population measures. Impact of total P.penetrans populations was determined using early and late season nematode population measures. Four Time Classes. Yield loss functions were also determined from studies which contained four or more samples. DTNl, DTNZ, DTN3, DTN4 represent the difference between check and aldicarb treated total nematode population densities for the first to the fourth sampling respectively. Pre-plant nematode density, planting date, DTNl, DTNZ, DTN3, and DTN4 where used to predict percentage tuber yield reduction for each size class. Total nematode density as opposed to soil, or root nematode density was used because it was the only class for which significant regression predictors were available under Meta-Analysis Objective 3. Stepwise procedures were used to determine the model. ./4 O D—‘ D) rn I coei rep] POP‘ Tab cha' soi m gr po fc 100 Meta-Analysis Objective 4 Results Class Correlation Significant (PR>{R}=.15) Pearson correlation coefficients are reported (Table 26). Positive values represent an increase in yield loss with higher nematode population densities. Table 26. Pearson correlation coefficients for percentage change in B, A, Jumbo, and Total tuber size classes with soil, root, and total nematode population density for two time categories. Soil Root Total Variety Size Early Late Early Late Early Late Superior Superior Superior Superior Russet Burbank Russet Burbank Russet Burbank Russet Burbank Atlantic Atlantic Atlantic . . . Atlantic . 0.87 . 0.95 . 0.83 . - no significant correlation . . . -0.41 0.44 . 0.44 0.42 0.72 0.78 0.59 0.33 0.50 0.72 . . -O.52 -0.72 . -0.47 0.40 -0.58 -0.72 . . o o o u o o . 0.82 . 0.90 . 0.72 6C4b'wriC45'w'ic4b‘w Regression Nematode populations in roots appeared to have the greatest impact on tuber yield for Superior, while total population density had the greatest impact on tuber yield for Russet Burbank. Two time classes. Four analyses resulted in statistically significant (PR>F .15) regression results (Table 27). Complete regression information is provided (Appendix I, Tables 1,2,3, and 4). Tab] perc @112. m m Soi Rod Tot: Eli Soi R00 19; for Con Tat Ta] ss Lg -sssw 101 Table 27. Impact of in—season nematode densities on percentage yield loss by cultivar, B, A, Jumbo, and Total and size categories. Tuber size categorv Model B A Jumbo Total Superior RA 2 PR > F RA 2 PR > F RA 2 PR > F RA 2 PR > F Soil . . . . . . . . Root . . 0.82 0.0311 . . 0.63 0.1353 Total . . 0.35 0.1420 . . . . Rus Burb Soil . . . . . . . . Root . . . . . . . . Total . . . . 0.96 0.0375 . . . - no significant regression Four Time Classes. Significant results were obtained for Superior (Table 28) but not for Russet Burbank. Complete regression results for percentage yield loss for B, A, Jumbo, and Total tuber yield are provided (Appendix J, Tables 1,2,3, and 4 respectively). Table 28. Summary of stepwise regression results for in- season total nematode population density on percentage tuber vield reductionq, Tuber size Beta COEFFICIENT SIGN Class PROB > F R-SQUARE A29 DTN1 DTN2 DTN3 DTN4 B 0.003 0.73 . + . . . A 0.004 0.83 . . + . + Jumbo - 0.077 0.58 + . . - . Total 0.006 0.81 . + . . + . - no significant regression per grc reg org re] deI If“? in 19 Re pc CC 102 Meta-Analysis Obiective 5 Impact of Aldicarb and P.penetrans on Plant Development Specific methodologies and results for research performed under Meta-Analysis Objective 5 are provided. Specific Methodology The impact of P.penetrans population density on potato growth and development was analyzed using correlation and regression analysis. Absolute and relative changes in plant organ growth were determined. Regression equations for relative partitioning between major sinks were also developed. Data Base In order to study the impact of aldicarb and P.penetrans on potato plant development, an additional information source was tapped. During the period (1985- 1987) research was conducted at the Montcalm County Potato Research Farm to provide a data base for validation of potato modeling efforts. Data collected in 1986 was confounded by poor germination in the spring and flooding in the fall and was excluded from this study. The Model Validation Data Base includes weekly or biweekly measurements of nematode populations in soil, stolon, and roots (Appendix B, Table 7) as well as plant growth parameters such as above ground, below ground, root, stem, stolon, and tuber biomass (Appendix B,Table 8). The difference between nematode population density in non-treated and aldicarb treated plots (delta) is assumed to be the impacting portion of the nematode population. 103 Changes in plant growth parameters were expressed as delta values (treated mass - non-treated mass) or as relative values (non-treated mass/treated mass). Partitioning was expressed as (mass of sink/total plant mass). Correlation Analysis Correlation analysis was performed to quantify the degree of linear relationship between nematode population measurements and growth impact measurements. Days after planting (DAP), Delta soil, root, total, stolon nematodes (DSOIL, DROOT, DTOT, DSTOL respectively), were correlated with delta and relative below ground, above ground, root, stolon, and tuber biomass. Regression Analysis Regression analyses were used to determine the impact of aldicarb and P.penetrans on both absolute changes in plant growth parameters and relative partitioning of photosynthates to plant organs. Plant Growth Parameters. stepwise regression techniques were employed to determine nonlinear time relationships between nematode populations and change in delta and percentage plant growth parameters. Independent variables available for selection included delta soil, root, and total nematode counts as well as first through third order time (days after planting) measurements. Plppp_£gppipippipg. The affect of aldicarb and ELpeppppgpg on metabolite partitioning between the above ground, below ground, and tuber portions of the plant. The 104 ratio of below ground, above ground, and tuber biomass to total biomass was determined. The difference in relative partitioning between aldicarb and non-treated plots was then determined for graphic explanation of aldicarb / P.penetrans impacts. Stepwise regression techniques were employed to determine nonlinear time and nematode population impacts on plant partitioning. Variables available for selection included delta soil, root, and total nematode counts as well as first through third order time (days after planting) measurements. For below ground partitioning measures reciprocal first through third order (days after planting) measures were used as time indicators. Because a reliable nematode population model is not currently available, the analysis was repeated using only time variables. Meta-Analysis Objective 5 Results Correlation There were five significant correlations associated with delta plant growth parameters and eight significant correlations associated with percentage plant growth parameters. Delta Plant Growth. Significant (PR > IR}=.15) Pearson Correlation Coefficients are reported (Table 29). Positive values represent less mass in non-treated plots. 105 Table 29. Pearson correlation coefficients for days after planting, delta (soil, root, total, and stolon) nematode population densities with plant growth parameters expressed as a difference. Plant Nematodepparameters Parameters DAP DSOIL DROOT DTOT DSTOL Below Ground . . . . . Above Ground . 0.65 . . . Root . . . . . stolon . 0.54 . . . Tuber 0.76 0.88 . 0.55 . U G Stem . . . . . — no significant correlation Relative Plant Growth. For plant growth measurements expressed as a ratio significant (PR > {Rl=.15) Pearson Correlation coefficients are reported (Table 30). Negative values represent less mass in non-treated plots. Table 30. Pearson correlation coefficients for days after planting, delta (soil, root, total, and stolon) nematode population densities with percentage plant growth parameters. Plant Nematode parameters Parameters DAP DSOIL DROOT DTOT DSTOL Below Ground -0.50 . -0.76 -0.53 Above Ground . -0.65 . . 0.77 Root . . -0.69 -0.61 . stolon . . . . . Tuber . -0.51 . . . UG Stem . . . . . . - no significant correlation Regression Two significant regressions were obtained for delta plant growth parameters. Three significant regressions were associated with percentage plant growth parameters. Delta Plant Growth. Regression results for delta plant growth are provided (Table 31). Complete regression results for delta abOVe ground, and tuber growth parameters 106 are provided (Appendix K, Tables 1 and 2 respectively). Table 31. Summary of stepwise regression results for the impact of aldicarb and P.penetrans on delta plant growth parameters. Plant Growth Beta coefficient sign Parameter Prob>F r-s sgpare QAP DAP2 DAP3 DSOIL DROOT DTOT Below Ground . . . . . . Above Ground 0.0066 0.58 . . . + . . Root . . . . . . . Stolon . . . . . . . . Tuber 0.0019 0.68 . . . + . . U G Stem . . . . . - no significant regression Relative Plant Growth. Regression results are summarized (Table 32). Complete regression results for percentage below ground, above ground, and stolon growth parameters are provided (Appendix K, Tables 3,4, and 5 respectively). Table 32. Summary of stepwise regression results for the impact of aldicarb and P.penetrans on percentage plant growth parameters. Plant Growth Beta coefficient sign Parameter Prob>F r-sgpare QAP DAP2 DAP3 DSOIL DROOT DTOT Below Ground 0.0063 0. 58 . . . . — Above Ground 0.0411 0.38 . . . ' - . . Root . . . . . . . . Stolon 0.0177 0.48 . . . . - . Tuber . . . . . . . U G Stem . . . . . . . - no significant regression Plant Partitioning with Nematode Terms. Significant regression results were obtained for each of the partitioning sinks (Table 33). ~Days after planting as a cubic term was not selected by any analysis. Complete regression results for partitioning to below ground 107 aldicarb, below ground check, above ground aldicarb, above ground check, tuber aldicarb, and tuber check are provided (Appendix K, Tables 6,7,8,9,10,and 11 respectively). Table 33. Summary of regression results for the impact of delta nematode population parameters on partitioning ratio. Beta coefficient sign sink Treat PR > F R—square 9A2 DAP2 DSOIL DROOT DTOT Below Ground* Ald 0.0001 0.90 + + . . . Below Ground* Chk 0.0001 0.96 + + . - . Above Ground Ald 0.0003 0.87 . - . + . Above Ground Chk 0.0001 0.85 . — . + . Tuber Ald 0.0001 0.91 + . + . . Tuber Chk 0.0001 0.90 + + . * For below ground measures time variables (DAP,DAP2) equal (l/DAP, 1/DAP2) respectively Plant Partitioning wlo Nematode Terms. Significant results were obtained for each partitioning sink (Table 34). Complete regression results for partitioning to above ground aldicarb, above ground check, tuber aldicarb, and tuber check are provided (Appendix K, Tables 12,13,14, and 15 respectively). Table 34. Summary of regression results for partitioning ratio as a function of days after plantino. Beta coefficient sign Sink Treat PR > F R-square DAP DAP2 DAP3 Below Ground* Ald 0.0001 0.88 . + Below Ground* Chk 0.0001 0.91 . + Above Ground Ald 0.0001 0.83 . + Above Ground Chk 0.0002 0.76 . + . Tuber Ald 0.0001 0.90 + . . Tuber Chk 0.0001 0.88 + * For below ground measures time variables (DAP,DAP2,DAP3) equals (l/DAP, 1/DAP2, l/DAP3) respectively 108 These regression results for below ground, above ground, and tuber biomass partitioning ratio were graphically compared with the observed ratios (Figures 19,20, and 21 respectively). The difference in relative partitioning between aldicarb and non-treated plots was then determined and plotted versus days after planting (Figure 22). 0.9— 0.3 — 0.73 9 I»— § 0.5 - g E o s — 9 I— ; 0.4 - C! < O. 0.3 0.2 0.1 o I I I fl I I I l I I I '1‘ 20 40 SD 80 ‘100 120 DAYS AFTER PLANT l PG D ALDK‘ARB OISE-WED + ALDICARB EST'MMTED 0 CHECK GISERV'ED A CHECK ESTIMATED Figure 18. Above ground to total biomass partitioning ratio vs. days after planting 109 0.28 —e 0.28 — 0.24 — 022— 0.2 — 0.18 — 0.16 - 0.14 _ 0.12 _. PARTIT IONING RATIO 0.08—- 0.051 0.04 0.02 — I I 20 40 so an 100 120 DAYS AFTER OLANTI-IG D ALDICARB OBSERVED + ALDICARB ESTIMATED 0 CHECK OQSERVED A CHECK ESrIMATED Figure 20. Below ground to total biomass partitioning ratio vs. days after planting PARTITIONING RATIO I I f —1 20 40 60 80 100 120 DAYS AFTER PLANT me D Amman oesenvso + ALDICARB ESTIMATED 0 CHECK oqsenvso A CHECK ESTIMATED Figure 21. Tuber to total biomass partitioning ratio vs. days after planting 110 0.08 -— o + . + d 0.06 — f + ‘f 0.04 — + + OJ 2‘ 2 0.02 — + 9 o + a a "I” < D U a O u D U CI LIJ -U.02 -I a + E _ 0 g 0.04j o UJ & -o.os ° 5 0 ~U.08- o -o.1 — -0.12—- O - 0 ‘ 14 I I I I I | I I I I 2o 40 so on ma 120 DAYS AFTER PLANTING D BELOW GROUND + AmVE GROUND 0 TUBER Figure 22. Difference in aldicarb and check treatment partition ratios vs. days after planting. Discussion Variability in Study Findings Superior. Superior B yields were unaffected by either initial nematode population or aldicarb application. Superior A yields were significantly decreased by the higher initial nematode population and increased by aldicarb application. For Jumbo tubers, treatment was significant but initial nematode population was not. Differences in total tuber yield were statistically significant for both treatment and initial nematode population, with treatment increasing yield and initial nematode population decreasing yield. lll Russet Burbank. Increase in B tuber yields was significantly associated with high initial nematode populations. This increase in B yield is associated with significant decreases in A and J yields for both high initial nematode population and non-treated control. Total yield is decreased and tuber size class distribution shifts toward smaller sizes. Atlantic. For B, A, and Total yield, the only significant differences occurred for high vs. low initial nematode population in aldicarb treated soils. Jumbo yields were significantly higher for both aldicarb treated and low initial nematode population soils. Impact of Aldicarb on Tuber Yield Analysis for the impact of aldicarb on tuber yield was divided into average yield loss, cumulative probability distributions, and regression analysis. Average Yield Loss. There were yield decreases for all size classes of Superior potatoes with Jumbos experiencing the greatest losses. For Russet Burbank there was an increase in B yield. Yield losses between A and Total were relatively balanced while the B yield increase apparently came at the expense of Jumbo yield. Atlantic experienced a slight increase in A yield. This was associated with decreases in B, J, and Total yields. Atlantic appears to be the most resistant variety to P.penetrans infestations. 