Citation
Bioeconomic modelling of hard clam growout in Florida

Material Information

Title:
Bioeconomic modelling of hard clam growout in Florida the replacement decision
Creator:
Holiman, Stephen Glenn, 1955-
Publication Date:
Language:
English
Physical Description:
vi, 157 leaves : ill., col. photos ; 29 cm.

Subjects

Subjects / Keywords:
Clam culture -- Florida ( lcsh )
Shellfish culture -- Florida ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1993.
Bibliography:
Includes bibliographical references (leaves 153-155).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Stephen Glenn Holiman.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright Stephen Glenn Holiman. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
030445615 ( ALEPH )
31303697 ( OCLC )

Downloads

This item has the following downloads:


Full Text
BIOECONOMIC MODELLING OF HARD CLAM GROWOUT
IN FLORIDA: THE REPLACEMENT DECISION
By
STEPHEN GLENN HOLIMAN
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1993




ACKNOWLEDGEMENTS
I would like to thank Dr. Thomas Spreen for his patient guidance and nurturing those many times when I began to lose focus and direction. Thanks go to Dr. Eric Thunberg for the financial support, the enthusiasm and the patience during the lost summer of family medical problems. Thanks also go out to Dr. Charles Adams for his office wit and assistance with the clam budgets, Dr. William Boggess and Dr. Richard Weldon for their patient response to financial questions, and to Dr. Charles Cichra for bearing up under numerous "well, economists do it this way."
The crew at Project Ocean, notably Dr. David Vaughan and Leslie Sturmer, also deserve thanks for various things not the least of which were those 100-odd deliciously sweet clams my wife and I got to eat. A huge thank-you also goes out to the anonymous commercial clam farm which graciously provided the growth data.
The biggest thanks, however, must go to my wife and parents. My wife was as patient as she could be given all that I have put her through. My father, Stanley M. Holiman, while not quite understanding the Ph.D. process, was always there for support. My biggest regret is that my mother, Bonnie Gene Holiman, could not be here to see it all complete. She claimed she did not want me to finish because then we would have to move away. But I know that both she and my dad were proud of what I was doing and that her spirit is still here with us, even as I write this. Thank you all.
ii




TABLE OF CONTENTS
pag.e
ACKNOWLEDGEMENTS...................................... i
ABSTRACT................................................ v
CHAPTER
1INTRODUCTION.................................. 1
The Management Problem: Replacement Under
Risk...................................... 4
Research Objectives............................ 6
Organization of the Dissertation.................... 7
2 HARD CLAM AQUACULTURE........................ 8
Introduction.................................. 8
Hatchery Phase............................... 8
Nursery Phase................................ 13
Growout Phase............................... 20
Harvest.................................... 21
3 BIOECONOMIC MODEL............................ 23
Economic Model.............................. 23
Hard Clam Growth and Mortality Models.............. 30
Bivalve Growth Literature...................... 31
Growth Model............................... 33
Estimation of Mortality........................ 40
4 RESULTS AND DISCUSSION......................... 43
Growth Simulation............................. 43
Mortality................................... 48
Economic Evaluation........................... 50
Single Plant Production Optimal Rotations............. 51
Multiplot Production Optimal Rotations............... 59
Production Performance Comparison................ 70
Economic Performance........................ 75
iii




Income Statements .. .. .. .. .. .. .. ... .. ... ...77
Net Present and Annualized Values. .. .. .. .. .. .. ..79
5 SUMMARY AND CONCLUSIONS...................... 91
Summary................................... 91
Conclusions and Implications...................... 93
Limitations and Suggestions For Further
Research.................................. 98
APPENDIX
1 PRODUCTION AND ENVIRONMENTAL DATA USED IN
GROWTH MODEL ESTIMATION................ 102
2 ENVIRONMENTAL VALUES USED IN GROWTH
SIMULATION ............................. 147
3 HARD CLAM EXPECTED MEAN PRICES................ 150
4 INITIAL INVESTMENT REQUIREMENTS FOR HARD CLAM
BOTT~OM BAG GROWOUT..................... 151
5 PRODUCTION COSTS FOR HARD CLAM BOTTOM
BAG GROWOUT............................ 152
REFERENCES............................................. 153
BIOGRAPHICAL SKETCH.................................... 156
iv




Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
BIGECONOMIC MODELLING OF HARD CLAM GROWOUT
IN FLORIDA: THE REPLACEMENT DECISION By
Stephen Glenn Holiman
December 1993
Chairman: Dr. Thomas H. Spreen
Major Department: Food and Resource Economics Department
The choice of production method and timing are examined to determine their effect on optimal growout and replacement scheduling of hard clams given probabilistic clam growth and variable monthly clam prices. Hard clam growth is modelled as a function of initial clam size and water temperature, salinity and dissolved oxygen. The environmental parameters take on probabilistic values, hence determining probabilistic growth. Growth function coefficients are estimated for 2 clam planting densities and the resultant functions used to simulate clam growth of 10 and 15 millimeter clam seed planted at 62.5 and 75 clams per square foot over a 34-month period beginning in each of the 12 calendar months. Growth simulation results are then combined with mortality assumptions, production restrictions (lease size and planting capacity) and size-dependent price expectations to estimate expected returns. These returns are then used to determine the best production method (seed size and planting density), planting and replacement
v




schedule and maximum returns. The expected returns are then used to estimate the net present value of the lease.
Results indicate that, where possible, growers should purchase larger seed and plant at the higher density and that, regardless of production method used, plant and harvest scheduling requires special attention. Growout times, replacement schedules and expected revenues vary by site location as determined by the specific environmental conditions of each site. The returns indicate that current lease fee requirements substantially undervalue the productive potential of the lease. The potential returns of hard clam growout are site specific, thus suggesting the importance of site selection and suggesting variable lease fees dependent upon site potential. Justifying such a fee structure, however, requires additional research on identifying the specific environmental profiles that generate different economic potential.
vi




CHAPTER 1
INTRODUCTION
Aquaculture is increasingly seen as a means of augmenting the supply of commercially important aquatic species. Under certain conditions, aquaculture may possess a comparative advantage over wild harvest (Shang, 1981). Wild populations may be widely dispersed due to natural ecological dynamics or as a result of harvest pressure. Dispersion affects per unit harvest costs by increasing search time, labor requirements, fuel use, etc. Aquaculture may produce lower per unit costs by concentrating the target species in a confined and more accessible location. Genetic selection combined with controlled feeding and environmental conditions can improve yields relative to natural production. Aquaculture can also allow suppliers to mitigate seasonally fluctuating wild catch and guarantee delivery with greater certainty than when dependent upon wild harvest.
The Florida hard clam aquaculture industry is an example of an emerging aquaculture industry. Two species of hard clam are native to Florida, Mercenaria mercenaria, the northern hard clam, and Mercenaria campechiensis, the southern hard clam (Vaughan et al., 1988). Natural territories of the two species overlap. M. mercenaria is found from the Gulf of St. Lawrence, Canada, to the northern Gulf of Mexico, with the center of abundance from Massachusetts to Virginia. M. campechiensis
1




2
is found from Cape May, New Jersey, to Campeche, Mexico, with the center of abundance in southwest Florida. Some hybridization occurs where the species overlap, but the two species typically prefer different habitats, with M. mercenaria being an estuarine inter- to subtidal species and M. campechiensis preferring deeper, higher salinity waters (Malouf and Bricelj, 1989). Commercially exploited populations are typically M. mercenaria, as this is the more abundant species and it lives in more accessible waters. Hard clams can live for 23 years or more and achieve a length in excess of 135 millimeters (mm) (Malouf and Bricelj, 1989).
Wild hard clams have historically been found in Florida waters, but large scale harvest and culture have typically been confined to mid-Atlantic and north Atlantic coastal regions (Manzi and Castagna, 1989). From 1973 to 1983, annual Florida hard clam landings from wild stocks averaged 107.54 thousand pounds of meat compared to total U.S. average annual landings of 14.362 million pounds (Adams et al., 1991). See Table 1-1. In the early 1980s, however, a large natural set of clams in the Indian River Lagoon resulted in average annual harvests of 1.366 million pounds of clam meat from 1984 through 1987 (Adams et al., 1991). Total annual U.S. landings over the same period averaged 13.664 million pounds of meat. This represented an increase in Florida landings as a percentage of total U.S. landings from less than 1% during 1974-83 to almost 10% during 1984-87. Harvests eventually declined to 711 thousand pounds in 1988, however, due to a combination of harvest pressure and changing environmental conditions. Total U.S. hard clam landings in 1988 were 12.371 million pounds of meat,




3
Table 1-1. Hard Clam Landings, 1973-88.
Share
Year Florida U.S. of U.S.
--- Thousand pounds of meat--- Percent
1973 139 14,505 0.95
1974 94 14,665 0.64
1975 74 14,995 0.49
1976 61 15,251 0.40
1977 148 14,690 1.00
1978 126 13,295 0.94
1979 72 12,058 0.59
1980 62 13,370 0.46
1981 117 18,118 0.64
1982 145 12,855 1.12
1983 145 14,186 1.02
1984 1377 14,749 9.33
1985 1441 16,697 8.63
1986 1448 11,793 12.28
1987 1197 11,418 10.48
1988 711 12,371 5.47
Source: 1973-84, NMFS; 1985-88, FDNR.




4
with Florida production comprising 5.7% of the total. Florida's wild clam harvest has not rebounded to mid-1980 levels.
The ability of Florida waters to support such large wild harvests and the ability of regional markets to absorb the harvests caused many to consider the potential of culturing clams in Florida. Particular note was given to the suitability of Florida's natural environmental conditions relative to hard clam production. Clam growth is temperature sensitive. Lower water temperatures depress growth while the converse is true, to a point, for warmer water temperatures (Manzi and Castagna, 1989). An accelerated growth rate should allow increased clam biomass production per unit of area per unit of time and, consequently, increased revenues.
The Management Problem: Replacement Under Risk
The decision to undertake hard clam aquaculture requires the evaluation of numerous issues. As will be described in Chapter 2, hard clam aquaculture is an integrated process consisting of hatchery, nursery and growout phases. A commercial culturist may choose to operate at all levels of production or decide to specialize on a particular level. Hence, a choice of level of integration must be made. Similarly, multiple production technologies exist for each production stage, so a technology adoption decision is required. Next, for each choice of production technology, options exist relative to production scheduling, size of clam seed, planting densities, monitoring schedules, etc. Further, decisions are required on marketing strategy. Specific strategies




5
require decisions concerning whether to target seasonal markets, what size clams to market, and where to market.
As will become evident in Chapter 2, the productive capacity, capital requirements and labor intensity of the hatchery and nursery phases of hard clam aquaculture make them less suitable than growout culture for operations owned, managed, and worked by a single individual. The research presented in this dissertation focuses on growout culture. Management options for hard clam growout include the selection of seed size, choice of planting method, clam density and replacement scheduling.
The evaluation of management options is complicated where risk is encountered. Risk is inherent in any system where outcomes are not guaranteed. This applies to both production and financial outcomes. Stochastic production may result in variable output quality and quantity. Variable output produces variable revenue. Stochastic prices further increase the variability of revenue. Production uncertainty has particular relevance in hard clam aquaculture as prices are higher in the smaller legal size categories and lower for larger sized clams. Consumer preference results in a price structure where a price penalty is incurred for larger size classes. Hence, the culturist is concerned that his clams not grow out of the higher-priced size classes.
The culturist must determine where risk is likely to occur. Then, risk reducing options must be identified. The adoption of specific options will then be dependent on the relative costs of implementation and the subsequent rewards of the reduced risk.




6
Research on the costs and returns of hard clam growout aquaculture has been limited to systems and growing conditions representative of the South Atlantic region (Adams et al., 1991). No comprehensive work has been conducted on systems and conditions specific to Florida. Current studies (Adams et al., 1991; Thunberg and Adams, 1990) do not incorporate clam price and yield variability other than through basic sensitivity analysis. Sensitivity analysis simply changes an outcome without examining the likelihood of that outcome actually occurring. The true consideration of risk reflects both its impact and the likelihood of occurrence.
A decision to undertake hard clam growout aquaculture in Florida, therefore, requires knowledge of appropriate production systems, sources of operation risk, the effects of risk on operation design and replacement scheduling, and estimates of the costs and returns of hard clam growout under risk.
Research Objectives
The objectives of this research are
1. Develop a bioeconomic model of Florida hard clam growout.
2. Generate cost and return estimates of hard clam growout under different scenarios
of growth and price variability to provide insights to optimal operation design and
management.




7
Organization of the Dissertation
Chapter 2 provides a description of hard clam aquaculture in Florida. Theoretical consideration, data sources, and the bioeconomic model are presented and discussed in Chapter 3. The results of the application of the bioeconomic model are given in Chapter 4. A summary, conclusions, study limitations and suggestions for future research comprise Chapter 5.




CHAPTER 2
HARD CLAM AQUACULTURE
Introduction
Hard clam aquaculture consists of three phases--hatchery, nursery and growout (Manzi and Castagna, 1989). Only the first two phases typically entail controlled environments where specific growing conditions are maintained. In the following discussion, current practices and key issues for Florida hard clam aquaculture are described.
Hatchery Phase
It has been estimated that 40% of hatchery operating costs are for the production of algae or one-celled plants (phytoplankters) for hard clam food (Hartman, 1989). Two primary methods of algal culture, the Glancy and Milford methods, are used for hard clam aquaculture (Castagna and Manzi, 1989). The Glancy Method filters or clarifies seawater to remove predatory zooplankters and large phytoplankters. Treated seawater is kept in shallow, gently aerated tanks, exposed to natural or artificial light. The algae is normally fed to the clams within 48 hours before larger, less digestible phytoplankters
8




9
dominate. The Glancy Method works best in moderate climates and where an abundance of natural phytoplankton exists.
The Milford Method relies on the controlled production of selected species of phytoplankters using sterile media and growth promotents. Pure cultures of single algal species are produced. Total harvest and replacement of the algae is practiced to prevent contamination.
Since seawater is the medium for both clam and algal growth, the success of a hard clam hatchery is highly dependent upon the availability of water of a suitable quality. The variables of specific concern to the culturist are water temperature, salinity, dissolved oxygen, chemical or bacterial contamination, and algal and zooplankton content. Water quality can be manipulated by the culturist, but it may not be costeffective to do so. Larval rearing requires water temperatures of 25-300 C (Adams et al., 1991), salinity of 26-27 parts per thousand (ppt) (Eversole, 1987), and dissolved oxygen levels of 6.8-7.4 milligrams per liter (Eversole, 1987). Larval growth is fastest at 300 C, a temperature which also promotes high bacterial contamination (Menzel, 1989). Proper salinity and dissolved oxygen levels are more critical for larval and juvenile clams than for older clams. Older clams are able to remain closed for longer periods and rely on various metabolic mechanisms to reduce oxygen requirements during periods of environmental stress (Eversole, 1987).
In the hatchery phase, sexually mature hard clams are induced to spawn and produce fertilized eggs. Broodstock are initially selected from wild stock possessing desired characteristics such as large size, or of a special color form or marking pattern




10
called notata. Notata markings are brown zig zag patterns in the shell as pictured in Figure 2-1. Notata patterns are usually present in only 1 % of wild populations and are used as a means of cultured product identification (Vaughan et al., 1988). The presence of notata markings can also be used as a marketing tool to help consumers distinguish cultured from wild clams. Also, the use of notata markings discourages poaching as large numbers of notata clams are an indication of cultured origins. The benefits of notata breeding may be temporary, though, as escape and breeding by cultured clams increases the presence of notata markings in wild clam stocks.
Figure 2-1. Hard clam showing notata pattern.
Spawning is induced by thermal shock, a process of alternatively raising and lowering the water temperature (Hartman, 1989). Broodstock are placed on a spawning table containing 3-4 inches of clean seawater at 200 C and left undisturbed until all are open and actively siphoning water. The water temperature is then gradually raised to 300




11
C, left for 30 minutes and then lowered to 200 C for an additional 30 minutes. This process is repeated until all clams spawn. Gonadal material from sacrificed adult clams may be added to the water to further induce reluctant spawners. Egg production ranges from 2-30 million eggs per female (Hartman, 1989).
The fertilized eggs are placed in cone-shaped fiberglass or plastic containers of clean seawater. Within 24 hours of fertilization, the clam larvae, also known as veligers, develop shells and swim freely (Vaughan et al., 1988). Clam larvae do not actively feed for the first 48 hours after fertilization (Castagna and Manzi, 1989) and are not usually disturbed during this period.
A popular device for rearing clam larvae is the downweller (Castagna and Manzi, 1989). A downweller is a plastic or fiberglass cylinder with an open top and a sievecovered bottom as pictured in Figure 2-2. Several downwellers are placed in a large fiberglass reservoir filled with clean seawater. The top of the downweller extends above the reservoir waterline. Water flows into the top of each downweller through individual pipes and flows out through the bottom sieves. This system allows the free-swimming larvae to remain in the water column in contact with higher quality food and away from smothering sediment and sick or dead larvae. The larvae are sieved every two days. Sieving allows for the removal of dead larvae and contaminants and permits counting and size sorting.
Predation and fouling are problems that plague hard clam aquaculture from the earliest stages through harvest. Predation is the consumption of clams by other animals and fouling is the build-up of living organisms on the culture equipment, resulting in




12
i nf I ow
cu I turereservoir cylinder
s i eve
outflow
Figure 2-2. Downweller. smothering, impeded water flow and reduced food access. As mentioned previously, hatchery seawater is filtered, clarified or sterilized to remove predatory zooplankters. Fouling is controlled through reducing water contaminant content, frequent water changes and regular equipment cleaning.
Between 8 and 14 days after fertilization, the clam larvae develop a muscular foot and reach the final larval or pediveliger stage (Vaughan et al., 1988). The larvae lose the ability to swim, but remain mobile through the use of their foot. The clam larvae are called set or post-set and enter the nursery phase.




13
Nursery Phase
The goal of the nursery phase is the production of sufficient quantities of seed for final growout. Most growout methods require 7-10 millimeter (mm) or larger seed (measured along the longest axis) (Manzi and Castagna, 1989). Growth rates vary with clam genetics, culture method and growing conditions. Post-set can be expected to reach 1-3 mm at three months, 3-9 mm at six months and 10-15 mm at nine months (Vaughan et al., 1988).
Both onshore and field nursery systems exist. Onshore systems provide the greatest amount of access, control and predator protection, but may do so at considerable land and facility costs. Ambient temperature seawater is used for onshore systems. Water temperature adjustments may be made, however, to maintain optimal temperatures of 20-28' C (Vaughan et al., 1989). Food requirements may be met solely by natural seawater or through a combination of natural sources and cultured supplements. Field nursery systems are generally cheaper, but allow for less access and control as the clams are exposed to ambient water temperatures, algal content, etc. Field nursery systems are preferred by Florida hard clam culturists (Vaughan and Cresswell, 1989).
A substantial degree of control is required in the nursery phase, especially to reduce predation. The costs of seed production and the subsequent value of clam seed make high survival and rapid growth economic imperatives. Land-based nursery systems use methods similar to those used by hatcheries to control predation. Field systems require a different approach and typically rely on some type of physical barrier between young clams and predators. Dominant hard clam predators in Florida include rays




14
(Dasyatis spp. and Gymnura micruva), sheepshead (Archosargus probatocephalus), blue crabs (Callinectes spp.), and various mollusc species (Melongena corona, Fasciolaria spp., Euleura cadata and Thais haemostoma) (Vaughan and Cresswell, 1989). Unprotected hard clam plots in a Florida and Georgia study suffered 100% mortality, of which, 90% was attributed to blue crabs (Eversole, 1987). Vulnerability to predation is inversely proportional to age, as young clams lack the size or shell thickness to prevent crushing, opening or boring by predators.
Fouling also requires special attention. Major fouling organisms in Florida are sponges (Cliona spp., Haliclonia spp., and Halochondria spp.), sea squirts (Molgula occidentalis and Styela plicata), hydroids (Obelia spp.), barnacles (Balanus spp.), algae (Gracilaria spp.), and various mollusc species (Crepidula fornicata, Crassotrea virginica, C. rhizophorae, Modiolus spp. and Branchiodontes spp.) (Vaughan and Cresswell, 1989). Control methods vary with the culture method used and include various combinations of scrubbing, sun drying and turning the equipment over to smother the fouling organisms.
In many operations, the nursery and hatchery phases overlap as hatcheries retain the post-set in their larval rearing containers. This reduces stress and allows greater control over growing conditions. Both downwellers and upwellers are used. Upwellers differ from downwellers in the direction of water flow, with water flowing from the reservoir to the rearing cylinder rather than from the cylinder to the reservoir, as in downwellers. See Figure 2-3. Upwellers vary according to whether seawater is pushed (active flow) or pulled (passive flow) through the clams. Water exits each cylinder through a top drain. Upwellers are more common than downwellers in nursery culture.




15
Active upwellers are recommended for clams less than 3 mm (Manzi and Castagna, 1989). With proper flow rates, the post-set are suspended just above the sieve by the force of the flow and exposure to algae is maximized. Post-set must be evenly distributed over the sieve for equal food access.
Raceways are a traditional land-based nursery method (Manzi and Castagna, 1989). Raceways are long tanks or troughs of epoxy-coated wood, fiberglass or concrete. Seawater is pumped into one end of the raceway and exits from the other end.
Both shallow and deep raceway systems exist. A shallow system consists of a single layer of clams with just enough water to cover the clams. Deep systems use racks or tiers of trays to create multiple clam layers. A continuous flow of seawater is required. Water quality decreases as the distance from the point of inflow increases. Water algal content is highest at the point of inflow and lowest at the point of outflow. Elevated sediment and waste levels at the end of the raceway (point of outflow) may impede feeding and respiration. Hadley and Manzi (1984) showed a correlation of clam growth with the distance from the inflow, with highest growth occurring in clams nearest the inflow and lowest growth occurring the farthest from the inflow. Flow problems may restrict raceway systems to small post-set culturing.
Post-set under 3 mm are best raised in a land nursery system due to potential predation and smothering problems. Larger post-set perform well in field systems (Castagna, 1983; Castagna, 1986; Vaughan and Cresswell, 1989; Vaughan et al., 1989). Field nursery systems depend on the ability of natural water systems to support post-set growth. This eliminates the need to pump seawater or provide cultured algae. Vertical




16
inflow
culture reservoir
cylinder overflow
outflow
post-set
mass
sieve
Passive Flow Upweller
outf I ow
post-set si eve mass
forced i nf I ow I Active Flow Upwel ler Figure 2-3. Upwellers.




17
field systems are multi-tiered structures that place post-set in the water column near higher concentrations of phytoplankton and away from silt and benthic predators (Manzi and Castagna, 1989). Excessive fouling can be problematic. System examples are rafts, cages and racks of suspended trays or nets (Vaughan et al., 1989; Manzi and Castagna, 1989). A rack support structure is shown in Figure 2-4. Vertical systems are the most space efficient methods for culturing large post-set. The use of vertical systems may be restricted, though, as they may be a navigational hazard.
Dimensions: width. 32"
length: 66"
height: 18'
ground clearance. 6"
Rebar Frame
1I!esh Bag
I Ground Support Figure 2-4. Rack system.
Horizontal field systems rely on culturing post-set on the water bottom. Greater attention must be given to siltation and predation. Fouling can similarly be a problem.




18
Examples of horizontal systems are trays, bottom bags and the flexible belt system (Vaughan et al., 1989). A bottom bag and flexible belt are shown in Figure 2-5. Tray systems use shallow plastic, fiberglass or wooden trays filled with 2 inches of sand or gravel. A mesh covers the tray to exclude predators. Excessive fouling is scraped off the tray and mesh.
Bottom bags are mesh bags held in place by metal stakes. The mesh weave varies with clam size. The bags may be sewn shut or have one side closed with PVC pipe to allow easier access to the clams. A flotation device may be placed in the bag to aid sedimentation, after which the float is removed. Fouling is controlled by turning the bags over.
The flexible belt system consists of a pair of parallel plastic ropes holding individual plastic mesh bags in a pod or modular arrangement attached with PVC pipe closures (Vaughan et al., 1989). Post-set are placed in small mesh bags which are then placed into the individual plastic mesh units. Each bag unit is removable for maintenance or harvest. The entire belt is anchored to the substrate and fouling is reduced by turning the belt over, as with bottom bags.
Growth trials in Florida (Vaughan and Cresswell, 1989) using 3-9 mm post-set cultured at 1,800 clams per square foot showed trays to be superior to bottom bags and cages, with bottom-bag and flexible-bag culture growth intermediate to that of bottom nets and cages. Choice of tray substrate, sand versus gravel, showed no effect on growth, but resulted in survival rates of 95% for sand and 45% for gravel.




19
Float
Rebar Stake
Mesh Bag
PVC Closure BOTTOM BAG
PVC Pipe
I
Mesh Bag Rope
FLEXIBLE BELT
Figure 2-5. Bottom bag and flexible belt.




20
The length of the nursery phase varies with stock genetics, culture method, environmental conditions, food abundance and desired seed size.
Growout Phase
The growout phase takes seed clams and raises them to market size. Increasing water and food demands makes land-based clam growout systems economically impractical. Instead, culturists rely upon natural water systems to meet clam requirements. During growout, culturists are primarily concerned with providing protection from predation while not impeding water flow or food access. Containers, meshes and densities are selected such that adjustments are not required during the growout phase.
The predominant growout methods are tray, bottom net and soft tray or bag culture. Growout trays are similar to those used in field nursery systems except tray dimensions and mesh sizes change to reflect the larger clam size. Sand is the preferred substrate and clams are generally planted at 50-100 clams per square foot (Vaughan et al., 1988). The mesh is kept clean by periodic scrapping to remove fouling organisms.
Bottom nets are the least expensive growout method in both material costs and maintenance (Vaughan et al., 1988). Clams are broadcast in plots over the water bottom and then covered with mesh nets. The nets may be held by an iron frame or staked. Net or plot dimensions vary according to preference, location, management ability, etc., but are usually 25-50 feet long and 8-12 feet wide. Occasionally, mesh is laid under the clams to reduce escape and facilitate harvesting. Bottom nets must be regularly checked




21
for oversiltation, fouling and predation. Fouling must be physically removed or the nets periodically replaced with clean nets. Fouled nets are sun-dried to kill the fouling organisms.
Soft-tray or soft-bag growout is similar to soft-bag nursery culture (Vaughan et al., 1988). Clams are placed into mesh bags which are then staked to the bottom. A larger mesh size and bag are used for growout than for nursery culture. Bags are usually four feet square or four feet by eight feet. Fouling is again controlled by flipping the bags over.
The flexible belt system can be used for hard clam growout. Mesh size is larger than for nursery use, and a bag insert is usually not required. The belt is serviced and maintained in the same manner as in nursery use.
Equipment durability is an important consideration in the choice of a particular growout method as growout may take from 3-4 years in cold northern waters, and 1.5 to 2.5 years in warmer waters (Eversole, 1987). Bags, trays and nets must be chosen such that they are capable of extended use and not require frequent repair or replacement.
Harvest
Florida hard clams can be legally harvested for consumptive sale when they measure 7/8 inches in width for sale outside the state and one inch for sale inside the state (Adams et al., 1991). Measurements are made across the hinge as indicated in Figure 2-6. This equates to a two-inch or 50-mm-long clam.




22
Width
//Height
Length
Figure 2-6. Hard clam measurement axes.
The harvesting method used is dictated by the growout method practiced. Tray and bag culture allow for total harvest of the containment device. The tray or bag is manually or mechanically lifted from the bottom and legal clams removed. Bottom-plant methods require a different approach, as the clams are not in any container. Some form of rake, tong or mechanical harvest is required. State law may restrict the use of specific harvest methods, thereby determining the choice of growout method. Mechanical harvest requires a special permit in Florida.




CHAPTER 3
BIOECONOMIC MODEL
Economic Model
The objective of a hard clam aquaculture operation is assumed to be profit maximization. Management options for hard clam growout include the selection of seed size, choice of planting method and clam density, and replacement scheduling. The task confronting the aquaculturist is to select a seed size, planting method, clam density, and planting and harvest schedule that maximizes net revenue.
The first three decision options are typically single incident decision choices--an option is selected at planting time and remains fixed throughout the production process. Clam density may be altered through periodic culling. Replacement scheduling, however, is a decision requiring periodic evaluation. At each stage of the growout process, a dichotomous decision choice is faced: to sell the existing stock and replant with new seed, or to keep the existing stock for another period.
Economic principles dictate that the decision to replace or keep existing stock be based on a comparison of the gains from keeping existing stock an additional period with the opportunity gains from replacement stock during the same period (Perrin, 1972). If the gains associated with retaining existing stock exceed the average net returns from
23




24
harvesting and replacing with new clam stock, then current stock should be retained. Otherwise, immediate harvest and replacement is warranted.
A hard clam is an appreciating asset (to a point) that generates a single, point-ofharvest return. Hard clams provide no stream of revenues prior to harvest. Future hard clam growth and prices cannot be forecast with total accuracy due to the inherent uncertainty of the various processes being examined. Growth and death processes are at best imperfectly describable, and their dependence on stochastic environmental conditions increases the uncertainty of achieving specific future outcomes or states. Price movements are likewise uncertain. The result is that future states can be predicted only in a probabilistic manner. Hence, rather than having a single net revenue state, F,, in each period s, there exist k net revenue states in each period s, Fk,,. Also, each net revenue state k in period s is realized with the probability/3ks.
Given the above considerations, the discrete-time net present value of hard clams replaced every s periods is
R(s, oo) 1 E (1 +r)- -3k, Fk,. M] (3.1)
1 -(1+r)- kEkwhere
R(s, oo) = net present value of an infinite stream of revenues from an asset replaced every s periods;
r = discount rate;
Fk,S = net revenue from asset in state k in period s; Bk,. = the probability of having state k in period s;




25
M = asset replacement cost;
K*= the set of all possible states.
The term in brackets is the net present value of a single asset cycle and the term outside the brackets converts this to an infinite chain. Equation 3.1 computes the net present value of an infinite annuity received every s periods.
Replacement literature uses the term "defender" to refer to the asset already in use and the term "challenger" to refer to the replacement asset (Perrin, 1972). If a single challenger (C) exists, (3. 1) is maximized with respect to replacement age s and the maximum present value R* calculated. Should multiple challengers exist (differing by productive capability), (3. 1) must be maximized for each challenger. The best challenger is that which generates the highest R* and is denoted by C*. At each production stage (period), the culturist has the choice of replacing the defender with C* or allowing the asset to grow an additional period. Crane (1979) shows that the replacement decision is based on a comparison of the infinite net revenue streams of each alternative, or
replace if R 1> (I~Y I~ Di'1]
jEj' (3.2)
keep if R < E (I+jr Di1
indifferent otherwise,
where
=i' the probability of having net revenue j if the defender's life is extended 1 period; =j, net revenue j from the defender if it's life is extended 1 period.
Equation 3.2 constitutes the replacement decision policy HI for the hard clam




26
aquaculturist. The decision space defines the set of all possible options available to the producer. Assuming that the operation will continue production in some manner, the decision space A has two elements: to keep, and to harvest and replace.
Management of hard clam aquaculture consists of a series of decisions to maximize net revenue. Dynamic programming is a useful optimization procedure for problems involving a sequence of interrelated decisions (Dreyfus and Law, 1977). Central to dynamic programming are the concepts of stage and state. A stage is the periodic division unit, often time, at which the system is evaluated and a policy decision required. State variables are observable or measurable conditions such as clam size, mortality, prices, etc. These provide the basis for periodic (stage) evaluation.
The action chosen at each stage allows the system to change from state to state according to the processes driving changes in the state variables. A harvest decision prompts restocking and next period's stock is then replacement seed. A decision not to harvest allows additional stock growth, mortality, and market risk.
In stochastic systems, transition from state to state follows probabilistic patterns. The description and solution of problems involving stochastic systems is simplified if the underlying probability process satisfies the Markovian assumption (Howard, 1981). The Markovian assumption is that the conditional probability of any future state, given decision a and past and present states, is dependent only on a and the present state of the process. Precisely,
P {X{X1 .t=l X = i,At P{Xt-1jX = i, At=} (3.3)
where




27
X, = state of the system at stage t; At = decision at stage t; Vt = {X0, A0, XI, A,, ..., Xt-1, At-1}. One-step transition probabilities are usually denoted by pij, indicating the probability of transition from state i, i = 1, 2, ..., N, to state j, j = 1, 2, ..., N, in one period (Hillier and Lieberman, 1986). Transition probabilities are stationary if they do not change over time. This implies that P{Xt+. = j 1 Xt = i} = P{X, = j I X0 = i} for each i, j, and n = 0, 1, 2, ..., and all t = 0, 1, 2, ... These n-step transition probabilities can be denoted by p(n) i and satisfy the properties p n) > 0, for all i and j, and n = 0, 1, 2, M (3.4)
pj pi( = 1 for all i, and n = 0, 1, 2,...
j =0
The n-step transition probabilities p(n)ij represent the probability that a process in state i will be in state j after n periods. They can be presented in matrix form
(n) (n) (ni)
POO POI ... PON
p) (n) ...
The Chapman-Kolmogorov equations give a method of solving the n-step transition probabilities (Hillier and Lieberman, 1986). In the transition from state i to state j, the process will be in some state k after exactly v steps, where v is less than n. This is shown by




28
M
p(n) Ey (3.6)
pQ = pfI p~, for all i, j, n, and 0 _< v < n. (3.6)
k=0
Equation 3.6 states that the process goes to state k after v steps and then to state j in n v steps and summing over all possible k must yield p(n)i For the special cases v = 1 and v = n 1, (3.6) becomes
M
n) ( ) (3.7)
k=0
and
M
(n) (3.8)-1
Pi" = p-lpkj (3.8)
k=0
for all i, j, n. Equations 3.7 and 3.8 show that the n-step transition probabilities can be obtained recursively from the one-step transition probabilities. If n = 2, (3.7) and (3.8) become
M
p1j =E P Pkj, for all i, j. (3.9)
k=0
The p(2)j are the elements of the matrix p(2) and they are obtained by multiplying the matrix of one-step transition probabilities by itself, or
p2) = p p = p2 (3.10)
It then generally follows that the matrix of n-step transition probabilities can be obtained from
p) P* P .. *p = pn (3.11)
SPPn-1 = pn-'P.




29
The decision policy H defined by (3.2) is a stationary policy as the action it specifies at time t depends only on the state of the system at time t and is nonrandom (Dreyfus and Law, 1977).
The maximum expected total discounted net revenue of a hard clam growout system starting in state i, following decision policy 11, and evolving for t time periods is given by
m
Yj'= max {Gi., + PE P1i(a) Yt(} i=0,1,2,...,n (3.12)
a j=O
where
Gi,, = expected reward in state i at time t given action ot; p = discount factor;
Pij(a) = probability of transition from state i at time t to state j at time t+ 1 when action oa is taken.
As t approaches infinity, Yt, converges to Y1, where Y, is the expected total discounted revenue of the system starting in state i and continuing indefinitely when an optimal policy is followed (Dreyfus and Law, 1977). Equation 3.12 then becomes
m
Yi = max {Gia + PFj Pij(a) Y }; i=0,1,2,...,n. (3.13)
a j=0
Dreyfus and Law (1977) prove that the solution to (3.13) is optimal and that the stationary policy H driving (3.13) is itself optimal. The optimum hard clam replacement schedule is therefore obtained when (3.13) is satisfied.




30
Equation 3.13 is solved by the application of successive approximations to (3.12). This procedure is described by Dreyfus and Law (1977). The first step involves assigning a terminal value to Y% where t = N, the terminal stage. This terminal value is typically zero. This allows for the solution of the one-period model, YN4., to be solved from YN.1. = max. {Gj.}, i = 0, 1, 2, ... This represents the expected total discounted revenue starting in state i in stage N-i and evolving for one period. This represents the first approximation to the optimal policy.
Next, yN-2. is solved using the yN-10, YN- 1, ..., Y" from YN~1 giving the second approximation to the optimal policy. This represents the expected total discounted revenue starting in state i and evolving for two periods. This process is repeated using the recursive relationships until the optimal policy is evident. This occurs when the optimal policies of successive iterations are identical. A key consideration in this method is that it does not specify how large N should be (Dreyfus and Law, 1977). It may be necessary to apply the method of successive approximations to (3.12) using different values of N.
Hard Clam Growth and Mortality Models
As previously indicated, hard clam revenue is determined by clam size, price and quantity of clams. Knowledge of growth and survival is therefore necessary to predict expected revenues. Growth models simulate the movement of clams across size categories, while mortality models determine survival and, hence, clam quantities available for sale.




