Bio-economic Factors Affecting Feasibility of Floating In-pond Raceway Systems: A Stella Modeling Approach by Rachel McLemore Regan A thesis submitted to the Graduate Faculty of Auburn University in partial fulfillment of the requirements for the Degree of Master of Science Auburn, Alabama August 6, 2011 Copyright 2011 by Rachel McLemore Regan Approved by Terrill R. Hanson, Chair, Associate Professor of Fisheries and Allied Aquacultures James Stoeckel, Assistant Professor of Fisheries and Allied Aquacultures Jesse A. Chappell, Assistant Professor of Fisheries and Allied Aquacultures ii Abstract Floating in-pond raceway (FIPR) systems have the potential to efficiently compete with conventional ponds in their ability to produce large quantities of fish profitably. However, little is known about the biological and economic capabilities of these systems, so this thesis investigates a bio-economic modeling approach to help researchers better understand the system and determine what factors most affect its potential. Findings from model simulations provide insights into what field research areas are most important to conduct. A bio-economic model was developed to simulate the stocking, grow-out and harvest of hybrid striped bass (HSB) in an FIPR. HSB were chosen as a model species because FIPR field trials in West Alabama were already incorporating this species and because their relatively high market value ($7.00/kg and up). System/species feasibility was analyzed in terms of HSB biomass production and net returns through the process of completing four modeling objectives with water temperature dependent fish growth at the core of the developed model. Objectives explored include stocking density and density dependent mortality rate relationships, influence of stochastic loss events, fingerling stocking size and date of stocking relationship to final harvest, and using the bio- economic model to construct longer term production plans that maximize net returns. At this stage, data produced by the bio-economic model simulations is not intended for direct use by producers in making production decisions, but to explore what bio-economic factors most affect the feasibility of the floating in-pond raceway systems and to use this iii information to prioritize where further field research is needed to improve model accuracy and value as a management tool. Eventually, as field trials fill current gaps in knowledge, the refined model will be useful in making on-farm decisions to improve the production efficiency and operation profitability. Results from the four objectives indicate that understanding fish growth rate as it relates to seasonal water temperature and species specific biology is essential to producing an accurate model. Secondly, a better understanding of density dependent mortality is needed to determine stocking densities that optimize system performance and maximize net returns. Third, stochastic mortality loss events were identified as critically important factors affecting economic feasibility and further field research is needed to refine the frequency and magnitude parameters in the model. Finally, analysis using the bio-economic model indicated that overwintering HSB in the FIPR is expensive, especially in the case of holding larger fish. As growth parameters used in the model are refined through field research the model can become a useful tool for producers and researchers in making profit maximizing HSB FIPR system management decisions. iv Acknowledgements The author would like to express her appreciation to Dr. Terry Hanson for his assistance and guidance throughout her studies. The author would like to further thank Dr. James Stoeckel and Dr. Jesse Chappell for their advice and direction towards the fulfillment of her degree. Finally, she would like to her husband Shawn and her parents for their encouragement and support. v Contents Abstract ........................................................................................................................................... ii Acknowledgements ........................................................................................................................ iv List of Tables ................................................................................................................................ vii List of Figures ................................................................................................................................. x List of Equations ........................................................................................................................... xii Introduction ..................................................................................................................................... 1 Literature Review............................................................................................................................ 3 Floating In-Pond Raceway Systems (FIPRs) ............................................................................ 6 Stella? Bio-economic Modeling System ............................................................................... 12 Objectives ..................................................................................................................................... 15 General Methods ........................................................................................................................... 18 Stocking and Stocking Density ............................................................................................... 18 Weekly Growth Rate............................................................................................................... 20 Biomass Production ................................................................................................................ 24 Average Fish Size ................................................................................................................... 25 Fish Lost to Mortality ............................................................................................................. 26 Market Sized Harvest .............................................................................................................. 27 Feed Required ......................................................................................................................... 29 Feed Costs ............................................................................................................................... 29 Fingerling Costs ...................................................................................................................... 30 vi Electricity Costs ...................................................................................................................... 31 Receipts at Harvest ................................................................................................................. 32 Costs and Net Return .............................................................................................................. 33 Results ........................................................................................................................................... 38 Objective 1 .............................................................................................................................. 38 Objective 1 Methods ................................................................................................... 38 Objective 1 Results ..................................................................................................... 42 Objective 1 Discussion ............................................................................................... 45 Objective 2 .............................................................................................................................. 50 Objective 2 Methods ................................................................................................... 50 Objective 2 Results ..................................................................................................... 54 Objective 2 Discussion ............................................................................................... 56 Objective 3 .............................................................................................................................. 62 Objective 3 Methods ................................................................................................... 62 Objective 3 Results ..................................................................................................... 64 Objective 3 Discussion ............................................................................................... 69 Objective 4 .............................................................................................................................. 73 Objective 4 Methods ................................................................................................... 73 Objective 4 Results ..................................................................................................... 75 Objective 4 Discussion ............................................................................................... 77 Summary ....................................................................................................................................... 80 References ..................................................................................................................................... 83 Appendices .................................................................................................................................... 86 vii List of Tables Table 1. Depreciation schedule for HSB FIPR system land, equipment, and machinery investments required for one 3.2 ha (8 acre) pond in West Alabama. .................................... 36 Table 2. Annual fixed costs (interest payments and depreciation). .............................................. 37 Table 3. Bio-economic parameters for Objective 1. ..................................................................... 40 Table 4. Cumulative mortalities for each mortality curve and stocking density. ......................... 41 Table 5. Hybrid striped bass stocking density, total number of fish stocked and total fingerling cost per FIPR raceway system. ............................................................................................... 42 Table 6. Comparison of harvested biomass and net returns for each mortality curve and stocking density. ..................................................................................................................... 44 Table 7. Comparison of costs for each mortality curve. .............................................................. 45 Table 8. Catastrophic loss variables for a 36 week grow out period. ........................................... 53 Table 9. Bio-economic parameters for Objective 2. .................................................................... 53 Table 10. Net returns and standard deviations for 50% magnitude of loss scenarios. ................. 54 Table 11. Other net return values for 50% magnitude of loss. ..................................................... 55 Table 12. Net return for 10% magnitude loss scenarios. .............................................................. 56 Table 13. Other net return averages for grow-out periods with 10% magnitude of loss. ............. 56 Table 14. Bio-economic parameters for Objective 3. .................................................................. 63 Table 15. Size and weight of HSB fingerlings, their prices and cost to stock the FIPR system used in Objective 3 scenarios (includes a constant $950 per order shipping cost). ................ 64 Table 16. Weeks of grow out, net returns and net return per week for varying fingerling length-stocking date combinations. ........................................................................................ 67 viii Table 17. Production weight at harvest (kg) and production quantity per week of grow-out for each fingerling size and month scenario. ................................................................................ 68 Table 18. Bio-economic model parameters for Objective 4. ........................................................ 74 Table 19. Five year production plan from the bio-economic model using the most profitable choice variables for HSB in a FIPR system beginning March 1 Year 1 and ending the last day of Year 5, Feb 28. ............................................................................................................. 75 Table 20. Five year production plan from the bio-economic model using the most profitable choice variables for HSB in a FIPR system FIPR system beginning April 1 Year 1 and ending the last day of Year 5, March 31. ................................................................................ 76 Table 21. Enterprise budget for one 65 m3 FIPR system HSB crop stocked at 300 fish/m3 stocking density in a 3.32 ha pond stocked on February 1 and harvested 36 weeks later. ..... 87 Table 22. Enterprise budget for one 65 m3 FIPR system HSB crop stocked at 400 fish/m3 stocking density in a 3.32 ha pond stocked on February 1 and harvested 36 weeks later. ..... 88 Table 23. Enterprise budget for one 65 m3 FIPR system HSB crop stocked at 500 fish/m3 stocking density in a 3.32 ha pond stocked on February 1 and harvested 36 weeks later. ..... 89 Table 24. Enterprise budget for one 65 m3 FIPR system HSB crop stocked at 600 fish/m3 stocking density in a 3.32 ha pond stocked on February 1 and harvested 36 weeks later. ..... 90 Table 25. Enterprise budget for one 65 m3 FIPR system HSB crop stocked at 700 fish/m3 stocking density in a 3.32 ha pond stocked on February 1 and harvested 36 weeks later. ..... 91 Table 26. Enterprise budget for one FIPR system HSB crop initially stocked at 500 fish/m3 in a 3.32 ha pond stocked on Feb 1 and harvested 36 weeks later. Catastrophic loss event occurring at indicated week with 50% of HSB removed at loss week. .................................. 92 Table 27. Enterprise budget for one FIPR system HSB crop initially stocked at 500 fish/m3 in a 3.32 ha pond stocked on Feb 1 and harvested 36 weeks later. Catastrophic loss event occurring at indicated week with 10% of HSB removed at loss week. .................................. 93 Table 28. Production, receipts, total costs and net return for 50% magnitude of loss scenarios. . 94 Table 29. Production, receipts, total costs and net return for 10% magnitude of loss scenarios. 94 Table 30. Enterprise budget for a five year production strategy stocking only 7? fingerlings in one 65 m3 FIPR system HSB crop stocked at 500 fish/m3 stocking density in a 3.32 ha pond and first stocking on March 1. ....................................................................................... 95 ix Table 31. Enterprise budget for a five year production strategy initially stocking a 3? fingerling (Years 1 and 2) and subsequently stocking a 7? fingerling (Years 3, 4 and 5) in one 65 m3 FIPR system HSB crop stocked at 500 fish/m3 stocking density in a 3.32 ha pond and first stocking on April 1. ......................................................................................... 96 x List of Figures Figure 1. Basic Stella bio-economic model components. ............................................................ 14 Figure 2. Stella stocking components. .......................................................................................... 19 Figure 3. Stocking and restocking the HSB reservoir components. ............................................. 20 Figure 4. Assumed relationship between maximum weight gained per week and fingerling size. ................................................................................................................................... 22 Figure 5. Monthly water temperature and percentage of realized growth. ................................... 22 Figure 6. Relationship between water temperature and maximum (or optimal) growth rate. ...... 23 Figure 7. Growth and temperature adjustment component. .......................................................... 23 Figure 8. Initial and subsequent biomass production................................................................... 25 Figure 9. Average fish size component......................................................................................... 26 Figure 10 Components of mortality and its relationship to stocking density. ............................. 27 Figure 11 Stella harvest of HSB reservoir .................................................................................... 28 Figure 12. Stella biomass lost component. .................................................................................. 28 Figure 13 Stella cumulative feed required .................................................................................... 29 Figure 14. Feed cost component. .................................................................................................. 30 Figure 15. Fingerling costs component. ........................................................................................ 31 Figure 16. Electricity cost component. ......................................................................................... 32 Figure 17. Receipts component..................................................................................................... 33 Figure 18. Stella cost components. .............................................................................................. 34 Figure 19. Stella fixed cost component........................................................................................ 35 xi Figure 20. Stella receipts, total costs and net return components. ............................................... 35 Figure 21. Basic Stella model components including three mortality curve choices. ................. 41 Figure 22. Relationship between cumulative mortality and stocking density under density independent and density dependent scenarios. ................................................................. 42 Figure 23. Relationship between stocking density and final biomass for three mortality curves. ............................................................................................................................... 44 Figure 24. Relationship between stocking density and net returns for three mortality curves. ... 45 Figure 25. Stella catastrophe components ................................................................................... 52 Figure 26. Number of fish in FIPR and Net Return over time, no loss event.1 ........................... 58 Figure 27. Loss event of 50% magnitude occurring 5 weeks after stocking.2 ............................. 59 Figure 28. Loss event of 50% magnitude occurring 19 weeks after stocking.3 .......................... 59 Figure 29. Loss event of 50% magnitude occurring 32 weeks after stocking. ............................. 60 Figure 30. Comparing five fingerling lengths at stocking to growth over time when stocked on February 1 (week 5) and harvested at ? 681 grams. .................................................... 70 Figure 31. Comparing five fingerling lengths at stocking to growth over time when stocked on April 1 (week 513) and harvested at ? 681 grams. ...................................................... 