112 Cumulative Probability Distribution. While mean yield loss does give some information regarding the impacts of aldicarb on potato variety yield, a more complete picture can be gained by inspection of a cumulative probability distribution (CPD). In addition to central tendency, a CPD also provides information about the distribution of results around that tendency. As such, the cumulative probability distribution provides the clearest representation of aldicarb impact variability. It also forces the decision maker to be explicit in regard to the level of uncertainty associated with decision making. Pre—season Model. This model was determined based on decision making information a grower would have at time of planting. The variation explained by cultivar, initial nematode count, and planting date was low. Cultivar was the most significant variable in the model. Planting date was significant for delta total weight. Post-Season Model. This model was determined based on information available after.the growing season for use in retrospective estimation of potential yield losses. Addition of growing season length and total weight of aldicarb treated potatoes increased model r-square values. Stepwise. Stepwise procedures provided good results for estimation of total yield loss but failed to provide for estimation of B and Jumbo yield size classes. Size class estimates could be obtained using mean proportion values discussed previously. 113 Summapy of Regression Estimates. Of the three models tested (preseason, postseason GLM, postseason stepwise), the stepwise procedure provided the best overall results (Table 35). Table 35. Conparison of models for percentage tuber yield loss estimation. Tuber Superior Rus Burb Atlantic size Pre-Season Postseason Stepwise Stepwise Stepwise Class 31:; PR > F 31:; PR > F 333g PR > F 3:1; PR > F 31:; PR > F B 0. 58 0. 0007 0. 61 0. 0035 0. 39 0.0840 . . 0. 38 0.1412 A 0.4.......9000300640001807000003074002660..8200307 Junbo 0.73 0. 0715 Total 0. 49 0. 0008 0. 66 0. 0001 0. 86 0. 0001 0. 76 0. 0133 0. 97 0. 0088 There was no clear pattern in variables selected by this procedure (Table 36). The coefficient on planting date was more often negative than positive. This would indicate that late planting decreases tuber yield loss. The beta estimate for average initial nematode population was more often positive, indicating increased yield loss with increasing number of nematodes. The beta estimate for growing season length was inconclusive. The coefficient on total weight of treated tubers was more often positive, indicating that relative nematode impact increases with increasing yield. Table 36. Summary of beta coefficient signs for variables selected by regression procedures. Beta coefficient sign Tuber Size APO PJD GSL TTW ._§l駧___ REE 29§ NEE EQ§ .EQ _Q§ flE_ 29$ B 0 3 3 0 1 1 0 l A 2 0 3 2 1 l O l Jumbo 0 0 1 0 0 0 1 0 Total 1 3 2 2 1 2 0 2 SUM 3 6 9 4 3 4 l 4 114 Impact of Aldicarb on Temporal Variation in P.penetrans Population Although significant regressions were obtained, model predictive ability was limited. Model r-square values ranged from 0.20 to 0.86. Sample frequency, spacing, autocorrelation, and heterogeneous variance are all significant threats to the validity of these results. This analysis yielded two interesting results. The first is an apparently decreased ability to predict nematode population density, as measured by model r-square, for aldicarb treated plots. The second is substantially higher regression R- square values for nematode population density on Russet Burbank when compared to nematode population density on Superior. Decreased r—square values are probably due to non- standardized variables. The relative variation (as measured by model r-square) in nematode population was greater for aldicarb treated than check soils, although the total variation (as measured by average residuals) was less for aldicarb treated soils. Average residual values show that the absolute variation explained in population is higher for aldicarb treated soils even though the relative variation explained is lower. Increased ability to predict nematode populations for Russet Burbank is probably due to the average number of samples taken in each of the original studies. There appears to be a relationship between the average number of 115 samples taken per study and the regression r-square (Table 37). The relationship between variability added and predictive ability increase may not be favorable for studies with few Samples. Table 37. Relationship between average number of soil, root, and total nematode samples reported per study and ability to explain nematode population variation. SOIL ROOT TOTAL SAMP CHK ALD SAMP CHK ALD SAMP CHK ALD Cultivar (STY RA2 R02 (STY R02 R02 [STY R02 RAZ Russet Burbank 4.5 0.66 0.48 4.5 0.86 0.45 4.7 0.53 0.81 Superior 3.2 0.31 0.32 2.3 0.47 . 3.4 0.45 0.20 Atlantic 1.7 . . 1.0 . 1.3 . no significant regression Minimum Sample Fregpency. Two pieces of information are useful in determining the minimum number of samples to be taken during a growing season. They are the apparent third order relationship between nematode population and time, and the relationship between average number of samples taken and regression r—square values. The relationship between nematode population and time appears to be third order. If nematode sampling information is going to be used for population modeling, then logic would dictate that in addition to a pre-plant sample at least four (n+1) samples should be taken during the growing season. The apparent relationship between average number of samples and resultant regression r~square would indicate that if five in season nematodes samples were taken model r- square values would be approximately .70. 77’ Sample spacing. If samples were equally spaced then 116 time series analysis could be used for forecasting. Time series analysis requires equally spaced samples with a minimum data set of 50 points. The 50 points would not have to be in the same season but should be made at equal intervals after the planting date, and be taken during growing seasons of equal length. Nematode population density measures used in this study were not equally spaced, preventing this type of analysis. Autocorrelation. In regression analysis autocorrelation violates the independence of samples assumption. Regression analysis assumes that measures are independent. The nematode population at a given time is a function of the nematode population at previous times. There is a high degree of autocorrelation in this data set. Impact of Temporal Variation in P.penetrans on Tuber Yield Division of nematode sampling data into time classes eliminated some of the autocorrelation problems associated with sampling data, but also lowered the number of observations available for analysis. Even with the smaller data set, there were significant results with good predictability for delta A and total yield for Superior and delta total yield for Russet Burbank. These results appear to be better in terms of average r-square values across size categories to earlier regression models. Collapse of many sampling dates into four time classes caused some unaccounted for error, also the error associated with III I 77’ population modeling had a multiplicative affect on model 117 reliability. It appears that if a good model of Pppppgppppg population dynamics was available, then impact on yield estimates could be improved using this type of analysis. An increase in the number of samples taken during the growing season would increase ability to determine a window for maximum nematode impact on yield by allowing a greater number of time classes. Impact of Temporal Variation in P.penetrans on Plant Development The ability to explain differences in plant organ growth between treated and non-treated plots was limited. Differences in partitioning can be estimated and support the observation that aboveeground portions of non-treated plants die earlier in the season than those of treated plants. This decrease in above ground partitioning is associated with an increase in relative tuber partitioning as carbohydrates are moved from above ground portions of the plant to tubers for non-treated plots. In treated plots carbohydrates are partitioned into above ground plant portions until late in growing season resulting in greater total available biomass and tuber yield. There was no apparent impact of treatment on relative partitioning to below ground portions of the plant indicating that the plant does not respond to root injury by increasing partitioning to below ground portions. F___________________________________1 118 Model Parameterization The optimal model would simulate nematode population dynamics in soils and roots and then link nematode populations with physiological damage to potato plant organs or relative plant partitioning. Relative plant partitioning functions are only available for Russet Burbank and will have to be incorporated into plant growth simulation models to determine the applicability of these relationships for decision making. The next best model would track nematode population dynamics and correlate population levels in set time or environmental intervals with end of season yield change. The Impact of Temporal Variation on Yield section of this paper was an attempt to parameterize this type of model. Some significant results were obtained. Lack of a reliable nematode population model limits the applicability of this type of model. The third level of model would use summary growing season information to predict the impact of aldicarb'and P.penetrans on tuber yield. This is level of modeling available data currently supports. The stepwise regression procedure provided a good estimate of change in total tuber for each variety but not size categories. This estimate of change in tuber yield is coupled with the mean proportion of size class yield to total yield to estimate size categories. 119 Table 38. Summary of estimation procedures for impact of aldicarb on tuber vield. MEAN GLM GLM STEPWISE YIELD PRE- POST- POST- 4 OR > LOSS - SEASON SEASON SEASON SAMPLES Superior n count 18 18 18 18 9 r-square 0.49 0.65 0.86 0.81 Prob > F 0.00 0.00 0.00 0.00 Ave Residual 21.5 15.9 12.5 11.5 46.6 Min Residual 6.7 1.6 0.4 0.6 1.7 Max Residual 49.7 51.6 40.0 40.2 123.5 Russet Burbank n count 9 . 9 9 9 r—square 0.49 0.65 0.76 Prob > F 0.00 0.00 0.01 Ave Residual 32.2 40.8 30.6 15.7 Min Residual 0.4 1.4 4.9 1.8 Max Residual 75.0 93.6. 59.9 43.0 Atlantic n count 7 7 7 7 r-square 0.49 0.65 0.38 Prob > F 0.00 0.00 0.14 Ave Residual 25.0 19.0 36.1 3.6 Min Residual 14.2 6.3 12.2 0.0 Max Residual 39.7 35.4 64.5 8.5 120 Based on these results, three subroutines were developed for estimation of the impact of aldicarb and P.penetrans on tuber yield in SUBSTOR. An INCLUDE file, PPENE.INC, was used for data dictionary, variable initialization, and common blocks. IPPENE reads PPFILE.PAR and initializes program variables. PPIMPACT calculates estimated yield based on the results‘from the integrated research review previously discussed. OUTYLD writes output file containing summary yield information. PPENE.INC contains variables used in determination of aldicarb and P.penetrans impacts on yield. DATA DICTIONARY VARIABLE DEFINITION SSYLD SUBSTOR SIMULATED YIELD (KG/HA) ABYLD ESTIMATED ALDICARB B YIELD ABYLD ESTIMATED ALDICARB A YIELD AJYLD ESTIMATED ALDICARB J YIELD CTYLD ESTIMATED CHECK TOTAL YIELD CBYLD ESTIMATED CHECK B YIELD CAYLD ESTIMATED CHECK A YIELD CJYLD ESTIMATED CHECK J YIELD ALDVAL VALUE OF ALDICARB TREATED YIELD ($/AC) CHKVAL VALUE OF NON-TREATED YIELD $/AC PDD ESTIMATED DEGREE DAY BASE 10 AT ISOW INIPP INITIAL Pp/lOO cm SOIL BVAL VALUE OF B POTATOES $/CWT AVAL VALUE OF A POTATOES $/CWT JVAL VALUE OF JUMBO POTATOES s/CWT DTOT DELTA TOTAL YIELD DUE TO Pp INPP35 UNIT NUMBER FOR Pp PARAMETER FILE GSL GROWING SEASON LENGTH IN DAYS OUT36 UNIT NUMBER FOR OUTYIELD FILE JPLANT PLANTING DATE . PPFILE NAME OF INPUT FILE CONTAINING Pp PARAMETERS OUTYIELD NAME OF FILE FOR SUMMARY YIELD DATA PPFLAG INDICATES IF Pp YIELD LOSS FUNCTIONS ARE TO BE INCLUDED 1 = YES REAL INIPP,BVAL,AVAL,JVAL,SSYLD,ABYLD,AAYLD,AJYLD, +CTYLD,CBYLD,CAYLD,CJYLD,ALDVAL,CHKVAL,DTOT,PDD INTEGER INPP35,GSL,OUT36,PPVAR,JPLANT 121 CHARACTER OUTYIELD*10,PPFILE*10,PPFLAG*1 COMMON /PPENE/INIPP,BVAL,AVAL, JVAL, SSYLD,ABYLD,AAYLD, +AJYLD,CTYLD,CBYLD,CAYLD, CJYLD,ALDVAL,CHKVAL,DTOT, PDD, +INPP35,GSL,OUT36,PPVAR,JPLANT,OUTYIELD,PPFILE,PPFLAG IPPENE IPPENE is called from the MAIN program and initializes the aldicarb/P.penetrans yield impact model. IPPENE reads input parameter file number 35 named PPFILE.PAR which contains values indicating potato variety planted, initial nematode populations/100 cm3 soil, and the values of B, A, and J tubers per hundred weight. It then opens output file 36 for summary yield information and returns to the MAIN program. PPIMPACT PPIMPACT is called from program MAIN after the simulation loop has been exited. It estimates yield reductions associated without the use of aldicarb based on information obtained in the integrated research review. First the SUBSTOR yield is converted from metric tonnes per hectare to hundred weight per acre. SSYLD=SSYLD*(2.2046/247.105) A growing season length of 140 days was assumed. GSL = 140 The IF THEN ELSEIF ELSE statement is used to distinguish between varieties for which yield loss functions were developed. Values of 1,2,and 3 stand for Superior, Russet Burbank, and Atlantic respectively. DTOT indicates percentage yield loss estimated without the application of aldicarb. 