<%BANNER%>

Effects of a Prescribed Fire on Soil Nutrient Pools in the Pine Rockland Forest Ecosystem

Permanent Link: http://ufdc.ufl.edu/UFE0043515/00001

Material Information

Title: Effects of a Prescribed Fire on Soil Nutrient Pools in the Pine Rockland Forest Ecosystem
Physical Description: 1 online resource (287 p.)
Language: english
Creator: Nguyen, Chung T
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: calcareous -- fire -- florida -- phosphorous -- pine -- prescribed -- rocklands -- slash -- soils -- solubility -- south
Interdisciplinary Ecology -- Dissertations, Academic -- UF
Genre: Interdisciplinary Ecology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The Pine Rockland forest originating from limestone substrates is listed as one of the most endangered ecosystems in the United States, and has harbored the South Florida slash pine. Changes of fire regime and landscape fragmentation have caused a declination of Pine Rockland forest ecosystem. The objectives of this research were to (1) determine nutrient limitation and effects of prescribed fire on nutrient availability; (2) determine impacts of fire intensity and soil moisture on soil nutrient pools; and (3) predict availability of P after a fire. Results of analysis of foliar nutrient contents and DRIS indices indicated that P is the most limiting factor and potassium is the second limiting nutrient in the Pine Rockland forest; whereas, nitrogen is in marginal to limitation. Prescribed fire significantly increased soil pH, EC, and extractable P, Mg, K, Mn after 14 days, and extractable Fe and Ca in 270 and 360 days after the fire respectively, but not for extractable Cu and Zn. Soil NH4+ pool significantly increased immediately after the fire, whereas the soil NO3- pool was increased in post-fire 180 days. Fire intensity significantly decreased total contents of C, N, and K in residual ash, but not for total content of P, Ca, Mg, Fe, Mn, and Zn. The fire intensity significantly impacted soil pH, EC, and soil extractable contents of PO4-P, NO3-N, NH4-N, Mg, K, Fe, Mn, Cu, and Zn. However,soil moisture after a fire significantly changed soil pH, EC, and soil extractable NH4-N, NO3-N, PO4-P, Fe, and time after a fire only significantly changed extractable content of Mn. HPO42-, H2PO4-, FeHPO4 (aq), MgHPO4 (aq), CaHPO4 (aq), MnHPO4 (aq), FeH2PO4+, CaH2PO4+, and CaPO4- were major compounds of P in the soil solution in the Pine Rockland. Prescribed fire significantly increased extractable concentrations of these P compounds after14 days, except for FeH2PO4+. Within a relatively low P availability, solubility of P in the soil solution was controlled by vivianite and MnHPO4, and undersaturated with Ca/Mg-P minerals. Changes of ionic activities in the soil solution following the fire shifted equilibrium of P between vivianite and MnHPO4.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Chung T Nguyen.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Li, Yuncong.
Local: Co-adviser: Munoz-Carpena, Rafael.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043515:00001

Permanent Link: http://ufdc.ufl.edu/UFE0043515/00001

Material Information

Title: Effects of a Prescribed Fire on Soil Nutrient Pools in the Pine Rockland Forest Ecosystem
Physical Description: 1 online resource (287 p.)
Language: english
Creator: Nguyen, Chung T
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: calcareous -- fire -- florida -- phosphorous -- pine -- prescribed -- rocklands -- slash -- soils -- solubility -- south
Interdisciplinary Ecology -- Dissertations, Academic -- UF
Genre: Interdisciplinary Ecology thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: The Pine Rockland forest originating from limestone substrates is listed as one of the most endangered ecosystems in the United States, and has harbored the South Florida slash pine. Changes of fire regime and landscape fragmentation have caused a declination of Pine Rockland forest ecosystem. The objectives of this research were to (1) determine nutrient limitation and effects of prescribed fire on nutrient availability; (2) determine impacts of fire intensity and soil moisture on soil nutrient pools; and (3) predict availability of P after a fire. Results of analysis of foliar nutrient contents and DRIS indices indicated that P is the most limiting factor and potassium is the second limiting nutrient in the Pine Rockland forest; whereas, nitrogen is in marginal to limitation. Prescribed fire significantly increased soil pH, EC, and extractable P, Mg, K, Mn after 14 days, and extractable Fe and Ca in 270 and 360 days after the fire respectively, but not for extractable Cu and Zn. Soil NH4+ pool significantly increased immediately after the fire, whereas the soil NO3- pool was increased in post-fire 180 days. Fire intensity significantly decreased total contents of C, N, and K in residual ash, but not for total content of P, Ca, Mg, Fe, Mn, and Zn. The fire intensity significantly impacted soil pH, EC, and soil extractable contents of PO4-P, NO3-N, NH4-N, Mg, K, Fe, Mn, Cu, and Zn. However,soil moisture after a fire significantly changed soil pH, EC, and soil extractable NH4-N, NO3-N, PO4-P, Fe, and time after a fire only significantly changed extractable content of Mn. HPO42-, H2PO4-, FeHPO4 (aq), MgHPO4 (aq), CaHPO4 (aq), MnHPO4 (aq), FeH2PO4+, CaH2PO4+, and CaPO4- were major compounds of P in the soil solution in the Pine Rockland. Prescribed fire significantly increased extractable concentrations of these P compounds after14 days, except for FeH2PO4+. Within a relatively low P availability, solubility of P in the soil solution was controlled by vivianite and MnHPO4, and undersaturated with Ca/Mg-P minerals. Changes of ionic activities in the soil solution following the fire shifted equilibrium of P between vivianite and MnHPO4.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Chung T Nguyen.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Li, Yuncong.
Local: Co-adviser: Munoz-Carpena, Rafael.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0043515:00001


This item has the following downloads:


Full Text

PAGE 1

1 EFFECTS OF A PRESCRIBED FIRE ON SOIL NUTRIENT POOLS IN THE PINE ROCKLAND FOREST ECOSYSTEM By CHUNG TAN NGUYEN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2011

PAGE 2

2 2011 Chung T. Nguyen

PAGE 3

3 To all those who supported and inspired me these past four years

PAGE 4

4 ACKNOWLEDGMENTS I would like to sincerely thank to my major advisor, Dr. Yuncong Li, and to my co advisor, Dr. Rafael Muoz Carpena, who instructed and supported me during my Ph.D. degree. Their thoughtfulness, inspiration, and constant guidance were cornerstones in the t imely completion of my Ph.D. program. m grateful to the supervisory services received from the members of my Ph.D. committee: Dr. James Snyder for getting the permission and his guidance on the establishment of the field burning experiment and sample col lection as well as valuable comments on reviewing my dissertation ; Dr. Bruce Schaffer for his guidance and knowledge on proposal, data statistical analyse s and presentation of research results; Dr. Nicholas Comerford for his advice on selection of suitabl e methods for analysis of soil chemical components in the laboratory. Particularly, I wish to have my sincere thanks to Dr. Kati Migliaccio for her guidance support, and assistance i n the establishment of the laboratory experiments for determination of soil moisture curve as well as calculation of soil moisture content I would like to thank to the technicians at the Tropical Research and Education Center (TREC) in Homestead, FL namely; Quigin Yu, Rosado Laura, Tina Dispenza, Michael Gutierrez, Chunfang Li, and David Li, who instructed me to set up the laborato ry experiment a nd constantly helped me to analyze soil and plant tissue samples in the laboratory I also wish to thank to Dr. Qingren Wang, Dr. Guodong Liu, and Dr. Xiaohui Fan Post Doctoral Associate at the Soil and Water Science laboratory at TREC for their guidance and patience during the time I learned to use the instruments for measurement of nutrient elements in the laboratory Finally, I would like to thank to the Vietna m Education Foundation for granting me the two year f ellowship. I wish particularly to thank to Dr. Stephen R. Humphrey, Director of Academic Programs of School of Natural Resources and Env ironment, University of Florida, for awarding me the match assistantship in my whole Ph.D. program.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ ............. 4 LIST OF TABLES ................................ ................................ ................................ ......................... 9 LIST OF FIGURES ................................ ................................ ................................ ..................... 11 LIST OF ABBREVIATIONS ................................ ................................ ................................ ...... 1 4 ABSTRACT ................................ ................................ ................................ ................................ .. 1 7 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ ................. 1 9 Pine Rockland Ecosystem ................................ ................................ ................................ ...... 1 9 Fire Intensity and Severity ................................ ................................ ................................ ..... 2 1 Ecological Role of Prescribed Fire in the Pine Rockland Forest ................................ ........... 2 3 Effects of Prescribed Fire on Soil Properties ................................ ................................ ......... 2 5 Soil Physical Properties ................................ ................................ ................................ ... 2 6 Chemical Properties ................................ ................................ ................................ ......... 2 7 Soil pH and EC ................................ ................................ ................................ .......... 2 7 Soil Organic Carbon ................................ ................................ ................................ .. 2 9 Nitr ogen ................................ ................................ ................................ ..................... 30 Phosphorus ................................ ................................ ................................ ................. 3 3 Other Nutrient s ................................ ................................ ................................ ........... 3 5 Research Rationale and Objectives ................................ ................................ ........................ 37 2 PHOSPHORUS LIMITATION IN THE PINE ROCKLAND FOREST ECOSYSTEM AND NUTRIENT AVAILABILITY AFTER PRESCRIBED FIRE ................................ .... 4 5 Materials and Methods ................................ ................................ ................................ ........... 50 Description of Study Site ................................ ................................ ................................ 50 Determination of P Limitation and C:N:P ratio ................................ ............................... 5 1 Field Burning Experiment and Sample Collection ................................ .......................... 5 2 Sample Preparation ................................ ................................ ................................ .......... 5 3 Chemical Analysis ................................ ................................ ................................ ........... 5 3 Statistical Analysi s ................................ ................................ ................................ ........... 5 4 Results and Discussion ................................ ................................ ................................ .......... 5 5 Phosphorus l imitation in Pine Foliage and Fuel L oad ................................ ..................... 5 5 C:N:P Ratios in Pine Foliage, Fuel Load, and Soil ................................ .......................... 5 6 Pools of C and N in Fuel Load after Fire ................................ ................................ .......... 5 7 Changes of Soil pH and EC after Fire ................................ ................................ ............. 5 8 Changes of Soil C, N and P Pools after Fire ................................ ................................ ... 5 9

PAGE 6

6 Effects of Fire on other Soil Nutrient s ................................ ................................ ............. 60 Summary ................................ ................................ ................................ ................................ 6 2 3 BURNING TEMPERATURE AND SOIL MOISTURE AFFECTED NUTRIENT POOLS IN CALCAREOUS SOILS UNDER THE PINE ROCKLAND FOREST .............. 7 3 Materials and Methods ................................ ................................ ................................ ........... 7 8 Incubation Experiment ................................ ................................ ................................ ..... 7 8 Soil Water Field Capacity ................................ ................................ ................................ 80 Estimation of Field Fire Temperature ................................ ................................ ............. 80 Chemical Analysis ................................ ................................ ................................ ........... 8 1 Statistical Analysis ................................ ................................ ................................ ........... 8 1 Results and Discussion ................................ ................................ ................................ ........... 8 2 Biomass and Total Nutrients in Residual Ashes ................................ .............................. 8 2 pH, EC, Water Soluble Nutrients in Residual Ashes ................................ ...................... 8 4 Interactions of Soil Nutrient Pools ................................ ................................ ................... 8 6 Effects of Burn Temperature on Soil Nutrient Pools ................................ ....................... 88 Effects of Soil Water Content on Soil Nutrient Pools ................................ ..................... 90 Effects of Incubation Time on Soil Nutrient Pools ................................ .......................... 92 Estimation of Field Fire Temperature from Residual Ashes ................................ ............ 92 Summary ................................ ................................ ................................ ................................ 9 4 4 PREDICTING PHOSPHORUS AVAILABILITY IN CALCAREOUS SOILS UNDER THE PINE ROCKLAND FOREST AFTER A PRESCRIBED FIRE .................. 10 6 Materials and Methods ................................ ................................ ................................ ......... 11 1 Simulation of Extractable Concentration s of P Compounds after the Fire .................... 11 1 Construction of Stabilit y Diagrams of Phosphorus after the Fire ................................ .. 11 2 Soil Moisture Curve for the Pine Rockland Soils ................................ .......................... 11 3 Model for Pred iction of P Availability after the Fire ................................ ..................... 1 14 Statistical Analysis ................................ ................................ ................................ ......... 11 7 Results and Discussion ................................ ................................ ................................ ........ 11 7 Phosphorus Speciation after the Fire ................................ ................................ ............. 11 7 Soil Solution Chemistry after the Fire ................................ ................................ ............ 11 8 Phosphate Phase Equilibria after the Fire ................................ ................................ ....... 1 20 Modeling a Soil Water Characteristic Curve for the Pine Rockland forest .................... 1 22 Model ing for Pred iction of P Availability after the Fire ................................ ................ 12 4 Summary ................................ ................................ ................................ ............................... 12 5 5 CONCLUSIONS ................................ ................................ ................................ ................... 1 3 9 APPENDIX A STUDY SITE AND SAMPLING PROCEDURES ................................ ............................. 1 4 9

PAGE 7

7 B P ROCEDURES OF CHEMICAL ANALYSIS AND QA/QC METHODS IN LABORA T ORY ................................ ................................ ................................ ................... 1 50 pH ................................ ................................ ................................ ................................ ......... 1 50 E lectrical C onductivity ................................ ................................ ................................ ........ 1 50 Ammonium ................................ ................................ ................................ .......................... 15 1 Nitrate ................................ ................................ ................................ ................................ .. 15 3 Extractable P hosphorus ................................ ................................ ................................ ........ 15 4 Extractable M etals ................................ ................................ ................................ ............... 1 5 8 Total C arbon and T otal N itrogen ................................ ................................ ......................... 16 2 Total P hosphorus and T otal M et al ................................ ................................ ....................... 16 3 C CALIBRATION CURVES OBTAINED FROM LABORATORY ANALYSES, AND STANDARD SOLUTIONS USED FOR METAL MEASUREMENT .............................. 165 D COVARIANCE STRUCTURE MODELS AND ANOVA RESULTS BY REPEA TED MEASURES ANALYSIS, AND RESULTS ........................... 167 Results of ANOVA a nalysis by omparison .................... 167 Ca 2+ ................................ ................................ ................................ ................................ 167 Mg 2+ ................................ ................................ ................................ ................................ 169 K + ................................ ................................ ................................ ................................ ..... 170 Fe 2+ ................................ ................................ ................................ ................................ .. 171 Mn 2+ ................................ ................................ ................................ ................................ 173 Zn 2+ ................................ ................................ ................................ ................................ 174 Cu 2+ ................................ ................................ ................................ ................................ .. 175 PO4 P ................................ ................................ ................................ ............................... 177 NH4 N ................................ ................................ ................................ .............................. 178 NO3 N ................................ ................................ ................................ .............................. 179 EC ................................ ................................ ................................ ................................ .... 181 pH ................................ ................................ ................................ ................................ ..... 182 TP (%) ................................ ................................ ................................ .............................. 183 TN (%) ................................ ................................ ................................ ............................. 185 TC (%) ................................ ................................ ................................ ............................. 186 Soil C:N ratio ................................ ................................ ................................ ................... 187 Soil N:P ratio ................................ ................................ ................................ .................... 189 Soil C:P ratio ................................ ................................ ................................ .................... 190 E PROCEDURES FOR THE LABORATORY INCUBATION EXPERIMENT .................. 192 F PROCEDURES FOR DETERMINATION OF SOIL FIELD CAPACITY ....................... 19 4 G RESULTS OF MULTIPLE REGRESSION ANALYSES BY PROC GLM PROCEDURES FOR THE LABORATORY INCUBATION EXPERIMENT .................. 195

PAGE 8

8 H RESULTS OF LINEAR REGRESSION ANALYSI S ON ASH pH, EC, TOTAL AND WATER SOLUBLE NUTRIENTS IN LABORAROTY RESIDUAL ASHES .................. 20 7 I RESULTS OF FITTING REGRESSION MODELS AND ANOVA ANALYSIS ON TP, TCa, TFe, T Mg IN RESIDUAL ASHES ................................ ................................ ....... 214 Results of f itting of regression m odels with 1 st 2 nd and 3 rd order among b urn temperature and c ontent of TP, TCa, TFe, or TMg in l aboratory s imulation a shes ............ 21 4 J MODEL .................... 22 1 K CALCULATION OF STOICHIOMETRY FOR ACTIVITY OF HPO 4 2 FROM Ca/Mg/Fe/Mn P MINERALS ................................ ................................ .............................. 22 3 L EXPERIMENT FOR DETERMINATION OF SOIL WATER CURVE ........................... 22 7 Procedures for the e xperiment ................................ ................................ ............................. 227 M VOLUMETRIC WATER CONTENTS OF PINE ROCKLAND SOIL OBTAINED FROM OBSERVATION EXPERIMENT AND FROM RETENTION CURVE MODEL ............ 230 N RESULTS OF FITTING OF VOLUMETRIC WATER CONTENT BY RETC MODEL 23 1 O RESULTS OF FITTING OF MULTIPLE REGRESSION MODELS F OR PREDICTION OF P AVAILABILITY AFTER THE FIRE ................................ ............... 24 7 P CHECKING OF ASSUMPTION S FOR THE SELECTED PREDICTIVE MODEL ........ 26 2 Q ROCKLAND SOIL S AFTER THE FIRE ................................ ................................ ........... 26 4 LIST OF REFERENCES ................................ ................................ ................................ ........... 26 7 BIOGRAPHICAL SKETCH ................................ ................................ ................................ ..... 28 7

PAGE 9

9 LIST OF TABLES Table page 2 1 A summary of analytical methods in laboratory ................................ ............................... 6 3 2 2 Thresholds of foliar N and P limitations in upland ecosystems ................................ ....... 6 3 2 3 Critical foliar N contents of slash pine forests ( Pinus elliottii var elliottii ) ..................... 6 4 2 4 Critical foliar P contents and N:P ratio of slash pine forests ................................ ........... 6 4 2 5 DRIS indices of N, P, K, Ca, and Mg in the South Florida slash pine ............................ 6 5 2 6 Nutrient contents of pine foliage and fuel load in the Pine Rockland forest ................... 6 5 2 7 Loss of biomass and pools of C and N in forest floor and understory vegetation (fuel load) due to fire ................................ ................................ ................................ ....... 6 6 2 8 Results of ANOVA analysis of soil nutrient pools following the prescribed fire ........... 6 7 2 9 times after the prescribed fire ................................ ................................ .......................... 6 7 3 1 Results of linear regression analysis on pH, EC, and total and water soluble nutrients in laboratory residual ashes ................................ ................................ ............... 9 6 3 2 R esults of multiple regression analysis on incubation time, soil moisture, burn t emperature and their interactions ................................ ................................ .................... 96 3 3 Estimation of a field fire temperature from laboratory and field residual ashes ............ 9 7 4 1 Average extracted concentration of soil nutrient elements, pH, and ionic strength after the fire used model ................................ .......................... 1 2 7 4 2 Solubility product constants and stoichiometry of Ca/Mg/Fe/Mn P minerals .............. 128 4 3 Results of ANOVA analysis and simulated concentrations of P compounds associated with Ca, Mg, Fe, and Mn after the fire ................................ ......................... 1 28 4 4 Results of ANOVA analysis and ionic activities of elements after the fire obtained f rom ................................ ................................ .................... 1 29 4 5 Correlation matrix of ionic activities of nutrient speciation program (n = 45) ................................ ................................ ........................... 1 29

PAGE 10

10 4 6 Stoichiometry between activity of HPO 4 2 and activities of Ca 2+ Fe 2+ Mg 2+ Mn 2+ for selected Ca/Mg/Fe/Mn P minerals ................................ ................................ .......... 1 30 4 7 Correlation relationships among pH + pHPO 4 2 pH 1/2pFe 2+ pH 1/2pMn 2+ and pH 1/2pCa 2+ in soil solution (n = 45) ................................ ................................ ... 1 30 4 8 Saturation Index (SI) for phosphate minerals in the Pine Rockland soils following the fire obtained from simulation results of Minteq speciation program ............................. 13 1 4 9 Fitted parameters of Retc model used to describe soil water characteristics of the Pine Rockland calcareous soils, and coefficients used to evaluate a goodness of fit of model ................................ ................................ ............................. 13 1 4 10 A summary of parameters of regression models fitted by response variable of P availability and three predictors ................................ ................................ ..................... 13 2 4 11 Pine Rockland forest ................................ ................................ ................................ ...... 13 3 4 12 Observation and simulation data of extractable P content after the fire at the ................................ ................................ ........................ 133 C 1 Standard solutions and detection limits used for metal measurement ............................. 166 F 1 Results of determination of soil field capacity ................................ ............................... 194 I 1 Results of ANOVA analysis on TP/TCa, TP/TMg, TP/TFe in residual ashes ............... 214 M 1 Results of volumetric water content fitted by Retc model ................................ .............. 230

PAGE 11

11 LIST OF FIGURES Figure page 1 1 Pine Rockland forest ecosystem ................................ ................................ ....................... 39 1 2 Temperature thresholds of typical soil components ................................ ......................... 40 1 3 Pool of soil organic matter ................................ ................................ ................................ 40 1 4 Variations of soil extractable N to fire severity gathered by eight studies ....................... 4 1 1 5 Pos s ible pathways relating to fate of NH 4 + and NO 3 after a fire ................................ .... 4 2 1 6 Pos s ible pathways relating to fate of ortho phosphate after a fire ................................ ... 4 3 1 7 Possible pathways relating to fate of nutrient cations after a fire ................................ .... 4 4 2 1 Blocks of Pine Rockland in the Long Pine Key ................................ .............................. 6 8 2 2 C:N ratios in soil, pine foliage, and fuel load of the Pine Rockland forests .................... 69 2 3 N:P ratios in soil, pine foliage, and fuel load of the Pine Rockland forests ..................... 69 2 4 C:P ratios in soil, pine foliage, and fuel load of the Pine Rockland forests ...................... 70 2 5 Rainfall and changes of pH and EC along with rainfall after the fire .............................. 71 2 6 Changes of soil N:P ratio by following the fire ................................ ............................... 7 2 2 7 Changes of soil C:P ratios by follow ing the fire ................................ .............................. 7 2 3 1 Losses of TC, TN, and biomass in fuel loads at different heating temperatures ............. 9 8 3 2 Effects of burn temperature on total K in residual ash in term of dry fuel load .............. 9 8 3 3 Effects of burn temperatures on water soluble nutrients in residual ash in term of dry fuel load ................................ ................................ ................................ ..................... 99 3 4 The interaction of burn temperature and incubation time on soil extractable Mn ........... 99 3 5 The interaction of soil moisture and incubation time on so il extractable nutrients ........ 100 3 6 The interaction of soil moisture and burn temperature on soil extractable nutrients ...... 10 1 3 7 Effects of burn temperature on extractable concentration of soil nutrients ................... 10 2

PAGE 12

12 3 8 Effects of soil moisture on extractable co ncentrations of soil nutrients ......................... 10 3 3 9 Effects of incubation time on extractable concentration of soil nutrients ...................... 10 4 3 10 Concentration s of TP, TCa, TMg, and TFe in field and laboratory ashes ..................... 10 5 3 11 Goodness of fit models for fire temperature prediction based on TP, TCa, TMg, and TFe in residual ash ................................ ................................ ................................ .. 10 5 4 1 Concentrations of ortho P species after the fire, relatively Olsen P contents ............... 13 4 4 2 Relationship of pH and phosphate mineral stability in soil solution before fire ........... 134 4 3 Relat ionship of pH and phosphate mineral stability in 14 days after fire ...................... 13 5 4 4 Relationship of pH and phosphate mineral stability in 30 days after fire ...................... 13 5 4 5 Relationship of pH and phosphate mineral stability in 90 days after fire ...................... 13 6 4 6 Relationship of pH and phosphate mineral stability in 180 days after fire .................... 13 6 4 7 Soil moisture curve of the Pine Rockland fitted by Retc model ................................ .... 13 7 4 8 A goodness of fit of model for soil moisture curve of the Pine Rockland forest .......... 137 4 9 A goodne ss of fit for the predictive model of P availability after the fire .................... 1 38 4 10 Verification of a predictive model for prediction of P availability after the fire ........... 1 38 A 1 Pine Rockland forest in the Long Pine Key, Everglades National Park ......................... 149 A 2 A 50 x 50cm sampling frame used to collect ashes after the fire ................................ ... 149 C 1 Calibration chart and curve obtained from NH 4 N measurement by AQ2 ..................... 165 C 2 Calibration chart and curve obtained from NO 3 N measurement by AQ2 ..................... 165 C 3 Standard solutions and calib ration curve obtained from P measurement ....................... 166 D 1 Covariance structure models applied for ANOVA analysis by repeated measures ........ 167 E 1 A 30 x 30cm sampling frame used to collect fuel load and soil samples ....................... 192 E 2 Procedures for preparation. a) soils, b ) fuel loads, c) bottles, d) muffle furnace ........... 192 E 3 Ash production at different heating temperatures ................................ ........................... 193

PAGE 13

13 E 4 Procedures for incubation. a) mixing ash products, b) applying ashes into samples, c) applying DI water for soil water content, and d) incubating oil samples ....................... 193 J 1 ................................ ................................ ... 221 J 2 b) possible solid species, and c) main solution components ................................ ............ 2 22 L 1 Establishment of the experiment: (a) Richard Plates and (b) Tempe Cells .................... 227 P 1 Plot of predicted values versus residuals ................................ ................................ ......... 262 P 2 Normal probability plot of residuals ................................ ................................ ............... 263

PAGE 14

14 LIST OF ABBREVIATIONS AB DTPA Ammonium Bicarbonate Diethylene Triamine Pentaacetic Acid C Carbon CCB Continuing Calibration Blank CCV Continuing Calibration Verification DCP Dicalcium Phosphate DCPD Dicalcium Phosphate Dihydrate DDI Water Double Deonized Water DI Water Deonized Water DNRA Dissimilation Nitrate Reduction to Ammonium DRIS Diagnosis and Recommendation Integrated System EC Electrical Conductivity EDTA Ethylene Diamine Tetraacetic Acid ET Evapotranspiration FAWN Florida Automated Weather Network HA Hydroxyapatite HID Hole in the Donut ICB Initial Calibration Blank ICV Initial Calibration Verification IS Ionic Strength LOI Loss of Ignition MCP Monocalcium Phosphate MSE Mean Standard Error

PAGE 15

15 N Nitrogen NEDD N (1 napththyl) Ethelenediamine Dihydrochloride NPP Normal Probability Plot OCP Octacalcium Phosphate OM Organic Matter P Phosphorus QA Quality Assurance QC Quality Control SI Saturated Index SOM Soil Organic Matter TAl Total Aluminum TC Total Carbon TCa Total Calcium TCP Tricalcium Phosphate TCu Total Copper TFe Total iron TK Total Potassium TM Total Metal TMg Total Magnesium TMn Total Manganese TN Total Nitrogen TP Total Phosphorus TREC Tropical Research and Education Center

PAGE 16

16 TZn Total Zinc USDA U.S Department of Agriculture USEPA U.S. Environmental Protection Agency USGS U.S. Geological Survey

PAGE 17

17 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EFFECTS OF A PRESCRIBED FIRE ON SOIL NUTRIENT POOLS IN THE PINE ROCKLAND FOREST ECOSYSTEM By Chung T. Nguyen December 2011 Chair: Yuncong Li Co Chair: Rafael Mu oz Carpena Major: Interdisciplinary Ecology The Pine Rockland forest originating from limestone substrates is listed as one of the most endangered ecosystems in the United States and has harbored the South Florida slash pine Changes of fire regime and landscape fragmentation have caused a declination of Pine Rockland forest ecosystem T he objectives of this research were to (1) determine nutrient limitation and effects of prescribed fire on nutrient availability; (2) det ermine impacts of fire intensity and soil moisture on soil nutrient pools; and (3) predict availabilit y of P after a fire Results of analysis of foliar nutrient contents and DRIS indices indicated that P is the most limiting fac tor and potassium is the second limiting nutrient in the Pine Rockland forest ; whereas, nitrogen is in marginal to limitation. Prescribed fire significantly increased soil pH, EC, and extractable P, Mg, K, Mn after 14 days, and extractable Fe and Ca in 270 and 360 days after the fire respectively, but not for extractable Cu and Zn. Soil NH 4 + pool significantly increased immediately after the fire, whereas the soil NO 3 pool was increased in post fire 180 days. F ire intensity significantly decreased total contents of C, N, and K in residual ash, but not fo r total content of P, Ca, Mg, Fe, Mn, and Zn. The fire intensity significantly impacted soil pH,

PAGE 18

18 EC and soil extractable contents of PO 4 P, NO 3 N NH 4 N Mg, K, Fe, Mn, Cu and Zn. However, soil moisture after a fire significantly changed soil pH, EC, and soil extractable NH 4 N, NO 3 N, PO 4 P Fe and t ime after a fire only significantly changed extractable content of Mn. HPO 4 2 H 2 PO 4 FeHPO 4 (aq), MgHPO 4 (aq), CaHPO 4 (aq), MnHPO 4 (aq), FeH 2 PO 4 + CaH 2 PO 4 + and CaPO 4 were major compounds of P in the soil solution in th e Pine Rockland Prescribed fire significantly increased extractable con centrations of these P compounds after 14 days, except for FeH 2 PO 4 + Within a relatively low P avail ability, solubility of P in the soil solution was controlled by vivianite and MnHPO 4 and undersaturated with Ca/Mg P mineral s Changes of ionic activities in the soil solution following the fire shifted equilibrium of P between vivianite and MnHPO 4

PAGE 19

19 CHAPTER 1 INTRODUCTION Pine Rockland Ecosystem The Pine Rockland is a unique upland ecosystem originated from limestone substrates in South Florida and is one of the most endangered forest types in the world ( Florida Natural Area Inventory, 2008 ). In addition to three intact remna nt regions in South Florida (Figure 1 1a) including the Long Pine Key in the Everglades National Park, the Big Pine Key in the Florida Lower Keys, and a part of the Big Cypress National Preserve the Pine Rockland is also foun d in the Bahamas and Cuba (Snyder 1986; U.S Fish and Wildlife Service, 2007) Soils of t he South Florida Pine R ockland are characterized in the e ntisols o rder which is poorly developed mineral soil without natural horizons and absent of distinct pedogenic horizons ( Collins, 2002 ). P resence of fragments of limestone ( calcium carbonate ) in the soil of the Pine Rockland creates high soil pH ranging from 7 .4 to 8 .4 (Li, 2001). The soil surface is often irregular with solution holes accum mulated by organic matter and humus. Soils in the surface layer range from dark grayish brown to brown fine sands or fine sandy loams whereas the subsurface layers are light gray to yellowish red fine sand and brown to reddish brown sandy or clay loams (Sn yder, 1986). Soil d rainagel often occurs quickly due to the poros ity of limestone substrates; however, many areas of Pine Rockland are wet for a short period after heavy rains. During the rainy season, some of the P ine Rockland areas may be shallowly in undated by very slow flowing surface water (U.S. Fish and Wildlife Service, 2007 ). The Pine Rockland forest is a monospecific stand of the South Florida slash pine (Pinus elliottii var. densa ) with a diverse understory of palms, hardwoods, and herbs (Snyder, 1986) Pinus elliottii var. densa differs from the typical variety which is known as Pinus elliottii var. elliottii not only in geographical location, but in seedling development and wood density as well

PAGE 20

20 (Newton et al., 2005). The South Florida slash pine has longer needles and smaller cones than the typical slash pine variety. A dditionally, it has a denser wood and a thicker, longer taproot than those of the typical slash pine variety (Hill, 2001) It can also be distinguished from loblolly pine or other pines by the characterist ics of its needles, cones, bark and fruit. Needles of the South Florida slash pine have brooms at ends of rough twigs. These needles are approximately from 12 to 28 cm in length. Its cones range from 12.5 to 20 cm in len gth. Plates of bark of the slash pine are large, flat and orange brown. The fruit is a dark brown woody cone that is 12 to 20 cm long (Hill, 2001). The Pine Rockland forest is listed as one of the major endangered ecosystems in the United States ( Florida N atural Area Inventory, 2008 ). It is rated as G1 on the global conservation scale, which represents the most critically imperiled habitats because of extreme rarity or extreme vulnerability to extinction due to some natural or man made factors. More than 20 % of 225 native plant species in the Pine Rockland forest are found nowhere else in the world (Ross, 1995). Several plants and animals that are n ative to the Pine Rockland environmentally endangered species list as threatened or endangered (U.S. Fish and Wildlife Service, 2007). Human activities, land use change, and natural disasters such as hurricanes have caused a habitat degradation of the Pine R ockland forest ecosystem since the late 1800s. The biggest loss of the Pine Rockland habitat occurred in Miami Dade County, where approximately 98.5% of Pine Rockland (Fairchild Tropical Batonic Garden, 2008) was converted into urban and suburban areas by land use changes during the last century (Figure 1 1b). The Pine Rockland ecosystem is a fire dependent habitat, and is maintained by fires every 3 to 10 years for continued health of endemic plants (Snyder et al., 1990). Historically, lightning was a pri mary source of fire which usually burned the understory of the Pine Rockland forest with a minimal effect on the pine canopy. Natural fires benefited the Pine Rockland

PAGE 21

21 ecosystem by increasing contents of nutrients, decreasing competition of hardwood specie s with herbaceous plants, helping germination of pine seeds (Snyder, 1986), controlling exotic species, and creating more light for understory native plan t s (U.S. Fish and Wildlife Service, 2007). However, changes in fire regime and landscape fragmentation caused by humans have made the Pine Rockland ecosystem highly vulnerable. Rock plowing during the 1950s destroyed a large area of the Pine Rockland forest in Long Pine Key, called Hole in the Donut (HID), that has promoted a favorable environment for a ra pid invasion of exotic plant species, particularly Brazilian pepper (Li and Norland, 2001). The presence of massive amounts of Brazilian pepper in the HID is a threat to undisturbed Pine Rockland habitat in the Long Pine Key. Fire Intensity and Severity A commonly accepted term used to describ e ecological effects of fire is fire severity ( Neary et al. 2005). Fire severity has been defined as the qualitative effect which fire ha s on the soil and water system, and has been used to describe individual fires and fire regimes ( Govender et al. 2006; Neary et al. 1999; Simard, 1991 ). While f ire severity is largely dependent upon the nature and characteristics of fuels available for burning, f ire intensity has been more narrowly defined as rate of energy release d by an individual fire or temperature of burning environment ( Govender et al. 2006; Neary et al. 1999 ) It is determined as I = H x W x R, where I is fire intensity, H is heat yield (kJ g 1 ), W is mass of fuel combusted (g m 2 ), and R is spread rate of heat fire front (m s 1 ) (Bilgili et al. 2003; DeBano et al. 1998 ). Fire regime describes general characteristics of fire that occurs within a specific vegetation type or ecosystem. Three basic kinds of fire regime expressing levels of fire severity are understory fire regime, stand replacement, and mixed fire regime (Neary et al., 2005). Of fire behavior characteristics, spread rate of fire intensity is related to both fuel loading and fuel

PAGE 22

22 moisture (Bilgili et al., 2003). The d irect ef fect of fire on below ground systems is a result of fire severity, which integrates above ground fuel loading ( a live and dead), soil moisture and subsequent soil temperature, and fire duration T he t ime for recovery of below ground systems not only depends on burning intensity, but also on previous disturbance and recruitment of organisms (Neary et al., 1999). The season of the fire influences rates of fuel consumption and fire intensity (Sparks et al., 2002), and a short fire interval drastically decreases biomass (Enslin et al., 2000; McCaw et al., 2002). Fire intensity is often affected by b urning season and fire frequency, but is not depend ent intensity on burning techniques (micro plot burn and macro plot burn) ( Smith et al. 1993) Fire can be determined by some different methods Henig Sever et al. ( 2001 ) estimated wildfire intensity bas ed on a sh pH and the soil micro arthropod community. They found that ash accumulation and an increase of ash layer pH were dire ctly related to fire intensity, a nd there was a positive correlation between fire intensity and mini mum diameter of burned branches, whereas a negative correlation was found between size of micro arthrop od community and fire intensity Gill et al. ( 2000 ) who used PIP (Probability of Igni tion at a Point) model to determine effects of fire regimes over 15 years in the Kakadu National Park in Australia found that early dry season fires had lower intensity than late dry season fires. Govender et al. ( 2006 ) used fuel loads, fuel moisture contents, rates of fire spread and heat yields of fuel to compute fire intensity for 21 years with five different return intervals (1, 2, 3, 4 and 6 years) and five different month s per year (Feb, Apr, Aug, Oct, Dec) Their results showed that f ire season had no significant difference of fire intensity between annual burns and burns in the biennial, triennial, quadrennial categories. They also found that fuel loads were significantly different between annual, 2 3 4 year burns and increased with incre asing rainf all over two previous years

PAGE 23

23 T he vegetation community dominated by slash pine in the P ine Rockland forest c an be considered as a v egetation type with an understory fire regime (Snyder et al., 1990) Fire s in the P ine Rockland forests are surface fires that consume only litter and understory vegetation because the pine canopy is usually too open to support a crown fire (Sah et al., 2004 & 2006) Accumulation of p ine needles on or near the ground associated with the open canopy of the pine t rees allows for rapid drying of fuels that is favorable for fires (Snyder, 1986). F ires d id not result in significant changes in species composition in the Pine Rockland forest because almost all species are perennials that survive and recover in place (Gu nderson and Snyder 1997) Ecological Role of Prescribed Fire in the Pine Rockland Forest Fire is a dynamic process that influences composition, structure, and patterns of a vegetation community and soil (Boerner, 19 82) N ot only does fire burn vegetation, residues, and micoorganisms in above ground layers, but it also heats below ground environments ( DeBano et at. 1998 ). Fire can significantly influence physical chemical, and biological properties of soil (Certini, 2005; Neary et al., 2005). Soil properties can experience short term, long term or permanent fire induced changes, depending primarily on types of property, severity and frequency of fires, and post fire climatic conditions. Fire can cause alter a tions of soil biogeochemical cycles, consume soil organic matter, ephemerally increase soil pH, and change availability of nutrien ts in soil (Gonzlez Prez et al., 2004; Neill et al., 2007; Prieto Fernandez et al., 1993) Consequently fire can shift the competit ive balance between plant spe cies, especially between native plants and invasive plants. Prescribed fire is defined as fire that is purposely set under specific weather conditions to reduce natural buildup of vegetation and litter in a fire adapted habitat (Knapp et al., 2009; Wade e t al., 1980). It is an effective tool to decrease fuel accumulation, manage wildlife habitats,

PAGE 24

24 control an invasion of exotic plant species, and maintain natural plant communities. Prescribed fires of low to moderate intensity often promote a renovation of dominant plant species by eliminating undesired species, whereas severe fires (wildfires) have negative impacts on soil because they cause a loss of organic matter and nutrients through volatilizatio n, damage soil structure and porosity, and change microbial composition (Certini, 2005; DeBano et at., 1998; Neary et al., 2005). Fire is an important factor of natural forest ecosystems and is required to maintain and to control, at least in part, the re lative dominance of hardwood plants in understory of the Pine Rockland forest (Snyder et al., 1990). In a bsence of fire, the pine canopy is likely to be replaced by dense hardwoods, resulting in loss of the characteristic herb flora (U.S Fish and Wildlife Service, 2007). Three major sources of fire associated with the Pine Rockland forest are prescribed fire, human caused wildfire, and lightning caused fire (Gunderson and Snyder, 1997). Lightning caused fires usually occur during the rainy season from May t hrough October. In contrast, human caused wildfires mostly happen in the dry season from November through May. Prescribed burnings in the Pine Rockland forest, so far, have often been done from November to April when burning conditions are mild H istor ical ly however, the main fire season occurred during the growing season (March October) (Knapp et al., 2009). The obvious ecological impact of fires on the Pine Rockland forest is to promote rapid regrowth and to increase flowering of the native vegetation co mmunities in the understory of the pine trees (Wade et al., 1980). Prescribed fire is one of the most widespread management actions in South Florida natural areas, where the Department of the Interior m anagers are responsible for the vast majority of th e remaining Pine Rockland forests (USGS, 2000) Responses of vegetation communities in the Pine Rockland forest ecosystem with seasons and intervals of prescribed burns were reported in

PAGE 25

25 some studies. Snyder (1986) utilized prescribed fire to burn two different sites of Pine Rockland in Long Pine Key in two different seasons (dry and wet). Site 1 was higher, more frequently burned pinelands while site 2 was lower, less frequently burned pi nelands. In that study fire intensity in wet season was lower than that in dry season in both site 1 and sit e 2. By conducting summer burne and winter burne experiments in the Big Pine Key, Snyder et al. (2005) showed that summer burns had more intense f ires than winter burns. Sah et al. (2006) used char height and temperature sensitive paints to determine fuel consumption of fire in the Pine Rockland and showed that surface fuel consumption was higher in summer than in winter. In an other study, Sah et al (2010) assessed mortality of Sout h Florida slash pine after fire and reported that the rate of mortality of pine trees after a summer burn was not different from that after a winter burn. Effects of Prescribed Fire on Soil Properties Fire is a fundamenta l, dynamic process that plays a dominant role in recycling organic matter (DeBano et al., 1998) and influences the composition, structure, and pattern of the vegetation community and soil (Neary et al., 2005; Knapp et al., 2009) Immediate r esponses of soil compo n ents to a fire happen as a result of chemical release from ash produced from biomass combustion (Hernandez et al. 1997 ; Monleon et al. 1997). L ong term impact s, which are usually subtle and can persist fo r years following a fire, occur as a c onsequence of changes in soil biogeochemical cycles (DeBano et al. 1998 & 2005) or arise from a relationship between hydrology, fire, soils, nutrient cycling, and site productivity (Monleon et al. 1996 & 1997 ; Neary et al. 1999). Fire comprehensively in fluence s on soil physical, chemical, mineralogical, and biological properties. P hysical properties that are impacted by fire include soil texture, clay content, soil structure, bulk density, and porosity (Certini, 2005; DeBano et al., 2005; Ulery and Graham; 1993a). T he most common chemical components affected by fire include soil organic matter (SOM) and carbon

PAGE 26

26 (C), soil organic nitrogen soil pH bu ffer capacity, cation exchange capacity, and availabilit y of a nutrient element (Knoepp et al., 2005), whereas biological component s most affected by fire are microbial and invertebrate biomass and community composition (Busse and DeBano 2005). Figure 1 2 describes temperature thresholds of typical soil compo nents usually affected by a fire. Soil P hysical P roperties Soil texture and structure: Soil texture is defined as the relative proportion of different sized inorganic components that are found in the 0.08 inch mineral fraction of soils (Brady and Weil, 2002; DeBano, 2000b) Components of soil texture are often not impacted by fire because of relatively high temperature thresholds. Clay, the finest textural fraction, can only be changed by soil heating temperature of about 400 0 C and is completely destroyed at soil temperatures of 700 0 C to 800 0 C, but these temper atures can rarely reach beyond several centimeters below the soil surface (Certini, 2005; DeBano et al., 2005). However, humus, an essential component of soil organic matter which acts as a glue to hold mineral soil particles together to form aggregates of soil structure (Brady and Weil, 2002), can be easily broken down as a result of combustion of soil organic matter at low and moderate temperatu res (DeBano et al., 1998 ; Giovannini et al., 1988). Bulk density and porosity: Bulk density is defined as the m ass of dry soil per unit bulk volume expressed in g/cm 3 and is related to porosity which is the volume of pores in a soil sample divided by bulk volume of the sample (Brady and Weil, 2002). Heating associated with fire often causes a collapse of soil str ucture that affects both porosity and pore size distribution in the surface soil. As a result, fire increases bulk density and reduces soil porosity of organic horizons (Neill et al., 2007), and promotes rapid runoff and erosion (DeBano et al., 1998 & 2005 ). Additionally, black carbon or charcoal created at temperatures between 250 and 500 0 C from an incomplete combustion of woody residues also contribute to the increase of bulk density in soil (Giovannini et al., 1988).

PAGE 27

27 Water repellency: T he important physi cal process associated with fire in the Pine Rockland is water repellency. Heat transfer ring downward into the soil surface during a fire creates water repellency. The water repellency is often caused by changes of SOM structure resulting from drying of organic matter and coating of mineral soil particles (DeBano, 2000a & b ) Erosion is a natural process that involves three different components : detachment, transport, and deposition. A decrease in cover due to combustion of above ground vegetat ion and litter during fire promotes a higher rate of erosion (DeBano et al., 2005). Chemical Properties Fire not only directly impacts soil organic matter C, N, P S nutrient cations, pH, and electrical conductivity (EC), but it also indirectly influences the exchange of cations absorbed on the surface of mineral soil particles and humus ( DeBano et al., 1998; Knoepp et al., 2005; Neary et al., 2005). F ire induced cha nges of nutrient elements in soil solution mainly derive from changes of soil pH, cation exchange capacity, and the C/N ratio following a fire. Fire never increase s the total amount of nutrients in soil, but can transform these elements making nutrients mo re available for plant and microorganism uptake. Soil pH and EC Soil pH is one of the most i m portant factors influencin g availability of soil nutrients ;fire induced c hanges of soil pH in turn promote alteration of nutrient availability. ( DeBano et al. 1998; Ubeda et al. 2005). Nutrient cations (Ca, Mg, and K) released by combustion of fuels and organic matter tend to move down ward into soils and result in elevation of surface soil pH (Dikici and Yilmax, 2006; Murphy et al., 2006a) In g eneral, fire quic k ly incre ases soil pH resulting from combustions of SOM and plant residues, and denaturation of organic acids in soils Neverthless, the increase in soil pH only occurs when fire temperature is greater than

PAGE 28

28 450 o C and fuels are absolutely burned to crea te white ash ( Arocena and Opio 2003; Macadam, 1987; Qian et al. 2009a & b ) and i s temporary because increasing the soil pH depends on original pH buffer in soils composition and amount of cations produced, and soil moisture content ( Knoepp et al., 2005). Khanna et al. ( 1994 ) observed that capacity of a sh to neutralize soil acidity was well correlated with sum of concentration s of K, Ca, and Mg in ash By a nalyzing a series of soils developed on different lithologies in Q. engelmannii P. ponderosa and mixed conifers forest Ulery et al. ( 1993 b) found that the topsoil pH could increase as much as 3 units immediately after fire This rise was essentially due to production of K and Na oxides, hydroxides, and carbonates from the fire In contrast, fire induced changes of soil pH are negligible in calcareous soils wh ich have a high pH buffer (Motav alli et al. 1995). In addition, a nions in ash are also important for neutralizing acidity of soil. Durgin et al. ( 1984 ) showed that concentrations of anions (CO 3 2 HCO 3 OH SO 4 2 ) dominant in black ash and grey as h increased alkalinity, and that EC and concentrations of these anions were highe r in grey ash than in black ash Of these anions, CO 3 2 and HCO 3 played an essential role in neutralizing soil acidity because of their abundance in ash products. After being releas ed into soil anions associated with ash are moved downward into soil by rain water. In soil solution s these anions are hydrolyzed and neu tralize H + ions r esult ing in an increase of soil pH. H ydrolysis of CO 3 2 in water that results in releasing hydroxyl ions (OH ) (Equation 1 1) as well as OH ions produced from combustion contribute chiefly to the increase of soil pH after a fire. Manganes e oxide (MnO 2 ) and iron hydroxide (Fe(OH) 3 ) are abundant in acidic soils. Reactions between HCO 3 and MnO 2 and Fe(OH) 3 (Equations 1 2 & 1 3) consume a large amount of H + ions that causes reduction of H + activity in soil solution and considerably increases soil pH in acidic soils.

PAGE 29

29 CO 3 2 + 2H 2 O H 2 CO 3 + 2OH Equation 1 1 1 MnO 2 + HCO 3 + 3H + MnCO 3 + 2H 2 O Equation 1 2 1 Fe(OH) 3 + HCO 3 + 2H + FeCO 3 + 3H 2 O Equation 1 3 1 Soil Organic Carbon Soil organic matter, which is defined as a wide range of organic substances, comprises of intact plant and ani mal tissues, organisms, dead roots and ot her recognizable plant residues, and humus ( Brady and Weil, 2002; Essington, 2004). O rganisms (macrofauna, mesofauna, and microorganisms) living in soil and organic compounds associated with them are collectively called microbial biomass. Humus, 60 to 80% of SOM, is composed of aromatic ring type structures with a molecular weight from about 2,000 to 300,000, and is split into humic and nonhumic substances. I ll defined, complex, resistant compounds are called humic substances, whereas nonhumic substances refer to the group of identifiable biomolecules that are mainly produced by microbial action and are generally less resistant to breakdown. Furthermore, humic substances are classified into fulvic acid and humic acid, based on their solubility in acid and alkali (Figure 1 3). Soil organic matter is an important component of soil because it serves a key role not only in soil chemistry such as cation exchange capacity, cycling of nutrients, and water retention but also in soil physical an d biological properties (Certini, 2005; DeBano et al., 1998) Fire causes a reduction of SOM content (Covington and Sackett, 1984; Gonzales Perez et al., 2004) and alteration s of chemical characteristics of SOM (Gonzales Perez et al., 2004; Knicker, 2007) Depending on fire severity, SOM can be slightly distilled, partly charred, or completely oxidized (Hatten and Zabowski, 2009 & 2010) In a laboratory experiment by heating the 10 cm of topsoil from Pinus sylvestris forest to four different temperatures of 150, 220, 350, and 490 o C, Fern andez 1 Reddy and DeLaune, 2008

PAGE 30

30 et al. (1999) found that SOMs were not affected by th e lowest temperature, but were completely burned at the highest temperature, whereas 37% of SOM were lost at 220 o C. Approximately 85% of S OM was destroyed at between 22 0 o C and 35 0 o C. In addition, N eary et al. ( 2005 ) showed that about 99% of SOM were exclusively demolished at 450 o C for 2 hours or at 500 o C for a half hour Fire frequency also influences SOM because it diminishes accumulation of above ground material layers (Johnson and Curtis, 2001; Neary et al., 2005). Bird et al. (2000) reported that low frequency burning (every 5 years) resulted in an increase in soil C of approximately 10% compared with that of high frequency burning (every year). Along a gradient of fire frequency (every 1 4 year s) over a 12 year period with spring burns and summer burns, Neill et al. (2007) demonstrated that soil organic horizon thickness decreased more in summer burns than in spring burns, and only summer burning every 1 2 years reduced soil organic horizons. Re covery of SOM after fire generally is due to net primary productivity of secondary ecological successions. Fernandez et al. (1999) showed that SOM content was totally recovered in two years after fire. Nitrogen The i mmediate response of N pool to fire is loss of organic N in the forest floor through volatilization during combustion ( Wan et al., 2001 ; Vose et al. 1999 ). High fire temperature rapidly oxidizes components of N in organic matter, volatilizes organic N compounds contained in forest floor and soil organic matter, and thereafter releases NH 4 N (Christensen and Muller, 1975; Kutiel et al. 1989; Marion et al. 1 991). In general the total amount of volatilized organic N is directly proportional to the amount of SOM destroyed during fire Soil organic N is reduced as a result of combustion; however, con tents of inorganic N ions increase considerably following fi re (Choromanska and Deluca 2001; Knoepp and Swank, 1993 ; Raison et al. 1985) Fire induced production of NH 4 N is related to the decomposition of secondary amide

PAGE 31

31 groups and amino acids in soils. By analyzing thermal decomposition of proteins and N rich o rganic matter, Russel et al. (1974) proved that production of NH 4 N was involved in the decomposition of secondary amide groups and amino acids which were easily decomposed at heating temperatures above 100 0 C, and heat resistant N compounds can be volatilized up to 400 0 C. In addition, ash produced from above ground combustion also contains a substantial amount of NH 4 + N (Qian et al., 2009 a & b). Its downward movement into soils results in increasing content of NH 4 + N in soil solution (Lavoie et al ., 2010; Monleon et al., 1997; Schoch and Binkley, 1986). Unlike ammonium, fire induced increase of soil NO 3 was small immediately after fire, but could reach a maximum of approximately threefold of its preburn level within 0.5 1 year after fire, and then decreased thereafter (Bauhus et al., 1993; Covington and Sackett, 1992; Wan and Luo 2001). Prieto Fernandez et al. (1993) demonstrated that total inorganic N was increased in both the surface layer (0 5cm) and subsuface layer (5 10cm) of a P. pinaster forest one month after fire while soil NO 3 content was only increased in the subsurface layer. Choromanska et al. (2001) showed that concentrations of mineralizable N and NH 4 N in surface soil (0 10cm) of a Ponderosa pine forest were significantly incre ased immediately after fire, and the mineralizable N was significantly reduced after nine months. Weston and Attiwill (1990) quantified the concentration of fire induced inorganic N in the topsoil of E. regnans forest in which inorganic N reached three tim es higher than its original concentration over the first 205 days, but returned to its pre fire level after 485 days. Fire intensity and fire frequency also impact the soil N pool. Combustion of soil organic N starts at temperature s above 100 0 C, but its l oss really occur s at temperature of about 2 00 o C and organic N components are entirely volatilized at temperature above 500 o C. While 75 to

PAGE 32

32 100% of N is lost at temperatures of 400 to 500 o C, t emperatures of 300 to 400 o C can cause loss of 50 to 75% of N, an d between 25 to 50% of N is volatilized at temperatures of 200 to 300 o C (Knoepp et al., 2005). Total amounts of NH 4 N and NO 3 N prod uced as a result of fire generally increases with increasing fire intensity. In a collection of eight studies, Neary et al. (2005) showed that the more severe a fire was, the higher the inorganic N content produced. T otal amounts of inorganic N produced by low severity is less than 10 ppm (mg N kg soil), but soil inorganic N amounts reach from 16 ppm to 43 ppm with medium and high fire severity (Figure 1 4) Covington and Sackett (1986), who conducted a study of repeated burning at 1 2 and 5 y ear intervals in a Ponderosa pine forest in the Nort hern Arizona, found that there were no significant differences in TN among treatments of burning frequencies, but available concentrations of inorganic N were higher on the repeated burning sites than on unburned controls. Monleon et al. (1996) conducted b urns on Ponderosa pine sites at three burn return intervals of 4 months, 5 years, and 12 years. Only surface soils (0 5cm) had significant changes of N, with increases of inorganic N and soil C:N ratio in the 4 month burned sites, reduction of total N and soil C:N ratio at the 5 year burned sites, and no effect on soil C:N ratio at the 12 year burned sites. After entering into soil solution, NH 4 + and NO 3 participate in processes of N biogeochemical cycle (Figure s 1 5 ) where f ate of soil NH 4 N is different from NO 3 N Ammonium is oxidized into nitrate by bacteria l oxidation, absorbed into negatively charged surfaces of minerals by cation excha nge, fixed in clay colloids by fixation process, volatilized in air as ammonia, immobilized into soil bioma ss through microbial activities, or leached into groundwater or surface flow (Cleemput, 1998 ; Megonigal et al., 2004). B io availability of nitrate almost depends on the soil microbial community. Nitrate is converted back into ammonium by

PAGE 33

33 dissimilation nitr ate reduction to ammonium (DNRA) (Burgin and Hamilton, 2007 ) denitrified into nitrogen gas (Korom, 1992 ) or leached into groundwater. Phosphorus Phosphorous is probably the second most limiting nutrient in natural forest ecosystems after N C ommon compounds of P often fall into one of two main groups: those containing calciu m; and those containing iron or aluminum. Calcium phosphate compounds become more soluble as soil pH decreases; hence, they tend to dissolve and disappear in acid soils. These co mpounds are quite stable and very insoluble at higher pH, and so become dominant forms of P compounds present in neutral and alkaline soils. Solubility of calcium phosphate compounds are in order of monocalcium P > Dicalcium P > octacalcium P > tricalcium P > oxy apatite > hydroxyl apatite > carbonate apatite > flourapatite (Brady and Weil, 2002; Pierzynski et al., 2005 ). In contrast, hydroxy l phosphate minerals of iron or aluminum, such as strengite (FePO 4 .2H 2 O), vivianite (Fe 3 (PO 4 ) 2 8H 2 O), and variscite (AlPO 4 .2H 2 O), have very low solubility in strongly acid ic soils and become more soluble as soil pH rises. These minerals are unstable in alkaline soils, but they are prominent in acid soils (Brady and Weil, 2002; Pierzynski et al., 2000) In both acid ic an d alkali ne soils, P tends to undergo sequential reactions that produce P containing compounds of lower solubility and less bioavailability for plant uptake Orthophosphate which is defined as inorganic P ions including H 3 PO 4 H 2 PO 4 HPO 4 2 and PO 4 3 is a bio available form of P in soil solution. S peciation of these ions in soil solution is dependent on soil pH (Figure 1 6) and is characterized b y three equilibrium constants (Equations 1 4, 1 5 & 1 6). Of these ions, H 2 PO 4 and HPO 4 2 are two dominant species of orthophosphate in natural forest soils. While H 2 PO 4 is a dominant for m of orthophosphate in soils with a pH of 4 to 7.2, HPO 4 2 is only present in alkaline soils with pH greater than 7.2 ( Pierzynski et al. 2005)

PAGE 34

34 [H 3 PO 4 2 PO 4 ] + [H + ] logK sp1 = 2.15 Equation 1 4 2 [H 2 PO 4 4 2 ] + [H + ] logK sp2 = 7.20 Equation 1 5 2 [HPO 4 2 4 3 ] + [H + ] logK sp3 = 12.35 Equation 1 6 2 Fire has a significant impact on increasing contents of P in soils N ot only does fire supp ly a large amount of ortho phosp hate into soil from ash ( Durgin et al. 1984; Khanna et al. 1994; Qian et al., 2009b) but it also converts soil organic P into ortho phosphate as well (Cade Menum et al. 200 2 ; Certini, 2005; Sharpley and Moyer 2000) T his results in an increase of P bio availability in soils after fire ( Giardina et al. 2000; Knoepp et al. 2005; Romanya et al. 1994 ; Serrasolsas, 1995b ). Residence time of inorganic P ions after fire depends entirely on the presence of cations (Al, Fe, Ca, Mg, Mn ) in soil solution. Figure 1 6 describes possible pathways relating to the fate of inorganic P ions in soil solution after fire. Precipitation wit h cations of Al, Fe, Ca, Mg, or Mn and a dsorption of P ions on surface of soil part icles are two main processes that respond immediately to the availability o f inorganic P ions in soil solution. In acidic soils, inorganic P ions bind to Al, Fe, and Mn oxides through chemisorptions, whereas in neutral or alkaline soils, it binds to Ca /Mg minerals or precipitates as discrete Ca /Mg phosphate minerals (Brady and Weil, 2002; Neary et al. 200 5; Pierzynski et al. 2000 ). P recipitation of P can be defined as the formation of discrete, insoluble compounds in soils, and can be viewed as the reverse of mineral dissolution. The most common ly precipitated P compounds are constituted by reactions among orthophosphates and cations of Ca, Al, and Fe. In alkali ne soils, HPO 4 2 ions quickly react with calcium to form dicalcium phosphate dehydrate (C aHPO 4 .2H 2 O) which later reverts to other more stable Ca phosphates such as octacalcium phosphate (Ca 4 H(PO 4 ) 3 .2.5H 2 O), and then to flour apatite (Ca 10 (PO 4 ) 6 F 2 ). In acidic soils, r eactions of H 2 PO 4 ions with dissolved cations ( Fe 3+ Al 3+ and Mn 4+ ) form hydroxyl pho sphate 2 Essington, 2004

PAGE 35

35 precipitates In addition to precipitation, a morphous or crystalline oxides which are dominant in acidic soils are capable of absorbing inorganic P ions, fixing them on the soil surface, and creating discrete solid phase s of Fe P and Al P P hosphate fixation processes mostly occur when H 2 PO 4 react s with or is adsorbed to the surfaces of insoluble oxides of iron, aluminum, or manganese, such as gibbsite (Al 2 O 3 3H 2 O) and goethite (Fe 2 O 3 3H 2 O), and 1:1 type silicate clays. H 2 PO 4 ion s are attracted to positive charges on the surfaces of iron and aluminum oxides or on the broken edges of kaolinite clays to form outer sphere complexes or may replace a structural hydroxyl to form an inner sphere complex with surfaces of oxides (or clays). Other Nutrient s Many recent studies have reported that fires (prescribed fire and wildfire) significantly increased cation contents especially Ca, K, Mg, Na ( Gray and Dighton, 2006; Groechl et al. 1993; Grove et al. 1986; Marion et al. 1991; Soto et al 1993) These studies mostly focused on determining alterations of nutrient cations in acidic and neutral soils. Brais et al. (2000) found that exchangable Ca and Mg contents following a light to moderate burn increased in the 0 10cm mineral layer of a boreal forest in Nort hwest Quebec. Grove et al. (1986) demonstrated that extractable nutrients (Ca, Mg, K, Na) at the top 3 cm of soil immediately increased after the fire, whereas only extractable K concentration was increased at the lower depths. In a st udy of a Q. rubra and P. grandidentata forest, Adams and Boyle (1980) found that availabilities of Ca, Mg, and K were significantly higher than pre burn levels after one month, but returned to pre burn levels in three months. Tomkins et al. (1991) describe d changes i n concentrations of K and Mg in surface soil of a Eucalyptus forest lasted up to 6 months after fire. In other studies, h owever, Neill et al. (2007) showed that fire had no significant effect on exchangable Ca, Mg, and K in a

PAGE 36

36 O ak pine forest. Vose et al. (1999) showed that there was no response of soil chemistry to burning in a Southern Appalachian pine hardwood ecosystems. Ash produced from fuel combustion contain s a large amount of cations (DeBano and Conrad, 1978; Qian et al., 2009b ) contri but ing to increas ed activity of cations in the soil solution (DeBano et al. 1998 ) Fire induced increases of ionic activities in soil solution are strictly dependent on types of cations, soil properties (Blank et al., 2007; Kutiel and Shaviv, 1992), accum ulation of fuel loading (Snyder, 1986), and degree of combustion (Durgin et al ., 1984, Raison et al., 1985). By comparing availability of nutrient elements in two different kinds of soils, Blank et al. (2007) concluded that fire in duced increases in availa bility of K, Ca, Mg, and Fe differed between Nevada soils and Utah soils, which were explained by differences in soil moisture and elevation. By analyses of plant ash, Durgin et al (1984) showed that contents of Ca, Mg, and K were higher in grey ash than in black ash. Moreover, Raison et al. (1985) showed that Ca, Mg, and K were dominant cations in the plant ash and concentrations of these cations were greater in fine ashes (grey and white ash) than in partly combusted material. Fire induced transformation of elements was described by Boerner (1982), which included loss through volatilization during fire, deposition as ashes, and remaining in incompletely burned vegetation or detritus. Similar to N and P, after deposition into soi l solution, macro and micro nutrients undergo processes in their biogeochemical cycles. Figure 1 7 expresses main pathways relating to the fate of nutrient elements in the soil solution. Their fate often relates to processes of precipitation with orthopho sphates, adsorption by colloids through cation exchanges, loss by leaching and /or runoff, and redox changes. Behaviors of these elements mostly involve reactions with nitrates and orthophosphates produced from fire or in the soil solution.

