fertilizer is brought upriver for distribution to farmers. The 

 river seldom is closed to barge traffic by ice. Hence, our 

 results represent effects of a seasonal minimum in fleeting 

 activity. 



Mortality 



Mortality rates were calculated as percentages by dividing 

 the number of recaptures that had died between samplings by the 

 total number of recaptures. 



Dead shells were marked in order to develop a correction 

 factor for mortality rates. A marked mussel which dies during 

 the interval between recaptures may be more likely to be washed 

 out of the study plot or buried beneath sediment than a live 

 mussel which maintains itself in the substrate at the sediment- 

 water interface. A mortality rate based on the number of marked 

 shells found dead since last being captured thus underestimates 

 the actual mortality if dead shells are less likely to be 

 recaptured than live mussels. 



Comparison of mortality rates between plots assumes the 

 probability of recovery of dead shells is the same for each plot. 

 However, mortality rates are more likely to be underestimated in 

 the fleeted plots, where prop wash displaces dead shells, than in 

 control plots. Light, fragile shells are more likely to be 

 displaced than heavy shells. Therefore, separate correction 



15 



