![]() ![]() |
![]() |
UFDC Home | myUFDC Home | Help |
|
|
||||||||||||||||||||||||||||||||||||||||||||
Full Text | |||||||||||||||||||||||||||||||||||||||||||||
ECONOMIC AND ENVIRONMENTAL ANALYSIS OF TREE CROPS ON MARGINAL LANDS IN FLORIDA By MATTHEW HARVEY LANGHOLTZ 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 2005 Copyright 2005 by Matthew Langholtz ACKNOWLEDGMENTS I am grateful to my committee chair, Dr. Donald L. Rockwood, for his academic guidance and support, as well as Drs. Janaki R.R. Alavalapati, Douglas R. Carter, Alex Green, and Dr. P.K. Ramachandran Nair, for their tenacious work ethics and lifelong dedications to forestry and sustainable land-use systems. I particularly thank Dr. Nair for his assistance with my master's degree and assistance with acquiring funding for my Ph.D. program. This research was made possible by direct and indirect support from Woodward and Curran on behalf of Orlando/Orange County, Florida, Iluka Resources, Inc., the Florida Institute for Phosphate Research, and the Alumni Fellowship from the University of Florida College of Agricultural and Life Sciences. I thank my wife Maribel who has sacrificed for me to pursue this degree. I thank my family for their support, and particularly my sister Gabrielle for her assistance with the English language. Special thanks go to a long history of phytoremediation lab crews, Jared Mathey, Paul Proctor, Mark Torok, Richard Cardellino, Mauricio Arias, Geoff Filshe, Chris Cosby, Erin Maehr, Brian Becker, Bijay Tamang, and Luis Achugar. TABLE OF CONTENTS page A C K N O W L E D G M E N T S ................................................................................................. iii LIST OF TABLES ....................................................... ............ .............. .. vii LIST OF FIGURES ............................... ... ...... ... ................. .x ABBREVIATIONS AND ACRONYMS ..............................................................xii A B S T R A C T .............................................. ..........................................x iii CHAPTER 1 IN TR OD U CTION ............................................... .. ......................... .. B ack g rou n d ...................................... .............................. .... ......... ...... . R claim ed W after .............................................. ...... .... .............. .. P hosphate M ined L ands ............................................................. .....................2 T itanium M ined L ands ........................................ .......................................3 O bj ectiv e s ................................................................... ................................. . .4 Literature Review ............................................................................. ...... 4 Environmental Impacts of SRWC Production ............................................... 4 C arbon sequestration ........................................ .. .............. ...... .... .......... .. Phytoremediation and reclamation....................... ....................... 6 Slash Pine Productivity on M ined Lands ........................................ ...................7 P o licy ............... ............................................................. 7 E conom ics ............................................................. 8 Forest Financial Analysis ............. .............. .................... 11 Procedures ............... .............................. .........12 The Study A reas and Scope........................................... .......................... 12 M methodology ............. ..... .. .......... ................................. 13 Optimization of non-coppicing species......................... .. ............... 13 Optimization of coppicing species ................................. ....................... 14 Valuation of the non-timber benefits ........................................................ 16 Optimization of non-coppicing with inclusion of the non-timber benefits ............... .............. ........................ ........................ 17 Optimization of coppicing species with inclusion of non-timber benefits ............. .... ................... .......... ........... 18 2 EFFECT OF DENDROREMEDIATION INCENTIVES ON THE PROFITABILITY OF SHORT-ROTATION WOODY CROPPING OF Eucalyptus grandis ........................ .................... ... ........... ......... 20 In tro d u ctio n ......................................... .. ............................................2 0 M methodology ........................................................... .... .... ......... 23 Optim ization of Coppicing Species............... ............ ... ............... .. 23 Optimization of Coppicing Species with Inclusion of the Dendroremediation S erv ic e ............................................................................................ .... 2 4 M odel inputs ................................................................... ................ 2 6 R results and D discussion ......................................................... ... ...... ........... ...30 Sensitivity Analysis of Dendroremediation Incentive and Interest Rate..................35 C o n c lu sio n s..................................................... ................ 3 8 F future R research ....................................................................... 39 3 AN ECONOMIC ANALYSIS OF Eucalyptus SPP. AS SHORT-ROTATION WOODY CROPS ON CLAY SETTLING AREAS IN POLK COUNTY, F L O R ID A ............................................................................................................. 4 1 In tro d u ctio n .......................................................................................4 1 M e th o d o lo g y ....................................................................................................4 3 M odel Inputs ......................................................................................................48 G row th Function................................................... 48 C a rb o n V a lu e s ............................................................................................... 5 1 Market Assessment...............................................52 The established market: mulch ..............................................52 M ulchw ood price..................................................... 53 M ulchwood quantity ................. ...............................55 Potential m market: biom ass fuels............................................. 56 O operational C osts ............................................................59 Additional Non-Timber Benefits ........................ ...................................... 59 Below-ground C sequestration .........................................59 R eclam action incentives ............................................ .............. 61 Summary of Model Inputs and Assumptions .....................................................62 R results and Sensitivity A nalysis........... ...... .............................. .... .... ..............62 Conclusions................ ......... ... .. .... .. .... ................ 69 F future R research .......................................................................7 1 4 ECONOMICS OF SLASH PINE CULTURE ON TITANIUM MINED LANDS IN NORTH CENTRAL FLORIDA ............................................................... 73 Intro du action ............................................................................................. 7 3 M eth odology ................................ ........................................................... 7 5 E conom ic M odel .................. ................ ........................................... 75 G row th and Y field M odel ................................................. ................... ...... 76 M market A ssessm ent ............................................................ ............... 80 Silvicultural A lternatives....................................... ............... ............... 82 v C o st A ssu m ption s .............................................................. ......................93 Sim ulations ............................................................................................... .......95 R e su lts .................. ........................................................................9 8 E established Stands ............................... .. .. ......... .... ................ .. 98 Soil Amendments on Young Plantations.................................. ...............101 S en sitiv ity A n aly sis ...... .. ............................................ .......... .... ................ 103 C o n c lu sio n s.......................................................................................................... 1 0 5 F future R research ............................................................ ................... .... 106 5 C O N C L U SIO N S ................................ ........................ ................ ..... .......... 108 Sum m ary of Results....................... .... ................... .............. 108 SRWC Production with Reclaimed Water ...................................................... 108 SRWC Production on Clay Settling Areas ...................................................... 109 Slash Pine Production on Titanium Mined Lands............................................110 O overall Policy Im plications .......................................................... .............. 110 Future Research .................................... ............................ ... ......... 118 APPENDIX A BIOMASS CONVERSIONS.......................................................... ............. 121 B CHAPTER 4 LEV OUTPUTS ...........................................................................122 C P H O T O S ...................................................................... 12 7 L IST O F R E FE R E N C E S ......................................................................... ................... 130 BIOGRAPH ICAL SKETCH .............................................................. ............... 141 LIST OF TABLES Table p 1-1 Farmgate (production and harvest) costs for SRWCs and herbaceous biomass crops in Florida and other regions ........................................ ........................ 9 1-2 Sum m ary of study sites. ................................................ ............................... 13 2-1 Net returns and optimum stage lengths for a Eucalyptus grandis short-rotation woody crop system irrigated with reclaimed water. ..............................................31 2-2 Optimum LEVs, optimum stages per cycle, and optimum stage lengths for a range of dendroremediation values for Eucalyptus grandis irrigated with reclaim ed w ater in central Florida................................... ............................. ......... 32 2-3 Marginal increases in net returns ($ ha-1) per dollar ofN dendroremediation incentive for Eucalyptus grandis in central Florida. .........................................35 2-4 Changes in profit ($ ha-1) for Eucalyptus grandis in central Florida as interest rate increases from 4% to 5% and 5% to 6% ........................... ..................37 2-5 Estimated parameters and descriptors used in Eq. (2-9) of Eucalyptus grandis irrigated with reclaimed water in central Florida (R2>0.99). ..................................37 3-1 Number of observations, average DBH (cm), height (m) and inside-bark dry above-ground biomass yields of EG and EA ..................................................50 3-2 Mulch markets for Eucalyptus produced in Polk County. .....................................54 3-3 Estimated equivalent stumpage values for high and low transportation cost scenarios. All tons are green weight............... ............................... 55 3-4 Potential bioenergy markets for Eucalyptus produced in Polk County ...................58 3-5 LEV, optimum number of stages and optimum stage length for each stage by C benefit scenario and biomass price...................... ......................... 63 3-6 LEV ($ ha-1), optimum stage lengths, marginal benefit, and estimated below- ground benefit ($ ha-) by C sequestration incentive ($ Mg-1)..............................64 3-7 Change in LEV ($ ha-)) per 1% increase in interest rate.......................................65 3-8 Optimum harvest scheduling (stage lengths and number of stages per cycle) at interest rates of 4%, 7%, and 10%.. .......................................................66 3-9 LEVs and marginal impact on LEVs by changes in site preparation, planting and weeding costs.................... ............................... .. 67 3-10 Estimated discounted value of below-ground C benefits by C price, interest rate and grow th function. ........................... ........... ...... ...... ...... ...... 68 4-1 Merchantable standards of DBH (ddbh) and top diameter outside bark (di). ............79 4-2 Treatments included in the SRWC-84 and SRWC-84-2001 studies......................83 4-3 SRWC-84 age 5 and SRWC-84-2001 age 4 mined (SM) and unmined (UM) average heights, standard deviation and Duncan grouping ....................................84 4-4 SRWC-84 age 5 and SRWC-84-2001 age 4 mined (SM) and unmined (UM) average survival (%) and standard deviation by treatment.................................... 85 4-5 Average of top Duncan group survival of SRWC-84 and SRWC-84-2001 ..............92 4-6 2004 Average pine plantation establishment costs for the southeast U.S................93 4-7 Cost scenarios based on Smidt et. al (2005). ................................... ..................... 94 4-8 Land type, measurement age, measurement date, and number of 63 1/50th ha plots used in the analysis of established stands .............. .....................................95 4-9 Number of plots, average SI, SI standard deviation, average LEV, and standard deviation of LEV on mined and unmined lands................... ..................................99 4-10 SI (base age 25), survival, age of survival, cost of initial rotation, cost of subsequent rotations, LEV and optimum rotation........................................101 4-11 Volume, cost, LEV and IRR of comparative mined land simulations................... 102 4-12 Minimum growth response in SI needed to meet or exceed a base scenario's L E V ...............................................................................104 4-13 Maximum initial and subsequent rotation establishment costs tolerated to meet or exceed a base scenario's LEV ..................................... ......................... 104 5-1 Summary of LEVs ($ ha-) of EA production on CSAs............... ................... 110 5-2 Delivered costs of biomass for fuel, costs of electricity, and resulting divergence from costs of conventional electricity ............................................... 112 B-l Titanium mined plot measurement date, age, volume, SI and stand density.........123 B-2 Titanium mined LEV and optimum rotation age. .............................................124 B-3 Unmined plot measurement date, age, volume, SI and stand density.................. 125 B-4 Titanium mined LEV and optimum rotation age. .............................................126 LIST OF FIGURES Figure pge 2-1. Estimated high and low growth and yield functions for Eucalyptus grandis at W inter G arden, Florida.. .............................. ... ........................................28 2-2. Net returns ($ ha-1) as a function of dendroremediation incentive ........................36 3-1. Inside bark yields (dry Mg ha-) of EA and EG on a CSA near Lakeland, Florida for 5 treatm ents........................................................................ 49 3-2. Observed and predicted inside bark stem yields of EA. ................ .................. 51 3-3. Location and potential consumption of buyers of woody biomass from Polk C o u n ty ........................................................... ................ 5 6 4-1. Mean heights estimated by stem analysis from stands on 25 reclaimed and 25 unm ined sites (M they, 2001) ............ ........................................... ............... 74 4-2. Mean diameter inside bark (DIB) estimated by stem analysis from stands on 25 reclaimed and 25 unmined sites (M they, 2001). ............. ..................................... 74 4-3. Representative pulp, chip-and-saw, sawtimber, and total outside bark volumes (m 3 h a ) .......................................................................................8 0 4-4. South-wide pine stumpage prices quarterly averages from 1995-2005 (Timber M art South 2005). ........................ ........................ .. .... ...... ...............8 1 4-5. Average heights (m) by age (year) and treatment, SRWC-84 mined site ..............85 4-6. Average survival (%) by age (year) and treatment, SRWC-84 mined site .............86 4-7. Average heights (m) by age (year) and treatment, SRWC-84 unmined site............86 4-8. Average survival (%) by age (year) and treatment, SRWC-84 unmined site..........87 4-9. Average heights (m) by age (year) and treatment, SRWC-84-2001 mined site. .....88 4-10. Average survival by age (year) and treatment, SRWC-84-2001 mined site............88 4-11 Average heights (m) by age (year) and treatment, SRWC-84-2001 unmined site .............................................................. ................ 8 9 4-12. Average survival by age (year) and treatment, SRWC-84-2001 unmined site........89 4-13. Height (m) by treatment of SRWC-84-2001 (age 4) and SRWC-84 (age 5), satellite mined (SM ) and unmined (UM ).. ....................................... ...............90 4-14. Average heights (m) of subsoiled and not subsoiled treatments on SRWC-84- 200 1 m ined land.. ...................................................................... 9 1 4-15. Average survival (%) of subsoiled and not subsoiled treatments on SRWC-84- 200 1 m ined land.. ...................................................................... 9 1 4-16. Total predicted above-ground inside-bark volume (m3 ha-1) for sim ulations 1-9. .......................................................................98 4-17. LEV ($ ha-1) by SI (m, base age 25) for 34 and 29 stands (Table 4-8) on mined and unm ined lands, respectively.. ....................... .............................................100 4-18. LEV ($ acre-1) by SI (ft, base age 25) for 34 and 29 stands (Table 4-8) on mined and unm ined lands, respectively.. ....................... .............................................100 5-1. Additional cost of electricity (COE) (cents kWh-1) over COE from coal, for production on CSAs and under dendroremediation (WC2). ............. ...............113 5-2. Additional delivered cost of fuel (COF) ($ dry Mg-1) over COF coal equivalent, for production on CSAs and under dendroremediation (WC2) ..........................113 5-3. Estimated value of CO2 mitigation service, dendroremediation service, additional COF and COF coal equivalent ($ dry Mg-1), for production on CSAs and under dendrorem edition (W C2).................................................................. 115 5-4. External costs for 14 generation technologies (Roth & Ambs, 2004b). ..............116 C-1. Rapid infiltration basins near Winter Garden, Florida ................ .....................127 C-2. 1.75-year-old Eucalyptus grandis irrigated with reclaimed wastewater at the Water Conserv II facility near Winter Garden, Florida...................................127 C-3. Cogongrass (Imperata cylindrica) following herbicide treatment on a clay settling area near Lakeland, Florida. ........................................... ............... 128 C-4. A three-year-old Eucalyptus grandis stand on a clay settling area near Lakeland, Florida. ................................................................. 