PAGE 37

37 Research Ratio nale and Objectives Previous studies showed that fire has pl ayed an important role in changing nutrient contents in mineral and organic soils with acidic or neutral pH. Fire reduces TC and TN contents in the soil and forest floor as a consequence of combus tion of SOM and aboveground vegetation. Losses of C and N pools are proportional to an increas e in fire intensity. However, fire increases contents of inorganic N, orthophosphate, Ca, Mg, K, and other elements. Ash production and nutrient elements associated with residual ash have important impacts on changes of nutrient availability in soil solutio n. Effects of fire on vegetation and acidic soils in pine forests have been well documented in many recent studies: pine oak mixed forest (Kuddes Fisher and Arthur, 2002; Hubbard et al., 2004), loblolly pine (Schoch and Binkley, 1986; Binkley et al., 1992; Waldrop et al., 1992), longleaf pine (Wilson et al., 2002; Boring et al., 2004; Glitzenstein et at., 2003), longleaf pine slash pine (Abrahamson and Abrahamson, 1996; Lavoie et al., 2010). The Pine Rockland slash pine in South Florida is a nutrient poor and fire dependent forest ecosystem ( U.S Fish and Wildlife Service, 2007 ; USGS, 2000 ), and is calcareous soil with a high pH buffer and abundance of calcium magnesium, iron, and manganese Effects of fire on soil chemistry of the Pine Rockland ecosystem m ay differ from those of fire in acidic or neutral pH soils. Although ecological roles of fire on vegetation recovery in the Pine Rockland forest ecosystem have been evaluated (Ross et al., 2003; Sah et al., 2004 and 2006; Snyder 1986; Snyder et al., 2005), effects of fire on soil chemistry have not been documented. Research q uestions for this study were: (i) What is a status of P in the Pine Rockland forest: (ii) How will prescribed fire impact the availability of P, N and other elements in calcareous soils in the Pine Rockland ecosystem ? (iii) How can soil moisture and burning temperature affect availability of P, N, and other elements following prescribed fire ? (iv) What are the main forms of phosphate

PAGE 38

38 compounds and their concentrations following prescribe d fire? We hypothesized that P was a limiting factor in the Pine Rockland forest prescribed fire would improve availability of P and other nutrients and soil moisture was an important factor which contributed to availability of P, N, and nutrient element s after prescribed fire. The overall objective of my research was to evaluate impacts of fire on availability of nutrients in calcareous soils in the Pine Rockland forest. Specific objectives were: 1) d etermining the status of P in the Pine Rockland fores t, C:N: P ratios in soil and pine foliage and how prescribed fire could impact availability of P and other nutrient s (Chapter 2); 2 ) d etermining in the laboratory how fire intensity and soil moisture affect available soil P contents, and soil nutrient pools, and to estimate a fire temperature for a field burning experiment (Chapter 3) ; and 3) s imulating P species and their contents following a fire, determin ing solubility of phosphate minerals after the fire, and construct ing a best fit curve of soil moisture content for Pine the long term fire effect on P availability (Chapter 4).

PAGE 39

39 Figure 1 1 Pine Rockland forest ecosystem A ) Distribution of the Pine Rockland forests in South Florida (Adapted from Library Taxes A & M (2008)), B ) loss of the Pine Rockland Area in the Miami Dade County during the 1900s (Adapted from Fairchild Tropical Botanic Garden (2008)) Big Cypress National Preserve A B Long Pine Key Miami Rock Ridge Big Pine Key

PAGE 40

40 Figure 1 2. Temperature thresholds of typical soil components ( Adapted from Neary et al. ( 2005 ) ) Figure 1 3. Pool of soil organic matter ( Adapted from Brady and Weil ( 2002) )

PAGE 41

41 Figure 1 4 Variations of soil extractable N to fire severity gathered by eight studies (Adapted from Neary et al. (2005))

PAGE 42

42 Figure 1 5 Possible pathways relating to fate of NH 4 + and NO 3 after a fire ( Adapted from Pierzynski et al. ( 2000 ) ) Flow to surface water Groundwater Soil solution N (NH 4 + NO 3 ) Organic N: Soil biomass Soil organic matter Mineralization Ammonia Volatilization (NH 4 + NH 3 ) Bacterial changes (NO 3 N 2 gas) Clay Colloids (NH 4 + ) Leachin g Surface water ( Eutrophication ) Erosion, Runoff ( Sediment & soluble N ) Inputs Outputs Internal cycling Denitrification Ammonia Fixation Plant uptake Volatilization Adsorbed Bacterial oxidation (NH 4 + NO 3 ) Immobilization Nitrification Bacterial reduction (NO 3 NH 4 + ) DNR A NH 3 N 2 Cation exchange (NH 4 + ) Desorbed Combustion of SOM & fuels

PAGE 43

43 Figure 1 6 Possible pathways relating to fate of ortho phosphate after a fire ( Adapted from Pierzynski et al ( 2000) ) Phosphorous (P) in soil solution (H 2 PO 4 HPO 4 2 ) Organic P: Soil biomass Soil organic matter Soluble organic P Decaying plant residues Mineralization Immobilization Sorbed P Clays Al, and Fe oxides Secondary P minerals Ca, Al, Fe, and Mn Phosphates Primary P minerals Apatites Leachin g Flow to surface water Groundwater Surface water ( Eutrophication ) Erosion, Runoff ( Sediment and soluble P ) Inputs Outputs Internal cycling Desorption Precipitation Dissolution Dissolution Adsorption Plant uptake Combustion of SOM & fuels

PAGE 44

44 Figure 1 7. Possible pathways relating to fate of nutrient cations after a fire ( Adapted from Pierzynski et al ., 2000) Combustion of SOM & fuels Macro and micro nutrients in soil solution Organic Forms: Soil organic matter Soil biomass Mineralization Immobilization Colloids: Cation Exchange Secondary minerals Primary minerals Leachin g Redox Changes Groundwater Surface water Erosion, Runoff ( Sediment bound and soluble) Inputs Outputs Internal cycling Desorption Precipitation Dissolution Dissolution Adsorption Plant uptake Reduction Oxidation

PAGE 45

45 CHAPTER 2 PHOSPHORUS LIMITATION IN THE PINE ROCKLAND FOREST ECOSYSTEM AND NUTRIENT AVAILABILITY AFTER A PRESCRIBED FIRE Three basic methods have been used to diagnos e nitrogen (N) and phosphorus (P) deficiencies in fores t ecosystems by foliar analysis: 1) correlating nutrient concentration with plant growth and defining a critical level, 2) adjusting critical levels for stand age or development, and 3) evaluating a nutrient balance (Comerford and Fisher, 1984) Correlating nutrient content with growth has been used for indentifyi ng criti cal levels of a single nutrient. The nutrient balance approach assesses the sufficiency level of one nutrient in relation to other nutrients (Sumner, 1977). The correlating method of critical N and P levels for stand age is often implemented experi mentally by adding N and/or P fertilizers to soils in order to determine optimum growth contents of N and P for a certain ecosystem. In the nutrient balance approach, optimum N and P contents f or growth are determined by a dynamic relationship with other n utrients. The Diagnosis and Recommendation Integrated System (DRIS) is an effective tool utilized to evaluate sufficiency of N and P (Sumner, 1978 & 1979; Walworth and Sumner, 1987 ). Atmospheric N deposition often leads to N saturation and P constraint in natural forest ecosystems (Aber et al., 1989; Asner et al., 1997) Many studies have determined limitations of N and P as well a s symptoms of N saturation in different forests (de Visser et al., 1994; Fenn et al., 1996; Tessier and Raynal, 2003; Vitousek and Howarth, 1991) Deriving from the Redfield ratio (C:N:P = 106:16:1) for phytoplankton, many studies in greenhouses by adding N and P fertilizers have determine d N and/or P limitations for plants from natural forest ecosy stems through foliar N:P ratios. I t is assumed that under conditions of low P and high N supply, plants will take up relatively more N than P resulting in a higher foliar N:P ratio. In contrast, under conditions of low N and high P supply, plants will take up relatively more P than N and,

PAGE 46

46 therefore, the N:P ratio in the plant tissue will be relatively low. Based on values of foliar N and P for optimal growth, conclusions can be drawn regarding critical values of foliar N and P or the critical N:P value for forest ecosystems or crops. The N:P ratio has been widely used to determine nutitional limitations of N and P in terrestrial ecosystems throughout the world. This ratio is used as a diagnostic indicator of N saturation (Fenn et al., 1996) and nutrient li mitation of vegetative growth (Chaneton et al., 1996) The N:P ratio has not only been utilized to detect N saturation (Fenn et al., 1996; Fenn et al., 1998; Williams et al., 1996) but has be en also applied to i dentify thresholds of P limitation in different ter restrial ecosystems (Gusewell et al., 2003; Koerselman and Meuleman, 1996; Verhoeven et al., 1996) The foliar N:P ratio is usually considered a sensitive index of P limitation to vegetation gro wth, and an effective to ol for forest managers to determine N saturation and P limitation in forest ecosystems (Tessier and Raynal, 2003) Primary symptoms of N excess include a higher N:P ratio in foliage and a lower C:N rati o in soil (Fenn et al., 1998) Spatial and temporal variations have impacts on N saturation and P limitation because content of nutrients may be altered in the growing season, or by deposition and soil type. By studying the distribution and dynamics of N and P in grazed and ungrazed grass lands in Argentina over winter and spring seasons, Chaneton et al (1996) found that shoot N:P ratios were high in the early winter and low in t he late spring across two areas. Also, the system was relatively more P l imited in early winter, but more N limited in late spring. Fourqurean and Zieman (1992) determined variations in C, N, P contents i n seagrass at two spatial scales: locally (10 100 m) and regionally (10 100 km) through the Florida Bay. Their results showed that C:P and N:P ratios varied greatly both locally and regionally while C :N ratios showed less variation. T he seagrass ecosystem was P limited with leaf N:P ratio greater than 58.6 and N sa turated with leaf C:N ratio small er than 18.5.

PAGE 47

47 Carlyle and Namb iar (2001) evaluated the effects of three typical soils on N saturation and P limitation of a Pinus radiata plantation in Australia. They found that litter C:N ratios for N saturation were 53.2, 60.6, 82.6 for Yellow podsolic, Krasnozem, Siliceous sand so il types, whereas soil C:N ratios for N saturation in these soils were 31.3, 34.1, 37.7. Soil N:P ratios for P limitation ranged from 13 to 18.8 and did not differ much among three types of soils in both litter and soil. Fenn et al. (1998) used both foliar C:N and soil C:N ratios to evaluate N saturation of the mixed forest (Pine, Oak, Fern) in the San Gabriel Mountains, Northeast Los Angeles, CA. Indications for N saturation in the mixed forest were from 24 to 48 for foli ar C:N and from 18.8 to 26.6 for soil C:N. McGroddy et al. (2004) estimated global nutrient ratios for limitation of N and P in terrestrial ecosystems based on molar basis as described by Redfield (1934 and 1958), and reported that global C:N:P ratios for limitations of N and P were 1212:28:1 for foliage and 3007:45:1 for litter. In addition to foliar critical values, the Diagnosis and Recommendation Integrated System (DRIS) index has been widely use d for determining nutrient limitations of different ecosystems. It has been shown to considerable advantages on itself (Sumner, 1977, 1979). The DRIS index is used to assess limitations of not only N or P but also the extent of other elements. It utilizes a dynamic relationship and a wide range of ratios among nutrient elements to evaluate the sufficiency status of each nutrient (Bailey et al., 2000; Bailey et al., 1997a & b; Baldock and Schulte, 1996; Bangroo et al., 2010). The status of an individual nutr ient is diagnosed based on at least two and as many as eight other plant nutrients (Walworth and Sumner, 1987). The DRIS index which takes into account of a nutrient balance within plant tissues is determined from functions of foliar nutrient contents.

PAGE 48

48 The DRIS approach has been successfully used to interpret results of nutrient sufficiency or deficiency of many different crops such as corn, maize, potato, sweet potato, sugarcane, and rice (Bangroo et al., 2010). The biggest advantage of the DRIS method is that it automatically ranks excesses or deficiencies of nutrients in order of importance (Walworth and Sumner, 1987). However, this method also has disadvantages because a DRIS norm needs to be established for each ecosystem. In each ecosystem, DRIS indexe s of nutrients are estimated from its own DRIS norm. Since a norm is usually more easily determined easier for agronomic crops than for natural forests, the DRIS method is not commonly used in natural forest ecosystems. The o nly DRIS norm that has been est ablished for a forest plantation was for Loblolly pine in Georgia and the Carolinas (M.E. Sumner, personal communication, 2011). The DRIS method (Bailey et al., 1997a; Bangroo et al., 2010; Bethlenfalway et al., 1990; Elwali et al., 1985; Needham et al., 1990) calculates a DRIS index for each nutrient element which is the mean of deviation of a ratio containing a given nutrient from its respective optimum or DRIS norm (Bailey et al., 1997a &b). The norm obtained as the mean value for a particular form of expression for a high yielding or desirable population and it has its respective coefficient of variation (Bangroo et al., 2010). DRIS indices range from negative values to positive values, but sum to zero. Negative values indicate a deficiency, whereas positive values indicate a sufficiency of nutrients (Bangroo et al., 2010). A major factor controlling P availability in calcareous soils is the abundance of calcium. The a vailability of P is relatively low in calcareous soils, ranging from 2.1 mg P/kg so il in natural forests to 35.7 mg P/kg soil in agricultural areas with fertilizations (Afif et al., 1993 ; Chien et al., 2009; von Wandruszka, 2006; Yang et al., 2002) Effects of fire on availability of P and nutrient elements in calcareous soils have been evaluated in some studies. Ubeda et al.

PAGE 49

49 (2005) conducted a study on a grassland ecosystem developed on calcareous bedrock, and found that soil pH, total C and N, and extractable potassium in the top 0 5 cm soil were significantly increased immediately after fire, and that contents of C and N were still higher than their pre burn levels in one year after fire while soil pH and extractable K returned to their pre burn va lues after one year. Hernandez et al. (1997) conducted five burning experiments with different fire temperatures and soil textures on calcareous soils in Pinus halepensis and Pinus pinaster forests. Their results showed that soil pH values of five burned s ites were slightly higher than their unburned counterparts, whereas contents of extractable P, K, NH 4 N and NO 3 N in burned sites were significantly higher than those in unburned sites. In particular, available P conten t in sites burned at 300 0 C and 350 0 C reached 120% of that of unburned controls. The Pine Rockland forest developed on limestone bedrock is a nutrient poor and fire dependent ecosystem It is thought that P is a limiting factor in this ecosystem, and bio availability of micro nutrients may be relatively low as well Limitations of P and micronutrients in calcareous soils under the Pine Rockland forest are mainly due to abundances of Ca, Mg, Fe, and Mn. The ecological role of fire in the Pine Rockland has been recently tested (Ross et al., 20 03; Sah et al., 2004 and 2006; Snyder, 1986; Snyder et al., 2005) but the effects of fire on soil chemistry of this ecosystem have not been well documented. Therefore, a study that comprehensively assesses nutrient status and the role of fire on nutrient a vailability in the Pine Rockland is necessary. The objectives of this study were (i) to confirm P limitation based on analysis of pine foliage, and (ii) to evaluate effects of fire on changes in soil pH, EC, and soil nutrients following a prescribed fire.

PAGE 50

50 Materials and Methods Description of Study Site Th e study was conducted in a Pine Rockland Ecosystem in Long Pine Key, Everglades National Park South Florida slash pine ( Pinus elliottii var. densa ) trees (Appendix A) are dominant in the overstory of t he Pine Rockland ecosystem and have an average canopy height ranging from 20 to 24 m (Snyder et al. 1990) Tree density at the site was reported to be 453 to 1,179 pines/ha (Snyder, 1986). The understory contains a diverse assemblage of native plants and a rich herbaceous layer including saw palmettos, locust berry, willow bustic, beauty berry, broom grasses, and silver palms ( U.S. Fish and Wildlife Service, 2007). The t errain of Long Pine Key is flat and moderatel to well drained with an average elevation of approximately 2 m above sea level (Snyder, 1990). Surface soils are very shallow because limestone bedrock is at or near the surface. Most sites in this area are only wet for a short period following heavy rain events, but some sites may be inundated by very slow flowing surface water during the rainy season ( U.S. Fish and Wildlife Service, 2007) Soils in the Long Pine Key were formed with Miami oolitic limestone (Snyder, 1986; U.S. Fish and Wildlife Service, 2007) with various d epths from less than one cm (mostly rock outcrop) to over 10 cm. When deep soil more looks like mineral soil with a reddish brown color and contains a small amount of organic matter, soils in rock outcropping area often were d ominated by organic matter. The hydroperiod in the Long Pine Key ranges from about 20 to 60 days/year. Annual average rainfall and temperature in this area are approximately 143 cm and 2 5 0 C, respectively (Weather Channel, 2008) Average water table is 0.6 to 2.0 m below the soil surface in the dry season and 0.3 to 1.0 m below the surface in the wet season (Olmsted et al., 1983). Brazilian pepper ( Schinus terebinthifolius) and Burma reed ( Neyraudia reynaudiana ) are

PAGE 51

51 the two most widespread invaders in the a rea. The exotic Brazilian pepper has entirely occupied the Hole in the Donut, and has threatened the Pine Rockland on the Long Pine Key. The Burma reed which can be tolerant to fire has established slightly in the Pine Rockland of the Long Pine Key ( U.S. F ish and Wildlife Service, 2007 ). Currently, the Everglades National Park is carrying out an ongoing program of prescribed burning to control invasive plants in the Long Pine Key (Olmsted et al., 1983). Although prescribed fire was usually done in winter in the Pine Rockland in the past, the Park has begun to use a summer burning program since 1981 (Doren et al., 1993) Determination of P Limitation and C:N:P ratio A literature search for critical P value in slash pine forests as well as critical P values for P limitation in upland ecosystems was performed using the followin g key words: nitrogen content, nitrogen concentration, nitrogen deficiency, nitrogen limitation, phosphorus content, phosphorus concentration, phosphorus deficiency, phosphorus limitation, N:P, N/P, N:P ratio, N/P ratio, N to P ratio, nutrient concentratio n, nutrient content, nutrient deficiency, nutrient limitation, nitrogen fertilization and phosphorus fertilization (Tessier and Raynal, 2003). The results were tabulated for a generalized determination of global P limitation in upland ecosystems, and for a specific determination of critical P value in slash pine forests. Determination of global P restriction aimed to providing a broad view of P limitation that could occur in upland forest ecosystems. Limitation of P in the Pine Rockland forest ecosystems we re determined based on comparisons between the foliar P contents of the Pine Rockland slash pine and published critical values of pine foliage P in slash pine forests. Additionally, DRIS indices were established for P based on the DRIS norm created for th e Loblolly pine forest in the Southeastern United States. Although the DRIS norms for Loblolly may not be exactly that of slash pine, they are closely related species with similar

PAGE 52

52 geographical distributions. Needham et al. (1990) established the DRIS nor ms for N, P, K, Ca, and Mg for the Loblolly pine forest (Table 2 5). Relying on these norms, DRIS indices of N, P, K, Ca, and K for the Pine Rockland slash pine were calculated by the following two steps (Bangroo et al., 2010; Walworth and Sumner, 1987): F irst, construction of functions for each nutrient ratio pair: F(A/B) = ( 1) if A/B is greater or equal to a/b F(A/B) = (1 ) if A/B is smaller than a/b Where A/B is a nutrient ratio of pine foliage to be diagnosed such as ratios among N, P, Mg, Ca, and K; a/b is the optimum value or norm for the given ratio obtained from the Loblolly pine forest; CV is the coefficient of variation associated with the norm. Secondly, calculation of DRIS index for each nutrient: N ind ex = [f(N/P) + f(N/K) + f(N/Ca) + f(N/Mg)]/4 P index = [f(P/N) + f(P/K) + f(P/Ca) + f(P/Mg)]/4 K index = [f(K/N) + f(K/P) + f(K/Ca) + f(K/Mg)]/4 Mg index = [f(Mg/N) + f(Mg/P) + f(Mg/K) + f(Mg/Ca)]/4 Ca index = [f(Ca/N) + f(Ca/P) + f(Ca/K) + f(Ca/Mg)]/4 Ind ices of these five nutrients were listed in an increasing order. Negative index value of a nutrient would demonstrate a deficiency of the given nutrient, whereas positive index value of a nutrient would prove a sufficiency of the given nutrient. Field Burn ing Experiment and Sample Collection Three sites ( manegement blocks I1, I2, and Boy Scout Camp) within the Long Pine Key (Figure 2 1) were selected for sampling. The geographic coordinat e s of these blocks are approximately 25 0 23 0

PAGE 53

53 t he Boy Scout block which was entirely burned on November 19, 2008. Absence of fire for five years and drought associated with the event in the 2007 2008 La Nina provided favorable conditions for fire Prescribed fire was started at noon and completed approximately two hours after burning. The air temperatures averaged 28 0 C during the fire, and a slight precipitation occurred approximately one week prior to burning (Figure 2 5). Soil samples were rand omly collected at topsoil 0 5cm within nine different sampling dates, including pre burn, and 14 30, 90, 180, 270, 360, 450, and 540 days following the fire. Twenty soil samples were collected at each sampling period, and four samples were mixed together for each replicate (in total, five replicates for each sampling period). A 50 x 50 cm sampling frame (Appendix A) was used to collect fuel loads and residual ash with ten replicates. Fuel loads including litter, residues, and understory vegetation were ra ndomly collected one day before the fire in order to determine content of nutrients in fuel loads. Ash products created by combustion were immediately gathered after the fire by a vacuum machine. Twenty samples of pine foliage were randomly collected from twenty different pine trees (average heights of 8 10 m) in unburned blocks of I1 and I2 in January (Appendix A), beginning of the pine growing season (Harley et al., 2011). Sample Preparation Soil samples were air dried, ground, and sieved at 2 mm scree n. Plant roots and rocks were removed before sieving though 2 mm. Fuel samples and pine needles were cut into 1 to 2 cm segments, dried in oven at 65 0 C for 5 days, and ground before nutrient analysis Chemical Analysis Electrical conductivity (EC), pH, NH 4 N, NO 3 N, PO 4 P, total P (TP), total C (TC), total N (TN), and other elements were measured. Soil and ash pH and EC were measured using a

PAGE 54

54 pH/Conductivity meter. Soil P was extracted by 0.5M NaHCO 3 (Olsen et al., 1954), and measured using a spectrophoto meter with the automated colorimetric method. Ammonium and nitrate were extracted with 2M KCl, and measured using a discrete auto analyzer (AQ2, Seal Analytical, Mequon, WI) with USEPA methods of 350.1 and 353.2, respectively. Other nutrients (Ca, K, Na, F e, Mg, Mn, Zn, and Cu) were extracted by the Ammonium Bicarbonate Diethylene Triamine Pentaacetic Acid (AB DTPA) method, and measured on an Atomic Absorption spectrophotometer (AA6300, Shimadzu Scientific Instruments, Columbia, MD). Total C and N in soil, fuel load, and pine foliage were automatically measured by CNS auto analyzer (vario Max, Elementar Americas, InC., Mt. Laurel, NJ). Total P in soil, pine foliage, and fuel load were dry ashed, dissolved in 6M HCl acid, and measured using a spectrophotomet er. Analytical methods of these soil components and their detection limits were summarized in Table 2 1. Procedures of chemical analysis and QA/QC methods in laboratory for analyses of these chemical components are presented in Appendix B. Procedures for p reparation of standard solutions and standard calibration curves obtained from laboratory analyses are shown in Appendix C. Statistical Analysi s Repeated Measures method was used to determine significant effects of prescribed fire on nutrient contents. Significant differences at confidence levels of p < 0.05; p < 0.01, and p < 0.001 lized to determine significant differences by time for each nutrient component. Plots of residuals versus fitted values were performed to evaluate homogeneity of variances for nutrient components. If there was a variance inequality of any nutrient componen t, the data were transformed to obtain an equality of variances. Selection of a suitable model from analysis of within subject covariance structure is the most important step in ANOVA analysis by the Repeated Measures method. Analysis of

PAGE 55

55 covariance struct ure within subjects ensures that inferences about the mean are valid. Compound symmetry and 1 st order autoregressive covariance structure models were used for the ANOVA analysis. The compound symmetry model assumes that any two measurements on the same subject have the same covariance regardless of length of time between measurements. The 1 st order autoregressive model assumes that observations on the same subject that are closer in time are more highly corr elated than measurements at times that are farther apart. Correlation between two measurements in the 1 st order autoregressive model decreases exponentially with increasing in length of time between two measurements. The most appropriate covariance struct ure was selected by comparing Akaike's Information Criterion (AIC) and Schwarz's Bayesian Criterion (SBC) in two covariance structure models. A model with the smallest values of AIC and SBC was deemed the best. R esults of ANOVA analyses by the Repeated Mea sures are exhibited in Appendix D. Results and Discussion P hosphorus L imitation in Pine Foliage and Fuel Load Results of a literature search on nutrient limitation indicated that, in general, most of upland ecosystems are limited by P (Table 2 2). Publish ed N:P ratios of plant tissues in upland forest ecosystems ranged from 7 to 16 for N limitation, whereas P limitation was approximately from 14.17 to 29.41. However, results showed also that some upland forest ecosystems were limited by both N and P. Addit ionally, results of a literature search on critical N and P values of slash pine forests revealed that critical N value of pine foliage ranged from 0.87% to 1.30% (averagely 1.0%) (Table 2 3), and critical P value of pine foliage was from 0.07% to 0.125% ( averagely 0.09%) (Table 2 4). By pine foliar analysis, N and P contents of the Pine Rockland forest were 0.81% and 0.045%, respectively (Table 2 6). These findings indicated that P is a limited factor in the Pine Rockland forest, whereas N is marginal to l imitation or may be

PAGE 56

56 saturated in this forest ecosystem where has a high amount of atmospheric N deposition associated with the high lightning intensity (Li et at., 2002). On analysis of a dynamic relationship among pine foliar nutrients, DRIS indices of P, K, N, Mg, and Ca for the Pine Rockland slash pine were 46.13, 10.77, 0.89, 8.71, and 49.05, respectively (Table 2 5). Negative value of DRIS index associated with a nutrient indicated that the given nutrient was deficient while positive value of DRIS index expressed sufficiency of a given nutrient. The higher the negative value o f DRIS index, the more deficient the nutrient is. Obviously, result of DRIS index analysis proved that P is the most limiting nutrient and K is the second limiting nutrient in the Pine Rockland ecosystem. The n egative value ( 0.89) of N index close to zero could confirm that N is marginal limit ing Results of other nutrient analyses in pine foliage and fuel load (forest floor and understory vegetation) a re presented in Table 2 6. Generally, contents of other nutrients in pine foliage were higher than those in fuel load, except for Ca and Fe. Content of carbon was decreased from 51.7% in pine foliage to 46.6% in fuel load. Contents of N and P in pine foliage (0.81% and 0.045%) were approximately t wice N and P contents in fuel load (respectively 0.475% and 0.0 2%). Contents of Mg, K, Mn, and Zn in pine foliage were similar to those in fuel load. The results indicate that the release rates of C, N, P, Mg, K, Mn, and Zn in the forest floor due to decomposition process are different (Knoepp et al., 2005b). Release and retention of nutrients in litters are driven by biological, physical, and chemical processes in which N, Ca, and Mg in litters are being regulated by biological processes while K release is a physical process (Laskowski et al., 1995). C:N:P Ratios in Pine Foliage, Fuel Load, and Soil C:N:P ratios were 1162.5 : 17.9 : 1 for pine foliage, 2361.2 : 23.8 : 1 for fuel load, and 920 : 30.3 : 1 for soil. Figures 2 2, 2 3, 2 4 showed ratios of C:N, N:P, and C:P in soil, fuel load,

PAGE 57

57 and pine foliage. The order o f C:N was fuel load (99:1) > pine foliage (65:1) > soil (30:1). Photosynthesis creates a large amount of carbon in plant tissue, and provides carbon for soil C pool through turnover (Binkley et al., 1992; Gonzlez Prez et al., 2004) Although soil contains a large amount of N, most of soil N exist in organic forms that are not available for plant uptake. Ammonia and nitrate are the only two major forms of inorganic N which are bio available for plant uptake. In natural forest ecosyst ems, availability of ammonia and nitrate is often limited because two thes e forms of N are mostly dependent on processes of microbial mineralization, nitrogen biological fixation, or atmospheric deposition. Consequently, C:N ratios in pine foliage (65:1) a nd fuel load (99.2:1) were higher than soil C:N ratio (30.4:1). Since a large portion of N in plant tissues disappears during decay while abundance of carbon still remained in plant tissues, C:N ratio in fuel load was higher than that in pine foliage. Soi l N:P ratio of 30.3:1 was greater than N:P ratios of pine foliage (17.9:1) and fuel load (23.8:1); whereas soil C:P ratio of 920:1 was lower than pine foliar C:P ratio (1162.5:1) and fuel load C:P ratio (2361.2:1). Higher N:P ratio and lower C:N ratio in s oil to those in fuel load and pine foliage could also indicate that N may not be limited in the Pine Rockland. This inference was suitable wit h results of some previous studies in which soil C:N ratios for N saturation in three typical soils (Siliceous san d, Yellow podsolic, and Krasnozem) of pine plantation forest in Australia were 31.3, 34.1, and 37.7 (Carlyle and Nambiar, 2001), and soil C:N ratio for N saturation in the mixed forest (Pine, Oak, Fern) in California was a bout 26.6 ( Fenn et al., 1998). Pools of C and N in Fuel Load after Fire Prescribed fire did not affect the canopy of pine trees, but significantly decreased the understory biomass and pools of C and N in fuel load, at least immediately after the fire (Table 2 7). Approximately 90.2 % o f aboveground and understory biomass were consumed by the fire.

PAGE 58

58 Fire volatilized approximately 80% and 92% of N and C contents in fuel loads. Losses of NH 4 + and NO 3 associated with volatilization of fuel load N pool were 86% and 48%. The result showed that considerable amounts of fuel loads accumulated in the understory aboveground biomass over the fire interval were lost. The amount of biomass consumed by the fire was higher than percentage of fuel consumption in dry season burn reported by Snyder (198 6). This might be due to differences in nature of initial conditions and burning frequency before fire. Volatilizations of C and N pools in fuel loads were linearly to loss of biomass. Higher rate of C loss and lower rate of N loss were similar to results from the simulation experiment in laboratory. Changes of Soil pH and EC after Fire Prescribed fire impacted significantly on pH (p < 0.05) and EC (p < 0.001) at the 0 5 cm depth (Table 2 8). Soil EC increased dramatically after the fire, from 333 S/cm i n the pre burn to 698 S/cm in 14 days after the fire, and maintained a high value until 30 days, before returning to its pre fire level after 90 days (Table 2 9). Although soils in the Pine Rockland forest have high pH, prescribed fire slightly increased soil pH, from 7.55 in pre burn to 7.82 in 14 days after the fire before returning to its original value in 30 days later (Table 2 9). Fire induced increases in soil pH and EC in calcareous soils found here were similar to those of Ubeda et al. (2005). Soi l pH was close to the value reported by Snyder (1986). Release of inorganic ions from fuel combustion led to considerable increases in EC (Hernndez et al., 1997; Kutiel and Shaviv, 1992) Increase in pH, the so called liming effect, could be attri buted not only to accumulation of K and Na hydroxides, and Mg and Ca carbonates in post fire ash (Knicker, 2 007) but also due to the destruction of acid group in organic matter during the fire (Dumontet et al., 1996; Giovannini et al., 1988) Rainfall was a primary factor contributing to

PAGE 59

59 increases in soil pH and EC as a result of leaching of inorganic ions fro m the ash into the surface soil. Figure 2 5 showed changes of soil pH and EC along with rainfall events in one year following the fire. The ephemeral rise of soil pH following fire actually occurred after the first rainfall event. Changes of Soil C, N, an d P Pools after Fire Prescribed fire significantly affected soil C (p < 0.01) and N contents (p < 0.01) (Table 2 8). Although the fire virtually removed the pools of C and N in the forest floor and understory vegetation through volatilization, C and N pool s in the surface soil 14 days after the fire were measured as result of residual ash deposition. Comparisons in soil C, N, and P contents over time 9 indicate that prescribed fire significantly changed pools of soil C and N, but not soil P. Contents of TC and TN at 0 5cm depth significantly increased in 14 days after the fire. Content of TN was 0.95% in pre burn and increased to 1.42% after 14 days. Pre fire content of TC was 24% and its content was 40% after 14 days. Contents of TC and TN reached their original values in 30 days after the fire; nevertheless, these contents continued to decrease thereafter and had the lowest contents after 90 days. Although content of TP in soil was increased from 0.032% in pr e burn to 0.042% in 14 days later, the difference was not statistically significant. According to our observations, most of the native plants and vegetation under pine grew rapidly and bloomed in 90 days, and finished their cycles in 180 to 360 days follow ing the fire. Therefore, nitrogen demands for development and blooming of understory vegetation were very high during 90 days after the fire. This caused a dramatic decline in soil TN contents at 90 days after the fire, and made pools of soil N stable afte r 360 days. Increas e in TC and TN following the fire did not affect the soil C:N ratio, but it significantly altered soil C:P and N:P ratios (Table 2 8). Alterations of soil C:P and N:P ratios

PAGE 60

60 fluctuated with changes of soil C and N contents after the fire (Figures 2 6 and 2 7). Ratios of soil C:P and N:P following the fire increased at 14 days, tended to decrease gradually until 90 days, went up and reached a peak at 270 days, and then declined to its original values in 540 days. With a confidence leve l of 95%, significant difference of soil N:P and C:P ratios actually occurred between 90 and 270 days after the fire. Ammonium and nitrate were two major components contributing to increase in soil N pool after the fire. Prescribed fire increased extracta ble contents of NH4 + (p < 0.001) and NO 3 (p < 0.001) significantly (Table 2 8). Extractable NH 4 + concentration increased immediately after the fire, which remained until 30 days, and then declined thereafter. While content of ammonium in soil was raised immediately due to the direct release of ash products during the fuel combustion, an increase of n itrate resulted from nitrification. NO 3 concentration was not changed immediately after the fire, but increased significantly in the post 180 days. Soil NH 4 + pool was 40.6 mg kg soil in pre burn, increased approximately twofold after 14 days (99.6 mg kg soil), and then declined to the pre fire level after 30 days (53.6 mg kg soil). Although fire immediately increased soil NO 3 from 8.1 mg kg soil in pre burn to 11.3 mg kg soil in 14 days later, it was not significantly. Soil NO 3 pool increased at po st fire 180 days to approximately 25.1 mg kg soil, which was a t ripling of its pre fire value. Changes in soil NH 4 + and NO 3 pools in this study were similar to the meta analysis results of Wan et al. (2001) which assessed the terrestrial ecosystems. Significant increase in soil NO 3 content after 180 days was followed by heavy rainfall (Nardoto and Bustamante, 2003) which promoted the nitrification process, and resulted in a reduction of NH 4 + (Covington and Sackett, 1992; DeLuca and Zouhar, 2000; Knicker, 2007) Effects of Fire on Other Soil Nutrient s

PAGE 61

61 Analysis of ANOVA by Repeated Measures indicated that there were significant differences in extractable contents of Ca, Fe (p < 0.05), Mg, Mn, PO 4 P (p < 0.001), and K (p < 0.01) in soil after the fire, except for Cu and Zn (Table 2 8). Effects of fire on soil nutrient pools were variable. Table 2 9 shows a comparison of extractable nutrient contents over time by od. Prescribed fire significantly increased extractable concentrations of Olsen P, K, Mg, and Mn after 14 days, but extractable Fe and Ca increased after 270 and 360 days, respectively. Olsen P content increased from 4.2 mg kg soil in pre burn to 13.4 mg kg soil after 14 days. AB DTPA extractable K, Mg, and Mn increase d immediately from 52 to 144 mg kg soil (Mg), from 165 to 277 mg k g soil (K), and from 69 to 256 mg kg soil (Mn) in comparison between pre fire and and 14 days after the fire. Extractable Fe w as 178 mg kg soil in pre burn, and increased into 289 mg kg soil after 270 days, wh ereas Ca increased from 1060 mg kg soil in pre burn to 1162 mg kg soil at 360 days after the fire. Prescribed fire was very effective for increasing the pool size a nd contents of extractable K, Ca, Mg, Mn, and Fe (Blank et al., 2007; Lavoie et al., 2010) in the Pine Rockland forests, except for Cu and Zn. Olsen P content measured in the Pine Rockland forest was in a reported range of Olsen P concentration for calcareous soils (Afif et al., 1993b; Ryan et al., 1985a & b ) As described in Figure 2 5, rainfall played an important role in changes in soil nutrient pools following the fire. The rainy season st arted after 180 days and lasted until 360 days after the fire. Since soils have high organic matter, they often saturated and became anaerobic for a short period after continuous rainfall events during the rainy season. Under the anaerobic conditions, form s of Fe (III) such as Fe(OH) 3 goethite (FeOOH:H 2 O), FeCO 3 and hematite (Fe 2 O 3 ) were reduced into Fe (II) forms (Brady and Weil, 2002, p. 611; Reddy and Delaune, 2008, p. 409) which resulted in increases of extractable Fe 2+ after 270 days.

PAGE 62

62 Summary This study employed both foliar critical value and DRIS indices methods to determine status of P and other nutrients in the Pine Rockland forest. The finding indi cated that P is the most limiting factor in the Pine Rockland forest. The restriction of P led to e levation in pine foliar N:P ratio (17.9) which was higher than foliar N:P critical level (14 15) of slash pine forests reported by Comerford and Fisher (1984). However, pine foliar N:P ratio was lower than those in soil (30.3) and in fuel load (23.8). C:N: P ratios were 1162.5:17.9:1 for pine foliage, 2361.2:23.8:1 for fuel load, and 920:30.3:1 for soil. Lower C:N ratio and higher N:P ratio in soil than those in pine foliage and fuel load could provide evidence that N may be sufficient in the Pine Rockland f orest. Prescribed fire impacted significantly soil pH, EC, and soil nutrient pools at 0 5cm soil depth of the Pine Rockland forest. The fire increased immediately soil pH, EC, and extractable contents of P, Mg, K, Mn in post fire 14 days, but increased ext ractable content of Fe in post fire 270 days and that of Ca in post fire 360 days. However, the fire did not influence available contents of Cu and Zn. Increases of Fe and Ca contents depended on more rainfall events. Prescribed fire increased significantl y contents of TC and TN, but did not change TP content. Increases of TC and TN occu rred in post fire 14 days, and they returned to their pre fire levels in 30 days. Effects of fire on soil inorganic N pools were variable. Soil NH 4 + pool significantly incre ased immediately after the fire, whereas soil NO 3 pool was increased in post fire 180 days, corresponding with a dramatic decline of NH 4 + production. Increases of soil C and N pools did not vary soil C:N ratio, however; significant differences were observ ed for ratios of N:P and C:P.

PAGE 63

63 Table 2 1. A summary of analytical methods in the laboratory Parameter Method Instrument Detection limit pH and EC (ash) Ash : water = 1:50 pH/conductivity meter 0.1 pH (soil) Soil : water = 1:2.5 pH/conductivity meter 0.1 EC (soil) Soil : water= 1:2 pH/conductivity meter 0.1 NH 4 N (soil) Soil: 2M KCl = 1:10 discrete auto analyzer 0.02 NO 3 N (soil) Soil: 2M KCl = 1:10 D iscrete auto analyzer 0.006 NH 4 N (ash) Ash: water = 1:20 D iscrete auto analyzer 0.02 NO 3 N (ash) Ash: water = 1:20 D iscrete auto analyzer 0.006 TC Combustion CNS auto analyzer 0.3 TN Combustion CNS auto analyzer 0.06 Extractable P (soil) Olsen P Spectrophotometer 0.01 Soluble P (ash) Ash: water = 1:20 Spectrophotometer 0.01 Total P Ashing and 6N HCl Spectrophotometer 0.01 Extractable cations AB DTPA Spectrophotometer 0.001 0.02 S/m for EC and mg/L for nutrients. Table 2 2 Thresholds of foliar N and P limitations in upland ecosystems Studies Systems Locations Addition N:P ratio Limitatio n by Aerts et al. (1988) Heath The Netherlands N & P 29.41 P Alan et al. (2000) Carica papaya Laboratory N & P 7 N & P Bobbink (1991) Chalk grassland The Netherlands N & P 16.08 N & P Bowman (1994) Alpine dry meadow Colorado N & P 13 N Bowman (1994) Alpine wet meadow Colorado N & P 14 P Clarholm et al. (1995) Picea abies plantation Sweden N & P 9.8 N & P de Visser et al (1994) Coniferous forest Europe N 7 14.5 Not N Herbert et al. (1995) Montane forest Hawaii N & P 13.83 P Jacobson et al. (2001) Picea abies stand Sweden N & P 7.54 N Jacobson et al. (2001) Pinus sylvestris stand Sweden N & P 8.96 N Ljungstrom et al. (1995) Beech forest Europe P 14.17 At least P Mohren et al. (1986) Douglas fir forest The Netherlands N & P 22 25 P Mohren et al. (1986) Douglas fir forest The Netherlands N & P 8 N Tessier et al. 2003 Catskill Mountains New York, USA N & P 17.71 P Valentine et al. (1990) Pine plantation North Carolina N & P 10.42 N & P Adapted from Tessier et al. (2003)

PAGE 64

64 Table 2 3 Critical foliar N contents of slash pine forests ( Pinus elliottii var elliottii ) Studies Locations Critical N content (%) Barron Gafford et al. (2003) Southeastern Georgia, USA 0.9 1.2 Comerford et al. (1984) Southeastern Coastal Plain s ites, USA 0.96 1.01 Dickens et al. (2003) Southeastern Coastal Plain, Georgia 1.0 Fisher et al. (2000) Southeastern Georgia, USA 1.0 Gholz et al. (1985) Northern Florida, USA 0.87 1.03 McGarvey et al. (2004) Putnam and Levy Counties, Florida, USA 1.19 1.24 Richards et al. (1972) Australia 0.9 Teskey et al. (1994) Northeastern Florida, USA 0.93 1.16 Will et al. (2001) Southeastern, USA 0.95 1.30 Xu et al. (1995) Australia 0.94 1.19 Table 2 4 Critical foliar P contents and N:P ratio of slash pine forests Studies Locations Critical foliar Critical foliar P content (%) N:P ratio Barron Gafford et al. (2003) Southeastern Georgia, USA 0.095 0.125 Comerford et al. (1984) Southeastern Coastal Plains, USA 0.07 14 15 Dickens et al. (2003) Southeaste rn Coastal Plain, Georgia 0.09 Fisher et al. (2000) Southeastern Georgia, USA 0.09 Gholz et al. (1985) Northern Florida, USA 0.055 0.085 Richards et a l. (1972) Australia 0.075 0.08 Teskey et al. (1994) Northeastern Florida, USA 0.072 0.097

PAGE 65

65 Table 2 5 DRIS indices of N, P, K, Ca, and Mg in the South Florida slash pine Variable N/P P/K Ca/P Mg/P N/K Ca/N Mg/N Ca/Mg Ca/K Mg/K DRIS norm of Loblolly pine forest Mean 10.426 0.227 1.589 0.696 2.361 0.154 0.066 2.355 0.362 0.158 C.V. 11.6 16.6 28.4 28.5 18.1 24.5 24.5 25.25 35.2 31.9 Foliar nutrient ratios of South Florida slash pine Mean 17.907 0.296 9.144 3.228 5.303 0.509 0.181 2.835 2.727 0.954 f(A/B) 6.19 1.84 167.42 12.77 6.89 9.42 7.10 0.81 18.56 15.79 DRIS Index of South Florida slash pine Nutrient P K N Mg Ca Index 46.13 10.77 0.86 8.71 49.05 Source from Needham et al. (1990) and the norm and coefficient of variation (C.V.) of the Loblolly pine forest Table 2 6 Nutrient contents of pine foliage and fuel load in the Pine Rockland forest Parameter C N P Ca Mg K Fe Mn Zn Pine foliage (% of dry mass) Mean 51.7 0.805 0.045 0.405 0.144 0.156 0.0045 0.0024 0.0017 SD 0.42 0.102 0.006 0.099 0.022 0.030 0.0011 0.0009 0.0007 SE 0.09 0.023 0.001 0.022 0.005 0.007 0.0002 0.0002 0.0002 Min 51.0 0.703 0.036 0.263 0.103 0.101 0.0028 0.0011 0.0009 Max 52.8 1.040 0.059 0.611 0.191 0.215 0.0064 0.0044 0.0036 Fuel load (% of dry mass) Mean 46.6 0.475 0.020 0.640 0.131 0.172 0.0086 0.0023 0.0009 SD 0.43 0.053 0.003 0.127 0.055 0.026 0.0027 0.0011 0.0003 SE 0.15 0.019 0.001 0.040 0.017 0.008 0.0009 0.0003 0.0001 Min 46.1 0.435 0.018 0.510 0.085 0.143 0.0060 0.0014 0.0005 Max 47.3 0.600 0.025 0.905 0.277 0.210 0.0135 0.0052 0.0014 maximum value

PAGE 66

66 Table 2 7 Loss of biomass and pools of C and N in forest floor and understory vegetation (fuel load) due to fire Parameter Biomass TN TC NH 4 + NO 3 % loss Mean 90.19 80.13 91.59 86.44 48.21 SD 2.24 2.87 0.40 1.52 11.06 SE 0.71 1.17 0.16 0.62 4.52 Min 86.47 74.71 91.16 84.16 34.28 Max 93.82 83.08 92.02 88.36 64.98

PAGE 67

67 Table 2 8 Results of ANOVA analysis of soil nutrient pools following prescribed fire Variables Pr > F Significant level Variables Pr > F Significant level Variables Pr > F Significant level Ca 2+ 0.0497 Fe 2+ 0.0122 TC 0.0015 ** Mg 2+ < .0001 *** Mn 2+ < .0001 *** TN 0.0048 ** K + 0.0011 ** Zn 2+ 0.0989 NS TP 0.1139 NS PO 4 P < .0001 *** Cu 2+ 0.2317 NS C:N 0.2532 NS NH 4 N < .0001 *** pH 0.0187 N:P 0.0334 NO 3 N < .0001 *** EC < .0001 *** C:P 0.0277 Table 2 9 Days after fire Ca 2+ Mg 2+ K + Fe 2+ Mn 2+ Zn 2+ Cu 2+ PO 4 P NH 4 + NO 3 TP TN TC EC pH mg/kg soil % S/cm 0 1060ab 52b 165bc 178ab 69b 3.5a 3.4a 4.2b 40.6bc 8.1b 0.0322a 0.93abc 24abc 333c 7.55ab 1 4 905b 144a 277a 153b 256a 5.1a 4.1a 13.4a 99.6a 11.3ab 0.0423a 1.42a 40a 698a 7.82a 30 1035a b 80b 193abc 207ab 89b 5.4a 3.1a 4.4b 53.6b 2.4b 0.0301a 0.91abc 25abc 544ab 7.48b 90 972ab 40b 188abc 170ab 47b 3.3a 3.8a 1.2b 23.1bc 5.2b 0.0347a 0.69c 15c 370bc 7.45b 180 1094 ab 54b 236ab 158b 53b 5.3a 3.8a 4.0b 31.2bc 25.1a 0.0340a 0.84bc 23bc 383bc 7.56ab 270 989a b 56b 122c 289a 89b 4.2a 3.6a 3.2b 21.6bc 8.5b 0.0277a 0.98abc 28abc 348bc 7.48b 360 1162 a 53b 145bc 173ab 73b 5.3a 4.2a 4.6b 22.1bc 2.6b 0.0377a 1.28ab 34ab 328c 7.57ab 450 1049ab 62b 160bc 197ab 114b 4.0a 4.0a 2.7b 23.0bc 2.7b 0.0337a 0.96abc 33ab 453bc 7.72ab 540 972a b 34b 137bc 223ab 59b 2.4a 3.8a 3.0b 11.4c 10.5b 0.0317a 0.79bc 20bc 332c 7.58ab

PAGE 68

68 Figure 2 1 Blocks of Pine Rockland in Long Pine Key (Adapted from Everglades National Park)

PAGE 69

69 Figure 2 2 C:N ratios in soil, pine foliage, and fuel load of the Pine Rockland forests Figure 2 3 N:P ratios in soil, pine foliage, and fuel load of the Pine Rockland forests

PAGE 70

70 T he lower and upper whisker caps are 10 th and 90 th percentiles; the lower and upper boundaries of the box are 25 th and 75 th percentiles; solid lines are medians of ratios; and dotted lines are means of ratios Figure 2 4 C:P ratios in soil, pine foliage, and fuel load of the Pine Rockland forests

PAGE 71

71 Figure 2 5 Rainfall and changes of pH and EC along with rainfall after the fire 0 2 4 6 8 10 12 14 Rainfall (cm) Rainfall (cm) 0 14 30 90 180 270 360 450 5 4 0

PAGE 72

72 Figure 2 12: Changes of pH and EC after the fire along with rainfall Figure 2 6 Changes of soil N:P ratio by following the fire Figure 2 7 Changes of soil C:P ratios by following the fire 30 30

PAGE 73

73 CHAPTER 3 BURNING TEMPERATURE AND SOIL MOISTURE AFFECTED NUTRIENT POOLS IN CALCAREOUS SOILS UNDER THE PINE ROCKLAND FOREST ECOSYSTEM Post fire plant ash derived from both wildfire and prescribed fire is an important factor in the balance and cycling of nutrients in natural ecosystems (Wan et al., 2001; Yermakov and Rothstein, 2006). Composition and availability of ash nutrients depend entirely on fire intensity (fire temperature), plant species, nutrient enrichment of habitat, and accumulation of fuels (Qian et al., 2009b). During a fire, a portion of nutrient elements accumul ated in vegetation or su rface soil can be transferred to the atmosphere (Raison et al., 1985; Cachier et al., 1995); however, a large amount of nutrients rem ain in residual ash and are deposit ed downward to the soil surface (Neary et al., 2005; Marion et al., 1991). Environmental factors includ ing wind, rainfall, runoff, erosion, or leaching can contribute to deposition and/or redistribution of ash nutrients into the soil. Deposition of resid ua l ashes results in increased nutrient contents which may have important effects on nutrient status of soils after a fire. Increases of post fire nutrient contents in soils have been recently observed in several studies (DeBano and Conrad, 1978; Lavoie et a l., 2010; Murthy et al., 2006a; Ubeda et al., 2005; Ulery and Graham, 1993a). Ulery et al. (1993b) showed that soil pH was increased after the fire by approximately three units due to large amounts of basic cations in wood ashes deposited into soil, wherea s Badia and Marti (2003) pointed out that soil pH, carbon ( C ) : nitrogen ( N ) ratio, content of available nutrients, and soil organic matter were increased with the addition of ashes. Basic cations [potassium (K ) magnesium ( Mg ) calcium ( Ca) ] accumulated in ashes also increases in soil surface pH (Murphy et al., 2006a; Murphy et al., 2006b). A study conducted by Khanna et al. (1994) found that neutralization of soil acidity correlated well with K, Ca, Mg contents in residual ashes. One important factor influ encing nutrient availability after a fire is fire intensity (Certini,

PAGE 74

74 2005; Neary et al., 2005), which is generally defined as fire temperature and duration ( Govender et al. 2006; Neary et al. 1999 ). Available contents of elements depend on volatilization points of elements and fuel consumption caused by fire temperature. Sulfur (S), C, and N are often oxidized completely at high burning temperatures (> 400 0 C) due to their relatively low vaporization points (Neary et al., 2005; Gray and Dight on, 2006). In contrast, elemental phosphorus (P) and metals having relatively high vaporization points (e.g., 1484 0 C for Ca and 1091 0 C for Mg) are not impacted by fire temperature. Loss of P in a field burn was reported by Raison et al (1985), but under co ntrolled laboratory conditions all contents of P were accumulated in ashes possibly due to limited air movement (Gray and Dighton, 2006). Responses of plants to fire with regard to nutrient availability in residual ashes vary among species as well as ecos ystems, which are attributed to differences in structure and composition of vegetation. Miao et al. (1997) showed that allocation of nutrients after a fire varied among species growing along a nutrient gradient in the Everglades. Moreover, translocation of nutrients during leaf senescence also constituted lower nutrient contents in dead leaves than in live leaves (Miao, 2004) whereas litter created more available nutrients than the same amoun t of plant tissues before burning (Debano and Conrad, 1978) In a laboratory burn experiment with fire intensity ranging from 150 to 550 0 C, Qian et al. (2009b) demons trated that avai lability of ash nutrients in a sawgrass and cattail ecosystems in the Everglades was significantly affected by fire intensity. There was more labile inorganic P remaining in sawgrass ashes than in cattail ashes; contents of NH 4 N and NO 3 N followed the same pattern with total N (TN) showing a significantly decreasing trend with the lowest contents at 450 and 550 0 C; there was a significant increase of ash pH along the increase of temperature gradient; availability of different metals varied w ith the fire intensity; but contents of total P (TP), total Ca (TCa), total

PAGE 75

75 Mg (TMg), and total aluminum (TAl) were not affected by burning temperatures. In another study, by assessing loss of nutrients in ashes between two different burn seasons under fie ld conditions, Snyder (1986) showed that TP, TCa, and TMg contents were not impacted by either wet season burn or dry season burn A significant loss of total K (TK) occurred in the wet season burn, but there was no loss of TK in the dry season burn. Anoth er factor determining availability of soil nutrients is soil water content. Soil moisture is an important characteristic of soil that affects chemical reactions in soil solution and provides nutrients for plant uptake. Availability of nutrient sources in t he rhizosphere depends on soil water content (Dunham and Nye, 1976) because the concentration and mobility of nutrients in soil solution as well as ion exchange from solid phases are influenced by soil water content. In natural forest conditions, soil mois ture fluctuates with temperature, rainfall, and above ground vegetation. Fluctuations of soil moisture can change soil solution chemistry and regulate the availability of nutrients in the soil solution and the distribution of plant species as well. Limitat ions of bio available forms of several plant nutrients have created severe problems for plant growth on calcareous soils. Some plants do n o t grow successfully on these soils because of deficiencies of P and/or other micro nutrients (Tyler, 1992) Phosphorus uptake by plants mainly controlled by P depletion and diffusion rates in the rhizosphere is greatly affected by soil moisture in calcareous soils (Gahoonia et al., 1994). The soil water content considerably influences mineral nutrient contents in soil solutions and their uptake by plants (Misra and Tyler, 1999). Moreover, bicarbonate (HCO 3 ) ion, mostly constituted from calcite (CaCO 3 ), is relatively high in calcareous soil solution. An increase in soil moisture causes an increase in HCO 3 concentration (Inskeep and Bloom, 1986; Mengel et al., 1984). The presence of bicarbonate ions in the soil solution can control activi ties of other ions, especially cations. Bicarbonate ions are not only

PAGE 76

76 considered as a primary factor responsible for plant chlorosis (Mengel, 1994; White and Robson, 1990), but they also interfere with plant uptake of other nutrients such as K and Mg (Misr a and Tyler, 1999). The effect of soil water content on nutrient content s of calcareous soil and in plant biomass was evaluated by Misra and Tyler (1999). They tested seven soil moisture levels, ranging from 35 to 85% of water holding capacity, and showed that extractable Ca, Mg, and Zn decreased while soil pH, solution HCO 3 P and Mn increased with increasing soil moisture; potassium had the highest concentration at 50 70% of water holding capacity; contents of P, Zn, and Mn which had relatively low solu bility and availability in calcareous soils varied in different plant species. Pine Rockland is one of the major endangered e cosystems in the United States, and rated as G1 on the global conservation scal e This represents one of the most critically imperiled habitats globally because of extreme rarity or because of extreme vulnerability to extinction due to natural or man made factors ( Florida Natural Area Inventory, 2008 ; Snyder et al., 1990 ) T he Pine Rockland forest ecosystem originated on limesto ne substrates and has been influenced by anthropogenic activities and fire suppression. Since the European settlement, approximately 98.5% of the Pine Rockla nd area in Miami Dade County was converted into urban, suburban, and agricultural areas ( Fairchild Tropical Batonic Garden, 2008) Intact ecosystems of the Pine Rockland currently exist in three remna nt regions in South Florida includ e the Long Pine Key in the Everglades National Park, the Big Pine Key in the Florida Lower Keys, and a part of the Big Cypress National Preserve (Snyder et al. 2 005; U.S Fish and Wildlife Service, 2007). Anthropogenic activities have led to significant changes of the Pine Rockland ecosystem in geographic extent (e.g., fragmentation of habitats), environmentally d riven factors (e.g., wildfire), biodiversity (e.g., invasion of exotic species), and nutrient biogeochemical cycling. In particular, limitations of available P and micro nutrients (Mn, Cu, and Zn) are threatening t he

PAGE 77

77 biodiversity of native plant communitie s in this ecosystem. Hundreds of hectares of the Pine Rockland forest have been abandoned or replaced entirely by Brazilian pepper ( Schinus terebinthifolius ) in the Hole in the Donut region of Everglades National Park (Loope and Dunevitz, 1981). Recent interests in preservation of intact Pine Rockland ecosystems, the Everglades National Park has resulted in the use of prescribed fire to manage the Pine Rockland forests (Olmsted et al., 1983). The Park has focused on controlling exotic plant speci es and appropriating fire management regimes with respect to maintenance, protection, restoration, and enhancement of the Pine Rockland forest ecosystem. Several recent studies have focused on evaluating effects of fire on restoration of understory vegetat ion i n Pine Rockland forests (Knapp et al., 2009; Robbin and Myers, 1992; Ross et al., 1994 and 2003; Sah et al., 2004 and 2006; Sah et al., 2010; Snyder et al., 2005; Sparks et al., 2002). Snyder (1986) assessed effects of prescribed fire o n restoration o f mass and nutrients in the understory vegetation in the Pine Rockland during two different burn seasons (dry season and wet season burn s ), and showed that the mass and nutrient contents of the understory vegetation varied with the burn season. This implie d that there might have been changes of soil nutrient contents after the fire relating to differences in soil water content between two burn seasons. With limited information regarding the content of nutrients in plant ashes and post fire soil nutrient sta tus, it is essential to quantify availability of nutrients in fire induced plant ashes, and to assess impacts of soil moisture and fire intensity on post fire soil nutrient status for the fire dependent Pine Rockland ecosystem. This initiates a better unde rstanding of biogeochemical processes associated with a fire. This is particularly important for determining the role of a prescribed fire in P biogeochemical process because P is the most limiting factor in the Pine Rockland ecosystem.