122 The fractional modifiers are used to distribute total yield between size categories. IF (PPVAR.EQ.'1') THEN DTOT = -.5461647l+.00151086*INIPP+.00520185*GSL+.00047833*SSYLD ABYLD AAYLD AJYLD CTYLD CBYLD CAYLD CJYLD ALDVA CHKVAL = CBYLD*BVAL+CAYLD*AVAL+CJYLD*JVAL ELSEIF (PPVAR.EQ.'2') THEN SSYLD*.043 SSYLD*.887 SSYLD*.07 SSYLD-SSYLD*DTOT CTYLD*.O48 CTYLD*.908 CTYLD*.044 = ABYLD*BVAL+AAYLD*AVAL+AJYLD*JVAL-PSTCOST DTOT = .2199-.0020*ISOW+.0005*INIPP+.0016*GSL-.0002*SSYLD ABYLD = SSYLD*.175 AAYLD = SSYLD*.697 AJYLD = SSYLD*.055 CTYLD = SSYLD-SSYLD*DTOT CBYLD = CTYLD*.239 CAYLD = CTYLD*.661 CJYLD = CTYLD*.045 ALDVAL = ABYLD*BVAL+AAYLD*AVAL+AJYLD*JVAL-PSTCOST CHKVAL = CBYLD*BVAL+CAYLD*AVAL+CJYLD*JVAL ELSE PDD = -l934.l4+13.99*JPLANT (estimate planting degree days) ‘ DTOT = 1.20559577+ .00093791* INIPP-.00084458*PDD- .00822848*GSL ABYLD = SSYLD*.071 AAYLD = SSYLD*.813 AJYLD = SSYLD*.116 CTYLD = SSYLD-SSYLD*DTOT CBYLD = CTYLD*.O7O CAYLD = CTYLD*.867 CJYLD = CTYLD*.063 ALDVAL ABYLD*BVAL+AAYLD*AVAL+AJYLD*JVAL-PSTCOST CHKVAL = CBYLD*BVAL+CAYLD*AVAL+CJYLD*JVAL Once yield estimations are complete OUTYLD and TRTSUM output subroutines are called. OUTYLD OUTYLD is called by PPIMPACT and writes yield summary information to output file 36. Total yield with and without aldicarb application is reported along with size category ~ 123 yields. The program returns to PPIMPACT. IBEfiHM TRTSUM is called from PPIMPACT at the end of the simulation. TRTSUM writes a summary of treatment information to output file 39. Output variables include cumulative rainfall, irrigation, number of irrigation events, total drainage out of soil profile, nitrate and pesticide leached, as well as yield information and treatment value. Summary The documentation of model development procedures, development of yield loss functions, and an expansion of SUBSTOR to include the value of aldicarb to potato 'production were the results of work completed under Thesis Objective 5. The estimated value term may be used as a measure of the benefit of aldicarb application to growers. Aldicarb application value was intended to be compared to application risk values determined under Thesis Objective 3. Work completed under Thesis Objective 4 also showed limitations in the current information base. Improvements were suggested. CHAPTER VI NITRATE AND ALDICARB LEACHING EXPERIMENT To obtain nitrate and aldicarb leaching information, two non-weighing lysimeters were installed at the Montcalm Potato Research Farm in 1986. In 1988, research was conducted using these lysimeters to test the impact of alternate management practices on nitrogen and aldicarb leaching. Results of this field experiment were used for comparison with simulation modeling results. Materials and Methods Two non—weighing lysimeters were installed in June of 1986 at the Michigan State University Potato Research Farm. The lysimeters are 48 inches wide, 68 inches long and 72 inches deep. They were constructed of welded 3/16 inch sheet metal. An epoxy material was sprayed on all lysimeter surfaces for rust reduction. Access chambers were located at the long end of the lysimeters to facilitate sample collection (Figure 23). Installation To determine the preinstallation soil profile, a six- foot deep soil core was taken. The soil was predominately a McBride sandy loam. Soil layers were removed separately Using a back hoe, and placed on plastic tarps to prevent mixing. Sufficient soil was removed so that the lysimeter tops were buried 12 inches below the soil surface. This burial 124 125 Lgsimeter filled with soil Manhole Access Sli ht d d d . ' . forgdralggg: nee e ' 2 liter graduated ' cglinder inside [7.3 liter pail Figure 23. Isometric projection Of non-weighing lysimeter. 126 depth allowed farm implement use. Once the lysimeters were placed into the soil and leveled, two inches of P stone were placed on the slanted metal drainage floors. A layer of drain tile cloth was placed over the stone to improve infiltration. After being sieved through a one half inch screen, the removed soil was placed directly on the drain tile cloth. Stones larger than one half inch were removed. The soil was periodically packed while the lysimeters were being filled. The average depth of each soil layer was calculated and the soil was replaced accordingly. After the lysimeters were filled and covered, four inches of water were applied to ensure settling in and around the lysimeters. ' Treatments The treatments in the nitrate/aldicarb leaching experiment included the impact of at planting fertilizer, in-season fertilizer, and irrigation management. Sampling for aldicarb and nitrates began January 18, 1988. Planting Russet Burbank potatoes were planted on May 11, 1988. Seed pieces were placed four inches deep and 12 inches apart using 34 inch row spacing. Aldicarb was applied as TEMIK 15G in the furrow at 20.0 lbs./acre. At planting, fertilizer was applied to both standard and conservation treatments (Table 39). 127 Table 39. Nitrate/aldicarb leaching experiment at-plant fertilizer treatments. Treatment N lb/acre P lbéacre K lbzacre Standard . 75 50 75 Conservation 28 56 84 Fertilizer All nitrogen was applied as urea. The standard management strategy received 75 lbs./acre nitrogen at planting followed by 69 lbs./acre 54 days after planting, and 55 lbs./acre 77 days after planting. The conservation management strategy received 28 lbs. nitrogen at planting followed by 54 lbs./acre 63 days after planting, and 31 lbs./acre 77 days after planting (Table 40). Table 40. Nitrate/aldicarb leaching experiment nitrogen fertilizer treatments (lbs.[acre). Standard Conservation Date Rate Cumulative Date Rate Cumulative 5-11 75 75 5-11 28 28 7-04 69 144 . . 28 . . 144 7-13 54 82 7-27 55 199 7-27 31 113 Irrigation If natural rainfall was insufficient to meet plant needs irrigation was applied using overlapping solid set sprinklers. The intent was to apply one inch per application to the standard treatment every three to five days and one half inch per application every two to three days to the conservation treatment. Actual irrigation rates were limited by the volume of available water (Table 41). The many irrigation volumes was less than intended. '128 Table 41. Nitrate/aldicarb leaching experiment irrigation treatments in (inches). Standard Conservation Date Rate Cumulative Rate Cumulative 06/14 0.3 0.3 0.3 0.3 06/20 0.9 1.2 . 0.3 06/26 1.0 2.2 1.0 1.3 06/29 . 2.2 0.5 1.8 07/03 0.5 2.7 0.5 2.3 o7/o4 . 2.7 0.5 2.8 07/07 1.1 3.8 0.8 3.6 07/12 . 3.8 0.5 4.1 07/15 0.3 4.1 0.3 4.4 07/20 1.4 5.5 . 4.4 07/21 . 5.5 0.5 4.9 07/26 0.7 6.2 0.4 5.3 07/28 0.1 6.3 . 5.3 08/01 1.2 7.5 . 5.3 08/04 . 7.5 0.2 5.5 08/11 0.7 8.2 . 5.5 Average 0.7 0.5 Leachate Analysis Leachate samples were collected and analyzed for concentrations of nitrate nitrogen and aldicarb. Sampling Leachate was collected in the lysimeter access hole using a nested collection device. A two liter graduated cylinder was placed inside a 17.3 liter pail inside a 125 liter plastic tub. This system was used so that the relative accuracy of measure would be consistent with the volume of the sample. Leachate collected was thoroughly mixed and two 100 cm9 polypropylene samples bottles were filled. Samples were kept cool and out of sunlight during transportation to the MSU campus where they were stored in a freezer at -15° C. Leachate volume was recorded. Leachate not used for analysis was taken Off—site for disposal. 129 Nitrate Nitrate analysis was performed using a Lachat flow injection analyzer QuikChem Method No. 12-107-04—1 A (Lachat, 1988). A 10 ppm reference standard was used for calibration. The minimum detection level was 0.01 ppm. The standard error of this procedure at 10 ppm was two percent. Aldicarb The frozen aldicarb samples were sealed in styrofoam coolers and sent by over-night mail to Rhone-Poulenc Ag Company. The samples were analyzed using high performance liquid chromatography. This method allows for the determination of carbamate residues (Aldicarb, Aldicarb- sulfoxide, and Aldicarb-sulfone) at one part-per-billion with a relative standard deviation of 10% at five parts—per— billion (Hudson, 1988). w Leachate samples were collected on 17 dates (Table 42). Analysis of leachate samples indicated that no aldicarb, aldicarb sulfoxide, or aldicarb sulfone leached out of the soil profile during the growing season. Aldicarb Degradation and Movement Simulations of aldicarb degradation in soil indicated that aldicarb is rapidly converted to aldicarb sulfoxide which is then slowly converted to aldicarb sulfone and non- toxic hydrolysis products (Figure 24). TIT—'1 130 Table 42. Nitrate/aldicarb leaching experiment sampling results for leachate volume, nitrate concentration, and aldicarb metabolite concentrations. Conservation Concentration Volume ppm ppb Date liters N03-ALQ ASQ ASE 18-Jan 26.9 35.1 nd nd nd 23-Feb 110.4 3.2 nd nd nd 25-Mar 31.6 2.5 nd nd nd 19-Apr 31.6 0.8 nd nd nd 08-Jun 8.6 9.8 nd nd nd 23-Jun 4.8 33.9 nd nd nd 03-Ju1 2.6 35.9 nd nd nd 12—Jul 0.9 30.4 nd nd nd 20-Jul 0.1 . . . . 26-Jul 0.1 . nd nd nd 3l-Jul 0.1 . . . . 11-Aug 0.1 . . . . 24-Aug 0.1 60.2 . . . 31-Aug 0.0 .. . . . 21-Sep 3.6 65.3 nd nd nd 28-Sep 41.8 100.4 nd nd nd 06-Oct 43.0 59.7 nd nd nd nd - not detected (<1ppb) * - Cracked sample bottle Standard Concentration Volume ppm liters 22.1 c m-bOHNHOHOl-‘Nmm owmowqpbwqomq but-N ppb HQ§:_ ALB A§Q A§H 42.5 nd nd nd 3.8 nd nd nd 22.1 nd nd nd 29.2 nd nd nd 34.3 nd nd nd 39.9 nd nd nd 39.8, nd nd nd 38.2 nd nd nd 105.4 nd nd nd 87.1 nd nd nd 81.2 nd nd nd 105.0 nd nd nd 60.7 * * * 120.2 nd nd nd 78.8 nd nd nd - Insufficient volume 2.6 — MASS (LBS/ACRE) l m l 0 7HVH$HVHV 120 140 160 180 D ALD + 200 220 DAY OF YEAR 50 0 Figure 24. Simulated mass of aldicarb (ALD), aldicarb-sulfoxide (ASO), aldicarb-sulfone (ASN) total mass degraded (DEG) vs. day of year. 131 Because of this rapid degradation the movement of aldicarb within the soil profile is best shown as percentage of total toxic residue (TTR) remaining in each soil layer. For simulations representing the lysimeter experiment aldicarb was applied into layer 1 of the 6 layer representation of McBride soil. By mid season there was little difference in the distributions of aldicarb in the soil of conservation and standard treatments (Figure 25). By the end of the simulated season the distribution of TTR in the soil had changed with a greater percentage of the TTR at leachable depths in lower layers of the standard management practice (Figure 26). PERCENT OF TTR IN LAYER f TN A I MBA—q, . 51 79 son. LAYER DEPTH (INCHES) [ZZ CONSERVATION [:3 STANDARD Figure 25. Mid-season (July 13) distribution of TTR in the soil profile for simulated conservation and standard treatments. 132 \\‘1 //A \ N // /A / \\ \ 24 33 g]. ”//// 6 1 u 51 79 SOIL LAYER DEPTH C INCFES) ZZ CONSERVAT ION [:3 sum Figure 26. End of season (Sept 23) distribution of TTR in soil layer for simulated conservation and standard treatments. simulated vs. Observed Results Yield results for simulated and observed compared favorably. Simulation analysis of leaching parameters was ended on the sixth of October. Simulation results indicated a greater mass of nitrate leached than observed. A small amount of TTR leached out of both the conservation and standard treatment soil profile. Ability to compare aldicarb results is limited. An estimate of simulation accuracy using the analysis minimum detection limit and amount of water leached out of the lowest layer of the soil profile indicated an over estimation of aldicarb mass leached (Table 43). 133 Table 43. Comparison of per acre simulated and observed results for risk-benefit parameters. Conservation Standard Parameter Simulated Observed Simulated Observed Yield cwt 237.0 241.5 253.8 244.8 NO3- lbs 40.84 0.26 56.54 1.25 TTR lbs 0.0162 — 0.4 -< 3 D 0.3 —. g 1 U 0.2 — °~1 'l/ o T I l l l T ‘17 I 110 150 175 155 194 205 224 237 244 255 DAY OF YEAR o museum/mow + STANDARD Figure 27. Nitrate/aldicarb leaching experiment observed nitrate mass leached vs. day of year. 50 so — G‘ E < 40 — \ II) In .1 U D g 30 - l) < M] J W 3 20 —. ’: Z 10 a o .. I. I“ I I 77 l T f f l T l I T 120 140 150 180 200 220 240 250 DAY OF YEAR Cl CONSERVATION + STANDARD Figure 28. Nitrate/aldicarb leaching experiment simulated nitrate mass leached vs. day of year. 136 Part of the difference in total mass leached between simulated and observed may be found in the relationship between cumulative drainage volume and cumulative nitrate mass leached. Simulated results indicated little difference between management strategies (Figure 29). Observed results indicated greater masses leached per unit of drainage in the standard management strategy than in the conservation management strategy (Figure 30). Mixing in the soil profile may be mitigating changes in nitrate concentration at lower soil layers. CMATIVE NITRATE MASS [LBS/ACRE) - o 'T—rfi i I fi' I I r I I I I fir f If 0.000-300,47D.SED-590.61O.ESO.5704700.740.750.7613.B40.890.91U.910.991.1S1.20 mum (INCHES) . sumac . CONSEHVATION Figure 29. Nitrate/aldicarb leaching experiment observed cumulative nitrate mass leached vs. cumulative drainage. 137 N I IRATE LEACHED (LBS/ ACRE) u D l I I I I I 0 2 4 6 DRAINAGE (INCHES) I CONSERVATION . sump.) Figure 30. Nitrate/aldicarb leaching experiment simulated cumulative nitrate mass leached vs. cumulative drainage. The relationship between simulated aldicarb total toxic residue mass leached and day of year was similar to that of simulated nitrate mass leached (Figure 31). Leaching events occurred on day 170 and day 230. No aldicarb residues were observed in lysimeter leachate. 