144 



Fishery Bulletin 98(1) 



trols (Fig. 3; proportional hazards, X"=13.50,P<0.001: 

 ). Hemocyte and parasite den- 



relative risk=ei055''05i74 



sity were jointly analyzed as covariates in the propor- 

 tional hazards model. For injected crabs, the decline 

 in ln( total hemocyte density) was significantly asso- 

 ciated with mortality (In day of death = 0.875 + 0.145 

 In total hemocyte density - 0.017 In Parasite density 

 + 0.409 W; X"=4.47 with 1 df, P<0.05). Hemocyte 

 density (untransformed), and parasite density (In, 

 and untransformed) were not associated with mor- 

 tality (x'-, P=0.07, 0.61, and 0.47, respectively); thus, 

 decreases in hemocyte density (In), not parasite den- 

 sity, were associated with imminent death. 



Direct observations from crabs used to maintain 

 infections and experimental results indicated that the 

 parasite was detectable in the hemolymph approxi- 

 mately two weeks after injection (Fig. 4). Although 

 the parasite could be detected as early as one week 

 after inoculation, detectability (the percentage of 

 infected crabs exhibiting detectable parasites in the 

 hemolymph) was relatively low (3,0-35%) after 14 to 



18 days, reaching 80-85% after 26 to 32 days, and 

 100% after 35 days. (Detectability was based solely on 

 inoculated animals that developed infections. The four 

 crabs from the mortality-II experiment that did not 

 present infections, hereafter refeiTed to as "immune" 

 crabs, were excluded from the analysis of detectabil- 

 ity. ) Proliferation and gi'owth of the parasite followed 

 a similar pattern as detectability, and the two vari- 

 ables were clearly related (Table 1; Fig. 4). 



Growth of the parasite showed a marked increase 

 in the mean density of vermiform plasmodia over days 

 18 to 26 (Table 1). The mean density of trophonts 

 increased markedly over days 32 to 35. Note, how- 

 ever, that to avoid mortalities from other causes (e.g. 

 secondary infections), sampling could not be done on 

 a daily basis. 



Plasmodia were found within the hearts of 93% 

 (n=14/15) of the injected crabs in the early infection 

 experiment. Plasmodia were found in 4 of 5 crabs 

 as early as day 3 (Table 2l. Uninucleate trophonts 

 were observed in the heart on and after day 7. Rel- 



