RATE OF DISAPPEARANCE OF 

 CARCASSES AFTER SPAWNING 



Carcasses of fish found on each survey were 

 distinctively marked by amputating one or more 

 of the fins and were replaced where they were 

 found. Any fin-clipped fish found on a survey 

 was recorded. Despite the fact that surveys 

 were made at 7-day intervals and 3,378 

 carcasses were marked, only 41, or 1 percent, 

 were subsequently found, indicating a rapid 

 disappearance (table 3). On the smaller tribu- 

 taries scavengers seemed to be responsible, 

 while on Brooks River the fast current caused 

 carcasses to be washed downstream into deep 

 water. The successive survey counts of dead 

 fish on a given stream may therefore be re- 

 garded as additive for minimum total enumera- 

 tion purposes. 



EFFECTS OF BEAR PREDATION 



Bears were numerous and fed actively on 

 salmon in all tributariesof Brooks Lake during 

 the early and peak spawning periods. Toward 

 the end of spawning in each stream, despite 

 an abundance of salmon, bears moved to 

 berry fields on high ground away from the 

 stream valleys. This seasonal shift in diet 

 from salmon to berries has been described 

 at Karluk Lake on Kodiak Island by Clark 

 (1957). Our program in 1957 was not originally 

 planned to study bear predation, but some 

 knowledge was gained. 



From the evidence of fresh signs of bears 

 (feces, footprints, clawmarks) and of partially 

 eaten salmon carcasses, bear activity in order 

 of prevalence by stream was as follows: Hidden 

 Creek, Up-a-tree Creek, One Shot Creek, 

 Headwater Creek, and Brooks River. 



Bears could be heard from the streamside 

 field station on Brooks River almost every 

 night during September as they fought with 

 each other and pursued salmon near the weir. 

 This was a temporary situation, however, 

 resulting from the concentration of fish and 

 the presence of the weir, which prevented fish 

 from escaping easily. Under natural conditions 

 bears probably have little effect on sockeye 



salmon in large streams such as Headwater 

 Creek and Brooks River. 



Hidden Creek, the most heavily frequented 

 by bears, was carefully surveyed on September 

 3 to evaluate the effect bears had on spawning. 

 The creek, about 10 feet wide and flowing 

 about 10 cubic feet per second, is similar 

 in size and terrain to Up-a-tree and One Shot 

 Creeks. The creek was surveyed upstream to 

 beyond the limit of spawning. Most of the banks 

 were covered with brush and high grass, but 

 every 10 or 15 feet the vegetation was flattened, 

 indicating "feeding tables" where bears had fed 

 on salmon. It could not be determined if these 

 were the result of fishing by many bears or 

 of great activity by a few. Fresh bear feces 

 was abundant and contained both berries and 

 salmon bones. 



Most of the carcasses mutilated by bears 

 had deep tooth marks, or the posterior section 

 of the body was missing (fig. 7). On the Hidden 

 Creek survey, several live spent salmon were 

 seen with large bite marks on the median and 

 posterior portions of the body. Spawning suc- 

 cess was recorded whenever it was possible. 

 The appearance of the gonads in the anterior 

 portion of the body usually provided clues to 

 whether fish had spawned successfully. 



Of 290 dead sockeye salmon found on the 

 Hidden Creek survey, 188 were females and 102 

 males. Bears had mutilated 247, of which 10 

 females and 9 males were unspawned, and 3 

 females and 13males were partly spawned. The 

 remaining 212 fish were completely spawned. 

 Spawning was apparently at or slightly past its 

 peak, since more carcasses (489) were found on 

 this survey than on any preceding or succeeding 

 survey. Although this was only a single sample 

 from one stream. I believe it is a measure of 

 maximum bear damage on the Brooks system. 

 The spawning population in Hidden Creek was 

 more heavily affected by bears than the popu- 

 lations in the other tributaries, as shown by 

 comparative general observations of bear 

 predation during stream surveys. 



I conclude that even under conditions of 

 intense bear activity on a small stream where 

 salmon are abundant and vulnerable, bears 

 catch mostly spawned out fish and have little 

 effect on ultimate production. 



10 



