Harmful algal blooms: Causes, impacts and detection

Abstract

Blooms of autotrophic algae and some heterotrophic protists are increasingly frequent in coastal waters around the world and are collectively grouped as harmful algal blooms (HABs). Blooms of these organisms are attributed to two primary factors: natural processes such as circulation, upwelling relaxation, and river flow; and, anthropogenic loadings leading to eutrophication. Unfortunately, the latter is commonly assumed to be the primary cause of all blooms, which is not the case in many instances. Moreover, although it is generally acknowledged that occurrences of these phenomena are increasing throughout the world's oceans, the reasons for this apparent increase remain debated and include not only eutrophication but increased observation efforts in coastal zones of the world. There is a rapidly advancing monitoring effort resulting from the perception of increased impacts from these HABs, manifested as expanding routine coastal monitoring programs, rapid development and deployment of new detection methods for individual species, toxins, and toxicities, and expansion of coastal modeling activities towards observational forecasts of bloom landfall and eventually bloom prediction. Together, these many efforts will provide resource managers with the tools needed to develop effective strategies for the management and mitigation of HABs and their frequently devastating impacts on the coastal environment.

Full text

REVIEW PAPER
Harmful algal blooms: causes, impacts and detection
Received: 5 February 2003 / Accepted: 5 May 2003 / Published online: 30 July 2003
 Society for Industrial Microbiology 2003
Abstract Blooms of autotrophic algae and some het-
erotrophic protists are increasingly frequent in coastal
waters around the world and are collectively grouped as
harmful algal blooms (HABs). Blooms of these organ-
isms are attributed to two primary factors: natural
processes such as circulation, upwelling relaxation, and
river flow; and, anthropogenic loadings leading to
eutrophication. Unfortunately, the latter is commonly
assumed to be the primary cause of all blooms, which is
not the case in many instances. Moreover, although it is
generally acknowledged that occurrences of these phe-
nomena are increasing throughout the world’s ocean s,
the reasons for this apparent increase remain debated
and include not only eutrophication but increased
observation efforts in coastal zones of the world. There
is a rapidly advancing monitoring effort resulting from
the perception of increased impacts from these HABs,
manifested as expanding routine coastal monitoring
programs, rapid development and deployment of new
detection methods for individual species, toxins, and
toxicities, and expansion of coastal modeling activities
towards observational forecasts of bloom landfall and
eventually bloom prediction. Together, these many
efforts will provide resource managers with the tools
needed to develop effective strategies for the manage-
ment and mitigation of HABs and their frequently
devastating impacts on the coas tal environment.
Keywords Harmful algal blooms Æ Detection Æ
Molecular techniques Æ Remote sensing Æ Modeling
Introduction
Large accumulations of phytoplankton, macroalgae
and, occasionally, colorless heterotrophic protists are
increasingly reported throughout the coastal areas of all
continents. Aggregations of these organisms can dis-
color the water giving rise to red, mahogany, brown, or
green tides, can float on the surface in scums, cover
beaches with biomass or exudates (foam), and deplete
oxygen levels through excessive respiration or decom-
position. Alternatively, certain species in harmful algal
blooms (HABs) can exert their effects through the syn-
thesis of compounds (e.g., toxins) that can alter cellular
process of other organisms from plankton to humans.
The most severe, and therefore memorable, effects of
HABs include fish, bird, and mammal (including
human) mortalities, respiratory or digestive tract prob-
lems, memory loss, seizures, lesions an d skin irritation,
as well as losses of coastal resources such as submerged
aquatic vegetation and benthic epi- and in-fauna.
For certain toxin producing species, significant
impacts occur at population densities of only several
hundred cells per liter. For example, Dinophysis need
only be present at 100 s of cells l
)1
to induce diarrhetic
symptoms, as they are concentrated by shellfish and then
ingested by human consumers. Pfiesteria piscicida and
P. shumwayiae are associated with fish lesions, skin and
eye irritation, and short-term neurocognitive disorders
[59], and need only reach levels of 250 zoospores l
)1
to
be of concern. Toxin-producing species are found in
other groups besides the dinoflagellates, including
raphidophytes, diatoms, cyanobacteria, and several
J Ind Microbiol Biotechnol (2003) 30: 383–406
DOI 10.1007/s10295-003-0074-9
Kevin G. Sellner Æ Gregory J. Doucette
Gary J. Kirkp atrick
K. G. Sellner (&)
Chesapeake Research Consortium,
645 Contees Wharf Road, Edgewater,
MD 21037, USA
E-mail: sellnerk@si.edu
Tel.: +1-410-7981283
Fax: +1-410-7980816
G. J. Doucette
Marine Biotoxins Program,
Center for Coastal Environmental
Health and Biomolecular Research,
NOAA/National Ocean Service,
219 Fort Johnson RoadCharleston,
SC 29412, USA
G. J. Kirkpatrick
Mote Marine Laboratory,
1600 Ken Thompson Parkway,
Sarasota, FL 34236, USA

other groups with fewer toxic representatives (e.g., pry-
mnesiophytes). The primary groupings of HAB toxins
according to syndrome include paralytic shellfish poi-
sons (PSP), neurotoxic shellfish poisons (NSP), amnesic
shellfish poisons (ASP), diarrhetic shellfish poisons
(DSP), azaspiracid shellfish poisoning (AZP), ciguatera
fish poison ing (CFP), and cyanobacteria toxin poisoning
(CTP). Represent ed in this diverse group are neurotox-
ins, carcinogens, and a number of other compounds,
chemistries (e.g., free radical formation), and sympto-
mologies that affect living resources or humans exposed
to the caus ative organisms or to their toxins following
concentration by filter-feeding bivalves or planktivorous
fish. Several recent reviews provide a detailed treatment
of the range of algal toxins and their effects [29,75,179].
Reasons for the increasing interest in HABs include
not only public safety concerns associated with pro-
tecting human health, but also adverse effects on living
resources of many coastal systems, economic losses
attributed to reduced tourism, recreation, or seafood
related industries, and costs required to maintain public
advisory services and monitoring programs for shellfish
toxins, water quality, and plankton composition. A
recent study [65] estimated approximately US $49 mil-
lion was lost annually to HAB-related impacts in the
United States over a 5 year study period (1987–1992).
Further, many areas ideal for establishing productive
and profitable wild shellfisheries (e.g., Alaska and
Georges Bank) remain closed year-round due to per-
sistent toxicity of the resource resulting from repeated
toxin exposure and/or an inability to depurate accu-
mulated toxin from the contaminated shellfish. The
potential production for an Alaskan shellfishery has
been estimated at US $50 million annually, a consid-
erable economic benefit that cannot be realized. Simi-
larly, the Georges Bank surf clam fishery has been
closed since 1989 due to continuing PSP toxicity and a
United States roe-on scallop industry for this area has
consequently not been developed.
A global response to anthropogenic loadings?
There is no doubt that HABs are occurring in more
locations than ever before (Fig. 1) and new sightings are
reported regularly. Several researchers have argued that
this trend is due to increasi ng eutrophication throughout
the world [152] and there are several classic examples
relating HAB frequency to anthropogenic activities. For
example, red tides in Tolo Harbor, Hong Kong, showed
a remarkable increase from 1976 to 1986 that strongly
paralleled a local rise in population density (Fig. 2a);
nitrogen (N) and phosphorus (P) also increased 25- and
6-fold, respectively, over the same period [74]. In a
similar time series, population and industrial develop-
ment in the Seto Inland Sea region of Japan [103] for the
period 1968–1976 coincided wi th about a 6-fold increase
in the number of HABs each year, to a maximum of 300,
while N and P increased more than 30- and 5-fold,
respectively (Fig. 2b). Implementation of better man-
agement practices in the early 1970s that reduced
chemical oxygen demand (COD) has resulted in a sub-
stantial decline in bloom frequency to one-half the
maximum number, N concentrations only 13-fold higher
than initially (i.e., 4 lM), and P at pre-development
levels. In another example, increased nutrient loading to
the southern North Sea since World War II has resulted
in prolonged spring phytoplankton maxima, with the
colonial bloom-forming alga Phaeocystis following the
initial diatom bloom. This successional shift is attributed
to silicon-limitation leading to collapse of the diatom
assemblage opening a niche for Phaeocystis [134]. With
overall elevated nutrient levels and colony formation
promoted under nitrate replete conditions [126], this
haptophyte, considered a sub-optimal food source for
zooplankton grazers, can dominate surface waters.
Elevated nutrient loading has also bee n proposed as
the primary reason for increa sing HAB s in a number of
other systems. Low salinity coastal waters throughout
the world are experiencing substantial increases in
halotolerant cyanobacteria in response to elevated
nutrient loading stemming from human activities. For
example, coastal embayments in Brazil are highly
enriched and support large Microcystis aeruginosa pop-
ulations [44]. This potentially toxic species is also
increasing in headwaters of the Chesapeake Bay
[P. Tango, personal communication]. Pseudo-nitzschia,
the diatom genus responsible for domoic acid produc-
tion and amnesic shellfish poisoning, has increased
dramatically over the last 50 years along the Louisiana
coast, strongly correlated with nitrate loading in the
Mississippi River (Fig. 3; see [108, 170]). Very recently,
it has been argued that precipitation-driven coastal
runoff and associated anthropogenic nutrient loads may
be responsible for Pseudo-nitzschia blooms off central
California, and that upwelling might not be as impor-
tant as is often suggested [73]. Mass introduction of
nutrients from North Carolina hog farm waste ponds
resulted in an approximate 6-fold increase in Pfiesteria
piscicida zoospores (from Fig. 12 in [26]), mimicking
laboratory observations of zoospore responses to inor-
ganic and organic enrichment. Elevated nutrient inputs
into the Northern Adriatic, Aegean, and Black Sea have
resulted in increasing frequencies of HAB-related events
[23, 98, 148]. Nutrient inputs accompanied by harbor
construction have led to longer residence times and
recirculating embayment gyres, maintaining HABs for
extended periods in small harbor s of southern Spain
[42]. Cyanobacterial pigments, likely from the summer
dominant taxa Nodulari a and Aphanizomenon, dramat-
ically increased in Baltic Sea sediments in the 1960s
(Fig. 4), after relatively low levels for the previous
7,000 years [113]. As anthropogenic loading was mini-
mal several thousand years ago, these data suggest
cyanobacterial abundance may have increased in
response to post World War activities (i.e., industrial
expansion and nutrient loading), previously documented
as a period for rapid increases in total N and P [77].
384

Nodularia was also a primary contributor to the nutri-
ent-rich Peel-Harvey estuary in Australia [85], prior to
river diversion and salinity increases that severely lim-
ited the growth of this cyanobacterium. Macroalgae
respond similarly to these environmental changes.
Increasing macroalgae in shallow, highly nutrient-en-
riched New England estuaries is well documented,
leading to losses of submerged aquatic vegetation
[64,177]. For oceanic environments, LaPointe [76] sum-
marized the recent literature to conclude that increases
in ambient nutrient levels in nutrient-limited reef envi-
ronments favor frondose macroalgal growth and, in
some cases, overgrowth of corals and coral mortality. It
should be noted, however, that this view is not universal,
as Szmant [161] concluded that overgrowth and coral
degradation are likely due to a number of factors besides
nutrient enrichment. Whether nutrient-driven or not,
increasing macroalgae could also favor growth of at-
tached benthic dinoflagellates responsible for ciguatera
poisonings [21].
Increasing input s of N and P are often associated
with declining silicon contributions due to a number of
river management strategies such as dam construction
[67], leading to lower available silico n for diatom
production and greater contributions of non-silicon-
requiring species like Phaeocystis, dinoflagellates, and
prymnesiophytes. These authors have presented an
excellent summary of such impacts for the Black and
Baltic Seas, while Anderson et al. [13] have also reviewed
the effects of altered N/P and P/Si ratios in a number of
other systems.
There is increasing discussion on the potential role of
aquaculture and mariculture in HAB development.
Cultured shellfish and finfish populations produce huge
amounts of feces, pseudofeces, and other excretory
products rich in N and P important to algal growth.
Particulates, either defecated materials or uneaten fish
foods (only 30% of added fish food is harvested as fish
biomass, see [47]), settle to the bottom and, through
remineralization, yield soluble N (either oxidized nitrate,
Fig. 1 Global distribution of
harmful algal bloom (HAB)
toxins and toxicities (from
[178])
385

nitrite, or reduced ammonium) and P. If the system is
highly flushed, potential utilization by autotrophs in
overlying waters is likely limited [121]. However, several
shallow, poorly flushed areas with extensive mariculture
operations report increasi ng and unique HABs not
observed prior to introduction of the managed fish
stocks [188], suggestive of culture operation -induced
HAB expansion. Several HABs in Spanish rias have also
been attributed to growth resulting from utilization of
remineralized detritus settled from bivalve rope cultures
[128]; however, the initial nutrient input was from
upwelling sources and hence these HAB events represent
a combined effect of natural and anthropogenic
processes.
These exam ples suggest that HAB occurrence is
strongly associated with human activities in coastal
zones. Moreover, those cases outlined above represent
only a few of the numerous citations available, and
articles by Anderson [5], Smayda [152], Van Dolah [178],
and Anderson et al. [13] should be consulted for addi-
tional locations and discussions.
Natural processes are equally important
The preceding discussion would lead one to believe that
human activities and the associated increase in nutrient
loadings are likely the primary reason for HABs
occurring in our world’s oceans. In fact, this is not the
case, and the scientific community has a respon sibility to
indicate the importance of natural events in bloom for-
mation. Oceanic an d estuarine circulation and river flow
greatly influence the abundance and distribu tion of
plankton and the combined physical (e.g., currents,
upwelling, etc.) -chemical (e.g., salinity, nutrients, etc.)
factors of these systems, coupled with unique life cycles
and behaviors of some HAB taxa, result in blooms that
impact coastal ecosystems and populations. There are
numerous examples indicating the importance of these
processes.
As discussed above, Pseudo-nitzschia spp. off Loui-
siana are apparently increasi ng as a direct result of
nitrate delivery from the Mississippi River watershed, a
phytoplankton response to non-point source introduc-
tion of fertilizers. In several other areas with recurrent
Pseudo-nitzschia and domoic acid-related problems,
however, nutrient supply is a natural process, involving:
(1) storm-associated forcing as either rain-induced river
discharge or wind-induced mixing of deep nitrate poo ls
into surface waters as in Prince Edward Island in 1988
[153] and Puget Sound in 1997 [167]; (2) wind-induced
coastal upwelling off California and Washington states,
and the Iberian peninsula [2, 36, 143, 168]; (3) physical
transport and deposition of healthy phytoplankton
populations into the United States Pacific Northwest
[169]; and (4) physically-controlled thin layer formation
and maintenance as in East Sound, Wash. and Monterey
Bay, Calif. [40, 127].
Basin scale circulation, and not anthropogenic forces,
also act as effective vectors for distributing bloom taxa,
leading to coastal blooms and adverse impacts on living
resources, and in some HAB species, unique character-
istics of the life cycle combine with regional physics
enabling successful proliferation. Alexandrium spp. in
the Gulf of Maine (Fig. 5) are transported from the Bay
of Fundy along the New England coast in two separate
coastal currents, the Eastern and Western Maine
Coastal Currents, part of the Gulf of Maine circulation
[8]. Shellfish intoxication is due primarily to introduc-
tion of these populations during downwelling favorable
wind conditions, followed by southwesterly alongshore
transport. Add itionally, introduction of the PSP-
producing Alexandrium population into the Gulf of
Maine in 1972 was driven by meteorology: a hurricane
brought coastal Nova Scotian populations south across
the Scot ian shelf into the northern Gulf of Maine [6].
The life cycle of Alexandrium spp. also aids in the
successful establishment of bloom populations. This
dinoflagellate produces a resting stage—a cyst—during
periods of suboptimal growth conditions. The cyst sinks
to the bottom and, after an obligate dormancy period,
Fig. 2a, b Anthropogenic-induced HAB in coastal systems. a Tolo
Harbor, Hong Kong [74]; bars population, line HABs per year.
b Seto Inland Sea, Japan [103]; arrow implementation of waste
reduction practices for the coastal discharges
386