PAGE 78

78 The main objective of this chapter was to assess effects of prescribed fire on nutrient availability in soil during different burn seasons simulated under controlled condition of laboratory incubation. Specific tasks included: 1) determining effects of burn temperatures on p H, EC, and nutrient contents in residual ashes; 2) determining a relationship between burn temperature and extractable concentration of soil nutrients; 3) determining whether this relations hip is affected by soil water content; 4) determining whether the r elationship is impacted by time after burn; and 5) estimating a field fire temperature based on laboratory simulated results. Materials and Methods Incubation Experiment A burn experiment was conducted under controlled condition s of laboratory incubation to determine how fire intensity and soil water content impact soil nutrient pools after a fire. Treatments consisted of four burn temperatures (250, 350, 450, and 550 0 C), three levels of soil moisture (35, 70, and 100% of soil fie ld capacity), six different lengths of incubation time (14, 30, 60, 90, 120, and 180 days). There were three replicates of each treatment combination. Soil samples were put into plastic bottles (approximately 800 cm 3 ; 10 cm of length x 8 cm of width x 10 c m of height), and were incubated at 25 0 C for each incubation time. The incubation temperature was the average annual temperature in the Everglades National Park over fifty years (The Weather Channel, 2008). The four burn temperatures of fuel load samples w ere based on temperatures reported by many previous studies and soil chemical properties, i.e., nitrogen (N) starts to volatilize at temperatures above 200 o C (Certini, 2005; Neary et al., 2005) and plant tissues and litters are entirely burned and converte d i nto white ash at temperatures from 500 to 550 o C (Gray and Dighton, 2006; Qian et al., 2009b). The three soil moisture levels selected, represented a range

PAGE 79

79 that can be found in Pine Rockland throughout the year. The Long Pine Key in the Everglades Natio nal Park that was burned in five years ago was chosen for collecting samples. A 30 x 30 cm sampling frame was used to collect soil samples and fuel load samples in the field. The fuel load included pine needles and understory parts All plants litter, and residues inside the frame were gathered for the fuel load samples. Soils from the top 0 5 cm depth in the same frame with the fuel load samples were collected. Twenty four samples were collected for the incubation experiment. Soils in each sam pling frame sample were divided into nine subsamples each placed in a bottle for the laboratory incubation experiment. Additionally, six other sampling frame samples of soils and fuel loads were collected for determination nutrient contents in the fuel lo ad samples developing a soil moisture curve and determination of soil field capacity. S oil samples were air dried, ground, and sieved through a 2 mm mesh screen Plants, litter, and residues of fuel loads were cut into approximately 1 2 cm pieces, and dr ied in an oven at 65 0 C for 5 days. Mixed soils (150 g) were put into plastic bottles for the incubation experiment. Subsamples were used to develop the soil moisture curve and the determination of the field capacity. Approximately 28 g of oven dried, mixed tissues and residues were put into 500 mL glass beakers, and heated for 6 hours for each of the four temperatures. Residual ash produced by heatin g was weighed to calculate perce ntages of ash production for a given temperature. Ash at the same heating temperature was mixed well, and then divided into fifty four equal parts and placed on top of the soil in the sample bottle s. Deionized water was appli ed to soil samples at each soil water content. During the incubation, while the bottles were covered with plastic Ziploc bags, bottle caps were left open. Ziploc bags were opened and closed immediate ly every 24 hours. This allowed the sample to be exposed to oxygen to ensure an aerobic

PAGE 80

80 condition for soils inside the bottles. A procedure for the laboratory incubation experiment is shown in Appendix E. Soil Water F ield C apacity Soil field capacity was estimated by a gravimetric method. Four replicates of 100 g air dried, sieved soil were put into 250 mL plastic cups with small holes in the bottom. Soils in these cups were repacked and continuously flooded with tap water for 2 hours. The fl ooded soils were then drained for 24 hours under laboratory condition s The drained soils were weighed to determine the weight of the soil. Thereafter, t he drained soils were dried in o ven at 105 o C for 24 ho urs, cooled to room temperature and reweighed to determine the weight of oven dried soil for each replicate (Brady a nd Weil, 2002; Liu et al., 2007 ). Calculation of the amount of water for each soil water content treatment is shown in Appendix F Estimation of a Field Fire Temperature F ire temperature s during a field fire can be computed bas ed on a relationship between nutrient contents in laborato ry residual ash and nutrient contents in field ash Qian et al. (2009b) used piecewise equations and contents of TC, TN, and TP in residual ash from both the laboratory simulation and the field to estimate the field fire temperature. Because of relatively high vaporization points of elemental P (774 0 C), Ca (1484 0 C), Mg (1091 0 C), and Fe (2862 0 C) (Wikipedia, 2010), contents of TP, TCa, TMg, and TFe in residual ashes are not impacted by fire temperatures (Qian et al., 2009b; Snyder, 1986). In addition, inorganic P ions are capable of binding to compounds of Fe, Ca, and Mg; therefore, residual ash TP, TCa, TMg, and TFe were selected as predictor variables for an estimation of a field fire temperature. Regression models ( 1 st 2 nd and 3 rd order ) were fitted with burn temperature as the response variable and TP, TCa, TMg, or TFe contents in residual ash as the independent

PAGE 81

81 variable A field fire temperature was estimated from the selected regression models and TP, TCa, TMg, or TFe content in ash collected from the field. The field ash was the ash which was collected in the field study as described in Chapter 2. A fire field temperature was an average value of estimated temperatures from the regression models and the field ash Furthermore, ANOVA analyses of TP/TCa, TP/TMg, and TP/TFe in field and laboratory ashes with the dependent variable of burn temperature were performed to check if there was any loss of P, Ca, Mg, and Fe in field ashes (Appendix I). Chemical Analysis Soil c hemical components in this chapter were analyzed using the same methods described in the section of chemical analysis in Chapter 2. Statistical Analysis M ultiple regression analysis usi ng the PROC GLM procedures (SAS Institute Inc., Cary, North Carolina) was used for evaluating the effect s of soil water content burn temperature, incubation time, and interactions among these factors on soil nutrient pools. The fitting process of t he multiple regression analy sis was first carried out by a full model including main effects, two way and three way interactions. A reduced model was re fitted if there was any insig nificant component in the full model. In addition, a plot of residuals versus predicted values, a normal probability plot (NPP), and Durbin Watson test were done simultaneously with the regression analyses to check homogeneity of error variances, normal distribution of error variances, and independence of the errors for analytical components, respectively. If there was any heterogeneity of error variances for an independent variable, the data were transformed to obtain an equality of error variances. M were performed to determine outliers of the data that could affect results of the multiple regression

PAGE 82

82 analysis. Al l results of multiple regression analysis are shown in Appendix G L inear regression analysis using the PROC GLM procedures was utilized to assess effects of burn temperature on ash pH, electrical conductivity (EC), and total and water soluble nutrients in laboratory residual ash. Results of linear regression analyses are shown in Appendix H. The method of a stepwise, multiple linear r egression of 1 st 2 nd and 3 rd order was used to fit regression models between burn temperature and TP, TCa, TFe, or TMg co ntents in residual ash Selection criteria for choosing the best models to estimate a field fire temperature were based on the highest R square, the lowest coefficient of variation and lowest root MSE (mean standard error). Results of fitting multiple reg ression models are shown in Appendix I Results and Discussion Biomass and Total Nutrients in Residual Ashes Significant differences were observed for effects of heating temperatures on losses of biomass and TN and TC contents in the fuel loads ( Table 3 1 and Figure 3 1). Most of the fuel load biomass, TN, and TC were lost via volatilization at 450 and 550 0 C. At these temperatures, the fuel load materials were converted into white ash (Appendix E) with an average ash production of 8.3% at 450 0 C co mpared with 6.7% of ash products at 550 0 C. Black ash (charcoal) was generated at 350 0 C with an average ash production of 19.1%, whereas heating to 250 0 C produced approximately 43.6% of a reddish black ash. Volatilization of TC and TN was proportional to bi omass loss. Volatilization of TN ranged from 24.5% at 250 0 C to 91.8% at 550 0 C. Approximately 33.4% of TN was lost by heating to 350 0 C while loss of TN was about 81% at 450 0 C. Volatilization of TC ranged from 51.2 to 96.1% at 250 and 550 0 C, respectively, w hereas up to 81 and 94.3% of TC was lost at 350 and 450 0 C, respectively.

PAGE 83

83 R esults of the linear regression analysis revealed that increases in burn t emperatures did not impact total content s of nutrients in ash except for TC (p < 0.001), TN (p < 0.001), and TK (p < 0.001) (Table 3 1 ). With comparatively low concentration in soils, Cu could not be d etermined in residual ash bec ause it was below the detection limit. Due to their relatively high melting points (over 600 0 C), high heat of vaporization (over 12 8 kJ/mol), and limited air movement in the muffle furnace, contents of TP, TCa, TMg, TFe, TMn, and TZn in term s of dry fuel load content did not decrease with increas ing burning temperature. This indicated that TP, TCa, TMg, TFe, TMn, and TZn accumulated i n residual ash within increase d in burn temperature. Nevertheless, because of the low melting point (63.38 0 C) and heat of vaporization (76.9kJ/mol) (Wikipedia, 2010), TK content was significantly reduced as burn temperature increased in term of dry fuel lo ad content. Contents of TK significantly decreased between 250, 350 and 450 0 C, but n o decrease was observed for TK content s at 550 0 C (Figure 3 2) The dry weight of the fuel loads (~ 331 g/m 2 ) was in the range of dry weight of pre burn above ground vegetation in a dry season burn described by Snyder (1986), which was from 200 to 491 g/m 2 Significant impacts of burn temperatures were observed on TC and TN contents in plant ash A high burn temperature (over 450 0 C) resulted in removal of substantial a mounts of C and N pools from the fuel loads. Our results were similar to the rate of biomass loss reported by Snyder (1986) and Gillon et al. (1995), and the rates of volatilized C and N due to increasing burn temperatures (Qian et al., 2009b). There were no significant effects of burn temperature on TP, TCa, TMg, TFe, TMn, and TZn contents, which were due to their relatively high volatilization temperatures. However, the significant decrease of TK as burn temperatures increased may be due to its relatively low melting point and the nature of understory vegetation. A significant loss of K in post burn field ash in the Pine Rockland was also quantified by Snyder

PAGE 84

84 (1986). Raison et al. (1985) showed that the order of magnitude of volatilization was N > K > P > Ca, in which significant loss of non particulate K could occur at high temperature s Qian et al. (2009b) point ed out that field collected ash contained 76% less total K compa red with laboratory burned ash containing dead cattail leaves at temperatures over 450 0 C with dead cattail leaves. Moreover, loss of K (p < 0.01) due to increased temperature under controlled condition s in the laboratory was also observed by Gillon et al. (1995). pH, EC, and Water Soluble Nutrients in Residual Ashes Results of linear re gression analysis indicated that pH, EC, and the water soluble nu trient contents in residual ash significantly varied with burn temperatures, except for water soluble Ca and Mg (Table 3 1). Ash pH and EC trend ed to increase along the temperature gradient. Ash pH significantly increased from acid ic with heating temperatures at or below 250 0 C to alkaline at temperatures greater than or equal to 350 0 C (Figure 3 3c). The average pH was 5.4 at 250 0 C, significantly increased to 8. 1 at 350 0 C and finally stabilized at 450 and 550 0 C. No significant difference of ash EC was below or at 250 0 C (~ 214.3 S/cm); however, ash EC significantly increased at 350 0 C ( 813.3 S/cm), continued to increase at 450 0 C (1090 S/cm), and reached a peak a t 550 0 C (1253.3 S/cm) (Figure 3 3d). Residual ash pH and EC increased as of heating temperature increased, as a result of inorganic ion releases from the fuel combustion. S i gnificant increases in residual ash pH and EC were similar to results of recent st udies (Certini, 2005; Qian et al., 2009b). These increases were attributed to releases of base cations from organic matter combustion which could change soil pH and EC and impact the availabilities of soil nutrient pools (DeBano et al., 1998; Knoepp et al. 2005; Murphy et al., 2006a). Nevertheless, significant increases of ash pH and EC occurred only at high temperatures ( 350 0 C) which coincided with a complete fuel combustion (Arocena and Opio, 2003; Qian et al., 2009b ).

PAGE 85

85 Significant effects of burn temperatures were observed on water soluble N an d P in residual ash (Table 3 1 ). Concentration s of NH 4 N were significantly reduce d along the temperature gradient from 250 to 550 0 C, with the lowest concentration s at 450 and 550 0 C (Figure 3 3a) Heating to 250 0 C produced 13.25 g NH 4 N /g dry fuel load, whereas 1.09g NH 4 N /g dry fuel loads after heating to 550 0 C. Approximately 10.95 and 2.74 g NH 4 N were produced at heating temperatures of 350 and 450 0 C, respectively. Concentrations of NO 3 N in residual ash were significantly reduced at a heating temperature o f 3 50 0 C, from 2.4972 g/g at 250 0 C to 0.3743 g/g at 350 0 C No significance was observed for NO 3 N in ash at burn temperatures greater than or equal to 350 0 C (Figure 3 3b) The decline of water soluble NH 4 N as burn temperature increased indicated that ammonia oxidized during heating, and was lost through volatilization. The significant reduction of water soluble NO 3 N only at 350 0 C was different from results of other studies such as Gray and Dighton (20 06) and Qian et al. (2006b). This may be attributed to differences in the composition of understory species (Lavoie et al., 2010). Burning at high temperatures (350, 450, and 550 0 C) resulted in significantly less water soluble P than that at lower temperatures (Figure 3 3a). The water soluble P contents were 60.55 g/g dry fuel load at 250 0 C, 17.48 g/g dry fuel load at 350 0 C 2.32 g/g dry fuel load at 450 0 C, and 0.36 g/g dry fuel load at 550 0 C. The significant decrease of the water soluble P at t he higher burn temperatures ( 450 0 C) was similar to results reported by Qian et al. (2009b). This implied that a low fire intensity generally could produce more available P than an intense fire. With increasing burn temperatures, inorganic P ions tended t o bind to basic oxides ( Ca, Mg, Fe, Mn) in residual ash and then formed insoluble phosphates (Raison et al., 1985) Additionally, a dramatic increase in the TP/ water soluble P ratio in residual ash at higher temperatures (2.8, 10,

PAGE 86

86 75, and 524 g/g dry fuel load at 250, 350, 450, and 550 0 C respectively) indicated that inorganic phosphate became unavailable with an intense fire (Gray and Dighton, 2006). Burn temperature significantly affec ted water soluble metals in ash but the responses of each metal along the temperature gradient in term s of dry fuel load conten t were very different The c oncentration of water soluble K was significantly lower at 250 0 C than at 350, 450, or 550 0 C. The water soluble K was 429.9 g/g dry fuel load at 250 0 C, whereas K concentrations were 737.5, 950.8, and 955.4 g/g at 350, 450, and 550 0 C, respectively (Figure 3 3b). W ater soluble Mn and Zn contents gradually decreased as burn temperatures increased but Zn concentration was not significantly different between 350, 450, and 550 0 C (Figure 3 3b) There was no difference for water soluble Mn contents between 450 and 550 0 C (Figure 3 3a) Differences in melting points, boiling points, and heat of vaporization of plant nutrients generally resulted in different res pon ses to concentrations of water soluble nutrients in residual ash as burn temperatures increased. The fuel combustion converted organic compounds into basic cations and basic oxides (Qian et al., 2009b; Ulery et al., 1993a & b; Ulery et al., 1995). Interact ions of Soil Nutrient Pools Results of statistical analysis showed that the three way interaction among soil moisture, burn temperature, and incubation time were n o t significant; however, two way interaction s among these factors were significant for some nutrients (Table 3 2). The interaction between incubation time (or period) and soil moisture significantly affected extractable NH 4 N (p < 0.001), PO 4 P, K, Fe (p < 0.05), and Zn (p < 0.01). The interaction between soil moisture and b u rn temperature was significant for soil EC (p < 0.01), extractable NO 3 N, PO 4 P, Fe (p < 0.001), and Mn (p < 0.05) while the interaction between incubation time and burn temperature was only significant for extractable Mn (p < 0.01).

PAGE 87

87 Under the laboratory condition, the extractable Mn concentration was significantly higher at 250 0 C than at 350, 450, or 550 0 C at the same incubation time (Figure 3 4 ). Increasing the incubation time resulted in reduction of extractable Mn. T here was no statistical ly significan t effect of temperatures on the extrac table Mn after 180 days. E xtractable P concentration decr eased not only with increasing incubation time but also with increasing soil water content (Figure 3 5a) The highest P concentr ation of 8.4 mg/kg was observed a t soil water content of 35% field capacity in 30 days, whereas the lowest P was 4.0 m g/kg at 100% of field capacity after 180 days. The extractable NH 4 + concentrations increased with an increase in the soil water content but it decreased when the incubation time was increased, especially after 30 days (Figure 3 5c) Concentration of NH 4 + reached a peak (184 mg/kg soil) at a soil water content of 100 % field capacity in 14 days, and its lowest concentration (38 mg/kg soil) occurred at the soil wa ter content of 35% field capacity after 120 days. Responses of K, Fe, and Zn to the soil moisture over time were variable. The extractable K content was significantly reduced after 120 days and dramatically increased after 180 days at the same soil water c ontent (Figure 3 5b) Changes of the extractable Fe fluctuated with soil water content and time. In general, the extractable Fe was higher at 35 a nd 100% of field capacity than at 70% of field capaci ty at the same incubation time (Figure 3 5d) Concentrati on of e xtractable Zn did not change until 120 days, but significantly decreased after 180 days (Figure 3 5e) The burn temperature at 350 0 C and soil water content 35% of field capacity resulted in the highest extractable P content (10.6 mg/kg) whereas the lowest content ( 2.9 mg/kg ) occurred at 550 0 C and 100% of field capacity (Figure 3 6a) At the same soil water content the burn temperature of 350 0 C always resulted in a higher concentration of P The combined effect of burn temperature and soil moistu re o n the extractable NO 3 was different from those of P. The s oil water

PAGE 88

88 content of 70% field capacity always constituted greater amount of NO 3 N when the temperature was kept constant. The highest NO 3 ( approximately 203 mg/kg soil ) occur red at temperatur es 350 or 550 0 C with soil water content of 70% field capacity (Figure 3 6b ). Responses of Fe and Mn to the interaction between burn temperature and soil moisture varied A s ignificant increase of extractable Mn only occurred at the burn temperature of 250 0 C at all soil water contents The extractable Mn concentration at 250 0 C reached approximately 70 mg/kg soil, whereas Mn concentration ranged from 46 to 52 mg/kg soil at the other burn temperatures (Figure 3 6c ) The extractable Fe content was highest at 55 0 0 C and 100% of field capacity with 201.5 mg/kg soil and lowest at 250 0 C and 70% of field capacity wi th 148.5 mg/kg soil (Figure 3 6d ). Similarly, changes of EC showed the same pattern as NO 3 N. A temperature of 450 0 C and soil water content of 70% field c apacity resulted in the highest EC ( 1265.9 S/cm ) ( Figure 3 6e ). Effects of Burn Temperature on Soil Nutrient Pools B urn temperature significantly impacted soil pH, extractable NH 4 N, NO 3 N, P, K, Fe, Mn, Zn, Cu (p < 0.001), EC (p < 0.01), and extractable Mg (p < 0.05) (Table 3 2). As a result of availability of the water so luble nutrients in residual ash at different burn temperatures, changes of pH, EC, and extractable concentrations of soil NH 4 N, NO 3 N, P, K, Mg, Fe, Mn, Zn, and Cu along the temperature gradient had generally the same pattern as those of the water soluble nutrients in the laboratory residual ash. The burn temperature significantly increased extractable Mg, K, Fe, Cu, and soil pH. Generally, the burn temperature at or over 450 0 C produced more availability of Mg, K (Figure 3 7a), Fe, Cu (Figure 3 7b), and led to an increase of soil pH (Figure 3 7c). The extractable Mg was around 65 mg/kg soil at or below 350 0 C, and approximately 70.8 mg/kg soil at or higher 450 0 C. The extractabl e K was significantly different among 250, 350, and 450 0 C, but no significance was

PAGE 89

89 found between 450 and 550 0 C. The extractable K concentrations were 339.7, 401.2, 415.9, and 421.0 mg/kg soil at the burn temperatures of 250, 350, 450, and 550 0 C, respective ly. Responses of soil extractable Fe and pH to the burn temperatures had the same trend as those of extractable K. The concentrations of Fe were 158.9, 166.2, 180.2, and 181.8 mg/kg soil with respect to the burn temperature gradient, whereas soil pH was 7. 45, 7.58, 7.65, 7.67 at burn temperatures of 250, 350, 450, and 550 0 C, respectively. Extracta ble Cu increased linearly with increases in burn temperature, ranging from 2 mg/kg at 250 0 C to 2.78 mg/kg soil at 550 0 C In contrast, an increase in burn temperature resulted in decrease in extractable NH 4 N, Mn, and Zn (Figure 3 7d). A significant reduction of NH 4 N occurred when the burn temperature increased from 350 to 450 0 C. No significant change was observed for NH 4 N contents between 250 and 350 0 C or 450 and 550 0 C. The concentration s of NH 4 N ranged from 102 to 104.9 mg/kg soil at or below 350 0 C and were approximately 91 mg/kg soil at or over 450 0 C. The extractable Zn content which linearly decreased as burn temperatures increased were 4.01, 3.75, 3. 56, 3.09 mg/kg at 250, 350, 450, and 550 0 C, respectively. The burn temperature significantly decreased the extractable Mn at 350 0 C while no significant difference was found for extractable Mn between 350, 450 and 550 0 C. Changes of extractable PO 4 P, NO 3 N (Figure 3 7e), and EC (Figure 3 7f) along the burn temperature gradient were different from other soil nutrient elements, whose concentrations had the highest values at 350 0 C. E xtractable P content was 7.2 mg/kg soil at 250 0 C, reached its maximum concentr ation of 9.04 mg/kg soil at 350 0 C, and decreased gradually to 6.06 and 3.31 mg/kg soil at 450 and 550 0 C, respectively. The extractable NO 3 N content at 350 0 C was 168 mg/kg whereas NO 3 N contents were 122.2, 132.0, and 129.6 mg/kg soil at 250, 450, and 550 0 C, respectively. The highest value of EC was 1134.1 S/cm at 350 0 C, whereas the values of EC were 959.2, 1089.1 and 1040.8 S/cm at 250, 450, and 550 0 C.

PAGE 90

90 Various responses of soil nutrient pools along the temperatur e gradient depended on availability of nut rient s in the residual ash which were attributable to differences in their melting points, boiling points, and heating points of vaporization. Unlike in a field fire, leaching and plant uptake did not influence the soil nutrient pools under the controlled condition s in the laboratory, alterations of soil nutrient pools with re spect to the burn temperature t ended to be the same as in the residual ash. Thus, changes of the nutrient pools were directly related to the amount of nutrients in ash deposited into the soil (Badia and Marti, 2003). Effects of Soil Water Content on Soil Nutrient Pools Overall, soil water content significantly impacted soil pH, EC (p < 0.01), and extractable NH 4 N, NO 3 N, PO 4 P, Fe (p < 0.001), but had no effect on extractable Mg, K, Mn, Zn, or Cu (Table 3 2). Under the laboratory conditions extractable P significantly decreased with an increase in soil water content (Figure 3 8a). The highest P concentration was 7.36 mg/kg at soil water content of 35% field capacity, whereas the lowe st P was 5.39 mg/kg at 100% of field capacity. A soil water content of 70% field capacity resulted in 6.46 mg/kg soil of extractable P. However, an increase in the soil water content significantly increased extractable NH 4 + (Figure 3 8b). The greatest conc entration of NH 4 N was 119.3 mg/kg soil when the soil moisture was 10 0% of field capacity, whereas the lowest concentration was 71.7 mg/kg soil occurred at the soil water content of 35% field capacity. The soil with water content of 70% field capacity had 101.3 mg NH 4 N/kg soil. Responses of NO 3 N to soil water content showed the same tendency as EC Soil extractable NO 3 N and EC reached a peak at a soil water content of 70% field capacity. The extractable NO 3 concentration was 128.4 mg/kg with soil water content at 35% field capacity, significantly increased to 188.7 m g/kg at 70% of field capacity, and decreased to 96.7 mg/kg

PAGE 91

91 when soil water content was 100% of field capacity (Figure 3 8c) The EC values were 976.3 and 988.2 S/cm with soil water content s of 35 and 100% field capacity, respectively, whereas EC was 1202.8 S/cm at 70% field capacity (Figure 3 8d) Extractable Fe was significantly lower at a soil water content of 70% field capacity (162.2 mg/kg soil ) than at the soil moisture contents of 35% (169.9 mg/kg soil) or 100% field capacity (183.3 mg/kg soil ) (Figure 3 8e) Although soil pH at a soil water content of 70% field capacity was lower than at 35% of field capacity, the dif ference was not statistically (Figure 3 8f) Soil pH significantly in creased when soil water content was at 100% field capacity due to increasing CO 2 concentration under anaerobic condition (Glinski et al., 1992; Yan et al., 1996). The soil with water content of 100% field capacity had the highest con centrations of extracta ble NH 4 N and Fe, and pH. The high soil water content promoted a denitrification process which only occurred at the soil water content greater than 60% of air filled pore space (van Cleemput, 1998) When soils become more anaerobic, the high soil moisture was favorable for Fe 2+ oxidizing bacteria to enhance Fe 3+ reduction and produce more Fe 2+ (Megonigal et al., 2004 ) Nitrate accumulated over time as a result of the nitrification process (Misra and T yler, 1999; Tyler, 1996); however, availability of NO 3 N in term s of soil water content was also dependent on Fe 2+ concentration because an increase of Fe 2+ content in the soil solution could enhance the rate of nitrate reduction (van Cleemput, 1998) Therefore, the con centration of NO 3 N was considerably lower at soil water content of 100% field capacity that at 35 or 70% field capacity as a result of an increase in the concentration of Fe 2+ C omparatively high concentrations of P at the low soil moisture denoted that higher soil water content favored reactions among P with Ca and Mg to form Ca/Mg P precipitates.

PAGE 92

92 Effects of Incubation Time on Soil Nutrient Pools Th e incubation time ( or period) significantly affected soil EC and extractable P O 4 P NH 4 N, NO 3 N, Mg, K, Mn, Cu, and Zn (p < 0.001), but not extractable Fe or soil pH (Table 3 2). Ammonia content linearly declined over the time, and stabilized after 120 days. Conversely, a reduction of NH 4 + resulted in an increase of extractable NO 3 N which was positive ly correlated with incubation time (Figure 3 9 a). Significant changes of soil EC and extractable P, Mg, K, Mn, Zn, and Cu actually occurred at 120 days or thereafter. The extractable P O 4 P, Zn (Figure 3 9 b) and Mg, Mn (Figure 3 9 c) were significantly decreased when the incubation time was raised from 120 to 180 days while the extractable Cu (Figure 3 9 b) was increased after 120 days. The extractable K was greatly reduced at 120 days comp ared to 90 days, but higher than its original v alue after 180 days (Figures 3 9 c). Soil EC was significantly higher at 120 days compared to its original value at 14 days, and was then st able after 120 days (Figure 3 9 d). Increasing in incubation time resul ted in a reduction of extractable P, NH 4 N, Mg, Mn, and Zn, and an incre ase of soil EC and extractable K, Cu, and NO 3 N. Changes of soil pH, EC, and extractable nutrients occurred a t or after 120 days indicated that time is an important factor contributing to availability of soil nutrient pools following a fire. Soil nutrient concentrations changed simultaneously at 120 days indicated that there was a dynamic interaction among soil chemical components and biogeochemical reactions at this time. Estimation of Field Fire Temperature from Residual Ashes N o significant effect of fire temperature was observed for TP/TCa, TP/TMg, or TP/TFe ratios. This indicated that P, Ca, Mg, and Fe accumulated in both laboratory simulated ash and in ash collected from the field. Hence, contents of TP, TCa, TFe, and TMg in residual ash could be utilized as predictor variables for estimating field fire tem peratures. Figure 3 10 showed TP, TCa,

PAGE 93

93 TMg, and TFe contents accumulated in laboratory residual ash and in fiel d ash. Accumulation of residual ash TP, TCa, TMg, and TFe increased as burn temperatures increased. The TP was 0.30 mg/g ash at 250 0 C and increas ed to 2.6 mg/g ash at 550 0 C. The TFe content was 0.48 mg/g ash at 250 0 C and 4.8 mg/g ash at 550 0 C. Calcium content was highest in residual ash with 20.2 mg/g ash at 250 0 C and 155.89 mg/g ash at 550 0 C, whereas Mg content was lower than that of Ca, having 1.6 mg/g ash at 250 0 C and 12.0 mg/g ash after heating to 550 0 C. Concentrations of TP, TCa, TMg, and TFe in the field ashes were 1.11, 76.96, 6.15, and 1.82 mg/g ash (Table 3 3). Fitted linear regression models of TP, TCa, TMg, and TFe selected for predict ion of a field fire temperature were presented in Table 3 3. These models demonstrated a good relationship between burn temperature s and fitted temperatures with R square values over 0.95 (Figure 3 1 1). From simulated temperatures, an average field fire te mperature was estimated to be approximately 370 0 C. The high R square values in the goodness of fit models and no significant difference observed in ratios of TP/TCa, TP/TMg, and TP/TFe indicated that P, Ca, Mg, and Fe contents in ash were good in dicators f or predicting a field fire temperature. The estimated temperature of the field fire was similar to the results reported by Qian et al. (2009b) and Snyder (1986), but was higher than the temperature measured by Sah et al. (2006) and lower than the temperatu re determined by Snyder et al. (2005). The difference in fire temperatures may have been due to differences in understory vegetation communities and the burned seasons among studies. The fire experiments of Snyder et al. (2005) and Sah et al. (2006) were c arried out in July and December and conducted in the Big Pine Key and the National Deer Refuge in the Florida Keys, respectively, which were dominated by West Indian tropical hardwoods and a diverse herb layer. The present study was implemented in the Long Pine Key in October which h as mostly dominated by grasses and palm trees in the understory vegetation.

PAGE 94

94 Summary Results of the study showed that the burn temperature, soil water content, and time after burn had significant impacts on soil nutrient pools following a fire. The interaction among incubation time, soil moisture content, and bu rn temperature did not impact the availability of the nutrients; however, the interaction bet ween soil water content and burn temperature significantly affected soil EC, extractable NO 3 N, PO 4 P, Fe, and Mn while the interaction between incubation time and burn temperature was only significant for extractable Mn. A burn temperature of 350 0 C and so il water content of 35% field capacity resulted in the highest extractable P content The highest NO 3 concentration occurred at or over 350 0 C and 70% field capacity. The interaction between the burn temperature of 350 0 C or higher and soil water content of 70% field capacity resulted in the highest EC. The highest extractable Fe was at or over 450 0 C and 100% field capacity. The burn temperature at below 350 0 C produced the greater concentration of extractable Mn over soil moisture and time after burn. B ur n temperature gradient which modified the quantity of elemental nutrients in r esidual ash, significantly changed soil nutrient pools. An increase in the burn temperature resulted in an increase of soil pH and extractable Mg, K, Fe, Cu, and the decrease of extractable NH 4 N, Mn, and Zn, whereas the burn temperature of 350 0 C constituted the highest concentrations of PO 4 P, NO 3 N, and EC. The soil water content had no effect on the extractable Mg, K, Mn, Zn, and Cu; however, it significantly changed soil pH, EC, and availability of NH 4 N, NO 3 N, and PO 4 P. Increasing in the soil water content increased the soil pH and availability of NH 4 N, and decreased the availability of PO 4 P. The extractable NO 3 N and soil EC were highest at soil water content of 70% field capacity, whereas this moisture produced the lowest extractable Fe. Incubation time did not influence extractable Fe and soil pH, but it significantly changed the extractable NH 4 N, NO 3

PAGE 95

95 N, P, Mg, K, Mn, Zn, Cu, and EC. The increase of incubation time caused a linear reduction of extractable NH 4 N and an increase of NO 3 N. Significant alterations of soil EC and availability of P, Mg, K, Mn, Zn, and Cu occurred at or after 120 days of burning in which the extractable P, Mg, Mn, and Zn were decreased w hile those of K and Cu were increased. The burn temperature significantly impacted extractable contents of all soil nutrients indicating that difference of fire temperatures between different burn seasons will produce different availabilities of soil nutr ient pools after a fire. The significant effects of the interactions between soil water content and burn temperature on extractable NO 3 N, PO 4 P, Fe, and between incubation time and burn temperature on extractable Mn implied that soil water content and tim e after a fire play important role s in the availability of soil nutrient pools following a fire. The dramatic increase in the ratio of TP to inorganic P in residual ash as fire intensity increased indicated that a low intensity fire generally can produce more available P than a high intensity fire because inorganic P ions tend to b i nd to basic oxides (Ca, Mg, Fe, Mn) in residual ash, and then form insoluble phosphates when the fire intensity increases The reduction in extractable P c ontent as an increase in the soil water content indicated that higher soil water content favored for reactions among P with Ca and Mg or Fe and Mn to form Ca/Mg P precipitates or Fe/Mn P hydroxyl compounds

PAGE 96

96 Table 3 1 Results of linear regression analysis on pH, electrical conductivity (EC), and total and water soluble nutrients in laboratory residual ash Total (T) Sig. level Water soluble Sig. level pH EC N *** NH 4 N *** ** *** C *** NO 3 N ** P NS PO4 P *** Ca NS Ca NS Mg NS K *** K *** Mg NS Fe NS Mn *** Mn NS Zn *** Zn NS p < 0.05, ** p < 0.01, *** p < 0.001, NS: not significant Table 3 2 Results of statistical analysis on incubation time (or period), soil water content burn temperature, and their interactions Variables Period (P) Moisture (M) Temperature (T) P*M M*T P*T P*M*T pH NS ** *** NS NS NS NS EC *** ** ** NS ** NS NS NH 4 N *** *** *** *** NS NS NS NO 3 N *** *** *** NS *** NS NS PO 4 P *** *** *** *** NS NS Mg *** NS NS NS NS NS K *** NS *** NS NS NS Fe NS *** *** *** NS NS Mn *** NS *** NS ** NS Zn *** NS *** ** NS NS NS Cu *** NS *** NS NS NS NS p < 0.05, ** p < 0.01, *** p < 0.001, NS: not significant

PAGE 97

97 Table 3 3. Estimation of a field fire temperature from laboratory and field residual ashes Predictor variables The fitted linear regression models were selected for prediction of P availability after a fire Field burning ashes Content ( g g ) Simulated T ( 0 C ) TP Y = 171.26 + 0.30356*TP 0.00014573*(TP) 2 + 3.2363E 8*(TP) 3 1110.2 372.9 TCa Y = 214.04 + 0.00199*TCa 76964.2 367.2 TFe Y = 240.22 + 0.063*TFe 1821.8 355.0 TMg Y = 154.2 + 0.07195*TMg 0.000008*(TMg) 2 + 3.9495E 10*(TMg) 3 6149.5 385.9 Average field fire temperature : 370 0 C Y is a function of the laboratory heated temperatures, or a burn temperature, with independent variable of TP, TCa, TFe, or TMg (g g of ash) from ash products; T is a temperature

PAGE 98

98 Figure 3 1. Losses of total carbon (TC), total nitrogen (TN), and biomass in fuel load at different burn temperatures Figure 3 2. Effects of burn temperature on total K in residual ash in term of dry fuel load

PAGE 99

99 Figure 3 3 Effects of burn temperatures on water soluble nutrients in residual ash in term of dry fuel load A ) PO 4 P, NH 4 N, Mn; B ) NO 3 N, K, Zn; C) pH; and D ) EC Figure 3 4. The effect of burn temperature and incubation time on soil extractable Mn 14 B A C D

PAGE 100

100 Figure 3 5 The effect of soil moisture and incubation tim e on soil extractable nutrients. A ) PO 4 P, B) K, C ) NH 4 N, D ) Fe, and E ) Zn E A B C D 1 4 14 14 14 14

PAGE 101

101 Figure 3 6 The effect of soil moisture and burn temperature o n soil extractable nutrients. A ) PO 4 P, B ) NO 3 N, C) Mn, D ) Fe, and E ) EC E A B C D

PAGE 102

102 A B C D E F Burn Temperature ( 0 C) Figure 3 7 Effects of burn temperature on extractable concentration of soil nutrients. A) Mg, K; B) Fe, Cu; C) pH; D ) NH 4 N, Mn, Zn; E ) PO 4 P, NO 3 N; and F ) EC

PAGE 103

103 A C D B E F Soil moisture (% of field capacity) Figure 3 8 Effects of soil moisture on extractable concentrations of soil nutrients. A ) PO 4 P, B ) NH 4 N, C ) NO 3 N, D) EC, E) Fe, and F ) pH

PAGE 104

104 Incubation time (days) Figure 3 9 Effects of incubation time on extractable c oncentration of soil nutrients. A ) NH 4 N and NO 3 N; B ) PO 4 P, Zn, and Cu ; C ) K, Mg, and Mn; and D ) EC D 14 14 14 14 A B C

PAGE 105

105 Figure 3 10. Concentrations of TP, TCa, TMg, and TFe in laboratory and field ashes Figure 3 11. Goodness of fit models for fire temperature prediction based on TP, TCa, TMg, and TFe in residual ash 250 350 450 550 250 350 450 550 Fitted temperature of models ( 0 C) Laboratory heating temperature ( 0 C) Fitted temperature for TP (R square = 0.9971) Fitted temperature for TCa (R square = 0.9526) Fitted temperature for TFe (R square = 0.9623) Fitted temperature for TMg (R square = 0.9955)

PAGE 106

106 CHAPTER 4 PREDICTING PHOSPHORUS AVAILABILITY IN CALCAREOUS SOILS IN A PINE ROCKLAND FOREST AFTER A PRESCRIBED FIRE Phosphorous (P) is probably the second most limiting nutrient in natural forest ecosystems after nitrogen (N) (DeBano et al., 1998; Pierzynski et al., 2000). Common inorganic compounds of P in so ils often fall into one of two main groups: those containing calcium; and those containing iron and aluminum (Brady and Weil, 2002). Calcium phosphate compounds become more soluble as soil pH decreases; hence, they tend to dissolve and disappear in acid soils. These compounds are quite stable and less insoluble at higher pH, and become dominant forms of P compounds present in neutral and alkaline soils. In contrast to calcium phosphates, hydroxyl phosphate minerals of iron and aluminum, strengite (FePO 4 .2H 2 O) and variscite (AlPO 4 .2H 2 O) have very low solubility in strongly acid soils and become more soluble as soil pH rises. These minerals are unstable in alkaline soils, but they are predominant in acid ic soils. In both acid ic and alkali ne soils, therefore, phosphorus tends to undergo seq uential reactions that produce P containing compounds of lower solubility and less bioavailability for plant uptake. P hosphate minerals, as means of reducing P bio availability in soils, are described in more detail by Pierzynski et al. (2005). Orthophosph ate which consist of H 3 PO 4 H 2 PO 4 HPO 4 2 and PO 4 3 is a bioavailable P form in soil solution. The presence of these ions in soil solution is a function of soil pH (Brady and Weil, 2002). H 2 PO 4 and HPO 4 2 are two dominant P species in most natural fores t soils. Species of H 2 PO 4 are predominant in soils with neutral pH ranging from 4 to 7.2, whereas HPO 4 2 is dominant in soils with pH greater than 7.2 (Pierzynski et al., 2005). Fire often produces a substantial amount of P for soils not only by providing orthophosphate ions from ash residual deposit during a fire but also converting soil organic P compounds into orthophosphate as well (Cade Menum et al., 2002; Certini, 2005; Shar pley and

PAGE 107

107 Moyer, 2000). Ash released from fuel combustion contain a large amount of orthophosphate (Khanna et al., 1994; Durgin and Vogelsang 1984) resulting in an increase of P bioavailable in acidic ( Giardina et al., 2000; Knoepp et al. 2005; Serrasolsas 1995b) and in calcareous alkaline soils (Ubeda et al., 2005; Hernandez et al., 1997). The q uality and quantity of orthophosphate in residual ash depend wholly on fire intensity during a fire. Romanya et al. (1994) pointed out that the content of soil ext ractable P after the fire was higher than its pre fire level. Surface reactions and precipitation are two main mechanisms immediately affecting the fate of orthophosphate ions in calcareous soils after a fire (von Wandruszka, 2006) Main products of surface reactions are dicalcium phosphate (DCP) and octacalcium phosphate (OCP). In calcareous soils, calcium carbonate ( calcite) plays an essential role in P retention when concentration of P is relatively high in the soil solution, but non carbonate clays are important to retention of P at lower P concentration s (von Wandruszka, 2006; Zhou and Li, 2001) Afif et al. (1993) found that P availability in calcareous soils is negatively correlated with the amount of lime, whereas Castro and Torrent (1994) indicated that P retention increased with the ratio of Fe oxides to CaCO 3 Hamad et al. (1992) reported that 1.6 % of a coating rate between Fe 2 O 3 and calcite increased P retention by 9 fold. By comparing the relative importance of surf ace reaction and precipitation in P retention, Tunesi et al. (1999) co ncluded that precipitation was the dominant mechanism in reduction of P availability in soils with high exchangeable cations. Precipitation of phosphorus can be defined as a formation o f discrete, insoluble P compounds in soils, and can be viewed as a reverse of mineral dissolution. With the high pH of calcareous soils, P binds to Ca/Mg minerals or precipitates as discrete Ca/Mg phosphate (Brady and Weil, 2002; Knoepp et al., 2005). The a bundance of soluble Ca ions controls activities of inorganic P ions in soil solution after a fire resulting in formation of dicalcium phosphate

PAGE 108

108 dehydrate (CaHPO 4 .2H 2 O) which later reverts to others more stable Ca phosphates such as octacalcium phosphate ( Ca 4 H(PO 4 ) 3 .2.5H 2 O), then hydroxyapatite ( Ca 5 (PO 4 ) 3 (OH)) and finally transform s into apatite (Ca 10 (PO 4 ) 6 F 2 ). Additionally, soil water content, which influences soil solution chemistry, also contributes considerably to availability of P after a fire. Under a field condition, soil water content often fluctuates with rainfall and air temperature. Fluctuations of soil moisture affect availa bility of nutrients in calcareous soils (Gahoonia et al., 1994; Misra and Tyler, 1999). Calcareous soils are generally characterized by high contents of calcite (CaCO 3 ) and bicarbonate ions (HCO 3 ). Increasing soil moisture causes an increase in concentra tions of soil solution HCO 3 ions (Inskeep and Bloom, 1986; Mengel et al., 1984). A soil moisture curve is a useful tool in determining hydraulic properties of an unsaturated soil. It describes a relationship between soil water content ( ) and soil water potential ( ), and is often utilized to predict soil water storage and plant water supply, and is especially effective for determining field capacity and wilting point of a given soil. Water holding capacity of soil is dependent on soil porosity and nature of bonding in soils; therefore, the soil moisture curve is specific to different types of soils, and depends on soil text ure and structure, organic matter content, and composition of solution phase (Muoz Carpena et al., 2002; Pham and Fredlund, 2008). Although shape of a water retention curve can be characterized by several models, the Mualem van Genuchten model is known as fitting well in different soils (Schaap and van Genuchten, 2006). Availability of P in calcareous soils depends on persistence of metastable calcium phosphate minerals (Fixen et al., 1983). By continuously applying fertilizers in two different calcareous soils in Colorado through three years, Fixen et al. (1983) proved that solubility of P was controlled by octacalcium phosphate (OCP) if concentration of Olsen P exceeded 35 mg/kg soil,

PAGE 109

109 whereas in a range of 10 to 25 mg Olsen P/kg soil, tricalcium phosphat e (TCP) inhibited solubility of P. In calcareous soils, transformation of Ca P minerals derives from monocalcium phosphate (MCP) to hydroxyl apatite (HA); therefore, a solubility order of Ca P minerals are MCP > dicalcium phosphate dihydrate (DCPD) > TCP > HA (Brady and Weil, 2002, p. 612; Pierzynski, et al., 2005). However, kinetics of Ca P transformation could be influenced by several factors. Transformation of DCPD to OCP can slow with the presence of magnesium (Bell and Black, 1970), and decreases with the existence of soluble organic matter (Moreno et al., 1960). Pine Rockland soils have a low content of carbonate clay and a high content of non carbonate clay. According to a USDA soil survey in 1958, the pineland Rockdale soil contained approximately 9 6% of non carbonate clay and 4% of carbonate clay. Zhou and Li (2001) quantified that the Pine Rockland soil contained the total clay of approximately 730 g/kg soil in which the content of non carbonate clay was 707 g/kg soil and the amount of carbonate cl ay was 23 g/kg soil. Because of relatively low clay content, clay minerals are not significant in adsorption of P, but precipitation among P with Ca and Mg is the dominant process affecting retention of P in calcareous soils in the Pine Rockland (Zhang et al., 2001). In addition, since available contents of Fe and Mn are rather high in calcareous soils in South Florida (Hanlon et al., 1996; Zinati et al., 2001), reactions between orthophosphate ions with dissolved Fe and Mn ions forming hydroxyl phosphate p Identification of phosphate minerals as a result of weathering P bearing minerals in parent materials has been documented by previous studies (Lindsay et al., 1989; Karath anasis, 1991). A direct determination of naturally occurring phosphate minerals in calcareous soils in Pine Rockland forests is generally very difficult because of the relatively low P content and complexity of phosphate minerals. Minteq, a geochemical ass essment tool for environment

PAGE 110

110 systems, applies equilibrium speciation models to simulate species of soil chemical components of determining a magnitude order of p hosphate minerals controlling availability of P in soil solution (Allison et al., 1991). A phosphate solubility diagram, which is constructed from ionic solubility of phosphate minerals in soil environments. The diagram is particularly useful for determining a relative stability of phosphate compounds and minerals in soils at various pH values. Not only is P equilibrium a function of pH in the diagram, but it also depe nds on the ionic activities of Ca 2+ Mg 2+ Mn 2+ and Fe 2+ in soil solution (Karathanasis, 1991). In calcareous soils, phosphate mineral solubility is mostly controlled by Ca/Mg P minerals (Fixen et al., 1983; Zhang et al., 2001). Although Mn P equilibrium relationships are often super saturations with respect to MnHPO 4 and MnPO 4 :1.5H 2 O, these minerals could also participate in controlling solubility of P in calcareous soils when soil solution concentration of P is very low (Karathanasis, 1991, Lindsay et al ., 1989). Effects of prescribed fire on vegetation in Pine Rockland forest have been well documented (Ross et al., 2003; Sah et al., 2004 and 2006; Snyder 1986; Snyder et al., 2005). Results of these studies showed that the fire regime, fuel loads, fire in tensity, and burn season significantly impacted the post fire recovery of the understory vegetation, but were inconclusive for soil P dynamics. The main objectiv e of this study was to predict the long term impact of fire on P availability in calcareous soi ls in a Pine Rockland forest. Specific tasks included: 1) speciation program; 2) describing a relationship of soil pH and phosphate minerals to phosphate solubility after a fire by establishing stability diagrams of phosphorus; 3) modeling a best fit

PAGE 111

111 predictive model for P availability after a fire. Materials and Methods Simulation of Extractable Concentration of P compounds after the Fire Results from the field burning experiment described in Chapter 2 were utilized to simulate concentrations of P compounds and construct stability diagrams of phosphorus after the fire. Table 4 1 show s soil chemical components used to characterize P compounds and their model. These components comprised soil pH, ionic strength (IS), and concentrations of extracted PO 4 P, Ca 2+ Mg 2+ K + Na + Fe 2+ Mn 2+ Zn 2+ Cu 2+ NH 4 N and NO 3 N. Aluminum was excluded be cause of its low concentration at high pH. Ionic strength (IS) was estimated from electrical conductivity (EC) using the Marion Babcock equation: Log (IS ) = 1.159 + Log (EC), where IS is in unit of mol/m 3 and EC is in dS/m or mS/cm (Essington, 2004). The simulation process was carried out three primary steps including (i) specification of CO 2 partial pressure; (ii) specification of possible solid species included into the speciation models; and (iii) input of extracted concentrations of solution components. The atmospheric CO 2 partial pressure was 0.00038 atm or 3.42 for a logarithm of CO 2 partial pressure, which is l. This pressure is the most suitable in equilibrium between calcite and CO 2 (gas) at high soil pH, especially in calcareous soils. The possible solid species were tricalcium phosphate (Ca 3 (PO 4 ) 2 am1, am2 beta), octocalcium phosphate (Ca 4 H(PO 4 ) 3 :2.5H 2 O) monetite (CaHPO 4 ); dicalcium phosphate dihydrate (CaHPO 4 :2H 2 O), calcite (CaCO 3 ), hydroxyapatite, Mg 3 (PO 4 ) 2 MgHPO 4 :3H 2 O, Mn 3 (PO 4 ) 2 MnHPO 4 strengite, and vivianite, which could be involved in the speciation of P compounds and in controlling

PAGE 112

112 availability of P as well. The extracted solution components, as described in Table 4 1, were cations and anions in soil so lution participating directly in speciation processes. The initial input for the speciation process is displayed in Appendix J. A total of forty five data points speciation model. All simulation processes were carried out at 25 0 C, the average temperature in the Everglades National Park over 30 years. Extractable concentrations of P compounds wer e computed based on their percentage distributions with respect to the extracted P contents, which resulted from the equilibrium speciation model. Additionally, ionic activities of soil solution components and saturation indices of Ca/Mg/Mn/Fe P minerals w ere obtained from results of the speciation process. Construction of Stability Diagrams of Phosphorus after the Fire Table 4 2 presented relevant information for eleven phosphate minerals and two orthophosphate ions that are potential in calcareous soils. Equilibrium reactions and solubility product constants of Ca P minerals were obtained from Pierzynski et al. (2005), whereas those of other minerals were received from the Minteq geochemical program. Relationships among the activity of HPO 4 2 with activit ies of H + Ca 2+ Mg 2+ Fe 2+ and Mn 2+ for the selected minerals were established based on equilibrium reactions and solubility product constants. Calculations of these relationships are fully described in Appendix K. The following example shows steps involv ed in comput ing a relationship between activities of HPO 4 2 and Ca 2+ for the monocalcium P mineral. Ca(H 2 PO 4 ) 2 :H 2 O Ca 2+ + 2H 2 PO 4 + H 2 O K s = 10 1.15 2(H 2 PO 4 HPO 4 2 + H + ) K s = (10 7.2 ) 2 Ca(H 2 PO 4 ) 2 :H 2 O Ca 2+ + 2HPO 4 2 + 2H + + H 2 O K s = 10 15.55 K s = 10 15.55 = +

PAGE 113

113 15.55 = log(Ca 2+ ) + 2log(HPO 4 2 ) + 2log(H + ) 15.55 = pCa 2+ + 2pHPO 4 2 + 2pH pHPO 4 2 = 7.775 0.5pCa 2+ pH Stability diagrams of phosphorus plotted between pHPO 4 2 versus pH with presence of the selected phosphate minerals were constructed using average values of ionic activities of HPO 4 2 Ca 2+ Fe 2+ Mg 2+ and Mn 2+ in soil solution following the fire. These ionic activities used to develop the diag addition, a correlation matrix among pH + pHPO 4 2 pH 1/2pFe 2+ pH 1/2pMn 2+ pH 1/2pCa 2+ and pH 1/2pMg 2+ was also considered to determine a further relationship among pH and activ ity of HPO 4 2 with ionic activities of Ca 2+ Mg 2+ Fe 2+ and Mn 2+ in soil solution. Soil Moisture Curve for Pine Rockland Soils Determination of a soil moisture curve for the Pine Rockland soil was done in the laboratory. The Tempe cell and Richard plate methods were used to measure water quality in soils at different suctions. Fourteen different pressures (suctions), which included 0, 0.49, 0.98, 1.47, 2.45, 3.43, 4.90, 6.67, 14.71, 29.41, 58.82, 98.04, 500, and 800 kPa, were applied for determination of soil moisture curve. A Tempe cell method was applied at lower suctions (less than 100 kPa), whereas a Richard plate method was used at higher suction s (higher than 100 kPa). The soil moisture curve was f itted by the Retention curve (Retc) model based on volumetric water content (the amount of water per cubic unit of soil) and soil suction or matrix potential (the amount of water held by capillarity against gravity force) (Muoz Carpena et al., 2006; vanGe nuchten et al., 1991). Procedures for a measurement of the volumetric water content in the laboratory are described in Appendix L. The shape and equation of the soil water retention curve was mainly characterized by the

PAGE 114

114 Mualem van Genuchten equation: ( ) = r + w here ( ) is a water retention curve (L 3 /L 3 ), is suction pressure (cm of water), s is saturated water content (L 3 /L 3 ), and r measure of pore size distribution (vanGenuchten et al., 1991). Initial characterization of soil for the Retc model was a gravely loam (Al Yahyai et al., 2006). Hydraulic p roperties of an unsaturated soil in the Retc model were based on the modified Mualem van Genuchten formulation as described by Schaap and vanGenuchten (2005). The observed water content from the laboratory experiment and the fitted water content from the m odel are presented in Appendix M, and results and parameters used to fit the water retention curve are shown in Appendix N. A goodness of fit of model was examined by three different methods: (i) coefficient of determination; (ii) coefficient of efficiency ; and (iii) index of agreement (Legates and McCabe, 1999; Krause et al., 2005). Coefficient of determination (R 2 correlation coefficient, and expresses a proportion of a total variance in the observation data which ca n be explained by the model. Coefficient of determination was obtained from a linear regression between the observational data (O) and the fitted data (P). Coefficient of efficiency (E) is a ratio of a mean square error of variance of the observation data, subtracted from unity (Legates and McCabe, 1999), and expressed as: E = 1.0 observational mean. Index of agreement (d) represents a ratio of a mean square error and potential error (Krause et al., 20 05), and is described as: d = 1 Higher values of R 2 d, and E indicate a better agreement of the model. Model ing for Prediction of P Availability after the Fire Data received from results of the laboratory simulation experiment in Chapter 3 were

PAGE 115

115 used to establish a predictive model for P availability after a fire. Data set of the soil water content (35, 70, 100% of field capacity), the burn temperature of fuel loa ds or fire intensity (250, 350, 450, 550 0 C) and the incubation time ( time after burn ) (14, 30, 60, 90, 120, and 180 days) were utilized to develop a multiple regression model for evaluating a long term effect of fire on P availability. Extractable P concen tration was assigned as a response variable, whereas predictors (independent variables) consisted of soil moisture content, burn temperature, and incubation time. The method of a stepwise, multiple regression analysis using PROC GLM procedures (SAS Institu te Inc., Cary, North Carolina) were applied for fitting the predictive models of P availability. Fitting of the regression models was based on a combination of orders (1 st 2 nd 3 rd ), two way and three way interactions among three selected predictors. A full regression model of each combination was fitted to determine significant components to be included in the model. A reduced model would be considered if there was any component in the full model that was not significant. Insignificant components were d etermined by F not significantly important in the full model, a full model without the intercept was also fitted. Selection criteria for choosing the best model to predict availability of phosphorus after a fire was based on a model having a highest R square value, and lowest values of variation coefficient and root MSE (mean standard error). Results of fitting of multiple regression models were enclosed in Appendix O. In addition, three basic assumptions of multiple regression analysis : constant variance, normality of errors, and independence of errors were examined to validate a selected predictive model. A plot of predicted values versus residuals was performed to check for homogeneity of variance for the selected model. A normal probability plot of residuals (NNP plot) was used to detect non normality of errors. Independence of errors was checked by a

PAGE 116

116 Durbin Watson test statistic (Appendix P). The selected predicti ve model was further evaluated by analysis of goodness of fit measures. The goodness of fit measures for the selected model was assessed by three various methods as described in the previous section. In efforts to verify the selected model, a field burnin g experiment was carried out on the Pine Rockland forest at the Tropical Research and Education Center (TREC), University of Florida, Homestead, FL. A 5 acre plot of Pine Rockland forest was burned on May 5 th 2010. Post fire ash production was collected immediately after the fire to analyze contents of TN and TC in the ashes. Post fire soil samples were collected at 7, 14, 21, 30, 60, and 90 days with 20 replicates for each sampling time in order to measure extractable P concentrations. Moreover, a rain g auge was also installed into the burned location to record precipitation after the fire. Contents of ash TN and TC were used to estimate a field fire temperature (Qian et al., 2009b), whereas rainfall data recorded from rain gauge w ere used to estimate soi l moisture content following the fire. contents of TN and TC in ashes (Qian et al., 2009b) and a linear regression relationship between contents of TN and TC and heating temperatures of the laboratory incubation experiment described in Chapter 3. The soil water contents in 0 5cm topsoil at the sampling periods after the fire were computed from differences of water amounts between precipitation recorded from the rain gauge and evapotranspiration (ET) obtained from the website of the Florida Automated Weather Network (FAWN) at TREC (FAWN, 2011). The soil moisture content was calculated from daily rain fall If the water amount of the daily rain event was higher than the water amount of soil field capacity estimated from the soil retention curve, water amount at the field capacity was counted for that day. Detailed information of calculation of soil water content is described in Appendix Q.

