From what has been said up to now, it must be clear that an enhancement programme needs rigorous design, and requires not only in-depth knowledge about the life history of the species, but also an economic evaluation of the activity through the intertemporal flow of benefits and costs. While planning a restocking programme, the following topics should be considered:
Enhancement programmes need careful experimental design. Whatever the seeding technique selected, the appropriate scale of the experiment must be clearly defined according to the desired objectives (Sainsbury et al., 1997).
The wide variety of physical, environmental, biological, economic and social circumstances encountered in shellfish production requires that enhancement programmes should be site-specific. A good strategy is to design such experiments on relatively small spatial scales at first in order to allow a "hypothesis-falsifying" procedure to be followed (see McAllister and Peterman, 1992; Walters, 1997; Castilla and Defeo, 2001 and Section 3.5), incorporating control sites/replicates of e.g. selected spat densities.
In order to conduct enhancement experiments, each experimental unit must be adequately replicated in order that monitoring of growth, survival and production according to specific environmental and habitat characteristics, in such a way that estimates can be established within limits of statistical confidence. In this way, estimates of variability could also be used to evaluate the success of the experiment under uncertainty. Small-scale, replicated plots can be used to evaluate alternative scenarios (e.g. different stocking densities of the enhanced population) and effects of habitat quality (substratum, hydrodynamics). Basic ecological considerations, such as predator-prey interactions and the effects on the benthic community of massive transplantation/seeding of organisms, should be analysed before extrapolating results to a larger scale.
If adequately replicated, enhanced pilot scale sub-populations in experimental plots can be used to evaluate the success of mixed management strategies within an adaptive framework. For example, small areas could be closed to fishing or even subjected to different intensities of fishing in order to assess the potential benefits both of a rotational management scheme and restocking with seeded juveniles (Brand et al., 1991). If areas of similar productivity are considered, the experiment might be successful even on a short-term basis.
The increasing demand for seafood places emphasis on large-scale, commercially oriented, technology and intensive enhancement programmes. Thus, results obtained on an experimental, local or pilot scale, must then be evaluated at larger scales (see May, 1994 for useful concepts relating ecological questions and spatial scales). For example, a large-scale transplantation of spat must consider the ecological implications of such a perturbation on conspecific organisms (e.g. the effects of "genetic contamination" by interbreeding of hatchery stock, which might be less adapted to the environment, with local stocks), and on the benthic community as a whole (see Castilla, 1988; Bailly, 1991; Brand et al., 1991). Schiel (1993) gives one of the most useful examples of the experimental evaluation of commercial-scale enhancement of a shellfish population. He described a large-scale experiment in which growth/survival of seeded abalone Haliotis iris was assessed at a range of sites.
As in any ecological experiment, it is difficult to trace a rigorous sampling design in large-scale enhancement operations for the following reasons:
The varying nature of each site (habitat quality) precludes the definition of adequate control areas and replicates. In many cases areas are so large that replication is almost impossible.
Difficulties occur in filtering out the effects of a restocking programme from natural fluctuations of the stock or alternative management initiatives. Concomitant changes in fishing effort could also confound the results of the seeding process.
Stock enhancement is a long-term goal and must be evaluated accordingly. However, it is difficult to wait for a commercially harvestable size to evaluate success. Some projections, based on knowledge obtained from short-term experiments, might help overcome this obstacle.
The experimental design of enhancement exercises therefore requires careful attention to metapopulation dynamics and recruitment processes. In this vein, the existence of a metapopulation offers an opportunity to perform large-scale enhancement experiments in order to evaluate the capacity of the species to restock population subunits previously depleted by fishing or other disturbance (such as red tide outbreaks). Transplanting could be particularly useful where metapopulations have clearly defined "source" or "sink" characteristics (Shepherd and Brown, 1993). In order to conduct active enhancement of a shellfish metapopulation, and define harvest refuges serving as sources of individuals for replenishment or transplanting, information on the early stages of the life cycle is critical. Information on habitat quality or adult density alone is not enough to assure a higher probability of success, and Lipcius, Stockhausen and Eggleston (2001) discounted determining the site of the reserve by chance without information on transport processes of larvae.