can excyst (break open) to release vegetative cells that
swim to the surface to re-seed bloom populations
(Fig. 6). This resting stage therefore provides this din o-
flagellate with a unique competitive advantage over
populations that cannot persist under poor conditions,
and migration into surface circulation cells ensures
transport throughout a region. Some other dinoflagel-
lates and flagellates can produce cysts, and resting stages
are also documented for some diatoms (spores) and
cyanobacteria (akinetes), ensuring re-introduction of
vegetative populations into overlying waters of high
irradiance for potential growth, accumulation, and
blooms.
Development and transport in other large current
systems independent of anthropogenic contributions is
also characteristic of several other bloom species. The
N-fixing cyanobacteria, Trichodesmium spp., are car-
ried throughout the tropical and subtropical latitudes
in oligotrophic systems with the potential for bloom
development in at least one locale, the west Florida
shelf, a function of wind-delivered iron-rich dust from
the Sahara [186]. The N-fixing cyanobacteria, including
Trichodesmium, grow well in generally N-limited oce-
anic waters due to their ability to fix atmospheric
nitrogen, N
2
. Populations of this taxon in the eastern
Gulf of Mexico may pre-condition Florida shelf waters
Fig. 3a, b Time series of
nutrient and Pseudo-nitzschia in
the Mississippi delta and shelf.
a Nitrate-nitrogen increase at
two locations in the Mississippi
River delta (adapted from
[170]); b densities of Pseudo-
nitzschia spp. in surficial
sediments of five cores collected
on the Mississippi shelf
(adapted from [108])
Fig. 4 Historical record of
cyanobacteria in the Gotland
Deep, Baltic Sea (from [113]).
Filled squares Myxoxanthophyll,
filled diamonds zeaxanthin, filled
triangles echinenone
387

for subsequent blooms of the neurotoxic red tide
organism Karenia brevis, an almost annual bloom-
former along the western coast of Florida. Again,
circulation determines impacts locally and afar: in
1987, K. brevis originating in the eastern Gulf of
Mexico were transported to North Carolina in the
Gulf Stream, devastating local estuaries and associated
industries with total losses estimated at US $25 million
[163, 165 ].
Recurrent introductions of nutrients and several
harmful taxa into nearshore environments are associated
with a number of areas typifie d by upwelling/downwel-
ling regimes. As noted above for Pseudo-nitzschia in
California and Was hington State, the coasts of France,
Spain, and Portugal experience upwelling/downwelling-
induced exposures to several toxic algal species, includ-
ing Dinophysis, Karenia mikimotoi, Alexa ndrium affine,
Gymnodinium catenatum,andLingulodinium polyedrum
[4, 45, 50, 84, 97, 119]. K. mikimotoi is also a common
dominant in upwelling centers along the Benguela region
of the South African coast [112] and Chang [33] has
reported upwelling induced bloom formation and sub-
sequent mass mortalities caused by Gymnodinium brevi-
sulcatum (K. brevisulcatum). On much smaller scales,
wind-induced upwelling in estuaries and coastal bays
can promote growth of HAB species: in the Chesapeake
Bay, wind-induced tilting of the pycnocline (Fig. 7) and
introduction of bottom recycled nutrients resulted in
blooms of mixed dinoflagellate species [147] and
cyanobacteria in summer stratified waters of the Gulf of
Finland [71]. Spring tide-induced destratification also
yields algal blooms, as demonstrated by a large bloom of
Cochlodinium that followed a water column mixing event
in the York River estuary, Chesapeake Bay [60].
Fig. 5 Circulation in the Gulf of
Maine, northeastern United
States. Alexandrium
populations initially entered the
basin in 1972 south of Nova
Scotia and are now found in the
mouth of the Bay of Fundy
(BOF). Populations are
advected south and west in the
eastern Maine coastal current
(EMCC) to (1) encyst just south
of Penobscott Bay and settle to
subsequently excyst and seed
new populations further south
in the western Maine coastal
current (WMCC) or river
mouths south of Penobscott, (2)
occasionally be carried as
vegetative populations from the
EMCC to the WMCC, or (3) be
carried offshore (adapted from
[8, 19])
Fig. 6 The life cycle of Alexandrium, a dinoflagellate with cyst
resting stages (1) that can act as reservoirs for new population
growth (adapted from D.M. Anderson, personal communication).
The resting stages rupture (excyst) to yield swimming cells (2)
which continue to divide to produce a vegetative population (3). As
nutrients are depleted, division slows and gametes are formed that
fuse to form a zygote and then a cyst (4, 5)
388

Wind-driven flows can also introduce oceanic or shelf
populations into coastal embayments, leading to harm-
ful and/or toxic blooms. Advection of shelf populations
of K. mikimo toi, Dinophysis acuta,andD. acuminata into
Irish embayments has been repeatedly observed [89, 102,
117, 118], many rife with mussel culture, leading to
harvesting closures and economic hardship due to con-
tamination of the resource. Moita et al. [97] argue for
northerly winds distributing southerly G. catenatum
populations along the west coast of Portugal in a coastal
current, following wind-driven breakdown of a persis-
tent front off Cape South Vicente. In the north (Galicia),
southerly winds forced warm, offshore waters into the
Ria de Vigo in 1985, leading to a G. catenatum bloom
[49]. Similar wind-driven flooding of Spanish rias has
introduced harmful taxa into these environments,
resulting in toxicity of raft mussels. Gentien et al. [53]
have constructed a simple model indicating that Bay of
Brest Gymnodinium nagasakiense populations can be
readily advected far to the south by strong northwesterly
winds in April, reproducing field observed population
distributions.
Hydrological events, such as rain-induced buoyant
plume formation or delivery of micronutrients, also
favor HAB development. In the Chesapeake Bay Loftus
et al. [82] reported increases of dinoflagellate biomass to
over 300 lg chlorophyll l
)1
following a heavy rainfall,
with the populations aggregating in the thin buoyant
lens of fresher water immediately below the surface.
Mid-coast, inshore red tides along Florida’s west coast
might persist longer due to rainfall and river flow from
central Florida [39], with no reason given for the
expanded durations. Grane
´
li et al. [57] suggested that
selenium and cobalt elution from local soils during
heavy rains might have been partially responsible for
blooms of Chrysochromu lina polylepis in the Skag gerak
and Kattegat.
Other large-scale meteorological events can lead to
bloom formation. El Nin
˜
o driven lower-than-normal sea
temperatures of New Zealand’s northeast coast have
been linked to recurrent spring Mesodinium/Noct iluca
blooms giving way to summer raphidophyte and dino-
flagellate blooms, particularly K. mikimotoi in the latter
group [124]. This succession contrasts with normal years
of spring diatoms, summer dinoflagellates, and diatoms
once more in the fall. PSP and CFP increases in the
Indo-Pacific have also been linked to El Nin
˜
o events
[61, 86]. North Atla ntic Oscillations (NAO) have also
been implicated as drivers for upwelling-induced blooms
along Spain’s coast, generated by increased alongshore
winds developing as a result of greater temperature
differences between the land and sea [48]. In the Galician
coast of northwest Spain, this effect should lead to
increased abundance of G. catenatum, a strong vertical
migrator capable of utilizing deeper remineralized
nutrients from the decomposition of post-bloom sedi-
mented materials. Belgrano et al. [20] have correlated
primary productivity, chlorophyll a, and three Dinoph-
ysis species in a Swedish fjord with the positive phase of
the NAO index (milder, warmer winters, and higher
salinities) in the 1980s and suggest that summer blooms
of C. polylepis and G. aureolum in 1988 were partially
attributable to this decadal phenomenon.
Aggregation at density discontinuities may also be
important in bloom formation and maintenance. Franks
[51] has described in detail the actual process of cell
accumulation at frontal systems. Several HABs in
frontal regions, maintained on the downwelling sides of
fronts through active migration, are well documented.
Holligan [66] described K. mikimotoi accumulations at
the Ushant front in the western English Channel.
Frontal accumulations of Nodularia and Aphanizomenon
have also been described at the mouth of the Gulf of
Finland [71, 96]. In the Chesapeake Bay and its tribu-
taries, at least four species are associated with such
surface fronts, including Prorocentrum minimum, Gyro-
dinium uncatenum, Heterocapsa rotundata, and Gymn-
odinium pseudopalustre [149, 171, 172, 174]. The
importance of frontal accumulations in shellfish toxicity
is exemplified in the Iberian Peninsula, where G. caten-
atum accumulates both at a downwelling front between
the poleward slope current comprised of naked dino-
flagellates, and through migration at the convergence
[46]. The accumulated cells are trapped close to
shore and enter the rias to intoxicate mussel rafts. In
Argentina, Alexandrium tamarense accumulates in
coastal fronts associated with subantarctic waters off
Patagonia and are subsequently transported inshore
during wind reversals [31], thereafter causing serious
PSP toxin contamination of shellfis h in the region.
As noted above for Pseudo-nitzschia, thin layers of
many phytoplankton species including several ha rmful
taxa, may be previously unrecognized recurrent features
of coastal systems. These layers are unique and can be,
but need not be, associated with the pycnocline or
nutricline. Aside from Pseudo-nitzschia, Dinophysis and
Alexandrium have now been observed in these narrow
but horizontally wide features, along the West coast
Fig. 7 Cross-bay tilting of the pycnocline (represented as a line
between gray and black regions) induced by winds shifting from
calm to weak westerly winds (fi) to strong southerly winds („)
along the axis of the Chesapeake Bay. The tilting drives sub-
pycnocline remineralized nutrients into the euphotic zone (white
line 1% light level) leading to phytoplankton blooms (see [147]).
Wind-induced upwelling along western coasts of the major
continents leads to similar nutrient introduction and elevated
surface production, often harmful algal species
389

of the United States (Fig. 8) and in Swedish waters
[40, 58, 127]. The thin layers have some integrity and
appear to remain intact for some time and distance. In
East Sound, Wash., a thin layer of the diatom Ps eudo-
nitzschia persisted as an intense feature along the entire
length of the 12 km fjord for 3 days between wind
events, only to reappear after passage of the winds [127].
Further, a 7 day time series of finescale hourly profiles in
Monterey Bay, Calif. indicated that Pseudo-nitzschia
could form thin layers in open coastal waters that have
similar intensity, thickness, and persistence to those
observed in East Sound [40]. Although the Pseudo-
nitzschia layers in both systems showed little sign of
sinking (e.g., they were associated with a relative narrow
density range throughout the period), the depth at which
the Pseudo-nitzschia layer occurred varied by more than
10 m in response to internal waves in Monterey Bay and
by subduction by the inflow of lighter waters in East
Sound. These vertical shifts in depth were large enough
to radically change light availability and the potential
for contact with the benthos. It is interesting to speculate
on the possible aperiodic role of these layers in seeding
inshore areas and suspension feeders with vegetative
cells, cysts/spores, and toxin, providing episodic and
undetectable seeding for events that have no apparent
seed population, perhaps explaining domoic acid poi-
sonings in razor clams and Dungeness crabs in Oregon
and Washington coasts (observed by Taylo r and Horner
[162]) in October–November 1991.
There is considerable evidence for sub-surface max-
ima of several taxa oc casionally contributing to HABs
and adverse effects (e.g., shellfish toxicity) upon delivery
to inshore areas. Offshore Dinophysis populations in
discrete layers can serve as ‘‘seed’’ for surface blooms/
intoxications in Spain and Sweden, entering into shallow
depths through either upwelling or other wind driven
events (e.g., [80]). A pycnocline-associated K. mikimotoi
population off the Bay of Biscay, France resides there
year-round [84], using remineralised ammonium for
growth in situ. This may be the same population
reported at the seasonal thermocline in the western
English Channel, leading to eventual surface blooms at
the Ushant front [66]. During wi nter-early spring in the
Chesapeake Bay, the dinoflagellate Prorocentrum mini-
mum is carried northwards in thin layer aggregations
just below the pycnocline, resurfacing through occa-
sional destratification events, shoaling at the flanks, and
in association with mixing events at the northern
extreme of the deep trough of the bay, to form the an-
nual spring maximum of this species [173].
The findings discussed above argue strongly for the
dual role of natural processes and anthropogenic forcing
in HAB formation. Bloom events driven by circulation,
Fig. 8 Thin layers of harmful algae observed in East Sound (Wash.)
in August 1997 overlain on vertical density structure. Alexandrium
catenella and Dinophysis acuminata were the dominant net
plankton at the depth of their thin layer (making up 72% and
48% of the net plankton, respectively). In contrast, the thin layer of
Chaetoceros convolutus/concavicornis occurred at the same depth as
the much more abundant Chaetoceros debilis (making up only
0.3% the net plankton). While the concentrations of Chaetoceros
convolutus/concavicornis are low, they are just below the two cells
ml
)1
level reported to cause problems in fish [40]
390