31
Bivalve Growth Literature
Hard clams have seasonal growth (Eversole, 1987). Seasonal environmental variations include changes in air and water temperatures, salinity, dissolved oxygen and other parameters. The moderate temperatures of the spring and fall produce the fastest growth (Eversole, 1987). High summer temperatures result in the slowest growth and severe winter temperatures similarly depress growth (Eversole, 1987). Changes in the environmental parameters affect hard clams directly by altering metabolic, feeding and respiration rates (Eversole, 1987; Van Heiningen, 1992; Malouf and Bricelj, 1989) and indirectly through affecting phytoplankton availability.
Seasonal growth imposes a degree of production risk on hard clam growout as clam growth becomes dependent upon uncertain environmental conditions. Knowledge of the impacts of specific environmental parameters would allow the incorporation of production risk into the bioeconomic model through the estimation of growth given probabilistic environmental conditions.
Despite the recognition of environmental impacts on clam growth, little empirical work exists on quantifying these impacts. Some authors acknowledge the importance of environmental influences in determining clam growth, but make no attempt to quantify these effects. Askew (1978) and Loesch and Haven (1973) model growth simply as a function of initial size. Lough (1975) uses the function Y = b0 + b1(T) + b2(S) + b3(T2) + b4(S2) + b5(T*S) to specify the effects of temperature (T) and salinity (S) on the percentage growth of hard clams and two other bivalve species. Parameter estimates were made for only two and ten-day-old clam larvae. Results indicated a cessation of




32
growth at temperatures and salinity above 32.5 0C and 27 ppt, respectively. Metabolic changes in older clams raise questions on the applicability of these results to larger clams. As clams mature, optimal temperature and salinity ranges change, and tolerance levels increase as older clams are able to remain closed under adverse conditions (Eversole, 1987). The increased tolerance to adverse conditions may allow clams to continue to grow during periods of environmental stress. Growth would, however, likely be less than under favorable conditions.
In a paper describing Virginia private oyster culture, Bosch and Shabman (1989) model growth as a function of initial weight, season and salinity. Their model has the form W, = Woe', j k, Where W, and W0 are final and initial oyster weights, e is the base of the natural logarithm, and a,, bj and ck are the seasonal, salinity and instantaneous growth rate (a function of initial weight) effects, respectively. The paper does not attempt to validate the growth function, focusing instead on using it as a tool in simulation modelling of oyster production. Results, though, indicate that improved knowledge of salinity effects holds great promise for increasing oyster culture profitability. Parameter estimates in this model were independently determined from different studies and combined for application to Virginia oyster culture. The linear form of the model is ln(W!W0) = aibjck. The model is unsuitable for estimation of the individual effects of the various environmental parameters. At best, a single coefficient could be estimated, representing the combined effects of season, salinity and initial weight.




33
Food (phytoplankton) is discussed by several authors as a major factor in determining clam growth (Epifanio, 1979; Menzel, 1989; Malouf and Bricelj, 1989). Most researchers focused on the growth effects of specific diets, showing that certain algal diets were more beneficial than others. Malouf and Bricelj (1989) discussed the impact of clearance rates--the volume of water filtered completely free of food particles per unit time--on clam growth. Growth was shown to be a function of clearance rate which, in turn, was affected by water temperature, food concentration and food quality. Phytoplankton quality and abundance is also affected by various environmental conditions such as atmospheric temperature, rain, or excessive run-off (Ryther, 1986). These factors influence algal growth and, hence, its availability as food. Growth Model
Clam growth can be described as
G = f(I, A, S, T, D, N, C, M) (3.14)
where
G = clam growth or final size; I = initial clam size;
A = clam age;
S water salinity;
T = water temperature;
D = water dissolved oxygen; N = food profile or algal content of the water;




34
C = water flow or current characteristics;
M = other factors.
The selection of specific independent variables included in the growth model regression was based on considerations of data availability. Specifically, data on food availability and water flow characteristics were unavailable, thus these factors could not be included in the regression analysis. Techniques exist for measuring these parameters--water speed and direction can be measured, chlorophyll levels are an indication of algal abundance, light refraction meters can measure turbidity or sediment load--but monitoring of such parameters is currently not undertaken with any regularity by either aquaculturists or water management officials.
Hard clams are measured by shell size (height or length) and not by total weight or meat mass. Although clam weight or meat mass may decrease under adverse growing conditions, shell size only decreases as a result of shell blunting in extremely old and large clams. Thus, clam growth is nonnegative and an acceptable hard clam growth function is required to mimic this condition. This was accomplished through the use of a log-linear growth function.
The regression model for clam growth in period i was
lnY, = 00 + OllnAi + 0321nSi + 031lnTi + 041nDi + Ei (3.15)
where
lnY, = natural log of the ratio of a clam's final size at the end of period i over its initial size at the beginning of period i;
B0 = a constant;




35
lnAj natural log of the clam's age at the beginning of period i; InS, natural log of the mean water salinity over period i; InT1 natural log of the mean water temperature over period i; InD, natural log of the mean water dissolved oxygen over period i; E= residual for period i.
Non-decreasing shell size results in the ratio "final size/initial size" to be no less than 1, the natural log of which is 0. Thus, the dependent variable, Y,, is 0 or positive and preserves the nonnegative growth requirement. The model is also similar in form to the linear form of the Bosch-Shabman model but allows for estimation of the individual effects of different environmental parameters.
Production data were obtained from a commercial hard clam operation located in the Indian River Lagoon near Melbourne, Florida, and used to estimate the hard clam growth function coefficients. The data covered production from June 1990 to October 1992 and contained 729 observations of monthly growth and mortality averages. Size measurements were made along the longest axis (length) of the clams. There were an average of 25 observations per month. Individual plantings ranged from 360,000 to 5 million clams. Clams ranged in age from 1 to 37 months old (post-plant age). All seed were 10 mm. New plantings occurred monthly and in any given month 23-30 different plants existed simultaneously on the lease. Planting densities prior to January 1991 were 80 clams per square foot and 60 clams per square foot during and after January 1991.
The production data also included environmental data from June 1990 to October 1992 and consisted of daily observations of water temperature, salinity and dissolved




36
oxygen. Monthly averages were computed from the daily figures. The production data set is given in Appendix 1.
The coefficients of the model were first estimated by ordinary least squares (OLS) regression and the results are shown in Table 3-1. Separate regressions were run for 60 and 80 clams per square foot.
Plots of the standard errors of (3.15) against individual independent variables indicated heteroskedastic disturbances. Heteroskedasticity produces consistent but inefficient parameter estimates in the general linear model. Further, since clam growth is restricted to be nonnegative, the dependent variable in (3.15) is a limited dependent variable. Maddala and Nelson (1975) show that ignoring heteroskedasticity in limiteddependent-variable models produces inconsistent parameter estimates. Equation (3.15) was tested for heteroskedasticity using the SPEC procedure in SAS (SAS, 1985). The specifics of the SPEC procedure are outlined in White (1980). The results of the SPEC procedure rejected the null hypothesis of no heteroskedasticity. To correct for heteroskedasticity, the model was estimated using weighted least squares (WLS) procedures. Additionally, since Y is a limited dependent variable, heteroskedastic tobit estimation was conducted. Heteroskedastic tobit procedures are described in Maddala (1983). The results of these efforts are also shown in Table 3.1.
Efficiency criteria, however, may be of little concern when the ultimate intention is growth simulation. Forecasting quality, or the ability to produce clam growth in a realistic time framework becomes more crucial. New clam sizes at the end of each growout period (month) were calculated from




Table 3-1. Hard clam growth model parameter estimates from OLS, WLS and heteroskedastic tobit regressions. Standard errors are in parentheses.
Estimation Standard
Procedure Density Error 80 61 82 B3 B4
1a 60 0.0664 0.6577 -0.0507 0.1736 -0.2742 -0.1019
(0.1755) (0.0068) (0.0409) (0.0458) (0.0245)
80 0.0415 0.2316 -0.0550 0.0629 -0.0718 -0.0085
(0.0699) (0.0029) (0.0161) (0.0154) (0.0079)
2b 60 0.0685 0.9751 -0.0681 0.1267 -0.3063 -0.1324
(0.0963) (0.0061) (0.0319) (0.0422) (0.0190)
80 0.0455 -0.0154 -0.0377 0.1018 -0.0492 -0.0083
(0.0610) (0.0037) (0.0107) (0.0127) (0.0044)
3C 60 0.3489 1.0137 -0.0770 0.3007 -0.4489 -0.2161
(0.1546) (0.0070) (0.0490) (0.0682) (0.0300)
80 0.3344 0.0422 -0.0724 0.2417 -0.1820 -0.0210
(0.1432) (0.0053) (0.0349) (0.0324) (0.0156)
a OLS; b WLS; c heteroskedastic tobit.




38
finsi = insi eo Aia' Sf 2 Ti Di (3.16)
where
finsi = clam size at the end of period i;
insi = clam size at the beginning of period i; e = base of the natural log.
Equation (3.16) was used to determine whether a particular estimation technique produced parameter estimates that resulted in acceptable clam growth. Acceptable growth was defined as attaining marketable size (50 mm long) within 2 2.5 years, a growout period common to Indian River producers.
Despite estimation using data truncated at 0, growth simulation will not guarantee nonnegative clam growth. It is therefore necessary to impose a no-shrinkage restriction. If growth simulation results in a clam shrinking, then the resultant final size is reset at equal to the clam's initial size.
The results of test simulations for 80 clams per square foot are shown in Table 3-2. Only OLS procedures produced parameter estimates that adequately reflected observed data on growth performance. Thirty-month final clam sizes were 50.567 mm at 80 clams per square foot and 56.114 mm at 60 clams per square foot (results not shown) using OLS parameter estimates. Comparative sizes for WLS and heteroskedastic Tobit were 36.602 mm and 30.721 mm, respectively, for clams planted at 80 clams per square foot. The OLS estimation produced the best model and was therefore used to simulate hard clam growth in the remaining portions of this research.




39
Table 3-2. Hard clam final size (mm) simulation results using ordinary least squares (OLS), weighted least squares (WLS) and heteroskedastic tobit (Tobit) estimation techniques for 10 mm seed planted in January at 80 clams per square foot.
Method of Estimation
Month OLS WLS Tobit
1 12.004 11.493 12.065
2 13.804 12.696 13.492
3 15.757 13.995 15.280
4 17.887 15.383 17.434
5 20.065 16.798 19.620
6 22.249 18.153 21.638
7 24.587 19.703 24.125
8 26.635 21.076 25.763
9 28.597 22.571 27.374
10 30.276 23.782 27.997
11 31.641 24.678 27.997
12 32.455 24.978 27.997
13 33.535 25.646 27.997
14 34.367 25.882 27.997
15 35.962 26.730 28.123
16 37.661 27.661 29.479
17 39.708 28.713 29.399
18 41.490 29.707 29.882
19 42.905 30.441 29.882
20 44.342 31.428 29.983
21 45.891 32.740 30.599
22 46.967 33.986 30.634
23 46.967 33.986 30.634
24 47.060 34.196 30.634
25 47.060 34.196 30.634
26 47.252 34.196 30.634
27 47.873 34.542 30.634
28 48.615 34.999 30.634
29 49.773 35.605 30.634
30 50.567 36.136 30.634




40
A comparison of the performance of the 60-clam and 80-clam models, with plants in each of the 12 calendar months, is shown in Table 3-3. The 60-clam-density model produced larger but more variable clam growth. The difference in the performance of the two models is likely a demonstration of crowding effects. Higher densities reduce mobility, access to food, and access to fresh water, thereby negatively affecting growth. The net effect of crowding, however, might not be uniform across all months due to monthly variations in water quality. The simulation results provide evidence of the positive/negative effects specific months have on the growth performance of clams. Specifically, a growth bias toward fall and winter conditions and away from spring and summer conditions is indicated. Timing is apparently at issue, with clams unable to overcome the negative effects of planting in less favorable months. Estimation of Mortality
Little empirical work exists on developing mortality models for estimating bivalve mortality rates. In situations where populations are modelled from the postlarval. stage without consideration of reproduction, the numbers of surviving individuals is due to mortality only. Mortality of seed clams is often many times that of adults (Eversole, 1987). In the absence of adverse environmental conditions, predation is the major cause of hard clam mortality (Eversole, 1987). Susceptibility to predation decreases with increasing clam size and commercial growout methods utilize effective physical barriers to reduce predation. Exposure to environmental conditions outside tolerance ranges also produces increased mortality. As discussed previously with reference to growth conditions, tolerance ranges as they relate to mortality increase for older clams, however,




41
Table 3-3. Average expected 30-month sizes (mm) from simulation model testing using 10 mm seed.
Planting Density
Month 60 clams/sq. ft. 80 clams/sq. ft.
January 65.92 56.16
February 59.09 54.19
March 55.26 52.91
April 50.28 51.19
May 47.33 50.25
June 48.45 50.41
July 51.16 51.13
August 56.87 53.21
September 63.76 56.74
October 70.88 58.36
November 73.15 58.53
December 70.78 57.62
as they are able to maintain closure longer, effectively shutting out certain adverse conditions, and are able to utilize various metabolic mechanisms undeveloped in juvenile clams (Eversole, 1987).
Most models describing populations with an absence of reproduction are usually expressed in terms of an instantaneous mortality rate (Allen et al., 1984) which, despite the name, computes an annual or monthly mortality. Askew (1978) links oyster mortality to size and computes monthly mortality from annual data, using the assumption that short term rates concur with long term rates. Data showed mortality peaks in winter and early summer. Askew acknowledged the potential impact of season on mortality, but concluded insufficient evidence existed to warrant inclusion in his analysis. The




42
observed association of mortality with size implied an inverse relationship as mortality rates decreased with increasing size.
Evaluation of the mortality data used in this research did not produce a workable model of environmentally-dependent mortality. Hard clam mortality was thus calculated in a deterministic manner based on age. First, a cumulative 30-month mortality was identified from discussions with hard clam research specialists (David Vaughan, Harbor Branch Oceanographic Institute, Ft. Pierce, Florida, and Leslie Sturmer, Project Ocean, Cedar Key, Florida, personal communications). Next, a weighting system for mortality was determined from examination of available mortality data linking mortality with age. The resulting weight system placed greater weights on younger clams, indicating greater early mortality and lower mortality of older clams. The weight system was then used to generate periodic (monthly) conversion factors that produced the given cumulative 30month mortality. The monthly conversion factors are given in Table 3-4. The factors sum to one over 30 months and actual periodic monthly mortality is computed by multiplying the cumulative 30-month mortality by the respective conversion factor. Table 3-4. Periodic (monthly growout age) mortality conversion factors.
Period Conversion factor
1 0.10
2 0.08
3 0.06
4-11 0.04
12-17 0.03
18-34 0.02




CHAPTER 4
RESULTS AND DISCUSSION
Growth Simulation
Hard clam growth was simulated using equation (3.14) and environmental data collected from the St. John's River Water Management District (SJRWMD). The SJRWMD data consisted of daily water temperature, salinity and dissolved oxygen readings from different locations in the Indian River Lagoon. To test the robustness of the growth model, three sites were selected from which monthly averages were computed. The selection of the three sites was based on comparison of site conditions, with the selected sites representing mean, above mean, and below mean conditions. Mean monthly values for each site are listed in Appendix 2. Each environmental parameter was assumed to have a normal distribution. Using parameter means and standard deviations, monthly averages were assumed to take on three possible values: the mean, greater than mean and less than mean. The extreme values were calculated by adding or subtracting two standard deviations to or from the mean, respectively. Exposure to a mean value had a 68 % probability, while exposure to each of the extreme values had a probability of 16%. A given environment consisted of some combination of mean, above mean and below mean values for the three parameters of interest. For
43




44
example, environment El. consisted of mean values for temperature, salinity and dissolved oxygen in month m, where m = 1, 2,..., 12. Environment E2. had mean temperature and salinity and greater than mean dissolved oxygen in month m, etc. Given three parameters, 3' or 27 possible environments existed each month.
The environmental probabilities for the 27 environmental states are given in Table 4-1. The probability of exposure to a particular environment in any month was determined by the product of the probabilities of receiving each individual parameter. Environmental states were defined in mean, above mean and below mean terms. For example, the probability of having El., indicating mean temperature, salinity and dissolved oxygen in month m, was 0.68' or 31.44%. The probability of having environment E2. was 0.68'*0.16 or 7.4%. The probability of exposure to a given environmental state was the same regardless of which stage or month transition occurred from. The probability of receiving mean values for all three parameters in the current or future growing periods was 31.44 % regardless of which month the production process was in. Actual parameter values for a given environmental state, however, and their impact on clam growth, varied from month to month. Mean water temperature for environmental state Ell was not the same as mean water temperature for E12, etc. January temperatures are typically different than February temperatures. The probability of exposure to a mean, above mean or below mean value in each month, however, remains the same.
Growth simulation began with a clam or bag of clams of a specific initial size, either 10 mm or 15 mm, and an age of 1, indicating the first growout month. Simulation




45
Table 4-1. Environmental probabilities used in hard clam growth simulation.
Dissolved
Environment Oxygen Salinity Temperature Probability
El X X X 0.3144
E2 X X + 0.0739
E3 X X 0.0739
E4 X + X 0.0739
E5 X + + 0.0174
E6 X + 0.0174
E7 X X 0.0739
E8 X + 0.0174
E9 X 0.0174
ElO0 + X X 0.0739
Ell + X + 0.0174
E12 + X 0.0174
E13 + + X 0.0174
E14 + + + 0.0040
El5 + + 0.0040
E16 + X 0.0174
E17 + + 0.0040
E18 + 0.0040
E19 X X 0.0739
E20 X + 0.0174
E21 X 0.0174
E22 + X 0.0174
E23 + + 0.0040
E24 + 0.0040
E25 X 0.0174
E26 + 0.0040
E27 0.0040
Code: X = mean value, (-) = mean minus two standard deviations, (+) = mean plus two standard deviations.




46
began with a January plant. January growth simulation produced a new clam or unit with 27 possible profiles or sizes. Although the first step began with seed of equal size, 27 (potentially) alternative final sizes were possible, the result of 27 different environments. The second step required that each of the 27 alternative units produced by step 1 be exposed to each of the 27 February, or period 2, environments. The outcome of an exposure defined by El (mean values for all three environmental parameters) in period 2 was calculated as the average size of the 27 period 1 clams transitioning through El in period 2. Similarly, the outcome of exposure to E2 (mean dissolved oxygen and salinity and above mean temperature) was the average of those same 27 period 1 units exposed to E2 in period 2, etc. Thus, the outcome of each sequential growth period was 27 "new" clams, where each clam represented an average of 27 potentially different clams transitioning through a given environment.
It is important to note that the 27 clams produced each period need not be of different sizes. The zero growth outcome of certain environments often produced identical representative clams. An example of the outcome of growth simulation is given in Table 4-2. Growth was simulated for 34 months. This produced a 27*34 matrix of ending clam sizes for each period. Each cell or entry in the matrix represented the average of 27 clams exposed to the environment designated by that particular row. For a given site, this procedure was repeated with each of the 12 calendar months as the initial growth period. This process was repeated for each initial seed size (10 and 15 mm) and planting density (60 and 80 clams per square foot). This produced 2*2*12 or 48 growth simulations per growout site.




47
Table 4-2. Example of hard clam growth (ending size) simulation results for a January plant using 15 mm seed planted at 80 clams per square foot.
Period
Environment 1 2 ... 13 14 15 ... 33 34
Size (mm)
1 18.60 21.97 53.26 56.55 60.22 85.02 85.08
2 18.26 21.55 52.28 55.45 59.15 85.02 85.08
3 19.06 22.57 54.59 58.09 61.68 85.02 85.63
4 18.78 22.27 53.78 57.31 61.19 85.02 85.50
5 18.44 21.84 52.80 56.20 60.11 85.02 85.08
6 19.25 22.88 55.13 58.87 62.68 85.51 86.70
7 18.38 21.60 52.64 55.59 58.94 85.02 85.08
8 18.05 21.18 51.68 54.51 57.89 85.02 85.08
9 18.84 22.19 53.96 57.10 60.37 85.02 85.08
10 18.54 21.94 53.10 56.46 60.14 85.02 85.08
11 18.20 21.51 52.13 55.36 59.07 85.02 85.08
12 19.01 22.54 54.44 58.00 61.59 85.02 85.40
13 18.72 22.24 53.62 57.22 61.11 85.02 85.26
14 18.38 21.81 52.65 56.11 60.02 85.02 85.08
15 19.19 22.84 54.97 58.78 62.59 85.33 86.46
16 18.33 21.57 52.49 55.50 58.86 85.02 85.08
17 17.99 21.15 51.53 54.43 57.81 85.02 85.08
18 18.79 22.16 53.81 57.01 60.28 85.02 85.08
19 18.68 22.02 53.49 56.66 60.33 85.02 85.08
20 18.34 21.59 52.51 55.56 59.25 85.02 85.08
21 19.15 22.62 54.83 58.20 61.79 85.02 85.99
22 18.86 22.31 54.01 57.42 61.30 85.14 85.86
23 18.52 21.88 53.03 56.31 60.21 85.02 85.08
24 19.33 22.92 55.37 58.99 62.79 85.77 87.06
25 18.46 21.64 52.87 55.70 59.04 85.02 85.08
26 18.12 21.23 51.91 54.62 57.99 85.02 85.08
27 18.92 22.23 54.20 57.22 60.47 85.02 85.08




48
Final monthly expected clam sizes were computed using the results of the growth simulation and the environmental probabilities. The expected size of a clam at the end of period i equaled the sum of the expected outcome of a clam exposed to environment j in period i times the probability of encountering environment j. Thus, each 27*34 matrix of final clam sizes, where each row represented a different environment, was reduced to a 1 *34 vector of expected final sizes. Each outcome represented an expected mean clam size. Size standard deviations were then calculated from the equation STD 2.151 0.0603 FINS (4.1)
where
STD standard deviation of hard clam size; FINS hard clam mean size.
Equation 4.1 was estimated using the commercial production data previously described.
Mortality
The periodic and cumulative mortalities used in the production simulation are shown in Table 4-3. Early mortality rates are higher than later ones, reflecting the higher natural mortality of younger and smaller clams. When 15-mm seed clams were used, the first two months of the mortality series were eliminated, reflecting the increased survival expectation of larger clams.




49
Table 4-3. Monthly and cumulative mortality used in hard clam growth simulation.
Mortality (%)
Month Monthly Total
1 2 2
2 1.6 3.6
3 1.2 4.8
4 0.8 5.6
5 0.8 6.4
6 0.8 7.2
7 0.8 8
8 0.8 8.8
9 0.8 9.6
10 0.8 10.4
11 0.8 11.2
12 0.6 11.8
13 0.6 12.4
14 0.6 13
15 0.6 13.6
16 0.6 14.2
17 0.6 14.8
18 0.4 15.2
19 0.4 15.6
20 0.4 16
21 0.4 16.4
22 0.4 16.8
23 0.4 17.2
24 0.4 17.6
25 0.4 18
26 0.4 18.4
27 0.4 18.8
28 0.4 19.2
29 0.4 19.6
30 0.4 20
31 0.4 20.4
32 0.4 20.8
33 0.4 21.2
34 0.4 21.6




50
Economic Evaluation
The results of the hard clam growth simulation were used to determine optimal production design and expected returns. Two production scenarios were modelled. The first scenario treated an acre as a single unit. The entire acre was planted at one time and all clams were harvested in a single month. Hereafter, this scenario will be referred to as single plant production. The second scenario imposed a monthly marketing requirement. Producers might require or prefer a monthly harvest and revenue stream in order to meet labor restrictions or expenditure requirements. Similarly, market outlets might require monthly supplies from growers to satisfy monthly consumer demand. Therefore, some positive quantity of clams was required to be sold monthly. Planting restrictions were not imposed, so an entire acre could still be planted at one time as a single unit. The acre must be harvested, however, in monthly units or plots, so this method is termed multiplot production. Production assumptions are based on Adams et al. (1993). Some assumptions common to both scenarios were
(a) The operation lease consisted of a two-acre submerged tract. The site is
situated such that it is never exposed during low tide. Each acre is
identical in terms of productive capacity.
(b) A staggered production schedule is followed. Each acre is planted and
harvested successively. The first acre is planted the first year and the second acre planted the next. Expected growout periods under a given production regime are identical, thus producing harvests in successive
years.
(c) 750 4-foot-square growout bags are planted per acre.
(d) Clam seed is stocked at 1000 (62.5 clams per square foot) or 1200 (75
clams per square foot) clams per bag. These represent recommended
average and maximum densities (Adams et al., 1993).




51
(e) The current month is January and the operation has the option of beginning
production (making the initial plant) in the current month or in any of the
future 11 months.
Information on hard clam prices was required to determine the expected returns of growing clam stock. Hard clam price data were obtained from the Florida Department of Natural Resources (FDNR) and a commercial fish house in the Indian River Lagoon area. The combined data set consisted of monthly average wholesale (dockside) prices from January 1986 to April 1993 for littleneck, topneck, cherrystone and chowder clams. Although the data covered a span of eight years, relatively few price observations were available as price reporting was voluntary. Price forecasts for subsequent portions of this research were based on average monthly prices for the four size categories. Under current Florida law, clams less than 50 mm long (one inch across the hinge) are not legal for sale in Florida. Licensed dealers may sell 45-mm (7/8 inch across the hinge) clams outside the state. Price data on the smaller clams was unavailable. The assumption is also made that the clam producer is not a licensed dealer and is therefore prohibited from selling 45-mm clams. All undersized clams were assigned a price of zero. Mean monthly prices for the various size categories are given in Appendix 3.
Single Plant Production Op2timal Rotations
The results of the hard clam growth simulation were combined with seed quantities and mortality rates to estimate expected revenues. The prescribed densities translated to initial seed plants of 750,000 and 900,000 clams per acre. Surviving clam numbers were determined using initial plant quantities and the previously described mortality rates.




52
Sixty-eight percent of surviving clams were assumed to possess mean size, 16% mean plus two standard deviations, and 16% mean minus two standard deviations. Hard clam prices are size dependent. Applying the standard deviation to the mean expected size generated a size spread within each age group, thus allowing surviving clams of a given age class to potentially fall into different size categories. Expected gross revenues were then calculated from total clam numbers within each size category times the expected price.
Next, expected net discounted revenues were computed by subtracting periodic maintenance costs and using an annual real discount rate of 4%. These revenues were then examined to identify the growout period that produced maximum expected net discounted revenues. Each production method (density and seed size) and planting month (January through December) was evaluated independently. Hereafter, the term "production method" should be understood to refer to a specific seed density (clams per square foot) and seed size (mm). Thus, optimal growout periods were identified for 10 and 15 mm seed planted at 62.5 and 75 clams per square foot beginning in January, February, March, etc. Revenues over the course of the 34-month growout followed general patterns of first increasing and then decreasing in value. This was due to stocks first growing into and then out of more valuable size categories. Changes were not uniformly up or down, though, as mortality or price movements often offset the effects of growth gains. The revenues produced by these optimal growout periods represented the best potential revenues possible given expected growth, mortality and price movements. Optimal growout periods are listed in Tables 4-4, 4-5 and 4-6.




53
Table 4-4. Site 1 optimal growout periods (months) for single plant production by production method and planting month.
Planting Seed density/size
Month 62.5/10 62.5/15 75/10 75/15
January 28 19 25 16
February 34 22 31 19
March 30 23 30 21
April 29 23 29 20
May 33 23 31 21
June 34 22 30 20
July 33 21 31 19
August 32 20 28 18
September 30 18 24 17
October 28 17 26 16
November 27 16 25 16
December 27 16 24 16
Table 4-5. Site 2 optimal growout periods (months) for single plant production by production method and planting month.
Planting Seed density/size
Month 62.5/10 62.5/15 75/10 75/15
January 32 23 32 19
February 31 24 31 19
March 33 24 33 21
April 34 32 34 22
May 28 31 33 21
June 34 30 32 20
July 33 29 31 14
August 32 25 30 19
September 29 20 29 18
October 34 26 28 17
November 34 18 27 17
December 33 21 26 17




54
Table 4-6. Site 3 optimal growout periods (months) for single plant production by production method and planting month.
Planting Seed density/size
Month 62.5/10 62.5/15 75/10 75/15
January 32 20 32 19
February 31 24 31 19
March 30 23 33 21
April 34 24 34 22
May 33 27 33 21
June 32 26 32 20
July 31 26 31 14
August 34 28 30 19
September 33 20 29 18
October 31 26 28 17
November 29 17 27 17
December 29 17 26 17
The next step required connecting consecutive production cycles so that optimal
rotations could be identified. A rotation is defined as a pattern of sequential plant, growout, harvest, replant, etc., decisions. A given production schedule or cycle (as
defined by the plant month and growout period) resulted in the next cycle beginning in
a particular month. Unless the optimal growout period was 23 months, each growout
schedule resulted in the next plant beginning in a different month than the previous cycle. For example, a January-Year 1-plant growing for 23 months would be harvested in December, Year 2, allowing replanting in January, Year 3. Simultaneous harvest and
replant was not allowed in the model due to the time and labor requirements of harvesting an entire acre, so the effective turnaround time was the growout period plus
one month. If planting occurred in January and the growout period were 24 months,




55
harvest would occur in January of the third year and replanting would occur in February. Then, the optimal February-plant growout period would be followed.
This "second-cycle" start was identified for each starting month. Consecutive plants were then strung together. An example might be: January plant, 28-month growout, May harvest, June plant, 32-month growout, February harvest, March plant, 24-month growout, etc. In all cases stabilization or a repetitive cycle emerged. This occurred when any subsequent planting month was the same as any previous planting month. For example, any 23-month growout produced immediate stabilization. As indicated above, a January-plant growing for 23 months resulted in a December harvest and subsequent January replant. The effective turnaround time is 24 months, hence replanting always occurs in the same month. Some repetitive cycles involved multiple plants, while others eventually attained a 24-month repetitive cycle, but required several plants before falling into the cycle.
Once planting patterns were identified, expected revenues were examined to determine optimal rotations. This required comparing the revenue potential of all combinations of immediate replacement and delayed cycles. A cycle with immediate replacement would be that as described in the previous paragraph. Immediate (next month) replacement would follow any harvest. A delayed cycle would include varying quantities of down months. A down month is one in which the acre is empty; no growing clams exist on the acre and neither harvest nor planting occurs. A down month was indicated if revenues could be increased by waiting. For instance, February conditions may be more conducive to rapid growth than January conditions. Hence, a new operation would make initial plants in February and, if harvest ever occurred in




56
December, would delay replanting one month until February. Multiple down months were also a possibility.
An example of an optimal rotation might be to plant in January, grow for 18 months, harvest in August, wait two months until November, plant, and then follow the optimal November cycle. Or, the optimal decision might be to wait four months after the August harvest and replant in January. Hence, although actual growout lasts only 18 months, the effective turnaround time is 24 months and the acre is empty for four months.
Evaluation of the returns of all possible combinations of delayed and immediate replacement production allows the identification of the rotation that produces maximum expected net revenues. Optimal rotations were identified for each production method for each month of the year. Then, these 12 rotations (distinguished by the 12 calendar months) were compared to determine, given the opportunity to begin production in any month, the rotation that produced maximum expected net revenues. Optimal rotations for each production method are given in Table 4-7. Production results generated by the optimal rotations are given in Table 4-8. As can be seen where the number of clams sold is less than the number of clams survive, certain production schedules produce undersized clams.
The difference between the procedures described in the above paragraphs is subtle yet significant. The first procedure addresses the question of, if the operation somehow finds itself in a given month with a planting option, what planting schedule should be initiated. Unexpected growth and harvest, mortality, natural disaster, etc. might result in an empty acre in any month. The second procedure addresses the question of when




Table 4-7. Optimal planting rotations for single plant production by production method.
Plant month/growout period (months) Seed density/size (mm) 1 Wait 2 Wait 3 Wait 4
Site 1
62.5/10 Jan/28 5 Nov/27 1 Apr/29 no Nov repeat
62.5/15 Feb/22 1 Feb repeat
75/10 Sep/24 3 Jan/25 6 Sep repeat
75/15 Feb/19 no Oct/16 3 Jun/20 3 Jun repeat
Site 2
62.5/10 Feb/31 2 Dec/33 2 Dec repeat
62.5/15 Jan/23 no Jan repeat
75/10 Feb/31 4 Feb repeat
75/15 Feb/19 4 Feb repeat
Site 3
62.5/10 Jan/32 3 Jan repeat
62.5/15 Jan/20 3 Jan repeat
75/10 Feb/31 4 Feb repeat
75/15 Feb/19 4 Feb repeat




Table 4-8. Optimal rotation production results for single plant production.
Average Size # Clams # Clams # Clams
Seed density/size Aize (months) Size (mm) STD Start Survive Sold
Site 1 Plant month
62.5/10 January 28 63.96 6.01 750,000 606,000 606,000
November 27 64.47 5.92 750,000 609,000 609,000
April 29 50.16 5.17 750,000 603,000 506,520
62.5/15 February 22 61.79 5.88 750,000 624,000 624,000
75/10 September 24 50.39 5.19 900,000 741,600 622,944
January 25 51.84 5.28 900,000 738,000 619,920
75/15 February 19 63.99 6.01 900,000 759,600 759,600
October 16 62.12 5.90 900,000 772,200 772,200
June 20 64.37 6.03 900,000 756,000 756,000
Site 2
62.5/10 February 31 53.07 5.35 750,000 597,000 501,480
December 33 62.05 5.89 750,000 591,000 591,000
62.5/15 January 23 62.91 5.94 750,000 621,000 621,000
75/10 February 31 51.15 5.23 900,000 716,400 601,776
75/15 February 19 61.80 5.88 900,000 759,600 759,600
Site 3
62.5/10 January 32 61.77 5.87 750,000 594,000 594,000
62.5/15 January 20 62.11 5.90 750,000 630,000 630,000
75/10 February 31 51.09 5.23 900,000 716,400 601,776
75/15 February 19 61.77 5.87 900,000 759,600 759,600




59
a new lease should begin operation. For example, at Site 1 using 10 mm seed planted at 62.5 clams per square foot, a February plant requires a 34-month growout to maximize net revenue. See Table 4-4. If expected conditions occur, however, the grower should never be faced with a planting decision in February as the optimal planting rotation is January-November-April-November. See Table 4-7. Once optimal rotations are determined, rotations not included in the optimal design are relevant only if unexpected conditions occur.
Multiplot Production Optimal Rotations
A simple management adjustment to a monthly marketing constraint would be to follow the rotation pattern determined for single plant production and harvest over a 12month period rather than in a single month. To do so, however, would incur additional mortality and produce potential revenue losses as clams grew into less profitable size categories. Additionally, as the acre is planted as a unit, replanting would be affected as insufficient space would exist for the new seed.
Correct determination of proper management strategy when faced with a monthly harvest requirement, however, requires a change in perspective since the production or management question is now different than that with single plant production. With single plant production the question is, given both planting and marketing freedom, what is the best rotation to follow. All seed is assumed essentially identical with respect to growth potential. Optimal plant and growout times are therefore identical for all seed. When given the freedom or option to plant and harvest all clams simultaneously, the correct decision would be to do so. All clams are allowed to attain economic maturity. The




60
correct production question now is, however, given that clams must be marketed each month, when is the best time to plant seed such that they will be available for sale in each calendar month. The intuitive impact of this restriction is that clams will be planted and harvested in months other than those prescribed by the single plant production analysis.
The monthly marketing constraint imposes 12 harvests annually. No restriction was placed on when clams were planted; all clams could be planted during the same month, or any other combination. Actual planting schedules depend upon revenue comparisons of the various options. The clams would be harvested, however, in monthly units. Thus, the production unit, as driven by the harvest unit, is now a fraction of an acre. This requires that the acre be subdivided into units or plots, each representing a separate harvest and, possibly, a separate plant. To simplify the analysis, it was assumed that all plots be equal in size and, hence, planted with equal numbers of clam seed. As previously mentioned, the monthly harvest constraint did not specifically impose quantity restrictions, but rather only required positive monthly sales. All plots were therefore identical and the analysis did not address the question of optimal monthly planting quantities subject to minimum or maximum sales constraints. The analysis could be modified to account for specific volume sales restrictions.
Planting schedules are driven by harvesting requirements since, as previously stated, the economic question is when is the best time to plant seed for harvest in month i. The clam growth simulation results and net revenue estimations used in the first analysis were appropriate for use in determining when harvests began and, hence, when




61
plants were required. An initial harvest rule was selected such that harvest began (given the option of planting in each of the 12 calendar months) when the expected net revenues from a given plant exceeded all other potential revenues for that month from any other plant. This determination required a comparison of not only all potential revenues for that month in that year, but also for that month in all future years. Future year evaluations incorporated a consideration of the revenue potential of not only a given plant 12 months later, but also that of other plants that might not have been planted when the initial observation is made. Since production or growth potential was repetitive, only one set of simulation results had to be examined to encompass all future years. With growout simulated for 34 months using 12 different starting months, the simulation results covered a 45-month period. At one site with a particular production method, a January plant (the first possible plant) with a 28-month growout was determined to be the best way to produce clams for harvest in May. Harvest therefore began in May and the harvest cycle or year was May to April. Harvest was not begun in April because April revenues were maximized by an August plant (with a 32-month growout). Since January was the reference or first possible planting month, an August planting opportunity had yet to occur. April revenues were therefore maximized by waiting until August. This relationship repeated for all other months prior to the first May harvest month: revenues could be increased by waiting. May revenues were greatest, however, with a January plant and, thus, production began in January and produced a first harvest two years later in May.




