458 



Fishery Bulletin 101(2) 



facility in Chincoteague, Virginia. At the bleeding facil- 

 ity, we randomly selected a predetermined number ( 10 in 

 1999, 30 in 2000 and 2001) of newly matured male horse- 

 shoe crabs (identified by pristine shell condition and the 

 presence of boxing-glove lower claws [Shuster*] ) from all 

 of the horseshoe crabs obtained in that day's trawls. We 

 selected newly matured male horseshoe crabs to minimize 

 covariance in our study. These horseshoe crabs were not 

 bled and served as a control in the experiment. They were 

 packed in coolers labeled "unbled," and set aside. The same 

 number of newly matured male horseshoe crabs were then 

 randomly selected from the remaining horseshoe crabs and 

 underwent BioWhittaker's normal bleeding process. Upon 

 completion of the bleeding process, the horseshoe crabs 

 were packed in coolers labeled "bled." 



All coolers containing horseshoe crabs, both bled and 

 unbled, were immediately packed in an air-conditioned ve- 

 hicle and transported to the Virginia Seafood Agricultural 

 Research and Extension Center in Hampton, Virginia. The 

 horseshoe crabs were removed from the coolers and the 

 unbled horseshoe crabs were marked with external tags 

 to distinguish them from the bled horseshoe crabs. These 

 markings were unobtrusive and did not cause any undue 

 stress to the unbled horseshoe crabs. All of the horseshoe 

 crabs were placed in four replicated, flow-through holding 

 tanks, and equal numbers of bled and unbled horseshoe 

 crabs were held in each tank. The horseshoe crabs remained 

 in the tank system at Hampton for two weeks. Horseshoe 

 crabs were maintained in appropriate conditions (Brown 

 and Clapper, 1981), and monitored daily. Horseshoe crabs 

 that died during the two-week period were removed and 

 returned to the ocean at the time of their death. 



At the conclusion of each two-week period, the status 

 of each horseshoe crab (dead or alive) was recorded. All 



•* Shuster, C. N., Jr 1999. Managing the horseshoe crab re- 

 source: it's the adult age that counts, 32 p. Virginia Institute 

 of Marine Science, College of William and Marv, (Gloucester 

 Point, VA. 



surviving horseshoe crabs were removed from the tank, 

 placed in coolers, packed in an air-conditioned vehicle, re- 

 turned to BioWhittaker's bleeding facility in Chincoteague, 

 Virginia, and returned to the Atlantic Ocean in accordance 

 with BioWhittaker's standard operating procedures. This 

 procedure was repeated eight times during summers 1999, 

 2000, and 2001. The results from each of the replicates 

 were combined, and the overall percentage mortality was 

 calculated for the bled and unbled groups. 



Using Fisher's exact test, we evaluated statistical sig- 

 nificance of differences in mortality between the bled and 

 unbled horseshoe crabs (Mehta and Patel, 1999). We then 

 calculated a 957? confidence interval for average differen- 

 tial mortality using the common odds ratio in the statistical 

 program StatXact (Mehta and Patel, 1999). 



Results 



A Fisher's exact test for statistical significance showed 

 differences between mortality rates in bled and unbled 

 horseshoe crabs (P=2.085E-04). Bled horseshoe crabs 

 (n=200) had an overall mortality rate of S'J't compared 

 to the 0.5% mortality rate of the unbled horseshoe crabs 

 («=200; P<0.001) (Table 1). Thus, this study estimates 

 average differential mortality between bled and unbled 

 horseshoe crabs to be 7.5'y. The 95^^ confidence interval 

 for this average differential mortality ranges from 0.14% 

 to 38.1% as calculated with the common odds ratio (Mehta 

 and Patel, 1999). 



Discussion 



Our results indicate that horseshoe crab mortality due to 

 bleeding is relatively low. Two small-scale studies had pre- 

 viously estimated postbleedmg mortality. Rudloe (1983), 

 observing bled and unbled horseshoe crabs in a penned 

 cove in Florida, found that bleeding increased mortality by 