meteorology, or natural nutrient loading (e.g., upw elled
nutrients, river discharge from relatively spa rsely
inhabited regions) will likely occur regardless of human
intervention. In contrast, HAB species that are strongly
influenced by factors derived from human activities that
impact land, water, or air are potentially manageable,
providing the political will is present to commit the
resources needed to manage loads associated with wa-
tershed and coastal develop ment. In either case, the
impact of HABs on coastal communities is significant
and has resulted in efforts to pro-actively reduce the
environmental and public health threat from these
events by enhancing our ability to detect blooms, toxins,
and toxicities.
HAB detection: current and future possibilities
The preceding sections provide an overview of the many
possible causes and effects of HABs in the coastal zone .
As is the case with any natural- or anthropogenic-driven
phenomenon that represents a potential hazard to the
health of humans, wildlife, or ecosystems, effective
management and mitigation strategies are essential for
reducing the hazards associated with HAB events. While
no single approach can address all possible impacts,
timely detection of harmful algal species and the toxins
they produce represents a critical component of most
HAB management plans. Such information, if made
available early in the process of HAB initiation/devel-
opment, can provide coastal resource managers, fisher-
men, aquaculture operators, and public health officials
with the data needed to reco mmend or ta ke actions for
minimizing the effects of HABs. Moreover, organism
and toxin detection capabilities are also critical tools for
researchers studying HAB population and toxin
dynamics, and developing models needed to forecast and
predict these events. The following sectio n describes
some of the current and future approaches to detecting
HAB species as well as their toxins.
Detection of HAB species
The classical approach for detecting and enumerating
phytoplankton species, including those referred to as
harmful and/or toxic, is direct observation by light
microscopy of live or preserved material (see [154] and
chapters therein). Although this technique provides
important visual confirmation of the presence of a spe-
cies in a water samp le and generates reasonably accurate
estimates of cell abund ance, it is generally considered to
be tedious and time-consuming while requiring an
appropriate level of experience/expertise in phyto-
plankton identification. Light microscopy is therefore of
limited use when real-time or near-real-time detection is
the objective. Nonetheless, several volunteer phyto-
plankton monitoring programs have incorporated the
use of portable field microscopes and training focused
exclusively on the recognition of pot ential HAB species
in order to assist coastal managers in the early detection
of possible bloom events in certain areas (e.g., [62]).
An alternative approach to detecting phytoplankton
cells also based on their morphological/optical proper-
ties and relying largely on the princip les of flow
cytometry was developed recently. The instrument,
referred to as the flow cyto meter and microscope
(FLOWCAM, Fig. 9; http://www.fluidimaging.com),
generates data for 12 different intrinsic characteristics
(e.g., size, chlorophyll and phycoerythrin content; note
that this approach does not involve labeling of the cells
in any way), as well as producing a photographic image
for each cell or particle that passes through it. An on-
board image processor can be ‘‘trained’’ to reco gnize
certain cell types, such as those representing potentially
harmful species, and stored images can be accessed at
any time following acquisition in order to confirm
identifications. The FLOWCAM is a portable unit that
can analyze particles ranging in size from 10 to 1,000 lm
(which accounts for the majority of harmful algal spe-
cies), accept either discrete samples or a continuous flow
of up to 10 ml min
)1
, and generate abundance data in
terms of numbers per liter for selected cell types. The
instrument’s ability to operate in continuous (i.e.,
pumped) sampling mode for extended periods on AC
power in a weatherproof enclosure suggests a strong
potential to monitor for the presence of harmful taxa
synoptically at multiple shore-based monitoring sites. A
submersible version of the FLOWCAM that can be
moored temporarily or permanently has recently become
commercially available and should further enhance the
potential HAB monitoring capabilities of this
instrument.
Particle size distributions, ranging from particles
0.7 lm to fish, can be determined with other method-
ologies as well. Gentien et al. [52] have developed a
particle size analyzer for particles from 0.7 to 400 lm
based on diffraction; in a profiler mode, vertical distri-
butions of HAB species in the Baltic and off the French
coast have been determine d. Acoustic profilers are also
available [e.g., Tracor acoustic profiling system (TAPS)],
permitting characteriza tion of macrozooplankton and
larger organisms that may graze or alternatively avoid
accumulations of harmful algae.
The most rapidly growing area of HAB species
detection involves the targetin g of specific molecules,
such as chemical moieties located on the cell surface and
various components of an organism’s genome. These
classes of molecules lend themselves well to detection by
antibody or oligonucleotide probes, respectively, using
methods derived from previously developed biomedical
applications. In the case of cell-surface targets, the most
common approach has employed conventional proto-
cols for the immunization and subsequent boosting of a
host animal (e.g., rabbit, mouse) with chemically-fixed,
whole cells of a given algal species to produce either
polyclonal or monoclonal antibodies (see review by
Vrieling and And erson [182]). The antibodies generated
391

are then screened for reactivity against the target species
as well as a range of closely- and distantly-related phy-
toplankton taxa to confirm the specificity of the recog-
nition. Since the immunogen presented to the host
animal is an uncharacterized mixture of cell surface
antigens displayed by intact cells, rather than a single,
purified compound, the resulting antibodies are
produced against one (monoclonal) or more (polyclonal)
unidentified constituents present on the cell surface at
the time of harvesting and fixation. Such constituents
may include polysaccharides [133], proteins [101], and
lipopolysaccharides [140], or combinations thereof.
Because the composition of cell surface antigens will
vary with an alga’s physiological status, screening of the
antibody should also include testing against the target
species grown under different culture conditions to
confirm similar labeling across a range of cell metabolic
states (e.g., [14, 110]). Once an antibody has been
characterized in the laboratory (e.g., titered and con-
firmed to be specific for the target species, limited by the
cultures available for testing cross-reactivity), field
applications can be developed.
Similarly, lectin-cell surface polysaccharide binding
has been used to detect several harmful taxa and various
cell morphologies associated with different stages in the
life cycles of some dinoflagellates [3, 54, 72]. The lectin, a
non-immunogenic carbohydrate-binding protein, is
generally a natural plant product specifically recognizing
a monosaccharide or simple oligosaccharide and, when
labeled with a fluorescent reporter (e.g., fluorescein iso-
thiocyanate), permits discrimination of specific taxa
based on surf ace carbohydrate composition. Although
lectins are inexpensive and readily available, there are no
reports of lectin-based detection of HAB species
currently being used in the field.
There are two strategies currently being employed for
the detection of harmful algal species with antibodies
and lectins, involving either epifluorescence microscropy
or flow cytometry. In both cases, species-specific anti-
bodies recognizing cell surface antigens are applied to
intact cells in conjunction with a fluorophore-based
reporting system, yielding a fluore scent signal from
target cells labeled with an antibody that can be detected
with appropriate instrumentation. The use of fluores-
cence to detect antibody-antigen reactions is collectively
referred to as immunofluorescence and the application
of techniques based on this approach for phytoplankton
research has been critically reviewed by Vrieling and
Anderson [182]. It should be noted that antibodies
directed against intracellular molecules [e.g., tubulin,
Rubisco, PCNA (proliferating cell nuclear antigen)] in
phytoplankton, including harmful species, have been
Fig. 9 Docktop flow cytometer
and microscope (FLOWCAM)
system developed by Fluid
Imaging Technologies (FIT)
(top left). Flow cytometer
component (top right) and
optical sensor component
(bottom left) are contained in a
weatherproof housing that can
be equipped with wireless
internet access. Sampling of up
to 10 ml min
)1
can be pre-
programmed or triggered based
on real-time fluorescence/
scatter signal and images of
each processed particle (bottom
right) can be obtained. Photos
courtesy of C. Sieracki and
W. Thibaudeau, FIT
392

produced and applied using immunofluorescence-based
detection [79]. Although such antibodies generally rec-
ognize specific, well-characterized prot eins and several
of these antigens can be visualized within the same
sample, their use is aimed more at studies of phyto-
plankton ecology/physiology rather than species identi-
fication. Moreover, development of field applications
has favored antibodies targeting cell surface antigens, a
trend that likely reflects their ease of preparation an d the
lack of requirement for the permeabilization of cells that
would be needed to expose intracellular antigens to an
antibody.
A number of researchers have developed antibodies
against cell surface antigens specific for a wide range of
harmful taxa (reviewed by Anderson [7]). Examples of
algal groups investigated using antibodies include
dinoflagellates (e.g., Alexandrium spp. [1, 138]; Gymn-
odinium spp. [101, 110]; Gyrodinium spp. [183], [185]),
diatoms (e.g., Pseudo-nitzschia spp. [18, 110]), raphido-
phytes (e.g., Chattonella spp. [176]), and pelagophytes
(e.g., Aureococcus anophagefferens [9]). From a HAB
monitoring perspective, both microscopic and flow
cytometric immunofluorescence-based approaches have
been applied or evaluat ed. In the case of the small
(ca. 2 lm diameter), relatively nondescript brown tide
organism, A. anophagefferens, Anderson et al. [11]
reported that cells labeled with a species-specific poly-
clonal antibody could be detected at concent rations as
low as 10–20 cellsÆml
)1
using epifluorescence micros-
copy. This method was used to map the distribution of
A. anophagefferens throughout the coastal waters of the
northeast United States in order to identify areas with a
potential for brown tide outbre aks. The epifluorescence
technique and, most recently, a high throughput (96-well
plate format), enzyme-linked immunosorbant assay
(ELISA) using a monoclonal antibody directed to a cell
surface antigen [30] have been employed by brown tide
monitoring programs conducted in this region. Inter-
estingly, ELISA-based methods for cell detection have
yet to see widespread use, but can be expected to grow in
popularity given their potential to greatly enhance the
speed of analysis and sample throughput while reducing
variability between samples. One notable caveat is the
elimination of a visual confirmation of labeled cell
morphology that is possible with epifluorescence
microscopy-based methods.
Several studies have explored the potential of
immunofluorescence-based, flow cytometric methods for
the detection of natural HAB populations (see review by
Peperzak et al. [111]). One of the first such studies was
reported by Vrieling et al. [183], who found that anti-
body-labeled cells of the ichthyotoxic dinoflagellate,
Gyrodinium aureolum, collected from the North Sea
could be identified via flow cytometry, yet quantification
as compared to light microscope counts was poor due to
loss of cells during sample processing. Other researchers
attempting to quantify Alexandrium spp. in field samples
have also reported problems with quantification result-
ing from cell loss [139]. Thus, while the technique of
immuno-flow cytometry shows promise as an automated
means of detecting antibody-labeled HAB species, issues
related to the loss of cells during staining, and thus poor
quantification of cell concentrations must still be
addressed and have apparently precluded incorporation
of this approach into routine HAB monitoring efforts.
In addition to cell surface antigens, the other class of
target molecules that has been employed for highly
specific detection of HAB taxa is the nucleic acids. In
particular, components of the ribosomal RNA genes
(rDNA) and their transcriptional products, the corre-
sponding ribosomal RNA (rRNA) molecules possess
several characteristics that make these cellular constitu-
ents highly amenable to such applicati ons. Genes coding
for rRNA are present in all living organisms and thus
large, public domain sequence databases are available
(e.g., http://rdp.cme.msu.edu/html/) to facilitate robust
comparisons between newly vs. previously sequenced
taxa. Ribosomal gene sequences contain regions that
range from highly conserved to highly variable, which
allows for the identification of target areas that can
distinguish taxa at various levels, including strains,
species, genera, and increasingly broad phylogenetic
groupings. Moreover, the ribosomes, located in the
cytoplasm and comprised largely of rRNA, represent
easily accessible, generally abundant targets for the
oligonucleotide probes used to bind these molecules.
However, as noted above for phytoplankton cell surface
antigens, rRNA levels can vary as a function of algal
physiological status (e.g., Anderson et al. [12]). It is thus
also imperative that labeling intensities of target species
be compared under a range of both favorable and
unfavorable condition s in the laboratory prior to the
development of field applications.
For detecting harmful algal species, the small (18S)
and large (24S) subunit rRNA molecules have been most
frequently used as the target of oligonucleotide pro-
bes—pieces of synthetic DNA that recognize a given
target sequence within the rRNA molecule. In all cases,
even though a probe is designed to be specific for one or
more algal taxa based on the available sequence data, its
binding must be empirically verified as the target region
may be inaccessible due to folding of the rRNA mole-
cule upon itself. There are two primary approaches for
the use of oligonucleotide probes in the detection of
HAB species. The first is referred to as either whole cell
hybridization (WC) or fluorescence in-situ hybridization
(FISH), in which the probe penetrates into chemically
fixed, intact cells, hybridizes or binds to its target
sequence on the rRNA molecules, and is then visualized
via a fluorescent reporter either attached directly to the
probe or applied during a secondary labeling step.
Similar to immunofluorescence method s described
above, algal cells labeled using FISH protocols can be
examined directly by epifluorescence microscopy or
analyzed using automated methods such as flow
cytometry. Also analogous to algal cell surface antigens,
the abundance of ribosomes within a cell, and thus
labeling intensity, generally varies in proportion to
393