62
The initial harvest therefore began when revenues were greatest relative to all other production options for that particular month (year neutral). Determination of the best planting time for the next and all subsequent monthly harvests was then based on comparison of potential revenues for that specific month and year. The distinction between the first step--that of identifying the initial harvest--and this step is important. The first step determined both the month and the year of initial harvest. Harvest then became anchored to that year and the subsequent identification of maximum revenues became year-specific. Revenue comparison thus considered only clams present at that point in time and not clams in future years. Despite the narrowing of focus, once harvest began, revenues chosen represented overall maximums similar to the initial harvest.
Maximum revenues were identified and then tracked back to determine planting schedules. This identified an annual planting schedule consisting of 12 plants. These were then linked to determine the required acre-subdivision by requiring identical annual planting schedules and comparing planting requirements with plot availability as determined by harvests. Same-month harvest and replanting was allowed with multiplot production due to the reduced labor requirements. With 750 bags per acre and a minimum of 12 plots per acre, at most 63 bags would be harvested in any month. Replanting during the same month would not be a problem. The number of required plots per acre are given in Table 4-9. Total plot requirements were determined on alease or 2-acre basis. Per acre plot requirements were then computed by dividing this total by 2. Fractional per acre plot numbers were indicative of one production unit (plot) spread over two acres. The number of bags per plot are given in Table 4-10. This was




63
Table 4-9. Number of required plots per acre for multiplot hard clam production.
Seed density/seed size
62.5/10 62.5/15 75/10 75/15
Site 1 17 12 14 12
Site 2 17 12.5 17 12
Site 3 16.5 12 16.5 12
Table 4-10. Number of growout bags per plot for multiplot hard clam production.
Seed density/seed size
62.5/10 62.5/15 75/10 75/15
Site 1 45 63 54 63
Site 2 45 60 45 63
Site 3 46 63 46 63
computed by dividing the number of bags per acre, 750, by the number of plots per acre. In some instances, rounding requirements in the computation of the number of bags per acre resulted in more than 750 bags planted per acre. Per acre bag totals are given in Table 4-11. The number of required plots were then used to scale production and revenues to reflect the decrease in the unit of analysis from one acre to one plot. The multiplot production numbers are given in Tables 4-12, 4-13 and 4-14.
Table 4-11. Total bags planted per acre for multiplot hard clam production.
Seed density/seed size
62.5/10 62.5/15 75/10 75/15
Site 1 765 756 756 756
Site 2 765 750 765 756
Site 3 759 756 759 756




Table 4-12. Site 1 multiplot production results.
Growout Average Size # Clams # Clams # Clams
Seed density/size (months) Size (mm) STD, Start Survive Sold
Plant month Harvest month
62.5/10 January 28 May 63.95 6.00 45,000 36,360 36,360
January 29 June 65.65 6.10 45,000 36,180 36,180
January 30 July 65.91 6.12 45,000 36,000 36,000
January 31 August 66.17 6.13 45,000 35,820 35,820
January 32 September 66.49 6.15 45,000 35,640 35,640
January 33 October 66.62 6.16 45,000 35,460 35,460
January 34 November 66.94 6.18 45,000 35,280 35,280
February 34 December 61.83 5.87 45,000 35,280 35,280
August 32 April 63.78 5.99 45,000 35,640 35,640
November 27 February 62.47 5.91 45,000 36,540 36,540
December 25 January 53.86 5.39 45,000 36,900 30,996
December 27 March 61.85 5.87 45,000 36,540 36,540
Total annual sales: 425,916
62.5/15 January 17 June 63.46 5.97 63,000 53,676 53,676
January 18 July 64.09 6.01 63,000 53,424 53,424
January 19 August 64.72 6.05 63,000 53,172 53,172
January 20 September 65.25 6.08 63,000 52,920 52,920
January 21 October 65.59 6.10 63,000 52,668 52,668
January 22 November 66.37 6.15 63,000 52,416 52,416
February 22 December 61.78 5.87 63,000 52,416 52,416
February 23 January 66.19 6.14 63,000 52,164 52,164
March 23 February 67.06 6.19 63,000 52,164 52,164
April 23 March 64.14 6.01 63,000 52,164 52,164
December 16 April 63.83 5.99 63,000 53,928 53,928
December 17 May 66.97 6.18 63,000 53,676 53,676
Total annual sales: 634,788




Table 4-12- -continued.
Growout Average Size # Clams # Clams # Clams
Seed density/size (months) Size (mm) STD Start Survive Sold
Plant month Harvest month
75/10 January 25 February 51.83 5.27 64,800 53,136 44,634
February 25 March 50.89 5.21 64,800 53,136 44,634
February 28 June 53.80 5.39 64,800 52,358 43,981
February 30 August 54.18 5.41 64,800 51,840 43,546
March 25 April 50.95 5.22 64,800 53,136 44,634
March 26 May 51.82 5.27 64,800 52,877 44,417
April 27 July 50.54 5.19 64,800 52,618 44,199
September 24 September 50.39 5.18 64,800 53,395 44,852
October 24 October 50.44 5.19 64,800 53,395 44,852
October 27 January 53.82 5.39 64,800 52,618 44,199
November 24 November 50.26 5.18 64,800 53,395 44,852
December 24 December 50.00 5.16 64,800 53,395 44,852
Total annual sales: 533,652
75/15 January 16 May 62.80 5.93 75,600 64,714 64,714
January 17 June 64.84 6.05 75,600 64,411 64,411
January 19 August 67.20 6.20 75,600 63,806 63,806
February 17 July 62.42 5.91 75,600 64,411 64,411
February 19 September 63.99 6.00 75,600 63,806 63,806
February 20 October 65.21 6.08 75,600 63,504 63,504
February 21 November 66.52 6.16 75,600 63,202 63,202
April 20 December 62.65 5.92 75,600 63,504 63,504
May 20 January 62.97 5.94 75,600 63,504 63,504
June 20 February 64.37 6.03 75,600 63,504 63,504
August 19 March 64.80 6.05 75,600 63,806 63,806
December 16 April 63.70 5.99 75,600 64,714 64,714
Total annual sales: 766,886




Table 4-13. Site 2 multiplot production results.
Growout Average Size # Clams # Clams # Clams
Seed density/size (months) Size (mm) STD Start Survive Sold
Plant month Harvest month
62.5/10 January 27 April 53.25 5.36 45,000 36,540 30,694
January 29 June 57.83 5.63 45,000 36,180 30,391
January 31 August 58.15 5.65 45,000 35,820 30,089
February 27 May 50.48 5.19 45,000 36,540 30,694
February 29 July 52.44 5.31 45,000 36,180 30,391
February 31 September 53.06 5.34 45,000 35,820 30,089
February 32 October 53.28 5.36 45,000 35,640 29,938
February 33 November 53.42 5.37 45,000 35,460 29,786
February 34 December 54.31 5.42 45,000 35,280 29,635
March 34 January 52.63 5.32 45,000 35,280 29,635
November 28 March 55.40 5.49 45,000 36,360 30,542
December 26 February 50.52 5.19 45,000 36,720 30,845
Total annual sales: 362,729
62.5/15 January 23 January 62.91 5.94 60,000 49,680 49,680
January 24 February 66.32 6.14 60,000 49,440 49,440
February 24 March 64.31 6.02 60,000 49,440 49,440
March 24 April 62.74 5.93 60,000 49,440 49,440
March 25 May 67.12 6.19 60,000 49,200 49,200
August 25 October 62.76 5.93 60,000 49,200 49,200
November 18 June 63.79 5.99 60,000 50,880 50,880
November 21 September 67.20 6.20 60,000 50,160 50 160
December 18 July 62.88 5.94 60,000 50,880 50,880
December 19 August 63.20 5.96 60,000 50,640 50,640
December 22 November 64.86 6.06 60,000 49,920 49,920
December 23 December 65.29 6.08 60,000 49,680 49,680
Total annual sales: 598,560
C7%
ON




Table 4-13--continued.
Growout Average Size # Clams # Clams # Clams
Seed density/size (months) Size (mm) STD Start Survive Sold
Plant month Harvest month
75/10 January 26 March 50.28 5.18 54,000 44,064 37,014
January 27 April 51.71 5.26 54,000 43,848 36,832
February 27 May 50.52 5.19 54,000 43,848 36,832
February 28 June 50.74 5.20 54,000 43,632 36,651
February 29 July 50.88 5.21 54,000 43,416 36,469
February 30 August 51.04 5.22 54,000 43,200 36,288
February 31 September 51.15 5.23 54,000 42,984 36,107
February 32 October 51.18 5.23 54,000 42,768 35,925
March 32 November 50.17 5.16 54,000 42,768 35,925
March 33 December 50.47 5.19 54,000 42,552 35,744
March 34 January 51.21 5.23 54,000 42,336 35,562
December 26 February 50.41 5.18 54,000 44,064 37,014
Total annual sales: 436,363
75/15 January 17 June 62.70 5.93 75,600 64,411 64,411
January 18 July 63.97 6.05 75,600 64,109 64,109
January 19 August 64.78 6.05 75,600 63,806 63,806
January 21 October 66.12 6.13 75,600 63,202 63,202
January 22 November 67.18 6.20 75,600 62,899 62,899
February 19 September 61.79 5.87 75,600 63,806 63,806
March 21 December 63.23 5.96 75,600 63,202 63,202
April 21 January 62.74 5.93 75,600 63,202 63,202
June 20 February 62.35 5.90 75,600 63,504 63,504
June 21 March 67.20 6.20 75,600 63,202 63,202
November 17 April 63.86 6.00 75,600 64,411 64,411
December 17 May 63.77 5.99 75,600 64,411 64,411
Total annual sales: 764,165




Table 4-14. Site 3 multiplot production results.
Growout Average Size # Clams # Clams # Clams
Seed density/size (months) Size (mm) STD Start Survive Sold
Plant month Harvest month
62.5/10 January 26 March 51.89 5.27 46,000 37,536 31,530
January 32 September 61.77 5.87 46,000 36,432 36,432
January 33 October 62.75 5.93 46,000 36,248 36,248
January 34 November 62.75 5.93 46,000 36,064 36,064
February 26 April 50.00 5.16 46,000 37,536 31,530
February 27 May 52.33 5.30 46,000 37,352 31,376
February 28 June 54.19 5.41 46,000 37,168 31,221
February 30 August 54.81 5.45 46,000 36,800 30,912
March 28 July 50.85 5.21 46,000 37,168 31,221
March 33 December 53.42 5.37 46,000 36,248 30,448
March 34 January 56.68 5.56 46,000 36,064 30,294
December 26 February 50.00 5.16 46,000 37,536 31,530
Total annual sales: 388,806
62.5/15 January 20 September 62.11 5.89 63,000 52,920 52,920
January 21 October 63.25 5.96 63,000 52,668 52,668
January 22 November 63.28 5.96 63,000 52,416 52,416
January 23 December 64.18 6.02 63,000 52,164 52,164
February 23 January 62.14 5.89 63,000 52,164 52,164
March 23 February 62.43 5.91 63,000 52,164 52,164
March 24 March 65.28 6.08 63,000 51,912 51,912
July 24 July 62.77 5.93 63,000 51,912 51,912
September 21 June 62.21 5.90 63,000 52,668 52,668
November 17 April 62.13 5.89 63,000 53,676 53,676
December 17 May 62.47 5.91 63,000 53,676 53,676
December 20 August 66.33 6.14 63,000 52,920 52,920
Total annual sales: 631,260
00




Table 4-14 --continued.
Growout Average Size # Clams # Clams # Clams
Seed density/size (months) Size (mm) STD Start Survive Sold
Plant month Harvest month
75/10 January 26 March 50.11 5.17 55,200 45,043 37,836
January 27 April 51.53 5.25 55,200 44,822 37,650
February 27 May 50.30 5.18 55,200 44,822 37,650
February 28 June 50.60 5.20 55,200 44,602 37,466
February 29 July 50.80 5.21 55,200 44,381 37,280
February 30 August 50.98 5.22 55,200 44)160 37,094
February 31 September 51.09 5.23 55,200 43,939 36,909
March 31 October 50.07 5.16 55,200 43,939 36,909
March 32 November 50.13 5.17 55,200 43,718 36,723
March 33 December 50.41 5.18 55,200 43,498 36,538
April 33 January 50.08 5.16 55,200 43,498 36,538
December 26 February 50.60 5.20 55,200 44,043 37,836
Total annual sales: 446,429
75/15 January 17 June 62.31 5.90 75,600 64,411 64,411
January 18 July 63.82 5.99 75,600 64,109 64,109
January 19 August 64.58 6.04 75,600 63,806 63,806
January 21 October 66.13 6.13 75,600 63,202 63,202
January 22 November 67.02 6.19 75,600 62,899 62,899
February 19 September 61.76 5.87 75,600 63,806 63,806
March 21 December 63.07 5.95 75,600 63,202 63,202
April 21 January 62.82 5.93 75,600 63,202 63,202
June 20 February 62.45 5.91 75,600 63,504 63,504
June 21 March 64.54 6.04 75,600 63,202 63,202
November 17 April 63.57 5.98 75,600 64,411 64,411
December 17 May 63.55 5.98 75,600 64,411 64,411
Total annual sales: 764,165




70
Production Performance Comparison
Stabilized production results were used to make performance comparisons between single plant and multiplot production. An operation reached stabilization when both acres had been brought into production, production cycles had achieved optimal patterns (as described in previous sections) and operating loan requirements were zero (as determined by the full economic evaluation discussed in the next section). Optimal repetitive cycles varied by site, production method and marketing constraint. Optimal cycles defined when plants and, more importantly, harvests occurred. Multiplot production specifically required monthly and, once begun, annual harvests. Hence, although individual plants might grow for up to 34 months, optimal repetitive cycles were annual in nature. Single plant production had no such restrictions and, thus, allowed cycle length to be determined by growth and economic dynamics. It was therefore possible to have years where no harvest (or plant) occurred. For example, all single plant production scenarios using 10 mm seed produced cycles that included years in which no harvest occurred due to the length of the growout periods. Site 1 cycles covered a 5-year period of four harvests followed by one year with no harvest. Cycles at sites 2 and 3 using 10 mm. seed covered three years with two harvests followed by a year of inactivity. All scenarios using 15 mm seed produced annual harvests.
Cycle length considerations were used in the determination of summary statistics for model comparison. The consideration of average production allows more direct comparison between the different production methods as it incorporates the negative effect of inactive years due to extended growout times. Not all summary statistics




71
required averaging of this type. Average clam size and growout time were actual performance values and did not require consideration of cycle length.
Average clam size and growout times for both single plant and multiplot production are given in Table 4-15. No consistent patterns emerged when comparing single plant and multiplot results; a given seed size and density might produce larger clams under single plant production than under multiplot production at one site and smaller clams at another site. Average growout times were similarly nonconsistent. Within a given method (single plant versus multiplot) using a given seed size, increased growout times produced larger clams. Five of the 12 multiplot scenarios (three sites and four production options per site) had longer average growout times yet produced larger clams in seven scenarios. Thus, in two occasions, shorter multiplot average growout times produced larger average clams. This could be explained by, as the multiplot production mix typically included both shorter and longer growout periods for different plants, the impact of the larger clams (produced by the longer growout periods) was greater than that of the smaller clams (produced by the shorter growout periods). A comparison of the annual number of clams planted, survive and sold is shown in Table 4-16. Patterns were again difficult to discern. Multiplot production generally resulted in greater numbers of clams planted (10 of 12 situations) and sold (9 of 12 situations). Plant quantities varied due to differences in total bags per acre and the effects of extended growout periods and plot subdivision. Percentage survival, however, could be higher/lower yet produce fewer/greater sales. The production of undersized clams was indicated when the quantity of clams sold was less than the quantity survived. With only




72
Table 4-15. Average final clam size (mm) and growout time (months) for single plant and multiplot hard clam production.
Seed density/ Average Average
Seed size Size Growout time
Single plant production
Site 1 62.5/10 57.32 28
62.5/15 61.79 22
75/10 51.12 24.5
75/15 64.37 20
Site 2 62.5/10 62.05 33
62.5/15 62.91 23
75/10 51.15 31
75/15 61.80 19
Site 3 62.5/10 61.77 32
62.5/15 62.11 20
75/10 51.09 31
75/15 61.77 19
Multiplot production
Site 1 62.5/10 63.79 30.2
62.5/15 64.95 20.1
75/10 51.58 25.8
75/15 64.29 18.7
Site 2 62.5/10 53.73 30.1
62.5/15 64.45 22.2
75/10 50.81 29.6
75/15 64.14 19.4
Site 3 62.5/10 55.12 29.8
62.5/15 63.22 21.3
75/10 50.56 29.4
75/15 63.80 19.4




Table 4-16. Average annual numbers of clams planted, survive and sold for single plant and multiplot production.
Seed density! % Sold % Sold
Seed size Plant Survive Sold Survival (of plant) (of survive)
Single plant production
Site 1 62.5/10 600,000 484,800 446,208 80.8 74.4 92.0
62.5/15 750,000 624,000 624,000 83.2 83.2 100.0
75/10 720,000 591,840 497,146 82.2 69.1 84.0
75/15 900,000 756,000 756,000 84.0 84.0 100.0
Site 2 62.5/10 500,000 394,000 334,320 78.8 66.9 84.9
62.5/15 750,000 621,000 621,000 82.8 82.8 100.0
75/10 600,000 477,600 401,184 79.6 66.9 84.0
75/15 900,000 759,600 759,600 84.4 84.4 100.0
Site 3 62.5/10 500,000 396,000 396,000 79.2 79.2 100.0
62.5/15 750,000 630,000 630,000 84.0 84.0 100.0
75/10 600,000 477,600 401,184 79.6 66.9 84.0
75/15 900,000 759,600 759,600 84.4 84.4 100.0




Table 4-16--continued.
Seed density/ ###%% Sold % Sold
Seed size Plant Survive Sold Survival (of plant) (of survive)
Multiplot production
Site 1 62.5/10 540,000 431,640 425,916 79.9 78.9 98.7
62.5/15 756,000 634,788 634,788 84.0 84.0 100.0
75/10 777,600 635,299 533,652 81.7 68.6 84.0
75/15 907,200 766,886 766,886 84.5 84.5 100.0
Site 2 62.5/10 540,000 431,820 362,729 80.0 67.2 84.0
62.5/15 720,000 598,560 598,560 83.1 83.1 100.0
75/10 648,000 519,480 436,363 80.2 67.3 84.0
75/15 907,200 764,165 764,165 84.2 84.2 100.0
Site 3 62.5/10 552,000 442,152 388,806 80.0 70.4 87.9
62.5/15 756,000 631,260 631,260 83.5 83.5 100.0
75/10 662,400 530,465 446,429 80.1 67.4 84.0
75/15 907,200 764,165 764,165 84.2 84.2 100.0




75
one exception, undersized clams were harvested with the same production methods in both single plant and multiplot production.
It should be noted that the above comparisons are intended to simply describe some of the production results. Since determinations of preferred or optimal cycles were based on economic criteria, explanations and an understanding of why specific results might occur require considerations of the cost and revenue impacts of the various production options.
Economic Performance
The hard clam growth simulation and the identification of optimal rotations described in the previous sections identified and described potential challengers. A logical application of this information and the simulation models employed would be to take an existing growout operation, identify the profile of existing stock, and determine harvest and replacement schedules for existing stock given knowledge of challengers and the expected revenues from existing stock. Despite being extremely situation specific, such an application would be descriptive of the use to which this research might be applied as a management or extension tool.
A second application, however, involves applying the previously described optimal rotations to a new growout operation and determining the expected returns and accrued benefits of the resultant operation. This second application was chosen for the remainder of the research. The rationale for this selection was that, if expected conditions occurred, such an evaluation provided more insight into what an optimally managed operation might be worth. With the first application, once existing stock is




76
replaced and expected conditions again occur, an existing operation would likely fall into an identical new-operation production pattern. The onset of the pattern would, however, be delayed according to the continued growout requirements of existing stock. If expected conditions do not occur, i.e. expected environmental conditions fail to occur, expected conditions fail to produce expected growth, or expected prices fail to be realized, then the system is in a constant state of flux requiring continuous evaluation, and the system likely never attains repetitive patterns.
The optimal rotations for both production methods described both planting and harvesting patterns for one half of the lease (one acre). This pattern was then repeated on the second acre, thereby determining planting and harvest management of the entire lease. These management patterns were used to determine asset replacement and operating expenditure requirements. The expenditure requirements were then used to determine fixed and variable costs according to guidelines described in Adams et al. (1993). The operation described in Adams et al. and modelled here assumes that the owner-operator is employed full time in some other occupation and that the culture operation exists as a part-time activity. Alternative income activities are assumed limited to minimum wage jobs. Asset and production requirements were combined with expected revenues in a cashflow model to determine annual loan requirements, finance payments, taxes, net positions, etc. Financial assumptions common to both production scenarios were
(a) Sixty-five percent of initial investment requirements and asset replacement
is financed by borrowed capital.
(b) All operating loans are fully financed by borrowed funds.




77
(c) All debt capital has a 9.5 % real annual rate of interest over a ten-year loan
period.
(d) Depreciation on capital assets is computed using straight-line depreciation.
(e) The wage rate is $4. 10 per hour.
(f) Monitoring costs are $120 per acre per month.
(g) Growout bags are harvested at the rate of two bags per hour. This
includes pulling, grading and bagging (sales bags).
The debt capital interest rate assumes a 4 % real interest rate and a 5.5 % risk premium. Initial investment costs are listed in Appendix 4 and production costs are listed in Appendix 5.
Income statements
Income statements are developed using the results of the cashfiow analysis. Average expectations of stabilized production, as described in the previous section on production comparisons, were determined. Again, an operation reached stabilization when full production of the lease had begun, production cycles had achieved optimal repetitive patterns and operating loan requirements were zero. Values used in the income statements thus represented averages over the repetitive cycle.
Returns net of variable and fixed costs represent the returns to the land and owner labor, management and capital. The opportunity cost of owner labor was calculated by determining the annual number of labor hours required and assuming the next best alternative was a minimum wage job. The opportunity cost of capital was calculated at 11. 5 % of owned equity. This interest rate included a 2 % premium over the loan rate.




78
The next step, required identifying the residual claimant and quantifying the residual. The residual claimant is that component to which final net returns or residual accrues. Both management and land might be considered valid residual claimants. Selection of the residual claimant requires a thorough understanding of the productive process examined, and consideration of the goals of the research and the ability to realistically isolate the contributions of specific inputs. Although management of a small hard clam growout operation is a new endeavor, precedent exists for estimating management fees in other agricultural fields as a percentage of gross or net returns. Management fees ranging from three to 10 percent are used in grain, vegetable and citrus operations with the higher percentages usually awarded for the more labor intensive crops (various personal communications). Most management payments likely include base salaries with performance incentives. The lack of a history of hard clam management and the difficulty of the identification of available alternatives, however, makes detailed estimation of management fees difficult. Percentage calculation is, therefore, appealing.
Valuation of the land is equally problematic. The argument could be made that, since a lease fee structure already exists, these costs should be deducted, thereby attributing the residual to management. Current lease fees, however, are only $20 per acre per year, and thus likely represent only filing or paperwork fees. This fee thus fails to represent the unique productive capacity of the land. Considerations of alternative commercially productive use values is likely limited to oyster culture, an emerging operation similar to clam culture and one in which evaluations of economic potential are similarly incomplete.




79
Given the above considerations, the decision was made to make the land the residual claimant. Production results indicated that returns were driven by site specific characteristics. The same management philosophy or decision rules were applied across all production options at all three sites. Actual management in terms of specific planting and harvest schedules varied across sites, but the rules determining or identifying optimal rotations were identical. Returns, however, varied by site. Thus, the land was chosen as the residual claimant. Returns to management were calculated as 3% of gross revenues. Residual returns represented the returns to the lease or two acres of water bottom. The results of this evaluation for single plant production are shown in Tables 4-17, 4-18 and 4-19. Multiplot production financial results are shown in Tables 4-20, 4-21 and 4-22.
Examination of the income statements showed that, while differences in performance existed, all scenarios produced significant residual returns ranging from a low of $4,000 per lease under multiplot production at Site 2 using 10-mm seed planted at 62.5 clams per square foot, to a high of $33,400 per lease at Sites 2 and 3 under single plant production using 15-mm seed planted at 75 clams per square foot. Net present and annualized values
The figures presented in the income statements represented expected performance of the operation after operating loans are retired and the planting schedules achieve an optimal cycle. Thus, the income statements fail to incorporate performance information prior to the achieved stability. Two other measurements, the net present value and the annualized value, capture these off-year effects and are therefore more descriptive. The




80
Table 4-17. Site 1 single plant production income statement.
Production method (seed density/seed size) 62.5/10 62.5/15 75/10 75/15
Revenue $56,000 $75,900 $62,100 $97,100
Expenses
Variable cost 13,200 19,700 15,100 22,800
Fixed cost
Overhead 1,000 1,000 1,000 1,000
Debt interest 3,100 2,200 3,000 2,300
Depreciation 8,300 6,200 7,800 6,200
Taxes 8,400 13,100 11,200 17,900
Net return to labor, management, 22,000 33,700 24,000 46,900
capital and land
Opportunity cost of labor 2,900 3,600 2,900 3,600
Opportunity cost of capital 7,300 5,000 5,700 15,700
Opportunity cost of management 1,700 2,300 1,900 2,900
Net return to land 10,100 22,800 13,500 24,700




81
Table 4-18. Site 2 single plant production income statement.
Production method (seed density/seed size) 62.5/10 62.5/ 15 75/10 75/15
Revenue $50,600 $75,500 $52,200 $97,500
Expenses
Variable cost 11,500 19,700 13,000 22,800
Fixed cost
Overhead 1,000 1,000 1,000 1,000
Debt interest 2,900 2,200 2,700 2,200
Depreciation 7,200 6,200 7,200 6,200
Taxes 9,100 13,000 10,500 18,300
Net return to labor, management, 18,900 33,400 17,800 47,000
capital and land
Opportunity cost of labor 2,400 3,600 2,400 3,600
Opportunity cost of capital 6,600 4,900 3,700 7,100
Opportunity cost of management 1,500 2,300 1,600 2,900
Net return to land 8,400 22,600 10,100 33,400




82
Table 4-19. Site 3 single plant production income statement.
Production method (seed density/seed size) 62.5/10 62.5/15 75/10 75/15
Revenue $50,800 $80,900 $52,200 $97,500
Expenses
Variable cost 11,500 19,700 13,000 22,800
Fixed cost
Overhead 1,000 1,000 1,000 1,000
Debt interest 2,900 2,200 2,700 2,200
Depreciation 8,300 6,200 7,200 6,200
Taxes 10,100 14,500 10,500 18,300
Net return to labor, management, 17,000 37,300 17,800 47,000
capital and land
Opportunity cost of labor 2,400 3,600 2,400 3,600
Opportunity cost of capital 3,600 5,600 3,700 7,100
Opportunity cost of management 1,500 2,400 1,600 2,900
Net return to land 9,500 25,700 10,100 33,400




83
Table 4-20. Site 1 multiplot production income statement.
Production method (seed density/seed size) 62.5/10 62.5/15 75/10 75/15
Revenue $48,400 $72,100 $60,400 $89,600
Expenses
Variable cost 11,100 21,300 14,700 21,400
Fixed cost
Overhead 1,000 1,000 1,000 1,000
Debt interest 2,800 2,200 3,100 2,200
Depreciation 7,600 6,200 8,500 6,200
Taxes 7,300 12,400 9,300 16,500
Net return to labor, management, 18,600 28,800 23,800 42,300 capital and land
Opportunity cost of labor 3,000 3,700 3,300 3,700
Opportunity cost of capital 7,600 6,800 8,300 11,400
Opportunity cost of management 1,500 2,200 1,800 2,700
Net return to land 6,500 16,200 10,400 24,500




84
Table 4-21. Site 2 multiplot production income statement.
Production method (seed density/seed size) 62.5/10 62.5/15 75/10 75/15
Revenue $41,400 $69,500 $49,100 $88,700
Expenses
Variable cost 11,000 17,500 12,700 21,400
Fixed cost
Overhead 1,000 1,000 1,000 1,000
Debt interest 2,800 2,300 2,800 2,200
Depreciation 7,600 6,100 7,600 6,200
Taxes 5,300 11,900 7,000 16,300
Net return to labor, management, 13,700 30,700 18,000 41,600
capital and land
Opportunity cost of labor 3,000 3,500 3,000 3,700
Opportunity cost of capital 5,500 9,600 6,700 9,300
Opportunity cost of management 1,200 2,100 1,500 2,700
Net return to land 4,000 15,500 6,700 25,900




85
Table 4-22. Site-3 multiplot production income statement.
Production method (seed density/seed size) 62.5/10 62.5/15 75/10 75/15
Revenue $44,300 $72,800 $50,200 $88,700
Expenses
Variable cost 11,200 18,300 12,900 21,400
Fixed cost
Overhead 1,000 1,000 1,000 1,000
Debt interest 2,900 2,300 2,900 2,200
Depreciation 7,700 6,200 7,700 6,200
Taxes 6,000 12,600 7,200 16,300
Net return to labor, management, 15,500 32,400 18,500 41,600
capital and land
Opportunity cost of labor 3,000 3,700 3,000 3,700
Opportunity cost of capital 5,500 8,900 6,900 9,300
Opportunity cost of management 1,300 2,200 1,500 2,700
Net return to land 5,700 17,600 7,100 25,900




86
net present value represents net returns--calculated from initial start-up--discounted over 20 years at 10.2%. The present value per acre represents the amount a person should be willing to pay to obtain a 2-acre 20-year lease. The annualized value represents an average annual return to the lease over the same 20-year period and is comparable to the annual lease fee--$40 for two acres--for performance evaluation. The annualized value is the amount a person should be willing to pay annually for the 2-acre lease. The discount rate used represented an average of the loan rate and the opportunity cost of capital rate weighted by the capital asset financing ratio previously specified. The present value of the land for all production scenarios is given in Table 4-23 and the annualized values are given in Table 4-24.
A comparison of the present value per lease (2 acres of land) for single plant and multiplot production show values from single plant production to exceed those from multiplot production in all cases. Differences are minor, however, at Site 1 planting 15mm seed at 75 clams per square foot. An examination of the income statements of the two scenarios shows that, while the gross revenues for single plant production exceed those of multiplot production, $97,100 to $89,600, a considerable portion of this difference, $4,300, is negated when the opportunity cost of capital is accounted for. Further evaluation of the two situations using production information not presented here attributed the higher opportunity cost of capital with single plant production to a faster rate of equity build-up. Specifically, on January 1 of any given year, the date on which evaluations were based, both acres were fully planted under single plant production whereas only 1.33 acres were planted under multiplot production. At this particular site,




87
Table 4-23. Present value of a two-acre lease across all sites and production options.
Seed density/seed size
62.5/10 62.5/15 75/10 75/15
Site 1
Single plant $19,700 $145,200 $42,300 $141,300
Multiplot -9,600 83,300 13,800 140,900
Site 2
Single plant 11,900 143,400 51,000 220,300
Multiplot -21,300 39,100 -5,500 147,900
Site 3
Single plant 47,400 166,100 51,000 220,300
Multiplot -8,400 74,100 -3,900 147,900
Table 4-24. Annualized value of a two-acre lease across all sites and production options.
Seed density/seed size
62.5/10 62.5/15 75/10 75/15
Site 1
Single plant $1,700 $12,200 $3,600 $11,900
Multiplot -800 7,000 1,200 11,800
Site 2
Single plant 1,000 12,000 4,300 18,500
Multiplot -1,800 3,300 -500 12,400
Site 3
Single plant 4,000 14,000 4,300 18,500
Multiplot -700 6,200 -300 12,400




88
there was a concentration of monthly plants early in the year. Three plants were scheduled for January and four for February. Thus, seven plots were empty at the point of evaluation.
The results confirm expectations that harvest restrictions impair the economic performance of the operation. Imposing harvest restrictions leads to either or both planting and harvesting clams other than when otherwise prescribed in the absence of such restrictions. This can produce economically immature and overage clams as well as increased mortality. Monthly marketing reduced revenues sufficiently that in only one instance, at Site 1 planting at 75 clams per square foot, were 10-mm seed profitable. Revenues were never sufficient to overcome initial negative net balances.
Larger seed always outperformed smaller seed, while higher planting densities almost always outperformed lower planting densities. The one exception occurred at Site I where the lower density using 15-mm clams outperformed the higher density. Reasons can again be traced to equity build-up. See Table 4-17. The greatest returns were achieved by using the largest seed planted at the highest density at Sites 2 and 3. The major cost impact of using larger seed came from higher seed costs. These costs are more than recouped, though, through faster turnover of stock. Similarly, although growth was affected by planting density with higher densities producing more conservative growth, the negative effects of the higher density were not sufficient to offset the gains attributed to larger volume sales.
Selection of the best site varied with production method. Each site represented physically distinct locations possessing unique environmental profiles in terms of




89
expected water temperature, salinity and dissolved oxygen. The results indicate that site selection is an important factor to consider in hard clam growout. How to effectively incorporate this information in management decisions is less clear. While ambient environmental conditions varied from site to site, determination of combinations that can be classified as productively similar or different is difficult. Also, identical productive or economic potential depended upon both production method and harvest restriction. Sites 2 and 3 produced identical economic outcomes under single plant production at 75 clams per square foot, but not at 62.5 clams per square foot. Further, under multiplot production, Sites 2 and 3 produced identical returns using 15-mm seed planted only at the higher density. Thus, the incorporation of environmental considerations is difficult.
Trends evidenced by the net present values of the various production options are repeated in the annualized values. The primary benefit of the annualized values is the ease of comparison with current lease fees. As can be seen from Table 4-23, with the exception of the plantings under multiplot production using 10-mm seed, all other scenarios produce values much larger than the current fee of $40 (for two acres).
A final analysis examined the impact of lower prices. Prices were uniformly (across all size categories) reduced 10, 20 and 30%. The evaluation was performed on Site 2, single plant production. Net present value and annualized value results are given in Table 4-25. As can be seen from the results, the price reduction eventually totally erodes the profitability of production using 10-mm seed. Production using 15-mm seed, however, still produces considerable returns, though net present value decreases with price reductions at a rate of two-to-one or greater.




90
Table 4-25. Net present value (NPV) and annualized value (AV) of a two-acre lease at Site 2 under uniform price reduction. Price Seed density/seed size
Reduction 62.5/10 62.5/15 75/10 75/15
10 %
NPV -$9,100 $107,700 $28,400 $174,500
AV -800 9,000 2,400 14,700
20 %
NPV -30,200 72,100 5,700 128,100
AV -2,500 6,100 500 10,800
30 %
NPV -51,100 37,700 -17,500 82,900
AV -4,300 3,200 -1,500 7,000




CHAPTER 5
SUMMARY AND CONCLUSIONS
summau
The determination of optimal hard clam growout production design incorporating the selection of seed size, planting density, plant scheduling and replacement timing depends on consideration of the complicated interactions of monthly clam growth and price relationships. Optimal scheduling depends on the ability to predict future conditions of growth, mortality and price. Growth prediction for this research was accomplished through the development of a hard clam growth function modelling periodic growth as a function of initial clam size, age, and water temperature, salinity and dissolved oxygen. Growth function parameter estimates were made for two planting densities. Stochastic hard clam growth was simulated for two sizes of seed clam planted at two densities using environmental values from three sites in the Indian River area of Florida. Monthly mortality was imposed in a deterministic manner using a terminal base mortality and weight system such that the mortality of younger clams was greater than that of older clams.
The results of the hard clam growth simulation were used to estimate expected net revenues by incorporating average monthly clam prices for four size categories.
91




92
Clam prices were determined from time series price data. The expected revenues were then examined to determine optimal planting months and growout times. Two marketing arrangements were examined, the first allowing all clams to be planted and harvested as a single unit (single plant production), and the second requiring monthly harvests (multiplot production). Evaluations of the results produced specific plant, harvest and replacement schedules that varied with production method (seed size and density) and site location. Optimal rotations determined total operational input requirements and cost and revenue flows. These were then used to determine the residual value of the lease, the present value of the residual stream and an annualized value of the residual stream.
Comparison of the economic performance of each production method at each site showed single plant production to always outperform multiplot production. Larger seed planted at the higher density outperformed all other options in all but one case. Single plant production allowed clams to be planted and harvested such that expected net returns were maximized. Multiplot production caused clams to be planted and/or harvested in months other than those indicated by less restrictive optimization. The higher returns to larger seed planted at greater densities were attributed to a shorter growout period generating faster turnover and higher volume sales. The one case where the lower density outperformed the higher density was attributed to higher opportunity costs resulting from faster equity build-up with the higher density.




93
Conclusions and Implications
The results of this study must be put in proper perspective. Various assumptions were required to construct a representative operation. Individual features are presented as neither concrete absolutes nor unrealistic options. They are reasonable assumptions as determined by scholarly research and through contacts with industry professionals. Actual values or requirements for individual producers are expected to be both higher and lower. The purpose of the research is not to determine precisely what can be accomplished, but rather to provide a foundation for establishing and directing further areas of emphasis and consideration. Where consideration has not been given to seed density, planting month, site location, etc., this research provides justification for such. It is in this context that the results should be viewed.
The economic potential of hard clam growout is a complicated interaction of growth, as dictated by various environmental factors, and variable prices, as determined by clam size and month of harvest. Higher/lower densities may produce slower/faster growth while smaller/larger seed produce longer/shorter growout times. These only have economic relevance when combined with cost and revenue considerations. It is inaccurate and insufficient to say that a faster growth rate or a larger seed is preferred without knowledge of what effect the growth rate or seed size has on when clams reach market size and what prices might be expected. Further, while prices might be uniform across a region, growth conditions likely vary from site to site making optimal operation design specific to a given site.




