7 MARCH 2008
On Phytoplankton Trends
How are phytoplankton at coastal sites around the world
responding to ongoing global change?
Victor Smetacek and James E. Cloern
Phytoplankton-unicellular algae in the
surface layer of lakes and oceans-fuel
the lacustrine and marine food chains
and play a key role in regulating atmospheric
carbon dioxide concentrations. How will ris-
ing carbon dioxide concentrations in the air
and surface ocean in turn affect phytoplank-
ton? Answering this question is crucial for
projecting future climate change. However,
because phytoplankton species populations
appear and disappear within weeks, assessing
change requires high-resolution monitoring
of annual cycles over many years. Such long-
term studies at coastal sites ranging from estu-
aries and harbors to open coastlines and
islands are yielding bewildering variability,
but also fundamental insights on the driving
forces that underlie phytoplankton cycles (1).
An example of regularity is provided by a
45-year data set from weekly phytoplankton
monitoring in Lake Windermere, England,
which shows that the diatom species Asterio-
nella formosa dominates phytoplankton
biomass from autumn to spring but is virtually
absent during summer; this species drives
silicon cycling in the lake (2). In contrast,
weekly data collected in Narragansett Bay in
Rhode Island since 1959 reveal that the phyto-
plankton react with wide fluctuations in com-
position and timing of the annual biomass
peak to driving forces ranging from large-
scale hydrography to the temperature depend-
ence of zooplankton growth (3).
How does phytoplankton performance in
other regions fit along this spectrum from
monotonous regularity to the verge of chaos
(4)? And what will be, or is, the impact of cli-
Early studies of annual cycles of phyto-
plankton (5-7) led to the idea that seasonal
changes in biomass and species could be
attributed to shifts in nutrient and light avail-
ability, with grazing pressure increasing in
importance in the aftermath of the spring
bloom. This annual cycle and its
driving forces came to be accepted
as the rule by ecologists and
modelers, although deviations
were reported (8). However, this
view assumes that phytoplankton
can grow faster than their con-
(viruses), and parasites. This is not supported
by physiological considerations. Many uni-
cellular phytoplankton consumers divide
faster than their prey in the presence of an
adequate food supply. Clearly, the mecha-
nisms keeping consumers and pathogens at
bay also need to be considered.
The example of San Francisco Bay (see the
figure) illustrates how shifting grazing pres-
sure can radically change phytoplankton
annual cycles. For two decades, a spring
diatom bloom was followed each year by low
phytoplankton biomass in the presence of high
nutrient concentrations, a result of strong top-
down control by clams and other bottom-
dwelling suspension feeders during summer
and autumn. This annual cycle changed ab-
ruptly in 1999 with the appearance of an
autumn bloom, a new pattern that has persisted
and led to increases in phytoplankton biomass
and dinoflagellate blooms in an era of reduced
nutrient inputs. These changes were caused by
a shift in the northeast Pacific Ocean from its
warm phase to its cold phase, which led to
massive immigrations of flatfish and crus-
taceans into San Francisco Bay, where they
reduced clam abundance and weakened their
grazing control of phytoplankton growth (9).
Another interesting case is exhibited by
data collected off the island of Heligoland
in the North Sea since 1967. Here, a 1.5°C
rise in winter water temperatures over the
past decades resulted in a delay of the spring
bloom by some weeks. The shift is attrib-
uted to grazing pressure exerted by larger
winter and spring zooplankton populations,
partly due to an earlier appearance of the
plankton-feeding larvae of benthic inverte-
brates (10). Similar shifts have been re-
ported from Narragansett Bay. These obser-
vations imply that temperature rather than
the food supply controls grazing pressure
in the spring.
In some coastal ecosystems, nutrient
(nitrogen and/or phosphorus) concentrations
have declined substantially over the past one
or two decades, indicating clear responses to
improved wastewater treatment and practices
to reduce agricultural runoff. However, in
contrast to lakes and the Black Sea (11),
coherent phytoplankton responses were not
apparent, either as synchronous reductions in
biomass or shifts to communities characteris-
tic of low-nutrient habitats.
