sign in sign up
<<
Home , Algal bloom, Stratification (water)

HARMFUL ALGAL BLOOMS

HARMFUL ALGAL BLOOMS IN STRATIFIED ENVIRONMENTS

BY PATRICK GENTIEN, PERCY DONAGHAY, HIDEKATSU YAMAZAKI, ROBIN RAINE, BEATRIZ REGUERA, AND THOMAS OSBORN

Successful management of the coastal able. A major challenge is to understand standing, and subsequent modeling of environment requires sound knowledge the processes governing the population harmful populations (Figure 1). Cou- of the biological, chemical, and physical dynamics to a level of detail that allows pling physical effects such as turbulence, processes governing the growth, reten- mathematical formulations that are rela- shear, and advection with biological tion, dispersion, and transport of water tively simple, yet reproduce the salient behavior (migration, physiological adap- constituents, including . The features of HAB population dynamics. tation) holds the key to understanding growing problem posed by the existence A feature common to many oceanic vertical distributions, bloom dynamics, of harmful plankton species now de- harmful algal events is that the phyto- and patterns of toxicity. Some of these mands a thorough understanding of the plankton populations, composed mainly physical processes are not yet defi ned at life histories and population dynamics of dinofl agellates, build up to their high- the appropriate scale, yet may be crucial of these organisms. Without this knowl- est in subsurface layers in the formation of harmful blooms. edge, managers can neither assess the (Figure 1). These layers are usually as- Furthermore, models have thus far been impact of anthropogenic activity on the sociated with water-column stratifi ca- restricted by our inability to gauge the environment and biota, nor understand tion (Figure 2). Stratifi ed water is en- interactions between the biology of algal the relationship between harmful phyto- countered in systems, coastal taxa and underlying physical processes. plankton species and the harmful events embayments, and as well as We describe here the major features of (commonly referred to as harmful algal in retentive zones, such as gyres, in the blooms in subsurface layers, considering blooms [HABs]) caused by them, and open . The temporal scales (days to the biological response to upper-mixed- climate. Without this understanding, the weeks) and spatial scales (decimeters to layer dynamics, and the population dy- desirable goals of prediction and miti- meters in thickness) of subsurface lay- namics of blooms. gation of harmful events are unattain- ers pose problems for sampling, under-

Th is article has been published in , Volume 18, Number 2, a quarterly journal of Th e Oceanography Society. Copyright 2005 by Th e Oceanography Society. All rights reserved. Reproduction of any portion of this article by photo- 172 Oceanography Vol.18, No.2, June 2005 copy machine, reposting, or other means without prior authorization of Th e Oceanography Society is strictly prohibited. Send all correspondence to: [email protected] or Th e Oceanography Society, PO Box 1931, Rockville, MD 20849-1931, USA. Typical assemblage collected in the (Killary Harbour, Ireland) composed of Cerati- um lineatum, Protoperidinium depressum, Pro- toperidinium stenii, and one cell of Dinophysis acuminata. Note an empty theca of C. lineatum and a of a pennate .

Oceanography Vol.18, No.2, June 2005 173 UPPER OCEAN and mixing energy with time. quences on , trig- AND MIXEDLAYER DYNAMICS The formation of a surface mixed gering the annual life cycles of many The term “” is often ap- layer has pronounced effects on plank- organisms in the . In addition, con- plied to the thin band of water, some tonic organisms’ environment. Foremost tinued heating of the surface mixed tens of meters thick, that extends from amongst these is promotion of the dia- layer through summer renders the water the ocean’s surface to the . tom in temperate climates. column increasingly stable, yet has the The mixed layer is the boundary layer In winter, the mixed layer deepens due to additional consequence of preventing between the atmosphere and the deep increased winds and net surface cooling, the mixing of nutrients from below. ocean, and results from the balance be- entraining water from below that had Therefore, nutrients in the surface mixed tween heating from solar radiation and been isolated from surface processes dur- layer are exhausted during early summer. surface cooling. Winds help mix the heat ing the summer. Entrainment and deep- These warm, nutrient-depleted condi- through the layer by generating waves, ening of the mixed layer usually peak tions are ideal for the proliferation of inducing Langmuir circulation, produc- several months after the winter solstice. fl agellates, including dinofl agellates. As ing , initiating near-inertial During the spring, however, buoyancy many HAB species are dinofl agellates, frequency motions, and generating tur- input during daytime solar heating over- it is during temperate-climate summers bulence. In the coastal ocean, freshwater rides the turbulent mixing tendencies such as those of northwestern Europe, runoff from the land contributes directly from wind and surface cooling. Phyto- that many HAB events occur. The annual to the upper layer, and in the shallow plankton are, therefore, maintained in cycle closes by cooling during autumn. parts of the coastal ocean, tidal energy the surface layer and respond to the fa- Hidden within this rather simplistic can contribute to mixing in the upper vorable light and nutrient environment description of seasonal cycles are many layer. The mixed layer therefore repre- by blooming. processes that occur over much shorter sents an integration of the variations in The spring bloom has huge conse- Patrick Gentien (patrick.gentien@ifremer. fr) is Director of CREMA, a Joint Re- search Unit between Ifremer and CNRS, L’Houmeau, France. Percy Donaghay is Senior Research Scientist, Graduate School of Oceanography, University of Rhode