94
Comparisons of the results of this study with others is difficult because of differences in production and financial assumptions. Thunberg and Adams (1990) estimate fifth-year net returns for two million seed (one million clams sold), a 3-year growout, and a $0.14 market price at $71,840. The comparable result from the current research is $17,800 to $24,000 for single plant production using 10-mm clams planted at 75 clams per square foot (900,000 clams planted and 602,000 to 623,000 sold) and sold at $0.10 per clam (see Tables 4-6, 4-17, 4-18 and 4-19). Ignoring the clam volume differential of the two studies, the results by Thunberg and Adams include $40,000 attributable to the higher price and do not subtract taxes which would be in excess of $20,000 under the assumptions of the current study. A more direct comparison is possible with Adams et al. (1993), as they assume a 2-year growout and $0.10 per clam. Net returns from clams planted at 75 clams per square foot are $32,807 for 765,000 clams sold. The comparable results from this study are $47,000 for single plant production (756,000 to 760,000 clams sold) and approximately $42,000 for multiplot production (764,000 to 767,000 clams sold), both using 15-mm seed planted at 75 clams per square foot (see Tables 4-12, 4-17, 4-18, 4-19, 4-20, 4-21 and 4-22). Thus, comparisons of this research with other studies indicate that returns can be increased by careful attention to production method and timing.
The results of this study also indicate that, in the absence of mitigating circumstances, the grower should plant larger seed and at the highest recommended density. Fifteen-millimeter seed may not be available in all areas, or it may be available, but only at prohibitive prices. While this research did not conduct seed price sensitivity




Full Text

PAGE 1

BIOECONOMIC MODELLING OF HARD CLAM GROWOUT IN FLORIDA: THE REPLACEMENT DECISION By STEPHEN GLENN HOLIMAN A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1993

PAGE 2

ACKNOWLEDGEMENTS I would like to thank Dr. Thomas Spreen for his patient guidance and nurturing those many times when I began to lose focus and direction. Thanks go to Dr. Eric Thunberg for the financial support, the enthusiasm and the patience during the lost summer of family medical problems. Thanks also go out to Dr. Charles Adams for his office wit and assistance with the clam budgets, Dr. William Boggess and Dr. Richard Weldon for their patient response to financial questions, and to Dr. Charles Cichra for bearing up under numerous "well, economists do it this way." The crew at Project Ocean, notably Dr. David Vaughan and Leslie Sturmer, also deserve thanks for various things not the least of which were those 100-odd deliciously sweet clams my wife and I got to eat. A huge thank-you also goes out to the anonymous commercial clam farm which graciously provided the growth data. The biggest thanks, however, must go to my wife and parents. My wife was as patient as she could be given all that I have put her through. My father, Stanley M. Holiman, while not quite understanding the Ph.D. process, was always there for support. My biggest regret is that my mother, Bonnie Gene Holiman, could not be here to see it all complete. She claimed she did not want me to finish because then we would have to move away. But I know that both she and my dad were proud of what I was doing and that her spirit is still here with us, even as I write this. Thank you all. 11

PAGE 3

TABLE OF CONTENTS ACKNOWLEDGEMENTS . . . . . . . . 11 ABSTRACT ...... ..................... .. .. .. ... V CHAPTER 1 INTRODUCTION . . . . . . . 1 The Management Problem: Replacement Under Risk .................................. 4 Research Objectives . . . . . . 6 Organization of the Dissertation . . . . 7 2 HARD CLAM AQUACULTURE . . . . . 8 Introduction . . . . . . . 8 Hatchery Phase . . . . . . . 8 Nursery Phase ............................. 13 Growout Phase ..... .. ...... .. ....... ... 20 Harvest ..... ..................... .. 21 3 BIOECONOMIC MODEL . . . . . .... 23 Economic Model . . . . . . 23 Hard Clam Growth and Mortality Models . . . 30 Bivalve Growth Literature . . . . .... 31 Growth Model . . . . . . . 33 Estimation of Mortality . . . . . 40 4 RESULTS AND DISCUSSION .................... .. 43 Growth Simulation ............. .. .. ... .. 43 Mortality . . . . . . . . 48 Economic Evaluation . . . . . . 50 Single Plant Production Optimal Rotations . . 51 Multiplot Production Optimal Rotations . . . 59 Production Performance Comparison . . ... 70 Economic Performance . . . . . 75 iii

PAGE 4

Income Statements . . . ........... 77 Net Present and Annualized Values .............. 79 5 SUMMARY AND CONCLUSIONS ............ ...... .. 91 APPENDIX Summary ................................ 91 Conclusions and Implications . . . . . 93 Limitations and Suggestions For Further Research . . . . . . . 98 1 PRODUCTION AND ENVIRONMENT AL DATA USED IN GROWTH MODEL ESTIMATION . . . 102 2 ENVIRONMENT AL VALUES USED IN GROWTH SIMULATION ........................... 147 3 HARD CLAM EXPECTED MEAN PRICES ............. 150 4 INITIAL INVESTMENT REQUIREMENTS FOR HARD CLAM BOTTOM BAG GROWOUT . . . . 151 5 PRODUCTION COSTS FOR HARD CLAM BOTTOM BAG GROWOUT ......................... 152 REFERENCES . . . . . . . . . .... 153 BIOGRAPHICAL SKETCH ................................. 156 IV

PAGE 5

Abstract of Dissertation Presented to the Graduate School o f the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy BIOECONOMIC MODELLING OF HARD CLAM GROWOUT IN FLORIDA: THE REPLACEMENT DECISION By Stephen Glenn Holiman December 1993 Chairman: Dr. Thomas H. Spreen Major Department: Food and Resource Economics Department The choice of production method and timing are examined to determine their effect on optimal growout and replacement scheduling of hard clams given probabilistic clam growth and variable monthly clam prices. Hard clam growth is modelled as a function of initial clam size and water temperature, salinity and dissolved oxygen. The environmental parameters take on probabilistic values, hence determining probabilistic growth. Growth function coefficients are estimated for 2 clam planting densities and the resultant functions used to simulate clam growth of 10 and 15 millimeter clam seed planted at 62.5 and 75 clams per square foot over a 34-month period beginning in each of the 12 calendar months. Growth simulation results are then combined with mortality assumptions, production restrictions (lease size and planting capacity) and size-dependent price expectations t o estimate expected returns. These returns are then used to determine the best production method (seed size and planting density), planting and replacement V

PAGE 6

schedule and maximum returns. The expected returns are then used to estimate the net present value of the lease. Results indicate that where possible, growers should purchase larger seed and plant at the higher density and that, regardless of production method used, plant and harvest scheduling requires special attention. Growout times, replacement schedules and expected revenues vary by site location as determined by the specific environmental conditions of each site. The returns indicate that current lease fee requirements substantially undervalue the productive potential of the lease. The potential returns of hard clam growout are site specific thus suggesting the importance of site selection and suggesting variable lease fees dependent upon site potential. Justifying such a fee structure, however requires additional research on identifying the specific environmental profiles that generate different economic potential. VI

PAGE 7

CHAPTER 1 INTRODUCTION Aquaculture is increasingly seen as a means of augmenting the supply of commercially important aquatic species. Under certain conditions, aquaculture may possess a comparative advantage over wild harvest (Shang, 1981). Wild populations may be widely dispersed due to natural ecological dynamics or as a result of harvest pressure. Dispersion affects per unit harvest costs by increasing search time, labor requirements, fuel use, etc. Aquaculture may produce lower per unit costs by concentrating the target species in a confined and more accessible location. Genetic selection combined with controlled feeding and environmental conditions can improve yields relative to natural production. Aquaculture can also allow suppliers to mitigate seasonally fluctuating wild catch and guarantee delivery with greater certainty than when dependent upon wild harvest. The Florida hard clam aquaculture industry is an example of an emerging aquaculture industry. Two species of hard clam are native to Florida, Mercenaria mercenaria, the northern hard clam, and Mercenaria campechiensis, the southern hard clam (Vaughan et al., 1988). Natural territories of the two species overlap. M. mercenaria is found from the Gulf of St. Lawrence, Canada, to the northern Gulf of Mexico, with the center of abundance from Massachusetts to Virginia. M. campechiensis 1

PAGE 8

2 is found from Cape May, New Jersey, to Campeche Mexico, with the center of abundance in southwest Florida. Some hybridization occurs where the species overlap, but the two species typically prefer different habitats with M. mercenaria being an estuarine interto subtidal species and M. campechiensis preferring deeper, higher salinity waters (Malouf and Bricelj, 1989). Commercially exploited populations are typically M. mercenaria, as this is the more abundant species and it lives in more accessible waters. Hard clams can live for 23 years or more and achieve a length in excess of 135 millimeters (mm) (Malouf and Bricelj, 1989). Wild hard clams have historically been found in Florida waters, but large scale harvest and culture have typically been confined to mid-Atlantic and north Atlantic coastal regions (Manzi and Castagna, 1989). From 1973 to 1983 annual Florida hard clam landings from wild stocks averaged 107.54 thousand pounds of meat compared to total U.S. average annual landings of 14 362 million pounds (Adams et al., 1991). See Table 1-1. In the early 1980s however a large natural set of clams in the Indian River Lagoon resulted in average annual harvests of 1.366 million pounds of clam meat from 1984 through 1987 (Adams et al. 1991). Total annual U.S. landings over the same period averaged 13. 664 million pounds of meat. This represented an increase in Florida landings as a percentage of total U S. landings from less than 1 % during 1974-83 to almost 10% during 1984-87. Harvests eventually declined to 711 thousand pounds in 1988 however due to a combination of harvest pressure and changing environmental conditions. Total U.S. hard clam landings in 1988 were 12.371 million pounds of meat,

PAGE 9

3 Table 1-1. Hard Clam Landings, 1973-88. Share Year Florida U.S. of U.S. ---Thousand pounds of meat--Percent 1973 139 14,505 0.95 1974 94 14,665 0.64 1975 74 14,995 0.49 1976 61 15,251 0.40 1977 148 14,690 1.00 1978 126 13 295 0.94 1979 72 12 058 0.59 1980 62 13,370 0.46 1981 117 18,118 0.64 1982 145 12,855 1.12 1983 145 14 186 1.02 1984 1377 14,749 9.33 1985 1441 16 697 8.63 1986 1448 11 793 12.28 1987 1197 11 418 10.48 1988 711 12 371 5.47 Source: 1973-84 NMFS ; 1985-88 FDNR.

PAGE 10

4 with Florida production comprising 5. 7% of the total. Florida's wild clam harvest has not rebounded to mid-1980 levels. The ability of Florida waters to support such large wild harvests and the ability of regional markets to absorb the harvests caused many to consider the potential of culturing clams in Florida. Particular note was given to the suitability of Florida's natural environmental conditions relative to hard clam production. Clam growth is temperature sensitive. Lower water temperatures depress growth while the converse is true, to a point, for warmer water temperatures (Manzi and Castagna, 1989). An accelerated growth rate should allow increased clam biomass production per unit of area per unit of time and, consequently, increased revenues. The Management Problem: Replacement Under Risk The decision to undertake hard clam aquaculture requires the evaluation of numerous issues. As will be described in Chapter 2, hard clam aquaculture is an integrated process consisting of hatchery, nursery and growout phases. A commercial culturist may choose to operate at all levels of production or decide to specialize on a particular level. Hence, a choice of level of integration must be made. Similarly, multiple production technologies exist for each production stage, so a technology adoption decision is required. Next, for each choice of production technology, options exist relative to production scheduling, size of clam seed, planting densities, monitoring schedules, etc. Further, decisions are required on marketing strategy. Specific strategies

PAGE 11

5 require decisions concerning whether to target seasonal markets, what size clams to market, and where to market. As will become evident in Chapter 2, the productive capacity, capital requirements and labor intensity of the hatchery and nursery phases of hard clam aquaculture make them less suitable than growout culture for operations owned, managed, and worked by a single individual. The research presented in this dissertation focuses on growout culture. Management options for hard clam growout include the selection of seed size, choice of planting method, clam density and replacement scheduling. The evaluation of management options is complicated where risk is encountered. Risk is inherent in any system where outcomes are not guaranteed. This applies to both production and financial outcomes. Stochastic production may result in variable output quality and quantity. Variable output produces variable revenue. Stochastic prices further increase the variability of revenue. Production uncertainty has particular relevance in hard clam aquaculture as prices are higher in the smaller legal size categories and lower for larger sized clams. Consumer preference results in a price structure where a price penalty is incurred for larger size classes. Hence, the culturist is concerned that his clams not grow out of the higher-priced size classes. The culturist must determine where risk is likely to occur. Then, risk reducing options must be identified. The adoption of specific options will then be dependent on the relative costs of implementation and the subsequent rewards of the reduced risk.

PAGE 12

6 Research on the costs and returns of hard clam growout aquaculture has been limited to systems and growing conditions representative of the South Atlantic region (Adams et al., 1991). No comprehensive work has been conducted on systems and conditions specific to Florida. Current studies (Adams et al., 1991; Thunberg and Adams, 1990) do not incorporate clam price and yield variability other than through basic sensitivity analysis. Sensitivity analysis simply changes an outcome without examining the likelihood of that outcome actually occurring. The true consideration of risk reflects both its impact and the likelihood of occurrence. A decision to undertake hard clam growout aquaculture in Florida, therefore, requires knowledge of appropriate production systems, sources of operation risk, the effects of risk on operation design and replacement scheduling, and estimates of the costs and returns of hard clam growout under risk. Research Objectives The objectives of this research are 1. Develop a bioeconomic model of Florida hard clam growout. 2. Generate cost and return estimates of hard clam growout under different scenarios of growth and price variability to provide insights to optimal operation design and management.

PAGE 13

7 Organization of the Dissertation Chapter 2 provides a description of hard clam aquaculture in Florida. Theoretical consideration, data sources and the bioeconomic model are presented and discussed in Chapter 3. The results of the application of the bioeconomic model are given in Chapter 4. A summary conclusions study limitations and suggestions for future research comprise Chapter 5.

PAGE 14

CHAPTER 2 HARD CLAM AQUACULTURE Introduction Hard clam aquaculture consists of three phases--hatchery, nursery and growout (Manzi and Castagna, 1989). Only the first two phases typically entail controlled environments where specific growing conditions are maintained. In the following discussion, current practices and key issues for Florida hard clam aquaculture are described. Hatchery Phase It has been estimated that 40 % of hatchery operating costs are for the production of algae or one-celled plants (phytoplankters) for hard clam food (Hartman, 1989). Two primary methods of algal culture, the Glancy and Milford methods, are used for hard clam aquaculture (Castagna and Manzi, 1989). The Glancy Method filters or clarifies seawater to remove predatory zooplankters and large phytoplankters. Treated seawater is kept in shallow, gently aerated tanks, exposed to natural or artificial light. The algae is normally fed to the clams within 48 hours before larger, less digestible phytoplankters 8

PAGE 15

9 dominate. The Glancy Method works best in moderate climates and where an abundance of natural phytoplankton exists. The Milford Method relies on the controlled production of selected species of phytoplankters using sterile media and growth promotents. Pure cultures of single algal species are produced. Total harvest and replacement of the algae is practiced to prevent contamination. Since seawater is the medium for both clam and algal growth, the success of a hard clam hatchery is highly dependent upon the availability of water of a suitable quality. The variables of specific concern to the culturist are water temperature, salinity, dissolved oxygen, chemical or bacterial contamination, and algal and zooplankton content. Water quality can be manipulated by the culturist, but it may not be cost. effective to do so. Larval rearing requires water temperatures of 25-300 C (Adams et al., 1991), salinity of 26-27 parts per thousand (ppt) (Eversole, 1987), and dissolved oxygen levels of 6. 87.4 milligrams per liter (Eversole, 1987). Larval growth is fastest at 30 C, a temperature which also promotes high bacterial contamination (Menzel, 1989). Proper salinity and dissolved oxygen levels are more critical for larval and juvenile clams than for older clams. Older clams are able to remain closed for longer periods and rely on various metabolic mechanisms to reduce oxygen requirements during periods of environmental stress (Eversole, 1987). In the hatchery phase, sexually mature hard clams are induced to spawn and produce fertilized eggs. Broodstock are initially selected from wild stock possessing desired characteristics such as large size, or of a special color form or marking pattern

PAGE 16

10 called notata. Notata markings are brown zig zag patterns in the shell as pictured in Figure 2-1. Notata patterns are usually present in only 1 % of wild populations and are used as a means of cultured product identification (Vaughan et al., 1988). The presence of notata markings can also be used as a marketing tool to help consumers distinguish cultured from wild clams. Also, the use of notata markings discourages poaching as large numbers of notata clams are an indication of cultured origins. The benefits of notata breeding may be temporary, though as escape and breeding by cultured clams increases the presence of notata markings in wild clam stocks. Figure 2-1 Hard clam showing notata pattern. Spawning is induced by thermal shock, a process of alternatively raising and lowering the water temperature (Hartman, 1989). Broodstock are placed on a spawning table containing 3-4 inches of clean seawater at 20 C and left undisturbed until all are open and actively siphoning water. The water temperature is then gradually raised to 30

PAGE 17

11 C, left for 30 minutes and then lowered to 20 C for an additional 30 minutes. This process is repeated until all clams spawn. Gonadal material from sacrificed adult clams may be added to the water to further induce reluctant spawners. Egg production ranges from 2-30 million eggs per female (Hartman, 1989). The fertilized eggs are placed in cone-shaped fiberglass or plastic containers of clean seawater. Within 24 hours of fertilization, the clam larvae, also known as veligers, develop shells and swim freely (Vaughan et al., 1988). Clam larvae do not actively feed for the first 48 hours after fertilization (Castagna and Manzi, 1989) and are not usually disturbed during this period. A popular device for rearing clam larvae is the downweller (Castagna and Manzi, 1989). A downweller is a plastic or fiberglass cylinder with an open top and a sieve covered bottom as pictured in Figure 2-2. Several downwellers are placed in a large fiberglass reservoir filled with clean seawater. The top of the downweller extends above the reservoir waterline. Water flows into the top of each downweller through individual pipes and flows out through the bottom sieves. This system allows the free-swimming larvae to remain in the water column in contact with higher quality food and away from smothering sediment and sick or dead larvae. The larvae are sieved every two days. Sieving allows for the removal of dead larvae and contaminants and permits counting and size sorting. Predation and fouling are problems that plague hard clam aquaculture from the earliest stages through harvest. Predation is the consumption of clams by other animals and fouling is the build-up of living organisms on the culture equipment, resulting in

PAGE 18

12 inflow reservoir c u I tu,e --i--cy I i nde, -----1 r ----------sieve t t t outflow I I I Figure 2-2. Downweller. smothering, impeded water flow and reduced food access. As mentioned previously, hatchery seawater is filtered, clarified or sterilized to remove predatory zooplankters. Fouling is controlled through reducing water contaminant content, frequent water changes and regular equipment cleaning. Between 8 and 14 days after fertilization, the clam larvae develop a muscular foot and reach the final larval or pediveliger stage (Vaughan et al., 1988). The larvae lose the ability to swim, but remain mobile through the use of their foot. The clam larvae are called set or post-set and enter the nursery phase.

PAGE 19

13 Nursery Phase The goal of the nursery phase is the production of sufficient quantities of seed for final growout. Most growout methods require 7-10 millimeter (mm) or larger seed (measured along the longest axis) (Manzi and Castagna, 1989). Growth rates vary with clam genetics, culture method and growing conditions. Post-set can be expected to reach 1-3 mm at three months, 3-9 mm at six months and 10-15 mm at nine months (Vaughan et al., 1988). Both onshore and field nursery systems exist. Onshore systems provide the greatest amount of access, control and predator protection, but may do so at considerable land and facility costs. Ambient temperature seawater is used for onshore systems. Water temperature adjustments may be made, however, to maintain optimal temperatures of 20-28 C (Vaughan et al., 1989). Food requirements may be met solely by natural seawater or through a combination of natural sources and cultured supplements. Field nursery systems are generally cheaper, but allow for less access and control as the clams are exposed to ambient water temperatures, algal content, etc. Field nursery systems are preferred by Florida hard clam culturists (Vaughan and Cresswell, 1989). A substant i al degree of control is required in the nursery phase, especially to reduce predation. The costs of seed production and the subsequent value of clam seed make high survival and rapid growth economic imperatives. Land-based nursery systems use methods similar to those used by hatcheries to control predation. Field systems require a different approach and typically rely on some type of physical barrier between young clams and predators. Dominant hard clam predators in Florida include rays

PAGE 20

14 (Dasyatis spp. and Gymnura micruva), sheepshead (Archosargus probatocephalus), blue crabs (Callinectes spp.), and various mollusc species (Melongena corona, Fasciolaria spp., Euleura cadata and Thais haemostoma) (Vaughan and Cresswell, 1989). Unprotected hard clam plots in a Florida and Georgia study suffered 100% mortality, of which, 90% was attributed to blue crabs (Eversole, 1987). Vulnerability to predation is inversely proportional to age, as young clams lack the size or shell thickness to prevent crushing, opening or boring by predators. Fouling also requires special attention. Major fouling organisms in Florida are sponges (Cliona spp., Haliclonia spp., and Halochondria spp.), sea squirts (Molgula occidentalis and Styela plicata), hydroids (Obelia spp.), barnacles (Balanus spp.), algae (Gracilaria spp.), and various mollusc species (Crepidula fornicata, Crassotrea virginica, C. rhizophorae, Modiolus spp. and Branchiodontes spp.) (Vaughan and Cresswell, 1989). Control methods vary with the culture method used and include various combinations of scrubbing, sun drying and turning the equipment over to smother the fouling organisms. In many operations, the nursery and hatchery phases overlap as hatcheries retain the post-set in their larval rearing containers. This reduces stress and allows greater control over growing conditions. Both downwellers and upwellers are used. Upwellers differ from downwellers in the direction of water flow, with water flowing from the reservoir to the rearing cylinder rather than from the cylinder to the reservoir, as in downwellers. See Figure 2-3. Upwellers vary according to whether seawater is pushed (active flow) or pulled (passive flow) through the clams. Water exits each cylinder through a top drain. Upwellers are more common than downwellers in nursery culture.

PAGE 21

15 Active upwellers are recommended for clams less than 3 mm (Manzi and Castagna, 1989). With proper flow rates, the post-set are suspended just above the sieve by the force of the flow and exposure to algae is maximized. Post-set must be evenly distributed ove r the sieve for equal food access. Raceways are a traditional land-based nursery method (Manzi and Castagna, 1989). Raceways are long tanks or troughs of epoxy-coated wood, fiberglass or concrete. Seawater is pumped into one end of the raceway and exits from the other end. Both shallow and deep raceway systems exist. A shallow system consists of a single layer of clams with just enough water to cover the clams. Deep systems use racks or tiers of trays to create multiple clam layers. A continuous flow of seawater is required. Wate r quality decreases as the distance from the point of inflow increases. Water algal content is highest at the point of inflow and lowest at the point of outflow. Elevated sediment and waste levels at the end of the raceway (point of outflow) may impede feeding and respiration. Hadley and Manzi (1984) showed a correlation of clam growth with the distance from the inflow, with highest growth occurring in clams nearest the inflow and lowest growth occurring the farthest from the inflow. Flow problems may restrict raceway systems to small post-set culturing. Post-set under 3 mm are best raised in a land nursery system due to potential predation and smothering problems. Larger post-set perform well in field systems (Castagna, 1983; Castagna, 1986; Vaughan and Cresswell, 1989; Vaughan et al., 1989). Field nursery systems depend on the ability of natural water systems to support post-set growth. This eliminates the need to pump seawater or provide cultured algae. Vertical

PAGE 22

inflow sieve--+-16 culture reservoir cylinder overflow l -----------------------------1 ... Passi ve Flow Upwel ler outflow post-set :mass outflow t t t sieve post-set mass forced inf low Acti v e Fl ow Upwel ler Figure 2-3. Upwellers.

PAGE 23

17 field systems are multi-tiered structures that place post-set in the water column near higher concentrations of phytoplankton and away from silt and benthic predators (Manzi and Castagna, 1989). Excessive fouling can be problematic. System examples are rafts, cages and racks of suspended trays or nets (Vaughan et al., 1989; Manzi and Castagna, 1989). A rack support structure is shown in Figure 2-4. Vertical systems are the most space efficient methods for culturing large post-set. The use of vertical systems may be restricted, though, as they may be a navigational hazard. Dimensions : width : 32" length : 66" height : 18" ground clearance : 6" Figure 2-4. Rack system. _____ Rebar Frame lfesh Bag ~ Ground Support Horizontal field systems rely on culturing post-set on the water bottom. Greater attention must be given to siltation and predation. Fouling can similarly be a problem.

PAGE 24

18 Examples of horizontal systems are trays, bottom bags and the flexible belt system (Vaughan et al., 1989). A bottom bag and flexible belt are shown in Figure 2-5. Tray systems use shallow plastic, fiberglass or wooden trays filled with 2 inches of sand or gravel. A mesh covers the tray to exclude predators. Excessive fouling is scraped off the tray and mesh. Bottom bags are mesh bags held in place by metal stakes. The mesh weave varies with clam size. The bags may be sewn shut or have one side closed with PVC pipe to allow easier access to the clams. A flotation device may be placed in the bag to aid sedimentation, after which the float is removed. Fouling is controlled by turning the bags over. The flexible belt system consists of a pair of parallel plastic ropes holding individual plastic mesh bags in a pod or modular arrangement attached with PVC pipe closures (Vaughan et al., 1989). Post-set are placed in small mesh bags which are then placed into the individual plastic mesh units. Each bag unit is removable for maintenance or harvest. The entire belt is anchored to the substrate and fouling is reduced by turning the belt over, as with bottom bags. Growth trials in Florida (Vaughan and Cresswell, 1989) using 3-9 mm post-set cultured at 1,800 clams per square foot showed trays to be superior to bottom bags and cages, with bottom-bag and flexible-bag culture growth intermediate to that of bottom nets and cages. Choice of tray substrate, sand versus gravel, showed no effect on growth, but resulted in survival rates of 95 % for sand and 45 % for gravel.

PAGE 25

19 Float Rebar Stake tlesh Bag PVC Closure BOTTOl1 BAG PVC Pipe / tlesh Bag Rope FLEXIBLE BELT Figure 2-5. Bottom bag and flexible belt.

PAGE 26

20 The length of the nursery phase varies with stock genetics, culture method, environmental conditions, food abundance and desired seed size. Growout Phase The growout phase takes seed clams and raises them to market size. Increasing water and food demands makes land-based clam growout systems economically impractical. Instead culturists rely upon natural water systems to meet clam requirements. During growout, culturists are primarily concerned with providing protection from predation while not impeding water flow or food access. Containers, meshes and densities are selected such that adjustments are not required during the growout phase. The predominant growout methods are tray, bottom net and soft tray or bag culture. Growout trays are similar to those used in field nursery systems except tray dimensions and mesh sizes change to reflect the larger clam size. Sand is the preferred substrate and clams are generally planted at 50-100 clams per square foot (Vaughan et al. 1988) The mesh is kept clean by periodic scrapping to remove fouling organisms. Bottom nets are the least expensive growout method in both material costs and maintenance (Vaughan et al., 1988). Clams are broadcast in plots over the water bottom and then covered with mesh nets The nets may be held by an iron frame or staked. Net or plot dimensions vary according to preference, location, management ability, etc., but are usually 25-50 feet long and 8-12 feet wide. Occasionally, mesh is laid under the clams to reduce escape and facilitate harvesting. Bottom nets must be regularly checked

PAGE 27

21 for oversiltation, fouling and predation. Fouling must be physically removed or the nets periodically replaced with clean nets. Fouled nets are sun-dried to kill the fouling organisms. Soft-tray or soft-bag growout is similar to soft-bag nursery culture (Vaughan et al., 1988). Clams are placed into mesh bags which are then staked to the bottom. A larger mesh size and bag are used for growout than for nursery culture. Bags are usually four feet square or four feet by eight feet. Fouling is again controlled by flipping the bags over. The flexible belt system can be used for hard clam growout. Mesh size is larger than for nursery use, and a bag insert is usually not required. The belt is serviced and maintained in the same manner as in nursery use. Equipment durability is an important consideration in the choice of a particular growout method as growout may take from 3-4 years in cold northern waters, and 1.5 to 2.5 years in warmer waters (Eversole, 1987). Bags, trays and nets must be chosen such that they are capable of extended use and not require frequent repair or replacement. Harvest Florida hard clams can be legally harvested for consumptive sale when they measure 7 /8 inches in width for sale outside the state and one inch for sale inside the state (Adams et al. 1991). Measurements are made across the hinge as indicated in Figure 2-6. This equates to a two-inch or 50-mm-long clam.

PAGE 28

/ / / / / / / / Height / -----Length Figure 2 6. Hard clam measurement axes Width I I I I I 22 The harvesting method used is dictated by the growout method practiced. Tray and bag culture allow for total harvest of the containment device. The tray or bag is manually or mechanically lifted from the bottom and legal clams removed. Bottom-plant methods require a different approach as the clams are not in any container. Some form of rake tong or mechanical harvest is required. State law may restrict the use of specific harvest methods thereby determining the choice of growout method. Mechanical harvest requires a special permit in Florida

PAGE 29

CHAPTER 3 BIOECONOMIC MODEL Economic Model The objective of a hard clam aquaculture operation is assumed to be profit maximization. Management options for hard clam growout include the selection of seed size, choice of planting method and clam density, and replacement scheduling. The task confronting the aquaculturist is to select a seed size, planting method, clam density, and planting and harvest schedule that maximizes net revenue. The first three decision options are typically single incident decision choices--an option is selected at planting time and remains fixed throughout the production process. Clam density may be altered through periodic culling. Replacement scheduling, however, is a decision requiring periodic evaluation. At each stage of the growout process, a dichotomous decision choice is faced: to sell the existing stock and replant with new seed, or to keep the existing stock for another period. Economic principles dictate that the decision to replace or keep existing stock be based on a comparison of the gains from keeping existing stock an additional period with the opportunity gains from replacement stock during the same period (Perrin, 1972). If the gains associated with retaining existing stock exceed the average net returns from 23

PAGE 30

24 harvesting and replacing with new clam stock, then current stock should be retained. Otherwise, immediate harvest and replacement is warranted. A hard clam is an appreciating asset (to a point) that generates a single, point-of harvest return. Hard clams provide no stream of revenues prior to harvest. Future hard clam growth and prices cannot be forecast with total accuracy due to the inherent uncertainty of the various processes being examined. Growth and death processes are at best imperfectly describable, and their dependence on stochastic environmental conditions increases the uncertainty of achieving specific future outcomes or states. Price movements are likewise uncertain. The result is that future states can be predicted only in a probabilis t ic manner. Hence, rather than having a single net revenue state, Fs, in each period s, there exist k net revenue states in each period s, Fk,s Also, each net revenue state k in period s is realized with the probability /3k s Given the above considerations, the discrete-time net present value of hard clams replaced every s periods is R(s,oo) = l [ ( 1 +rfs {3 F M] 1 ( 1 ) -s L k,s k s +r kEk (3.1) where R(s, oo) = net present value of an infinite stream of revenues from an asset replaced every s periods; r = discount rate; Fk s = net revenue from asset in state k in period s; Bk s = the probability of having state k in period s;

PAGE 31

25 M = asset replacement cost; K. = the set of all possible states. The term in brackets is the net present value of a single asset cycle and the term outside the brackets converts this to an infinite chain. Equation 3 1 computes the net present value of an infinite annuity received every s periods. Replacement literature uses the term "defender" to refer to the asset already in use and the term "challenger" to refer to the replacement asset (Perrin, 1972) If a single challenger (C) exists (3 1) is maximized with respect to replacement age s and the maximum present value R calculated. Should multiple challengers exist (differing by productive capability) (3 1) must be maximized for each challenger. The best challenger is that which generates the highest R* and is denoted by c. At each production stage (period), the culturist has the choice of replacing the defender with c or allowing the asset to grow an additional period. Crane (1979) shows that the replacement decision is based on a comparison of the infinite net revenue streams of each alternative or 1 replace if R > 1-(1 +r t 1 keep if R. < l 1-(1 +rr 1 indifferent otherwise where [~ (l+rr 1 0 I D 11 L.., J J j Ej [1: (1 +rr 1 oj I D j l] jEj' (3.2) oi 1 = the probability of having net revenue j if the defender's life is extended 1 period; D i I = net revenue j from the defender if it's life is extended 1 period. Equation 3.2 constitutes the replacement decision policy II for the hard clam

PAGE 32

26 aquaculturist. The decision space defines the set of all possible options available to the producer. Assuming that the operation will continue production in some manner, the decision space A has two elements: to keep, and to harvest and replace. Management of hard clam aquaculture consists of a series of decisions to maximize net revenue. Dynamic programming is a useful optimization procedure for problems involving a sequence of interrelated decisions (Dreyfus and Law, 1977). Central to dynamic programming are the concepts of stage and state. A stage is the periodic division unit, often time, at which the system is evaluated and a policy decision required. State variables are observable or measurable conditions such as clam size, mortality, prices, etc. These provide the basis for periodic (stage) evaluation. The action chosen at each stage allows the system to change from state to state according to the processes driving changes in the state variables. A harvest decision prompts restocking and next period's stock is then replacement seed. A decision not to harvest allows additional stock growth, mortality, and market risk. In stochastic systems, transition from state to state follows probabilistic patterns. The description and solution of problems involving stochastic systems is simplified if the underlying probability process satisfies the Markovian assumption (Howard, 1981). The Markovian assumption is that the conditional probability of any future state, given decision ex and past and present states, is dependent only on ex. and the present state of the process. Precisely, (3.3) where

PAGE 33

Xt = state of the system at stage t; At = decision at stage t ; 27 One-step transition probabilities are usually denoted by Pii indicating the probability of transition from state i, i = 1 2, ... N to state j j = 1 2 ... N, in one period (Hillier and Lieberman 1986) Transition probabilities are stationary if they do not change over time. This implies that P{X 1 +n = j I X 1 = i} = P{X 0 = j I X 0 = i} for each i, j, and n = 0 1, 2 .. and all t = 0, 1 2 These n-step transition probabilities can be denoted by p < 0 \ and satisfy the properties P t 0 for all i and j and n = 0, 1, 2 ... M L Pt > = I for all i and n = 0, 1, 2, ... j= O (3.4) The n-step transition probabilities p< 0 \ represent the probability that a process in state i will be in state j after n periods. They can be presented in matrix form p ~) ( n ) P o i ( n ) P o N ( n ) (n ) ( n ) p ( n ) P10 P11 P1N (3.5) = ( n ) ( n ) p: !>N o PNI The Chapman-Kolmogorov equations give a method of solving the n-step transition probabilities (Hillier and Lieberman 1986). In the transition from state i to state j, the process will be in some state k after exactly v steps, where v is less than n. This is shown by

PAGE 34

28 M (n) "t'"' (v) (n-v) & all d O < < Pii = L..., Pik J>ki 1or 1, J, n, an v n. (3.6) k:O Equation 3. 6 states that the process goes to state k after v steps and then to state j in n v steps and summing over all possible k must yield p< 0 \. For the special cases v = 1 and v = n 1, (3.6) becomes M Pt)= L (3.7) k:O and (3.8) for all i, j, n. Equations 3.7 and 3.8 show that then-step transition probabilities can be obtained recursively from the one-step transition probabilities. If n = 2, (3. 7) and (3.8) become M Pt = L pik pkj' for all i, j. (3.9) k:O The pC2\ are the elements of the matrix p(2) and they are obtained by multiplying the matrix of one-step transition probabilities by itself, or p(2) = p p = p2_ (3.10) It then generally follows that the matrix of n-step transition probabilities can be obtained from p (n) = p p .. p = p n = ppn 1 = pn lp_ (3.11)

PAGE 35

29 The decision policy I1 defined by (3.2) is a stationary policy as the action it specifies at time t depends only on the state of the system at time t and is nonrandom (Dreyfus and Law, 1977). The maximum expected total discounted net revenue of a hard clam growout system starting in state i, following decision policy IT, and evolving for t time periods is given by m Y/ = max {Gia + PL P/a) Yr'} i=0,1,2, ... ,n a j=O where Gia = expected reward in state i at time t given action a; p = discount factor; (3.12) Pia) = probability of transition from state i at time t to state j at time t+ 1 when action a is taken. As t approaches infinity, Y\ converges to Yi, where Yi is the expected total discounted revenue of the system starting in state i and continuing indefinitely when an optimal policy is followed (Dreyfus and Law, 1977). Equation 3.12 then becomes m Yi = max {Gi a + p L P/a) Y); i=0,1,2, ... ,n. (3.13) a j = O Dreyfus and Law (1977) prove that the solution to (3.13) is optimal and that the stationary policy I1 driving (3.13) is itself optimal. The optimum hard clam replacement schedule is therefore obtained when (3.13) is satisfied.

PAGE 36

30 Equation 3.13 is solved by the application of successive approximations to (3 12). This procedure is described by Dreyfus and Law (1977) The first step involves assigning a terminal value to Y\ where t = N the terminal stage. This terminal value is typically zero. This allows for the solution of the one-period model, yN \, to be solved from yN \ = max a { Gi a }, i = 0 1, 2 This represents the expected total discounted revenue starting in state i in stage N 1 and evolving for one period. This represents the first approximation to the optimal policy N t Y N -2 1 ed th yN 1 yN-1 yN 1 f yN 1 th d ex, i 1s so v usmg e 0 1 m rom i g1vmg e secon approximation to the optimal pol i cy This represents the expected total discounted revenue starting in state i and evolving for two periods. This process is repeated using the recursive relationships until the optimal policy is ev i dent. This occurs when the optimal policies of successive iterations are identical. A key consideration in this method is that it does not specify how large N should be (Dreyfus and Law, 1977). It may be necessary to apply the method of successive approximations to (3.12) using different values of N. Hard Clam Growth and Mortality Models As previously indicated hard clam revenue is determined by clam size, price and quantity of clams Knowledge of growth and surv i val is therefore necessary to predict expected revenues Growth models simulate the movement of clams across size categories while mortality models determine survival and hence clam quantities available for sale.

