86 
Fishery Bulletin 110(1) 
Larval samples collected during net tows as part of 
the California Cooperative Oceanic Fisheries Investi- 
gations (CalCOFI) program provide the only available 
fishery-independent data on Pacific mackerel abundance 
in the California Current system (Crone et al., 2009). 
These data have been used to estimate a time series 
of daily larval production at hatching (Lo et ah, 2009; 
Lo et al., 2010) by a method somewhat similar to the 
daily egg production method (Lasker, 1985). Production 
is calculated on the basis of development and mortality 
rates of larvae, which are estimated from water tem- 
perature, larval abundance, and size. The time series 
of larval production may be incorporated into the stock 
assessment as an index of the stock size that produced 
the larvae (Dorval et al., 2007). However, the analysis 
was not included in the most recent stock assessment 
for Pacific mackerel (Crone et al., 2009), partly because 
of the concern that measurements within the current 
CalCOFI sampling pattern may not be representative 
of conditions for the entire range of the stock. Sampling 
was conducted over all, or nearly all, of the expected 
range of the stock during the early years of the Cal- 
COFI survey (1951-80) but now occurs only in an area 
from Avila Beach (35.1°N) to San Diego, California 
(32.7°N). We refer to this area hereafter as the “core 
area” (Fig. 1). 
We used CalCOFI data to model the likelihood of 
capturing Pacific mackerel larvae as a function of wa- 
ter temperature, zooplankton displacement volume, 
geostrophic flow (i.e., flow resulting from the pressure 
gradient force and Coriolis force), longitude, day of year, 
and the commercial-passenger-fishing-vessel (CPFV) 
index (a proxy for stock size). The model was a spatially 
continuous function (i.e., data were “smoothed” rather 
than presented as raw catch data) intended to provide 
predictions of the areas where Pacific mackerel were 
likely to occur near the time of spawning each year. The 
objectives of the study were 1) to explain why distribu- 
tions of Pacific mackerel may have changed through 
time; 2) to compare trends in habitat quality in the core 
CalCOFI area and Mexican waters, where possible; 3) to 
determine how habitat conditions may affect larval pro- 
duction or other survey estimates that are potentially 
used in stock assessments; and 4) to allow for improved 
allocation of sampling effort in future surveys. 
Materials and methods 
Survey data 
Pacific mackerel larvae and oceanographic data were 
collected during CalCOFI cruises from 1951 to 2008. 
Samples were collected in an approximate grid pattern 
centered on the southern California Bight and rotated 
-30° off the meridian so that the grid was oriented 
with (parallel to) the shoreline. The spatial and tem- 
poral extent of CalCOFI sampling varied during this 
period. The sampling pattern initially extended from the 
California— Oregon border to the tip of Baja California. 
Figure 1 
Approximate range of Pacific mackerel (Scomber japon- 
icus) in the northeast Pacific. Light gray dots indicate 
locations where samples were collected as part the 
California Cooperative Oceanic Fisheries Investiga- 
tions (CalCOFI) program and that were used in the 
study. The area enclosed by the black line indicates 
the core CalCOFI area, which was sampled in all years 
of the survey. Black dots indicate the current 66-sta- 
tion sampling pattern that is conducted each quarter. 
Data were collected in U.S. and Mexican waters during 
most sampling years from 1951 to 1984. Thereafter, 
sampling was restricted to U.S. waters. The core area 
from Avila Beach to San Diego in the California Bight 
was sampled consistently through the entire time series. 
Sampling was conducted approximately monthly from 
1950 to 1960 and in 1966, 1969, 1972, 1975, and 1984; 
it was conducted quarterly from 1961 to 1965, 1978 
to 1980, and 1985 to 2008 with target dates of Janu- 
ary, April, July, and October. Sampling was conducted 
only triennially from 1966 to 1984, except that cruises 
were conducted in January, April, and June 1968. Only 
cruises conducted from April through September were 
included in our study. This period corresponds with 
the spawning period of Pacific mackerel, as indicated 
by the occurrence of their eggs in net samples. It has 
recently become possible to identify Pacific mackerel 
eggs (specifically to distinguish them from Pacific hake 