0 Island, Narragansett, RI, USA. Hidekatsu Yamazaki is Professor, Department of 10 Ocean Sciences, Tokyo University of Marine

20 Science and Technology, Tokyo, Japan. Rob- in Raine is Senior Research Scientist, Martin Depth (m) Depth 30 Ryan Institute, National University of 40 Ireland, Galway, Ireland. Beatriz Reguera is 12 14 16 0 0,4 0,8 0 50 100 Particulate load Dinoflagellates (%) Senior Research Scientist, Instituto Español de Oceanografi a,Vigo, Spain. Th omas R. Figure 1. Vertical distribution of temperature (degrees C), particulate total volume (rela- tive unit), and fractional cell of dinofl agellates (percent) off the “Pertuis Osborn is Professor of Physical Oceanog- d’Antioche,” France (simplifi ed and redrawn from Gentien et al., [1995]). In this example, raphy, Department of Earth and Planetary Dinophysis acuminata formed as part of the dinofl agellate assemblage in the pycnocline, but was absent elsewhere in the . Th e distribution of this dinofl agellate was associ- Sciences, Th e Johns Hopkins University, ated with an increase in organic aggregates. Baltimore, MD, USA.

174 Oceanography Vol.18, No.2, June 2005 0 2 2 4 the permanently stratifi ed subtropical 1 1 8 2 1 gyres, episodic entrainment of nutri- 10 4 4 2 ents into the mixed layer associated with 8 4 1 20 4 4 evolving mesoscale eddies can stimulate 8 4 4 2 1 2 2 blooms and subsequent 30 1 grazing and growth events within the Depth (m) Depth 40 . In a similar fashion to its

50 creation in spring, erosion of the sea- sonal thermocline in autumn is episodic 60 0 10 20 30 40 50 60 or discontinuous with sharp deepening Distance (km) events often caused by simultaneous Figure 2. fl uorescence (relative units) along a transect in Bantry Bay (SW Ireland). high winds and convective heat loss at Shaded areas represent values greater than 10. Th e major species in the assemblage was Kar- enia mikimotoi (alias Gyrodinium aureolum). Redrawn and modifi ed from Raine et al. (1993). the sea surface.

POPULATION DYNAMICS Each phytoplankton species has a dif- ferent combination of characteristics that defi nes its ecological niche and de- termines its distribution and behavior. The challenge is to defi ne unique adap- temporal or spatial scales. For example, of hours to days. Formation of the sea- tations of individual HAB species that a daily cycle can be observed during the sonal thermocline in spring is not a account for its survival and persistence, onset of stratifi cation in spring. A weak, gradual, monotonic process, but rather and in some instances its dominance, shallow mixed layer is formed during the a series of low-wind heating periods al- during bloom events. By understand- day, which is destroyed each evening by ternating with mixing events caused by ing their adaptations, it should be pos- convective mixing due to nighttime ra- stronger winds associated with the pas- sible to describe and predict patterns diative cooling. Physically, the seasonal sage of storms. It is important to note of species abundance as a function of mixed layer therefore develops earlier as buoyancy is effi ciently ventilated down- wards; consequently, the summer sea surface maximum temperature is greater, The growing problem posed by the existence and the spring bloom develops earlier, of harmful plankton species now demands a than if there were no diurnal cycle in the heat input to the surface ocean. The in- thorough understanding of the life histories creasing in shallow layers also and population dynamics of these organisms. leads to shallower absorption of short- wave radiation and hence warming of a thinner layer of surface water, which then leads to further stratifi cation. that these time scales closely match the hydrographic processes, nutrient distri- Punctuating these regular cycles are dominant scales of growth and behavior butions, and community interactions, episodic events, usually on time scales of phytoplankton and zooplankton. In preferably by the use of models in an