71 xii List of Equations Equation 1: Number stocked ......................................................................................................... 18 Equation 2: Stocking ..................................................................................................................... 19 Equation 3: Restocking ................................................................................................................. 19 Equation 4: Realized Growth. ....................................................................................................... 24 Equation 5: Biomass added. ......................................................................................................... 24 Equation 6: Growth added in grams for average fish size. .......................................................... 25 Equation 7: Feed required ............................................................................................................ 29 Equation 8: Cost of feed .............................................................................................................. 30 Equation 9: Total cost of fingerlings: ........................................................................................... 31 Equation 10: Fingerling cost ......................................................................................................... 31 Equation 11: Fingerling shipping cost .......................................................................................... 31 Equation 12: Electricity used ....................................................................................................... 32 Equation 13. Receipts calculation. ................................................................................................ 33 1 Introduction Over the past 15 years, floating in-pond raceway (FIPR) systems have been developed as complements and even as possible alternatives to traditional pond aquaculture production systems. FIPR systems combine cage and raceway techniques by floating an enclosed raceway inside a traditional aquaculture pond and using injected air pressure to produce a stream flow through the raceway structure. Although a number of research projects and field trials have been conducted to test FIPR systems and their potential as a technique for fish culture, there is still much to be learned about them. In an attempt to better understand the production potential and economic potential of this system, a bio-economic computer model was developed using Stella software to simulate production of hybrid striped bass (HSB) (female white bass (Morone chrysops) X male striped bass (Morone saxatilus)) in a FIPR system. This species was chosen for the bio-economic model development because it had been reared with some success in FIPR trials in West Alabama. By modeling a FIPR system, we hope to gain insight into the biological, physical and economic factors that influence feasibility of FIPR systems for HSB. The current bio-economic model uses data on survival, growth, stocking numbers and sizes from the literature and expert opinion. Adjustable bio-economic model parameters include stocking densities, temperature dependent growth rates, feed and fingerling costs and market prices. The bio-economic model automatically harvests fish when the average size is ?0.68 kg. Profits are calculated based on receipts minus variable and fixed costs (net return) resulting from quantities harvested and expenses incurred during the production process. The bio-economic model tracks cumulative net returns over time. It also tracks fixed costs including construction, 2 equipment and machinery depreciation, loan interest amounts, taxes, insurance, repairs and maintenance. Enterprise budgets were developed from the bio-economic model results and follow standard farm enterprise budget practices so results are comparable to other aquaculture enterprise budgets. Based on limited research with FIPR systems in West Alabama, it appears that HSB in these systems can have high survival rates, better feed conversion ratios and be easier to harvest than in pond aquaculture systems. However, there is little research that quantifies the biological and economic differences between the FIPR and pond production systems for HSB. Building and managing FIPR systems for experimental research purposes is both costly and time consuming. By using a computer simulation bio-economic model, researchers can manipulate a variety of interconnected environmental, biological, and economic variables to predict, with some degree of certainty, the expected total yield, production costs, survivorship, net returns, etc., for any point during the grow out period. The modeling of aquaculture production systems is difficult and not perfect, but with more data from research systems, model parameters can be refined for improved predictive abilities. 3 Literature Review With more Americans consuming fish than ever before, the southeastern United States has the opportunity to provide for the increased demand for aquaculture products. However, southeastern fish farmers are finding it increasingly difficult to keep their farms profitable. Fluctuating and increasingly higher domestic input costs and increasing competition of low priced import substitutes from Asia are keeping profit margins razor thin for many US fish producers (Hanson and Sites, 2011). For southeastern fish farmers to remain competitive, they need to consider alternative production systems that may be more efficient than their present systems. The traditional method of culturing fish in the southeastern United States uses shallow earthen ponds which are relatively inexpensive to construct and straightforward to manage. In west Alabama ponds are built so the surrounding watershed provides water to fill them through runoff, therefore decreasing reliance on a deepwater aquifer that is expensive to use. Once filled, ponds are drained infrequently (every 10 to 15 years). Ponds can self-regulate their water quality through naturally occurring biological processes such as photosynthesis, denitrification and other reactions (Brune et al., 2003b). Ponds can provide supplemental food to developing fish through naturally occurring insect larvae, algae and other nutrient sources. Pond production also has its challenges. Ponds require ample land for the pond to be constructed and the ponds must be strategically placed to take advantage of the topography and retain sufficient water for the ponds. For ponds to be easily harvested, they need to be shallow and rectangular for harvest seines to be most efficiently used. In hilly areas, building proper 4 ponds requires more earth moving which translates to higher construction costs. Ponds also require periodic maintenance and upkeep. Although naturally occurring biological processes can help maintain pond water quality these same processes are much less controllable at high production intensity levels. Ponds need daily monitoring to make sure they are maintaining tolerable ranges of dissolved oxygen, pH, water temperatures to support fish health and growth. Weather events, such as droughts or storms can affect production as well as cause thermal stratification and turnovers in the water. Natural events can make pond production difficult because they are uncontrollable, unpredictable, and increase risk of mortality associated with fluctuating water quality. There are other disadvantages for pond production concerning care for the fish being produced. Fish in a pond are more dispersed and more difficult to see or inventory compared to other types of enclosed systems. Even when feeding a floating feed pellet it can be difficult to identify potential health or disease problems (Brune et al., 2003a; Masser and Lazur 1997). Control over an aquaculture system is the most important managerial aspect and obtaining more control over all biological aspects of a production system often means higher associated costs. A system that is both cost effective and highly controllable is desirable. In an attempt to obtain more complete control over fish production systems, raceway systems were developed and have been much used in the trout industry. Raceway systems maintain higher water quality by flowing clean water through the system continuously, but, in general, very little water is reused or recycled in these systems. These systems are ideal for freshwater fish that require cold and highly oxygenated water such as rainbow trout (Oncorhynchus mykiss) and in areas where clean water is abundant, such as near a river, stream or spring (Lim and Webster, 2006). Raceway systems are advantageous for culturing species 5 that are sensitive to water quality and dissolved oxygen, but have one particular disadvantage; they require flowing water at all times and are therefore not conducive for areas such as the coastal plain where streams are intermittent, especially in the summer months. Nor are they conducive where there are other uses of the water that may contaminate water supply to the facility. Cage culture is another production system type used in many parts of the world. Cages are usually made of rigid mesh boxes that float in the water (Alabama Aquaculture Best Management Practice (BMP); Masser, 1997; Masser and Lazur, 1997). Water and other material can flow passively through the cage but the fish cannot escape. The Southern Regional Aquaculture Center states that the maximum pounds of fish produced in such systems is maximized at 14 lb/ft3 and this quantity is highly influenced by species choice and pond conditions (Masser, 1997). Cages can be built and maintained with relatively low costs and can be fully harvested by extracting and emptying the cage of fish. Fish can be more easily observed during feeding when compared to ponds. But cages have several disadvantages. Because water is not pumped through the cage, stocking densities are limited based on any natural flow in the water, or how quickly fresh water can flow through the cage. Decreased water quality in the form of low dissolved oxygen and high ammonia levels can be a problem for cages that are stocked at high densities or when many multiple cages are placed in the same vicinity. Fish wastes accumulate on the substrate below the cages and degrade the sediments and overall water quality. Recirculating systems require the least amount of water of any aquaculture production system due to their water reuse. However, recirculating systems do not rely on natural components to maintain water quality; instead they rely on equipment such as mechanical filters, 6 water pumps, temperature controls and aerators to maintain water quality (Masser and Lazur, 1997). This equipment requirement increases construction costs compared to ponds and requires higher levels of management skill to keep them in continuous operation at efficient levels. Water quality is usually the most limiting factor in these systems (Lazur and Britt, 1997). Fish in recirculating systems are highly visible and can be more easily monitored. Disease recognition and treatment is more easily managed in recirculating systems resulting in higher production rates. Because recirculating systems use less water than other systems to grow the same amount of fish and can be built above ground, they usually require less land for construction and can be constructed closer to markets. Each type of fish production system has its advantages and disadvantages. Traditional pond production systems are relatively inexpensive to build but costly and time-consuming to completely harvest. Discovering and treating diseases and accurately determining feeding rates is difficult in ponds when compared to other systems. Raceway systems combat these issues because fish are held in relatively small enclosure areas, but their continuous need for water is so great that they are often eliminated as a viable option for warm water aquaculture (Masser and Lazur, 1997). Cage culture may provide easier waste collection, more effective disease recognition and treatment, more efficient feeding, and in some cases, easier harvesting, but is limited by stocking densities because of decreased water quality. Floating In-Pond Raceway Systems (FIPRs) In the early 1990?s a culture system was developed that combined the advantages of pond, raceway and cage systems into a caged raceway structure located in a pond. Early systems were developed by researchers at Auburn University, and it was called the floating in-pond raceway or FIPR (Masser and Lazur, 1997; Yoo, Masser and Hawcroft, 1995). The pond 7 provides a constant source of flowing oxygenated water via airlift pumps through a raceway with a mesh barrier on the end, creating a raceway like enclosure (Masser and Lazur, 1997). Fish contained inside needed to be able to withstand a slow current of water in order to survive and thrive. FIPR systems incorporate airlift pumps for both aeration and water flow. Another important part of FIPR construction and a major advantage is a surface cover to protect fish from bird predation and to provide shade (Chappell, 2009; Hanson, 2009; Masser and Lazur, 1997). FIPR systems utilize the benefits of raceway, cage and pond production systems. They allow for stocking densities and production levels similar to raceways but require much less water, since the outflow of FIPR water is into the pond and is re-circulated through this water body and is eventually re-flowed through the FIPR. Some FIPR systems also incorporate waste removal at the end of the raceway, where traditional cages and raceways usually have no waste collection (Chappell, 2009; Brown, 2010). When the water leaves the FIPR structure it goes back into the pond where the natural de-nitrification process can occur and water quality can be improved and maintained naturally, without the added costs of filtration. There are other managerial benefits to FIPR systems. Since all the fish are in an enclosed space, feeding to satiation is much more feasible and fish are more likely to have access to feed compared to pond systems where fish are more dispersed (Chappell, 2009). Early disease recognition and treatment is also easier since fish are more visible and contained (Masser and Lazur, 1997; Yoo, Masser and Hawcroft, 1995). Studies show that FIPR systems have better water quality than the surrounding pond?s surface water in terms of dissolved oxygen, temperature and consistency (Hartleb, 2004; Masser and Lazur, 1997; Morrison et al., 1995). Hartleb (2004) found that water in the FIPR system was consistently a few degrees lower than in the supporting pond?s surface water. Water 8 temperature is especially important in the southeastern U.S. where fish are subject to extremely high water temperatures nearing 30?C in summer. The airlift pumps provide oxygenated water at all times, supplying water that is higher in oxygen than the water before it was pumped through the FIPR system (Masser and Lazur, 1997). Higher dissolved oxygen levels and more consistent water temperatures that are optimal for specific fish species can result in more efficiency, higher production and higher net returns. Another benefit of the FIPR system is that they can be used in ponds that would be otherwise unsuitable for fish production and harvest. Where production ponds must be shallow so they might be seined for partial harvesting, FIPR systems can be placed in deeper ponds and harvesting can be done with a seine inside the system structure or with a crane that removes the whole unit (D'Abramo and Frinsko, 2008). Deeper ponds contain lower water temperatures in their depths and may be preferred for growing some species of fish such as hybrid striped bass (D'Abramo and Frinsko, 2008; Hodson, 1989; Volkman, Kohler and Kohler, 2004). Another advantage of FIPR systems is that farmers can diversify their species produced in a single pond by putting different species in each raceway, or stocking each raceway at different times to diversify cash flows (Chappell, 2009; Odom, 2009). Beside the biological benefits, there are numerous economic benefits as well. After the first season of Brown?s research in a fixed in-pond raceway system they were able to harvest nearly 19,000 pounds of fish per acre (Brown, Chappell and Hanson 2009; Brown, 2010). It is important to adjust stocking rates according to the size of the pond supporting the FIPR and Masser (1994) recommended stocking no more than 14,800 fish/ha of pond (Masser and Lazur, 1997). 9 There are drawbacks to the FIPR system as well. The initial costs of building a FIPR system can be high depending on the materials used; and fixed FIPR system can be very expensive if built with concrete, but many less expensive materials can also be used that decrease this initial cost. Also, fish stocked very densely combined with any problem with the airlift pump system could be catastrophic for the farmer and should be addressed beforehand by including back-up pumps and emergency power generation as part of the standard FIPR equipment and machinery needs of system. The time window to prevent a great loss is much shorter in more intensive production systems than in less intensive pond systems and risk mitigation strategies are critical. A similar type of in-pond raceway system, called the fixed in-pond raceway system, functions on the same principals as the floating raceways except it is permanently built into the bottom of the pond with concrete or plastic walls. There is a considerably higher construction cost associated with building a fixed IPR compared to a FIPR system. The air lift pumps and resulting water flows for the fixed IPR systems are powerful enough to circulate water through the entire pond. In a west Alabama fixed IPR research project a pond was drained and six raceways were built on a concrete pad with cement block walls. There was a baffle to direct water down the length of the six acre pond around the length of the bend, and then back to the paddlewheels at the front opening of the fixed IPR system (Brown, Chappell & Hanson 2009). The initial trial of this fixed IPR system showed promise, producing a high weight of catfish and two other co-culture species, tilapia and paddlefish. Another species that shows potential for culture in the FIPR is the hybrid striped bass (HSB) or sunshine bass (Fullner, Gottschalk and Pfeifer, 2007; Hanson, 2009; Hodson and Hayes, 1989). HSB are produced by fertilizing eggs from white bass, Morone chrysops, with 10 sperm of striped bass, M. saxatilis. The white bass is a freshwater species found in the Mississippi basin and other gulf coast states and can reach a weight of 1.81- 2.27 kg. The striped bass is an anadromous species occurring along the Atlantic and Gulf Coasts of the U.S. but have been successfully introduced into lakes and reservoirs throughout the U.S. and along the Pacific coast and can reach up to 31.75 kg. Since striped bass spawn in fresh water, some introduced populations have been able to reproduce in landlocked bodies of water despite being isolated from the ocean. Aquaculture of sunshine bass has been increasing over the past few years because of a general decline of the striped bass fishing industry and because of the hybrid?s vigorous growth and its success in lakes and reservoirs as a game fish (Hodson and Hayes, 1989). HSB have a wide tolerance of environmental conditions. Their optimal growth temperatures are between 25 and 27?C but can survive temperatures between 4 and 33?C. They can also survive in salinities up to 25 ppt. In production ponds, HSB usually reach a harvest weight of 0.680 to 1.13 kg in 18 to 24 months. Traditionally, hybrid striped bass were produced using a three-phase system. In Phase I or the nursery phase, eggs are hatched and raised to 30-45 day old fingerlings where each fingerling weighs approximately 1 gram. These fingerlings are then removed and stocked for Phase II in ponds from 30,000 to 44,000 per hectare. When these animals reach 125 to 225 grams they are graded and restocked into the final Phase III grow-out ponds at densities between 7,000 and 11,000 per hectare and grown to market size which is usually around 900 grams (D'Abramo, L. R., et. al., 2002; D'Abramo and Frinsko, 2008; Hodson and Hayes, 1989). Based on a survival of 80%, farmers can expect yields between 7,200 and 7,800 kg/ha. Research in the last 10 years has led to a more streamlined method of HSB production called ?direct stock? (D'Abramo and Frinsko, 2008; D'Abramo et al., 2002). The direct stock 11 method eliminates the Phase II component of the three phase system and stocks larger, graded fingerlings directly into the grow-out pond. D?Abramo?s protocol states that fingerlings should be stocked after they reach 3 grams in the initial phase. The larger and more uniform the fingerlings are when they are stocked led to larger and more uniform market sized fish (D?Abramo, et. al, 2002; Chappell, 2009). One disadvantage of HSB is that they are cannibalistic, especially with larger fish cannibalizing smaller fingerlings. Therefore, it is important to effectively size grade the fish before stocking them for grow-out so that they are similarly sized and do not cannibalize each other. Cannibalism can also be prevented by feeding fish to satiation. Another disadvantage of farming HSB compared to other warm water fish, such as channel catfish, in southeastern states is that they are less tolerant of high water temperatures and require higher dissolved oxygen levels. Water flowing through the FIPR is oxygenated with airlift pumps and uses water from approximately one meter below the water surface, so it is both cooler in temperature and higher in oxygen than the surface pond water. This can provide a favorable environment for HSB even during the hot summer months of the year. Alternatives to growing HSB in FIPR systems until harvest size are the possibility of stocker production in the FIPR system until they are past their most vulnerable stage and then grown out in ponds, or even stocked in reservoirs for recreational fishing. A study in Europe described HSB as an excellent fish for culture in pond circulation systems, saying they over wintered with low losses, and withstood handling pressures well (Fullner, Gottschalk and Pfeifer, 2007). HSB could also be raised from fingerling to market size in a FIPR system. 