128 C-5. Dredge mining by Iluka Resources, Inc., near Green Cove Springs, Florida........129 C-6. Eight-year-old slash pine stands on mined (left) and unmined lands. .................129 ABBREVIATIONS AND ACRONYMS C carbon COE cost of electricity COF cost of fuel CSA clay settling area DBH diameter at breast height DFSS dedicated feedstock supply system DIB diameter inside bark EA Eucalyptus amplifolia EG Eucalyptus grandis FASOM Forest and Agricultural Sector Optimization Model FIPR Florida Institute of Phosphate Research FONC first order necessary condition ha hectare IPCC Intergovernmental Panel on Climate Change IRR internal rate of return kWh kilowatt hour LCOE levelized cost of electricity LEV land expectation value LHS left hand side MAI mean annual increment Mg metric ton MSY maximum sustained yield N nitrogen NTB non-timber benefit P Phosphorus REPI Renewable Energy Production Incentive RHS right hand side RIB rapid infiltration basin RPS Renewable Portfolio Standard SI site index SM satellite mined SOC soil organic carbon SRWC short-rotation woody crop TPA trees per acre TPH trees per hectare UM unmined WC2 Water Conserve II WUI wildland urban interface 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 ECONOMIC AND ENVIRONMENTAL ANALYSIS OF TREE CROPS ON MARGINAL LANDS IN FLORIDA By Matthew Langholtz December 2005 Chair: Donald L. Rockwood Major Department: Forest Resources and Conservation Tree crops can be used to remove contaminants from reclaimed wastewater, restore ecological functions of phosphate and titanium mined lands and to provide renewable energy in Florida. The economic feasibility of these potential tree crop systems, the value of environmental services they provide and opportunities to make up the current difference between minimum feasible and current market prices are investigated. Profitability measured as land expectation value (LEV) of 128 scenarios of Eucalyptus grandis cropping irrigated with reclaimed wastewater ranged from -$2,343 to +$2,762 ha-1 and was greatly reduced by high interest rates, high irrigation costs, and low yields. Each $1 kg-1 N increment in a dendroremediation incentive increases profit by $223-$376 ha-1, depending on interest rate and site productivity. Optimum management requires harvests every 2.6 to 4.0 years and replanting after two or three harvests, though the optimum number of stages per cycle would increase with improved coppice growth. LEVs of Eucalyptus amplifolia cropping on phosphate-mined lands in central Florida ranged from $762 to $6,507 ha-1 assuming interest rates of 10% and 4%, respectively, establishment costs of $1,800 ha-1, planting costs of $1,200 ha-l, high yields, and a stumpage price of $20 dry Mg-1, excluding CO2 mitigation incentives. Incorporating CO2 mitigation incentives increased LEV, particularly when incentives recognize the CO2 emissions reduced by biofuels use. Optimum management necessitates harvests every 2.5 to 3.5 years and replanting after two or five harvests. Average LEVs of slash pine (Pinus elliottii) stands established on titanium mined lands varied widely with productivity, but on average were profitable and similar to those of unmined lands. Optimum management is comparable to that of conventional slash pine culture in northeast Florida. Early-rotation responses to soil amendments suggest that growth and survival can be improved by fertilizer and subsoil treatments, respectively. Plantation establishment costs including soil amendment as high as $423 to $878 ha-1 ($171 to $355 acre-1) are economically viable depending on growth response. CHAPTER 1 INTRODUCTION Background Landowners in Florida are interested in profitable land-use options. Short-rotation woody crops (SRWC) in Florida have competitive growth rates and production costs compared to other states (Rahmani et al., 1997) and can provide multiple environmental benefits. Currently, three opportunities in Florida have the potential to contribute to forest production: 1) reclaimed water from the Water Conserv II (WC2) facility in Orlando, 2) clay settling areas (CSAs) on former phosphate mined lands, such as the Kent site in Lakeland, and 3) reclaimed titanium mined lands such as those associated with the Iluka mining operation near Green Cove Springs. The financial feasibility of producing Eucalyptus spp. or slash pine (Pinus elliottii) on these marginalized lands and irrigated with reclaimed water requires further research. Reclaimed Water The population of Florida is expected to nearly double in the next 30 years, from 15.9 million permanent residents in 2000 to a projected 30.1 million in 2030 (Bureau of Economic and Business Research, 2001), which will result in an increase in both water consumption and wastewater production. As Florida faces increasing pressure on water resources, wastewater presents both a challenge of disposal and an opportunity for reuse. The WC2 facility near Winter Garden is an innovative water reuse program that has achieved international recognition for its water conservation and reuse methods. Waste water from the Orlando area, following treatment, is pumped to ornamental nurseries and to 8,600 acres (3,480 ha) of citrus groves on sandhills of Orange and Lake Counties (State of Florida, 2003). In light of increasing competition from overseas production, particularly Brazil, future trends of citrus production in Florida are uncertain. The preliminary $760 million on-tree value of all Florida citrus for the 2001 season is the lowest since the 1985-86 season, while production was also down 7% from 1999-2000 (Florida Agricultural Statistics Service, 2003). SRWC production using reclaimed water could provide a crop alternative for citrus producers and other landowners on sandhill soils in Florida. Furthermore, SRWCs can be used to extract nitrogen, phosphorous and chlorides from the reclaimed water prior to its infiltration to the aquifers. Research at WC2 between 1998 and 2001 suggests that E. grandis and cottonwood (Populus deltoides) can yield 13.1 and 10.9 Mg ha-1 year1, respectively, and have great potential to extract NO3-N and NH4-N from reclaimed water (Rockwood et al., 2002a). Phosphate Mined Lands Central Florida produces 75% of the nation's and 25% of the world's phosphate supply (IMC Phosphates, 2002). This phosphate is used primarily in agriculture, and also in a range of consumer products. In the mining process, the surface soil is removed and put aside as "overburden" and clays are separated from phosphates and then sent to CSAs. There are about 162,000 hectares (400,000 acres) of phosphate-mined lands in Florida (Segrest, 2003). The Lakeland area contains over 38,000 hectares (95,000 acres) of CSA and/or overburden soil, as a result of phosphate mining. These CSAs, classified as clayey Haplaquents, can be a valuable resource for biomass production (Mislevy et al., 1989). Reclamation and reuse of mined landscapes are major foci of the Florida Institute of Phosphate Research (FIPR), an independent State research agency that has spent almost $11 million on research related to this topic. One project documented growth of cottonwood, E. grandis, and E. amplifolia on a CSA in Lakeland, FL. After 15 months cottonwood reached average heights of 4.7 and 6.1 m on double row planting (8,400 trees ha-1) and single row planting (4,200 trees ha-1) configurations, respectively; E. grandis after 9 months reached average heights of 3.8 and 2.8 m on double row planting and single row planting, respectively; E. amplifolia after 9 months reached average heights of 2.3 and 3.1 m on double row and single row planting, respectively (Rockwood et al., 2002b). These initial results suggest that cottonwood, E. grandis, and E. amplifolia are suitable for conditions on CSAs in central Florida. FIPR continues to fund research related to SRWC production on CSAs, underscoring the demand for such research. CSAs, titanium mined lands, and other marginal lands would have a low opportunity cost, as most of these ownerships are idle. Tree crop production could provide a valuable land-use alternative to these areas. Titanium Mined Lands Near Green Cove Springs, Iluka Resources Inc. has been producing titanium minerals and zircon since 1972 using dredge mining and satellite dry mining. These reclaimed mines are used extensively for slash pine plantations to produce timber products, reclaim soil productivity, and re-establish wildlife habitat. Following research of slash pine productivity and culture on mined lands by Mathey (2001) and Proctor (2002), an economic analysis of silvicultural options on reclaimed titanium mines can help landowners make sound forest management decisions. Objectives The general objective of this research is to determine the feasibility of tree crops grown on reclaimed mine lands or using reclaimed water in northern and central Florida. Specific objectives include the following: 1. Evaluate financial and environmental aspects of SRWC production on lands irrigated with reclaimed wastewater. 2. Evaluate financial and environmental aspects of SRWC production on CSAs. 3. Determine the financial viability of slash pine production on reclaimed titanium mined lands. Literature Review Environmental Impacts of SRWC Production As a production-oriented land-use option, SRWCs have the potential of providing environmental services such as reductions in C02 emissions, carbon sequestration, and soil stabilization. Carbon sequestration Atmospheric concentrations of CO2 have increased from 280 parts per million (ppm) in the year 1850 to 370 ppm near the end of the 20th century. This increase has been attributed to the use of fossil fuels for energy and has been associated with increased global temperatures. By the year 2100, CO2 concentrations are expected to rise to between 540 and 970 ppm, and temperatures are expected to increase by 1.4 OC to 5.80C (IPCC, 2001b). If temperatures increase, various environmental changes will occur, including rising sea levels, loss of coastline, changes in ocean currents, changes in precipitation, and a variety of associated changes in agricultural production, habitat changes and shifts, and disease distribution (IPCC, 2001a). As with other types of forests, SRWC plantations sequester atmospheric carbon as they grow, store the carbon on the site and in the products they produce (until the stand and products are oxidized), and provide alternatives to products that produce CO2 emissions. For these reasons, the production of SRWCs would have implications for reducing atmospheric CO2. Models by Heath et al. (1993) and Turner et al. (1995) show that U.S. forests will continue to sequester atmospheric carbon for the next 40 years. Barker et al. (1995) evaluate the potential of the U.S. Conservation Reserve Program to offset greenhouse gas emissions in the United States through carbon sequestration. Their simulations suggest that intensive afforestation on environmentally sensitive cropland could sequester about 16 Tg C equivalent. In addition to carbon sequestration, they also identified the potential reduction of CO2 emissions through the production of biofuels and fuelwood, and identified this possibility as needing research. While the production of SRWCs has various potential environmental benefits, a feasibility analysis must consider potentially negative environmental implications as well. According to a review of research up to 1998, the establishment of SRWC plantations is beneficial to some wildlife species but is detrimental to others, and these impacts need to be taken into consideration in the planning of plantation establishment at the landscape level (Tolbert & Wright, 1998). Environmental impacts of the establishment of SRWCs during the first year are not unlike those of the production of annual crops (Joslin & Schoenholtz, 1997). SRWC plantations are expected to improve surface runoff and groundwater quality when compared to annual crops following the first establishment year (Thornton et al., 1998). The land condition prior to establishment should also be considered, and SRWC establishment may be an improvement over post-mining conditions. In Sweden, Salix is valued for its benefits in promoting wildlife diversity and its capacity to phytoremediate cadmium contaminated soils (Perttu, 1998), as well as its minimal need for herbicides and its contribution to soil organic matter (Ledin, 1998). The environmental benefits of SRWCs and resistance to insect pests and weed species are reiterated by Sage (1998). Indeed, Abrahamson et al. (1998) propose that SRWC production systems in New York are ecologically and environmentally sustainable and that the limit to production is economic viability. They conclude that environmental and ecological benefits of the system should act as an impetus for developments needed to overcome the economic constraints of the system. Phytoremediation and reclamation Biomass production systems can be associated with phytoremediation objectives, as is being done at a research site at WC2 in Orlando, and an arsenic contaminated site in Archer. Eucalyptus sp. and cottonwood are identified as SRWC candidate species with potential to accumulate nutrients and mitigate problems associated with urban waste and stormwater runoff (Rockwood et al., 1995b; Pisano, 1998; Moffat et al., 2001). Moffat et al. conclude that approximately 100 m3 ha-1 yr-1 is an ideal application rate of effluents, and their results suggest that sewage sludge should be applied to every rotation of the SRWC rather than annually. Corseuil and Moreno (2001), Watson et al. (1999), Gommers et al. (2000), and Vervaeke et al. (2001) discuss research related to the use of willow in phytoremediation. Thompson et al. (1998) describe changes in poplar respiration following exposure to 2,4,6-trinitrotoluene, and Heinonsalo et al. (2000) describe the effect ofPinus sylvestris root growth on soil hydrocarbon oxidation. Goor et al. (2001) assess the potential to use willow SRWC systems as a land-use alternative for farmland contaminated by the Chemobyl nuclear power plant disaster. A study by Bungart and Huttl (2001) of SRWC systems in post-mine landscapes indicates that SRWC production is an ideal land-use alternative for post-mining landscapes. Slash Pine Productivity on Mined Lands Slash pine growth on mine tailings north of Starke was constrained by high bulk density and low organic matter content, as well as extremes of soil moisture (Darfus & Fisher, 1984). Slash pine growth on mine tailings can be improved by minimizing variation in topography or adding inexpensive manure such as waste humate or sewage sludge. Alternatively, wetter and dryer areas of the tailings could be planted with cypress (Taxodium spp.) or longleaf pine (P. palustris), respectively. Mathey (2001) found site indexes and growth patterns of established slash pine plantations on mined and unmined lands to be similar. Proctor (2002) assessed growth responses to fertilizer and subsoiling in young plantations. The effectiveness of reclamation practices for mined lands may be assessed by the use of growth and yield models for slash pine plantations on mined and unmined sites. Pienaar and Rheney (1995) and Fang and Bailey (2001) developed height growth models accounting for intensive silvicultural treatments in slash pine plantations. Policy Some objectives of SRWC production include mitigation of CO2 emissions, improvements in air and water quality, employment generation, and other societal benefits. Policies that seek to reduce or internalize external environmental costs to society at large have great implications for SRWC production. Hohenstein and Wright (1994), Wright and Hughes (1993), and Graham et al. (1992) describe the potential for SRWCs to offset CO2 emissions in the U.S. Tuskan (1998) identifies research needs related to SRWCs in the U.S., including long-term use of fertilizers and irrigation and development of improved harvesting methods. Ehrenshaft and Wright (1991) describe a SRWC database management system, and modeled projections by Fischer and Schrattenholzer (2001) suggest that bioenergy could supply 15% of global primary energy by the year 2050. Economics The economic feasibility of SRWC production has been examined. Turhollow (1994) presented cost estimates for 1989 and 2010 for supplying biomass via five cropping strategies in five regions of the U.S. One of these strategies for the Midwest and South used SRWCs. Turhollow proposed that energy crops must sell at between $43 and $60 dry Mg-1 in 1989 and $30 and $43 dry Mg-1 in 2010 to be economically viable. Rahmani et al. (1997) described production costs of SRWCs in Florida as consisting of a) farmgate costs (production and harvest costs), and b) transportation costs. They estimated Florida eucalyptus farmgate costs at $32.00-$39.00 dry Mg-1 and yields at 20-31 dry Mg ha-1 yr-1. These cost estimates were generated using levelized costs and the AGSYS Budget Generator, which calculates inputs costs such as labor, machinery, fertilizer, etc., based on a database of costs of machinery and materials. The method of cost estimation did not affect the range of estimated farmgate costs. SRWC farmgate cost estimates in Florida ranging from $16-47 dry Mg-1 compare favorably with herbaceous biomass crops in Florida and are likely to be lower than other regions of the U.S. (Table 1-1). These numbers show that Florida has a competitive advantage in the production of SRWCs. Table 1-1. Farmgate (production and harvest) costs for SRWCs and herbaceous biomass crops in Florida and other regions. Crop $/Dry Mg $/Dry ton Florida, SRWCs: Leucaena 16-47 15-43 Eucalyptus spp. 32-39 29-36 Florida, herbaceous: Sugarcane 23-25 21-32 Elephantgrass 24-32 22-29 Other Regions: Poplar 33-132 39-120 Willow 30-110 27-100 TVA Estimatesa 32-69 29-63 Source: Rahmani et al. 1997; a Adapted from Downing and Graham (1996). Transportation costs of SRWC biomass in Florida have been estimated at $3.10 dry Mg-1 (Rahmani et al., 1997), assuming an average distance of 32 km (20 miles) and moisture content of 15%. This transportation cost was lower than estimates for herbaceous biomass crops, which ranged from $7.85-$12.28 dry Mg-1, attributable to moisture contents ranging from 20-75%. Turhollow et al. (1996) estimated transportation costs for herbaceous biomass crops ranging from $8.37-$13.95 Mg-1 with dry matter contents of 50% and 30%, respectively. Transportation estimates for Florida are competitive with these out-of-state costs. Projected prices and quantities of SRWCs are functions of the amount and quality of land, expected yields, production costs, and profit potential. Downing and Graham (1996) described potential SRWC-production scenarios in the Tennessee Valley Authority Region, which consists of parts of Tennessee, Kentucky, Virginia, North Carolina, Georgia, Alabama, and Mississippi. They defined farmgate costs for SRWCs, including sweetgum (Liquidambar styraciflua), sycamore (Platanus occidentalis), and poplar (Populus spp.), under a variety of soil- and land-value categories. Under projected yields ranging from 5.4-9.7 dry Mg ha-1 year-', farmgate costs ranged from $32-51 on former cropland, and from $48-69 dry Mg-1 on former pastureland; SRWC production costs ranged from $31.90-69.30 dry Mg-1. A farmgate price ranging from $44-55 dry Mg-1 would be needed to ensure profits similar to current land uses. Increasing SRWC biomass yields 25% decreased farmgate prices 20%. On a national level, SRWCs will probably help meet growing demands for pulpwood production. Currently, SRWCs are produced on fewer than 80,000 ha (200,000 acres) in the U.S., with most of this production in the Pacific Northwest, though a much greater area has potential for SRWC production. Alig et al. (2000) studied the economic potential of SRWCs on agricultural land in the U.S. They estimate that 0.6-1.1 million ha (1.5 to 2.8 million