growth rate. The extent to which fluctuations in ribo-
some levels under different growth conditions affect the
labeling of target cells must therefore be investigated
experimentally to aid in interpretation of data from
natural populations (e.g., [14, 111]).
The whole cell hybridization approach has been
developed and applied extensively for the detection of
many harmful algae, including dinofl agellates (e.g., Al-
exandrium spp. [1]; Dinophy sis spp. [120]; Karenia spp.
(C. Mikulski, personal communication, [91]); Pfiesteria
spp. [135]), diatoms (e.g., Pseudo-nitzschia spp. [92, 93,
110, 143]), and raphidophytes (e.g., Heterosigma ak-
ashiwo [175]; Fibrocapsa japonica [175]) (see Fig. 10 for
an example). Perhaps the best example of the use of the
WC technique to monitor harmful algal species has been
reported by workers in New Zealand [123, 124, 125]. In
this case, WC-formatted probes for Alexandrium spp.
and Pseudo-nitzschia spp. have been integrated into the
country’s two-tiered biotoxin monitoring programs for
industry and public health. Probes for additional HAB
species present in New Zealand’s coastal waters,
including the dinoflagellates Karenia spp. and the
raphidophytes Heterosigma and Fibrocapsa , are also
being tested in the WC format to assess their suitability
for inclusion in the country’s phyto plankton monitoring
programs [123, 125]. The probe results for toxic algal
species represent the first tier, which provides risk
assessment information for decision making by shellfish
harvesters, while the second tier involves testing of
shellfish for biotoxin contamination. The laboratory
responsible for condu cting the WC assays (Cawthron
Institute; http://www.cawthron.org.nz/phytoplankton_
lab.htm) is approved by Internat ional Accreditation
New Zealand (recognized under ISO-IEC Guide 25).
The use of WC-formatted probes for routine phyto-
plankton monitoring is also being explored by other
countries with severe problems related to HABs (e.g.,
[34]).
A recently developed technology compatible with the
detection of FISH-labeled microbial cells is laser scan-
ning solid phase cytometry. This approach involves the
filtration and labeling of cells with fluorescently-tagged
rRNA probes followed by scanning of cells on filter
membranes (rather than in solution as for flow cytom-
etry) using laser excitation. Protocols are currently being
developed for use of this semi-automated method in the
detection of HAB species such as Alexandrium minutum
and Pseud o-nitzschia spp., ultimately within the context
of routine HAB monitoring programs [41].
The second approach to applying oligonucleotide
probes for harmful algal species detection is the sand-
wich hybridization (SH) method, which involves chem-
ical lysis of the algal cells to release rRNA target
molecules, foll owed by binding of the target by a species-
specific ‘‘capture’’ probe immobilized to a solid support
(e.g., bead, membrane, etc.), and then hybridization of a
‘‘signal’’ probe to another region of the rRNA. The
latter is res ponsible for visualizing the captured rRNA
using a colorimetric, fluorometric, or chemiluminescent
reporting system and this reaction chemistry can be
configured in a variety of ways (see Fig. 10 for example).
While the SH method precludes the direct microscopic
observation of labeled target cells, this technique allows
for rapid, high throughput sample analysis and has been
effectively automated in a variety of formats. Initial
application of SH assays for the detection of harmful
taxa was reported by Scholin et al. [142, 143] for toxic
diatoms of the genus Pseudo-nit zschia. In this case, the
capture probes were covalently linked to nylon beads
and the signa l probe-based reporting system produced a
colored reaction on the surface of the beads that was
proportional to the number of target cells. These
researchers have since developed an automated labora-
tory method that employs a robo tic sample processor
and has evolved from a system using the nylon beads
described above embedded in so-called ‘‘plastic analyti-
cal cards’’, which suffered from a number of limitations,
to a 96-well plate-based format with the capture probes
immobilized on polystyrene prongs that are moved by a
robotic arm through a pre-programmed series of wells
containing the sample lysate and assay reagents [144].
Note that preparation of the sample by chemically lysing
the algal cells following their capture on a filter is
performed manually.
This rapid, high throughput SH sample processing
technology, including pre-packaged reagents specific
for the colorimetric-based detection of a number of
Fig. 10 Environmental sample processor (ESP; left) developed by
the Monterey Bay Aquarium Research Institute (MBARI) for the
automated, in situ conduct of rRNA sandwich hybridization (SH)
assays and sample archival capabilities for whole cell (WC)
hybridizations. Array photo (top right) shows positive SH response
across triplicate channels for the toxic diatom, Pseudo-nitzschia
australis, while WC image (bottom right) shows corresponding
sample treated with P. australis probe and demonstrating presence
of P. austrailis cells. Photos courtesy of C. Scholin (MBARI)
394

individual HAB species (Pseudo-nitzschia spp., Alex-
andrium spp., and several raphidophytes), is now com-
mercially available (Fig. 11; Saigene, Seattle, Wash.;
http://www.saigene.com/Technology/ahab.htm). Detec-
tion of these same three groups of harmful algae by
automated SH assays is being tested in field trials by
New Zealand researchers, with the aim of incorporatin g
this technique into their existing phytoplankton moni-
toring programs [123]. Other investigators are currently
evaluating this method for use in studies of natural HAB
populations, such as the PSP-producing dinoflagellates
Alexandrium spp. (e.g., [13]) and the domoic acid pro-
ducing diatoms Pseudo-nitzschia spp. (e.g., [107, 144])
(Fig. 10).
In addition to colorimetric-based reporting systems
such as that just described, the binding of an oligonu-
cleotide probe to its rRNA target in solution (i.e., fol-
lowing algal cell lysis) can be detected through the use of
electrochemical methods. This type of reporting system
is the centerpiece for a new generation of devices cur-
rently being developed and aimed at establishing por-
table, field-based detection capabilities for ha rmful algal
species (e.g., [81, 90]). As in the case of the automated
laboratory SH processor outlined above, the species
detected can be changed simply by employing a different
suite of taxo n-specific probes. Sample preparation [e.g.,
chemical-based cell lysis, polymerase chain reaction
(PCR) amplification, etc.] does, however, remain a
manual process. Nonetheless, the benefits of portable
detection units to HAB monitoring programs, especially
within the aquaculture industry and for use on board
vessels, are clear and their use will likely become more
commonplace in the future.
As noted above, use of both antibody and nucleic
acid probes for HAB species detection evolved from
existing biomedical applications . Similarly, the relatively
new and rapidly advancing field of array -based detection
is now being adopted for HAB species. One such
approach involving the use of membrane-based arrays is
aimed at establishing the capability for real-time, in situ
detection of harmful algal species (and their toxins; see
below) on moored platfor ms [146]. In this case, the array
consists of nanoliter volumes of rRNA capture probe
spotted and stabilized onto a membrane. An SH assay,
including harvesting and filtration of the algal cells,
production of a cell lysate, application of the rRNA-
containing lysate to an array, and detection of bound
target molecules via a signal probe/chemiluminescent-
based reporting system, is complete ly automated and
performed autonomously on board a dedicated instru-
ment called the environmental samp le processor (ESP;
Fig. 10). The ESP is deployed as part of a stationary
mooring and signals generated by the SH assay, as well
as ancillary data from the mooring (e.g., salinity, tem-
perature, in vivo fluorescence, etc.), are transmitted in
real time to land- or ship-based laboratories for addi-
tional processing and data manipulation. Archival
capabilities, permitting samples to be tested by WC
hybridization and toxin assay methods upon retrieving
the instrument and returning to the laboratory, have
also been designed into the ESP. Successful trial
deployments of the ESP have been conducted in both
Monterey Bay, Calif. and in Casco Bay, Me. (C. Scholin
et al. personal communication) using arrays configur ed
for both Pseudo-nitzschia spp. and Alexandrium spp.,
allowing simultaneous detection of either group. One
can envision the concurrent deployment of multiple ESP
units configured for detection of several HAB species
providing an advanced, synoptic view of bloom devel-
opment and dissipation within a region as a supplement
to existing monitoring programs.
Finally, a rapidly emerging approach to the detection
of HAB species targeting the unique genetic signatures
of these organisms is the application of PCR-based
methods, which have been used widely for the detection
of other microbes such as bacteria and viruses (e.g.,
[160]). Several studies have used taxon-specific PCR
primers to amplify selected regions of target genes
(e.g., the rRNA gene cluster) of HAB species from a
standard genomic DNA preparation, followed by
detection of the resulting amplicon by various tech-
niques, including gel electrophoresis and staining/blot-
ting protocols (e.g., Alexandrium spp. [109],
Gymnodinium [55], Pfiesteria spp. [136]) and fluorescent
fragment detection (e.g., Pfiesteria [35]). Such PCR-
based methods can be highly sensitive and have been
successfully employed to amplify and detect single veg-
etative cells and cysts of harmful species [24]. More
recently developed techniques, such as quantitative
competitive PCR, real-time PCR, and time step PCR,
have been quickly adopted by the research community
for detecting certain HAB species (e.g., Pfiesteria
[25, 137, 189], Microcystis [105]). While these methods
have the potential to yield quantitative information on
algal cell concentrations, this capability has yet to be
realized for harmful taxa occurring in natural samples;
Fig. 11 Universal processor manufactured by Saigene for the
automated performance of sandwich hybridization assays to detect
HAB species. Samples are loaded into the first row of wells and a
prong strip attached to each of the processor arms carries the
probes and rRNA target through the assay reagents to yield a
colorimetric signal for positive samples in the last row of wells.
Photo courtesy of R. Gordon, Saigene
395

nevertheless, such approaches are reported to be capabl e
of rapid, highly specific and sensitive detection [25]. Of
particular note are recent developments in real-time
PCR that have yielded portable instrumentation suitable
for use in the field (e.g., Cepheid, Sunnyvale, Calif.;
http://www.cepheid.com) that will likely be incorporated
into HAB monitoring efforts and research programs in
the near future.
HAB toxin detection
The toxins produced by harmful algal species include a
broad spectrum of compounds, ranging in size from
several hundred to over 1,000 Da and varying in solu-
bility from highly water-soluble to fat-soluble. The ma-
jor classes of generally well-characterized toxins include
the saxitoxins (PSP), brevetoxins (NSP), domoic acid
(ASP), okadaic acid/dinophysistoxins (DSP), azaspir-
acids (AZP), ciguatoxins (CFP), and microcystins/ana-
toxins/cylindrospermop sin (CTP). In most cases, a toxin
class consists of a family or group of structurally and
functionally related compounds, with individual toxin
derivatives exhibiting an intrinsic toxic potency that can
differ from that of their congeners by over three orders
of magnitude (e.g., PSP toxins [104]). HAB toxins occur
not only in the algal species producing them, but also in
a variety of other organisms throughout aquatic or
marine food webs as a result of trophic transfer pro-
cesses (e.g., Scholin et al. [145]). In the latter case, a toxin
can be metabolized or biotransformed into a structurally
different compound that may be of either higher or
lower toxicity than the original toxin molecule. The
broad chemical and structural diversity of algal toxins,
coupled with differences in intrinsic potency and their
susceptibility to biotransformation, account for many of
the challenges associated with the detection of these
compounds.
Methods used for detecting algal toxins can be
grouped into three main areas, including chemical
analyses and in vitro and in vivo assays and all of these
have been recently reviewed [32, 63]. While in vivo
mammalian bioassays exist for several of the major toxin
groups [43], this approach shows no potential for high
throughput, automated, or in situ application and is
thus outside the scope of this review. In the case of
chemical analyses, the literature is replete with high
performance liquid chromatography (HPLC)-based
methods employing either UV- (including individual or
scanned ranges of wavelengths) or fluorescence-based
detection of either the native toxin molecule or a
chemically derivatized form of the toxin (see chapters
in [63]). Versions of such methods are currently used for
regulatory purposes (e.g., domoic acid [15]) as well as
investigations of toxin production in both the laboratory
(e.g., Anderson et al. [10]) and field (e.g., [153]). More
recently, mass spectrometers have been employed as
detectors coupled to liquid chromatographic separation
methods for the identification of HAB toxins [116]. Mass
spectrometers yield a highly specific, mass-based detec-
tion and, if operated in tandem mode (i.e., LC-MS/MS)
such that the toxin molecule is fragmented to produce a
series of diagnostic daughter ions, this approach can
provide valuable confirmation of a toxin’s presence in
natural samples (e.g., [16]). The latter is especially criti-
cal when dealing with harmful species not previously
known to be toxic in a given region (e.g., Pan et al.
[106]).
Many advances in HPLC instrumentation and col-
umn technology have reduced the time of analysis and
the use of programmed autosamplers has automated the
sample injection process; however, HPLC analyses
remain sequential and thus do not permit the concurrent
injection of multiple samples on a single instrument
required to truly achieve high throughput testing. In the
case of mass spectrometry, technological innovations
have resulted in the development of small, modular
instruments, and researchers at the Center for Ocean
Technology (http://cot.marine.usf.edu/) have success-
fully incorporated these mass spectrometers (Fig. 12)
into autonomous underwater vehicles (AUVs). While
prototype testing of these systems is still underway, the
potential for adapting existing methods for certain algal
toxins, especially those known to occur dissolved in
seawater (e.g., brevetoxin, domoic acid), suggests that
future designs could be configured for the in situ
detection of toxins on board AUVs. Such instruments
would provide mobile, real-time detection capabilities
for HAB toxins that would greatly benefit both moni-
toring and research programs.
The second group of HAB toxin detection methods
comprises in vitro assays and can be divide d broadly
into functional- and structural-based approaches. The
Fig. 12 Underwater quadrupole mass spectrometer contained in an
autonomous underwater vehicle. Developed by the Center for
Ocean Technology (COT), this first deployable version is aimed at
detection and quantification of volatile organic compounds and
dissolved gases. Current efforts include development of capabilities
for tracing both anthropogenic and natural chemicals (e.g., HAB
toxins) using networks of autonomous underwater vehicles
(AUVs). Photo courtesy of D. Fries, COT
396

former rely on detection of a toxin’s biochemical activity
while the latter depend on recognition of chemical
structure at the molecular level. These two categories of
in vitro assays for HAB toxins have been the subject of
several recent reviews [32, 166, 180]. There are advan-
tages and disadvantages to assays based on either the
functional or structural approach. Given that functiona l
assays for a toxin (e.g., receptor binding assays) are
based on binding by its biological receptor, and that the
affinity of this interaction is proportional to the toxin’s
intrinsic potency, the assay response reflects the inte-
grated or net toxicity of all congeners present in a
sample that are bound by the same class of receptor. The
same is true for toxins that have been modified struc-
turally (e.g., biotransformations), providing receptor
recognition remains unaltered. Nonetheless, such assays
cannot be used to identify a toxin(s), only to detect and
measure a particular toxic activity. Structural assays
(e.g., immunoassays) require the conformational inter-
action of a toxin with a recognition factor and are thus
susceptible to any changes to the toxin molecule that
would interfere with this interaction. In the absence of
structural modifications, these assays generally display a
high degree of specificity for the toxin class they were
designed for, yet detection of multiple toxin congeners
depends on the degree to which the assay recognition
factor (e.g., antibody, in the case of ELISAs) cross-
reacts with these various chemical derivatives. Both
functional and structural in vitro assays are susceptible
to non-specific binding of non-target material, which
must be accounted for in the assay de sign and imple-
mentation.
Among the various functional assays developed for
the detection of HAB toxins, including cyto toxicity
assays (e.g., Manger et al. [87]), enzyme inhibition assays
(e.g., Della Loggia et al. [38]), and receptor binding
assays (e.g., Van Dolah et al. [181]), there are cases in
which these tests have been incorporated into existing
HAB toxin monitoring progra ms (e.g., Suarez-Isla et al.
[159]). Yet, some of the same features that make such
assays useful for estimating toxic activity and protecting
the public from consuming contaminated seafood, are
actually impediments to formatting these methods for
in situ toxin detection in HAB species. In particular,
retaining the biological activity of a cell line or a
receptor preparation (required for toxin recognition)
under adverse conditions outside the laboratory remains
an obstacle to the development of in situ functional
assays that has yet to be overcome. Receptor assays for
both domoic acid [115] and the saxi toxins [114] con-
ducted in the laboratory have, however, been used to
test archival samples collected on board the in situ ESP
platform in conjunction with material processed for SH
and WC assays for the associated HAB species (see
above). Such integrated detection of both HAB species
and toxins is essential for the accurate assessment of
HAB-related risks and studies of HAB dynamics, due to
the potentially wide fluctuations in algal toxicity as a
function of physiological status [17]. Nonetheless, toxin
measurements on archival samples must ultimately be
replaced by determinations performed on board in situ
platforms such as the ESP in order to achieve real-time
or near real-time resolution of HAB development and
toxicity.
In comparison to the functional approach, structure-
based assays such as immunoassays are collectively ro-
bust techniques that lend themselves well to use in the
field and likely (in the future) to deployment on in situ
platforms. Antibody-based assay s (e.g., ELISAs) have
been developed for a variety of HAB toxi ns and many of
these tests are now commercially available (see [32,
100]). While most ELISA testing is currently performed
in the laboratory, generally in a high throughput 96-well
plate format, a product distributed by Jellett Biotek
(http://www.jellettbiotek.ca/) called the MIST Alert for
PSP toxins (to be re-issued as the Jellett Rapid Test for
PSP toxins; J. Jellett, personal communication) tests
single samples on a lateral flow immunochromato-
graphic platform similar to that used for home preg-
nancy test kits (Fig. 13). The MIST Alert system
produces qualitative (i.e., positive/negative) results in
less than 20 min and has undergone extensive testing
against the AOAC mouse bioassay, presently the regu-
latory standard for PSP toxin testing [78]. In addition,
this product has also been evaluated successfully for use
with plankton samples (MS submitted, [151]), as has
another version of the MIST Alert for domoic acid (to
be re-issued as the Jellett Rapid Test for ASP toxins)
(MS submitted, [150]). The portability of this system
makes it suitable for the rapid detection of certain algal
toxins in field settings, although the sample throughput
and quantification capabilities are limited.
Fig. 13 MIST Alert kit for paralytic shellfish poisons (PSP) toxins
(to be re-issued as the Jellett Rapid Test for PSP toxins) developed
by Jellett Biotek, showing cassettes with positive (top) and negative
(bottom) test strips. No T line indicates that toxin is present in a
sample, while a visible T line indicates a toxin level below detection
limit. The C line is a control line showing that the reagents have
been sufficiently activated to provide a valid test result. The
regulatory limit for PSP toxins is 80 lg STX equivalents 100 g
)1
.
Photo courtesy of J. Jellett, Jellett Rapid Testing
397