PAGE 37

31 Bivalve Growth Literature Hard clams have seasonal growth (Eversole, 1987). Seasonal environmental variations include changes in air and water temperatures, salinity, dissolved oxygen and other parameters. The moderate temperatures of the spring and fall produce the fastest growth (Eversole, 1987). High summer temperatures result in the slowest growth and severe winter temperatures similarly depress growth (Eversole, 1987). Changes in the environmental parameters affect hard clams directly by altering metabolic, feeding and respiration rates (Eversole, 1987; Van Heiningen, 1992; Malouf and Bricelj, 1989) and indirectly through affecting phytoplankton availability. Seasonal growth imposes a degree of production risk on hard clam growout as clam growth becomes dependent upon uncertain environmental conditions. Knowledge of the impacts of specific environmental parameters would allow the incorporation of production risk into the bioeconomic model through the estimation of growth given probabilistic environmental conditions. Despite the recognition of environmental impacts on clam growth, little empirical work exists on quantifying these impacts. Some authors acknowledge the importance of environmental influences in determining clam growth, but make no attempt to quantify these effects. Askew (1978) and Loesch and Haven (1973) model growth simply as a function of initial size. Lough (1975) uses the function Y = b 0 + b 1 (T) + bi(S) + blT 2 ) + biS 2 ) + b/T*S) to specify the effects of temperature (T) and salinity (S) on the percentage growth of hard clams and two other bivalve species. Parameter estimates were made for only two and ten-day-old clam larvae. Results indicated a cessation of

PAGE 38

32 growth at temperatures and salinity above 32.5C and 27 ppt respectively. Metabolic changes in older clams raise questions on the applicability of these results to larger clams. As clams mature, optimal temperature and salinity ranges change, and tolerance levels increase as older clams are able to remain closed under adverse conditions (Eversole, 1987). The increased tolerance to adverse conditions may allow clams to continue to grow during periods of environmental stress. Growth would, however, likely be less than under favorable conditions. In a paper describing Virginia private oyster culture, Bosch and Shabman (1989) model growth as a function of initial weight, season and salinity. Their model has the form Wt = W 0 e\\\, where Wt and W 0 are final and initial oyster weights, e is the base of the natural logarithm and tlj, bj and ck are the seasonal, salinity and instantaneous growth rate (a function of initial weight) effects, respectively. The paper does not attempt to validate the growth function, focusing instead on using it as a tool in simulation modelling of oyster production. Results, though, indicate that improved knowledge of salinity effects holds great promise for increasing oyster culture profitability. Parameter estimates in this model were independently determined from different stud i es and combined for application to Virginia oyster culture. The linear form of the model is ln(W/W 0) = a;bfk The model is unsuitable for estimation of the individual effects of the various environmental parameters. At best, a single coefficient could be estimated, representing the combined effects of season salinity and initial weight.

PAGE 39

33 Food (phytoplankton) is discussed by several authors as a major factor in determining clam growth (Epifanio, 1979; Menzel, 1989; Malouf and Bricelj, 1989). Most researchers focused on the growth effects of specific diets, showing that certain algal diets were more beneficial than others. Malouf and Bricelj (1989) discussed the impact of clearance rates--the volume of water filtered completely free of food particles per unit time--on clam growth. Growth was shown to be a function of clearance rate which, in tum, was affected by water temperature, food concentration and food quality. Phytoplankton quality and abundance is also affected by various environmental conditions such as atmospheric temperature, rain, or excessive run-off (Ryther, 1986). These factors influence algal growth and, hence, its availability as food. Growth Model Clam growth can be described as G = f(I, A, S, T, D, N, C, M) where G = clam growth or final size; I = initial clam size; A = clam age; S = water salinity; T = water temperature; D = water dissolved oxygen; N = food profile or algal content of the water; (3.14)

PAGE 40

C = water flow or current characteristics; M = other factors. 34 The selection of specific independent variables included in the growth model regression was based on considerations of data availability. Specifically, data on food availability and water flow characteristics were unavailable, thus these factors could not be included in the regression analysis. Techniques exist for measuring these parameters--water speed and direction can be measured, chlorophyll levels are an indication of algal abundance, light refraction meters can measure turbidity or sediment load--but monitoring of such parameters i s currently not undertaken with any regularity by either aquaculturists or water management officials. Hard clams are measured by shell size (height or length) and not by total weight or meat mass. Although clam weight or meat mass may decrease under adverse growing conditions, shell size only decreases as a result of shell blunting in extremely old and large clams. Thus, clam growth is nonnegative and an acceptable hard clam growth function is required to mimic this condition. This was accomplished through the use of a log-linear growth function. The regression model for clam growth in period i was (3.15) where lnY; = natura l log of the ratio of a clam's final size at the end of period i over its initial size at the beginning of period i; B 0 = a constant;

PAGE 41

35 lnAi = natural log of the clam's age at the beginning of period i; lnSi = natural log of the mean water salinity over period i; lnT; = natural log of the mean water temperature over period i; lnDi = natural log of the mean water dissolved oxygen over period i; Ej = residual for period i. Non-decreasing shell size results in the ratio "final size/initial size" to be no less than 1, the natural log of which is 0. Thus, the dependent variable, Y;, is 0 or positive and preserves the nonnegative growth requirement. The model is also similar in form to the linear form of the Bosch-Shabman model but allows for estimation of the individual effects of different environmental parameters. Production data were obtained from a commercial hard clam operation located in the Indian River Lagoon near Melbourne, Florida, and used to estimate the hard clam growth function coefficients. The data covered production from June 1990 to October 1992 and contained 729 observations of monthly growth and mortality averages. Size measurements were made along the longest axis (length) of the clams. There were an average of 25 observations per month. Individual plantings ranged from 360,000 to 5 million clams. Clams ranged in age from I to 37 months old (post-plant age). All seed were 10 mm. New plantings occurred monthly and in any given month 23-30 different plants existed simultaneously on the lease. Planting densities prior to January 1991 were 80 clams per square foot and 60 clams per square foot during and after January 1991. The production data also included environmental data from June 1990 to October 1992 and consisted of daily observations of water temperature, salinity and dissolved

PAGE 42

36 oxygen. Monthly averages were computed from the daily figures. The production data set is given in Appendix 1. The coefficients of the model were first estimated by ordinary least squares (OLS) regression and the results are shown in Table 3-1. Separate regressions were run for 60 and 80 clams per square foot. Plots of the standard errors of (3 .15) against individual independent variables indicated heteroskedastic disturbances. Heteroskedasticity produces consistent but inefficient parameter estimates in the general linear model. Further, since clam growth is restricted to be nonnegative, the dependent variable in (3.15) is a limited dependent variable. Maddala and Nelson (1975) show that ignoring heteroskedasticity in limited dependent-variable models produces inconsistent parameter estimates. Equation (3.15) was tested for heteroskedasticity using the SPEC procedure in SAS (SAS, 1985). The specifics of the SPEC procedure are outlined in White (1980). The results of the SPEC procedure rejected the null hypothesis of no heteroskedasticity. To correct for heteroskedasticity, the model was estimated using weighted least squares (WLS) procedures. Additionally, since Y is a limited dependent variable, heteroskedastic tobit estimation was conducted. Heteroskedastic tobit procedures are described in Maddala (1983). The results of these efforts are also shown in Table 3.1. Efficiency criteria, however, may be of little concern when the ultimate intention is growth simulation. Forecasting quality, or the ability to produce clam growth in a realistic time framework becomes more crucial. New clam sizes at the end of each growout period (month) were calculated from

PAGE 43

Table 3-1. Hard clam growth model parameter estimates from OLS, WLS and heteroskedastic tobit regressions. Standard errors are in parentheses. Estimation Standard Procedure Density Error Bo 61 62 63 64 l8 60 0.0664 0.6577 -0.0507 0.1736 -0.2742 -0.1019 (0.1755) (0.0068) (0.0409) (0.0458) (0.0245) 80 0.0415 0.2316 -0.0550 0.0629 -0.0718 -0.0085 (0.0699) (0.0029) (0.0161) (0.0154) (0.0079) 2b 60 0.0685 0.9751 -0.0681 0.1267 -0.3063 -0.1324 (0.0963) (0.0061) (0.0319) (0.0422) (0.0190) 80 0.0455 -0.0154 -0.0377 0.1018 -0.0492 -0.0083 (0.0610) (0.0037) (0.0107) (0.0127) (0.0044) 3 c 60 0.3489 1.0137 -0.0770 0.3007 -0.4489 -0.2161 (0.1546) (0.0070) (0.0490) (0.0682) (0.0300) 80 0.3344 0.0422 -0.0724 0.2417 -0.1820 -0.0210 (0.1432) (0.0053) (0.0349) (0.0324) (0.0156) a OLS; b WLS; c heteroskedastic tobit.

PAGE 44

fi fl. AfJ, SfJ TfJ DfJ, msi msi e i ; ; ; where finsi = clam size at the end of period i; insi = clam size at the beginning of period i; e = base of the natural log. 38 (3.16) Equation (3.16) was used to determine whether a particular estimation technique produced parameter estimates that resulted in acceptable clam growth. Acceptable growth was defined as attaining marketable size (50 mm long) within 2 2.5 years, a growout period common to Indian River producers. Despite estimation using data truncated at 0, growth simulation will not guarantee nonnegative clam growth. It is therefore necessary to impose a no-shrinkage restriction. If growth simulation results in a clam shrinking, then the resultant final size is reset at equal to the clam's initial size. The results of test simulations for 80 clams per square foot are shown in Table 3-2. Only OLS procedures produced parameter estimates that adequately reflected observed data on growth performance. Thirty-month final clam sizes were 50.567 mm at 80 clams per square foot and 56.114 mm at 60 clams per square foot (results not shown) using OLS parameter estimates. Comparative sizes for WLS and heteroskedastic Tobit were 36.602 mm and 30.721 mm, respectively, for clams planted at 80 clams per square foot. The OLS estimation produced the best model and was therefore used to simulate hard clam growth in the remaining portions of this research.

PAGE 45

39 Table 3-2. Hard clam final size (mm) simulation results using ordinary least squares (OLS), weighted least squares (WLS) and heteroskedastic tobit (Tobit) estimation techniques for 10 mm seed planted in January at 80 clams per square foot. Method of Estimation Month OLS WLS Tobit 1 12.004 11.493 12.065 2 13.804 12.696 13.492 3 15.757 13.995 15.280 4 17.887 15.383 17.434 5 20.065 16.798 19.620 6 22.249 18.153 21.638 7 24.587 19.703 24.125 8 26.635 21.076 25.763 9 28.597 22.571 27.374 10 30.276 23.782 27.997 11 31.641 24.678 27.997 12 32.455 24.978 27.997 13 33.535 25.646 27.997 14 34.367 25.882 27.997 15 35.962 26.730 28.123 16 37.661 27.661 29.479 17 39.708 28.713 29.399 18 41.490 29.707 29.882 19 42 905 30.441 29.882 20 44.342 31.428 29.983 21 45.891 32.740 30.599 22 46.967 33.986 30.634 23 46.967 33.986 30.634 24 47.060 34.196 30.634 25 47.060 34.196 30.634 26 47.252 34.196 30.634 27 47.873 34.542 30.634 28 48.615 34.999 30.634 29 49.773 35.605 30.634 30 50.567 36.136 30.634

PAGE 46

40 A comparison of the performance of the 60-clam and 80-clam models, with plants in each of the 12 calendar months, is shown in Table 3-3. The 60-clam-density model produced larger but more variable clam growth. The difference in the performance of the two models is likely a demonstration of crowding effects. Higher densities reduce mobility, access to food, and access to fresh water, thereby negatively affecting growth. The net effect of crowding, however, might not be uniform across all months due to monthly variations in water quality. The simulation results provide evidence of the positive/negative effects specific months have on the growth performance of clams. Specifically, a growth bias toward fall and winter conditions and away from spring and summer conditions is indicated. Timing is apparently at issue, with clams unable to overcome the negative effects of planting in less favorable months. Estimation of Mortality Little empirical work exists on developing mortality models for estimating bivalve mortality rates. In situations where populations are modelled from the postlarval stage without consideration of reproduction, the numbers of surviving individuals is due to mortality only. Mortality of seed clams is often many times that of adults (Eversole, 1987). In the absence of adverse environmental conditions, predation is the major cause of hard clam mortality (Eversole, 1987). Susceptibility to predation decreases with increasing clam size and commercial growout methods utilize effective physical barriers to reduce predation. Exposure to environmental conditions outside tolerance ranges also produces increased mortality. As discussed previously with reference to growth conditions, tolerance ranges as they relate to mortality increase for older clams, however,

PAGE 47

41 Table 3-3. Average expected 30-month sizes (mm) from simulation model testing using 10 mm seed. Planting Density Month 60 clams/sq. ft. 80 clams/ sq. ft. January 65.92 56.16 February 59.09 54.19 March 55.26 52.91 April 50.28 51.19 May 47.33 50.25 June 48.45 50.41 July 51.16 51.13 August 56.87 53.21 September 63.76 56.74 October 70.88 58.36 November 73.15 58.53 December 70.78 57.62 as they are able to maintain closure longer, effectively shutting out certain adverse conditions, and are able to utilize various metabolic mechanisms undeveloped in juvenile clams (Eversole, 1987). Most models describing populations with an absence of reproduction are usually expressed in terms of an instantaneous mortality rate (Allen et al., 1984) which, despite the name, computes an annual or monthly mortality. Askew (1978) links oyster mortality to size and computes monthly mortality from annual data, using the assumption that short term rates concur with long term rates. Data showed mortality peaks in winter and early summer. Askew acknowledged the potential impact of season on mortality, but concluded insufficient evidence existed to warrant inclusion in his analysis. The

PAGE 48

42 observed association of mortality with size implied an inverse relationship as mortality rates decreased with increasing size. Evaluation of the mortality data used in this research did not produce a workable model of environmentally-dependent mortality. Hard clam mortality was thus calculated in a deterministic manner based on age. First, a cumulative 30-month mortality was identified from discussions with hard clam research specialists (David Vaughan, Harbor Branch Oceanographic Institute Ft. Pierce Florida and Leslie Sturmer, Project Ocean, Cedar Key, Florida, personal communications). Next, a weighting system for mortality was determined from examination of available mortality data linking mortality with age The resulting weight system placed greater weights on younger clams indicating greater early mortality and lower mortality of older clams The weight system was then used to generate periodic (monthly) conversion factors that produced the given cumulative 30month mortality. The monthly conversion factors are given in Table 3-4. The factors sum to one over 30 months and actual periodic monthly mortality is computed by multiplying the cumulative 30 month mortality by the respective conversion factor. Table 3-4 Periodic (monthly growout age) mortality conversion factors. Period Conversion factor 1 0 10 2 0 08 3 0.06 4-11 0.04 12-17 0.03 18-34 0.02

PAGE 49

CHAPTER 4 RESULTS AND DISCUSSION Growth Simulation Hard clam growth was simulated using equation (3.14) and environmental data collected from the St. John's River Water Management District (SJRWMD). The SJRWMD data consisted of daily water temperature, salinity and dissolved oxygen readings from different locations in the Indian River Lagoon. To test the robustness of the growth model, three sites were selected from which monthly averages were computed. The selection of the three sites was based on comparison of site conditions, with the selected sites representing mean above mean, and below mean conditions. Mean monthly values for each site are listed in Appendix 2. Each environmental parameter was assumed to have a normal distribution. Using parameter means and standard deviations, monthly averages were assumed to take on three possible values: the mean, greater than mean and less than mean. The extreme values were calculated by adding or subtracting two standard deviations to or from the mean, respectively. Exposure to a mean value had a 68 % probability, while exposure to each of the extreme values had a probability of 16%. A given environment consisted of some combination of mean above mean and below mean values for the three parameters of interest. For 43

PAGE 50

44 example, environment Elm consisted of mean values for temperature, salinity and dissolved oxygen in month m, where m = 1, 2, ... 12. Environment E2m had mean temperature and salinity and greater than mean dissolved oxygen in month m, etc. Given three parameters, 3 3 or 27 possible environments existed each month. The environmental probabilities for the 27 environmental states are given in Table 4-1. The probability of exposure to a particular environment in any month was determined by the product of the probabilities of receiving each individual parameter. Environmental states were defined in mean, above mean and below mean terms. For example, the probability of having Elm, indicating mean temperature, salinity and dissolved oxygen in month m, was 0.68 3 or 31.44%. The probability of having environment E2m was 0.68 2 *0.16 or 7.4%. The probability of exposure to a given environmental state was the same regardless of which stage or month transition occurred from. The probability of receiving mean values for all three parameters in the current or future growing periods was 31.44 % regardless of which month the production process was in. Actual parameter values for a given environmental state, however, and their impact on clam growth, varied from month to month. Mean water temperature for environmental state El I was not the same as mean water temperature for El 2 etc. January temperatures are typically different than February temperatures The probability of exposure to a mean, above mean or below mean value in each month, however, remains the same. Growth simulation began with a clam or bag of clams of a specific initial size, either 10 mm or 15 mm, and an age of 1, indicating the first growout month. Simulation

PAGE 51

45 Table 4-1. Environmental 12robabilities used in hard clam growth simulation. Dissolved Environment Oxygen Salinity Tem~rature Probability El X X X 0.3144 E2 X X + 0.0739 E3 X X 0.0739 E4 X + X 0.0739 E5 X + + 0.0174 E6 X + 0.0174 E7 X X 0.0739 E8 X + 0.0174 E9 X 0.0174 EIO + X X 0.0739 Ell + X + 0.0174 E12 + X 0 0174 E13 + + X 0.0174 E14 + + + 0.0040 E15 + + 0.0040 E16 + X 0.0174 El7 + + 0.0040 E18 + 0.0040 E19 X X 0.0739 E20 X + 0.0174 E21 X 0.0174 E22 + X 0.0174 E23 + + 0.0040 E24 + 0 0040 E25 X 0.0174 E26 + 0.0040 E27 0.0040 Code: X = mean value, (-) = mean minus two standard deviations, ( +) = mean plus two standard deviations.

PAGE 52

46 began with a January plant. January growth simulation produced a new clam or unit with 27 possible profiles or sizes. Although the first step began with seed of equal size, 27 (potentially) alternative final sizes were possible, the result of 27 different environments. The second step required that each of the 27 alternative units produced by step 1 be exposed to each of the 27 February, or period 2, environments. The outcome of an exposure defined by El (mean values for all three environmental parameters) in period 2 was calculated as the average size of the 27 period 1 clams transitioning through El in period 2. Similarly, the outcome of exposure to E2 (mean dissolved oxygen and salinity and above mean temperature) was the average of those same 27 period 1 units exposed to E2 in period 2, etc. Thus, the outcome of each sequential growth period was 27 "new" clams, where each clam represented an average of 27 potentially different clams transitioning through a given environment. It is important to note that the 27 clams produced each period need not be of different sizes. The zero growth outcome of certain environments often produced identical representative clams. An example of the outcome of growth simulation is given in Table 4-2. Growth was simulated for 34 months. This produced a 27*34 matrix of ending clam sizes for each period. Each cell or entry in the matrix represented the average of 27 clams exposed to the environment designated by that particular row. For a given site, this procedure was repeated with each of the 12 calendar months as the initial growth period. This process was repeated for each initial seed size (10 and 15 mm) and planting density (60 and 80 clams per square foot). This produced 2*2*12 or 48 growth simulations per growout site.

PAGE 53

47 Table 4-2. Example of hard clam growth (ending size) simulation results for a January plant using 15 mm seed planted at 80 clams per square foot. Period Environment 1 2 13 14 15 33 34 Size (mm) 1 18.60 21.97 53.26 56.55 60.22 85.02 85.08 2 18.26 21.55 52.28 55.45 59.15 85.02 85.08 3 19.06 22.57 54.59 58.09 61.68 85.02 85.63 4 18.78 22.27 53.78 57.31 61.19 85.02 85.50 5 18.44 21.84 52.80 56.20 60.11 85.02 85.08 6 19.25 22.88 55.13 58.87 62.68 85.51 86.70 7 18.38 21.60 52.64 55.59 58.94 85.02 85.08 8 18.05 21.18 51.68 54.51 57.89 85.02 85.08 9 18.84 22.19 53.96 57.10 60.37 85.02 85.08 10 18.54 21.94 53.10 56.46 60.14 85.02 85.08 11 18.20 21.51 52.13 55.36 59.07 85.02 85.08 12 19.01 22.54 54.44 58.00 61.59 85.02 85.40 13 18.72 22.24 53.62 57.22 61.11 85.02 85.26 14 18.38 21.81 52.65 56.11 60.02 85.02 85.08 15 19.19 22.84 54.97 58.78 62.59 85.33 86.46 16 18.33 21.57 52.49 55.50 58.86 85.02 85.08 17 17.99 21.15 51.53 54.43 57.81 85.02 85.08 18 18.79 22.16 53.81 57.01 60.28 85.02 85.08 19 18.68 22.02 53.49 56.66 60.33 85.02 85.08 20 18.34 21.59 52.51 55.56 59.25 85.02 85.08 21 19.15 22.62 54.83 58.20 61.79 85.02 85.99 22 18.86 22.31 54.01 57.42 61.30 85.14 85.86 23 18.52 21.88 53.03 56.31 60.21 85.02 85.08 24 19.33 22.92 55.37 58.99 62.79 85.77 87.06 25 18.46 21.64 52.87 55.70 59.04 85.02 85.08 26 18.12 21.23 51.91 54.62 57.99 85.02 85.08 27 18.92 22.23 54.20 57.22 60.47 85.02 85.08

PAGE 54

48 Final monthly expected clam sizes were computed using the results of the growth simulation and the environmental probabilities. The expected size of a clam at the end of period i equaled the sum of the expected outcome of a clam exposed to environment j in period i times the probability of encountering environment j. Thus, each 27*34 matrix of final clam sizes, where each row represented a different environment, was reduced to a 1 *34 vector of expected final sizes. Each outcome represented an expected mean clam size. Size standard deviations were then calculated from the equation STD = 2.151 0.0603 FINS where STD = standard deviation of hard clam size; FINS = hard clam mean size. (4.1) Equation 4.1 was estimated using the commercial production data previously described. Mortality The periodic and cumulative mortalities used in the production simulation are shown in Table 4-3. Early mortality rates are higher than later ones, reflecting the higher natural mortality of younger and smaller clams. When 15-mm seed clams were used, the first two months of the mortality series were eliminated, reflecting the increased survival expectation of larger clams.

PAGE 55

49 Table 4-3. Monthly and cumulative mortality used in hard clam growth simulation. Mortality ( % ) Month Monthly Total 1 2 2 2 1.6 3.6 3 1.2 4.8 4 0.8 5.6 5 0.8 6.4 6 0.8 7.2 7 0.8 8 8 0.8 8.8 9 0.8 9.6 10 0.8 10.4 11 0.8 11.2 12 0.6 11.8 13 0.6 12.4 14 0.6 13 15 0.6 13.6 16 0.6 14.2 17 0.6 14.8 18 0.4 15.2 19 0.4 15.6 20 0.4 16 21 0.4 16.4 22 0.4 16.8 23 0.4 17.2 24 0.4 17.6 25 0.4 18 26 0.4 18.4 27 0.4 18.8 28 0.4 19.2 29 0.4 19.6 30 0.4 20 31 0.4 20.4 32 0.4 20.8 33 0.4 21.2 34 0.4 21.6

PAGE 56

50 Economic Evaluation The results of the hard clam growth simulation were used to determine optimal production design and expected returns. Two production scenarios were modelled. The first scenario treated an acre as a single unit. The entire acre was planted at one time and all clams were harvested in a single month. Hereafter, this scenario will be referred to as single plant production. The second scenario imposed a monthly marketing requirement. Producers might require or prefer a monthly harvest and revenue stream in order to meet labor restrictions or expenditure requirements. Similarly, market outlets might require monthly supplies from growers to satisfy monthly consumer demand. Therefore, some positive quantity of clams was required to be sold monthly. Planting restrictions were not imposed, so an entire acre could still be planted at one time as a single unit. The acre must be harvested, however, in monthly units or plots, so this method is termed multiplot production. Production assumptions are based on Adams et al. (1993). Some assumptions common to both scenarios were (a) The operation lease consisted of a two-acre submerged tract. The site is situated such that it is never exposed during low tide. Each acre is identical in terms of productive capacity. (b) A staggered production schedule is followed. Each acre is planted and harvested successively. The first acre is planted the first year and the second acre planted the next. Expected growout periods under a given production regime are identical, thus producing harvests in successive years. (c) 750 4-foot-square growout bags are planted per acre. (d) Clam seed is stocked at 1000 (62.5 clams per square foot) or 1200 (75 clams per square foot) clams per bag. These represent recommended average and maximum densities (Adams et al., 1993).

PAGE 57

51 (e) The current month is January and the operation has the option of beginning production (making the initial plant) in the current month or in any of the future 11 months. Information on hard clam prices was required to determine the expected returns of growing clam stock. Hard clam price data were obtained from the Florida Department of Natural Resources (FDNR) and a commercial fish house in the Indian River Lagoon area. The combined data set consisted of monthly average wholesale (dockside) prices from January 1986 to April 1993 for littleneck, topneck, cherrystone and chowder clams. Although the data covered a span of eight years, relatively few price observations were available as price reporting was voluntary. Price forecasts for subsequent portions of this research were based on average monthly prices for the four size categories. Under current Florida law, clams less than 50 mm long (one inch across the hinge) are not legal for sale in Florida. Licensed dealers may sell 45-mm (7/8 inch across the hinge) clams outside the state. Price data on the smaller clams was unavailable. The assumption is also made that the clam producer is not a licensed dealer and is therefore prohibited from selling 45-mm clams. All undersized clams were assigned a price of zero. Mean monthly prices for the various size categories are given in Appendix 3. Single Plant Production Optimal Rotations The results of the hard clam growth simulation were combined with seed quantities and mortality rates to estimate expected revenues. The prescribed densities translated to initial seed plants of 750,000 and 900,000 clams per acre. Surviving clam numbers were determined using initial plant quantities and the previously described mortality rates.

PAGE 58

52 Sixty-eight percent of surviving clams were assumed to possess mean size, 16% mean plus two standard deviations, and 16% mean minus two standard deviations. Hard clam prices are size dependent. Applying the standard deviation to the mean expected size generated a size spread within each age group, thus allowing surviving clams of a given age class to potentially fall into different size categories. Expected gross revenues were then calculated from total clam numbers within each size category times the expected price. Next, expected net discounted revenues were computed by subtracting periodic maintenance costs and using an annual real discount rate of 4 % These revenues were then examined to identify the growout period that produced maximum expected net discounted revenues. Each production method (density and seed size) and planting month (January through December) was evaluated independently. Hereafter, the term "production method" should be understood to refer to a specific seed density (clams per square foot) and seed size (mm). Thus, optimal growout periods were identified for 10 and 15 mm seed planted at 62.5 and 75 clams per square foot beginning in January, February, March, etc. Revenues over the course of the 34-month growout followed general patterns of first increasing and then decreasing in value. This was due to stocks first growing into and then out of more valuable size categories. Changes were not uniformly up or down, though, as mortality or price movements often offset the effects of growth gains. The revenues produced by these optimal growout periods represented the best potential revenues possible given expected growth, mortality and price movements. Optimal growout periods are listed in Tables 4-4, 4-5 and 4-6.

PAGE 59

53 Table 4-4. Site 1 optimal growout periods (months) for single plant production by production method and planting month. Planting Seed density/size Month 62.5/10 62.5/15 75/10 75/15 January 28 19 25 16 February 34 22 31 19 March 30 23 30 21 April 29 23 29 20 May 33 23 31 21 June 34 22 30 20 July 33 21 31 19 August 32 20 28 18 September 30 18 24 17 October 28 17 26 16 November 27 16 25 16 December 27 16 24 16 Table 4-5. Site 2 optimal growout periods (months) for single plant production by production method and planting month. Planting Seed density/size Month 62.5/10 62.5/15 75/10 75/15 January 32 23 32 19 February 31 24 31 19 March 33 24 33 21 April 34 32 34 22 May 28 31 33 21 June 34 30 32 20 July 33 29 31 14 August 32 25 30 19 September 29 20 29 18 October 34 26 28 17 November 34 18 27 17 December 33 21 26 17

PAGE 60

54 Table 4-6. Site 3 optimal growout periods (months) for single plant production by production method and planting month. Planting Seed density/size Month 62.5/10 62.5/15 75/10 75/15 January 32 20 32 19 February 31 24 31 19 March 30 23 33 21 April 34 24 34 22 May 33 27 33 21 June 32 26 32 20 July 31 26 31 14 August 34 28 30 19 September 33 20 29 18 October 31 26 28 17 November 29 17 27 17 December 29 17 26 17 The next step required connecting consecutive production cycles so that optimal rotations could be identified. A rotation is defined as a pattern of sequential plant, growout harvest, replant, etc decisions. A given production schedule or cycle (as defined by the plant month and growout period) resulted in the next cycle beginning in a particular month. Unless the optimal growout period was 23 months, each growout schedule resulted in the next plant beginning in a different month than the previous cycle. For example, a January-Year I-plant growing for 23 months would be harvested in December, Year 2, allowing replanting in January, Year 3 Simultaneous harvest and replant was not allowed in the model due to the time and labor requirements of harvesting an entire acre, so the effective turnaround time was the growout period plus one month. If planting occurred in January and the growout period were 24 months,

PAGE 61

55 harvest would occur in January of the third year and replanting would occur in February. Then, the optimal February-plant growout period would be followed. This "second-cycle" start was identified for each starting month. Consecutive plants were then strung together. An example might be: January plant, 28-month growout, May harvest, June plant, 32-month growout, February harvest, March plant, 24-month growout, etc. In all cases stabilization or a repetitive cycle emerged. This occurred when any subsequent planting month was the same as any previous planting month. For example, any 23-month growout produced immediate stabilization. As indicated above, a January-plant growing for 23 months resulted in a December harvest and subsequent January replant. The effective turnaround time is 24 months, hence replanting always occurs in the same month. Some repetitive cycles involved multiple plants, while others eventually attained a 24-month repetitive cycle, but required several plants before falling into the cycle. Once planting patterns were identified, expected revenues were examined to determine optimal rotations. This required comparing the revenue potential of all combinations of immediate replacement and delayed cycles. A cycle with immediate replacement would be that as described in the previous paragraph. Immediate (next month) replacement would follow any harvest. A delayed cycle would include varying quantities of down months. A down month is one in which the acre is empty; no growing clams exist on the acre and neither harvest nor planting occurs. A down month was indicated if revenues could be increased by waiting. For instance, February conditions may be more conducive to rapid growth than January conditions. Hence, a new operation would make initial plants in February and, if harvest ever occurred in

PAGE 62

56 December, would delay replanting one month until February. Multiple down months were also a possibility. An example of an optimal rotation might be to plant in January, grow for 18 months, harvest in August, wait two months until November, plant, and then follow the optimal November cycle. Or, the optimal decision might be to wait four months after the August harvest and replant in January. Hence, although actual growout lasts only 18 months, the effective turnaround time is 24 months and the acre is empty for four months. Evaluation of the returns of all possible combinations of delayed and immediate replacement production allows the identification of the rotation that produces maximum expected net revenues. Optimal rotations were identified for each production method for each month of the year. Then, these 12 rotations (distinguished by the 12 calendar months) were compared to determine, given the opportunity to begin production in any month, the rotation that produced maximum expected net revenues. Optimal rotations for each production method are given in Table 47. Production results generated by the optimal rotations are given in Table 4-8. As can be seen where the number of clams sold is less than the number of clams survive, certain production schedules produce undersized clams. The difference between the procedures described in the above paragraphs is subtle yet significant. The first procedure addresses the question of, if the operation somehow finds itself in a given month with a planting option, what planting schedule should be initiated. Unexpected growth and harvest, mortality, natural disaster, etc. might result in an empty acre in any month. The second procedure addresses the question of when

PAGE 63

Table 4-7. Optimal planting rotations for single plant production by production method. Plant month/growout period (months) Seed density/size (mm) Wait 2 Wait 3 Wait 4 Site 1 62.5/10 Jan/28 5 Nov/27 Apr/29 no Nov repeat 62.5/15 Feb/22 1 Feb repea t 75/10 Sep/24 3 Jan/25 6 Sep repeat 75/15 Feb/19 no Oct/16 3 Jun/20 3 Jun repeat Site 2 62.5/10 Feb/31 2 Dec/33 2 Dec repeat 62.5/15 Jan/23 no Jan repeat 75/10 Feb/31 4 Feb repeat 75/15 Feb/19 4 Feb repeat Site 3 62.5/10 Jan/32 3 Jan repeat 62.5/15 Jan/20 3 Jan repeat 75/10 Feb/31 4 Feb repeat 75/15 Feb/19 4 Feb repeat

PAGE 64

Table 4-8. OQtimal rotation 12roduction results for single Qlant Qroduction. Average Size # Clams # Clams # Clams Seed densit~/size Age (months) Size (mm) STD Start Survive Sold Site 1 Plant month 62.5/10 January 28 63.96 6.01 750,000 606,000 606,000 November 27 64.47 5.92 750,000 609,000 609,000 April 29 50.16 5.17 750,000 603,000 506,520 62.5/15 February 22 61.79 5.88 750,000 624,000 624,000 75/10 September 24 50.39 5.19 900,000 741,600 622,944 January 25 51.84 5.28 900,000 738,000 619,920 75/15 February 19 63.99 6.01 900,000 759,600 759,600 October 16 62.12 5.90 900,000 772,200 772,200 June 20 64.37 6.03 900,000 756,000 756,000 Site 2 62.5/10 February 31 53.07 5.35 750,000 597,000 501,480 December 33 62.05 5.89 750,000 591,000 591,000 62.5/15 January 23 62.91 5.94 750,000 621,000 621,000 75/10 February 31 51.15 5.23 900,000 716,400 601,776 75/15 February 19 61.80 5.88 900,000 759,600 759,600 Site 3 62.5/10 January 32 61.77 5.87 750,000 594,000 594,000 62.5/15 January 20 62.11 5.90 750,000 630,000 630,000 75/10 February 31 51.09 5.23 900,000 716,400 601,776 75/15 Febru~ 19 61.77 5.87 9001000 7591600 7591600 \JI 00

PAGE 65

59 a new lease should begi n operation. For example, at Site 1 using 10 mm seed planted at 62.5 clams per square foot, a February plant requires a 34-month growout to maximize net revenue. See Table 4-4. If expected conditions occur, however, the grower should never be faced with a planting decision in February as the optimal planting rotation is January-November-April-November. See Table 4-7. Once optimal rotations are determined, rotatio n s not included in the optimal design are relevant only if unexpected conditions oc c ur. Multiplot Production Opt i mal Rotations A simple management adjustment to a monthly marketing constraint would be to follow the rotation pattern determined for single plant production and harvest over a 12month period rather than in a single month. To do so, however, would incur additional mortality and produce potential revenue losses as clams grew into less profitable size categories. Additionally, as the acre is planted as a unit, replanting would be affected as insufficient space woul d exist for the new seed. Correct determination of proper management strategy when faced with a monthly harvest requirement, howe v er, requires a change in perspective since the production or management question is now different than that with single plant production. With single plant production the questi o n is, given both planting and marketing freedom, what is the best rotation to follow. All seed is assumed essentially identical with respect to growth potential. Optimal plant and growout times are therefore identical for all seed. When given the freedom or option to plant and harvest all clams simultaneously, the correct decision would be to do so All clams are allowed to attain economic maturity. The

PAGE 66

60 correct production ques t ion now is, however, given that clams must be marketed each month, when is the best time to plant seed such that they will be available for sale in each calendar month. The intuitive impact of this restriction is that clams will be planted and harvested in months other than those prescribed by the single plant production analysis. The monthly mar k eting constraint imposes 12 harvests annually. No restriction was placed on when clams were planted; all clams could be planted during the same month, or any other combination. Actual planting schedules depend upon revenue comparisons of the various options. The clams would be harvested, however, in monthly units. Thus, the production unit, as driven by the harvest unit, is now a fraction of an acre. This requires that the acre be subdivided into units or plots, each representing a separate harvest and, possibly, a separate plant. To simplify the analysis, it was assumed that all plots be equal in s i ze and, hence, planted with equal numbers of clam seed. As previously mentioned, the monthly harvest constraint did not specifically impose quantity restrictions, but rather only required positive monthly sales. All plots were therefore identical and the analysis did not address the question of optimal monthly planting quantities subject to minimum or maximum sales constraints. The analysis could be modified to account for specific volume sales restrictions. Planting schedules are driven by harvesting requirements since, as previously stated, the economic question is when is the best time to plant seed for harvest in month i. The clam growth simulation results and net revenue estimations used in the first analysis were appropriate for use in determining when harvests began and, hence, when

PAGE 67

61 plants were required. An initial harvest rule was selected such that harvest began (given the option of planting in each of the 12 calendar months) when the expected net revenues from a given plant exceeded all other potential revenues for that month from any other plant. This determination required a comparison of not only all potential revenues for that month in that year, but also for that month in all future years. Future year evaluations incorporated a consideration of the revenue potential of not only a given plant 12 months later, but also that of other plants that might not have been planted when the initial observation is made. Since production or growth potential was repetitive, only one set of simulation results had to be examined to encompass all future years. With growout simulated for 34 months using 12 different starting months, the simulation results covered a 45-month period. At one site with a particular production method, a January plant (the first possible plant) with a 28-month growout was determined to be the best way to produce clams for harvest in May. Harvest therefore began in May and the harvest cycle or year was May to Apri l. Harvest was not begun in April because April revenues were maximized by an August plant (with a 32-month growout). Since January was the reference or first possible planting month, an August planting opportunity had yet to occur. April revenues were therefore maximized by waiting until August. This relationship repeated for all other months prior to the first May harvest month: revenues could be increased by waiting. May revenues were greatest, however, with a January plant and, thus, production began in January and produced a first harvest two years later in May.