Shepherd and Brown (1993) provided a preliminary example on how to apply metapopulation theory to South Australian abalone stocks; the first requirement being to define the complex of substock units. They integrated within this the concept of a "minimum viable population" in order to develop a cost-effective management framework for such a complex of stock units. This and other studies on metapopulations are used here to define a tentative guideline on how to apply metapopulation theory for the purposes of experimental enhancement of shellfish productivity. The main steps could be summarized as follows:
Define subareas by extension and number, according to the intrinsic characteristics of the resource and the fishery (i.e. by scale of aggregation of the resource and behaviour and subregional access rights of the fishers). Mapping the fishing grounds and stock abundance should precede the design of a system for acquiring information on the spatial dynamics of settlement over the long-term (see Caddy and Garcia, 1986). Subareas should be easy to identify for fishers and researchers, and should facilitate the collection of spatially accurate information (Cabrera and Defeo, 2001).
Estimate the times of settlement and recruitment. Identify potential sources of larvae and discern between source and sink areas. Care should be taken to evaluate larval connectivity between the discrete areas already defined. Monitor local recruitment of postlarvae and the degree of replenishment of different grounds. It is important to consider here the duration of the larval dispersal stage: those with a shorter larval development period may be more suitable candidates for hatchery rearing and subsequent seeding.
Identify key environmental variables, notably intensity and direction of currents that could explain prevailing larval dispersal and settlement. As each ground may be exposed to a distinct regime of primary production, nutrients, food availability, predation and disturbance, these between-ground differences should, if possible, be quantified.
Quantify spatial and temporal variations in density of recruits and adults (defined as sexually mature individuals), over a reasonable time frame and by site (e.g. source and sink areas). In each naturally seeded area, acquire information on resource users, local stock dynamics (growth and mortality). Accurate definition of spawning and recruitment timing is critical to providing fine-tuning of the appropriate timing for conducting stock enhancement programmes. Indeed, timing and durations of settlement of many species were often specific and quite short; especially in temperate latitudes (see e.g. Robinson and Tully, 2000). Thus, choosing the specific habitat and time of year for enhancement of a benthic species may be the key to success, and for most species, information on specific environmental requirements is generally lacking. Seasonal lows in abundance of previously established cohorts may represent the most suitable time for releasing juveniles in order to minimize inter- and intraspecific competition and predation, thus highlighting the value of careful ecological study of the potential release sites.
Compare growth and mortality information from fished and unfished grounds, in order to isolate effects of density and fishing intensity from those induced by environmental gradients in habitat quality. Growth and mortality patterns should be quantified through time (e.g. under different densities) and in space (e.g. by fishing grounds or LPs) in order to evaluate variations in density-dependent processes and habitat quality. Growth rates of transplanted/seeded individuals must be compared to those of the natural stock under different densities. Estimates of age-specific natural mortality (see Chapter 2) are particularly useful for detecting these critical periods when natural mortality from predation or other causes drops sharply from high values for spat to older animals. In order to have some idea of growth rates and development times from egg to mature adult, some information on environmental factors is critical, and it will be useful to keep time series of relevant water temperatures and wind conditions (e.g. Caddy, 1979c, Botsford, 2001). If development is protracted, and there are high rates of density-dependent mortality (e.g. cannibalism), then culture will be labour-intensive and economically prohibitive. Biometric relationships such as length vs. total weight/muscle weight should be determined to predict the expected meat yield from a mean individual size or age, thus allowing some economic projections for cultivation times.
If fishing rights are assigned geographically, quantify spatial variations in fishing intensity using e.g. a composite production modelling approach that includes simultaneous levels of production and fishing intensity from areas with variable intensities of harvesting but a similar basic ecosystem.
Estimate appropriate target and limit reference points (sensu Caddy and Mahon, 1995) for each LP. Complementary management strategies, such as catch quotas, number of fishers allowed to fish/area of ground, and minimum individual harvestable sizes, should also be agreed upon. A minimum viable population and optimal fishing mortality or harvest rate should be based on simple yield simulations from known growth and mortality rates, or empirically using the composite production modelling approach.
Depending on the inherent spatial characteristics of the metapopulation (see Shepherd and Brown, 1993) and prevailing fishing intensity, an effort should be made to identify existing spawning refugia and nursery areas (for motile species: Herrenkind et al., 1997). In essence, the refugia should be large enough to diminish the risk of stock collapse despite prolonged recruitment failure in LPs due to adverse environmental conditions. Spatio-temporal variability in abundance of stock, larvae and subsequent recruitment, as well as in the prevailing hydrographic regimes, should be considered when evaluating number and/or size of refugia or other spatially explicit management tool (e.g. MPAs). The dimensions of the area protected should be large enough for stock rebuilding purposes within and outside its boundaries.
