ling rainbow trout were killed in the field, de - 

 spite changes made in the electrode spacing. 

 In Harlow Creek, a trout stream of similar size 

 and with the same type of control structure 

 (Type C), only 6 small rainbow trout were found 

 dead in the field. Low water velocities in Elm 

 River and sufficiently high velocities in Harlow 

 Creek account for the difference in number of 

 fatalities . 



Stream velocity is probably the most im- 

 portant physical characteristic affecting fish 

 mortality from electrocution. Generally, an 

 excessive kill occurs when water velocities are 

 low, unless the pattern of fish movement is re- 

 stricted to a section of the stream where the 

 fish can be trapped before they enter the elec- 

 trical field. In the Pilgrim River, the flow is 

 sluggish after the spring runoff, and since the 

 white sucker did not have a well-defined up- 

 stream route of migration 1,572 were killed in 

 the electrical field. 



In the selection of control sites, every ef- 

 fort has been made to place the devices in 

 locations harmless to the migrating fish. The 

 most important consideration, however, has 

 been the blocking of the sea lampreys from 

 spawning grounds. If the site must be selected 

 where conditions will cause some fish kill, the 

 mortality should be considered part of the cost 

 of operations. 



Although fish losses were high at some 

 barriers, we believe they were generally in- 

 significant in relation to the total runs and to 

 the benefits of lamprey control. Among the 10 

 important species, including the suckers, the 

 total mortality was only 29.8 percent. This 

 figure includes individuals taken alive in the 

 trap and those found dead in and below the field . 

 It was not possible to determine the number of 

 fish and lampreys that were blocked and turned 

 back by the barriers to enter other streams. 

 Observations indicate that certain species, such 

 as suckers and smelt, spawn in some of the 

 streams below the barriers. 



Biological characteristics of the 

 lamprey spawning runs 



To supplement the original study on 

 the sea -lamprey spawning runs in Michigan 

 (Applegate 1950), the collection of data has 

 been continued with the control program . 



Over the present range of the species 

 in the Great Lakes basins, the seasonal 

 spawning migrations extend from March 

 through August. Early migrants enter a few 

 streams with estuarine waters in March, and 

 upstream movement begins about mid-April, 

 when water temperatures rise and remain 

 above 40° F. As water temperatures in- 

 crease, the upstream movement gains momen- 

 tum and usually reaches a peak about the end 

 of May or the first week of June . In the weekly 

 record for 1954 (table 5 and fig. 5), we have 

 combined the number of sea lampreys cap- 

 tured in each period from the productive 

 streams of Lake Superior and northern Lake 

 Michigan. In Lake Michigan, 54 percent of 

 the total run in the 7 streams with electrical 

 barriers had been captured by May 21; the 

 corresponding figure for 15 streams in Lake 

 Superior was 28 percent By July 9, the 

 spawning migration of sea lampreys in the 

 streams of northern Green Bay had ended. 

 On Lake Superior, the spawning runs in a 

 few streams continued into August. 



Several index streams were selected 

 in 1953 and 1954 as sources of information 

 on the biological characteristics of the 

 spawning runs. The data on the sex of sea 

 lampreys in these streams showed that the 

 percentage of males increased from 1953 to 

 1954. (table 6). In the 5 index streams ob- 

 served in 1953, the sex composition of 1,777 

 lampreys was 99 males to 100 females. In 

 1954 (when the number of index streams on 

 Lake Superior was increased to 9), the sex 

 ratio was 140 males to 100 females for 3,939 

 sea lampreys. The sex ratio of 3,530 sea 

 lampreys sampled in 1954 in the 5 index 



18 



