Ainley and others 



Chapter 34 



Offshore Occurrence Patterns in Central California 



all birds within 300 m of one fore-quarter, i.e., that side of 

 the bow on which glare was least, but did not include birds 

 following or attracted to the ship (see Tasker and others 

 1984; not a problem in the case of murrelets). Almost all 

 murrelets were detected on the water, thus, we did not 

 correct for bird flight speed or direction (Spear and others 

 1992). We looked forward of the 300-m transect boundary 

 to ensure that murrelets were detected before the ship 

 caused them to dive or take flight. Our counts were not 

 affected by ocean turbulence, as wind speeds were generally 

 less than 15 knots and the 10-m above water vantage point 

 of the flying bridge precluded birds from being obscured by 

 waves. The average ( s.d.) wind speed was 9.0 6 knots. 

 Similarly, counts were not affected by poor visibility as we 

 experienced no fog. 



Two observers censused the 300-m band continuously 

 whenever the ship was under way (cruising speed 14-15 km/ 

 h). We divided censuses into 15-minute (ca. 3.5 km) segments; 

 at the start of each segment, we recorded position, distance- 

 to-shore, water depth, weather (wind speed and direction, 

 cloud cover) and, using the ship's electronic systems, sea- 

 surface temperature, and salinity. We determined thermocline 

 depth and slope from "CTD'"s conducted by ship's personnel 

 at 8-km intervals along our cruise track (CTD is "conductivity 

 and temperature with depth" and refers to the probe used to 

 measure these factors.) Within the region bounded by the 

 coast, 26 km offshore and at latitudes 37 02' N and 37 20' 

 N (fig. 1), we logged 1863 15-min transects. 



National Marine Fisheries Service personnel conducted 

 trawls at night to estimate the concentrations of micronektonic 

 (i.e., ones <10 cm in length) crustaceans and fish. The trawl 

 surveys, done in the May-June period, were designed to 

 assess prevalence of juvenile rockfish (Sebastes spp.). These 

 and similar sized organisms, all of which the nets targeted, 

 all would include suitable murrelet prey. One track of five 

 trawl stations at increasing bottom depth (25 - 480 m) occurred 

 in the southern part of the study area, off Davenport, and 

 another occurred in the northern part, off Pescadero (see fig. 

 1A). In most years, two sweeps of the trawl stations were 

 made during the murrelet census period, but in a few years 

 only one was made. For a given year, we averaged trawl 

 results combining all sweeps and both station lines by depth 

 stratum in 5 intervals. Herein, we present average percent 

 composition of the 5 trawl intervals among the four most 

 prevalent species groups: euphausiids (mostly Thysanoessa 

 spinifera), anchovies (Engraulis mordax), juvenile rockfish, 

 and myctophiids (mostly Tarletonbaena crenularis). 



We mapped murrelet sightings using an Arc-Info 

 Geographic Information System (ESRI, Inc., Redlands, 

 California). We analyzed the physical habitat features listed 

 above in an attempt to explain murrelet occurrence patterns 

 by both multiple regression on murrelet density and logistic 

 regression on presence/absence of murrelets on census 

 segments. Data were log-transformed before analysis. Our 

 initial model also included the following interaction terms as 

 independent variables: distance-to-shore x distance-to-nesting- 



area; distance-to-shore x depth; and distance-to-shore x 

 distance-to-shelf-break. Besides the physical habitat features, 

 we calculated distance-to-nesting-area for each transect 

 segment using the southern boundary of the region where 

 most nests or grounded fledglings have been found (see fig. 

 1A; see Carter and Erickson 1992). We conducted the 

 backwards stepwise analysis first for all years and all seasons, 

 and then for early and late spring separately. We considered 

 results to be significant at P < 0.05. 



We could not use prey abundance as a factor in the 

 regressions owing to the statistically inappropriate spacing of 

 the trawl stations relative to the bird censuses. We made 

 qualitative comparisons between murrelet and prey distributions, 

 considering both the species composition of trawls and the 

 number of organisms caught per trawl by depth stratum. 



Finally, we investigated the distribution of murrelets 

 relative to distance-to-shore, as this parameter is integral to 

 assessment of murrelet numbers by other researchers in 

 many regions to the north. Besides determining the proportion 

 of murrelets seen at 1000-m intervals from shore, we also 

 normalized the data by using search effort (number of 

 murrelets seen divided by the number of transects at each 

 distance interval x 100). 



Results 



Although our June surveys were conducted in the middle 

 of the murrelet nesting season (Carter and Erickson 1 992), 

 we saw no fledglings. Our surveys earlier in the spring 

 coincided with the early courtship period. 



Most murrelet sightings were within 7 km of shore (fig. 

 2) (median <5 km, x s.d. 5.5 5.4 km). The largest number 

 of sightings occurred 3-5 km offshore; normalizing the data 

 by search effort revealed no change in this pattern. We saw 

 one Marbled Murrelet 24 km offshore, the farthest from 

 shore that one was seen. This was near to the edge of the 

 continental shelf break. 



In spite of the wide area of our search, almost all murrelet 

 sightings were within 10 km of Ano Nuevo. Only once did 

 we see Marbled Murrelets anywhere else in the region, 

 Bodega Bay to Carmel, in any of the years (see below). On a 

 smaller scale within 10 km of Ano Nuevo substantial 

 annual variation was apparent (compare cells of fig. 1). On 

 some cruises, we saw few murrelets: during early spring, 

 only 5 in April 1988 and none in March 1994; and during 

 late spring, none in June 1986, two in 1989, and one in 1992, 

 even though sampling was adequate. In spite of even more 

 sampling tracks, we saw only 5 murrelets in June 1988 and 6 

 in June 1993. Murrelets were much more numerous during 

 other cruises: 18 in April 1987; 19 in March 1992; and 16 in 

 March 1993. Highest numbers occurred in late spring (June): 

 27 in 1987, 16 in 1990, 45 in 1991, and 28 in 1994. The 

 March 1993 census was the only one in which we saw 

 substantial numbers of murrelets farther than 10 km from 

 Ano Nuevo; 12 more were recorded off Santa Cruz, a point 

 located just to the right of the margin in figure 1. 



364 



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



