664 



Fishery Bulletin 102(4) 



Squid fishing vessels could not always be distinguished 

 as light boats or seiners and therefore were recorded as 

 "squid fishing vessels." 



The numbers of squid fishing vessels showed large 

 skew in their frequency distribution. These data were 

 transformed by x' = log 10 (x+l). Similarly, proportion 

 lunar phase was transformed by x' = arcsin(V.r), and 

 detected light pixels were transformed by x' = log ]0 (.v+l) 

 to correct for skew (Zar, 1984). These transformations 

 produced normally distributed data acceptable for re- 

 gression analysis. With these transformed variables, 

 multiple stepwise regression (forward selection) was 

 performed with the software S-Plus 2000 (MathSoft 

 Inc., Cambridge, MA) to examine the effects of squid 

 fishing vessels and the proportion lunar phase on de- 

 tected light pixels. Squid fishing vessels and proportion 

 lunar phase showed very little correlation (r=-0.09). 



Fishery characteristics, 1992-2000 



For quantitative analysis of the fishery data, we aggre- 

 gated the nightly satellite data (i.e., light pixels detected 

 on the water) into calendar quarters, as suggested by 

 the within-year distribution of squid landings in the 

 bight (Butler et al., 1999). To standardize conditions of 

 light detection, we excluded all data after 28 May 2000, 

 because this was the starting date of mandatory shield- 

 ing of the high intensity lights of the lightboats. This 

 regulation was enforced by California's Department of 

 Fish and Game to reduce light pollution by the light- 

 boats. The shields did not totally obscure the lightboats 

 from detection by the satellites (authors' pers. obs.) but 

 made the emitted light less bright, and, hence, less 

 detectable by the satellites. Thus, our data for fishing 

 effort spanned calendar quarters from Jul-Sep 1992 to 

 Jan-Mar 2000. We included a quarter for analysis if it 

 contained 10 or more nights of acceptable images. By 

 these criteria, we described effort for 24 of the 31 cal- 

 endar quarters. The mean number of nights per quarter 

 was 26 (range=10-72 nights). 



The quantity (kg) and location of landed market squid 

 were recorded by California Department of Fish and 

 Game (CDFG) throughout the 1992-2000 study pe- 

 riod and were made available to the authors. During 

 this study period, squid fishing in the bight occurred 

 exclusively at night (Vojkovich, 1998). The squid were 

 landed at port within several hours after being caught; 

 therefore the landings for a given day corresponded to 

 the previous night's effort. Squid fishermen reported 

 the locations of their hauls by CDFG fishing blocks. We 

 defined catch taken from the Southern California Bight 

 as that from blocks 651-896 and 1032-1035 (Fig. 1). 

 Blocks 651-896 are typically 10' latitude x 10' longitude 

 and can be used to locate regions of high catch. Blocks 

 1032-1035 are large latitudinal bands, generally 30' 

 wide, that encompass blocks 651-896. We used blocks 

 1032-1035 in calculating the total catch in the bight, 

 but not in depicting the location of the catch. 



To construct the abundance index of landings per 

 unit of effort (LPUE). we first estimated the number of 



squid fishing vessels for each night of satellite data, us- 

 ing the regression results of the ground-truthing work 

 (see "Results" section). We then summed the nightly 

 estimated number of vessels for each calendar quarter. 

 For those nights for which we had estimated numbers of 

 vessels, we also summed the landed catch within each 

 calendar quarter. To arrive at LPUE for the quarter, 

 we divided the summed landings by the corresponding 

 summed effort. 



Environmental data 



We used the multivariate ENSO index (MEI) to indicate 

 overall environmental conditions over the course of the 

 1992-2000 study period. The MEI is a multivariate 

 index that incorporates sea level pressure, surface zonal 

 and meridional wind components, sea surface tempera- 

 ture, surface air temperature, and cloudiness (Wolter 

 and Timlin, 1998). The MEI index is calculated for the 

 tropical Pacific (i.e., between 10°N and 10°S, from Asia 

 to the Americas), and its monthly values appear on 

 the website http://www.cdc.noaa.gov/~kew/MEI/table. 

 html. 2 



Analysis of the location of fishing effort over the 

 course of the traditional squid fishing season in the 

 bight led to an investigation of oceanographic data for 

 waters surrounding Santa Cruz Island in March. Spe- 

 cifically, we examined sea temperature from two sourc- 

 es. First, we obtained sea surface temperature for all 

 satellite nights in March 1993-2000 from the Physical 

 Oceanography Distributed Active Archive Center (PO. 

 DAAC) at California Institute of Technology (Pasadena, 

 CA). These data were reported for 18x18 km grids, 

 which were approximately the size of the 10'xlO' fish- 

 ing blocks. We selected the grid that covered block 686 

 to represent the northern shore of the island, and that 

 which covered block 708 to represent the southern shore 

 (Fig. IB). For each year in the 1993-2000 period, we 

 calculated mean March temperature for both blocks. 



The second source of sea temperature was the da- 

 tabase maintained by the California Cooperative Oce- 

 anic Fisheries Investigations (CalCOFI). Since 1950. 

 the CalCOFI program has conducted quarterly survey 

 cruises along transects perpendicular to the southern 

 California coast. This system of transects incorporates 

 66 geographically fixed stations. At each station, a 

 conductivity-temperature-depth (CTD) instrument is 

 deployed. Details on survey methods appear on the web- 

 site http://www-mlrg.ucsd.edu/calcofi. html. :1 along with 

 the publicly accessible database. For April 1993-2000, 

 we obtained temperatures at sea surface and at 75 me- 

 ters depth at two stations (Fig. 1A): 83.42 (northeast 

 of Santa Cruz Island: 34.18°N, 119.51°W) and 83.51 

 (southwest of Santa Cruz Island; 33.88 D N, 120.13 W). 



- NOAA-CIRES Climate Diagnostics Center website. I Ac- 

 cessed 3 November 200.3.1 



3 California Cooperative Oceanic Fisheries Investigations 

 website. [Accessed 3 November 200.3.1 