Oceanography Vol.18, No.2, June 2005 175 operational mode. dynamics in several of the terms of this simple modeling. Yet, the interaction Box 1 shows an equation for the evo- equation, but coupled processes cover between the light environment and a lution of the local number of organisms a broad range of temporal and spatial species that is able to vertically migrate, per unit volume. When considering a scales, with many non-linearities. In or the decoupling between photosynthe- particular HAB species within its com- other words, one cannot deduce the be- sis and growth, has shown the approach munity, there is a set of state variables, havior of the system by considering the to be too simplistic. Similarly, the use of

Ci, corresponding to the concentration effect of one parameter in isolation. For nutrients in predicting growth should be of a species, and a corresponding set of example, our understanding of under- used with caution as some HAB popula- equations governing many direct and water irradiance and the relationship tions rely solely on ammonium fl uxes coupled biological, chemical, and physi- between and irradiance from the pycnocline (Le Corre et al., cal processes. Environmental condi- has made the light-controlled growth 1993) while they are usually grown in tions and processes affect population of phytoplankton highly amenable to vitro on nitrate.

BOX 1: THE GENERIC EQUATION

A robust mathematical equation for the local number of organisms per unit volume can be written in the following form: ∂C = µC mC ∇.(C v) ∇.(Cu ) ∂t

∂C is the time rate of change of the state variable C, the number of cells of a particular HAB species per unit volume. ∂t Population dynamics is defi ned as the change of C in space and time.

µC represents growth by cell division. The growth rate is determined by intrinsic genetic factors, and modulated by envi- ronmental factors, such as nutritional and light history, turbulence, temperature, and salinity.

mC is the direct loss of organisms through mortality. This term includes processes such as grazing, mechanical damage, and death from infections by viruses or other pathogens.

∇.(Cu ) includes the three-dimensional, time-variable transport of cells by the water fl ow, for example, mean circulation, tidal currents, wind drift, and turbulence (often identifi ed as turbulent ). “Disappearance” of blooms due to off- shore fl ow would appear in this term.

∇.(C v) is the motion of organisms relative to the water, described as velocity, v. This term includes swimming, sinking, or rising due to buoyancy and slippage relative to the local fl ow that arises for a variety of reasons such as size, tropism, and shape (Donaghay and Osborn, 1997).

176 Oceanography Vol.18, No.2, June 2005 Many models are based on the physi- Density stratifi cation motions a vertical profi le of horizontal velocity ological characteristics of a single species in a predominantly horizontal fashion. at a comparable resolution to the verti- considered in isolation. These models The vertical variation in this horizontal cal scale of thin layers and fi ne structure. then attempt to predict its presence and fl ow, called shear, converts distributions This measurement must be taken in con- abundance based on some knowledge of of any property, including the plankton, junction with the variation in horizontal environmental conditions. For example, into sharp, vertical gradients. Physical and vertical distributions of biological, models of species’ behavior have focused oceanographers call such meter-scale chemical, and physical fi elds.

