on the migration of sockeye salmon. As has l)een 

 brought out earlier, the travel time was 8 to 9, 

 7 to 8, and 7 to 8 days for the three groups taggcMl 

 in 1954 before Rocky Reach Dam was constructed. 

 In 1963, after Rocky Reach Dam was completed, 

 the most frequent time was 9 days. On the basis 

 of these travel times, we may assume that the 

 sockeye salmon arrive at the mouth of the Okanogan 

 Rivor, SO miles above Rock Island Dam or roughly 

 hallway on the 154-mile distance from Rock Island 

 Dam to Zosel Dam, on the fourth or fifth day after 

 I)assing Rock Island Dam. Accordingly, the re- 

 maining 3 to 5 days of the typical 8- or 9-day total 

 migration time are spent negotiating the 74-mile 

 route from the confluence to Zosel Dam. 



Under this assumption, we can retrace the migra- 

 tion of certain segments of the runs in several years. 

 In 1937, for example, fish were abundant at Rock 

 Island Dam beginning July 13. Had the Okanogan- 

 bound segment of this run moved without delay, we 

 would have expected it at Zosel Dam beginning 

 July 21. Yet, the counts at Zosel Dam remained 

 practically nil until July 31—3 days after the 

 temperature of the Okanogan River began a sharp 

 decline. Prior to July 28, the temperature had been 

 either relatively stable or climbing sharply; either 

 condition apparently delayed the fish — some as 

 long as 10 days. 



The events of 1937 were essentially repeatet! in 

 1944; fish counts increased significantly at Rock 

 Island beginning July 13, 1944, but not at Zosel 

 Dam until August 3 — 5 days after the water tem- 

 perature decreased on July 29 and up to 13 days 

 later than expected. Then, following several days 

 of high counts, the number of sockeye salmon 

 arriving at Zosel Dam dropped sharply, finally 

 reaching a low of two fish on August 9. These low 

 counts corresponded with rising or stable water 

 temperatures. The count surged again on August 

 10 — 3 days after a sharp drop in water temperature. 

 Fish migration was similar in 1952. Counting of 

 fish began at Zosel Dam on July 15, but no sockeye 

 salmon were seen until the evening of July 20, when 

 seven were taken in the traps. The count increased 

 markedly the next day, and sockeye salmon were 

 relatively abundant the following 9 days. This 

 increased abundance of fish began 5 days after the 

 beginning of a temperature drop which eventually 

 lasted 9 days. Then, beginning July 25, the tem- 

 perature rose steadily until .\ugust 5, when it began 

 to decline. This second rise in temperature brought 



a second period of low counts which lasted until 

 August 10 — 5 days after the temperature fell on 

 August 5. Movement was suppressed a third time 

 by a general 6-day rise in temperature from August 

 8 to 13. A decrease of tempiM-ature on .\ugust 14 

 resulted in a surge of fish at Zosel Dam on August 

 18 — 4 days later. Thereafter, the temperature 

 tlioppcd steadily and, judging by the counts, fish 

 migration through tiie area was unimpeded. 



The counts at Zosel Dam in 1953 generally re- 

 flected the counts at Rock Island. This agreement 

 probably occurred because increases in water tem- 

 perature were short (2-4 days) and were followed by 

 falling temperatures during the time when the 

 greater portion of the run was migrating through 

 the critical area. 



A late-season temperature rise on August 6-7, 

 1953, coincided with the arrival of tagged fish. 

 This event provides the first opportunity to study 

 the effects of water temjicrature in terms of a marked 

 segment of the population. For this analysis, we 

 need to refer to figure 5, which depicts the Okanogan 

 River temperatures and the movement of tagged 

 fish from Rock Island to Zosel Dam. On the as- 

 sumption that the normal travel time from Rock 

 Island Dam to the confluence of the two rivers is 

 4 to 5 days, we reason that some individuals from 

 the lot tagged on July 31 reached the confluence 

 on August 4 or 5, liefore tlie temperature rise of 

 August (1-7. These early arrivals peaked initially 

 at Zosel Dam on August 8. The lising water tem- 

 peratures on August 6-7 suppressed entry of the 

 later arrivals into the Okanogan River until a drop 

 of temperature on August 8. This decrease of 

 temperature resulted in another surge of tagged fish 

 at Zosel Dam on August II to 12, and gave the 

 count of tagged fish at Zosel a bimodal distribution 

 not evident at Rock Island. Apparently, few fish 

 from the lot tagged on August 1 arrived at the river 

 mouth before the temperature rose; most arrived 

 during the rise of August 6 to 7 and therefore did 

 not appear at Zosel Dam until August 12—4 days 

 after the temperature fell on August 8. Similarly, 

 few fish from the lot tagged on August 2 arrived at 

 the confluence l)cfore the temperature rise suppressed 

 their entry. Most of the.se fish arrived at Zosel Dam 

 on August 12 to 13 — 4 or 5 days after the tem- 

 perature drop. 



The movement of the various tagged segments of 

 the 1951 run can also be retraced from figure 5. 



144 



U.S. I'isn AND WII.DUFK SERVIfn 