acres) would generate about 10 to 16 Tg year-', equivalent to about 40% of current U.S. hardwood pulpwood production. They used the Forest and Agricultural Sector Optimization Model (FASOM), an intertemporal, price-endogenous model, linking the U.S. forest and agricultural sectors. The SRWC was assumed to be hybrid poplar, using data from the U.S. Department of Energy Oak Ridge National Laboratory. Most of the current U.S. cropland was determined to be suitable for SRWC poplar production: 0.5, 13.7, 34.0, 5.7, and 35.1 million ha (1.2, 33.9, 84.0, 14.0, and 86.8 million acres) in the Pacific Northwest, Lake States, corn belt, Southeast, and South Central regions, respectively (Walsh et al., 1998). Under the FASOM projections, SRWC plantation area is 0.9, 1.1, 0.6, and 1.1 million ha (2.1, 2.8, 1.5, and 2.6 million acres) in the first, second, third, and fifth [sic] decades, respectively. Even at peak production, SRWCs are projected to occupy less than 1% of cultivated U.S. cropland area. This modeling was done to estimate the potential supply of wood fiber to the pulp-and-paper sector from SRWCs. However, the authors also mention the potential for SRWCs to produce non-pulp products such as veneer. This and fuel for bioenergy are alternative products that have the potential for increasing the demand for SRWCs. The results from Alig et al. (2000) indicate that the contributions of hardwood biomass could be relatively high when compared to the area of land involved. Interestingly, these increased yields could reduce U.S. forest plantation area and allow more U.S. forestland to be converted into agricultural production. Forest Financial Analysis Methods for determining the optimum forest harvest cycle length, or "rotation age," have improved over the past century as they have progressively internalized an increasing number of factors. The most basic of forest management objectives has been to produce as much forest product as possible by maximizing mean annual increment (MAI). However, this method fails to account for the establishment costs and the time value of money, i.e., "discount rate." The Fisher solution for forest optimization, while an improvement from maximizing MAI, failed to capture the opportunity costs associated with occupying the land (Rideout & Hessein, 1997). This limitation was addressed by the Faustmann model by projecting an infinite number of rotations in determining LEV that is included as a "land rent" cost in determining the optimum rotation age (Chang, 1984; Chang, 1998; Chang, 2001). Hartman (1976) modified the Faustmann solution to include the value of environmental services in determining LEV and optimum rotation age. While Medema and Lyon (1985) adapted the Faustmann solution to find LEV and optimum harvesting cycles for coppicing tree species, Smart and Burgess (2000) incorporated the valuation of environmental services in determining the LEV and optimum rotation age of SRWC willow production systems. Many environmental services that forests provide are non-market services or "externalities" for which timber producers have historically not been compensated. The internalization of these market externalities could provide incentives for landowners to manage their forests for the provision of environmental services. Environmental services have been categorized as "intrinsic" vs. "instrumental" (Farber et al., 2002) and as providing regulation, habitat, production, and information values (de Groot et al., 2002). The innovative forestry systems described in this research (e.g., mine reclamation and phytoremediation) are designed specifically for the instrumentally oriented environmental services such as regulation of soil and air quality and production of wood products or energy. The compensation for carbon sequestration services has been shown to lengthen the economically optimum rotation age (Plantinga & Birdsey, 1994; Stainback, 2002). Research regarding the impact of compensation for reduced carbon emissions on optimum rotation age (achieved through fossil fuel displacement, as opposed to compensation for increased carbon storage in the biomass or soil) is lacking. Procedures The Study Areas and Scope This research is relevant to Florida sandhill sites irrigated with reclaimed water, phosphate mined lands of central Florida, and titanium-mined lands of northeast Florida. Financial analyses were done on forestry scenarios represented by the following three sites: 1. Water Conserv II study site: WC2 near Winter Garden receives secondary treated effluent from the City of Orlando Water Reclamation Facility and Orange County South Regional Reclamation Facility. The water contains mean NO3-N and Cl- concentrations of 6.92 mg L-1 and 86 mg L1, respectively, and is currently supplied to approximately 70 agricultural customers free of charge irrigating 4,450 ha of citrus plantations. A 2.8-hectare study site at the WC2 facility is characteristic of sandhill Entisol soils where SRWCs could be produced using reclaimed wastewater. 2. Kent study site: a 57-hectare CSA in Lakeland is the site for research related to reclamation of phosphate mined lands. As an anthropogenic soil (see Phosphate Mined Lands), the soil is a clayey Haplaquent (Mislevy et al., 1989), with a pH of 7.0, with little organic matter. 3. Iluka study site: Titanium mined lands are represented by studies at Iluka mining company at Green Cove Springs, 55 km south of Jacksonville. This mine has been producing titanium minerals and zircon since 1972. The operations include a dredge mine, a satellite dry mine and associated mineral separation plants. Land in the area is used extensively for slash pine plantations. The Iluka site is situated on spodosol soils. Data sets utiltized include SRWC-72 from Winter Garden (1998-2003), SRWC-90 and plots in operational areas from Lakeland (2001- 2005), and SRWC-84 and SRWC- 84-2001 from Green Cove Springs (1999-2005) (Table 1-2). Table 1-2. Summary of study sites. Site Study Situation Species Dates Covered WC2, Winter Garden SRWC-72 Reclaimed Water EG, March 1998 to May CW 2003 Kent, Lakeland SRWC-90 CSA EA, EG May 2001 to January 2005 Iluka, Green Cove SRWC-84 Reclaimed Titanium SP November 1999 to Springs SRWC-84-2000 Mined vs. Unmined Plots January 2005 SP=Slash pine, EG=Eucalyptus grandis, EA=Eucalyptus amplifolia, CW=cottonwood. Methodology Optimization of non-coppicing species A financial feasibility analysis was applied through the use of the Faustmann solution to determine LEV and optimum rotation ages and coppice stage lengths as described above (Forest Financial Analysis). The basic Faustmann solution for a non- coppicing even-aged stand is defined as LEV= V(t) e C 1-1) -er*t \ where LEV is the land expectation value (i.e., the land value as defined by a forestry scenario repeated in perpetuity), V(t) is the value of the stand at time t (i.e., price times volume), C is cost of stand establishment at the beginning of the rotation, r is the interest rate, and t is the optimum rotation age. Optimum rotation age is determined by taking the derivative ofEq. (1-1), setting it equal to zero, and solving for t. The first order necessary condition (FONC) for the Faustmann solution is found by taking the derivative of Eq. (1-1) and rearranging, resulting in V(t)**'1- (1-2)- V'(t)= r*V(t)+r* _V( (1-2) which is equivalent to V'(t) = r*V(t)+r *LEV (1-3) or V'(t) =() r (1-4) V(t) +LEV Eq. (1-2) states that the FONC for the optimization of the Faustmann solution is the time t where the marginal benefit in growth represented by the left-hand side (LHS), just equals the opportunity cost of the forest capital and the land rent shown in right-hand side (RHS). This can alternatively be stated as the time t where the ratio of marginal benefit (growth) to opportunity costs of the forest capital and the land rent just equals the given interest rate r (Eq. (1-4)). This non-coppicing form of the Faustmann solution was used for the analysis of slash pine on titanium-mined sites. Optimization of coppicing species Determining optimum rotations of SRWC coppice systems at the Kent site and WC2 site differed from determining optimum rotations of non-coppicing systems. The optimum age of each coppice harvest must be determined as well as the optimum time to replant (i.e., the optimum number of stages before replanting). Following the terminology used by Smart and Burgess (2000), a coppice stage length describes the period of time between coppice harvests, while a coppice cycle length describes the period of time and/or number of coppice stages between replanting of the trees. Medema and Lyon (1985) modified the Faustmann formula (Eq. (1-1)) to solve for multiple coppice stage lengths given a fixed number of coppice stages (n): y n1 -e-* .rj=I) J LEV=- =_*y-- (1-5) 1-eJ-^ where to=0 n = the number of coppice stages, s, V(t) = the value as a function of time (as a function of growth of stage s), r = the real interest rate (excluding inflation), t = time, the rotation age in years of stage s, and Cs = costs of stage s discounted to the start of the stage. Eq. (1-5) defines LEV as the sum of the benefits of each coppice stage discounted to the present minus the sum of the costs of each coppice stage discounted to the present, for a fixed number of coppice stages projected in perpetuity. Cs indicates a cost that may be replanting the coppice cycle, or may be a different cost associated with each coppice stage, such as weeding costs. Estimates for the prices and costs associated with SRWC production at the Kent site and WC2 site came from the preliminary work done by the Common Purpose Institute at the Kent site. Solving for the optimum stage lengths and cycle lengths of the Kent site and WC2 site was a two-part solution. First, n was fixed and the optimum stage lengths were determined for the fixed number of stages per cycle (i.e., n was consecutively fixed for 1- 4). As with the non-coppicing optimization, the optimum stage length for each individual stage was the point at which the marginal benefit of the continued biomass growth over the next unit of time is just equal to the marginal opportunity cost of the forgone benefit due to not harvesting, plus the marginal cost of delaying all future coppice stages and cycles. The marginal cost of delaying all future coppice stages and cycles is defined as the LEV of the subsequent coppice stages multiplied by interest rate r. Next, the optimum number of stages was found by determining at what value of n an additional coppice stage to the coppice cycle (i.e., LEV of n+1 minus LEV of n) has a marginal benefit less than zero. Stated differently, the optimum number of stages is that which provides the highest LEV. As the number of stages n varies, the optimum stage lengths can also vary. Valuation of the non-timber benefits Two non-timber benefits (NTBs) included in the determination of LEV and optimum rotation ages were: 1. Phytoremediation ofwastewater. The Southern Regional Water Reclamation Facility of Orange County records costs associated with wastewater treatment, and the Florida Department of Environmental Protection regulates standards for wastewater treatment used to irrigate non-edible crops. Sewage water treatment costs were used to estimate the value of the phytoremediation services at WC2. 2. C sequestration, and offset of CO2 emissions. CO2 is a greenhouse gas covered in the trading policy of the Chicago Climate Exchange, Inc. and the International Carbon Bank and Exchange. These sources and others were used to provide ranges of potential values of C sequestration and CO2 emission reductions. As described above, several variables influence calculation of LEV and optimum rotation ages of tree crops at titanium mined lands, phosphate mined lands, and lands irrigated with reclaimed wastewater, including production costs, yields, product prices, interest rates and values of environmental services. The sensitivity of LEV and optimum rotation ages to variation of each of these factors was tested. Optimization of non-coppicing with inclusion of the non-timber benefits To identify the divergence between private and social value maximization, the methodology described by Hartman (1976) was used to include the values of mine reclamation and carbon sequestration in the analysis of slash pine on titanium mined sites. Hartman (1976) used an integration to account for social amenities associated with standing forest: t NTB(t) = NTB (n) e r*dn (1-6) 0 where NTB(t) was the present value of a stream of NTBs of one rotation quantified by the integration of the discounted value of these benefits according to stand age n. The NTB defined by Eq. (1-6) was an additional benefit to be added to the numerator of Eq. (1-1): jNTB(n) e *dn + V(t) e *t C LEV = o (1-7) 1 -e-*t Deriving the F.O.C. for optimality of the basic Hartman model (Eq. (1-7)) was the same as the derivation of the F.O.C. of the Faustmann model: (NTB(t)*e' + V'(t)* e' V (t) r*e re)* ( e' ) KNTB(n)*e-r'dn+V(t)*e-rt C *(r*e-ret (1= (1-8) NTB(t) e + V '(t) et = r *V(t)e +r N -(n) e Fd -r(t* e -* (1-9) NTB(t)+V'(t)= r *V(t)+r*LEV (1-10) NTB(t) + V'(t) =r (1-11) V(t) +LEV The NTB remains on the LHS of Eq. (1-10) as an additional marginal benefit, and an additional reason to delay the harvest of the stand. Similarly, the NTB in the numerator of the LHS of Eq. (1-11) serves to delay the time t at which the ratio of benefits to costs equals the interest rate r. Optimization of coppicing species with inclusion of non-timber benefits To internalize the values of mine land reclamation, wastewater phytoremediation, carbon sequestration, and reduction in CO2 emissions associated with Eucalyptus spp. culture at the WC2 and Kent sites, the methodology described by Smart and Burgess (2000) was used to include the social amenity in the analysis. The NTB of a given coppice stage can be defined as NTB, = (iNTB (t))e *e-))dt (1-12) where the NTB of stage s was the definite integral of the flow of the benefits for the duration of the stage, discounted from the time of the harvest of the previous stage. This NTB can be added to Eq. (1-5), the equation for the LEV of coppicing species: I nI r* s ty l 1j) ()-r*y l 1 ) LEV V(ts)*e e- J_ +NTB, *e -C, eg J1 LEV = (1-13) 1- e' rM' Eq. (1-13) accounts for the value of a NTB derived from keeping the trees in the field, i.e., delaying harvest. Conversely, one potential NTB of the SRWC, the reduction in CO2 emissions due to displacement of fossil fuels, was associated with the harvest of the trees. This NTB, which takes place at the time of the harvest of the trees, was treated 19 as an addition to the V(t), term in Eq. (1-13). This harvest-associated NTB served to decrease optimum stage length, counteracting the increase of optimum stage length due to NTBs derived from standing trees. This interaction was assessed in Chapter 3. CHAPTER 2 EFFECT OF DENDROREMEDIATION INCENTIVES ON THE PROFITABILITY OF SHORT-ROTATION WOODY CROPPING OF Eucalyptus grandis Introduction Water resources, traditional forest products, fire management, recreation, and wildlife are the five main elements of particular concern in the wildland-urban interface (WUI) (Macie & Hermansen, 2003). They note: "Municipal waste facilities in rapidly developing areas face difficulties with handling and treating increased waste loads...allocating high-quality, abundant flows of water and managing forest ecosystems at large watershed scales remain key challenges." Trees within and surrounding urban centers can provide a variety of environmental services including sequestering atmospheric carbon dioxide, enhancing biodiversity, providing aesthetics and recreation, and remediating urban wastewater. Nutrients from urban wastewater and other sources cause eutrophication and degradation of aquatic ecosystems. Increasing concentrations of nitrogen (N) and phosphorus (P) are compromising water quality in Florida. The population of Florida is expected to nearly double in the next 30 years, from 15.9 million permanent residents in the year 2000 to a projected 30.1 million in 2030 (Bureau of Economic and Business Research, 2001), which will result in an increase in both water consumption and wastewater production. As Florida faces increasing pressure on water resources, wastewater presents both a challenge of disposal and an opportunity for reuse. Trees can mitigate nutrient loading by extracting N from reclaimed wastewater, thus improving both water quality and tree growth, and reducing fertilizer inputs. This chapter assesses economic impacts of incentives to use fast-growing trees to remove N from reclaimed wastewater discharged from an urban center. Florida Administrative Code Chapter 62-610 mandates primary treatment (removal of biosolids), secondary treatment (removal of dissolved elements), and basic disinfection at sewage treatment plants (State of Florida, 2004b). Reclaimed water leaving the Southern Regional Water Reclamation Facility of Orange County, Florida contains 7 ppm nitrate N (P. Duel, Orlando Wastewater Treatment Plant Manager, pers. comm., February 2004). Following treatment, the reclaimed water can be used for irrigation. For example, 132,500 m3 (35 million gallons) day1 of reclaimed water is pumped 35 km from sewage treatment plants in Orlando and surrounding areas to the Water Conserv II, a reclaimed water distribution facility, where 60% of this reclaimed water is applied to 1,700 ha of citrus groves, ornamental nurseries, and golf courses. The remaining 53,000 m3 (14 million gallons) day'1 of reclaimed water is pumped into open sand pits called rapid infiltration basins (RIBs), where the water percolates into the Florida aquifer (State of Florida, 2003; Rabbani & Munch, 2000). The use of trees to extract contaminants from soil or water is defined as dendroremediation (Rockwood et al., 2004). An example is using tree plantations as a tertiary or "finishing" treatment to remove N from reclaimed water (Aronsson & Perttu, 2001; Labrecque et al., 1997; Perttu, 1998; Rosenqvist et al., 1997; Pisano, 1998; e.g. Licht & Isebrands, 2005). Dendroremediation is used to address urban waste problems in Sweden and Finland (Ettala, 1987), the United Kingdom (Alker et al., 2002), Canada (Gordon et al., 1989) and Hong Kong (Wong & Lueng, 1989). As an alternative to releasing reclaimed water in RIBs, it could be dendroremediated by SRWCs. Research at Water Conserv II between 1998 and 2001 suggests that E. grandis irrigated with reclaimed water can yield about 13 dry Mg ha-1 year-, and extract over 300 kg nitrate nitrogen (N) ha-1 year- (Rockwood et al., 2001). As the citrus industry in Orange County is projected to decline, biomass crops present an alternative that can produce wood for rough sawtimber, landscape mulch, or biomass for renewable energy. In addition to dendroremediation of reclaimed water, SRWC production can generate employment (Borsboom et al., 2002) and sequester carbon in above- and below-ground biomass, and soil organic carbon (Eriksson et al., 2002). If the Florida Department of Environmental Protection mandates renewable portfolio standards, SRWC biomass may be used to displace fossil fuels in electricity generation, providing additional benefits including reduction ofCO2, NOx, and SOx emissions and diversification of domestic energy resources (Segrest et al., 1998; Stricker et al., 2000; Roth & Ambs, 2004a). Environmental economists suggest incorporating environmental benefits and costs as an effective strategy using market forces work to correct externalities (Van Kooten & Bulte, 2000). In the face of increased wastewater treatment standards, producers of wastewater search for cost effective mitigation strategies. On the other hand, tree growers who could use wastewater as an input in their production process may be willing to provide a service by utilizing wastewater. The optimum use of wastewater (on a voluntary basis) by a tree grower depends on the marginal cost and