400 
Fishery Bulletin ! 13(4) 
Table 2 
The 5 extracted principal components of climate-related meteorological and hydrological variables that are con- 
sidered to have influenced the variability of catches of Gulf menhaden ( Brevoortia patronus) in beam plankton 
trawl (BPL) surveys from 1981 through 2008 in the northern Gulf of Mexico. E-W=east-west; N-S=north-south; 
N:P=nitrogen to phosphorus ratio. Data were divided into 2 climate regions: central (Louisiana, Mississippi, Ala- 
bama, and Florida panhandle) and western (Texas). 
Regions 
Variables 
^st 
Extracted principal components 
2 nd 3 rd 4 th 
5 th 
North Atlantic 
North Atlantic Oscillation 
0.467 
0.549 
Offshore waters 
E-W wind direction 
0.408 
-0.592 
0.308 
N-S wind direction 
-0.635 
0.474 
-0.356 
Central region 
Cube-root-transformed N:P ratio 
-0.375 
0.479 
0.761 
Sea level 
0.535 
0.585 
River flow 
0.583 
-0.494 
0.469 
Precipitation 
0.795 
-0.372 
Palmer Drought Severity Index 
0.771 
-0.419 
Air temperature 
0.666 
0.684 
Western region 
Precipitation 
0.788 
-0.314 
Palmer Drought Severity Index 
0.822 
-0.378 
Cube-root-transformed river flow 
0.663 
0.398 
-0.439 
Air temperature 
0.760 
0.482 
Sea level 
0.611 
0.638 
0.327 
Eigenvalue 
4.269 
2.605 
2.433 
1.368 
1.008 
Variance explained (%) 
30.489 
18.604 
17.377 
9.774 
7.203 
Cumulative variance explained (%) 
30.489 
49.093 
66.470 
76.244 
83.447 
air temperature from the western and central regions, 
sea level from the western and central regions, and 
PDSI from the western region. 
Correlation analysis of commercial landings with 
environmental variables showed that numbers and pro- 
portions of age-1 Gulf menhaden were correlated with 
air temperature and sea level in the western region (r< 
-0.485, P<0.002): high numbers (10 million fish/vtw) 
and proportions (74%/vtw) were found during periods 
of low (0.87°C below average) air temperature and 
(6954 mm) sea level. Low (6.2 million fish/vtw) num- 
bers of age-1 Gulf menhaden were correlated with high 
(0.47°C above average) air temperature and (7258 mm) 
sea level (r< -0.446, P<0.003) in the central region. 
High (74%/vtw) proportions of age-1 Gulf menhaden in 
fishery landings were correlated with AMO cold years 
(r= -0.497, P=0.000). 
Multiple regression analyses were performed to de- 
termine which of the 3 extracted components individu- 
ally or combined best resembled the variability of fish- 
ery-dependent landings of Gulf menhaden. According to 
the regression models developed from AIC, a significant 
predictor (second component) with a negative regres- 
sion coefficient explained 46% of the variance contained 
in the numbers of age-1 Gulf menhaden (Table 3, Fig. 
4B). In addition, 2 significant predictors (first and sec- 
ond components) with positive and negative regression 
coefficients, respectively, explained 34% of the variance 
contained in the proportion of age-1 Gulf menhaden in 
landings (Table 3, Fig. 4C). 
Discussion 
Differences in responses of menhaden to decadal and 
annual meteorological and hydrological regimes were 
evident in this study, providing additional support 
for the notion that climate plays a role in structuring 
suitable habitat and, therefore, in survival of larvae 
and juveniles of menhaden. The specific meteorological 
and hydrological conditions identified as determinates 
of abundance estimates, however, differed from those 
identified in past studies. We found that a high recruit- 
ment of menhaden postlarvae (3.6/BPL haul), juveniles 
(38.3/seine haul), and age-1 fish in landings (10 mil- 
lion/vtw) was associated with cold, wet spawning and 
recruitment seasons characterized by high PDSI (0.45) 
and coastal river flow (401.6 m 3 /s above average), by 
a Mississippi River N:P ratio of 2.3, and by low sea 
level (7065 mm); low recruitment occurred in warm, 
dry seasons. In contrast, Guillory et al. (1983) found 
that elevated juvenile recruitment was associated with 
cold, dry winters and poor recruitment with warm, wet 
winters. Contradictory results between studies may be 
related to 1) differences in data selected for analysis, 
2) gear types and associated differences in life history 
stages and habitats, and 3) spatial and temporal dif- 
ferences in the characterization of meteorological and 
hydrological conditions that drive recruitment of men- 
haden species. 
Data reviewed in the Guillory et al. (1983) and Go- 
voni (1997) studies span what the authors described as 