12 Stella? Bio-economic Modeling System Field research of a FIPR system for a specific species can be costly and time consuming. However, an alternative modeling method can be used to simulate preliminary growth studies and economic analyses. Models can help users understand complex biological systems and enable the user to assess relationships between variables in the system. They also enable the user to optimize the output or flow through the modeled system. Costanza (1998) reported that a group of ?novice? modelers were able to build their own models in a short period of time and able to answer a wide range of ecological and economic questions about the systems they had been studying (Costanza and Gottlieb, 1998). Modeling can also be used as a tool to direct research efforts in the field toward more relevant research areas, saving time and reducing costs of multiple field trials. The program chosen for this project is called Stella?, made by ISEE Systems. Stella is a dynamic modeling program that uses an icon-based graphic interface to build models. HSB and channel catfish have been raised in FIPR system trials in West Alabama, providing some baseline data for bio-economic model construction. The resulting Stella bio-economic model will help identify parameters needing additional research as well as determine which factors are most influential on FIPR feasibility. This bio-economic modeling template could be used as a beginning point for modeling the production of other fish species. Baseline parameters and bio-economic model assumptions were derived from the scientific literature, and expert opinions of researchers and producers. The bio-economic model tracked the growth of HSB from initial stocking size to harvest and removed animals from the system at a mortality rate typical of HSB systems in West Alabama. Using the programmed parameters and assumptions, the bio-economic model calculated variable costs and fixed costs. 13 It also calculated receipts and net returns based on sales of harvested HSB minus production and fixed costs. The bio-economic model calculated results as often as the user decided; i.e. provide results for any time period desired. The bio-economic model could be set to calculate a certain number of time periods, and these sets of data were called ?runs?. For example, the bio- economic model can be set to run for a year with the time step equaling one day, the total ?run? would equal 365 days. Stella will then track the changes in a tabular form and in a graphical form which allows the user to see the results immediately. The bio-economic model contains variable parameters that can be adjusted by the user. This allows the user to make adjustments to the bio-economic model?s specifications and immediately see the resulting implications. For example, the feed price is one adjustable factor. The price of feed for HSB in the FIPR system is $1.00/kg for 48% protein feed. The price per kg of feed influences the total cost of feed and therefore net returns. The user can see and record these dependent variable values. Then the user may leave every parameter the same and adjust the feed price to another level, say $2.00/kg, and run the bio-economic model again. Now the costs will increase and profits will decrease over time and the user can see the magnitude of the change instantly. Like the Microsoft Excel spreadsheet software, the user develops the equations and relationships between components of the bio-economic model. Unlike Excel, Stella can calculate the equations repetitively to simulate and track changes over time. Stella also uses graphics to illustrate different types of functions, and each graphic contains a number, equation or graph depending on what the user programs it to have. As the bio-economic model is built, a map is formed of the functions and their relationships. Below is an example of the basic Stella graphic types: 14 Figure 1. Basic Stella bio-economic model components. The Converter can be a number or equation and can be connected to make different relationships, much like a formula in an Excel software cell. Arrows connects the Converter to the Flow Pipe. The Flow Pipe is used to insert numbers into a Reservoir. Stella recalculates the equations in the Converter with each time step, which can be daily, weekly, monthly, etc., and recalculates for the duration of the bio-economic model run, which can be any chosen length of time. The Reservoir collects and stores the numbers over the entire length of the bio-economic model, where the other Stella elements simply recalculate. Red arrows point from a converter or reservoir to another converter or flow pipe, and they indicate that the equation in the origin is used in the graphic the arrow is pointing to. A Ghost is used to copy any of the tools and paste them somewhere else in the system, mostly to keep the map more organized. They always have dotted outlines and indicate that only that part of the bio-economic model and the formula in it is copied. More specific Stella bio-economic model elements will be provided and discussed in the methods section. 15 Objectives The overall goal of this study was to develop a bio-economic computer model depicting HSB production in floating in-pond raceways. The bio-economic model includes biological relationships between the culture environment, HSB characteristics and the economic setting to measure outcomes such as production, receipts, costs and net returns. Biological relationships of the culture environment include water temperature and its effect on HSB growth; species characteristics include not only the temperature dependent growth rates, but also the mortality associated with stocking densities and crop duration. The economic setting includes uncertainty, such as the risk of unexpected death events, input/output prices and initial stocking weight-date effects on harvest quantity and net returns. Development of a bio-economic model with these relationships will be useful as a research and decision making tool for researchers and producers, and can be adapted for other species. Data from the literature, a FIPR system field trial in west Alabama and expert opinion was used in the computer software STELLA? to model biological production, costs, receipts and net returns for HSB in a FIPR system. The objectives and hypotheses set for the bio- economic model development were to: I. Determine the importance of knowing the shape of a density dependent mortality curve. It was hypothesized that that the shape of the density dependent mortality curve would significantly affect production and net returns. Secondly, maximized production would not always correspond to maximized net returns. Specifically, this bio-economic modeling objective was to determine the effect of the relationship between mortality and 16 stocking density on production and net returns by varying the HSB stocking density in the FIPR system at set intervals of 300, 400, 500, 600 and 700 fish/ m3 using fixed, linear and exponential mortality relationships on fish biomass production and net returns above all costs. II. Evaluate the effects of stochastic mortality events, on production and net returns. It was hypothesized that decreasing the magnitude of stochastic loss events from 50% to 10% would greatly improve production and net return and increase the chance of positive net return. Additionally, HSB FIPR crops with loss events occurring later in the grow-out period should experience lower net returns due to higher accumulated variable costs, than crops with loss events occurring earlier in the grow-out period. This bio-economic modeling objective sought to assess the economic consequences of stochastic loss events in FIPR systems, and determine the economic benefits realized from decreasing both the magnitude and frequency of stochastic loss events. This objective developed a component of the bio-economic model to generate stochastic loss events with varying frequencies of loss (1%, 10%, 20% and 40% through any given grow-out period) and two magnitudes (50% and 10% loss of current crop abundance). Effects were measured in terms of production biomass and net returns. III. Examine the influence of initial stocking size and initial stocking date on crop duration, production and net returns in the HSB FIPR system. Growth rate is dependent on seasonal water temperatures typical of the southeastern United States. Since water temperature directly affects growth rates, the choice of fingerling size stocked at a particular time can be an important aspect in determining the profitability of HSB FIPR systems. It was hypothesized that larger fingerlings (7?) stocked any month of the year 17 should reach market size in less time than smaller fingerlings (6?, 5?, 4? or 3?) stocked in the same month and resulting production and net returns should be greater than for smaller fingerlings. Secondly, stocking larger fingerlings early in the year (January through April) would result in greater crop production and net returns because they would reach market size before the winter. IV. Develop a method to analyze long-term HSB FIPR production planning to maximize net return potential from the bio-economic model. It was hypothesized that the largest fingerling size available (7?) would be the best choice to stock in any month of the year in order to maximize net return over an extended multi-year, multi-crop production scenario (five years). 18 General Methods The bio-economic model was designed to represent a grow-out facility for HSB after being stocked with fingerlings purchased from an outside source. Mortality, biomass, crop value, feed costs, etc. were all calculated in weekly time steps. The time steps for the bio- economic model were weeks and therefore, results were produced on a weekly basis. Stocking and Stocking Density Stocking density can be adjusted for any given bio-economic model objective by using a slider bar. The range of values for Stocking Density is from 0.5 to 700 fish/m3 based on the scientific literature search and expert opinion. The Number Stocked converter multiplies the number chosen on the stocking slider bar with the number of raceways used (one in all cases in this thesis) and multiplies by the volume of each raceway (65 m3 in all cases in this thesis) (Figure 2 and Equation 1). This allows the user to choose the approximate size of the farm by choosing the appropriate number of raceways. The FIPR volume is set at 65 m3 based on the volume of prototype raceways in West Alabama. Volume of Raceway in ft3 converts metric to English units, thus the bio-economic model can be readily understood by producers. Equation 1: Number stocked Number stocked = (Volume of Raceway in m3)*(Stocking Density per m3)* (Number of Raceways) 19 Figure 2. Stella stocking components. The Number Stocked converter controls the inflow of animals to the grow-out reservoir. The week of first stocking is set by Stocking Week which can be adjusted by a slider bar. When time in the bio-economic model equals Stocking Week then Number Stocked is added to the FIPR reservoir (Equation 2 and Figure 3). When Harvest occurs and the FIPR Reservoir empties, restocking occurs after a set Restocking Delay period passes (which is controlled by the bio- economic model user). The Restocking Delay simulates the # weeks required for a producer to obtain and restock fingerlings (Equation 3). Equation 2: Stocking If TIME = Stocking Week ? 1 OR Stocking Week=0 AND TIME = 0 THEN Number stocked ELSE 0 Equation 3: Restocking IF TIME < or = Restocking Delay THEN 0 ELSE IF Time= Restocking Delay + Restock Time THEN Number Stocked ELSE 0 20 Figure 3. Stocking and restocking the HSB reservoir components. Weekly Growth Rate Weekly growth is the most important component in the Stella model because it directly or indirectly influences all of the other model components. There are three main model converters that control the HSB growth and they are: Growth Added to Previous Size, Average Weekly Water Temperature and Temperature Growth Adjustment. Although growth rates vary among weeks, depending on water temperature and fish size, all fish within a crop grow at the same rate during any given week. The Temperature Growth Adjustment can be switched off by the user with the Temperature Growth Switch (Equation 4). The Growth Added to Previous Size is a graphical converter that sets the maximum amount of biomass (grams) that can be added to each HSB in the current week based on a converter, Previous Size Grams, that was the size of the HSB in the FIPR from the previous week (Figure 4). Essentially, the Growth Added to Previous Size converter is the maximum amount of weekly growth that is expected to occur under optimal conditions for any given sized animal. Optimal water temperature for HSB growth has been determined to be between 23 and 25?C (Hodson, 1989), with growth rates decreasing as temperatures exceed or fall below the 21 optimal temperature range. The Average Weekly Water Temperature converter defines the model?s weekly water temperatures based on average weekly water temperatures from records for a fixed in-pond raceway system experiment in West Alabama (Figure 5) (Brown, 2010). . Since fish growth is greatly influenced by water temperature, the Temperature Growth Adjustment converter designates the percentage of realized optimal growth within a temperature range of 9 to 35? C, (Figure 5). When water temperature is optimal, between 23-25? C for HSB, the Temperature Growth Adjustment converter produces 1, or 100% of optimal growth. At suboptimal temperatures, the converter produces a value less than one which designates the percentage of maximum growth (Growth Added to Previous Size) that will occur at that temperature (Figure 5 and Figure 7). The Actual Growth Kg converter then uses the Temperature Growth Adjustment to determine the actual growth to add to each HSB, For example, water temperatures in June exceed the optimal temperature range for growth. Therefore, if a fish weighs 200 g and water temperature is 30 C, the Actual Growth will be Growth Added to Previous Size (18.5 grams) * 85% (Temperature Growth Adjustment) / 1000 (conversion from grams to kg) = 0.016 kg (Figure 4, Figure 5 and Figure 6). The data produced by Actual Growth is fed into to the Biomass added reservoir flow pipe. With each weekly time step the Biomass Added flow pipe multiplies Actual Growth by Number of HSB in FIPR to calculate the weekly increase in biomass. The Biomass in kg reservoir also has an initial function to account for the initial stocking weight of the fish (Equation 4). 22 Figure 4. Assumed relationship between maximum weight gained per week and fingerling size. Figure 5. Monthly water temperature and percentage of realized growth. 0 10 20 30 40 50 60 0 100 200 300 400 500 600 700 W e ight G a ine d pe r w e e k ( g) S i z e of f i nge r l i ng ( g) 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 0 5 10 15 20 25 30 35 J F M A M J J A S O N D P e r c e nt of O pt im a l G r ow t h R e a li z e d W a t e r T e m pe r a t ur e ( C ) M ont h of Y e a r W a t e r Te m p e r a t u r e R e a l i z e d G r o w t h 23 Figure 6. Relationship between water temperature and maximum (or optimal) growth rate. Figure 7. Growth and temperature adjustment component. