Japanese Hatchery-based Stock Enhancement:
Lessons for the Chesapeake Bay Blue Crab

Implications for Chesapeake Bay Blue Crab Enhancement

In Japan and elsewhere, public and fisher perceptions often cause government agencies to place production demands over goals related to scientific evaluation of fisheries enhancement. For instance, consider a relatively simple experiment to evaluate enhancement by only releasing P. trituberculatus in alternate years. Given the rapid recruitment of hatchery-released P. trituberculatus to the fishery, this type of experiment should be effective in evaluating whether release years result in higher yields than non-release years. But consider a fisher or taxpayer's perspective: what is the benefit of not releasing crabs for half the years of a given period? The emphasis on production is also clearly seen in the initial goals put forth by the Japan Sea Ranching Program, which emphasize production of larvae and young juveniles and then sequentially move towards evaluation of hatchery contribution. This may seem like a logical series, but recall that most development of juvenile production for P. trituberculatus occurred over two decades ago. Rather than developing means with which to quantitatively evaluate hatchery releases, national and prefectural centers have instead emphasized research on mass production of other more difficult species such as squid, grouper, eel, and bluefin tuna. Thus, production science is substantially advanced, but scientific evaluations of hatchery-based enhancement remain underdeveloped. The jeopardy in not simultaneously considering the series of objectives necessary for hatchery enhancement is apparent in lack of evaluation of the stage at release for P. trituberculatus (i.e., C-1 vs. C-4) during the 1970s and 1980s, which arguably resulted in many years of wasted effort in trying to enhance Japanese crab fisheries. Such criticisms similarly apply to hatchery programs elsewhere in the world, where enhancement tactics and goals are incompletely considered despite large public investments (Solemdal et al. 1984; Secor and Houde 1998; Lichatowich 1999; Hilborn and Eggers 2000; Secor et al. 2000a).

In applying lessons from the Japanese case studies to hatchery-based enhancement in the Chesapeake Bay, the authors diverged in opinion on the feasibility of hatchery-based enhancement. In part, this was driven by the manner in which the disciplines of fisheries science, ecology, and aquaculture diverge on uncertainty in stock enhancement evaluation in Japan and elsewhere. D. Secor believes that sufficient information and perspective can be drawn from the Japan case studies to make the judgment that stock enhancement is not a desirable goal in the management of Chesapeake Bay blue crabs. A. Hines and A. Place believe that additional scientific investigations are needed to evaluate feasibility issues. Points of important agreement were reached on the merits of continued research on the artificial propagation of crabs, which will contribute valuable information on reproductive physiology, and through release experiments of juvenile crabs to support habitat and fishery management aims, as well as to provide a means to begin to evaluate the feasibility of stock enhancement.

Artificial Propagation and Chesapeake Bay Blue Crabs

D.H. Secor

Differences in the life cycle and relative productivity between Chesapeake Bay blue crab and Japanese P. trituberculatus are substantial and informative in evaluating the issue of hatchery enhancement. The yield and value of the Chesapeake Bay blue crab fishery is two to three orders of magnitude greater than those case studies developed for P. trituberculatus. Fishery independent studies on P. trituberculatus indicate relatively short distances (<30 km) between spawning, nursery, and feeding areas for a given P. trituberculatus stock (Ariyama 2000; Karakawa 2001). In a two-year mark-recapture study, Karakawa (2001) released 80 mm CW crabs into the Seto Inland Sea and subsequently recovered tagged crabs within 20 km of their release location. In contrast, female blue crabs that occur in the Maryland waters must undertake migrations that range between 110 km (mouth of the Potomac) and 300 km (head of the Bay) to marine waters to spawn. Similarly, juveniles that occur in the upper bay all originate from marine waters outside the Chesapeake Bay. Therefore, while hatchery contribution in Japan can be examined on a local scale (lagoon, embayment, segment of Inland Sea), the Chesapeake Bay must be considered on an ecosystem-wide basis. Given this essential difference in scale, one is forced to conclude that it is infeasible to supplement Chesapeake Bay's productive and large fishery.