Genetic factors must be taken into account. Classically, it has often been assumed that past enhancement programmes have been successful if populations appear to have been restored in their area of introduction. Testing allele frequency and mtDNA in hatchery stock, and comparing it with that from supposedly successful transplants may however paint a different picture. Thus, Burton and Tegner (2000) found that a red abalone population in California planted in 1979 which supposedly supported the fishery there during the 1980s, resembled other robust natural populations in the region in its genotypic frequencies, and showed no genetic signature of the broodstock used in the transplants. Although the test was not considered conclusive, it does not suggest discarding the previous generalization that hatchery outplants of abalone attempted to date appear to have been unsuccessful. One of the problems of cultivating shellfish for transplanting was illustrated by a genotyping of individual abalone larvae produced in a hatchery (Selvamani, Degnan and Degnan, 2001). Despite attempts to normalize the share of sperm from a number of males used to fertilize eggs in culture, DNA markers revealed that a single father fertilized almost all eggs reaching larval stage. This suggests the need for highly controlled breeding practices to ensure that the genetic diversity of the broodstock for out-planting to the field is maintained. Evidence from finfish culture has already warned of the dangers that repeated enhancement using a narrow genotype will adversely affect species resilience over the long term, and the same message evidently applies to abalone and other invertebrates produced in culture.
While it is easy to import a technology from elsewhere, in many cases enhancement programmes fail when technical problems substantially increase processing costs and lead to serious economic losses. As already mentioned, high costs of production of spat, and high predation on them when released onto the grounds, are critical factors. Other technical problems mentioned are environmental impacts due to seasonally extreme conditions (e.g. ice cover in high latitude waters, summer hypoxia in shallow tropical waters and lagoons, or heavy wave action) and processing constraints (the byssus of Mytilus edulis clogs the sorting and cleaning machinery). These kinds of technical problems may lead to a drop in production or compromise enhancement programmes (Kristensen and Hoffmann, 1991). Technical feasibility in rearing larvae, juveniles or adults may also constitute major bottlenecks in enhancement activities. Progress therefore requires a combination of technical applications in the methodology, and ecological acumen (see Peterson, Summerson and Luettich Jr., 1996 for a test of alternative transplantation techniques in scallops).
Facilitation of collection of sufficient numbers of larvae and juveniles from the wild for on-growing is of utmost importance: the timing of collector placement, and collector design, are essential to maximize seed collection during peak settlement periods. A key problem in enhancement programmes is the precise timing of release of juveniles or spat into the wild, in order to minimise mortality rates and costs. Release time should ideally be set after the critical stage in the life history has passed, where this is characterized for example by specialized diet or susceptibility to predation. In general, the longer the rearing time before release, the higher probability of survival. However, this increases production costs, so a trade-off between ecological and economic factors will need to be made in determining the optimum individual size for restocking (Tegner, 1989; Larkin, 1991).
The capacity to rear larval stages through to commercial size before releasing, and thus the appropriate duration of rearing for ranching is critical, i.e. whether the specimens are to be released in a recent post-settlement stage, or as adults, must be evaluated. Economic considerations are critical here, as well as ecological issues (competition, predation). For example, if some early stage is especially vulnerable to predation, it would be better to collect and release juveniles after this critical stage, particularly if natural mortality declines above a given size.
Another constraint may arise when shipping organisms to the transplanted sites. Schiel (1993) found that the greatest stress to transplanted abalone Haliotis iris occurred in packing and transport, and here the density per shipping tube needs to be carefully evaluated. The same author reported mortality rates of up to 47 percent in tubes where juveniles of Haliotis iris were packed at densities of 1000/tube. Many fragile organisms must be transported in aerated seawater and released at sea at well-defined experimental sites to assure the success of the operation. Peterson, Summerson and Luettich Jr. (1996) evaluated the success of alternative transplant methods for adult bay scallops Argopecten irradians, using five different sets of environmental conditions for a 6-h time of transfer from the source to the destination site. They found negligible mortality rates during travel times of up to four h with high flow speed and therefore high oxygen concentration, and this minimized the risks of stress and mortality from handling and transport.