PAGE 117

117 Statistical Analysis Significant differences i n concentrations of P compounds and ionic activities of solution components following the fire were examined by the Repeated Measures method as described in t difference of extractable concentrations of P compounds at different sampling times after the fire. Additionally, a regression correlation matrix among activities of solution nutrient elements and among pH + pHPO 4 2 pH 1/2pFe 2+ pH 1/2pMn 2+ pH 1/ 2pCa 2+ and pH 1/2pMg 2+ using the method of principle component analysis were performed to check correlation relationship among them in soil solution Results and Discussion Phosphorus Speciation after the Fire Minteq program showed fifteen different P compounds in which nine P species were dominant in the calcareous soils including two orthophosphate species of HPO 4 2 and H 2 PO 4 and seven other species of FeHPO 4 (aq), MgHPO 4 (aq), CaHPO 4 (aq), MnHPO 4 (aq), FeH 2 PO 4 + CaH 2 PO 4 + and CaPO 4 Simulated concentrations of these species were significantly different after the fire, except FeH 2 PO 4 + (p = 0.1756) (Table 4 3). Extractable concentrations of HPO 4 2 H 2 PO 4 and ortho phosphat e simulated from the Minteq are shown in Figure 4 1. Contents of HPO 4 2 following the fire, which were always much higher than those of H 2 PO 4 largely fluctuated with changes of extractable P content; whereas the concentration of H 2 PO 4 was less varied along with extractable P concentration. The concentration of HPO 4 2 wa s 0.23 mg/kg soil prior to burning level and significantly increased to 0.72 mg/kg soil 14 days after the fire, and returned to the pre burn value in 30 days. C ontent of HPO 4 2 0.073 mg/kg soil, significantly decreased in 90 days which was close to conte nt of H 2 PO 4 Orthophosphate content, which is

PAGE 118

118 defined as a sum of HPO 4 2 and H 2 PO 4 also varied in the same pattern as extractable P content. Table 4 3 shows the average concentrations and results of the multiple comparisons of P compounds associated with Ca, Mg, Fe, and Mn following the fire. Most of these species were significantly increased in 14 days and decreased to their pre fire values in 30 days; howev er c oncentration of CaH 2 PO 4 + only significantly decreased in 90 days, and returned to its initial value in 180 days after the fire. Among these species, FeHPO 4 (aq) and CaHPO 4 (aq) were two predominant compounds of P in soils with pre burn contents of 1.57 mg/kg soil for FeHPO 4 (aq) and 1.44 mg/kg soil for CaHPO 4 (aq). Although the pre burn concentration of MnHPO 4 (aq) was 0.4 mg/kg soil, it reached to 3.68 mg/kg soil in 14 days which was the highest content compared to the content of other P compounds at that time. The fire significantly impacted contents of orthophosphate and primary P compounds associated with Ca, Mg, Fe, and Mn. Although both HPO 4 2 and H 2 PO 4 exi st in the soils, HPO 4 2 is dominant in the soil solution of the Pine Rockland calcar eous soils which have soil pH greater than 7.2 (Pierzynski et al., 2005). Obviously, the deposit of ash production contributed to increasing in the concentrations of orthophosphate and other P compounds, resulting in an increase in the extractable P conten t in 14 days after the fire (Hernandez et al., 1997). In contrast, a rapid development of understory plant communities which demanded a large amount of ortho P in the soil solution caused a dramatic reduction of HPO 4 2 and H 2 PO 4 contents 90 days after the fire Soil Solution Chemistry after the Fire Ionic activities of nutrient elements in soil solution following the fire were obtained from elements are shown in Table 4 indicates a negative logari thm of ionic activities of nutrient elements. T he fire significantly affected the activities of nutrient elements, excluding

PAGE 119

119 pZn 2+ (p = 0.1043). Fire significantly increased the ionic activities of Mg 2+ K + Mn 2+ and NH 4 + in 14 days after the fire and NO 3 activity in 180 days, but significantly decreased the activities of Cu 2+ HPO 4 2 and H 2 PO 4 in 14 days after the fire. The activities of Ca 2+ and Fe 2+ fluctuated with time as a consequence of seasonal variations. Ionic activities of nutrient elements in soil solution were highly correlated with their extracted concentrations. Significant changes in the activities of these elements almost occurred concurrently as those of their extracted contents. The results showed that the activity of P was very low in the soil solution (around 10 for pHPO 4 2 and 10.5 for pH 2 PO 4 respectively). In contrast, activities of Ca 2+ (pCa 2+ ~ 1.75), Mg 2+ (pMg 2+ ~ 2.95), Fe 2+ (pFe 2+ ~ 2.7), and Mn 2+ (pMn 2+ ~ 3.24) were very high. This result indicated that the activity of P in soil solution of the Pine Rockland calcareous soils was strongly controlled not only by solubility of Ca/Mg phosphate minerals but also solubility of iron and manganese phosphate min erals. Table 4 5 shows a correlation matrix for ionic activities of the nutrient elements in the soil solution. Coefficients of correlation among the ionic activities were similar to results reported by Zhang et al. (2001), which had a negative correlation among pH with activities of Ca 2+ Mg 2+ Mn 2+ and K + and a positive correlation among pH with activities of Fe 2+ and phosphate ionic species. Soil solution pH was highly correlated with pCu 2+ pHPO 4 2 and pH 2 PO 4 (0.931, 0.620, and 0.871, respectively). The positive correlation between pH and pFe 2+ (0.279) and the negative correlations among pH with pCa 2+ ( 0.062), pMg 2+ ( 0.356), and pMn 2+ ( 0.445) indicated that increasing in soil pH reduces the activity of Fe 2+ and raises the activities of Ca 2+ Mg 2+ and Mn 2+ pHPO 4 2 which are the dominant phosphate ionic species in the Pine Rockland calcareous soils, positively correlated with pCa 2+ (0.342) and pH, but negatively correlated with pMg 2+ ( 0.185), pFe 2+ ( 0.427), and pMn 2+ ( 0.645). This result sugges ts that changes in phosphate are controlled by calcium phosphate minerals if the pH is raised (Zhang et

PAGE 120

120 al., 2001). Furthermore, a higher correlation between pMn 2+ and pHPO 4 2 than those among pCa 2+ pMg 2+ and pFe 2+ with pHPO 4 2 demonstrated that manganes e has the most magnitude on solubility of phosphate ionic species in the soil solution of the Pine Rockland forest. Phosphate Phase Equilibria after the Fire Initial solubility relationships between the activity of HPO 4 2 and the activities of Ca 2+ Fe 2+ Mg 2+ and Mn 2+ for the selected Ca/Mg/Fe/Mn P minerals are presented in Table 4 6. The activity of HPO 4 2 in the soil solution controlled by minerals of monocalcium P, tricalcium P (beta), octacalcium P, hydroxyapatite, vivianite, Mg 3 (PO 4 ) 2 and Mn 3 (PO 4 ) 2 was dependent on soil pH; whereas relationships among activity of HPO 4 2 with monetite, dicalcium dihydrate P, MgHPO 4 :3H 2 O, and MnHPO 4 were independent of soil pH. A high coefficient of correlation between pHPO 4 2 and pH revealed that P equilibrium in soils was a function of soil pH. Additionally, solution parameters including Ca, Mg, Fe, and Mn could play a significant role in controlling P solubility in soil (Karathanasis, 1991) because they are exemplified by th e strong regression relationships among pH + pHPO 4 2 with pH 1/2pFe 2+ (0.954), pH 1/2pMn 2+ ( 0.898), pH 1/2pCa 2+ ( 0.819), and pH 1/2pMg 2+ ( 0.758) (Table 4 7). Solubility diagrams are particularly useful for determining a relative stability of phosp hate compounds and minerals in soils at different pH values (Essington, 2004). Figures 4 2, 4 3, 4 4, 4 5, and 4 6 showed compositions in the soil solution with relation to the stability of phosphate minerals in calcareous soils of the Pine Rockland pre bu rn and 14, 30, 90, and 180 days after the fire. These stability diagrams expressed a relationship between pHPO 4 2 versus pH with presence of the selected phosphate minerals. Monocalcium P mineral disappeared from diagrams because it does n o t exist in the s oil solution with the high soil pH. These diagrams were developed using the activities of Ca 2+ Mg 2+ Fe 2+ Mn 2+ HPO 4 2 and pH as reported in Table 4

PAGE 121

121 4. Red diamond points on the diagrams displayed solubility of P relating to the phosphate minerals. The point plotted above a mineral solubility isotherm indicated that solubility of P was supersaturated with respect to the given mineral, and solution conditions could form stable or meta stable states of the given mineral. The point plotted below a solubilit y isotherm of a certain mineral described that P solubility was under saturated relative to the given mineral. The point plotted on or very near a solubility line of a certain mineral showed that P solubility was in equilibrium with the given mineral. Ge nerally, solubility of P was under saturated with Ca/Mg P minerals including monetite, DCPD, TCP, OCP, HA, Mg 3 (PO 4 ) 2 MgHPO 4 :3H 2 O, and Mn 3 (PO 4 ) 2 The fire shifted solubility of P in equilibrium between vivianite and MnHPO 4 Pre burn solubility of P was in equilibrium with vivianite and under saturated with MnHPO 4 (Figure 4 2), but the soil solution was in equilibrium with MnHPO 4 and under saturated with vivianite in post fire 14 days (Figure 4 3). Particularly, there was an shift between two minerals of Mn 3 (PO 4 ) 2 and Mg 2 (PO 4 ) 2 14 days after the fire. The solution was in equilibrium with both vivianite and MnHPO 4 30 days after the fire (Figure 4 4). Although solubility of P returned to its pre burn state 90 days after the fire (Figure 4 5), the diagram of P s olubility in post fire 180 days was exactly the same as that in the pre burn condition (Figure 4 6). The role of Mn and Fe on P solubility seemed to be important in Pine Rockland calcareous soil where there is a very low P concentration and high soil pH. Increases in the activities of HPO 4 2+ after the fire promoting the formation of precipitates between P with Mn 2+ and Fe 2+ resulted in shifting the equilibrium between vivianite and MnHPO 4 This was similar to the result of Zhang et al. (2001), where MnHPO 4 in the soil solution was closely in equilibrium without the application of fertilizers, and saturated after applying fertilizers. Precipitations between P with Ca and Mg to form Ca P and Mg P minerals

PAGE 122

122 were not likely to occur after the fire because the so lubility points of P always fell below the solubility lines of Ca and Mg P minerals. The pattern of the stability diagrams was similar to results of previous studies (Fixen et al., 1982; Karathanasis, 1991; Zhang et al., 2001). The order of magnitude con trolling solubility of P and increasing in the activity of P in the soil solution was vivianite > MnHPO 4 > HA > TCP > OCP > Monetite > DCPD > MgHPO 4 :3H 2 O > Mg 3 (PO 4 ) 2 > Mn 3 (PO 4 ) 2 The magnitude was demonstrated by saturation indices (SI) of these minerals as shown in Table 4 8. These SI values P mineral (Ca 4 H(PO 4 ) 3 :2.5H 2 O ), but instead has Ca 4 H(PO 4 ) 3 :3H 2 O; therefore, the saturation index of the octacalcium P mineral was no t presented in Table 4 8. The positive, negative, and zero values of SI indicated oversaturation, undersaturation, and equilibrium, respectively. A given mineral with a lower negative SI value has a higher priority to react with orthophosphate species in o rder to form precipitation. Equilibrium will shift to other mineral having a higher negative SI value after the given mineral reaches in equilibrium and becomes oversaturation. Modeling a Soil Water Characteristic Curve for the Pine Rockland forest Soil water characteristic curve s represent a relationship between volumetric water content and matric suction. Parameters of the Retc model utilized to express soil water characteristics of the Pine Rockland soils are presented in Table 4 9. The fitted val ues of the saturated water content ( s ), the residual water content ( r 1.89, and 0.87, respectively. Figure 4 7 shows the shape and equation of the fitted soil wate r retention curve with R 2 value of 0.9214. The relatio nship between the volumetric water content and the suction for the Pine Rockland soils was described by the following equation: y = 0.064ln(x) + 0.6454, where y is the volumetric water content (v/v); and x is suction (kPa).

PAGE 123

123 The fitted parameters and shap e of the curve were very similar to the results of Al Yahyai et al. (2006 ) and Muoz Carpena et al. (2002), which described soil water characteristics of Krome calcareous soil in the agricultural areas in the South Florida. However, there was a shifting of the volumetric water content. Al Yahyai et al. (200 6 ) and Muoz Carpena et al. (2002) showed that the volumetric water content (v/v) ranged from 0.15 to 0.50, whereas the volumetric water content in this study was from 0.22 to 0.66. This shift could be ex plained by the fact that the Pine Rockland forest soils contained more organic matters than the Krome very gravelly loam soil. Therefore, the Pine Rockland soils exhibited a higher water holding capacity at low suction and a lower water holding capacity at high suction than those of the calcareous rock plowed soil (Krome soil). Effects of organic matter and bulk density on performance of the soil water retention curve were documented by Cornelis et al. (2001). Stuffs of the Pine Rockland forest floor were a ccumulated for five years; hence, the Pine Rockland soils contained a high content of organic matter, approximately 46.6% which was not reported in this study. Bulk density of the repacked soil used for determination of the retention curve was 1.1g cm 3 w hich was in a range of from 1.0 to 1.4 g cm 3 for the rocky outcrop soils reported by USDA Soil Survey of Dade County (1996). Evaluation of a goodness of fit measure was useful for express ing good ness of fit between observed and predicted values (Al Yahya i et al., 2006; Cornelis et al., 2001; Regalado et al., 2005). It can be observed from Figure 4 8 that there was a strong relationship between the observed volumetric water content and the fitted volumetric water content with coefficient of determination ( R 2 ) of 99.4%. Moreover, coefficient of efficiency (E) a nd index of agreement (d) were very high : 0.994 for E and 0.998 for d (Table 4 9). Thus, the Retention Curve model with

PAGE 124

124 the modified Mualem van Genuchten formulation as described by Schaap and van Genu chten (2005) pe rformed very well with the gravelly loam soils of the Pine Rockland forests. Modeling for Prediction of P Availability after the Fire Table 4 10 showed a meta analysis and results of the stepwise, multiple regression analysis of the fitted shows components included in the models. Among twenty four fitted models, the model 17 which included intercept, the 1 st order of soil moisture and burning temperature, the 2 nd order o f time (days after a fire) and burning temperature, and the interactions between soil moisture and time and between soil moisture and burning temperature was selected for a predictive model of P availability after a fire. As shown in Table 4 10, the chosen model performed the best based on the highest coefficient of determination ( 0.8523 ) the lowest coefficient of variation (15.76%), and the smallest RMSE value (1.009). The relationship between extractable P content a fter a fire and predictors was describe d by the following equation: Extractable P content = 3.38 27.19* M + 0.06579* T 0.000143* D 2 0.000115* T 2 + 0.05008* M*D + 0.03983* M*T where M is the soil moisture content (v/v); T is the fire temperature ( 0 C); and D is time or day s after the fire. A normal probability plot having a straight line demonstrated that the developed model had a normality of errors. Although the vertical spread of residuals did not equally distribute in both sides of the zero line, there was no evidence for an inequality o f variances for the chosen model. The expected value of Durbin Watson (D) is approximately 2.0 for independence of errors, but an acceptable range for the D value is from 1.5 to 2.5. A critical value of D of the selected model was 1.41, which was very clos e to the acceptable range. Therefore, the developed model basically met a normality of errors, equality of variances, and independence of errors of the regression analysis.

PAGE 125

125 Based on a relationship of contents of TN and TC in ashes between the laboratory in burning experiment, an average fire temperature of 0 C (Table 4 11). Average volumetric water contents, observed P contents, and simulated P concentrations at six different times of 12. By comparing simulation data with observation data, it was found that contents of simulated P agreed well with concentrations of measured P with an R 2 value of 0.811 9 (Figure 4 10). Relatively high R 2 values implied that the developed model fit s very well with observational data, and could perform a good simulation for available P content following a fire. Comparison of agreement index (d) with E and R 2 revealed that a higher value of d was attributable to overcoming an insensitivity of E and R 2 to differences in the fitted and observed means and variances (Krause et al., 2005). Summary Phosphorus is the most limiting nutrient in the Pine Rockland ecosystem due to abun dance of Ca and high contents of Mg, Fe, and Mn. Availability of P in this ecosystem is dependent on persistence of metastable Ca P minerals and phosphate minerals of Mg, Fe, and ibrium speciation, HPO 4 2 H 2 PO 4 FeHPO 4 (aq), MgHPO 4 (aq), CaHPO 4 (aq), MnHPO 4 (aq), FeH 2 PO 4 + CaH 2 PO 4 + and CaPO 4 were majority compounds of P in the soil solution of the Pine Rockland calcareous soils. The prescribed fire had a significant impact on contents of these P compounds in soil solution, except for FeH 2 PO 4 + HPO 4 2 was predominant species of orthophosphate in the Pine Rockland soils. Its contents fluctuated with extractable P concentrations following a fire, and significantly increased in 14 days after the fire.

PAGE 126

126 The prescribed fire greatly impacted soil solution chemistry. It significantly increased ionic a ctivities of Mg 2+ K + Mn 2+ and NH 4 + in post fire 14 days and activity of NO 3 in post fire 180 days, but significantly decreased activities of Cu 2+ HPO 4 2 and H 2 PO 4 in 14 days after the fire. Activities of Ca 2+ and Fe 2+ which was not influenced by fire fluctuated with time as a result of seasonal variations. Soil solution pH was highly correlated with pCu 2+ pHPO 4 2 and pH 2 PO 4 The positive correlation between pH and pFe 2+ and the negative correlations among pH with pC a 2+ pMg 2+ and pMn 2+ indicated that increasing in soil pH reduces the activity of Fe 2+ and raises activities of Ca 2+ Mg 2+ and Mn 2+ Findings indicated that the activity of orthophosphate ions in the soil solution was mainly controlled by solubility of v ivianite and MnHPO 4 minerals. Solubility of P was generally under saturated with Ca/Mg P minerals. The prescribed fire shifted solubility of P in equilibrium between vivianite and MnHPO 4 The magnitude order controlling solubility of P as an increase of P activity in the soil solution included vivianite > MnHPO 4 > HA > TCP > OCP > Monetite > DCPD > MgHPO 4 :3H 2 O > Mg 3 (PO 4 ) 2 > Mn 3 (PO 4 ) 2 The soil water characteristic curve of the Pin e Rockland calcareous soils fits very well with the modified Mualem van Genuc hten model. The relationship between volumetric water content and suction in the Pine Rockland soils was expr essed by the equation : y = 0.064ln(x) + 0.6454, where y is a volumetric water content (v/v) and x is a suction (kPa). A long term effect of a fi re on availability of P was modeled by results of simulation from the laboratory incubation experiment with three predictors : soil moisture, fire intensity, and time after the fire. Results of multiple regression analysis showed that the following predicti ve model could perform the best prediction for P availability following a fire: Extractable P content = 3.38 27.19* M + 0.06579* T 0.000143* D 2 0.000115* T 2 + 0.05008* M*D + 0.03983* M*T

PAGE 127

127 Table 4 1 Average extracted concentration of soil nutrient elements, pH, and ionic strength after the fire used equilibrium speciation model Days after fire pH EC (S/cm) IS (mol/L) PO 4 P Ca 2+ Mg 2+ K + Na + Fe 2+ Mn 2+ Zn 2+ Cu 2+ NH 4 + NO 3 mg/kg soil 0 7.55 ab 333 c 0.0048 c 4.2 b 1060 ab 52 b 165 bc 291 bc 178 ab 69 b 3.5 a 3.4 a 40.6 bc 8.1 b 14 7.82 a 698 a 0.0101 a 13.4 a 905 b 144 a 277 a 359 a 153 b 256 a 5.1 a 4.1 a 99.6 a 11.3 ab 30 7.48 b 544 ab 0.0078 ab 4.4 b 1035 ab 80 b 193 abc 358 ab 207 ab 89 b 5.4 a 3.1 a 53.6 b 2.4 b 90 7.45 b 370 bc 0.0053 bc 1.2 b 972 ab 40 b 188 abc 310 abc 170 ab 47 b 3.3 a 3.8 a 23.1 bc 5.2 b 180 7.56 ab 383 bc 0.0055 bc 4.0 b 1094 ab 54 b 236 ab 277 c 158 b 53 b 5.3 a 3.8 a 31.2 bc 25.1 a 270 7.48 b 348 bc 0.0050 bc 3.2 b 989 ab 56 b 122 c 301 abc 289 a 89 b 4.2 a 3.6 a 21.6 bc 8.5 b 360 7.57 ab 328 c 0.0047 c 4.6 b 1162 a 53 b 145 bc 278 c 173 ab 73 b 5.3 a 4.2 a 22.1 bc 2.6 b 450 7.72 ab 453 bc 0.0065 bc 2.7 b 1049 ab 62 b 160 bc 313 abc 197 ab 114 b 4.0 a 4.0 a 23.0 bc 2.7 b 540 7.58 ab 332 c 0.004 8 c 3.0 b 972 ab 34 b 137 bc 307 abc 223 ab 59 b 2.4 a 3.8 a 11.4 c 10.5 b IS (I onic strength ) was calculated from EC (electrical conductivity) described by Marion Babcock equation: Log (IS) = 1.159 + Log (EC), where IS is in units of mol/m 3 and EC is in dS/m or mS/cm (Essington, 2004). Means followed by the same letter within the same column are not significant different

PAGE 128

128 Table 4 2 Solubility product constants and stoichiometry of Ca/Mg/Fe/Mn P minerals Name Stoichiometry K s Monetite CaHPO 4 + H + Ca 2+ + H 2 PO 4 10 0.30 Monocalcium P Ca(H 2 PO 4 ) 2 :H 2 O Ca 2+ + 2H 2 PO 4 + H 2 O 10 1.15 Dicalcium dihydrate P CaHPO 4 :2H 2 O + H + Ca 2+ + H 2 PO 4 + 2H 2 O 10 0.63 Tricalcium P (beta) Ca 3 (PO 4 ) 2 (beta) + 4H + 3Ca 2+ + 2H 2 PO 4 10 10.18 Octacalcium P Ca 4 H(PO 4 ) 3 :2.5H 2 O + 5H + 4Ca 2+ + 3H 2 PO 4 + 2.5H 2 O 10 11.76 Hydroxyapatite Ca 5 (PO 4 ) 3 (OH) + 7H + 5Ca 2+ + 3H 2 PO 4 + H 2 O 10 14.46 Vivianite Fe 3 (PO 4 ) 2 :8H 2 O 3Fe 2+ + 2PO 4 3 + 8H 2 O 10 Mg 3 (PO 4 ) 2 Mg 3 (PO 4 ) 2 3Mg 2+ + 2PO 4 3 10 MgHPO 4 :3H 2 O MgHPO 4 :3H 2 O Mg 2+ + PO 4 3 + H + + 3H 2 O 10 Mn 3 (PO 4 ) 2 Mn 3 (PO 4 ) 2 3Mn 2+ + 2PO 4 3 10 MnHPO 4 MnHPO 4 Mn 2+ + PO 4 3 + H + 10 H 2 PO 4 H 2 PO 4 HPO 4 2 + H + 10 HPO 4 2 HPO 4 2 H + + PO 4 3 10 Sou rce: Peirzynski et al. (2005); from Minteq s is solubility product constant of minerals Table 4 3. Results of ANOVA analysis and simulated concentrations of P compounds associated with Ca, Mg, Fe, and Mn after the fire Days after fire FeH 2 PO 4 + FeHPO 4 (aq) MgHPO 4 (aq) MnHPO 4 (aq) CaHPO 4 (aq) CaH 2 PO 4 + CaPO 4 Results of ANOVA analysis NS *** *** *** *** *** Simulated concentration (mg kg soil) 0 0.096 a 1.57 b 0.17 b 0.40 b 1.44 bc 0.034 a 0.15 b 14 0.132 a 3.46 a 1.16 a 3.68 a 3.20 a 0.043 a 0.65 a 30 0.100 a 1.42 b 0.27 b 0.62 b 1.40 bc 0.037 a 0.14 b 90 0.031 a 0.43 b 0.05 b 0.09 b 0.45 c 0.012 b 0.04 b 180 0.072 a 1.20 b 0.20 b 0.29 b 1.59 bc 0.038 a 0.17 b 270 0.132 a 1.51 b 0.10 b 0.30 b 0.81 bc 0.026 a 0.07 b 360 0.087 a 1.49 b 0.20 b 0.43 b 1.74 b 0.039 a 0.20 b 450 0.043 a 0.93 b 0.12 b 0.38 b 0.85 bc 0.014 a 0.17 b 540 0.076 a 1.32 b 0.07 b 0.23 b 0.94 bc 0.021 a 0.11 b ** *** 0.01, and 0.001

PAGE 129

129 Table 4 4. Days after fire pH pCa 2+ pMg 2+ pK + pFe 2+ pMn 2+ pZn 2+ pCu 2+ pNH 4 + pNO 3 pHPO 4 2 pH 2 PO 4 Result of ANOVA analysis ** *** ** *** NS ** *** *** *** ** 0 7.55 ab 1.75 ab 2.95 a 2.46 ab 2.70 ab 3.24 a 4.48 a 5.13 ab 2.75 bc 3.97 bc 9.99 c 10.32 b 14 7.82 a 1.83 a 2.43 b 2.21 c 2.79 a 2.55 b 4.37 a 5.47 a 2.36 d 3.91 bc 10.47 a 11.09 a 30 7.48 b 1.75 ab 2.65 ab 2.35 abc 2.62 ab 3.00 ab 4.28 a 5.11 b 2.60 cd 4.50 a 10.16 abc 10.44 b 90 7.45 b 1.75 ab 3.02 a 2.36 abc 2.68 ab 3.23 a 4.49 a 4.95 b 2.95 b 4.17 ab 10.01 bc 10.26 b 180 7.56 ab 1.70 b 2.86 a 2.27 bc 2.72 a 3.19 a 4.27 a 5.08 b 2.83 bc 3.54 c 10.03 bc 10.39 b 270 7.48 b 1.74 ab 2.79 ab 2.55 a 2.44 b 2.94 ab 4.48 a 5.02 b 2.97 b 3.98 bc 10.38 ab 10.66 ab 360 7.57 ab 1.67 b 2.81 ab 2.47 ab 2.67 ab 3.07 ab 4.28 a 5.06 b 2.98 ab 4.44 a 10.08 bc 10.46 b 450 7.72 ab 1.74 ab 2.78 ab 2.43 ab 2.62 ab 2.87 ab 4.43 a 5.28 ab 2.96 b 4.43 a 10.32 abc 10.84 ab 540 7.58 ab 1.75 ab 2.99 a 2.49 a 2.55 ab 3.12 ab 4.64 a 5.09 b 3.25 a 3.86 bc 10.26 abc 10.65 ab ** *** 0.01, and 0.001 log arithm of ionic activities of Ca 2+ Mg 2+ Mn 2+ Fe 2+ and HPO 4 2 Table 4 5 pH pCa 2+ pMg 2+ pK + pFe 2+ pMn 2+ pZn 2+ pCu 2+ pNH 4 + pNO 3 pHPO 4 2 pCa 2+ 0.062 pMg 2+ 0.356 0.101 pK + 0.180 0.132 0.495 pFe 2+ 0.279 0.227 0.449 0.523 pMn 2+ 0.445 0.455 0.623 0.268 0.028 pZn 2+ 0.102 0.160 0.005 0.116 0.035 0.008 pCu 2+ 0.931 0.067 0.448 0.226 0.321 0.491 0.109 pNH 4 + 0.220 0.505 0.445 0.474 0.230 0.498 0.214 0.395 pNO 3 0.078 0.086 0.208 0.128 0.016 0.104 0.009 0.033 0.024 pHPO 4 2 0.620 0.342 0.185 0.122 0.427 0.645 0.039 0.553 0.218 0.018 pH 2 PO 4 0.871 0.184 0.287 0.010 0.133 0.619 0.074 0.796 0.243 0.049 0.926

PAGE 130

130 Table 4 6 Stoichiometry between activity of HPO 4 2 and activities of Ca 2+ Fe 2+ Mg 2+ Mn 2+ for the selected Ca/Mg/Fe/Mn P minerals Name Stoichiometry Monetite pHPO 4 2 = 6.9 pCa 2+ Monocalcium P pHPO 4 2 = 7.775 0.5pCa 2+ pH Dicalcium dihydrate P pHPO 4 2 = 6.57 pCa 2+ Tricalcium P (beta) pHPO 4 2 = 2.11 1.5pCa 2+ + pH Octacalcium P pHPO 4 2 = 3.28 4/3pCa 2+ + 2/3pH Hydroxyapatite pHPO 4 2 = 2.38 5/3pCa 2+ + 4/3pH Vivianite pHPO 4 2 = 6.53 1.5pFe 2+ + pH Mg 3 (PO 4 ) 2 pHPO 4 2 = 0.71 1.5pMg 2+ + pH MgHPO 4 :3H 2 O pHPO 4 2 = 5.825 pMg 2+ Mn 3 (PO 4 ) 2 pHPO 4 2 = 0.4365 1.5pMn 2+ + pH MnHPO 4 pHPO 4 2 = 13.05 pMn 2+ log arithm of ionic activities of Ca 2+ Mg 2+ Mn 2+ Fe 2+ and HPO 4 2 Table 4 7 Correlation relationships among pH + pHPO 4 2 pH 1/2pFe 2+ pH 1/2pMn 2+ a nd pH 1/2pCa 2+ in soil solution (n = 45) Solution variables pH + pHPO 4 2+ pH 1/2pFe 2+ pH 1/2pMn 2+ pH 1/2pCa 2+ pH 1/2pFe 2+ 0.9544 pH 1/2pMn 2+ 0.8975 0.8394 pH 1/2pCa 2+ 0.8191 0.8783 0.8539 pH 1/2pMg 2+ 0.7576 0.7231 0.9098 0.8610

PAGE 131

131 Table 4 8. Saturation index (SI) for phosphate minerals in the Pine Rockland soils followi ng the fire obtained from simulation results of the Minteq speciation program Days after fire Vivianite MnHPO 4 HA TCP Monetite DCPD MgHPO 4 :3H 2 O Mg 3 (PO 4 ) 2 Mn 3 (PO 4 ) 2 0 0.000 0.102 1.407 4.037 4.899 5.180 7.091 14.948 15.114 14 0.574 0.000 2.097 4.637 5.409 5.690 7.092 14.021 13.878 30 0.000 0.006 1.849 4.925 5.206 6.271 6.921 14.593 14.981 90 0.000 0.147 1.641 4.112 4.815 5.096 7.082 15.250 15.576 180 0.000 0.146 1.100 3.841 4.814 5.094 7.007 14.756 15.304 270 0.000 0.243 2.603 4.777 5.183 5.464 7.316 15.513 15.426 360 0.000 0.082 1.125 3.876 4.859 5.140 7.080 14.876 15.011 450 0.000 0.122 1.477 4.189 5.133 5.414 7.247 14.866 14.622 540 0.000 0.356 2.021 4.454 5.119 5.400 7.462 15.820 15.632 SI = logarithm of the ratio of the ionic activity product (IAP) to equilibrium solubility product constant (K s ), SI = log(IAP) log(K s ); positive, negative, and zero SI values indicate oversaturation, undersaturation, and equilibrium, respectively. Saturation index of each variable at time point was an average value of replicates at the given time point. Table 4 9. Fitted parameters of Retc model used to describe soil water characteristics of the Pine Rockland calcareous soils, and coefficients used to evaluate a goodness of fit of model Retention curve model Measures of goodness of fit of model Parameter Value Coefficient Value s 0.41 R 2 0.994 r 0.07 E 0.994 0.08 d 0.998 n 1.89 m 0.47 s is the saturated water content; r is the the model fitting parameters R 2 is the coefficient of determination; E is the coefficient of efficiency; and d is the index of agreement

PAGE 132

132 Table 4 10 A summary of parameters of regression models fitted by response variable of extractable P and three predictors Model Day (D) Moisture (M) Temperature (T) D M D T M T D M T I Parameters from fitted models Components having insignificant 0.05 in the model 1 st 2 nd 3 rd 1 st 2 nd 3 rd 1 st 2 nd 3 rd R 2 CV RMSE Pr > F 1 x x x x 0.5715 26.65 1.7064 <.0001 2 x x x x x x x 0.6180 25.35 1.6227 <.0001 D*T 3 x x x x x x 0.6180 25.29 1.6189 <.0001 4 x x x x x x x x 0.6191 25.37 1.6243 <.0001 D*M, D*T, D*M*T 5 x x x x x 0.6088 25.53 1.6343 <.0001 6 x x x x 0.6562 23.88 1.5286 <.0001 7 x x x x x x x 0.7429 20.79 1.3311 <.0001 8 x x x x x x x x 0.7450 20.76 1.3290 <.0001 D*M*T 9 x x x x 0.7068 22.05 1.4116 <.0001 10 x x x x x x x 0.7598 20.10 1.2868 <.0001 D*M 11 x x x x x x 0.7595 20.06 1.2845 <.0001 12 x x x x x x x x 0.7619 20.06 1.2843 <.0001 D*M, D*T, D*M*T 13 x x x x x 0.7503 20.39 1.3056 <.0001 14 x x x x x x x 0.8090 17.93 1.1476 <.0001 M, M*M 15 x x x x x 0.7124 21.89 1.4013 <.0001 16 x x x x x x x x x x 0.8554 15.71 1.0055 <.0001 I, D, M*M, D*T 17 x x x x x x x 0.8523 15.76 1.0090 <.0001 18 x x x x x x x x x x x 0.8565 15.69 1.0042 <.0001 D, M*M, D*T, D*M*T 19 x x x x x x x x x x Model could not be fitted by 3 rd order of moisture 20 x x x x x x x x x 0.8566 15.60 0.9990 <.0001 D, M, D*D, M*M, D*D*D 21 x x x x 0.6468 24.20 1.5492 <.0001 22 x x x x x x x x x x x x 0.9031 12.92 0.8272 <.0001 D, D*D, M*M, D*D*D, D*T 23 x x x x x x x 0.8279 17.01 1.0890 <.0001 24 x x x x x x x x x x x x x 0.9042 12.88 0.8247 <.0001 D, D*D, M*M, D*D*D, D*T, D*M*T RMSE: Root mean square error, R 2 : R Square; CV: Coefficient of Variation; 1 st 2 nd 3 rd : orders in the linear regression; I: Intercept

PAGE 133

133 Table 4 11. Rockland forest Variables The fitted models from the laboratory residual ashes Field collected ashes Content ( % ) Simulated T ( 0 C) TC Y = 75.27233 0.0897* (T) 1.02 521.3 TN Y = 4.16540 + 0.03148 *(T) 0.00004139 *(T) 2 28.53 518.9 Average simulated T: 520 0 C concentrations of TN (%) and TC (%) in ash production simulated from the laboratory incubation with independent variable of heating temperature (T) Table 4 12 Observation and simulation data of extractable P content after the fire at the Pine Rockland forest Days after fire Average volumetric water content at 5cm topsoil Simulated P content Observed P content v/v mg/kg soil 7 0.00000 6.52 9.16 14 0.02555 6.35 6.98 21 0.07011 6.08 6.55 30 0.14230 5.69 6.77 60 0.08632 5.71 4.85 90 0.11078 5.15 3.51

PAGE 134

134 Figure 4 1 Concentrations of ortho P species after the fire, relatively Olsen P contents Figure 4 2. Relationship of pH and phosphate mineral stability in soil solution before the fire 30

PAGE 135

135 Figure 4 3 Relationship of pH and phosphate mineral stability in 14 days after the fire Figure 4 4 Relationship of pH and phosphate mineral stability in 30 days after the fire

PAGE 136

136 Figure 4 5 Relationship of pH and phosphate mineral stability in 90 days after the fire Figure 4 6 Relationship of pH and phosphate mineral stability in 180 days after the fire

PAGE 137

137 R = 0.994, p < 0.001 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Fitted volumetric water content Observed volumetric water content Figure 4 7. Soil moisture curve of the Pine Rockland fitted by the Retc model Figure 4 8. A goodness of fit model for soil moisture curve of the Pine Rockland forest y = 0.064ln(x) + 0.6454 R = 0.9214 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0 100 200 300 400 500 600 700 800 Volumetric water content Suction (kPa)

PAGE 138

138 R = 0.8523, p < 0.001 0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 Observed P concentration (mg/kg soil) Fitted P concentration (mg/kg soil) Figure 4 9 A goodness of fit for the predictive model of P availability after the fire Figure 4 10 Verification of the predictive model for P prediction after the fire R = 0.8119, p = 0.0142 7 14 21

PAGE 139

139 CHAPTER 5 CONCLUSIONS The Pine Rockland forest is a unique upland ecosystem originated from limestone substrates, and pr eserved one of the most endangered forest types in the world South Florida slash pine (Pinus elliottii var. densa ), with a diverse native plants in the understory vegetation (Snyder et al., 1990). It is found in three rem nant regions in South Florida including the Long Pine Key in the Everglades National Park, the Big Pine Key in the Florida Lower Keys, and a part of the Big Cypress National Preserve (Snyder, 1986). The increasing influence of human activities on the lands cape has changed many critical ecosystem processes, particularly on the hydrological balance, fire regime, and rapid invasion of exotic species. The Pine Rockland forest is a poor nutrient and fire dependent ecosystem ( U.S Fish and Wildlife Service, 200 7; USGS, 2000 ) which is often maintained by fires in every 3 to 10 years for a continued health of endemic plants (Snyder et al., 1990). Fire is an environmentally important factor that requires for controlling, at least a part, relative dominance of hardw ood plants in understory of the Pine Rockland forest. The importance of fire as a part of this ecosystem has long been recognized (Snyder et al., 2005; Wade et al., 1980). The e cological roles of fire on vegetation recovery in the Pine Rockland forest have been recently evaluated (Knapp et al., 2009; Robbin and Myers, 1992; Ross et al., 1994 & 2003; Sah et al., 2004 & 2006; Sah et al., 2010; Snyder, 1986; Sparks et al., 2002); however, effects of fire on soil nutrient pools have not been documented. The ove rall goal of this study was to evaluate effects of prescribed fire on soil nutrient pools in the Pine Rockland forest ecosystem. We hypothesized that phosphorus (P) is a limiting factor, and prescribed fire would improve deficiency of P and create more ava ilability of soil nutrient elements in the Pine Rockland forest. The specific objectives and summary of results are outlined below.

PAGE 140

140 Objective 1 : It was to d etermine whether P is limited in the Pine Rockland ecosystem due to abundance of calcium in calcareous soils, and evaluate how prescribe fire could improve availability of soil nutrient pools. The specific tasks of this objective were to 1) determine P limitation basing on pine foliar P and other nutrients, and foliar DRIS indices; 2) evaluate effects of fire on pools of C and N in fuel load and soil; 3) assess impacts of fires on soil pH, EC, and soil nutrient pools; and 4) compare C:N:P ratios among pine foliage, fuel load and soil. To achieve these specific tasks, a field burning experiment was conducted in the Long Pine Key in the Everglades National Park on November 19 th 2008. In addition, samples of pine foliage were collected at the beginning of the growing season of pine tree, and fuel load samples were collected in one day before the burning experiment. Determination of P restriction was based on c omparisons between the foliar P of the Pine Rockland slash pine and published critical values of pine foliage P in slash pine forests as well as on pine foliar DRIS indices. The foliar P concentration in the Pine Rockland was 0.045 % of dry biomass while the published critical P concentration of slash pine forest was from 0.07% to 0.125% of dry biomass. The foliar N concent ration in the Pine Rockland was approximately 0.81% of dry biomass which was closely to the critical N concentration of the slash pine forest, ranging from 0.87% to 1.3% of dry biomass. DRIS indices of P, K, N, Mg, and Ca in the Pine Rockland slash pine we re 46.13, 10.77, 0.89, 8.71, and 49.05, respectively. These results demonstrated that P is the most limiting factor and N is marginal to limitation or may be saturated in the Pine Rockland forest ecosystem. The P limitation led to rise the pine foliar N:P ratio of 17.9 that was higher than the critical foliar N:P ratio of the Slash pine forest. The C:N:P ratios of pine foliage and fuel load were 1162.5:17.9:1 and 2361.2:23.8:1 respectively, whereas those of soil were 920:30.3:1. The lower

PAGE 141

141 C:N ra tio and higher N:P ratio in soil than those in pine foliage could provide more evidence that N is sufficient in the Pine Rockland forest ecosystem. The fire did not impact on soil C:N ratio, but it significantly altered soil C:P and N:P ratios. The soil C: P and N:P ratios fluctuated with changes of soil C and N concentration s after the fire. Alterations of the soil C:P and N:P ratios depended on more the growing season of pine trees and seasonal variation than the fire. The fire consumed approximately 90.2 % of the forest floor and understory biomass, caused losses of 80% of N pool and 92% of C pool in the forest floor and understory biomass, and volatilized 86% of NH 4 + pool and 48% of NO 3 pool that associated with the N pool in the forest floor and underst ory biomass In contrast, the prescribed fire significantly affected on soil C and N pools at the top 0 5cm soil from 0.95 to 1.42% for TN and from 24 to 40% for TC in 14 days after the fire, respectively. Although the concentration s of TC and TN reached their original values in 30 days after the fire, these concentration s continued to decrease thereafter and had the lowest values after 90 days. Ammonium and nitrate were two major components contributing to increase in soil N pool af ter the fire. The extractable NH 4 + in the top 0 5cm soil increased immediately after the fire, remained until 30 days, and then declined in thereafter. However, the extractable NO 3 was not changed immediately after the fire, but increased significantly in the post fire 180 days as a result of the nitrification process. The fire induced increases in soil EC, pH, and the extractable concentration s of P, Mg, K, and Mn occurred concurrently in 14 days after the fire, whereas the increase in the extractable Ca and Fe occurred in 270 days and 360 days respectively as a result of the variation on soil moisture content between the dry season and the rainy season. Objective 2 : It was to assess the effect of the fire to soil nutrient pools following by different b urn seasons on calcareous soils in the Pine Rockland forest ecosystem. The specific

PAGE 142

142 tasks of this study included 1) determining effects of different fire temperatures on pH, EC, and nutrients in residual ashes; 2) determining a relationship between burn te mperature and soil nutrient pools in the calcareous soils; 3) determining if the relationship is affected by soil moisture content; 4) assessing if the relationship is impacted by time after burn; and 5) estimating a fire temperature for the field burning experiment basing on concentrations of TP, TMg, TCa, TFe in residual ashes. The burning experiment was conducted under the controlled condition of laboratory incubation to determine how burn temperature and soil moisture content influenced on soil nutrient pools after the fire. Four burning temperatures of fuel loads which included 250, 350, 450, and 550 0 C used to determine effects of burn temperature on nutrient concentrations in ashes and soils. The soil moisture content consisted of 35, 100, and 70% of soil field capacity, representing a range of soil moisture that can be found in the Pine Rockland throughout the year. Soil samples were incubated at 25 0 C in six different time intervals of 14, 30, 60, 90, 120, and 180 days. I ncrease in burn t emperatures significantly impacted on TC, TN, and TK in ashes in term of dry fuel load but did not change concentrations of TP, TCa, TMg, TFe, TMn, and TZn d ue to their relatively high volatilization temperatures. Losses of TN and TC via volatilization occurred mostl y at 450 and 550 0 C, and were proportional to biomass loss. Concentration of TK significantly decreased at the burn temperatures of 350 and 450 0 C, but no further decrease was found for concentration of TK at heating to 550 0 C. The volatilizations of TN and T K in the residual ash along the burn temperature gradient resulted in significant changes of the water soluble concentrations of NH 4 N, NO 3 N, and K in the residual ash. The water soluble c oncentration of NH 4 N significantly reduce d as the burn temperature increased In contrast, the concentrations of NO 3 N in residual ashes significantly increased at 250 0 C, but n o significance was

PAGE 143

143 observed at or greater than 350 0 C. T he water soluble K was significantly lower in 250 0 C than those in 350, 450, and 550 0 C. The temperature induced changes of water dissolved pH and EC, and the water soluble nutrients in residual ashes were significantly along the burn temperature gradient. The ash pH significantly increased from acidic with burn temperatures at or below 250 0 C to a lkaline at the temperatures at or over 350 0 C. Although no significant difference was found for ash EC with the temperature at or below 250 0 C, the EC was linearly increased as the burn temperature increased above 250 0 C. However, the water soluble P decreased when heating to 350 0 C or higher. Responses of water soluble concentrations of other nutrient elements along the temperature gradient in term of dry fuel load were very variously. The water soluble Ca and Mg reached their highest values at 350 0 C, whereas the water soluble Mn and Zn were gradually reduced as the burn temperature increased. The different availability of nutrients in the ash resulted in significant changes of soil pH EC and extractable NH 4 N, NO 3 N, P, Mg, K, Fe, Mn, Zn, Cu. General ly, the burn temperature at or over 450 0 C led to an increase of soil pH, and produced more available Mg, K, Fe, and Cu On the other hand an increase in the burn temperatures caused a decline of extractable NH 4 N, Mn, and Zn. The extractable P, NO 3 N, and EC with respect to the burn temperatures reached a peak after heating to 350 0 C The soil moisture content significantly impacted on soil pH, EC and extractable NH 4 N, NO 3 N, P, and Fe but was not for extractable Mg, K, Mn, Zn, and Cu. U nder the c ontrolled condition in the laboratory, extractable P decreased with an increase in the soil moisture. I ncreasing in the soil moisture increased in the extractable NH 4 + Responses of NO 3 N and EC to the soil moisture had the same tendency that reached a peak with the soil water content of 70%

PAGE 144

144 field capacity. However, soil pH and extractable Fe varied in the opposite pattern to those of EC and NO 3 N from which their lowest values were on t he soil water content of 70% field capacity. The time after the fire significantly influenced on soil EC and extractable P, NH 4 N, NO 3 N, Mg, K, Mn, Cu, and Zn, but not for extractable Fe and pH. Ammonia was linearly declined with an increase of the time, and stabilized after 120 days. Conversely, the reduction of NH 4 + concentration resulted in the increase of NO 3 N concentration through the nitrification process. Significant changes of soil EC and extractable P, Mg, K, Mn, Zn, and Cu actually occurred at 120 days or thereafter. Non significant difference of TP/TCa, TP/Mg, and TP/TFe in residual ashes indicated that TP, TCa, TMg, and TFe were good predictors for precisely estimating a field fire temperature. Based on the regression relationship between the simulation temperatures and concentrations of TP, TCa, TMg, and TFe in residual ashes, the fire temperature for the field burning experiment was estimated to be approximately 370 0 C. Objective 3 : The goal was to predict long term impacts of fire on P availability in calcareous soils under the Pine Rockland forest. Four specific tasks consisted of 1) determining P relationship of soil pH and phosphate minerals to phosp hate solubility after a fire by establishing stability diagrams of phosphorus; 3) modeling soil water characteristic curve for Pine predictive model for prediction of P availability after a fire. The P species and the solubility diagrams of P were determined basing on the results from the field burning experiment while the predictive model was fitted from the simulation results in the laboratory experiment with predi ctors of soil moisture content, burn temperature, and incubation time.

PAGE 145

145 4 2 H 2 PO 4 FeHPO 4 (aq), MgHPO 4 (aq), CaHPO 4 (aq), MnHPO 4 (aq), FeH 2 PO 4 + CaH 2 PO 4 + and CaPO 4 are the dominant P compounds in the Pine Rockland calcareous soils. Concentrations of these species were significantly different following the fire, except FeH 2 PO 4 + HPO 4 2 was a predominant species of orthophosphate, and fluctuated with Olsen P concentr ations. In general, these species significantly increased in 14 days and returned to their pre burn values in 30 days after the fire. Nevertheless, concentration s of HPO 4 2 CaHPO 4 (aq), and CaH 2 PO 4 + continued to decrease until 90 days after the fire. Ionic activities of nutrients in soil solution highly correlated with their extracted concentrations. The fire significantly affected the activities of nutrient elements, excluding pZn 2+ Fire significantly increased the ionic activities of Mg 2+ K + Mn 2+ and NH 4 + in post fire 14 days and activity of NO 3 in post fire 180 days, but significantly declined the activities of Cu 2+ HPO 4 2 and H 2 PO 4 in 14 days after the fire. The activities of Ca 2+ and Fe 2+ fluctuated with time as the consequence of seasonal variations. Ionic activities of nutrient elements in soil solution highly correlated with their extracted concentrations. The solution pH had a negative correlation with activities of Ca 2+ Mg 2+ Mn 2+ and K + and a positive correlation with activities of Fe 2+ and orthophosphate species. The activity of HPO 4 2 in the soil solution controlled by minerals of MC P TCP, OC P HA, v ivianite Mg 3 (PO 4 ) 2 and Mn 3 (PO 4 ) 2 was dependent on soil pH; whereas relationships among activity of HPO 4 2 with m onetite DCD P MgHPO 4 :3H 2 O and MnHPO 4 were independent on soil pH. T he activity of orthophosphate ions in the soil solution was mainly controlled by solubility of vivianite and MnHPO 4 minerals. The solubility of P in the soil solution

PAGE 146

146 was expressed by the magnitude order of vivianite > MnHPO 4 > HA > TCP > OCP > Monetite > DCPD > MgHPO 4 :3H 2 O > Mg 3 (PO 4 ) 2 > Mn 3 (PO 4 ) 2 The soil water characteristic curve of the Pine Rockland calcareous soils was expressed by the equation: y = 0.064ln(x) + 0.6454, where y is a volumetric water content (v/v) and x is suction (kPa). A long term effect of fire on availability of P could predict by the following model: Extractable P = 3.38 27.19* M + 0.06579* T 0.000143* D 2 0.000115* T 2 + 0.05008 *M*D + 0.0 3983 *M*T 0 or day after a fire (days). Research Synthesis : The bioavailability of P in the Pine Rockland forest ecosystem is very low due to the nature of the calcareous soils. Phosphorus is the most limiting factor in this ecosystem with the average Olsen P concentration of 4.2 mg/kg soil. The prescribed fire cou ld increase threefold in available P (13.4 mg/kg soil) in the top 0 5cm soil in 14 days after the fire Because of its relatively low concentration in the soil solution, the solubility of P in the soil solution is in equilibrium with the vivianite mineral and undersaturated with all Ca/Mg P minerals. The prescribe fire shifted the equilibrium from the vivianite into MnHPO 4 in 14 days after the fire and returned its original equilibrium sta tus in 18 0 days. A dramatic increase in ratio of TP to inorganic P in residual ash as fire intensity raised implied that low fire intensity generally can produce more available P than high intense fire because inorganic P ions tend to b i nd to basic cations (Ca, Mg, Fe, Mn) in residual ash es and then form insoluble phosphates when the fire intensity is increased The reduction in Olsen P as an increase in the soil moisture indicated that higher soil moisture favored for reactions among P with Ca and Mg or Fe and Mn to form Ca/Mg P precipitates or Fe/Mn P hydroxyl com pounds The h igher N:P ratio and lower C:N ratio in soil to those in pine foliage could indicate that

PAGE 147

147 N may not be limited in the Pine Rockland forest where have had a high rate of nitrogen deposition associated with the high lightning intensity. Although the fire virtually removed a large amount of N pool in the forest floor and understory vegetation it increased soil N H 4 N and NO 3 N pools. However, the concentration of NH 4 N in soil trended to reduce with an increase in the fire intensity and was higher at the high soil moisture content, whereas the nitrification process promoted the increase in the soil NO 3 N pool which could occur in several months after a fire. The prescribe fire could rise availability of Mn, Zn, and Cu. The high fire intensity create d less the available Mn and Zn and more availability of Cu than the low fire intensity. The soil fire could cause a reduction of soil extractable Mn, and Zn, and increment of soil extractable Cu. The relatively high correlation between the pH and the activity of Cu in soil solution demonstrated that availability of Cu depends considerably on the high soil pH of the Pine Rockland forest. Significance of This St udy and Suggestions of Further Study : This study provided basic information for predicting a dynamics of soil nutrient pools after a fire in calcareous soils under the Pine Rockland forest ecosystems in South Florida. The findings can also be used for similar ecosystems elsewhere. The following are the major contributions from my study: (1) t he highest negative value of DRIS index of P and much lower foliar P compared to critical P concentrations reported for slash pine has indicated that phosphorus is the most limiting nutrient in the Pine Rockland forest, which has not been documented. Modeling tools provide an opportunity to predict a long term impact of a prescribed fire on P availability in calcareous so il of the Pine Rockland forests; (2) a lthough many studies have evaluated the role of prescribed fire on restoration of the understory vegetation of the Pine Rockland forest ecosystem, this is the first time that such a detailed study involved in exploring how fire intensity affects on availability o f P and other

PAGE 148

148 nutrients in post fire ash and soil. Findings on an inorganic P reduction in residual ash as an increase in fire intensity as well as a decrease of soil extractable P after a fire due to an increment of soil moisture content would be benefici al to ecological fire managers, who could choose a suitable burn season to produce a maximum bio avail able P amount for plant uptake; and (3) i t is being thought that phosphate solubility is mainly controlled by Ca/Mg P minerals in calcareous soils; howeve r, this study found that the solubility of P in the Pine Rockland soil is controlled by Fe/Mn P minerals, particularly vivianite mineral, due to relatively low P concentration and a high content of iron. This is also the first time that solubility of P was evaluated for the Pine Rockland calcareous soil. This finding could help soil scientists to have a broader view on predicting solubility of phosphate minerals in calcareous soils. The Pine Rockland forest is the nutrient p oor ecosystem. Low availability o f micro nutrients, particularly Zn and Cu can cause a growth limitation for the vegetation communities and the slash pine in the Pine Rockland forest as well. Although the field burning study showed that the prescribe d fire did not influence availability of Zn and Cu the laboratory study demonstrated that fire intensity significantly impacted extractable Zn and Cu. Additionally fire intensity and soil moisture content also significantly affected extractable Fe and Mn. This study did not include evaluation of response of plant community to prescribed fire. For sure, changes of P and other nutrients after fire directly affect survival and growth of plants. Therefore, a further study which has a linkage between fire induced change of soil micro nutrients and post fire responses of understory vegetation to these nutrients should be carried out.

PAGE 149

149 APPENDIX A STUDY SITE AND SAMPLING PROCEDURES Fi gure A 1. Pine Rockland f orest in the Long Pine Key, Everglades National Park Figure A 2. A 50 x 50cm sampling frame used to collect ashes after the fire

PAGE 150

150 APPENDIX B PROCEDURES OF CHEMICAL ANALYSIS AND QA/QC METHODS IN LABORATORY pH Procedures: Soil and ash pH were measured using the AR 60 Dual Channel pH/Conductivity meter. Soil pH was extracte d with rate of 10g soil and 25mL DDI water. 10 g o f soils were weighted into 50 mL plastic bottles and 25 mL DDI water added followed by shak ing at 180 o scillations p er m inute for 15 minutes by a reciprocating shaker. The shaking solutions were filtered through a Whatman paper No. 42 Similarly, ash pH was determined by extracting 1g ash and 50 mL DDI water (Qian et al., 2009a & b) Calibration of pH meter: pH meter was calibrated daily using three pH buffer standard solutions o f 4.0, 7.0, and 10.0 at ambient temperature of 25 0 C before measurement. Calibration curve method was the first order with a calibrated slope of approximately 99.6% (an acceptable range of slope from 98.0% to 100%) Quality a ssurance (QA)/quality control (QC) method: Three methods for controlling quality included repeated measurement of soil samples, NSI standard QC solution, and pH buffer standard solution. After every 20 samples, a duplicate of the 20 th sample, standard QC solution, and pH buffer solutio n were measured to check the meter. The QC standard solution received from the NSI company had value of 7.20 (an acceptable range from 7.00 to 7.40). The pH buffer solution applied for the QA/QC was 7.0. Electrical C onductivity Procedures: E lectrical condu ctivity was measured on a 1:2 dilution (10g soil: 20 mL DDI water) for soil and 1:50 dilution (1g ash: 50 mL DDI water) for ash on the AR 60 Dual Channel pH/Conductivity meter. 10 g of so il samples were weighed in 50 mL plastic bottles, and then added by 2 0 mL DDI water. Solutions were shaken in 10 minutes, allowed suspension to stand

PAGE 151

151 for 2 hours, thereafter filtered through a Whatman paper No. 42. A similar method was used for measurement of EC on ash samples (Qian et al., 2009b). Calibration of conductivity meter: A reference standard solution of 1,412 S/cm from the NSI was used to calibrate conductivity meter before measurement daily. Standard settings of the instrument were cell constant of 1.0/cm, default temperature of 25 0 C, and temperature coefficient of 0.00%/ 0 C. If the meter is set up with cell constant of 1.001/cm or 0.9999/cm and temperature coefficient of 0.00%/ 0 C, calibration process meets the standard criter ia for the instrument. QA/QC method: Similarly to pH measurement, EC was cont rolled by three measurements: duplicates of every 20 th sample; standard QC reference solution of EC, and calibrated standard solution. After every 20 samples were measured, repetition of the 20 th sample, standard QC reference solution of 349 S/cm (an acce ptable range from 310 to 387 S/cm), and calibrated standard solution of 1,412 S/cm were performed to check the conductivity meter. Ammonium Procedures: Ammonium in soils was extracted by 2M KCl extracting solution, and measured by the AQ2 discrete auto analyzer. An extracting solution of 2M KCl was created by adding 149.1g KCl into a 1000 mL volumetric flash with 900 mL DDI water, stirring the solution until KCl completely dissolved, and then diluted volume into 1000 mL with DDI water. After 2g of soils were weighed into 50 mL plastic bottles, 20 mL of the extracting solution were added into the bottles, shaken at 180 oscillations p er m inute for 30 minutes, and then filtered through Whatman paper No. 42 (Mulvaney, 1996). Filtrate solutions were used to measure extractable concentrations of ammonium and nitrate. A similar procedure was used for ash samples, but extraction rate of 2g ash and 40 mL DDI water and a shaking time of 2 hours at 180 oscillations p er m inute (Qian et al., 2009b).

