Environmental Engineering Reference
In-Depth Information
Figure 4.
Vertical distributions of (a) total prokaryotes, (b) flagellated protozoa and (c) viral-
like particles determined by epifluorescence microscopy in samples collected on 7 Jul 98 (CAR-
32). Only means presented for clarity's sake. Relative standard error at any given depth is
typically
<
20% of the mean. Broken horizontal line denotes O
2
/H
2
S interface.
variations in viral inventories corresponded most closely with flagellate abun-
dance and BNP in the redoxcline (Table 3), while varying independently of all
other measured variables throughout the water column (Tables 1 and 3). These
trends support the interpretation that rates of free virus and flagellate production
are controlled by prokaryotic production, or in other words, that parasites and
bacterivores proliferate when hosts/prey actively grow.
3.5 Biological Production in the Redoxcline
While depth distributions of BNP always exhibit peaks within the redox-
cline (Fig. 2c), this production typically amounts to a minor fraction (
x
=
19%) of surface BNP (Table 2), and this production is insufficient to sup-
port the enriched inventories of prokaryotes, protozoa and viruses within the
redoxcline. Marine redoxclines are known to support elevated levels of dark
DIC assimilation, attributable to chemoautotrophy [19,21,31,49,50,56,60,61].
In the Cariaco, depth of the 80-100 m thick chemoautotrophic layer varies in
response to the interface's position and most DIC assimilation occurs where O
2
is undetectable (Fig. 7). Like photoautotrophic production in the surface layer,
chemoautotrophic production within the redoxcline has varied about 20-fold
over our observation period (0.16 to 3.33gCm
−
2
d
−
1
;n=16cruises).