Once in the wild, estimations of survival through recapture rates provide a way of monitoring success. Controlled release onto shellfish habitat, microtagging, and the development of a monitoring plan, represent three important methodological aspects directed at evaluating the technical feasibility of an enhancement programme. If microtagged, the precise location of recaptured animals is needed to evaluate the extent of local movement and the capacity of the seeded animals to restock target or adjacent areas (Addison and Bannister, 1994).
The economic significance of enhancement programmes has still not been fully evaluated, perhaps due to the difficulties in estimating total profits and costs derived from the seeding activity. One economic bottleneck derives from high hatchery costs: these programmes can be prohibitively expensive even for high unit-value resources (Addison and Bannister, 1994). One way to reduce total costs is by releasing juveniles at an early stage. However, this in turn could increase mortality rates after release, because of higher predation rates and susceptibility to environmental variations. This clearly constitutes a bioeconomic trade-off, and must be evaluated accordingly. The maximum cost that could be justified for evaluating new enhancement practices, should be in proportion to the expected benefits and impacts.
Hilborn (1998) reviewed the economic performance of nine marine stock enhancement projects for fish, turtle and lobsters involving restocking. He noted that no project evaluated showed clear evidence of a resulting increase in abundance as a result, but then few were planned in such a way that success criteria or economic performance could be evaluated. His suggestions were that systematic marking of released individuals would help establish survival and population enhancement, with explicit control areas incorporated in a proper statistical design, and subject to prior peer review by experts. The economics of stocking should be compared with other approaches such as habitat protection or improved management of the wild stock, and in this connection, evaluating the various benefits to the stock and ecosystem of protected areas requires close consideration (see e.g. Dixon and Sherman, 1991).
Bioeconomic analysis must be defined at the planning stage of the enhancement programme and must be specific to a local (among grounds) and regional (among countries) basis. Especially the former is crucial for benthic species with marked spatial variations in carrying capacity, recruitment, and growth and mortality rates, which constitute input variables affecting the economic analysis of stock enhancement. Moreover, some economic inputs might differ on a regional scale (e.g. opportunity costs of labour and capital) and thus economic analyses must not be overgeneralized. An analysis of marketing is also needed, because the choice of the species to be enhanced will be guided by demand/supply market laws. Different product types (whole weight, muscle weight), and the corresponding unit prices must be also included in the economic analysis, according to variations in the local/international demand.
In enhancement operations, relatively short sampling periods are used to estimate abundance, growth and survival rates through time (Schiel, 1993). Thus, economic projections should be employed to estimate the net present value (NPV) of the enhancement activity: abundance, growth and survival estimates derived from the short-term project must be extrapolated to the period at which organisms will be available to harvesting. An enhancement programme will be economically efficient if it maximizes the NPV of the yield obtained, which could be estimated as:
where TR and TC are respectively the total revenues and costs in time t, and d is the discount rate. Total revenues are obtained by multiplying the unit price of the products (e.g. whole weight, muscle weight, shells) by the estimated catch according to specific growth and survival rates. Total costs in each year are mainly based on costs of rearing individuals through a selected "seeding stage". The discount rate d considers the future value of the money invested.
An increasing discount rate diminishes the value of any future yield. Even though traditional values should approach an interest rate of ca. five percent, discount rates could generally takes higher values (up to ten percent), mainly as a result of uncertainty about future yields derived from the enhancement activity. As in fisheries, there are high uncertainty levels about changes in costs and prices, stock magnitude, growth and survival rates, and the prevailing economic and market regional situation. Therefore, d will tend to increase still further due to a probable expectation about falling prices or rapid depletion of the enhanced stock, implying that high exploitation rates in the short-term will be preferred over a long-term sustainable goal. This is particularly important in open-access regimes, in which free-rider behaviour and externalities commonly occur (Seijo, Defeo and Salas, 1998).
The high variability and difficulties in the estimation of some inputs (supply of spats, recapture, survival and growth rates and economic variables), together with a costly and low intensity of sampling through time, add uncertainty to the estimation of the NPV. Thus, different sources of uncertainty should be included when estimating the economic feasibility of the enhancement operation, e.g. different scenarios of growth and survival rates, and prices and costs could be used as inputs to estimate benefit and costs and the corresponding NPV of the enhancement plan. Moreover, different d values should be used to reflect dissimilar intertemporal preferences in resource use (e.g. different minimum harvestable sizes according to market demand).