138 0.022 0.02 ~ 0.018 — 0.015 — 0.014 — 0.012 — 0.01 — 0.005 — TTR LEACHED (LBS/ACRE) 0,006 — 0.004 — 0.002 M ALLLLLL. I W I H I H w H I F I I I I 130 150 170 190 210 230 250 270 DAY OF YEAR D CONSERVATION + STANDARD Figure 31. Nitrate/aldicarb leaching experiment simulated total toxic residue mass leached vs. day of year. The percentage differences between estimated and observed values show the relative accuracy of simulation results (Table 46). Percentage differences for nitrate mass leached did not compare favorably. Accuracy Of percentage difference for total toxic metabolite fell between values for yield and nitrate mass leached. Table 46. Comparison of simulated and observed percentage decrease in yield, nitrated leaching, and aldicarb leaching parameters associated with a switch to the conservation strategy. Percentage Decrease Parameter Simulated Observed Yield 1 Nitrate 28 791 TTR 26 15 1 - estimated based on minimum detection limit and leachate volume 139 Implications of Lvsimeter Exneriment Results Comparison of observed versus simulated results for the lysimeter experiment indicated that the risk-benefit analysis simulation results were fairly accurate for McBride soil in 1988. This increases the credibility of the risk- benefit analysis results but can not be considered a full model validation. Estimations of the impact of management practices on nitrate are questionable. For the lysimeter experiment, the model over estimated the mass of nitrate moving out of the soil profile. The nitrate leached overestimation assumption is further backed by the magnitude of leaching losses in comparison to the amounts applied. The reliability of the aldicarb model is uncertain. Simulated aldicarb mass leached was greater than observed, although the relative impact of management strategies on mass leached was small. If aldicarb mass leached is overestimated it may be explained by unaccounted volatilized mass. CHAPTER VII SIMULATION MODELING FOR REGIONAL RISK-BENEFIT ANALYSIS Work completed under Thesis Objective 5 represents the application of results from Thesis Objectives 1-4. Revised SUBSTOR was used to estimate impacts of selected potato production management factors, including irrigation scheduling, nitrogen fertilizer application, and aldicarb application on potato yield, nitrate leaching and aldicarb leaching. This was done under each set of environmental conditions, soil type and land use identified in the prototype study area. Crop yield, nitrogen costs and aldicarb costs were used to determine management strategy profitability. Mass of nitrate leached below the soil profile was used as a measurement of ground water nitrate contamination risk for each management strategy. Mass of aldicarb leached below the soil profile was used as a measurement of ground water aldicarb contamination risk for each management strategy. Materials and Methods Results from study Thesis Objective 1 indicated that potato production occurred on five soil types between 1986 and 1988. For each soil type and year combination, the impacts of alternate management strategies were compared using the revised version of SUBSTOR. SUBSTOR modifications were described in Chapter IV and Chapter V. 141 SUBSTOR Revised SUBSTOR was used for risk-benefit analysis simulations. The model was used to estimate the impacts of irrigation and nitrogen application on nitrate movement and aldicarb movement through simulation of soil water movement, plant growth and development, and plant water uptake. The impact of aldicarb on plant development was not a part of the revised version of SUBSTOR used. The impact of aldicarb on potato tubers was estimated at the end of each simulation. This was done using the model developed in Chapter V. The impact of delayed aldicarb application on tuber yield was not estimated (Table 47). Table 47. Sensitivity of revised SUBSTOR in relation to production system variables. Output Input timing and application rate Parameter Irrigation Nitrogen Aldicarb Water movement Yes Yes No Nitrate Movement Yes Yes NO Aldicarb Movement Yes Yes Yes Plant Growth Yes Yes No Tuber Yield Yes Yes Yes1 1 For at-plant application Of aldicarb only Forty SUBSTOR simulations were performed. Twenty of these simulations were used for comparative risk—benefit analysis. These twenty simulations represented the ten sets of environmental conditions (soils and weather) determined under Thesis Objective 1 and the two management strategies (standard and experimental) determined under Thesis Objective 2. Aldicarb was applied at planting in both standard and experimental management strategies. Twenty 142 simulations represented the standard and conservation management strategies with aldicarb applied at plant emergence. These twenty simulations were used to show the impact of delayed application on leaching, but could not be used for risk-benefit analysis because impact on yield was undetermined. Estimated yields, nitrate mass leached, and aldicarb mass leached were recorded. Aldicarb mass leached was reported as the sum of aldicarb plus aldicarb-sulfoxide plus aldicarb-sulfone (total toxic residue (TTR)). Weather. Weather information was obtained from a Licor 2000 data logger and CR-21 weather station. Each day, readings were taken for maximum and minimum temperature, precipitation, and solar radiation. This data was recorded and used for SUBSTOR modeling. Soils. Soil profile properties were defined for five soils using Soil Conservation Service 232 forms. These forms describe Chemical and physical characteristics of soils. If not available in the SCS 232 form values were estimated using procedures defined in CERES-Maize model documentation. Management Strategies Data files representing irrigation management strategies were developed for 1986 - 1988. Standard treatments involved one inch of irrigation water applied every three to five days, whereas conservation treatments had one half inch applied every two to three days. The nitrogen fertilizer applications determined under 143 Thesis Objective 2 were used for the risk-benefit analysis system. Aldicarb application was simulated by modification of the ALDIC.PAR input file described in Chapter IV. For the risk—benefit analysis system, aldicarb was applied at planting for standard and conservation management strategies. Additional simulations were'performed showing the impact of delayed aldicarb application on mass leached. Benefit. Management system benefit was estimated using the price of nitrogen fertilizer, the cost of aldicarb, and the market value of tubers (Table 48). Table 48. Market prices used in management strategy benefit analvsis. Year Nitro en lb.1 Aldicarb $13.0 lb.ai.2 Tubers CWT3 1986 0.18 47.00 8.00 1987 0.15 47.00 4.50 1988 0.19 ’ 47.00 8.40 1 Mason Elevator 2 Grower Services 3 Spud Pack Economic value was estimated for the standard and conservation management strategies using at—plant aldicarb application. Management strategy value was calculated by multiplying the marketable tuber yield by market value and then subtracting nitrogen and aldicarb costs. Risk. Risk to ground water from agricultural non-point source contamination was estimated by simulation of the mass of nitrate nitrogen, aldicarb, and aldicarb metabolites leaching out Of the soil profile. Nitrate risk measures were estimated for both standard and conservation (at-plant aldicarb) management schemes. Aldicarb risk measures were estimated for standard and conservation management with at- 144 plant and at-emergence aldicarb application. B§§EI£§ During the three years for which simulation analysis was performed, there was large variation in growing season precipitation. The fall of 1986 was inordinately wet. The 1987 growing season was moderate, with a fairly even rainfall distribution. The 1988 growing season was dry. This variation in precipitation distribution was fortunate, impact of alternate management practices could be estimated for a wet, a normal, and a dry growing season (Figure 32). Irrigation volume for conservation and standard management strategies changed for each year of the study with the conservation management strategy being more sensitive to rainfall distribution than the standard management strategy (Table 49). CLWLATIVE RAINFALL ( lNCI-ES) 0 :4 = ' ‘ : ' t I I I I I ‘I’ I lil ‘r I‘F I I I I I I ‘r f I 120 14B 150 180 200 220 240 250 290 300 320 310 DAY OF YEAR 0 1988 + 1387 o 1985 Figure 32. Cumulative rainfall during 1986, 1987, and 1988. I -' 145 Table 49. Cumulative water applied in standard and conservation management strategies. Irrigation (inches) Year Rain Standard Conservation 1986 34.8 13.0 9.5 1987 17.4 16.0 11.0 1988 11.4 17.0 13.5 Simulation Simulation results indicated that weather, soil type, and alternate management practices impacted water, nitrate, aldicarb leaching, and the yield Of marketable tubers (Table 50) 0 Table 50. 1986-1988 Simulation results for five soil types and two management strategies. Tparhed Market tuber (in.) (lbs./acre) (CWT/acre) Year Soil Type Managl‘Water NO3- Aldicarb Aldicarb Check 1986 Epoufette Cons 28 237 1986 Epoufette Stan 32 292 1986 Grayling Cons 26 203 0.4179 305 252 1986 Grayling Stan 31 240 05495 . 354 297 1986 Mancelona Cons 27 151 0.7584 220 178 1986 Mancelona Stan 31 194 1.0126 264 216 1986 McBride Cons 27 92 0.5069 339 283 1986 McBride Stan 31 131 0.6603 366 309 1986 Montcalm Cons 27 165 0.2952 315 261 1986 Montcalm Stan 31 204 0.4000 340 284 1987 Grayling Cons 8 153 0.1025 415 354 1987 Grayling Stan 12 161 0.2508 545 484 1987 Mancelona Cons 9 82 0.3322 389 330 1987 Mancelona Stan 13 112 0.5364 451 390 1987 McBride Cons 8 33 0.1618 477 415 1987 McBride Stan 12 47 0.3493 571 510 1987 Montcalm Cons 8 78 0.0484 440 379 1987 Montcalm Stan 12 112 0.1311 506 444 1988 McBride Cons 8 26 0.0869 470 408 1988 McBride Stan 11 27 0.1391 571 511 1 Management strategy Stan —Standard Cons -Conservation 146 Estimated grower profitability was affected by year, soil type, and management practices (Table 51). The average value of aldicarb to growers was $307 for the conservation management strategy and $327 for the standard management strategy. Table 51. Estimated management strategy profitability in dollars by year. soil type. and management practice. Conservation Standard Year Soil Type Aldicarb No Aldicarb Aldicarb No Aldicarb 1986 Epoufette 2038 1701 2045 1700 1986 Grayling 2366 1989 2749 2340 1986 Mancelona 1686 1397 2029 1692 1986 McBride 2638 2237 2845 2436 1986 Montcalm 2446 2061 2637 2236 1987 Grayling 1798 1571 2376 2148 1987 Mancelona 1681 1463 1953 1725 1987 McBride 2077 1845 2493 2265 1987 Montcalm 1911 1683 2200 1968 1988 McBride 3873 3399 4711 4254 For the conservation management strategy, simulation results indicated that in 1986 for all soils except McBride more nitrogen was leached out of the soil profile than was applied as fertilizer. In the standard management strategy with exceptions of Mancelona and McBride more nitrate nitrogen was leached out of the profile (Table 52). This would indicate that leaching caused by heavy fall rains extracted residual soil nitrogen from the soil profile. With the exception of Grayling sand, percentage nitrate leaching in 1987 and 1988 was closer to the expected 17 to 54 percent loss predicted by Hubbard (1984) and Hallberg (1986). 147 Table 52. Summary of nitrate mass leached and percentage of mass applied (%AP) by year, soil tvpe. and management strategy. Conservation Standard Year Soil Type Mass 352 Mass 3A2 1986 Epoufette 237 158 292 146 1986 Grayling 203 135 240 120 1986 Mancelona 151 101 194 97 1986 McBride 92 61 131 66 1986 Montcalm 165 110 204 102 1987 Grayling 153 102 161 81 1987 Mancelona 82 55 112 56 1987 McBride 33 22 47 24 1987 Montcalm 78 52 112 56 1988 McBride 26 17 27 14 Average 122 81 152 76 Estimated aldicarb mass movement out of the soil profile was also impacted by management practices (Table 53). Aldicarb mass leached decreased with the use of conservation irrigation management practices, but increased with emergence application. The increase in mass leached associated with emergence application of aldicarb may be due to greater masses in the soil at the end of the season which are susceptible to movement with heavy fall rains. 148 Table 53. Summary of aldicarb TTR mass leached and percentage of mass applied (%AP) by aldicarb application timingl management practice, soil type, and year. At Plant At Emergence Research Standard Research Standard Mass fig: Mass £53 Mass 3A2 Mass 352 1986 Epoufette 0.4510 15 0.6096 20 0.4848 16 0.6575 22 1986 Grayling 0.4179 14 0.5495 18 0.4630 15 0.6037 20 1986 Mancelona 0.7584 25 1.0126 34 0.6869 23 0.9656 32 1986 McBride 0.5069 17 0.6603 22 0.6869 23 0.6928 23 1986 Montcalm 0.2952 10 0.4000 13 0.3273 11 0.4470 15 1987 Grayling 0.1025 3 0.2508 8 0.1309 4 0.3334 11 1987 Mancelona 0.3322 11 0.5364 18 0.4353 15 0.7025 23 1987 McBride 0.1618 5 0.3493 12 0.2093 7 0.4563 15 1987 Montcalm 0.0484 2 0.1311 4 0.0515 2 0.1701 6 1988 McBride 0.0869 3 0.1391 5 0.1201 4 0.1924 6 AVG 0.3161 11 0.4639 15 0.3596 12 0.5221 17 Risk-Benefit Analysis Of the forty simulations performed for risk-benefit analysis twenty may be directly compared. Comparable management strategies for risk-benefit analysis were conservation and standard irrigation and nitrogen strategies with aldicarb applied at planting to both management strategies. The use of conservation management practices decreased the profitability of potato production but also decreased risk to ground water contamination (Table 54). The impact of alternate management practices differed for each year of the study. In 1986, yield loss associated with the conservation strategy was the least while the percentage decrease in nitrate and aldicarb mass leached was the greatest. In 1988, yield loss associated with the conservation management strategies was the greatest while 149 impact on nitrate and aldicarb mass leaching was the least. Table 54. Decrease in profit and leaching measures (standard - conservation) associated with a switch to the conservation management strategy. Leached Profit (in.) (lb.[acre) Year Soil Type (fizacre) E39 N03- TTR 1986 Epoufette 7 4 55 0.1586 1986 Grayling 383 5 37 0.1316 1986 Mancelona 343 4 43 0.2542 1986 McBride 207 4 39 0.1534 1986 Montcalm 191 4 39 0.1048 1987 Grayling 578 4 8 0.1483 1987 Mancelona 272 4 30 0.2042 1987 McBride 416 4 14 0.1875 1987 Montcalm 290 4 34 0.0827 1988 McBride 839 3 1 0.0522 Average 352 4 30 0.1477 These management practice impact values may also be expressed as a percent decrease associated with a switch to the conservation strategy (Table 55). Table 55. Percentage decrease in profit and leaching measures (1.0-conservation/standard) associated with a switch to the conservation management strategv. Profit Reached Decrease (in.) (lb.zacre) Year Soil Type (Sgacre) H20 NO3- TTR 1986 Epoufette <1 13 19 26 1986 Grayling 14 16 15 24 1986 Mancelona 17 13 22 25 1986 McBride 7 13 30 23 1986 Montcalm 7 13 19 26 1987 Grayling 24 33 5 59 1987 Mancelona 14 31 27 38 1987 McBride 17 33 3O 54 1987 Montcalm 13 33 30 63 1988 McBride 18 27 4 38 AVG 13 23 20 38 The impact of a regional shift to the conservation management strategy was displayed by integrating changes in profitability, nitrate mass and aldicarb mass over the prototype study area (Table 56). If the total mass of nitrate leached was transported into water supplies, it would be sufficient to raise the nitrate concentration of 210,567 gallons of pure water to the health advisory level. If the total mass of aldicarb leached was transported into water supplies, it would be sufficient to raise the aldicarb concentration of 52,475,563 gallons of pure water to the health advisory level. Further degradation of both compounds would be expected in the unsaturated zone below the soil profile, so actual risk to ground water would be decreased. Table 56. Total decrease in profits, nitrate mass leached, and aldicarb mass leached associated with a switch to conservation management strategy in the prototype studv area. ($) (1bs.) leg; Soil Type Acres Profit NO3- TTR 1986 Epoufette 3 21 165 0.4758 1986 Grayling 9 3447 333 1.1844 1986 Mancelona 24 8232 1032 6.1008 1986 McBride 100 20700 3900 15.3400 1986 Montcalm 8 1528 312 0.8384 1987 Grayling 2 1156 16 0.2966 1987 Mancelona 8 2176 240 1.6336 1987 McBride 57 23712 798 10.6875 1987 Montcalm 30 8700 1020 2.4810 1988 McBride 91 76349 91 4.7502 TOTAL 332 146021 7907 43.7883 151 Summapy The conservation management practice was associated with decreases in yields, nitrate mass leached, and aldicarb mass leached. Differences in yield and masses leached were inversely proportional and seem to be a function of weather characteristics. Large yield differences and small leaching differences were associated with 1988, a dry season. Smaller yield differences and larger leaching differences were associated with 1986, a wet year. The conservation management strategies resulted in average profitability decrease Of $352/acre, an average nitrate reduction of 30 lbs/acre, and an average aldicarb reduction of 0.1477 lbs./acre. CHAPTER VIII DISCUSSION AND RECOMMENDATIONS Discussion and recommendations are organized based on thesis objectives. Regional Data Base The digitization portion of the geocoding process was completed using complete polygon techniques which require the majority of boundaries to be digitized twice. Operator time would be reduced if arc-node digitization procedures were used. Arc-node digitization also provides a cleaner output file for display. GIS analysis proved valuable in determining spatial variation in factors important to regional potato production. ERDAS Matrix operations using soil type,, land use, and weather could be expanded to include land ownership (Platt map). A mail survey of growers could then be used to more accurately parameterize simulations representing grower practices. W The conservation management strategies used in this study was only one of many possible alternatives. Lack of information on yield impact of emergence applied aldicarb limited the scope of risk-benefit analysis. Application of one-half inch per application every two to three days may be insufficient to meet plant needs. 