NOTE Beikson and Hurton Potential causes of mortality for Linwius polyphenws 



295 



them in a bleeding rack, and disinfecting the surface of 

 their arthrodial membrane with a 70% ethanol-soaked 

 cotton swab. An 18-gauge needle was inserted through 

 the arthrodial membrane into the cardiac sinus to ex- 

 tract the predetermined amount of blood. The animals 

 were then placed in a holding container without water 

 for 15-20 hours. Time out of water approximated the 

 duration that horseshoe crabs may be out of water after 

 bleeding at some biomedical bleeding facilities (Grogan'; 

 Walls^). Bled horseshoe crabs were also not immediately 

 put back in water to prevent them from absorbing wa- 

 ter and regaining their blood volume. Temperatures to 

 which horseshoe crabs were exposed were recorded us- 

 ing a temperature logger. Air temperature during this 

 time was stable at 21°C. Subsequently, the animals 

 were returned to their holding tanks and mortality was 

 monitored for two weeks. 



Experiment 2 



Horseshoe crabs from the August collection were used 

 for experiment 2. Because of difficulty in obtaining the 

 desired 200 study animals, slightly fewer ( 195) horseshoe 

 crabs were selected for this experiment. This sample of 

 110 males and 85 females was used to test the effects of 

 various levels of blood extraction on mortality in the pres- 

 ence of external stressors. These animals represented the 

 "higher-stressed" group because they were exposed to 

 additional stressors to which the first group was not. The 

 external stressors (i.e., air exposure, increased tempera- 

 ture, etc.) were those derived from simulated transport 

 and holding procedures during a simulated biomedical 

 bleeding process. Because there is great variability in 

 transport and holding procedures and conditions, we 

 simulated a protocol that would demonstrate horseshoe 

 crab response under relatively poor conditions. 



During the week of acclimation to the aquaculture 

 system, the horseshoe crabs were held in the same con- 

 ditions and handled in the same manner as group 1. 

 Horseshoe crabs were assigned to one of five bleeding 

 treatments used in experiment 1. The bleeding process 

 of experiment 2 involved removing horseshoe crabs from 

 their holding tank and placing them in holding contain- 

 ers located outside where they were exposed to air, sun, 

 and increased temperatures. During this 6-hour period, 

 the air temperature rose from 21°C to 29°C to simulate 

 time on the deck of a trawler. Next, the animals were 

 moved into a small moving truck, which was not air- 

 conditioned. This phase simulated transportation to the 

 bleeding facility and holding time until placement into 

 the coldroom of the laboratory. During this 4-hour pe- 

 riod, the outdoor temperature increased to 31°C and the 

 temperature inside the closed truck peaked at 36°C. The 

 animals were then transferred to the HCRC's tank room 

 where they remained for 16 hours at 21°C. This phase 

 mimicked the holding time in a coldroom prior to bleed- 

 ing. The animals were bled according to their assigned 

 treatment and in the same manner as in experiment 1. 

 During this 8-hour period, the room temperature was 

 22°C. Once the bleeding was completed, the horseshoe 



crabs were moved back into the truck for 13 hours to 

 simulate holding time and transport to the boat. The 

 overnight temperature inside the truck dropped from 

 24°C to 20"C. In the morning, the horseshoe crabs were 

 returned to the water in their recirculating aquaculture 

 tanks and mortality was monitored for two weeks. 



Statistical analyses 



Fisher's exact test was used to evaluate 1) differences in 

 mortality of unbled horseshoe crabs between the higher- 

 stressed and lower-stressed groups, and 2) differences 

 in mortality of bled crabs between the lower-stressed 

 and higher-stressed groups. Logistic regression was 

 used to examine the association between mortality and 

 bleeding in the higher-stressed group. Data were also 

 assessed to see if an increase in the volume bled was 

 correlated with an increase in mortality by pooling the 

 frequency of mortalities between males and females in 

 each bleeding treatment and testing these values against 

 bleeding percentage by employing a regression with a 

 fitted quadratic curve. 



Results 



Experiment 1 



No horseshoe crab mortality occurred in any of the five 

 treatments in the lower-stressed group, indicating that 

 post-bleeding mortality in Limulus did not arise solely 

 from the effects of blood loss. Hence, our initial hypoth- 

 esis that substantial blood loss was the principle cause 

 of mortality in horseshoe crabs was rejected. 



Experiment 2 



There were a total of 14 mortalities distributed through- 

 out the bleeding treatments among horseshoe crabs 

 under higher stress conditions (Table 1). All mortalities 

 occurred within the first seven days of the study. Bled 

 horseshoe crabs had an overall mortality rate of 8.3% 

 compared to the 2.6% mortality rate of unbled crabs, 

 suggesting a relationship between mortality and bleed- 

 ing under higher-stressed conditions (;?=195; P=0.0088) 

 (Table 1). The bleeding variable was significant in the 

 logistic regression model (P=0.G160); yet, sex was not a 

 significant variable (n=156; P=0.6100); seven deaths in 

 total occurred in each sex category (Table 1). Mortality 

 rates reached 13.6% for males bled 30% of blood volume 

 and female mortality was as high as 29.4% in the sub- 

 group bled 40% of blood volume (Table 1). With male 

 and female mortalities pooled, the frequency of mortality 

 increased as the percentage of blood taken from the crabs 

 was increased (;! = 5; P=0.006; r-=0.994) (Table 1). 



Comparison of the two experiments 



With no mortalities in the unbled treatment of the lower- 

 stressed group and one mortality in the unbled treat- 