PAGE 68

62 The initial harvest therefore began when revenues were greatest relative to all other production options for that particular month (year neutral). Determination of the best planting time for the next and all subsequent monthly harvests was then based on comparison of potential revenues for that specific month and year. The distinction between the first step--that of identifying the initial harvest--and this step is important. The first step determined both the month and the year of initial harvest. Harvest then became anchored to that year and the subsequent identification of maximum revenues became year-specific. Revenue comparison thus considered only clams present at that point in time and not clams in future years. Despite the narrowing of focus, once harvest began, revenues chosen represented overall maximums similar to the initial harvest. Maximum revenues were identified and then tracked back to determine planting schedules. This identified an annual planting schedule consisting of 12 plants. These were then linked to determine the required acre-subdivision by requiring identical annual planting schedules and comparing planting requirements with plot availability as determined by harvests. Same-month harvest and replanting was allowed with multiplot production due to the reduced labor requirements. With 750 bags per acre and a minimum of 12 plots per acre, at most 63 bags would be harvested in any month. Replanting during the same month would not be a problem. The number of required plots per acre are given in Table 4-9. Total plot requirements were determined on alease or 2-acre basis. Per acre plot requirements were then computed by dividing this total by 2. Fractional per acre plot numbers were indicative of one production unit (plot) spread over two acres. The number of bags per plot are given in Table 4-10. This was

PAGE 69

63 Table 4-9. Number of required plots per acre for multiplot hard clam production. Seed density/seed size 62 5/10 62.5/15 75/10 75/15 Site 1 17 12 14 12 Site 2 17 12.5 17 12 Site 3 16.5 12 16.5 12 Table 4-10 Number of growout bags per plot for multiplot hard clam production. Seed density/seed size 62 5/10 62.5/15 75/10 75/15 Site 1 45 63 54 63 Site 2 45 60 45 63 Site 3 46 63 46 63 computed by dividing the number of bags per acre 750, by the number of plots per acre. In some instances rounding requirements in the computation of the number of bags per acre resulted in more than 750 bags planted per acre. Per acre bag totals are given in Table 4-11. The number of required plots were then used to scale production and revenues to reflect the decrease in the unit of analysis from one acre to one plot. The multiplot production numbers are given in Tables 4-12, 4-13 and 4-14 Table 4-11. Total bags planted per acre for multiplot hard clam production. Seed density/seed size 62.5/10 62 5/15 75/10 75/15 Site 1 765 756 756 756 Site 2 765 750 765 756 Site 3 759 756 759 756

PAGE 70

Table 4-12. Site 1 multiQlot 2roduction results. Growout Average Size # Clams # Clams # Clams Seed densit}'./ size {months} Size {mm} STD Start Survive Sold Plant month Harvest month 62.5/10 January 28 May 63.95 6.00 45,000 36,360 36,360 January 29 June 65.65 6.10 45,000 36,180 36,180 January 30 July 65.91 6.12 45,000 36,000 36,000 January 31 August 66.17 6.13 45,000 35,820 35,820 January 32 September 66.49 6.15 45,000 35,640 35,640 January 33 October 66.62 6.16 45,000 35,460 35,460 January 34 November 66.94 6.18 45,000 35,280 35,280 February 34 December 61.83 5.87 45,000 35,280 35,280 August 32 April 63.78 5.99 45,000 35,640 35,640 November 27 February 62.47 5.91 45,000 36,540 36,540 December 25 January 53.86 5.39 45,000 36,900 30,996 December 27 March 61.85 5.87 45,000 36,540 36,540 Total annual sales: 425,916 62.5/15 January 17 June 63.46 5.97 63,000 53,676 53,676 January 18 July 64.09 6.01 63,000 53,424 53,424 January 19 August 64.72 6.05 63,000 53,172 53,172 January 20 September 65.25 6.08 63,000 52,920 52,920 January 21 October 65.59 6.10 63,000 52,668 52,668 January 22 November 66.37 6.15 63,000 52,416 52,416 February 22 December 61.78 5.87 63,000 52,416 52,416 February 23 January 66.19 6.14 63,000 52,164 52,164 March 23 February 67.06 6.19 63,000 52,164 52,164 April 23 March 64.14 6.01 63,000 52,164 52,164 December 16 April 63.83 5.99 63,000 53,928 53,928 December 17 May 66.97 6.18 63,000 53,676 53,676 Total annual sales: 634,788

PAGE 71

Table 4-12--continued. Growout Average Size # Clams # Clams # Clams Seed densit}'./size {months} Size {mm} STD Start Survive Sold Plant month Harvest month 75/10 January 25 February 51.83 5.27 64,800 53,136 44,634 February 25 March 50.89 5.21 64,800 53,136 44,634 February 28 June 53.80 5.39 64,800 52,358 43,981 February 30 August 54.18 5.41 64,800 51,840 43,546 March 25 April 50.95 5.22 64,800 53,136 44,634 March 26 May 51.82 5.27 64,800 52,877 44,417 April 27 July 50.54 5.19 64,800 52,618 44,199 September 24 September 50.39 5 18 64,800 53,395 44,852 October 24 October 50.44 5.19 64,800 53,395 44,852 October 27 January 53.82 5.39 64,800 52,618 44,199 November 24 November 50.26 5.18 64,800 53,395 44,852 December 24 December 50.00 5.16 64,800 53,395 44,852 Total annual sales: 533,652 75/15 January 16 May 62.80 5.93 75,600 64,714 64,714 January 17 June 64.84 6.05 75,600 64,411 64,411 January 19 August 67.20 6.20 75,600 63,806 63,806 February 17 July 62.42 5.91 75,600 64,411 64,411 February 19 September 63.99 6.00 75,600 63,806 63,806 February 20 October 65.21 6.08 75,600 63,504 63,504 February 21 November 66.52 6.16 75,600 63,202 63,202 April 20 December 62.65 5.92 75,600 63,504 63,504 May 20 January 62.97 5.94 75,600 63,504 63,504 June 20 February 64.37 6.03 75,600 63,504 63,504 August 19 March 64.80 6.05 75,600 63,806 63,806 December 16 April 63.70 5.99 75,600 64,714 64,714 Total annual sales: 766 886 O'I VI

PAGE 72

Table 4-13. Site 2 multiQlot Qroduction results. Growout Average Size # Clams # Clams # Clams Seed densit~/size {months} Size {mm} STD Start Survive Sold Plant month Harvest month 62.5/10 January 27 April 53.25 5.36 45,000 36,540 30,694 January 29 June 57.83 5.63 45,000 36,180 30,391 January 31 August 58.15 5.65 45,000 35,820 30,089 February 27 May 50.48 5.19 45,000 36,540 30,694 February 29 July 52.44 5.31 45,000 36,180 30,391 February 31 September 53.06 5.34 45,000 35,820 30,089 February 32 October 53.28 5.36 45,000 35,640 29,938 February 33 November 53.42 5.37 45,000 35,460 29,786 February 34 December 54.31 5.42 45,000 35,280 29,635 March 34 January 52.63 5.32 45,000 35,280 29 635 November 28 March 55.40 5.49 45,000 36,360 30 542 December 26 February 50.52 5.19 45,000 36,720 30,845 Total annual sales: 362,729 62.5/15 January 23 January 62.91 5.94 60,000 49,680 49,680 January 24 February 66.32 6.14 60,000 49,440 49,440 February 24 March 64.31 6 02 60,000 49,440 49,440 March 24 April 62.74 5.93 60,000 49,440 49,440 March 25 May 67.12 6.19 60,000 49,200 49,200 August 25 October 62.76 5.93 60,000 49,200 49 200 November 18 June 63.79 5.99 60,000 50,880 50,880 November 21 September 67.20 6.20 60,000 50,160 50,160 December 18 July 62.88 5.94 60,000 50,880 50,880 December 19 August 63.20 5.96 60,000 50,640 50,640 December 22 November 64.86 6.06 60,000 49,920 49,920 December 23 December 65.29 6.08 60,000 49,680 49,680 Total annual sales: 598,560 O'I O'I

PAGE 73

Table 4-13--continued. Growout Average Size # Clams # Clams # Clams Seed density/size {months} Size {mm} STD Start Survive Sold Plant month Harvest month 75/10 January 26 March 50.28 5.18 54,000 44,064 37,014 January 27 April 51.71 5.26 54,000 43,848 36,832 February 27 May 50.52 5.19 54,000 43,848 36,832 February 28 June 50.74 5.20 54,000 43,632 36,651 February 29 July 50.88 5.21 54,000 43,416 36,469 February 30 August 51.04 5.22 54,000 43,200 36,288 February 31 September 51.15 5.23 54,000 42,984 36,107 February 32 October 51.18 5.23 54,000 42,768 35,925 March 32 November 50.17 5.16 54,000 42,768 35,925 March 33 December 50.47 5.19 54,000 42,552 35,744 March 34 January 51.21 5.23 54,000 42,336 35,562 December 26 February 50.41 5.18 54,000 44,064 37,014 Total annual sales: 436,363 75/15 January 17 June 62 70 5.93 75,600 64,411 64,411 January 18 July 63.97 6.05 75,600 64,109 64,109 January 19 August 64.78 6.05 75,600 63,806 63,806 January 21 October 66.12 6.13 75,600 63,202 63,202 January 22 November 67.18 6.20 75,600 62,899 62,899 February 19 September 61.79 5.87 75,600 63,806 63,806 March 21 December 63.23 5.96 75,600 63,202 63,202 April 21 January 62.74 5.93 75,600 63,202 63,202 June 20 February 62.35 5.90 75,600 63,504 63,504 June 21 March 67.20 6.20 75,600 63,202 63,202 November 17 April 63.86 6.00 75,600 64,411 64,411 December 17 May 63.77 5.99 75,600 64,411 64,411 Total annual sales: 764 165 O'I -:i

PAGE 74

Table 4-14. Site 3 multiQlot 2roduction results. Growout Average Size # Clams # Clams # Clams Seed densit)'./size {months} Size {mm} STD Start Survive Sold Plant month Harvest month 62.5/10 January 26 March 51.89 5.27 46,000 37,536 31,530 January 32 September 61.77 5.87 46,000 36,432 36,432 January 33 October 62.75 5.93 46,000 36,248 36,248 January 34 November 62.75 5.93 46,000 36,064 36,064 February 26 April 50.00 5.16 46,000 37,536 31,530 February 27 May 52.33 5.30 46,000 37,352 31,376 February 28 June 54.19 5.41 46,000 37,168 31,221 February 30 August 54.81 5.45 46,000 36,800 30,912 March 28 July 50.85 5.21 46,000 37,168 31,221 March 33 December 53.42 5.37 46,000 36,248 30,448 March 34 January 56.68 5.56 46,000 36,064 30,294 December 26 February 50.00 5.16 46,000 37,536 31,530 Total annual sales: 388,806 62.5/15 January 20 September 62.11 5 89 63,000 52,920 52,920 January 21 October 63.25 5.96 63,000 52,668 52,668 January 22 November 63.28 5.96 63,000 52,416 52,416 January 23 December 64.18 6.02 63,000 52,164 52,164 February 23 January 62.14 5.89 63,000 52,164 52,164 March 23 February 62.43 5.91 63,000 52,164 52,164 March 24 March 65.28 6.08 63,000 51,912 51,912 July 24 July 62.77 5.93 63,000 51,912 51,912 September 21 June 62.21 5.90 63,000 52,668 52,668 November 17 April 62.13 5.89 63,000 53,676 53,676 December 17 May 62.47 5.91 63,000 53,676 53,676 December 20 August 66.33 6.14 63,000 52,920 52,920 Total annual sales: 631,260 0\ 00

PAGE 75

Table 4-14--continued. Growout Average Size # Clams # Clams # Clams Seed densit~/size {months} Size {mm} STD Start Survive Sold Plant month Harvest month 75/10 January 26 March 50.11 5.17 55,200 45,043 37,836 January 27 April 51.53 5.25 55,200 44,822 37,650 February 27 May 50.30 5.18 55,200 44,822 37,650 February 28 June 50.60 5.20 55,200 44,602 37,466 February 29 July 50.80 5.21 55,200 44,381 37,280 February 30 August 50.98 5.22 55,200 44,160 37,094 February 31 September 51.09 5.23 55,200 43,939 36,909 March 31 October 50.07 5.16 55,200 43,939 36,909 March 32 November 50.13 5.17 55,200 43,718 36,723 March 33 December 50.41 5.18 55,200 43,498 36,538 April 33 January 50.08 5.16 55,200 43,498 36,538 December 26 February 50.60 5.20 55,200 44,043 37,836 Total annual sales: 446,429 75/15 January 17 June 62.31 5.90 75,600 64,411 64,411 January 18 July 63.82 5.99 75,600 64,109 64,109 January 19 August 64.58 6.04 75,600 63,806 63,806 January 21 October 66.13 6 13 75,600 63,202 63,202 January 22 November 67.02 6.19 75,600 62,899 62,899 February 19 September 61.76 5.87 75,600 63,806 63,806 March 21 December 63.07 5.95 75,600 63,202 63,202 April 21 January 62.82 5.93 75,600 63,202 63,202 June 20 February 62.45 5.91 75 600 63,504 63,504 June 21 March 64.54 6.04 75,600 63,202 63,202 November 17 April 63.57 5.98 75,600 64,411 64,411 December 17 May 63.55 5.98 75,600 64,411 64,411 Total annual sales: 764 165 0\

PAGE 76

70 Production Performance Comparison Stabilized production results were used to make performance comparisons between single plant and multiplot production. An operation reached stabilization when both acres had been brought into production, production cycles had achieved optimal patterns (as described in previous sections) and operating loan requirements were zero (as determined by the full economic evaluation discussed in the next section). Optimal repetitive cycles varied by site, production method and marketing constraint. Optimal cycles defined when plants and, more importantly, harvests occurred. Multiplot production specifically required monthly and, once begun, annual harvests. Hence, although individual plants might grow for up to 34 months, optimal repetitive cycles were annual in nature. Single plant production had no such restrictions and, thus, allowed cycle length to be determined by growth and economic dynamics. It was therefore possible to have years where no harvest (or plant) occurred. For example, all single plant production scenarios using 10 mm seed produced cycles that included years in which no harvest occurred due to the length of the growout periods. Site 1 cycles covered a 5-year period of four harvests followed by one year with no harvest. Cycles at sites 2 and 3 using 10 mm seed covered three years with two harvests followed by a year of inactivity. All scenarios using 15 mm seed produced annual harvests. Cycle length considerations were used in the determination of summary statistics for model comparison. The consideration of average production allows more direct comparison between the different production methods as it incorporates the negative effect of inactive years due to extended growout times. Not all summary statistics

PAGE 77

71 required averaging of this type. Average clam size and growout time were actual performance values and did not require consideration of cycle length. Average clam size and growout times for both single plant and multiplot production are given in Table 4-15. No consistent patterns emerged when comparing single plant and multiplot results; a given seed size and density might produce larger clams under single plant production than under multiplot production at one site and smaller clams at another site. Average growout times were similarly nonconsistent. Within a given method (single plant versus multiplot) using a given seed size, increased growout times produced larger clams. Five of the 12 multiplot scenarios (three sites and four production options per site) had longer average growout times yet produced larger clams in seven scenarios. Thus, in two occasions, shorter multiplot average growout times produced larger average clams. This could be explained by, as the multiplot production mix typically included both shorter and longer growout periods for different plants, the impact of the larger clams (produced by the longer growout periods) was greater than that of the smaller clams (produced by the shorter growout periods). A comparison of the annual number of clams planted, survive and sold is shown in Table 4-16. Patterns were again difficult to discern. Multiplot production generally resulted in greater numbers of clams planted (10 of 12 situations) and sold (9 of 12 situations). Plant quantities varied due to differences in total bags per acre and the effects of extended growout periods and plot subdivision. Percentage survival, however, could be higher/lower yet produce fewer/greater sales. The production of undersized clams was indicated when the quantity of clams sold was less than the quantity survived. With only

PAGE 78

72 Table 4-15. Average final clam size (mm) and growout time (months) for single plant and multiplot hard clam production. Seed density/ Average Average Seed size Size Growout time Single eiant eroduction Site 1 62.5/10 57.32 28 62.5/15 61.79 22 75/10 51.12 24.5 75/15 64.37 20 Site 2 62.5/10 62.05 33 62.5/15 62.91 23 75/10 51.15 31 75/15 61.80 19 Site 3 62.5/10 61.77 32 62.5/15 62.11 20 75/10 51.09 31 75/15 61.77 19 MultiElot Eroduction Site 1 62.5/10 63.79 30.2 62.5/15 64.95 20.1 75/10 51.58 25.8 75/15 64.29 18.7 Site 2 62.5/10 53.73 30 1 62 5/15 64.45 22.2 75/10 50.81 29.6 75/15 64.14 19.4 Site 3 62.5/10 55.12 29 8 62.5/15 63.22 21.3 75/10 50.56 29.4 75/15 63.80 19.4

PAGE 79

Table 4-16. Average annual numbers of clams planted, survive and sold for single plant and multiplot production. Seed density/ # # # % % Sold % Sold Seed size Plant Survive Sold Survival (of plant) (of survive) Single plant production Site 1 62 5/10 600,000 484,800 446,208 80.8 74.4 92.0 62.5/15 750,000 624,000 624,000 83.2 83.2 100.0 75/10 720,000 591 840 497,146 82.2 69.1 84.0 75/15 900,000 756 000 756,000 84.0 84.0 100.0 Site 2 62.5/10 500,000 394,000 334,320 78.8 66.9 84.9 62.5/15 750,000 621,000 621,000 82.8 82.8 100.0 75/10 600,000 477,600 401,184 79.6 66.9 84.0 75/15 900,000 759,600 759,600 84.4 84.4 100.0 Site 3 62.5/10 500,000 396,000 396,000 79.2 79.2 100.0 62.5/15 750,000 630,000 630,000 84.0 84.0 100.0 75/10 600,000 477,600 401 184 79.6 66.9 84.0 75/15 900,000 759,600 759,600 84.4 84.4 100.0

PAGE 80

Table 4-16--continued. Seed density/ # # # % % Sold % Sold Seed size Plant Survive Sold Survival (of plant) (of survive) Multiplot production Site 1 62.5/10 540,000 431,640 425,916 79.9 78.9 98.7 62.5/15 756,000 634,788 634,788 84.0 84.0 100.0 75/10 777,600 635,299 533,652 81.7 68.6 84.0 75/15 907,200 766,886 766,886 84.5 84.5 100.0 Site 2 62.5/10 540,000 431,820 362,729 80.0 67.2 84.0 62.5/15 720,000 598,560 598,560 83.1 83.1 100.0 75/10 648,000 519,480 436,363 80.2 67.3 84.0 75/15 907,200 764,165 764,165 84.2 84.2 100.0 Site 3 62.5/10 552,000 442,152 388,806 80.0 70.4 87.9 62.5/15 756,000 631,260 631,260 83.5 83.5 100.0 75/10 662,400 530,465 446,429 80.1 67.4 84.0 75/15 907,200 764,165 764,165 84.2 84.2 100.0

PAGE 81

75 one exception, undersized clams were harvested with the same production methods in both single plant and multiplot production. It should be noted that the above comparisons are intended to simply describe some of the production results. Since determinations of preferred or optimal cycles were based on economic criteria, explanations and an understanding of why specific results might occur require considerations of the cost and revenue impacts of the various production options. Economic Performance The hard clam growth simulation and the identification of optimal rotations described in the previous sections identified and described potential challengers. A logical application of this information and the simulation models employed would be to take an existing growout operation, identify the profile of existing stock, and determine harvest and replacement schedules for existing stock given knowledge of challengers and the expected revenues from existing stock. Despite being extremely situation specific, such an application would be descriptive of the use to which this research might be applied as a management or extension tool. A second application, however, involves applying the previously described optimal rotations to a new growout operation and determining the expected returns and accrued benefits of the resultant operation. This second application was chosen for the remainder of the research. The rationale for this selection was that, if expected conditions occurred, such an evaluation provided more insight into what an optimally managed operation might be worth. With the first application, once existing stock is

PAGE 82

76 replaced and expected conditions again occur, an existing operation would likely fall into an identical new-operation production pattern. The onset of the pattern would, however, be delayed according to the continued growout requirements of existing stock. If expected conditions do not occur, i.e. expected environmental conditions fail to occur, expected conditions fail to produce expected growth, or expected prices fail to be realized, then the system is in a constant state of flux requiring continuous evaluation, and the system likely never attains repetitive patterns. The optimal rotations for both production methods described both planting and harvesting patterns for one half of the lease (one acre). This pattern was then repeated on the second acre, thereby determining planting and harvest management of the entire lease. These management patterns were used to determine asset replacement and operating expenditure requirements The expenditure requirements were then used to determine fixed and variable costs according to guidelines described in Adams et al. (1993). The operation described in Adams et al. and modelled here assumes that the owner operator is employed full time in some other occupation and that the culture operation exists as a part-time activity. Alternative income activities are assumed limited to minimum wage jobs. Asset and production requirements were combined with expected revenues in a cashflow model to determine annual loan requirements, finance payments, taxes net positions, etc. Financial assumptions common to both production scenanos were (a) Sixty-five percent of initial investment requirements and asset replacement is financed by borrowed capital. (b) All operating loans are fully financed by borrowed funds.

PAGE 83

77 (c) All debt capital has a 9.5 % real annual rate of interest over a ten-year loan period. (d) Depreciation on capital assets is computed using straight-line depreciation. (e) The wage rate is $4.10 per hour. (t) Monitoring costs are $120 per acre per month. (g) Growout bags are harvested at the rate of two bags per hour. This includes pulling, grading and bagging (sales bags). The debt capital interest rate assumes a 4% real interest rate and a 5.5% risk premium. Initial investment costs are listed in Appendix 4 and production costs are listed in Appendix 5. Income statements Income statements are developed using the results of the cashflow analysis. Average expectations of stabilized production, as described in the previous section on production comparisons, were determined. Again, an operation reached stabilization when full production of the lease had begun, production cycles had achieved optimal repetitive patterns and operating loan requirements were zero. Values used in the income statements thus represented averages over the repetitive cycle. Returns net of variable and fixed costs represent the returns to the land and owner labor, management and capital. The opportunity cost of owner labor was calculated by determining the annual number of labor hours required and assuming the next best alternative was a minimum wage job. The opportunity cost of capital was calculated at 11.5% of owned equity. This interest rate included a 2% premium over the loan rate.

PAGE 84

78 The next step, required identifying the residual claimant and quantifying the residual. The residual claimant is that component to which final net returns or residual accrues. Both management and land might be considered valid residual claimants. Selection of the residual claimant requires a thorough understanding of the productive process examined, and consideration of the goals of the research and the ability to realistically isolate the contributions of specific inputs. Although management of a small hard clam growout operation is a new endeavor, precedent exists for estimating management fees in other agricultural fields as a percentage of gross or net returns. Management fees ranging from three to 10 percent are used in grain, vegetable and citrus operations with the higher percentages usually awarded for the more labor intensive crops (various personal communications). Most management payments likely include base salaries with performance incentives. The lack of a history of hard clam management and the difficulty of the identification of available alternatives, however, makes detailed estimation of management fees difficult. Percentage calculation is, therefore, appealing. Valuation of the land is equally problematic. The argument could be made that, smce a lease fee structure already exists, these costs should be deducted, thereby attributing the residual to management. Current lease fees, however, are only $20 per acre per year, and thus likely represent only filing or paperwork fees. This fee thus fails to represent the unique productive capacity of the land. Considerations of alternative commercially productive use values is likely limited to oyster culture, an emerging operation similar to clam culture and one in which evaluations of economic potential are similarly incomplete.

PAGE 85

79 Given the above considerations, the decision was made to make the land the residual claimant. Production results indicated that returns were driven by site specific characteristics. The same management philosophy or decision rules were applied across all production options at all three sites. Actual management in terms of specific planting and harvest schedules varied across sites, but the rules determining or identifying optimal rotations were identical. Returns, however, varied by site. Thus, the land was chosen as the residual claimant. Returns to management were calculated as 3 % of gross revenues. Residual returns represented the returns to the lease or two acres of water bottom. The results of this evaluation for single plant production are shown in Tables 4-17, 4-18 and 4-19. Multiplot production financial results are shown in Tables 4-20, 4-21 and 4-22. Examination of the income statements showed that, while differences in performance existed, all scenarios produced significant residual returns ranging from a low of $4,000 per lease under multiplot production at Site 2 using 10-mm seed planted at 62.5 clams per square foot, to a high of $33,400 per lease at Sites 2 and 3 under single plant production using 15-mm seed planted at 75 clams per square foot. Net present and annualized values The figures presented in the income statements represented expected performance of the operation after operating loans are retired and the planting schedules achieve an optimal cycle. Thus, the income statements fail to incorporate performance information prior to the achieved stability. Two other measurements, the net present value and the annualized value, capture these off-year effects and are therefore more descriptive. The

PAGE 86

80 Table 4-17. Site 1 single plant production income statement. Production method (seed density/seed size) 62.5/10 62.5/15 75/10 75/15 Revenue $56,000 $75,900 $62,100 $97,100 Expenses Variable cost 13,200 19,700 15,100 22,800 Fixed cost Overhead 1,000 1,000 1,000 1,000 Debt interest 3,100 2,200 3,000 2,300 Depreciation 8,300 6,200 7,800 6,200 Taxes 8,400 13,100 11,200 17,900 Net return to labor, management, 22,000 33,700 24,000 46,900 capital and land Opportunity cost of labor 2,900 3,600 2,900 3,600 Opportunity cost of capital 7,300 5,000 5,700 15,700 Opportunity cost of management 1,700 2,300 1,900 2,900 Net return to land 10,100 22,800 13,500 24,700

PAGE 87

81 Table 4-18. Site 2 single plant production income statement. Production method (seed density/seed size) 62.5/10 62.5/15 75/10 75/15 Revenue $50,600 $75,500 $52,200 $97,500 Expenses Variable cost 11,500 19,700 13,000 22,800 Fixed cost Overhead 1,000 1,000 1,000 1,000 Debt interest 2,900 2,200 2,700 2,200 Depreciation 7,200 6,200 7,200 6,200 Taxes 9,100 13,000 10,500 18,300 Net return to labor, management, 18,900 33,400 17,800 47,000 capital and land Opportunity cost of labor 2,400 3,600 2,400 3,600 Opportunity cost of capital 6,600 4,900 3,700 7,100 Opportunity cost of management 1,500 2,300 1,600 2,900 Net return to land 8,400 22,600 10,100 33,400

PAGE 88

82 Table 4-19. Site 3 single plant production income statement. Production method (seed density/seed size) 62.5/10 62.5/15 75/10 75/15 Revenue $50,800 $80,900 $52,200 $97,500 Expenses Variable cost 11,500 19,700 13,000 22,800 Fixed cost Overhead 1,000 1,000 1,000 1,000 Debt interest 2,900 2,200 2,700 2,200 Depreciation 8,300 6,200 7,200 6,200 Taxes 10,100 14 500 10,500 18,300 Net return to labor, management, 17,000 37,300 17,800 47,000 capital and land Opportunity cost of labor 2,400 3,600 2,400 3,600 Opportunity cost of capital 3,600 5,600 3,700 7,100 Opportunity cost of management 1,500 2,400 1,600 2,900 Net return to land 9 500 25,700 10,100 33,400

PAGE 89

83 Table 4 20. Site 1 multiplot production income statement. Production method (seed density/seed size) 62 5/10 62.5/15 75/10 75/15 Revenue $48 400 $72,100 $60 400 $89 600 Expenses Variable cost 11 100 21 300 14 700 21,400 Fixed cost Overhead 1,000 1 000 1,000 1,000 Debt interest 2 800 2 200 3,100 2,200 Deprec i ation 7 600 6,200 8 500 6,200 Taxes 7 300 12 400 9,300 16 500 Net return to labor management 18 600 28 800 23 800 42,300 capital and land Opportunity cost of labor 3 000 3,700 3 300 3,700 Opportunity cost of capital 7 600 6 800 8,300 11,400 Opportunity cost of management 1 500 2,200 1,800 2,700 Net return to land 6 500 16 200 10,400 24,500

PAGE 90

84 Table 4-21. Site 2 multiplot production income statement. Production method (seed density/seed size) 62.5/10 62.5/15 75/10 75/15 Revenue $41,400 $69,500 $49,100 $88,700 Expenses Variable cost 11,000 17,500 12,700 21,400 Fixed cost Overhead 1,000 1,000 1,000 1,000 Debt interest 2,800 2,300 2,800 2,200 Depreciation 7,600 6,100 7,600 6,200 Taxes 5,300 11,900 7,000 16,300 Net return to labor management 13 700 30,700 18,000 41 600 capital and land Opportunity cost of labor 3,000 3,500 3,000 3,700 Opportunity cost of capital 5,500 9,600 6 700 9,300 Opportunity cost of management 1,200 2,100 1,500 2,700 Net return to land 4,000 15,500 6,700 25,900

PAGE 91

85 Table 4-22. Site multiplot production income statement. Production method (seed density/seed size) 62.5/10 62.5/15 75/10 75/15 Revenue $44,300 $72,800 $50,200 $88,700 Expenses Variable cost 11,200 18,300 12,900 21,400 Fixed cost Overhead 1,000 1,000 1,000 1,000 Debt interest 2,900 2,300 2,900 2,200 Depreciation 7,700 6,200 7,700 6,200 Taxes 6,000 12,600 7,200 16,300 Net return to labor, management, 15,500 32,400 18,500 41,600 capital and land Opportunity cost of labor 3,000 3,700 3,000 3,700 Opportunity cost of capital 5,500 8,900 6,900 9,300 Opportunity cost of management 1,300 2,200 1,500 2,700 Net return to land 5,700 17,600 7,100 25,900

PAGE 92

86 net present value represents net returns--calculated from initial start-up--discounted over 20 years at 10.2 % The present value per acre represents the amount a person should be willing to pay to obtain a 2-acre 20-year lease. The annualized value represents an average annual return to the lease over the same 20-year period and is comparable to the annual lease fee--$40 for two acres--for performance evaluation. The annualized value is the amount a person should be willing to pay annually for the 2-acre lease. The discount rate used represented an average of the loan rate and the opportunity cost of capital rate weighted by the capital asset financing ratio previously specified. The present value of the land for all production scenarios is given in Table 4-23 and the annualized values are given in Table 4-24. A comparison of the present value per lease (2 acres of land) for single plant and multiplot production show values from single plant production to exceed those from multiplot production in all cases. Differences are minor, however, at Site 1 planting 15mm seed at 75 clams per square foot. An examination of the income statements of the two scenarios shows that, while the gross revenues for single plant production exceed those of multiplot production, $97,100 to $89,600, a considerable portion of this difference, $4,300, is negated when the opportunity cost of capital is accounted for. Further evaluation of the two situations using production information not presented here attributed the higher opportunity cost of capital with single plant production to a faster rate of equity build-up. Specifically, on January 1 of any given year, the date on which evaluations were based, both acres were fully planted under single plant production whereas only 1.33 acres were planted under multiplot production. At this particular site,

PAGE 93

87 Table 4-23. Present value of a two-acre lease across all sites and production options. Seed density/seed size 62.5/10 62.5/15 75/10 75/15 Site 1 Single plant $19,700 $145,200 $42,300 $141,300 Multiplot -9,600 83,300 13,800 140,900 Site 2 Single plant 11,900 143,400 51,000 220,300 Multiplot -21,300 39,100 -5,500 147,900 Site 3 Single plant 47,400 166,100 51,000 220,300 Multiplot -8,400 74,100 -3,900 147,900 Table 4-24. Annualized value of a two-acre lease across all sites and production options. Seed density/seed size 62.5/10 62.5/15 75/10 75/15 Site 1 Single plant $1,700 $12,200 $3,600 $11,900 Multiplot -800 7,000 1,200 11,800 Site 2 Single plant 1,000 12,000 4,300 18,500 Multiplot -1,800 3,300 -500 12,400 Site 3 Single plant 4,000 14,000 4,300 18,500 Multiplot -700 6 ,2 00 -300 12,400

PAGE 94

88 there was a concentration of monthly plants early in the year. Three plants were scheduled for January and four for February. Thus, seven plots were empty at the point of evaluation. The results confirm expectations that harvest restrictions impair the economic performance of the operation. Imposing harvest restrictions leads to either or both planting and harvesting clams other than when otherwise prescribed in the absence of such restrictions. This can produce economically immature and overage clams as well as increased mortality. Monthly marketing reduced revenues sufficiently that in only one instance, at Site 1 planting at 75 clams per square foot, were 10-mm seed profitable. Revenues were never sufficient to overcome initial negative net balances. Larger seed always outperformed smaller seed, while higher planting densities almost always outperformed lower planting densities. The one exception occurred at Site 1 where the lower density using 15-mm clams outperformed the higher density. Reasons can again be traced to equity build-up. See Table 4-17. The greatest returns were achieved by using the largest seed planted at the highest density at Sites 2 and 3. The major cost impact of using larger seed came from higher seed costs. These costs are more than recouped, though, through faster turnover of stock. Similarly, although growth was affected by planting density with higher densities producing more conservative growth, the negative effects of the higher density were not sufficient to offset the gains attributed to larger volume sales. Selection of the best site varied with production method. Each site represented physically distinct locations possessing unique environmental profiles in terms of

PAGE 95

89 expected water temperature, salinity and dissolved oxygen. The results indicate that site selection is an important factor to consider in hard clam growout. How to effectively incorporate this information in management decisions is less clear. While ambient environmental conditions varied from site to site, determination of combinations that can be classified as productively similar or different is difficult. Also, identical productive or economic potential depended upon both production method and harvest restriction. Sites 2 and 3 produced identical economic outcomes under single plant production at 75 clams per square foot, but not at 62.5 clams per square foot. Further, under multiplot production, Sites 2 and 3 produced identical returns using 15-mm seed planted only at the higher density. Thus, the incorporation of environmental considerations is difficult. Trends evidenced by the net present values of the various production options are repeated in the annualized values. The primary benefit of the annualized values is the ease of comparison with current lease fees. As can be seen from Table 4-23, with the exception of the plantings under multiplot production using 10-mm seed, all other scenarios produce values much larger than the current fee of $40 (for two acres). A final analysis examined the impact of lower prices. Prices were uniformly (across all size categories) reduced 10, 20 and 30%. The evaluation was performed on Site 2, single plant production. Net present value and annualized value results are given in Table 4-25. As can be seen from the results, the price reduction eventually totally erodes the profitability of production using 10-mm seed. Production using 15-mm seed, however, still produces considerable returns, though net present value decreases with price reductions at a rate of two-to-one or greater.

PAGE 96

90 Table 4-25. Net present value (NPV) and annualized value (AV) of a two-acre lease at Site 2 under uniform price reduction. Price Seed density/seed size Reduction 62.5/10 62.5/15 75/10 75/15 10 % NPV -$9,100 $107,700 $28,400 $174,500 AV -800 9,000 2,400 14,700 20 % NPV -30,200 72,100 5,700 128,100 AV -2,500 6,100 500 10,800 30 % NPV -51,100 37,700 -17,500 82,900 AV -4,300 3,200 -1,500 7,000

PAGE 97

CHAPTER 5 SUMMARY AND CONCLUSIONS Summary The determination of optimal hard clam growout production design incorporating the selection of seed size, planting density, plant scheduling and replacement timing depends on consideration of the complicated interactions of monthly clam growth and price relationships. Optimal scheduling depends on the ability to predict future conditions of growth, mortality and price. Growth prediction for this research was accomplished through the development of a hard clam growth function modelling periodic growth as a function of initial clam size, age, and water temperature, salinity and dissolved oxygen. Growth function parameter estimates were made for two planting densities. Stochastic hard clam growth was simulated for two sizes of seed clam planted at two densities using environmental values from three sites in the Indian River area of Florida. Monthly mortality was imposed in a deterministic manner using a terminal base mortality and weight system such that the mortality of younger clams was greater than that of older clams. The results of the hard clam growth simulation were used to estimate expected net revenues by incorporating average monthly clam prices for four size categories. 91

PAGE 98

92 Clam prices were determined from time series price data. The expected revenues were then examined to determine optimal planting months and growout times. Two marketing arrangements were examined, the first allowing all clams to be planted and harvested as a single unit (single plant production), and the second requiring monthly harvests (multiplot production). Evaluations of the results produced specific plant, harvest and replacement schedules that varied with production method (seed size and density) and site location. Optimal rotations determined total operational input requirements and cost and revenue flows. These were then used to determine the residual value of the lease, the present value of the residual stream and an annualized value of the residual stream. Comparison of the economic performance of each production method at each site showed single plant production to always outperform multiplot production. Larger seed planted at the higher density outperformed all other options in all but one case. Single plant production allowed clams to be planted and harvested such that expected net returns were maximized. Multiplot production caused clams to be planted and/or harvested in months other than those indicated by less restrictive optimization. The higher returns to larger seed planted at greater densities were attributed to a shorter growout period generating faster turnover and higher volume sales. The one case where the lower density outperformed the higher density was attributed to higher opportunity costs resulting from faster equity build-up with the higher density.

PAGE 99

93 Conclusions and Implications The results of this study must be put in proper perspective. Various assumptions were required to construct a representative operation. Individual features are presented as neither concrete absolutes nor unrealistic options. They are reasonable assumptions as determined by scholarly research and through contacts with industry professionals. Actual values or requirements for individual producers are expected to be both higher and lower. The purpose of the research is not to determine precisely what can be accomplished, but rather to provide a foundation for establishing and directing further areas of emphasis and consideration. Where consideration has not been given to seed density, planting month, site location, etc., this research provides justification for such. It is in this context that the results should be viewed. The economic potential of hard clam growout is a complicated interaction of growth, as dictated by various environmental factors, and variable prices, as determined by clam size and month of harvest. Higher/lower densities may produce slower/faster growth while smaller/larger seed produce longer/shorter growout times. These only have economic relevance when combined with cost and revenue considerations. It is inaccurate and insufficient to say that a faster growth rate or a larger seed is preferred without knowledge of what effect the growth rate or seed size has on when clams reach market size and what prices might be expected. Further, while prices might be uniform across a region growth conditions likely vary from site to site making optimal operation design specific to a given site.