Transport A major challenge is to understand the processes In order for a population to build up lo- cally to signifi cant cell densities, growth governing the population dynamics to a level of terms must exceed loss terms. Transport detail that allows mathematical formulations that due to mean circulation, tidal currents, and wind drift is one of the most impor- are relatively simple, yet reproduce the salient tant loss terms. For example, the disap- features of HAB population dynamics pearance of coastal blooms may simply be due to offshore fl ow. A precise knowl- edge of the fl ow fi eld is therefore essen- mainly on their growth, grazing, and, vertical variations in salinity, tempera- tial to the prediction of the occurrence of in a few cases, mortality. Consequently, ture, and density “fi ne structure.” Thin a toxic event at the coast. This knowledge nutrition limitation has often been con- layers of concentrated chemical and must be acquired at the horizontal and sidered to be the major factor governing biological concentrations are the chemi- vertical scales of phytoplankton patches. population dynamics of harmful . cal and biological equivalents of the fi ne While the horizontal scales are accessible As a result, the paradigm that eutrophi- structure in temperature, salinity, and from three-dimensional hydrodynamical cation is the cause of HABs has emerged. density. The important caveat is that models, it appears that the required ver- The approach is, however, an over-sim- biological and chemical layers are also tical scale (tens of centimeters) needs a plifi cation of the generic equation shown forced by biochemical processes in ad- better understanding of the distribution in Box 1, which largely ignores species- dition to physical processes. However, of the energy at the same scale. HABs specifi c behavioral attributes like tro- while the coupling of processes may (or frequently occur in thin layers, which are pism, migration, and response to grazers. may not) bind biochemical layers to lay- limit layers between two overlying water Diffi culties in parameterizing these and ers defi ned by temperature, salinity, or masses, the movements of which are not other terms have led to large uncertain- density, it is shear that has a subtle but yet fully understood. Understanding and ties in model predictions. highly important effect on plankton dis- subsequent modeling of the advection tributions. When taken in conjunction of HABs will be essential to prediction PHYSICAL CONTROL with horizontal gradients of various wa- based on operational oceanography. PROCESSES ter components such as phytoplankton, Populations of phytoplankton are often shear has a major role in forming thin Turbulence and Mixing concentrated locally in either patches layers and fi ne structure with horizontal Phytoplankton populations exhibit a de- or layers. These are governed by phys- scales of kilometers. When studying phy- gree of patchiness governed by physical ics in that the population is subject to toplankton distributions, a crucial, fi rst- processes. For example, while shear tends transport, dispersion, and turbulence. order measurement therefore includes to spread phytoplankton patches into

Oceanography Vol.18, No.2, June 2005 177 thin layers where local concentrations Turbulence occurs everywhere in the for the consideration of the infl uences are maintained, mixing tends to decrease ocean, but its intensity varies consider- of turbulence on trophodynamics, one local concentrations. These opposing in- ably in time and space. The past two must realize that a gap exists between the fl uences may have profound effects; their decades of microstructure observations measurement needs of physical oceanog- infl uence on grazing has been have yielded a signifi cant body of infor- raphers and the applications of interest shown to depend on the concentration mation on the of oceanic turbu- to biologists. Normally, what physical of toxins to which grazers are exposed. lence. In general, turbulence exhibits a oceanographers call an “instantaneous” Small-scale agitation may profoundly layered structure and patchiness, just like dissipation is an average dissipation rate infl uence phytoplankton populations. plankton. Its intensity is normally higher over a certain spatial scale, somewhere Most of the effects are due to collisions near the sea surface, the ocean fl oor, and around 0.5 to 5 m. Physical oceanogra- phers are interested in an extensive aver- age of such “instantaneous” dissipation A feature common to many oceanic harmful rates. On the other hand, microorgan- isms experience a true instantaneous ve- algal events is that the phytoplankton locity strain fi eld. The difference between populations, composed mainly of , the “local” dissipation rate experienced by cells and a one-meter-average dissipa- build up to their highest concentrations in tion rate commonly used by physicists is subsurface layers... depicted in Figure 4.

Aggregation in Subsurface Layers between cells; therefore, the degree of in the thermocline. Twenty-fi ve to thirty- Aggregations of phytoplankton in thin agitation depends on cell density. The meter patches of turbulence are found layers at the seasonal thermocline, or at incidence of collision is non-linear and in the thermocline; however, turbulent other within the water col- cannot be predicted in a simple way. layers in a thermocline are usually patchy, umn, can be comprised of a range of Among the effects related to cell col- with a typical thickness of a few meters species. can be present in high lisions are cell cycle arrest leading to (Figure 3). A “patch” is not a single over- numbers, but often these are senescent reduction in growth rate, cell lysis (the turning event from top to bottom, rather, populations that have aggregated at the rupture and destruction of the cell) lead- it is a section of dissipation exceeding a density discontinuity during settlement ing to increased mortality, and bloom fi xed threshold, for example, 10-8 W/kg. towards the seabed. Diatoms common in fl occulation leading to a decrease in lo- Thicker patches are usually associated early summer, such as Leptocylindrus or cal concentration (Jackson, 1990). In the with stronger turbulence and last longer , can be found in very high latter case, plankton population sensi- than thinner ones. Inertial wave shear numbers. More frequently, thin layers tivity is highly dependent on the cells’ causes persistent patches that can last consist of very high densities of dino- physiological status. It has been shown for several hours. Persistent patches are fl agellates. Karenia mikimotoi or various that a burst of turbulence induced, for also found at interfaces of intrusions, for Ceratium species are common in the instance, by a wind event may cause the example, the California undercurrent. Celtic Sea, on the Ushant front, or in the termination of a bloom through sinking Physical oceanographers can now Kattegat– area between Swe- of the aggregated cells. This is a good ex- measure (with some confi dence) the den, Denmark, and Norway (Figure 5). ample of the interaction between physi- rate of dissipation of turbulent kinetic Around the southwest of Ireland, toxic cal and biological events. energy ε. While existing data are useful Dinophysis species (Figure 6a) have been