In terms of progress toward the goal of developing in
situ sensors for HAB toxins employing in vitro-type
assays, there are (to the authors’ knowledge) no such
systems curre ntly in place. The most promising strate-
gies for in situ toxin detection (in addition to those noted
above involving mass spectrometry) appear to be those
based on structural recognition of toxin molecules, such
as antibody-based tests . In fact, the configuration of
toxin immunoassays is often analogous to assays for
HAB species detection using oligonucleotide probes (see
above) which have already been formatted for in situ
applications. Work is now underway to develop an
immunoassay-based method for detection of domoic
acid on board the ESP platform described above (G.J.
Doucette et al. unpubli shed data) and other investiga-
tors are pursuing alternative approaches (e.g., toxin
biosensors; [88]) that may also be deployed for remote
detection of HAB toxins in the future.
In concluding this section it must be emphasized that,
in addition to detection methods for HAB species and
toxins, parallel development of technologies for the
collection and concent ration of potentially dilute ana-
lytes such as algal cells and their toxins (especially the
latter), as well as the automation of sample preparation
protocols, are critical. As discussed above, wide fluct u-
ations in the abundance of harmful algal species and
their toxin levels over a variety of temporal/spatial scales
are well documented and can pose a challenge to
obtaining sufficient material for analysis. Thus, in order
to fully realize the potential for in situ detection of HAB
species and toxins to address a range of applications, it is
critical to engage managers, researchers/engineers, and
industry in dialogue to identify the information needs
(type, frequency, etc.), the most appropriate technolo-
gies (organism/toxin detection, sample collection/pro-
cessing, etc.) to obtain these data, and the most efficient
means to manufacture and bring robust, reliable prod-
ucts to market.
Optical Detection
Remote sensi ng for the detection of surface pigment,
reflectance, or temperature signatures has been utilized
for HAB detection for the last 30 years. The large spatial
scale and high frequency of observations provided by
remote sensing makes it appealing as a means of
detecting and assessing HAB features [28, 37, 164].
Steidinger and Haddad [155] demonstrated the utility of
a satellite for detecting HABs using imagery from the
coastal zo ne color scanner (CZCS), and the interaction
of some hydrographic features and algal blooms can be
synoptically assessed for near-surface waters using ocean
color sensors [99]. The advanced very high resolution
radiometer (AVHRR) has been used to find blooms of
phytoplankton that scatter light or that occur in highly
turbid water [27, 56, 69, 157]. The sea-viewing, wide-
field-of-view sensor (SeaWiFS) currently collects global
chlorophyll concentration data on nearly a daily basis.
The United States NOAA CoastWatch program now
acquires and processes SeaWiFS imagery for HAB
monitoring utilizing patterns of chlorophyll anomalies
[158].
Utilizing spectral reflectance (ocean color) and an
ocean color inversion model [131], the phytoplankton
community composition associated with HAB events
has been estimated [130]. Additionally, the model is
sensitive to optical variations within algal groups related
to cell-specific pigment variations making it possible to
assess algal physiology at the same time.
Light absorbance spectra of HABs are maximal in the
blue (and to a lesser extent, the red) portion of the visible
spectrum. Absorbance attributable to accessory pig-
ments is difficult to discern and quantify because chlo-
rophyll a dominates the signal and there is spectral
variance imparted by variation in how pigments are
packaged [68, 94, 95, 132]. The success of absorbance-
based optical techniques to discriminate among distinct
taxonomic groups depends upon the ability of the ap-
proaches to differentiate subtle absorbance characteris-
tics of accessory pigments. A step forward for use of this
approach was the development of microphotometric
measurements of single cell absorbance spectra; these
provided end member spectra for the numerical
decomposition of mixed-species cultures [83]. Modeled
contributions assigned to either species displayed trends
consistent with the actual proportions contributed to the
spectrum by each algal culture. The utility of this ap-
proach for identification of algal taxa depends on the
capabilities for acquiring high-resolution microphoto-
metric data with low signal-to-noise ratios.
Using particulate absorbance spectra from a diverse
range of phytoplankton, noxious bloom-forming dino-
flagellates have been delineated from other algae
through discriminant analysis [68, 94, 129]. Millie et al.
[94] combined fourth-derivative analysis of particulate
absorbance spectra with a similarity algorithm to
discriminate spectra of the Florida (USA) red-tide
dinoflagellate Karenia brevis (Davis) within hypothetical
mixed culture assemblages. When applied in the eastern
Gulf of Mexico, a significant, linear relationship existed
between the derivative spectrum-based similarity index
and the fraction of chlorophyll biomass contributed by
K. brevis [70]. An automated, shipboard HAB detector,
incorporating the aforementioned derivative spectrum-
based similarity index, has refined in situ acquisition of
the required hyperspectral absorbance data for unat-
tended, in situ detection of K. brevis (Fig. 14). This ap-
proach is being adapted to provide detection and
mapping of K. brevis utilizing AUVs.
Platforms and arrays
Sampling the marine environment is difficult due to
large spatial and temporal variations in chemistry,
biology, and physics. Classically, off-pier or shipboard
sampling has yielded single point determinations in
398

time and space and, therefore, limited harmful alga-
specific sampling except for the rarer, obvious HAB
event. With development of the suite of identification
methods for species and toxins described above, there
has been an increasing commitment to transforming
these techniques to in-water capabilities that could be
packaged with a suite of platforms and sampling sys-
tems permitting simultaneous detection of environment,
species, and impact.
Moorings of numerous instrument types, suspended
from fixed structures, floats, or buoys, are now routine
in oceanography and limnology. These arrays often in-
clude autonomous technologies for measuring currents
(ADCPs), temperature, conductivity, and depth sensors
(CTD systems), fluorescence (pigments and colored
dissolved organic matter), optical properties, seston/
turbidity, and several nutrient species (nitrate, ammo-
nium, phosphate, iron) with data stored internally or
transmitted to shore through wireless communications.
Some acoustic procedures are also on-line (e.g., TAPS),
permitting estimations of size distributions of organisms
in a particular water parcel. These standard packages,
fixed in location, have now been refined to permit ver-
tical and spatial sampling, either through tethered
sampling over depth or the release of active or passive
samplers from the fixed moori ng. The ESP described
above is one such moored HAB sensor array that, when
combined with the above technologies, provides ocean-
ographic conditions accompanying the HAB distribu-
tions passing the package.
Another option is moving the package through the
water, rather than sampling the water advected past a
fixed mooring location. Towed sensor packages are most
routine for this type of gear, and generally include CTD,
fluorescence, and occasionally taxa-specific sampling
capacities like sippers for small seston and dissolved
samples, nets for zooplankton, and recording devices
(video, cameras, FLOWCAMs, etc.) for grabbing pic-
tures of suspended material. Hydrowire-deployed or
free-falling, nearly neutrally buoyant sensor packages
have also been developed, permitting fine scale vertical
resolution of a number of water column parameters.
As alluded to above, a promising approach is
attaching sensor packages to robotic undersea vehicles
(RUVs) or AUVs. These tethered or far-ranging plat-
forms often permit sampling at scales and locations not
readily accessible with routine shipboard or moored
sampling. The in-water sampling historically completed
by field oceanographers is now replaced by large area
sensor detection, managed by electronics engineers and
equipment maintenance specialists, with oceanographers
receiving telemetered data for rapid assimilation and
interpretation. Further, aerial and satellite sampling
with ap propriate sensor packages for salinity, tempera-
ture, wave heights, and some pigments ensures rapid
data accumulation over much greater spatial scales than
previously possible; the repeated overflight schedules
eases some temporal limitations and increases sample
number, providing greater sample density than other-
wise attainable.
As sensors for HAB species and toxins move from
the laboratory bench to miniature sensor systems, and
wet chemistries are replaced by chip-based arrays,
deployment of multi-purpose platforms for ecosystem
Fig. 14 Contour map of the
distribution of the Florida red
tide, Karenia brevis, based on
taxa similarity index
determined by the R.V. "Breve
Buster" during the 2001
ECOHAB: Florida Process
Cruise Leg B. Data were
collected on 24 and 25 October
2001. Water was pumped by the
ship’s seawater system from
2 m below the surface. Cell
counts, by microscope
enumeration, were conducted
on board the ship. Cell counts
labeled Pre were collected prior
to the beginning of data
collection by the Breve Buster,
included to highlight the
presence of an offshore patch of
K. brevis detected by the
instrument. The dashed line is
composed of small black dots
indicating where data were
collected by the Breve Buster
399

assessment and HAB detection will becom e routine in
most monitoring programs.
Modeling and forecasting
As described above, advances in optical instrumentation
may provide rapid spatial coverage for HABs and,
coupled with data-assimilative modeling, may provide
the components necessary for building an automated
HAB detection and forecasting system. The desired
information must be isolated and extracted from the
measured bulk optical signals of the observed water
mass. A multi-platform optical approach utilizing re-
mote sensing and in situ mo ored technologies [37, 141] is
promising a capability to provide the near real-time
observations over ecologically relevant spatial and
temporal scales. Data-assimilation methods fuse these
observational networks to optically based models, pro-
viding a capability for detection and forecasting [22].
Recent advances in HAB research ha ve extended well
beyond the detection methods described above. Models
and forecasts of HABs are rapidly advancing so that
there are now well-developed biological models linked
with general circulation models to recreate distributions
of HABs in several environments. This is generally a
difficult task, as the growth of often sparse harmful
species into sizable portions of a mixed phytoplankton
assemblage implies an intimated quantifi able under-
standing of all processes impacting growth rate of the
harmful taxon and its competitive neighbors.
There are several general circulation models for the
Gulf of Maine that have been linked to population-
specific models for the PSP-producing dinoflagellate
Alexandrium, with the most advanced housed at the
Woods Hole Oceanographic Institution [156]. By cou-
pling currents inherent to the basin with river flows,
meteorological forcing, excystment, and net growth
estimations, populations of Alexandrium are formed and
transported in the western portion of the Gulf of Maine.
A comparison of model results and field observations
yields good correlations, expanding prospects for
development of early-warning capabilities for expected
intoxication of shellfish populations inshore and in the
offshore Georges Bank region.
Other efforts are also under development in several
other geographic areas with different HAB taxa. Off
Florida, Walsh and colleagues [187] have been devel-
oping a three dimensional biophysical model for Karenia
brevis, using an approach similar to that in the north-
eastern United States, tying required net growth rates to
regional circulation to ensure bloom densities are esti-
mated for the western Florida shelf. In contrast to the
Gulf of Maine-Alexandrium compl ex, however, nutrients
are quite dilute and there is no known cyst (resting stage)
for K. brevis to ensure reinoculation of surface waters
with vegetative, reproductive populations. This model is
still in development, so practical application is some
time in the future.
A simple model has been devised for the Gulf of
Mexico for projecting or forecasting K. brevis trajectories
and therefore possible landfall site s in the region. A
K. brevis algorithm has been identified for the eastern
Gulf of Mexico, which permits location of the red-tide
organism in SeaWiFS images from the area. Using pre-
dicted wind fields, surface K. brevis populations are
forecast over several days and broadcast as bulletins to a
user community respon sible for monitoring and safe-
guarding public resources in the Gulf [158]. Although
restricted to surface detectable populations, the forecasts
have been remarkably successful, resulting in continued
refinement and wider application around the Gulf of
Mexico.
Other promising applications are near. Coastal
upwelling-wind driven HAB models are well-developed
for France, Spain, Portugal, and South Africa and are
actively used for research in these areas. Transformation
and application as tools for coastal monitoring still re-
main to be completed. Similarly, wind-induced intru-
sions of HAB-rich coastal waters have been modeled in
southwestern Ireland [118] and successes of recently
deployed inexpensive thermistors in Irish bays indicate
that these models may soon be refined as general mete-
orologically- and tidally-forced HAB models with easily
traced temperatures as surrogates for HAB intrusions.
Further, the Gulf of Maine modeling approach for
cross-shelf transport and circula tion is bein g applied to
the west coast of Ireland, in the hope that models
developed in one environment might be applied to other
systems with less investment than required to develop a
completely new model . This cross-system transfer of
models is integral to future international cooperation, as
in programs like GEOHAB (Global Ecology and
Oceanography of Harmful Algal Blooms).
Acknowledgements The authors express their gratitude to a number
of colleagues who provided data, information, comment, photo-
graphic material, access to unpublished results, and figures with
very short notice: S. Azevedo, D. Caron, P. Donaghay, Q. Dortch,
M. Estrada, D. Fries, P. Gentien, R. Gordon, R. Horner, J. Jellett,
B. Keafer, J. Kleindinst, R. Kudela, G. Nolan, E.-L. Poutanen,
N. Rabalais, B. Reguera, J. Rensel, K. Saito, C. Scholin, C. Sie-
racki, R.E. Turner, F. Van Dolah, and J. Walsh.
References
1. Adachi M, Sako Y, Ishida Y (1996) Identification of the toxic
dinoflagellates Alexandrium catenella and A. tamarense (Din-
ophyceae) using DNA probes and whole-cell hybridization.
J Phycol 32:1049–1052
2. Adams NG, Lesoing M, Trainer VL (2000) Environmental
conditions associated with domoic acid in razor clams on the
Washington coast. J Shellfish Res 19:1007–1015
3. Aguilera A, Lo
´
pez-Rodas V, Gonza
´
lez-Gil S, Costas E (1995)
Use of FITC-labelled lectins to identify dinoflagellate species.
In: Lassus P, Arzul G, Erard E, Gentien P, Marcaillou C (eds)
Harmful marine algal blooms. Lavoisier, Paris, pp 707–715
4. Amorim A, Moita MT, Oliveira P (2002) Dinoflagellate
blooms related to coastal upwelling plumes off Portugal.
Abstract, 10th International Conference on Harmful Algae,
St. Pete Beach, Fla., p 12
400