marginal productivity of wastewater use. In the face of incentives for using wastewater, however, it is likely that tree growers could use wastewater at a level higher than that of voluntary use. This approach can be a win-win situation for wastewater producers and tree growers. Dendroremediation of municipal wastewater by willow SRWCs in Sweden is economically feasible, despite a growing season of six months (Rosenqvist et al., 1997), much shorter than that in Florida. This chapter, the first known study of the impact of dendroremediation incentives on management and profitability of a SRWC system, considers eucalyptus tree crops as a remediation strategy. An economic optimization model of a SRWC biomass production system that includes an incentive for dendroremediation of N in reclaimed water investigates how this incentive would influence land expectation value (LEV), an attribute of profitability, and optimal management of the associated SRWC production system. Methodology Optimization of Coppicing Species The basic Faustmann formula for a non-coppicing even-aged stand is defined by Eq. (1-1). The optimum rotation age t* is determined by taking the derivative of Eq. (1-1), setting it equal to 0, and solving for t (Chang, 1984). Eq. (1-5) defines the Faustmann formula as modified by Medema and Lyon (1985) for coppicing forest systems used to determine both the optimum duration of each stage as well as the optimum number of stages per cycle. Estimates for the prices and costs associated with SRWC production at Water Conserv II come from the preliminary work done by Rockwood et al. (2002a). An example of the generalized Eq. (1-5) fixed for two stages (n=2) is shown in Eq. (2-1): (V(t)*e (C,+C,))+ LEV= (V(t Ce(- )C (2-1) le(-r (t,+t2)) - where tl and t2 are the duration of stage one and stage two, respectively, Cp is the cost of planting at the beginning of the cycle, Cw is the cost of weeding at the beginning of the stage, C, is the annual maintenance cost, and Cr is the cost of irrigation establishment at the beginning of the operation. Optimization of Coppicing Species with Inclusion of the Dendroremediation Service Hartman (1976) internalized non-timber benefits (NTBs) into the Faustmann formula. The methodology described by Smart and Burgess (2000) is used to internalize the dendroremediation service associated with cultivating coppicing species such as Eucalyptus spp. irrigated with reclaimed water. There are two ways to account for NTBs. If a NTB is achieved at harvest, it is considered a stock benefit, while a continuous amenity is calculated as a flow benefit. A dendroremediation service might be payable following removal of N from the site with harvest of the biomass. In this scenario, the NTB would be treated as a stock benefit and accounted for much like a timber benefit as defined in Eq. (2-2). NTBs = NTB (t) (2-2) This NTB value can then be added to Eq. (1-5), the equation for net returns of coppicing species, as shown in Eq. (2-3). An example of Eq. (2-3) fixed for two stages is shown in Eq. (2-4). S IV((ts)e +NTBs e( -Cs eC) 1-e' *y ) (V(t)*e(l) + NTB*e r (CC +C + V(t)*e r*(t+t2)) NTB2s e(r*(tl+t2))- )) 1 -e(-r*(tl+t2)) (2-4) l-eC Cr Alternatively, if the dendroremediation service were deemed beneficial as the N is continuously accumulated in the growing trees, the NTB is treated as a flow benefit. The NTB of a given stage calculated as a flow can be defined as NTB d((NTB t) e(-rt)dt 0 (2-5) where the NTB of stage s is the definite integral of the flow of the benefits discounted to the beginning of the stage, for the duration of the stage. This NTB value can then be added to Eq. (1-5) as shown in Eq. (2-6). An example of Eq. (2-6) fixed for two stages is shown in Eq. (2-7). SZ IV(t e)*e + NTBF e r (t- 1) r Cs 1-e J- (V(t)* (-rl) NTBI -(C, +C,)) + V(ti2 -r*(t+t2)) NTBF *e( r) e( 1- e(- r(t+t2)) The NTB to be included in the determination of profit and optimum coppice management is the dendroremediation treatment of the reclaimed water. The Southern LEV TLEV LEV r*tl)) (2-6) (2-7) C C L.EV (2-3) ~ Regional Water Reclamation Facility records costs associated with wastewater treatment. These local sewage water treatment costs are used to estimate the shadow price of removing additional N from reclaimed water. The exact marginal cost of N removal is unknown. In an economic assessment of dendroremediation of municipal wastewater by willow in Sweden, Rosenqvist et al. (1997) determine that the costs for removing N and P of conventional treatment are $10- $27 kg-1 N (in 1994 USD). At the Southern Regional Water Reclamation Facility, pre- treatment waste contains 25 ppm ammonia N, and post-treatment reclaimed water contains 7 ppm nitrate N, resulting in a decrease of 18 milligrams liter- N; the total cost of sewage water treatment is $0.88 per 1,000 gallons (P. Duel, Orlando Wastewater Treatment Plant Manager, pers. comm., February 2004). Based on these values, the total cost of wastewater treatment is $12.92 kg-1 N, or $1.29 kg-1 N if 10% of the total cost is assumed associated with N removal. This cost estimate could be higher, since removal of additional N becomes increasingly costly, or it could be assumed that the value of removal of additional N should be lower, since the willingness to pay for the removal of additional N has not been substantiated. This analysis assumes a range of values from $0-$3.50 kg- N removed. Model inputs E. grandis (EG) was identified as producing more biomass and accumulating more N than Populus deltoides when irrigated with reclaimed water (Rockwood et al., 2002a). While P. deltoides is dormant during the winter months of central Florida (November through February), EG grows year-round, offering continuous dendroremediation services not possible with deciduous species (Rockwood et al., 2001). Though not a native species, EG is non-invasive in Florida and has been produced commercially in south central Florida since the 1970s without spreading (Rockwood, 1996). While other species could be considered in the future, the scope of this chapter is limited to EG, because to date it has demonstrated the greatest potential for dendroremediation of reclaimed water in central Florida. The methodology described here can be applied to other candidate species for which irrigated growth and yield data are available. Height and DBH data were taken between 0 and 26 months for EG in central Florida at a density of 3,500 trees ha-1 irrigated with 17 mm reclaimed water day-. Due to a lack of growth and yield data for irrigated EG in central Florida beyond 26 months, the above observations are extended using unpublished data for EG from Belle Glade, Florida to estimate high and low growth and yield functions (Carter and Rockwood, personal comm., 2004). The trees in Belle Glade were effectively irrigated because soil moisture was made adequate by controlling water in irrigation canals, and the growth rate of the two sites were similar through the first 26 months. This extended data set is used as a baseline for estimating a range of possible yield functions in the sensitivity analysis described below. Nonlinear regression is used to fit the data to an Arrhenius functional form: B(a) = b e (2-8) where B(a) is dry Mg stemwood and bark biomass ha-1 as a function of stand age a in years for the first stage, and b and c are the estimated parameters 118.9 and 2.73 for the low growth function and 154.0 and 2.92 for the high growth function, respectively (Figure 2-1). These functions, yielding 16 and 19 dry Mg ha-1 year1 stem biomass or 27 and 32 dry Mg ha-1 year1 total above ground biomass1 for the low and high growth functions assuming a rotation age of 3.6 years, are at the high range of estimated unirrigated Eucayptus spp. production of 20-31 dry Mg ha- year' described by Rahmani and Hodges et al. (1997). 150 4 100 0 I 0 5 10 15 stand age (years) High Growth Estimate Low Growth Estimate Figure 2-1. Estimated high and low growth and yield functions for Eucalyptus grandis at Winter Garden, Florida, irrigated with 17 mm day- reclaimed water. Yields of subsequent stages (i.e., yields of the coppice stages following the initial growth) are uncertain. In central Florida, the season in which the trees are harvested influences EG coppice productivity (Webley et al., 1986). Though coppice yields eventually decline, research of P. deltoides suggests that the yield of the second stage (i.e., first coppice regrowth) is generally higher than that of the initial growth stage (Hansen et al., 1983). While growth of the second stage of EG might be higher than that of the first stage due to the benefits derived from the previously established root system, coppice mortality associated with wind throw or weed competition might also increase, reducing yields. Due to a lack of data of coppice stages for EG irrigated with reclaimed 1 Assumes a factor of 1.7 to convert stem inside bark to total aboveground biomass (Mg ha-1) (Segrest, 2002). water, in this analysis yields of 80%, 65%, and 30% of the growth of the initial stage are assumed for the second, third, and fourth stages, respectively. These estimates are based on casual observation and are consistent with the methodology described by Medema and Lyon (1985). To incorporate a dendroremediation benefit, the amount ofN assimilated in biomass growth is estimated. Analysis ofbiomass samples of EG irrigated with reclaimed water at Water Conserv II indicates that leaves, stem bark, branches, and stem wood contain 1.39%, 0.28%, 0.27% and 0.09% nitrate N, respectively (Rockwood et al., 2001). Accounting for different rates of accumulation of these four components of tree biomass, N accumulation functions are shown in Eq. (2-8), where parameters b and c are 54.8 and -0.24 for the high growth estimate and 51.4 and -0.22 for the low growth estimate, respectively. While N accumulation rates of coppice stages might decrease with reduced biomass productivity or increase with higher leaf/stem ratios, actual rates of N accumulation by coppice stages are unknown. This model assumes the same N accumulation function for coppice stages as for the original growth stage. Based on previous work relating to SRWC production in Florida (Rockwood et al., 2002a; Segrest et al., 1998; Rahmani et al., 1997), the following model inputs are assumed: planting cost, $500.00 ha-1; weed control following a coppice harvest, $50.00 ha-1; annual maintenance fee, $50.00 ha-1; and the price of woody biomass (for mulch), $20.00 dry Mg-1. This model assumes simulated real interest rates of 4% and 6% and two different costs for irrigation (microemitter) installation, $2,471 ha-1 and $3,707 ha-l. The high price for the irrigation system needed to distribute the reclaimed water, which is correlated with the price of gasoline, is a cost not incurred by conventional forestry systems in Florida. Results and Discussion Table 2-1 illustrates net returns assuming no dendroremediation incentive, a dendroremediation incentive treated as a stock benefit, and a dendroremediation incentive treated as a flow benefit2, for high and low growth estimates, at a dendroremediation value of $1.50 kg-1 N, an interest rate of 4%, price of wood of $20 dry Mg-1, and an irrigation installation cost of $2,471 ha-1, for 1, 2, 3, and 4 coppice stages. By identifying the number of stages that yields the highest profit, the optimum number of stages is two and three for the high and low growth models, respectively. This process was repeated for each scenario of irrigation cost, growth and yield function, and interest rate combinations to determine optimum profit, optimum number of stages per cycle, and optimum stage durations (Table 2-2). Resulting LEVs range from -$2,343 to +$2,726 ha-1, less than LEVs of a SRWC system in the United Kingdom reported by Smart and Burgess (2000) of $3,931, $6,168 and $14,814 ha-1 for market only, low NTB and high NTB model scenarios, respectively (stumpage price of $31 dry Mg-1, establishment cost of $1,538 ha-1 and an exchange rate of $1.54/ in November 2000). If the cost of the irrigation system were assumed sunk, LEVs reported here would range from $1,364 to $5,233 ha-1, comparable to those of Smart and Burgess (2000). To compare these findings with Florida production costs calculated by a previous study, the model was used to find minimum stumpage prices needed to achieve LEVs of 2 If treated as a stock benefit, the dendroremediation service is provided when the nitrogen is taken from the site (removed with harvested biomass); if treated as a flow benefit, the service is continuously provided as the tree grows and accumulates nitrogen. $1,235 ha-1 and $2,470 ha-1, representing LEVs of conventional forestry (Borders & Bailey, 2001) and Florida agricultural land (Reynolds, 2005), respectively. Stumpage prices of $26 and $30 dry Mg-1 are required to match LEVs of $1,235 ha1 and $2,470 ha-1, respectively, assuming irrigation establishment costs of $3,707 ha-1, the high growth model and an interest rate of 5%. Rahmani et al. (1997) report Eucalyptus spp. farm gate production costs for Florida of $32-$39 dry Mg-1, less than the $48-$52 dry Mg-1 farm gate costs estimated here assuming a harvest cost of $22 dry Mg-1 (Rahmani et al., 1998). A higher cost of production is expected given the cost of irrigation establishment. Excluding irrigation costs C, from the model yields stumpage prices of $15 and $19 dry Mg-1 and farmgate prices of $37 and $42 dry Mg-1 needed to match LEVs of $1,235 ha-1 and $2,470 ha-1 respectively, closer to the estimates by Rahmani and Hodges et al. (1997). Table 2-1. Net returns and optimum stage lengths assuming 1, 2, 3, and 4 stages for a Eucalyptus grandis short-rotation woody crop system irrigated with reclaimed water in central Florida. High Growth: No N benefit N Benefit as Stock N Benefit as Flow # Stages LEV Optimum Stage LEV Optimum Stage LEV Optimum Stage per cycle ($ ha1)Lengths ($ ha-1) Lengths ($ ha-1) Lengths 1 $653 4.4 $1,068 4.2 $1,134 4.2 2* $1,4054.0, 3.6 $1,888 3.8, 3.6 $1,952 3.8,3.4 3 $1,3164.0, 3.6, 3.1 $1,824 3.8, 3.4, 2.9 $1,887 3.8, 3.4, 2.9 4 $1,1204.1,3.7, 3.3, 0.1 $1,610 3.9, 3.5, 3.1, 0.1 $1,674 3.9, 3.5, 3.1, 0.1 Low Growth: No N benefit N Benefit as Stock N Benefit as Flow # Stages LEV Optimum Stage LEV Optimum Stage LEV Optimum Stage per cycle ($ ha )Lengths ($ ha ) Lengths ($ ha ) Lengths 1 -$932 4.7 -$563 4.4 -$499 4.4 2 -$85 4.1, 3.7 $364 3.8, 3.4 $425 3.8, 3.4 3* -$63 4.0, 3.7, 3.2 $414 3.8, 3.4, 2.9 $475 3.8, 3.4, 2.9 4 -$258 4.2,3.8,3.4,0.1 $198 3.9, 3.5, 3.1, 0.1 $259 3.9, 3.5, 3.1, 0.1 Note: These calculations include an interest rate of 4%, price of wood of $20 dry Mg1, irrigation installation cost of $2,471 ha-l, and a dendroremediation value of $1.50 kg-1 N (where applicable). An "*" indicates the optimum number of stages per cycle as show by the highest net returns. The NTB is calculated as both a stock and flow benefit. Table 2-2. Optimum LEVs, optimum stages per cycle, and optimum stage lengths for a range of dendroremediation values for Eucalyptus grandis irrigated with reclaimed water in central Florida. Dendroremediation Benefit as a Stock Dendroremediation Benefit as a Flow LEV ($ ha') Optimum Stage Lengths (years) LEV ($ ha') Optimum Stage Lengths (years) Low Growth 4.0, 3.7, 3.2 4.0, 3.6, 3.1 3.9, 3.5, 3.0 3.8, 3.4, 2.9 3.7, 3.3, 2.8 3.6, 3.2, 2.8 3.5, 3.1, 2.7 3.4, 3.0, 2.6 4.0, 3.7, 3.2 4.0, 3.6, 3.1 3.9, 3.5, 3.0 3.8, 3.4, 2.9 3.7, 3.3, 2.8 3.6, 3.2, 2.8 3.5, 3.1, 2.7 3.4, 3.0, 2.6 High Low High Low Growth Growth Growth Growth $1,405 $1,584 $1,767 $1,952 $2,141 $2,333 $2,528 $2,726 $169 $349 $531 $717 $906 $1,098 $1,293 $1,491 -$63 $112 $291 $475 $662 $854 $1,050 $1,250 -$1,299 -$1,123 -$944 -$761 -$574 -$382 -$186 $15 4.0,3.6 3.9,3.5 3.9, 3.4 3.8, 3.4 3.7,3.3 3.7, 3.2 3.6, 3.2 3.5, 3.1 4.0, 3.6 3.9, 3.5 3.9, 3.4 3.8, 3.4 3.7, 3.3 3.7, 3.2 3.6, 3.2 3.5, 3.1 4.0, 3.7, 3.2 4.0, 3.6, 3.1 3.9, 3.5, 3.0 3.8, 3.4, 2.9 3.7, 3.3, 2.8 3.6, 3.2, 2.8 3.5, 3.1, 2.7 3.4, 3.1, 2.6 4.0, 3.7, 3.2 4.0, 3.6, 3.1 3.9, 3.5, 3.0 3.8, 3.4, 2.9 3.7, 3.3, 2.8 3.6, 3.2, 2.8 3.5, 3.1, 2.7 3.4, 3.1, 2.6 $ kg-' N A $0.00 $0.50 $1.00 $1.50 $2.00 $2.50 $3.00 $3.50 B $0.00 $0.50 $1.00 $1.50 $2.00 $2.50 $3.00 $3.50 High Growth $1,405 $1,563 $1,724 $1,888 $2,056 $2,227 $2,401 $2,579 $169 $327 $488 $653 $820 $991 $1,166 $1,343 Low Growth -$63 $92 $251 $414 $581 $753 $930 $1,111 -$1,299 -$1,144 -$985 -$822 -$654 -$482 -$306 -$124 High Growth 4.0, 3.6 3.9, 3.5 3.9, 3.4 3.8, 3.4 3.7, 3.3 3.7, 3.2 3.6, 3.1 3.5, 3.1 4.0, 3.6 3.9, 3.5 3.9, 3.4 3.8, 3.4 3.7, 3.3 3.7, 3.2 3.6, 3.2 3.5, 3.1 Table 2-2. Continued Dendroremediation Benefit as a Stock LEV ($ha-1) Optimum Stage Lengths (years) $ kg-' N C $0.00 $0.50 $1.00 $1.50 $2.00 $2.50 $3.00 $3.50 D $0.00 $0.50 $1.00 $1.50 $2.00 $2.50 $3.00 $3.50 High Growth -$195 -$91 $15 $124 $236 $355 $476 $601 -$1,430 -$1,326 -$1,220 -$1,113 -$1,000 -$881 -$759 -$635 Low Growth -$1,108 -$1,006 -$902 -$795 -$685 -$571 -$455 -$336 -$2,343 -$2,242 -$2,137 -$2,030 -$1,920 -$1,807 -$1,691 -$1,571 High Growth 3.9, 3.5 3.8, 3.5 3.7, 3.4 3.7,3.3 3.6, 3.2, 2.8 3.5, 3.2, 2.7 3.5, 3.1, 2.7 3.4, 3.0, 2.6 3.9, 3.5 3.8, 3.5 3.7, 3.4 3.7, 3.3 3.6, 3.2, 2.8 3.5, 3.2, 2.7 3.5, 3.1, 2.7 3.4, 3.0, 2.6 Dendroremediation Benefit as a Flow LEV ($ha1) Optimum Stage Lengths (years) Low Growth 3.9, 3.6, 3.2 3.8, 3.5, 3.1 3.7, 3.4, 3.0 3.6, 3.3, 2.9 3.5, 3.2, 2.8 3.5, 3.2, 2.8 3.4, 3.1, 2.7 3.3, 3.0, 2.6 3.9, 3.6, 3.2 3.8, 3.5, 3.1 3.7, 3.4, 3.0 3.6, 3.3, 2.9 3.5, 3.2, 2.8 3.5, 3.2, 2.8 3.4, 3.1, 2.7 3.3, 3.0, 2.6 High Growth -$195 -$69 $58 $187 $320 $459 $601 $746 -$1,430 -$1,305 -$1,112 -$1,048 -$916 -$777 -$635 -$490 Low Growth -$1,108 -$986 -$862 -$734 -$605 -$472 -$336 -$198 -$2,343 -$2,221 -$2,097 -$1,970 -$1,840 -$1,708 -$1,572 -$1,433 High Growth 3.9, 3.5 3.8,3.5 3.7, 3.4 3.7, 3.3 3.6, 3.3, 2.8 3.5, 3.2, 2.7 3.5, 3.1, 2.7 3.4, 3.1, 2.6 3.9, 3.5 3.8, 3.5 3.7, 3.3 3.7, 3.3 3.6, 3.3, 2.8 3.5, 3.2, 2.7 3.5, 3.1, 2.7 3.4, 3.1, 2.6 Low Growth 3.9, 3.6, 3.2 3.7, 3.5, 3.1 3.7, 3.4, 3.0 3.6, 3.3, 2.9 3.5, 3.2, 2.9 3.5, 3.2, 2.8 3.4, 3.1, 2.7 3.3, 3.0, 2.6 3.9, 3.6, 3.2 3.8, 3.5, 3.1 3.7, 3.4, 3.0 3.6, 3.3, 2.9 3.5, 3.2, 2.9 3.5, 3.2, 2.8 3.4, 3.1, 2.7 3.3, 3.0, 2.6 Note: LEVs are shown for both stock and a flow benefits under high and low growth models assuming A) an interest rate of 4% and an irrigation installation cost of $2,471 ha-1; B) an interest rate of 4% and an irrigation installation cost of $3,707 ha-1; C) an interest rate of 6% and an irrigation installation cost of $2,471 ha-1; and D) an interest rate of 6% and an irrigation installation cost of $3,707 ha-. The optimum economic rotation lengths for cycles with single stages shown in Table 2-1 are longer than maximum sustained yield (MSY) ages of 2.7 and 2.9 years for the low and high growth models, respectively. Economic optimum rotation ages of conventional forest plantations are typically shorter than the age of MSY (Samuelson, 1976). However, low stumpage prices relative to high regeneration costs can extend optimal economic rotation past the rotation of MSY, especially with SRWC species as demonstrated by Binkley (1987). Though inclusion of an environmental service provided by a standing forest extends optimal economic rotation (Hartman, 1976), incentives for dendroremediation, best achieved by rapidly growing stands, favors shorter rotations (Table 2-2). While increasing interest rates tends to shorten optimum coppice stage lengths as the opportunity cost of standing biomass increases, it also favors increasing the number of coppice stages per cycle, thus minimizing regeneration costs. At a 4% interest rate, the optimum number of stages per cycle is two and three for simulations using the high and low growth functions, respectively. At a 6% interest rate, the optimum number of stages per cycle is two for high growth rates with a dendroremediation incentive less than $2 kg-1 N and three for the remaining scenarios. Optimum stage length duration