178 Oceanography Vol.18, No.2, June 2005 found at cell densities as high as 124,000 carried out using single water bottles, As specifi c profi ling systems with new de- cell/l. Here, D. acuminata is considered and it was extremely fortuitous that the tection systems have developed, thin lay- to be one of the species present as a thin bottle had been lowered to precisely the ers have been reported more frequently. layer. Sampling at that time, in 1992, was correct depth to sample within the layer. Observations made with specifi c

Figure 4. Vertical profi le of dissipation rates com- puted from ∂v/∂z. Open circles are 1-m-average dissipation rate values computed from a con- ventional spectral method. Dots represent un- Figure 3. Microstructure profi le data obtained from a free-fall instrument smoothed local dissipation data and the asterisks Camel II. Th e instrument measures two shears (∂u, ∂z, ∂v/∂z) and vertical represent the average local dissipation rate over 1 temperature gradient (∂T/∂z) as well as the temperature profi le. Although m. SOURCE: Yamazaki and Lueck (1990). the upper surface layer is still in a stratifi ed condition, turbulence is ac- tive in the upper 10 m. A 5-m uniform temperature interface is associated with a turbulent patch at 28 m depth. A strong turbulent patch occurred at 170 m depth, where the California is found. Th e dissipation rate associated with 0.1/s rms velocity shear is roughly 7.5 x 10-8 W/kg; see for example, 103 m depth. SOURCE: Yamazaki and Lueck (1990).

Oceanography Vol.18, No.2, June 2005 179 Figure 5. Th is map of North- western Europe (Partensky and Sournia, 1986) provides a history of the distribution of blooms of the dinofl agellate Karenia mikimotoi. Th is species is a “fi sh- killer” and produces a suite of exotoxins inhibiting membrane ATP-ases. During summer, these blooms are localized in strongly stratifi ed areas. Th e east coast of Denmark and Kattegat–Skagerrak waters are under the infl uence of freshwater outfl ows from the Baltic and plumes from the Elbe and Rhine Rivers. A buoyant jet fl ows along the Norwegian coast. All along the Atlantic coast of northern Europe, including the locations of tidal fronts, blooms have been observed which are now linked to jet-like fl ows asso- ciated with physical dynamics at the coastal boundary of stratifi ed regions (size of triangles refl ects observed cell densities).

profi lers have been made in the Gulf of 2) to improve knowledge with the help them to grow in sub-optimal conditions Finland, in the Bay of Biscay (France), of a suite of newly designed instrumen- encountered in deep layers. Layers with and offshore San Juan Island (Wash- tation. However, specifi c effort has to be very high cell densities are closely asso- ington state, U.S.). Studies in Europe made to develop new observing systems ciated with the water column’s vertical have mainly targeted D. acuminata and that would allow identifi cation of phyto- structure, and a layer’s thickness is re- Gymnodinium catenatum (Figure 6b) in plankton groups, taxa, and even toxicity. lated to shear. High positive shear pro- different dinofl agellate dominated as- motes phytoplankton layers that are less semblages. In the San Juan Island study, What are the Adaptive that 2 m thick, whereas more diffuse thin a near-bottom, high-density, thin layer Advantages of Growing in Thin layers that are 5 to 10 m thick are associ- of Pseudo-nitzschia spp. persisted for sev- Layers? ated with physical discontinuities where eral days. International cooperation be- Some phytoplankton species, including shear is lower. tween U.S. and Europe is ongoing (Box HABs, have selected strategies allowing Formation of thin layers can benefi t