5. Anderson DM (1989) Toxic algal blooms and red tides: a
global perspective. In: Okaichi T, Anderson DM, Nemoto T
(eds) Red tides. Biology, environmental science and toxicol-
ogy. Elsevier, New York, pp 11–16
6. Anderson DM (1994) Red tides. Sci Am 271:52–58
7. Anderson DM (1995) Identification of harmful algal species
using molecular probes: an emerging perspective. In: Lassus P,
Arzul G, Erard E, Gentien P, Marcaillou C (eds) Harmful
marine algal blooms. Lavoisier, Paris, pp 3–13
8. Anderson DM (1997) Bloom dynamics of toxic Alexandrium
species in the northeastern U.S. Limnol Oceanogr 42:1009–
1022
9. Anderson DM, Kulis D, Cosper EM (1990) Immunofluores-
cent detection of the brown tide organism Aureococcus ano-
phagefferens. In: Cosper EM, Bricelj VM, Carpenter EJ (eds)
Novel phytoplankton blooms: causes and impacts of recurrent
brown tides and other unusual blooms. Springer, Berlin Hei-
delberg New York, pp 213–228
10. Anderson DM, Kulis DM, Keafer BA, Sullivan JJ, Hall S,
Lee C (1990) Dynamics and physiology of saxitoxin produc-
tion by the dinoflagellates Alexandrium spp. Mar Biol
104:511–524
11. Anderson DM, Keafer BA, Kulis DM, Waters RM, Nuzzi R
(1993) An immunofluorescent survey of the brown tide
chrysophyte Aureococcus anophagefferens along the northeast
coast of the United States. J Plankton Res 15:563–580
12. Anderson DM, Kulis DM, Keafer BA, Berdalet E (1999)
Detection of the toxic dinoflagellate Alexandrium fundyense
(Dinophyceae) with oligonucleotide and antibody probes:
variability in labeling intensity with physiological condition.
J Phycol 35:870–883
13. Anderson DM, Glibert PM, Burkholder JM (2002) Harmful
algal blooms and eutrophication: nutrient sources, composi-
tion, and consequences. Estuaries 25: 704–726
14. Anderson DM, Keafer BA, Kulis DM, Connell L, Scholin CA
(2002) Application of molecular probes in studies of Alex-
andrium in the Gulf of Maine: success and problem areas.
Abstract, 10th International Conference on Harmful Algae,
St. Pete Beach, Fla., p 14
15. AOAC (1991) Domoic acid in mussels, liquid chromato-
graphic method, first action (1991). Official methods of anal-
ysis. Association of Official Analytical Chemists, Washington,
D.C., section 991.26
16. Bargu S, Powell CL, Coale SL, Busman M, Doucette GJ,
Silver MJ (2002) Krill: a potential vector for domoic acid in
marine food webs. Mar Ecol Prog Ser 237:209–216
17. Bates SS (1998) Ecophysiology and metabolism of ASP toxin
production. In: Anderson DM, Cembella AD, Hallegraeff
GM (eds) Physiological ecology of harmful algal blooms.
NATO ASI series, vol G41, Springer, Berlin Heidelberg New
York, pp 405–426
18. Bates SS, Leger C, Keafer BA, Anderson DM (1993) Dis-
crimination between domoic-acid-producing and non-toxic
forms of the diatom Pseudonitzschia pungens using immuno-
fluorescence. Mar Ecol Prog Ser 100:185–95
19. Beardsley R, Butman B, Geyer R, Smith P (1997) Physical
oceanography of the Gulf of Maine: an update. In: Wallace G,
Braasch E (eds) Proceedings of the Gulf of Maine ecosystem
dynamics scientific symposium and workshop, RARGOM
Report 97-1. Hanover, N.H., pp 39–52
20. Belgrano A, Lindahl O, Hernroth B (1999) North Atlantic
Oscillation primary productivity and toxic phytoplankton in
the Gullmar Fjord, Sweden (1985–1996). Proc R Soc Lond B
266:425–430
21. Bird P, Hewson I, Watkinson A, Dennison W (2000) Effects
of increased nutrients on ciguatera associated dinoflagellates
of the Great Barrier Reef. Abstract, 9th International
Harmful Alagl Blooms Conference, Hobart, Tasmania
22. Bissett WP, Schofield O, Glenn S, Cullen JJ, Miller WL,
Plueddemann AJ, Mobley CD (2001) Resolving the impacts
and feedbacks of ocean optics on upper ocean ecology.
Oceanography 14:30–49
23. Bodenau N (1993) Microbial blooms in the Romanian area of
the Black Sea and contemporary eutrophication conditiones.
In: Smayda TJ, Shimizu Y (eds) Toxic phytoplankton blooms
in the sea., Elsevier, New York, pp 203–209
24. Bolch CJS (2001) PCR protocols for genetic identification of
dinoflagellates directly from single cysts and plankton cells.
Phycologia 40:162–167
25. Bowers HA, Tengs T, Glasgow HB Jr, Burkholder JM,
Rublee PA, Oldach DW (2000) Development of real-time
PCR assays for rapid detection of Pfiesteria piscicida and re-
lated dinoflagellates. Appl Environ Microbiol 66:4641–4648
26. Burkholder JM, Glasgow HB Jr (1997) Pfiesteria piscicida and
other Pfiesteria-like dinoflagellates: behavior, impacts, and
environmental controls. Limnol Oceanogr 42:1052–1075
27. Capone DG, Subramaniam A, Montoya JP, Voss M, Hum-
borg C, Johansen AM, Siefert RL, Carpenter EJ (1998) An
extensive bloom of the N
2
-fixing cyanobacterium Trichodes-
mium erythraeum in the central Arabian Sea. Mar Ecol Prog
Ser 172:281–292
28. Carder KL, Steward RG (1985) A remote-sensing reflectance
model of a red-tide dinoflagellate off west Florida. Limnol
Oceanogr 30:286–298
29. Carmichael WW (2001) Health effects of toxin-producing
cyanobacteria: the CyanoHABs. Human Ecol Risk Assess
7:1393–1407
30. Caron DA, Dennett MR, Moran DM, Schaffner RA, Lons-
dale DJ, Gobler CJ, Nuzzi R, McLean TI (2003) Development
and application of a monoclonal antibody technique for
counting Aureococcus anophagefferens, an alga causing
recurrent brown tides in the mid-Atlantic United States. Appl
Environ Microbiol (in press)
31. Carreto JI, Montoya NG, Colleoni ADC, Akselman R (1998)
Alexandrium tamarense blooms and shellfish toxicity in the
Argentine Sea: a retrospective view. In: Reguera B, Blanco J,
Ferna
´
ndez ML, Wyatt T (eds) Harmful algae. UNESCO,
Vigo, pp 131–134
32. Cembella AD, Doucette GJ, Garthwaite I (2003) In vitro as-
says for phycotoxins. In: Hallegraeff GM, Anderson DM,
Cembella AD (eds) Manual on harmful marine microalgae,
2nd edn. IOC-UNESCO, Paris (in press)
33. Chang FH (1999) Gymnodinium brevisulcatum sp. nov.
(Gymnodiniales, Dinophyceae), a new species isolated during
the 1998 summer toxic bloom in Wellington Harbor, New
Zealand. Phycologia 38:377–384
34. Cho ES, Hur HJ, Byun HS, Lee SG, Rhodes LL, Jeong CS,
Park JG (2002) Monthly monitoring of domoic acid producer
Pseudo-nitzschia multiseries (Hasle) Hasle using species-spe-
cific DNA probes and WGA lectins and abundance of Pseudo-
nitzschia species (Bacillariophyceae) from Chinhae Bay,
Korea. Bot Mar 45:364–372
35. Coyne KJ, Hutchins DA, Hare CE, Cary SC (2001) Assessing
temporal and spatial variability in Pfiesteria piscicida distri-
butions using molecular probing techniques. Aquat Microb
Ecol 24:275–285
36. Cuadrado MMS, Diaz F, Fdez-Riverola F, Corchado JM,
Torres JM (2002) Sea surface temperatures of northwest coast
of the Iberian Peninsula using AVHRR data in order to detect
Pseudo-nitzschia spp. blooms. Abstract, 10th International
Conference on Harmful Algae, St. Pete Beach, Fla., p 63
37. Cullen JJ, Ciotti AM, Davis RF, Lewis MR (1997) Optical
detection and assessment of algal blooms. Limnol Oceanogr
42:1223–1239
38. Della Loggia R, Sosa S, Tubaro A (1999) Methodological
improvement of the protein phosphatase inhibition assay for
the detection of okadaic acid in mussels. Nat Tox 7:387–392
39. Dixon LK, Steidinger KA (2002) Karenia brevis bloom
dynamics in the eastern Gulf of Mexico with respect to rainfall
and riverine flow. Abstract, 10th International Conference on
Harmful Algae, St. Pete Beach, Fla., p 73
40. Donaghay PL, Van Holliday D, Rines JEB, Sullivan JM,
McManus MA(2003) The importance of finescale structure
(thin layers) in controlling the dynamics and biodiversity of
401

plankton populations in stratified coastal waters. In: Kerfoot
C (ed) Limnol Oceanogr special issue (in press)
41. Eller G, Medlin L (2002) Molecular probes for the rapid
detection of toxin marine microalgae. Abstract, 10th Inter-
national Conference on Harmful Algae, St. Pete Beach, Fla.,
p83
42. Estrada M, Camp J (2002) Environmental sustainability: the
case of nutrients and harmful algal blooms in the Catalan
coast (NW Mediterranean). In: Lipiatou E (ed) Thresholds of
environmental sustainability: the case of nutrients. Research
in Enclosed Seas, vol 11. EUR 20170, pp 57–58
43. Fernandez ML, Cembella AD (1995) Mammalian bioassays.
In: Hallegraeff GM, Anderson DM, Cembella AD (eds)
Manual on harmful marine microalgae, IOC-UNESCO,
Paris, pp 213–228
44. Ferra
˜
o-Filho AS, Domingos P, Azevedo SMFO (2002)
Influences of Microcystis aeruginosa Ku
¨
tzing bloom on zoo-
plankton population in Jacarepagua
´
Lagoon (Rio de Janeiro,
Brazil). Limnologica 32:295–308
45. Figueras FG, Wyatt T, Alvarez-Salgado XA, Jenkinson IR
(1995) Advection, diffusion, and patch development in the
Rias Baixas. In: Lassus P, Arzul G, Erard E, Gentien P,
Marcaillou C (eds) Harmful marine algal blooms, Lavoisier,
Paris, pp 579–584
46. Figueiras FG, Alvarez-Salgado XA, Castro CG, Villarino ML
(1998) Accumulation of Gymnodinium catenatum Graham cells
in western Iberian Shelf waters in response to poleward flowing
slope currents. In: Reguera B, Blanco J, Fernandez ML, Wyatt
T (eds) Harmful algae. UNESCO, Vigo, Spain, pp 114–117
47. Folke C, Kautsky N, Troell M (1997) Salmon farming in
context: response to Black et al. J Environ Manage 50:95–103
48. Fraga S, Bakun A (1993) Global climate change and harmful
algal blooms: the example of Gymnodinium catenatum on the
Galician coast. In: Smayda TJ, Shimizu Y (eds) Toxic phy-
toplankton blooms in the sea. Elsevier, New York, pp 59–65
49. Fraga S, Anderson DM, Bravo I, Reguera B, Steidinger KA,
Yentsch CM (1988) Influence of upwelling relaxation in
dinoflagellate and shellfish toxicity in Ria de Vigo, Spain.
Estuarine Coastal Shelf Sci 27:349–361
50. Fraga S, Reguera B, Bravo I (1990) Gymnodinium catenatum
bloom formation in the Spanish Rias. In: Grane
´
li E, Sund-
strom B, Edler L, Anderson D (eds) Toxic marine phyto-
plankton. Elsevier, New York, pp 149–154
51. Franks PJS (1997) Spatial patterns in dense algal blooms.
Limnol Oceanogr 42:1297–1305
52. Gentien P, Lunven M, Le Haıˆ tre M, Duvent JL (1995) In situ
depth profiling of particle sizes. Deep Sea Res 42:1297–1312
53. Gentien P, Lazure P, Raffin B (1998) Effect of meteorological
conditions in spring on the extent of a Gymnodinium cf
nagasakiense bloom. In: Reguera B, Blanco J, Ferna
´
ndez ML,
Wyatt T (eds) Harmful algae, UNESCO, Vigo, pp 200–203
54. Glasgow HB, Springer JJ, Allen CI, Burkholder JM (2002)
Utilization of lectin binding assays to differentiate toxic and
nontoxic isolates of Pfiesteria and other HAB dinoflagellates.
Abstract, 10th International Conference on Harmful Algae,
St. Pete Beach, Fla., p 105
55. Godhe A, Otta SK, Rehnstam-Holm A-S, Karunasagar I,
Karunasagar I (2001) Polymerase chain reaction in detection
of Gymnodinium mikimotoi and Alexandrium minutum in field
samples from southwest India. Mar Biotechnol 3:152–162
56. Gower JFR (1994) Red tide monitoring using AVHRR
HRPT imagery from a local receiver. Remote Sensing Environ
48:309–318
57. Grane
´
li E, Paasche E, Maestrini SY (1993) Three years after
the Chrysochromulina polylepis bloom in Scandinavian waters
in 1988: some conclusions of recent research and monitoring.
In: Smayda TJ, Shimizu Y (eds) Toxic phytoplankton blooms
in the sea. Elsevier, New York, pp 23–32
58. Grane
´
li E, Gisselson L-A, Carlsson P, Pallon J (2002) Din-
ophysis blooms in the deep euphotic zone of the Baltic Sea: do
they grow in the dark? Abstract, 10th International Confer-
ence on Harmful Algae, St. Pete Beach, Fla., p 113
59. Grattan LM, Oldach D, Morris JG (2001) Human health risks
of exposure to Pfiesteria piscicida. Bioscience 51:853–857
60. Haas LW, Hastings SJ, Webb KL (1981) Phytoplankton re-
sponse to a stratification-mixing cycle in the York River
estuary during late summer. In: Neilson BJ, Cronin LE (eds)
Estuaries and nutrients. Humana Press, Clifton, N.J., pp 619–
636
61. Hales S, Weinstein P, Woodward A (1999) Ciguatera (fish
poisoning), El Nin
˜
o, and Pacific sea surface temperatures.
Ecosyst Health 5:20–25
62. Hall S (1999) Volunteer phytoplankton program. In: Martin
JL, Haya K (eds) Proceedings of the Sixth Canadian Work-
shop on Harmful Marine Algae. Can Tech Rep Fish Aquat
Sci 2261, p 30
63. Hallegraeff GM, Anderson DM, Cembella AD (eds) (2003)
Manual on harmful marine microalgae, 2nd edn. Monographs
on Oceanographic Methodology, vol 11. IOC-UNESCO, Paris
64. Hauxwell J, Cebrian J, Furlong C, Valiela I (2000) Macroalgal
canopies contribute to eelgrass (Zostera marina) decline in
temperate estuarine ecosystems. Ecology 82:1007–1022
65. Hoagland P, Anderson DM, Kaoru Y, White AW (2002) The
economic effects of harmful algal blooms in the United States:
estimates, assessment issues, and information needs. Estuaries
25:819–837
66. Holligan PM (1979) Dinoflagellate blooms associated with
tidal fronts around the British Isles. In: Taylor DL, Seliger
HH (eds) Toxic dinoflagellate blooms. Elsevier, New York,
pp 249–256
67. Humborg C, Conley DJ, Rahm L, Wulff F, Cociasu A,
Ittekkot V (2000) Silicon retention in river basins: far-reaching
effects on biogeochemistry and aquatic food webs in coastal
marine environments. Ambio 29:45–50
68. Johnsen G, Samset O, Granskog L, Sakshaug E (1994) In vivo
absorption characteristics in 10 classes of bloom-forming
phytoplankton: taxonomic characteristics and responses to
photoadaptation by means of discriminant and HPLC anal-
ysis. Mar Ecol Prog Ser 105:149–157
69. Kahru M, Horstmann U, Rud O (1994) Satellite detection of
increased cyanobacteria blooms in the Baltic Sea: natural
fluctuation or ecosystem change? Ambio 23:469–472
70. Kirkpatrick GJ, Millie DF, Moline MA, Schofield OM (2000)
Optical discrimination of a phytoplankton species in natural
mixed populations. Limnol Oceanogr 45:467–471
71. Kononen K, Kuparinen J, Ma
¨
kela
¨
K, Laanemets J, Pavelson
J, No
˜
mmann S (1996) Initiation of cyanobacterial blooms in a
frontal region at the entrance to the Gulf of Finland, Baltic
Sea. Limnol Oceanogr 41:98–112
72. Kremp A, Anderson DM (2002) The use of lectins to detect
life cycle stages in two species of cyst forming dinoflagellates.
Abstract, 10th International Conference on Harmful Algae,
St. Pete Beach, Fla., p 156
73. Kudela RM, Cochlan W, Roberts A (2002) Spatial and tem-
poral patterns of Pseudo-nitzschia spp. in central California
related to regional oceanography. Abstract, 10th International
Conference on Harmful Algae, St. Pete Beach, Fla., p 157
74. Lam CWY, Ho KC (1989) Red tides in Tolo Harbor, Hong
Kong. In: Okaichi T, Anderson DM, Nemoto T (eds) Red
tides. biology, environmental science and toxicology. Elsevier,
New York, pp 49–52
75. Landsberg JH (2002) The effects of harmful algal blooms on
aquatic organisms. Rev Fish Sci 10:1–113
76. Lapointe BE (1997) Nutrient thresholds for bottom-up con-
trol of macroalgal blooms on coral reefs in Jamaica and
southeast Florida. Limnol Oceanogr 42:1119–1131
77. Larsson U, Elmgren R, Wulff E (1985) Eutrophication and
the Baltic Sea. Ambio 14:9–14
78. Laycock MV, Jellett JF, Belland ER, Bishop PC, The
´
riault
BL, Russell-Tattrie AL, Quilliam MA, Cembella AD, Rich-
ards RC (2001) Mist Alert: a rapid assay for paralytic
shellfish poisoning toxins. In: Hallegraeff GM, Blackburn SI,
Bolch CJ, Lewis RJ (eds) Harmful algal blooms 2000. IOC-
UNESCO, Paris, pp 254–256
402