152 153 Future irrigation management practices should be more carefully defined and linked to soil water content. Precipitation forecasts could be used in an expected precipitation value format (probability of rainfall multiplied by expected precipitation) to optimize soil water relationships. Current SUBSTOR routines for irrigation at soil water threshold fill the whole soil profile and should be modified to fill the soil profile to the irrigation management depth. Future conservation should include an integrated research review addressing the impact of management strategies on crop yield and compound movement. Information obtained could then be used in a quadratic programing format for crop management optimization including profitability and ground water risk. Aldicarb Movement and Degradation The aldicarb movement and degradation model was limited by lack of information on systemic uptake of aldicarb and great variability in degradation rate estimates. One year of lysimeter data is insufficient for proper validation of model functions. An increase in precipitation and leaching in 1988 would have improved the reliability of simulated versus observed comparisons. Mass values of zero are difficult to compare. A major problem faced in integration of aldicarb movement and degradation function with existing models was the level of existing program documentation. There are 154 three main categories of documentation: source code, documentation of code operation, and documentation of information and processes used in code development. With standard FORTRAN code it is very difficult to understand the implications of parameter modification. Documentation of code operation allows for understanding of how parameter modifications impact simulation results but does little to link simulation modeling to the processes being simulated. If information and processes used in code development are documented, then simulation modeling becomes a condensation of the current level of system interaction understanding. The impact of variable modification is known as well as the source of coefficients which mOdify the variables. When combined with a hierarchal modeling structure, integrated research review and meta-analysis techniques may be successfully used for third level documentation. Addition research is needed showing the fate of aldicarb at the soil surface. Does aldicarb evaporate in solution with soil water, precipitate at the soil surface, or volatilize and leave the application site in a gaseous state? Studies dealing with systemic uptake of aldicarb focused on the concentration remaining in tubers at harvest. Additional information is needed on the mass uptake of aldicarb by the potato plant during the growing season. Is aldicarb or its metabolites taken up in proportion to concentration in translocated soil waters, excluded, or 155 preferentially absorbed? Research review results showed large variation in estimates for decay rates. Decay rate is very important in estimation of risk to ground waters. Three sites Of study are needed: the biologically active root zone, the unsaturated zone, and in ground water. Growth chamber experiments should be conducted for ranges of microbial population density, solution pH, temperature, organic matter content, and soil texture. In addition to statistical hypothesis testing, probability distributions should be developed to emphasize the uncertainty associated with degradation rate estimation. A data base should be developed using integrative research review methods containing the results of field experiments showing the impact of management practices on potato plant growth and development, soil water balance, nitrate movement, and aldicarb movement. This data base should represent a range of management practices and site locations. For each field experiment in the data base SUBSTOR simulations should be performed. Independent management variables should then be analyzed in conjunction with output dependant variables using residual analysis. Residual values (simulated - observed) should be graphed versus each independent variable. This procedure can be used to provide a quantified measure of simulation accuracy, and to show which model functions are the least accurate. 156 AldicarbZRoot—lesion Nematode Impact Integrated research review and meta-analysis techniques where highly valuable for determination of the yield impact model. A reliable nematode population model could not be developed because an insufficient number of samples taken per season were available and violation of statistical assumptions. Aldicarb was applied at planting in all studies used in the integrated research review. This limited ability to estimate the impact of at plant emergence aldicarb application. The impact of aldicarb on plant partitioning was estimated for Russet Burbank potatoes. Partitioning ratios showed a decrease in partitioning to above ground portions of the plant without aldicarb. Aldicarb may be stimulating growth or suppressing Potato Early Die disease. Partitioning information was not integrated into SUBSTOR because it was available for only one cultivar and would have required a substantial change in program operation. Additional data needs to be collected for the determination of the impact of aldicarb and P.penetrans on crop production. Five or more evenly spaced nematode population density samples are needed to represent nematode population dynamics. Daily air and soil temperature measurements should be taken. Leaf, stem, root, and tuber biomass should be recorded for each sampling date. Collection of this data should allow for quantitative estimation of in-season nematode population impacts on plant 157' growth, and tuber yield. Growth chamber experiments should be conducted to provide information for distributed delay modeling of P.penetrans populations. Growth chamber temperatures should be regulated to represent mean and mean plus and minus one standard deviation of diurnal temperature. Population density should be recorded by temperature treatment, life stage, and location (soil or root). The impact Of delayed aldicarb application on plant growth and tuber yield is currently unknown. Multiple year field experiments should be performed testing the impact of no aldicarb application, aldicarb applied at planting, and aldicarb applied at plant emergence on tuber yield for several cultivars. Risk-Benefit analysis Risk benefit analysis system results are subject to a great deal of uncertainty. Of the five soil types on which potatoes were produced during the three years Of the study, data for risk-benefit model validation were available for one soil type during one year. This problem was further compounded by the fact that 1988 was a dry year and little leaching occurred. Simulated and observed yield estimates did not compare favorably. simulated yield indicated much greater losses associated with the conservation management strategies than lysimeter experiment results. The over—all methodology for regional risk—benefit analysis worked quite well. GIS analysis provided the 158 necessary information on soil type and land use for simulation modeling of risks and benefits in heterogenous environments. If simulation modeling results were validated, then the integration Of geographic information systems with simulation modeling could prove to be a very useful tool for estimation of the regional impact of alternate management practices on associated risks and benefits. Cost and time requirements prohibit the use Of non- weighing lysimeters for validation on multiple soil types. Non-weighing lysimeters can not be used to determine concentrations within the soil profile. Procedures have been developed using tensiometers and suction lysimeters for quantitative pesticide analysis. In 1986, Sandra C. Cooper published procedures for the design and installation Of a monitoring network for measuring the movement Of aldicarb and its residues in the unsaturated and saturated zones. These procedures were developed under the auspices Of the U.S. Geologic Survey, Water Resource Division and should be considered in the design of validation experiments. §2EE§£Y The risk-benefit analysis system developed in this thesis can be used to analyze the impact of irrigation, nitrogen, and aldicarb management practices but can not be used to estimated the impact of crop rotation or other pest mangement practices. Additional research is needed to determine Optimal potato produciton managment practices. 159 This risk-benefit analysis system uses mass of nitrate and aldicarb leached out of the lowest layer of the soil profile as a measure of risk. This measure of risk could be improved by expansion of model abilities to include movement Of nitrate and aldicarb through the remainder of the unsaturated zone, into ground water supplies, and into drinking water. Then the presence of nitrate and aldicarb in drinking water must be linked to its impact on health and environmental quality. Another question that needs to be addressed is that of private cost vs. public benefit. The cost of switching to a management strategy which decreases grower profitability is Of private concern. Contamination of ground water is a public concern. Who should absorb thecost of protecting ground water quality? Are growers responsible for protecting water or should the public contribute in mitigating profitability decreases. 160 LITERATURE CITED Acosta, N. and R.B. Malek. 1979. Influence of Temperature on Population Development of 8 Species Pratylenchus on Soybean. J. Nematol. 11(3):229-232. Awad, M T. 1984. Movement of Aldicarb in Different Soil Types. Bulletin Environmental Contamination Toxicology. 32:377—382. Back, R.C., R.R. Romine, and J.L. Hansen. 1984. A Rating System For Predicting the Appearance Of Temik Aldicarb Residues in Potable Water. Environmental Toxicology and Chemistry. 3:589—597. Bank, S., and R.J. Tyrrell. 1984. Kinetics and Mechanism of Alkaline and Acidic Hydrolysis of Aldicarb. J. Agric. Food Chem. 32:1223-1232. Bernard, E.C. and C.W. Laughlin. 1976. Relative susceptibility of selected cultivars of potato to Pratylenchus penetrans. J. Nematol. 8:239-242. Biehn, W.L., A. Hawkins, and P.M. Miller. 1971. Effect of 1,3-Dichloropropene + Chloropicrin on Infection of Potatoes by Verticillium albo-atrum. Plant Disease Reporter 55(11):968-971. Bird, G.W., M. Sarette, C. Coley, E. Meister. 1978. On-Line Pest-Crop Ecosystem Simulation for Pratylenchus penetrans and Solanum Tuberosum. Unpublished documentation. Bird, G.W. 1986. 1986 Potato Nematode Research Report. 1986 Michigan Potato Research Report. 18:97-106. Bird, G.W. 1987. Michigan State University Department of Entomology,Personal Communication. Bromilow, R.H.,R.J. Kaker, M.A.H. Freeman and K. Goroga. 1980. The Degradation of Aldicarb and Oxamyl in Soil. Pesticide Science 11:371-378. Bromilow, R.H., and M. Leistra. 1980. Measured and Simulated Behavior of Aldicarb and its Oxidation Products in Fallow Soils. Pesticide Science. 11:389f395. Brown, M., R.M. Riedel, and R.C. Rowe. 1980. Effects of Pratylenchus penetrans or Verticillium dahliae on growth and yield of Solanum tuberosum cv. Superior in Microplots. J. Nematol. 12(4):216 (abstract). 161 Bunyan, P.J., M.J. Van Den Heuvel, P.I. Stanley, E.N. Wright. 1981. An Intensive Field Trial and Multi-Site Surveillance Exercise on the Use of Aldicarb to Investigate Methods for the Assessment of Possible Environmental Hazards Presented by New Pesticides. AgrO Ecosystems 7:239-262. Burpee, L.L. and J.R.Bloom. 1978. The influence Of Pratylenchus penetrans on the incidence and severity of Verticillium wilt of potato. J. Nematol. 10:95-99. Butler, K.S. 1978. Irrigation in the Central Sands of Wisconsin, Potentials and Impacts. University Of Wisconsin-Extension Geological and Natural History Survey, February 1978, p.4 Chapman, R.A., and C.M. Cole. 1982. Observations on the Influence Of Water and Soil pH on the Persistence on Insecticides. J. Environ. Sci. Health 5:487-405. Cooper, H.M. 1984. The integrative research review - A systematic approach. Applied Social Research Methods Series. Vol. 2. Beverly Hills: Sage Publications, Inc. Cooper, S.C. 1986. Design and installation of a monitoring network for measuring the movement of aldicarb and its residues in the unsaturated and saturated zones, Lee County, Georgia. Proceedings of the Agricultural Impacts on Ground Water - a Conference- sponsored by the National Water Well Association. August 11-13. pp. 194-223 v Dickerson, O.J., H.M. Darling, and E.D. Griffin. 