For the sake of example, I utilized harvest recapture rates from the Japanese case studies and predicted how many C-4 crabs would be required to supplement by 10% a depressed yield of Chesapeake blue crabs (100 million crabs; c. 50% current level) (Table 2). Estimates ranged from 100 million to 1 billion juveniles! (The Osaka Bay estimate of harvest recapture rate was excluded since it was driven by extremely high harvest rate -92% per year -which was two times that predicted for the Chesapeake Bay -47% per year.) I also applied a simple model of early juvenile mortality (Miller 2001) together with a harvest model (equation 1) specific to Chesapeake Bay blue crab. Again, hundreds of millions to billions of released C-4 juveniles would be required to make only small contributions (10%) to a depressed Chesapeake stock. As a reminder, current levels of C-4 juvenile production for all of Japan, which leads the world in production of juveniles (involvement of 18 prefectural centers and 1 national station), is 28-42 million per year (Ariyama 2000). In summary, the manager of Japan's single largest P. trituberculatus hatchery (Tamano National Station) was correct in his reaction to the idea of hatchery enhancement in the Chesapeake Bay: "It is very difficult to go against such a large catch!" (K. Maruyama, Director, Tamano Station, Japan Sea-Farming Association, Okayama pers. comm.).

That Japan has carried on a stock enhancement program for more than 20 years without clear evidence of its effectiveness at first blush seems remarkable. On the other hand, there are few systems in the world for which stock enhancement has been shown definitively. Lack of evidence for stock enhancement in Japan and elsewhere lies with inadequate science, but of equal importance is the amount information and uncertainty attendant in demonstrating stock enhancement. Fishery enhancement (put-grow-and-take) can in large part be evaluated by identifying released individuals as they are captured in a fishery (Okamoto unpubl; see Measurement of Hatchery Contribution, p. 17). Stock enhancement requires that (1) the reproductive stock has been supplemented; (2) the offspring of released individuals effectively supplement the next generation of crabs; (3) that the addition of individuals to the reproductive stock and subsequent generations are in fact supplementing the wild population rather than replacing it; and (4) negative genetic and ecological interactions between released and wild individuals are minimal. The last two issues have created much controversy for Pacific salmon, stemming from uncertainty inherent in trying to evaluate stock enhancement by comparing release and harvest numbers over generations (Hilborn and Eggers 2000). As we have discussed above for the Japanese case studies, one cannot disentangle hatchery supplementation from changes in habitat productivity, ecological and genetic interactions, and changes in exploitation patterns without substantial ancillary information.

To establish stock enhancement, one would need to follow the genealogy of released crabs and conduct broad based ecosystem-level studies on the interactions between hatchery and wild crabs and the capacity of the current ecosystem to support additional crabs (e.g., habitat productivity and predator-prey interactions). Thus, enhancement must be investigated on a system-wide basis. Given the complexity and uncertainty in establishing stock enhancement, it is no wonder that comprehensive programs of scientific evaluation are rare. Indeed, the costs of demonstrating stock enhancement of blue crabs in the Chesapeake Bay would be prohibitive, beyond the scope of any of the current programs that monitor living resources in Chesapeake Bay.

There are typically two issues that hatchery enhancement proposes to redress: (1) too few fish (i.e., overfishing, climatic productivity cycles), or (2) too little habitat. In either case, one should be careful to evaluate the extent of either problem and determine whether hatchery enhancement is an efficacious approach. In the case of blue crabs in the Chesapeake Bay, if overfishing is driving lower abundances and catches, then it would seem that addressing the source of the problem by reducing fishing would be most efficacious. If habitat is the problem, then hatcheries could hypothetically serve to bypass critical habitats, but here again we are not addressing the problem with a direct solution (i.e., habitat restoration). Rather we are proposing perpetual hatchery stocking as a surrogate for lost habitat (Lichatowich 1999). Thus, hatchery enhancement as a solution engenders considerable complexity and uncertainty because it is an indirect intervention, one that does not directly address root causes.

