Three basic measures were used to monitor the response of the bulk phytoplankton community to the addition of iron: (1) Total chlorophyll concentration, which can increase either because the total number of cells has increased, or because the chlorophyll per cell has increased; (2) Primary productivity, which is a measure of the rate of photosynthesis of the collected phytoplankton assemblage; and (3) Photochemical energy conversion efficiency, a relatively new technique that gives us a ``snapshot'' of the efficiency with which the photosynthetic mechanisms of the phytoplankton are functioning (and by inference, whether or not the cells are iron starved) (Martin et al., 1994; Kolber et al., 1994).
All three measures revealed quite clearly that the
phytoplankton in the iron-enriched patch were stimulated when
compared with the assemblages outside the patch (Figs 5 & 6).
Moreover, the results indicate that all size categories of the
phytoplankton were stimulated by the added iron, and stimulated
within the first 24 hours of iron addition (Fig. 6). This timing
and overall response contrasts strikingly with the results reported
from the bottle incubation experiments described above, in which
the response always appeared to be delayed by several days (Fig. 3)
and was usually dominated by one particular group---namely the
pennate diatoms. Moreover, although the stimulation in the patch
was almost instantaneous, the total increase in chlorophyll was
rather modest, and appeared to level off after the first two days
(Fig. 7a). The total drawdown of CO
in the iron-enriched
patch was only 10% of that which should have been possible had all
the N and P been assimilated (Watson et al. 1994). As Dr. Richard
Barber, the chief scientist on the cruise, so nicely put it,
``Apparently the phytoplankton in the patch hadn't read the
literature.''
Iron-enrichment bottle experiments were done in parallel with
the patch experiment, and some striking differences in the
phytoplankton response were observed (Fig. 7). Over the first
three days of the bottle experiments (Fig. 7b) the control bottles
responded to the same degree as the iron-enriched bottles
(suggesting that it was inadvertently contaminated with iron), and
the magnitude of the response in both treatments was roughly
equivalent to that in the patch (Fig. 7a). On the fourth day,
however, the chlorophyll in the bottles began to ``take off,'' as
has been seen repeatedly in previous experiments (Fig. 3), whereas
the chlorophyll concentrations in the patch leveled off at 0.6 mg
m
. After the fourth day we see explosive growth in the
bottles (Fig. 7c), with chlorophyll in the iron-enriched bottle
increasing over 10-fold, and reaching more than twice the levels in
the control.
One cannot help but wonder if the tremendous growth observed in the bottles by day 6 would have been observed in the patch if it had not been subducted. We suspect not, for several reasons. We know that iron was steadily lost from the patch, and it is likely that the supply ran out before a major bloom could develop. This is supported by the apparent leveling off of chlorophyll concentration in the patch after the second day, and the fact that nitrate concentrations in the water were not depleted significantly. Also, there is reason to suspect that macrozooplankton grazing (which was absent, or significantly reduced, in the bottles) was very active in the patch, repressing the development of higher chlorophyll levels (Martin et al., 1994). The importance of the comparison of the bottle and patch experiments---which is to my mind one of the most significant dimensions of this work---lies in its ability to reveal the degree to which bottle incubations distort our image of how the natural ecosystems function. Much of what we have learned about primary productivity in the sea has come from bottle experiments, thus these comparisons are critical to our understanding the biases in our data.
An expedition is planned for May 1995 in which the IRONEX experiment will be repeated and expanded---continuously fertilizing the patch with iron this time, and carefully examining the role of zooplankton grazing in regulating phytoplankton biomass. We are optimistic that a more complete picture of the dynamics of this ecosystem will emerge with this revised experimental design. The bottom line at this point, however, is that the addition of iron to the equatorial Pacific ecosystem stimulates phytoplankton productivity---thus the basic premise of the iron hypothesis has been validated.