PAGE 152

152 The USEPA standard method of 350.1 was used to measure extractable NH 4 N in soil and ash. Principle of this method is that alkaline phenol and hypochlorite react with ammonia to form indophenol blue. Blue color is intensified by sodium nit roprusside, and measured at wavelength of 650 660 nm on the AQ2 discrete auto analyzer. The range of this method is from 0.02 to 2.0 mg NH 4 N L 1 with a level of detection of 0.007 mg NH 4 N L 1 Preparation of color reagents: Four color reagents applying for NH 4 N measurement by the AQ2 auto analyzer consisted of EDTA buffer, sodium phenate, sodium hypochlorite, and sodium nitroprusside. EDTA (ethylenediamine tetraacetic acid) buffer reagent is used to prevent interferences associated with precipitation pr oblems due to high contents of Ca and Mg in sample solutions. The other reagents are used to develop blue color for standard solutions and sample solutions. Procedures for color reagent preparation are described more details in the USEPA method of 350.1. Calibration method: An auto standard concentration of 2 mg NH 4 N L 1 was placed at the cup position # 1 on the roundtable of the AQ2 instrument. A s tandard calibration curve was set up daily at nine target concentrations of 0%, 2.5%, 5%, 10%, 25%, 50%, 75% and 100% of the standard concentration (2mg NH 4 N L 1 ). The establishment of calibration curve was based on the relationship between observed absorbance s and the target concentrations (Appendix C) QA/QC method: QA/QC was set up with two auto bracketing controls. The first one was CCV (continuing calibration verification) and CCB (continuing calibration blank). The concentration of CCV was 1 mg NH 4 N L 1 and the CCB is DDI water. The CCV and CCB were performed after every ten samples measured and after completion of calibration setup. Acceptable ranges of CCV and CCB were 0.9 1.1 mg NH 4 N L 1 and 0.1 to 0.1 mg NH 4 N L 1 The other control s included duplicates and spikes which were checked within a frequency of

PAGE 153

153 every 20 samples measured. Concentration of ammonium added into spike samples was 0.5 mg NH 4 N L 1 received from a spiking stock concentration of 20 mg NH 4 N/ L 1 which was set up at the position # 57 of the roundtable of the AQ2 instrument. Nitrate Procedures: Nitrate extraction was the same me thod as described for the NH 4 N extraction. The USEPA standard method of 353.2 was used to measure nitrate for samples. Principle of this method is that nitrate reacts with sulfanilamide to form a diazonium compound which couples with N (1 napththyl) ethe lenediamine dihydrochloride (NEDD) to create a reddish purple azo dye. This color is measured spectrophotometrically at a wavelength of 520 nm. The range of this method is from 0.03 to 4.5 mg NO 3 N L 1 with a detection limit level of 0.006 mg NO 3 N L 1 Preparation of color reagents: Working buffer and sulfanilamide NEDD are two main color reagents used for nitrate measurement by the AQ2 discrete auto analyzer. Preparation of these reagents is described step by step in the USEPA method of 353.2. The work ing buffer reagent is a mixed solution containing EDTA which can prevent interferences of iron, copper, and other metals in sample solutions. The sulfanilamide NEDD reagent is also a mixed solution comprising of sulfanilamide and NEDD that are used to dev elop color for nitrate. Calibration method: An auto standard concentration of 5 mg NO 3 N L 1 was put at the cup position # 1 on the roundtable of the AQ2 instrument. A calibration curve was set up daily based on a relationship between nine target concentra tions and their observed absorbance s The target concentrations included 0, 0.0333, 0.05, 0.1333, 0.25, 1.25, 2.5, 5, and 0 mg NO 3 N L created from percentages of the standard concentration of 5 mg NO 3 N L 1 (Appendix C)

PAGE 154

154 QA/QC method: Procedures of QA/Q C for nitrate measurement was established the same as the ammonium measurement. But concentration of CCV was 1 mg NO 3 N L 1 and concentration adding for spike sample was 0.5 mg NO 3 N L 1 from a spiking stock concentration of 50 mg NO 3 N L 1 which was set up at the position # 57 of the roundtable. Calculation: Concentrations of NH 4 N and NO 3 N were calculated as follows: [NH 4 + ] or [NO 3 ] (mg kg soil or ash) = w here X is a content of NH 4 N or NO 3 N obtained from AQ2 auto analyzer, Y is a volume of the extracting solution or DDI water added into soil samples (20 mL ) or ash samples (40 mL ). Extractable P hosphorus Tests for soil P availability are mainly based on understanding of chemical P forms existing in soils. Some chemical solutions have been widely used to extract potential P forms in soils. Bray and Kurtz (1945), so called Bray P 1 test, developed a method f or P extraction in the acid to neutral soils. Mehlich (1953) developed a P test, which is called Mehlich 1, for soils with low CEC in the South and Mid Atlantic. Soltanpour and Workman (1979) developed AB DTPA (Ammonium Bicarbonate Diethylene Triamine Pe ntaacetic Acid) to extract P in calcareous soils. In addition, Olsen P test (Olsen et al. 1954) and Mehlich 3 test (Mehli ch, 1984) were utilized to extract soil P within a wider range of soil including acid soils and alkaline soils. Four basic reactions can remove P from the solid phase of soil (Elrashidi, 2009), including (i) dissolving actions of acids; (ii) anion repl acement to enhance desorption; (iii) complexing of cations binding P; and (iv) hydrolysis of cations binding P. Selection of a P test must take into account the chemical forms of P in soils. Calcium and magnesium phosphates are the major P minerals in calc areous soils. Tests of AB DTPA, Olsen P, and Mehlich 3 can be applied to extract soil P in calcareous soils (Elrashidi, 2009; Kamprath et al., 2000); however,

PAGE 155

155 Olsen P test is primarily recommended for P extraction in calcareous soils because of its own adv antages (Orphanos, 1978; Vuc ns et al., 2008). The Olsen P test contains a single extractant of sodium bicarbonate (NaHCO 3 ) This extractant reduces calcium in solution through precipitation of calcium carbonate, controls dissolution of Ca phosphates, and removes dissolved and absorbed P on calcium carbonate and Fe oxide surfaces (Elrashidi, 2009). Although the AB DTPA test contains ammonium bicarbonate, this test also includes a multi element extractant (DTPA). This multi element extractant often causes an interference of P extraction and gives an underestimation of result s Bates (1990) used five different extractants (NaHCO 3 AB DTPA, Bray Kurtz P1, Bray Kurtz P2, and Mehlich 3) to test P availability. Availability of P was tested on 88 different soil ty pes having soil pH range from 5.0 to 7.6. Result s of Bates (1990) showed that two alkaline extractants (NaHCO 3 and AB DTPA) were equally effective and superior in P availability to three acidic extractants (Bray Kurtz P1, Bray Kurtz P2, and Mehlich 3). Cor relation coefficients (r 2 values) of extractable P to plant P uptake were 0.74, 0.73, 0.54, 0.65, and 0.66 for NaHCO 3 AB DTPA, Bray Kurtz P1, Bray Kurtz P2, and Mehlich 3 respectively P rediction of plant P uptake was improved after adding soil pH into r egression models, resulting in R 2 values of 0.80, 0.80, 0.70, 0.73, and 0.75 for NaHCO 3 AB DTPA, Bray Kurtz P1, Bray Kurtz P2, and Mehlich 3 respectively Compari n g P availability by three different extractants (Olsen test, AB DTPA test, Pi test) in the same soil and environmental conditions, Ahmad et al. (2006) showed that the Olsen test had the highest available P content (16.43 mg/kg), the Pi test (P extracted by Fe impregnated filter paper after 16h of shaking) had 10.6 mg P/kg, whereas the AB DTPA method gave the lowest available P contents (6.37 mg/kg). Nevertheless, Olsen test may underestimate or

PAGE 156

156 overestimate availability of P in calcareous soils. Chien et al. (2009) found that Olsen P was underestimated when calcareous soils contained more than 20% of gypsum, whereas Castro and Torrent (1995) proved that the Olsen P test overestimated P availability in soils relatively rich in carbonate and poor in noncarbonat e clay and Fe oxides. Soils in the Pine Rockland forest have high contents of noncarbonate clay and Fe oxides and low carbonate content. Content of carbonate clay was from 10.4 to 36.0 g/kg soil and noncarbonate clay content was 699 to 715 g/kg soil (Zhou and Li, 2001) According to the USDA soil survey (1958), the Pine Rockland soils contained 16.5 g of carbonate clay/kg soil and 447 g of noncarbonate clay/kg soil. Gypsum content is relatively low in the Pine Rockland because sulfate in the Pine Rockland is derive d mos tly from wet deposition. In this study, Olsen P test (Olsen et al., 1954) was chosen for assessing P availability in calcareous soils under the Pine Rockland forests. Procedures: Soil P was extracted by 0.5M NaHCO 3 solution (Olsen et al., 1954) using a ra tio of 2g soil to 40 mL 0.5M NaHCO 3 solution, and measured by the Beckman DU 640 spectrophotometer with an automated colorimetric method. The 0.5M NaHCO 3 extracting solution was prepared by placing 42g of NaHCO 3 into a 1000 mL volumetric flash containing 950 mL DDI water. After NaHCO 3 was completely dissolved by stirring, pH of extracting solution was adjusted into 8.5 by adding 4M NaOH, and diluted the solution to volume of 1 liter. 2g of soils were weighed in 50 mL plastic bot tles. 1g of activated carbon was also added into sample bottles to remove dark brown color created from reactions between charcoal in soil samples and the extracting solution. Blank samples and blank samples with activated carbon were performed repeatedly after every 20 soil samples measured for QA/QC checking. 40 mL of the extracting solution were added into soil samples and blank samples. These bottles were

PAGE 157

157 shaken at 180 oscillations per minute for 30 minutes, and filtered by Whatman paper No. 42 (Kuo, 19 96). Filtrates were used to measure extractable P contents. A similar method was used to extract P from ash samples with rate of 2g ash and 40 mL DDI water (Qian et al., 2009b). Preparation of P standard solutions: Seven P standard solutions preparing for calibration of spectrophotometer included 0, 0.1, 0.2, 0.4, 0.6, 0.8, and 1.0 mg P L 1 Volumes of 0, 0.5, 1, 2, 3, 4, 5 mL of 10mg P L 1 stock standard solution were transferred into 50 mL flash, and diluted to the volume by blank solution (0.5M NaHCO 3 ex tracting solution) to make above P standard solutions. Preparation of color reagent: Color reagent was prepared by the following steps: (1) Slowly transferring 152.8 mL concentrated H 2 SO 4 to a 1 liter beaker with 400 mL DDI water (2) Putting 20g of a mmonium m olybdate into 1000 mL flash with 300 mL DDI water, (3) Slowly transferring (1) into (2), (4) Adding 100 mL of Antimony Po tassium Tartrate (5%) into (3), (5) Diluting (4) into 1 liter by DDI water. Color reagent was daily prepar ed by adding L ascorbic acid to the mixed solution at rate of 1.5g of L a scorbic acid to 100 mL of the mixed solution. Calibration curve method and determination of P concentration: 5 mL of standard solutions and filtrates (sample solutions) were transferred into 20 mL vials, followed by addition of 1drop of P Nitrophenol and 0.5 mL of 5M H 2 SO 4 0.61 mL of the color reagent (rate of 1 mL color reagent: 9 mL sample solution) to develop the blue color for standard and sample solutions. After adding color reagent, the blue color in standard and sample solutions which was stabilized after 45 minutes was measured at 880 nm by spectrophotometer. Calibration curve was established basing a linear relationship between known concentrations of seven standard solutions and their obtained absorbances (Appendix C) Concentrations of P in sample solutions

PAGE 158

158 (mg P L ) were calculated from standard calibration curve and absorbance obtained from sample solutions. Contents of P in soil or ash samples were estimated as follows: P (mg P kg soil or ash) = w here X is the concentration of P in the filtrates of soil or ash samples, Y is the average P content of the blank samples with activated carbon, and Z is the average P conce ntrations of the blank samples without activated carbon. QA/QC method: Since blue color in sample solutions is stable from 45 to 60 minutes after adding the color reagent, only twenty samples were measured at the same time. In addition to duplicates and b lank samples, the blank standard solution was also measured after every 20 samples in order to recalibrate the instrument. Extractable Metals Melich 1, Melich 3, and AB DTPA are currently three methods widely applied to extract plant available contents o f metals. AB DTPA method developed by Soltanpour and Workman (1979) is globally utilized to extract macro and micro nutrients in calcareous soils. AB DTPA extractant with pH buffer of 7.6 works very well with calcareous soils, and is recommended by the UF /IFAS Extension Soil Testing Laboratory to extract plant available contents of macro and micro nutrients for calcareous soils in South Florida. The AB DTPA extracting solution contains two main components including ammonium bicarbonate which is used to ex tract chiefly macronutrients, and DTPA, a chelating agent, which is used to extract micronutrients. Many studies reported that the AB DTPA method has given good results on extractable contents of K, Fe, Mn, Cu, and Zn (Havlin and Soltanpour, 1981; Leggett and Argyle, 1983; O'Connor, 1988; Soltanpour and Workman, 1979) Moreover, the AB DTPA extracting solution was also assessed to be good for extracting calcium and magnesium. By comparing effectiveness

PAGE 159

159 of the AB DTPA method wi th other traditional methods in different types of calcareous soils, Lucena and Bascones (1993) concluded that AB DTPA extractant was not only valid for micronutrients but also good for calcium extraction in calcareous soils. By evaluating the effective ca pacity of AB DTPA method in extraction of nutrient elements in South Florida calcareous soils using compost amendments, Hanton et al. (1996) proved that nutrients and heavy metals could be monitored successfully using AB DTPA extractant, and AB DTPA extrac tant was a good indicator to measure Ca and Mg in the South Florida calcareous soils. In this study, therefore, AB DTPA method was selected to extract contents of nutrient elements. Preparation for AB DTPA extracting solution: 1.25 mL of NH 4 OH added into a 2500 mL glass bottle containing 1750 mL of DDI water. Approximately 4.925 g of DTPA was added into the bottle, and swirled the solution until DTPA was completely dissolved. Approximately 197.65g of NH 4 HCO 3 was added into the solution, swirled the solution to dissolve NH 4 HCO 3 and brought the volume into 2500 mL by DDI water. The mixed solution (the AB DTPA extracting solution) was adjusted to pH of 7.6 by adding either concentrated hydrochloric acid if the solution pH was higher than 7.6 or concentrated amm onium hydroxide if the buffer pH of the extracting solution was lower than 7.6 (Reed and Martens, 1996). Extraction procedures: 10g of soil samples were weighed and put into 50 mL plastic bottles. 20 mL of the AB DTPA extracting solution were dispensed into the bottles. The solution bottles were shaken for 15 minutes at 180 rpm by a reciprocating shaker, and then filtered through a filter paper of Whatman No. 42. 5 mL of the filtrate was transferred into another 50 mL plastic bottle, added 0.5 mL of concentrated HNO 3 into the bottle (10 mL extracted solution:1 mL HNO 3 ), and thereafter diluted it into 50 mL by DDI water. The solution acidified by HNO 3 was analyzed by the AA 6300 Atomic Absorption Spectrophotometer to measure

PAGE 160

160 contents of elements (C a, Mg, K, Na, Fe, Mn, Cu, and Zn). Ash samples were extracted by DDI water with 2g ash and 40 mL DDI water (Qian et al., 2009b). These samples were shaken in 2 hours at 180 rpm, and filtered by Whatman No. 42. Other extraction steps were similar to soil ex traction. Preparation of blank solution: Blank solution for AB DTPA method was prepared by adding 25 mL of concentrated HNO 3 and 250 mL of the AB DTPA extracting solution into a 2500 mL plastic bottle, and then diluting it into the volume of 2500 mL by DDI water. Instrument calibration: Standard calibration curve and standard addition are two major methods which can be applied to calibrate the AA 6300 Atomic Absorption Spectrophotometer. It is recommended that method of standard calibration curve shows a good linearity in the low concentration area; therefore, method of the 1 st order calibration curve was utilized to calibrate the instrument for each element measured. Objective of the calibration curve method is to use known standard concentrations of an element to obtain target concentrations of unknown solutions from soil samples. The 1 st order calibration curve for each element measurement was established by five different standard solutions. The standard solutions were prepared from a stock standard solution of 10 mg L 1 and blank solution. A stock standard solution (1000 mg L 1 ) of each element was diluted into a stock standard solution of 100 mg L 1 by DDI water, and the 100 mg L 1 standard solution was then diluted into a stock standard solution o f 10 mg L by DDI water. Since concentrations of Ca, Mg, K, Na, Fe, Mn, Zn, and Cu in filtrate solutions were relatively different, concentrations of five standard solutions for each element were applied at different standard concentrations ( Appendix C ). B asing on specific concentrations of five standard solutions for each element, different volumes of the 10 mg L 1 standard solution were added into 50 mL flashes, and then diluted into 50 mL by the blank solution.

PAGE 161

161 QA/QC method: Setting of QA/QC aimed to validate a created calibration curve and verify effects of pretreatment and liquid properties. Validation of the created calibration curve was carried out by ICV, ICB, CCV, CCB, and re slope of the calibration curve. Spike test a nd duplicate were applied to verify influences of the pretreatment after every 20 samples were measured. ICV (initial calibration verification) and ICB (initial calibration blank) were performed immediately after the calibration curve was created. ICB occu rred after ICV performance was intended to check that blank value decreased exactly. CCV (continuing calibration verification) and CCB (continuing calibration blank) were automatically performed with a frequency of every 20 samples measured. Concentrations of ICV and CCV were different from the standard concentrations used to create the calibration curve, but were in a concentration range of the calibration curve. Measured values of ICV, ICB, CCV, and CCB which ranged between 90% and 110% of their prepared criteria. Measurement was stopped if CRDL (Contract required detection limit) was exceeded the CCV, and CCB. I n addition, re slope of the calibration curve was also established in frequency of every 20 samples measured. Re slope measurement was carried out at the highest standard concentration used to constitute the calibration curve. The re slope was the sensitiv ity correction measurement, and slope of the calibration curve was corrected basing on measured absorbance. Establishment of measurement time: Pre pray time, integration time, and response time are three important factors that impact on QA/QC and measureme nt results. The pre pray time is time between two measurements, or interval of time that sampler of the instrument is rinsed by DDI water. The pre pray time was set up at 18 seconds (5 seconds by default in the instrument). Eighteen seconds was enough time to clean all residuals of previous solution. The integration

PAGE 162

162 time is time that needs for sampler to take sample solution and pray the solution into burner, or time that needs to measure a sample. The integration time was set up at 8 seconds (6 seconds by default). The response time is the repeated number of the integration. The integration was repeated in two times. Concentration of each sample was average value of two integration times. Concentrations of nutrient elements in soils were calculated by the f ollowing equation: Ca (mg Ca kg soil) = Where X is a concentration of Ca obtained from the atomic absorbance spectrophotometer. Concentrations of other elements were estimated as same as Ca calculation. Total N itrogen and T otal C arbon Contents of TC and TN were automatically measured by the Elemental CNS auto analyzer. 200 mg (0.2g) of soil, ash, or plant tissue samples were weighed in crucibles, and put in roundtable of the CN S instrument in order. Sampler of the instrument picked up crucibles, and measured automatically contents of TC and TN. Calibration of CNS instrument: CNS instrument was daily calibrated, and performed before samples measured. A standard calibration was se t up by seven standard samples including three blanks (three empty crucibles) and four standards (four crucibles with 250mg of glutamic acid). The glutamic acid with 9.52% N and 40.78% C is a standard chemical used to calibrate the instrument and to check QA/QC during the measurement. QA/QC setup: CNS operation is mainly depended on combustion process to measure contents of TC and TN. During the combustion process, CNS instrument uses a chemical mixture installed into a reduction tube of the instrument to burn all of soil or plant tissues at approximately 1000 0 C to measure carbon and nitrogen. QA/QC setup aimed to check out of the chemical mixture in the reduction tube. Duplicates of samples and Runin samples (250 mg of glutamic acid) were

PAGE 163

163 performed in ever y 20 samples measured in order to check existence of the chemical mixture. If there is no more chemical in the reduction tube, TN (%) in Runin sample is higher than 9.52 %, and TN (%) in duplicate is also higher than that of the same sample of duplicate. Total Phosphorus and Total Metal Procedures of digestion: The dry ashed method was applied to digest soil, ash, and plant tissue samples. 0.5 g of samples were weighed in 50 mL glass beakers and heated in a muffle furnace. Heating process was followed by heating samples at 250 0 C in 30 minutes, and continuing to increase temperature to 550 0 C and maintaining at this temperature for 4 hours, thereafter reducing temperature below 100 0 C before removing samples from the furnace. Burned soil samples were digested by 6M HCl following eight steps: firstly moistening burned soil samples in beakers by 3 drops of DDI water; secondly adding 20 mL of 6M HCl into the beakers and placing them on a hot plate at a controlled temperature (from 100 to 120 0 C) until dry, approxim ately 3 hours; thirdly after the beakers were dried, heating the beakers on the hot plate at a high temperature from 260 to 280 0 C for 30 minutes; fourthly removing the beakers from the hot plate and allowing them cool down to room temperature; fifthly mois tening the soil in beakers with 2 3 mL of DDI water, and then adding 2.25 mL of 6M HCl; sixthly putting simultaneously four beakers into the hot plate with the high temperature, and removing them out the plate when boiling was started; seventhly, transferr ing solutions in the beakers into 50 mL volumetric flasks, washing the beakers at least three times with DDI water, and bring the flasks into 50 mL by DDI water; finally filtering the solution with Whatman paper No. 42. The filtrates were used to measure c ontents of phosphorus and metals. For ash and plant tissue samples, digestion process was similar to soil digestion, except for steps of 2, 3, 4, and 6. Blank solution, standard solutions, and measurement of TP and TM: Blank solution was

PAGE 164

164 created by adding 112.5 mL 6M HCl into a 2500 mL plastic bottle, and bringing to volume of the bottle with DDI water. Preparation of standard solutions for TP and TM was the same steps as described by sections of extractable P and extractable metals, but the above blank so lution was used to create standard solutions for TP and TM measurements. Total contents of P and metals were determined by methods as presented in QA/QC sections of extractable P and metals. QA/QC method: In addition to QA/QC setup for P measurement as de scribed in the section of extractable P, a standard soil of SQCO 019 from NSI was applied for QA/QC in TP measurement. The standard soil sample having a standard concentration of 1040 mg P kg soil, in companion with duplicate and blank sample, was checked after every 20 samples measured. QA/QC setup for TM measurement was similar to section of extractable metals.

PAGE 165

165 APPENDIX C CALIBRATION CURVES OBTAINED FROM LABORATORY ANALYSES, AND STANDARD SOLUTIONS USED FOR METAL MEASUREMENT Calibration Chart Type Observed Calculated Target % Error S1 0.0314 0.0228 0.0000 S90 0.0731 0.0516 0.0500 3.2250 S91 0.1026 0.1044 0.1000 4.3537 S92 0.1597 0.2065 0.2000 3.2354 S93 0.3298 0.5106 0.5000 2.1118 S94 0.6062 1.0045 1.0000 0.4506 S95 0.8858 1.5043 1.5000 0.2884 S96 1.1581 1.9910 2.0000 0.4521 S0 0.0328 0.0204 0.0000 Polynomial Order: 1 Correlation Coefficient: 0.9999 Carryover: 0.1 Date & Time: Wed May 05 13:53:18 2010 Calibration Graph Figure C 1. Calibration chart and curve obtained from NH 4 N m easurement by AQ2 Calibration Chart Type Observed Calculated Target % Error S1 0.0014 0.0215 0.0000 S90 0.0050 0.0029 0.0333 91.3896 S91 0.0102 0.0384 0.0500 23.2672 S92 0.0212 0.1129 0.1333 15.3413 S93 0.0436 0.2646 0.2500 5.8429 S94 0.1941 1.2839 1.2500 2.7089 S95 0.3867 2.5879 2.5000 3.5154 S96 0.7352 4.9477 5.0000 1.0455 S0 0.0047 0.0009 0.0000 Polynomial Order: 1 Correlation Coefficient: 0.9997 Carryover: 0.5 Date & Time: Wed Apr 21 08:03:37 2010 Calibration Graph Figure C 2. Calibration chart and curve obtained from NO 3 N m easurement by AQ2

PAGE 166

166 Absorbance Standard concentration (mg P/L) 0.0017 0.0 0.0722 0.1 0.1316 0.2 0.2477 0.4 0.3628 0.6 0.4781 0.8 0.5834 1.0 Figure C 3. Standard solutions and calibration curve obtained from P measurement by spectrophotometer Table C 1. Standard solutions and detection limits used for metal measurement Elements Standard concentrations appli ed for calibration curve method (mg/L) Detection limit (mg/L) 0 0.2 0.4 0.5 0.6 0.8 1.0 2.0 3.0 5.0 Ca x x x x x 0.02 K x x x x x 0.0025 Na x x x x x 0.001 Fe x x x x x 0.025 Mg x x x x x 0.001 Mn x x x x x 0.01 Zn x x x x x 0.002 Cu x x x x x 0.008 y = 1.7257x 0.0200 R = 0.9992 0.0 0.2 0.4 0.6 0.8 1.0 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 Concentration (mg P/L) Absorbance Standard concentration (mg P/L)

PAGE 167

167 APPENDIX D COVARIANCE STRUCTURE MODELS AND ANOVA RESULTS BY REPEATED MEASURES ANALYSIS Figure D 1. Covariance structure models applied for ANOVA analysis by repeated measures Results of ANOVA Analysis by Repeated Measures Method Ca 2+ The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable Ca Covariance Structure Compound Symmetry Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations

PAGE 168

168 Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 449.1 AIC (smaller is better) 453.1 AICC (smaller is better) 453.5 BIC (smaller is better) 452.4 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 32 2.25 0.0497 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 1059.76 48.4434 32.9 21.88 <.0001 Time 14 904. 61 48.4434 32.9 18.67 <.0001 Time 30 1035.04 48.4434 32.9 21.37 <.0001 Time 90 972.20 48.4434 32.9 20.07 <.0001 Time 180 1094.06 48.4434 32.9 22.58 <.0001 Time 270 988.70 48.4434 32.9 20.41 <.0001 Time 360 1161.85 48.4434 32.9 23.98 <.0001 Time 450 1048.94 48.4434 32.9 21.65 <.0001 Time 540 971.95 48.4434 32.9 20.06 <.00 01 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for Ca2_ This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of F reedom 36 Error Mean Square 11733.83 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 225.88 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 1161.85 5 360 B A 1094.06 5 180 B A 1059.76 5 0 B A 1048.94 5 450 B A 1035.04 5 30 B A 988.70 5 270 B A 972.20 5 90 B A 971.95 5 540 B 904.61 5 14

PAGE 169

169 Mg 2+ The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Varia ble Mg Covariance Structure Autoregressive Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 346.9 AIC (smaller is better) 350.9 AICC (smaller is better) 351.2 BIC (smaller is better) 350.1 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 23.3 9.75 <.0001 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 52.1450 11.6969 28.5 4.46 0.0001 Time 14 143.66 11.6969 28.5 12.28 <.0001 Time 30 80.1241 11.6969 28.5 6.85 <.00 01 Time 90 40.2032 11.6969 28.5 3.44 0.0018 Time 180 53.6432 11.6969 28.5 4.59 <.0001 Time 270 55.6509 11.6969 28.5 4.76 <.0001 Time 360 53.4054 11.6969 28.5 4.57 <.0001 Time 450 61.7985 11.6969 28.5 5.28 <.0001 Time 540 33.7322 11.6969 28.5 2.88 0.0074 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for Mg2_ This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ.

PAGE 170

170 Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 688.3541 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 54.71 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 143.66 5 14 B 80.12 5 30 B 61.80 5 450 B 55.65 5 270 B 53.64 5 180 B 53.41 5 360 B 52.15 5 0 B 40.20 5 90 B 3 3.73 5 540 K + The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable K Covariance Structure Autoregressive Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Base d Degrees of Freedom Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 398.4 AIC (smaller is better) 402.4 AICC (smaller is better) 402.7 BIC (smaller is better) 401.6 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 22.1 5.14 0.0011

PAGE 171

171 Least Squares Me ans Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 164.79 22.8170 33.5 7.22 <.0001 Time 14 277.33 22.8170 33.5 12.15 <.0001 Time 30 193.48 22.8170 33.5 8.48 <.0001 Time 90 187.56 22.8170 33.5 8.22 <.0001 Time 180 236.42 22.8170 33.5 10.36 <.0001 Time 270 121.79 22.8170 33.5 5.34 <.00 01 Time 360 144.82 22.8170 33.5 6.35 <.0001 Time 450 159.86 22.8170 33.5 7.01 <.0001 Time 540 136.82 22.8170 33.5 6.00 <.0001 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for K_ This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 2609.794 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 106.53 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 277.33 5 14 B A 236.42 5 180 B A C 193.48 5 30 B A C 187.56 5 90 B C 164.79 5 0 B C 159.86 5 450 B C 144.82 5 360 B C 136.82 5 540 C 121.79 5 270 Fe 2+ The Mixed Procedur e Model Information Data Set WORK.CN2 Dependent Variable Fe Covariance Structure Autoregressive Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5

PAGE 172

172 Number of Observations Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 406.7 AIC (smaller is better) 410.7 AICC (smaller is better) 411.1 BIC (smaller is better) 410.0 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 22.3 3.31 0.0122 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 177.87 25.5837 33.8 6.95 <.0001 Time 14 153.31 25.5837 33.8 5.99 <.0001 Time 30 207.18 25.5837 33.8 8.10 <.0001 Time 90 169.80 25.5837 33.8 6.64 <.0001 Time 180 158.06 25.5837 33.8 6.18 <.00 01 Time 270 288.91 25.5837 33.8 11.29 <.0001 Time 360 172.77 25.5837 33.8 6.75 <.0001 Time 450 196.97 25.5837 33.8 7.70 <.0001 Time 540 222.59 25.5837 33.8 8.70 <.0001 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for Fe2_ This test c ontrols the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 3284.141 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 119.5 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 288.91 5 270 B A 222.59 5 540 B A 207.18 5 30 B A 196.97 5 450 B A 177.87 5 0 B A 172.77 5 360 B A 169.80 5 90 B 158.06 5 180 B 153.31 5 1 4

PAGE 173

173 Mn 2+ The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable Mn Covariance Structure Autoregressive Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Method Satterthwaite Class Level Info rmation Class L evels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 375.5 AIC (smaller is better) 379.5 AICC (smaller is better) 379.9 BIC (smaller is better) 378.7 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 19.1 16.04 <.0001 Least Squares Mea ns Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 69.2158 17.0049 30.5 4.07 0.0003 Time 14 256.37 17.0049 30.5 15.08 <.0001 Time 30 89.2194 17.0049 30.5 5.25 <.0001 Time 90 46.9500 17.0049 30.5 2.76 0.0097 Time 180 53.1232 17.0049 30.5 3.12 0.0039 Time 270 89.3215 17.0049 30.5 5.25 <.0001 Time 360 72.6488 17.0049 30.5 4.27 0.0002 Time 450 113.92 17.0049 30.5 6.70 <.00 01 Time 540 58.8743 17.0049 30.5 3.46 0.0016 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for Mn2_ This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ.

PAGE 174

174 Alpha 0.05 E rror Degrees of Freedom 36 Error Mean Square 1444.268 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 79.247 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 256.37 5 14 B 113.92 5 450 B 89.32 5 270 B 89.22 5 30 B 72.65 5 360 B 69.22 5 0 B 58.87 5 540 B 53.12 5 180 B 46.95 5 90 Zn 2+ The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable Zn Covariance Structure Autoregressive Subject Eff ect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. The Mixed Procedure Fit Statistics 2 Res Log Likelihood 154.1 AIC (smaller is better) 158.1 AICC (smaller is better) 1 58.5 BIC (smaller is better) 157.4 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F

PAGE 175

175 Time 8 21.1 1.99 0.0989 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 3.4637 0.7582 35.1 4.57 <.0001 Time 14 5.1102 0.7582 35.1 6.74 <.0001 Time 30 5.3852 0.7582 35.1 7.10 <.0001 Time 90 3.2660 0.7582 35.1 4.31 0.0001 Time 180 5.3352 0.7582 35.1 7.04 <.0001 Time 270 4.1550 0.7582 35.1 5.48 <.0001 Time 360 5.2907 0.7582 35.1 6.98 <.0001 Time 450 4.0349 0.7582 35.1 5.32 <.0001 Time 540 2.3814 0.7582 35.1 3.14 0.0034 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for Zn2_ This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate th an REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 2.880462 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 3.5391 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 5.385 5 30 A 5.335 5 180 A 5.291 5 360 A 5.110 5 14 A 4.155 5 270 A 4.035 5 450 A 3.464 5 0 A 3.266 5 90 A 2.381 5 540 Cu 2+ The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable Cu Covariance Structure Autoregressive Subject Effect Subject Es timation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5

PAGE 176

176 Number of Observ ations Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 89.7 AIC (smaller is better) 93.7 AICC (smaller i s better) 94.1 BIC (smaller is better) 92.9 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 20.5 1.46 0.2317 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 3.4427 0.3086 35.5 11.16 <.0001 Time 14 4.1178 0.3086 35.5 13.34 <.0001 Time 30 3.06 44 0.3086 35.5 9.93 <.0001 Time 90 3.8265 0.3086 35.5 12.40 <.0001 Time 180 3.8131 0.3086 35.5 12.36 <.0001 Time 270 3.5783 0.3086 35.5 11.59 <.0001 Time 360 4.2436 0.3086 35.5 13.75 <.0001 Time 450 3.9561 0.3086 35.5 12.82 <.0001 Time 540 3.7901 0.3086 35.5 12.28 <.0001 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for Cu2_ This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate th an REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 0.476144 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 1.4389 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 4.2436 5 360 A 4.1178 5 14 A 3.9561 5 450 A 3.8265 5 90 A 3.8131 5 180 A 3 .7901 5 540 A 3.5783 5 270 A 3.4427 5 0 A 3.0644 5 30

PAGE 177

177 PO4 P The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable P Covariance Structure Compou nd Symmetry Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 158.6 AIC (smaller is better) 162.6 AICC (smaller is better) 1 63.0 BIC (smaller is better) 161.9 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 32 16.19 <.0001 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 4.2394 0.8289 33.8 5.11 <.0001 Time 14 13.3829 0.8289 33.8 16.15 <.0001 Time 30 4.4067 0.8289 33.8 5.32 <.0001 Time 90 1.2249 0.8289 33.8 1.48 0.1488 Time 180 3.95 28 0.8289 33.8 4.77 <.0001 Time 270 3.1879 0.8289 33.8 3.85 0.0005 Time 360 4.5643 0.8289 33.8 5.51 <.0001 Time 450 2.7131 0.8289 33.8 3.27 0.0025 Time 540 3.0140 0.8289 33.8 3.64 0.0009 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for PO4_P This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ.

PAGE 178

178 Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 3.435366 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 3.865 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 13.383 5 14 B 4.564 5 360 B 4.407 5 30 B 4.239 5 0 B 3.953 5 180 B 3.188 5 27 0 B 3.014 5 540 B 2.713 5 450 B 1.225 5 90 NH4 N The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable NH4_N Covariance Structure Compound Symmetry Subject Effect Subject Estimation Method REML Residual Variance Method Profile F ixed Effects SE Method Model Based Degrees of Freedom Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations Number of Observations Re ad 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 316.7 AIC (smaller is better) 320.7 AICC (smaller is better) 321.1 BIC (smaller is better) 319.9 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F T ime 8 32 12.98 <.0001

PAGE 179

179 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 40.5855 7.2474 35.3 5.60 <.0001 Time 14 99.5711 7.2474 35.3 13.74 <.0001 Time 30 53.5506 7.2474 35.3 7.39 <.0001 Time 90 23.0816 7.2474 35.3 3.18 0.0030 Time 180 31.2149 7.2474 35.3 4.31 0.0001 Time 270 21.6083 7.2474 35.3 2.98 0.0052 Time 360 22.1453 7.2474 35.3 3.06 0.0043 Time 450 22.99 57 7.2474 35.3 3.17 0.0031 Time 540 11.3979 7.2474 35.3 1.57 0.1247 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for NH4_N This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 262.6211 Critical Value of Studentized Range 4.662 79 Minimum Significant Difference 33.793 Means with the same letter are not significantly different. Tukey Grou ping Mean N Time A 99.57 5 14 B 53.55 5 30 C B 40.59 5 0 C B 31.21 5 180 C B 23.08 5 90 C B 23.00 5 450 C B 22.15 5 360 C B 21.61 5 270 C 11.40 5 540 NO3 N The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable NO3_N Covariance Structure Autoregressive Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Method Satterthwait e Class Le vel Information Clas s Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations

PAGE 180

180 Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res L og Likelihood 256.2 AIC (smaller is better) 260.2 AICC (smaller is better) 260.6 BIC (smaller is better) 259.4 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 21.6 5.22 0.0010 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 8.1121 3.1089 35.9 2.61 0.0131 Time 14 11.3295 3.1089 35.9 3.64 0.0008 Time 30 2.4052 3.1089 35.9 0.77 0.4442 Time 90 5.1645 3.1089 35.9 1.66 0.1054 Time 180 25.1373 3.1089 35.9 8.09 <.0001 Time 270 8.5222 3.1089 35.9 2.74 0.0095 Time 360 2.6389 3.1089 35.9 0.85 0.4016 Time 450 2.7202 3.1089 35.9 0.87 0.3874 Time 540 10.5369 3.1089 35.9 3.39 0.0017 The ANOVA Procedure Tukey's Stud entized Range (HSD) Test for NO3_N This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 48.33865 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 14.498 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 25.137 5 180 B A 11.330 5 14 B 10.537 5 540 B 8.522 5 270 B 8.112 5 0 B 5.165 5 90 B 2.720 5 450 B 2.639 5 360 B 2.405 5 30

PAGE 181

181 EC The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable EC Covariance Structure Compound Symmetry Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 444.8 AIC (sma ller is better) 448.8 AICC (smaller is better) 449.2 BIC (smaller is better) 448.0 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 32 8.27 <.0001 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 333.33 42.8200 35.5 7.78 <.0001 Time 14 697.64 42.8200 35.5 16.29 <.0001 Time 30 543.84 42.8200 35.5 12.70 <.0001 Time 90 370.08 42.8200 35.5 8.64 <.0001 Time 180 383.34 42.8200 35.5 8.95 <.0001 Time 270 348.22 42.8200 35.5 8.13 <.0001 Time 360 328.16 42.8200 35.5 7.66 <.0001 Time 450 452.72 42.8200 35.5 10.57 <.0001 Time 540 331.50 42.8200 35.5 7.74 <.0001 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for EC This test controls the Type I experimentwise error rate, b ut it generally has a higher Type II error rate than REGWQ.

PAGE 182

182 Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 9167.775 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 199.66 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 697.64 5 14 B A 543.84 5 30 B C 452.72 5 4 50 B C 383.34 5 180 B C 370.08 5 90 B C 348.22 5 270 C 333.33 5 0 C 331.50 5 540 C 328.16 5 360 pH The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable pH Covariance Structure Autoregressive Subject Effect Subject Estimation Method RE ML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number o f Observations Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 18.1 AIC (smaller is better) 14.1 AICC (smaller is better) 13.7 BIC (smaller is better) 14.8 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 21.7 3.04 0.0187

PAGE 183

183 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 7.5480 0.06889 36 109.57 <.00 01 Time 14 7.8220 0.06889 36 113.54 <.0001 Time 30 7.4780 0.06889 36 108.55 <.0001 Time 90 7.4460 0.06889 36 108.09 <.0001 Time 180 7.5600 0.06889 36 109.74 <.0001 Time 270 7.4840 0.06889 36 108.64 <.0001 Time 360 7.5720 0.06889 36 109.91 <.0001 Time 450 7.7160 0.06889 36 112.01 <.0001 Time 540 7.5820 0.06889 36 110.06 <.0001 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for pH This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 0.023732 Critical Value of Studentized Range 4.66279 Minimum Significa nt Difference 0.3212 Means w ith the same letter are not significantly different. Tukey Grouping Mean N Time A 7.82200 5 14 B A 7.71600 5 450 B A 7.58200 5 540 B A 7.57200 5 360 B A 7.56000 5 180 B A 7.54800 5 0 B 7.48400 5 270 B 7.47800 5 3 0 B 7.44600 5 90 TP (%) The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable TP Covariance Structure Autoregressive Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Mod el Based Degrees of Freedom Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations

PAGE 184

184 Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 423.1 AIC (smaller is better) 427.1 AICC (smaller is better) 427.5 BIC (smaller is better) 426.3 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 20.6 1.60 0.1845 Least Squares Mea ns Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 321.82 31.8915 34.7 10.09 <.0001 Time 14 422.93 31.8915 34.7 13.26 <.0001 Time 30 301.09 31.8915 34.7 9.44 <.0001 Time 90 347.14 31.8915 34.7 10.88 <.0001 Time 180 340.12 31.8915 34.7 10.66 <.0001 Time 270 276.50 31.8915 34.7 8.67 <.0001 Time 360 376.98 31.8915 34.7 11.82 <.0001 Time 450 337.04 31.8915 34.7 10.57 <.00 01 Time 540 316.92 31.8915 34.7 9.94 <.0001 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for TP____ This test controls the Type I experimentwise e rror rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 0.000051 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 0.0149 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 0.042293 5 14 A 0.037698 5 360 A 0.034714 5 9 0 A 0.034012 5 180 A 0. 033704 5 450 A 0.031182 5 0 A 0.031692 5 540 A 0.030109 5 30 A 0.027650 5 270

PAGE 185

185 TN (%) The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable TN____ Covariance Structure Compound Symmetry Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Meth od Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations Number of Observations Read 45 Number of Observations Used 45 Number of Observations Not Used 0 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 19.9 AIC (smaller is better) 23.9 AICC (smaller i s better) 24.2 BIC (smaller is better) 23.1 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 32 3.55 0.0048 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 0.9303 0.1189 34.4 7.83 <.0001 Time 14 1.4151 0.1189 34.4 11.91 <.0001 Time 30 0.9056 0.1189 34.4 7.62 <.0001 Time 90 0.6 896 0.1189 34.4 5.80 <.0001 Time 180 0.8386 0.1189 34.4 7.06 <.0001 Time 270 0.9835 0.1189 34.4 8.28 <.0001 Time 360 1.2838 0.1189 34.4 10.80 <.0001 Time 450 0.9575 0.1189 34.4 8.06 <.0001 Time 540 0.7871 0.1189 34.4 6.62 <.0001 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for TN____

PAGE 186

186 This test controls the Type I experimentwise error rate, but it g enerally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 0.070635 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 0.5542 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 1.4151 5 14 B A 1.2838 5 360 B A C 0.9835 5 270 B A C 0.9575 5 450 B A C 0.9517 5 0 B A C 0.9056 5 30 B C 0.8386 5 180 B C 0.7871 5 540 C 0.6896 5 90 TC (%) The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable TC (%) Covariance Structure Autoregressive Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Meth od Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 263.3 AIC (smaller is better) 267.3 AICC (smaller is better) 267.6 BIC (smaller is better) 266.5 Type 3 Tests of Fixed Effects Num Den

PAGE 187

187 Effect DF DF F Value Pr > F T ime 8 21.5 4.89 0.0015 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 24.1156 3.4326 35.8 7.03 <.0001 Time 14 39.4621 3.4326 35.8 11.50 <.0001 Time 30 25.2949 3.4326 35.8 7.37 <.0001 Time 90 15.0340 3.4326 35.8 4.38 <.0001 Time 180 22.9942 3.4326 35.8 6.70 <.0001 Time 270 28.0598 3.4326 35.8 8.17 <.0001 Time 360 33.7697 3.4326 35.8 9.84 <.0001 Time 450 32.94 31 3.4326 35.8 9.60 <.0001 Time 540 19.8249 3.4326 35.8 5.78 <.0001 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for TC____ This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 58.93609 Critical Value of Studentized Range 4.662 79 Minimum Significant Difference 16.009 Means with the same letter are not significantly different. Tukey G rouping Mean N Time A 39.462 5 14 B A 33.770 5 360 B A 32.943 5 450 B A C 28.060 5 270 B A C 25.295 5 30 B A C 28.516 5 0 B C 22.994 5 180 B C 19.825 5 540 C 15.034 5 90 Soil C:N ratio The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable C_N_ratio Covariance Structure Compound Symmetry Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Method Satterthwait e Class Le vel Information Clas s Levels Values Time 9 0 14 30 90 180 270 360 450 540

PAGE 188

188 Subject 5 1 2 3 4 5 Number of Observations Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res L og Likelihood 249.8 AIC (smaller is better) 253.8 AICC (smaller is better) 254.2 BIC (smaller is better) 253.1 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 32 1.36 0.2532 Least Squares Means S tandard Effect Time Estimate Error DF t Value Pr > |t| Time 0 26.8136 2.8571 35.5 9.38 <.0001 Time 14 28.1136 2.8571 35.5 9.84 <.0001 Time 30 29.2028 2.8571 35.5 10.22 <.0001 Time 90 21.9202 2.8571 35.5 7.67 <.0001 Time 180 27.3310 2.8571 35.5 9.57 <.0001 Time 270 28.2735 2.8571 35.5 9.90 <.0001 Time 360 26.7711 2.8571 35.5 9.37 <.0001 Time 450 34.7003 2.8571 35.5 12.15 <.0001 Time 540 25.3755 2.8571 35.5 8.88 <.0001 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for C_N_ratio This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 40.81496 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 13.322 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 34.700 5 450 A 29.203 5 30 A 28 .274 5 270 A 28.114 5 14 A 27.331 5 180 A 26.814 5 0 A 26.771 5 360

PAGE 189

189 A 25.375 5 540 A 21.920 5 90 Soil N:P ratio The Mixed Procedure Model Information Data Set WORK.CN2 Dependent Variable N_P_ratio Covariance Structure Autoregressive Subject Effect Subject Estimation Method REML Residual Variance Method Profile Fixed Effects SE Method Model Based Degrees of Freedom Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations Number of Observations Read 45 Number of Observations Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 253.0 AIC (smaller is better) 257.0 AICC (smaller is better) 257.3 BIC (sma ller is better) 256.2 Type 3 Tests of Fixed Effects Num Den Effect DF DF F Value Pr > F Time 8 19.3 2.74 0.0334 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 29.1075 3.1579 28.8 9.22 <.0001 Time 14 33.6452 3.1579 28.8 10.65 <.0001 Time 30 30.0741 3.1579 28.8 9.52 <.0001 Time 90 22.9960 3.1579 28.8 7.28 <.0001 Time 180 25.2722 3.1579 28.8 8.00 <.0001 Time 270 35.7968 3.1579 28.8 11.34 <.0001 Time 360 33.3465 3.1579 28.8 10.56 <.0001 Time 450 28.2369 3.1579 28.8 8.94 <.0001 Time 540 24.7085 3.1579 28.8 7.82 <.0001 The ANOVA Procedure

PAGE 190

190 Tukey's Studentized Range (HSD) Test for N_P_ratio This test controls the Type I experimentwise error rate, but it generally ha s a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 49.94508 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 14.737 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 35.797 5 270 B A 33.645 5 14 B A 33.346 5 360 B A 30.074 5 30 B A 29.108 5 0 B A 28.237 5 450 B A 25.272 5 180 B A 24.709 5 540 B 22.996 5 90 Soil C:P ratio The Mixed Procedure Model Information Data Set WORK.CN1 Dependent Variable C_P_ratio Covariance Structure Autoregressive Subject Effect Subject Estimation Method REML Residual Variance Method Profil e Fixed Effects SE Method Model Based Degrees of Freedo m Method Satterthwaite Class Level Information Class Levels Values Time 9 0 14 30 90 180 270 360 450 540 Subject 5 1 2 3 4 5 Number of Observations Number of Observations Read 45 Number of Observat ions Used 45 Convergence criteria met. Fit Statistics 2 Res Log Likelihood 509.8 AIC (smaller is better) 513.8 AICC (smaller is better) 514.2 BIC (smaller is better) 513.1 Type 3 Tests of Fixed Effects

PAGE 191

191 Num Den Effect DF DF F Value Pr > F Time 8 20.7 2.82 0.0277 Least Squares Means Standard Effect Time Estimate Error DF t Value Pr > |t| Time 0 771.83 106.55 34.5 7.24 <.0001 Time 14 942.59 106.55 34.5 8.85 <.0001 Time 30 842.29 106.55 34.5 7.91 <.0001 Time 90 503.21 106.55 34.5 4.72 <.00 01 Time 180 693.08 106.55 34.5 6.50 <.0001 Time 270 1011.41 106.55 34.5 9.49 <.0001 Time 360 893.09 106.55 34.5 8.38 <.0001 Time 450 992.34 106.55 34.5 9.31 <.0001 Time 540 624.15 106.55 34.5 5.86 <.0001 The ANOVA Procedure Tukey's Studentized Range (HSD) Test for C_P_ratio This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 36 Error Mean Square 56925.29 Critical Value of Studentized Range 4.66279 Minimum Significant Difference 497.52 Means with the same letter are not significantly different. Tukey Grouping Mean N Time A 1011.4 5 270 B A 992.3 5 450 B A 942.6 5 14 B A 893.1 5 360 B A 842.3 5 30 B A 771.8 5 0 B A 693.1 5 180 B A 624.2 5 540 B 503.2 5 90

PAGE 192

192 APPENDIX E PROCEDURES FOR THE LABORATORY INCUBATION EXPERIMENT Figure E 1. A 30 x 30cm sampling frame used to collect fuel load and soil samples Figure E 2. Procedures for preparation. a) s oil s, b) f uel l oads, c ) bottles, d) m uffle f urnace a b d c

PAGE 193

193 Figure E 3. Ash p roduction at different heating t emperatures Figure E 4. Procedures for incubation. a) mixing ash products, b) applying ashes into samples, c) applying DI wat er for soil water content, and d) incubating oil samples a b d a c

PAGE 194

194 APPENDIX F PROCEDURES FOR DETERMINATION OF SOIL FIELD CAPACITY Table F 1. Results of determination of soil field capacity Replicates Air dried weight (g) Drained soil weight (g) Oven dried soil weight (g) Field capacity (%) Water content in air dry soil (%) a b c d = (b c)*100/c e = (a c)*100/c 1 100.09 199.57 88.04 126.68 13.69 2 100.04 199.07 88.14 125.86 13.50 3 100.06 192.17 83.78 129.37 19.43 4 100.03 193.02 86.83 122.30 15.20 Average 100.1 196.0 86.7 126.1 15.5 The water amounts at 35%, 70%, and 100% of field capacity applied for the incubation experiment were calculated as follows: W ater amount in 150g air dried soil: 150g soil x 15.5% = 23.25 ml H 2 O 150 g air dried soil contains: 126.75 g oven dried soil + 23.25 ml H 2 O 35% of field capacity : 126.1% x 35% = 44.1% 70% of field capacity: 126.1% x 70% = 88.27% 100% of field capacity: 126.1% x 100% = 126.1% Water amount added into 150g ai r dried soil samples at 35% of field capacity : 44.1% x 126.75 g soil 23.25 ml H 2 O = 33 ml Water amount added into 150g air dried soil samples at 70% of field capacity : 88.2 7% x 126.75 g soil 23.25 ml H 2 O = 88 ml Water amount added into air d ried soil samples at 100% of field capacity : 126. 1% x 126.75 g soil 23.25 ml H 2 O = 136 ml

PAGE 195

195 APPENDIX G RESULTS OF MULTIPLE REGRESSION ANALYSES BY PROC GLM PROCEDURES FOR THE LABORATORY INCUBATION EXPERIMENT The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: P P Sum of Source DF Squares Mean Square F Value Pr > F Model 7 891.880275 127. 411468 48.29 <.0001 Error 208 548.790959 2.638418 Corrected Total 215 144 0.671234 R S quare Coeff Var Root MSE P Mean 0.619073 25.37268 1.624321 6.401850 Source DF Type I SS Mean Square F Value Pr > F Period 1 107.0590408 107.0590408 40.58 <.0001 Moisture 1 137.8747888 137.8747888 52.26 <.0001 Temperature 1 578.4344802 578.4344802 219.24 <.0001 Period*Moisture 1 13.1956077 13.1956077 5.00 0.0264 Moisture*Temperature 1 53.7227858 53.7227858 20.36 <.0001 Period*Temperature 1 0.0403640 0.0403640 0.02 0.9017 Period*Moistu*Temper 1 1.5532081 1.5532081 0.59 0.4438 Dependent Variable: P P Sum of Source DF Squares Mean Square F Value Pr > F Model 6 890.327067 148.387845 56.35 <.0001 Error 209 550.344167 2.633226 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE P Mean 0.617995 25.34770 1.622722 6.401850 Source DF Type I SS Mean Square F Value Pr > F Period 1 107.0590408 107.0590408 40.66 <.0001 Mo isture 1 137.8747888 137.8747888 52.36 <.0001 Temperature 1 578.4344802 578.4344802 219.67 <.0001 Period*Moisture 1 13.1956077 13.1956077 5.01 0.0262 Moisture*Temperature 1 53.7227858 53.7227858 20.40 <.0001 Period*Temperature 1 0.0403640 0.0403640 0.02 0.9016 Dependent Variable: P P Sum of Source DF Squares Mean Square F Value Pr > F Model 5 890.286703 178.057341 67.94 <.0001 Error 210 550.384531 2.620879 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE P Mean 0.617967 25.28820 1.618913 6.401850

PAGE 196

196 Source DF Type I SS Mean Square F Value Pr > F Period 1 107.0590408 107.0590408 40.85 <.0001 Moisture 1 137.8747888 137.8747888 52.61 <.0001 Temperature 1 578.4344802 578.4344802 220.70 <.0001 Period*Moisture 1 13.1956077 13.1956077 5.03 0.0259 Moisture*Temperature 1 53.7227858 53.7 227858 20.50 <.0001 The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: NH4 NH4 Sum of Source DF Squares Mean Square F Value Pr > F Model 7 251552.4944 35936.0706 86.33 <.0001 Error 208 86579.0761 416.2456 Corrected Total 215 338131.5705 R Square Coeff Var Root MSE NH4 Mean 0.743949 20.94201 20.40210 97.42185 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 82518.9500 82518.9500 198.25 <.0001 Period 1 156699.0187 156699.0187 376.46 <.0001 Temperature 1 5586.9453 5586.9453 13.42 0.0003 Moisture*Period 1 6131.3155 6131.3155 14.73 0.0002 Moisture*Temperature 1 1.8245 1.8245 0.00 0.9473 Period*Temperature 1 425.7091 425.7091 1.02 0.3130 Moistu*Period*Temper 1 188.7313 188.7313 0.45 0.5015 Dependent Variable: NH4 NH4 Sum of Source DF Squares Mean Square F Value Pr > F Model 6 251363.7631 41893.9605 100.91 <.0001 Error 20 9 86767.8074 415.1570 Corrected Total 215 338131.5705 R Square Coeff Var Root MSE NH4 Mean 0.743390 20.91461 20.37540 97.42185 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 82518.9500 82518.9500 198.77 <.0001 Period 1 156699.0187 156699.0187 377.45 <.0001 Temperature 1 5586.9453 5586.9453 13.46 0.0003 Moisture*Period 1 6131.3155 6131. 3155 14.77 0.0002 Moisture*Temperature 1 1.8245 1.8245 0.00 0.9472 Period*Temperature 1 42 5.7091 425.7091 1.03 0.3124 Dependent Variable: NH4 NH4 Sum of Source DF Squares Mean Square F Value Pr > F

PAGE 197

197 Model 4 250936.2295 62734.0574 151.81 <.0001 Error 211 87195.3410 413.2481 Corrected Total 215 338131.5705 R Square Coeff Var Root MSE NH4 Mean 0.742126 20.86647 20.32850 97.42185 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 82518.9500 82518.9500 199.68 <.0001 Period 1 156699.0187 156699.0187 379.19 <.0001 Temperature 1 5586.9453 5586.9453 13.52 0.0003 Mo isture*Period 1 6131.3155 6131.3155 14.84 0.0002 The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: NO3 NO3 Sum of Source DF Squares Mean Square F Value Pr > F Model 7 746665.373 106666.482 32.77 <.0001 Error 208 677062.137 3255.106 Corrected Tot al 215 1423727.510 R Square Coeff Var Root MSE NO3 Mean 0.524444 41.36232 57.05354 137.9360 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 27599.8311 27599.8311 8.48 0.0040 Period 1 646375.5199 646375.5199 198.57 <.0001 Temperature 1 510.3588 51 0.3588 0.16 0.6925 Moisture*Period 1 255.7087 255.7087 0.08 0.7795 Moisture*Temperature 1 64 287.6899 64287.6899 19.75 <.0001 Period*Temperature 1 1506.5502 1506.5502 0.46 0.4971 Moistu*Period*Temper 1 6129.7148 6129.7148 1.88 0.1715 Dependent V ariable: NO3 NO3 Sum of Source DF Squares Mean Square F Value Pr > F Model 6 740535.659 123422.610 37.76 <.0001 Error 209 683191.852 3268.861 Corrected Total 215 1423727.510 R Square Coeff Var Root MSE NO3 Mean 0.520139 41.44962 57.17395 137.9360 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 27599.8311 27599.8311 8.44 0.0041 Period 1 646375.5199 646375.5199 197.74 <.0001 Temperature 1 510.3588 510.3588 0.16 0.6932 Moisture*Period 1 255.7087 255.7087 0.08 0.7800 Moisture*Temperature 1 64287.6899 64287.6899 19.67 <.0001

PAGE 198

198 Period*Temperature 1 1506.5502 1506.5502 0.46 0.4980 Dependent Variable: NO3 NO3 Sum of Source DF Squares Mean Square F Value Pr > F Model 4 738773.400 1846 93.350 56.89 <.0001 Error 211 684954.110 3246.228 Corrected Total 215 142 3727.510 R S quare Coeff Var Root MSE NO3 Mean 0.518901 41.30587 56.97568 137.9360 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 39788.1244 39788.1244 12.26 0.0006 Period 1 646375.5199 646375.5199 199.12 <.0001 Temperature 1 52047.3065 52047.3065 16.03 <.0001 Moisture*Temperature 1 64287.6899 64287.6899 19.80 <.0001 The GLM Procedure Number of Observations Read 216 Number o f Observations Used 216 Dependent Variable: Mg Mg Sum of Source DF Squares Mean Square F Value Pr > F Model 7 43113.9441 6159.1349 20.71 <.0001 Error 208 61867.6150 297.4405 Corrected Total 215 104981.5591 R Square Coeff Var Root MSE Mg Mean 0.410681 25.33742 17.24646 68.06717 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 147.79143 147.79143 0.50 0.4817 Pe riod 1 40944.17542 40944.17542 137.66 <.0001 Temperature 1 1318.37570 1318.37570 4.43 0.0365 Moisture*Period 1 42.37133 42.37133 0.14 0.7062 Moisture*Temperature 1 49.10095 49.10095 0.17 0.6849 Period*Temperature 1 414.05598 414.05598 1.39 0.2394 Moistu*Period*Temper 1 198.07330 198.07330 0.67 0.4154 Dependent Variable: Mg Mg Sum of Source DF Squares Mean Square F Value Pr > F Model 6 42915.8708 7152.6451 24.09 <.0001 Error 209 62065.6883 296.9650 Corrected Total 215 104981.5591 R Square Coeff Var Root MSE Mg Mean 0.408794 25.31716 17.23267 68.06717 Source DF Type I SS Mean Square F Value Pr > F

PAGE 199

199 Moisture 1 147.79143 147.79143 0.50 0.4813 Period 1 40944.17542 40944.17542 137.88 <.0001 Temperature 1 1318.37570 1318.37570 4.44 0.0363 Moisture*Period 1 42.37133 42 .37133 0.14 0.7060 Moisture*Temperature 1 49.10095 49.10095 0.17 0.6847 Period*Temperature 1 4 14.05598 414.05598 1.39 0.2390 Dependent Variable: Mg Mg Sum of So urce DF Squares Mean Square F Value Pr > F Model 3 42410.3426 14136.7809 47.90 <.0001 Error 212 62571.2166 295.1472 Corrected Total 215 104981.5591 R Square Coeff Var Root MSE Mg Mean 0.403979 25.23955 17.17985 68.06717 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 147.79143 147.79143 0.50 0.4800 Period 1 40944.17542 40944.17542 138.72 <.0001 Temperature 1 1318.37570 1318.37570 4.47 0.0357 The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: K K Sum of Source DF Squares Mean Square F Value Pr > F Model 7 232723.7821 33246.2546 13.42 <.0001 Error 208 515 350.5578 2477.6469 Corrected Total 215 748074.3399 R Square Coeff Var Root MSE K Mean 0.311097 12.61950 49.77597 394.4368 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 803.6009 803.6009 0.32 0.5696 Period 1 33001.7488 33001.7488 13.32 0.0003 Temperature 1 180391.6127 180391.6127 72.81 <.0001 Moisture*Period 1 10530.4672 10530.4672 4.25 0.0405 Moisture*Temperature 1 106.7017 106.7017 0.04 0.8358 Period*Temperature 1 494.3652 494.3652 0.20 0.6556 Moistu*Period*Temper 1 7395.2857 7395.2857 2.98 0.0855 Dependent Variable: K K Sum of Source DF Squares Mean Square F Value Pr > F Model 6 225328.4964 37554.7494 15.01 <.0001 Error 209 522745.8435 2501.1763 Corrected Total 215 748074.3399 R Square Coeff Var Root MSE K Mean

PAGE 200

200 0.301211 12.67928 50.01176 394.4368 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 803.6009 803.6009 0.32 0.5714 Period 1 33001.7488 33001.7488 13.19 0.0004 Temperature 1 180391.6127 180391.6127 72.12 <.0001 Moisture*Period 1 10530.4672 10530.4672 4.21 0.0414 Mo isture*Temperature 1 106.7017 106.7017 0.04 0.8366 Period*Temperature 1 494.3652 494.3652 0.20 0.6571 Dependent Variable: K K Sum of Source DF Squares Mean Square F Value Pr > F Model 4 224727.4296 56181.8574 22.65 <.0001 Error 211 523346.9103 2480.3171 Corrected Total 215 748074.3399 R Square Coeff Var Root MSE K Mean 0.300408 12.62630 49.80278 394.4368 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 803.6009 803.6009 0.32 0.5698 Period 1 33001.7488 33001.7488 13.31 0.0003 Temperature 1 180391.6127 180391.6127 72.73 <.0001 Moisture*Period 1 10530.4672 10530.4672 4.25 0.0406 The GLM Procedure Number of Observations Read 216 Number of Observ ations Used 216 Dependent Variable: Fe Fe Sum of So urce DF Squares Mean Square F Value Pr > F Model 7 32383.33717 4626.19102 17.48 <.0001 Error 208 55042.04072 264.62520 Corrected Total 215 87425.37789 R Square Coeff Var Root MSE Fe Mean 0.370411 9.469961 16.26730 171.7779 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 5771.52855 5771.52855 21.81 <.0001 Period 1 14.45253 14.45253 0.05 0.8154 Temperature 1 18468.35837 18468.35837 69.79 <.0001 Moisture*Peri od 1 1049.12153 1049.12153 3.96 0.0478 Moisture*Temperature 1 6185.11206 6185.11206 23.37 <.0001 Period*Temperature 1 758.26269 758.26269 2.87 0.0920 Moistu*Period*Temper 1 136.50145 136.50145 0.52 0.4734 Dependent Variable: Fe Fe Sum of Source DF Squares Mean Square F Value Pr > F

PAGE 201

201 Model 6 32246.83572 5374.47262 20.36 <.0001 Error 209 55178.54217 264 .01216 Corrected Total 215 87425.37789 R Square Coeff Var Root MSE Fe Mean 0.3 68850 9.458985 16.24845 171.7779 So urce DF Type I SS Mean Square F Value Pr > F Moisture 1 5771.52855 5771.52855 21.86 <.0001 Period 1 14.45253 14.45253 0.05 0.8152 Temperature 1 18468.35837 18468.35837 69.95 <.0001 Moisture*Period 1 1049.12153 1049.12153 3.97 0.0475 Moisture*Temperature 1 6185.11206 6185.11206 23.43 <.0001 Period*Temperature 1 758.26269 758.26269 2.87 0.0916 Dependent Variable: Fe Fe Sum of Source DF Squares Mean Square F Value Pr > F Model 5 31488.57303 6297.71461 23.64 <.0001 Error 210 55936.80486 266.36574 Corrected Total 215 87425.37789 R Square Coeff Var Root MSE Fe Mean 0.360177 9.501054 16.32071 171.7779 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 5771.52855 5771.52855 21.67 <.0001 Period 1 14.45253 14 .45253 0.05 0.8160 Temperature 1 18468.35837 18468.35837 69.33 <.0001 Moisture*Period 1 10 49.12153 1049.12153 3.94 0.0485 Moisture*Temperature 1 6185.11206 6185.11206 23.22 <.0001 The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: Mn Mn Sum of Source DF Squares Mean Square F Value Pr > F Model 7 18161.76282 2594.53755 37.50 <.0001 Error 208 14390.60068 69.18558 Corrected Total 215 32552.36350 R Square Coeff Var Root MSE Mn Mean 0.557925 15.43393 8.317787 53.89287 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 175.24893 175.24893 2.53 0.1130 Period 1 5356.27765 5356.27765 77.42 <.0001 Temperature 1 11448.86482 11448.86482 165.48 <.0001 Moisture*Period 1 106.70997 106.70997 1.54 0.2157 Moisture*Temperature 1 385.46962 385.46962 5.57 0.0192

PAGE 202

202 Period*Temperature 1 647.42388 647 .42388 9.36 0.0025 Moistu*Period*Temper 1 41.76794 41.76794 0.60 0.4380 Dependent Variable: Mn Mn Sum of Source DF Squares Mean Square F Value Pr > F Mo del 6 18119.99489 3019.99915 43.73 <.0001 Error 209 14432.36862 69.05440 Corrected Total 215 32552.36350 R Square Coeff Var Root MSE Mn Mean 0.556641 15.41929 8.309897 53.89287 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 175.24893 175.24893 2.54 0.1127 Period 1 5356.27765 5356.27765 77.57 <.0001 Temperature 1 11448.86482 11448.86482 165.79 <.0001 Moisture*Period 1 106.70997 106.70997 1.55 0.2152 Moisture*Temp erature 1 385.46962 385.46962 5.58 0.0191 Period*Temperature 1 647.42388 647.42388 9.38 0.0025 Dependent Variable: Mn Mn Sum of Source DF Squares Mean Square F Value Pr > F Model 5 18013.28491 3602.65698 52.04 <.0001 Error 210 14539.07859 69.23371 Corrected Total 215 32552.36350 R Square Coeff Var Root MSE Mn Mean 0.553363 15.4393 0 8.320680 53.89287 Source DF Type I SS Mean Square F Value Pr > F Mo isture 1 175.24893 175.24893 2.53 0.1131 Period 1 5356.27765 5356.27765 77.37 <.0001 Temperature 1 11448.86482 11448.86482 165.37 <.0001 Moisture*Temperature 1 385.46962 385.46962 5.57 0.0192 Period*Temperature 1 647.42388 647.42388 9.35 0.0025 The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: Zn Zn Sum of Source DF Squares Mean Square F Value Pr > F Model 7 120.8635977 17.2662282 34.07 <.0001 Error 208 105.3983368 0.5067228 Corrected Total 215 226.2619345 R Square Coeff Var Root MSE Zn Mean 0.534176 19.77035 0.711845 3.600567 Source DF Type I SS Mean Square F Value Pr > F

PAGE 203

203 Moisture 1 0. 00862878 0.00862878 0.02 0.8963 Period 1 90.57580624 90.57580624 178.75 <.0001 Temperature 1 23.44641073 23.44641073 46.27 <.0001 Moisture*Period 1 5.04079205 5.04079205 9.95 0.0018 Mo isture*Temperature 1 0.13988819 0.13988819 0.28 0.5999 Period*Temperature 1 1.29992540 1.29992540 2.57 0.1107 Moistu*Period*Temper 1 0.35214631 0.35214631 0.69 0.4054 Dependent Variable: Zn Zn Sum of Source DF Squares Mean Square F Value Pr > F Model 6 120.5114514 20.0852419 39.70 <.0001 Error 209 105.7504831 0.5059832 Corrected Total 215 226.2619345 R Square C oeff Var Root MSE Zn Mean 0.532619 19.75592 0.711325 3.600567 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 0.00862878 0.00862878 0.02 0.8962 Period 1 90.57580624 90.57580624 179.01 <.0001 Temperature 1 23.44641073 23.44641073 46.34 <.0001 Moisture*Period 1 5.04079205 5.04079205 9.96 0.0018 Moisture*Temperature 1 0.13988819 0.13988819 0.28 0.5996 Period*Temperature 1 1.29992540 1.29992540 2.57 0.1105 Dependent Variable: Zn Zn Sum of Source DF Squares Mean Square F Value Pr > F Model 4 119.0716378 29.7679094 58.60 <.0001 Error 211 107.1902967 0.5080109 Co rrected Total 215 226.2619345 R Square Coeff Var Root MSE Zn Mean 0.526256 19.79546 0.712749 3.600567 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 0.00862878 0.00862878 0.02 0.8964 Period 1 90.57580624 90.57580624 178.30 <.0001 Temperature 1 23.44641073 23.44641073 46.15 <.0001 Moisture*Period 1 5.04079205 5.04079205 9.92 0.0019 The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: Cu Cu Sum of Source DF Squares Mean Square F Value Pr > F Model 7 22.72708882 3.24672697 36.38 <.0001 Error 208 18.56494385 0.08 925454 Corrected Total 215 41.29203267