While there can be no imperative for hatchery enhancement of Chesapeake Bay blue crabs, there is a second more reasonable question -is there a role for artificial propagation of C. sapidus? The answer, informed by our Japanese experience, is a resounding yes. In Japan, the releases of tagged P. trituberculatus crabs have provided valuable information on growth and mortality rates, habitat requirements, migration patterns and local abundance of crabs that would not have otherwise been attainable. Propagation techniques are well developed for P. trituberculatus and are seeing rapid development for Chesapeake Bay blue crabs (A. Place pers. comm.). Thus, it would be quite feasible to use released crabs as environmental probes by releasing different stage crabs into varying habitats and conditions in the Chesapeake Bay (e.g. Secor and Houde 1995; Secor et al. 2000b). Such experimental releases of blue crabs could test hypotheses related to the importance of the littoral zone and submerged aquatic vegetation as nursery habitat. Released crabs could provide improved estimates of vital rates and test the efficiency of fishing and sampling gears and thereby support better stock assessments and fishery regulations. Much remains unknown about the reproductive biology and early life history of blue crabs that could become understood through experimental rearing and related laboratory-based studies. Vital attributes such as fecundity, spawning behavior and frequency, growth rate, molt increment, longevity, and juvenile energetic requirements and behaviors are better known in P. trituberculatus than in C. sapidus, due in part to artificial propagation (Ariyama 2000; Hamasaki 2000).

It is not surprising that the triumvirate of impacts to fisheries -loss of habitat, climate, and exploitation -prevail in Japan just as they do in the Chesapeake Bay. Of particular concern in Japan is the loss of shoreline habitat due to continued seawall construction and shoreline modifications. This was especially apparent in Osaka Bay (Figure 15), where scores of kilometers of sandy littoral beach habitat, considered critical nursery habitat for P. trituberculatus, have been converted to rocky inter-tidal environments and seawalls to curtail shoreline erosion adjacent to urban and industrial centers. It seems unlikely that hatchery releases of P. trituberculatus will keep pace with rates of nursery habitat loss and high exploitation: an important lesson for the Chesapeake Bay?

Lessons from Japanese Hatchery-based Stocking of Portunid Crabs

A.H. Hines and A.R. Place

The Japanese hatchery-based stocking program for portunid crabs has devoted a great deal of effort to development of juvenile production, while proportionately very little effort has been allocated to evaluating the effectiveness of releasing those juveniles for stock enhancement. Comparison of the limited studies of enhancement effects in the subsystems described here provide useful insights into the effect of system scale. They also indicate the difficulties of working with a species that has a complex life cycle and that is difficult to tag. Since the studies of enhancement effects are certainly not definitive, extensive, long-term, or even conducted with the same methods, their use is more relevant for guiding additional research than for proscribing success or failure of stock enhancement in other systems. In addition, the Japanese process of hatchery releases is designed primarily for direct support of the fishery catch, and not for enhancing the spawning stock. Still, we should cautiously consider the major lessons. Clearly, there is obvious need to have research assessing enhancement activities with just as much effort and funding for assessing effectiveness as for production. Just as clearly, there is enormous need of good laboratory experimental systems for basic research and fishery management of the Callinectes sapidus population that has been so productive historically and is in such sharp decline currently. This research needs to be coupled with immediate and sustained improved fishery management and habitat restoration.

While the Seto Inland Sea does not appear to be as productive for Portunus trituberculatus as Chesapeake Bay is for Callinectes sapidus, the Japanese system has a much different bathymetry and lacks the extensive shallow water nurseries that characterize Chesapeake Bay. Most other estuarine systems of the U.S. that have C. sapidus fisheries also lack the historical productivity of Chesapeake Bay. Adjoining Delaware Bay produces about 15% of Chesapeake Bay landings and the sound system of North Carolina currently produces about 80% of the Chesapeake (as the catch in the Bay has declined to new lows and that in North Carolina has increased). The reasons for the high Chesapeake productivity or the limits to production for the other systems are not understood. Perhaps the question should not be why doesn't stocking cause the Japanese system to produce like Chesapeake Bay, but rather is the stocking system adequate to sustain or enhance fishery production in Japan, and what factors limit fishery production more generally?