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 5 10 15 20 25 30 35 O pt im a l G r ow t h R e a li z e d W a t e r T e m pe r a t ur e ( C ) 24 Equation 4: Realized Growth. IF Batch Biomass in kg > 0 AND Temperature growth switch = 1 THEN Growth added to previous size * Temperature Growth Adjustment/1000 ELSE IF Batch Biomass in kg > 0 THEN Growth added to previous size/1000 ELSE 0 Biomass Production The biomass components of the Stella bio-economic model determine the production of the FIPR system. The Initial Biomass flow pipe delivers biomass into the Batch Biomass in Kg reservoir. When Stocking or Restocking occurs the flow pipe calculates the amount of biomass to insert based on the number of fingerlings stocked and the initial weight of the fingerlings. The user control page has a slider bar where the Initial Stocking Weight in Grams can be selected. The initial week?s growth is determined by the First week of growth converter. After the first week of growth, biomass is continually added based on the growth from the Realized Growth Kg converter through the Biomass Added flow pipe (Equation 5 and Table 7). Equation 5: Biomass added. IF Number of HSB in FIPR > 0 THEN Actual Growth * Number of HSB in FIPR ELSE 0 25 Figure 8. Initial and subsequent biomass production. Average Fish Size The Average Fish Size g reservoir accumulates the average realized growth per fish (Equation 6). When a harvest occurs, the Average Size Dumping out flow pipe removes the data stored in the reservoir and the process begins again with the next stocking (Figure 9). Equation 6: Growth added in grams for average fish size. IF Initial Biomass > 0 THEN Initial Stocking Wt grams ELSE IF Harvesting > 0 THEN 0 ELSE Realized Growth kg *1000 26 Figure 9. Average fish size component. Fish Lost to Mortality The mortality rate for the FIPR system is controlled by the Mortality Rate converter. Mortality Rate contains a graphical function related to Stocking Rate. The Total Deaths converter multiplies the Number of Animals Stocked by the Mortality Rate to calculate the cumulative number of fish that will die and be removed from the system over the course of the grow-out period. Since the literature gives little indication of when HSB mortality is more likely to occur, the bio-economic model distributes mortality evenly across the grow-out period by taking the total deaths and dividing them by the number of weeks in the grow-out of a batch. The user must assign a value for the number of weeks of grow-out using the Average Growing Season slider before running the bio-economic model. For example, if 1000 fish are stocked, the grow-out period is 50 weeks long and the mortality rate is 10%, then two fish will be lost per week for a total of 100 fish lost, or 10% of 1000 fish stocked. The Deaths flow pipe then removes that number of fish every week from the Number of HSB in the FIPR system (Figure 10). 27 Figure 10. Components of mortality and its relationship to stocking density. Market Sized Harvest The Stella bio-economic model automatically harvests the crop from the HSB reservoir when the Average Fish Size reaches or exceeds a user selected target weight (0.681 kg) per fish (Figure 11). When the Average Fish Size equals or exceeds the target weight, fish are moved from the Number of HSB in FIPR reservoir to the Harvest reservoir. When a harvest occurs, the Biomass reservoir drops to zero to show a harvest event. A similar mechanism is used to show a harvest in the Batch Biomass reservoir, when the Average Fish Size reaches or exceeds the selected target weight, then the Batch Reservoir is emptied by the Biomass Lost flow pipe (Figure 12). 28 Figure 11. Stella harvest of HSB reservoir. Figure 12. Stella biomass lost component. 29 Feed Required The amount of feed required in a production cycle is calculated based on a daily percentage of body weight, 1.5% per day, and the converter Temperature Growth Adjustment which varies the amount of required feed depending on the water temperature in the same manner as before (Figure 13 and Equation 7). Used feed accumulates in a reservoir over the course of the grow-out period. Feed Required is only calculated when there are fish in the reservoir and not during the restocking delay period (Equation 7). Equation 7: Feed required Feed Required = IF Number of HSB in FIPR = 0 THEN 0 ELSE IF time = 0 THEN Initial Stocking Wt grams/1000 ELSE Biomass in kg/Number of HSB in FIPR Figure 13 Stella cumulative feed required Feed Costs Feed costs are calculated for each individual crop and accumulated for multiple crops over time. The Feed Price per Kg is selected manually in the interface menu (Equation 8). The Cost of Feed converter functions as the equation for the Weekly Feed in-flow pipe (Equation 13). Batch Feed Costs are not calculated during the restocking delay period (Figure 14). 30 Equation 8: Cost of feed Cost of Feed = Feed Price per Kg * Feed Required. Figure 14. Feed cost component. Fingerling Costs The price for each fingerling is automatically adjusted for size categories as the user selects the fingerling size to stock. Each fingerling size in inches corresponds to weight in grams which corresponds to a specific fingerling price. There is a Crop Fingerling Costs reservoir to track the cost of fingerlings for each stocking event and a cumulative reservoir, CF Costs, to accumulate all fingerling costs across multiple crops (Figure 15). Fingerling costs are calculated only at stocking or restocking. Fingerlings are purchased from off-farm commercial hatcheries and therefore there is a fixed shipping cost that is also calculated in this section and added into the total costs of production. 31 Equation 9: Total cost of fingerlings: Total Cost of Fingerlings = If Number stocked = Stocking OR Number stocked = Restocking THEN Total Cost of Fingerlings ELSE 0 Equation 10: Fingerling cost Fingerling Cost = If time = 0 then Total Cost of Fingerlings else 0 Equation 11: Fingerling shipping cost Total Shipping Cost = Mileage from Keo AR * Costs per mile $ Figure 15. Fingerling costs component. Electricity Costs Electricity costs are estimated based on the number of raceways, the size of the pump and the price of electricity per kilowatt hour. Each raceway uses a 1.5 horsepower airlift pump which runs 24 hours a day. By estimating the number of kilowatt hours the airlift pump uses per week, the electricity usage can be calculated (Equation 12). Electricity costs per kilowatt hour (kWh) can vary, so that parameter is adjustable with a slider bar. Electricity costs accumulate and are stored a reservoir (Figure 16). Electricity costs are only calculated during grow-out and 32 not during the restocking delay period. Both batch and cumulative electricity costs are calculated. Equation 12: Electricity used Electricity Used = Number of raceways * Cost of Electricity per kWh* 188 kWh. 1 HP = .746 kilowatts 1.5 HP Motor * .746 kilowatts * 24 hrs. a day * 7 days = 188 kWh/week per motor Figure 16. Electricity cost component. Receipts at Harvest Receipts are the funds received from the sales of the HSB harvested. When the HSB reach market size the Receipts converter multiplies the kilograms of biomass in the biomass reservoir to the predetermined price per kg of fish (Figure 17 and Equation 13). Because of the constraints of the Receipts converter, the crop only has worth when the HSB reach a size of ? 0.680 kg (1.5 lb). 33 Equation 13. Receipts calculation. IF Average fish size kg > .680 OR Average fish size kg = .680 THEN Batch Biomass in kg * Price per kg of market fish ELSE 0. Figure 17. Receipts component. Costs and Net Return Three specified variable costs are calculated by the bio-economic model based on the assumptions made for each run: feed, fingerlings and electricity, as already discussed. These costs are stored together in the reservoir called, Batch Specified Costs and accumulated over time in Cumulative Specified Costs. Additional variable costs for elements not specifically described in the model, such as chemicals, fuel, labor, maintenance, etc., are calculated based on economic analyses of hybrid striped bass production (D?Abramo et al., 2002 and D?Abramo and Frinkso, 2008). Their studies indicated that the additional variable costs were approximately 43% of the total variable costs. The converter, Other Variable Costs multiplies the accumulated Batch Specified Costs by 3/7. Total Variable Costs sums the Batch Specified Costs and the Other Variable Costs. 34 Figure 18. Stella cost components. Fixed costs are typically accessed on a yearly basis, but since some HSB FIPR crops require more than a year to reach harvest size and since the bio-economic model calculates on a weekly time step, the fixed costs are charged weekly at a rate of $81 per month. See the detailed depreciation schedule (Table 1 and Table 2). A reservoir is used, Batch Fixed Cost, to accumulate fixed cost over the grow-out period. When harvest occurs, Batch Fixed Cost empties and the process begins again (Figure 19). 35 Figure 19. Stella fixed cost component. Total costs are calculated in the Total Costs converter where Batch Specified Costs and Batch Fixed Costs are added together. The Net Return component is a converter which takes the value produced from Receipts and subtracts Total Costs to calculate profits made above all costs (Figure 20). Figure 20. Stella receipts, total costs and net return components. 36 Table 1. Depreciation schedule for HSB FIPR system land, equipment, and machinery investments required for one 3.2 ha (8 acre) pond in West Alabama. R e p a i r s A n n u a l P e r c e n t U s e f u l a s a % o f r e p a i r s & U s e i n Li f e , S a l v a g e A n n u a l I n te r e s t o n n e w c o s t m a i n te n a n c e I te m U n i t C o s t/ u n i t N u m b e r F I P R C o s t Y e a r s V a l u e D e p r e c i a ti o n I n v e s tm e n t (%) ($ ) A . C a p i ta l c o s t L a n d ha 750 3 . 9 NA P o n d c o n s t ru c t i o n ha 2 , 4 7 1 3 . 2 100% 8 , 0 0 0 15 1 , 0 0 0 467 450 10 53 G ra v e l 667 1 100% 667 0 33 W e l l (4 5 0 g p m , 2 5 -h p e l e c t ri c m o t o r) 2 5 , 0 0 0 1 20% 5 , 0 0 0 20 100 245 255 25 63 O F F ICE BU IL D IN G 2 0 ' x 4 0 ' s q ft 40 800 5% 1 , 6 0 0 20 50 78 83 10 8 S h o p ea 3 0 , 0 0 0 1 5% 1 , 5 0 0 20 250 63 88 10 8 S h o p t o o l s a n d e q u i p m e n t ea 5 , 0 0 0 1 5% 250 10 25 23 14 10 3 S u b to ta l (e x c l u d i n g l a n d ) 1 7 , 0 1 7 1 , 4 2 5 874 922 134 B . Eq u i p m e n t F l o a t i n g In -P o n d Ra c e w a y (i n c l . A i rl i ft p u m p ) ea 2 0 , 0 0 0 1 100% 2 0 , 0 0 0 10 2 , 5 0 0 1 , 7 5 0 1 , 1 2 5 10 200 H a rv e s t N e t ea 500 1 100% 500 5 0 100 25 10 10 D o c k , fl o a t i n g ea 800 1 100% 800 8 0 100 40 5 5 T ru c k s , 3 / 4 t o n , u s e d ea 1 6 , 0 0 0 1 10% 1 , 6 0 0 6 250 225 93 6 16 T ra c t o rs , 4 5 -6 5 h p , u s e d ea 2 0 , 0 0 0 1 10% 2 , 0 0 0 6 250 292 113 10 33 Bo o m a t t a c h m e n t fo r h a rv e s t i n g ea 1 0 , 0 0 0 1 10% 1 , 0 0 0 7 5 142 50 11 16 P T O a e ra t o rs , e m e rg e n c y a e ra t i o n ea 3 , 5 0 0 1 17% 583 10 25 56 30 10 6 E l e c t ri c a e ra t o rs , 1 0 h p ea 3 , 5 0 0 1 100% 3 , 5 0 0 5 500 600 200 10 70 Bu s h h o g / m o w e r ea 4 , 5 0 0 1 5% 225 5 75 30 15 20 9 F e e d b i n n o t u s e d b e c a u s e b a g s o n l y ea 5 , 0 0 0 1 0% 0 20 0 0 0 10 0 D O m e t e r a n d a c c e s s o ri e s ea 2 , 0 0 0 1 10% 200 10 0 20 10 50 10 Co m p u t e r ea 1 , 5 0 0 1 5% 75 5 0 15 4 10 2 S u b to ta l 3 0 , 4 8 3 3 , 6 0 5 3 , 3 3 0 1 , 7 0 4 376 TO TA L 4 7 , 5 0 0 5 , 0 3 0 4 , 2 0 4 2 , 6 2 7 510 37 Table 2. Annual fixed costs (interest payments and depreciation). U s e f u l S a l v a g e Li f e V a l u e L o a n Re p a y m e n t fo r Ca p i t a l It e m s C o s t (Y e a r s ) ($ ) 2 2 , 9 5 9 20 1 , 1 4 8 1 7 , 0 1 7 20 851 A v e ra g e a n n u a l P & I = 1 , 9 9 9 L o a n Re p a y m e n t fo r E q u i p m e n t Cu m u l a t i v e i n t e re s t = 1 3 , 3 4 7 7 1 , 9 0 7 Cu m u l a t i v e p ri n c i p a l p y m t = 3 0 , 4 8 3 7 4 , 3 5 5 6 , 2 6 1 A n n u a l In t e re s t a n d D e p re c i a t i o n D e p r e c i a ti o n 3 , 0 5 5 fo r 3 0 0 s u rfa c e a c re s 4 , 2 0 4 p e r a c re 7 , 2 5 8 p e r h a 24 p e r 4 . 0 4 7 h a (1 0 a c re s ) 60 242 D e p re c i a t i o n + In t e re s t fo r Y e a r 1 = Cu m u l a t i v e i n t e re s t = Cu m u l a t i v e p ri n c i p a l p m t = T o t a l 1 s t y r P & I = A n n u a l In t e re s t p a y m e n t = D e p re c i a t i o n fo r Y e a r 1 = 38 Results The results portion is separated into four sections, each pertaining to an individual objective. Each section includes the specific methods, results and a discussion pertaining to the particular objective. A summary of all four objectives follows the results portion. Objective 1 Determine the importance of knowing the shape of a density dependent mortality curve relating cumulative mortality to initial stocking density. Hypotheses: 1) Production and net return patterns will differ between the fixed, linear and exponential mortality curves even when cumulative mortality at the lowest and highest stocking densities is held constant. 2) Maximum net returns will not always correspond to maximum production. Objective 1 Methods A bio-economic model was developed to assess the relationship between stocking density, final production, and net returns. From the literature, the expected cumulative mortality for pond systems producing HSB ranges from 0-25%, i.e. 0-25% of fish stocked will be lost to one of many potential perils, including disease, water quality, cannibalism, bird predation, etc. during a crop?s production cycle (D?Abramo et. al., 1994; Kelly and Kohler, 1996; Kemeh and Brown, 2001). It was assumed that FIPR systems would have a relatively low cumulative 39 mortality rate due to increased control over water quality, fish inventory and feeding compared to pond systems. Therefore, the baseline cumulative mortality was set at 10% (i.e. 10% of stocked HSB die over the course of a production run). At some point, an increase in the stocking density of a FIPR system will increase the cumulative mortality above 10% due to crowding and carrying capacity limitation of the raceway environment. However, the relationship between cumulative mortality and stocking density has not been empirically determined for FIPRs. This objective uses the Stella program to biologically and economically evaluate whether changing the shape of the density dependent mortality curve results in significant differences in production and net receipts. One density independent (fixed) and two density dependent (linear, exponential) mortality curve components were added to the bio-economic model and used to simulate the relationship between cumulative mortality and stocking density (Figure 21). Each mortality curve component in the bio-economic model can be turned on (and the others turned off) with a special switch function located on the bio-economic model user interface page. Each switch causes the specified percentage of stocked fish to be removed (i.e. die) from the HSB reservoir over the course of a grow-out period based on the type of mortality curve switch that was switched on. The density independent curve was held constant at 10% cumulative mortality across all stocking densities (Figure 22 and Table 4). The density dependent curves each yielded a 10% cumulative mortality at the lowest stocking density (300 fish / m3) and a 50% cumulative mortality at the highest stocking density (700 fish / m3). However, they differed in the trajectory between the lowest and highest mortality rates (Figure 22 and Table 4). Since there was no prior research found that indicated when during a grow-out period mortality actually occurred, fish, were removed weekly as a function of grow-out period length 40 and assumed total mortality. For example, if total mortality was 10% and the length of grow-out period was 40 weeks, then 0.25% (10% ? 40 weeks) of the total number of fingerlings stocked was removed each week. Thus, weekly mortality added up to the desired cumulative mortality at the end of the production season. When discussing a bio-economic model it is important to define the parameter values and assumptions as their control and influence plays an important role in determining the outcome of the simulation. Parameters held constant in the Stella bio-economic model for this experiment can be found in Table 3. Under these bio-economic model parameters fingerlings grew to the final harvest size in 36 weeks. See general methods section for detailed description of other bio- economic model assumptions, including the bio-economic model growth components (Figure 4 and Figure 6). The number of HSB stocked in the FIPR for each stocking density and the cost to stock are in Table 5. Table 3. Bio-economic parameters for Objective 1. The bio-economic model generated data for each of three mortality curves at five stocking densities for one grow out period (initial stocking to market size) per model run (15 P a r a m e t e r V a l u e F I P R W a t e r V ol u m e 65 m 3 S t oc ki n g d a t e 1- F e b I n i t i a l f i n g e r l i n g w e i g h t 63 g I n i t i a l f i n g e r l i n g l e n g t h 17 .9 c m F i n g e r l i n g p r i c e $0 .7 0 H a r v e s t s i z e 68 1 g M a r ke t p r i c e $7 .1 6/ kg F e e d pr i c e $1 .0 0/ kg E l e c t r i c i t y c os t $0 .0 7/ kW h F i x e d c os t s $8 1/ w e e k V a r i a bl e c os t s 3/ 7 of s pe c i f i e d c os t s 41 total runs). Data resulting from each run included biomass produced, variable costs of production, fixed costs, receipts and net returns (receipts minus variable and fixed costs). Cumulative production and net returns were compared between mortality curves. Receipts, variable costs and fixed costs were compiled into detailed enterprise budgets (Appendix Table 21 through Table 25). Figure 21. Basic Stella model components including three mortality curve choices. Table 4. Cumulative mortalities for each mortality curve and stocking density. S t o c k i n g d e n s i t y ( f i s h / m 3 ) F i x e d L i n e a r E x p o n e n t i a l 300 10% 10% 10% 400 10% 20% 15% 500 10% 30% 22% 600 10% 40% 33% 700 10% 50% 50% M o r t a l i t y C u r v e 42 Figure 22. Relationship between cumulative mortality and stocking density under density independent and density dependent scenarios. Table 5. Hybrid striped bass stocking density, total number of fish stocked and total fingerling cost per FIPR raceway system. Objective 1 Results Final biomass produced for each mortality curve and stocking density combination is shown in Figure 23 and Table 6. Final biomass was highest when the fixed mortality curve was applied and increased linearly with increasing stocking density (Figure 23 and Table 6). Final 0% 10% 20% 30% 40% 50% 60% 300 400 500 600 700 C u m u l at i ve M or t al i t y ( %) S t oc k i n g D e n s i t y ( f i s h / m 3 ) F i xe d L i ne a r E xpone nt i a l S t o c k i n g d e n s i t y ( f i s h / m 3 ) N u m b e r o f F i s h s t o c k e d P r i c e f o r F i n g e r l i n g s ( $ ) 300 1 9 ,5 0 0 $ 1 3 ,6 5 0 400 2 6 ,0 0 0 $ 1 8 ,2 0 0 500 3 2 ,5 0 0 $ 2 2 ,7 5 0 600 3 9 ,0 0 0 $ 2 7 ,3 0 0 700 4 5 ,5 0 0 $ 3 1 ,8 5 0 43 biomass also increased with increasing stocking density when the linear mortality curve was applied, but the rate of increase slowed at higher stocking densities. When the exponential mortality curve was applied, final biomass increased as stocking density increased from 300 to 600 fish / m3, but subsequently decreased as stocking density increased further to 700 fish / m3. Net returns for each mortality curve and stocking density combination are presented in Figure 24 and Table 6). The fixed mortality curve yielded the highest net returns overall, with net returns increasing linearly as stocking density increased (Figure 24 and Table 6). The linear and exponential mortality curves yielded increasing net returns from 300 to 500 fish/m3 but decreasing net returns as stocking density increased further to 700 fish/m3 (Figure 24 and Table 6). Total costs increased with stocking density for all mortality curves. Variable costs, fixed costs and total costs for each stocking density and mortality curve are summarized in Table 7. Detailed enterprise budgets for each stocking density and mortality curve combination are provided in Table 21 through Table 25 in the appendix. 