PAGE 204

204 R Square Coeff Var Root MSE Cu Mean 0.5 50399 12.60532 0.298755 2.370071 So urce DF Type I SS Mean Square F Value Pr > F Moisture 1 0.02681326 0.02681326 0.30 0.5842 Period 1 4.41933481 4.41933481 49.51 <.0001 Temperature 1 17.90643658 17.90643658 200.62 <.0001 Moisture*Period 1 0.17778274 0.17778274 1.99 0.1596 Moisture*Temperature 1 0.13881834 0.13881834 1.56 0.2138 Period*Temperature 1 0.00088223 0.00088223 0.01 0.9209 Moistu*Period*Temper 1 0.05702086 0.05702086 0.64 0.4250 Dependent Variable: Cu Cu Sum of Source DF Squares Mean Square F Value Pr > F Model 6 22.67006795 3.77834466 42.41 <.0001 Error 209 18.62196471 0.08910031 Corrected Total 215 41.29203267 R Square Coeff Var Root MSE Cu Mean 0.549018 12.59442 0.298497 2.370071 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 0.02681326 0.02 681326 0.30 0.5839 Period 1 4.41933481 4.41933481 49.60 <.0001 Temperature 1 17. 90643658 17.90643658 200.97 <.0001 Moisture*Period 1 0.17778274 0.17778274 2.00 0.1593 Moisture*Temperature 1 0.13881834 0.13881834 1.56 0.2134 Period*Temperature 1 0.00088223 0.00088223 0.01 0.9208 Dependent Variable: Cu Cu Sum of Source DF Squares Mean Square F Value Pr > F Model 3 22.35258464 7.45086155 83.40 <.0001 Error 212 18.93944803 0.08933702 Corrected Total 215 41.29203267 R Square Coeff Var Root MSE Cu Mean 0.541329 12.61114 0.29 8893 2.370071 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 0.02681326 0.02681326 0.30 0.5844 Period 1 4.41933481 4.41933481 49.47 <.0001 Temperature 1 17.90643658 17.90643658 200.44 <.0001 The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: EC EC Sum of

PAGE 205

205 Source DF Squares Mean Square F Value Pr > F Model 7 4571017.69 653002.53 18.79 <.0001 Error 208 7227531.94 34747.75 Co rrected Total 215 11798549.63 R Square Coeff Var Root MSE EC Mean 0.387422 17.65571 186.4075 1055.792 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 19336.360 19336.360 0.56 0.4565 Period 1 4061714.949 4061714.949 116.89 <.0001 Temperature 1 107900.023 107900.023 3.11 0.0795 Moisture*Period 1 34147.301 34147.301 0.98 0.3227 Moisture*Temperature 1 286682.223 286682.223 8.25 0.0045 Period*Temperature 1 24052.440 24052.440 0.69 0.4064 Moistu*Period *Temper 1 37184.391 37184.391 1.07 0.3021 Dependent Variable: EC EC Sum of Source DF Squares Mean Square F Value Pr > F Model 6 4533833.30 755638.88 21.74 <.0001 Error 209 7264716.33 34759.41 Corrected Total 215 11798549.63 R Square Coeff Var Root MSE EC Mean 0.384270 17.65867 186.4388 1055.792 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 19336.360 19336.360 0.56 0.4566 Pe riod 1 4061714.949 4061714.949 116.85 <.0001 Temperature 1 107900.023 107900.023 3.10 0.0796 Moisture*Period 1 34147.301 34147.301 0.98 0.3228 Moisture*Temperature 1 286682.223 286682.223 8.25 0.0045 Period*Temperature 1 24052.440 24052.440 0.69 0.4064 Dependent Variable: EC EC Sum of Source DF Squares Mean Square F Value Pr > F Model 4 4475633.56 1118908.39 32.24 <.0001 Error 211 7322916.07 34705.76 Corrected Total 215 11798549.63 R Square Coeff Var Root MSE EC Mean 0.379338 17.64504 186.2948 1055.792 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 305914.342 305914.342 8.81 0.0033 Period 1 4061714.949 4061714.949 117.03 <.0001 Temperature 1 381998.978 381998.978 11.01 0.0011 Moisture*Temperature 1 286682.223 286682.223 8.26 0.0045

PAGE 206

206 The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: pH pH Sum of Source DF Squares Mean Square F Value Pr > F Model 7 1.54419845 0.22059978 9.54 <.0001 Error 208 4.80867887 0.02311865 Corrected Total 215 6.35287731 R Square Coeff Var Root MSE pH Mean 0.243071 2.003844 0.152048 7.587 824 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 0.13799103 0.13799103 5.97 0.0154 Period 1 0.01659215 0.01659215 0.72 0.3979 Temperature 1 1.33492676 1.33492676 57.74 <.0001 Moisture*Period 1 0.03874446 0.03874446 1.68 0.1969 Moisture*Temperature 1 0.00000387 0.00000387 0.00 0.9897 Period*Temperature 1 0.00006768 0.00006768 0.00 0.9569 Moistu*Period*Temper 1 0.01587250 0.01587250 0.69 0.4083 Dependent Variable: pH pH Sum of Source DF Squares Mean Square F Value Pr > F Model 6 1. 52832595 0.25472099 11.03 <.0001 Error 209 4.82455136 0.02308398 Corrected Total 215 6.35287731 R Square Coeff Var Root MSE pH Mean 0.240572 2.002341 0.151934 7.587824 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 0.13799103 0.13799103 5.98 0.0153 Period 1 0.01659215 0.01659215 0.72 0.3975 Temperature 1 1.33492676 1.33492676 57.83 <.0001 Moisture*Period 1 0.03874446 0.03874446 1.68 0.1966 Moisture*Temperature 1 0.00000387 0.00000387 0.00 0.9897 Period*Temperature 1 0.00006768 0.00006768 0.00 0.9569 Dependent Variable: pH pH Sum of Source DF Squares Mean Square F Value Pr > F Model 3 1.48950994 0.49650331 21.64 <.0001 Error 212 4.86336738 0.02294041 Corrected Total 215 6.35287731 R Square Coeff Var Root MSE pH Mean 0.234462 1.996105 0.151461 7.587824 Source DF Type I SS Mean Square F Value Pr > F Moisture 1 0.13799103 0.13799103 6.02 0.0150 Period 1 0. 01659215 0.01659215 0.72 0.3960 Temperature 1 1.33492676 1.33492676 58.19 <.0001

PAGE 207

207 APPENDIX H RESULTS OF LINEAR REGRESSION ANALYSIS ON ASH pH, EC, TOTAL AND WATER SOLUBLE NUTRIENTS IN LABORATORY RESIDUAL ASHES The GLM Procedure Number of Observations Read 12 Number of Observations Used 12 Dependent Variable: pH pH Sum of Source DF Squares Mean Square F Value Pr > F Model 1 7.95704167 7.95704167 10.82 0.0082 Error 10 7.35205000 0.73520500 Corrected Total 11 15.30909167 R Square Coeff Var Root MSE pH Mean 0.519759 11.70435 0.857441 7.325833 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 7.95704167 7.95704167 10.82 0.0082 Standard Parameter Estimate Error t Value Pr > |t| Intercept 4.412500000 0.91950299 4.80 0.0007 Temperature 0.007283333 0.00221390 3.29 0.0082 Dependent Variable: EC EC Sum of Source DF Squares Mean Square F Value Pr > F Model 1 1727546.017 1727546.017 107.20 <.0001 Error 10 161144.233 16114.423 Corrected Total 11 1888690.250 R Square Coeff Var Root MSE EC Mean 0.914679 15.06290 126. 9426 842.7500 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 1727546.017 1727546.017 107.20 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 514.7166667 136.1307711 3.78 0.0036 Temperature 3.3936667 0.3277644 10.35 <.0001 Dependent Variable: Ca Ca Sum of Source DF Squares Mean Square F Value Pr > F Model 1 271458.100 271458.100 2.51 0.1445 Error 10 108 3216.982 108321.698 Corrected Total 11 1354675.082 R Square Coeff Var Root MSE Ca Mean 0.200386 101.6885 329.1226 323.6576

PAGE 208

208 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 271458.0996 271458.0996 2.51 0.1445 Standard Parameter Estimate Error t Value Pr > |t| Intercept 861.7610600 352.9446882 2.44 0.0348 Temperature 1.3452586 0.8497 909 1.58 0.1445 Dependent Variable: K K Sum of Source DF Squares Mean Square F Value Pr > F Model 1 442760.5605 442760.5605 85.10 <.0001 Error 10 52029.7890 5202.9789 Corrected Total 11 494790.3495 R Square Coeff Var Root MSE K Mean 0.894845 9.611067 72.13168 750.5064 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 442760.5605 44276 0.5605 85.10 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 63.28162428 77.35260654 0.82 0.4324 Temperature 1.71806201 0.18624319 9.22 <.0001 Dependent Variable: Mg Mg Sum of Source DF Squares Mean Square F Value Pr > F Model 1 2927.16979 2927.16979 0.85 0.3769 Error 10 34239.42369 3423.94237 Corrected Total 11 37166.59348 R Square Coeff Var Root MSE Mg Mean 0.078758 52.82394 58.51446 110.7726 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 2927.169791 2927.169791 0.85 0.3769 Standard Parameter Estimate Error t Value Pr > |t| Intercept 166.6502868 62.74977071 2.66 0.0241 Temperature 0.1396942 0.15108369 0.92 0.3769 Dependent Variable: Mn Mn Sum of Source DF Squares Mean Square F Value Pr > F Model 1 201 9.922713 2019.922713 33.03 0.0002 Error 10 611.602143 61.160214 Corrected Total 11 2631.524857 R Square Coeff Var Root MSE Mn Mean 0.767586 57.96844 7.820500 13.49096

PAGE 209

209 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 2019.922713 2019.922713 33.03 0.0002 Standard Parameter Estimate Error t Value Pr > |t| Intercept 59.90845976 8.38655 152 7.14 <.0001 Temperature 0.11604375 0.02019244 5.75 0.0002 Dependent Variable: Zn Zn Sum of Source DF Squares Mean Square F Value Pr > F Model 1 3.88962675 3.88962675 29.73 0.0003 Error 10 1.30810550 0.13081055 Corrected Total 11 5.19773225 R Square Coeff Var Root MSE Zn Mean 0.748331 55.53474 0.361677 0.651263 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 3.88962675 3.88962675 29.73 0.0003 Standard Parameter Estimate Error t Value Pr > |t| Intercept 2.688156930 0.38785581 6.93 <.0001 Temperature 0.005092234 0.00093385 5.45 0.0003 Dependent Variable: NH4 NH4 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 299.7569696 299.7569696 123.34 <.0001 Error 10 24.3024330 2.4302433 Corrected Total 11 324.0594026 R Square Coeff Var Root MSE NH4 Mean 0.925006 22.25270 1.558924 7.005548 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 299.7569696 299.7569696 123.34 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 24.88684491 1.67175949 14.89 <.0001 Temperature 0.04470324 0.00402512 11.11 <.000 1 Dependent Variable: NO3 NO3 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 7.20895014 7.20895014 14.79 0.0032 Error 10 4.87267081 0.48726708 Corrected Total 11 12.08162095 R Square Coeff Var Root MSE NO3 Mean 0.596687 83.30160 0.698045 0.837973

PAGE 210

210 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 7.20895014 7.20895014 14.79 0.0032 Standard Parameter Estimate Error t Value Pr > |t| Intercept 3.610976511 0.74857007 4.82 0.0007 Temperature 0.0069 32508 0.00180234 3.85 0.0032 Dependent Variable: Ex t_P Ext P Sum of Source DF Squares Mean Square F Value Pr > F Model 1 5746.024625 5746.024625 43.81 <.0001 Error 10 1311.697690 131.169769 Corrected Total 11 7057.722315 R Square Coeff Var Root MSE Ext_P Mean 0.814147 56.76776 11.45294 20.17507 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 574 6.024625 5746.024625 43.81 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 98.46359568 12.28190679 8.02 <.0001 Temperature 0.19572131 0.02957136 6.62 <.0001 Dependent Variable: Biomass_____ Biomass (%) Sum of Source DF Squares Mean Square F Value Pr > F Model 1 2231.136240 2231.136240 53.63 <.0001 Error 10 416.038427 41.603843 Corrected Total 11 2647.174667 R Square Coeff Var Root MSE Biomass_____ Mean 0.842837 8.001949 6.450104 80.60667 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 2231.136240 2231.136240 53.63 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 31.82266667 6.91696603 4.60 0.001 0 Temperature 0.12196000 0.01665410 7.32 <.0001 Dependent Variable: TN____ TN (%) Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.51695640 0.51695640 112.06 <.0001 Error 10 0.04613026 0.00461303 Co rrected Total 11 0.56308667 R Square Coeff Var Root MSE TN____ Mean 0.918076 23.64038 0.067919 0.287302

PAGE 211

211 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 0.51695640 0.51695640 112.06 <.0001 Standard Parameter Estimate Er ror t Value Pr > |t| Intercept 1.0298 78613 0.07283529 14.14 <.0001 Temperature 0.001856442 0.00017537 10.59 <.0001 Dependent Variable: TC____ TC (%) Sum of Source DF Squares Mean Square F Value Pr > F Model 1 1171.505481 1171.505481 40.02 <.0001 Error 10 292.737411 29.273741 Corrected Total 11 1464.242892 R Square Coeff Var Root MSE TC____ Mean 0.800076 50.14524 5.410521 10.78970 Source DF T ype I SS Mean Square F Value Pr > F Temperature 1 1171.505481 1171.505481 40.02 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 46.13946621 5.80213773 7.95 <.0001 Temperature 0.08837441 0.01396991 6.33 <.0001 Dependent Variable: TP TP Sum of Source DF Squares Mean Square F Value Pr > F Model 1 23.86685765 23.86685765 4.01 0.0732 Error 10 59.56438879 5.95643888 Corrected Total 11 83.43124644 R Square Coeff Var Root MSE TP Mean 0.286066 1.404285 2.440582 173.7954 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 23.86685765 23.86685765 4.01 0.0732 Standard Parameter Estimate Error t Value Pr > |t| Intercept 168.7497940 2.61723226 64.48 <.0001 Temperature 0.0126140 0.00630155 2.00 0.0732 Dependent Variable: TCa TCa Sum of Source DF Squares Mean Square F Value Pr > F Model 1 3049.260 3049.260 0.01 0.9268 Er ror 10 3437782.296 343778.230 Corrected Total 11 3440831.556 R Square Coeff Var Root MSE TCa Mean 0.000886 4.975485 586.3260 11784.30

PAGE 212

212 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 3049.260167 3049.260167 0.01 0.9268 Stand ard Parameter Estimate Error t Value Pr > |t| Intercept 11727.26890 628.7646332 18.65 <.0001 Tempe rature 0.14258 1.5138873 0.09 0.9268 Dependent Variable: TFe TFe Sum of Source DF Squares Mean Square F Value Pr > F Model 1 642.654225 642.654225 1.78 0.2114 Error 10 3604.009713 360.400971 Corrected Total 11 4246.663939 R Square Coeff Var Root MSE TFe Mean 0.151332 6.941913 18.98423 273.4726 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 642.6542254 642.6542254 1.78 0.2114 Standard Parameter Estimate Error t Value Pr > |t| Intercept 247.2905784 20.35831813 12.15 <.0001 Temperature 0.0654550 0.04901707 1.34 0.2114 Dependent Variable: TK TK Sum of Source DF Squares Mean Square F Value Pr > F Model 1 613953.7554 613953.7554 38.74 <.0001 Error 10 158498.2924 15849.8292 Corrected Total 11 772452.0478 R Square Coeff Var Root MSE TK Mean 0.794811 11.57171 125.8961 1087.964 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 613953.7554 613953.7554 38.74 <.0001 Standard Parameter Estimate Error t Value Pr > |t | Intercept 1897.213207 135.0085317 14.05 <.0001 Temperature 2.023123 0.3250623 6.22 <.0001 Dependent Variable: TMg TMg Sum of Source DF Squares Mean Square F Value Pr > F Mo del 1 2255.62677 2255.62677 2.80 0.1253 Error 10 8059.04145 805.90414 Corrected Total 11 10314.66822 R Square Coeff Var Root MSE TMg Mean 0.218681 3.279768 28.38845 865.5627

PAGE 213

213 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 2255.626766 2255.626766 2.80 0.1253 Standard Parameter Est imate Error t Value Pr > |t| Inter cept 914.6136899 30.44322202 30.04 <.0001 Temperature 0.1226275 0.07329867 1.67 0.1253 Dependent Variable: TMn TMn Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.2208524 0.2208524 0.02 0.8966 Error 10 124.3686407 12.4368641 Corrected Total 11 124.5894931 R Square Coeff Var Root MSE TMn Mean 0.001773 7.99517 5 3.526594 44.10903 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 0.22085242 0.22085242 0.02 0.8966 Standard Parameter Estimate Error t Value Pr > |t| Intercept 44.59439038 3.78185056 11.79 <.0001 Temperature 0.00121340 0.00910563 0.13 0.8966 Dependent Variable: TZn TZn Sum of Source DF Squares Mean Square F Value Pr > F Model 1 2.96681365 2.96681365 3.49 0.0913 Error 10 8.49959201 0.84995920 Corrected Total 11 11.46640567 R Square Coeff Var Root MSE TZn Mean 0.258740 12.36420 0.921932 7.456465 Source DF Type I SS Mean Square F Value Pr > F Temperature 1 2.96681365 2.96681365 3.49 0.0913 Standard Parameter Estimate Error t Value Pr > |t| Intercept 5.677532034 0.98866227 5.74 0.0002 Temperature 0.004447332 0.00238042 1.87 0.0913

PAGE 214

214 APPENDIX I RESULTS OF FITTING REGRESSION MODELS AND ANOVA ANALYSIS ON TP, TCa, TMg, AND TFe IN RESIDUAL ASH ES Table I 1. Results of ANOVA a nalysis on TP/TCa, TP/TMg, TP/TFe in residual ashes Variables 65 0 C 250 0 C 350 0 C 450 0 C 550 0 C Field ash Pr > F Sig. level TP/TCa 0.0165 a 0.0149 a 0.0142 a 0.0147 a 0.0168 a 0.0197 a 0.3680 NS TP/TMg 0.1863 a 0.1885 a 0.2028 a 0.2087 a 0.2177 a 0.1889 a 0.0695 NS TP/TFe 1.2194 a 0.6273 a 0.6849 a 0.5616 a 0.5457 a 0.9039 a 0.1105 NS Results of Fitting of Regression Models with 1 st 2 nd and 3 rd order among Burn Temperature (Response Variable) and Content of TP, TCa, TFe, or TMg in Laboratory Simulation Ashes (Explanatory Variable) Linear regression model for TP with The REG Procedure Dependent Variable: Temperature Temperature Number of Observations Read 12 Number of Observations Used 12 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 1 147781 147781 666.03 <.0001 Error 10 2218.85073 221.88507 Corrected Total 11 150000 Root MSE 14.89581 R Square 0.9852 Dependent Mean 400.00000 Adj R Sq 0.9837 Coeff Var 3.72395 Parameter Estimates Parameter Standard Variabl e Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 222.54987 8.10979 27.44 < .0001 TP TP 1 0.12254 0.00475 25.81 <.0001 Quadratic regression model for TP The REG Procedure Dependent Variable: Temperature Temperature Number of O bservations Read 12 Number of Observations Used 12 Analysis of Variance Sum of Mean

PAGE 215

215 Source DF Squares Square F Value Pr > F Model 2 147807 73903 303.28 <.0001 Error 9 2193.13084 243.68120 Corrected Total 11 150000 Root MSE 15.61029 R Square 0.9854 Dependent Mean 400.00000 Adj R Sq 0.9821 Coeff Var 3.90257 Parameter Estimates Parameter Standard Variable L abel DF Estimate Error t Value Pr > |t| Intercept Intercept 1 218.90934 14.06407 15.57 <.0001 TP TP 1 0.13063 0.02540 5.14 0.0006 TP_2 2nd power of TP 1 0.00000277 0.00000852 0.32 0.7527 Cubic regression model for T P The REG Procedure Dependent Variable: Temperature Temperature Number of Observations Read 12 Number of Observations Used 12 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 3 149560 49853 905.58 <.0001 Error 8 440.41061 55.05133 Corrected Total 11 150000 Root MSE 7.41966 R Square 0.9971 Dependent Mean 400.00000 Adj R Sq 0.9960 Coeff Var 1.85491 Parameter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 171.26427 10.76967 15.90 <.0001 TP TP 1 0.30356 0.03294 9.22 <.0001 TP_2 2nd power of TP 1 0.00014573 0.00002566 5.68 0.0005 TP_3 3rd power of TP 1 3.236256E 8 5.735489E 9 5.64 0.0005 Linear regression model for TCa The REG Procedure Dependent Variable: Temperature Temperatur e Number of Observations Read 12 Number of Observations Used 12 Analysis of Variance Sum of Mean

PAGE 216

216 Source DF Squares Square F Value Pr > F Model 1 142892 142892 201.04 <.0001 Error 10 7107.50412 710.75041 Corrected Total 11 150000 Root MSE 26.65990 R Square 0.9526 Dependent Mean 400.00000 Adj R Sq 0.9479 Coeff Var 6.66498 Parameter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 214.03657 15.20667 14.08 <.0001 TCa TCa 1 0.00199 0.000 14024 14.18 <.0001 Quadratic regression model for TCa The REG Procedure Dependent Variable: Temperature Temperature Number of Observations Read 12 Number of Observations Used 12 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 2 142898 71449 90.54 <.0001 Error 9 7102.43012 789.15890 Corrected Total 11 150000 Root MSE 28.09197 R Square 0.9527 Dependent Mean 400.00000 Adj R Sq 0.9421 Coeff Var 7.02299 Parameter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 215.83354 27.54952 7.83 <.0001 TCa TCa 1 0.00192 0.00081046 2.37 0.0416 TCa_2 2nd pow er of TCA 1 3.55409E 10 4.432363E 9 0.08 0.9378 Cubic regression model for TCa The REG Procedure Dependent Variable: Temperature Temperature Number of Observations Read 12 Number of Observations Used 12 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 3 144007 48002 64.08 <.0001

PAGE 217

217 Error 8 5993.02001 749.12750 Corrected Total 11 150000 Root MSE 27.37019 R Square 0.9600 Dependent Mean 400.00000 Adj R Sq 0.9451 Coeff Var 6.84255 Parame ter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Inter cept 1 177.39808 41.44888 4.28 0.0027 TCa TCa 1 0.00419 0.00202 2.07 0.0721 TCa_2 2nd power of TCA 1 3.09069E 8 2.604977E 8 1.19 0.2695 TCa_3 3rd power of TCA 1 1.18099E 13 9.70459E 14 1.22 0.2583 Linear regression model for TFe The REG Procedure Dependent Variable: Temperature Temperature Number of Observations Read 12 Number of Observations Used 12 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 1 144343 144343 255.18 <.0001 Error 10 5656.50831 565.6 5083 Corrected Total 11 150000 Root MSE 23.78342 R Square 0.9623 Depe ndent Mean 400.00000 Adj R Sq 0.9585 Coeff Var 5.94585 Parameter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 240.21913 12.13193 19.80 <.0001 TFe TFe 1 0.06300 0.00394 15.97 <.0001 Quadratic regression model for TFe The REG Procedure Dependent Variable: Temperature Temperature Number of Observations Read 12 Number of Observations Used 12 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 2 145100 72550 133.26 <.0001

PAGE 218

218 Error 9 4899.80275 544.42253 Corrected Total 11 150000 Root MSE 23.33286 R Square 0.9673 Dependent Mean 400.00000 Adj R Sq 0.9601 Coeff Var 5.83322 Parameter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 223.27192 18.66269 11.96 <.0001 TFe TFe 1 0.08515 0.01918 4.44 0.0016 TFe_2 2nd power of TFE 1 0.00000415 0.00000352 1.18 0.2686 Cubic regression model for TFe The REG Procedure Dependent Variable: Temperature Temperature Number of Observations Read 12 Number of Observations Used 12 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 3 146935 48978 127.86 <.0001 Error 8 3064.62438 383.07805 Corrected Total 11 150000 Root MSE 19.57238 R Square 0.9796 Dependent Mean 400.00000 Adj R Sq 0.9719 Coeff Var 4.89309 Parame ter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 176.96912 26.31742 6.72 0.0001 TFe TFe 1 0.17762 0.04521 3.93 0.0044 TFe_2 2nd power of TFE 1 0.00004372 0.00001832 2.39 0.0441 TFe_3 3rd power of TFE 1 4.635843E 9 2.118035E 9 2.19 0.0600 Linear regression model for TMg The REG Procedure Dependent Variable: Temperature Temp erature Number of Observ ations Read 12 Number of Observations Used 12 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F

PAGE 219

219 Model 1 147118 147118 510.46 <.0001 Error 10 2882.06965 288.20697 Corrected Total 11 150000 Root MSE 16.97666 R Square 0.980 8 Dependent Mean 400.00000 Adj R Sq 0.9789 Coeff Var 4.24416 Parameter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 214.53110 9.56060 22.44 <.0001 TMg TMg 1 0.02696 0.00119 22.59 <.0001 Quadratic regression model for TMg The REG Procedure Dependent Variable: Temperature Te mperature Number of Obse rvations Read 12 Number of Observations Used 12 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 2 147122 73561 230.04 <.0001 Error 9 2878.00185 319.77798 Corrected Total 11 150000 Root MSE 17.88234 R Square 0.9 808 Dependent Mean 400.00000 Adj R Sq 0.9765 Coeff Var 4.47058 Parameter Estimates Parameter Standard Variable Label DF Estimate Error t Value Pr > |t| Intercept Intercept 1 216.18294 17.77404 12.16 <.0001 TMg T Mg 1 0.02620 0.00684 3.83 0.0040 TMg_2 2nd power of TMG 1 5.549943E 8 4.92077E 7 0.11 0.9127 Cubic regression model for TMg The REG Procedure Dependent Variable: Temperature Temperature Number of Observations Read 12 Number of Observations Used 12 Analysis of Variance Sum of Mean Source DF Squares Square F Value Pr > F Model 3 149332 49777 596.18 <. 0001 Error 8 667.95165 83.49396

PAGE 220

220 Corrected Total 11 150000 Root MSE 9.13750 R Square 0.9955 Dependent Mean 400.00000 Adj R Sq 0.9939 Coeff Var 2.28438 Parameter Estimates Parameter Standard Variable Label DF Esti mate Error t Value Pr > |t| Intercept Intercept 1 154.20021 15.08735 10.22 <.0001 TMg TMg 1 0.07195 0.00956 7.53 <.0001 TMg_2 2nd power of TMG 1 0.00000802 0.00000159 5.05 0.0010 TMg_3 3rd power of TMG 1 3.94955E 10 7.6767E 11 5.14 0.0009

PAGE 221

221 APPENDIX J EQUILIBRIUM SPECIATION MODEL Figure J 1.

PAGE 222

222 Figure J 2. Initial species input into the M speciation model a) fixed species, b) possibl e solid species, and c) main solution components a b c

PAGE 223

223 APPENDIX K CALCULATION OF STOICHIOMETRY FOR ACTIVITY OF HPO 4 2 FROM C a /M g /F e /M n P MINERALS Monetite CaHPO 4 + H + Ca 2+ + H 2 PO 4 K s = 10 0.30 H 2 PO 4 HPO 4 2 + H + K s = 10 7.2 CaHPO 4 Ca 2+ + HPO 4 2 K s = 10 6.9 K s = 10 6.9 = 6.9 = log(Ca 2+ ) + log(HPO 4 2 ) 6.9 = pCa 2+ + pHPO 4 2 pHPO 4 2 = 6.9 pCa 2+ Monocalcium P Ca(H 2 PO 4 ) 2 :H 2 O Ca 2+ + 2H 2 PO 4 + H 2 O K s = 10 1.15 2(H 2 PO 4 HPO 4 2 + H + ) K s = (10 7.2 ) 2 Ca(H 2 PO 4 ) 2 :H 2 O Ca 2+ + 2HPO 4 2 + 2H + + H 2 O K s = 10 15.55 K s = 10 15.55 = 15.55 = log(Ca 2+ ) + 2log(HPO 4 2 ) + 2log(H + ) 15.55 = pCa 2+ + 2pHPO 4 2 + 2pH pHPO 4 2 = 7.775 0.5pCa 2+ pH Dicalcium P CaHPO 4 :2H 2 O + H + Ca 2+ + H 2 PO 4 + 2H 2 O K s = 10 0.63 H 2 PO 4 HPO 4 2 + H + K s = 10 7.2 CaHPO 4 :2H 2 O Ca 2+ + HPO 4 + 2H 2 O K s = 10 6.57 K s = 10 6.57 = 6.57 = log(Ca 2+ ) + log(HPO 4 2 )

PAGE 224

224 6.57 = pCa 2+ + pHPO 4 2 pHPO 4 2 = 6.57 pCa 2+ Tricalcium P Ca 3 (PO 4 ) 2 (beta) + 4H + 3Ca 2+ + 2H 2 PO 4 K s = 10 10.18 2(H 2 PO 4 HPO 4 2 + H + ) K s = (10 7.2 ) 2 Ca 3 (PO 4 ) 2 (beta) + 2H + 3Ca 2+ + 2HPO 4 2 K s = 10 4.22 K s = 10 4.22 = 4.22 = 3log(Ca 2+ ) + 2log(HPO 4 2 ) 2log(H + ) 4.22 = 3pCa 2+ + 2pHPO 4 2 2pH pHPO 4 2 = 2.11 1.5pCa 2+ + pH Octacalcium P Ca 4 H(PO 4 ) 3 :2.5H 2 O + 5H + 4Ca 2+ + 3H 2 PO 4 + 2.5H 2 O K s = 10 11.76 3(H 2 PO 4 HPO 4 2 + H + ) K s = (10 7.2 ) 3 Ca 4 H(PO 4 ) 3 :2.5H 2 O + 2H + 4Ca 2+ + 3HPO 4 2 + 2.5H 2 O K s = 10 9.84 K s = 10 9.84 = 9.84 = 4log(Ca 2+ ) + 3log(HPO 4 2 ) 2log(H + ) 9.84 = 4pCa 2+ + 3pHPO 4 2 2pH pHPO 4 2 = 3.28 4/3pCa 2+ + 2/3pH Hydroxyapatite Ca 5 (PO 4 ) 3 (OH) + 7H + 5Ca 2+ + 3H 2 PO 4 + H 2 O K s = 10 14.46 3(H 2 PO 4 HPO 4 2 + H + ) K s = (10 7.2 ) 3 Ca 5 (PO 4 ) 3 (OH) + 4H + 5Ca 2+ + 3HPO 4 2 + H 2 O K s = 10 7.14 K s = 10 7.14 = 7.14 = 5log(Ca 2+ ) + 3log(HPO 4 2 ) 4log(H + )

PAGE 225

225 7.14 = 5pCa 2+ + 3pHPO 4 2 4pH pHPO 4 2 = 2.38 5/3pCa 2+ + 4/3pH Vivianite Fe 3 (PO 4 ) 2 :8H 2 O 3Fe 2+ + 2PO 4 3 + 8H 2 O K s = 10 37.76 2( H + + PO 4 3 ) 2 HPO 4 2 K s = (10 12.35 ) 2 Fe 3 (PO 4 ) 2 :8H 2 O + 2H + 3Fe 2+ + 2HPO 4 2 + 8H 2 O K s = 10 13.06 K s = 10 13.06 = 13.06 = 3log(Fe 2+ ) + 2log(HPO 4 2 ) 2log(H + ) 13.06 = 3pFe 2+ + 2pHPO 4 2 2pH pHPO 4 2 = 6.53 1.5pFe 2+ + pH Mg 3 (PO 4 ) 2 Mg 3 (PO 4 ) 2 3Mg 2+ + 2PO 4 3 K s = 10 23.28 2 (H + + PO 4 3 HPO 4 2 ) K s = (10 12.35 ) 2 Mg 3 (PO 4 ) 2 + 2H + 3Mg 2+ + 2HPO 4 2 K s = 10 1.42 K s = 10 1.42 = 1.42 = 3log(Mg 2+ ) + 2log(HPO 4 2 ) 2log(H + ) 1.42 = 3pMg 2+ + 2pHPO 4 2 2pH pHPO 4 2 = 0.71 1.5pMg 2+ + pH MgHPO 4 :3H 2 O MgHPO 4 :3H 2 O Mg 2+ + PO 4 3 + H + + 3H 2 O K s = 10 18.175 H + + PO 4 3 HPO 4 2 K s = 10 12.35 MgHPO 4 :3H 2 O Mg 2+ + HPO 4 + 3H 2 O K s = 10 5.825 K s = 10 5.825 = 5.825 = log(Mg 2+ ) + log(HPO 4 2 ) 5.825 = pMg 2+ + pHPO 4 2

PAGE 226

226 pHPO 4 2 = 5.825 pMg 2+ Mn 3 (PO 4 ) 2 Mn 3 (PO 4 ) 2 3Mn 2+ + 2PO 4 3 K s = 10 23.827 2 (H + + PO 4 3 HPO 4 2 ) K s = (10 12.35 ) 2 Mn 3 (PO 4 ) 2 + 2H + 3Mn 2+ + 2HPO 4 2 K s = 10 0.873 K s = 10 0.873 = 0.873 = 3log(Mn 2+ ) + 2log(HPO 4 2 ) 2log(H + ) 0.873 = 3pMn 2+ + 2pHPO 4 2 2pH pHPO 4 2 = 0.4365 1.5pMn 2+ + pH MnHPO 4 MnHPO 4 Mn 2+ + PO 4 3 + H + K s = 10 25.4 H + + PO 4 3 HPO 4 2 K s = 10 12.35 MnHPO 4 Mn 2+ + HPO 4 K s = 10 13.05 K s = 10 13.05 = 13.05 = log(Mn 2+ ) + log(HPO 4 2 ) 13.05 = pMn 2+ + pHPO 4 2 pHPO 4 2 = 13.05 pMn 2+

PAGE 227

227 APPENDIX L EXPERIMENT FOR DETERMINATION OF SOIL WATER CHARACTERISTIC CURVE Figure L 1. Establishment of the experiment: (a) Richard Plates and (b) Tempe Cells Procedures for the Experiment The 1 bar, 1400/1405 100 Tempe pressure cells, core cylinders, and 1 bar ceramic plates were utilized to determin e volumetric water amount with pressures less than 1 bar (1020 cm H 2 O). The volume of the core cylinder was appro ximately 415.7 cm 3 with height of 53 mm and diameter of 100 mm. The height of the selected cylinder corresponded to height of soil in the field. The bottom of the cylinder was secured by mesh cloth and rubber band to prevent loss of soil. Air dried soils w ere repacked into cores of cylinder s using a layering approach to prevent uneven distribution of soil bulk density. Sample preparation: Cores with soils were put into a container and soaked with a mixed a b

PAGE 228

228 solution of 0.005M Calcium Sulfate Di hydrate (CaSO 4 :2H 2 O) and Thymol (0.25 g Thymol/L). This solution aimed to reduce biological activities in soils. Since soils contained a high organic content, cores with soils were soaked for one week. Soil was slowly wet by placing a shallow solution, increased gradua lly height of the mixed solution in the container in every day, eventually soaked with the mixed solution just below the top of the core in last day. A gradual soakage avoided to immerse suddenly soils which could cause a compressed gas and disrupt soil st ructure. Tempe cell setup: one day before setting up Tempe cell. While submerged in the mixed solution, a ceramic plate was pla ced in bottom piece of Tempe cell, and then the core was place d on the ceramic plate. After the nserted into the cell, the core was placed in upside down position on the plate so that the mesh cloth w as on the top and removed from the core. There should be no bubbles visible on the bottom of the Tempe cell. If any bubble was present in the bottom of the Tempe cell, process of Tempe cell setup was repeated from the first step. Tempe cells were put on a Tempe cell stand with graduate cylinders placed underneath (Figure 4 3b). The top of Tempe cell s were covered by wet paper towels to prevent evaporative loss of water while soils in cores wer e drained to reach equilibrium, a nd graduated cylinders were also covered by parafilm to prevent evaporation. After water was no long er draining from soils the cores were 3b). P ressure s applied for Tempe cell method w ere 0, 0.49, 0.98, 1.47, 2.45, 3.43, 4.90, 6.67, 14.71, 29.41, 58.82, 98.04 kPa The experiment started with the lowest pressure and increased gr adually with higher pressures. T otal amount of water at each pressure was determined when water was no longer fallen from Tempe cell s and measured by a water manometer. G raduated cylinder s w ere empty after water volume s of a given pressure were recorded. S oils removed after

PAGE 229

229 the last pressure were used for Richard plate method with higher pressures. Fifteen bar Richard plates were moistened at least one day before the experiment of Richard plate started. Twelve rings with a h eight of 0.5 cm and diameter of 3 cm were placed togeth er on each plate, and then put above soils into the rings (Figure 4 3c). Three Richard plates were installed simultaneously into a compressor with a secured seal. The plates were connected to outside b y tubes from which water amount could be measured at each pressure applied. Pressures used for Richard plate method were 500, and 800 kPa ( 5 and 8 bars respectively) After removing from the last pressure, soils were dried in Oven at 105 0 C for 24 hours. W eight of so ils before and after drying was also recorded to measure water amount remained in soil s after applying the 8 bar pressure

PAGE 230

230 APPENDIX M VOLUMETRIC WATER CONTENTS OF PINE ROCKLAND SOIL OBTAINED FROM OBSERVATION EXPERIMENT AND FROM RETENTION CURVE MODEL Table M 1. Results of volumetric water content fitted by Retc model Suction (pressure) Volumetric water content from observation Volumetric water content fitted by Retention Curve model (RETC) ------------------------------------------------------------------------------------------------------------------------------------------------cm H 2 O kPa R1 R2 R3 R4 R1 R2 R3 R4 .. 5 0.49 0.6177 0.6160 0.6425 0.6261 0.6167 0.6202 0.6425 0.6308 10 0.98 0.6160 0.6160 0.6420 0.6254 0.6126 0.6114 0.6415 0.6198 15 1.4 7 0.6153 0.6109 0.6420 0.6165 0.6081 0.6025 0.6403 0.6088 25 2.4 5 0.6032 0.5943 0.6386 0.5984 0.5988 0.5853 0.6374 0.5883 35 3.4 3 0.5893 0.5753 0.6348 0.5756 0.5894 0.5695 0.6342 0.5702 50 4. 90 0.5686 0.5472 0.6249 0.5453 0.5758 0.5488 0.6289 0.5472 68 6.6 7 0.5474 0.5217 0.6153 0.5171 0.5606 0.5278 0.6221 0.5245 150 14. 71 0.4998 0.4604 0.5965 0.4536 0.5067 0.4648 0.5887 0.4596 300 29. 41 0.4260 0.3867 0.5246 0.3817 0.4474 0.4064 0.5333 0.4016 600 58. 82 0.4057 0.3564 0.4459 0.3528 0.3862 0.3513 0.4555 0.3478 1000 98 04 0.3839 0.3365 0.4106 0.3372 0.3435 0.3143 0.3928 0.3118 5099 500 00 0.2286 0.2186 0.2202 0.2086 0.2322 0.2186 0.2242 0.2189 8159 800 .00 0.1781 0.1809 0.1911 0.1885 0.2069 0.1967 0.1889 0.1975

PAGE 231

231 APPENDIX N RESULTS OF FITTING OF VOLUMETRIC WATER CONTENT BY RETENTION CURVE MODEL Replicate 1 ******************************************************************* Analysis of soil hydraulic properties * Example 1: Forward Problem * Mualem based restriction, M=1 1/N Analysis of retention data only MType= 3 Method= 3 ********************************** ********************************* INITial values of the coefficients ================================== No Name INITial value Index 1 ThetaR .0650 1 2 ThetaS .4100 1 3 Alpha .0750 1 4 n 1.8900 1 5 m .4709 0 6 l .5000 0 7 Ks 1.0000 0 Observed data ============= Obs. No. Pressure head Water content Weighting coefficient 1 .000 .6203 1.0000 2 5.000 .6177 1.0000 3 10.000 .6160 1.0000 4 15.000 .6153 1.0000 5 25.000 .6032 1.0000 6 35.000 .5893 1.0000 7 50.000 .5686 1.0000 8 68.000 .5474 1.0000 9 150.000 .4998 1.0000 10 300.000 .4260 1.0000 11 600.0 00 .4057 1.0000 12 1000.000 .3839 1.0000 13 5099.000 .2286 1.0000 14 8158.000 .1781 1.0000 NIT SSQ ThetaR ThetaS Alpha n 0 1.39291 .0650 .4100 .0750 1.8900 1 .58235 .2551 .6473 .0947 1.0050 2 .51816 .0474 .6461 .2056 1.0077 3 .50709 .0197 .6458 .2242 1.0080 4 .50177 .0068 .6457 .2331 1.0082 5 .49916 .0005 .6456 .2375 1.0083 wcr is less then 0.001: Changed to fit with wcr=0.0

PAGE 232

232 NIT SSQ ThetaS Alpha n 0 1.78016 .4100 .0750 1.8900 1 .28994 .5314 .0066 1.0050 2 .17330 .6117 .3071 1.1371 3 .02750 .6494 .0922 1.1210 4 .02320 .6536 .0796 1.1291 5 .02005 .6560 .0702 1.1365 6 .01767 .6572 .0626 1.1431 7 .01259 .6465 .0304 1.1620 8 .01039 .6311 .0123 1.2029 9 .00378 .6202 .0104 1.2426 10 .00367 .6196 .0105 1.2464 11 .00367 .6197 .0106 1.2460 12 .00367 .6198 .0106 1.2460 13 .00367 .6198 .0106 1.2459 Correlation matrix ================== Theta Alpha n 1 2 3 1 1.0000 2 .6713 1.0000 3 .4030 .8614 1.0000 RSquated for regression of observed vs fitted values = .98711721 ================================================================= Nonlinear least squares analysis: final results =============================================== 95% Confidence limits Variable Value S.E.Coeff. T Value Lower Upper ThetaS .61977 .00900 68.90 .6000 .6396 Alpha .01057 .00263 4.02 .0048 .0164 n 1.24595 .02218 5 6.18 1.1971 1.2948 Observed abd fitted data ======================== NO P log P WC obs WC fit WC dev 1 .1000E 04 5.0000 .6203 .6198 .0005 2 .5000E+01 .6990 .6177 .6167 .0010 3 .1000E+02 1.0000 .6160 .6126 .0034 4 .1500E+02 1.1761 .6153 .6081 .0072 5 .2500E+02 1.3979 .6032 .5988 .0044 6 .3500E+02 1.5441 .5893 .5894 .0001 7 .5000E+02 1.6990 .5686 .5758 .0072 8 .6800E+02 1.8325 .5474 .5606 .0132 9 .1500E+03 2.1761 .4998 .5067 .0069 10 .3000E+03 2.4771 .4260 .4474 .021 4 11 .6000E+03 2.7782 .4057 .3862 .0195 12 .1000E+04 3.0000 .3839 .3435 .0404 13 .5099E+04 3.7075 .2286 .2322 .0036 14 .8158E+04 3.9116 .1781 .2069 .0288 Sum of squares of observed versus fitted values =============================================== Unweighted Weighted Retention data .00367 .00367

PAGE 233

233 Cond/Diff data .00000 .00000 All dat a .00367 .00367 Soil hydraulic properties (MType = 3) ===================================== WC P logP Cond logK Dif logD .0016 .3312E+13 12.520 .1045E 28 28.981 .8904E 13 13. 050 .0032 .1978E+12 11.296 .1659E 25 25.780 .4218E 11 11.375 .0063 .1181E+11 10.072 .2632E 22 22.580 .1998E 09 9.699 .0126 .7050E+09 8.848 .4176E 19 19.379 .9465E 08 8.024 .0190 .1356E+09 8.13 2 .3111E 17 17.507 .9041E 07 7.044 .0253 .4210E+08 7.624 .6626E 16 16.179 .4484E 06 6.348 .0316 .1699E+08 7.230 .7106E 15 15.148 .1552E 05 5.809 .0379 .8096E+07 6.908 .4937E 14 14.307 .4283E 0 5 5.368 .0443 .4326E+07 6.636 .2542E 13 13.595 .1010E 04 4.996 .0506 .2514E+07 6.400 .1051E 12 12.978 .2124E 04 4.673 .0569 .1557E+07 6.192 .3678E 12 12.434 .4091E 04 4.388 .0632 .1014E+07 6.006 .1128E 11 11.948 .7354E 04 4.133 .0696 .6886E+06 5.838 .3106E 11 11.508 .1250E 03 3.903 .0759 .4834E+06 5.684 .7834E 11 11.106 .2029E 03 3.693 .0822 .3491E+06 5.543 .1835E 10 10.736 .3 168E 03 3.499 .0885 .2583E+06 5.412 .4034E 10 10.394 .4785E 03 3.320 .0949 .1951E+06 5.290 .8401E 10 10.076 .7025E 03 3.153 .1012 .1501E+06 5.176 .1669E 09 9.778 .1006E 02 2.997 .1075 .1173E+06 5.069 .3179E 09 9.498 .1410E 02 2.851 .1138 .9295E+05 4.968 .5837E 09 9.234 .1938E 02 2.713 .1202 .7461E+05 4.873 .1037E 08 8.984 .2619E 02 2.582 .1265 .6056E+05 4.782 .1790E 08 8.747 .3485E 02 2.458 .1328 .4966E+05 4.696 .3006E 08 8.522 .4573E 02 2.340 .1391 .4110E+05 4.614 .4930E 08 8.307 .5925E 02 2.227 .1455 .3430E+05 4.535 .7910E 08 8.102 .7589E 02 2.120 .1518 .2884E+05 4.460 .1244E 07 7.905 .9618E 02 2.017 .1581 .2443E+05 4.388 .1920E 07 7.717 .1207E 01 1.918 .1644 .2082E+05 4.319 .2914E 07 7.536 .1502E 01 1.823 .1708 .1786E+05 4. 252 .4353E 07 7.361 .1854E 01 1.732 .1771 .1540E+05 4.188 .6410E 07 7.193 .2271E 01 1.644 .1834 .1335E+05 4.125 .9311E 07 7.031 .2761E 01 1.559 .1897 .1163E+05 4.065 .1336E 06 6.874 .3336E 01 1.477 .1960 .1017E+05 4.007 .1893E 06 6.723 .4006E 01 1.397 .2024 .8936E+04 3.951 .2655E 06 6.576 .4783E 01 1.320 .2087 .7881E+04 3.897 .3684E 06 6.434 .5679E 01 1.246 .2150 .6977E+ 04 3.844 .5062E 06 6.296 .6710E 01 1.173 .2213 .6198E+04 3.792 .6894E 06 6.162 .7891E 01 1.103 .2277 .5523E+04 3.742 .9307E 06 6.031 .9238E 01 1.034 .2340 .4937E+04 3.693 .1246E 05 5.904 .1077E+00 .968 .2403 .4426E+04 3.646 .1656E 05 5.781 .1251E+00 .903 .2466 .3979E+04 3.600 .2185E 05 5.660 .1447E+00 .840 .2530 .3586E+04 3.555 .2863E 05 5.543 .1668E+00 .778 .2593 3239E+04 3.510 .3727E 05 5.429 .1917E+00 .717 .2656 .2933E+04 3.467 .4822E 05 5.317 .2195E+00 .659 .2719 .2662E+04 3.425 .6201E 05 5.208 .2507E+00 .601 .2783 .2421E+04 3.384 .7930E 05 5.1 01 .2854E+00 .544 .2846 .2206E+04 3.344 .1009E 04 4.996 .3242E+00 .489 .2909 .2013E+04 3.304 .1277E 04 4.894 .3672E+00 .435 .2972 .1841E+04 3.265 .1608E 04 4.794 .4149E+00 .382 .3036 .1686E+04 3.227 .2016E 04 4.696 .4678E+00 .330 .3099 .1547E+04 3.189 .2516E 04 4.599 .5263E+00 .279

PAGE 234

234 .3162 .1421E+04 3.153 .3127E 04 4.505 .5909E+00 .228 .3225 .1307E+04 3.116 .3871E 04 4.412 .6621E+00 .179 .3289 .1204E+04 3.081 .4774E 04 4.321 .7406E+00 .130 .3352 .1111E+04 3.046 .5865E 04 4.232 .8269E+00 .083 .3415 .1026E+04 3.011 .7181E 04 4.144 .9218E+00 .035 .3478 .9480E+03 2.977 .8763E 04 4.057 .1026E+01 .011 .3542 .8772E+03 2.943 .1066E 03 3.972 .1140E+01 .057 .3605 .8124E+03 2.910 .1293E 03 3.889 .1266E+01 .102 .3668 .7531E+03 2. 877 .1563E 03 3.806 .1403E+01 .147 .3731 .6987E+03 2.844 .1885E 03 3.725 .1554E+01 .191 .3794 .6487E+03 2.812 .2266E 03 3.645 .1719E+01 .235 .3858 .6027E+03 2.780 .2719E 03 3.566 .1899E +01 .279 .3921 .5602E+03 2.748 .3254E 03 3.488 .2097E+01 .322 .3984 .5211E+03 2.717 .3886E 03 3.411 .2313E+01 .364 .4047 .4849E+03 2.686 .4630E 03 3.334 .2549E+01 .406 .4111 .4513E+ 03 2.654 .5505E 03 3.259 .2809E+01 .448 .4174 .4203E+03 2.624 .6534E 03 3.185 .3092E+01 .490 .4237 .3914E+03 2.593 .7742E 03 3.111 .3403E+01 .532 .4300 .3646E+03 2.562 .9157E 03 3.038 .3744E+01 .573 .4364 .3396E+03 2.531 .1081E 02 2.966 .4119E+01 .615 .4427 .3164E+03 2.500 .1275E 02 2.894 .4530E+01 .656 .4490 .2947E+03 2.469 .1502E 02 2.823 .4981E+01 .697 .4553 2744E+03 2.438 .1767E 02 2.753 .5479E+01 .739 .4617 .2554E+03 2.407 .2077E 02 2.683 .6027E+01 .780 .4680 .2376E+03 2.376 .2439E 02 2.613 .6633E+01 .822 .4743 .2210E+03 2.344 .2861E 02 2.543 .7303E+01 .864 .4806 .2053E+03 2.312 .3355E 02 2.474 .8047E+01 .906 .4870 .1906E+03 2.280 .3932E 02 2.405 .8873E+01 .948 .4933 .1768E+03 2.247 .4607E 02 2.337 .9794E+01 .991 .4996 .1637E+03 2.214 .5398E 02 2.268 .1082E+02 1.034 .5059 .1514E+03 2.180 .6326E 02 2.199 .1198E+02 1.079 .5123 .1397E+03 2.145 .7416E 02 2.130 .1329E+02 1.123 .5186 .1287E+03 2.110 .8699E 02 2.061 .1476E+02 1.169 .5249 .1182E+03 2.073 .1021E 01 1.991 .1644E+02 1.216 .5312 .1083E+03 2.035 .1201E 01 1.921 .1836E+02 1.264 .5376 .9889E+02 1.995 .1414E 01 1.850 .20 58E+02 1.314 .5439 .8989E+02 1.954 .1668E 01 1.778 .2316E+02 1.365 .5502 .8131E+02 1.910 .1974E 01 1.705 .2619E+02 1.418 .5565 .7310E+02 1.864 .2344E 01 1.630 .2978E+02 1.474 .5628 .652 3E+02 1.814 .2795E 01 1.554 .3410E+02 1.533 .5692 .5766E+02 1.761 .3351E 01 1.475 .3938E+02 1.595 .5755 .5035E+02 1.702 .4046E 01 1.393 .4598E+02 1.663 .5818 .4327E+02 1.636 .4930E 01 1.307 .5443E+02 1.736 .5881 .3637E+02 1.561 .6083E 01 1.216 .6564E+02 1.817 .5945 .2960E+02 1.471 .7639E 01 1.117 .8126E+02 1.910 .6008 .2289E+02 1.360 .9848E 01 1.007 .1047E+03 2.020 .6071 .1610E+02 1.207 .1328E+00 .877 .1444E+03 2.160 .6134 .9001E+01 .954 .1971E+00 .705 .2322E+03 2.366 .6166 .5096E+01 .707 .2646E+00 .577 .3475E+03 2.541 .6182 .2903E+01 .463 .3320E+00 .479 .4932E+03 2.693 .6191 .1386E+01 .142 .4177E+00 .379 .7372E+03 2.868 .6197 .2179E+00 .662 .6014E+00 .221 .1664E+04 3.221 .6198 .3432E 01 1.464 .7353E+00 .134 .3203E+04 3.506 .6198 .0000E+00 .1000E+01 .000 End of problem ==============

PAGE 235

235 Replicate 2 ******************************************************************* Analysis of soil hydraulic properties * Example 1: Forward Problem * Mualem based restriction, M=1 1/N Analysis of retention data only MType= 3 Method= 3 ******************************************************************* INITial values of the coefficients ================================== No Name INITial value Index 1 ThetaR .0650 1 2 ThetaS .4100 1 3 Alpha .0750 1 4 n 1.8900 1 5 m .4709 0 6 l .5000 0 7 Ks 1.0000 0 Observed data ============= Obs. No. Pressure head Water content Weighting coefficient 1 .000 .6160 1.0000 2 5.000 .6160 1.0000 3 10.000 .6160 1.0000 4 15.000 .6109 1.0000 5 25.000 .5943 1.0000 6 35.000 .5753 1.0000 7 50.000 .5472 1.0000 8 68.000 .5217 1.0000 9 150.000 .460 4 1.0000 10 300.000 .3867 1.0000 11 600.000 .3564 1.0000 12 1000.000 .3365 1.0000 13 5099.000 .2286 1.0000 14 8158.000 .1781 1.0000 NIT SSQ ThetaR ThetaS Alpha n 0 1.22174 .0650 .4100 .0750 1.8900 1 .63378 .2351 .6374 .076 0 1.0050 2 .56150 .0117 .6361 .1809 1.0078 3 .55845 .0044 .6361 .1854 1.0079 4 .55695 .0008 .6360 .1876 1.0080 5 .55676 .0004 .6360 .1879 1.0080 wcr is less then 0.001: Changed to fit with wcr=0.0 NIT SSQ ThetaS Alpha n 0 1.58104 .4100 .0750 1.8900 1 .32522 .5304 .0081 1.0050 2 .12561 .6084 .5078 1.1143

PAGE 236

236 3 .03525 .6422 .1639 1.1088 4 .02766 .6486 .1411 1.1189 5 .02221 .6536 .1228 1.1279 6 .01825 .6571 .1084 1.1358 7 .01532 .6594 .0968 1.1428 8 .01310 .6606 .0873 1.1490 9 .00825 .6500 .0467 1.1665 10 .00519 .6355 .0231 1.2002 11 .00179 .6272 .0209 1.2238 12 .00176 .6273 .0211 1.2251 13 .00176 .6272 .0211 1.2251 14 .00176 .6272 .0211 1.2251 Correlation matrix ================== Theta Alpha n 1 2 3 1 1.0000 2 .7328 1.0000 3 .4187 8394 1.0000 RSquated for regression of observed vs fitted values = .99411768 ================================================================= Nonlinear least squares analysis: final results =============================================== 95% Confidence limits Variable Value S.E.Coeff. T Value Lower Upper ThetaS .62724 .00742 84.57 .61 09 .6436 Alpha .02113 .00371 5.69 .0130 .0293 n 1.22511 .01239 98.91 1.1978 1.2524 Observed abd fitted data ======================== NO P log P WC obs WC fit WC dev 1 .1000E 04 5.0000 .6160 .6272 .0112 2 .5000E+01 .6990 .6160 .6202 .0042 3 .1000E+02 1.0000 .6160 .6114 .0046 4 .1500E+02 1.1761 .6109 .6025 .0084 5 .2500E+02 1.3979 .5943 .5853 .0090 6 .3500E+02 1.5441 .5753 .5695 .0058 7 .5000E+02 1.6990 .5472 .5488 .0016 8 .6800E+02 1.8325 .5217 .5278 .0061 9 .1500E+03 2.1761 .4604 .4648 .0044 10 .3000E+03 2.4771 .3867 .4064 .0197 11 .6000E+03 2.7782 .3564 .3513 .0051 12 .1000E+04 3.0000 .3365 .3143 .0222 13 .5099E+04 3.7075 .2286 .2186 .0100 14 .8158E+04 3.9116 .1781 .1967 .0186 Sum of squares of observed versus fitted values =============================================== Unweighted Weighted Retention data .00176 .00176 Cond/Diff data .00000 .00000 All data .00176 .00176

PAGE 237

237 Soil hydraulic properties (MType = 3) =============================== ====== WC P logP Cond logK Dif logD .0016 .1567E+14 13.195 .1012E 30 30.995 .4402E 14 14.356 .0032 .7207E+12 11.858 .2706E 27 27.568 .2707E 12 12.568 .0064 .3315E+11 10. 521 .7234E 24 24.141 .1664E 10 10.779 .0128 .1525E+10 9.183 .1934E 20 20.714 .1023E 08 8.990 .0192 .2518E+09 8.401 .1955E 18 18.709 .1139E 07 7.944 .0256 .7015E+08 7.846 .5170E 17 17.286 .6293E 07 7.201 .0320 .2603E+08 7.416 .6558E 16 16.183 .2370E 06 6.625 .0384 .1158E+08 7.064 .5227E 15 15.282 .7002E 06 6.155 .0448 .5840E+07 6.766 .3023E 14 14.520 .1750E 05 5.757 .0512 .3227E+ 07 6.509 .1382E 13 13.859 .3870E 05 5.412 .0576 .1912E+07 6.282 .5284E 13 13.277 .7791E 05 5.108 .0640 .1197E+07 6.078 .1753E 12 12.756 .1457E 04 4.836 .0704 .7841E+06 5.894 .5189E 12 12.285 .2567E 04 4.591 .0768 .5327E+06 5.727 .1397E 11 11.855 .4306E 04 4.366 .0832 .3733E+06 5.572 .3476E 11 11.459 .6928E 04 4.159 .0896 .2686E+06 5.429 .8081E 11 11.093 .1076E 03 3.968 .0960 1977E+06 5.296 .1772E 10 10.751 .1621E 03 3.790 .1024 .1484E+06 5.171 .3696E 10 10.432 .2379E 03 3.624 .1088 .1134E+06 5.055 .7369E 10 10.133 .3411E 03 3.467 .1152 .8795E+05 4.944 .1413E 09 9.850 .4791E 03 3.320 .1216 .6917E+05 4.840 .2614E 09 9.583 .6607E 03 3.180 .1280 .5508E+05 4.741 .4688E 09 9.329 .8962E 03 3.048 .1344 .4434E+05 4.647 .8170E 09 9.088 .1198E 02 2.922 .1408 .3606E+05 4.557 .1388E 08 8.858 .1579E 02 2.802 .1472 .2960E+05 4.471 .2302E 08 8.638 .2057E 02 2.687 .1536 .2450E+05 4.389 .3737E 08 8.427 .2649E 02 2.577 .1600 .2043E+05 4.310 .5948E 08 8.226 .3376E 02 2.472 .1664 .1716E+05 4.235 .9298E 08 8.032 .4263E 02 2.370 .1728 .1451E+05 4.162 .1429E 07 7.845 .5335E 02 2.273 .1792 .1235E+05 4.091 .2162E 07 7.665 .66 24E 02 2.179 .1856 .1056E+05 4.024 .3225E 07 7.492 .8162E 02 2.088 .1920 .9083E+04 3.958 .4745E 07 7.324 .9986E 02 2.001 .1984 .7850E+04 3.895 .6893E 07 7.162 .1214E 01 1.916 .2048 .681 5E+04 3.833 .9898E 07 7.004 .1466E 01 1.834 .2112 .5942E+04 3.774 .1405E 06 6.852 .1761E 01 1.754 .2176 .5202E+04 3.716 .1975E 06 6.704 .2104E 01 1.677 .2240 .4572E+04 3.660 .2748E 06 6.561 .2501E 01 1.602 .2304 .4032E+04 3.606 .3790E 06 6.421 .2958E 01 1.529 .2368 .3568E+04 3.552 .5180E 06 6.286 .3484E 01 1.458 .2432 .3167E+04 3.501 .7022E 06 6.154 .4085E 01 1.389 .2496 .2820E+04 3.450 .9444E 06 6.025 .4771E 01 1.321 .2560 .2518E+04 3.401 .1261E 05 5.899 .5552E 01 1.256 .2624 .2254E+04 3.353 .1672E 05 5.777 .6436E 01 1.191 .2688 .2024E+04 3.306 .2202E 05 5.657 .7436E 01 1.129 .2752 .1821E+04 3.260 .2881E 05 5.540 .8564E 01 1.067 .2816 .1642E+04 3.215 .3748E 05 5.426 .9832E 01 1.007 .2880 .1484E+04 3.171 .4847E 05 5.315 .1126E+00 .949 .2944 .1344E+04 3.128 .6235E 05 5.205 .1285E+00 .891 .3008 .1219E+04 3.086 .7978E 05 5.098 .1463E+00 .835 .3072 .1108E+04 3.045 .1016E 04 4.993 .1662E+00 .779 .3136 .1009E+04 3.004 .1287E 04 4.890 .1883E+00 .725 .3200 .9206E+03 2.964 .1624E 04 4.790 .2129E+00 .672 .3264 .8410E+03 2.925 .2039E 04 4.691 .2402E+00 .619 .3328 .7694E+03 2.886 .2550E 04 4.593 .2705E+00 .568