1964. Pathogenicity and population trends of Pratylenchus penetrans on potato and corn. Phytopathology 54:317- 322. _ Dorsch, M.M., R.R.R.Scragg, A.J. McMichael, P.A. Baghurst, and K.F. Dyer. 1984. Congenital Malformations and Maternal Drinking Water Supply in Rural South Australia: A Case—Control Study. American Journal of Epidemiology 119(4):473-486. Drake, S. 1981. Cause experiment and science. p.ix. Chicago: The University of Chicago Press. Dunn, R.A. 1972. Importance Of depth in soil, presence of host roots and role of eggs as compared to vermiform stages in overwintering Of Pratylenchus penetrans at Ithaca, New York. J. Nematol. 4:221-222 (abstract). Erving, J. 1986. Unpublished Summary of Michigan Public Health Data. 162 Exner, M.E.,Spalding R.F. 1979. Evolution of Contaminated Ground water in Holt County, Nebraska. Water Resources Research. Vol 15 no 1, 1979. pp.139-147. Ferris, J.M. 1970. Soil Temperature Effects on Onion Seedling injury by Pratylenchus penetrans. J. Nematol. 2(3):24a-251. ' Francl, L.J., L.V. Madden, R.C. Rowe, and R.M. Riedel. 1987 Potato Yield Loss Prediction and Discrimination Using Preplant Population Densities of yerticillium dahliae and Pratylenchus penetrans. Phytopathology 77(4):579— 584. Hallberg, G.R. 1986. From Hoes to Herbicides-Agriculture and Ground water Quality. Journal of Soil and Water November-December. Conservation. Row Treatment of Hawkins, A. and P.M. Miller. 1971a. Potatoes with Systemics for Meadow Nematode (Pratylenchus penetrans) Control. Am. Potato J. 48:21-24.‘ Hawkins, A. and P.M. Miller“ 1971b. Row fumigation with Vorlex for control of a meadow nematode (Pratylenchus penetrans) in potatoes. Am. Pot. J. 48:65-68. 1984. Shallow Ground water Quality Beneath an Hubbard, R.K. Intensive Multiple-Cropping System using Central Pivot Irrigation. J. Environmental Quality 13(1):156-161. 'Hudson, J.R. 1988. Rhone Poulenc Ag Co. PO Box 12014 2 T.W. Alexander drive, Research Triangle Park, NC. 27709, Personal Communication. C.A., and J.R. Kiniry. 1986. CERES-Maize A simulation Jones, model of maize growth and development. College Station: Texas A & M University Press. Jones, R.L., R.C. Back. 1984. Monitoring Aldicarb Residues in Florida Soil Water. Environmental Toxicology and 3:9-20. Chemistry. Jones, R.L., J.L. Hanson, R.R. Romine, and T.E. Marquardt. 1986. Unsaturated Zone Studies of the Degradation and Movement of Aldicarb and Aldoxycarb Residues. Environmental Toxicology and Chemistry 5:361-372. and W.F. Mai. 1968. Influence of soil mositure Nematologica 14:101-122. Kable, P.F. on Pratylenchus penetrans. 163 Kelley, R., G.R. Hallberg, L.G. Johnson, R.D.Libra, C.A.Thompson, R.C. Splinter, and M.G. DeTroy. 1986. Pesticides in Ground Water in Iowa. Proceedings of Agricultural Impacts on Ground Water. August 11-13, 1986. Sponsored by The National Water Well Association. pp.622-647. Klingler, J. 1972. The Effect of Single and Combined Heat and Carbon Dioxide Stimulation at Different Ambient Temperatures on the Behavior of two Plant Parasitic Nematodes. J. Nematol. 4(2):95-100. Kimpinski, J. 1979. Root-lesion nematodes in potatoes. Am. Potato J. 56:79-86. Kimpinski, J. and C.B. Willis. 1981. Influence of Soil Temperature and pH on Pratylenchus penetrans and P.crenatus in Alfalfa and Timothy. J. Nematol. l3(3):333-338. Kimpinski, J. 1982. The Effect of Nematicides on Pratylenchus penetrans and Potato Yields. Am. Potato J. 59:327-335. Kotcon, J.B., D.I. Rouse, and J.E. Mitchell. 1985. Interactions of Verticillium dahliae, Colletotrichum coccodes, Rhizoctonia solani, and Pratylenchus penetrans in Early Dying Syndrome of Russet Burbank Potatoes. Phytopathology 75(1):68—74. Kotcon, J.B. and R. Loria. 1986. Influence of Pratylenchus penetrans on Plant Growth and Water Relations in Potato. J. Nematol 18(3):385-392. Kruska, R. 1986. Unpublished analysis of Michigan Public Water Well Records. Lachat. 1988. Multi-Channel Colorimeter Operating Instruction. Lachat 10500 North Port Washington road, Mequon, Wisconsin, 43092. Leistra, M.,A.Dekker, and A.M.M. van der Burg. 1984. Leaching of Oxidation Products of Aldicarb from Greenhouse Soils to Water Courses. Arch. Environ. Contam. Toxicol. 13:327-344. Lemley, A.T., and Wei-Zhu Zhong. 1984. Hydrolysis of Aldicarb, Aldicarb Sulfoxide, and Aldicarb Sulfone at Parts per Billion Levels in Aqueous Mediums. J. Agric. Food Chem. 32:714-719. Li-Tse Ou, K.Sture, V. Edvardsson, and P.C. Rao. 1985a. Aerobic and Anaerobic Degradation of Aldicarb in Soils. J. Agric. Food Chem. 33:72-78. 164 Li-Tse Ou, S.V. Edvardsson, J.E.Thomas, and P.C. Rao. 1985b. Aerobic and Anaerobic Degradation of Aldicarb Sulfone in Soils. J. Agric. Food Chem. 33:545-548. Maitlen, J.C., and D.M. Powell. 1982. Persistence of Aldicarb in Soil Relative to the Carry-Over of Residues into Crops. J. Agric. Food Chem. 30:589-592 Mamiya, Y. 1971. Effect of Temperature on the Life Cycle of Pratylenchus penetrans on Cryptomeria Seedlings and Observations on its reproduction. Nematologica 17(1):82-92 Martin, M.J., R.M. Riedel, and R.C. Rowe. 1982. Verticillium dahliae and Pratylenchus penetrans: Interactions in the Early Dying Complex of Potato in Ohio. Phytopathology 72(6): 640-644. McWilliams, L. 1984. Ground water Pollution in Wisconsin: A Bumper Crop Yields Growing Problems. Environment 26(4):25-34. Michigan Agricultural Statistics Service. 1988. 201 Federal Building, P.C.Box 20008, Lansing, Mi.,48901. (517) 334— 6001. Michigan Department of Agriculture. 1986. County Food and Agricultural Development Statistics. Miller, P.M. 1974. Effect of Soil Temperature on Control of Pratylenchus penetrans by Three Contact Nematocides. Plant Dis. Rep. 58(8):708-710. Noling, J.W. G.W. Bird, and E.J. Grafius. 1984. Joint Influence of Pratylenchus penetrans (Nematoda) and Leptinotarsa decemlineata (Insecta) on Solanum tuberosum Productivity and Pest Population Dynamics. Journal of Nematology. 16(3):230-234 01thof, H.A. 1983. Reaction of six potato cultivars to Pratylenchus penetrans. Canadian Journal of Plant Pathology 5:285-288. 01thof, H.A., E.D. McGarvey, and M. Chiba. 1985. Oxamyl in the control of Pratylenchus penetrans on Potatoes. Canadian Journal of Plant Pathology 7:155-160. 01thof, H.A. 1986. Reaction of Six Solanum tuberosum Cultivars to Pratylenchus penetrans. J. Nematol. l8(1):54-58. Oostenbrink, M. 1958. An inoculation trial with Pratylenchus penetrans in potatoes. Nematologica 3:30-33. 165 Patterson, S.M.T. and G.B. Bergeson. 1967. Influence of Temperature, Photoperiod, and Nutrition on Reproduction, Male-Female—Juvenile Ratio, and Root to Soil Migration of Pratylenchus penetrans. Plant Disease Reporter 51(2):78-82. Pionke, H.B. and J.B. Urban. 1985. Effect of Agricultural Land Use on Ground water Quality in a Small Pennsylvania Watershed. Ground Water 23(1):68-80 Riedel, R.M., R.C. Rowe, and M.J. Martin. 1985. Differential Interactions of Pratylenchus crenatus, P.penetrans, and P.scribneri with Verticillium dahliae in Potato Early Dying Disease. Phytopathology 75(4):419-422. Rhone-Poulenc. 1988. Temik Brand Aldicarb Pesticide Briefing Book. Rhone-Poulenc Ag Company P.O. Box 12014, 2 T.W. Alexander Drive Research Triangle Park, N.C. Rothschild, E.R., R.J. Manser, M.P.Anderson. 1982. Investigation of Aldicarb in Ground water in Selected Areas of the Central Sand Plain of Wisconsin. Ground Water 20(4): 437- -445. Rowe, R.C., R.M. Riedel, and M.J. Martin. 1985. Synergistic Interactions Between Verticillium dahliae and Pratylenchus penetrans in Potato Early Dying Disease. Phytopathology 75(4):412-418. Singh, B. and G.S.Sekhon. 1979. Nitrate Pollution of Ground water from Farm use of Nitrogen Fertilizer - A Review. Agriculture and Environment 4:207-225. Smelt, J.H., A. Dekker, M.Leistra, and N.W.H. Houx. 1983. Conversion of Four Carbamoyloximes in Soil Samples From Above and Below the Soil Water Table. Pesticide Science 14:173-181 Supak, J.R., A.R. Swoboda, and J.B. Dixion. 1978. Adsorption of Aldicarb by Clays and Soil Organo-Clay Complexes. Soil Sci. Soc. Am. J. 42:244-248 Swartz, M.S. 1986. Unpublished summary of Montcalm Potato Research Reports Swartz, M.S. 1987. Unpublished documentation of SUBSTOR potato growth and development model. Tabachnik, B.G., and L.S. Fidell. 1983. Using Multivariate Statistics. New Yorszarper & Row Publishers. Townshend, J.L. 1972. Influence of Edaphic Factors on Penetrations of Corn Roots by Pratylenchus penetrans and Pratylenchus minus in three Ontario Soils. Nematologica 18(2). 201- -212. 166 Townshend, J.M, and C.F. Marks. 1977. Temperature and Expression of Damage Caused by Pratylenchus penetrans in Flue Curred Tobacco. Nematologica 23(1):29-32. Townshend, J.L. 1978. Infectivity of Pratylenchus penetrans on Alfalfa. J. Nematol. 10(4):318-323. United States Department of Agriculture, Soil Conservation Service. 1960. Soil Survey: Montcalm County Michigan. Series 1949 no# 11. Vitosh, M.L., J.W. Noling, G.W. Bird, and R.W. Chase. 1980. The joint Action of Nitrogen and Nematicides on Pratylenchus penetrans and Potato Yield. American Potato Journal. 57:101-111. Vitosh, M.L. 1987. Michigan State University Department of Crop and Soil Science, Personal Communication. Wagenet, R.J. and J.L. Hudson. 1986. Predicting the Fate of Nonvolatile Pesticides in the Unsaturated Zone. Journal of Environmental Quality 15(4)315-322. Webster's New World Dictionary, 2d ed. 1979, S.V. "meta.“ Wong, K. and J.M. Ferris. 1968. Factors influencing the population fluctuations of Pratylenchus penetrans in soil. III. Host plant species. Phytopatholog 58:662-665. ' Zaki, M.H., D. Morgan, and D. Harris. 1982. Pesticides in Ground water: The Aldicarb Story in Suffolk County, NY. American Journal of Public Health. 72(12):1391-1395 APPENDIX A STUDY AREA LAND USE 1986-1988 Tommi. 03 name new 53» Co cosmocmwmcawt m6 .1 050C Z>>OZ¥ZD_ awe/«P: Imde. >mmhm2wu f“ 24913 167 D (I <1 I U fit 0 LU _J O. O. 5 2x00 $025.15 8me mm: pcm. mmtm spam to cosmucmmmtdmt m6 .N 0.59”. y. .. (m. 168 Emisoao 2262012: 8 I W i mZQmm>Om ZmOu. mwOHoo 5504. 169 ngwm>0m I 2x09 mmofifioafl APPENDIX B PRATYLENCHUS penetrans DATA BASE 170 Table 1. Pratylenchus penetrans data base pre-plant measures. Study P.p.[100 cc Soil N0# Cultivar 1gp; Jdate 0010 Aldicarb Check 1 SUP 1986 133 20.3 32.0 38.0 2 SUP 1982 148 43.1 12.0 16.0 3 SUP 1982 149 43.1 16.0 9.0 4 SUP 1982 133 223.3 14.0 7.0 5 SUP 1981 134 131.0 3.6 3.6 6 SUP 1981 134 131.0 2.0 3.6 7 SUP 1981 v 134 131.0 5.6 2.0 8 SUP 1981 134 131.0 3.2 5.2 9 SUP 1980 134 164.3 45.2 28.6 10 SUP 1979 148 46.6 30.4 28.2 11 SUP 1979 148 46.6 23.0 42.4 12 SUP 1979 148 46.6 42.0 54.6 13 SUP 1978 121 139.2 6.0 14.0 14 SUP 1978 121 139.2 12.0 24.0 15 SUP 1978 121 139.2 24.0 22.0 16 SUP 1977 130 0.0 24.0 43.0 17 SUP 1977 130 0.0 24.0 43.0 18 SUP 1977 130 0.0 24.0 43.0 19 RB 1986 141 116.0 3.2 20.2 20 R8 1982 122 0.0 17.0 7.0 21 R8 1982 122 0.0 24.0 11.0 22 RB 1982 122 0.0 29.0 45.0 23 RB 1982 122 0.0 54.0 67.0 24 RB 1977 130 0.0 51.0 55.0 25 RB 1977 130 0.0 51.0 55.0 26 RB 1977 130 0.0 51.0 55.0 27 RB 1985 129 131.0 3.8 3.8 28 ATL 1983 131 94.9 9.9 11.0 29 ATL 1983 131 94.9 2.5 4.0 30 ATL 1983 131 94.9 23.0 21.5 31 ATL 1983 131 94.9 47.0 15.5 32 ATL 1982 116 0.0 11.0 23.0 33 ATL 1982 116 0.0 57.0 51.0 34 ATL 1981 136 171.4 18.0 16.0 171 Table 2. Pratylenchus penetrans data base harvest measures. Study J B CWT/Acre A CUT/Ac J CUT/Ac Knob CUT/Ac Total CWT/Ac No# date Treat Check Treat Check Treat Check Treat Check Treat Check 2 252 11.0 11.0 193.0 177.0 13.0 7.0 217.0 195.0 3 250 13.0 12.0 193.0 177.0 13.0 7.0 219.0 196.0 4 250 18.0 19.0 205.0 163.0 9.0 3.0 232.0 185.0 5 237 11.2 10.4 343.5 289.6 36.8 22.7 391.5 322.7 6 237 11.6 9.1 367.7 307.3 65.3 38.5 444.6 354.9 7 237 20.7 20.2 277.5 249.5 9.1 7.5 307.3 277.2 8 237 14.8 13.0 326 3 276 7 38.2 15.6 379.3 305.3 9 230 214.2 117 4 2.3 0.0 10 243 11.7 9 7 220.8 191.8 17.8 9.3 250.3 210.8 11 243 10.0 9 4 246.6 221.0 25.6 13.4 282.2 243.8 12 243 10.6 10.4 281.7 252.2 32.5 17.3 324.8 279.9 13 233 10.8 8.4 264.6 212 0 4.1 4.3 279.5 224.7 14 233 9.4 6 9 285.3 235.0 10.2 7.2 304.9 249.1 15 233 8.4 8.0 301.1 224.5 22.5 5.8 332.0 238.3 16 265 172.0 119.0 17 265 196.0 105.0 18 265 174.0 86.0 19 281 31.3 32.8 277.8 157.8 38.5 29.8 411.1 271.1 20 252 88.0 99.0 231.0 166.0 8.0 3.0 16.0 6 0 343.0 274.0 21 252 63.0 73.0 313.0 287.0 38.0 27.0 24.0 14 0 428.0 394.0 22 252 76.0 95.0 216.0 193.0 5.0 1.0 20.0 4 0 318.0 293.0 23 252 64.0 71.0 293.0 254.0 21.0 12.0 27.0 10.0 405.0 347.0 24 265 256.0 200.0 25 265 245.0 213.0 26 265 270.0 194.0 27 267 3.3 4.4 334.5 212.2 42.2 8.7 26.9 14.6 444.1 280.3 28 271 28.0 28.0 291.0 300.0 14.0 6.0 333.0 334.0 29 271 28.0 30.0 329.0 366.0 46.0 25.0 403.0 421.0 30 271 30.0 32.0 353.0 372.0 48.0 24.0 431 0 428.0 31 271 34.0 25.0 339.0 365.0 38.0 22.0 411.0 412.0 32 252 34.0 30.0 360.0 348.0 61.0 28.0 455.0 406.0 33 252 36.0 31.0 354.0 326.0 42.0 18.0 432.0 375.0 34 254 13.2 13.2 275.1 272.0 81.2 45.2 369.5 330.4 Table 3. Pratylenchus penetrans data base agronomic measures. St N Citation N__0#k khag7 kgiha aR__OtationN _# sup! 1 2 G. N. BIRD, 1986 2 7 G. H. BIRD, 1982 3 8 G. U. BIRD, 1982 4 9 G. H. BIRD, 1982 5 84 ALFAL 10 H. C. OLSON, 1984 6 252 ALFAL 10 H. C. OLSON,1984 7 84 CORN 11 H. C. OLSON,1984 8 252 CORN 11 H. C. OLSON,1984 9 13 H. C.OLSON,1980 10 0 14 J. NOLING, 1981 11 56 14 J. NOLING, 1981 12 168 14 J. NOLING, 1981 13 84 14 J. NOLING, 1981 14 168 14 J. NOLING, 1981 15 336 14 J. NOLING, 1981 16 86 15 M.L.VITOSH,1980 17 168 15 M.L.VITOSH,1980 18 336 15 M.L.VITOSH,1980 19 1 Model VALIDATION,1986 20 84 CORN 16 M.L.VITOSH,1982 21 253 CORN 16 H. L. V1T0$H,1982 22 84 ALFAL 16 N. L. VITOSH,1982 23 253 ALFAL 16 H. L. VITOSH,1982 24 86 15 H. L. VITOSH,1980 25 168 15 H. L. VITOSH, 1980 26 336 15 H. L. VITOSH, 1980 27 18 Model VALIDATION, 1985 28 84 CORN 4 H. L. VITOSH, 1983 29 253 CORN 4 N. L. VITOSH, 1983 30 84 ALFALFA 5 H. L. VITOSH, 1983 31 253 ALFALFA 5 M. L. VITOSH, 1983 32 253 CORN 16 H. L. VITOSH, 1982 33 253 ALFALFA 16 N. L. VITOSH, 1982 34 12 r u BIRD 1981 173 Table 4. Pratylenchus penetrans data base in-season nematode population densities for Superior. SOIL ROOT TOTAL Study P.9(100 cc P.2[1.0 9. Soil + Root No# Jdate DD10 Treat Check Treat Check Treat Check 1 174 370.1 7,0 4 . 