44 Figure 23. Relationship between stocking density and final biomass for three mortality curves. Table 6. Comparison of harvested biomass and net returns for each mortality curve and stocking density. - 5,000 10,000 15,00 0 20,00 0 25,000 30,00 0 35,00 0 300 400 500 600 700 B i om as s P r od u c e d ( k g) S t oc k i n g D e n s i t y ( f i s h / m 3 ) F i xe d L i ne a r E xpo ne nt i a l F i s h / m 3 B i o m a s s ( k g ) N e t R e t u r n ( $ ) B i o m a s s ( k g ) N e t R e t u r n ( $ ) B i o m a s s ( k g ) N e t R e t u r n ( $ ) 300 1 2 ,7 2 6 4 2 ,1 1 1 1 2 ,7 2 6 4 2 ,1 1 1 1 2 ,7 2 5 4 2 ,1 1 1 400 1 6 ,9 6 8 5 8 ,1 1 2 1 5 ,3 0 1 4 8 ,3 0 9 1 6 ,1 4 7 5 3 ,2 8 5 500 2 1 ,2 1 0 7 4 ,1 1 3 1 7 ,0 4 3 4 9 ,6 0 5 1 8 ,6 6 7 5 9 ,1 5 5 600 2 5 ,4 5 2 9 0 ,1 1 4 1 7 ,9 5 1 4 5 ,9 9 9 1 9 ,6 3 4 5 5 ,8 9 2 700 2 9 ,6 9 4 1 0 6 ,1 1 5 1 8 ,0 2 6 3 7 ,4 9 1 1 8 ,0 2 6 3 7 ,4 9 1 F i x e d L i n e a r E x p o n e n t i a l 45 Figure 24. Relationship between stocking density and net returns for three mortality curves. Table 7. Comparison of costs for each mortality curve. Objective 1 Discussion The relationship between stocking density and mortality curve type had important impacts on production and profitability. In the case of the fixed mortality curve, which is independent of stocking density, production and net returns increased as stocking rate increased indicating an optimal stocking density at 700 fish/m3 (Table 6). Although this mortality curve would be ideal for producers, it is the least realistic of the three curve types because it implies $0 $2 0, 00 0 $40,000 $6 0, 00 0 $80,000 $100,000 $120,000 300 400 500 600 700 N e t R e t u r n S t oc k i n g D e n s i t y ( f i s h / m 3 ) F i xe d L i ne a r E xpo ne nt i a l F i s h / m 3 F i x e d L i n e a r E x p o n e n t i a l F i x e d L i n e a r E x p o n e n t i a l F i x e d L i n e a r E x p o n e n t i a l 300 4 5 ,1 3 6 4 5 ,1 3 6 4 5 ,1 3 6 3 ,8 7 2 3 ,8 7 2 3 ,8 7 2 4 9 ,0 0 8 4 9 ,0 0 8 4 9 ,0 0 8 400 5 9 ,5 0 8 5 7 ,3 7 7 5 8 ,4 5 8 3 ,8 7 2 3 ,8 7 2 3 ,8 7 2 6 3 ,3 8 0 6 1 ,2 4 9 6 2 ,3 3 0 500 7 3 ,8 8 0 6 8 ,5 5 2 7 0 ,6 2 8 3 ,8 7 2 3 ,8 7 2 3 ,8 7 2 7 7 ,7 5 2 7 2 ,4 2 4 7 4 ,5 0 0 600 8 8 ,2 5 1 7 8 ,6 6 2 8 0 ,8 1 3 3 ,8 7 2 3 ,8 7 2 3 ,8 7 2 9 2 ,1 2 4 8 2 ,5 3 4 8 4 ,6 8 5 700 1 0 2 ,6 2 3 8 7 ,7 0 6 8 7 ,7 0 6 3 ,8 7 2 3 ,8 7 2 3 ,8 7 2 1 0 6 ,4 9 6 9 1 ,5 7 9 9 1 ,5 7 9 V a r i a b l e C o s t s ( $ ) F i x e d C o s t s ( $ ) T o t a l C o s t s ( $ ) 46 that higher stocking rates have no effect on mortality in FIPR systems. The designer of the prototype FIPR system suggested that stocking densities higher than 500 fish/m3 would produce a higher stress environment for HSB than lower stocking densities because of potentially lower water quality conditions and ultimately could result in a lower survival rate (Chappell, 2010). However, we do not know the exact stocking density in which density dependent mortality will begin to effect production and profitability. Increasing stocking density will eventually cause production and net returns to experience diminishing marginal returns; in other words production and net returns will increase as stocking density increases but at a decreasing rate until density dependent mortality effects actually reduce net return and profitability. For the density-dependent linear mortality curve, biomass increased as stocking density increased but at decreasing rate (Figure 23). For example, when the stocking density was increased from 300 fish/m3 to 400 fish/m3 the biomass produced increased by 2,575 kg. But when the stocking density was increased from 600 fish/m3 to 700 fish/m3 the biomass produced increased by only 75 kg (Figure 23 and Table 6). Adding 100 kilograms of additional fingerlings does not guarantee >2,000 kilograms of increased production. The gain in production decreases with increasing stocking density. The decreasing rate of biomass production is due to the increasing mortality rate (Table 4). For the density-dependent exponential mortality curve, biomass did not always increase with increasing stocking density. Biomass production peaked at a stocking density of 600 fish/m3 and then decreased at the 700 fish/m3 stocking density. The biomass produced at the highest stocking density was the same for both the linear and exponential mortality curves because the mortality rate was set to 50% for both curves at 700 fish/m3. 47 The stocking density that maximized net return at harvest for both the linear and exponential mortality curves was 500 fish/m3 (Figure 24 and Table 6). At stocking densities beyond 500 fish/m3, the increased mortality rates greatly reduced receipts at harvest. The reduction in receipts was not matched with reductions in overall production costs (Appendix Table 21 through Table 25. In brief, the marginal costs of greater production inputs were greater than the marginal returns at the 600 and 700 fish/m3 stocking rates. Although both the linear and exponential curves had maximized net returns at a stocking density of 500 fish/m3 it should be noted that their net return values were different; the exponential curve had net returns approximately $10,000 more than the linear curve. This detail emphasizes the impact that the shape of the mortality curve can have on these FIPR systems, and how important it is to understand these curves. Increased production did not always indicate increased net return for density dependent cumulative mortality curves. Large fish that died late in the production cycle had consumed a considerable quantity of feed and thus added considerable cost to the crop, but did not produce any sales receipts, only production costs. This explains why net returns began to decrease while biomass continued to increase from stocking 500 and 600 fish/m3. When the stocking density of 700 fish/m3 was modeled, both production and net returns decreased because the mortality rate was 50%. Higher total costs for higher stocking densities are seen in Table 7 for all three mortality curves. As stocking density increased, variable costs increased mainly in the form of feed costs (Appendix Table 24 and Table 25). For the fixed mortality curve, the increase in stocking density and therefore feed costs was accompanied by increased receipts (Table 7) and increased net returns. For example, at a stocking density of 700 fish/m3 and a fixed mortality curve, feed 48 costs totaled $37,574 and receipts were $212,611 (Appendix Table 25). The net return above total costs was $106,115. The ratio of receipts to feed costs was 5.63 (212,611 / 37,754). In other words, for every dollar spent on feed, $5.63 was produced as receipts. This was not the case with the linear and exponential mortality curves. The ratio of feed costs to receipts for a stocking density of 700 fish/m3 for linear and exponential mortality curves was 4.73. The extra feed cost did not produce the same ratio of receipts. The increased feed costs were not matched by the same increase in receipts and therefore, profitability was lower. The results of this objective demonstrate that future field research is needed to determine which mortality curve best characterizes HSB in a FIPR systems across a realistic range in stocking densities. Data from field experiments could be used to adjust the Stella bio-economic model parameters to be more realistic and better suited as a tool for making production decisions. First, a field trial would need to establish the stocking density at which mortality begins to increase. For example, if mortality does not increase above 10% until fish are stocked at or above 700 fish/m3 then the fixed mortality rate curve would be most accurate. But if density dependence kicked in at 300 fish/m3 and the linear mortality curve turned out to be accurate, then 500 fish/m3 would be the optimal choice to maximize net returns, Figure 24. Field research needs to be conducted to give us a better understanding of when losses occur, i.e. what mortality curve best fits successively higher stocking rates. To determine this we would need to conduct field trial experiments at stocking densities used in this objective and measure cumulative mortality at each. With resulting production and mortality data we would be able to fit a density dependent mortality curve type to empirical data for each stocking density and incorporate these results into the developed bio-economic model. 49 Another area of future research might be to look at a different relationship altogether; one that relates stocking density to growth rate. In this analysis, the fish grew at the same rate at each stocking density and only mortality was increased or decreased. Instead of only analyzing the effect that stocking density has on mortality in this FIPR system, another study might consider the effect of stocking density on growth rate and, therefore, overall production and profitability. If field studies show that mortality is less affected than growth rate when increasing stocking densities in these systems, then management strategies for maximizing production and profitability may be quite different. It might be possible to obtain data on both objectives in one well run research and field trial effort. 50 Objective 2 Evaluate the effect on production and net returns of reducing the magnitude of catastrophic losses from 50% to 10% by purchasing risk mitigation equipment. Hypotheses: 1) Decreasing magnitude of stochastic events from 50% to 10% will greatly improve production and net return and increase the chance of positive net returns. 2) Crops with loss events occurring later in the grow-out period will experience lower net returns due to higher accumulated variable costs, than crops with loss events occurring earlier in the grow-out period. Objective 2 Methods In this bio-economic modeling process, risk for FIPR system production is expressed in terms of fish lost through stochastic disaster events. Loss events are composed of two parts: 1) the frequency or probability of one or more loss events occurring; and 2) the magnitude of loss during each loss event. As the frequency and magnitude of stochastic loss events increase, it is expected that the average net return should decrease. For this objective, a bio-economic model component was developed to examine the effects of frequency and magnitude of unexpected loss events, i.e., beyond ?natural? losses, on total production and net returns. A review of the literature found no data for expected levels of risk or catastrophic loss for HSB FIPR systems. The prototype FIPR system in West Alabama experienced a high loss event due to unknown causes in one of its first field trial runs which prompted this portion of the bio- economic model to be developed. Generally, in fish production systems, there are many potential risk factors that can cause stochastic mortality events such as electrical outages, disease 51 events, weather events and poor water quality. Densely stocked systems, such as in FIPR systems, require higher levels of management oversight to assure water quality is maintained and disease outbreaks are treated efficiently. Understanding potential risk factors and their effect on net return in FIPR systems is essential in assessing their economic feasibility. This objective does not focus on a specific cause of loss, but instead examines a range of risk factor values for a FIPR system, The risk factors utilized in this objective were four different frequency of loss values, (40%, 20%, 10% and 1% chance of a loss event occurring during a grow-out period) and two different magnitudes of stochastic catastrophic events (50% and 10% loss of fish) (Table 8). Four frequency of loss values were chosen for this objective because a typical value for FIPR systems is unknown and also because expert opinion suggests that with time and experience, managers can often reduce risk in fish production systems. The 50% magnitude of loss scenarios represented a high loss event whereas the 10% scenarios represented a lower loss event for comparison. A total of eight scenarios were examined. To assess these risk factor effects in FIPR systems, a stochastic component was added to the Stella bio-economic model to randomly initiate loss events. A Monte Carlo function that randomly produced a one or a zero to create a converter called Catastrophe Dice (Figure 25) was utilized. The Catastrophe Chance converter, linked to a slider bar on the user control page, allows the user to select the percent chance that a value of one, rather than zero, will be generated by the Catastrophe Dice, Figure 25. When Catastrophe Dice produces a one rather than a zero a catastrophic loss is initiated. The percent of fish lost during a catastrophic event is controlled by the Catastrophe Magnitude converter. Catastrophic Death multiplies the number 52 of fish currently in the raceway by the percent loss (Catastrophic Magnitude) and adds the result to the death outflow (Figure 25). Figure 25. Stella catastrophe components Because the bio-economic model component used for this objective was stochastic, rather than deterministic, the model was run 500 times for the eight scenarios and net returns at harvest was calculated each time. The net return values from the 50% loss scenarios were either very high or very low rather than being normally distributed because any crop where a loss event occurred had a substantial loss in net returns. Crops with no loss event had a high uniform net return value. Therefore the average and standard deviation of the positive net return values and the average and standard deviation of the negative net return values were calculated separately for each scenario. Receipts, variable costs and fixed costs were compiled into detailed enterprise budgets for marginal analysis. 53 Table 8. Catastrophic loss variables for a 36 week grow out period. Several parameters were held constant in the bio-economic model for this objective (Table 9.). Optimal growth occurred at water temperatures between 23-25?C and no growth occurred below 11?C or above 34?C (See Figure 4 and Figure 6 in general methods). Additional variable and fixed costs were added based on similar cost items for a 1.4 ha pond (see general methods for more detail). The crop was harvested when the average fish size reached or exceeded 0.681 kg (D?Abramo et. al., 2008). The bio-economic model parameters resulted in grow-out duration of 36 weeks from initial stocking to harvest at market size. Table 9. Bio-economic parameters for Objective 2. % C h a n c e o f D i s a s t e r p e r g r o w o u t p e r i o d % C h a n c e o f D i s a s t e r p e r w e e k 40% 1 .1 1 % 50% or 10% 20% 0 .5 6 % 50% or 10% 10% 0 .2 8 % 50% or 10% 1% 0 .0 3 % 50% or 10% % L o s s O c c u r r i n g P a r a m e t e r V a l u e F I P R W a t e r V ol u m e 65 m 3 S t oc ki n g d a t e 1- F e b I n i t i a l f i n g e r l i n g w e i g h t 63 g I n i t i a l f i n g e r l i n g l e n g t h 17 .9 c m f i n g e r l i n g p r i c e $0 .7 0 H a r v e s t s i z e 68 1 g T ot a l M or t a l i t y 10% M a r ke t p r i c e $7 .1 6/ kg F e e d pr i c e $1 .0 0/ kg E l e c t r i c i t y c os t $0 .0 7/ kW h F i x e d c os t s $8 1/ w e e k V a r i a bl e c os t s 3/ 7 of s pe c i f i e d c os t s 54 Objective 2 Results Positive and negative net return averages and standard deviations were calculated from 500 model runs for each loss scenario (Table 10 and Table 12). For the 50% magnitude of loss scenarios, frequency of loss scenarios of either 40% or 20% were most likely to have negative net returns and had the lowest average negative net returns (Table 4). These same frequency loss levels had the lowest average positive net returns along with the highest standard deviations (Table 10). The scenarios with 10% and 1% frequency of loss had lower chances of negative net returns occurring and higher average positive net returns. For the scenario with 1% frequency of loss, there were no negative net returns; only three loss events occurred in the 500 runs which allowed for the highest average of positive net return and lowest standard deviation out of the four scenarios. Table 10. Net returns and standard deviations for 50% magnitude of loss scenarios. *Since the grow out period is 36 weeks, each week has an equal chance of a stochastic loss occurring. F r e q u e n c y o f L o s s p e r G r o w O u t P e r i o d F r e q u e n c y o f L o s s p e r W e e k * C h a n c e o f N e g a t i v e N e t R e t u r n A v e r a g e o f N e g a t i v e N e t R e t u r n ( $ ) A v e r a g e o f P o s i t i v e N e t R e t u r n ( $ ) 40% 1 .1 1 % 6% - 2 0 ,5 2 1 + / - 5 ,7 8 1 5 7 ,0 2 3 + / - 2 8 ,0 7 5 20% 0 .5 6 % 2% - 2 3 ,1 2 6 + / - 7 ,4 2 8 6 4 ,1 3 4 + / - 2 2 ,9 4 0 10% 0 .2 8 % 0 .4 0 % - 1 9 ,2 8 5 + / - 2 ,7 9 9 6 8 ,9 3 5 + / - 1 7 ,3 1 7 1% 0 .0 3 % 0 .0 0 % - - 7 3 ,8 6 6 + / - 3 ,8 9 9 S t a n d a r d D e v i a t i o n o f N e g a t i v e N e t R e t u r n ( $ ) S t a n d a r d D e v i a t i o n o f P o s i t i v e N e t R e t u r n ( $ ) 55 Net return values for the 50% magnitude of loss at 40% (40% / 36 weeks in grow out period = 1.11% per week) frequency of loss trials can be put into several category levels: 1) either very high returns, at exactly $74,113; or 2) low returns, between $14,000 and $0; or 3) very low returns, between -$17,000 and -$39,661 (Table 11). When no loss occurred, net returns were uniformly $74,113 but when one loss event occurred the average net return value was between $11,093 and $12,393 (Table 11). A 50% loss event could reduce net returns values to the $10,000 - $14,000 range. Lower net returns were associated with trials having multiple loss events and negative net returns were associated with trials that had both multiple loss events and loss events occurring in the latter half of the production cycle. Table 11. Other net return values for 50% magnitude of loss. When the bio-economic model scenarios with a magnitude of loss of 10% rather than 50% were simulated, there were no scenarios that had a negative net return (Table 12). As the frequency of loss decreased, the average net returns increased. Compared to the net returns for the 50% magnitude of loss scenarios, the 10% magnitude of loss scenarios had a lower standard deviation (Table 12). Without a loss event occurring, net returns for the 10% magnitude of loss scenarios was $74,113. Net returns for 10% frequency of loss trials did not fall into the same high, low positive or negative categories that the 50% trials did (Table 13). The lowest net return value that occurred was $48,234. 