520 
Fishery Bulletin 11 5(4) 
Table 1 
Start and end dates for oceanographic cruises and num¬ 
ber of sampling stations (N) used as data sources for 
analyses of seasonal and regional variation in egg sizes 
of Argentine anchoita (Engraulis anchoita) in the south¬ 
eastern Brazilian Bight. 
Date 
Oceanographic cruise 
Start 
End 
N 
Seasonal 
DeproasI 
2/7/01 
2/13/01 
19 
Deproas II 
7/12/01 
7/19/01 
17 
Deproas III 
1/5/02 
1/24/02 
72 
DeproasIV 
8/3/02 
8/21/02 
66 
Regional 
FINEP I 
11/29/75 
12/18/75 
140 
EPM Sardinha 
1/10/88 
1/30/88 
78 
V. Hensen /JOPS 
12/28/90 
1/11/91 
89 
Sardinha I 
12/8/91 
12/18/91 
110 
Sardinha II 
1/9/93 
1/18/93 
108 
by using discriminant analysis, as described by Favero 
et al. (2015). Only a subsample of about 100 eggs per 
sampling station was measured. 
Satellite data 
Sea-surface temperature (SST) corresponds with lev- 
el-3 gridded images obtained from the advanced very 
high resolution radiometer (AVHRR) on board NOAA 
satellites and processed by the Pathfinder Project. The 
Pathfinder data set is the result of a collaboration be¬ 
tween the NOAA National Oceanographic Data Cen¬ 
ter and the University of Miami Rosenstiel School of 
Marine and Atmospheric Science, and is distributed at 
NASA’s Physical Oceanography Distributed Active Ar¬ 
chive Center (AVHRR Pathfinder Level 3 Daily SST, 
vers. 5: daytime, PODAAC-PATHF-DYD50, website, 
and nighttime, POBAAC-PATHF-DYM50, website, ac¬ 
cessed December 2015) on a global scale, with a lin¬ 
ear gridded projection, and within a spatiotemporal 
resolution of 4 kmx4 kmxl month. Day and night-time 
data were used to compute SST monthly means. Sur¬ 
face chlorophyll-a concentration (CHL) was acquired by 
level-3 mapped images obtained from the Sea-viewing 
Wide Field-of-view Sensor on board the SeaStar satel¬ 
lite. These data were processed by the NASA Goddard 
Space Flight Center by using the Ocean Chlorophyll 4 
algorithm, vers. 4 (O’Reilly et al., 2000), and distrib¬ 
uted at NASA’s Ocean Color Web (chlor_a, website, 
accessed December 2015) on a global scale, with cy¬ 
lindrical equidistant projection, and a spatiotemporal 
resolution of 9 kmx9 kmxl month. Because the CHL 
is logarithmically distributed within the oceans, we 
chose to work with the loglO-transformed CHL. We 
used monthly means of SST and surface CHL to char¬ 
acterize the typical summer and winter scenarios of 
2001 and 2002, to correspond with the periods when 
the cruises were undertaken (Table 1). Monthly means 
were chosen to represent our study area for 2 reasons: 
1) monthly means serve as a low-pass filter that re¬ 
moves high-frequency external processes that could 
be involved in the variability of SST and surface CHL 
(e.g., processes driven by the atmosphere), and 2) un¬ 
like the averages for the specific days of the cruises, 
monthly means include processes with a lag between 
their cause and effect (e.g., the time lag between the 
supply of nutrients to the upper ocean and the growth 
of primary production for a given area). 
Data analyses 
Egg abundance was calculated as 
N=(x x d)/V (Tanaka, 1973), 
where N - egg abundance (number of eggs per square 
meter at each sampling station); 
x - the number of eggs sampled; 
d = the maximum depth sampled in meters; and 
V = the volume of water filtered in cubic meters. 
The mean abundance was calculated for all the sam¬ 
pling stations, not just for those where eggs were col¬ 
lected. The calculation of the frequency of egg occur¬ 
rence was based on Guille (1970). 
One-way analysis of variance was used to test the 
differences in the egg major and minor axes, egg vol¬ 
umes, temperature, and salinity (at 10-m depth), when 
compared by season (winter and summer) and by area 
(areas 1, 2 and 3) within each year. A posteriori Tukey’s 
honest significant difference test was performed in or¬ 
der to find means that were significantly different from 
each other. Temperature-salinity diagrams in relation 
to egg volume from ichthyoplankton tows were plotted 
to infer the distribution of egg sizes for Argentine an¬ 
choita with respect to water mass for each year, area, 
and season. We used the 10-m measurement for the 
abiotic data because the eggs of anchovies mainly occur 
close to the surface, in the upper 20 m (Tanaka, 1992; 
Sabates et al., 2008). R software, vers. 3.2.1( R Core 
Team, 2015) was used for these analyses. 
Results 
Seasonal variation 
Oceanographic conditions Mean seawater temperature 
measured at a depth of 10 m varied significantly with¬ 
in each period analyzed (F- 8.732, df=3, P<0.01). Mean 
summer temperatures were higher than those of winter 
for both years, but the lowest individual temperature 
values were measured during summer (Table 2), ow¬ 
ing to the SACW intrusion that occurred in both years 
surveyed, 2001 and 2002 (Gogalo et al., 2011; Araujo, 
2013). The standard deviation (SD) in temperature pro- 