While their stock enhancement effects are poorly assessed, Japanese hatchery techniques demonstrate that several species of portunid crabs can be reared in large cultures to early juvenile instars, giving credibility to the idea that this can be done with Callinectes sapidus. The Japanese array of hatcheries has successfully concentrated on Portunus trituberculatus as one of its focal species for massive production of juveniles. However, these hatcheries depend on a diverse array of multiple species that are raised sequentially over a staggered series of reproductive seasons to sustain year-round activity. They do not specialize in only one species. The rearing methods of various hatcheries all had obvious similarities, but also differed significantly in their methods of production and use of larval/juvenile food -a major component of the rearing process. They also differed substantially in their methods for "secondary rearing" of the C-1 to C-4 instars. Hatcheries also all seemed to have a significant problem in transferring early crab instars between systems (hatchery to secondary rearing, secondary rearing to release site). Hatchery collection methods of concentrating juvenile crabs in Japan are harsh (millions of early instar juveniles in tanks on the order of 200 m3 are concentrated into 1 m3 transport vessels) and produce high rates of damage and limb autonomy that may reduce growth rates and survival of released crabs. Improving techniques for this step could improve production and post-release survival.

Both P. trituberculatus and C. sapidus have complex life cycles in which their larvae are exported away from the adult population and into the offshore plankton where genetic stocks of adjoining subsystems are mixed. After settlement, juveniles above a critical size become benthic and remain within subsystems. Japanese research showed that it is important to grow crabs to release at a size when they are benthic and not very mobile, so that they are less vulnerable to predators than the earliest stages that are planktonic or frequently swim into the water column. For P. triberculatus, this is C-4 - C-5 at about 20 mm CW. This size coincides with 20 mm CW C. sapidus that leave the SAV beds of the lower bay and settle into the subestuaries to feed and grow to maturity (Hines et al. 1987, Pile et al. 1996, Moksnes et al. 1997 ), except that it takes about C-7 to C-9 to attain this size in C.sapidus ( Pile et al. 1996, Moksnes et al. 1997). Once in the subestuaries, juvenile C. sapidus seem to stay within the systems and grow to maturity (Hines et al. 1987, Hines et al. 1995). The major source of mortality for the juvenile C. sapidus in the upper bay subestuaries like Rhode River seems to be cannibalism by large crabs (Hines and Ruiz 1995), and with the reduced levels of the population of adults, juveniles currently suffer very little mortality (based on 12+ years of tethering studies). So release of 20 mm juveniles in these systems might enjoy substantially higher survival than P. trituberculatus juveniles in Japan, thus possibly reducing the levels of juvenile stocking needed to have an equivalent effect. Further, initial results of recent experiments in the North Carolina Sounds indicate that stocking of C-1 - C-2 juveniles was successful at significantly increasing abundances in small embayments for at least several weeks (D. Eggleston, North Carolina State University, pers. comm.)

After growing to maturity and mating in the subsystems, both crab species then become mobile again and move to utilize other habits for egg production and release of dispersive larvae. Upon attaining maturity, P. trituberculatus migrates off shore, in an analogous fashion to C. sapidus migrating to the mouths of estuaries. The scale/distance of the migration depends on the system for both species. C. sapidus completes its life cycle by migrating 200+ km distances from the upper portion of big Chesapeake Bay but over much shorter distances throughout the rest of Chesapeake Bay and in other estuaries. The ovigerous females of both species can be found within portions of the juvenile habitat, as well as moving into separate habitat that favors release and survival of larvae. The life cycles are thus similar in that they have motile larval and early post-settlement stages and motile adult stages, while the juvenile instars remain fairly sedentary within subsystems where they grow to maturity and mate. While P. trituberculatus is a coastal species, C. sapidus utilizes estuaries for its benthic life stages.