40% 20% 10% 1% A v e r a g e p o s i t i v e n e t r e t u r n v a l u e s o f 1 o r m o r e l o s s e s 1 1 ,0 9 3 1 1 ,5 4 9 1 1 ,2 1 5 1 2 ,3 9 3 L o w e s t p o s t i v e n e t r e t u r n v a l u e 2 ,2 2 7 5 ,0 4 9 2 ,2 2 7 1 2 ,3 5 8 L o w e s t n e g a t i v e n e t r e t u r n v a l u e - 3 9 ,6 6 1 - 3 8 ,0 6 7 - 2 1 ,2 6 4 --- Fr e q u e n c y o f L o s s 56 Table 12. Net return for 10% magnitude loss scenarios. *Since the grow out period is 36 weeks, each week has an equal chance of a stochastic loss occurring. Table 13. Other net return averages for grow-out periods with 10% magnitude of loss. Objective 2 Discussion Unanticipated loss events caused by electrical outages, disease outbreaks or managerial error are not uncommon in aquaculture production systems but the likelihood of these events occurring and their effect on production and net returns in FIPR systems are undocumented in the literature. By assessing a range of values for frequency of loss and magnitude of loss, inferences can be made into the effect of risk on production and net returns for FIPR systems. By comparing the results of the 50% magnitude of loss scenarios to the 10% magnitude of loss scenarios we can better understand the effects of risk mitigation strategies on costs and net returns. It was hypothesized that the FIPR bio-economic model scenarios with the higher magnitudes and frequencies of loss (fish death) events would have greater variability and lower F r e q u e n c y o f L o s s F r e q u e n c y o f L o s s p e r W e e k * C h a n c e o f N e g a t i v e N e t R e t u r n A v e r a g e o f N e g a t i v e N e t R e t u r n S t a n d a r d D e v i a t i o n o f N e g a t i v e N e t R e t u r n A v e r a g e o f P o s i t i v e N e t R e t u r n 40% 1 .1 1 % 0% --- --- 6 9 ,4 1 1 + / - 7 ,2 1 2 20% 0 .5 6 % 0% --- --- 7 1 ,7 4 1 + / - 5 ,1 6 6 10% 0 .2 8 % 0% --- --- 7 2 ,9 4 8 + / - 4 ,0 6 3 1% 0 .0 3 % 0% - - - --- 7 4 ,0 6 3 + / - 787 S t a n d a r d D e v i a t i o n o f Pos i t i v e N e t R e t u r n ( $ ) 40% 20% 10% 1% A v e r a g e p o s i t i v e n e t r e t u r n v a l u e s o f 1 o r m o r e l o s s e s 5 9 ,6 0 1 6 0 ,9 3 4 6 0 ,2 4 4 6 1 ,6 5 7 L o w e s t p o s t i v e n e t r e t u r n v a l u e 4 8 ,2 3 4 4 9 ,9 8 8 3 9 ,1 5 0 6 1 ,5 4 6 L o w e s t n e g a t i v e n e t r e t u r n v a l u e --- --- --- --- Fr e q u e n c y o f L o s s 57 net returns than scenarios with lower magnitudes of loss per loss event and lower frequencies (probabilities) of loss events occurring. This hypothesis was proven correct. In general, a 50% magnitude of loss event was catastrophic for net returns. Although the chance of negative net returns was low, ranging from 0%-6%, trials with the least detrimental losses still experienced a decrease in net returns of approximately $61,000. Some crop trials had a negative net return of approximately -$40,000. Why the difference in net returns for just one crop with a 50% loss? In some trials, a loss event occurred more than once, causing a decrease in HSB numbers and therefore, receipts. When the fish were sold at harvest, net return was much lower due to the greatly decreased number of HSB in the FIPR. The timing of the catastrophic loss also had impact on the net returns. If the loss event occurred early in the grow-out period net returns were higher, conversely if the loss occurred near the end of the grow-out period, the net returns were much lower. The reason for this difference is in the variable costs, mainly feed. Fish that were lost to catastrophe had consumed feed from the time they were stocked. Therefore, when they were removed from the system (loss event mortality) before they could be sold, those costs were unrecoverable. When a loss event occurred near the end of a grow-out period, all the feed that had been consumed by the HSB was lost. Figure 26 illustrates a crop scenario where no loss event occurs. In Figure 27, the loss event occurred early in grow-out period and the final net return was approximately $12,000. Feed costs were reduced in this case as there were fewer fish to feed from an earlier stage of the production cycle. In Figure 28, the loss event occurred in the middle of the grow-out period and the net returns were approximately $10,600. In Figure 29, the loss event occurred near the end of the grow-out period and the net returns were approximately $7,000. Negative net returns occurred when there were two or more loss events in one grow-out period. 58 Receipts were much greater, almost double, for the 10% magnitude of loss scenarios compared to the 50% magnitude of loss scenarios, and when net returns are compared, the 10% loss scenarios are more than three times more profitable. The increase in net returns for the 10% magnitude of loss scenario came from increased survival and production over the 50% scenario rather than a difference in production costs. Cost of production for the two magnitude scenarios were similar, so the difference between the net return levels is derived from the improved survival and increased production. Detailed enterprise budgets comparing an early, middle and late occurrence of a loss event can be found in the Appendix Table 26 and Table 27. Production data comparing an early, middle and late occurrence of a loss event can be found in Appendix Table 28 and Table 29. Figure 26. Number of fish in FIPR and Net Return over time, no loss event.1 1 Initial stocking occurred in week 5 of the simulation run. - 80,000 - 60 ,0 00 - 40,000 - 20 ,0 00 0 20 ,0 00 40,000 60 ,0 00 80,000 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 0 10 20 30 40 N e t r e t ur n ( $) F is h i n F I P R G r ow - out T i m e ( w e e ks ) N um be r of f i s h N e t R e t ur n 59 Figure 27. Loss event of 50% magnitude occurring 5 weeks after stocking.2 2 Initial stocking occurred in week 5 of the simulation run. Figure 28. Loss event of 50% magnitude occurring 19 weeks after stocking.3 3 Initial stocking occurred in week 5 of the simulation run. - 80,000 - 60 ,0 00 - 40,000 - 20 ,0 00 0 20 ,0 00 40,000 60 ,0 00 80,000 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 0 10 20 30 40 N e t r e t ur n ( $) F is h i n F I P R G r ow - out T i m e ( w e e ks ) N um be r of f i s h N e t R e t ur n - 80,000 - 60 ,0 00 - 40,000 - 20 ,0 00 0 20 ,0 00 40,000 60 ,0 00 80,000 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 0 10 20 30 40 N e t r e t ur n ( $) F is h i n F I P R G r ow - out T i m e ( w e e ks ) N um be r of f i s h N e t R e t ur n 60 Figure 29. Loss event of 50% magnitude occurring 32 weeks after stocking. 4 4 Initial stocking occurred in week 5 of the simulation run. A loss event at the 50% magnitude level caused the net return values to vary widely between trials, with no loss events at one end of the spectrum and trials with one or more loss events at the other end of the spectrum. The net return values were either a reoccurring maximum value when no loss event occurred, or a very low positive or negative net return value in the event of one or more loss events. The net return data was bifurcated rather than bell- shaped, with values either very high, or very low. The wide range of potential net returns with a high catastrophic loss indicates the importance of mitigating or reducing risk in FIPR systems before they happen. The lowest amount of risk possible is ideal but probably not realistic because there is always an element of uncertainty (risk) in fish production systems. Investing in risk mitigation equipment such as automatic emergency aeration systems or hiring more highly skilled workers can be expensive, but the benefits of reducing risk are difficult to ignore and may be worth the additional cost. The 10% magnitude of loss scenarios had a much smaller range of - 80,000 - 60 ,0 00 - 40,000 - 20 ,0 00 0 20 ,0 00 40,000 60 ,0 00 80,000 0 5,000 10,000 15,000 20,000 25,000 30,000 35,000 0 10 20 30 40 N e t r e t ur n ( $) F is h i n F I P R G r ow - out T i m e ( w e e ks ) N um be r of f i s h N e t R e t ur n 61 net return values and all of the net returns were positive. In the event of loss occurring, the reduction of net returns was much less than with the higher 50% magnitude of loss. Results from the bio-economic model scenarios with the lower magnitudes and frequencies of loss had the lowest variability and highest net returns as hypothesized. Even with a 40% chance of loss, the lowest net return was $48,223, much higher than the lowest positive net return for 50% magnitude of loss scenarios (-$39,661), so the hypothesis was found to be valid. Also, it was hypothesized that HSB FIPR crops with loss events occurring later in the grow-out period will experience lower net returns due, to higher variable costs, than crops with loss events occurring earlier in the grow-out period. In fact it was found that crops with loss events occurring early in the grow-out period did experience higher net returns than crops experiencing loss events later in the grow-out period. The results of this objective serve as an indicator of areas for future research in FIPR systems. Field research could examine both the frequency of unpredicted loss events occurring as well as measuring mortality levels from loss events to provide frequency and magnitude loss information that would make the bio-economic model more realistic and useable by producers. The results also serve as an indicator to the importance of understanding the risks associated with FIPR systems. If the frequency of loss events is likely to be high and the loss from an event is likely to be great, a producer may want to look at alternative, less risky production systems or, alternatively, if the frequency and magnitude of loss events can be reduced and managed by good producers then the FIPR system can provide high returns. 62 Objective 3 Examine the influence of initial stocking size and initial stocking date on crop duration, production and net returns in the HSB FIPR system under a water temperature-dependent growth regime. Hypotheses: 1) Larger fingerlings will produce higher net returns than smaller fingerlings stocked in any given month because the larger fingerlings will reach market size in less time. 2) Net returns will be greater for larger fingerlings stocked early in the year as opposed to late in the year due to shortened grow-out period requirements. Objective 3 Methods This objective models the production and net returns stemming from specific HSB fingerling size and stocking date combinations to determine their influence on profitability of FIPR systems. For this objective several parameters were held constant, Table 14 Optimal growth occurred at water temperatures between 23?C and 25?C and no growth occurred below 11?C or above 34?C. (Refer to the general methods section that contains a description and graph of the growth rate relationship to water temperature over a 52 week period). Additional variable and fixed costs were added based on similar costs for a 1.4 ha pond. (See general methods section for detail). The crop was harvested when the fish size reached or exceeded 0.681 kg (D?Abramo et. al., 2008). 63 Table 14. Bio-economic parameters for Objective 3. Unlike the previous objectives where an additional bio-economic model component was added to account for mortality or risk, this objective used existing bio-economic model components. The fingerling size and stocking month parameters of the bio-economic model were varied to produce production and net return data for each scenario (Table 15). HSB fingerlings were purchased according to length, in inches, rather than by weight and the relative lengths for each weight are included into the cost and production components of the Stella bio- economic model (Table 15). The first day of each month was used as the stocking date parameter for each scenario. The HSB fingerling sizes used were 3?, 4?, 5?, 6? and 7? and each size was stocked in each month from January through December. The bio-economic model calculated net returns for each size and month combination (total of 60 scenarios). Results were used to develop a table indicating production time (weeks until harvest size was reached) production quantity, and net returns at harvest. The cost to stock the FIPR with HSB fingerlings at a stocking density of 500 fish/m3 varied with each fingerlings size (Table 15). Fingerling pricing was held constant at $0.10 per inch throughout the year with a shipping cost of $950 per order. P a r a m e t e r V a l u e F I P R W a t e r V ol u m e 65 m 3 H a r v e s t s i z e 681 g T ot a l M or t a l i t y 10% M a r ke t pr i c e $7.16/ kg F e e d pr i c e $1.00/ kg E l e c t r i c i t y c os t $0.07/ kW h F i x e d c os t s $81/ w e e k V a r i a bl e c os t s 3/ 7 of s pe c i f i e d c os t s 64 Table 15 Size and weight of HSB fingerlings, their prices and cost to stock the FIPR system used in Objective 3 scenarios (includes a constant $950 per order shipping cost). Objective 3 Results Grow-out duration, net returns at harvest and weekly net returns for each fingerling-size / stocking month combination are provided in Table 16. Because net return at harvest does not reflect the length of time required for receipts to be generated, the average net return per week was calculated in order to better compare scenarios. The net return per week value represents the profit earned per week of grow-out effort. In some instances, net returns were very high, but the length of time required was also high whereas other scenarios had slightly lower net return but half the grow-out time needed (Table 2). For example, 3? fingerlings stocked in July returned $77,117 in 64 weeks whereas 7? fingerlings stocked in February returned $74,113 in only 36 weeks. As fingerling size increased the range of weeks required for grow-out for each stocking month of the year decreased (Table 17). Each fingerling size had initial stocking months with high profitability and months with low profitability (Table 16). For most months of the year the 7? fingerling was the most profitable choice in terms of net returns and time compared to the other fingerling sizes. They exhibited net returns above $60,000 when stocked in January ? March or September-December, but had net returns below $30,000 when stocked in April or L e n g t h ( i n ) L e n g t h ( c m ) W e i g h t ( g ) $ / f i n g e r l i n g C o s t t o s t o c k F I P R ( $ ) 3 7 .6 6 0 .3 0 $ 1 0 ,7 0 0 4 1 0 .2 14 0 .4 0 $ 1 3 ,9 5 0 5 1 2 .7 25 0 .5 0 $ 1 7 ,2 0 0 6 1 5 .2 43 0 .6 0 $ 2 0 ,4 5 0 7 1 7 .7 8 63 0 .7 0 $ 2 3 ,7 0 0 65 May, (Table 16). The 6? fingerling size was generally the least profitable choice with only one stocking month of the year (September) having net returns over $60,000. The 4? fingerling size was also not as profitable as the 7? fingerling but did exhibit one highly profitable month, August, when the net returns were approximately $80,000 at harvest. The 3? fingerling was a highly profitable choice (net returns > $60,000) when stocked December through August excluding February. On average, the 3? fingerling had the highest net returns at $62,779, but also averaged the longest grow-out period lengths at 72 weeks (Table 16). In contrast, the 6? fingerling had the lowest net returns average ($36,095) but also a lower average grow-out period length (60 weeks). The 7? fingerling had the shortest grow-out period length (46 weeks) and highest average net returns per week of grow-out ($1,268). The fingerling size and stocking month with the highest average net return per week was the 7? fingerling when stocked in March and the lowest average net return per week combination was the 7? fingerling stocked one month later in April. Biomass production results for each scenario are provided in Table 3 and ranged from a high of 21,210 kg for the 7? fingerling stocked in February to a low of 17,691 kg for the 3? fingerling stocked in November, Table 17. Production quantities varied between the scenarios because the bio-economic model harvests the HSB when they reach or exceed 0.681 kg each. In some scenarios the HSB were nearly 0.681 kg in one week then exceeded that weight in the following week and then were harvested. Those crops had higher biomass compared to crops where the fish size reached exactly 681 grams in a certain week and were not allowed the extra week of growth that other crops were given. To better compare the differences in total biomass produced from each scenario, the production per cycle, that is the weight produced divided by the number of weeks required for the HSB to reach harvest size was calculated, Table 3. This 66 allows comparison between scenarios beginning with larger fingerlings needing less time to reach harvest size to smaller fingerlings needing more time to reach harvest. On average the initial 7? stocked fingerlings produced the highest weekly biomass at 452 kg per week, whereas the initially stocked 3? fingerling produced an average of 261 kg per week. 67 Table 16. Weeks of grow out, net returns and net return per week for varying fingerling length -stocking date combinations. M on t h w k s o f g r o w - out N e t R e t u r n ( $ ) N e t R e t u r n p e r w e e k ( $ ) w k s o f g r o w - out N e t R e t u r n ( $ ) N e t R e t u r n p e r w e e k ( $ ) w k s o f g r o w - out N e t R e t u r n ( $ ) N e t R e t u r n p e r w e e k ( $ ) w k s o f g r o w - out N e t R e t u r n ( $ ) N e t R e t u r n p e r w e e k ( $ ) w k s o f g r o w - out N e t R e t u r n ( $ ) N e t R e t u r n p e r w e e k ( $ ) J 74 60,117 812 72 53,648 745 69 34,964 507 66 15,562 236 40 69,049 1,726 F 69 59,703 865 67 53,466 798 65 43,591 671 61 16,284 267 36 74,113 2,059 M 66 63,689 965 63 52,227 829 61 43,110 707 57 16,514 290 32 73,450 2,295 A 64 64,557 1,009 61 56,854 932 59 52,490 890 55 27,400 498 51 6,289 123 M 65 68,907 1,060 63 65,993 1,048 59 56,154 952 55 46,078 838 52 29,904 575 J 65 74,656 1,149 63 69,301 1,100 60 61,307 1,022 56 55,694 995 51 42,938 842 J 64 77,117 1,205 62 71,826 1,158 60 69,228 1,154 56 59,148 1,056 47 46,579 991 A 66 71,939 1,090 62 80,025 1,291 59 71,534 1,212 55 59,322 1,079 52 55,756 1,072 S 86 40,168 467 83 17,425 210 60 67,196 1,120 55 68,037 1,237 47 60,864 1,295 O 85 52,386 616 83 39,707 478 81 25,627 316 61 47,368 777 52 67,424 1,297 N 82 57,305 699 80 49,794 622 78 38,469 493 74 10,401 141 49 68,663 1,401 D 79 62,806 795 76 49,677 654 74 38,860 525 70 11,334 162 45 69,099 1,536 AVG 72 62,779 894 70 54,995 822 65 50,211 797 60 36,095 631 46 55,344 1,268 3" F i n g e r l i n g 4" F i n g e r l i n g 5" F i n g e r l i n g 6" F i n g e r l i n g 7" F i n g e r l i n g 68 Table 17. Production weight at harvest (kg) and production quantity per week of grow-out for each fingerling size and month scenario. M on t h P r odu c t i on ( kg ) P r odu c t i on pe r w e e k P r odu c t i on ( kg ) P r odu c t i on pe r w e e k P r odu c t i on ( kg ) P r odu c t i on pe r w e e k P r odu c t i on ( kg ) P r odu c t i on pe r w e e k P r odu c t i on ( kg ) P r odu c t i on pe r w e e k J 18,387 248 19,106 265 18,601 270 19,640 298 20,322 508 F 18,402 267 19,123 285 20,075 309 19,647 322 21,210 589 M 19,155 290 18,787 298 19,745 324 19,284 338 20,715 647 A 18,953 296 18,930 310 20,220 343 19,476 354 19,429 381 M 18,911 291 19,758 314 19,363 328 19,952 363 20,314 391 J 18,854 290 19,242 305 19,286 321 20,241 361 19,666 386 J 18,621 291 18,977 306 20,001 333 20,169 360 19,592 417 A 18,531 281 19,833 320 19,646 333 19,348 352 20,450 393 S 18,740 218 17,736 214 18,964 316 19,882 361 20,531 437 O 17,763 209 17,908 216 18,622 230 18,856 309 20,579 396 N 17,691 216 18,314 229 19,170 246 18,853 255 20,643 421 D 18,734 237 18,288 241 19,168 259 18,861 269 20,618 458 AVG 18,562 261 18,833 275 19,405 301 19,517 329 20,339 452 3" 4" 5" 6" 7" 69 Objective 3 Discussion There are several factors that affect grow-out times and net returns resulting from the different fingerling sizes and stocking months. The growth function is the backbone of the bio- economic model simulations because of the impact it has on production and ultimately profitability. In the bio-economic model, growth rate varies with water temperature and with average size of the HSB crop. January, February, November and December are the coldest months and thus least favorable for HSB growth (Figure 5). (See general methods section for detailed descriptions of growth functions). The relationship between fingerling size and net returns changes dramatically between stocking months. While 7? fingerlings were the most profitable size for February stocking, they are the least profitable size for April stocking (Table 16 and Table 17). When stocked in February, the 7? fingerlings were harvested before the start of the winter period, and net returns were very high. Overwintering HSB in the FIPR system will often mean lowered net returns. HSB must be fed during the cold months of the year to maintain their prior weight. This lowers net returns because the costs of feeding the HSB through a winter period are not matched by increases in biomass value. When stocked April 1st, the 7? fingerlings almost reach harvest size before the cold winter period, but do not achieve the harvest weight threshold. When winter begins, the 7? fingerlings are approximately 600 grams and they must be fed to maintain their weight. The other fingerling sizes must be overwintered as well but they are smaller and require less feed to be maintained so are more profitable than the 7? fingerling when stocked in April (Figure 31). Therefore, the 3? fingerlings had the highest net return at harvest and the highest net return per week because they required the least amount of feed during the winter period (Table 16). 