ranges from 2.6-4.0 years. Increasing the interest rate from 4% to 6% or increasing the dendroremediation incentive by $1 kg-1 N decreases optimum stage lengths by about 0.1 years (Table 2-2). These results are consistent with those of Smart and Burgess (2000) who observe that decreasing yields or increasing the discount rate decreases LEV and thus the opportunity cost of the land, extending the coppice cycle to delay regeneration costs, while having a negligible effect on optimum stage lengths. Sensitivity Analysis of Dendroremediation Incentive and Interest Rate This model was used to assess the sensitivity of profitability to changes in the dendroremediation incentive and the interest rate. Under all scenarios, the dendroremediation incentive had a positive nearly-linear relationship with profitability. Average marginal increases in profitability per dollar ofN dendroremediation incentive according to growth function (high or low), benefit (stock or flow) and interest rate (4% or 6%) are shown in Table 2-3. Assuming an interest rate of 4%, a $1 kg-1 N increment in the dendroremediation incentive caused a marginal increase in profitability of about $376 and $335 assuming flow and stock dendroremediation benefits, respectively, with little or no influence from the cost of irrigation or the growth model. Assuming an interest rate of 6%, a $1 kg-1 N increase in dendroremediation incentive caused a marginal increase in profitability of about $264 and $223 assuming flow and stock dendroremediation benefits, respectively. Profit sensitivity to dendroremediation incentive is shown in Figure 2-2. Table 2-3. Average marginal increases in net returns ($ ha-1) per dollar of N dendroremediation incentive according to growth function (high or low), benefit (stock or flow) and interest rate (4% or 6%) for Eucalyptus grandis in central Florida. Marginal benefits are insensitive to changes in irrigation cost. Scenario: 4% 6% High growth, flow benefit $377 $268 High growth, stock benefit $336 $227 Low growth, flow benefit $375 $261 Low growth, stock benefit $334 $220 $3,000 $2,500 $2,000 $1,500 S $1,000 " > $500 $0 -$500 -$1,000 -$1,500 $000 $050 $1 00 $1 50 $200 $250 $300 $350 Dendroremediation Incentive ($/kg N) A -I=4%, G=H, B=F 1 =4%, G=H, B=S S- -I=4%, G=L, B=F 1=4%, G=L, B=S -I=6%, G=H, B=F =6%, G=H, B=S -- ) I=6%, G=L, B=F 1=6%, G=L, B=S Figure 2-2. Net returns ($ ha-1) as a function of dendroremediation incentive ($ kg-1 N) (I=interest rate, H=high growth function, L=low growth function, F=flow benefit model, S=stock benefit model), assuming an irrigation cost of $2,471 per hectare. Changes in profitability ($ ha-) as interest rate increases from 4% to 5% and from 5% to 6% for the high and low growth functions at dendroremediation incentives of $0, $2, and $4 kg-1 N are shown in Table 2-4. The changes are the same assuming either flow or stock benefit models. Profitability is highly sensitive to changes in the interest rate, especially assuming the high growth function shown in Figure 2-1. Assuming a dendroremediation incentive of $2 kg- N, an increase in the interest rate from 4% to 5% causes a marginal decrease in net returns by $1,028 and $717 for the high and low growth assumptions, respectively, while an increase in interest rate from 5% to 6% causes a marginal decrease in profit by $791 and $550 for the high and low growth assumptions, respectively. Sensitivity to interest rate is not influenced by the price of irrigation or the type of benefit (stock or flow) assumed. Table 2-4. Changes in profit ($ ha-) for Eucalyptus grandis in central Florida as interest rate increases from 4% to 5% and 5% to 6% for high and low growth functions shown in Figure 2-1, at dendroremediation incentives of $0, $2, and $4 kg-1 N. Changes in profit are the same assuming either flow or stock benefit dendroremediation functions. Interest rate Interest rate Dendroremediation increase from 4% increase from 5% Growth Function Incentive ($ kg-' N) to 5% to 6% $0.00 -$907 -$693 High growth function $2.00 -$1,028 -$791 $4.00 -$1,149 -$888 $0.00 -$593 -$446 Low growth function $2.00 -$717 -$550 $4.00 -$842 -$654 Using Data Fit 8.0, model outputs under the range of assumptions were condensed into LEV prediction Eq. (2-9), where I is the real interest rate, g is -1 < g < 1, where -1 and 1 represent the low and high growth, respectively, as represented in Figure 2-1, Yis the price of irrigation establishment ($ ha-1), v is dummy variable 0 or 1 for calculation as stock and flow benefits, respectively, and Nis the value of the dendroremediation benefit ($ kg-1) (R2>0.99). This prediction equation could be used to predict LEV in the absence of modeling software. Estimated parameters and statistical descriptors are shown in Table 2-5. LEV(I, g,Y,v,N)= (/ *eA + g*/* 2 *e Y) + (2-9) (/4 e -,* +6 v + g *.7 e#) *.N+E Table 2-5. Estimated parameters and descriptors used in Eq. (2-9) of Eucalyptus grandis irrigated with reclaimed water in central Florida (R2>0.99). Constants Value Standard t-ratio Error Po 9339.94 31.82 293.50 /P 27.34 0.07 367.88 f2 1903.38 20.34 93.59 f3 23.76 0.26 93.15 f4 759.74 10.73 70.83 Ps 20.45 0.30 67.73 P6 41.18 0.86 47.73 Pf7 0.02 0.00 0.00 /8s 84.32 5.85 14.41 Conclusions In Florida, environmental and/or treatment costs associated with municipal wastewater disposal will intensify as urban populations grow. Using SRWC plantations to dendroremediate wastewater can provide various environmental services and societal benefits in the WUI and should be considered as wastewater remediation option. Our results suggest that financial compensation for dendroremediation services would be required to make the system economically feasible for private landowners. Calculations of net returns for 128 SRWC dendroremediation scenarios ranged from -$2,343 to +$2,762 ha-1 and are greatly reduced by high interest rates, high irrigation costs, and low growth functions. Each $1 kg-1 N increase in the dendroremediation incentive increases profit by $223-$376 ha-1, depending on interest rate and site productivity. $1 kg-1 N is probably less than the price to achieve the same service at a wastewater treatment plant. A 1% increase in interest rate can reduce profit by $446-$1,149 depending on the scenario. Increasing the interest rate from 4% to 6% or increasing the dendroremediation incentive by $1 kg-1 N decreases optimum stage lengths by about 0.1 year, which may not be operationally significant. However, the decision of whether to select two or three stages per cycle is influenced by the growth and yield function, which could be increased through improvements in weed control. Ceterisparibus, higher growth of the first stage decreases stage length and number of stages per cycle, though improving growth of the second and third stages, which may be possible through weed and vine control, would favor longer stage duration, more stages per cycle, and could increase profitability. High costs of irrigation establishment greatly reduce net returns. Microemitter irrigation establishment cost about $2,471 ha-1 in 2000 and has gone up to about $3,707 ha-1 in 2004 due to increasing fuel prices. However, costly microemitter irrigation systems are designed to conserve water, which might not be necessary in the case of distributing reclaimed water, possibly providing an opportunity to apply less expensive irrigation systems. Additionally, citrus growers who want to take advantage of previously existing irrigation systems would not incur the cost of irrigation, effectively increasing profitability by $2,471-$3,707 ha-1. Compensation to landowners for the dendroremediation service could be considered payable either when the trees and N are harvested (stock benefit), or periodically as the trees grow and accumulate N (flow benefit). Because of the short optimum cycle lengths, differences between the stock and flow benefit net returns are smaller than they would be in conventional forest rotation ages. Accounting for the dendroremediation service as a flow benefit rather than a stock benefit increases profits by $61-$64 ha-1 and $138-$148 ha-1 assuming dendroremediation incentives of $1.50 and $3.50 kg1- N, respectively. Accounting for the dendroremediation benefit as a stock would probably be easier to administer. Municipalities that use SRWCs to dendroremediate wastewater could also use this model to account for the value of the dendroremediation service they might achieve. Future Research Dendroremediation will be feasible for municipal waste facilities in the WUI if either tree farmers are paid enough for their dendroremediation service to make the system economically viable, or if the net cost for municipalities to dendroremediate with tree crops is cheaper than the alternative cost of treatment. Because case-specific costs, yields, and prices may vary from the assumptions made in this chapter, broad conclusions about the feasibility of dendroremediation of reclaimed water cannot be derived from these results. With more information, particularly with regards to coppice growth and valuation of the dendroremediation service, the model described here can be used to make localized feasibility assessments. A study to assess the willingness to pay for reductions of N contamination from reclaimed water would elucidate the value of the dendroremediation service. An assessment of the willingness to accept exotic species in central Florida should be considered in a feasibility analysis of dendroremediation using EG. This model should be extended to include the most immediate environmental benefits, such as dendroremediation ofP in the reclaimed water, C sequestration, and, under scenarios where the biomass is used for bioenergy, mitigation of atmospheric CO2 due to displacement of fossil fuels. The improved biomass production attributable to the N, P, and K in the reclaimed water, and the savings associated with reduced fertilizer use should also be internalized in this model. CHAPTER 3 AN ECONOMIC ANALYSIS OF Eucalyptus SPP. AS SHORT-ROTATION WOODY CROPS ON CLAY SETTLING AREAS IN POLK COUNTY, FLORIDA Introduction Central Florida produces 75% of the nation's and 25% of the world's phosphate supply, primarily used for fertilizer (IMC Phosphates, 2002). There are about 162,000 hectares (400,000 acres) of phosphate-mined lands in Florida (Segrest, 2003). In the mining process, clays are washed from phosphate ore and pumped into clay settling areas (CSAs). Polk County, Florida contains over 38,000 hectares (94,000 acres) of CSAs and/or overburden soil, as a result of phosphate mining. These CSAs, classified as clayey Haplaquents (Mislevy et al., 1989), are characterized by high bulk density, poor drainage, high levels of P, K, and micronutrients, pH of 7.0-8.3, and are commonly dominated by cogongrass (Imperata cylindrica), an invasive exotic species in Florida. CSAs are largely left idle because of operational difficulties but may be a valuable resource base for biomass production. Ongoing research and operational trials on a 50-hectare CSA near Lakeland, FL, suggests that CSAs can be used for the production of SRWCs. This chapter assesses the economic viability of this practice. One environmental service that would be provided by the production of SRWCs on CSAs is atmospheric CO2 mitigation. Global carbon trading has increased from 13 Tg CO2 in 2001 to 70 MMg CO2 in 2003 (Ecosystem Marketplace, 2004), a trend that is likely to continue following the international ratification of the Kyoto Protocol on February 16, 2005. The establishment of tree plantations on non-forested CSAs has the potential to sequester carbon by increasing the amount of C per area of land (Booth, 2003). Chaturvedi (2004) emphasizes the importance of considering the C density of land prior to the carbon sequestering land use practice. He states "The most clear benefit in carbon sequestration terms would be if the plantation were somehow established in a desert with an existing standing stock of virtually zero Mg C/ha." An advantage of SRWC production on CSAs is the near-zero C density of the land prior to plantation establishment, as the land is bare of vegetation with little accumulation of soil organic carbon (SOC) following mining. Even on 20-40 year-old CSAs, C density is likely to remain low if forest cover is not established. Research suggests that SRWCs sequester and maintain SOC (Joslin & Schoenholtz, 1997). On a 60-year-old CSA in central Florida SOC of a 2.5-year-old E. grandis (EG) plantation was 214% and 304% greater at depths of 0-30 and 30-60 cm, respectively, than SOC quantities found in adjacent areas dominated by cogongrass (Wullschleger et al., 2004). In addition to C sequestration in situ, if used as a dedicated feedstock supply system (DFSS), SRWC plantations can mitigate atmospheric CO2 by displacement of CO2 emissions associated with the combustion of fossil fuels (Sims, 2002; Marland, 2000; Schlamadinger & Marland, 1996). The displacement of fossil fuels by biomass fuels can be an effective way to mitigate atmospheric CO2 because 1) CO2 emissions can be continuously reduced, rather than reaching an eventual plateau of C accumulation in standing biomass, 2) the long-term cost per Mg of CO2 is cheaper with displacement rather than sequestration, as land remains available for continued production in the future, and 3) reductions of net CO2 emissions are not as risk prone as C sequestered in situ, which is susceptible to future events such as fire or land-use change (Eriksson et al., 2002). This chapter investigates the impact of CO2 mitigation incentives on management and profitability of SRWC DFSSs on CSAs in Polk County, Florida. An economic optimization model of a SRWC biomass production system that includes an incentive for atmospheric CO2 mitigation is built and used to investigate how this incentive would influence land expectation value (LEV) and optimal management of the SRWC production system. Methodology As described in Chapter 1, Eq. (1-1) defines LEV, net returns of a non-coppicing forestry system projected in perpetuity. This equation is modified in Eq. (1-5) to allow for coppicing forestry systems, which includes n number of growth stages (initial growth stage and subsequent coppice stages). Eq. (1-5) is used to calculate LEVs under a range of model input assumptions, exclusive of environmental externalities. To assess the divergence between private and societal benefits derived from the system, LEVs are then compared to those calculated by Eq. (1-13), which incorporates a non-timber benefit (NTB) for each growth stage s. Quantification and incorporation of the NTB requires a functional form which reflects the nature of the benefit provided by the forestry system. In this scenario, the externality to be incorporated is atmospheric CO2 mitigation. Trees sequester atmospheric CO2 in woody biomass as they grow. The value of standing aboveground C at time t for coppice stage s, assuming stage growth function g(t), carbon content of 47% by weight (Peter et al., 1996), and multiplying by 1.7 to convert stem inside bark to total aboveground biomass (Mg ha-) (based on Segrest, 2002; Patzek & Pimentel, 2005) a can be estimated as Cs (t)=g(t)*C *0.8 (3-1) where g(t) is the growth function for growth stage s as a function of time and Cp is the price of carbon. Once carbon is sequestered there is no further benefit from it, so the derivative of Eq. (3-1) is used to calculate the marginal benefit of the C sequestration service, yielding: CBsA f((CAt))*) )dt (3-2) where the aboveground C sequestration benefit of stage s is the definite integral of the flow of the carbon benefit discounted to the beginning of the stage, for the duration of the stage. Central to the concept of carbon sequestration is the life span of the sequestered carbon, either in the ecosystem, or in products derived from harvests from the ecosystem (Murray, 2003). As wood products burn or decay, sequestered carbon is re-emitted to the atmosphere in the form of C02, countering the benefit achieved by the sequestered C. This societal cost of the decay or oxidation of the sequestered carbon must be calculated and subtracted from Eq. (3-2). The rate of re-emission depends on the end use of the wood products. The two most likely products identified by a SRWC market survey in Polk County (below) are mulch and biofuel. The decay of C sequestered in these two products is accounted for differently. Eq. (3-3) represents the societal cost of CO2 emissions from the decay of mulch harvested from stage s at age t, where y is the life of the biomass in years assuming linear decay, discounted first to the end of the growth stage at discount rate r. For example, for y=5, 1/5th of the harvested mulch would decay during each of five years. Subtracting the right hand side of Eq. (3-3) from the right hand side of Eq. (3-2) assuming that mulch decays in five years (Duryea et al., 1999; Duryea, 1999) yields Eq. (3-4), the integration of the marginal value of above-ground C sequestration discounted to the beginning of the growth stage, minus the societal cost of CO2 emissions associated with mulch decay discounted first to the end of the growth stage and then discounted to the beginning of the growth stage. Though actual mulch decay may be non-linear and may take longer than five years, the decay function in Eq. (3-3) was chosen to simplify the analysis and provide a conservative estimate of the net C sequestration benefit. CS t)= C-r*) (- ,r*t) (3-3) y r NTB [ C (t())*e(t)t [1-es) *e(-t) (3-4) This NTB calculated in Eq. (3-4) is then included in the optimization model for each growth stage of the mulch scenario and discounted to the beginning of the coppice cycle. Eq. (3-4) is incorporated in Eq. (1-5), the equation for net returns of a coppice cycle having n number of growth stages, as shown in Eq. (3-5), where V(t) is the growth function for stage s times biomass price. To elucidate the discounting of each benefit and cost in the model, an example of Eq. (3-5) fixed for two stages is shown in Eq. (3-6), including annual maintenance cost Ca and a one-time establishment cost C,. V(()* e +, + C (t)) *e(-rt) dt *e(r JltJ1 V ^t)*P ^ ^Ye^[+ dt I, s-I CSA (-r*5) -r*y t J (-r* ltJ-1) LEVimch (t)= *e( CJi (- 3-5) 1 -r*, J 1-e 01 1 A A (I (-r*5) e (3-6) E,- (t) l- *(xy)- A e, e- e (-r*5)e r- s in the sseqent rtt 5 r LEI V,, C, (3-6) resulting in no net emissions from biomass combustion, and displacing the use of fossil fuels with closed-loop biofuel reduces net CO2 emissions. Thus, bioenergy from DFSSs produces no net CO2 emissions, eliminating the need to calculate the costs of post-harvest biomass C decay. However, recognizing that there are fossil fuel inputs to the cultivation, harvest, and processing of SRWC DFSSs consuming up to 10% of the energy produced by the bioenergy system (Forsberg, 2000; Heller et al., 2004; Klass, 1998), 10% of the carbon sequestration benefit achieved at stage age t is discounted to the beginning of the stage and subtracted from the carbon benefit calculated by Eq. (3-2), yielding Eq. (3-7): NTBF (b (t))*e(t) dt -[(0.1*C ())]*e (3-7) 0 The net NTB calculated for each growth stage for the biofuel scenario in Eq. (3-7) is then added to Eq. (1-5), resulting in Eq. (3-8). Eq. (3-9) is an example of Eq. (3-8) fixed for two growth stages. V(t)*e jre J d(Cbs (t))*e(-r *t))dt *e( j t-1) s=1 0.1* Cb, (t))*e( C e(r J 1 e J_ r j )lJ LEVboftel/ (t) (3-8) V(t)*e, rt ))+ !