Weber and McClatchie: Effect of environmental conditions on the distribution of Scomber /aponicus 
93 
Figure 5 
Mean annual values of predictor variables (A-D) in the core area (black line), northern Mexican waters 
north of California Cooperative Oceanic Fisheries Investigations line 95 (dashed line), and southern 
Mexican waters (dotted line). The dot-dashed line in E indicates the commercial-passenger-fishing-vessel 
(CPFV) index, which was applied as an annual adjustment to all samples. Shading indicates modeled 
probability of capturing Pacific mackerel (Scomber japonicus) larvae; darker shading indicates greater 
probability. (F) Proportion of sampled sites where one or more larvae actually were captured. (G) Overall 
mean predicted values for the model. (H) Mean predicted values at the median CPFV and day of year. 
larvae could be predicted by using zooplankton dis- 
placement volume, geostrophic How, and temperature 
as predictors of physical habitat in the California Cur- 
rent system. However, modeling habitat preferences 
was complicated by the fact that two types of spawners 
occur: those that exhibit peak spawning during April 
in the SCB at about 15.5°C, and a smaller group that 
exhibits peak spawning in August near Punta Euge- 
nia, Mexico, at 20°C or greater. Lo et al. (2010) also 
reported the existence of a secondary period of high 
larval abundance that occurs near Punta Eugenia in 
August determined from a combination CalCOFI data 
and data from the Investigaciones Mexicanas de la 
Corriente de California (IMECOCAL) program. The 
IMECOCAL program began in 1997 and collects many 
of the same types of data as CalCOFI between Ensenada 
and approximately Lazaro Cardenas, Baja California 
(30.5°N). It is unknown whether the bimodal distribu- 
tions in temperature and seasonal preferences reported 
here and by Lo et al. (2010) were caused by heterogeneity 
within the northernmost Pacific mackerel stock in the 
northeast Pacific or by captures of some Pacific mack- 
erel from the more southern stock that occur near the 
tip of Baja California (Roedel, 1952). In either case, the 
relatively continuous distributions of temperatures and 
days in which Pacific mackerel were captured indicate 
that a gradient of behaviors existed. 
We took an empirical approach to modeling multiple 
spawner types by fitting interactions between latitude 
and temperature, and latitude and day of year, to fit a 