79. Lin S, Carpenter EJ (1996) An empirical protocol for whole-
cell immunofluorescence of marine phytoplankton. J Phycol
32:1083–1094
80. Lindahl O (1993) Hydrodynamical processes: a trigger and
source for flagellate blooms along the Skagerrak coasts? In:
Smayda TJ, Shimizu Y (eds) Toxic phytoplankton blooms in
the sea, Elsevier, New York, pp 775–781
81. Litaker W, Sundseth R, Wojciechowski M, Bonaventura C,
Henkens R, Tester P (2000) Electrochemical detection of
DNA or RNA from harmful algal bloom species. In: Halle-
graeff GM, Blackburn SI, Bolch CJ, Lewis RJ (eds) Harmful
algal blooms 2000. IOC-UNESCO, Paris, pp 242–245
82. Loftus ME, Subba Rao DV, Seliger HH (1972) Growth and
dissipation of phytoplankton in Chesapeake Bay. I. Response
to a large pulse of rainfall. Chesapeake Sci 13:282–299
83. Lohrenz SE, Fahnenstiel GL, Kirkpatrick GJ, Carroll CL,
Kelly KA (1999) Microphotometric assessment of spectral
absorption and its potential application for characterization
of harmful algal species. J Phycol 35:1438–1446
84. Loyer S, Gentien P, Lazure P, Menesguen A (2002) Modelling
(3-D) of Karenia mikimotoi in the Bay of Biscay, French coast:
vertical, spatial, and temporal 6-years validation. Abstract,
10th International Conference on Harmful Algae, St. Pete
Beach, Fla., p 176
85. Lukatelich RJ, McComb AJ (1986) Nutrient levels and the
development of diatom and blue-green algal blooms in a
shallow Australian estuary. J Plankton Res 8:597–618
86. Maclean JL (1989) Indo-Pacific red tides, 1985–1988. Mar
Pollut Bull 20:304–310
87. Manger RL, Leja LS, Lee SY, Hungerford JM, Hokama Y,
Dickey RW, Granade HR, Lewis R, Yasumoto T, Wekell MM
(1995) Detection of sodium channel toxins: directed cytotox-
icity assays of purified ciguatoxins, brevetoxins, saxitoxins,
and seafood extracts. J Assoc Off Anal Chem Int 78:521–527
88. Marquette CA, Coulet PR, Blum LJ (1999) Semi-automated
membrane based chemiluminescent immunosensor for flow
injection analysis of okadaic acid in mussels. Anal Chim Acta
398:173–182
89. McMahon T, Raine R, Silke J (1998) Oceanographic control
of harmful phytoplankton blooms around southwestern Ire-
land. In: Reguera B, Blanco J, Ferna
´
ndez ML, Wyatt T (eds)
Harmful algae. UNESCO, Vigo, pp 128–130
90. Medlin LK, Kerkmann K, Huljic S, Eller G, Lange M (2002)
Application of molecular probes for the detection of harmful
algae on DNA-microchips. Abstract, 10th International
Conference on Harmful Algae, St. Pete Beach, Fla., p 193
91. Reference deleted
92. Miller PE, Scholin CA (1998) Identification and enumeration
of cultured and wild Pseudo-nitzschia (Bacillariophyceae)
using species-specific LSU rRNA-targeted fluorescent probes
and filter-based whole cell hybridization. J Phycol 34:371–382
93. Miller PE, Scholin CA (2000) On detection of Pseudo-nitzs-
chia (Bacillariophyceae) species using whole cell hybridization:
sample fixation and stability. J Phycol 36:238–250
94. Millie DF, Schofield OM, Kirkpatrick GJ, Johnsen G, Tester
PA, Vinyard BT (1997) Detection of harmful algal blooms
using photopigments and absorption signatures: a case study
of the Florida red tide, Gymnodinium breve. Limnol Oceanogr
42:1240–1251
95. Millie DF, Schofield OME, Kirkpatrick GJ, Johnsen G,
Evens TJ (2002) Using absorbance and fluorescence spectra to
discriminate microalgae. Euro J Phycol 37:313–322
96. Moisander PH, Rantaja
¨
rvi E, Huttunen M, Kononen K
(1997) Phytoplankton community in relation to salinity fronts
at the entrance to the Gulf of Finland, Baltic Sea. Ophelia
46:187–203
97. Moita MY, da Graca Vilarinho M, Palma AS (1998) On the
variability of Gymnodinium catenatum Graham blooms in
Portugese waters. In: Reguera B, Blanco J, Ferna
´
ndez ML,
Wyatt T (eds) Harmful algae. UNESCO, Vigo, pp 118–121
98. Moncheva S, Gotsis-Skretas O, Pagou K, Krastev A (2001)
Phytoplankton blooms in Black Sea and Mediterranean
coastal ecosystems subjected to anthropogenic eutrophication:
similarities and differences. Estuarine Coastal Shelf Sci
53:281–295
99. Muller-Karger FE, Walsh JJ, Evans RH, Meyers MB (1991)
On the seasonal phytoplankton concentration and sea surface
temperature cycles of the Gulf of Mexico as determined by
satellites. J Geophys Res 96:12645–12665
100. Naar J, Weidner A, Baden D (2002) Competitive ELISA an
accurate, quick and effective tool to monitor brevetoxins in
environmental and biological samples. Abstract, 10th Inter-
national Conference on Harmful Algae, St. Pete Beach, Fla.,
p 210
101. Nagasaki K, Uchida A, Ishida Y (1991) A monoclonal anti-
body which recognizes the cell surface of red tide alga
Gymnodinium nagasakiense. Nippon Suisan Gakkaishi
57:1211–1214
102. O’Boyle S, Nolan G, Raine R (2001) Harmful phytoplankton
events caused by variability in the Irish Coastal Current along
the west of Ireland. In: Hallegraeff GM, Blackburn SI, Bolch
CJ, Lewis RJ (eds) Harmful Algal Blooms 2000, IOC-UNE-
SCO, Paris, pp 143–148
103. Okaichi T (1989) Red tide problems in the Seto Inland Sea,
Japan. In: Okaichi T, Anderson DM, Nemoto T (eds) Red
tides. Biology, environmental science and toxicology. Elsevier,
New York, pp 137–144
104. Oshima Y, Blackburn SI, Hallegraeff GM (1993) Compara-
tive study on paralytic shellfish toxin profiles of the dinofla-
gellate Gymnodinium catenatum from three different countries.
Mar Biol 116:471–476
105. Ouellette AJA, Boyer GL, Wilhelm SW (2002) Quantitative
PCR and sequence analysis for determination of microbial
community structure and the detection of toxic Microcystis in
Lake Erie. Abstract, 10th International Conference on
Harmful Algae, St. Pete Beach, Fla., p 222
106. Pan Y, Parsons ML, Busman M, Moeller PDR, Dortch Q,
Powell CL, Doucette GJ (2001) Pseudo-nitzschia sp. cf.
pseudodelicatissima—a confirmed producer of domoic acid
from the northern Gulf of Mexico. Mar Ecol Prog Ser
220:83–92
107. Parsons ML, Scholin CA, Miller PE, Doucette GJ, Powell
CL, Fryxell GA, Dortch Q, Soniat T (1999) Pseudo-nitzschia
species (Bacillariophyceae) in Louisiana coastal waters:
molecular probe field trials, genetic variability, and domoic
acid analyses. J Phycol 35:1368–1378
108. Parsons ML, Dortch Q, Turner RE (2002) Sedimentological
evidence of an increase in Pseudo-nitzschia (Bacillariophyceae)
abundance in response to coastal eutrophication. Limnol
Oceanogr 47:551–558
109. Penna A, Magnani M (1999) Identification of Alexandrium
(Dinophyceae) species using PCR and rDNA-targeted probes.
J Phycol 35:615–621
110. Peperzak L, Sandee B, Scholin C, Miller P, van Nieuwerburgh
L (2000) Application and flow cytometric detection of anti-
body and rRNA probes to Gymnodinium mikimotoi (Dino-
phyceae) and Pseudo-nitzschia multiseries (Bacillariophyceae).
In: Hallegraeff GM, Blackburn SI, Bolch CJ, Lewis RJ (eds)
Harmful algal blooms 2000. IOC-UNESCO, Paris, pp 206–
209
111. Peperzak L, Vrieling EG, Sandee B, Rutten T (2000) Immuno
flow cytometry in marine phytoplankton research. Sci Mar
64:165–181
112. Pitcher GC, Roesler CS, Nelson G (2002) The importance of
surface boundary layer characteristics and advection in the
development of red tide in the southern Benguela upwelling
system. Abstract, 10th International Conference on Harmful
Algae, St. Pete Beach, Fla., p 232
113. Poutanen E-L, Nikkila
¨
K (2001) Carotenoid pigments as
tracers of cyanobacterial blooms in recent and post-glacial
sediments of the Baltic Sea. Ambio 30:179–183
114. Powell CL, Doucette GJ (1999) A receptor binding assay for
paralytic shellfish poisoning toxins: recent advances and
applications. Nat Tox 7:393–400
403