( (cb (*t))*e dt rt) + o [(0.1*Cbs (t)] *e( -1 t 2 -1 LEV (t)= -e-*(tl)) l-e C, (3-9) Thus, Equations (3-5) and (3-8) are used for incorporating C externalities in mulch and biofuel production scenarios, respectively. These models, with Eq. (1-5) for optimization without incorporation of externalities, are used to calculate LEV and optimum age of each of n number of growth stages. The process is repeated iteratively adding an additional growth stage for each scenario until the marginal benefit of the additional stage is negative, identifying the optimum number of growth stages per coppice cycle and associated LEVs. Finally, the sensitivity of these LEVs to variation in the below model inputs is assessed. Model Inputs Growth Function Lacking published growth and yield functions of SRWCs produced on CSAs needed for inclusion in this model, measurements were taken from a trial of EG and E. amplifolia (EA) on a CSA near Lakeland, Florida (Rockwood et al., 2005). Established between May and July of 2001, SRWC-90 was planted at densities of 4,200 (single row) and 8,400 (double row) trees per hectare, unfertilized and fertilized on May 20, 2002 with 150 kg ammonium nitrate ha-1. Height and DBH measurements were taken August 20, 2002, July 16, 2003, December 23, 2003, August 27, 2004, and January 11, 2005. Number of surviving trees, average height, and average DBH per plot by progeny after 3.5 years on January 11, 2005, are shown in Table 3-1. A modified volume prediction equation developed by Max and Burkhart (Bredenkamp, 2000) was a good predictor of volumes of 66 destructively sampled trees (R2>0.99) and used to convert height and DBH measurements to per-hectare yields assuming specific gravity of 0.40 (Rockwood et al., 1995a). Per-hectare inside-bark aboveground yields (dry Mg ha-1) of EG and EA under five treatments are shown in Table 3-1 and Figure 3-1. Decreasing rates of productivity were observed on January 11, 2005 at 3.5 years of age, suggesting an optimizable function could be fit to the data for use in the model. 70 EA 1 60 EA4 EA2 EG 4 S50 EA3 S4 EG5 EA4 EA 5 240 -0 EG 3 EG 12 10 -E/ 0 E A 2 EG 3 +EG 4 0 1 2 3 4EG Age (years) -- EG 5 Figure 3-1. Inside bark yields (dry Mg ha-1) of EA and EG on a CSA near Lakeland, Florida for 5 treatments: 1) 4,200 trees per hectare, unfertilized, 2) 8,400 trees per hectare, unfertilized, 3) 4,200 trees per hectare, fertilized with 150 kg ha-1 ammonium nitrate on May 20 2002 at 11 months, 4) 8,400 trees per hectare, fertilized as treatment 3, and 5) same as treatment 2. EA was identified as a likely candidate species due to a) greater frost resistance than EG, which allows flexibility to plant in late summer during increased rainfall with minimum frost damage to small trees the subsequent winter and b) higher yields than EG despite being planted two months later. An air photo from 1995 revealed that Treatments 1 and 2 had been established on areas of the CSA where cogongrass was more densly established than the other treatments, probably explaining their lower yields. Treatments 3 and 4 were identified as being representative of moderately low and moderately high yields when compared to SRWC yields from other areas of the CSA. Nonlinear regression was used to fit the yield data to the functional form: B(a) = e b+c*n(a)-d(a) (3-10) Table 3-1. Number of observations, average DBH (cm), height (m) and inside-bark dry above-ground biomass yields (Mg ha-') by progeny of EG and EA planted at densities of 4,200 (single) and 8,400 (double) trees per hectare, unfertilized (0) and fertilized (1) on May 20, 2002, with 150 kg ammonium nitrate ha-1, and measured January 11, 2005, at 3.5 years. "a" and "b" indicate lowest and highest yielding progenies within treatments, respectively. Single 0 Double 0 Single 1 Double 1 Double 0 (2) Progeny N DBH H Ms ha- N DBH H Ms ha' N DBH H Ms ha' N DBH H Ms ha' N DBH H Ms ha- 11 5.3a 7.5a 12 6.6 8.5 9 7.7b 8.9 10 7.6 9.4b 9 7 8.9 12 4.7 5.5 11 4.2 5.2 10 3.7a 4.6a 12 5.9 6.5 11 5.7 6.5 11 7.3b 7.8b 19.3a 27.2 30.4b 27 20.2 11 5.6 5.4a 12.1 10.4 22.3b 18 4 7 20 5.5b 8.6 21 3.9 6.9 20 5.4 8.6b 20 3.7 6.8 18 3.8 5.6 22 2.6a 4.0a 23 3.8 5.6 23 3.8 5.6 23 4.1b 5.9b 21 3.6 5.3 10.8a 23.8 14.6 27.2 b 12.5 9.7 3.2 11.2b 10.2 11.1 8.2 10 4.5a 6.9a 9 8.3 9.7 10 6.5 9.2 9 8.5 10.3b 11 6.2 8.4 EA 10 6.5 a 7.5a 12 7.7 9 9 7.4 8.7 10 8.7 10.0b 12 8.3 9.5 12 8.2 9.5 13.7 15 6.0a 9.1 a 36.1 15 9.4 12.6 25.5 14 9.4b 12.6b 36.5b 12 7.9 11.9 19.5 15 8 11.3 16.0a 20 6.1a 8.3a 31.8 19 7.6 10.4 23.2 22 7.7 10.4 32.7 23 6.9 9.7 35.3 24 8.8b 11.2b 38.8b 22 8.1 10.7 29.4a 12 5.0o 8.2a 19.3a 77.3 21 7.0b 10 65.9b 78.2b 17 5.4 8.9 28.9 38.4 19 6.9 10.2b 57.8 55.9 16 6.5 9.7 54.5 46.1a 20 5.7a 7.9a 31.6a 54.8 20 5.9 8.4 32.7 57.2 19 7.5 10 55.5 46.6 24 7.4 9.9 68.2 94.2b 22 8.1b 10.6b 68.9b 73.7 23 7.1 9.5 58.6 3242 3469 4064 4200 4223 4904 4907 5025 5033 5091 5108 51 where B(a) is dry stemwood biomass (Mg ha-') as a function of stand age a in years for the first stage, and b, c and dare the estimated parameters 2.57, 4.00 and 1.20 for EA 3 and 2.76, 3.67 and 0.92 for EA 4, respectively (Figure 3-2). Using a factor of 1.7 for total above-ground biomass, maximum sustained yields are 17 and 32 dry Mg ha-1 year-1, comparable to 20-31 dry Mg ha-1 year estimated for Eucalyptus in Florida (Rahmani et al., 1997) but higher than the estimated 9-17 dry Mg ha-1 yr- estimated by Klass (1998), who observes that yields could be improved with SRWC development in the sub-tropical south. -EA3 100 Average E -J i'5-DI1 S80 EA 4 EA 4 Average c- Average i 60 E-1 -4 4J0 x Predicted 40- m-*EE Cio EA3 > ..- EA 3 Average 20- .0 ..-- -- Predicted 0 ..EA4 0 1 2 3 4 5 Age (years) Figure 3-2. Observed and predicted inside bark stem yields of EA treatments 3 and 4, 4,200 and 8,400 trees per hectare, fertilized, and low and high progeny yields for each treatment. Carbon Values The Kyoto Protocol was ratified by 140 nations on February 16, 2005, strengthening ongoing efforts to reduce greenhouse gas emissions. While estimates for world carbon prices range from $4 to $27 Mg-1 C, $10 Mg-1 C is identified as a likely value (Vogt et al., 2005; Best & Wayburn, 2001). C prices assumed in this model range from $0 to $35 Mg1 C. Market Assessment To identify products and prices to be used in this analysis, a SRWC market assessment was made in July 2004 in and around Polk County, Florida. On-site and phone interviews were done with individuals from the Florida Division of Forestry, mulch industries, nurseries, electricity generation facilities, and potential biomass producers. Though not all companies interviewed were willing to share market information, a range of price values and demand quantity were derived from the interviews. Potential products from woody biomass grown on CSAs in Polk County include mulch, energy, timber, pallets, and fiberboard. The most likely products are 1) mulch, having an existing multi-million of dollar market in Polk County annually, and 2) feedstock for electricity generation, a prospective market with much potential for expansion. Following is a summary of these two most relevant woody biomass markets. The established market: mulch Mulch production is a major industry in central Florida, involving companies such as Seaboard Supply in Ft. Green, Greenleaf Products, Inc. in Haines City, Florida Fence Post Co. in Ona, Forest Resources Inc. in Tampa and Aaction Mulch in Fort Myers. These companies produce mulch from various sources, including sawmill waste of cypress and pine, sand pine harvests and forest thinnings (Garry Zipper, pers. com., July 15, 2004), eucalyptus plantations in south central Florida, and melaleuca eradication harvests in southern Florida. Mulch consumers look for a product that will resist decay, has a desirable appearance, and is reasonably priced. Cottonwood, lacking in decay resistance, is undesirable as a mulch product. EG is a desirable material due to its red heartwood, attractive scent, and resistance to rot and termites (Mike Milliken, pers. com., August 2nd, 2004). Some mulch users express concern about over-harvesting of cypress and want an alternative to cypress mulch products (Bobby Robins, pers. com., July 15, 2004). While demand for cypress mulch could serve as an incentive for sustainable cypress management and the establishment of cypress plantations, eucalyptus mulch marketed as "cypress-free" is likely to appeal to consumers who are concerned about loss of cypress trees. Mulchwood price Mike Milliken (pers. com., August 3, 2004) of Greenleaf Products suggested $14 green ton-' stumpage price (up from $10 green ton1 in 2002), assuming availability of minimum supply to produce 144,000 bags, requiring 2,600 green tons (Appendix Eq. 1). Dwight Knight of Seaboard Supply stated he would be willing to pay $33 green Mg-1 ($30 green ton-1), delivered (the mill is 48 km [30 miles] south of Lakeland), unprocessed (pers. com., August 5, 2004). As of August 2004, transportation costs are $1.30 loaded km ($2.10 mile-1), with each load carrying 21 Mg (25 tons) (Eric Hoyer, pers. com., July 12, 2004), which equals $0.06 green Mg-1 km1 ($0.08 green ton'1 mile-'). Assuming transportation of 48 km (30 miles) at $0.06 green Mg-1 km-1 ($0.08 green ton1 mile-1), a harvesting cost of $17.64 Mg-1 ($16.00 ton-'), and a delivered price of $33 green Mg-1 ($30 ton-'), this scenario suggests an equivalent stumpage value of about $12 green Mg-1 ($12 green ton-) ($33-$17.64-(48*$0.06)=-$12.48 Mg-1), depending on the transportation distance. Hypothetical high and low transportation cost scenarios and associated stumpage values are shown in Table 3-3. Table 3-2. Mulch markets for Eucalyptus produced in Polk County. Stumpage Volume price (green (green Mg Ha [acreage] Mg-1 [ton-1 ]: [tons]): Note: Location: neededT Greenleaf (a) $9 [$8] 435 [500 ] Minimum purchase, to be Haines n/a per mixed with other City, FL purchase products. Greenleaf(b) $15 [$14] 2,357 Minimum amount needed Haines 3,500 [2,600] for a run of bags. City, FL [8,600] Greenleaf (c) $15-18 [$14-$16] >9,072 [10,000] $13 [$12] up to 22,680 [25,000] year' Approximate amount per week, equivalent to about 122,469 green Mg (135,000 tons) year'. Minimum amount needed for Lowes or Home Depot to list a new line item, and to set up an on- site operation. Could purchase up to 235,900 green Mg [260,000 tons] year' Based on $33 green Mg-' ($30 green ton ') delivered price assuming a shipping cost of $0.08 ton-1 mile-', 30 miles shipping, and harvesting cost of $16 ton1. Haines City, FL Ft Green, FL 7,100 [17,500] 650 [1,600] Seaboard Supply tApproximate acreage needed for sustained production over a year, assuming growth of 34 green Mg ha-1 (15 tons acre ') year' (i.e., annual demand divided by annual production). Table 3-3. Estimated equivalent stumpage values for high and low transportation cost scenarios. All tons are green weight. High Cost Scenario Low Cost Scenario 64 km @ $0.06 Mg- km1 = 32 km @ $0.06 Mg1 km = -$3.84 Mg'- -$1.92 Mg-1 Transportation (40 miles @ $0.08 ton mile' (20 miles $0.08 ton mile -$3.36 ton-) -$1.68 ton') Harvest Cost (conventional -$18 Mg-1 (-$16 ton ') -$9 Mg-1 (-$8 ton') equipment) Price (delivered) +$28 Mg-1 (+$25 ton-) +$33 Mg-1 (+$30 ton') Equivalent Stumpage +$6.16 Mg-~ ($5.64 ton-) +$22.40Mg-' ($20.32 ton-) Value Mulchwood quantity Knight may purchase up to 22,700 green Mg (25,000 tons) yearf. Assuming yields of 34 green Mg ha-1 (15 tons acre-) year-', 647 ha (1,600 acres) of CSAs might be cultivated to meet the demand for this particular mulch mill. While significant, acreage needed to supply this particular plant would occupy a relatively small portion of the estimated (8,094 ha) 20,000 acres of CSAs in Polk County. Milliken (pers. com., August 2, 2004) affirmed that Greenleaf Products is capable of purchasing 50 semi loads per day (about 24 green Mg [26 tons] of eucalyptus load-1) for 200 days per year totaling about 240,000 green Mg (260,000 tons) per year. Producing this amount, assuming 34 green Mg ha-1 (15 tons acre-) yearf, could occupy about 7,100 of 8,100 ha (17,500 of 20,000 acres) of CSAs in Polk County. Mulch is currently produced in part from byproducts from sawmills and small- diameter trees from forest thinnings (Linda Kiella, Garry Zipper, pers. com., July 15, 2004). Because of the demand for mulch, sawmills convert waste into a product, and forest managers in some cases can reduce costs associated with forest management, forest fuel load control, and eradication of melaleuca (Melaleuca quinquenervia). It is uncertain how much of the current biomass market (wastewood, thinnings, melaleuca control, etc.) might be displaced if additional biomass is grown on CSAs. However, according to Milliken (pers. com., August 2nd, 2004), the market is constrained by supply of desirable material, not demand. Potential market: biomass fuels The biomass market for energy generation, while speculative, is potentially very large. Power generation plants that are using or could use bioenergy include Ridge Generating Station in Auburndale; Lakeland Electric in Lakeland; Big Bend Power Plant near Apollo Beach; and Tampa Electric Polk Power station near Mulberry (Figure 3-3). - -U 50X 4 Figure 3-3. Location and potential consumption of buyers of woody biomass from Polk County. Ridge Energy currently charges a tipping fee to receive biomass, ranging from $9 green Mg-1 ($8 ton-) for low-ash biomass that is pre-chipped up to $38 green Mg-1 ($35 ton-1) for high-ash unprocessed biomass. Ridge Energy might be able to accept 635 Mg (700 tons) day-1 of DFSSs for free (i.e., no tipping fee) if it contains less than 6% ash (no roots and a minimum of soiling) and if it is processed (i.e., chipped to smaller than 3") (Phil Tuohy, pers. com., July 27, 2004). If the biomass is processed and delivered for free, additional economic incentives would need to be applied to make biomass production economically viable. Lakeland Electric of Lakeland produces 2.8 million MW hours of electricity. As Lakeland Electric has had to raise rates during 2004, further rate increases associated with using renewable energy would be difficult to impose. However, bioenergy will be an attractive option if the Florida DEP mandates renewable energy production. If Lakeland Electric were to have a renewable portfolio standard (RPS) mandate for 4% renewables, they would need to generate 12.5 MW of renewable energy, the equivalent of 8-14 Mg (9-15 tons) of biomass hour- (20% moisture content), meaning about 54,000- 88,900 air-dry (20% MC on green weight basis) Mg (59,000-98,000 tons) or 85,000- 143,000 green Mg (94,000-158,000 tons) year-'. Matt McArdle, a biofuels industry specialist contracted by Lakeland Electric, calculates a potential biofuel demand of 63,500 green Mg (70,000 tons) year- and possibly using up to 127,000 green Mg (140,000 tons) year-', and suggests a likely price of $11 green Mg-1 ($10 ton-) delivered (pers. com., August 27, 2004). The Big Bend Power Plant near Apollo Beach is another possible biomass buyer if a RPS is mandated and could consume up to 45,000 green Mg (50,000 tons) year-'. The Tampa Electric Polk Power station near Mulberry, while a possible candidate, is more likely to use herbaceous biomass crops due to blocking of one of the flurry feed systems of the gasifier in a trial with woody biomass in 2001 (Jeff Curry and McCardle, pers. com., July 27 and August 27, 2004) (Table 3-4). Because of relatively cheap conventional power generation fuels, utilities in central Florida currently pay from $-39 to $11 green Mg-1 ($-35 to +$10 ton-) delivered for biomass. However, existing government incentives for renewable energy that could improve the profit margin of biomass for energy include the Renewable Energy Production Incentive (REPI)1 and the Section 45 Tax Credit for utilities that pay federal income taxes. REPI, authorized under Section 1212 of the Energy Policy Act of 1992, is designed to promote increases in the generation and utilization of electricity from renewable energy sources (U.S.Department of Energy, 2005). REPI offers 1.760 kWh-1, and the Section 45 Tax Credit offers a reduction in taxes of 2.760 kWh-1. Assuming a heat rate of 11,500 BTUs kWh-1 and 9,343 BTUs kg-1 (4,238 BTUs lb-1) woody biomass at 50% MC on a green weight basis, REPI would be worth $14.29 green Mg-1 ($12.97 ton-) or $28.59 dry Mg-1 ($25.94 ton-1) delivered, and similarly the Section 45 Tax Table 3-4. Potential bioenergy markets for Eucalyptus produced in Polk County. Prices are delivered. Estimated Price Quantity (green Mg-1 (green Mg Ha [acres] [ton-']): [tons]): Note: Location: needed 63,500- Lakeland 127,000 2,000-4,000 Lakeland $11 [$10] 127,000 Delivered Price Lakeland, FL 2,000-,000 Electric [70,000- [5,000-10,000] 140,000] 45,400 1,400 Big Bend $11 [$10] 5 0 Delivered Price Apollo Beach, FL [3500 [50,000] [3,500] 91,000- Economic 2,400-5,300 Ridge 181,000 incentives 24053 Ridge $0 [$0]incentives Aubumdale, FL [6,000- Energy [100,000- could be 13,000 200,000] applied. 1 Likely to favor Tampa $11 [$10] n/a herbaceous Mulberry, FL n/a Electric b. biomass. a Approximate acreage needed for sustained production over a year, assuming growth of 15 green tons acre-' year' (i.e., annual demand divided by 15). 1 Imp \ "\ \ .eere.energy.gov/wip/program/repi.html, 02-15-2005 Credit would be worth $44.85 dry Mg-1 ($40.69 ton-) delivered. Based on this survey, stumpage prices for eucalyptus would range from $11-$44 dry Mg1 ($5-$20 green ton', or $10-$40 dry ton' assuming 50% moisture content on a green weight basis). Three biomass prices assumed in this analysis are $10, $20 and $30 dry Mg1. Operational Costs Operational costs on CSAs are higher than those of conventional forestry, as working conditions on sites with heavy clays and/or cogongrass infestation are problematic. A commercial trial of SRWC production on a CSA near Lakeland, Florida incurred costs of $1,800 ha-1 for site preparation and $1,200 ha-1 planting cost (C, and Cp in Equations (3-6) and (3-9)). To assess the sensitivity of LEV to changes in operational costs, values of $900 and $1,800 ha1 for site preparation and $600-$1,200 ha-1 for planting were used, assuming decreasing costs with increased commercialization and economies of scale. In light of apparent growth response to weed control, a weeding cost (Cw in Equations (3-6) and (3-9)) of $0 and $200 ha-1 with the beginning of each growth stage was tested. The model was run assuming interest rates of 4%, 7% and 10%. Additional Non-Timber Benefits Below-ground C sequestration Because the response of below-ground (SOC) accumulation to harvest scheduling is not known, below-ground C sequestration is not incorporated in this model. However, below-ground C sequestration can be estimated and added to calculated LEV as an additional NTB. Root systems of EG grown in a clay settling area in central Florida were 40% of the total biomass (Segrest, 2002), or equivalent to the above-ground inside-bark growth function. Under sustained yield SRWC management, it could be assumed that biomass in root systems peak during the coppice stage that produces the greatest above- ground biomass, and remains steady in subsequent coppice stages and cycles, where decay of dead root systems is replaced by re-growth. Anecdotal evidence suggests that greatest yields at the Kent site occur during the first coppice stage and decline in subsequent coppice stages, as described in Chapter 2. Therefore, the value of C sequestration in root systems for the first coppice stage (s=1) at time t can be defined as Eq. (3-11) and remain constant for the life of the plantation. CR1 (t)= g(t)* 0.47 Pc (3-11) The derivative of Eq. (3-11) is the value of the carbon sequestered in roots discounted to plantation age 0: CBI =[ CR (t) e(r) dt (3-12) d_0 Information about SOC accumulation on CSAs in Florida is limited. Wullschleger et al. (2004) found that on a 25-year-old CSA, SOC under 2.5-year-old plantation of EG at a planting density of 9,800 trees ha-1 accumulated 151 and 96 Mg ha-1 more than SOC under cogongrass in soil depths of 0-30 cm and 30-60 cm, respectively. Their model of soil carbon dynamics estimated that a SRWC EG plantation contributes to the storage of an additional 274 Mg C ha-1 after 25 years, reaching an additional 354 Mg C ha-1 after 50 years. A polynomial function fitted to the data simulation takes the following form: SOC(t)= -0.1668*t2 +15.084*t (3-13) where SOC (Mg ha-1) is expressed