After estimating uncertainty in the input variables, a criterion for choice among estimates is needed to provide options to a decision-making body. This could be done through decision theory. Decision analysis applied in fisheries (Hilborn, Pickitch and Francis, 1993; Hilborn and Peterman, 1996) considers alternative bioeconomic states of the fishery with the corresponding probabilities of occurrence, as a function of some possible policy actions. In this context, Bayesian inference allows the simultaneous consideration of multiple hypotheses and the integration of different types of information from many sources, reflecting scientific judgement as well as existing empirical data. Decision analysis could also be used to incorporate the above estimates of uncertainty into choices of enhancement actions. Data gathered in surveys conducted over experimental grounds on which enhancement programmes are taking place, as well as life history parameters derived from these data, could be used to provide a formal assessment of the enhanced stock, and in such cases the Bayesian approach is robust for estimating parameters, despite concerns over possible data outliers and mis-specification of priors (Millar, 2002; Myers et al., 2002).
In the above context, a decision table could be built on the basis of alternative enhancement actions and alternative hypotheses erected about parameter values and their corresponding probabilities of occurrence (Table 7.1). For example, high, medium and low levels of stocking densities ("alternative enhancement decisions") could be evaluated as a function of different hypotheses about resource performance (alternative scenarios of growth/survival rates; time needed to reach the minimum legal size, etc).
In some cases "experience may be insufficient for decision makers to be willing to attach numerical (cardinal) probabilities to the possible outcomes (states of nature)" (Schmid, 1989). Thus, decision tables could be created to account for different alternative system states (columns) and the possible decisions (rows), left with probabilities missing. The likelihood of outcomes could then be ranked only ordinally, and thus decision-makers could make a choice under uncertainty by expressing their subjective judgement about likelihoods in directional and qualitative terms. Schmid (1989) proposed three criteria for dealing with uncertainty and to guide decisions, without the need for explicit statements as to the probabilities of alternative parameter values: Maximin, Minimax and Maximax. These criteria vary according to degree of precaution. The Maximin criterion is a risk-averse approach that leads to selecting the maximum of all minimum outcomes. The Minimax Regret criterion is a less cautious approach that selects the minimum of the maximum regret, defined as the difference between the real benefit and the one that could have been obtained if the correct decision had been taken. Finally, the Maximax criterion is the most optimistic, in that it selects the highest outcome within the decision table. Once the table is built, the NPV of the activity is estimated for each combination of the enhancement actions and parameter values. These criteria were successfully applied in fisheries management (FAO, 1995; Pérez and Defeo, 1996; Seijo, Defeo and Salas, 1998; Defeo and Seijo, 1999) and could be easily adapted to enhancement problems. The reader must refer to the papers above-mentioned for a detailed application of these criteria.
Table 7.1 Key elements of a hypothetical decision table directed at evaluating alternative enhancement options. S1 is a hypothesis that implies a lower level of individual growth rate or a higher survival than S2 and S3. D1 to D3 represent alternative decisions concerning stocking densities. p values represent the probabilities of alternative hypotheses being true, and Oij represent the relative value of the outcome of a given stocking density i as applied to a given growth/survival rate j. Oij values could be regarded as representing net revenues obtained by each enhancement option. Finally, V1 to V3 represent the expected values of each action across all alternative hypotheses. A variance term might be added to each expected V value (after FAO, 1995; Hilborn and Peterman, 1996; Defeo and Seijo, 1999).
|
Alternative |
Alternative hypotheses about parameter |
Expected |
||
| |
H1 |
H2 |
H3 |
|
| |
P1 |
p2 |
p3 |
|
|
D1 (50 ind·m-2) |
O11 |
O21 |
O31 |
V1 |
|
D2 (100 ind·m-2) |
O12 |
O22 |
O32 |
V2 |
|
D3 (150 ind·m-2) |
O13 |
O23 |
O33 |
V3 |
Given the high uncertainty usually found in the majority of the parameters of an enhancement model, a precautionary approach must be considered suitable for evaluating the economic feasibility of the operation. Thus, not only the lower levels of the confidence intervals of the parameters should be used as inputs to estimates of the NPV, but also the criterion that gives a cautious approach (e.g. MaximIn Defeo and Seijo, 1999).
