KAKLUK RIVER RED-SALMON INVESTIGATION 



65 



focused our attention, first, on number of parents, because of the immediate practical 

 value of the results to be obtained on pressing problems of conservation. The 

 Bureau of Fisheries is empowered to secure an adequate escapement of spawning 

 fish and can find means for attaining this end; but it has no influence over the seasonal 

 conditions that favor or are detrimental to egg-laying, incubation, and hatching, or 

 to the incidence of disease, the attacks of parasites, or the depredations of predatory 

 forms, all of which exact a heavy toll during life in fresh water. 



Although the first duty of the bureau to assure adequate spawning reserves, 

 we shall not neglect other data that can be obtained in the prosecution of our inves- 

 tigations, nor the testing of these as to their predictive value. Annually we shall 

 secure the census of fish that ascend the river to spawn. This does not furnish the 

 actual number of spawning fish, inasmuch as many die without spawning on reaching 

 the lake or its tributaries; but it is believed closely to approximate the number of 

 actual spawners and represents the closest determination of their number of which 

 we are capable. As to the subsequent history of the brood, we entertain no hope 

 of being able to estimate the numbers of fry emerging from the gravels, nor of the 

 numbers of fry and fingerlings during their residence in the lake. Attempts to esti- 

 mate the proportions of the downstream migration appear more promising, although 

 the difficulties in the way seem formidable. After the fingerlings have reached the 

 sea, the only clues that can be obtained as to the size of the growing colonies are such 

 as are furnished by the precociously maturing grilse. The No. 1 grilse — those that 

 mature during the same season in which they reach the sea (figs. 25, 26, and 27) — have 

 little or no value for this purpose. They are recruited almost exclusively from a small 

 group of fingerlings, which migrate seaward in their fourth year, and such predictions 

 as could be based on the numbers in which they appear from year to year would 

 furnish evidence concerning only the 6 4 and 7 4 groups, which furnish relatively unim- 

 portant constituents of the run. To be of any considerable value, predictions must 

 deal with the probable size of the 5 3 group, which normally comprises about 80 per 

 cent of the run. This group is largely withdrawn from observation from the time 

 when it descends the river to the sea, in its third year, until it returns as 5-year fish. 

 Only a small percentage of it matures precociously one year earlier and returns to 

 form the 4 3 group. 



This group varies greatly in size with different years, and the question for us to 

 solve is whether the magnitude of- its occurrence in any year gives reliable evidence 

 of the size of the brood to which it belongs, which will largely mature the following 

 year as group 5 3 and will form the bulk of that year's run. If approximately the same 

 percentage of each brood matures precociously as 4 3 fish, the numbers of the latter 

 will have predictive value ; but if the percentage varies widely in response to external 

 conditions that differ from year to year and affect the age of maturing, the number 

 of 4 3 grilse can not be relied on as an indicator of the size of the next year's run. 

 This important question can be answered only after observations extending over a 

 term of years. 



A good example of the manner in which we seek to make use of the lines of 

 evidence above described is found in our attempts to predict the probable size of 

 the 1927 Karluk run. As we have already indicated, these lines of evidence are at 

 present to be considered as on trial. They are subjects for investigation and are 

 68335—27 5 