PAGE 100

94 Comparisons of the results of this study with others is difficult because of differences in production and financial assumptions. Thunberg and Adams (1990) estimate fifth-year net returns for two million seed (one million clams sold), a 3-year growout, and a $0 14 market price at $71,840. The comparable result from the current research is $17,800 to $24,000 for single plant production using 10-mm clams planted at 75 clams per square foot (900,000 clams planted and 602,000 to 623,000 sold) and sold at $0.10 per clam (see Tables 4-6 4-17, 4-18 and 4-19). Ignoring the clam volume differential of the two studies, the results by Thunberg and Adams include $40,000 attributable to the higher price and do not subtract taxes which would be in excess of $20,000 under the assumptions of the current study. A more direct comparison is possible with Adams et al. (1993), as they assume a 2-year growout and $0.10 per clam. Net returns from clams planted at 75 clams per square foot are $32,807 for 765,000 clams sold. The comparable results from this study are $47,000 for single plant production (756,000 to 760 000 clams sold) and approximately $42,000 for multiplot production (764,000 to 767 000 clams sold), both using 15-mm seed planted at 75 clams per square foot (see Tables 4-12, 4-17 4-18, 4-19, 4-20, 4-21 and 4-22). Thus, comparisons of this research with other studies indicate that returns can be increased by careful attention to production method and timing. The results of this study also indicate that, m the absence of mitigating circumstances, the grower should plant larger seed and at the highest recommended density. Fifteen-millimeter seed may not be available in all areas, or it may be available, but only at prohibitive prices. While this research did not conduct seed price sensitivity

PAGE 101

95 analysis, the magnitude of difference between 10-mm and 15-mm results would indicate that some upward leeway on larger seed prices still exists. Limitations on planting density will be dictated by site-specific environmental conditions. The grower must determine maximum planting densities for his particular site from either research recommendations or personal test trials. The residual values attributed to the lease computed in this study indicate that current lease fees substantially undervalue the productive capacity of the land. Given the complexity of the determination of residual values and their variability from site to site, it is unlikely that the full residual value could ever be recaptured as rent. The failure to properly value the land nevertheless represents a substantial revenue loss to Florida residents, the true owners of the land. Further the variability of residual values by site provides justification for variable lease fees based upon the productive capacity of each specific site. The identification of productively unique sites, however, remains a problem. The costs of such a determination may not justify the benefits. A uniform lease fee arrangement may be the preferred option. This research suggests, however, that the lease fee be greater than current requirements. The magnitude of the difference between existing lease fees and the residual values estimated by this research raise interesting policy implications. Current fees may be low due to ignorance of the true value of the land, or they may be set intentionally low so as to allow access by low income groups residing in coastal communities. If the intention were to allow access by low income groups, increased lease fees based on the productive capacity of the lease could reduce access by increasing start-up costs, thereby

PAGE 102

96 erecting a barrier to entry. This problem could be eliminated, however, and increased revenues captured through the imposition of a royalty or production tax. The culturist would pay a fee based upon the actual output of the operation. Start-up costs would remain the same, thus allowing continued access by original target groups, and the public would reap increased benefits from its ownership of the resource. The effects of the marketing constraint, as demonstrated by multiplot production, provide evidence of the practical management and economic implications of such a constraint. The results indicate that restrictive marketing requirements may have considerable impact on the economic performance of a hard clam growout operation as evidenced by Site 2 and Site 3 Market restrictions may also have little impact, as evidenced by Site 1. Actual marketing constraints might be less restrictive than those modelled, requiring less frequent sales or more restrictive, such as requiring both monthly sales and volume commitments. Less restrictive marketing requirements would be expected to produce returns approaching single plant production, while more res t rictive marketing requirements would be expected to further erode the profitability of hard clam culture. Nevertheless, in the event of such restrictions, a production option might be linkage with producers at other "environmentally unique" sites producing clams on different harvest cycles Through cooperative agreements, supply requirements might be met while allowing production to more closely follow less restrictive site specific optimal plant and replacement schedules. Although environment-related loss can be severe, the relatively low start-up costs and rapid turnover time from plant to market allow fairly rapid recovery from adverse

PAGE 103

97 events. Of greater concern to the industry is market stability. The clam market has seen prices decrease from $0.17 per littleneck clam in the 1980's (Adams et al., 1991) to $0.10 in 1993 (Adams et al., 1993). This research allowed price to vary by month as determined by historical data, with littleneck prices varying from $0.10 to $0.13 per clam. As recently as October, 1993, however, littleneck prices were $0.08 in the Indian River, Florida, area and were $0.08 to $0.10 in the Cedar Key, Florida, area (various personal communications). While current studies have been incapable of keeping abreast of the downward price movements the price sensitivity analysis in the previous chapter demonstrates the potential impact of reduced prices on the value of a lease. In the absence of increased market demand or supply disruptions in other productive regions, the influx of additional cultured supplies by Florida producers will likely exert additional downward price pressure. The hope is that these problems can be offset through increased marketing efforts directed towards increasing consumer awareness and demand. Recent developments with respect to legal size categories raise further questions about market stability. The current proposal before the Florida legislature would allow 5/8-inch cultured clams (approximately 31 mm long) to be sold within the state (marine Fisheries Commission, 1993). The impact of such a proposal is unclear and would depend upon how the marketplace reacts in terms of demand for both the new and existing legal size categories. Without question, Florida producers would be capable of producing the smaller clams in shorter growout periods than that required to produce larger clams. What is less clear, however, is whether prices would dictate that the smaller clams be produced. Current prices do not uniformly decrease with size.

PAGE 104

98 Topnecks occasionally command higher prices than littlenecks, depending upon the season and supply. Each size category serves a particular market niche. A specialty market targeting the smaller clams may develop such that higher prices are commanded. Also, demand for the smaller clams would likely absorb a portion of existing demand for larger clams. Markets will still exist, however, for larger clams. Price dynamics would determine the ultimate production mix, as in this research where it was occasionally indicated that topnecks, a larger clam, be produced rather than littlenecks, the smallest legal clam. Limitations and Suggestions for Further Research While the growth function developed for this research performed satisfactorily, additional work is required in further developing and refining the hard clam growth function. Of specific concern is the fact that in some instances larger clams (70-90 mm) exhibited 5 and 6-mm monthly growth spurts, a phenomenon not supported in literature (Eversole, 1987). While clams this large never factored into economic decisions due to harvest indications at smaller sizes, these growth spurts indicate a potential flaw in the growth function. While the flaw may or may not be relevant to the size stage or growth window of concern to culturists, its existence nevertheless begs attention. Thus, considerations of functional form and/or the incorporation of additional independent variables would be beneficial. The effects of current, tidal orientation, phytoplankton abundance, suspended solids, etc., warrant investigation. Also, a density variable would be useful so that a single growth function could be applied to multiple plant densities.

PAGE 105

99 Once environmentally-dependent growth relationships are defined, additional work is required in identifying environmental profiles that can be expected to produce similar growth, thus allowing classification of lease sites by growth and, hence, economic potential. This would allow the creation of variable lease fees, differing by economic potential. The development of a hard clam mortality function would allow stochastic mortality to be incorporated in a manner similar to growth, thus more realistically depicting stock changes. This was an intended component of this research, but an examination of the available data failed to identify any concrete relationships. While the impact of severe cold weather is unlikely in Florida, the results of this research indicated a bias against clam planting and growout in warmer months. The negative impact of warm weather planting and growout took the form of stagnant growth and increased mortality as determined by extended growout times. This is likely a simplistic summation of the environmental impact on mortality and more explicit linkage needs to be identified. Existing price data were insufficient to allow sophisticated incorporation of price expectations. This research was forced to use average monthly prices which were derived from relatively few field observations. Additional data would allow more accurate predictions. Certain risks exist for the prospective hard clam aquaculturist. This research assumed a specific familiarity with both the methods of clam culture and with marine environments as a whole. In the absence of such familiarity, a learning curve is expected

PAGE 106

100 such that mistakes will be made leading to decreased economic performance due to increased mortality, mishandling of clams, and misreading of economic signals. The results of this research were based on the assumption of stable expected environmental conditions. These conditions are long term phenomena and actual conditions may expose clams to less favorable growing environments. Of specific concern is exposure to extreme conditions such as might exist in the event of a storm or hurricane. The method of culture described provides protection from predation. Bag anchoring is a guard against mild current and tidal flow. Hurricane conditions, however, are capable of extensive clam mortality and equipment damage. The possibility of severe weather should be appreciated and incorporated into site selection. Further environmental threats exist from sewage contamination, excessive rain and run-off effects, natural toxic algal blooms, etc. Excessive rain creates multiple problems in that it reduces salinity levels and increases pollutant run-off into bays. The likelihood of adverse conditions will vary from location to location and should be considered when selecting a site. A final point of consideration deals with the issue of replanting undersized clams. It is not unusual for growers to periodically harvest legal clams and replant those clams that have yet to attain market size. This research did not allow replanting undersized clams. The production assumption was that all clams in a given plant be harvested simultaneously as a single unit. When the harvest of undersized clams was indicated, the undersized clams were discarded. Undersized clams were produced when specific conditions produced stunted or slow growing clams such that either a marketable size was

PAGE 107

101 not achievable for all clams under the time constraints established by the models or additional growout produced net losses from current market-sized. In most instances, replanting would likely not allow greater numbers of clams than those described in this research to flow through the system. Some undersized clams may be sufficiently slow growers or stunted as a result of a particular production method that market size is not attainable in any reasonable timeframe. Some production scenarios simulated by this research failed to produce marketable clams in 34 months. Further, the existence of replanted clams may impede replanting with new seed by monopolizing an already limited growing space. Also the added costs of culling and replanting may exceed any increased revenues. Culling and replanting undersize clams is a costlier, more time consuming process than simply harvesting and discarding. Replanting is similarly more time consuming than initial planting. The gains from eventual sales of undersized clams may, therefore, not justify the costs. It is unlikely, then, that replanting undersized clams would significantly improve the results of this research.

PAGE 108

APPENDIX 1 PRODUCTION AND ENVIRONMENT AL DATA USED IN GROWTH MODEL ESTIMATION

PAGE 109

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density JUN90 JUL88 23 47 47 0 5 35.96 27.56 80 AUG88 22 43 43 0 5 35.96 27.56 80 SEP88 21 40 41 1 5 35.96 27.56 80 OCT88 20 37 39 2 6 35.96 27.56 80 NOV88 19 36 38 2 6 35.96 27.56 80 DEC88 18 30 33 3 5 35.96 27.56 80 JAN89 17 32 33 1 6 35.96 27.56 80 FEB89 16 30 30 0 5 35.96 27.56 80 MAR89 15 29 30 1 6 35.96 27.56 80 APR89 14 27 29 2 3 35.96 27.56 80 MAY89 13 25 28 3 6 35.96 27.56 80 JUN89 12 22 25 3 4 35.96 27.56 80 JUL89 11 23 25 2 6 35.96 27.56 80 AUG89 10 21 25 4 6 35.96 27.56 80 SEP89 9 21 25 4 6 35.96 27.56 80 OCT89 8 26 27 1 6 35.96 27.56 80 NOV89 7 22 24 2 2 35.96 27.56 80

PAGE 110

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density DEC89 6 19 25 6 3 35.96 27.56 80 FEB90 4 19 20 1 3 35.96 27.56 80 MAR90 3 16 18 2 3 35.96 27.56 80 JUL90 JUL88 24 47 48 1 5 31.33 28.04 80 AUG88 23 43 45 2 6 31.33 28.04 80 SEP88 22 41 44 3 6 31.33 28.04 80 OCT88 21 39 39 0 5 31.33 28.04 80 NOV88 20 38 40 2 7 31.33 28.04 80 DEC88 19 33 35 2 6 31.33 28.04 80 JAN89 18 33 36 3 6 31.33 28.04 80 FEB89 17 30 34 4 7 31.33 28.04 80 MAR89 16 30 33 3 9 31.33 28.04 80 APR89 15 29 34 5 5 31.33 28.04 80 MAY89 14 28 34 6 4 31.33 28.04 80 JUN89 13 25 30 5 4 31.33 28.04 80 JUL89 12 25 30 5 7 31.33 28.04 80 AUG89 11 25 30 5 6 31.33 28.04 80

PAGE 111

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density SEP89 10 27 32 5 6 31.33 28.04 80 OCT89 9 25 26 1 6 31.33 28.04 80 NOV89 8 24 25 1 2 31.33 28.04 80 DEC89 7 25 26 1 3 31.33 28.04 80 FEB90 5 20 21 1 3 31.33 28.04 80 MAR90 4 18 20 2 3 31.33 28.04 80 JUN90 1 12 14 2 3 31.33 28.04 80 AUG90 JUL88 25 48 49 1 6 27.75 29.29 80 AUG88 24 45 46 1 6 27.75 29.29 80 SEP88 23 44 45 1 6 27.75 29.29 80 OCT88 22 39 39 0 5 27.75 29.29 80 NOV88 21 40 41 1 7 27.75 29.29 80 DEC88 20 35 37 2 6 27.75 29.29 80 JAN89 19 36 39 3 6 27.75 29.29 80 FEB89 18 34 36 2 6 27.75 29.29 80 MAR89 17 33 33 0 7 27.75 29.29 80 APR89 16 34 34 0 5 27.75 29.29 80

PAGE 112

Obs Plant Initial Date Date Age Size (mm) MAY89 15 34 JUN89 14 30 JUL89 13 30 AUG89 12 30 SEP89 11 32 OCT89 10 26 NOV89 9 25 DEC89 8 26 FEB90 6 21 MAR90 5 20 JUN90 2 14 JUL90 1 10 SEP90 JUL88 26 49 AUG88 25 46 SEP88 24 45 OCT88 23 39 NOV88 22 41 Final FS Size (mm) Growth STD 34 0 4 32 2 5 31 1 7 31 1 6 32 0 6 30 4 6 27 2 3 27 1 3 24 3 3 23 3 3 20 6 4 16 6 4 49 0 6 46 0 6 45 0 6 41 2 5 41 0 7 Salinity Temp 27.75 29.29 27.75 29.29 27.75 29.29 27.75 29.29 27.75 29.29 27.75 29.29 27.75 29.29 27.75 29.29 27.75 29.29 27.75 29.29 27.75 29.29 27.75 29.29 26.31 28.83 26.31 28.83 26.31 28.83 26.31 28.83 26.31 28.83 Dissolved Oxygen 4.75 4.75 4.75 4.75 4.75 Plant Density 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 ...... 0 O'I

PAGE 113

Obs Plant Initial Date Date Age Size (mm) DEC88 21 37 JAN89 20 39 FEB89 19 36 MAR89 18 33 APR89 17 34 MAY89 16 34 JUN89 15 32 JUL89 14 31 AUG89 13 31 SEP89 12 32 OCT89 11 30 NOV89 10 27 DEC89 9 27 FEB90 7 24 MAR90 6 23 JUN90 3 20 JUL90 2 16 Final FS Size (mm) Growth STD 38 1 6 39 0 6 37 1 6 33 0 7 34 0 5 34 0 4 32 0 5 31 0 7 31 0 6 32 0 6 30 0 6 28 1 3 27 0 3 25 1 3 24 1 3 22 2 4 18 2 4 Salinity Temp 26 31 28.83 26.31 28.83 26 31 28.83 26.31 28.83 26.31 28.83 26.31 28.83 26.31 28.83 26.31 28.83 26.31 28.83 26.31 28.83 26 31 28.83 26.31 28.83 26.31 28 83 26.31 28.83 26.31 28.83 26.31 28.83 26.31 28.83 Dissolved Oxygen 4.75 4.75 4.75 4 75 4 75 4 75 4.75 4 75 4.75 4.75 4 75 4 75 4.75 4.75 4.75 4.75 4.75 Plant Density 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 ,_. 0 -J

PAGE 114

Obs Plant Initial Date Date Age Size (mm) AUG90 1 10 OCT90 JUL88 27 49 AUG88 26 46 SEP88 25 44 OCT88 24 41 NOV88 23 41 DEC88 22 38 JAN89 21 39 FEB89 20 37 MAR89 19 33 APR89 18 34 MAY89 17 34 JUN89 16 34 JUL89 15 34 AUG89 14 31 SEP89 13 32 OCT89 12 30 Final FS Size (mm) Growth STF 14 4 4 49 0 6 49 3 6 49 5 6 43 2 5 43 2 7 42 4 6 42 3 6 37 0 6 33 0 7 36 2 5 36 2 4 36 2 5 36 2 7 31 0 6 32 0 6 33 3 6 Salinity Temp 26.31 28.83 22.32 26.11 22.32 26.11 22.32 26.11 22.32 26.11 22.32 26.11 22 32 26.11 22.32 26.11 22.32 26 11 22.32 26.11 22.32 26.11 22.32 26.11 22.32 26.11 22.32 26.11 22.32 26.11 22.32 26.11 22.32 26.11 Dissolved Oxygen 4.75 5.76 5.76 5.76 5.76 5 76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 5.76 Plant Density 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 ...... 0 00

PAGE 115

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density NOV89 11 28 30 2 3 22.32 26.11 5.76 80 DEC89 10 27 27 0 3 22.32 26.11 5.76 80 FEB90 8 24 25 1 3 22.32 26.11 5.76 80 MAR90 7 23 24 1 3 22 32 26.11 5.76 80 JUN90 4 20 22 2 4 22.32 26.11 5.76 80 JUL90 3 18 19 1 4 22.32 26.11 5.76 80 AUG90 2 14 15 1 3 22.32 26.11 5.76 80 SEP90 1 10 11 1 2 22 32 26.11 5.76 80 NOV90 JUL88 28 49 49 0 6 23.47 22.12 5.87 80 AUG88 27 49 49 0 6 23 47 22.12 5.87 80 SEP88 26 49 49 0 6 23.47 22 12 5.87 80 OCT88 25 43 45 2 5 23 47 22.12 5.87 80 NOV88 24 43 44 1 7 23.47 22.12 5 87 80 DEC88 23 42 43 1 6 23.47 22 12 5.87 80 JAN89 22 42 43 1 6 23.47 22.12 5.87 80 FEB89 21 37 39 2 6 23.47 22 12 5.87 80 MAR89 20 33 35 2 7 23.47 22.12 5.87 80 0 \0

PAGE 116

Obs Plant Initial Date Date Age Size (mm) APR89 19 36 MAY89 18 36 JUN89 17 36 JUL89 16 36 AUG89 15 31 SEP89 14 32 OCT89 13 33 NOV89 12 30 DEC89 11 27 FEB90 9 25 MAR90 8 24 JUN90 5 22 JUL90 4 19 AUG90 3 15 SEP90 2 11 OCT90 1 10 DEC90 JUL88 29 49 Final FS Size (mm) Growth STD 37 1 5 36 0 4 36 0 5 37 1 7 32 1 6 32 0 6 33 0 6 32 2 3 27 0 3 25 0 3 24 0 3 23 1 4 21 2 4 17 2 3 13 2 2 12 2 2 49 0 6 Salinity Temp 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 23.47 22.12 24.05 19.1 Dissolved Oxygen 5.87 5.87 5.87 5.87 5.87 5.87 5.87 5.87 5.87 5.87 5.87 5.87 5.87 5.87 5.87 5.87 7.18 Plant Density 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 ...... ...... 0

PAGE 117

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density AUG88 28 49 49 0 6 24.05 19.1 7.18 80 SEP88 27 49 49 0 6 24.05 19 1 7.18 80 OCT88 26 45 46 1 7 24.05 19.1 7.18 80 NOV88 25 44 46 2 8 24.05 19.1 7.18 80 DEC88 24 43 43 0 6 24.05 19.1 7.18 80 JAN89 23 43 43 0 6 24.05 19.1 7.18 80 FEB89 22 39 39 0 6 24.05 19.1 7 18 80 MAR89 21 35 37 2 8 24.05 19.1 7.18 80 APR89 20 37 37 0 5 24.05 19.1 7.18 80 MAY89 19 36 36 0 6 24.05 19.1 7.18 80 JUN89 18 36 36 0 5 24.05 19.1 7.18 80 JUL89 17 37 37 0 5 24.05 19.1 7.18 80 AUG89 16 32 32 0 6 24.05 19.1 7.18 80 SEP89 15 32 34 2 6 24.05 19.1 7.18 80 OCT89 14 33 35 2 6 24.05 19.1 7.18 80 NOV89 13 32 32 0 3 24.05 19.1 7.18 80 DEC89 12 27 27 0 3 24.05 19.1 7.18 80

PAGE 118

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density FEB90 10 25 25 0 3 24.05 19.1 7.18 80 MAR90 9 24 24 0 3 24.05 19.1 7.18 80 JUN90 6 23 24 1 4 24.05 19.1 7.18 80 JUL90 5 21 23 2 5 24.05 19.1 7.18 80 AUG90 4 17 20 3 7 24.05 19.1 7.18 80 SEP90 3 13 15 2 5 24.05 19.1 7.18 80 OCT90 2 12 15 3 5 24.05 19.1 7.18 80 NOV90 1 10 14 4 4 24.05 19.1 7.18 80 JAN91 AUG88 29 49 51 2 5 24.52 19.74 6.05 80 SEP88 28 49 49 0 6 24.52 19.74 6.05 80 OCT88 27 46 47 1 6 24.52 19.74 6.05 80 NOV88 26 46 47 1 6 24.52 19.74 6.05 80 DEC88 25 43 43 0 6 24.52 19.74 6.05 80 JAN89 24 43 44 1 6 24.52 19.74 6.05 80 FEB89 23 39 41 2 6 24.52 19.74 6.05 80 MAR89 22 37 39 2 5 24.52 19.74 6.05 80 APR89 21 37 37 0 5 24.52 19.74 6.05 80

PAGE 119

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density MAY89 20 36 37 1 5 24.52 19.74 6.05 80 JUN89 19 36 36 0 5 24.52 19.74 6.05 80 JUL89 18 37 38 1 6 24.52 19.74 6.05 80 AUG89 17 32 33 1 5 24.52 19.74 6.05 80 SEP89 16 34 36 2 6 24.52 19.74 6.05 80 OCT89 15 35 37 2 4 24.52 19.74 6.05 80 NOV89 14 32 34 2 4 24.52 19.74 6.05 80 DEC89 13 27 28 1 4 24.52 19.74 6.05 80 FEB90 11 25 27 2 4 24.52 19.74 6.05 80 MAR90 10 24 26 2 4 24.52 19.74 6.05 80 JUN90 7 24 26 2 4 24.52 19.74 6.05 80 JUL90 6 23 26 3 4 24.52 19.74 6.05 80 AUG90 5 20 20 0 3 24.52 19.74 6.05 80 SEP90 4 15 18 3 2 24.52 19.74 6.05 80 OCT90 3 15 16 1 3 24.52 19.74 6.05 80 NOV90 2 14 16 2 3 24.52 19.74 6.05 80 DEC90 1 10 12 2 2 24.52 19.74 6.05 80

PAGE 120

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density FEB91 AUG88 30 51 51 0 5 23.6 19.33 6.5 80 SEP88 29 49 49 0 6 23.6 19.33 6.5 80 OCT88 28 47 47 0 6 23.6 19.33 6.5 80 NOV88 27 47 47 0 6 23.6 19.33 6.5 80 DEC88 26 43 45 2 6 23.6 19.33 6.5 80 JAN89 25 44 44 0 6 23.6 19.33 6.5 80 FEB89 24 41 41 0 6 23.6 19.33 6.5 80 MAR89 23 40 42 2 5 23.6 19.33 6.5 80 APR89 22 37 37 0 5 23.6 19.33 6.5 80 MAY89 21 37 40 3 5 23.6 19.33 6.5 80 JUN89 20 36 40 4 5 23.6 19.33 6.5 80 JUL89 19 38 38 0 6 23.6 19.33 6.5 80 AUG89 18 33 36 3 5 23.6 19.33 6.5 80 SEP89 17 36 36 0 6 23.6 19.33 6.5 80 OCT89 16 37 38 1 4 23.6 19.33 6.5 80 NOV89 15 34 37 3 4 23.6 19.33 6.5 80 DEC89 14 27 28 1 4 23.6 19.33 6.5 80

PAGE 121

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density FEB90 12 27 29 2 4 23.6 19.33 6.5 80 MAR90 11 26 28 2 4 23.6 19.33 6.5 80 JUN90 8 26 30 4 4 23.6 19.33 6.5 80 JUL90 7 26 30 4 4 23.6 19.33 6.5 80 AUG90 6 20 25 5 3 23 6 19.33 6.5 80 SEP90 5 18 23 5 2 23.6 19.33 6.5 80 OCT90 4 16 22 6 3 23.6 19.33 6.5 80 NOV90 3 16 21 5 3 23.6 19.33 6.5 80 DEC90 2 12 18 6 2 23.6 19.33 6.5 80 JAN91 1 10 15 5 2 23.6 19.33 6.5 80 JAN91 1 10 15 5 2 23.6 19.33 6.5 60 MAR91 OCT88 29 47 48 1 6 26.4 19.95 6.3 80 NOV88 28 47 51 4 6 26.4 19.95 6.3 80 DEC88 27 45 47 2 6 26.4 19.95 6.3 80 JAN89 26 44 47 3 6 26.4 19.95 6.3 80 FEB89 25 41 42 1 5 26.4 19.95 6.3 80 MAR89 24 42 43 1 5 26.4 19.95 6.3 80

PAGE 122

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density APR89 23 37 38 1 5 26.4 19.95 6.3 80 MAY89 22 40 41 1 5 26.4 19.95 6.3 80 JUN89 21 40 41 1 6 26.4 19.95 6.3 80 JUL89 20 38 39 1 6 26.4 19 95 6.3 80 AUG89 19 36 40 4 6 26.4 19.95 6.3 80 SEP89 18 36 38 2 4 26.4 19.95 6.3 80 OCT89 17 38 39 1 4 26.4 19.95 6.3 80 NOV89 16 37 38 1 4 26.4 19.95 6.3 80 DEC89 15 28 31 3 4 26.4 19.95 6.3 80 FEB90 13 29 30 1 4 26.4 19.95 6 3 80 MAR90 12 28 30 2 3 26.4 19.95 6.3 80 JUN90 9 28 28 0 5 26.4 19.95 6.3 80 JUL90 8 28 28 0 4 26.4 19.95 6.3 80 AUG90 7 25 26 1 4 26.4 19.95 6.3 80 SEP90 6 23 25 2 4 26.4 19.95 6.3 80 OCT90 5 22 22 0 4 26.4 19.95 6.3 80 NOV90 4 21 21 0 4 26.4 19.95 6.3 80

PAGE 123

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density DEC90 3 18 22 4 3 26 4 19.95 6.3 80 JAN91 2 15 16 1 3 26.4 19.95 6.3 80 JAN91 2 15 16 1 3 26.4 19.95 6.3 60 APR91 OCT88 30 48 48 0 6 25 91 23.59 5.83 80 NOV88 29 51 51 0 6 25.91 23.59 5.83 80 DEC88 28 47 47 0 6 25.91 23.59 5.83 80 JAN89 27 47 47 0 6 25.91 23.59 5.83 80 FEB89 26 42 46 4 6 25.91 23.59 5.83 80 MAR89 25 43 43 0 5 25 91 23.59 5.83 80 APR89 24 38 42 4 5 25.91 23 59 5.83 80 MAY89 23 41 44 3 5 25.91 23.59 5.83 80 JUN89 22 41 43 2 5 25.91 23.59 5.83 80 JUL89 21 39 39 0 6 25 91 23.59 5.83 80 AUG89 20 40 40 0 6 25.91 23.59 5.83 80 SEP89 19 38 41 3 6 25.91 23.59 5.83 80 OCT89 18 39 39 0 4 25.91 23.59 5.83 80 NOV89 17 38 40 2 4 25.91 23.59 5.83 80

PAGE 124

Obs Plant Initial Date Date Age Size (mm) DEC89 16 31 FEB90 14 30 MAR90 13 30 JUN90 10 28 JUL90 9 28 AUG90 8 26 SEP90 7 25 OCT90 6 22 NOV90 5 21 DEC90 4 22 JAN91 3 16 JAN91 3 16 MAR91 1 10 MAY91 OCT88 31 48 DEC88 29 47 JAN89 28 47 FEB89 27 46 Final FS Size (mm) Growth STD 33 2 4 32 2 4 32 2 4 32 4 3 31 3 5 28 2 4 27 2 4 25 3 4 24 3 4 23 1 4 20 4 3 20 4 3 12 2 3 50 2 6 47 0 6 50 3 6 49 3 6 Salinity Temp 25.91 23.59 25.91 23.59 25.91 23.59 25.91 23.59 25.91 23.59 25.91 23.59 25.91 23 59 25.91 23.59 25.91 23.59 25.91 23.59 25.91 23.59 25.91 23.59 25.91 23.59 29.18 27.5 29.18 27.5 29.18 27.5 29.18 27.5 Dissolved Oxygen 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.83 5.14 5.14 5.14 5.14 Plant Density 80 80 80 80 80 80 80 80 80 80 80 60 60 80 80 80 80 ....... ....... 00

PAGE 125

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density MAR89 26 43 46 3 5 29.18 27.5 5.14 80 APR89 25 38 42 4 5 29.18 27.5 5.14 80 MAY89 24 41 44 3 5 29.18 27.5 5.14 80 JUN89 23 43 45 2 5 29.18 27.5 5.14 80 JUL89 22 39 42 3 6 29.18 27.5 5.14 80 AUG89 21 40 40 0 6 29.18 27.5 5.14 80 SEP89 20 41 44 3 6 29.18 27.5 5.14 80 OCT89 19 39 40 1 4 29.18 27.5 5.14 80 NOV89 18 40 40 0 4 29.18 27.5 5.14 80 DEC89 17 33 35 2 4 29.18 27.5 5.14 80 FEB90 15 32 34 2 4 29.18 27.5 5.14 80 MAR90 14 32 34 2 4 29 18 27.5 5.14 80 JUN90 11 32 36 4 3 29 18 27.5 5.14 80 JUL90 10 31 35 4 5 29.18 27.5 5.14 80 AUG90 9 28 30 2 4 29.18 27.5 5.14 80 SEP90 8 27 30 3 4 29.18 27.5 5.14 80 OCT90 7 25 28 3 4 29.18 27.5 5.14 80

PAGE 126

Obs Plant Initial Final Date Date Age Size (mm) Size (mm) NOV90 6 24 28 DEC90 5 23 25 JAN91 4 20 24 JAN91 4 20 22 MAR91 2 12 14 APR91 1 10 12 JUN91 OCT88 32 50 50 JAN89 29 50 50 FEB89 28 49 49 MAR89 27 46 46 APR89 26 42 42 MAY89 25 44 44 JUN89 24 45 45 JUL89 23 42 42 AUG89 22 40 41 SEP89 21 44 46 OCT89 20 40 41 FS Growth STD Salinity 4 4 29.18 2 4 29.18 4 3 29.18 2 3 29.18 2 3 29.18 2 3 29.18 0 6 25.94 0 6 25.94 0 6 25.94 0 5 25 94 0 5 25.94 0 5 25.94 0 5 25.94 0 6 25.94 1 6 25.94 2 5 25.94 1 4 25.94 Dissolved Temp Oxygen 27.5 5.14 27.5 5.14 27.5 5.14 27.5 5.14 27.5 5.14 27.5 5.14 28.08 1.78 28.08 1.78 28.08 1.78 28.08 1.78 28.08 1.78 28.08 1.78 28.08 1.78 28.08 1.78 28.08 1.78 28.08 1.78 28.08 1.78 Plant Density 80 80 80 60 60 60 80 80 80 80 80 80 80 80 80 80 80 ...... N 0

PAGE 127

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density NOV89 19 40 42 2 5 25.94 28.08 1.78 80 DEC89 18 35 37 2 4 25.94 28.08 1.78 80 FEB90 16 34 35 I 4 25.94 28.08 1.78 80 MAR90 15 34 35 I 4 25.94 28.08 1.78 80 JUN90 12 36 38 2 4 25.94 28.08 1.78 80 JUL90 11 35 35 0 5 25 94 28.08 1.78 80 AUG90 10 30 33 3 4 25.94 28.08 1.78 80 SEP90 9 30 33 3 4 25.94 28.08 1.78 80 OCT90 8 28 29 1 4 25.94 28.08 1.78 80 NOV90 7 28 31 3 4 25.94 28.08 1.78 80 DEC90 6 25 27 2 4 25.94 28.08 1.78 80 JAN91 5 24 25 1 3 25 94 28.08 1.78 80 JAN91 5 22 23 1 3 25 94 28.08 1.78 60 MAR91 3 14 16 2 3 25.94 28.08 1.78 60 APR91 2 12 14 2 3 25.94 28.08 1.78 60 MAY91 1 10 12 2 3 25.94 28.08 1.78 60 JUL91 OCT88 33 50 50 0 6 23.04 29 3.74 80 ...... N ......

PAGE 128

Obs Plant Initial Date Date Age Size (mm) JAN89 30 50 FEB89 29 49 MAR89 28 46 APR89 27 42 MAY89 26 44 JUN89 25 45 JUL89 24 42 AUG89 23 41 SEP89 22 46 OCT89 21 41 NOV89 20 42 DEC89 19 37 FEB90 17 35 MAR90 16 35 JUN90 13 38 JUL90 12 35 AUG90 11 33 Final FS Size (mm) Growth STD 50 0 6 49 0 6 46 0 6 42 0 5 44 0 6 45 0 5 42 0 5 41 0 4 46 0 4 41 0 6 42 0 4 37 0 4 35 0 4 35 0 4 38 0 5 35 0 4 33 0 4 Salinity Temp 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 23.04 29 Dissolved Oxygen 3.74 3.74 3.74 3.74 3 74 3.74 3.74 3.74 3.74 3.74 3.74 3.74 3.74 3.74 3.74 3.74 3.74 Plant Density 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 ..... N N

PAGE 129

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density SEP90 10 33 33 0 5 23.04 29 3.74 80 OCT90 9 29 29 0 4 23.04 29 3.74 80 NOV90 8 31 31 0 4 23.04 29 3.74 80 DEC90 7 27 29 2 4 23.04 29 3.74 80 JAN91 6 25 27 2 4 23.04 29 3.74 80 JAN91 6 23 25 2 4 23.04 29 3.74 60 MAR91 4 16 16 0 3 23.04 29 3 74 60 APR91 3 14 16 2 3 23.04 29 3.74 60 MAY91 2 12 15 3 2 23.04 29 3.74 60 JUN91 1 10 13 3 2 23.04 29 3.74 60 AUG91 MAY89 27 44 44 0 6 18.76 29.34 3.78 80 JUN89 26 42 46 4 7 18.76 29.34 3.78 80 JUL89 25 42 42 0 5 18.76 29.34 3.78 80 AUG89 24 41 41 0 4 18.76 29.34 3.78 80 SEP89 23 46 46 0 4 18.76 29.34 3.78 80 OCT89 22 41 41 0 6 18.76 29.34 3.78 80 NOV89 21 42 42 0 4 18.76 29.34 3.78 80

PAGE 130

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density DEC89 20 37 37 0 4 18.76 29.34 3.78 80 FEB90 18 35 35 0 4 18.76 29 34 3.78 80 MAR90 17 35 35 0 4 18.76 29.34 3.78 80 JUN90 14 38 38 0 5 18.76 29.34 3.78 80 JUL90 13 35 35 0 4 18.76 29.34 3.78 80 AUG90 12 33 33 0 4 18.76 29.34 3.78 80 SEP90 11 33 33 0 5 18.76 29.34 3.78 80 OCT90 10 29 29 0 4 18.76 29.34 3.78 80 NOV90 9 31 31 0 4 18.76 29.34 3.78 80 DEC90 8 29 29 0 4 18.76 29.34 3.78 80 JAN91 7 27 29 2 4 18.76 29.34 3.78 80 JAN91 7 23 25 2 4 18.76 29.34 3.78 60 MAR91 5 16 16 0 3 18.76 29.34 3.78 60 APR91 4 16 18 2 3 18.76 29.34 3.78 60 MAY91 3 15 16 1 2 18.76 29.34 3.78 60 JUN91 2 13 13 0 2 18 76 29.34 3.78 60 JUL91 1 10 12 2 2 18.76 29.34 3.78 60

PAGE 131

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density SEP91 JUN89 27 45 46 1 5 22.02 28.54 3.94 80 JUL89 26 42 43 1 4 22.02 28.54 3.94 80 AUG89 25 41 43 2 5 22.02 28.54 3.94 80 SEP89 24 46 46 0 5 22.02 28.54 3.94 80 OCT89 23 41 41 0 5 22.02 28.54 3.94 80 NOV89 22 42 43 1 3 22.02 28.54 3.94 80 DEC89 21 37 38 1 4 22.02 28.54 3.94 80 FEB90 19 35 37 2 5 22.02 28.54 3.94 80 MAR90 18 35 36 1 5 22.02 28.54 3.94 80 JUN90 15 38 39 1 4 22.02 28.54 3.94 80 JUL90 14 35 37 2 4 22.02 28.54 3.94 80 AUG90 13 33 33 0 3 22.02 28.54 3.94 80 SEP90 12 33 35 2 3 22.02 28.54 3.94 80 OCT90 11 29 30 1 3 22.02 28.54 3.94 80 NOV90 10 31 31 0 5 22.02 28.54 3.94 80 DEC90 9 29 29 0 3 22.02 28.54 3.94 80 JAN91 8 27 29 2 2 22.02 28.54 3.94 80

PAGE 132

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density JAN91 8 23 25 2 4 22.02 28.54 3.94 60 MAR91 6 16 16 0 3 22.02 28.54 3.94 60 APR91 5 18 19 1 3 22.02 28.54 3.94 60 MAY91 4 15 16 1 2 22.02 28.54 3.94 60 JUN91 3 13 14 1 2 22.02 28.54 3.94 60 JUL91 2 12 14 2 2 22.02 28.54 3.94 60 AUG91 1 10 12 2 2 22.02 28.54 3.94 60 OCT91 JUN89 28 46 46 0 4 18.25 25 4.90 80 JUL89 27 43 43 0 6 18.25 25 4.90 80 AUG89 26 43 43 0 4 18.25 25 4.90 80 SEP89 25 46 46 0 3 18.25 25 4.90 80 OCT89 24 41 42 1 6 18.25 25 4.90 80 NOV89 23 43 43 0 5 18.25 25 4.90 80 DEC89 22 38 39 1 4 18.25 25 4.90 80 FEB90 20 37 38 1 5 18.25 25 4.90 80 MAR90 19 36 37 1 5 18.25 25 4.90 80 JUN90 16 39 40 1 5 18.25 25 4.90 80