Enhancement practices have been applied to protect, maintain or improve shellfish populations. Because of the increasing number of enhancement programmes around the world, a scientific approach to evaluate their effectiveness in stock rebuilding is essential. However, the extent to which stock enhancement programmes contributed to natural populations of shellfishes has not been adequately assessed. Indeed, even though intuitively attractive, restocking programmes have been pursued with little evaluation of their success or failure (Addison and Bannister, 1994). Some reasons arise from: (a) the absence of biological knowledge of the species; (b) the lack of definition of clear objectives from the beginning of the planning stage; (c) experimental inadequacies resulting from an undefined methodological framework; and (d) technical problems associated with the supply, maintenance and rearing of spat (Cowx, 1994).
The following steps summarize the information provided earlier in this document and could be considered when assessing success of any enhancement plan:
(1) Determine the initial number and size structure of seeded organisms, together with the sites of placement. If possible, mark or differentiate them from wild animals. Use control, unseeded sites for comparative purposes. Characterize each site as precisely as possible.
In order to evaluate the success of any seeding experiment, seeded animals should be microtagged (Wickins, Beard and Jones, 1986), thus permitting the identification of hatchery-reared animals during subsequent field sampling and monitoring of commercial landings. It is commonly difficult to discern whether hatchery-reared animals have survived in addition to, or at the expense of, natural stock. However, Schiel (1993) suggested an effective and cheap means of tagging abalone indelibly by allowing abalones to feed sequentially on different algae before releasing. A continuous switching between algae generates alternating bands of different colours on the apex of the shell that can be seen for several years. At least this is applicable to abalone stocks.
A target density should ideally be estimated for seeding. It will be based on previously acquired knowledge about the SRR and carrying capacity of the system. A range of sites and, if possible, densities at each site, should be used to test hypotheses related to habitat quality and variations in the carrying capacity in each habitat. Data must be interpreted quantitatively in order to assess among-site variations in growth and survival rates and the success of active restocking. Unseeded sites should be useful controls for comparative purposes. Some measures of the effectiveness of the restocking process should also be quantified. Site-specific survival and individual growth rates of released animals, from the beginning of the seeding process to the time at which the individuals become available to harvesting, could be used for this purpose.
(2) Estimate abundance variations through successive and periodic sampling. Estimate survival and individual growth rates and compare them with those of the wild stock.
Length-frequency analysis should be clearly the best way to provide estimates of growth and survival rates. Overall growth and mortality patterns must be compared to those of unseeded sites and also among seeded sites. ANOVA procedures should be useful for this purpose.
As mentioned above, an enhancement plan is essentially long-term. However, it will be difficult to carry on sampling for years in order to estimate population dynamics features (growth, survival) until the harvestable size is reached. This is almost impossible for long-lived shellfish. Thus, projections of growth and mortality rates must be done from the seeding stage to the length at which organisms reach the minimum legal size.
(3) Estimate the number of microtagged organisms that survived to the harvestable size (biological samplings) and the relative contribution of the enhancement operation to the global landings from the whole area (by sampling landings and markets).
The success of stock enhancement programmes should be evaluated by field sampling (target fishing close to the release sites) and by monitoring fishery landings. Stock enhancement, if effective, can be detected from the concomitant increase of fishing yields reported by fishers’ logbooks. Concerning this important issue, Kitada, Taga and Kishino (1992, and references therein) reviewed four groups of methods for estimating of the effectiveness of enhancement programmes on the basis of tag recoveries, which can be summarized as follows:
(a) Estimation of total recoveries from fishers’ reports. Tag recoveries are intuitively attractive because of low costs of acquiring information (Crowe, Dobson and Lee, 2001). However, the proportion of recaptured animals tends to underestimate the survival rate of seeds and the consequent measure of effectiveness of the enhancement programme, for several reasons (Addison and Bannister, 1994):
A substantial part of the catch escapes monitoring.
Landings away from the release site, and thus with low probability of recapture, tend not to be monitored.
Estimates of abundance from tagging are based on some marked to unmarked ratio. However, dissimilar behaviour between marked and unmarked animals (survival rates of the former group tend to be lower), together with a generally low chance of recapture because of low percentages of marked animals, usually lead to underestimates of abundance (Hilborn and Walters, 1992).
These limitations could be mitigated by intensive sampling in the field and of landings (Kitada, Taga and Kishino, 1992). Bannister and Pawson (1991) showed that microtagged Homarus americanus in field samples at releasing sites constituted up to 50 percent of the catch on specific days and ten percent over a season, including egg-bearing females. This fact unambiguously shows that hatchery-reared animals survived to maturity and contributed to the breeding stock. However, scientific results concerning this point for most examples are usually either nonexistent or inconclusive.