1 209 668.6 17 0 182.0 1 240 933.0 3 0 124,0 1 251 1026.8 4 0 126.0 2 196 557.8 4.0 47.0 3 197 1103 5 0.0 3.0 4.0 48.0 4.0 51.0 3 250 1925.1 2.0 44.0 4 192 998.5 2.0 16.0 3.0 92 0 5.0 108.0 4 250 1925.1 2.0 50.0 5 174 635.0 0.0 3.6 3.2 14.8 3.2 18.4 5 208 1063.0 0.8 6.4 2.8 84.8 3.6 91.2 5 237 1476 0 1.2 18.8 6 174 635 0 0.8 1.2 0.4 17.2 1 2 18.4 6 208 1063.0 0.0 6.4 0.4 72.8 0.4 79.2 6 237 1476.0 1.2 20.0 7 208 1063.0 0.0 6.4 0.4 80.0 0.4 86 4 7 237 1476.0 0.0 30.0 8 208 1063.0 0 4 6.8 0.0 55.6 0.4 62 4 8 237 1476.0 1.6 17.7 9 182 940 3 8.4 48.6 9 209 1398.1 1.8 172.4 9 230 1692.1 2.0 180.1 10 165 212.3 36.0 41.0 10 167 233.7 17.4 20.6 2.6 19.2 20.0 39.8 10 188 442.7 2.0 19.6 11.0 56.2 13.0 75.8 10 216 723.2 85.0 78.2 6.4 178.2 91.4 256.4 10 243 995.5 2.6 39.0 6.0 94.6 8.6 133.6 11 165 212.3 24.0 39.2 11 167 233.7 9.6 20.2 7.0 18.6 16.6 38 8 11 188 442.7 3.0 12.4 14.0 24.2 17.0 36.6 11 216 723.2 23.6 120.8 5.6 205.2 29.2 326.0 11 243 995.5 4.2 55.6 25.2 181.6 29.4 237.2 12 165 212.3 36.8 50.0 12 167 233.7 12.0 19.4 2.8 24.6 14.8 44.0 12 188 442.7 2.8 20.6 10.8 36.0 13.6 56.6 12 216 723.2 25.8 80.4 9.0 175.0 34.8 255.4 12 243 995 5 8.4 40.6 5.6 213.8 14.0 254.4 13 145 241.2 4.0 6.0 13 176 373 5 8.0 12.0 0.0 68.0 8.0 80.0 13 193 543 8 0.0 20.0 0.0 88.0 0.0 108.0 13 213 769.0 0.0 44.0 0.0 100.0 0.0 144.0 13 233 983.3 4.0 16.0 6.0 30.0 10.0 46.0 14 145 241 2 4.0 6.0 14 176 373 5 4.0 88.0 8.0 62.0 12.0 150.0 14 193 543.8 4.0 34.0 2.0 70.0 6.0 104.0 14 213 769.0 0.0 20.0 0.0 158.0 0.0 178.0 14 233 983 3 0.0 68.0 4.0 32.0 4.0 100.0 174 Table 4 (cont). Pratylenchus penetrans data base in- season nematode population densities for Superio or. 501 L ROOT TOTAL Study P.2(100 cc 1.0 . Soil + Root N0# Jdate 0010 Treat Check Treat Check Treat Check 15 145 241. 2.0 87 15 176 373.5 0.0 10.0 16.0 58.0 16. 0 68.0 15 193 543.8 6.0 20.0 4.0 102.0 10. 0 122.0 15 213 769.0 0.0 32.0 48.0 100.0 48. 0 132.0 15 233 983.3 2.0 38.0 0.0 42.0 2. 0 80.0 16 181 709.5 1. 0 166.0 16 199 959.7 13. 0 155.0 16 213 1154.3 15.0 118.0 16 220 1251.6 10. 0 132.0 16 265 1877.1 5. 0 67.0 17 181 709.5 1. 0 166.0 17 199 959.7 13. 0 155.0 17 213 1154.3 15.0 118.0 17 220 1251.6 10. 0 132.0 17 265 1877.1 5. 0 67.0 18 181 709.5 1. 0 166.0 18 199 959.7 13. 0 155.0 18 213 1154.3 15.0 118.0 18 220 1251.6 10. 0 132.0 18 265 1877.1 5. 0 67.0 Table 5. Pratylenchus penetrans data base in-season nematode population densities for Russet Burbank. Study P.2[1OO cc P.9(1.0 9. N0# Jdate 0010 Treat Check Treat Check Treat Check 19 149 143 3 .8 11.2 1.0 14.6 1.8 25.8 19 155 190.5 0.8 11.8 0.8 44.2 1.6 56.0 19 161 231.7 2.4 15.8 1.0 44.6 3.4 60.4 19 174 335.2 1.0 20.0 0.8 52.0 1.8 72.0 19 188 461.8 2.0 63.4 0.2 50.4 2.2 113.8 19 202 629.8 1.8 62.4 1.8 69.4 3.6 131.8 19 216 782.4 1.0 185.8 0.6 147.2 1.6 333.0 19 230 917.2 3.2 106.2 5.8 135.0 9.0 241.2 19 245 1014.3 0.8 129.6 3.4 235.0 4.2 364.6 19 258 1074.3 1.2 83.0 0.8 269.0 2.0 352.0 19 280 1196.9 0.6 102.6 20 130 0.0 16.0 6.0 20 196 974.4 1.0 70.0 20 252 2014.8 31.0 132.0 21 130 0.0 4. 7.0 21 196 974.4 1.0 36.0 21 252 2014.8 10.0 104.0 22 130 0.0 22.0 16.0 22 196 974.4 2.0 50.0 22 252 2014.8 14.0 268.0 ' 23 130 0.0 4.0 8.0 23 196 974.4 2.0 35.0 23 252 2014.8 13.0 150.0 24 181 709.5 9.0 128.0 24 199 959.7 22.0 222.0 24 213 1154.3 8.0 73.0 24 220 1251.6 15.0 202.0 24 265 1877.1 30.0 50.0 25 181 709.5 9.0 128.0 25 199 959.7 22.0 222.0 25 213 1154.3 8.0 73.0 25 220 1251.6 15.0 202.0 25 265 1877.1 30.0 50.0 26 181 709.5 9.0 128.0 26 199 959.7 22.0 222.0 26 213 1154.3 8.0 73.0 26 220 1251.6 15.0 202.0 26 265 1877.1 30.0 50.0 27 157 276.0 0.2 3.7 0.4 12.1 0.5 15.8 27 170 329.9 0.1 2.5 0.0 12.3 0.1 14.8 27 182 422.1 0.0 6.5 0.1 9.7 0.2 16.1 27 197 568.6 0.5 16.3 0.3 30.0 0.8 46.3 27 211 709.3 0.7 35.7 0.2 23.2 1.0 58.9 27 239 924.6 1.2 38.6 0.4 16.5 1.6 55.1 27 254 1082.7 0.4 45.6 27 267 1153.0 1.7 44.3 SOIL ROOT 175 TOTAL Soil + Root 176 Table 6. Pratylenchus penetrans data base in-season nematode population densities for Atlantic. SOIL ROOT TOTAL Study P.9(100 cc 9,211.0 g. Soil + Root N0# Jdate 0010 Treat Check Treat Check Treat Check 28 214 1421.0 0.0 29.7 18.8 0.3 18.8 30.0 28 271 2359.8 14.7 60.8 29 214 1421.0 0.7 13.7 18.8 0.1 19.5 13.8 29 271 2359 8 16.3 50.3 30 214 1421.0 0.3 60.7 19.4 0.4 19.7 61.1 30 271 2359.8 10.7 95.8 31 214 1421 0 0.7 35.0 15.2 0.1 15.9 35.1 31 271 2359.8 25.3 127.0 32 130 0.0 14.0 5.0 32 196 974.4 2.0 34.0 32 252 2014.8 19.0 132.0 33 130 0.0 13.0 12.0 33 252 2014.8 12.0 237.0 3 0 34 223 1276.3 1.0 78.0 34 254 1673.0 4.0 .0 177 lation data. Model validation data base nematode Table 7. ROOT SOIL+RO0T STOLON SOIL CHK ALD CHK ALD CHK ALD YEAR DATE DAP 1987 121 $338.0. 2231nal. 4323/4 00858 . . . . . 40050 00530088 0 12 45 60 72 101 122 -9 1987 133 1987 166 1987 181 1987 193 1987 222 1987 243 1985 120 8 8 1 3 8 1 SIM—368.5 111/Q45 511.280.:0 0.0.nU-0.-l.al. «J37unwnl.5 2290 sum $01324 . . . “wooonuunm 7.8.4 1.7.6.6 3 . . . 37.—590.58.5/4 133/41M. 21.05724] . . . coconut-”UL 81.3“2058 2,45 8123 111 70271947 578911356 1.1112222 massages 99999999 111111111 CHK - NON-TREATED CONTROL ALD - ALDICARB TREATED Model validation data base plant growth data Table 8. DRY HEIGHTS IN GRAMS UG STEM TUBER ROOT STOLON ALD CHK CHK CHK MDCM Am P A D r a e V: 0 12 1987 1987 1987 45 1987 60 1987 72 1987 101 1987 122 APPENDIX C BASIC DESCRIPTIVE STATISTICS FOR PPDB Table 1. Descriptive statistics cultivar Superior. 178 for varibales in P.p data base - PJD PLANTING JDATE PDD PLANTING D010 PTN PREPLANT TEMIK NEMATOOE PCH PREPLANT CHECK NEMATOOE HJD HARVEST JDATE GSL GROWING SEASON LENGTH TBH TEMIK 8 HT CBH CHECK 8 HT TAU TEMIK A HT CAH CHECK A HT TJH TEMIK JUMBO UT CJU CHECK JUMBO UT TTH TEMIK TOTAL UT CTH CHECK TOTAL HT DBHT B X YIELD LOSS DAHT A X YIELD LOSS DJHT J 2 YIELD LOSS DTUT T X YIELD LOSS STS SAMPLE TEMIK SOIL SCS SAMPLE CHECK SOIL STR SAMPLE TEMIK ROOT SCR SAMPLE CHECK ROOT STT SAMPLE TEMIK TOTAL 155_ LABEL _§_ fll§§ MEAN STD MIN MAX 18 0 130.67 15.37 104.00 149.00 18 0 87.47 67.07 0.00 223.30 18 0 19.06 13.03 2.00 45.20 18 0 23.73 17.05 2.00 54.60 18 0 244.67 11.52 230.00 265.00 18 0 114.00 22.78 95.00 161.00 13 5 12.40 3.51 8.40 20.70 13 5 11.35 4.01 6.90 20.20 15 3 280.89 80.03 193.00 493.00 15 3 236.20 77.77 117.40 449.00 14 4 21.39 17.18 2.30 65.30 14 4 11.33 9.91 0.00 38.50 16 2 281.65 80.11 172.00 444.60 16 2 224.54 76.69 86.00 354.90 13 5 0.10 0.10 -0.06 0.27 15 3 0.16 0.09 0.08 0.45 14 4 0.47 0.25 -0.05 1.00 16 2 0.22 0.12 0.10 0.51 42 25 8.13 15.42 0.00 85.00 42 25 31.93 27.67 1.20 120.80 35 32 6.10 9.08 0.00 48.00 35 32 77.06 57.10 14.80 213.80 54 13 11.32 14.56 0.00 91.40 54 13 117.97 66.28 13.40 326.00 SCT SAMPLE CHECK TOTAL a r... ...»... . _ . . .. 1 .... v .... a, . . 1 . . . 1- ...w .. Table 2. Descriptive statistics for variable cultivar Russet Burbank. 179 in P.p. data base- 153_ LABEL _!_ El§§ MEAN s10 MIN MAx PJD PLANTING JDATE 9 0 127.22 12.38 122.00 141.00 PDD PLANTING 0010 9 0 27.44 54.59 0.00 131.00 PTN PREPLANT TEMIK NEHATODE 9 0 31.56 20.88 3.20 54.00 PCN PREPLANT CHECK NEMATOOE 9 0 35.44 24.68 3.80 67.00 HJD HARVEST JDATE 9 0 261.22 10.05 252.00 281.00 GLs GROHING SEASON LENGTH 9 0 145.00 12.07 136.00 161.00 18H TEMIK 8 NT 6 3 54.27 31.33 3.30 88.00 C8u CHECK B NT 6 3 62.53 36.96 4.40 99.00 IAH TEMIK A HT 6 3 277.55 46.25 216.00 334.50 CAH CHECK A NT 6 3 211.67 50.60 157.80 287.00 TJH TEMIK JUMBO HT 6 3 25.45 16.44 5.00 42.20 CJH CHECK JUMBO NT 6 3 13.58 12.16 1.00 29.80 TNH TEMIK KNOBBY HT 5 4 22.78 4.74 16.00 27.00 CNH CHECK KNOBBY NT 5 4 9.72 4.71 4.00 14.60 TTH TEMIK TOTAL HT 9 0 346.69 78.21 245.00 444.10 CTU CHECK TOTAL HT 9 0 274.04 66.83 194.00 394.00 DBHT 8 x YIELD Loss 6 3 ~o.17 0.10 -0.33 -o.05 OAHT A z YIELD Loss 6 3 0.23 0.15 0.08 0.43 DJHT J X YIELD LOSS 6 3 0.53 0.25 0.23 0.80 DNHT K X YIELD LOSS 5 4 0.59 0.15 0.42 0.80 OTuT T x YIELD Loss 9 o 0.20 0.11 0.08 0.37 STS SAMPLE TEMIK SOIL 27 19 4.97 7.75 0.00 31.00 scs SAMPLE CHECK SOIL 27 19 62.07 66.47 2.50 268.00 STR SAMPLE TEMIK ROOT 18 28 1.20 1.44 0.00 5.80 SCR SAMPLE CHECK ROOT 18 28 69.45 76.96 9.70 269.00 STT SAMPLE TEMIK TOTAL 33 13 8.7 9.47 0 10 30.00 SCT SAMPLE CHECK TOTAL 33 13 123.89 100.12 14 80 364.60 ' 511m . '3'!!! '2. 180 Table 3. Descriptive statistics for variables in P.p. Data base - cultivar Atlantic. M LABEL _N_ as; MEAN STD MIN MAx PJD PLANTING JDATE 7 0 127.4 8.02 116.00 136.00 POD PLANTING D010 7 0 78.71 60.59 0.00 171.40 PTN PREPLANT TEMIK NEMATCDE 7 0 24.06 20.35 2.50 57.00 PCN PREPLANT CHECK NEMATOJE 7 0 20.29 14.97 4.00 51.00 HJD HARVEST JDATE 7 0 263.14 9.82 252.00 271.00 GLS GROHING SEASON LENGTH 7 0 135.71 8.04 118.00 140.00 TBU TEMIK 8 HT 7 0 29.03 7.66 13.20 36.00 CBN CHECK 8 HT 7 0 27.03 6.51 13.20 32.00 TAU TEMIK A HT 7 0 328.73 33.19 275.10 360.00 CAN CHECK A NT 7 0 335.57 38.02 272.00 372.00 TJH TEMIK JUMBO HT 7 0 47.17 20.67 14.00 81.20 CJH CHECK JUMBO NT 7 0 24.03 11.76 6.00 45.20 TTH TEMIK TOTAL HT 7 0 404.93 41.61 333.00 455.00 CTN CHECK TOTAL HT 7 0 386.63 40.78 330.40 428.00 OBNT B X YIELD LOSS 7 0 0.05 0.12 -0.07 0.26 DAUT A X YIELD LOSS 7 0 -0.02 0.07 -0.11 0.08 DJHT J X YIELD L053 7 0 0.50 0.06 0.42 0.57 DTNT T X YIELD LOSS 7 0 0.04 0.07 -0.04 0.13 STS SAMPLE TEMIK SOIL 12 5 10.55 8.36 0.00 25.30 SCS SAMPLE CHECK SOIL 12 5 71.58 67.27 5.00 237.00 STR SAMPLE TEMIK ROOT 5 12 14.44 8.24 0.00 19.40 SCR SAMPLE CHECK ROOT 5 12 7.58 16.44 0.10 37.00 STT SAMPLE TEMIK TOTAL 8 9 10.48 8.65 1.00 19.70 SCT SAMPLE CHECK TOTAL 8 9 41.80 24.88 13.00 78.00 - :_:"0 7 't.“ APPENDIX D ANALYSIS OF VARIANCE RESULTS Table 1. for J Analysis of variance results variable: B yield. Source DE SUM OF SOUARES Model 11 20135.23 Error 42 10571.84 C Total 5; 30707.08 F Value 7.27 r-Square 0.66 Pr > F 0.0001 BNT MEAN 26.91 Source 05 F Value Pr > F PPCODE 1 4.56 0.0386 TRE 1 0.12 0.7312 CUL 2 33.87 0.0001 PPCODE*TRE 1 0 . 00 0 .9809 PPCODE‘CUL 2 2.86 0.0686 TRE‘CUL 2 0.90 0.4162 PPCODE*TRE*CUL 2 0.03 0.9685 Table 2. AnalySis of variance results for deggndant variable: A vield. Source 0: SUM OF SQUARES Model 11 179106.35 Error 56 313065.02 C Total 61 492171.37 F Value 2.91 r-Square 0.36 Pr > F 0.0042 ANT MEAN 250.62 Source DE F Value Pr > F PPCODE 1 3.88 0.0537 TRE 1 5.36 0.0243 2 9.69 0.0002 PPCODE‘TRE 1 0.24 0.6264 PPCODE*CUL 2 0.52 0.5979 TRE*CUL 2 0.96 0.3903 PPCODE’TRE*CUL 2 0.11 0.8937 181 Table 3. Analysis of variance results for ‘ ' variable: Jumbo yield. Source 2: SUM OF SQUARES Model 11 8084.83 Error 56 10926 90 C Total 6: 19011.74 F Value 3.77 r-Square 0.42 PR > F 0.0005 JNT MEAN 20.14 Source 05 F Value Pr > F PPCODE 1 3.45 0.0685 TRE 1 14.80 0.0003 CUL 2 9.73 0.0002 PPCOOE*TRE 1 0.19 0.6656 PPCODE*CUL 2 0.48 0.6241 TRE*CUL 2 1.21 0.3059 PPCODE*TRE*CUL 2 0.08 0.9211 Table 4. Analysis of variance results for “ variable: Total yield. Source 0: SUM OF SQUARES Model 11 267161.22 Error 56 413065.25 C Total 61 680226.46 F Value 3.29 r-Square 0.39 PR > F 0.0016 TNT MEAN 303.38 Source Of F Value Pr > F PPCODE 1 3.50 0.0665 TRE 1 7.53 0.0081 CUL 2 11.13 0.0001 PPCCOE*TRE 1 0.15 0.7012 PPCOOE*CUL 2 0.77 0.4676 TRE*CUL 2 0.52 0.5972 PPCODE‘TRE*CUL 2 0.10 0.9044 182 APPENDIX E PERCENTAGE YIELD REDUCTION USING PRESEASON INFORMATION Table 1. Results of preseason regression analysis for percentage B tuber yield loss. Source _fi Model Error C Total DV1 DV2 APO PJD Parameter Intercept DV1 MN 4 1 5 1 1 1 1 DV2 APO PJD SS 0.3268 r-Sggare 0.58 0.2322 F Value Pr > F 0.5591 7.39 0.0007 0.1360 12.30 0.0021 0.1642 14.85 0.0009 0.0090 0.82 0.3754 0.0175 1.58 0.2220 Estimate T For H0: PR>{T} 0.1353 0.50 0.6233 0.3124 5.15 0.0001 0.2410 4.05 0.0006 0.0011 0.83 0.4186 -0.0027 -1.26 0.2220 Table 2. Results of preseason regression analysis for percentage A tuber yield loss. Source Model Error C Total DV1 DV2 APO PJD Parameter Intercept DV1 DV2 APO PJD O F SS 0.244650 0.257046 0.501697 0.031659 0.210184 0.001342 0.001463 Estimate 0.340994 -0.061507 -0.251830 -0.000482 -0.000784 T For H0: Pr > {Ti 1.25 0.2238 -1.04 0.3088 -4.21 0.0003 0.7103 -0.36 0.7207 Table 3. Results of preseason regression analysis for percentage Total tuber yield loss. Source _5 SS r-Sggare Model 4 0.244812 0.49 Error 27 0.250441 F Value Pr > F C Total 31 0.495254 6.60 0.0008 DV1 1 0.053313 5.75 0.0237 DV2 1 0.102765 11.08 0.0025 APO 1 0.004827 0.52 0.4768 PJD 1 0.083906 9.05 0.0056 Parameter Estimate T For H0: Pr > :T: Intercept 0.710082 3.89 0.0006 DV1 0.062287 1.39 0.1756 DV2 -0.120866 -2.37 0.0254 APO -0.000552 -0.49 0.6281 PJD -0.004188 -3.01 0.0056 _w APPENDIX F PERCENTAGE YIELD REDUCTION USING POST SEASON INFORMATION Table 1. Results of post season regression analysis for percentage B tuber yield loss. S r-Sguare Model 6 0.3392 0.61 Error 19 0.2198 F Value Pr > F C Total 25 0.5591 4.86 0.0035 DV1 1 0.1360 11.76 0.0028 DV2 1 14.19 0.0013 APO 1 0.0090 0.78 0.3871 PJD 1 0.0175 1.51 0.2336 GSL 1 0.0000 0.01 0.9362 TTN 1 0.0123 1.06 0.3154 Parameter Estimate T For H0: PR>{T} Intercept -0.1838 -0.21 0.8350 DV1 0.3534 2.48 0.0226 DV2 0.2326 3.78 0.0013 APO 0.0013 0.92 0.3675 PJD -0.0019 -0.72 0.4814 GSL‘ 0.0004 0.10 0.9192 TTN 0.0004 1.03 0.3154 Table 2. Results of post season regression analysis for percentage A tuber yield loss. Source DE SS r-S are Model 6 0 2511 0.64 Error 19 0.1436 F Value Pr > F C Total 25 0 3947 5.54 0.0018 DV1 1 0 0168 2.23 0.1514 DV2 1 0 2101 27.80 0.0001 APO 1 0.0092 1.23 0 2813 PJD 1 0.0010 0.14 0.7139 GSL 1 0.0046 0.62 0.4425 TTN 1 0.0090 1.20 0.2874 Parameter Estimate T For H0: PR>{T} Intercept -0.4487 '0.64 0.5312 DV1 0.0322 0.28 0.7824 DV2 -0.2615 -5.26 0.0001 APO -0.0007 -0.58 0.5657 PJD 0.0009 0.43 0.6733 GSL 0.0032 0.97 0.3463 TTN 0.0003 1.09 0.2874 :'.:-I« .I “.1 ... «cine!!! .58211162 I Table 3. Results of post season regression analysis for percentage total tuber yield loss. Source DE SS r-Sguare Model 6 0.3265 0.66 Error 25 0.1687 F Value Pr > F C Total 31 0.4952 8.06 0.0001 DV1 1 0.0533 7.90 0.0095 DV2 1 0.1027 15.23 0.0006 APO 1 0.0048 0.72 0.4057 PJD 1 0.0839 12.43 0.0017 GSL 1 0.0567 8.41 0.0077 TTN 1 0.0249 3.70 0.0660 Parameter Estimate T For H0: Pr > T Intercept -1.0097 -1.94 0.0639 DV1 0.2048 3.35 0.0026 DV2 -0.1542 -3.42 0.0021 APO 0.0001 0.04 0.9675 PJD 0.0014 0.71 0.4872 GSL 0.0059 3.48 0.0019 TTN 0.0005 1.92 0.0660 ‘1 vii-'1'! J‘ I)“. n 1.1.1115.qu - _' 3.1.81.0!!! 1’35“ APPENDIX G PERCENTAGE YIELD REDUCTION USING STEPWISE PROCEDURE ON POSTSEASON INFORMATION Table 1. Results of stepwise regression analysis for percentage 8 tuber yield loss on Superior. Source _fi SS r-Sguare Model 2 0.04425814 0.39 Error 10 0.06901162 F Value Pr > F TOTAL 12 0.11326975 3.21 0.0850 Parameter Estimate F Value Pr > F Intercept 2.4642718? PJD -0.00920442 6.34 0.0305 GSL -0.01068368 3.65 0.0852 Table 2. Results of stepwise regression analysis for percentage A tuber yield loss on Superior. Source 95 SS r-Sguare Model 1 0.02323454 0.70 Error 11 0.00985222 F Value Pr > F Total 12 0.03308676 25.94 0.0003 Parameter Estimate F Value Pr > F Intercept 0.69449382 PJD -0.00401257 25.94 0.0003 Table 3. Results of stepwise regression analysis for percentage total tuber yield loss on Superior. Source _fi SS £;§ggggg Model 3 0.18393066 0.84 Error 12 0.03367982 F Value Pr > F Total 15 0.21761048 21.84 0.0001 Parameter Estimate F Value Pr > F Intercept -0.89946955 APO 0.00260579 5.44 0.0378 GSL 0.00845263 48.99 0.0001 TTN 0.00047707 4.17 0.0638 Table 4. Results of stepwise regression analysis for percentage A tuber yield loss on Russet Burbank. Source _fi SS r-Sguare Model 1 0.08052799 0.74 Error 4 0.02745202 F Value Pr > F Total 5 0.10798001 11.73 0.0266 Parameter Estimate Intercept -1.242186 PJD 0.012063 187 188 Table 5. Results of stepwise regression analysis for percentage 3 tuber yield loss on Russet Burbank. Source DE SS r-Sguare Model 2 0.04947403 0.54 Error 6 0.0428843? F Value Pr > F Total 8 0.0923583? 