as a function of time t (years) after SRWC plantation establishment on a CSA. Eq. (3-13) is then used in the calculation of the NPV of the carbon benefit (Eq. (3-14)). The actual SOC sequestration process is likely to be more complicated than Eq. (3-13) suggests. However, lacking better data, Equations (3-12) and (3-14) can be used to estimate the additional benefit of below-ground (root + SOC) C sequestration benefits. CBsoc = (SOC(t *e dt (3-14) Reclamation incentives As a result of high bulk density, high pH, and the invasion of cogongrass, CSAs are slow to naturally revegetate and are difficult to put into agricultural or forestry production. Tree plantations can contribute to ecosystem restoration of degraded lands by facilitating natural regeneration (Haggar et al., 1997; Lamb, 1998; Lugo, 1997; Powers et al., 1997; Parrota, 1992; Parrota et al., 1997) especially in areas dominated by cogongrass (Otsamo, 2000; Kuusipalo et al., 1995). The establishment of SRWCs on CSAs can reduce soil bulk density, exclude cogongrass, and facilitate the establishment of natural regeneration of native tree species and ecosystem functions. Chapter 378 of the 2004 State of Florida Statutes includes provisions for reimbursement of CSA reclamation costs, ranging from $4,942 -$9,884 ha-1 ($2,000-$4,000 acre-'), funded from taxes on the phosphate mining industry (State of Florida, 2004a). Because it is not known if SRWC establishment would be recognized as a form of CSA reclamation, and because payment would not be a function of stand growth, mined land reclamation incentives are not included in this model. However, providing this reclamation compensation to SRWC systems would contribute to the LEV of SRWC production on CSAs. Summary of Model Inputs and Assumptions The model was run for the three scenarios (no NTB, C sequestration in mulch production, and C sequestration/CO2 displacement in biomass production), under all combinations of interest rates (4% and 7%), site preparation costs ($900 and $1,800 ha-1), planting costs ($600 and $1,200 ha-1), weed control costs ($0 and $200 ha-1), growth functions (low and high) and biomass stumpage prices ($10, $20 and $30 dry Mg-1 assuming whole-tree above-ground harvesting) for a fixed C sequestration incentive of $5 Mg- totaling 288 runs, allowing as many growth stages as needed until LEV begins to decline, assuming growth stages decline by 20% per stage. Additionally, sensitivity of LEV and harvest scheduling to C prices of $15, $25 and $35 was tested at a base scenario, as was increasing the cost of capital to 10%. LEVs exclude below-ground C sequestration benefits, the values of which are estimated independently below. Results and Sensitivity Analysis LEVs increase with growth rate and biomass stumpage price. Under all combinations of assumptions under a fixed C price of $5 Mg-1 C, LEVs range from $-2,789 to $4,616 ha-1 and $-224 to $18,121 ha-1 assuming stumpage prices of $10 and $30 Mg-1, respectively, comparable to LEVs of a SRWC system in the United Kingdom reported by Smart and Burgess (2000) of $3,931, $6,168 and $14,814 ha-1 for market only, low NTB and high NTB model scenarios, respectively (stumpage price of $31 dry Mg-1, establishment cost of $1,538 ha-1 and an exchange rate of $1.54 per in November 2000). Table 3-5 shows LEVs, optimum number of stages per cycle, and optimum stage lengths by C benefit scenario and stumpage price assuming a base scenario of 4% interest rate, $1,800 ha-1 site preparation cost, $1,200 ha-1 planting cost and a carbon price of $5 Mg-1 C. Under these assumptions, marginal increases in LEV per dollar increment in stumpage price range from $264-$293 and $588-$629 under the low growth and high growth functions, respectively. Marginal benefits of increasing stumpage price are greater with the high growth function, as benefits of increased yield are magnified over multiple rotations. The shortest optimal initial growth stage is 2.6 years under conditions of highest stumpage price and interest rate and lowest operational costs, and the longest optimal initial growth stage with a positive LEV is 3.5 years under conditions of high operational costs, low interest rate and low stumpage prices. Ceterisparibus, increasing stumpage price decreases optimum stage lengths and optimum stages per cycle, as the opportunity cost of the value of the stand increases. Incorporating the C incentive in the mulch product scenario increases optimum stage lengths, while applying the incentive in the biofuel scenario decreases optimum stage lengths due to reduced post-harvest emissions penalties, though differences in stage lengths are less than 1/10th of a year (Table 3-6). Table 3-5. LEV, optimum number of stages and optimum stage length for each stage by C benefit scenario and biomass price assuming a base scenario of 4% interest rate, $1,800 ha1 site preparation cost, $1,200 ha-1 planting cost, no post- establishment weeding cost, and a carbon price of $5 Mg- C. $10 dry Mg-1 $20 dry Mg-1 $30 dry Mg-1 NTB Growth LEV Optimum LEV Optimum LEV Optimum ($/ha) harvest age ($/ha) harvest age ($/ha) harvest age (years) (years) (years) None Low -1,967 3.1, 3.1, 3.2, 674 2.9, 2.9, 2.8, 3,722 2.8, 2.8, 3.3,3.4 2.6 2.6 C(M) Low -1,883 3.1, 3.1, 3.2, 771 2.9, 2.9, 2.8, 3,828 2.8, 2.8, 3.2,3.3 2.6 2.6 C(B) Low -1,424 3.0, 3.1, 3.1, 1,320 2.9, 2.9, 2.8, 4,448 2.8,2.8, 3.1, 2.9 2.6 2.6 None High 619 3.4, 3.4, 3.3, 6,507 3.2, 3.1, 2.9 12,960 3.2,3.0 3.0 C(M) High 810 3.4, 3.4, 3.3, 6,715 3.2, 3.1, 2.9 13,140 3.2,3.0 3.0 C(B) High 1,832 3.4, 3.4, 3.3, 7,869 3.2, 3.1, 2.9 14,419 3.1,3.0 2.9 Raising incentives for C sequestration increases LEV (Table 3-6). Under a base scenario of $20 dry Mg-1 stumpage price, interest rate 4%, site preparation $1,800 ha-1, planting cost $1,200 ha-1, high growth function and no post-establishment weeding, increasing the price of C from $0 to $35 ha-1 increased LEVs from $6,507 to $7,965 and $6,507 to $16,422 ha-1 for the mulch and biofuel scenarios, respectively. The marginal increase in LEV per dollar increment in C price is a constant $42 in the mulch scenario. Conversely, the marginal benefit in the biofuel scenario was both higher and more responsive to increases in C price, ranging from a marginal increase of $272 to $292 at $5 and $35 Mg-1 C, respectively. This reflects that the biofuel model is less penalized by post-harvest decay of sequestered C, thus increasing incentives for biofuel production rather than in situ sequestration. Table 3-6. LEV ($ ha-1), optimum stage lengths, marginal benefit, and estimated below- ground benefit ($ ha-1) by C sequestration incentive ($ Mg-1) under a base scenario of $20 dry Mg-1 stumpage, interest rate 4%, site preparation $1,800 ha-l, and planting cost $1,200 ha-1. Optimum Stage Marginal Benefit (ALEV Below-ground $ Mg-1 C LEV ($ ha-1 ) Lengths (years) per $1 C Incentive) ($ ha1) Mulch scenario 0 $6,507 3.2,3.1, 2.9 n/a n/a 5 $6,715 3.2, 3.1, 2.9 $42 $1,163 15 $7,131 3.3, 3.1, 2.9 $42 $3,492 25 $7,548 3.3, 3.2, 2.9 $42 $5,819 35 $7,965 3.3, 3.2, 2.9 $42 $8,097 Biofuel scenario 0 $6,507 3.2,3.1, 2.9 n/a n/a 5 $7,869 3.2, 3.1, 2.9 $272 $1,163 15 $10,598 3.2, 3.1, 2.8 $273 $3,492 25 $13,505 3.1,3.0 $291 $5,819 35 $16,422 3.1,3.0 $292 $8,097 The marginal reduction of LEV per percent increase in the cost of capital between 4% and 7%, assuming a C price of $5 Mg-1 C, is -$23 under the least profitable scenario and -$2,928 under optimum assumptions. For a base scenario of $1,800 ha-1 site preparation cost, $1,200 ha-1 planting cost, carbon price of $5 Mg-1 C, high growth function and no weeding costs, the marginal impact of increasing interest rates between 4% and 10% ranged from -$192 to -$2,581 (Table 3-7). More profitable scenarios are penalized more by higher interest rates. Increasing interest rates had little effect on optimum stage lengths (Table 3-8). Increases in interest rates from 4% to 7% and from 7% to 10% decreased optimum stage lengths by 1/10th of a year or less. At increases from 7% to 10% the model selected for optimization with an additional growth stage. This effect is consistent with results from Smart and Burgess (2000), who observe that in SRWC biomass systems the opportunity cost of the standing biomass is low relative to Table 3-7. Change in LEV ($ ha-1) per 1% increase in interest rate assuming $1,800 ha-1 site preparation cost, $1,200 ha-1 planting cost, carbon price of $5 Mg-1 C, high growth function and no weeding costs, without C sequestration incentives, C sequestration for the mulch production scenario, and C sequestration for the biofuel production scenario. $10 dry Mg-' $20 dry Mg-' $30 dry Mg' % LEV ($ ALEV/+1% LEV ($ ALEV/+1% LEV ($ ALEV/+ Interest ha1) Interest ha-') Interest ha-') 1% Rate Interest No NTB 4% $619 $6,507 $12,960 7% -$798 -$472 $2,413 -$1,365 $5,864 -$2,365 10% -$1,375 -$192 $762 -$550 $3,057 -$936 Mulch 4% $810 $6,715 $13,140 Scenario 7% -$616 -$475 $2,608 -$1,369 $6,029 -$2,370 10% -$1,213 -$199 $946 -$554 $3,239 -$930 Biofuel 4% $1,832 $7,869 $14,419 Scenario 7% -$88 -$640 $3,197 -$1,557 $6,677 -$2,581 10% -$880 -$264 $1,315 -$627 $3,611 -$1,022 the opportunity cost of the land, and thus increasing interest rate does not shorten rotations as it would with a conventional system, but rather LEVs are reduced, lowering the opportunity cost of the land relative to the marginal benefit of the stand growth, and stage lengths remain relatively unaffected, while the coppice cycle is extended to delay the cost of replanting. Table 3-8. Optimum harvest scheduling (stage lengths and number of stages per cycle) at interest rates of 4%, 7%, and 10% assuming $1,800 ha-1 site preparation cost, $1,200 ha-' planting cost, carbon price of $5 Mg-1 C, high growth function and no weeding costs, without C sequestration incentives, C sequestration for the mulch production scenario, and C sequestration for the biofuel production scenario. $10 dry Mg-1 $20 dry Mg-1 $30 dry Mg-1 % Optimum Optimum Optimum Optimum Optimum Optimu Interest number of stage number of stage number of stage Rate stages per lengths stages per lengths stages per lengths cycle (years) cycle (years) cycle (years) No NTB 4% 4 3.4,3.4, 3 3.2,3.1, 2 3.2,3.0 3.3,3.0 2.9 7% 4 3.3,3.3, 3 3.2,3.1, 2 3.1,3.0 3.3, 3.1 2.9 10% 5 3.2,3.2, 3 3.1,3.1, 3 3.0,3.0 3.3, 3.2, 2.9 2.8 2.9 Mulch 4% 4 3.4,3.4, 3 3.2,3.1, 2 3.2,3.0 Scenario 3.3,3.0 2.9 7% 4 3.3,3.3, 3 3.2,3.1, 2 3.1,3.0 3.3,3.1 2.9 10% 5 3.3, 3.3, 3 3.1,3.1, 3 3.1, 3.0 3.3, 3.2, 2.9 2.8 3.8 Biofuel 4% 4 3.4,3.4, 3 3.2,3.1, 2 3.1,3.0 Scenario 3.3,2.9 2.9 7% 4 3.3,3.3, 3 3.2,3.1, 2 3.1,3.0 3.2,3.0 2.9 10% 5 3.2, 3.3, 3 3.1,3.1, 3 3.1, 3.0 3.2,3.1, 2.9 2.7 2.5 Increases in operational costs decrease LEV (Table 3-9). Increases in site preparation, which are one-time up-front costs, have a dollar-for-dollar reduction in LEV. LEVs decrease $3 per dollar increase in planting costs, with slightly higher marginal impacts at higher stumpage prices, reflecting shorter coppice cycles and increased planting frequency. Weed control may be needed to insure high yields, though the exact impact of weed control on growth is not known. LEV is reduced $8 for every dollar increase in weed control cost applied at the beginning of each growth stage. Marginal impacts shown in Table 3-9 are the same under the three NTB scenarios, except for the im , ), marginal impact of weeding increases from -$8 to -$9 in the biofuels scenario assuming $30 Mg-1, reflecting the shorter optimal stage lengths and more frequent weeding. Table 3-9. LEVs and marginal impact on LEVs by changes in site preparation, planting and weeding costs, assuming a C price of $5 Mg-1, 4% interest rate, high growth function and no NTB. $10 dry Mg-' $20 dry Mg-' $30 dry Mg-' Input LEV ALEV/ LEV ALEV/ LEV ALEV/ Values ($ ha') AInput ($ ha') AInput ($ ha-1) AInput ($ ha') Site preparation $900 $1,519 $7,407 $13,860 (low) Site preparation $1,8001 $619 -$1 $6,507 -$1 $12,960 -$1 (high) Planting (low) $600 $2,354 $8,963 $15,754 Planting (high) $1,2001 $619 -$3 $6,507 -$4 $12,960 -$5 Weeding (low) $0o $619 $6,507 $12,960 Weeding (high) $200 -$937 -$8 $4,831 -$8 $11,261 -$8 Base scenario assumptions. The value of below-ground C sequestration, exogenous in this model, was estimated separately (Table 3-10). The estimated value of SOC sequestration, comprising the majority of the below-ground carbon benefit, is influenced only by C price and interest rate. The value of C sequestration in roots is additionally influenced by the growth and yield function. The SOC model by Wullschleger et al. (2004) yields 341 Mg ha-1 from 0-60 cm depth at 45 years, at a rate of 7.5 Mg SOC ha-1 year-,whichis greater than 136.3 Mg SOC ha-1, the average for longleaf-slash pine stands to 1 meter depth reported by Heath et al. (2003). The rate of accumulation is an order of magnitude more than sequestration rates reported from tree plantations on agricultural lands (Garten, 2002) but is closer to the 1-3 Mg SOC ha-1 year- sequestration rate reported in the top 30 cm of reclaimed minesoils over 25 years in Ohio (Lal & Akala, 2001), and might be influenced by the longer growing season and deeper measurement depth. Estimating carbon sequestered in roots as equivalent to 40% of the total biomass or 67% of the above ground biomass (based on Segrest, 2002) yields 15 and 31 Mg C ha-1 after three years for the EA 3 and EA 4 growth curves, respectively. This is more than the 6.6 and 7.4 Mg ha- 1 of below-ground organic matter after three years with the cultivation of sycamore (Plantanus occidentalus) in Tennessee and Mississippi, respectively, reported by Tobert and Thornton et al. (2000) though higher rates of sequestration are to be expected with a longer growing season and faster growing Eucalyptus spp.. Assuming a C price of $5 Mg-1, total estimated below-ground C benefits range from $650 ha-1 (low growth function and 10% interest rate) to $1,172 ha-1 (high growth function and 4% interest rate). Raising the C price to $15 and $25 Mg-1 approximately increases the below ground C benefit by 3 and 5 times, respectively. Table 3-10. Estimated discounted value of below-ground C benefits by C price, interest rate and growth function. Roots Estimated Total Below-ground C Benefit C price Interest Growth Minimum Maximum SOC Minimum Maximum ($ Mg1) Rate Function $5 4% Low $67 $74 $1,014 $1,081 $1,088 High $123 $158 $1,014 $1,137 $1,172 7% Low $64 $71 $751 $815 $822 High $117 $149 $751 $868 $900 10% Low $61 $67 $589 $650 $656 High $111 $140 $589 $700 $729 $15 4% Low $202 $223 $3,042 $3,244 $3,265 High $368 $473 $3,042 $3,410 $3,515 7% Low $192 $212 $2,254 $2,446 $2,466 High $350 $446 $2,254 $2,604 $2,700 10% Low $183 $202 $1,768 $1,951 $1,970 High $333 $421 $1,768 $2,101 $2,189 $25 4% Low $336 $372 $5,069 $5,405 $5,441 High $614 $789 $5,069 $5,683 $5,858 7% Low $320 $353 $3,757 $4,077 $4,110 High $584 $744 $3,757 $4,341 $4,501 10% Low $305 $336 $2,946 $3,251 $3,282 High $555 $702 $2,946 $3,501 $3,648 To compare these findings with production costs calculated by a previous study, this model was used to find minimum stumpage prices needed to achieve LEVs of $1,235 ha-1 and $2,470 ha-1, representing LEVs of conventional forestry (Borders & Bailey, 2001) and Florida agricultural land (Reynolds, 2005), respectively. Stumpage prices of $17 and $21 dry Mg-1 are required to match LEVs of $1,235 ha-1 and $2,470 ha-1, respectively, assuming site preparation costs of $1,800 ha-1, planting costs of $1,200 ha-1 and averaging the EA 3 and EA 4 growth functions, equivalent to -25 dry Mg ha-1 year1 and an interest rate of 5%. Rahmani et al. (1997) report Eucalyptus spp. farm gate production costs for Florida of $32-$39 dry Mg-1, slightly less than the $39-$43 dry Mg-1 farm gate costs estimated here assuming a harvest cost of $22 dry Mg-1 (Rahmani et al., 1998). A higher cost of production is expected given the cost of site preparation on CSAs. Conclusions Under these assumptions, even assuming high establishment and planting costs ($1,800 and $1,200 ha-1, respectively), a reasonable stumpage price ($20 dry Mg-1) and excluding C sequestration incentives, production of EA on CSAs in central Florida is profitable, with LEVs ranging from $762 to $6,507 ha-1 assuming interest rates of 10% and 4%, respectively. With the incorporation of a C sequestration benefit of $5 Mg-1 LEVs increase to $946 and $6,715 ha-1, while recognizing the CO2 mitigation benefits associated with the biofuel scenario increases LEVs to $1,315 and $7,869 ha-1 assuming interest rates of 10% and 4%, respectively. In addition, the societal value of below- ground C sequestration (roots + SOC at $5 Mg1 C) is likely to be $1,081-$1,172 ha-1 or $815-$900 ha-1 assuming discount rates of 4% and 7%, respectively. The influence of stumpage price, C sequestration benefit (CO2 mitigation scenario or C price) or interest rate (from 4% to 10%) on optimum stage lengths is less than one year, and is probably operationally unimportant. Increasing incentives for CO2 mitigation can increase or decrease optimum stage lengths by about 0.1 year in the mulch and biofuels scenarios, respectively. Harvesting on CSAs would likely be scheduled during the months of December-February when sites are more accessible and coppice response to harvest is best, and practical application of this model is more likely in evaluating the economic viability of the system rather than projecting optimum harvest scheduling to sub-year accuracy. However, this model could be used to suggest the optimum number of stages per cycle and optimal harvest scheduling by identifying the winter closest to the optimum harvest age. Because of the short growth stages, penalties for post-harvest CO2 emissions from product decay are discounted much less than those of conventional rotations of 20 or more years, countering benefits of in situ C sequestration, and underscoring the importance of recognizing the CO2 mitigation benefit of displacing fossil fuels in the biofuel scenario. These results emphasize both the potential for DFSSs on CSAs to mitigate atmospheric CO2 and for CO2 mitigation incentives to contribute to the profitability of SRWC production. Increases in LEV from CO2 displacement benefits are on par with increases gained from SOC sequestration, and to a lesser degree, in situ sequestration in above- and below-ground biomass. It would probably be impractical to provide incentives and penalties for the sequestration and decay of C for SRWC systems on a per- harvest basis, given the frequent harvest rate vis a vis conventional forestry systems. However, this model might be used to assess the present value of CO2 mitigation benefits over the life of the stand, providing the opportunity to offer incentives without monitoring of each biomass harvest. Though payment of C sequestration benefits independent of harvest monitoring could cause a divergence of private and socially optimum harvesting, these results suggest there is little difference in optimum harvest scheduling of private versus socially optimal SRWC production, and in fact both optimum stage lengths and stages per coppice cycle decrease in the biofuel production scenario, indicating that harvest monitoring might not be needed for a successful CO2 mitigation program. In the biofuel production scenario, probably the easiest way to incorporate CO2 mitigation benefits would be for utilities to pass on CO2 emissions reductions incentives to producers by increasing stumpage price. In light of uncertainty associated with SRWCs, potential financiers might expect a high rate of return on their investment. These results suggest that SRWCs can be profitable at interest rates of 10%, assuming some combination of adequate yields, stumpage prices, NTB incentives and/or operational costs are achieved. Future Research Research is needed to verify the assumptions made in this analysis. The most immediate need is for a better understanding of growth response to treatment options such as weeding and fertilization. With more information, particularly with regards to below-ground C sequestration, growth functions and coppice growth, this model can be used to make case-specific evaluations. A better understanding of long-term impacts of SRWC production on CSAs and eligibility for mined-land reclamation incentives would be beneficial, as would assessments of economic multiplier effects on communities in Polk County. Upcoming work of SFRC students regarding the use of SRWCs to control cogongrass and facilitate natural regeneration could contribute to this analysis. In light of 72 the 2004 hurricane season, a feasibility analysis incorporating risk assessment could be useful in assessing potential advantages of SRWCs to reduce the probability of hurricane damage. CHAPTER 4 ECONOMICS OF SLASH PINE CULTURE ON TITANIUM MINED LANDS IN NORTH CENTRAL FLORIDA Introduction Comparable to the phosphate mining industry in central Florida, titanium and zircon mining by Iluka Resources Inc and Dupont is prevalent in northeast Florida, with 1,600 ha (4,000 acres) mined in Clay County since the early 1970s. In the mining process, forest cover is removed, topsoil is retained, and through either dredge or dry mining, soil is processed, minerals are removed, and the homogenized soil is replaced. In response to concerns of environmental impacts of titanium mining, the Surface Mining Control and Reclamation Act of 1977 requires that mining operations re-apply topsoil on mined sites to restore wildlife habitat and hydrologic functions. Another significant contributor to the economy of northeast Florida is the forest products industry. In Clay County, Iluka and DuPont establish slash pine plantations on reclaimed mines to produce timber products and restore ecosystem functions. Unlike the experimental production of SRWCs on mined lands assessed in Chapter 3, slash pine culture on titanium mined lands in northeast Florida is well-established. Darfus and Fisher (1984) found young slash pine plantations established on mined lands in the mid- 1970s had poor survival and growth as a result of unleveled contours and disrupted soil moisture regimes. However, Mathey (2001) in a study of slash pine plantations established between 1978 and 1996 on lands mined by Iluka found no significant difference between site indices of reclaimed and unmined lands, though averages varied 74 slightly (21.2 m and 22.0 m respectively, base age 25 years), and stem analysis showed similar height and diameter at breast height (dbh) growth patterns on 8-, 10-, and 16-year- old stands on mined and unmined lands under identical management regimes (Figures 4-1 and 4-2). 