(b) Correlation between annual number of fingerlings and the corresponding landing weight. This method could be an option for shellfish with short life spans and relatively stable recruitment rates. However, recruitment tends to be highly variable and not related to the amount of the parental stock but to show environmentally driven fluctuations in early life stages. Even should the above assumption be valid, it is difficult to discern between increasing landings as the response to the enhancement programme, or as a result of natural recruitment. The situation is complicated when several age-classes are contributing to spawning.
(c) Prediction of recoveries by calculating yield per release based on the catch equation and simulation models. This method is a complement to the others mentioned, because the recovery rate of shellfish released is not taken into account. Once this measure is quantified, simulation models could be performed to evaluate the effectiveness of restocking.
(d) Sampling surveys of commercial landings and fish markets. Kitada, Taga and Kishino (1992) suggested that a proper estimate of recovery could not always be obtained by these three groups of methods. They proposed a two-stage sampling survey of markets of cooperative associations of fishers (primary sample unit) and landing days (secondary sample unit) to estimate the success of enhancement programmes. Measures of effectiveness included the ratio of marked animals in the landings and recovery rates. These estimates were then used to evaluate the economic feasibility of the programme.
(4) Perform an economic analysis of the activity through the estimation of the net present value of the intertemporal flow of benefits and costs. Use different discount rates to reflect dissimilar intertemporal preferences of society in resource use. Identify some possible bottlenecks that might have to be mitigated in order to reduce costs.
As detailed earlier in this Chapter, the economic success of any restocking programme must be assessed to evaluate its commercial viability. To this end, costs (variable and fixed) and economic revenues must be carefully estimated in order to have indicators as to the feasibility of the operation. Simple spreadsheet methods incorporating life history parameters have commonly been used for calculating mortality and growth of fish populations (see e.g. Sparre and Venema, 1992) using for example the Thompson and Bell procedure (Ricker, 1975). This approach has been employed for modelling abalone populations (see Sanders and Beinssen, 1998 and De Waal and Cook, 2001), who have extended it to incorporate a cost-benefit analysis. The economic effectiveness of a seeding operation under different conditions of survival, growth, labour costs and product sale prices can be investigated. De Waal and Cook (2001) show that ranching shellfish is only likely to be economically viable where mortality is not excessive and survival rate increases with age, which of course is generally the case (Caddy, 1991, 1996).
(5) Estimate uncertainty in the main inputs of the enhancement model, i.e. from growth and survival rates to unit prices and costs. Employ for this purpose alternative hypotheses for parameter values to predict outcomes from alternative enhancement (e.g. stocking densities) strategies in a decision analysis.
Uncertainty and risk analyses must be conducted to evaluate the bioeconomic feasibility of a stock enhancement programme. For example, the profit from a stock enhancement programme for a flatfish (the example is valid also for shellfish), as estimated by Kitada, Taga and Kishino (1992) was US$ 63 000, but the 95 percent confidence interval ranged between unprofitable and profitable [- US$ 4 000 to US$ 151 000].
Given the high variability in outcomes, a precautionary approach should be used to minimize risks. Some specific Reference Points (Caddy and Mahon, 1995; FAO, 1995, 1996) should be used as targets. In this specific case, Reference Points are not necessarily those derived from classic surplus production and yield per recruit models, and conventionally used to manage fisheries (e.g. MSY, FMAX, etc). Such models assume that recruitment is constant and rarely include input for recruitment variability, which can be one of the main sources of variability in invertebrate populations (Conan, 1986; Caddy, 1989b). In this specific case, variable stocking densities should be included as an option.
(6) Try to reduce uncertainty in input variables by achieving as accurate biological and economic data as possible as a result of a rigorous experimental design. Focus research on improving the performance of different enhancement strategies. Develop methods for optimizing the monitoring system.
Post-stocking evaluation has been largely neglected in enhancement programmes (Cowx, 1994). An enhancement programme requires explicit specification of the information needed to achieve enhancement objectives, taking into account all the processes (e.g. growth, mortality, prices, and market demand) required to ensure that these needs are met. Periodic evaluation and revision of the data collection and the results achieved is necessary. This should aid in reducing uncertainty in key variables, which in turn will affect the NPV from the activity. The evaluation should assess the long-term benefits of alternative stocking practices and regimes, and attempt to identify those factors precluding enhancement success.
