Sanchez-Rubio et al.: Occurrence of pelagic Sargassum in waters of the U S. Gulf of Mexico 
101 
Table 4 
Mean fall (September-November) occurrence, with 95% confidence intervals, of pelagic species of Sargassum 
over the continental shelf across the U.S. Gulf of Mexico and in the western and eastern regions after El 
Nino-Southern Oscillation (ENSO) events from 1982 through 2012. Occurrence represents the percentage 
of ichthyoplankton samples associated with pelagic Sargassum. The numbers in parentheses represent the 
total number of square degree areas (1 degree latitude by 1 degree longitude) that were compared. Numbers 
in brackets represent the number of square degree areas (1 degree latitude, 1 degree longitude) with higher 
percentages of Sargassum species after ENSO events. NS=nonsignificant. 
U.S. Gulf of Mexico 
Cold ENSO 
Neutral ENSO 
Warm ENSO 
Whole 
2.72 ±0.69% (53) 
1.84 ±0.69% (49) 
0.87 ±0.38% (53) 
P,=0.0001 
Western 
Eastern 
2.97 ±1.10% (24) 
2.51 ±0.88% (29) 
p 2 =ns 
2.70 ±1.21% (22) 
1.13 ±0.69% (27) 
P 2 =0.0087 
0.96 ±0.53% (24) 
0.80 ±0.55% (29) 
p 2 =ns 
P,=0.0077 
P ; =0.0022 
Whole 
2.63 ±0.67% (49) [29] 
2.72 ±0.69% (53) [40] 
1.84 ±0.69% (49) [15] 
1.70 ±0.65% (48) [26] 
0.87 ±0.38% (53) [4] 
0.92 ±0.41% (48) [10] 
Pq=0.0075 
p 3 =o 
P 3 =0.0084 
Western 
2.97 ±1.10% (24) [18] 
2.70 ±1.21% (22) [15] 
0.96 ±0.53% (24) [2] 
1.05 ±0.57% (22) [4] 
p 3 =ns 
P 3 =0.0008 
P 3 =0.Q049 
Eastern 
2.14 ±0.77% (27) [18] 
2.51 ±0.88% (29) [22] 
1.13 ±0.69% (27) [6] 
0.80 ±0.55% (29) [2] 
P 3 =0.0055 
P 3 =0.0001 
p 3 =ns 
P r values given by the Kruskal-Wallis H test for the analysis among ENSO events. 
P 2 -values given by the Mann-Whitney U test for the analysis within ENSO events. 
P 3 -values given by the Wilcoxon signed rank test for the analysis between ENSO events. 
Two different decadal regimes of weather-related hy¬ 
drographic characteristics associated with the coupling 
of AMO and NAO phases occurred over the time period 
represented in the SEAMAP samples. Overall percent¬ 
ages of samples with species of Sargassuryi across the 
USGOM under the coupled AMOc/NAOp and coupled 
AMOw/NAOn phases were 2.95% and 7.28%, respec¬ 
tively. Factors affecting the occurrence of Sargassum 
species under both regimes are those that influence 
growth (e.g., temperature, nutrient input), fragmenta¬ 
tion (reproduction) and dispersion (e.g., hurricane and 
eddies), and containment within a region (e.g., wind 
momentum and eddies). 
Productivity in the USGOM is strongly influenced by 
river flows from the Mississippi and Atchafalaya rivers. 
Sanchez-Rubio et al. (2011) found the AMO and NAO 
to be important drivers of climate-related features in¬ 
fluencing long-term hydrological conditions across 
coastal Louisiana and Mississippi. In their study, high 
river discharge occurred under the coupling of AMOc/ 
NAOp phases and lower volumes were associated with 
the coupled AMOw/NAOn phase. Low river discharge 
and a lowered nitrogen:phosphorus ratio (Sanchez- 
Rubio and Perry, 2015), conditions less favorable for 
Sargassum productivity, characterized the regime with 
greater Sargassum species occurrence; this finding in¬ 
dicates that factors associated with transport and re¬ 
tention may have played a greater role than growth 
in facilitating the occurrence of Sargassum species. 
Although physical processes associated with the dry 
regime undoubtedly played a major role in moving and 
maintaining the occurrence of Sargassum species, bio¬ 
logical processes may also have contributed to mainte¬ 
nance of the macroalgae in the western GOM. Lapointe 
et al. (2014) reported that the excretion of ammonium 
and soluble reactive phosphorus by high abundances of 
fishes associated with Sargassum macroalgae (filefishes 
and carangids) provides nutrients that help to sustain 
growth and biomass. They noted that new production 
of Sargassum macroalgae may occur in neritic waters 
of the western Atlantic Ocean and GOM as a result of 
this mutualistic association. Thus, both physical and 
biological processes may facilitate occurrence. 
Hurricane activity was higher in the North Atlan¬ 
tic Ocean, Caribbean region (Goldenberg et al., 2001), 
NERR, and GOM (this study), and the AMM index was 
higher and positive (this study) for the years after the 
mid 1990s climate shift. Vimont and Kossin (2007) 
found that the AMM was strongly related to hurricane 
activity on both decadal and interannual time frames 
and noted that the influence of the AMO on hurricane 
activity was manifested through the AMM. The influ- 