115. Powell CL, Ferdin ME, Busman M, Kvitek RG, Doucette GJ
(2002) Development of a protocol for determination of do-
moic acid in the sand crab (Emerita analoga): a possible new
indicator species. Toxicon 40:481–488
116. Quilliam MA (1996) Liquid chromatography-mass spec-
trometry of seafood toxins. In: Barcelo D (ed) Applications of
LC-MS in environmental science. Elsevier, Amsterdam,
pp 415–444
117. Raine R, Joyce B, Richard J, Pazos Y, Moloney M, Jones K,
Patching JW (1993) The development of a bloom of the
dinoflagellate Gyrodinium aureolum (Hulbert) on the south-
west Irish coast. ICES J Mar Sci 50:461–469
118. Raine R, Nolan G, Chamberlain T, McDermott G, Silke J
(2002) Monitoring and forecasting harmful phytoplankton
events in southwestern Ireland. Abstract, 10th International
Conference on Harmful Algae, St. Pete Beach, Fla., p 239
119. Reguera B, Bravo I, Marcaillou-leBaut C, Masselin P, Fer-
nandez ML, Miguez A, Martinez A (1993) Monitoring of
Dinophysis spp. and vertical distribution of okadaic acid on
mussel rafts in Ria de Pontevedra (NW Spain). In: Smayda
TJ, Shimizu Y (eds) Toxic phytoplankton blooms in the sea.
Elsevier, New York, pp 553–558
120. Rehnstam-Holm AS, Godhe A, Anderson DM (2002)
Molecular studies of Dinophysis (Dinophyceae) species from
Sweden and North America. Phycologia 41:348–357
121. Rensel JE, Whyte JNC (2003) Finfish mariculture and
harmful algal blooms. In: Hallegraeff GM, Anderson DM,
Cembella AD (eds) Manual on harmful marine microalgae.
Monographs on oceanographic methodology, vol 11, chapter
25. UNESCO, Paris
122. Rhodes L, Scholin C, Garthwaite I, Haywood A, Thomas A
(1998) Domoic acid producing Pseudo-nitzschia species de-
duced by whole cell DNA probe-based and immunochemical
assays. In: Reguera B, Blanco J, Fernandez ML, Wyatt T
(eds) Harmful algae. Xunta de Galicia & IOC-UNESCO,
Paris, pp 274–277
123. Rhodes L, Scholin C, Tyrrell J, Adamson J, Todd K (2000)
The integration of DNA probes into New Zealand’s routine
phytoplankton monitoring programmes. In: Hallegraeff GM,
Blackburn SI, Bolch CJ, Lewis RJ (eds) Harmful algal blooms
2000. IOC-UNESCO, Paris, pp 429–432
124. Rhodes L, Mackenzie AL, Kaspar HF, Todd KE (2001)
Harmful algae and mariculture in New Zealand. ICES J Mar
Sci 58:398–403
125. Rhodes L, Haywood A, Adamson J, Scholin C (2002) DNA
probes in whole cell format for the detection of Karenia spe-
cies in New Zealand waters. Abstract, 10th International
Conference on Harmful Algae, St. Pete Beach, Fla., p 242
126. Riegman R, Noordeloos AAM, Cade
´
e GC (1992) Phaeocystis
blooms and eutrophication of the continental coastal zones of
the North Sea. Mar Biol 112:479–484
127. Rines JEB, Donaghay PL, Dekshenieks MM, Sullivan JM,
Twardowski MS (2002) Thin layers and camouflage: hidden
Pseudo-nitzschia spp. (Bacillariophyceae) populations in a
fjord in the San Juan Islands, Washington, USA. Mar Ecol
Prog Ser 225:123–137
128. Rı
´
os AF, Fraga F, Figueiras FG, Pe
´
rez FF (1995) New and
regenerated production in relation to the proliferations of
diatoms and dinoflagellates in natural conditions. In: Lassus
P, Arzul G, Erard E, Gentien P, Marcaillou C (eds) Harmful
marine algal blooms. Lavoisier, Paris, pp 663–668
129. Roelke DL, Kennedy CD, Weidemann AD (1999) Use of
discriminant and fourth-derivative analyses with high resolu-
tion absorption spectra for phytoplankton research: limita-
tions at varied signal to noise ratio and spectral resolution.
Gulf Mexico Sci 17:75–86
130. Roesler CS, McLeroy-Etheridge SL (1998) Remote detection
of harmful algal blooms. SPIE Ocean Optics XIV 1:117–128
131. Roesler CS, Perry MJ (1995) In situ phytoplankton absorp-
tion, fluorescence emission, and particulate backscattering
spectra determined from reflectance. J Geophys Res
100:13279–13294
132. Roesler CS, Perry MJ, Carder KL (1989) Modeling in situ
phytoplankton absorption from total absorption spectra in
productive inland marine waters. Limnol Oceanogr 34:1510–
1523
133. Ross NW, Bates SS (1996) Electro-immunoblotting charac-
terization of Pseudo-nitzschia multiseries and P. pungens
antigens recognized by antibodies directed against whole cells.
J Appl Phycol 8:51–58
134. Rousseau V, Leynaert A, Daoud N, Lancelot C (2002) Dia-
tom succession, silicification, and silicic acid availability in
Belgian coastal waters (southern North Sea). Mar Ecol Prog
Ser 236:61–73
135. Rublee PA, Kempton JW, Schaefer EF, Burkholder JM,
Glasgow HB, Oldach D (1999) PCR and FISH detection ex-
tends the range of Pfiesteria piscicida in estuarine waters.
Virginia J Sci 50:325–336
136. Rublee PA, Kempton JW, Schaefer EF, Allen C, Harris J,
Oldach DW, Bowers H, Tengs T, Burkholder JM, Glasgow
HB (2001) Use of molecular probes to assess geographic dis-
tribution of Pfiesteria species. Environ Health Perspect
109:765–767
137. Saito K, Drgon T, Robledo JAF, Krupatkina DN, Vasta GR
(2002) Development of standard and quantitative-competitive
PCR-based diagnostic assays for Pfiesteria piscicida targeted
to the non-transcribed spacer of rDNA. Abstract, 10th
International Conference on Harmful Algae, St. Pete Beach,
Fla., p 248
138. Sako Y, Adachi M, Ishida Y (1993) Preparation and char-
acterization of monoclonal antibodies to Alexandrium species.
In: Smayda TJ, Shimizu Y (eds) Toxic phytoplankton blooms
in the sea. Elsevier, Amsterdam, pp 87–93
139. Sako Y, Murakami T, Adachi M, Uchida A, Ishida Y,
Yamaguchi M, Takeuchi T (1996) Detection of the toxic
dinoflagellate Alexandrium species by flow cytometry using a
monoclonal antibody. In: Yasumoto T, Oshima Y, Fukuyo Y
(eds) Harmful and toxic algal blooms. IOC-UNESCO, Sen-
dai, pp 463–466
140. Schmidt W, Drews G, Weckesser J, Fromme I, Borowiak D
(1980) Characterization of the lipopolysaccharides from eight
strains of the cyanobacterium Synechococcus. Arch Microbiol
127:209–215
141. Schofield O, Grzymski J, Bissett WP, Kirkpatrick GJ, Millie
DF, Moline M, Roesler CS (1999) Optical monitoring and
forecasting systems for harmful algal blooms: possibility or
pipe dream? J Phycol 35:1477–1496
142. Scholin CA, Buck KR, Britschgi T, Cangelosi G, Chavez FP
(1996) Identification of Pseudo-nitzschia australis (Bacillario-
phyceae) using rRNA-targeted probes in whole cell and
sandwich hybridization formats. Phycologia 35:190–197
143. Scholin CA, Miller P, Buck K, Chavez F, Harris P, Haydock
P, Howard J, Cangelosi G (1997) Detection and quantification
of Pseudo-nitzschia australis in cultured and natural popula-
tions using LSU rRNA-targeted probes. Limnol Oceanogr
42:1265–1272
144. Scholin CA, Marin III R, Miller PE, Doucette GJ, Powell CL,
Haydock P, Howard J, Ray J (1999) DNA probes and a
receptor-binding assay for detection of Pseudo-nitzschia
(Bacillariophyceae) species and domoic acid activity in cul-
tured and natural samples. J Phycol 35:1356–1367
145. Scholin CA, Gulland F, Doucette GJ, Benson S, Busman M,
Chavez FP, Cordaro J, DeLong R, DeVogelaere A, Harvey J,
Haulena M, Lefebrve K, Lipscomb T, Loscutoff S, Lowens-
tine LJ, Marin R III, Miller PE, McLellan WA, Moeller PDR,
Powell CL, Rowles T, Silvagni P, Silver M, Spraker T, Trainer
V, VanDolah FM (2000) Mortality of sea lions along the
central California coast linked to a toxic diatom bloom.
Nature 403:80–84
146. Scholin C, Marin R III, Massion E, Jensen S, Cline D, Roman
B, Doucette G (2002) Remote detection of HAB Species using
the Environmental Sample Processor (ESP): progress and
future directions. Abstract, 10th International Conference on
Harmful Algae, St. Pete Beach, Fla., p 254
404

147. Sellner KG, Brownlee DC (1990) Dinoflagellate-microzoo-
plankton interactions in Chesapeake Bay. In: Grane
´
li E,
Sundstro
¨
m B, Edler L, Anderson DM (eds) Toxic marine
phytoplankton. Elsevier, New York, pp 221–226
148. Sellner KG, Fonda-Umani S (1999) Dinoflagellate blooms
and mucilage production. In: Malone TC, Malej A, Harding
LW Jr, Smodlaka N, Turner RE (eds) Ecosystems at the land-
sea margin: drainage basin to coastal sea. Coastal Estuarine
Studies 55, American Geophysical Union, Washington, D.C.,
pp 173–206
149. Sellner KG, Lacouture RV, Cibik SJ, Brindley A, Brownlee
SG (1991) Importance of a winter dinoflagellate-microflagel-
late bloom in the Patuxent River estuary. Estuarine Coastal
Shelf Sci 32:27–42
150. Reference deleted
151. Reference deleted
152. Smayda TJ (1990) Novel and nuisance phytoplankton blooms
in the sea: Evidence for a global epidemic. In: Grane
´
li E,
Sundstro
¨
m B, Edler L, Anderson DM (eds) Toxic marine
phytoplankton, Elsevier, New York, pp 29–40
153. Smith JC, Cormier R, Worms J, Bird CJ, Quilliam MA,
Pocklington R, Angus R, Hanic L (1990) Toxic blooms of the
domoic acid containing diatom Nitzschia pungens in the
Cardigan River, Prince Edward Island. In: Grane
´
li E, Sun-
dstro
¨
m B, Edler L, Anderson DM (eds) Toxic marine phy-
toplankton. Elsevier, New York, pp 227–232
154. Sournia A (1978) Phytoplankton manual. Monographs on
oceanographic methodology, vol 6. UNESCO, Paris, pp 337
155. Steidinger KA, Haddad KD (1981) Biologic and hydro-
graphic aspects of red tides. Bioscience 31:814–819
156. Stock CA, McGillicuddy DJ Jr, Signell RP, Anderson DM
(2002) A quantitative modeling study of the initiation and
development of Alexandrium fundyense blooms in the western
Gulf of Maine. Abstract, 10th International Conference on
Harmful Algae, St. Pete Beach, Fla., p 270
157. Stumpf RP, Tyler MA (1988) Satellite detection of bloom and
pigment distributions in estuaries. Remote Sensing Environ
24:385–404
158. Stumpf RP, Culver ME, Tester PA, Tomlinson M, Kirkpa-
trick GJ, Pederson BA, Truby E, Ransibrahmanakul V, Sor-
acco M (2003) Monitoring Karenia brevis blooms in the Gulf
of Mexico using satellite ocean color imagery and other data.
Harmful Algae 2:147–160
159. Suarez-Isla BA, Sierralta J, Compagnon D, Fonseca M,
Loyola H (1997) Real life strategies for red tide management
in Chile: systematic application of receptor-based radioassays
for SPS and ASP toxins in international seafood safety pro-
grams. Abstract, 8th International Conference on Harmful
Algae, Vigo, Spain
160. Suzuki MT, Taylor LT, DeLong EF (2000) Quantitative
analysis of small-subunit rRNA genes in mixed microbial
populations via 5¢-nuclease assays. Appl Environ Microbiol
66:4605–4614
161. Szmant AM (2002) Nutrient enrichment on coral reefs: is it a
major cause of coral reef decline? Estuaries 25:743–766
162. Taylor FJR, Horner RA (1994) Red tides and other problems
with harmful algal blooms in Pacific Northwest coastal wa-
ters. In: Review of the Marine Environment and Biota of
Strait of Georgia, Puget Sound and Juan de Fuca Strait. Can J
Fish Aquat Sci Tech Rep 1948, BN Typographics West,
Victoria, B.C., pp 175–186
163. Tester PA, Fowler PK (1990) Brevetoxin contamination of
Mercenaria mercenaria and Crassostrea virginica: A manage-
ment issue. In: Grane
´
li E, Sundstro
¨
m B, Edler L, Anderson
DM (eds) Toxic marine phytoplankton. Elsevier, New York,
pp 499–503
164. Tester PA, Stumpf RP (1998) Phytoplankton blooms and
remote sensing: what is the potential for early warning.
J Shellfish Res 17:1469–1471
165. Tester PA, Stumpf RP, Vukovich FM, Fowler PK, Turner JT
(1991) An expatriate red tide bloom: transport, distribution
and persistence. Limnol Oceanogr 36:1053–1061
166. Towers NR, Garthwaite I (2001) Biological assay and detec-
tion methods for marine ‘‘shellfish’’ toxins. In: Massaro E (ed)
Neurotoxicology handbook, vol 1., Humana Press, London,
pp 269–291
167. Trainer VL, Adams NG, Bill BD, Anulacion BF, Wekell JC
(1998) Concentration and dispersal of a Pseudo-nitzschia
bloom in Penn Cove, Washington, USA. Nat Toxins 6:113–126
168. Trainer VL, Adams NG, Bill BD, Stehr CM, Wekell JC,
Moeller P, Busman M, Woodruff D (2000) Domoic acid
production near California coastal upwelling zones, June
(1998). Limnol Oceanogr 45:1818–1833
169. Trainer VL, Hickey BM, Horner RA (2002) Biological and
physical dynamics of domoic acid production off the Wash-
ington coast. Limnol Oceanogr 47:1438–1446
170. Turner RE, Rabalais NN (1991) Changes in the Mississippi
River this century: implications for coastal food webs. Bio-
science 41:140–147
171. Tyler MA (1984) Dye tracing of a subsurface chlorophyll
maximum of a red-tide dinoflagellate to surface frontal re-
gions. Mar Biol 78:285–300
172. Tyler MA, Heinbokel JF (1985) Cycles of red water and
encystment of Gymnodinium pseudopalustre in the Chesapeake
Bay: effects of hydrography and grazing. In: Anderson DM,
White AW, Baden DG (eds) Toxic dinoflagellates. Elsevier,
New York, pp 213–218
173. Tyler MA, Seliger HH (1978) Annual subsurface transport of
a red tide dinoflagellate to its bloom area: water circulation
patterns and organism distributions in the Chesapeake Bay.
Limnol Oceanogr 23:227–246
174. Tyler MA, Coats DW, Anderson DM (1982) Encystment in a
dynamic environment: deposition of dinoflagellate cysts by a
frontal convergence. Mar Ecol Prog Ser 7:163–178
175. Tyrrell JV, Bergquist PR, Bergquist PL, Scholin CA (2001)
Detection and enumeration of Heterosigma akashiwo and
Fibrocapsa japonica (Raphidophyceae) using rRNA-targeted
oligonucleotide probes. Phycologia 40:457–467
176. Uchida A, Nagasaki K, Hiroishi S, Ishida Y (1989) The
application of monoclonal antibodies to an identification of
Chattonella marina and Chattonella antiqua. Nippon Suisan
Gakkaishi 55:721–725
177. Valiela I, McClelland J, Hauxwell J, Behr PJ, Hersh D,
Foreman K (1997) Macroalgal blooms in