PAGE 133

Obs Plant Initial Final Date Date Age Size (mm) Size (mm) JUL90 15 37 38 AUG90 14 33 33 SEP90 13 35 35 OCT90 12 30 30 NOV90 11 31 31 DEC90 10 28 31 JAN91 9 29 29 JAN91 9 25 28 MAR91 7 16 17 APR91 6 19 19 MAY91 5 16 16 JUN91 4 14 16 JUL91 3 14 14 AUG91 2 12 13 SEP91 1 10 11 NOV91 JUN89 29 46 46 JUL89 28 43 43 FS Growth STD Salinity 1 4 18.25 0 3 18.25 0 3 18.25 0 3 18.25 0 4 18.25 3 4 18.25 0 3 18.25 3 4 18.25 1 3 18.25 0 3 18.25 0 2 18.25 2 2 18.25 0 2 18.25 1 2 18.25 1 2 18.25 0 4 21.43 0 6 21.43 Dissolved Temp Oxygen 25 4.90 25 4.90 25 4.90 25 4.90 25 4.90 25 4.90 25 4.90 25 4.90 25 4.90 25 4.90 25 4.90 25 4.90 25 4.90 25 4.90 25 4.90 19.74 6.49 19.74 6.49 Plant Density 80 80 80 80 80 80 80 60 60 60 60 60 60 60 60 80 80 ....... N -..J

PAGE 134

Obs Plant Initial Date Date Age Size (mm) AUG89 27 43 SEP89 26 46 OCT89 25 42 NOV89 24 43 DEC89 23 39 FEB90 21 38 MAR90 20 37 JUN90 17 40 JUL90 16 38 AUG90 15 33 SEP90 14 35 OCT90 13 30 NOV90 12 31 DEC90 11 31 JAN91 10 29 JAN91 10 28 MAR91 8 17 Final FS Size (mm) Growth STD 44 1 6 46 0 3 43 1 4 44 1 4 40 1 4 39 1 5 38 1 5 40 0 4 38 0 5 35 2 3 35 0 3 31 1 3 32 1 4 31 0 4 29 0 3 28 0 4 18 1 4 Dissolved Salinity Temp Oxygen 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 21.43 19.74 6.49 Plant Density 80 80 80 80 80 80 80 80 80 80 80 80 80 80 80 60 60 ..... N 00

PAGE 135

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density APR91 7 19 21 2 3 21.43 19.74 6.49 60 MAY91 6 16 18 2 4 21.43 19.74 6.49 60 JUN91 5 16 17 1 4 21.43 19.74 6.49 60 JUL91 4 14 16 2 3 21.43 19.74 6.49 60 AUG91 3 13 15 2 3 21.43 19.74 6.49 60 SEP91 2 11 13 2 2 21.43 19.74 6.49 60 OCT91 1 10 11 1 2 21.43 19.74 6.49 60 DEC91 JUN89 30 46 46 0 4 21.91 19.11 6.03 80 JUL89 29 43 43 0 6 21.91 19.11 6.03 80 AUG89 28 44 44 0 6 21.91 19.11 6.03 80 SEP89 27 46 46 0 3 21.91 19.11 6.03 80 OCT89 26 43 43 0 4 21.91 19.11 6.03 80 NOV89 25 44 45 1 4 21.91 19.11 6.03 80 DEC89 24 40 41 1 4 21.91 19.11 6.03 80 FEB90 22 39 39 0 5 21.91 19.11 6.03 80 MAR90 21 38 38 0 5 21.91 19.11 6.03 80 JUN90 18 40 40 0 4 21.91 19.11 6.03 80

PAGE 136

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density JUL90 17 38 38 0 5 21.91 19.11 6.03 80 AUG90 16 35 35 0 3 21.91 19.11 6.03 80 SEP90 15 35 35 0 3 21.91 19.11 6.03 80 OCT90 14 31 32 1 3 21.91 19.11 6.03 80 NOV90 13 32 32 0 4 21.91 19.11 6.03 80 DEC90 12 31 31 0 4 21.91 19.11 6.03 80 JAN91 11 29 29 0 3 21.91 19.11 6.03 80 JAN91 11 28 28 0 4 21.91 19.11 6.03 60 MAR91 9 18 19 1 4 21.91 19.11 6.03 60 APR91 8 21 22 1 3 21.91 19.11 6.03 60 MAY91 7 18 20 2 4 21.91 19.11 6.03 60 JUN91 6 17 18 1 4 21.91 19.11 6.03 60 JUL91 5 16 17 1 3 21.91 19.11 6.03 60 AUG91 4 15 17 2 3 21.91 19.11 6.03 60 SEP91 3 13 15 2 2 21.91 19.11 6.03 60 OCT91 2 11 12 1 2 21.91 19.11 6.03 60 NOV91 1 11 11 0 2 21.91 19.11 6.03 60

PAGE 137

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density JAN92 SEP89 28 46 46 0 3 21.38 16.14 6.39 80 OCT89 27 43 46 3 4 21.38 16.14 6.39 80 NOV89 26 45 46 1 4 21.38 16.14 6.39 80 DEC89 25 41 43 2 4 21.38 16.14 6.39 80 FEB90 23 40 42 2 5 21.38 16.14 6.39 80 MAR90 22 38 41 3 5 21.38 16.14 6.39 80 JUN90 19 40 43 3 4 21.38 16.14 6.39 80 JUL90 18 38 40 2 5 21.38 16.14 6.39 80 AUG90 17 35 40 5 5 21.38 16.14 6.39 80 SEP90 16 35 37 2 4 21.38 16.14 6.39 80 OCT90 15 32 37 5 4 21.38 16.14 6.39 80 NOV90 14 32 38 6 4 21.38 16.14 6.39 80 DEC90 13 31 36 5 4 21.38 16.14 6.39 80 JAN91 12 29 35 6 5 21.38 16.14 6.39 80 JAN91 12 28 34 6 4 21.38 16.14 6.39 60 MAR91 10 19 25 6 4 21.38 16.14 6.39 60 APR91 9 22 28 6 4 21.38 16.14 6.39 60

PAGE 138

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density MAY91 8 20 23 3 6 21.38 16.14 6.39 60 JUN91 7 18 23 5 4 21.38 16 14 6.39 60 JUL91 6 17 18 1 3 21.38 16.14 6.39 60 AUG91 5 17 20 3 3 21.38 16.14 6.39 60 SEP91 4 15 20 5 3 21.38 16.14 6.39 60 OCT91 3 12 15 3 2 21.38 16.14 6.39 60 NOV91 2 11 13 2 2 21.38 16.14 6.39 60 DEC91 1 10 11 1 2 21.38 16.14 6.39 60 FEB92 SEP89 29 46 48 2 3 22.33 18.53 5.5 80 OCT89 28 46 47 1 4 22.33 18.53 5.5 80 NOV89 27 46 47 1 4 22.33 18.53 5.5 80 DEC89 26 43 45 2 4 22.33 18.53 5.5 80 FEB90 24 42 45 3 4 22.33 18.53 5.5 80 MAR90 23 41 43 2 4 22.33 18.53 5.5 80 JUN90 20 43 47 4 4 22.33 18.53 5.5 80 JUL90 19 40 43 3 5 22.33 18.53 5.5 80 AUG90 18 40 43 3 3 22.33 18.53 5.5 80

PAGE 139

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density SEP90 17 37 42 5 5 22.33 18.53 5.5 80 OCT90 16 37 40 3 4 22.33 18.53 5.5 80 NOV90 15 38 40 2 5 22.33 18.53 5.5 80 DEC90 14 36 36 0 4 22.33 18.53 5.5 80 JAN91 13 35 38 3 2 22.33 18.53 5.5 80 JAN91 13 34 35 1 3 22.33 18.53 5 5 60 MAR91 11 25 30 5 5 22.33 18.53 5.5 60 APR91 10 28 32 4 4 22.33 18.53 5.5 60 MAY91 9 23 27 4 3 22.33 18.53 5.5 60 JUN91 8 23 28 5 4 22.33 18.53 5.5 60 JUL91 7 18 24 6 3 22.33 18.53 5.5 60 AUG91 6 20 23 3 3 22.33 18.53 5.5 60 SEP91 5 20 21 1 3 22.33 18.53 5.5 60 OCT91 4 15 20 5 3 22.33 18.53 5.5 60 NOV91 3 13 19 6 3 22.33 18.53 5.5 60 DEC91 2 11 13 2 2 22.33 18.53 5.5 60 JAN91 1 10 12 2 2 22.33 18.53 5.5 60

PAGE 140

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density MAR92 SEP89 30 48 48 0 3 21.42 19.72 5.7 80 OCT89 29 47 47 0 4 21.42 19 72 5.7 80 NOV89 28 47 47 0 4 21.42 19 72 5.7 80 DEC89 27 45 45 0 4 21.42 19.72 5.7 80 FEB90 25 45 45 0 4 21.42 19 72 5.7 80 MAR90 24 45 45 0 4 21.42 19.72 5.7 80 JUN90 21 47 48 1 5 21.42 19.72 5.7 80 JUL90 20 43 45 2 5 21.42 19 72 5.7 80 AUG90 19 43 44 1 3 21.42 19.72 5.7 80 SEP90 18 42 43 1 5 21.42 1 9 72 5.7 80 OCT90 17 40 41 1 3 21.42 19.72 5.7 80 NOV90 16 40 41 1 3 21.42 19.72 5.7 80 DEC90 15 36 38 2 3 21.42 19.72 5.7 80 JAN91 14 38 39 1 4 21.42 19.72 5 7 80 JAN91 14 35 37 2 3 21.42 19 72 5.7 60 MAR91 12 30 33 3 5 21.42 19.72 5.7 60 APR91 11 32 33 1 4 21.42 19.72 5.7 60

PAGE 141

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density MAY91 10 27 28 1 3 21.42 19.72 5.7 60 JUN91 9 28 30 2 4 21.42 19.72 5.7 60 JUL91 8 24 26 2 4 21.42 19.72 5.7 60 AUG91 7 23 24 1 3 21.42 19.72 5.7 60 SEP91 6 21 24 3 4 21.42 19.72 5.7 60 OCT91 5 20 23 3 3 21.42 19.72 5.7 60 NOV91 4 19 20 1 2 21.42 19 72 5.7 60 DEC91 3 13 14 1 2 21.42 19.72 5.7 60 JAN92 2 12 13 1 2 21.42 19.72 5.7 60 APR92 SEP89 31 48 48 0 3 24.55 21.45 5.91 80 OCT89 30 47 47 0 4 24.55 21.45 5.91 80 NOV89 29 47 47 0 4 24.55 21.45 5.91 80 DEC89 28 45 45 0 4 24.55 21.45 5.91 80 FEB90 26 45 45 0 4 24.55 21.45 5.91 80 MAR90 25 45 45 0 4 24.55 21.45 5.91 80 JUN90 22 48 48 0 5 24.55 21.45 5.91 80 JUL90 21 45 47 2 5 24.55 21.45 5.91 80

PAGE 142

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density AUG90 20 44 44 0 3 24.55 21.45 5.91 80 SEP90 19 43 44 1 5 24.55 21.45 5.91 80 OCT90 18 41 42 1 3 24.55 21.45 5.91 80 NOV90 17 41 42 1 3 24.55 21.45 5.91 80 DEC90 16 38 40 2 3 24.55 21.45 5.91 80 JAN91 15 39 40 1 4 24.55 21.45 5.91 80 JAN91 14 37 38 1 3 24.55 21.45 5.91 60 MAR91 14 33 35 2 5 24.55 21.45 5.91 60 APR91 12 33 34 1 4 24.55 21.45 5.91 60 MAY91 11 28 29 1 3 24.55 21.45 5.91 60 JUN91 10 30 31 1 4 24.55 21.45 5.91 60 JUL91 9 26 28 2 4 24.55 21.45 5.91 60 AUG91 8 24 24 0 3 24.55 21.45 5.91 60 SEP91 7 24 26 2 4 24.55 21.45 5.91 60 OCT91 6 23 25 2 3 24.55 21.45 5.91 60 NOV91 5 20 21 1 2 24.55 21.45 5.91 60 DEC91 4 14 15 1 2 24.55 21.45 5.91 60

PAGE 143

Obs Plant Initial Final PS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density JAN92 3 13 14 1 2 24.55 21.45 5.91 60 MAY92 SEP89 32 48 48 0 3 27.82 23.35 4.5 80 OCT89 31 47 47 0 4 27.82 23.35 4.5 80 NOV89 30 47 47 0 4 27.82 23.35 4.5 80 DEC89 29 45 45 0 4 27.82 23.35 4.5 80 FEB90 27 45 45 0 4 27.82 23.35 4.5 80 MAR90 26 43 43 0 4 27.82 23.35 4.5 80 JUN90 23 48 48 0 4 27.82 23.35 4.5 80 JUL90 22 47 47 0 4 27.82 23.35 4.5 80 AUG90 21 44 46 2 4 27.82 23.35 4.5 80 SEP90 20 44 46 2 4 27.82 23.35 4.5 80 OCT90 19 42 45 3 4 27.82 23.35 4.5 80 NOV90 18 42 44 2 4 27.82 23.35 4.5 80 DEC90 17 40 43 3 4 27.82 23.35 4.5 80 JAN91 16 40 43 3 4 27.82 23.35 4.5 80 JAN91 16 38 42 4 4 27.82 23.35 4.5 60 MAR91 14 35 36 1 4 27.82 23.35 4.5 60

PAGE 144

Obs Plant Initial Date Date Age Size (mm) APR91 13 34 MAY91 12 29 JUN91 11 31 JUL91 10 28 AUG91 9 24 SEP91 8 26 OCT91 7 25 NOV91 6 21 DEC91 5 15 JAN92 4 14 JUN92 SEP89 33 48 OCT89 32 47 NOV89 31 47 DEC89 30 45 FEB90 28 45 MAR90 27 43 JUN90 24 48 Final FS Size (mm) Growth STD 38 4 4 35 6 3 33 2 4 33 5 4 30 6 3 30 4 4 30 5 3 26 5 4 20 5 3 18 4 3 48 0 3 47 0 4 47 0 4 45 0 4 45 0 4 43 0 4 48 0 4 Salinity Temp 27.82 23.35 27.82 23.35 27.82 23.35 27.82 23.35 27.82 23.35 27.82 23.35 27.82 23.35 27.82 23.35 27.82 23.35 27.82 23.35 29.93 27.83 29.93 27.83 29.93 27.83 29.93 27.83 29.93 27.83 29.93 27.83 29.93 27.83 Dissolved Oxygen 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.5 4.83 4.83 4.83 4.83 4.83 4.83 4.83 Plant Density 60 60 60 60 60 60 60 60 60 60 80 80 80 80 80 80 80 ...... v,) 00

PAGE 145

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density JUL90 23 47 47 0 4 29.93 27.83 4.83 80 AUG90 22 44 46 2 4 29.93 27.83 4.83 80 SEP90 21 44 46 2 4 29.93 27.83 4.83 80 OCT90 20 42 45 3 4 29.93 27.83 4.83 80 NOV90 19 44 45 1 4 29.93 27.83 4.83 80 DEC90 18 43 45 2 4 29.93 27.83 4.83 80 JAN91 17 43 45 2 4 29.93 27.83 4.83 80 JAN91 17 42 44 2 4 29.93 27.83 4.83 60 MAR91 15 36 36 0 4 29.93 27.83 4.83 60 APR91 14 38 38 0 4 29.93 27.83 4.83 60 MAY91 13 35 35 0 3 29.93 27.83 4.83 60 JUN91 12 33 33 0 4 29.93 27.83 4.83 60 JUL91 11 33 33 0 4 29.93 27.83 4.83 60 AUG91 10 30 30 0 4 29.93 27.83 4.83 60 SEP91 9 30 30 0 3 29.93 27.83 4.83 60 OCT91 8 30 30 0 3 29.93 27.83 4.83 60 NOV91 7 26 28 2 4 29.93 27.83 4.83 60

PAGE 146

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density DEC91 6 20 22 2 3 29.93 27.83 4.83 60 JAN92 5 18 20 2 3 29.93 27.83 4.83 60 JUL92 SEP89 34 48 48 0 3 18.48 29.29 80 OCT89 33 47 47 0 4 18.48 29.29 80 NOV89 32 47 47 0 4 18.48 29.29 80 DEC89 31 45 45 0 4 18.48 29.29 80 FEB90 29 45 45 0 4 18.48 29.29 80 MAR90 28 43 43 0 4 18.48 29.29 80 JUN90 25 48 52 4 5 18.48 29.29 80 JUL90 24 47 52 5 5 18.48 29.29 80 AUG90 23 46 46 0 2 18.48 29.29 80 SEP90 22 46 46 0 3 18.48 29.29 80 OCT90 21 45 45 0 4 18.48 29.29 80 NOV90 20 45 45 0 4 18.48 29.29 80 DEC90 19 45 45 0 4 18.48 29.29 80 AN91 18 45 45 0 4 18.48 29.29 80 JAN91 18 44 44 0 4 18.48 29.29 60

PAGE 147

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density MAR91 16 36 36 0 4 18.48 29.29 60 APR91 15 38 38 0 4 18.48 29 29 60 MAY91 14 35 35 0 3 18.48 29.29 60 JUN91 13 33 34 1 4 18.48 29.29 60 JUL91 12 33 33 0 4 18.48 29.29 60 AUG91 11 30 30 0 4 18.48 29.29 60 SEP91 10 30 35 5 3 18.48 29.29 60 OCT91 9 30 32 2 3 18.48 29.29 60 NOV91 8 28 30 2 4 18.48 29.29 60 DEC91 7 22 22 0 3 18.48 29.29 60 JAN92 6 20 23 3 3 18.48 29.29 60 AUG92 SEP89 35 48 48 0 3 21.88 28.87 2.4 80 OCT89 34 47 47 0 4 21.88 28.87 2.4 80 NOV89 33 47 47 0 4 21.88 28.87 2.4 80 DEC89 32 45 45 0 4 21.88 28.87 2.4 80 FEB90 30 45 45 0 4 21.88 28.87 2.4 80 MAR90 29 43 43 0 4 21.88 28.87 2.4 80

PAGE 148

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density JUN90 26 52 52 0 5 21.88 28 87 2.4 80 JUL90 25 52 52 0 5 21.88 28 87 2.4 80 AUG90 24 46 46 0 3 21.88 28 87 2.4 80 SEP90 23 46 47 1 4 21.88 28 87 2.4 80 OCT90 22 45 46 1 4 21.88 28.87 2.4 80 NOV90 21 45 45 0 4 21.88 28.87 2.4 80 DEC90 20 45 45 0 4 21.88 28.87 2.4 80 JAN91 19 45 45 0 4 21.88 28.87 2.4 80 JAN91 19 44 45 1 4 21 88 28.87 2.4 60 MAR91 17 36 39 3 4 21 88 28.87 2.4 60 APR91 16 38 39 1 5 21.88 28.87 2.4 60 MAY91 15 35 38 3 5 21 88 28.87 2.4 60 JUN91 14 34 37 3 4 21.88 28 87 2.4 60 JUL91 13 33 34 1 4 21.88 28.87 2.4 60 AUG91 12 30 33 3 4 21.88 28.87 2.4 60 SEP91 11 35 36 1 3 21.88 28.87 2.4 60 OCT91 10 32 33 1 3 21 88 28.87 2.4 60

PAGE 149

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density NOV91 9 30 32 2 4 21.88 28.87 2 4 60 DEC91 8 22 26 4 3 21.88 28 87 2.4 60 JAN92 7 23 24 1 3 21.88 28 87 2.4 60 SEP92 SEP89 36 48 48 0 3 20.5 28 17 2.22 80 OCT89 35 47 47 0 4 20.5 28.17 2.22 80 NOV89 34 47 47 0 4 20.5 28 17 2.22 80 DEC89 33 45 45 0 4 20.5 28.17 2.22 80 FEB90 31 45 45 0 4 20.5 28.17 2.22 80 MAR90 30 43 43 0 4 20.5 28.17 2.22 80 JUL90 26 52 52 0 5 20.5 28.17 2.22 80 AUG90 25 46 46 0 3 20.5 28.17 2.22 80 SEP90 24 47 48 1 4 20.5 28.17 2.22 80 OCT90 23 46 47 1 4 20.5 28.17 2.22 80 NOV90 22 45 45 0 4 20.5 28.17 2.22 80 DEC90 21 45 45 0 4 20.5 28.17 2.22 80 JAN91 20 45 45 0 4 20.5 28.17 2.22 80 JAN91 20 45 45 0 4 20 5 28.17 2.22 60

PAGE 150

Obs Plant Initial Final PS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density MAR91 18 39 43 4 4 20.5 28.17 2.22 60 APR91 17 39 41 2 5 20.5 28.17 2.22 60 MAY91 16 38 41 3 5 20.5 28.17 2.22 60 JUN91 15 37 40 3 4 20.5 28 17 2.22 60 JUL91 14 34 35 1 4 20.5 28.17 2.22 60 AUG91 13 33 36 3 4 20.5 28.17 2.22 60 SEP91 12 37 37 0 4 20.5 28 17 2.22 60 OCT91 11 32 33 1 3 20.5 28.17 2.22 60 NOV91 10 32 34 2 4 20.5 28.17 2.22 60 DEC91 9 26 30 4 4 20.5 28.17 2.22 60 JAN92 8 24 26 2 3 20.5 28.17 2.22 60 OCT92 SEP89 37 48 48 0 3 18.96 21.65 4.40 80 OCT89 36 47 47 0 4 18.96 21.65 4.40 80 NOV89 35 47 47 0 4 18.96 21.65 4.40 80 DEC89 34 45 45 0 4 18.96 21.65 4.40 80 FEB90 32 45 45 0 4 18.96 21.65 4.40 80 MAR90 31 43 43 0 4 18.96 21.65 4.40 80

PAGE 151

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density JUL90 27 52 52 0 5 18.96 21.65 4.40 80 AUG90 26 46 48 2 3 18 96 21.65 4.40 80 SEP90 25 48 48 0 4 18.96 21.65 4.40 80 OCT90 24 47 48 1 2 18.96 21.65 4.40 80 NOV90 23 45 45 0 4 18.96 21.65 4.40 80 DEC90 22 45 45 0 4 18.96 21.65 4.40 80 JAN91 21 45 45 0 4 18.96 21.65 4.40 80 JAN91 21 45 45 0 4 18.96 21.65 4.40 60 MAR91 19 43 43 0 4 18 96 21.65 4.40 60 APR91 18 41 41 0 5 18.96 21.65 4.40 60 MAY91 17 41 41 0 5 18 96 21.65 4.40 60 JUN91 16 40 40 0 5 18.96 21.65 4.40 60 JUL91 15 35 35 0 5 18.96 21.65 4.40 60 AUG91 14 36 36 0 4 18.96 21.65 4.40 60 SEP91 13 37 37 0 4 18.96 21.65 4.40 60 OCT91 12 33 34 1 4 18.96 21.65 4.40 60 NOV91 11 34 34 0 4 18.96 21.65 4.40 60

PAGE 152

Obs Plant Initial Final FS Dissolved Plant Date Date Age Size (mm) Size (mm) Growth STD Salinity Temp Oxygen Density DEC91 10 30 30 0 4 18.96 21.65 4.40 60 JAN92 9 26 26 0 5 18.96 21.65 4.40 60

PAGE 153

APPENDIX 2 ENVIRONMENT AL VALUES USED IN GROWTH SIMULATION Temperature C) Site 1 Site 2 Site 3 Month Mean STD Mean STD Mean STD January 18.14 2.65 18.92 2.15 17.52 1.79 February 20.98 3.28 20.27 2.70 20.85 2.79 March 19.17 2.72 19.04 2.38 19.45 2.47 April 23.30 1.69 23.15 1.16 23.30 1.53 May 26.40 1.44 26.27 1.02 26.63 1.23 June 28.12 2.39 28.92 2.56 28.23 2.64 July 30.30 0.60 30.40 0.46 30.14 0.61 August 30.15 1.20 29.83 0.65 29.37 0.57 September 29.20 1.41 29.53 1.55 30.35 0.49 October 24.67 2.17 25.14 2.10 26.87 0.15 November 21.32 3.28 22.67 2.31 22.67 2.31 December 18.07 2.70 18.60 2.37 18.25 3.18 147

PAGE 154

148 Salinity (parts per thousand) Site 1 Site 2 Site 3 Month Mean STD Mean STD Mean STD January 27.70 2.33 28.50 2.04 26.25 2.06 February 26.75 3.18 26.00 4.11 25.05 2.81 March 27 20 3.94 25.80 3.70 26.12 3.17 April 26.42 2.08 26.42 2.20 26.17 2.34 May 26.50 2.38 25.37 1.06 26.33 2 08 June 26 12 1.93 25 95 2.42 26.67 1.53 July 24.25 6.40 23.83 5 78 23 60 5.81 August 20.25 8.84 20.17 5.80 20.07 5.59 September 25.50 2.78 25.33 1.53 26.50 3.54 October 24.82 2.69 24.76 2 74 25 23 3.22 November 25.25 3.23 25.33 0.58 24.83 0.29 December 25.17 4.75 23.62 4.96 24.50 6.36

PAGE 155

149 Dissolved Oxygen (milligrams per liter) Site I Site 2 Site 3 Month Mean STD Mean STD Mean STD January 7.92 1.59 8.20 1.36 8.73 1.72 February 7.33 0.75 7.52 1.09 7.43 1.56 March 7.40 0.68 7.20 0.43 7.20 0.94 April 7.77 3.03 7.85 3.46 8.15 3.56 May 6.75 0.66 7.25 0.63 7.93 1.10 June 6.53 0.69 6.65 0.87 6.57 1.37 July 5.98 1.23 5.82 1.55 6.24 1.13 August 7.45 1.34 6.67 1.90 6.80 2.98 September 6.87 1.01 6.60 0.36 7.40 0.28 October 7.35 1.42 7.56 1.48 7.07 0.59 November 6.80 0.43 6.73 0.81 7.10 0.10 December 8.93 1.37 8.20 0.29 7.65 0.35

PAGE 156

APPENDIX 3 HARD CLAM EXPECTED MEAN PRICES Littleneck Topneck Cherrystone Chowder Month 50-64 mm 64-77 mm 77-89 mm > 89 mm January $0.11 $0.12 $0.07 $0.05 February 0.12 0.13 0.07 0.05 March 0.11 0.12 0.07 0.04 April 0.12 0.12 0.07 0.04 May 0.12 0.11 0.07 0.05 June 0.11 0.12 0.08 0.04 July 0.11 0.10 0.08 0.05 August 0.10 0.12 0.08 0.05 September 0.13 0.12 0.07 0.05 October 0.10 0.12 0.07 0.06 November 0.10 0.13 0.07 0.07 December 0.12 0.13 0.08 0.06 150

PAGE 157

Equipment Bags Stakes Wet Suit Boat Motor Trailer Winch Truck Miscellaneous APPENDIX 4 INITIAL INVESTMENT REQUIREMENTS FOR HARD CLAM BOTTOM BAG GROWOUT Years of Life # Year 1 a 750/acre $6000 a 1500/acre 210 3 2 500 7 1 3000 2 1 3000 5 1 500 3 1 500 5 1 3000 5 1 700 Initial Site Survey 1 500 Total $17,910 Year 2 $6000 210 $6,210 a Bag life varies with the length of the hard clam growout period. If the growout period is two years or less, then bags last for two growouts or approximately 4 years. If the growout period is greater than two years, then the bags last for only one growout. 151

PAGE 158

APPENDIX 5 PRODUCTION COSTS FOR 2-ACRE HARD CLAM BOTTOM BAG GROWOUT Year 1 2 3 Variable Costs Seed a Supplies/ expendables 120 120 120 Fuel/oil Boat 900 1,100 1 100 Truck 300 300 300 Maintenance Boat/truck 1,000 1 000 1 000 Bags 65 130 130 Harvest Bags b Wages C Overhead Ex12ens e s Insurance 400 400 400 Permits 200 0 0 Bookkeeping/ Acct. Fee 500 500 500 Licenses 140 140 140 a Variable. Seed prices are $1.50 per 100 clams for 10 mm s ee d and $2.00 per 100 clams for 15-mm seed. Seed numbers vary with production method. b Variable. Harvest bags hold 250 clams and cost $0.20 each. c Variable. For single plant production $1500 per year if harvest occurs and $0 otherwise. For multiplot production, $0 in all years. 152

PAGE 159

REFERENCES Adams, Charles, James C. Cato, James E. Easley, Jr., Skip Kemp, William Mahan, John J. Manzi, Mike Oesterling, Robert Pomeroy, Eric Thunberg, David Vaughan, and Randal Walker. 1991. "Investing in Commercial Hard Clam Culture: A Comprehensive Guide to the South Atlantic States." Florida Sea Grant College Program Report# 104 (SGR-104). Adams, Charles, Stephen G. Holiman, and P. J. van Blokland. 1993. "Economic and Financial Considerations Regarding the Commercial Culture of Hard Clams in the Cedar Key Area of Florida." Staff Paper SP93-12. Food and Resource Economics Department, University of Florida, Gainesville, FL. Allen, P. G., L. W. Botsford, A. M. Schur, and W. E. Johnston. 1984. Bioeconomics of Aquaculture. Elsevier, Amsterdam. Askew, C. G. 1978. "A Generalized Growth and Mortality Model for Assessing the Economics of Bivalve Culture." Aquaculture 14:91-104. Bosch, Darrell, Leonard Shabman, and Geoffrey Knobl. 1989. "The Decline of Private Sector Oyster Culture in Virginia: Causes and Remedial Policies." Marine Resource Economics 6(3):227-243. Castagna, Michael. 1983. "Culture Methods for Growing the Clam Mercenaria mercenaria." Mems Assoc. Latinoam. Acuicult. 5(2):283-288. Castagna, Michael. 1986. "Clam Mariculture." In "An Overview of the Indian River Clamming Industry and the Indian River Lagoon," ed. Derek Busby. Florida Sea Grant Extension Technical Paper No. 44 Castagna, Michael, and John J. Manzi. 1989. "Clam Culture in North America: Hatchery Production of Nursery Stock Clams." In Clam Mariculture in North America, eds. J. J. Manzi and M. Castagna. Elsevier, Amsterdam. Crane, Donald M., Jr. 1979. "The Sugarcane Stubble Replacement Decision for South Florida." M.S. Thesis, University of Florida, Gainesville, FL. 153

PAGE 160

154 Dreyfus, Stuart E., and Averill M. Law. 1977. The Art and Theory of Dynamic Programming. Academic Press, New York, NY. Epifanio, C. E. 1979. "Growth in Bivalve Molluscs: Nutritional Effects of Two or More Species of Algae in Diets Fed to the American Oyster Crassostrea virginica (Gmelin) and the Hard Clam Mercenaria mercenaria (Linne)." Aquaculture 18: 1-11. Eversole, A. G. 1987. "Species Profiles: Life Histories and Environmental Requirements of Coastal Fishes and Invertebrates (South Atlantic)--Hard Clam." U.S. Fish and Wildlife Service Biological Report 82(11.75). U.S. Army Corps of Engineers, TR EL-82-4. 33 pp. Hadley, N. H., and J. J. Manzi. 1984. "Growth of Seed Clams, Mercenaria mercenaria, at Various Densities in a Commercial Scale Nursery System." Aquaculture 36:369-378. Hartman, Michael C. 1989. "Manual for the Design and Operation of a Low Budget Hatchery for the Hard Clam Mercenaria mercenaria in Florida." Florida Department of Agriculture and Consumer Services, Tallahassee, FL. Hillier, Frederick S., and Gerald J. Lieberman. 1986. Introduction to Operations Research. Holden-Day, Inc., Oakland, CA. Howard, Ronald A. 1981. Dynamic Probabilistic Systems. Volume 1: Markov Models. John Wiley and Sons, New York, NY. Loesch, Joseph G., and Dexter S. Haven. 1973. "Estimated Growth Functions and Size-Age Relationships of the Hard Clam, Mercenaria mercenaria, in the York River, Virginia." Veliger 16(1):76-81. Lough, R. Gregory. 1975. "A Reevaluation of the Combined Effects of Temperature and Salinity on Survival and Growth of Bivalve Larvae Using Surface Response Techniques." Fishery Bulletin 73(1):86-93. Maddala, G. S. 1983. Limited-Dependent and Qualitative Variables in Econometrics. Cambridge University Press, New York, NY. Maddala, G. S., and F. D. Nelson. 1975. "Switching Regression Models With Exogenous and Endogenous Switching." Proceedings of the American Statistical Association (Business Economics Section). pp. 423-6. Malouf, Robert E., and V. Monica Bricelj. 1989. "Comparative Biology of Clams: Environmental Tolerances, Feeding, and Growth." In Clam Mariculture in North America, eds. J. J. Manzi and M. Castagna. Elsevier, Amsterdam.

PAGE 161

155 Manzi, John J., and Michael Castagna. 1989. "Nursery Culture of Clams in North America." In Clam Mariculture in North America, eds. J. J. Manzi and M. Castagna. Elsevier, Amsterdam. Marine Fisheries Commission. 1993. "News Release, October 12." State of Florida, Tallahassee, FL. Menzel, Winston. 1989. "The Biology, Fishery and Culture of Quahog Clams, Mercenaria." In Clam Mariculture in North America, eds. J. J. Manzi and M. Castagna. Elsevier, Amsterdam. Perrin, R. K. 1972. "Asset Replacement Principles." Am. J. of Agric. Econ. 54:6067. Ryther, John H. 1986. "Biological and Environmental Factors Affecting the Clamming Industry." In "An Overview of the Indian River Clamming Industry and the Indian River Lagoon," ed. Derek Busby. Florida Sea Grant Extension Technical Paper No. 44. SAS Institute Inc. 1985. SAS User's Guide: Statistics. Version 5 Edition. SAS Institute Inc., Cary, NC. Shang, Yung C. 1981. Aquaculture Economics: Basic Concepts and Methods of Analysis. Westview Press, Boulder, CO. Thunberg, Eric M., and Charles M. Adams. 1990. "Evaluation of Aquaculture Investment: A Hard Clam Case Study." Staff Paper 389, Food and Resource Economics Department, University of Florida, Gainesville, FL. Van Heiningen, Judith N. 1992. Comparison oflnfluent and Effluent Water Quality for Land-Based Hard Clam Mercenaria mercenaria Nurseries. M.S. Thesis, University of Florida, Gainesville, FL. Vaughan, D. E., and R. LeRoy Cresswell. 1989. "Field Grow-out Techniques and Technology Transfer for the Hard Clam, Mercenaria mercenaria." Florida Department of Agriculture and Consumer Services, Tallahassee, FL. Vaughan, D., L. Cresswell, and M. Pardee. 1989. "A Manual for Farming the Hard Shell Clam in Florida." Florida Department of Agriculture and Consumer Services, Tallahassee, FL. White, Halbert. 1980. "A Heteroskedasticity-Consistent Covariance Matrix Estimator and a Direct Test for Heteroskedasticity." Econometrica 48(4):817-838.

PAGE 162

BIOGRAPHICAL SKETCH Stephen Glenn Holiman began his journey in life on February 11, 1955, in the back seat of the family car. This was the actual location of delivery. Debate continues as to whether this was also the location of conception. The car happened to be in Oakland, California, at the time and the local Naval hospital claimed responsibility for his delivery. Thus began life as a Navy brat with postings and relocations too numerous to mention. He enjoyed it, though, especially the three years in Hawaii, although he claims those years would have been much more memorable had they occurred after rather than before puberty. As he matured, he began to understand why the military moved people around so much: the kids were so much trouble that it was the only way to keep them one step ahead of the law. Stephen and his family eventually settled in Florida and Stephen attended the University of Florida where, despite being severely allergic to feathers, he majored in poultry science. He earned his bachelor's degree in 1979 and followed it up with a master's degree in 1980. Living in the same town for over four years, though, was almost more than he could bear so he decided to see the world via the Peace Corps and soon found himself beginning an incredibly satisfying journey that would keep him two 156

PAGE 163

157 and a half years in Malaysia and three and a half years in Thailand. The high point of his tour was his marriage to his lovely Thai wife, Sarinan. Having survived muggers, diarrhea, leeches and snakes (he especially enjoyed the cobra in the bathroom in Thailand), Stephen felt he was prepared to conquer anything, even a Ph.D. Six humbling years later, he refuses to answer the question of whether he'd do it all over again. Nevertheless, Stephen is now older, wiser and his life has been enriched by the birth of his now two-and-a-half-year-old son, Stephen Christopher Akaradej, and the forth-coming birth of child number two. And, he says, he really would do it all over again, even the Ph.D. All, that is, except for that one time in that seedy bar in Hadyai ...

PAGE 164

I certify that I have read this study and that in my op1mon it conforms to acceptable standards of scholarly presentation and is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy. ---7 ) l I j~ j -tjc~ J Thomas H. Spreen Chair Professor of Food and Resource Economics I certify that I have read this study and that in my op1mon it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. /I j~ [._ /Jr' ,J ~,,,--Eric M. Thunberg, Cochair Assistant Professor of Food and Resource Economics I certify that I have read this study and that in my acceptable standards of scholarly presentation and is fully ad~ as a dissertation for the degree of Doctor of Philosoph~ I opinion it conforms to ,....., .. <-1,1,".d quality, Charles M. Adams Associate Professor of Food and Resource Economics I certify that I have read this study and that in my opm1on it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy.~ vf4W I'll. Li~ Richard N. Weldon Assistant Professor of Food and Resource Economics

PAGE 165

I certify that I have read this study and that in my opm10n it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality as a dissertation for the degree of Doctor of Philosophy I ~ _1...u ;_'-..I,~ ,,._; ) ': .~\ .> ::.---' ~ William G. Boggess Professor of Food and Resource Economics I certify that I have read this study and that in my op1mon it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Charles Cichra Associate Professor of Forest Resources and Conservation This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate School and was acceptable as partial fulfillment of the requirements for the degree of Doctor of Philosophy. !1 December, 1993 JS tuk ;/ j' :j' Dean, College of Agriculture Dean, Graduate School