70 The 6? fingerling was often the worst choice of fingerling size for net returns (January, February, March, November and December). Compared to the other fingerling sizes, the six inch fingerlings are often largest when the winter period began, so those fish required more feed to be maintained; plus the cost to initially purchase 6? fingerlings added to overall operating costs compared to stocking costs for smaller fingerlings. The 7? fingerling costs more to purchase than the 6? fingerling but because it can reach market size faster, it often a more profitable choice. Figure 30. Comparing five fingerling lengths at stocking to growth over time when stocked on February 1 (week 5) and harvested at ? 681 grams. 0 100 200 300 400 500 600 700 800 0 10 20 30 40 50 60 70 80 F i s h w e i gh t ( g) G r ow O u t ( w e e k s ) 3" 4" 5" 6" 7" 71 Figure 31. Comparing five fingerling lengths at stocking to growth over time when stocked on April 1 (week 513) and harvested at ? 681 grams. Results showed a consistent trend with crops having smaller fish going into the overwintering period accumulating lower variable costs, primarily feed, during that period and were more profitable once grown out to harvest size. In cases where overwintering could be avoided all together, net returns were highest. Although the cost of purchasing 7? fingerlings was higher than the cost for smaller fingerlings, the net returns were also higher when the 7? fingerlings either did not need to go through the winter period or were overwintered when they were still relatively small. The bio-economic model data demonstrates that water temperature and initial fingerling size strongly impact production and net returns. Field research is needed to more accurately determine the relationship between grow out time and temperature regimes. This information can then be used to adjust Stella bio-economic model parameters and improve the accuracy of 0 100 200 300 400 500 600 700 0 10 20 30 40 50 60 70 80 90 F i s h w e i gh t ( g) G r ow O u t ( w e e k s ) 3" 4" 5" 6" 7" 72 growth, production and net return predictions. This objective assumed that purchased fingerlings can be accurately graded by inch categories and fish will grow to harvest size uniformly. Depending on the time of year, certain HSB fingerlings sizes may or may not be available for producers to purchase from hatcheries and producers may have to accept particular fingerling sizes that they had not anticipated receiving. In reality, fish purchased from producers can vary by age and size and the impacts of this variation should also be measured in field trials. Field trials using tightly graded fingerlings compared to ambient variation from the commercial hatchery would be meaningful and an important data addition for this bio-economic model and for the industry. The bio-economic model could be adjusted to include variability in growth due to variation in fingerling size to more accurately present HSB crops in the FIPR system. The growth component of the bio-economic model has such strong effects on production and net returns that any field trials generating data pertaining to HSB growth in FIPR systems as it relates to temperature would be a tremendous asset to improving the accuracy of the model. If field trials demonstrated that other aspects of production, such as protein content of feed, or stocking density, or water quality had similarly strong impacts on fish growth, then the model could be adjusted to those results as well. 73 Objective 4 Demonstrate how the bio-economic model can be used by researchers and/or producers to make longer term production decisions. Specifically, choose stocking date and initial fingerling size combinations based on data obtained from objective three to produce the highest net return to obtain the highest net return over a five year planning horizon. Hypothesis: 1) Stocking 7? fingerlings for each crop will allow more crops (and therefore more cumulative biomass) to be harvested in a five-year period than would be the case for crops stocked with fingerlings less than 7? in length. Objective 4 Methods With the variety of fingerling stocking sizes to choose from and the seasonal range in stocking dates, it is difficult to determine which combination will be the most profitable over a five year production period. This objective developed a method to use Objective 3 results to determine the most profitable five year production schedule. Profitability was measured through calculated net returns at harvest for all completed crops and accumulated costs for any unfinished crops for this time frame (Table 16). Stocking 7? fingerlings on March 1 was selected as the initial stocking scenario for the 5 year plan as it produced the highest weekly net returns in Objective 3. The dates of subsequent stockings were always chosen as the first day of the month following the date of the previous harvest and the fingerling size chosen for each subsequent stocking was based on which size would be most profitable when stocked on day, according to Objective 3 Table 16. Since the 5 year cycle started on March 1, the final day of the cycle was February 28. 74 According to the results of objective three, a crop with a start date of April 1 will be most profitable when stocked with 3? fingerlings. Therefore, this combination of stocking date and fingerling size was used to initiate a second 5-year production schedule for comparison to production schedule 1 (7? fingerlings stocked on March 1). For the subsequent crops, the optimal fingerling size choice was selected as in the first five year production schedule, where the fingerling with the highest net returns per week for the stocking date was stocked. Several bio-economic model parameters were held constant for this objective (Table 18). Optimal growth occurred at water temperatures between 23?C and 25?C and no growth occurred below 11?C or above 34?C. (Refer to the general methods section for complete description of growth rates)). Additional variable and fixed costs were added based on similar costs for a 1.4 ha pond. (See general methods section for detail). The crop was harvested when the fish size reached or exceeded 0.681 kg. Existing bio-economic model components were used and no additional components were developed to accomplish this objective. Table 18. Bio-economic model parameters for Objective 4. P a r a m e t e r V a l u e F I P R W a t e r V ol u m e 65 m 3 H a r v e s t s i z e 681 g T ot a l M or t a l i t y 10% M a r ke t pr i c e $7.16/ kg F e e d pr i c e $1.00/ kg E l e c t r i c i t y c os t $0.07/ kW h F i x e d c os t s $81/ w e e k V a r i a bl e c os t s 3/ 7 of s pe c i f i e d c os t s 75 Objective 4 Results The crop number, stocking date (month), fingerling stocking size, weeks of grow-out, year stocked and year harvested, biomass produced, net returns per week and net returns for the five year production schedule incorporating only 7? fingerlings are provided in Table 19. After the first stocking on March 1, the stocking and harvesting fell into a cycle of September harvest, October restocking, October harvest and November restocking. The first crop stocked in March Year 1 was harvested in October Year 1. The second crop was stocked on November 1 Year 1 with the harvest occurring 49 weeks later at the end of September Year 2. The third stocking occurred October 1 Year 2 and was harvested 52 weeks later in October Year 3. The subsequent stocking was on the 1st day of the following month, November 1 Year 3, and this pattern of harvest and re-stocking repeated itself thereafter (Table 19). The sixth stocking occurred on November 1 in Year 5 and since the crop could not be harvested until October in Year 6, net returns were negative because costs had accumulated with no receipts (although biomass production had occurred). Table 19. Five year production plan from the bio-economic model using the most profitable choice variables for HSB in a FIPR system beginning March 1 Year 1 and ending the last day of Year 5, Feb 28. C r op # S t oc ki n g D a t e F i n g e r l i n g S i z e W e e ks o f g r ow o u t Y e a r S t a r t e d Y e a r H a r v e s t e d B i om a s s P r od u c e d ( kg ) N e t R e t u r n pe r W e e k N e t R e t u r n ( $) 1 1- M a r 7" 32 1 1 20 ,8 32 2, 29 5 $7 3, 45 0 2 1- N ov 7" 49 1 2 20 ,7 58 1, 40 1 $6 8, 66 3 3 1- O c t 7" 52 2 3 20 ,6 95 1, 29 7 $6 7, 42 4 4 1- N ov 7" 49 3 4 20 ,7 58 1, 40 1 $6 8, 66 3 5 1- O c t 7" 52 4 5 20 ,6 95 1, 29 7 $6 7, 42 4 6 1- N ov 7" 17 5 6 2, 24 7 ( 2, 47 6) - $4 2, 09 5 T ot a l 10 5, 98 6 $3 03 ,5 29 76 The data for the five year production plan beginning with a 3? fingerling in March Year 1 is provided in Table 20. For the first two crops, stocking 3? fingerlings had the highest net returns per week. For the three following stockings the 7? fingerlings had the highest net return values per week, and therefore were the size stocked in crops three through five. For the final stocking on September 1 Year 5, the crop would not be harvestable until Year 6, therefore the crop had a negative value at the end of the calendar Year 5 because crop costs had accumulated with no receipts. Table 20. Five year production plan from the bio-economic model using the most profitable choice variables for HSB in a FIPR system FIPR system beginning April 1 Year 1 and ending the last day of Year 5, March 31. The 7? stocking plan (Table 19) had a total of five harvested crops and one unfinished crop at the end of year five. Total net returns over the five year period was $303,529 and total biomass produced was 105,986 kg. The mixed-size stocking plan (Table 20) had four harvested crops and one unfinished crop at the end of year five. Total net return over the five year period C r op # S t oc ki n g D a t e F i n g e r l i n g S i z e W e e ks o f g r ow o u t Y e a r S t a r t e d Y e a r H a r v e s t e d B i om a s s P r od u c e d ( kg ) N e t R e t u r n pe r W e e k N e t R e t u r n ( $) 1 1- A pr 3" 6 4 1 2 20 ,3 06 1, 00 9 $6 4, 55 7 2 1- J u l 3" 64 2 3 19 ,9 50 1, 20 5 $7 7, 11 7 3 1- O c t 7? 52 3 4 21 ,2 84 1, 29 7 $6 7, 42 4 4 1- S e p 7? 47 4 5 20 ,9 84 1, 29 5 $6 0, 86 4 5 1- S e p 7? 30 5 6 4, 01 9 ( 1, 78 0) - $5 3, 39 9 T ot a l 86 ,5 43 $2 16 ,5 63 77 was $16,563 and total biomass produced was 86,543 kg. The 7? stocking plan exceeded the mixed-size stocking plan in terms of net returns by $86,966 and in terms of biomass produced by 19,442 kg. For detailed enterprise budgets for each stocking plan, see Appendix Table 30 and Table 31. Objective 4 Discussion The purpose of this objective was to determine the most profitable production schedule for HSB FIPR systems using the developed Stella bio-economic model. However, it is recognized that many biological unknowns (educated guesses) are incorporated into this model so these specific results are not necessarily a production guide for producers to follow. This objective seeks to demonstrate one pragmatic way that the bio-economic model could be used to assist researchers and producers in making decisions about production. Knowledge of potentially higher net returns that could be achieved through choosing the most profitable fingerling size for a given restocking month is key to the development of a longer term production plan. The hypothesis that stocking 7? fingerlings for each crop of a five year production schedule beginning March 1 would be the most profitable fingerling size choice due to their high potential net return values over a shorter grow-out period was found to be true. For the initial March 1 stocking, the 7? fingerling was the optimal choice for highest net returns, highest net return per week and for shortest grow-out period. And for each subsequent crop stocking month, the 7? fingerling continued to be the best choice to maximize net return values because the restocking months occurred in the months where the 7? fingerlings was more profitable than the other fingerling choices, according to the results from objective three. 78 The results demonstrate that the initial fingerling size stocked can have great influence on net returns and profitability because of the variation in time required for grow-out. For the first production plan, five total crops were able to be harvested in five years, with a sixth crop started, because of the shortened amount of time required to grow-out the 7? fingerlings to market size. The mixed-size production plan, which began by stocking 3? fingerlings, was only able to have four completed crops in a five year period. By adding an additional harvest to the production schedule by initially stocking 7? fingerlings, and by avoiding over-wintering periods when the HSB are near harvest size, net returns and production over a five-year period were greater, indicating that time required for species specific fish grow-out is one of the most influential factors on net returns and production in the development of longer term production plans. The production plans developed for this objective are only two of many potential plans that could be tested with the bio-economic model. There are potentially many other plans that may increase net returns. The bio-economic model could also be used to make production decisions for unforeseen dynamics in HSB production. For example, if 7? fingerlings were not available to producers, or only certain sizes were available for restocking, the bio-economic model could be used to help make production decisions based on the fingerling size availability. Perhaps by waiting a few extra weeks for a different sized fingerling to be ready for purchase and stocking, a producer could increase his net return at harvest. The bio-economic model is a valuable tool that can help producers predict future returns for different stocking scenarios. Further tests using the bio-economic model would be beneficial. Once the growth component of the bio-economic model is refined through additional replicated research and/or field trials, additional alternative long-term production plans could be analyzed for profitability and production. A profit and production maximization schedule based on empirical data could 79 then be created to more accurately evaluate best strategies for stocking HSB in FIPR systems throughout the year. 80 Summary This research was directed toward identifying the bio-economic factors affecting the feasibility of floating in-pond raceway systems using a Stella bio-economic modeling approach. The developed bio-economic model was used to investigate factors related to density dependent mortality, stochastic loss events, fingerling and stocking date relationships and production and net return over a five year production horizon. In objective one, an important factor affecting the feasibility of the FIPR system was the relationship between stocking density and mortality. The shape of the mortality curve, either fixed, linear or exponential for increasing stocking density level, had a substantial impact on production and net return. It was assumed that mortality would increase when stocking density exceeded 300 fish/m3 and continue to increase up to 700 fish/m3. For the linear and exponential mortality curves, net returns were maximized at 500 fish/m3 but production was maximized at 700 fish/m3 and 600 fish/m3 respectively. Maximized production did not always correspond to maximized net returns. Future research should focus on the stocking density where density dependent mortality begins to have an increased effect and the shape of the mortality curve as stocking density levels increase. In objective two, the probability of a random catastrophic loss occurring was found to be as influential a factor as originally expected. By lowering the magnitude of a random disaster in the bio-economic model from 50% to 10% the feasibility of the FIPR system increased because at a 10% loss level average net returns were higher and the chance of negative net returns at harvest was eliminated. This showed the importance of mitigation equipment and the need for 81 their inclusion as an integral part of the production system from the beginning of production planning, and not after a loss event occurs. The timing of the stochastic event was also found to be an influential factor. When an event occurred early in the grow out period, net returns were not as low as with crops with stochastic losses occurring later. It is difficult to predict the probability of high loss events occurring under any circumstances as they are relatively rare and stochastic in nature, but field trials of the FIPR system could give us data that may help narrow the range of possibility surrounding the frequency and magnitude of losses from various types of loss events. For example, the time HSB can survive in a FIPR system when the airlift pump flow is interrupted or the percentage of fish that are able to survive a disease outbreak when a timely and effective treatment is applied. In objectives three and four, overwintering periods greatly affected net returns because of additional feed costs and, secondly, that fingerling size and stocking date combinations impacted the amount of economic loss for those overwintered fish. Fish overwintered at larger sizes were the most expensive to maintain, whereas fish overwintered at smaller sizes were the most cost effective. Over a five year period, by stocking larger fingerlings, the number of crops could be increased and resulting net returns were higher. Since the stocking month was varied (and therefore water temperatures fluctuated) between the scenarios, the results were greatly influenced by the bio-economic model?s water temperature dependent growth rate assumptions. Net return results are useful for comparing across objectives and to compare to other production systems. However, net returns are not the only indicators of feasibility; if high net returns can only be realized under very specific or highly improbable circumstances, the system may not truly be feasible. The results of this research indicate that high net returns can be realized under most of the circumstances examined except for the high magnitude of loss risk 82 scenarios. A few well managed HSB FIPR field trials that closely examine and evaluate stocking density dependent mortality, magnitudes and probability of stochastic loss and the relationship between growth and water temperature would enable this bio-economic model to be a useful tool in making production decisions for HSB production in FIPR systems. 83 References Alabama Aquaculture Best Management Practice (BMP). Managing cage culture systems. BPM No. 19. Brown, T.W., 2010. Intensive culture of channel catfish Ictalurus punctatus and hybrid catfish Ictalurus punctatus x Ictalurus furcatus in a commercial-scale, in-pond raceway system. Doctoral Dissertation. Department of Fisheries and Allied Aquacultures, Auburn University, Auburn, AL. Brown, T. W., Chappell, J. A., and Hanson, T. R. 2009. Commercial production of channel catfish and hybrid catfish utilizing an in-pond raceway system located in west alabama: An update. Proceedings of the World Aquaculture Society: Aquaculture America. Brune, D. E., G. Schwartz, A. G. Eversole, J. A. Collier, and T. E. Schwedler. 