PAGE 238

238 .3392 .7049E+03 2.848 .3177E 04 4.498 .3040E+00 .517 .3456 .6467E+03 2.811 .3943E 04 4.404 .3410E+00 .467 .3520 .5940E+03 2.774 .4876E 04 4.312 .3819E+00 .418 .3584 .5462E+03 2. 737 .6009E 04 4.221 .4271E+00 .369 .3648 .5028E+03 2.701 .7380E 04 4.132 .4768E+00 .322 .3712 .4634E+03 2.666 .9036E 04 4.044 .5316E+00 .274 .3776 .4274E+03 2.631 .1103E 03 3.957 .5920E +00 .228 .3840 .3945E+03 2.596 .1343E 03 3.872 .6584E+00 .182 .3904 .3645E+03 2.562 .1630E 03 3.788 .7314E+00 .136 .3968 .3370E+03 2.528 .1973E 03 3.705 .8116E+00 .091 .4032 .3118E+ 03 2.494 .2383E 03 3.623 .8999E+00 .046 .4096 .2886E+03 2.460 .2872E 03 3.542 .9969E+00 .001 .4160 .2673E+03 2.427 .3453E 03 3.462 .1103E+01 .043 .4224 .2476E+03 2.394 .4142E 03 3.383 .1221E+01 .087 .4288 .2295E+03 2.361 .4960E 03 3.305 .1350E+01 .130 .4352 .2128E+03 2.328 .5928E 03 3.227 .1492E+01 .174 .4416 .1973E+03 2.295 .7074E 03 3.150 .1648E+01 .217 .4480 1829E+03 2.262 .8427E 03 3.074 .1820E+01 .260 .4544 .1696E+03 2.229 .1003E 02 2.999 .2010E+01 .303 .4608 .1572E+03 2.197 .1191E 02 2.924 .2220E+01 .346 .4672 .1457E+03 2.163 .1413E 02 2.850 .2452E+01 .389 .4736 .1350E+03 2.130 .1676E 02 2.776 .2709E+01 .433 .4800 .1250E+03 2.097 .1985E 02 2.702 .2994E+01 .476 .4864 .1157E+03 2.063 .2349E 02 2.629 .3311E+01 .520 .4928 .1069E+03 2.029 .2778E 02 2.556 .3664E+01 .564 .4992 .9876E+02 1.995 .3286E 02 2.483 .4059E+01 .608 .5056 .9110E+02 1.960 .3885E 02 2.411 .4503E+01 .653 .5120 .8391E+02 1.924 .4593E 02 2.338 .5001E+01 .699 .5184 .7715E+02 1.887 .5433E 02 2.265 .5564E+01 .745 .5248 .7079E+02 1.850 .6429E 02 2.192 .6204E+01 .793 .5312 .6479E+02 1.812 .7615E 02 2.118 .69 33E+01 .841 .5376 .5913E+02 1.772 .9031E 02 2.044 .7770E+01 .890 .5440 .5377E+02 1.731 .1073E 01 1.969 .8738E+01 .941 .5504 .4869E+02 1.687 .1277E 01 1.894 .9865E+01 .994 .5568 .438 7E+02 1.642 .1525E 01 1.817 .1119E+02 1.049 .5632 .3929E+02 1.594 .1827E 01 1.738 .1277E+02 1.106 .5696 .3492E+02 1.543 .2198E 01 1.658 .1467E+02 1.166 .5760 .3074E+02 1.488 .2660E 01 1.575 .1700E+02 1.230 .5824 .2673E+02 1.427 .3243E 01 1.489 .1992E+02 1.299 .5888 .2287E+02 1.359 .3991E 01 1.399 .2367E+02 1.374 .5952 .1913E+02 1.282 .4977E 01 1.303 .2865E+02 1.457 .6016 .1549E+02 1.190 .6320E 01 1.199 .3562E+02 1.552 .6080 .1190E+02 1.076 .8252E 01 1.083 .4609E+02 1.664 .6144 .8312E+01 .920 .1129E+00 .947 .6392E+02 1.806 .6208 .4592E+01 .662 .1710E+00 .767 .1035E+03 2.015 .6240 .2573E+01 .410 .2331E+00 .632 .1555E+03 2.192 .6256 .1451E+01 .162 .2964E+00 .528 .2212E+03 2.345 .6266 .6842E+00 .165 .3781E+00 .422 .3310E+03 2.520 .6272 .1042E+00 .982 .5591E+00 .252 .7432E+03 2.871 .6272 .1590E 01 1.799 .6969E+00 .157 .1413E+04 3.150 .6272 .0000E+00 .1000E+01 .000 End of problem ==============

PAGE 239

239 Replicate 3 ******************************************************************* Analysis of soil hydraulic properties * Example 1: Forward Problem * Mualem based restriction, M=1 1/N Analysis of retention data only MType= 3 Method= 3 ******************************* ************************************ INITial values of the coefficients ================================== No Name INITial value Index 1 ThetaR .0650 1 2 ThetaS .4100 1 3 Alpha .0750 1 4 n 1.8900 1 5 m .4709 0 6 l .5000 0 7 Ks 1.0000 0 Observed data ====== ======= Obs. No. Pressure head Water content Weighting coefficient 1 .000 .6473 1.0000 2 5.000 .6425 1.0000 3 10.000 .64 20 1.0000 4 15.000 .6420 1.0000 5 25.000 .6386 1.0000 6 35.000 .6348 1.0000 7 50.000 .6249 1.0000 8 68.000 .6153 1.0000 9 150.000 .5965 1.0000 10 300.000 .5246 1.0000 11 600.000 .4459 1.0000 12 1000.000 .4106 1.0000 13 5099.000 .2286 1.0000 14 8158.000 .1781 1.0000 NIT SSQ ThetaR ThetaS Alpha n 0 1.82937 .0650 .4100 .0750 1.8900 1 .32231 .1971 .5334 .0052 1.0050 2 .30930 .0314 .5345 .0091 1.0075 3 .30688 .0082 .5345 .0100 1.0078 4 .30631 .0029 .5345 .0102 1.0079 5 .30603 .0003 .5345 .0103 1.0080 wcr is less then 0.001: Changed to fit with wcr=0.0 NIT SSQ ThetaS Alpha n 0 2.26798 .4100 .0750 1.8900 1 .47893 .4768 .0374 1.1716 2 .20402 .5174 .0136 1.1581

PAGE 240

240 3 .06753 .5754 .0053 1.1720 4 .00400 .6431 .0039 1.3016 5 .00091 .6432 .0033 1.3640 6 .00077 .6431 .0033 1.3703 7 .00077 .6431 .0033 1.3701 8 .00077 .6431 .0033 1.3701 Correlation matrix ============ ====== Theta Alpha n 1 2 3 1 1.0000 2 .5335 1.0000 3 .3107 .8575 1.0000 RSquated for regression of observed vs fitted values = .99768288 =================== ============================================== Nonlinear least squares analysis: final results =============================================== 95% Confidence limits Variable Value S.E.Coeff. T Value Lower Upper ThetaS .64310 .00324 198.19 .6360 .6502 Alpha .00333 .00027 12.18 .0027 .0039 n 1.37009 .01572 87.18 1.3355 1.4 047 Observed abd fitted data ======================== NO P log P WC obs WC fit WC dev 1 .1000E 04 5.0000 .6473 .6431 .0042 2 .5000E+01 .6990 .6425 .6425 .0000 3 .1000E+02 1.0000 .6420 .6415 .0005 4 .1500E+02 1.1761 .6420 .6403 .0017 5 .2500E+02 1.3979 .6386 .6374 .0012 6 .3500E+02 1.5441 .6348 .6342 .0006 7 5000E+02 1.6990 .6249 .6289 .0040 8 .6800E+02 1.8325 .6153 .6221 .0068 9 .1500E+03 2.1761 .5965 .5887 .0078 10 .3000E+03 2.4771 .5246 .5333 .0087 11 .6000E+03 2.7782 .4459 .4555 .0096 12 .1000E+04 3.0000 .4106 .3928 .0178 13 .5099E+04 3.7075 .2286 .2242 .0044 14 .8158E+04 3.9116 .1781 .1889 .0108 Sum of squares of obser ved versus fitted values =============================================== Unweighted Weighted Retention data .00077 .00077 Cond/Diff data .00000 .00000 All data .00077 .00077 Soil hydraulic properties (MType = 3) ===================================== WC P logP Cond logK Dif logD .0016 .3050E+10 9.484 .2320E 21 21.635 .1166E 08 8.933 .0033 .4688E+0 9 8.671 .5557E 19 19.255 .2145E 07 7.669 .0066 .7204E+08 7.858 .1331E 16 16.876 .3948E 06 6.404

PAGE 241

241 .0131 .1107E+08 7.044 .3189E 14 14.496 .7267E 05 5.139 .0197 .3701E+07 6.568 .7861E 13 13.105 3993E 04 4.399 .0262 .1701E+07 6.231 .7638E 12 12.117 .1338E 03 3.874 .0328 .9309E+06 5.969 .4456E 11 11.351 .3416E 03 3.466 .0394 .5688E+06 5.755 .1883E 10 10.725 .7350E 03 3.134 .0459 .3 750E+06 5.574 .6368E 10 10.196 .1405E 02 2.852 .0525 .2614E+06 5.417 .1830E 09 9.738 .2462E 02 2.609 .0591 .1901E+06 5.279 .4642E 09 9.333 .4039E 02 2.394 .0656 .1430E+06 5.155 .1068E 08 8.97 2 .6289E 02 2.201 .0722 .1105E+06 5.044 .2268E 08 8.644 .9388E 02 2.027 .0787 .8738E+05 4.941 .4512E 08 8.346 .1353E 01 1.869 .0853 .7038E+05 4.847 .8495E 08 8.071 .1895E 01 1.722 .0919 .5760E+05 4.760 .1526E 07 7.816 .2587E 01 1.587 .0984 .4780E+05 4.679 .2633E 07 7.580 .3458E 01 1.461 .1050 .4014E+05 4.604 .4387E 07 7.358 .4537E 01 1.343 .1116 .3407E+05 4.532 .7085E 07 7.150 .5855E 01 1.232 .1181 .2918E+05 4.465 .1113E 06 6.953 .7447E 01 1.128 .1247 .2521E+05 4.402 .1708E 06 6.768 .9350E 01 1.029 .1312 .2194E+05 4.341 .2562E 06 6.591 .1160E+00 .935 .1378 .1922E+05 4.284 .3769E 06 6.424 .1425E+00 .846 .1444 .1694E+05 4.229 .5447E 06 6.264 .1734E+00 .761 .1509 .1502E+05 4.177 .7744E 06 6.111 .2092E+00 .679 .1575 .1338E+05 4.126 .1085E 05 5.965 .2503E+00 .601 .1641 .1197E+05 4.078 .1499E 05 5.824 .2975E+00 .527 .1706 .1076E+05 4.032 .2045E 05 5.689 .3511E+00 .455 .1772 .9710E+04 3.987 .2758E 05 5.559 .4119E+00 .385 .1837 .8794E+04 3.944 .3680E 05 5.434 .4805E+00 .318 .1903 .7990E+04 3.903 .4861E 05 5.313 .5576E+00 .254 .1969 .7283E+04 3.862 .6361E 05 5.196 .6439E+00 .191 .2034 .6657E+04 3. 823 .8253E 05 5.083 .7403E+00 .131 .2100 .6102E+04 3.785 .1062E 04 4.974 .8474E+00 .072 .2166 .5607E+04 3.749 .1356E 04 4.868 .9663E+00 .015 .2231 .5165E+04 3.713 .1720E 04 4.764 .1098E +01 .040 .2297 .4768E+04 3.678 .2167E 04 4.664 .1243E+01 .094 .2362 .4410E+04 3.644 .2712E 04 4.567 .1402E+01 .147 .2428 .4087E+04 3.611 .3374E 04 4.472 .1578E+01 .198 .2494 .3795E+ 04 3.579 .4175E 04 4.379 .1770E+01 .248 .2559 .3530E+04 3.548 .5138E 04 4.289 .1980E+01 .297 .2625 .3288E+04 3.517 .6292E 04 4.201 .2210E+01 .344 .2691 .3068E+04 3.487 .7667E 04 4.115 .2460E+01 .391 .2756 .2867E+04 3.457 .9302E 04 4.031 .2733E+01 .437 .2822 .2682E+04 3.428 .1124E 03 3.949 .3029E+01 .481 .2887 .2512E+04 3.400 .1352E 03 3.869 .3351E+01 .525 .2953 2356E+04 3.372 .1620E 03 3.790 .3701E+01 .568 .3019 .2212E+04 3.345 .1935E 03 3.713 .4079E+01 .611 .3084 .2079E+04 3.318 .2302E 03 3.638 .4489E+01 .652 .3150 .1956E+04 3.291 .2730E 03 3.564 .4932E+01 .693 .3216 .1842E+04 3.265 .3227E 03 3.491 .5411E+01 .733 .3281 .1736E+04 3.239 .3804E 03 3.420 .5928E+01 .773 .3347 .1637E+04 3.214 .4470E 03 3.350 .6486E+01 .812 .3412 .1545E+04 3.189 .5239E 03 3.281 .7088E+01 .851 .3478 .1459E+04 3.164 .6125E 03 3.213 .7738E+01 .889 .3544 .1379E+04 3.140 .7142E 03 3.146 .8438E+01 .926 .3609 .1304E+04 3.115 .8308E 03 3.081 .9193E+01 .963 .3675 .1234E+04 3.091 .9642E 03 3.016 .1001E+02 1.000 .3740 .1168E+04 3.067 .1117E 02 2.952 .1088E+02 1.037 .3806 .1106E+04 3.044 .1291E 02 2.889 .11 83E+02 1.073

PAGE 242

242 .3872 .1047E+04 3.020 .1489E 02 2.827 .1285E+02 1.109 .3937 .9923E+03 2.997 .1715E 02 2.766 .1394E+02 1.144 .4003 .9404E+03 2.973 .1971E 02 2.705 .1513E+02 1.180 .4069 .891 5E+03 2.950 .2262E 02 2.646 .1640E+02 1.215 .4134 .8452E+03 2.927 .2591E 02 2.586 .1778E+02 1.250 .4200 .8014E+03 2.904 .2965E 02 2.528 .1926E+02 1.285 .4265 .7598E+03 2.881 .3388E 02 2.470 .2087E+02 1.320 .4331 .7205E+03 2.858 .3866E 02 2.413 .2261E+02 1.354 .4397 .6830E+03 2.834 .4406E 02 2.356 .2449E+02 1.389 .4462 .6475E+03 2.811 .5016E 02 2.300 .2652E+02 1.424 .4528 .6136E+03 2.788 .5706E 02 2.244 .2873E+02 1.458 .4594 .5813E+03 2.764 .6483E 02 2.188 .3113E+02 1.493 .4659 .5506E+03 2.741 .7361E 02 2.133 .3374E+02 1.528 .4725 .5212E+03 2.717 .8352E 02 2.078 .3658E+02 1.563 .4790 .4931E+03 2.693 .9471E 02 2.024 .3967E+02 1.599 .4856 .4662E+03 2.669 .1073E 01 1.969 .4306E+02 1.634 .4922 .4404E+03 2.644 .1216E 01 1.915 .4677E+02 1.670 .4987 .4156E+03 2.619 .1377E 01 1.861 .5085E+02 1.706 .5053 .3919E+03 2.593 .1559E 01 1.807 .5534E+02 1.743 .5119 .3690E+03 2.567 .1766E 01 1.753 .6031E+02 1.780 .5184 .3470E+03 2.540 .2000E 01 1.699 .6580E+02 1.818 .5250 .3258E+03 2.513 .2265E 01 1.645 .7192E+02 1.857 .5315 .3054E+03 2.485 .2568E 01 1.590 .7876E+02 1.896 .5381 .2856E+03 2.456 .2912E 01 1.536 .8643E+02 1.937 .5447 .2664E+03 2.426 .3306E 01 1.481 .9509E+02 1.978 .5512 .2478E+03 2.394 .3757E 01 1.425 .1049E+03 2.021 .5578 .2297E+03 2.361 .4276E 01 1.369 .1162E+03 2.065 .5644 .2121E+03 2. 327 .4875E 01 1.312 .1291E+03 2.111 .5709 .1949E+03 2.290 .5572E 01 1.254 .1442E+03 2.159 .5775 .1781E+03 2.251 .6387E 01 1.195 .1620E+03 2.209 .5840 .1616E+03 2.209 .7347E 01 1.134 .1831E +03 2.263 .5906 .1454E+03 2.163 .8488E 01 1.071 .2088E+03 2.320 .5972 .1293E+03 2.112 .9863E 01 1.006 .2405E+03 2.381 .6037 .1133E+03 2.054 .1155E+00 .938 .2808E+03 2.448 .6103 .9735E+ 02 1.988 .1365E+00 .865 .3338E+03 2.524 .6169 .8118E+02 1.909 .1636E+00 .786 .4069E+03 2.610 .6234 .6460E+02 1.810 .2001E+00 .699 .5154E+03 2.712 .6300 .4719E+02 1.674 .2534E+00 .596 .6978E+03 2.844 .6365 .2795E+02 1.446 .3449E+00 .462 .1098E+04 3.041 .6398 .1670E+02 1.223 .4323E+00 .364 .1626E+04 3.211 .6415 .1003E+02 1.001 .5126E+00 .290 .2301E+04 3.362 .6424 5124E+01 .710 .6057E+00 .218 .3460E+04 3.539 .6430 .9528E+00 .021 .7762E+00 .110 .8228E+04 3.915 .6431 .1774E+00 .751 .8763E+00 .057 .1730E+05 4.238 .6431 .0000E+00 .1000E+01 .000 End of problem ==============

PAGE 243

243 Replicate 4 ******************************************************************* Analysis of soil hydraulic properties * Example 1: Forward Problem * Mualem based restriction, M=1 1/N Analysis of retention data only MType= 3 Method= 3 ******************************************************************* INITial values of the coefficients ================================== No Name INITial value Index 1 ThetaR .0650 1 2 ThetaS .4100 1 3 Alpha .0750 1 4 n 1.8900 1 5 m .4709 0 6 l .5000 0 7 Ks 1.0000 0 Observed data ============= Obs. No. Pressure head Water content Weighting coefficient 1 .000 .6299 1.0000 2 5.000 .6261 1.0000 3 10.000 .6254 1.0000 4 15.000 .6165 1.0000 5 25.000 .5984 1.0000 6 35.000 .5756 1.0000 7 50.000 .5453 1.0000 8 68.000 .5171 1.0000 9 150.000 .4536 1.0000 10 300.000 .3817 1.0000 11 600.000 .3528 1.0000 1 2 1000.000 .3372 1.0000 13 5099.000 .2286 1.0000 14 8158.000 .1781 1.0000 NIT SSQ ThetaR ThetaS Alpha n 0 1.22998 .0650 .4100 .0750 1.8900 1 .69231 .2346 .6491 .0744 1.0050 2 .61262 .0032 .6477 .1836 1.0078 3 .61178 .0013 .6477 .1848 1.0079 4 .61136 .0004 .6477 .1854 1.0079 5 .61125 .0002 .6477 .1855 1.0079 wcr is less then 0.001: Changed to fit with wcr=0.0 NIT SSQ ThetaS Alpha n 0 1.58773 .4100 .0750 1.8900 1 .34390 .5354 .0076 1.0050 2 .11414 .6135 .5331 1.1088

PAGE 244

244 3 .03896 .6464 .1914 1.1062 4 .03745 .6509 .0454 1.1334 5 .02715 .6389 .0139 1.2040 6 .00334 .6379 .0248 1.2070 7 .00198 .6402 .0265 1.2183 8 .00198 .6400 .0262 1.2191 9 .00198 .6400 .0263 1.2190 10 .00198 .6400 .0263 1.2190 11 .00198 .6400 .0263 1.2190 Correlation matrix ================== Theta Alpha n 1 2 3 1 1.0000 2 .7468 1.0000 3 .4153 .8306 1.0000 RSquated for regression of observed vs fitted values = .99368809 ================================================================= Nonlinear least squares analysis: final results =============================================== 95% Confidence limits Variable Value S.E.Coeff. T Value Lower Upper ThetaS .64004 .00830 77.12 .6218 .6583 Alpha .02625 .00484 5 .43 .0156 .0369 n 1.21904 .01208 100.88 1.1924 1.2456 Observed abd fitted data ======================== NO P log P WC obs WC fit WC dev 1 .1000E 04 5.0000 .6299 .6400 .0101 2 .5000E+01 .6990 .6261 .6308 .0047 3 .1000E+02 1.0000 .6254 .6198 .0056 4 .1500E+02 1.1761 .6165 .6088 .0077 5 .2500E+02 1.3979 .5984 .5883 .0101 6 .3500E+02 1.5441 .5756 .5702 .0054 7 .5000E+02 1.6990 .5453 .5472 .0019 8 .6800E+02 1.8325 .5171 .5245 .0074 9 .1500E+03 2.1761 .4536 .4596 .0060 10 .3000E+03 2.4771 .3817 .4016 .0199 11 .6000E+03 2.7782 .3528 .3478 .0050 12 .1000E+04 3.0000 .3372 .3118 .0254 13 .5099E+04 3.7075 .2286 .2189 .0097 14 .8158E+04 3.9116 .1781 .1975 .0194 Sum of squares of observed versus fitted values =============================================== Unw eighted Weighted Retention data .00198 .00198 Cond/Diff data .00000 .00000 All data .00198 .00198 Soil hydraulic properties (MType = 3) ===================================== WC P logP Cond logK Dif logD

PAGE 245

245 .0033 .1111E+13 12.046 .7055E 28 28.151 .1096E 12 12.960 .0065 .4693E+11 10.671 .2237E 24 24.650 .7339E 11 11.134 .0131 .1982E+10 9.297 .7094E 21 21.149 .4914E 09 9.309 .0196 .3113E+09 8.493 .7924E 19 19.101 .5748E 08 8.240 .0261 .8372E+08 7.923 .2249E 17 17.648 .3291E 07 7.483 .0327 .3023E+08 7.480 .3014E 16 16.521 .1274E 06 6.895 .0392 .1315E+08 7.119 .2513E 15 15.600 .3849E 06 6.415 .0457 .6505E+07 6.813 .1509E 14 14.821 .9805E 06 6.009 .0522 .3536E+07 6.549 .7133E 14 14.147 .2204E 05 5.657 .0588 .2065E+07 6.315 .2807E 13 1 3.552 .4502E 05 5.347 .0653 .1277E+07 6.106 .9558E 13 13.020 .8530E 05 5.069 .0718 .8263E+06 5.917 .2896E 12 12.538 .1521E 04 4.818 .0784 .5554E+06 5.745 .7967E 12 12.099 .2578E 04 4.589 .0849 .3854E+06 5.586 .2021E 11 11.694 .4189E 04 4.378 .0914 .2748E+06 5.439 .4786E 11 11.320 .6566E 04 4.183 .0980 .2005E+06 5.302 .1068E 10 10.972 .9978E 04 4.001 .1045 .1493E+06 5.174 .2262E 10 10.646 .1476E 03 3.831 .1110 .1132E+06 5.054 .4578E 10 10.339 .2132E 03 3.671 .1176 .8723E+05 4.941 .8900E 10 10.051 .3015E 03 3.521 .1241 .6815E+05 4.833 .1669E 09 9.777 .4186E 03 3.378 .1306 .5392E+05 4.732 .3031E 09 9.518 .5713E 03 3.243 .1372 .4315E+05 4.635 .5347E 09 9.272 .7681E 03 3.115 .1437 .3489E+05 4.543 .9185E 09 9.037 .1019E 02 2.992 .1502 .2848E+05 4. 455 .1540E 08 8.812 .1334E 02 2.875 .1567 .2345E+05 4.370 .2527E 08 8.597 .1727E 02 2.763 .1633 .1946E+05 4.289 .4063E 08 8.391 .2212E 02 2.655 .1698 .1627E+05 4.211 .6413E 08 8.193 .2807E 02 2.552 .1763 .1369E+05 4.136 .9947E 08 8.002 .3529E 02 2.452 .1829 .1160E+05 4.064 .1519E 07 7.819 .4401E 02 2.356 .1894 .9878E+04 3.995 .2285E 07 7.641 .5446E 02 2.264 .1959 .8460E+ 04 3.927 .3389E 07 7.470 .6691E 02 2.175 .2025 .7282E+04 3.862 .4964E 07 7.304 .8165E 02 2.088 .2090 .6298E+04 3.799 .7183E 07 7.144 .9902E 02 2.004 .2155 .5471E+04 3.738 .1028E 06 6.988 .1194E 01 1.923 .2221 .4772E+04 3.679 .1455E 06 6.837 .1431E 01 1.844 .2286 .4179E+04 3.621 .2039E 06 6.691 .1707E 01 1.768 .2351 .3673E+04 3.565 .2831E 06 6.548 .2027E 01 1.693 .2416 3240E+04 3.510 .3895E 06 6.409 .2395E 01 1.621 .2482 .2866E+04 3.457 .5315E 06 6.275 .2817E 01 1.550 .2547 .2544E+04 3.406 .7194E 06 6.143 .3300E 01 1.481 .2612 .2265E+04 3.355 .9665E 06 6.015 .3852E 01 1.414 .2678 .2022E+04 3.306 .1289E 05 5.890 .4478E 01 1.349 .2743 .1809E+04 3.258 .1708E 05 5.768 .5189E 01 1.285 .2808 .1623E+04 3.210 .2248E 05 5.648 .5993E 01 1.222 .2874 .1460E+04 3.164 .2940E 05 5.532 .6899E 01 1.161 .2939 .1316E+04 3.119 .3823E 05 5.418 .7919E 01 1.101 .3004 .1189E+04 3.075 .4943E 05 5.306 .9064E 01 1.043 .3070 .1076E+04 3.032 .6358E 05 5.197 .1035E+00 .985 .3135 .9755E+03 2.989 .8137E 05 5.090 .1178E+00 .929 .3200 .8862E+03 2.948 .1036E 04 4.985 .1338E+00 .874 .3266 .8064E+03 2.907 .1313E 04 4.882 .15 16E+00 .819 .3331 .7350E+03 2.866 .1657E 04 4.781 .1715E+00 .766 .3396 .6709E+03 2.827 .2083E 04 4.681 .1935E+00 .713 .3461 .6133E+03 2.788 .2606E 04 4.584 .2179E+00 .662 .3527 .561 4E+03 2.749 .3249E 04 4.488 .2450E+00 .611 .3592 .5146E+03 2.711 .4035E 04 4.394 .2750E+00 .561 .3657 .4723E+03 2.674 .4994E 04 4.302 .3081E+00 .511

PAGE 246

246 .3723 .4339E+03 2.637 .6159E 04 4.210 .3446E+00 .463 .3788 .3990E+03 2.601 .7572E 04 4.121 .3850E+00 .415 .3853 .3674E+03 2.565 .9281E 04 4.032 .4294E+00 .367 .3919 .3385E+03 2.530 .1134E 03 3.945 .4785E+00 .320 .3984 .3122E+03 2.494 .1382E 03 3.859 .5324E+00 .274 .4049 .2881E+03 2.460 .1680E 03 3.775 .5919E+00 .228 .4115 .2661E+03 2.425 .2036E 03 3.691 .6573E+00 .182 .4180 .2459E+03 2.391 .2463E 03 3.609 .7294E+00 .137 .4245 .2273E+03 2.357 .2971E 03 3.527 .8087E+00 .092 .4310 .2103E+03 2.323 .3578E 03 3.446 .8960E+00 .048 .4376 .1946E+03 2.289 .4299E 03 3.367 .9921E+00 .003 .4441 .1801E+03 2.255 .5156E 03 3.288 .1098E+01 .041 .4506 .1667E+03 2.222 .6173E 03 3.210 .1215E+01 .085 .4572 .1543E+03 2.188 .7378E 03 3.132 .1344E+01 .128 .4637 .1429E+03 2.155 .8806E 03 3.055 .1486E+01 .172 .4702 .1323E+03 2.121 .1050E 02 2.979 .1643E+01 .216 .4768 .1224E+03 2.088 .1249E 02 2.903 .1817E+01 .259 .4833 .1132E+03 2.054 .1486E 02 2.828 .2010E+01 .303 .4898 .1047E+03 2.020 .1765E 02 2.753 .2225E+01 .347 .4964 .9677E+02 1.986 .2096E 02 2.679 .2464E+01 .392 .5029 .8935E+02 1.951 .2486E 02 2.604 .2730E+01 .436 .5094 .8241E+02 1. 916 .2949E 02 2.530 .3028E+01 .481 .5160 .7592E+02 1.880 .3497E 02 2.456 .3362E+01 .527 .5225 .6984E+02 1.844 .4147E 02 2.382 .3739E+01 .573 .5290 .6413E+02 1.807 .4919E 02 2.308 .4165E +01 .620 .5355 .5877E+02 1.769 .5838E 02 2.234 .4648E+01 .667 .5421 .5372E+02 1.730 .6934E 02 2.159 .5200E+01 .716 .5486 .4896E+02 1.690 .8247E 02 2.084 .5835E+01 .766 .5551 .4447E+ 02 1.648 .9825E 02 2.008 .6568E+01 .817 .5617 .4023E+02 1.604 .1173E 01 1.931 .7424E+01 .871 .5682 .3620E+02 1.559 .1404E 01 1.853 .8431E+01 .926 .5747 .3238E+02 1.510 .1687E 01 1.773 .9630E+01 .984 .5813 .2874E+02 1.458 .2036E 01 1.691 .1108E+02 1.044 .5878 .2527E+02 1.403 .2472E 01 1.607 .1285E+02 1.109 .5943 .2194E+02 1.341 .3022E 01 1.520 .1507E+02 1.178 .6009 1875E+02 1.273 .3732E 01 1.428 .1793E+02 1.254 .6074 .1566E+02 1.195 .4669E 01 1.331 .2174E+02 1.337 .6139 .1266E+02 1.102 .5951E 01 1.225 .2706E+02 1.432 .6204 .9707E+01 .987 .7800E 01 1.1 08 .3506E+02 1.545 .6270 .6763E+01 .830 .1072E+00 .970 .4871E+02 1.688 .6335 .3724E+01 .571 .1634E+00 .787 .7901E+02 1.898 .6368 .2080E+01 .318 .2239E+00 .650 .1189E+03 2.075 .6384 .1170E+01 .068 .2858E+00 .544 .1693E+03 2.229 .6394 .5493E+00 .260 .3662E+00 .436 .2535E+03 2.404 .6400 .8287E 01 1.082 .5460E+00 .263 .5685E+03 2.755 .6400 .1253E 01 1.902 .6845E+00 .165 .1077E+04 3.032 .6400 .0000E+00 .1000E+01 .000 End of problem ==============

PAGE 247

247 APPENDIX O RESULTS OF FITTING OF MULTIPLE REGRESSION MODELS FOR PREDICTION OF P AVAILABILITY AFTER THE FIRE Model 1: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 3 823.368310 274.456103 94.26 <.0001 Error 212 617.302924 2. 911806 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.5 71517 26.65482 1.706402 6.401850 So urce DF Type III SS Mean Square F Value Pr > F Day 1 107.0590408 107.0590408 36.77 <.0001 Moisture 1 137.8747888 137.8747888 47.35 <.0001 Temperature 1 578.4344802 578.4344802 198.65 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 15.34402853 0.55156361 27.82 <.0001 Day 0.01253608 0.00206743 6.06 <.0001 Moisture 7.13426676 1.03678 368 6.88 <.0001 Temperature 0.01463677 0.00103848 14.09 <.0001 Model 2: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 6 890.327067 148. 387845 56.35 <.0001 Error 209 550.344167 2.633226 Corrected Total 215 144 0.671234 R S quare Coeff Var Root MSE PO4_P Mean 0.617995 25.34770 1.622722 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day 1 20.4034436 20.4034436 7.75 0.0059 Moisture 1 117.5023736 117.5023736 44.62 <.0001 Temperature 1 191.0766954 191.0766954 72.56 <.0001 Day*Moisture 1 13.1956077 13.1956077 5.01 0.0262 Day*Temperature 1 0.0403640 0.0403640 0.02 0.9016 Moisture*Temperature 1 53.7227858 53.7227858 20.40 <.0001

PAGE 248

248 Standard Parameter Estimate Error t Value Pr > |t| Intercept 20.93804965 1.34782249 15.53 <.0001 Day 0.02472916 0.00888386 2.78 0.0059 Moisture 26.30277588 3.93751576 6.68 <.0001 Temperature 0.02629133 0.00308640 8.52 <.0001 Day*Moisture 0.03930053 0.01755609 2.24 0.0262 Day*Temperature 0.00000218 0.00001758 0.12 0.9016 Moisture*Temperature 0.03983191 0.00881852 4.52 <.0001 Model 3: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 5 890.28 6703 178.057341 67.94 <.0001 Error 210 550.384531 2.620879 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.617967 25.28820 1.618913 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day 1 50.9016376 50.9016376 19.42 <.0001 Moisture 1 117.5023736 117.5023736 44.83 <.0001 Temperature 1 241.6569864 241.6569864 92.20 <.0001 Day*Moisture 1 13.1956077 13.1956077 5. 03 0.0259 Moisture*Temperature 1 53.7227858 53.7227858 20.50 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 20.86634822 1.21420373 17.19 <.0001 Day 0.02385830 0.00541374 4.41 <.0001 Moisture 26.30277588 3.92827359 6.70 <.0001 Temperature 0.02611208 0.00271935 9.60 <.0001 Day*Moisture 0.03930053 0.0175148 8 2.24 0.0259 Moisture*Temperature 0.03983191 0.00879783 4.53 <.0001 Model 4: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 7 891.880275 127.411468 48.29 <.0001 Error 208 548.790959 2.638418 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.619073 25.3726 8 1.624321 6.401850 Source DF Type III SS Mean Square F Value Pr > F

PAGE 249

249 Day 1 9.66825839 9.66825839 3.66 0.0570 Moisture 1 57.12702555 57.12702555 21.65 <.0001 Temperature 1 95.63806239 95.63806239 36.25 <.0001 Day*Moisture 1 4.74425117 4.74425117 1.80 0.1814 Day*Temperature 1 1.52372996 1.52372996 0.58 0.4481 Moisture*Temperature 1 26.62311568 26.62311568 10.09 0.0017 Day*Moistur*Temperat 1 1.55320811 1.55 320811 0.59 0.4438 Standard Parameter Estimate Error t Value Pr > |t| Intercept 22.08227729 2.01102670 10.98 <.0001 Day 0.03862667 0.02017832 1.91 0.0570 Moisture 30.27450141 6.50620885 4.65 <.0001 Temperature 0.02915190 0.00484198 6.02 <.0001 Day*Moisture 0.08754011 0.06528224 1.34 0.1814 Day*Temperature 0.00003692 0.00004858 0.76 0.4481 Moisture*Temperature 0.04976123 0.01566511 3.18 0.0017 Day*Moistur*Temperat 0.00012060 0.00015718 0.77 0.4438 Model 5: The GLM Procedure Number of Observa tions Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 4 877.091096 219.272774 82.09 <.0001 Error 211 563.580138 2.670996 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.608807 25.52884 1.634318 6.401850 Source DF Typ e III SS Mean Square F Value Pr > F Day 1 107.0590408 107.0590408 40.08 <.0001 Moisture 1 104.4456910 104.4456910 39.10 <.0001 Temperature 1 241.6569864 241.6569864 90.47 <.0001 Mo isture*Temperature 1 53.7227858 53.7227858 20.11 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 19.93415203 1.15177493 17.31 <.0001 Day 0.01253608 0.00198010 6.33 <.0001 Moisture 23.06703217 3.68878268 6.25 <.0001 Temperature 0.02611208 0.00274523 9.5 1 <.0001 Moisture*Temperature 0.03983191 0.00888154 4.48 <.0001 Model 6: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 3 945.315822 315.105274 134.86 <.0001

PAGE 250

250 Error 212 495.355412 2.336582 Corrected Total 215 1440.671234 R Square Coeff Va r Root MSE PO4_P Mean 0.656163 23.87729 1.528588 6.401850 Source DF Type III SS Mean Square F Value Pr > F Da y*Day 1 142.0247409 142.0247409 60.78 <.0001 Moisture*Moisture 1 138.8002245 138.8002245 59.40 <.0001 Temperatu*Temperatur 1 664.4908564 664.4908564 284.39 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 11.68408692 0.28861199 40.48 <.0001 Day*Day 0.00007279 0.00000934 7.8 0 <.0001 Moisture*Moisture 12.53415352 1.62626176 7.71 <.0001 Temperatu*Temperatur 0.00001949 0. 00000116 16.86 <.0001 Model 7: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 6 1070.312331 178.385388 100.67 <.0001 Error 209 370.358903 1.772052 Corrected Total 215 1440.671234 R Square Coeff Va r Root MSE PO4_P Mean 0.742926 20.79375 1.331184 6.401850 Source DF Type III SS Mean Square F Value Pr > F Da y*Day 1 85.4132351 85.4132351 48.20 <.0001 Moisture*Moisture 1 170.5524952 170.5524952 96.25 <.0001 Temperatu*Temperatur 1 344.8446432 344.8446432 194.60 <.0001 Day*Moisture 1 10.6065295 10.6065295 5.99 0.0153 Day*Temperature 1 9.6659175 9.6659175 5.45 0.0205 Moisture*Temperature 1 86.1242099 86.1242099 48.60 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 12.45439184 0. 26756274 46.55 <.0001 Day*Day 0.00016818 0.00002422 6.94 <.0001 Moisture*Moisture 4 7.48169427 4.83988921 9.81 <.0001 Temperatu*Temperatur 0.00003781 0.00000271 13.95 <.0001 Day*Moist ure 0.03120982 0.01275683 2.45 0.0153 Day*Temperature 0.00002647 0.00001133 2.34 0.0205 Moisture*Temperature 0.04394971 0.00630422 6.97 <.0001 Model 8: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P

PAGE 251

251 Sum of Source DF Squares Mean Square F Value Pr > F Model 7 1073.297359 153.328194 86.81 <.0001 Error 208 367.373875 1.766221 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.744998 20.75951 1.328992 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day*Day 1 75.4488083 75.4488083 42.72 <.0001 Moisture*Moisture 1 157.0445674 157.0445674 88.92 <.0001 Temperatu*Temperatur 1 331.4466333 331.44 66333 187.66 <.0001 Day*Moisture 1 10.0224914 10.0224914 5.67 0.0181 Day*Temperature 1 11. 4659949 11.4659949 6.49 0.0116 Moisture*Temperature 1 73.4397442 73.4397442 41.58 <.0001 Day*Moistur*Temperat 1 2.9850282 2.9850282 1.69 0.1950 Standard Parameter Estimate Error t Value Pr > |t| Intercept 12.12467517 0.36834631 32.92 <.0001 Day*Day 0.00018867 0.00002887 6.54 <.0001 Moisture*Moisture 50.47681190 5.35307004 9.43 <.0001 Temperatu*Temperatur 0.00003894 0.00000284 13.70 <.0001 Day*Moisture 0.05851896 0.02456582 2.38 0.0181 Day*Temperature 0.00004217 0.0000165 5 2.55 0.0116 Moisture*Temperature 0.04975136 0.00771546 6.45 <.0001 Day*Moistur*Temperat 0.00008 630 0.00006639 1.30 0.1950 Model 9: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 3 1018.225111 339.408370 170.33 <.0001 Error 212 422.446123 1.992670 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.706771 22.05019 1.411620 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day*Day*Day 1 156.1173849 156.1173849 78.35 <.0001 Moistu*Moistu*Moistu 1 135.4036843 135.4036843 67.95 <.0001 Temper*Temper*Temper 1 726.7040422 726.7040422 364.69 <.0001 Standard Parameter Estimate Erro r t Value Pr > |t| Intercept 10.43604988 0.20863393 50.02 <.0001 Day*Day*Day 0.00000 041 0.00000005 8.85 <.0001 Moistu*Moistu*Moistu 26.43362465 3.20670650 8.24 <.0001 Temper*Temper*Te mper 0.00000003 0.00000000 19.10 <.0001

PAGE 252

252 Model 10: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 6 1094.609720 182.434953 110.18 <.0001 Error 209 346.061514 1.655797 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.759791 20.10009 1.286778 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day*Day*Day 1 79.8725036 79.8725036 48.24 <.0001 Moistu*Moistu*Moistu 1 126.2504032 126.2504032 76.25 <.0001 Temper*Temper*Temper 1 355.4645854 355.4645854 214.68 <.0001 Day*Moisture 1 0.4243149 0.4 243149 0.26 0.6132 Day*Temperature 1 6.8894035 6.8894035 4.16 0.0426 Moisture*Temperature 1 56 .2336509 56.2336509 33.96 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 9.67979570 0.22142435 43.72 <.0001 Day*Day*Day 0.00000065 0.00000009 6.95 <.0001 Moistu*Moistu*Moistu 69.21044644 7.92608585 8.73 <.0001 Temper*Temper*Temper 0.00000005 0.00000000 14.65 <.0001 Day*Moisture 0.00579373 0.01144506 0.51 0.6132 Day*Temperature 0.00001930 0.00000946 2.04 0.0426 Moisture*Temperature 0.02837741 0.00486943 5.83 <.0001 Model 11: The G LM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 5 1094.185405 218.837081 132.63 <.0001 Error 210 346.485829 1.649933 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.759497 20.06447 1.284497 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day*Day*Day 1 84.3509834 84.3509834 51.12 <.0001 Moistu*Moistu*Moistu 1 134.0385060 134.0385060 81.24 <.0001 Te mper*Temper*Temper 1 417.0538749 417.0538749 252.77 <.0001 Day*Temperature 1 13.1976723 13.1976723 8.00 0.0051 Moisture*Temperature 1 58.8949651 58.8949651 35.70 <.0001 Standard Parameter Estimate Error t Value Pr > |t|

PAGE 253

253 Intercept 9.67340578 0.22067246 43.84 <.0001 Day*Day*Day 0.00000063 0.00000009 7.1 5 <.0001 Moistu*Moistu*Moistu 68.00201588 7.54466722 9.01 <.0001 Temper*Temper*Temper 0.00000005 0. 00000000 15.90 <.0001 Day*Temperature 0.00002201 0.00000778 2.83 0.0051 Moisture*Temperature 0.02873462 0.00480949 5.97 <.0001 Model 12: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Square s Mean Square F Value Pr > F Model 7 1097.596108 156.799444 95.06 <.0001 Error 208 343.075126 1.649400 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.761864 20.06123 1.284290 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day*Day*Day 1 59.4819993 59.4819993 36.06 <.0001 Moistu*Moistu*Moistu 1 104.7684755 104.7684755 63.52 <.0001 Temper*Temper*Temper 1 342.0956488 342.0956488 207.41 <.0001 Day*Temperature 1 0.6711644 0.6711644 0.41 0.5242 Moisture*Temperature 1 29.6361113 29.6 361113 17.97 <.0001 Day*Moisture 1 1.0532899 1.0532899 0.64 0.4251 Day*Moistur*Temperat 1 2 .9863880 2.9863880 1.81 0.1799 Standard Parameter Estimate Error t Value Pr > |t| Intercept 10.08294676 0.37229831 27.08 <.0001 Day*Day*Day 0.00000060 0.00000010 6.01 <.0001 Moistu*Moistu*Moistu 65.94604742 8.27439800 7.97 <.0001 Temper*Temper*Temper 0.00000005 0.00000000 14.40 <.0001 Day*Temperature 0.00000805 0.00001261 0.64 0.5242 Moisture*Temperature 0.02429144 0.00573067 4.24 <.0001 Day*Moisture 0.01568610 0.01962926 0.80 0.4251 Day*Moistur*Temperat 0.00007592 0.0000564 2 1.35 0.1799 Model 13: The GLM Procedure Number of Observations Read 216 Number of Observations Used 2 16 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 4 1080.987733 270.246933 158.53 <.0001 Error 211 359.683501 1.704661 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.750336 20.39452 1.305627 6.401850

PAGE 254

254 Source DF Type III SS Mean Square F Value Pr > F Day*Day*Day 1 156.1173849 156.1173849 91 .58 <.0001 Moistu*Moistu*Moistu 1 139.5177658 139.5177658 81.84 <.0001 Temper*Temper*Temper 1 414.2182889 414.2182889 242.99 <.0001 Moisture*Temperature 1 62.7626218 62.7626218 36.82 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 9.79398200 0.22007666 44.50 <.0001 Day*Day*Day 0.00000041 0.00000004 9.57 <.0001 Moistu*Moistu*Moistu 69.25775473 7.6554845 1 9.05 <.0001 Temper*Temper*Temper 0.00000005 0.00000000 15.59 <.0001 Moisture*Temperature 0.02960 266 0.00487865 6.07 <.0001 Model 14: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 6 1165.443025 194.240504 147.50 <.0001 Error 209 275.228210 1 .316881 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.808958 17.92536 1.147555 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day 1 20.8431431 20.8431431 15.83 <.0001 Moisture 1 0.2778930 0.2778930 0.21 0.6464 Temperature 1 199.0061665 199.0061665 151.12 <.0001 Day*Day 1 55.8088432 55.8088432 42.38 <.0001 Moisture*Moisture 1 1.2033288 1.2 033288 0.91 0.3402 Temperatu*Temperatur 1 285.0625427 285.0625427 216.47 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 3.214021277 1.35954371 2.36 0.0190 Day 0.021556683 0.00541843 3.98 <.0001 Moisture 2.330572324 5.07337626 0.46 0.6464 Temperature 0.077266972 0.00628541 12.29 <.0001 Day*Day 0.000177825 0.00002732 6.51 <.0001 Moisture*Moisture 8.491978829 8.88362226 0.96 0.3402 Temperatu*Temperatur 0.000114880 0.00000781 14.71 <.0001 Model 15: The GLM Procedure Number of Observ ations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 4 1026.364907 256.591227 130.68 <.0001

PAGE 255

255 Error 211 414.306327 1.963537 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.712421 21.88840 1.401263 6.401850 Source DF Typ e III SS Mean Square F Value Pr > F Day 1 20.8431431 20.8431431 10.62 0.0013 Temperature 1 199.0061665 199.0061665 101.35 <.0001 Day*Day 1 55.8088432 55.8088432 28.42 <.0001 Te mperatu*Temperatur 1 285.0625427 285.0625427 145.18 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 4.696756889 1.46997719 3.20 0.0016 Day 0.021556683 0.00661637 3.26 0.0013 Temperature 0.077266972 0.00767503 10.0 7 <.0001 Day*Day 0.000177825 0.00003335 5.33 <.0001 Temperatu*Temperatur 0.000114880 0. 00000953 12.05 <.0001 Model 16: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 9 1232.401782 136.933531 135.44 <.0001 Error 206 208.269452 1. 011017 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.8 55436 15.70629 1.005493 6.401850 So urce DF Type III SS Mean Square F Value Pr > F Day 1 1.7259669 1.7259669 1.71 0.1928 Moisture 1 18.4410492 18.4410492 18.24 <.0001 Temperature 1 129.4842234 129.4842234 128.07 <.0001 Day*Day 1 55.8088432 55.8088432 55.20 <.0001 Moisture*Moisture 1 1.2033288 1.2033288 1.19 0.2766 Temperatu*Temperatur 1 285.0625427 285.0625427 281.96 <.0001 Day*Moisture 1 13.1956077 13.1956077 13.05 0.0004 Day*Temperature 1 0.0403640 0.0403640 0.04 0.8418 Moisture*Temperature 1 53.7227858 53.7227858 53.14 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 2.37999984 1.41806469 1.68 0.0948 Day 0.00936359 0.00716647 1.31 0.1928 Moisture 21.49908144 5.03391953 4.27 <.0001 Temperature 0.06561241 0.00579771 11.32 <.0001 Day*Day 0.00017782 0.00002393 7.43 <.0001 Moisture*Moisture 8.49197883 7.78387640 1.09 0.2766 Temperatu*Temperatur 0.00011488 0.00000684 16.79 <.0001 Day*Moisture 0.03930053 0.01087835 3.61 0.0004

PAGE 256

256 Day*Temperature 0.00000218 0.00001090 0.20 0.8418 Moisture*Temperature 0.03983191 0.0054642 6 7.29 <.0001 Model 17: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 6 1227.885799 204.647633 201.01 <.0001 Error 209 212.785435 1.018112 Corrected Total 215 1440.671234 R Square Coeff Var Ro ot MSE PO4_P Mean 0.852301 15.76131 1.009015 6.401850 Source DF Type III SS Mean Square F Value Pr > F Moisture 1 130.9220413 130.9220413 128.59 <.0001 Temperature 1 133.3864201 133.3864201 131.01 <.0001 Day*Day 1 103.4381905 103.4381905 101.60 <.0001 Temperatu*Temperatur 1 285.0625427 285.0625427 279.99 <.0001 Moisture*Day 1 30.7664605 30.7664605 30.22 <.0001 Moisture*Temperature 1 53.7227858 53.7227858 52.77 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 3.37854726 1.2413249 8 2.72 0.0070 Moisture 27.19046962 2.39777336 11.34 <.0001 Temperature 0.06579 166 0.00574795 11.45 <.0001 Day*Day 0.00014309 0.00001420 10.08 <.0001 Temperatu*Temper atur 0.00011488 0.00000687 16.73 <.0001 Moisture*Day 0.05008224 0.00911051 5.50 <.0001 Moisture*Temperature 0.03983191 0.00548340 7.26 <.0001 Model 18: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 10 12 33.954990 123.395499 122.37 <.0001 Error 205 206.716244 1.008372 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.856514 15.68574 1.004177 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day 1 0.1173653 0.1173653 0.12 0.7333 Moisture 1 18.4200832 18.4 200832 18.27 <.0001 Temperature 1 102.2390278 102.2390278 101.39 <.0001 Day*Day 1 55 .8088432 55.8088432 55.35 <.0001 Moisture*Moisture 1 1.2033288 1.2033288 1.19 0.2759

PAGE 257

257 Temperatu*Temperatur 1 285.0625427 285.0625427 282.70 <.0001 Day*Moisture 1 4.7442512 4.7442512 4.70 0.0312 Da y*Temperature 1 1.5237300 1.5237300 1.51 0.2204 Moisture*Temperature 1 26.6231157 26.6231157 26.40 <.0001 Day*Moistur*Temperat 1 1.5532081 1.5532081 1.54 0.2160 Standard Parameter Estimate Error t Value Pr > |t| Intercept 3.52422749 1.68986423 2.09 0.0383 Day 0.00453391 0.01328964 0.3 4 0.7333 Moisture 25.47080697 5.95946512 4.27 <.0001 Temperature 0.06275184 0. 00623201 10.07 <.0001 Day*Day 0.00017782 0.00002390 7.44 <.0001 Moisture*Moisture 8.49197883 7.77368836 1.09 0.2759 Temperatu*Temperatur 0.00011488 0.00000683 16.81 <.0001 Day*Moist ure 0.08754011 0.04035837 2.17 0.0312 Day*Temperature 0.00003692 0.00003004 1.23 0.2204 Moisture*Temperature 0.04976123 0.00968438 5.14 <.0001 Day*Moistur*Temperat 0.00012060 0.00009717 1.24 0.2160 Model 19: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 8 1234.106902 154.263363 154.59 <.0001 Error 207 206.564332 0.997895 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.856619 15.60404 0.998947 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day 1 0.75341303 0.75341303 0.76 0.3859 Moisture 0 0.00000000 Temperature 1 97.19754369 97.19754369 97.40 <.0001 Day*Day 1 0. 13954545 0.13954545 0.14 0.7088 Moisture*Moisture 0 0.00000000 Temperatu*Temperatur 1 84.60988180 84.60988180 84.79 <.0001 Day*Day*Day 1 0.28931306 0.28931306 0.29 0.5908 Moistu*Moistu*Moistu 0 0.00000000 Temper*Temper*Temper 1 68.37456463 68.37456463 68.52 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 49.88253197 B 5.78155920 8.63 <.0001 Day 0.01355201 0.01559659 0.87 0.3859 Moisture 2.33057232 B 4.41637799 0.53 0.5983 Temperature 0.46265629 0.04687846 9.87 <.0001 Day*Day 0.00007320 0.00019575 0.37 0.7088 Moisture*Moisture 8.49197883 B 7.73320010 1.10 0.2734 Temperatu*Temperatur 0.00112134 0.00012178 9.21 <.0001 Day*Day*Day 0.00000036 0.00000067 0.54 0.5908 Moistu*M oistu*Moistu 0.00000000 B

PAGE 258

258 Temper*Temper*Temper 0.00000084 0.00000010 8.28 <.0001 The X'X matrix has been found to be singular, and a generalized inverse was used to solve the normal equations. Terms who se estimates are followed by the letter 'B' are not uniquely estimable. Model 20: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 8 1234.106902 154.263363 154.59 <.0001 Error 207 206.564332 0.997895 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.856619 15.60404 0.998947 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day 1 0. 75341303 0.75341303 0.76 0.3859 Moisture 1 0.27789305 0.27789305 0.28 0.5983 Temperature 1 97.19754369 97.19754369 97.40 <.0001 Day*Day 1 0.13954545 0.13954545 0.14 0.7088 Mo isture*Moisture 1 1.20332876 1.20332876 1.21 0.2734 Temperatu*Temperatur 1 84.60988180 84.60988180 84.79 <.0001 Day*Day*Day 1 0.28931306 0.28931306 0.29 0.5908 Temper*Temper*Temper 1 68.37456463 68.37456463 68.52 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 49.88253197 5.78155920 8.6 3 <.0001 Day 0.01355201 0.01559659 0.87 0.3859 Moisture 2.33057232 4. 41637799 0.53 0.5983 Temperature 0.46265629 0.04687846 9.87 <.0001 Day*Day 0.00007320 0.00019575 0.37 0.7088 Moisture*Moisture 8.49197883 7.73320010 1.10 0.2734 Temperatu *Temperatur 0.00112134 0.00012178 9.21 <.0001 Day*Day*Day 0.00000036 0.00000067 0.54 0.5908 Temper*Temper*Temper 0.00000084 0.00000010 8.28 <.0001 Model 21: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 3 931.871588 310.623863 129.43 <.0001 Error 212 508.799646 2.399998 Correcte d Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.646832 24.19914 1.549193 6.401850

PAGE 259

259 Source DF Type III SS Mean Square F Value Pr > F Temperature 1 97.19754369 97.19754369 40.50 <.0001 Temperatu*Temperatur 1 84.60988180 84.60988180 35.25 <.0001 Temper*Temper*Temper 1 68.37456463 68.37456463 28.49 <.0001 Standard Parameter Esti mate Error t Value Pr > |t| Intercept 51.48846470 8.91350881 5.78 <.0001 Temperature 0.46265629 0.07270032 6.36 <.0001 Temperatu*Temperatur 0.00112134 0.00018886 5.94 <.0001 Temper*Temper*Temper 0.00000084 0.00000016 5.34 <.0001 Model 22: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 11 1301.065660 118.278696 172.84 <.0001 Error 204 139.60557 4 0.684341 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.903097 12.92203 0.827249 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day 1 0.00678248 0.00678248 0.01 0.9208 Moisture 1 18.44104915 18.44104915 26.95 <.0001 Temperature 1 92.22631556 92.22631556 134.77 <.0001 Day*Day 1 0.13954545 0.13954545 0.20 0.6521 Moisture*Moisture 1 1.20332876 1.20332876 1.76 0.1863 Temperatu*Temperatur 1 84.60988180 84.60988180 123.64 <.0001 Day*Day*Day 1 0.28931306 0.28 931306 0.42 0.5163 Temper*Temper*Temper 1 68.37456463 68.37456463 99.91 <.0001 Day*Moisture 1 13.19560773 13.19560773 19.28 <.0001 Day*Temperature 1 0.04036402 0.04036402 0.06 0.8084 Moisture*Temperature 1 53.72278578 53.72278578 78.50 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 44.28851086 4.82948654 9.17 <.0001 Day 0.00135892 0.0136501 3 0.10 0.9208 Moisture 21.49908144 4.14155457 5.19 <.0001 Temperature 0.45100 173 0.03884965 11.61 <.0001 Day*Day 0.00007320 0.00016211 0.45 0.6521 Moisture*Moistur e 8.49197883 6.40402547 1.33 0.1863 Temperatu*Temperatur 0.00112134 0.00010085 11.12 <.0001 Day*Day*Day 0.00000036 0.00000055 0.65 0.5163 Temper*Temper*Temper 0.00000084 0.00000008 10.00 <.0001 Day*Moisture 0.03930053 0.00894994 4.39 <.0001 Day*Temperature 0.00000218 0.00000896 0.24 0.8084 Moisture*Temperature 0.03983191 0.00449561 8.86 <.0001

PAGE 260

260 Model 23: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squares Mean Square F Value Pr > F Model 6 1192.822173 198.803696 167.64 <.0001 Error 209 247.849061 1.185881 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.827963 17.01042 1.088981 6.401850 Source DF Typ e III SS Mean Square F Value Pr > F Moisture 1 79.89154185 79.89154185 67.37 <.0001 Temperature 1 92.33296905 92.33296905 77.86 <.0001 Temperatu*Temperatur 1 84.60988180 84.60988180 71.35 <.0001 Te mper*Temper*Temper 1 68.37456463 68.37456463 57.66 <.0001 Moisture*Day 1 69.35301091 69.35301091 58.48 <.0001 Moisture*Temperature 1 53.72278578 53.72278578 45.30 <.0001 Standard Parameter Estimate Error t Value Pr > |t| Intercept 44.84300652 6.30550181 7.11 <.0001 Moisture 20.37940853 2.48291396 8.2 1 <.0001 Temperature 0.45118098 0.05113201 8.82 <.0001 Temperatu*Temperatur 0.00112134 0. 00013275 8.45 <.0001 Temper*Temper*Temper 0.00000084 0.00000011 7.59 <.0001 Moisture*Day 0.03264320 0.00426855 7.65 <.0001 Moisture*Temperature 0.03983191 0.00591797 6.73 <.0001 Model 24: The GLM Procedure Number of Observations Read 216 Number of Observations Used 216 Dependent Variable: PO4_P PO4 P Sum of Source DF Squar es Mean Square F Value Pr > F Model 12 1302.618868 108.551572 159.62 <.0001 Error 203 138.052366 0.680061 Corrected Total 215 1440.671234 R Square Coeff Var Root MSE PO4_P Mean 0.904175 12.88156 0.824658 6.401850 Source DF Type III SS Mean Square F Value Pr > F Day 1 0.39638915 0.39638915 0.58 0.4461 Moisture 1 18.42008325 18.42008325 27.09 <.0001 Temperature 1 90.84309497 90.84309497 133.58 <.0001 Day*Day 1 0.13954545 0.13954545 0.21 0.6510 Moisture*Moisture 1 1.20332876 1.2 0332876 1.77 0.1849 Temperatu*Temperatur 1 84.60988180 84.60988180 124.42 <.0001 Day*Day*Day 1 0 .28931306 0.28931306 0.43 0.5150 Temper*Temper*Temper 1 68.37456463 68.37456463 100.54 <.0001

PAGE 261

2 61 Day*Moisture 1 4.74425117 4.74425117 6.98 0.0089 Day*Temperature 1 1.52372996 1.52372996 2.24 0.1360 Moisture*Temperature 1 26.62311568 26.62311568 39.15 <.0001 Day*Moistur*Temperat 1 1.55320811 1.55320811 2.28 0.1323 Standard Parameter Estimate Error t Value Pr > |t| Intercept 43.14428321 4.87353178 8.85 <.0001 Day 0.01253858 0.0164233 3 0.76 0.4461 Moisture 25.47080697 4.89407723 5.20 <.0001 Temperature 0.44814 116 0.03877419 11.56 <.0001 Day*Day 0.00007320 0.00016160 0.45 0.6510 Moisture*Moistur e 8.49197883 6.38396743 1.33 0.1849 Temperatu*Temperatur 0.00112134 0.00010053 11.15 <.0001 Day*Day*Day 0.00000036 0.00000055 0.65 0.5150 Temper*Temper*Temper 0.00000084 0.00000008 10.03 <.0001 Day*Moisture 0.08754011 0.03314341 2.64 0.0089 Day*Temperature 0.00003692 0.00002467 1.50 0.1360 Moisture*Temperature 0.04976123 0.00795308 6.26 <.0001 Day*Moistur*Temperat 0.00012060 0.00007980 1.51 0.1323