Thus, physicochemical environmental
factors (such as temperature, light, and nutri-
ents) set the upper limits to biomass build-up
but do not explain why different phytoplank-
ton groups and species replace each other or
why their annual maxima occur when and
where they do. The fact that most species pro-
liferate for only a few weeks during specific
periods suggests that the timing of their
appearance is regulated by internal as well as
external factors, implying that their life cycles
are selected at longer time scales.
Because most phytoplankton species are
difficult to identify routinely under a light
microscope, it has widely been
assumed that many species are
opportunistic and respond to
favorable growth conditions at
any season. However, detailed
investigations of the “same”
species collected fromdifferent
periods of the year in the Bay of
Naples, Italy, using molecular
tools and backed up by detailed
visual examination and mating
experiments indicate that these comprise dis-
tinct, but cryptic, species (12, 13).
The results suggest that species-specific
life cycle properties have evolved in individ-
ual phytoplankton species in a manner analo-
gous to that in land plants (14). However, it
is unclear how environmental fluctuations
select phytoplankton species by regulating
processes such as sexuality, formation and
germination of resting stages, aggregation
and sinking, and deployment of defense
mechanisms that deter predators and com-
petitors and ward off infection.
The poleward retreat of phytoplankton
species with a minimum temperature require-
ment has been reported, as has the poleward
spread of warm-water species. However, most
species are temperature-tolerant, and whole-
sale poleward encroachment of entire coastal
ecosystems is unlikely. The response of phyto-
plankton to ongoing climate change also
depends on the geomorphology and hydrol-
ogy of the respective site. Thus, seasonal shifts
in snow-melt and rainfall in the catchment
areas change patterns of river discharge,
which in turn affect flushing and nutrient
delivery rates (15) but also disrupt life cycles
of dominant species geared to the former
flushing regime. Basin-scale oscillations in
the adjoining oceanic regime, modified by
global warming, can have similarly drastic
effects on coastal regions (9).
In addition to these effects, the structure of
coastal food webs down to the level of primary
producers will have changed as a result of
severe depletion of commercial fish and
shellfish stocks (16). Drastic changes in com-
munity structure due to removal oftop preda-
tors or herbivores have been shown in many ter-
restrial, lake, and marine benthic ecosystems,
but there are few examples from marine phyto-
plankton (11, 17). It is unlikely that marine
pelagic ecosystems differ fundamentally from
all the others. Long-term monitoring strategies
will thus have to be broadened to encompass
the full range of abiotic and biotic factors that
shape the annual cycles of phytoplankton.
Detailed interdisciplinary studies of the
complex interactions characteristic of coastal
ecosystems need to be undertaken before global
changes obliterate the remnants of the baseline.
1. American Geophysical Union Chapman Conference,
Long-Time Series Observations in Coastal Ecosystems:
Comparative Analyses of Phytoplankton Dynamics on
Regional to Global Scales, Rovinj (Croatia), 8 to 12
2. S. C. Maberly et al. Freshwater Biology31, 19 (1994).
3. T. J. Smayda, ICES J. Mar. Science55, 562 (1998).
4. E. Beninca et al., Nature451, 822 (2008).
5. V. Hensen, Komm. Wiss. Unters. Deutschen MeereV, 12-
16, 1 (1887).
6. E. L. Mills, Biological Oceanography: An Early History
1860-1970(Cornell Univ. Press, Ithaca, NY, 1989).
7. V. Smetacek, Estuaries8, 145 (1985).
8. G. A. Riley, Deep-Sea Res. 3(suppl.), 224 (1955).
9. J. E. Cloern et al., Proc. Natl. Acad. Sci. U.S.A. 104,
10. K. Wiltshire, talk presented at the Chapman Conference
on Long Time-Series Observations in Coastal Ecosystems.
11. A. E. Kideys, Science297, 1482 (2002).
12. A. Amato et al., Protist158, 193 (2007).
13. D. Sarno et al., J. Phycol. 41, 151 (2005).
14. R. A. Cheke, Science318, 577 (2007).
15. I. T. Stewart et al., J. Climate18, 1136 (2005).
16. R. A. Myers, B. Worms, Nature423, 280 (2003).
17. M. L. Pace et al., TREE14, 483 (1998).