A stocking program in Chesapeake Bay would/should short-circuit the dispersal and high mortality stages of larval development and early post settlement to release 20 mm CW crabs that remain within subestuarine systems until they grow to maturity. Once reaching maturity and mating, the large C. sapidus encounter the intense pot fishery as they become much more mobile and move out of the subestuaries. To promote transfer to the spawning stock, the spatial and temporal aspects of the pot fishery would have to be managed to allow corridors of migratory movement to the spawning grounds. Otherwise, it would become a put-and-take fishery.

The scale of the system and purpose of the stocking program is important. While incredible numbers (say 100 million or more) of juvenile Callinectes sapidus might be needed to enhance the entire Chesapeake Bay fishery, smaller estuaries or Chesapeake subestuaries might be enhanced with many fewer crabs. At the lower extreme of scale, for example, in the small Rhode River subestuary (a 585 ha embayment of the upper western shore), fishery-independent trawling and observations of fishing effort can estimate that the standing stock of fishery-legal crabs is now only ca. 2.2 x 103 crabs, and the fishery catch or production might be about 7.6 x 104 crabs. Assuming only a 10% survival of 20 mm sized crabs, it would require releasing about 7.6 x 105 juvenile crabs to double this fishery production in the Rhode River. If natural mortality of juveniles is lower (as our on-going tethering experiments indicate), then the augmentation effects would obviously increase production above this level. These considerations point to the value of testing hypotheses at varying smaller scales and at key life history stages with controlled experiments, and not attempting the obviously foolish effort to impact the entire Chesapeake Bay at once from the outset.

If the purpose of stocking is to enhance spawning stock, then additional considerations enter in. For Callinectes sapidus in Chesapeake Bay, there has been a concurrent, persistent and substantial reduction in the spawning stock, recruitment, and female size (Lipcius and Stockausen 2002). During recent years, the stock has become recruitment limited, apparently as a result of reduced size of the spawning stock. The baywide dredge survey indicates that the total spawning stock has declined sharply from a high of about 190 million females in 1990-1991 to a low of 42 million in 1998-1999 (Seitz et al. 2001); the past two years have declined markedly further to a level of approximately 25 million. One way to increase the spawning stock is to protect the mature and ovigerous females from fishing, either through sanctuaries or protected areas where females aggregate, or through regulating against fishing for females. Both approaches are being implemented in Chesapeake Bay to varying degrees, but these have had limited success thus far. The number of females protected in the lower Bay sanctuary is estimated to be about 8 million when the stock has been low (Seitz et al. 2001). Another way of increasing the stock might be through stock enhancement. Thus, to increase this low level of stock by 10% would require an enhancement of 800,000 females in the spawning sanctuary. Assuming that, say, 10% survival of 20 mm juveniles are released in a subsystem linked by a corridor to the sanctuary, then some 8 million juvenile females or 16 million juveniles of both sexes would be needed. The Tamano Station hatchery in Okayama, Japan produces about 10 million C-1 juvenile Portunus trituberculatus per year, thus indicating that such a level of production is feasible.

In addition to the stock enhancement issues of hatchery-based systems, other considerations are indicated for hatcheries. In spite of the current relatively low market price for wild crabs that precludes profitable culturing of hard crabs, we believe that some profitable niches do exist for culturing crabs. One example is the out-of-season production of soft shell crabs, which may be priced 5-7 times higher than hard crabs. To further explore this avenue the Center of Marine Biotechnology (University of Maryland Biotechnology Institute) has purchased a system from Green Solution Ltd. (Kfar Hess, Israel). This system consists of a large (20,000 L) closed recirculating tank containing two drum arrays embedded with up to 2,400 small chambers. The arrays are slowly rotated and the crab in each chamber is fed automatically when its chamber reaches the surface. Preliminary growth studies have indicated that market size for soft crabs can be attained in 2 months. This system is highly intensive (can maintain up 50 kg crabs/m3) and has been successfully tested with other decapods such as crayfish and lobsters.