Burger 



Chapter 29 



Marine Distribution, Abundance, Habitats in British Columbia 



adjacent to these two tracts of forest, while most of the very 

 low-density areas were adjacent to heavily logged watersheds 

 (Burger 1994). The relative importance of marine versus forest 

 habitat in explaining this distribution has not been determined. 



Population Changes 



Historical Numbers and Distribution 



Rodway (1990) and Rod way and others (1992) reviewed 

 the scanty evidence of long-term population changes in British 

 Columbia. There is anecdotal evidence of population declines 

 in the Strait of Georgia. Brooks (1926b) commented on the 

 scarcity of murrelets on the east coast of Vancouver Island 

 in 1925-1926, compared to numbers seen in 1920 and earlier, 

 and speculated that disease or crude oil might have been 

 responsible. Pearse (1946) reported a decline in murrelet 

 numbers in the Comox district, eastern Vancouver Island, 

 between 1917 and 1944 and attributed this to the removal of 

 coniferous forests. Rodway and others (1992) found no 

 significant trends in numbers of murrelets observed in 

 Christmas Bird Counts made at 22 sites, some extending 

 back as far as 1957, but relatively few counts were from 

 regularly used wintering areas. 



Recent Trends 



I assessed changes in densities of Marbled Murrelets in 

 the past 15-20 years by comparing surveys made from 1976 

 through 1982 with more recent surveys. Relatively few areas 

 can be compared, and the precise routes and methods of 

 some of the earlier surveys could not always be replicated. 

 The paucity of annually repeated surveys makes it difficult 

 to rule out inter-year fluctuations (perhaps related to oceanic 

 effects) as the cause of some of the changes reported here. 



Clayoquot Sound and Barkley Sound 



In June 1992 and 1993, Kelson and others (in press; 

 Kelson, pers. comm.) repeated the grid census made in 

 southeastern Clayoquot Sound in June 1982 by Sealy and 

 Carter (1984). In each year the entire area (294 km 2 ; Tofmo 

 Sound excluded) was surveyed once, over periods of several 

 days, by counting all murrelets seen within 250 m of a boat 

 traversing a u-shaped path through each 1-km by 1-km 

 block. The spatial distributions of the birds were broadly 

 similar in each count, but the total numbers declined from 

 4,522 in 1982 to 2,701 (60 percent of the 1982 total) in 1992 

 and 2,622 (58 percent) in 1993. Kelson and others (in press) 

 attributed the decline to significant reductions in old-growth 

 forests adjacent to Clayoquot Sound. 



Carter (1984) used a similar grid technique in June and 

 July 1980 in outer Trevor Channel and the adjacent Deer 

 Islands, one of the areas with consistently high densities in 

 Barkley Sound (Carter and Sealy 1984). In 10 morning 

 surveys he recorded a mean of 35 1 .6 birds (on the water and 

 flying; range 74-518) Repeating these censuses for the 

 morning periods in 1992 and 1993, Burger (1994) recorded 

 averages of only 153.0 (range 92-215; four surveys) from 2 



June to 26 July 1992 and 86.0 (4-194; six surveys) from 1 

 May to 25 July 1993. These means represent 44 percent and 

 25 percent of the 1980 mean, respectively. The low 1993 

 numbers were partly due to an early departure from this area 

 of murrelets in June, which was associated with persistent 

 warm water and possible low prey densities. 



In 1979, Carter (1984) counted murrelets in a 17.2-km 

 linear transect running through outer Trevor Channel and 

 along the open coast to Seabird Rocks. Burger (1994) counted 

 seabirds along a virtually identical 19.5-km route in 1987, 

 1989, 1991, and 1993 and found significantly lower murrelet 

 densities, with the mean values consistently less than 50 

 percent of the 1979 value (fig. 9; 1979 data versus pooled 

 data 1987-1993: Mann-Whitney test, U 7 , 3 = 79.5, P<0.01). 

 Other censuses suggest that 1979 was not an unusual year in 

 this area for the study period 1979-1982 (Carter 1984, Sealy 

 and Carter 1984a). These three replicated studies are consistent 

 in showing a significant decline in the densities of Marbled 

 Murrelets in Clayoquot and Barkley sounds. The changes 

 might be partly due to coarse-scale shifts in distribution, but 

 the Clayoquot Sound surveys covered a large area in which 

 distribution shifts of 1-10 km should have been detected. It 

 is likely that the 1992 and 1993 El Nino conditions caused 

 many murrelets to leave Clayoquot and Barkley Sound 

 temporarily. Support for this hypothesis came in spring 1994 

 when densities of murrelets in Trevor Channel, Barkley 

 Sound, were 2-3 times higher than they had been in 1992 and 

 1993 (Burger, unpubl. data). There is no simple correlation 

 between murrelet numbers and local sea temperatures, 

 however, because summer temperatures were also above 

 normal in early counts in 1979 and 1980 (but not 1982) and 

 in later counts in 1987 and 1989 (data from H. Freeland, in 

 Burger 1994). The effects of local ocean temperatures, 

 upwelling events, and El Nino conditions on the distribution 

 of Marbled Murrelets and their prey clearly need to be 

 investigated in detail to help explain the apparent declines. 



One likely cause of decline is the widespread loss of 

 valley-bottom old-growth forests in the surrounding areas 

 (Kelson and others in press, Sealy and Carter 1984). Between 

 1954 and 1990, an estimated 75 percent of the ancient rainforest 

 of southern Vancouver Island was logged, including extensive 

 tracts adjacent to Clayoquot and Barkley Sounds (Husband 

 and Frampton 1991), and much of this occurred in the past 

 decade. Gill-net fishing also killed appreciable numbers of 

 murrelets in Barkley Sound (Carter and Sealy 1984), but 

 does not occur here every year (Burger, pers. obs.) and is not 

 a factor in Clayoquot Sound (Kelson and others, in press). 

 Increased disturbances from sports fishing and recreational 

 boating might have displaced murrelets at a few localities. 



Queen Charlotte Islands 



Nearshore surveys made in May and June in 1977 

 (Vermeer and others 1983) provide comparative data for 

 some areas in which similar surveys were undertaken in 

 1990-1992 (table 3). The trends were not consistent; there 

 were increases in two areas with relatively low densities, but 





306 



USDA Forest Service Gen. Tech. Rep. PSW-152. 1995. 