2003. Intensification of pond aquaculture and high rate photosynthetic systems. Aquacultural Engineering 28(1-2):65-86. Chappell, J. PhD. Professor Auburn University. 2009. Personal Communication. Costanza, R., and S. Gottlieb. 1998. Modelling ecological and economic systems with STELLA: Part II. Ecological Modelling 112(2-3):81-84. D'Abramo, L. R., and M. O. Frinsko. 2008. Hybrid striped bass: Pond production of food fish. Souther Regional Aquaculture Center (3):August 14, 2009. 84 D'Abramo, L. R., C. L. Ohs, and T. R. Hanson. 2004. Effect of stocking weight and stocking density on production of hybrid striped bass (sunshine) in earthen ponds in the second phase of a 2-phase system. Journal of the World Aquaculture Society 35(1):33. D'Abramo, L. R., C. L. Ohs, T. R. Hanson, and J. B. Taylor. 2002. Production and economic analysis of two-phase and three-phase culture of sunshine bass in earthen ponds. North American Journal of Aquaculture 64(2):103-112. Fullner, G., T. Gottschalk, and M. Pfeifer. 2007. Experiments for the production of hybrid striped bass in in-pond circulation systems. Aquaculture International 15(3):241-248. Hanson, T. R. 2009. PhD. Professor Auburn University. 2009. Personal Communication Hanson, T.R. and D. Sites. 2011. 2010 U.S Catfish Database. Fisheries and Allied Aquacultures Departmental Series Nol. 4, Alabama Agaricultural Experiment Station, Auburn University, Auburn, AL http://www.aaes.auburn.edu/comm/pubs/fisheries/fish_4.pdf Hartleb, C. 2004. Floating raceways used to raise yellow perch at cranberry farms. Aquaculture Magazine 30(1):18-24. Hodson, R. G. 1989. Hybrid striped bass: Biology and life history. Southern Regional Aquaculture Center No. 300. Hodson, R. G., and M. Hayes. 1989. Hybrid stiped bass: Pond production of foodfish. Southern Regional Aquaculture Center Publication No. 303. 85 Lazur, A. M., and D. C. Britt. 1997. Pond recirculating production systems. Southern Regional Aquaculture Center Publication No. 455:July 16, 2009. Lim, C., and C. D. Webster editors. 2006. Tilapia biology, culture, and nutrition. Food Products Press, Binghampton, NY. Masser, M. P. 1997. Cage culture: Handling and feeding caged fish. Southern Regional Aquaculture Center :July 15 2009. Masser, M. P., J. Jensen, and J. Crews. Channel catfish production in ponds. Alabama Cooperative Extension System ANR-195. Masser, M. P., and A. Lazur. 1997. In-pond raceways. Southern Regional Aquaculture Center Publication No. 170:April 14, 2009. Morrison, J. R., W. L. Deavours, J. C. Jones, and M. A. Tabb. 1995. Early rearing of channel catfish fry in floating raceways and subsequent survival in ponds. Progressive Fish- Culturist 57(4):292-296. Odom, D. FIPR Manager. 2009. Personal Communication. Volkman, E. T., C. C. Kohler, and S. T. Kohler. 2004. Assessment of floating vertical raceways for the culture of phase-II hybrid striped bass. North American Journal of Aquaculture 66(2):125-132. Yoo, K. H., M. P. Masser, and B. A. Hawcroft. 1995. An in-pond raceway system incorporating removal of fish wastes. Aquacultural Engineering 14:175. 86 Appendices 87 Table 21. Enterprise budget for one 65 m3 FIPR system HSB crop stocked at 300 fish/m3 stocking density in a 3.32 ha pond stocked on February 1 and harvested 36 weeks later. I t e m F i xe d L i n e ar E xp on e n t i al 91,119 91,119 91,119 2. V A R I A B L E C O S T S F i n g e r l i n g s 14,761 14,761 14,761 F e e d 16,180 16,180 16,180 E l e c t r i c i t y 654 654 654 A l l U n s pe c i f i e d C os t s : s u c h a s r e pa i r s a n d m a i n t e n a n c e 13,541 13,541 13,541 c h e m i c a l s , f u e l , l a bor , a n d i n t e r e s t on ope r a t i n g c a pi t a l T O T A L V A R I A B L E C O S T 45,136 45,136 45,136 3. I N C O M E A B O V E V A R I A B L E C O S T S 45,983 45,983 45,983 4. F I X E D C O S T S D e pr e c i a t i on on c a pi t a l i t e m s 605 605 605 D e pr e c i a t i on on m a c h i n e r y a n d e qu i pm e n t 2,305 2,305 2,305 I n t e r e s t , t a x e s a n d i n s u r a n c e 962 962 962 T O T A L F I X E D C O S T S 3,872 3,872 3,872 5. T O T A L C O S T O F A L L S P E C I F I E D E X P E N S E S 49,008 49,008 49,008 6. N E T R E T U R N S T O L A N D A bov e v a r i a bl e c os t s 45,983 45,983 45,983 A bov e t ot a l c os t s 42,111 42,111 42,111 7. B R E A K E V E N P R I C E , $/ k g A bov e v a r i a bl e c os t s 3.55 3.55 3.55 A bov e t ot a l c os t s 3.85 3.85 3.85 8. B R E A K E V E N Q U A N T I T Y , k g A bov e v a r i a bl e c os t s 6,339 6,339 6,339 A bov e t ot a l c os t s 6,883 6,883 6,883 1. G R O S S R E C E I P T S M or t al i t y C u r ve T yp e 88 Table 22. Enterprise budget for one 65 m3 FIPR system HSB crop stocked at 400 fish/m3 stocking density in a 3.32 ha pond stocked on February 1 and harvested 36 weeks later. I t e m F i xe d L i n e ar E xp on e n t i al 121,492 109,557 115,616 2. V A R I A B L E C O S T S F i n g e r l i n g s 19,428 19,428 19,428 F e e d 21,574 20,082 20,839 E l e c t r i c i t y 654 654 654 A l l U n s pe c i f i e d C os t s : s u c h a s r e pa i r s a n d m a i n t e n a n c e 17,852 17,213 17,538 c h e m i c a l s , f u e l , l a bor , a n d i n t e r e s t on ope r a t i n g c a pi t a l T O T A L V A R I A B L E C O S T 59,508 57,377 58,458 3. I N C O M E A B O V E V A R I A B L E C O S T S 61,984 52,181 57,158 4. F I X E D C O S T S D e pr e c i a t i on on c a pi t a l i t e m s 605 605 605 D e pr e c i a t i on on m a c h i n e r y a n d e qu i pm e n t 2,305 2,305 2,305 I n t e r e s t , t a x e s a n d i n s u r a n c e 962 962 962 T O T A L F I X E D C O S T S 3,872 3,872 3,872 5. T O T A L C O S T O F A L L S P E C I F I E D E X P E N S E S 63,380 61,249 62,330 6. N E T R E T U R N S T O L A N D A bov e v a r i a bl e c os t s 61,984 52,181 57,158 A bov e t ot a l c os t s 58,112 48,309 53,285 7. B R E A K E V E N P R I C E , $/ k g A bov e v a r i a bl e c os t s 3.51 3.75 3.62 A bov e t ot a l c os t s 3.74 4.00 3.86 8. B R E A K E V E N Q U A N T I T Y , k g A bov e v a r i a bl e c os t s 8,358 8,059 8,210 A bov e t ot a l c os t s 8,902 8,602 8,754 1. G R O S S R E C E I P T S M or t al i t y C u r ve T yp e 89 Table 23. Enterprise budget for one 65 m3 FIPR system HSB crop stocked at 500 fish/m3 stocking density in a 3.32 ha pond stocked on February 1 and harvested 36 weeks later. I t e m F i xe d L i n e ar E xp on e n t i al 151,865 122,029 133,655 2. V A R I A B L E C O S T S F i n g e r l i n g s 24,095 24,095 24,095 F e e d 26,967 23,238 24,691 E l e c t r i c i t y 654 654 654 A l l U n s pe c i f i e d C os t s : s u c h a s r e pa i r s a n d m a i n t e n a n c e 22,164 20,566 21,188 c h e m i c a l s , f u e l , l a bor , a n d i n t e r e s t on ope r a t i n g c a pi t a l T O T A L V A R I A B L E C O S T 73,880 68,552 70,628 3. I N C O M E A B O V E V A R I A B L E C O S T S 77,985 53,477 63,027 4. F I X E D C O S T S D e pr e c i a t i on on c a pi t a l i t e m s 605 605 605 D e pr e c i a t i on on m a c h i n e r y a n d e qu i pm e n t 2,305 2,305 2,305 I n t e r e s t , t a x e s a n d i n s u r a n c e 962 962 962 T O T A L F I X E D C O S T S 3,872 3,872 3,872 5. T O T A L C O S T O F A L L S P E C I F I E D E X P E N S E S 77,752 72,424 74,500 6. N E T R E T U R N S T O L A N D A bov e v a r i a bl e c os t s 77,985 53,477 63,027 A bov e t ot a l c os t s 74,113 49,605 59,155 7. B R E A K E V E N P R I C E , $/ k g A bov e v a r i a bl e c os t s 3.48 4.02 3.78 A bov e t ot a l c os t s 3.67 4.25 3.99 8. B R E A K E V E N Q U A N T I T Y , k g A bov e v a r i a bl e c os t s 10,376 9,628 9,920 A bov e t ot a l c os t s 10,920 10,172 10,463 1. G R O S S R E C E I P T S M or t al i t y C u r ve T yp e 90 Table 24. Enterprise budget for one 65 m3 FIPR system HSB crop stocked at 600 fish/m3 stocking density in a 3.32 ha pond stocked on February 1 and harvested 36 weeks later. I t e m F i xe d L i n e ar E xp on e n t i al 182,238 128,533 140,576 2. V A R I A B L E C O S T S F i n g e r l i n g s 28,762 28,762 28,762 F e e d 32,360 25,648 27,153 E l e c t r i c i t y 654 654 654 A l l U n s pe c i f i e d C os t s : s u c h a s r e pa i r s a n d m a i n t e n a n c e 26,475 23,599 24,244 c h e m i c a l s , f u e l , l a bor , a n d i n t e r e s t on ope r a t i n g c a pi t a l T O T A L V A R I A B L E C O S T 88,251 78,662 80,813 3. I N C O M E A B O V E V A R I A B L E C O S T S 93,986 49,871 59,764 4. F I X E D C O S T S D e pr e c i a t i on on c a pi t a l i t e m s 605 605 605 D e pr e c i a t i on on m a c h i n e r y a n d e qu i pm e n t 2,305 2,305 2,305 I n t e r e s t , t a x e s a n d i n s u r a n c e 962 962 962 T O T A L F I X E D C O S T S 3,872 3,872 3,872 5. T O T A L C O S T O F A L L S P E C I F I E D E X P E N S E S 92,124 82,534 84,685 6. N E T R E T U R N S T O L A N D A bov e v a r i a bl e c os t s 93,986 49,871 59,764 A bov e t ot a l c os t s 90,114 45,999 55,892 7. B R E A K E V E N P R I C E , $/ k g A bov e v a r i a bl e c os t s 3.47 4.38 4.12 A bov e t ot a l c os t s 3.62 4.60 4.31 8. B R E A K E V E N Q U A N T I T Y , k g A bov e v a r i a bl e c os t s 12,395 11,048 11,350 A bov e t ot a l c os t s 12,939 11,592 11,894 1. G R O S S R E C E I P T S M or t al i t y C u r ve T yp e 91 Table 25. Enterprise budget for one 65 m3 FIPR system HSB crop stocked at 700 fish/m3 stocking density in a 3.32 ha pond stocked on February 1 and harvested 36 weeks later. I t e m F i xe d L i n e ar E xp on e n t i al 212,611 129,069 129,069 2. V A R I A B L E C O S T S F i n g e r l i n g s 33,429 33,429 33,429 F e e d 37,754 27,312 27,312 E l e c t r i c i t y 654 654 654 A l l U n s pe c i f i e d C os t s : s u c h a s r e pa i r s a n d m a i n t e n a n c e 30,787 26,312 26,312 c h e m i c a l s , f u e l , l a bor , a n d i n t e r e s t on ope r a t i n g c a pi t a l T O T A L V A R I A B L E C O S T 102,623 87,706 87,706 3. I N C O M E A B O V E V A R I A B L E C O S T S 109,987 41,363 41,363 4. F I X E D C O S T S D e pr e c i a t i on on c a pi t a l i t e m s 605 605 605 D e pr e c i a t i on on m a c h i n e r y a n d e qu i pm e n t 2,305 2,305 2,305 I n t e r e s t , t a x e s a n d i n s u r a n c e 962 962 962 T O T A L F I X E D C O S T S 3,872 3,872 3,872 4. T O T A L C O S T O F A L L S P E C I F I E D E X P E N S E S 106,496 91,579 91,579 6. N E T R E T U R N S T O L A N D A bov e v a r i a bl e c os t s 109,987 41,363 41,363 A bov e t ot a l c os t s 106,115 37,491 37,491 7. B R E A K E V E N P R I C E , $/ k g A bov e v a r i a bl e c os t s 3.46 4.87 4.87 A bov e t ot a l c os t s 3.59 5.08 5.08 8. B R E A K E V E N Q U A N T I T Y , k g A bov e v a r i a bl e c os t s 14,413 12,318 12,318 A bov e t ot a l c os t s 14,957 12,862 12,862 1. G R O S S R E C E I P T S M or t al i t y C u r ve T yp e 92 Table 26. Enterprise budget for one FIPR system HSB crop initially stocked at 500 fish/m3 in a 3.32 ha pond stocked on Feb 1 and harvested 36 weeks later. Catastrophic loss event occurring at indicated week with 50% of HSB removed at loss week. I t e m E ar l y ( 5 w k s ) M i d d l e ( 18 w k s ) L at e ( 32 w k s ) N o L os s 70,438 72,984 80,362 151,865 2. V A R I A B L E C O S T S F i n g e r l i n g s 24,095 24,095 24,095 24,095 F e e d 13,438 16,215 23,730 26,967 E l e c t r i c i t y 654 654 654 654 A l l U n s pe c i f i e d C os t s : s u c h a s r e pa i r s a n d m a i n t e n a n c e 16,366 17,556 20,777 22,164 c h e m i c a l s , f u e l , l a bor , a n d i n t e r e s t on ope r a t i n g c a pi t a l T O T A L V A R I A B L E C O S T 54,552 58,520 69,255 73,880 3. I N C O M E A B O V E V A R I A B L E C O S T S 15,886 14,465 11,107 77,985 4. F I X E D C O S T S D e pr e c i a t i on on c a pi t a l i t e m s 605 605 605 605 D e pr e c i a t i on on m a c h i n e r y a n d e qu i pm e n t 2,305 2,305 2,305 2,305 I n t e r e s t , t a x e s a n d i n s u r a n c e 962 962 962 962 T O T A L F I X E D C O S T S 3,872 3,872 3,872 3,872 5. T O T A L C O S T O F A L L S P E C I F I E D E X P E N S E S 58,424 62,392 73,127 77,752 6. N E T R E T U R N S T O L A N D A bov e v a r i a bl e c os t s 15,886 14,465 11,107 77,985 A bov e t ot a l c os t s 12,013 10,593 7,235 74,113 7. B R E A K E V E N P R I C E , $/ k g A bov e v a r i a bl e c os t s 5.55 5.74 6.17 3.48 A bov e t ot a l c os t s 5.94 6.12 6.52 3.67 8. B R E A K E V E N Q U A N T I T Y , k g A bov e v a r i a bl e c os t s 7,662 8,219 9,727 10,376 A bov e t ot a l c os t s 8,206 8,763 10,271 10,920 1. G R O S S R E C E I P T S W e e k of L os s 93 Table 27. Enterprise budget for one FIPR system HSB crop initially stocked at 500 fish/m3 in a 3.32 ha pond stocked on Feb 1 and harvested 36 weeks later. Catastrophic loss event occurring at indicated week with 10% of HSB removed at loss week. I t e m E ar l y ( 1 w k s ) M i d d l e ( 19 w k s ) L at e ( 31 w k s ) N o L os s 133,736 134,824 136,157 151,865 2. V A R I A B L E C O S T S F i n g e r l i n g s 24,095 24,095 24,095 24,095 F e e d 23,937 24,719 25,982 26,967 E l e c t r i c i t y 654 654 654 654 A l l U n s pe c i f i e d C os t s : s u c h a s r e pa i r s a n d m a i n t e n a n c e 20,865 21,200 21,742 22,164 c h e m i c a l s , f u e l , l a bor , a n d i n t e r e s t on ope r a t i n g c a pi t a l T O T A L V A R I A B L E C O S T 69,551 70,668 72,472 73,880 3. I N C O M E A B O V E V A R I A B L E C O S T S 64,185 64,156 63,685 77,985 4. F I X E D C O S T S D e pr e c i a t i on on c a pi t a l i t e m s 605 605 605 605 D e pr e c i a t i on on m a c h i n e r y a n d e qu i pm e n t 2,305 2,305 2,305 2,305 I n t e r e s t , t a x e s a n d i n s u r a n c e 962 962 962 962 T O T A L F I X E D C O S T S 3,872 3,872 3,872 3,872 5. T O T A L C O S T O F A L L S P E C I F I E D E X P E N S E S 73,423 74,540 76,344 77,752 6. N E T R E T U R N S T O L A N D A bov e v a r i a bl e c os t s 64,185 64,156 63,685 77,985 A bov e t ot a l c os t s 60,313 60,284 59,813 74,113 7. B R E A K E V E N P R I C E , $/ k g A bov e v a r i a bl e c os t s 3.72 3.75 3.81 3.48 A bov e t ot a l c os t s 3.93 3.96 4.01 3.67 8. B R E A K E V E N Q U A N T I T Y , k g A bov e v a r i a bl e c os t s 9,768 9,925 10,179 10,376 A bov e t ot a l c os t s 10,312 10,469 10,723 10,920 W e e k of L os s 1. G R O S S R E C E I P T S 94 Table 28. Production, receipts, total costs and net return for 50% magnitude of loss scenarios. Table 29. Production, receipts, total costs and net return for 10% magnitude of loss scenarios. E a r l y ( 5 w k s ) M i d d l e ( 1 8 w k s ) L a t e ( 3 2 w k s ) N o L o s s P r o d u c t i o n ( k g ) 9 ,8 9 3 1 0 ,2 5 1 1 1 ,2 8 7 2 1 ,3 2 9 R e c i e p t s ( $ ) 7 0 ,4 3 8 7 2 ,9 8 4 8 0 ,3 6 2 1 5 1 ,8 6 5 T o t a l C o s t s ( $ ) 5 8 ,4 2 4 6 2 ,3 9 2 7 3 ,1 2 7 7 7 ,7 5 2 N e t R e t u r n ( $ ) 1 2 ,0 1 3 1 0 ,5 9 3 7 ,2 3 5 7 4 ,1 1 3 W e e k o f L o s s E a r l y ( 5 w k s ) M i d d l e ( 1 8 w k s ) L a t e ( 3 2 w k s ) N o L o s s P r o d u c t i o n ( k g ) 1 8 ,7 8 3 1 8 ,9 3 6 1 9 ,1 2 3 2 1 ,3 2 9 R e c i e p t s ( $ ) 1 3 3 ,7 3 6 1 3 4 ,8 2 4 1 3 6 ,1 5 7 1 5 1 ,8 6 5 T o t a l C o s t s ( $ ) 7 3 ,4 2 3 7 4 ,5 4 0 7 6 ,3 4 4 7 7 ,7 5 2 N e t R e t u r n ( $ ) 6 0 ,3 1 3 6 0 ,2 8 4 5 9 ,8 1 3 7 4 ,1 1 3 W e e k o f L o s s 95 Table 30. Enterprise budget for a five year production strategy stocking only 7? fingerlings in one 65 m3 FIPR system HSB crop stocked at 500 fish/m3 stocking density in a 3.32 ha pond and first stocking on March 1. I t e m #1 M ar #2 N ov #3 O c t #4 N ov #5 O c t #6 N ov S U M 148,322 151,564 152,393 151,564 152,393 0 756,236 2. V A R I A B L E C O S T S F i n g e r l i n g s 24,095 24,095 24,095 24,095 24,095 24,095 144,570 F e e d 25,325 29,356 30,524 29,356 30,524 3,557 148,643 E l e c t r i c i t y 581 890 944 890 944 291 4,540 A l l U n s pe c i f i e d C os t s : s u c h a s r e pa i r s a n d m a i n t e n a n c e 21,429 23,289 23,813 23,289 23,813 11,975 127,609 c h e m i c a l s , f u e l , l a bor , a n d i n t e r e s t on ope r a t i n g c a pi t a l T O T A L V A R I A B L E C O S T 71,430 77,630 79,377 77,630 79,377 39,918 425,362 3. I N C O M E A B O V E V A R I A B L E C O S T S 76,892 73,933 73,017 73,933 73,017 - 39,9 18 330,874 4. F I X E D C O S T S D e pr e c i a t i on on c a pi t a l i t e m s 544 833 884 833 884 136 4,114 D e pr e c i a t i on on m a c h i n e r y a n d e qu i pm e n t 2,048 3,136 3,328 3,136 3,328 512 15,488 I n t e r e s t , t a x e s a n d i n s u r a n c e 850 1,301 1,381 1,301 1,381 1,073 7,287 T O T A L F I X E D C O S T S 3,442 5,270 5,593 5,270 5,593 1,721 26,889 5. T O T A L C O S T O F A L L S P E C I F I E D E X P E N S E S 74,872 82,901 84,970 82,901 84,970 41,639 452,251 6. N E T R E T U R N S T O L A N D A bov e v a r i a bl e c os t s 76,892 73,933 73,017 73,933 73,017 - 39,9 18 330,874 A bov e t ot a l c os t s 73,450 68,663 67,424 68,663 67,424 - 41,6 39 303,985 7. B R E A K E V E N P R I C E , $/ k g A bov e v a r i a bl e c os t s 3.33 3.33 3.33 3.33 15.06 0.58 A bov e t ot a l c os t s 3.49 3.56 3.57 3.56 16.12 0.61 8. B R E A K E V E N Q U A N T I T Y , k g A bov e v a r i a bl e c os t s 10,032 10,903 11,148 10,903 11,148 5,606 59,742 A bov e t ot a l c os t s 10,516 11,643 11,934 11,643 11,934 5,848 63,518 1. G R O S S R E C E I P T S C r op N u m b e r an d S t ar t D at e 96 Table 31. Enterprise budget for a five year production strategy initially stocking a 3? fingerling (Years 1 and 2) and subsequently stocking a 7? fingerling (Years 3, 4 and 5) in one 65 m3 FIPR system HSB crop stocked at 500 fish/m3 stocking density in a 3.32 ha pond and first stocking on April 1. I t e m #1 A p r #2 J u l y #3 O c t #4 Se p #5 Se p S U M 145,389 142,843 152,393 150,247 0 590,873 2. V A R I A B L E C O S T S F i n g e r l i n g s 9,990 9,990 24,095 24,095 24,095 92,265 F e e d 40,612 30,038 30,524 33,614 8,575 143,363 E l e c t r i c i t y 1,162 1,162 944 944 454 4,667 A l l U n s pe c i f i e d C os t s : s u c h a s r e pa i r s a n d m a i n t e n a n c e 22,185 17,653 23,813 25,137 14,196 102,983 c h e m i c a l s , f u e l , l a bor , a n d i n t e r e s t on ope r a t i n g c a pi t a l T O T A L V A R I A B L E C O S T 73,949 58,843 79,377 83,790 47,320 343,278 3. I N C O M E A B O V E V A R I A B L E C O S T S 71,440 84,000 73,017 66,457 - 47,3 20 247,594 4. F I X E D C O S T S D e pr e c i a t i on on c a pi t a l i t e m s 1,088 1,088 833 799 272 4,080 D e pr e c i a t i on on m a c h i n e r y a n d e qu i pm e n t 4,096 4,096 3,136 3,008 1,024 15,360 I n t e r e s t , t a x e s a n d i n s u r a n c e 1,700 1,700 1,624 1,786 1,393 8,202 T O T A L F I X E D C O S T S 6,884 6,884 5,593 5,593 2,689 27,642 5. T O T A L C O S T O F A L L S P E C I F I E D E X P E N S E S 80,833 65,727 84,970 89,383 50,009 370,921 6. N E T R E T U R N S T O L A N D A bov e v a r i a bl e c os t s 71,440 84,000 73,017 66,457 - 47,3 20 247,594 A bov e t ot a l c os t s 64,557 77,117 67,424 60,864 - 50,0 09 219,952 7. B R E A K E V E N P R I C E , $/ k g A bov e v a r i a bl e c os t s 3.33 3.33 3.33 3.33 8.46 A bov e t ot a l c os t s 3.64 3.72 3.57 3.56 8.94 8. B R E A K E V E N Q U A N T I T Y , k g A bov e v a r i a bl e c os t s 10,386 8,264 11,148 11,768 6,646 48,213 A bov e t ot a l c os t s 11,353 9,231 11,934 12,554 7,024 52,096 1. G R O S S R E C E I P T S