180 Oceanography Vol.18, No.2, June 2005 a

Figure 6. (a) Common species of the toxic phytoplankton genus Dinophysis. Clockwise from top left: D. tripos, D. acuta, D. caudata and D. acuminata. Along the European coastline of the , this genus is the cause of Diarrheic Shellfi sh Poisoning and has a major economic impact. Th ese species are also encountered on the Atlantic American coast but are not toxic. Comparative studies on toxinogenesis on both sides of the Atlantic would be valuable. (b) Gymnodinium catenatum is an unarmoured, marine, planktonic dinofl agellate species. It is a chain-forming, toxin-pro- ducing, red species associated with PSP events throughout the world. G. catenatum is a planktonic species. G. catenatum orients its swimming in shear fl ows, and swimming speed increas- es with chain length. G. catenatum actively concentrates at depths with low turbulence and shear. Th e fi rst G. catenatum red tide was reported from the Gulf of California. Populations of this species have been recorded from Mexico, Japan, Australia, Venezuela, the Philippines and Europe. G. catenatum produces a characteristic resting cyst. Th ese cysts can germinate after just two weeks of dor- mancy and initiate new populations. Cysts are not only a reseeding tool, but also a dispersal agent: G. catenatum was introduced to Australian waters via ships’ ballast water. b

phytoplankton taxa in a number of ways. First, formation of thin layers can opti- mize population growth rates by concen- trating the population where conditions are optimal for individual growth. For example, formation of a thin layer in the upper part of the nutricline can optimize growth rates of photosynthetic harmful algae by concentrating the population where growth rate is no longer limited by low nutrient concentrations in the surface layer, or by extremely low light

Oceanography Vol.18, No.2, June 2005 181 deeper in the water column. Second, it possible for a species to chemically localized area is more likely to ensure formation of thin layers can reduce modify its local environment, potentially successful mating. Finally, formation of mortality rates by allowing populations enhancing immigration rates by provid- thin layers can lead to physical-biologi- to achieve suffi cient densities whereby ing a chemical signal to migrating algae, cal interactions that can dramatically chemical defenses can be used to reduce or enhancing growth rates by increasing infl uence the transport and retention of losses and increase growth relative to nutrient uptake or reducing bioavailabil- harmful algae, as retention structures competitors. These chemical defenses ity of toxic metals or by modifying the can develop at density interfaces with include production and, in some cases, local properties of water, such as viscosi- suitable time (~10 days) and spatial (15 the release of compounds that induce ty, through excretion of external metabo- to 20 km) scales. grazer avoidance, suppress grazing ac- lites. Fourth, the release of chemicals also Formation of thin layers has two po- tivity, and/or kill zooplankton and mi- plays a key role in the complex life cycles tential disadvantages. First, formation of crozooplankton grazers within the layer. of many species, including harmful ones, thin layers can greatly increase mortality Third, formation of thin layers makes as the release of suffi cient gametes in a if chemical defenses are not adequate to

BOX 2: THE INTERNATIONAL DIMENSION

In September 2002, an American-European Union workshop led by P. Gentien (IFREMER), will use its specifi cally designed was held in Trieste, Italy to formulate a joint program on the profi ler composed of an in situ analyzer coupled with a study of Harmful Algal Blooms. As a result, the European fl uorescence video microscope to detect thin layers containing Commission and National Science Foundation have funded Dinophysis, as well as a novel in situ rheometer. L. Fernand (CE- two major three-year projects on HABs that have transatlantic FAS, U.K.) will map the chlorophyll thin layers and estimate partners. One of these, called HABIT, is specifi cally targeted horizontal diffusion. B. Reguera (IEO, Spain) is a specialist in at studying the maintenance and behavior of high-density the life cycle of Dinophysis and will estimate the nutrition and populations of Dinophysis in subsurface thin layers in strati- growth rate of Dinophysis populations. fi ed waters around Europe. HABIT is coordinated by biolo- International collaboration is an integral part of scientifi c re- gist R. Raine at the National University of Ireland, Galway, and search within Europe, which is funded by the EC under a series includes biologists and physicists from France, Spain, and the of Framework programs (HABIT is funded through the Sixth United Kingdom. Crucial to the study, however, is the partici- Framework Programme). The EU also supports established in- pation of T. Osborn and J. Katz of the John Hopkins University, ternational networks such as NEMEDA, which seek to mitigate Baltimore, Maryland. The U.S. team will make measurements the harmful effects of Dinophysis along the Atlantic seaboard of within thin layers of Dinophysis using their underwater ho- Europe and involve scientists from Portugal, Spain, France, and lographic camera (http://www.me.jhu.edu/~lefd/shc/digital. Ireland. Support for networks can be achieved either through htm). This system allows observation of processes actively oc- interregional programs (NEMEDA is supported by INTERREG curring at the critical scales of less than 1 cm. The French team, IIIb) or the Sixth Framework.