PAGE 262

262 APPENDIX P CHECKING OF ASSUMPTIONS F OR THE SELECTED PREDICTIVE MODEL Independence of Errors: Durbin Watson Test The REG Procedure Dependent Variable: PO4 _P PO4 P Collinearity Diagnostics --------------------------------Proportion of Variation --------------------------------Number Intercept Tempe rature Moisture Moist_Temp Moist_Day Day_Day Temp_Temp 1 0. 00008524 0.00002564 0.00022310 0.00024555 0.00145 0.00195 0.00008093 2 0.00010072 0.00007496 0.00016935 0.00053986 0.02072 0.08372 0.00041671 3 0.00012887 0.00038304 0.01015 0.00401 0.03007 0.06736 0.00317 4 0.01372 0.00006541 0.00245 0.01229 0.02478 0.00253 0.00565 5 0.00236 0.00056044 0.01966 0.04342 0.87712 0.80488 0.00185 6 0.06632 0.00155 0.89584 0.85790 0.04576 0.03948 0.10901 7 0.91730 0.99734 0.07150 0.08159 0.00009101 0.00007800 0.87983 The REG Procedure Dependent Variable: PO4_P PO4 P Durbin Watson D 1.410 Number of Observations 216 1st Order Autocorrelation 0.269 Figure P 1. Plot of predicted values versus residuals

PAGE 263

263 Figure P 2. Normal probability plot of residuals

PAGE 264

264 APPENDIX Q ROCKLAND SOIL S AFTER THE FIRE Where: ET: Evapotranspiration; FC: field capacity; and S WC: soil water content 0.421 is the volumetric water content (v/v) at fie ld capacity of the Pine Rockland s oil estimated from the soil moisture curve at suction of 1/3 bar ( 33.3 cbar or 33.3 kPa ). 2.1049 is the water amount (cm) of field capacity at 5cm topsoil (0.421 x 5 cm = 2.1049 cm) Days after fire Date Rainfall (cm) ET (cm) FC (cm) SWC (cm) FC at 5cm topsoil (%) Adjusted FC at 5 cm topsoil (%) Volumetric water content (v/v) a b c = 2.1049 if a > 2.1049 c = a if a < 2.1049 d = c b e = 100*d/ 5cm f(1) = e if c = 2.1 f(1) = e + 100*(c b)/5cm if c < 2.1 f(2) = f1 +100*(c1 b1)/5cm f(3) = f2 + 100*(c2 b2)/5cm g = f*0.421/100 4/20/10 0.4064 0.30 0.4064 0.1016 2.03 0 4/21/10 0.1524 0.33 0.1524 0 0 0 4/22/10 0 0.39 0 0 0 0 4/23/10 0 0.28 0 0 0 0 4/24/10 0 0.39 0 0 0 0 4/25/10 0 0.39 0 0 0 0 4/26/10 4.4958 0.18 2.1049 1.9220 38.44 38.44 4/27/10 0 0.39 0 0 0 30.72 4/28/10 0 0.37 0 0 0 23.40 4/29/10 0 0.18 0 0 0 19.75 4/30/10 0.1524 0.30 0.1524 0 0 16.70 5/1/10 0 0.30 0 0 0 10.60 5/2/10 0 0.35 0 0 0 3.69 5/3/10 0 0.37 0 0 0 0 5/4/10 0 0.41 0 0 0 0 Burned 5/5/10 0.0762 0.41 0.0762 0 0 0 0 5/6/10 0.0254 0.28 0.0254 0 0 0 0 5/7/10 0 0.37 0 0 0 0 0 5/8/10 0 0.33 0 0 0 0 0 5/9/10 0 0.45 0 0 0 0 0 5/10/10 0.02 0.43 0.02 0 0 0 0 5/11/10 0.3 0.30 0.3 0 0 0 0 7days 5/12/10 0 0.41 0 0 0 0 0 5/13/10 0 0.41 0 0 0 0 0

PAGE 265

265 5/14/10 0 0.35 0 0 0 0 0 5/15/10 0 0.37 0 0 0 0 0 5/16/10 0 0.30 0 0 0 0 0 5/17/10 0.16 0.30 0.16 0 0 0 0 5/18/10 5.26 0.28 2.1049 1.8204 36.41 36.41 0.153272 14days 5/19/10 0.02 0.20 0.02 0 0 32.74 0.137847 5/20/10 0 0.41 0 0 0 24.62 0.103630 5/21/10 0 0.43 0 0 0 16.08 0.067702 5/22/10 0 0.43 0 0 0 7.55 0.031774 5/23/10 0 0.45 0 0 0 0 0 5/24/10 0 0.39 0 0 0 0 0 5/25/10 6.64 0.33 2.1049 1.7798 35.60 35.60 0.149850 21days 5/26/10 0.56 0.35 0.56 0.2146 4.29 39.89 0.167915 5/27/10 0 0.41 0 0 0 31.76 0.133698 5/28/10 1.04 0.43 1.04 0.6133 12.27 44.02 0.185334 5/29/10 0.2 0.37 0.2 0 0 40.71 0.171378 5/30/10 0 0.37 0 0 0 33.39 0.140582 5/31/10 0 0.39 0 0 0 25.67 0.108076 6/1/10 0.1 0.33 0.1 0 0 21.17 0.089121 30days 6/2/10 0.56 0.28 0.56 0.2755 5.51 26.68 0.112319 6/3/10 0.02 0.43 0.02 0 0 18.55 0.078075 6/4/10 0 0.45 0 0 0 9.61 0.040436 6/5/10 2.22 0.43 2.1049 1.6782 33.56 43.17 0.181732 6/6/10 0.28 0.33 0.28 0 0 42.27 0.177933 6/7/10 0.06 0.39 0.06 0 0 35.74 0.150478 6/8/10 7.72 0.39 2.1049 1.7188 34.38 34.38 0.144718 6/9/10 0.02 0.43 0.02 0 0 26.24 0.110474 6/10/10 0.18 0.41 0.18 0 0 21.71 0.091412 6/11/10 1.24 0.33 1.24 0.9149 18.30 40.01 0.168441 6/12/10 0 0.30 0 0 0 33.92 0.142778 6/13/10 0.06 0.39 0.06 0 0 27.39 0.115323 6/14/10 3.04 0.28 2.1049 1.8204 36.41 36.41 0.153272 6/15/10 0.14 0.39 0.14 0 0 31.49 0.132553 6/16/10 0 0.45 0 0 0 22.55 0.094914 6/17/10 3.62 0.33 2.1049 1.7798 35.60 35.60 0.149850 6/18/10 5.12 0.22 2.1049 1.8814 37.63 37.63 0.158405 6/19/10 0 0.45 0 0 0 28.69 0.120766 6/20/10 0 0.47 0 0 0 19.34 0.081416 6/21/10 0 0.00 0 0 0 19.34 0.081416 6/22/10 0.18 0.47 0.18 0 0 13.59 0.057221 6/23/10 0.16 0.47 0.16 0 0 7.45 0.031343

PAGE 266

266 6/24/10 0 0.47 0 0 0 0 0 6/25/10 0 0.00 0 0 0 0 0 6/26/10 0.46 0.33 0.46 0.1349 2.70 2.70 0.011356 6/27/10 0.04 0.41 0.04 0 0 0 0 6/28/10 0 0.41 0 0 0 0 0 6/29/10 0 0.41 0 0 0 0 0 6/30/10 0.02 0.35 0.02 0 0 0 0 7/1/10 0 0.43 0 0 0 0 0 60days 7/2/10 0.14 0.26 0.14 0 0 0 0 7/3/10 1.12 0.26 1.12 0.8558 17.12 17.12 0.072058 7/4/10 0.72 0.22 0.72 0.4965 9.93 27.05 0.113860 7/5/10 0.8 0.22 0.8 0.5765 11.53 38.58 0.162397 7/6/10 0 0.39 0 0 0 30.85 0.129891 7/7/10 0.2 0.39 0.2 0 0 27.13 0.114224 7/8/10 0.28 0.37 0.28 0 0 25.42 0.107003 7/9/10 0 0.39 0 0 0 17.70 0.074497 7/10/10 0.68 0.33 0.68 0.3549 7.10 24.79 0.104376 7/11/10 0 0.47 0 0 0 15.45 0.065026 7/12/10 0 0.43 0 0 0 6.91 0.029098 7/13/10 3.16 0.33 2.1049 1.7798 35.60 35.60 0.149850 7/14/10 3.12 0.24 2.1049 1.8611 37.22 37.22 0.156694 7/15/10 1.52 0.39 1.52 1.1339 22.68 37.22 0.156694 7/16/10 0.12 0.28 0.12 0 0 33.93 0.142845 7/17/10 0 0.35 0 0 0 27.02 0.113761 7/18/10 3.78 0.35 2.1049 1.7595 35.19 35.19 0.148139 7/19/10 0 0.37 0 0 0 27.87 0.117344 7/20/10 0.96 0.39 0.96 0.5739 11.48 39.35 0.165666 7/21/10 0 0.43 0 0 0 30.82 0.129738 7/22/10 0 0.45 0 0 0 21.88 0.092099 7/23/10 10.36 0.24 2.1049 1.8611 37.22 37.22 0.156694 7/24/10 0.14 0.45 0.14 0 0 31.08 0.130842 7/25/10 0.12 0.39 0.12 0 0 25.76 0.108439 7/26/10 0 0.39 0 0 0 18.04 0.075933 7/27/10 0 0.47 0 0 0 8.69 0.036583 7/28/10 0 0.43 0 0 0 0.16 0.000655 7/29/10 0 0.39 0 0 0 0 0 7/30/10 2.68 0.35 2.1049 1.7595 35.19 35.19 0.148139 7/31/10 0 0.41 0 0 0 27.06 0.113922 8/1/10 4.54 0.37 2.1049 1.7391 34.78 34.78 0.146429 90days 8/2/10 3.04 0.24 2.1049 1.8611 37.22 37.22 0.156694 8/3/10 0.16 0.37 0.16 0 0 33.11 0.139370

PAGE 267

267 LIST OF R EFERENCES Aber, J.D., K.J. Nadelhoffer, P. Steudler, and J.M. Melillo. 1989. Nitrogen saturation in northern forest ecosystems BioScience 39:378 386. Abrahamson, W.G., and C.R. Abrahamson. 1996. Effects of fire on long unburned Florida uplands. Journal of Vegetation Science 7:565 574. Adams, P.W., and J.R. Boyle. 1980. Effects of fire on soil nutrients in clearcut and whole tree harvest sites in Central Michigan. Soil Sci. Soc. Am. J. 44:847 850. Afif, E., A. Matar, and J. Torrent. 1993. Availability of phosphate applied to calcareous Soils of West Asia and North Africa. Soil Sci. Soc. Am. J. 57:756 760. Ahmad, W., M.A. Gill, and T. Aziz. 2006. Bioavailability of inorganic phosphorus fractions in calcareous soils estimated by neubauer technique, iron impregnated filter paper, and chemical tests Communications in Soil Science and Plant Analysis 37:1185 1193. Allison, J.D., D.S. Brown, and K.J Novo Gradac. 1991. MINTEQA2/PRODEFA2, a geochemical assessment model for environmental systems : Version 3.0 User's Manual. EPA/600/3 91/021. Al Yahyai, R., B. Schaffer, F.S. Davies, and R. Munoz Carpena. 2006. Characterization of soil water retention o f a very gravelly loam soil varied with determination method Soil Science 171:85 93. Arocena, J.M., and C. Opio. 2003. Prescribed fire induced changes in properties of sub boreal forest soils. Geoderma 113:1 16. Asner, G.P., T.R. Seastedt, and A.R. Town send. 1997. The decoupling of terrestrial carbon and nitrogen cycles BioScience 47:226 234. Awasthi, R.P., S.K. Sharma, and V.P. Bhutani. 2000. Diagnosis a nd recommendation integrated system (DRIS) norms for peach ( Prunus persica L.) in Himachal Pradesh. Indian Journal of Horticulture 57(4): 277 280. Badia, D., and C. Marti. 2003. Plant ash and heat intensity effects on chemical and physical properties of two contrasting soils. Arid Land Research and Management 17:23 41. Bailey, J.S., J.A.M. Beattie, and D.J. Kilpatrick. 1997a. The diagnosis and recommendation integrated system (DRIS) for diagnosing the nutrient status of grassland swards: I. Model establishment. Plant and Soil 197:127 135. Bailey, J.S., A. Cushnahan, and J.A.M. Beattie. 1997b. The diagnosis and recommendation integrated system (DRIS) for diagnosing the nutrient status of grassland swards: II. Model calibration and validation. Plant and Soil 197:137 147.

PAGE 268

268 Bailey, J., R. Dils, R. Foy, and D. Patterson. 2000. The diagnosis and recommendation integrated system (DRIS) for diagnosing the nutrient status of grassland swards: III Practical applications. Plant and Soil 222:255 262. Baldock, J.O., and E.E. Schulte. 1996. Plant analysis with standardized scores combines d ris and sufficiency range approaches for corn Agron. J. 88:448 456. Bangroo, S.A., M.I. Bhat, T. Ali, M.A. Aziz, M.A. Bhat, and M.A. Wani. 2010. Diagnosis and recommendation integrated system (DRIS) a review International Journal of Current Research 10: 084 097. Barron Gafford, G.A., R.E. Will, E.C. Burkes, B. Shiver, and R.O. Teskey. 2003. Nutrient concentrations and contents, and their relation to stem growth, of intensively manage d Pinus taeda and Pinus elliottii stands of different planting densi ties Forest Science 49 (2): 291 300. Bates, T.E. 1990. Prediction of phosphorus availability from 88 Ontario soils using five phosphorus soil tests. Communications in Soil Science and Plant Analysis 21:1009 1023. Bauhus, J., P.K. Khanna, and R.J. Raison. 1993. The effect of fire on carbon and nitrogen mineralization and nitrification in an Australian forest soil. Australian Journal of Soil Research 31:621 639. Bell, L.C., and C.A. Black. 1970. Transformation of dibasic calcium phosphate dihydrate and octacalcium phosphate in slightly acid and alkaline soils Soil Sci. Soc. Am. J. 34:583 587. Bethlenfalvay, G.J., R.L. Franson, and M.S. Brown. 1990. Nutrition of mycorrhizal soybean evaluated by the diagnosis and recommendation inte grated system (DRIS). Agron. J. 82:302 304. Bilgili, E., and B. Saglam. 2003. Fire behavior in maquis fuels in Turkey. Forest Ecology and Management 184:201 207. Binkley, D., D. Richter, M.B. David, and B. Caldwell. 1992. Soil chemistry in a Loblolly/ Longleaf pine forest with interval burning Ecological Applications 2:157 164. Bird, M.I., E.M. Veenendaal, C. Moyo, J. Lloyd, and P. Frost. 2000. Effect of fire and soil texture on soil carbon in a sub humid savanna (Zimbabwe). Geoderma 94:71 90. Blank, R.R., F. Allen, and J.A. Young. 1994. Extractable anions in soils following wildfire in a sagebrush grass community Soil Sci Soc Am J 58:564 570. Blank, R.R., J. Chambers, B. Roundy, and A. Whittaker. 2007. Nutrient availability in rangeland soils: infl uence of prescribed burning, herbaceous vegetation removal, overseeding with bromus tectorum, season, and elevation Rangeland Ecology & Management 60:644 655.

PAGE 269

269 Boerner, R.E.J. 1982. Fire and nutrient cycling in temperate ecosystems BioScience 32:187 192. Boerner, R.E.J., K.L.M. Decker, and E.K. Sutherland. 2000a. Prescribed burning effects on soil enzyme activity in a southern Ohio hardwood forest: a landscape scale analysis. Soil Biology and Biochemistry 32:899 908. Boerner, R.E.J., S.J. Morris, E.K. Sutherland, and T.F. Hutchinson. 2000b. Spatial variability in soil nitrogen dynamics after prescribed burning in Ohio mixed oak forests. Landscape Ecology 15:425 439. Boring, L.R., J.J. Hendricks, C.A. Wilson, and R.J. Mitchell. 2004. Season of burn and nutrient losses in a longleaf pine ecosystem. International Journal of Wildland Fire 13:443 453. Brady, N.C., and R.R. Well. 2002. The nature and properties of soils ( thirteenth edition ). Upper Saddle River, New Jersy 07458: Prentice Ha ll. Brais, S., P. David, and R. Ouimet. 2000. Impacts of wild fire severity and salvage harvesting on the nutrient balance of jack pine and black spruce boreal stands. Forest Ecology and Management 137:231 243. Bray, R.H. and L.T. Kurtz. 1945. Determination of total, organic and available form of phosphorus in soil Soil Sci 59:39 5. Burgin, J.A., and Hamilton, K. S. 2007. Have we overemphasized the role of denitrification in aquatic ecosystems? a review of nitr ate removal pathways Front Ecological Environment 5:89 96. Busse, M.D., and L.F. DeBano. 2005. Soil Biology. p. 73 91. In Neary et al. (ed.). Wildland fire in ecosystems: effects of fire on soil and water USDA Forest Service, General Technical Report RMRS GTR 42, volume 4. Cachier, H., C. Liousse, P. Buat Menard, and A. Gaudichet. 1995. Particulate content of savanna fire emissions. Journal of Atmospheric Chemistry 22:123 148. Cade Menun, B.J., C.W. Liu, R. Nunlist, and J.G. McColl. 2002. Soil and li tter phosphorus 31 nuclear magnetic resonance spectroscopy: extractants, metals, and phosphorus relaxation times Journal of Environmental Quality 31:457 465. Carlyle, J.C., and E.K.S. Nambiar. 2001. Relationship between net nitrogen mineralization, pro perties of the forest floor and mineral soil, and wood production in Pinus radiata plantations. Can. J. For. Res. 31:889 898. Castro, B., and J. Torrent. 1994. Phosphate availability in calcareous Vertisols and Inceptisols in relation to fertilizer type and soil properties. Nutrient Cycling in Agroecosystems 40:109 119.

PAGE 270

270 Castro, B., and J. Torrent. 1995. Phosphate availability in calcareous Vertisols and Inceptisols in relation to fertilizer type and soil properties. Fertility Research 40:109 119. Certini G. 2005. Effects of fire on properties of forest soils: A review Oecologia 143:1 10. Chaneton, E.J., J.H. Lemcoff, and R.S. Lavado. 1996. Nitrogen and phosphorus cycling in grazed and ungrazed plots in a temperate subhumid grassland in Argentina. Journ al of Applied Ecology 33:291 302. Chien, S.H., E.R. Austin, and C.G. Calvo. 2009. Underestimation of available phosphorus by resin, bicarbonate and olsen tests in calcareous soils treated with gypsum Communications in Soil Science and Plant Analysis 40:1 924 1936. Choromanska, U., and T.H. DeLuca. 2001. Prescribed fire alters the impact of wildfire on soil biochemical properties in a ponderosa pine forest. Soil Sci Soc Am J 65:232 238. Christensen, N.L., and C.H. Muller. 1975. Effects o f fire on factor s controlling plant growth in adenostoma chaparral Ecological Monographs 45:29 55. Cleemput, V.O. 1998. Subsoils: Chemo and biological denitrification, N 2 O and N 2 emissions Nutrient Cycling in Agroecosystems 52:187 194. Collins, M.E. 2002. Soil orders of florida Department of Soil & Water Science, IFAS University of Florida. Comerford, N.B., and R.F. Fisher. 1984. Using foliar analysis to classify nitrogen deficient sites. Soil Sci. Soc. Am. J. 48:910 913. Cornelis, W.M., J. Ronsyn, M.V. Meirvenne and R. Hartmann. 2001. Evaluation of pedotransfer functions for predicting the soil moisture retention curve Soil Sci. Soc. Am. J. 65:638 648. Covington, W.W., and S.S. Sackett. 1984. The effect of a prescribed fire in southwestern ponderosa pine on or ganic matter and nutrients in woody debris and forest floor Forest Science 30:183 192. Covington, W.W., and S.S. Sackett. 1986. Effect of periodic burning on soil nitrogen concentrations in ponderosa pine. Soil Sci Soc Am J 50:452 457. Covington, W.W., and S.S. Sackett. 1992. Soil mineral nitrogen changes following prescribed burning in ponderosa pine. Forest Ecology and Management 54:175 191. DeBano, L.F. 2000a. Water repellency in soils : A historical overview Journal of Hydrology 231 232:4 32.

PAGE 271

271 DeBa no, L.F. 2000b. The role of fire and soil heating on water repellency in wildland environments : A Review. Journal of Hydrology 231 232:195 206. DeBano, L.F., and C.E. Conrad. 1978. The effect of fire on nutrients in a chaparral ecosystem. Ecology 59:489 4 97. DeBano, L.F., D.G. Neary, P.F. Folliott. 1998. Fire's effects on ecosystems John Wiley & Sons, Inc. DeBano, L.F., D.G. Neary, and P.F. Ffolliott. 2005. Soil physical properties p. 29 51. In Neary et al. (ed.). Wildland fire in ecosystems : Effects o f fire on soil and water. USDA Forest Service, General Technical Report RMRS GTR 42, volume 4. DeLuca, T.H., and K.L. Zouhar. 2000. Effects of selection harvest and prescribed fire on the soil nitrogen status of ponderosa pine forests. Forest Ecology and Management 138:263 271. deVisser, P.H.B., C. Beier, L. Rasmussen, K. Kreutzer, N. Steinberg, M. Bredemeier, K. Blanck, E.P. Farrell, and T. Cummins. 1994. Biological response of five forest ecosystems in the EXMAN project to input changes of water, nutrie nts and atmospheric loads. Forest Ecology and Management 68:15 29. Dickens, E.D., D.J. Moorhead, and B. McElvany. 2003. Pine plantation fertilization Better Crops 87:12 15. Dikici, H., and C.H. Yilmaz. 2006. Peat fire effects on some properties of an artificially drained peatland Journal of Environmental Quality 35:866 870. Doren, R.F., W.J. Platt, and L.D. Whitaker. 1993. Density and size structure of slash pine stands in the everglades region o f S outh Florida. Forest Ecology and Management. 59:295 312. Duever, M.J., J.E. Carlson, J.F. Meeder, L.C. Duever, L.H. Gunderson, L.A. Riopelle, T.A. Alexander, R.F. Myers, and D.P. Sprangler. 1979. Resource inventory and analysis of the B ig C ypress N ational Preserve. C enter for wetlands University of Florida; Gainesville, Florida. Dumontet, S., H. Dinel, A. Scopa, A. Mazzatura, and A. Saracino. 1996. Post fire soil microbial biomass and nutrient content of a pine forest soil from a dunal Mediterranean environment. Soil Biology and Biochemistry 28:1467 1475. Dungin, P.B., and P.J. Vogelsang. 1984. Dispersion of kaolinite by water extracts of douglas fir ash Can. J. Soil Sci. 64:439 443. Dunham, R.J., and P.H. Nye. 1976. The influence of soil water content on the uptake of ions by roots. III. phosphat e, potassium, calcium and magnesium uptake and concentration gradients in soil Journal of Applied Ecology 13:967 984.

PAGE 272

272 Egli, T., and J.R. Quayle. 1986. Influence of the carbon: nitrogen ratio of the growth medium on the cellular composition and the ability of the methylotrophic yeast hansenula polymorpha to utilize mixed carbon source s. J Gen Microbiol 132:1779 1788. Elrashidi, M. A. 2009. Selection o f an appropriate phosphorus test for soil s. USDA Natural Resource Conservation Service. Available at: ftp://ftp fc.sc.egov.usda.gov/NSSC/Analytical_Soils/phosphor.pdf Elwali, A.M.O., G.J. Gascho, and M.E. Sumner. 1985. DRIS norms for 11 nutrients in corn leaves Agron. J. 77:506 508. Enslin, B.W., A.L.F. Potgieter, H.C. Biggs, and R. Biggs. 2000. Long term effects of fire frequency and season on the woody vegetation dynamics of the Sclerocarya birrea/Acacia nigrescens savanna of the Kruger National Park. Koedoe 43:27 37. Essington, M.E. 2004. Soil and Water Chemistry: An intergrative approac h. CRC Press. Eugene, J.K. 2000. Relevance of soil testing to agriculture and the environmen t. Issue Paper. Council for Agricultural Science and Technology. Fairchild Tropical Batonic Garden. 2008. Available at: http://www.pineridgeorchids.com/sanctuary.htm. Fenn, M.E., M.A. Poth, J.D. Aber, J.S. Baron, B.T. Bormann, D.W. Johnson, A.D. Lemly, S.G. McNulty, D.F. Ryan, and R. Stottlemyer. 1998. Nitrogen excess in north american ecosystems: predisposing factors, ecosystem responses, and management strategies. Ecological Applications 8:706 733. Fenn, M.E., M.A. Poth, and D.W. Johnson. 1996. Evidence for nitrogen saturation in the San Bernardino Mountains in Southern California. Forest Ecology and Management 82:211 230. Fernndez, I., A. Cabaneiro, and T. Carballas. 1999. Carbon mineralization dynamics in soils after wildfires in two Galician forests. Soil Biology and Biochemistry 31:1853 1865. Fisher, R.F., and D. Blinkey. 2000 Ecology and management of forest soils Wiley, New York. Fixen, P.E., A.E. Ludwick, and S.R. Olsen. 1983. Phosphorus and potassium fertilization of irrigated alfalfa on calcareous soils : II. Soil phosphorus solubility relationships Soil Sci. Soc. Am. J. 47:112 117. Florida Automated Weather Network (FAWN) at TREC. 2011. Available at: http://fawn.ifas.ufl.edu/station.php?id=440 Florida Natural Area Inventory. 2008. Available at: http://www.fnai.org/index.cfm.

PAGE 273

273 Fourqurean, J.W., and J.C. Zieman. 1992. Phosphorus limitation of primary production in Florida Bay: Evidence from C: N: P ratios of the dominant seagrass Thalassia testudinum Limnology and Oceanography 37:162 171. Gahoonia, T., S. Raza, and N. Nielsen. 1994. Phosphorus depletion in the rhizosphere as influenced by soil moisture. Plant and Soil 159:213 218. Gholz, H.L., R.F. Fisher, and W.L. Prichett. 1985. Nutrient dynamics in slash pine plantation ecosystems Ecology 66:647 659. Giardina, C.P., R.L. Sanford, Jr., and I.C. Dockersmith. 2000. Changes in soil phosphorus and nitrogen during slash and burn clearing of a dry tropical forest Soil Sci Soc Am J 64:399 405. Gill, A.M., P.G. Ryan, P.H.R. Moore, and M. Gibson. 2000. Fire regimes of World Heritage Kakadu National Park., Australia Austral Ecology 25:616 625. Gillon, D., V. Gomendy, C. Houssard, J. Marechal, and J.C. Valette. 1995. Combustion and nutrient losses during laboratory burns International Journal of Wildland Fire 5:1 12. Giovannini, G., S. Lucchesi, and M. Giachetti. 1988. Effect of heating on some physical and chemical parameters related to soil aggregation and erodibility. Soil Science 146:255 261. Glinski, J., K. Stahr, Z. Stepniewska, and M. Brzezinska. 1992. Changes of redox and pH conditions in a flooded soil amended with glucose and nitrate under laboratory conditions. Zeitschrift fr Pflanzenernhrung und Bodenkunde 155:13 17. Glitzenstein, J.S., D.R. Streng, and D.D. Wade. 2003. Fire F requency effects on longleaf pine ( Pinus pa lustris P. Mille) V egetation in South Carolina and Northeast Florida, USA. Natural Areas Journal 23:22 37. Gonzlez Prez, J.A., F.J. Gonzlez Vila, G. Almendros, and H. Knicker. 2004. The effect of fire on soil organic matte r -a review. Environment Inter national 30:855 870. Govender, N., W.S.W. Trollope, and B.W.V. Wilgen. 2006. The effect of fire season, fire frequency, rainfall and management on fire intensity in savanna vegetation in South Africa. Journal of Applied Ecology 43:748 758. Gray, D.M., an d J. Dighton. 2006. Mineralization of forest litter nutrients by heat and combustion. Soil Biology and Biochemistry 38:1469 1477. Groeschl, D.A., J.E. Johnson, and D.W. Smith. 1993. Wildfire effects on forest floor and surface soil in a Table Mountain Pine Pitch Pine forest International Journal of Wildland Fire 3:149 154.

PAGE 274

274 Grove, T.S., A.M. O'Connell, and G.M. Dimmock. 1986. Nutrient changes in surface soils after an intense fire in Jarrah ( Eucalyptus Marginata Donn) forest. Austr alian Journal of Ecology 11:303 317. Gunderson, L.H., and J.R. Snyder. 1997. Fire patterns in the southern everglades. p. 291 305. In Davis et al. (ed.). Everglades: The ecosystem and its restoration CRC Press LLC. Gusewell, S., W. Koerselman, and J.T.A Verhoeven. 2003. Biomass N:P ratios as indicators of nutrient limitation for plant populations in wetlands Ecological Applications 13:372 384. Hammad, M.E., D.L. Rimmer, and J.K. Syers. 1992. Effect of iron oxide on phosphate sorption by calcite and ca lcareous soils. Journal of Soil Science 43:273 281. Hanlon, E.A., B. Schaffer, M. Ozores Hampton, and H.H. Bryan. 1996. Ammonium bicarbonate DTPA extraction of elements from waste amended calcareous soil. Communications in Soil Science and Plant Analysi s 27:2321 2335. Harley, G.L., H.D. Grissino Mayer, and S.P. Horn. 2011. The dendrochronology of Pinus Elliottii in the Lower Florida Keys: Chronology Development and Climate Response. Tree Ring Research 67:39 50 Harmel, R.D., P.K. Smith, and K.W. Migliaccio. 2010. M odifying goodness of fit indicators to incorporate both measurement and model uncertainty in model calibration and validation. Transactions of the ASABE 53:55 63. Hatten, J.A., and D. Zabowski. 2009. C hanges in soil organic matter pools and carbon mineralization as influenced by fire severity. Soil Sci. Soc. Am. J. 73:262 273. Hatten, J.A., and D. Zabowski. 2010. Fire severity effects on soil organic matter from ponderosa pine forest: A Laboratory Study. International Journal of Wildland Fire 19:613 623. Havlin, J.L., and P.N. Soltanpour. 1981. Evaluation of the NH 4 HCO 3 DTPA soil test for iron and zinc. Soil Sci. Soc. Am. J. 45:70 75. Henig Sever, N., D. Poliakov, and M. Broza. 2001. A novel metho d for estimation of wild fire intensity based on ash pH and soil microarthropod community. Pedobiologia 45:98 106. Hernndez, T., C. Garca, and I. Reinhardt. 1997. Short term effect of wildfire on the chemical, biochemical and microbiological properties of Mediterranean pine forest soils. Biology and Fertility of Soils 25:109 116. Hill, K. 2001. Slash Pine. Available at: http://www.sms.si.edu/irlspec/pinus_elliot.htm Http://en.wikipedia.org/wiki/Potassium: Citing_ Wikipedia

PAGE 275

275 Hubbard, R.M., J.M. Vose, B.D. Clinton, K.J. Elliott, and J.D. Knoepp. 2004. Stand restoration burning in oak pine forests in the southern Appalachians: effects on aboveground biomass and carbon an d nitrogen cycling. Forest Ecology and Management 190:311 321. Inskeep, W.P., and P.R. Bloom. 1986. Effects of soil moisture on soil pCO 2 soil solution bicarbonate, and iron chlorosis in soybeans Soil Sci. Soc. Am. J. 50:946 952. Johnson, D.W., and P.S. Curtis. 2001. Effects of forest management on soil C and N storage: meta analysis. Forest Ecology and Management 140:227 238. Kamprath, E.J., D.B. Beegle, P.E. Fixen, and S.C. Hodges. 2000. Relevance of soil testing to agriculture and the environment Council for Agricultural Science and Technology 15: 1 12 Karathanasis, A.D. 1991. Phosphate mineralogy and equilibria in two Kentucky Alfisols derived from ordovician limestones Soil Sci. Soc. Am. J. 55:1774 1782. Khanna, P.K., R.J. Raison, and R.A. Falkiner. 1994. Chemical properties of ash derived from eucalyptus litter and its effects on forest soils. Forest Ecology and Management 66:107 125. Knapp, E.E., B.L. Estes, and C.N. Skinner. 2009. Ecological effects of prescribed fire season: a literature review and synthesis for managers USDA Forest Service, General Technical Report PSW GTR 224. Knicker, H. 2007. How does fire affect the nature and stability of soil organic nitrogen and carbon? a review. Biogeochemistry 85:91 118. Knoepp, J.D., L.F. DeBano, and D.G. Neary. 2005. Soil chemistry. p. 53 71. In Neary et al. (ed.). Wildland fire in ecosystems: effects of fire on soil and water. USDA Forest Service, General Technical Report RMRS GTR 42, volume 4. Knoepp, J.D., and W.T. Swank. 19 93. Site preparation burning to improve southern appalachian pine hardwood stand: nitrogen responses in soil, soil water, and streams Can. J. For. Res. 23:2263 2270. Koerselman, W., and A.F.M. Meuleman. 1996. The vegetation N:P ratio: a new tool to detect the nature of nutrient limitatio n. Journal of Applied Ecology 33:1441 1450. Korom, F. S. 1992. Natural denitrification in the saturated zone: a review Water Resources Research, 28: 1657 1668. Krause, P., D.P. Boyle, and F. Base. 2005. Comparison of different efficiency criteria for hydrological model assessment. Adv. Geosci. 5:89 97. Kuddes Fischer, L.M., and M.A. Arthur. 2002. Response of understory vegetation and tree regeneration to a single prescribed fire in oak pine forests Natural Areas Journal 22:43 52.

PAGE 276

276 Kuo, S. 1996. Phosphorus. p. 869 919. In Sparks et al. (ed.). SSSA book series: methods of soil analysis, part 3 chemical method s. Soil Science Society of American, Inc. Madison, Wiscosin. Kutiel, P., and A. Shaviv. 1989. Effect of simulated forest fire on the availability of N and P in Mediterranean soils. Plant and Soil 120:57 63. Kutiel, P., and A. Shaviv. 1992. Effects of soil type, plant composition and leaching on soil nutrients following a simulated forest fire. Forest Ecolog y and Management 53:329 343. Laskowski, R., M. Niklinska, and M. Maryanski. 1995. The dynamics of chemical elements in forest litter. Ecology 76:1393 1406. Lavoie, M., G. Starr, M.C. Mack, T.A. Martin, and H.L. Gholz. 2010. Effects of a P rescribed fire on understory vegetation, carbon pools, and soil nutrients in a longleaf pine slash pine forest in florida Natural Areas Journal 30:82 94. Legates, D.R., and R.E. Davis. 1997. The continuing search for an anthropogenic climate change signal: limitations of correlation based approaches. Geophys. Res. Lett. 24:2319 2322. Legates, D.R., and G.J. McCabe, Jr. 1999. Evaluating the use of "goodness of fit" measures in hydrologic and hydroclimatic model validation. Water Resour. Res. 35:233 241. Leggett, G.E., and D.P. Argyle. 1983. The DTPA extractable iron, manganese, copper, and zinc from neutral and calcareous soils dried under different condition s. Soil Sci. Soc. Am. J. 47:518 522. Li, Y.C. 2001. Calcareous soil in miami dade county Fla. Coop Ext. Ser. Fac t Sheet. SL 183 Li, Y.C., A.K. Alva, D.V. Calvert and D. Banks. 2002. Atmospheric deposition of nitrogen in a high lightning intensity area. Commun. Soil Sci. Plant Anal. 33 (9&10): 1671 1677. Li, Y.C. and M. Norland. 2001. The role of soil fertility in invasion of Brazilian pepper ( Schinus terebinthifolius ) in Everglades National Park, FL. Soil Sci. 166:400 405. Lindsay, W.L., P.L.G. Vlek, and S.H. Chien. 1989. Phosphate minerals Minerals in Soil Environment 150:1089 1130. Liu, G., Y. Li, and A. Alva. 2007. Moisture quotients for ammonia volatilization from four soils in potato production region s. Water, Air, & Soil Pollution 183:115 127. Loope, L.L., D.W. Black, S. Black, and G.N. Avery. 1979. Distribution and abundanc e of flora in limestone Rockland Pine forests of Southeastern Florida. Report T 547. Everglades National Park: South Florida Research Center.

PAGE 277

277 Loope, L.L., and V.L. Dunevitz. 1981. Impact of fire exclusion and invasion of Schinus terebinthiforlius on limes tone Rockland Pine Forests of Southeastern Florida. Technical Report T 645. Everglades National Park, South Florida Research Center. Lucena, J.J., and I. Bascones. 1993. AB DTPA cation extraction in Spanish soil samples Communications in Soil Science and Plant Analysis 24:2427 2440. Macadam, A.M. 1997. Effects of broadcast slash burning on feuls and soil chemical properies in the sub boreal spruce zone o f Central British Columbia. Canadian Journal of Forest Research 17:1577 1584. Marion, G.M., J.M. Moreno, and W.C. Oechel. 1991. Fire severity, ash deposition, and clipping effects on soil nutrients in chaparral. Soil Science Society of America Journal 55:235 240. McCaw, W.L., J.E. Neal, and R.H. Smith. 2002. Stand characteristics a nd fuel accumulation in a sequence of Even Aged Karri ( Eucalyptus Diversicolor) stands in South West Western Australia. Forest Ecology and Management 158:263 271. McGarvey, R.C., T.A. Martin, and T.L. White. 2004. Integrating within crown variation in net photosynthesis in loblolly and slash pine families. Tree Physiology 24:1209 1220. McGroddy, M.E., T. Daufresne, and L.O. Hedin. 2004. Scaling of C:N:P stoichiometry in forests worldwide: implications of terrestrial Redfield type ratios Ecology 85:2390 2401. McMinn, J.W. 1970. Optimum depth and season fo r direct seeding slash pine in South F lorida. U.S. Forest Service Note SE 117. Megonigal, J.P., M.E. Hines, and P.T. Visscher. 2004. anaerobic metabolism: linkages to trace gases and aerobic processes. p. 317 424. In Schlesinger, W.H. (ed.). Biogeochemistry. Elsevier Pergamon, Oxford, UK. Mehlich, A. 1953. Determination of P, Cu, Mg, K, Na, and NH4. Mimeo. North Carolina Soil Test Divi sion, Raleigh. Mehlich, A. 1984. Mehlich 3 Soil test extractant: a modification of mehlich 2 extractant. Communications in Soil Science and Plant Analysis 15:1409 1416. Mengel, K. 1994. iron availability in plant tissues iron chlorosis on calcareous so ils. Plant and Soil 165:275 283. Mengel, K., M. Breininger, and W. Bbl. 1984. Bicarbonate, the most important factor inducing iron chlorosis in vine grapes on calcareous soil. Plant and Soil 81:333 344. Miao, S. 2004. Rhizome Growth and nutrient resorpt ion: mechanisms underlying the replacement of two clonal species in Florida Everglades. Aquatic Botany 78:55 66.

PAGE 278

278 Miao, S.L., and F.H. Sklar. 1997. Biomass and nutrient allocation of sawgrass and cattail along a nutrient gradient in the Florida Everglades. Wetlands Ecology and Management 5:245 264. Misra, A., and G. Tyler. 1999. Influence of soil moisture on soil solution chemistry and concentrations of minerals in the calcicole s Phleum phleoides and Veronica spicata grown on a limestone soil. Annals of Bo tany 84:401 410. Monleon, V.J., and K. Cromack jr. 1996. Long Term effects of prescribed underburning on litter decomposition and nutrient release in ponderosa pine stands in Central Oregon. Forest Ecology and Management 81:143 152. Monleon, V.J., and K. Cromack jr., and J.D. Landsberg. 1997. Short and long term effects of prescribed under burning on nitrogen availability in ponderosa pine stands in Central Oregon. Can. J. For. Res. 27:369 378. Moreno, E.C., W.L. Lindsay, and G. Osborn. 1960. Reactions of dicalcium phosphate dihydrate in soils Soil Science 90:58 68. Motavalli, P.P., C.A. Palm, W.J. Parton, E.T. Elliott, and S.D. Frey. 1995. Soil pH and organic C dynamics in tropical forest soils: evidence from laboratory and simulation studies. Soil Bi ology and Biochemistry 27:1589 1599. Mulvaney, R.L. 1996. Nitrogen inorganic forms. p. 1123 1200. In Sparks et al. (ed.). SSSA Book Series: Methods o f soil analysis, part 3 chemical methods Soil Science Society of American, Inc. Madison, Wiscosin. Muoz Carpena, R., Y. Li, and T. Olczyk. 2002. Alternatives of low cost soil moisture monitoring devices for vegetable production in south miami dade county fact sheet a ABE333 of the Department of Agricultural and Biological Engineering, University of Florida. Muoz Carpena. R, S. Shukla, and K. Morgan. 2006. Field devices for monitoring soil water content Regional Extension Bulletin no. SR IWM 2. The Irrigation Water Management Program Team of the Southern Regional Water Program. USDA CSREES. Murphy, J.D., D.W. Johnson, W.W. Miller, R.F. Walker, and R.R. Blank. 2006a. Prescribed fire effects on forest floor and soil nutrients in a sierra N evada Forest. Soil Science 171:181 199. Murphy, J.D., D.W. Johnson, W.W. Miller, R.F. Walker, E.F. Carroll, and R.R. Blank. 2006b. Wildfire effects on soil nutrients and leaching in a Tahoe Basin Watershed. Journal of Environmental Quality 35:479 489.

PAGE 279

279 Nardoto, G.B., and M.M.D.C. Bustamante. 2003. Effects of fire on soil nitrogen dynamics and microbial biomass in savannas of Central Brazil. Pesquisa Agropecuria Brasileira 38:955 962. Neary, D.G., C.C. Klopatek, L.F. DeBano, and P.F. Ffolliott. 1999. Fi re effects on belowground sustainability: A review and synthesis. Forest Ecology and Management 122:51 71. Neary, D.G., K.C. Ryan, L.F. DeBano. 2005. Wildland Fire in Ecosystems: Effects of fire on soils and water Gen. Tech. Rep. RMRS GTR 42 vol.4. Ogden, UT: U.S. Department of Agriculture, Forest Service, Rocky Mountain Research Station. 250 p. Available at: http://www.fs.fed.us/rm/pubs/rmrs_gtr042_4.pd f Needham, T.D., J.A. Burger, and R.G. Oderwald. 1990. Relationship between diagnosis and recommendation integrated system (DRIS) optima and foliar nutrient critical levels. Soil Sci. Soc. Am. J. 54:883 886. Neill, C., W.A. Patterson Iii, and D.W. Crary Jr. 2007. Responses of soil carbon, nitrogen and cations to the frequency and seasonality of prescribed burning in a Cape Cod Oak pine forest. Forest Ecology and Management 250:234 243. Newton, R. J., W. Tang, and S.M. Jain. 2005. Slash pine ( Pinus Ellio ttii Engelm.). Protocol for Somatic Embryogenesis in Woody Plants: 1 10. O'Connor, G.A. 1988. Use and misuse of the DTPA soil test J. Environ. Qual. 17:715 718. Olmsted, I., and L.L. Loope. 1984. Plant communities of Everglades National Park. p.167 184. I n Gleason, P.J. (ed). Environments of South Florida, past and present. Miami Geological Society; Coral Gables, Florida. Olmsted, I., W.B. Robertson, Jr., J. Johnson, and O.R. Bass, Jr. 1983. The vegetation of Long Pine Key, Everglades National Park. Report SFRC 83/05, South Florida Research Center, Everglades National Park; Homestead, Florida. Olsen, S. R., C.V. Cole, F.S. Watanabe, and L.A. Dean. 1954. Estimation of available phosphorus in soils by extraction with sodium bicarbonate. U.S. Department of Agriculture Circular 939. U.S. Government Printing Office, Washington, D.C. Oreskes, N., K. Shrader Frechette, and K. Belitz. 1994. Verification, validation, and confirmation of numerical models in the earth sciences Science 263:641 646. Orphanos, P. 1978. Extraction of phosphorus from the major soils of cyprus with water or bicarbonate. Plant and Soil 49:417 420. Pham, H.Q., and D.G. Fredlund. 2008. Equations for the entire soil water characteristic curve of a volume change soil. Canadian Geotechn ical Journal 45:443 453.

PAGE 280

280 Pierzynski, G.M., R.W. McDowell, and J.M. Sims. 2005. Phosphorus reactions and cycling in soils. p. 53 57. In Dale T. Westermann (ed.). Phosphorus: agriculture and the environment. Agronomy Monograph No.46. Pierzynski, G.M., J.T. Sims, and G.F. Vance. 2000. Soils and environmental quality (Second edition). Boca Raton, FL: CRC Press LLC. Prieto Fernandez, A., M.C. Villar, M. Carballas, and T. Carballas. 1993. Short term effects of a wildfire on the nitrogen status and its mineralization kinetics in a n Atlantic forest soil Soil Biology and Biochemistry 25:1657 1664. Qian, Y., S.L. Miao, B. Gu, and Y.C. Li. 2009a. Estimation o f postfire nutrient loss in the Florida Everglades. J. Environ. Qual. 38:1812 1820. Qian, Y., S.L. Miao, B. Gu, and Y.C. Li. 2009b. Effects o f burn temperature on ash nutrient forms and availability from cattail ( Typha domingensis ) and sawgrass ( Cladium jamaicense ) in the Florida Everglades. Environmental Quality 38:451 464. Raison, R.J ., P.K. Khanna, and P.V. Woods. 1985. Mechanism of element transfer to the atmosphere during vegetation fires Can. J. For. Res. 15:132 140. Rawls, W.J., D.L. Brakensiek, and K.E. Saxtonn. 1982. Estimation of soil water properties 25:1316 1320. Reddy, K.R., and R.D. DeLaune. 2008. Biogeochemitry of wetland: science and applications. Boca Raton, FL: CRC Press LLC. Redfield, A.C. 1934. On t he proportions of organic derivations in sea water and their relation to the composition of plankton p. 177 192. In Daniel R.J. (ed.). James Johnstone Memorial Volume University of Liverpool Press. Redfield, A.C. 1958. The biological control of chemical factors in the environment. Am Sci 46:205 221 Reed, S.T., and D.C. Martens. 1996. AB DTPA method p. 703 722. In Sparks et al. (ed.). SSSA Book Series: Methods of so il analysis, part 3 chemical methods Soil Science Society of American, Inc. Madison, Wiscosin. Regalado, C.M., A. Ritter, J. Alvarez Benedi, and R. Munoz Carpena. 2005. Simplified metho d to estimate the green ampt wetting front suction and soil sorptivity with the philip dunne falling head permeameter Vadose Zone J. 4:291 299. Richards, B., and D. Bevege. 1972. Principles and practice of foliar analysis as a basis for crop logging in pine plantations. Plant and Soil 36:109 119.

PAGE 281

281 Robbins, C.W., D.T. Westermann, and L.L. Freeborn. 1999. Phosphorus forms and extractability from three sources in a recently exposed calcareous subsoil. Soil Sci. Soc. Am. J. 63:1717 1724. Robbins, L.E., and R.L. Myers. 1992. Seasonal effects of prescribed burnin g in Florida: a review. Tall Timbers Research, Inc., Miscellaneous Publication No. 8, Tallahassee, FL. Romany, J., P.K. Khanna, and R.J. Raison. 1994. Effects of slash burning on s oil phosphorus fractions and sorption and desorption of phosphorus. Forest Ecology and Management 65:89 103. Ross, M.S. 1995. An action plan to conserve the native plants of F lorida Available at http://everglades.fiu.edu/serp/action/toc.html. Ross, M.S. C.L. Coultas, and Y.P. Hsieh. 2003. Soil productivity relationships and organic matter turnover in dry tropical forests of the Florida Keys. Plant and Soil 253:479 492. Ross, M.S., J.J. O'Brien, and L.d.S.L. Sternberg. 1994. Sea level rise and the reduc tion in pine forests in the Florida Keys. Ecological Applications 4:144 156. Russell, J.D., A.R. Fraser, J.R. Watson, and J.W. Parsons. 1974. Thermal decomposition of protein in soil organic matter. Geoderma 11:63 66. Ryan, J., D. Curtin, and M.A. Cheema 1985b. Significance of iron oxides and calcium carbonate particle size in phosphorus sorption by calcareous soil s. Soil Science Society of America Journal 49:74 76. Ryan, J., H.M. Hasan, M. Baasiri, and H.S. Tabbara. 1985a. Availability and transforma tion of applied phosphorus in calcareous Lebanese soils Soil Sci. Soc. Am. J. 49:1215 1220. Sah, J.P., M.S. Ross, S. Koptur, and J.R. Snyder. 2004. Estimating aboveground biomass of broadleaved woody plants in the understory of Florida Keys Pine forests. Forest Ecology and Management 203:319 329. Sah, J.P., M.S. Ross, J.R. Snyder, S. Koptur, and H.C. Cooley. 2006. Fuel loads, fire regimes, and post fire fuel dynamics in Florida Keys pine forests. International Journal of Wildland Fire 15:463 478. Sah, J.P., M.S. Ross, J.R. Snyder, and D.E. Ogurcak. 2010. Tree mortality following prescribed fire and a storm surge event in slash pine ( Pinus elliottii var. densa ) forests in the Florida Keys, USA. International Journal of Forestry Research vol. 2010: Article ID 204795 doi:10.1155/2010/204795 Schaap, M.G., and M.T. van Genuchten. 2006. A modified mualem van genuchten formulation for improved description of the hydraulic conductivity near saturation Vadose Zone J. 5:27 34.

PAGE 282

282 Schoch, P., and D. Binkley. 1986. Prescribed burning increased nitrogen availability in a mature loblolly pine stand. Forest Ecology and Management 14:13 22. Serrasolsas, I., and P.K. Khanna. 1995a. Changes in heated and autoclaved forest soils of S.E. Australia. I. Carbon and Nitr ogen. Biogeochemistry 29:3 24. Serrasolsas, I., and P.K. Khanna. 1995b. Changes in heated and autoclaved forest soils of S.E. Australia. II. Phosphorus and Phosphatase Activity. Biogeochemistry 29:25 41. Sharpley, A., and B. Moyer. 2000. Phosphorus forms in manure and compost and their release during simulated rainfall. Journal of Environmental Quality 29:1462 1469. Simard, S.J. 1991. Fire severity, changing scales, a nd how things hang together. International Journal of Wildland Fire 1:23 34. Smith, J.K., R.D. Laven, and P.N. Omi. 1993. Microplot sampling of fire behavior on Populus tremuloides stands in North Central Colorado. International Journal of Wildland Fire 3:85 94. Snyder, J.R. 1986. The impact of wet and dry season prescribed fire o n miami rock ridg e Pineland. Report SFRC 86/06. Everglades national Park: South Florida Research Center. Snyder, J.R., A. Herndon, and W.B. Robertson. 1990. South florida rocklands. p. 230 277. In Mayer et al. (ed.). Ecosystems of Florida. University of Central Florida Press; Orlando, Florida. Snyder, J.R., M.S. Ross, S. Koptur, and J.P. Sah. 2005. Developing ecological criteria for prescribed fire in South Florida Pineland ecosystems. USGS Open File Report of 2006 1062. Soltanpour, P.N., and S. Workma n. 1979. Modification of the NH4HCO3 DTPA soil test to omit carbon black Communications in Soil Science and Plant Analysis 10:1411 1420. Soto, B., R. Basanta, and F. Diaz Fierros. 1993. Influence of wildland fire on surface runoff from a hillslope Acta Geologica Hispanica 28:95 102. Sparks, J.C., R.E. Masters, D.M. Engle, and G.A. Bukenhofer. 2002. Season o f burn influences fire behavior and fuel consumption in restored shortleaf pine grassland communities. Restoration Ecology 10:714 722. Sumner, M.E. 1977. Application of beaufils' diagnostic indices to maize data published in the literature irrespective of age and conditions. Plant and Soil 46:359 369. Sumner, M.E. 1978. Interpretation of nutrient ratios in plant tissue. Communicatio ns in Soil Science and Plant Analysis 9:335 345.

PAGE 283

283 Sumner, M.E. 1979. Interpretation of foliar analysis for diagnostic purposes. A gronomy Joural 71:343 348. Sumner, M.E. 1994. Measurement of soil pH: problem s and solutions. Communications in Soil Science and Plant Analysis 25:859 879. Tedder, D.W., and F.G. Pohland. 1993. Emerging technologies in hazardous waste management iii American Chemical Society. Teskey, R.O., H.L. Gholz, and W.P. Cropper. 1994. Influence of climate and fertilization on ne t photosynthesis of mature slash pine. Tree Physiology 14:1215 1227. Tessier, J.T., and D.J. Raynal. 2003. Use of nitrogen to phosphorus ratios in plant tissue as an indicator of nutrient limitation and nitrogen saturation. Applied Ecology 40:523 534. To mkins, I.B., J.D. Kellas, K.G. Tolhurst, and D.A. Oswin. 1991. Effects of fire intensity on soil chemistry in a eucalypt forest. Australian Journal of Soil Research 29:25 47. Tunesi, S., V. Poggi, and C. Gessa. 1999. Phosphate adsorption and precipitation in calcareous soils: The role of calcium ions in solution and carbonate minerals. Nutrient Cycling in Agroecosystems 53:219 227. Tyler, G. 1992. Inability to solubilize phosphate in limestone soils key factor controlling calcifuge habit of plants. Plant and Soil 145:65 70. Tyler, G. 1996. Mineral nutrient limitations of calcifuge plants in phosphate sufficient limestone soil. Annals of Botany 77:649 656. Ubeda, X., M. Lorca, L.R. Outeiro, S. Bernia, and M. Castellnou. 2005. Effects of prescribed fire on soil quality in Mediterranean grassland (Prades Mountains, North East Spain). International Journal of Wildland Fire 14:379 384. Ulery, A.L., and R.C. Graham. 1993a. Forest fire effects on soil color and texture. Soil Sci. Soc. Am. J. 57:135 140. Ulery, A.L., R.C. Graham, and C. Amrhein. 1993b. Wood ash composition and soi l pH following intense burning Soil Science 156:358 364. Ulery, A.L., R.C. Graham, O.A. Chadwick, and H.B. Wood. 1995. Decade scale changes of soil carbon, nitrogen and exchangeable c ations under chaparral and pine. Geoderma 65:121 134. USDA Forest Service. 1990. Slash pine. Agriculture Handbook 654. Available at http://www.na.fs.fed.us/spfo/pu bs/silvics_manual/Volume_1/pinus/elliottii.htm USDA Natural Resources Conservation Service. 1996. Soil survey of dade county area, Florida.

PAGE 284

284 U.S. Fish and Wildlife Service. 2007. Pine rocklansd chapter in the south florida multi species recovery plan Available at: http://www.fws.gov/southeast/vbpdfs/commun/pr.pdf U.S Geological Survey. 2000. Fire ecology of south florida pinelands. Available at: http://fl.biology.usgs.gov/Center_Publications/Fact_Sheets/snyderflier.pdf vanCleemput, O. 1998. Subsoils: Chemo a nd biological denitrification, N 2 o and N 2 emissions. Nutrient Cycling in Agroecosystems 52:187 194. vanGenuchten, M.Th., F.J.Leji, and S.R. Yates. 1991. The RETC code for quantifying the hydraulic functions of unsaturated soils USEPA/600/2 91/065. Verhoeven, J.T.A., W. Koerselman, and A.F.M. Meuleman. 1996. Nitrogen or phosphorus limited growth in herbaceous, wet vegetation: relations with atmospheric inputs and management regimes. Trends in Ecology & Evolution 11:494 497. Vitousek, P.M., and R.W. Howarth. 1991. Nitrogen limitation on land and in the sea: how can it occur? Biogeochemistry 13:87 115. von Wandruszka, R. 2006. Phosphorus retention in calcareous soils and the effect of organic matter on its mobility. Geochemical Transactions 7:1 6. Vose, J.M., W.T. Swank, B.D. Clinton, J.D. Kn oepp, and L.W. Swift. 1999. Using stand replacement fires to restore southern Appalachian pine hardwood ecosystems: effects on mass, carbon, and nutrient pools. Forest Ecology and Management 114:215 226. Vuc ns R., I. L pen te, and J. Livmanis. 2008. Comparison of methods for the determination of phosphorus in carbonatic soils Latvian Journal of Agronomy 11:299 305. Wade, D., J. Ewel, and R. Hofstette. 1980. Fire in South Florida e cosystems. USDA Forest Service, General Technical Report, SE 17. Waldrop, T.A., D.L. White, and S.M. Jones. 1992. Fire regimes for pine grassland communities in the Southeastern United States. Forest Ecology and Management 47:195 210. Walworth, J.L. and M.E. Sumner. 1987. The diagnosis and recommendation integrated system In Stewart B.A. (ed.). Advances in soil science vol. 6. Springer, New York, pp 149 188. Wan, S., D. Hui, and Y. Luo. 2001. Fire effects on nitrogen pools and dynamics in terrestrial ecosystems: a meta analysis Ecologica l Applications 11:1349 1365. Wang, Q., Y.C. Li, and W. Klassen. 2005. Determination of cation exchange capacity on low to highly calcareous soils Communications in Soil Science and Plant Analysis 36:1479 1498. Weather Chanel, 2008. Avaialble at http://www.weather.com/

PAGE 285

285 Weston, C.J., and P.M. Attiwill. 1990. Effects of fire and harvesting on nitrogen transformations and ionic mobility in soils of Eucalyptus regnans forests of South Eastern Australia. Oecologia 83:20 26. White, P., an d A. Robson. 1990. Response of Lupins ( Lupinus angustifoliu s L.) and Peas ( Pisum sativum L.) to Fe defi ciency induced by low concentrations of Fe in solution or by addition of HCO 3 Plant and Soil 125:39 47. Will, R.E., G.A. Barron, E. Colter Burkes, B. Shiver, and R.O. Teskey. 2001. Relationship between intercepted radiation, net photosynthesis, respiration, and rate of stem volume growth of Pinus taeda and Pinus elliottii stands of different densities. Forest Ecology and Management 154:155 163. Williams, M.W., J.S. Baron, N. Caine, R. Sommerfeld, and R. Sanford. 1996. Nitrogen saturation in th e Rocky Mountains. Environmental Science & Technology 30:640 646. Wilson, C.A., R.J. Mitchell, L.R. Boring, and J.J. Hendricks. 2002. Soil nitrogen dynamic s in a fire maintained forest ecosystem: results over a 3 year burn interval. Soil Biology and Biochemistry 34:679 689. Xu, Z.H., J.A. Simpson, and D.O. Osborne. 1995. Mineral nutrition of slash pine in subtropical australia. ii. foliar nutrient response to fertilization. Nutrient Cycling in Agroecosystems 41:101 107. Yan, F., S. Schubert, and K. Mengel. 1996. Soil pH increase due to biological decarboxylation of organic anions. Soil Biology and Biochemistry 28:617 624. Yang, J.E., C.A. Jones, H.J. Kim, and J.S. Jacobsen. 2002. Soil inorganic phosphorus fractions and olsen p in phosphorus responsive calcareous soils: effects of fertilizer amount and incubation time. Communications in Soil Science & Plant Analysis 33:855 871. Yermakov, Z., and D. Rothstein. 2006. Changes in soil carbon and nitrogen cycling along a 72 year wildfire chronosequence in michigan jack pine forests. Oecologia 149:690 700. Zhang, M., A.K. Alva, Y.C. Li, and D.V. Calvert. 2001. Aluminum and iron fractions affecting phosphorus solubility and reactions in selected sandy soils Soil Science 166:940 948. Zhang, M.K., Z.L. He, D.V. Calvert, P.J. Stoffella, Y.C. Li, and E.A. Limb. 2002. Release potential of phosphorus in florida sandy soils in relation to phosphorus fractions and adsorption capacity. J. Environ. Sci. Health. A37 (5):793 809. Zhang, Y., J. Jiang, and M. Chen. 2008. MINTEQ modeling for evaluating the leaching behavior of heavy metals in mswi fly ash Journal of Environmental Sc iences 20:1398 1402.

PAGE 286

286 Zhou, M., and Y.C. Li. 2001. Phosphorus sorption characteristics of calcareous soils and limestone fro m the Southern Everglades and adjacent farmland s. Soil Sci. Soc. Am. J. 65:1404 1412. Zinati, G.M., Y.C. Li, and H.H. Bryan. 2001. Accumulation and fractionation of copper, iron, manganese and zinc in calcareous soils amended with composts J. Environ. Sci. Health. B36 (2):229 243.

PAGE 287

287 BIOGRAPHICAL SKETCH Chung T an Nguyen was born in the middle part of Vietnam and grew up in the countryside where is exposed to various agricultural activities. He graduated from the University of Agricu lture and Forestry (UAF) located in HoChi Minh City (formerly Sai Gon city) with Bachelor of Science in Forestry in 2000 H e worked for the Center for Information and Technology of the UAF from 2000 to 2001. He has been a lecturer at the University of Agriculture and Forestry since 2001 A ddition ally he served as the assistant o f vice P resident and the secretary of the Wetland Management project in the Mekong River Basin from 2001 to 2004. Chung Nguyen was awarded a full scholarship from the Vietna mese government for studying the m aster program in the United States in 2004, and c ompleted the Master degree in Environmental Sciences from the University of Colorado at Denver in 2006. After obtaining the M.Sc. degree, he was appointed as the assistant President of the University of Agriculture and Forestry from 2006 to 2007. He was gr anted a four year match assistantship from the School of Natural Resources and Environment, and a two year f ellowship from the Vietnam Education Foundation for studying the Ph.D program in Interdisciplinary Ecology at the University of Florida in 2007. Ch ung will be an assistant professor at the University of Agricu lture and Forestry after he re ceives the Ph.D. degree from the University of Florida in the fall of 2011