376
Romeroetal.
nism which could explain the fall maxima at sites
EAl and EA5.
Small, lightly silicified bicapitateNitzschia species (especially N. bicapitata), Thalassionema
nitzschioides var. parva and N. interruptestriata
(Plate I) were consistently present and abundant
throughout the sampling period. It is interesting to
note that over the sampling period, a more diverse
flora was observed at the southernmost site EA5.
A shift in species dominance from N. bicapitata
in spring to Pseudosolenia (= Rhizosolenia) calcar-avis (Plate I) in summer through early fall was
seen. P. calcar-avis is a large, warm water form
which occurs in both oceanic and near-shore samples worldwide (Sundstrom 1986). Previous studies showed that some species of the family
Rhizosoleniaceae adjust their buoyancy and migrate
vertically between surface water and nutrient-rich
water deeper in the water column (Villareal et al.
1993), and may concentrate at oceanic frontal
boundaries separating cold and warm waters
(Yoder et al. 1994), and may sink rapidly through
the water column to the sea floor (Sancetta et al.
1991). We speculate that this may have been the
case at our site EA5 (4 0 S), located in a transition
zone which separates cold, upwelled waters to the
north and warmer waters to the south. The organic
carbon contribution of large-volume cells like
P. calcaravis is noteworthy (Takahashi et al. 1990),
and may explain the higher correlation factor between diatom flux and C o'g flux south of the equator (r = 0.8 at site EA5) than north of it (r = 0.37 at
site EA1).
Western Equatorial Atlantic
360
50
50
WA7
~,
"0
'"
E
~
. .
~
180
~, 25
25 ~
-6
'"
"0
•
'"
E
E
OJ
~t
>
.§..
OJ
(ij
-S
6
x
0
0
0
::::l
x
150
250
350
450
x
<;::
::::l
::::l
]i 360
<;:: 50
WA3
50 '+=
0
(ij
E
le..
0
0
ro
,
I
is
i
I 25
t
t
180
~
25
I
I
0
0
0
Julian Days
150
250
350
450
Boreal
S
Su
F
W
Season
Fig. 5. Western Equatorial Atlantic: Comparison of total (X), opal (diamonds) and diatom (circles) fluxes in the
western equatorial Atlantic (WA 7 and WA3). X-axis as in Fig. 3 (see Table I for trap location).
Romeroetal.
nism which could explain the fall maxima at sites
EAl and EA5.
Small, lightly silicified bicapitateNitzschia species (especially N. bicapitata), Thalassionema
nitzschioides var. parva and N. interruptestriata
(Plate I) were consistently present and abundant
throughout the sampling period. It is interesting to
note that over the sampling period, a more diverse
flora was observed at the southernmost site EA5.
A shift in species dominance from N. bicapitata
in spring to Pseudosolenia (= Rhizosolenia) calcar-avis (Plate I) in summer through early fall was
seen. P. calcar-avis is a large, warm water form
which occurs in both oceanic and near-shore samples worldwide (Sundstrom 1986). Previous studies showed that some species of the family
Rhizosoleniaceae adjust their buoyancy and migrate
vertically between surface water and nutrient-rich
water deeper in the water column (Villareal et al.
1993), and may concentrate at oceanic frontal
boundaries separating cold and warm waters
(Yoder et al. 1994), and may sink rapidly through
the water column to the sea floor (Sancetta et al.
1991). We speculate that this may have been the
case at our site EA5 (4 0 S), located in a transition
zone which separates cold, upwelled waters to the
north and warmer waters to the south. The organic
carbon contribution of large-volume cells like
P. calcaravis is noteworthy (Takahashi et al. 1990),
and may explain the higher correlation factor between diatom flux and C o'g flux south of the equator (r = 0.8 at site EA5) than north of it (r = 0.37 at
site EA1).
Western Equatorial Atlantic
360
50
50
WA7
~,
"0
'"
E
~
. .
~
180
~, 25
25 ~
-6
'"
"0
•
'"
E
E
OJ
~t
>
.§..
OJ
(ij
-S
6
x
0
0
0
::::l
x
150
250
350
450
x
<;::
::::l
::::l
]i 360
<;:: 50
WA3
50 '+=
0
(ij
E
le..
0
0
ro
,
I
is
i
I 25
t
t
180
~
25
I
I
0
0
0
Julian Days
150
250
350
450
Boreal
S
Su
F
W
Season
Fig. 5. Western Equatorial Atlantic: Comparison of total (X), opal (diamonds) and diatom (circles) fluxes in the
western equatorial Atlantic (WA 7 and WA3). X-axis as in Fig. 3 (see Table I for trap location).