8.08 0.0250 Parameter Estimate F Value Pr > F Intercept -0.86??3480 Egg 0.00858743 8.08 0.0250 Table 6. Results of stepwise regression analysis for percentage 3 tuber yield loss on Atlantic. Source Qfi SS r-Sguare Model 1 0.03477225 0.38 Error 5 0.05700278 F Value Pr > F Total 6 0.09177503 3.05 0.1412 Parameter Estimate F Value Pr > F Intercept -0.047219 APO 0.004598 3.05 0.1412 Table 7. Results of stepwise regression analysis for Percentage A tuber yield loss on Atlantic. Source 95 SS r-Sggare Model 2 0.02198049 0.82 Error 4 0.00467399 F Value Pr > F Total 6 0.02665448 9.41 0.0307 Parameter Estimate F Value Pr > F Intercept 1.56674469 PJD -0.00692642 14.84 0.0183 GSL -0.00519901 8.40 0.0442 Table 8. Results of stepwise regression analysis for Percentage Jumbo tuber yield loss on Atlantic. Source 25 SS r-Sggare Model 2 0.01700639 0.72 Error 4 0.00620640 F Value Pr > F Total 6 0.02321279 5.18 0.0776 Parameter Estimate F Value Pr > F Intercept 2.0677198? PJD -0.00883309 10.34 0.0324 TTN -0.00109020 4.24 0.1084 Table 9. Results of stepwise regression analysis for percentage Total tuber yield loss on Atlantic. Source DE SS r-Sgyare Model 3 0.02832836 0.9? Error 3 0.00087822 F Value Pr > F Total 6 0.02920659 32.26 0.0088 Parameter Estimate F Value Pr > F Intercept 1.65479822 APO 0.00093791 3.45 0.1604 PJD -0.00577644 28.73 0.012? GSL -0.0066044? 52.81 0.0052 189 APPENDIX H POPULATION DENSITY REGRESSION RESULTS Table 1. Regression results for P.penetrans pgpglation density in Superior check soils. Source Model Error C Total Intercept PCN PJD SJD SJDZ SJD3 DF 5 36 41 Parameter 88 9829.55 21569.86 31399.41 Estimate 1414.9608 0.8695 ~0.1878 -21.3332 0.1068 -0.0002 r-Sggare 0.31 F Value Pr > F 3.28 0.0153 T For H0: Pr > {T1 1.231 0.2262 3.160 0.0032 -0.461 0.6479 1.141 0.2498 1.141 0.2613 -1.09? 0.2799 Table 2. Regression results for P.penetrans Qgpglation density in Superior aldicarb soils. Source Model Error C Total Parameter Intercept PTN PJD SJD SJDZ SJDS F 5 36 41 SS 3113.168 6641.364 9754.533 Estimate 913.6875 0.3592 0.4706 -15.4558 0.0806 -0.0001 r-Sggare 0.3192 F Value Pr > P 3.375 0.0133 T For H0: Pr > IT: 1.438 0.1591 1.72? 0.092? 2.009 0.0520 —1.534 0.1338 1.55? 0.1281 -1.008 0.1224 Table 3. Regression results for P.penetrans Qggglation density in Superior check root. Source Q: SS r-Sggare Model 5 51871.50 0.4? Error 29 58994.65 F Value Pr > F C Total 34 110866.16 5.10 0.0018 Parameter Estimate T For H0: Pr >ITI Intercept -38?5.3 -0.582 0.5648 PCN 1.68820 3.082 0.0045 PJD -0.01830 -0.024 0.9809 SJD 48.12584 0.486 0.6304 SJDZ -0.19235 -0.395 0.6954 SJD3 0.00025 0.31? 0.7539 190 Table 4. Regression results for P.pegetrans Qgpglation density in Superior aldicarb root. Source Model Error C Total Parameter Intercept PTN PJD SJD SJDZ SJD3 Estimate -?11.40 0.33714 -0.14818 10.56815 -0.05057 0.00008 r-Sguare 0.1? F Value Pr > F 1.19 0.3376 T For Ho- Pr > :1} -0.539 0.5940 2.31? 0.0278 -0.94? 0.3516 0.580 0.5944 -0.524 0.6042 0.511 0.6132 Table 5. Regression results for P.penetrans population density in Superior check total. Source DE SS r-Sggare Model 5 105279.945 0.45 Error 48 127575.992 P Value Pr > F C Total 53 232855.938 7.922 0.0001 Parameter Estimate T For H0: Pr > IT} Intercept -294.76 -0.082 0.9350 PCN 1.72541 3.39? 0.0014 PJD 0.49251 0.608 0.5458 SJO -6.112?1 -0 121 0.9045 SJDZ 0.07516 0.31? 0.752? SJD3 -0.00018 -0.503 0.6175 Table 6. Regression results for P.penetrans Qgpglation density in Superior aldicarb total. Source Model Error C Total Parameter Intercept PTN PJD SJD SJDZ SJD3 95 SS r-Sguare 5 2219.8205 0.20 48 9029.1520 F Value Pr > F 53 11248.9725 2.36 0.0450 Estimate T For HO: Pr > {T} 155.7447 0.162 0.8717 0.19691 1.099 0.2771 0.46261 2.049 0.0460 ‘3.85622 '0.285 0.7769 0.02212 0.349 0.7285 -0.00004 '0.412 0.6819 191 192 Table 7. Regression results for P.penetrans pppglation density in Russet Burbank check soils. Source 25 SS r-Sguare Model 5 75791.81 0.66 Error 21 39114.85 F Value Pr > F C Total 26 114906.67 8.14 0.0002 Parameter Estimate T For H0: Pr > 1T} Intercept 614.2715 0.627 0.5372 PCN 1.5133 2.936 0.0079 PJD '0.2393 '0.222 0.8264 SJD -12.1800 '0.744 0.4648 SJDZ 0.0741 0.902 0.3773 SJD3 -0.0001 -0.994 0.3314 Table 8. Regression results for P.pgnetrans population density in Russet Burbank aldicarb soils. Source _fi SS r-S are Model 5 747.654 0.48 Error 21 815.571 F Value Pr > F C Total 26 1563.226 3.850 0.0124 Parameter Estimate T For H0- Pr > IT: Intercept 165.1576 1.131 0.2710 PTN 0.00278 0.021 0.9834 PJD -0.41923 -2.203 0.0389 SJD -1.64485 -0.696 0.4942 SJOZ 0.00813 0.685 0.500? SJD3 -0.00001 -0.665 0.5132 Table 9. Regression results for P.pgnetrans pppplation density in Russet Burbank check roots. ‘11 Source Model Error 1 C Total 1 Parameter Intercept PCN PJD SJD SJDZ SJD3 5 2 7 SS 86160.043 14529.221 100689.264 Estimate -3943.9? 1.4165 3.7970 53.8211 -0.2813 0.0005 r-Sguare 0.86 F Value Pr > F 14.23 0.0001 T For H0: Pr > {T} -1.562 0.1442 2.534 0.0262 3.345 0.0058 1.418 0.1815 -1.483 0.1639 1.586 0.138? -5 AIM L 11m! 3 ' 3.2. 52.-31:11:11 .fl‘i EL? FILE: 1 T. l _ ...‘I. Table 10. Regression results for P.penetrans population density in Russet Burbank aldicarb roots. Source DF r-Sguare Model 5 15.96340 0.45 Error 12 19.33659 F Value Pr > F C Total 7 35.30000 1.98 0.1540 Parameter. Estimate T For HO- Pr > IT: Intercept 134.78 1.463 0.1692 PTN 0.06556 884 0.0840 PJD 0.0861? 1.576 0.1411 SJD -2 27402 -1.643 0.1263 SJDZ 0.01154 1.669 0.1210 SJD3 -0.00002 -1.679 0.1189 Table 11. Regression results for P.pgnetrans population density in Russet Burbank Check total. Source _5 SS r-Sgpare Model 5 168993.248 0.53 Error 27 151776.480 F Value Pr > F C Total 32 320769.729 6.01 0.000? Parameter Estimate T For H0: Pr > IT} Intercept 425.299 0.10? 0.9156 PCN 0.69121 1.055 0.3009 PJD 7.63305 3.443 0.0019 SJD ~30.4704 ~0.535 0.596? SJDZ 0.19341 0.702 0.4884 SJD3 -0.00036 -0.844 0.4063 Table 12. Regression results for P. pgnet rans population density in Russet Burbank aldicarb total. Source _fi SS r-Sguare Model 5 2316.64063 0.81 Error 2? 557.5090? F Value Pr > F C Total 32 2874.14970 22.44 0.0001 Parameter Estimate T For H0: Pr > {T} Intercept -292.542 -1.219 0.2334 PTN 0.300664 7.361 0.0001 PJD 0.206827 1.443 0.1606 SJD 4.224125 1.225 0.2312 SJDZ -0.022299 -1.336 0.1926 SJD3 0.000038 1.469 0.1535 193 APPENDIX I IMPACT OF IN-SEASON NEMATODE POPULATION DENSITY (TWO Classes) ON PERCENTAGE TUBER YIELD LOSS .r I 194 Table 1. Regression results for Superior percentage A yield loss based on early and late delta root nematode population density. Source 05 SS r-Sgpare Model 2 0.014989 0.82 Error 4 0.003210 F Value Pr > F C Total 6 0.0182 9.34 0.0311 Parameter Estimate T For H0: Pr > {T} Intercept 0.233590 7.219 0.0020 ER 0.000546 1.122 0.3246 LR -0.000835 -4.191 0.0138 Table 2. Regression results for Superior percentage total yield loss based on early and late delta root nematode population density. Source p5 SS r-Sguare Model 2 0.00895872 0.63 Error 4 0.00521270 F Value Pr > F C Total 6 0.01417143 3.44 0.1353 Parameter Estimate T For H0: Pr > {T} Intercept 0.257403 6.243 0.0034 ER 0.000121 0.196 0.8538 LR -0.000665 -2.617 0.0590 Table 3. Regression results for Superior percentage A yield loss based on early and late delta total nematode pppplation density. Source pg SS r-Sgpare Model 2 0.0840136 0.35 Error 9 0.1546780 F Value Pr > F C Total 11 0.2386916 2.44 0.1420 Parameter Estimate T For H0: Pr > {T} Intercept 0.213133 1.593 0.1457 ET 0.001242 1.794 0.1064 LT -0.000555 -0.?69 0.4618 Table 4. Regression results for Russet Burbank percentage Julbo yield loss based on early and late delta total nematode malation density. Source E SS r-Sgpare Model 2 0.18625 0.96 Error 2 0.00726 F Value Pr > F c Total 4 0.19352 25.61 0.0375 Parameter Estimate T For H0: Pr > 'IT'. Intercept 0.951758 11.831 0.0071 ET -0.00361? -5.115 0.0362 LT -0.001423 -4.732 0.0419 195 APPENDIX J IMPACT OF IN-SEASON TOTAL NEMATODE DENSITY ON PERCENTAGE YIELD LOSS ON SUPERIOR TUBERS _. _ I I I «I I 196 Table 1. Regression results for percentage 8 wt reduction based on early season total nematode population density. Source .fi SS r-Sguare Model 1 0.11795 0.73 Error ? 0.04359 F Value Pr > F C Total 8 0.16155 18.93 0.0033 Parameter Estimate T For M0: Pr > {T} Intercept 0.031151 0.659 0.5312 DTN1 0.001859 4.352 0.0033 Table 2. Regression results for percentage A wt reduction based on early mid and late season total nematode pppplation density. Source at SS r-Sgpare Model 2 0.12293 0.83 Error 6 0.02415 F Value Pr > F c Total 8 0.14728 15.14 0.0045 Parameter Estimate T For H0: Pr > {T} Intercept -0.181061 -1.580 0.1651 DTN2 0.003307 5.061 0.0023 DTN4 0.000787 1.653 0.1493 Table 3. Regression results for percentage Jumbo wt reduction based on preplant and late mid season total nematode pppplation density. Source 9: SS r-Sgpare Model 2 0.26320 0.58 Error 6 0.19434 £_yplpg Pr > F C Total 8 0.45755 4.06 0.0766 Parameter Estimate T For H0: Pr > 'T: Intercept 0.281040 1.348 0.2263 APO 0.015459 2.602 0.0406 DTN3 -0.001642 -1.?3? 0.1330 Table 4. Regression results for percentage Total wt redUction based on early mid and late season total nematode pppplation density. Source pf SS r-Sguare Model 2 0.10731 0.81 Error 6 0.0261? F Value Pr > F C Total 8 0.13168 13.209 0.0063 Parameter Estimate T For H0: Pr > IT: Intercept -0.166225 -0.1450 0.1971 DTN2 0.003185 4.873 0.0028 DTN4 0.000919 1.931 0.1017 IT‘ APPENDIX K IMPACT OF ALDICARB ON PLANT GROWTH PARAMETERS L _.3. . w...“ .141 197 Table 1. Regression results for delta gpove_g[9und growth on Russet Burbank. Source pg SS r-Sgpare Model 1 2334.18 0.58 Error 9 1706.73 F Value Pr > F C Total 10 4040.85 12.31 0.0066 Parameter Estimate F Value Pr > F Intercept 2.155202 DSOIL 1.208126 12.31 0.0066 Table 2. Regression results for delta tuber growth on Rpsset Burbank. Source 25 SS r-Sgpare Model 1 2794.51 0.68 Error 9 1331.68 F Valug Pr > F C Total 10 4126.19 18.89 0.0019 Parameter Estimate F Value Pr > F Intercept ~8.868531 DSOIL 1.321916 18.89 0.0019 Table 3. Regression results for percentage below ground growth. Source 25 SS r-Sgpare Model 1 0.18143 0.58 Error 9 0.13048 F Value Pr > F C Total 10 0.31191 12.51 0.0063 Parameter Estimate F Value Pr > F Intercept 1.163256 DROOT -0.012029 12.51 0.0063 Table 4. Regression results for percentage pbove ground growth. Source 9: SS r-Sggare Model 1 0.11140 0.38 Error 9 0.1767? F Value Pr > F C Total 10 0.02881 5.67 0.0411 Parameter Estimate F Value Pr > F Intercept 0.85379 DSOIL -0.00834 5.67 0.0411 Table 5. Regression results for percentage stolon growth on Russet Burbank. Source pi SS r-Sguare Model 1 0.94017 0.48 Error 9 1.00832 F Value Pr > F C Total 10 1.94850 8.39 0.0177 Parameter Estimate F Value Pr > F Intercept 1.325927 DROOT -0.027383 8.39 0.017? Table 6. Regression results for below ground partitioning in aldicarb treatments on Russet Burbank. Source 95 SS r-Sguare Model 1 0.05414 0.88 Error 10 0.00643 F Value Pr > F C Total 11 0.06058 70.72 0.0001 Parameter Estimate F Value Pr > F Intercept -0.001133 1(DAP2 201.0538 70.72 0.0001 Table ?. Regression results for below ground partitioning in check treatments on Russet Burbank. Source 9: SS r-Sguare Model 1 0.07970 0.91 Error 10 0.00??? F Value Pr > F C Total 11 0.08748 102.54 0.0001 Parameter Estimate F Value Pr > F Intercept -0.005222 1(DAP2 246.3776 102.54 0.0001 Table 8. Regression results for above ground partitioning in aldicarb treatments on Russet Burbank. Source pg SS r-Sguare Model 2 0.48155 0.87 Error 8 0.07002 F Value Pr > F C Total 10 0.55158 27.51 0.0003 Parameter Estimate F Value Pr > F Intercept 0.783252 DAP2 -0.000050 54.99 0.0001 DROOT 0.006510 5.40 0.0486 198 Table 9. Regression results for above ground partitioning in check treatments on Russet Burbank. Source 9: SS r-Sgpare Model 2 0.56562 0.88 Error 8 0.09931 F Value Pr > F C Total 10 0.66494 22.78 0.0005 Parameter Estimate F Value Pr > F Intercept 0.696241 DAP2 -0.000054 45.15 0.0001 DROOT 0.009451 8.03 0.0220 Table 10. Regression results for tuber partitioning in aldicarb treatment on Russet Burbank. Source pg SS r-Sgpare Model 2 0.72728 0.91 Error 8 :0.06880 F Value Pr > F C Total 10 0.79608 42.28 0.0001 Parameter Estimate F Value Pr > F Intercept ~0.286863 DAP 0.006744 19.28 0.0023 DSOIL 0.006370 3.05 0.1189 Table 11. Regression results for tuber partitioning in check treatments on Russet Burbank. Source pf SS r-Sguare Model 2 0.85890 0.91 Error 8 0.08736 F Value Pr > F C Total 10 0.94626 39.33 0.0001 Parameter Estimate F Value Pr > F Intercept -0.278207 DAP 0.006394 16.65 0.0061 DSOIL 0.009343 5.17 0.0526 Table 12. Regression results for above ground partitioning in aldicarb treatments w/o nematode parameters on Russet Burbank. Source _5 SS r-Sgpare Model 1 0.59569 0.84 Error 10 0.11756 F Value Pr > F C Total 11 0.71326 50.67 0.0001 Parameter Estimate F Value Pr > F Intercept 0.899615 DAP2 -0.000044 50.67 0.0001 Table 13. Regression results for above ground partitioning in check treatments w(o nematode parameters on Russet Burbank. Source _5 SS r-Sguare Model 1 0.62309 0.76 Error 10 0.20001 F Value Pr > F C Total 11 0.82310 31.15 0.0002 Parameter Estimate F Value Pr > F Intercept 0.861812 DAP2 -0.000045 31.15 0.0002 Table 14. Regression results for tuber partitioning in aldicarb treatments w/o nematode parameters on Russet Burbank. Source _5 SS r-Sguare Model 1 0.90453 0.90 Error 10 0.09504 F Value Pr > F C Total 11 0.9995? 95.1? 0.0001 Parameter Estimate F Value Pr > F Intercept -0.330985 DAP 0.008797 95.17 0.0001 Table 15. Regression results for tuber partitioning in check treatments w/o nematode pprameters on Russet Burbank. Source pg 55 r-sgpare Model 2 0.90453 0.90 Error 8 0.09504 F Value Pr > F C Total 10 0.9995? 95.17 0.0001 Parameter Estimate F Value Pr > F Intercept -0.330985 DAP 0.008797 95.17 0.0001 200 NICHIGQN STRTE UNIV. LIBRQRIES lllllHlHllllllllllllllllllllllllHHIIHIWHI 31293005635374