18 --- Reclaimed 16 14 -- Unmined 12 0 6 - 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Age of Tree (years) Figure 4-1. Mean heights estimated by stem analysis from stands on 25 reclaimed and 25 unmined sites (Mathey, 2001). 24 -- Reclaimed 22 - -- Unmined 16 14 S12 m 10 6 -- 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Age of Tree (years) Figure 4-2. Mean diameter inside bark (DIB) estimated by stem analysis from stands on 25 reclaimed and 25 unmined sites (Mathey, 2001). Iluka has a vested interest in the productivity of post-mining landscapes, as do private landowners who lease mining rights to Iluka. An assessment of the impacts of titanium mining on the economics of forestry would contribute to land management decisions in northeast Florida. Silvicultural practices such as fertilizing, bedding, and subsoiling may improve tree growth and forest profitability on mined lands while facilitating mined land restoration (Proctor et al., 2003). This chapter assesses the profitability of slash pine production on mined and unmined lands in northeast Florida and the economic viability of silvicultural treatments that might be used to improve production on mined lands. Methodology Economic Model As described in Chapter 1, Eq.(1-1) defines LEV, net returns of a forestry practice projected in perpetuity, where V(t) is the value function of the stand at age t, r is interest rate, and C is the sum of stand establishment costs discounted to the beginning of the rotation. This model is used to compare bare-land values of mined and unmined lands under slash pine production. In this case of a non-coppicing species, the Faustmann model remains fixed for one growth stage per cycle (i.e., one rotation). Calculating LEV on mined land vis-a-vis unmined land requires that stand establishment cost C be accounted for differently. Topsoil replacement and contouring costs are considered sunk, as these treatments are required by law regardless of the subsequent land use, and land clearing costs of the first rotation are excluded from the mined land simulations to accurately represent actual establishment costs on un-vegetated mined land. In these cases, LEV and optimum rotation age are calculated using Eq.(4-1), a variation of Eq.(1-1), in which initial costs C, are accounted for at the beginning of the projection and standard establishment costs C are assumed at the beginning of all subsequent rotations. Alternatively, cost savings at year zero can be added to Eq. (1-1), yielding the same result. LEV (V(t)- C)*e r*t LE1 V =t C (4-1) -e -rt To determine the financial viability of slash pine production on reclaimed titanium mined lands, Equations (1-1) and (4-1) were used to a) compare LEVs of established slash pine stands on 34 mined and 29 unmined sites, b) assess the economic viability of fertilizer and subsoil treatments on mined lands, c) estimate minimum economically feasible growth response under estimated treatment costs, and d) estimate maximum economically feasible treatment costs under predicted fertilizer responses. Use of Equations (1-1) and (4-1) requires that stumpage value V(t) be defined as product price times the yield function for each product. Growth and Yield Model The Plantation Management Research Cooperative at the University of Georgia, in conjunction with the forest industry, monitored slash pine growth and response to silvicultural treatments. Using data from this study, Pienaar and Rheney (1995) developed growth and yield models for slash pine plantations. The growth and yield functions require determination of average dominant height and basal area at time t. Average height H(t) is predicted by Eq. (4-2)1: SI*1.3679* 1-e(-007345*t) H (t)-= /36 [0.305] (4-2) (0.678 z + 0.546 z +1.395 *3 + 0.412* z I* 3)* te( 0691*t 1 Units in this chapter are shown in both metric and imperial units to facilitate interpretation within the context of the Florida forest industry. Squared brackets in equations indicate conversion from imperial to metric. where Slis site index (SI) in m at base age 25, and zl 1 if fertilized, 0 otherwise; z2= 1 if bedded, 0 otherwise; and z3 = 1 if herbicide is applied, zero otherwise. Tree survival (S) is estimated by S(t) =N *e(t21345 11345) (4-3) where S (stems ha-1) at year t2 is a function of the number of surviving trees N (stems ha-1) at year tl (alternatively, both S and N can be expressed as stems acre-1). Using Equations (4-2) and (4-3), basal area (B, m2 ha-) at time t is predicted by Eq. (4-4): B(t) = 3 394-3566) (t)(1 366+6 20) *S(t)(0 366+3 15[0.23] (4-4) ,(0.557* z4 +0.436* zl +2.134* z3 -0.354* zl* z2)*t* e(-009*t) Using the above three equations, predicted total volume (m3 ha-) at time t is estimated by Eq. (4-5): T(t) = H(t)82 *S(t)( 0017 03) *B(t)(1016 +00) *[0.07] (4-5) Equation (4-5), subsequent yield functions, and eventually Eq. (1-1) are largely a function of SI, which could be used to compare LEVs of mined lands to unmined lands. Mathey (2001) measured height and dbh in 116 1/50-hectare plots in 1-, 2-, 3-, 8- 10-, and 16-year-old plantations on mined and unmined lands managed by Iluka. However, Mathey found no significant difference between SIs of reclaimed and unmined sites. As variation in SI does not correspond to similar variation in LEV, and to apply a SI equation using parameters for slash pine, SIs were recalculated for 1/50-hectare plots on ten 8-, eight 10-, four 13-, two 15-, five 16- and five 21-year-old stands on mined lands, to determine LEVs under a range of SIs on mined land. SIs were calculated with Eq. (4-6): ( 2 0669 SI H* 0.91861e 0 10035A (4-6) where H is average height (m) of dominant and co-dominant trees at sample age A and SI is site index (m) at base age 25 (Bailey, 1982). Results of a Student's t-Test again showed no significant differences of SI between mined and unmined sites, averaging 19.7 m and 21.0 m respectively. To better reflect observed plot-specific volumes on mined and unmined lands Eq. (4-5) was adjusted following Davis and Johnson (1987) by Eq. (4-7): Tt) = Tt (t t) (4-7) where the total volume prediction equation T(t)p was multiplied by the observed volume T(to)o divided by total predicted volume also at time of observation to. T(to)o (m3) was calculated by Eq. (4-8) T(to) = 0.00616 *([0.394]* dbh)20578 *([3.28]* H)07468 *[0.0283](4-8) where dbh is dbh in cm and H is height in m (Brister et al., 1980)2. Volumes where then summed for each plot multiplied by 50 for volume per ha. Based on the adjusted total volume function derived from Eq. (4-7), product-specific volumes were then derived. Eq. (4-9) yields the volume of sawtimber V(t)saw (m3 ha-) : 0 'i/ 43 f~)06 l 01 / }, 43 8 0- 5* Q 8()*[ 541 384 69*S(t)-1 12 54dd )* 5 2 V(t) T(t) *e *[0.07] (4-9) 2 Alternatively, T(t)o can be calculated in ft3 by eliminating the numbers in brackets, with dbh and H as inches and feet, respectively. where dt is top merchantable diameter outside bark and and Q(t) is the quadratic mean of dbh, as expressed by Eq. (4-10): B (t) Q(t) =S (4-10) 0.005454*[0.0144] Chip-and-saw and pulpwood product classes were then estimated by Eqs. (4-11) and (4-12), respectively: S= T (t)- 52* (t)*[254] 84-0 69*S(t-) 12d (t)*[2 54].07] (4-11) V(t V (t)P = T(t)a *e 052 384 (t) V7 (t)W [0.07] (4-12) An example of pulp, chip-and-saw, sawtimber, and total outside bark volumes predicted using Eqs. (4-2)-(4-6) and adjusted to replicate observed volumes using Eq. (4-7)-(4-12) is shown in Figure 4-3. Northeast Florida merchantable standards used in this analysis are shown in Table 4-1. Table 4-1. Merchantable standards of dbh and top diameter outside bark (dr). dbh dt cm in cm in Sawtimber 24.4 9.6 21.8 7.6 Chip-and-Saw 16.8 6.6 9.1 3.6 Pulpwood 9.1 3.6 9.1 3.6 The value of the stand at time t was then be expressed as V (t) = ps,w V (t)sw+ Pns V (t) + Pp V (t)pW (4-13) 250 - 200 - 50 -- - 0 ---------- -- ------------ ----------- ---------------------------- 0 5 10 15 20 25 30 Time (years) Pulp SAdjusted Pulp Chip-and-saw dj ed Chip-and-saw Sawtmber Adjusted Sawtimber Total Predcted Volume ---- Adjusted Total Predcted Volume Figure 4-3. Representative pulp, chip-and-saw, sawtimber, and total outside bark volumes (m3 ha-1), predicted using Eqs (4-2)-(4-6) (solid lines) and adjusted to replicate observed volumes (dotted lines) using Eq. (4-7). where psaw, Pcns, and ppw are the price per volume of sawtimber, chip-and-saw, and pulpwood products, respectively. Product prices were defined to incorporate Eq. (4-13) into Eq. (4-1) to calculate LEV and optimum rotation age as described in Chapter 2. Market Assessment The conventional softwood forest products market in northeast Florida is much larger and more established than that of Eucalyptus spp. in south Florida. The forest industry has the highest economic impact in Florida of any agricultural crop and contributes over $16.6 billion annually (Hodges et al., 2004), with most of the state's pine inventory in northeast Florida (Carter & Langholtz, 2005). Assuming constant South-wide softwood demand, removals in Florida are projected to increase over the projection period due to relative abundance of supply as compared to other states. In northeast Florida from 2000-2020, removals are projected to increase slightly from 5.9 to 7.0 million m3 (210 to 250 million ft3), and inventory is projected to fluctuate between 59 and 67 million m3 (2.1 and 2.4 billion ft3). Even assuming South-wide removals decrease 1 percent annually, removals in northeast Florida are projected to remain fairly constant until 2020, fluctuating from 6.1 to 5.9 million m3 (216 to 209 million ft3) (Carter & Langholtz, 2005). The Iluka mining operation lies within 32 km (19 miles) of a Georgia- Pacific multi-product sawmill near Palatka and is expected to have access to timber markets for the foreseeable future. Mathey (2001) reported results of an economic assessment of slash pine production on Iluka's mined lands. Stumpage values used in said analysis of $93-$407 m-3 ($89- $391 green ton'1 assuming 1.04 green tons m-3) are inconsistent with the range of values reported from Timber Mart South over the past 10 years (Figure 4-4). Analysis in this chapter assumes values of $8.10, $26.62, and $41.27 m-3 ($20.19, $66.34 and $102.83 cord-1) forpp, pens, andppa, respectively (Timber Mart-South 1st Quarter 2005 average stumpage prices for Florida)3, typical of prices since 1995 (Figure 4-4). $50 $40 -----So-------- -----im w- $20 ------------ Pu Cip-n-aw $20 1 95 1Q 97 1Q 99 1Q01 Q 03 10 05 Figure 4-4. South-wide pine stumpage prices quarterly averages from 1995-2005 (Timber Mart South 2005). 3 Assumes about 2.5m3 cord-' (Appendix A Eq. 3). Silvicultural Alternatives Fertilizer amendments, weed control, and mechanical soil preparation can improve tree growth and contribute to pine plantation productivity (Dickens et al., 2002). However, the benefit of these treatments on mined lands is not well known. On mined lands in the Appalachian region, inorganic N fertilizer amendments increased herbaceous biomass production during the first growing season but did not affect hybrid loblolly pine (P. taeda) growth at 2 and 3 years of age. Increased seedling growth with organic amendments was more a function of moisture retention than soil nutrient availability (Schoenholtz et al., 1992). In a reforestation experiment testing the growth of three pine species on surface-mined sites in coalfields of southwest Virginia, fertilization had little effect on growth and was not as beneficial for tree establishment as an herbicide treatment (Torbert et al., 2000). Mathey (2001) and Proctor (2002) established field trials at Iluka testing the influence of fertilizer, herbicide, and subsoil treatments on slash pine and loblolly pine growth and survival on mined and unmined lands. SRWC-84, established December 9, 1999, tested 10 combinations of fertilizer/herbicide treatments (Table 4-2), and heights were measured at 1, 2, 3 and 5 years of age. SRWC-84-2001, established January 9, 2001, included treatments of 13 fertilizer/herbicide combinations (Table 4-2), as well as mycorrhization, humate incorporation, and subsoiling on the mined site, and was measured at 1, 2, and 4 years of age. In the SRWC-84 study, height and survival responses at 1, 2, 3, and 5 years of age were significantly different for both land type (satellite mined and unmined) and treatment (p< .0001). At age 5, trees averaged 4.2 m (14 ft) and 5.5 m (18 ft) tall on mined and unmined land, respectively, but survival was better on the mined land, averaging 85% and 50% on mined and unmined land, respectively. This trend of reduced growth but increased survival on mined land compared to unmined land is consistent with measurements of young stands by Mathey (2001). On the mined land, treatment G2 had the highest average height at 2, 3 and 5 years of age (Figure 4-5). A Duncan grouping analysis was used to identify treatments of similar growth and survival (Table 4-3). On the mined land at ages 2 and 3, treatments G2 and B2 were grouped with highest growth; at age 5 heights of treatments G2>B2>D2>MO were grouped highest, averaging 4.4 m tall, suggesting better response to post-establishment fertilizer application. On unmined land height, responses to treatments showed less variation (Figure 4-7) ranging from Table 4-2. Treatments included in the SRWC-84 and SRWC-84-2001 studies (Proctor, 2002). All fertilizer applications were applied at a rate of 40.3 kg N/ha (361bs N/ac). Treatment Description SRWC-84 SRWC-84- 2001 C Bedding only, no amendment X X G2 Granulite 5-3-0, broadcast in year 2 X X D2 DAP 18-46-0, broadcast in year 2 X X B2 16-4-8 with balanced micronutrients, X X broadcast in year 2 GOR Granulite 5-3-0, broadcast at planting, X X herbicide treatment DOR DAP 18-46-0, broadcast at planting, herbicide X X treatment BOR 16-4-8 with balanced micronutrients, X X broadcast at planting, herbicide treatment GORL Loblolly, Granulite 5-3-0, broadcast at X X planting, herbicide treatment HO Dry aluminum humate broadcast at planting at X X .35% by weight MO Mycorrhizal treatment at the time of planting, X X bedding only GOH Granulite 5-3-0, broadcast at planting, and X humate material at .35% DOH DAP 18-46-0, broadcast at planting, and X humate material at .35% BOH 16-4-8 with balanced micronutrients, and X humate material at .35% 4.8-5.7 m with all nine treatments sharing the same Duncan grouping by age 5 (Table 4- 3). Average heights by treatment of SRWC-84 and SRWC-84-2001 on January 15, 2005 are shown in Figure 4-13. The survival response to treatment on the mined land varied little, ranging from 78-97%, with two Duncan groups, both including the control (Figure 4-6), while survival on the unmined land varied by treatment from 8%-86% (Figure 4-8) resulting in five Duncan groups. Table 4-3. SRWC-84 age 5 and SRWC-84-2001 age 4 mined (SM) and unmined (UM) average heights, standard deviation and Duncan grouping ranked from tallest to shortest by treatment. SRWC-84 SM Age 5 G2, 4.6, 0.8, a B2, 4.5, 0.9, a D2, 4.2, 0.9, ab MO, 4.2, 1.1, ab GOR, 4.1, 0.9, bc HO, 4.1, 0.8, bc BOR, 4, 1, bc C, 3.9, 0.9, bc DOR, 3.7, 1.2, d GORL, 3.3, 0.6, d SRWC-84 UM Age 5 G2, 5.7, 0.8, a GORL, 5.7, 1.3, a D2, 5.5, 1, a C, 5.4, 0.9, a BOR, 5.4, 0.4, a DOR, 5.3, 0.6, a B2, 5.2, 1, a MO, 5, 0.8, a GOR, 4.8, 1, a 2001 SM Age 4 BOR, 2.9, 23, 0.5, a DOH, 2.8, 15, 0.7, ab DOR, 2.8, 25, 0.6, ab B2, 2.6, 49, 0.7, abc MO, 2.5, 40, 0.5, abcd G2, 2.5, 60, 0.8, abcd D2, 2.5, 44, 0.7, bcde C, 2.2, 61, 0.5, cdef GOH, 2.2, 17, 0.5, defg GORL, 2.2, 30, 0.7, fg HO, 2.1, 30, 0.6, fg BOH, 2.1, 16, 0.8, fg GOR, 1.8, 26, 0.6, 2001 UM Age 4 BOR, 4.0, 0.4, a D2, 3.7, 0.5, ab DOR, 3.7, 0.7, ab G2, 3.7, 0.5, ab B2, 3.6, 0.7, bc GOR, 3.3, 0.4, c GORL, 2.9, 0.3, d C, 2.8, 0.7, d Ranking Table 4-4. SRWC-84 age 5 and SRWC-84-2001 age 4 mined (SM) and unmined (UM) average survival (%) and standard deviation by treatment, ranked from highest to lowest. Letter "a" indicates shared highest Duncan group. Ranking SRWC-84 SM Age 5 1 GORL, 97, 5.9, a 2 DOR, 92, 1.2, 5.9, ab 3 BOR, 92, 5.9, ab 4 G2,92,3.4,ab 5 HO, 89, 5.9, ab 6 B2, 83, 3.4, ab 7 C, 82, 3.4, 0.9, ab 8 GOR, 81, 5.9, 0.9, b 9 MO, 81, 5.9, b 10 D2,78,3.4,b E I3 2 - SRWC-84 UM Age 5 GORL, 86, 7.7, a G2, 65, 4.5, b D2, 59, 4.5, b B2, 58, 7.7, b C, 48, 4.5, bc GOR, 31, 7.7, cd DOR, 31, 7.7, cd MO, 17, 7.7, de BOR, 8, 7.7, e 2001 SM Age 4 GOH, 94, 17, 0.5, a GORL, 92, 30, 0.7, a BOH, 89, 16, 0.8, a G2, 88, 60, 0.8, a C, 88, 61, 0.5, a HO, 86, 30, 0.6, ab DOH, 83, 15, 0.7, ab GOR, 81, 26, 0.6, ab MO, 74, 40, 0.5, ab B2, 69, 49, 0.7, ab DOR, 69, 25, 0.6, ab D2, 64, 44, 0.7, b BOR, 64, 23, 0.5, b 2001 UM Age 4 GORL, 81, 0.3, a G2, 73, 0.5, a C, 69, 0.7, a GOR, 66, 0.4, ab DOR, 61, 0.7, ab D2, 52, 0.5, b B2, 31, 0.7, c BOR, 25, 0.4, c - BOR B2 C DOR +KD2 --GOR --GORL G2 HO MO 1 2 3 4 5 Age (years) Figure 4-5. Average heights (m) by age (year) and treatment, SRWC-84 mined site. 86 100% -- 90% --- -- BOR 80% B2 70% C S60% DOR > --D2 50% 50% 40%- SGORL 30% G2 20% HO 10% MO 0% - 0 2 4 6 Age (years) Figure 4-6. Average survival (%) by age (year) and treatment, SRWC-84 mined site. 6 5 -- BOR -- B2 4 c E DOR -a 3 -- -- D2 -*- GOR 2 GORL -G2 1 MO 0 - 0 1 2 3 4 5 Age (years) Figure 4-7. Average heights (m) by age (year) and treatment, SRWC-84 unmined site. |