182 Oceanography Vol.18, No.2, June 2005 induce grazer avoidance, suppress graz- in the water column, and many standard lating to HABs in stratifi ed environments ing activity, and/or kill the grazers within measuring instruments (such as in situ will be discussed and fi nalized during the layer. In a similar fashion, the oppor- fl uorometers) are not able to observe the GEOHAB Open Science Meeting on tunity for infection by or viruses thin layers; hence, researchers have been HABs and Stratifi cation, to be held from will be greatly increased unless chemical unable to target them during sampling. 5 to 7 December 2005 at UNESCO head- defenses are adequate to suppress such It is only very recently that technologies quarters in Paris. attacks. Second, a thin layer may form at have been developed to a satisfactory REFERENCES Donaghay, P. and T.R. Osborn. 1997. Toward a theory of biological-physical control of harmful Several physical processes are crucial in the dynamics and impacts. and Oceanography 42(5, part 2):1283-1296. formation of harmful populations, but it is Gentien, P., M. Lunven, M. Lehaitre, and J.L. Du- vent. 1995. In-situ depth profi ling of particles in these areas where knowledge is lacking. sizes. Deep Sea Research 42:1297-1312. Jackson, G.A. 1990. A model of the formation of marine algal fl ocs by physical coagulation proc- esses. Deep Sea Research 37(8):1197-1211. Le Corre, P., S. L’Helguen, and M. Wafar. 1993. a depth that does not optimize growth level to allow the study of thin layers with source for uptake by Gyrodinium cf. and/or retention. For example, thin lay- the result that research into their main- aureolum in a tidal front. Limnology and Ocean- ography 38(2):446-451. ers may develop in subsurface, light- tenance is as yet in its infancy. Such stud- Partensky, F. and A. Sournia. 1986. Le dinofl agellé limited environments or in nutrient-de- ies require close collaboration involving Gyrodinium cf. aureolum dans le plankton de pleted surface layers. In either case, the physicists, modelers, and phytoplankton l’Atlantique nord: Identifi cation, écologie, toxi- cité. Cryptogamie Algologie 7:251-275. population growth of a photosynthetic ecologists. The outcome of these studies Raine, R., B. Joyce, J. Richard, Y. Pazos, M. Moloney, harmful species may be reduced relative will have application to many scientifi c K.J. Jones, and J. W. Patching. 1993. The devel- opment of a bloom of the dinofl agellate (Gyro- to a population concentrated in the nu- questions outside the domain of HABs. dinium aureolum [Hulburt]) on the south-west tricline or spread over the water column. For example, estimates of fl ux in Irish coast. ICES Journal of Marine Science As the focus in the study of harmful al- the deep ocean may vary by a factor of 50:461-469. Yamazaki, H. and R. Lueck. 1990. Why oceanic gae is on one single species, modeling 10 and depend on the species of phyto- dissipation rates are not lognormal? Journal of exercises should take into consideration plankton present. Research priorities re- Physical Oceanography 20(12):1907-1918. these different processes. For one given species, the ratio of advantages to disad- vantages will vary during the course of the bloom, permitting either an increase or decrease in population size It is only very recently that technologies

CONCLUSIONS have been developed to a satisfactory level Several physical processes are crucial in to allow the study of thin layers with the the formation of harmful populations, but it is in these areas where knowledge result that research into their maintenance is lacking. Acquisition of information on is as yet in its infancy. the relationship between HABs and strat- ifi cation has been slow and erratic. Sub- surface layers can be present at any depth

Oceanography Vol.18, No.2, June 2005 183