The Scottish Government

« Previous | Contents | Next »

Listen

11.0 SCOTTISH COASTAL WATERS AS OPEN SYSTEMS

11.1 Contrasting behavior of HAB species in Scottish coastal waters and elsewhere in Europe

The phytoplankton communities and bloom dynamics in Scottish and European coastal waters, particularly Scandinavian waters, are generally similar. The phytoplankton has a Boreal character and the annual bloom and species succession patterns in Scottish waters are similar to those throughout the North Sea. However, there are some regional HAB differences along north-south and east-west gradients that do not appear to be related to regional aquacultural patterns. Along the north-south latitudinal gradient, fish farming is most extensive in the north (Scotland, Norway) and shellfish culture most extensive in the south, from France to Spain. In Irish coastal waters, fish farming is less extensive than in Scotland and Norway, but shellfish cultivation is more intense.

The most distinctive HAB feature in Scottish coastal waters is the closure of shellfish harvesting grounds because of domoic acid intoxication and potential ASP toxicity attributed to blooms of Pseudo-nitzschia [Section 2.2]. In contrast, DSP toxicity of shellfish from ingestion of Dinophysis is relatively rare [Section 3.5]. These two patterns of shellfish toxicity in Scottish waters differ significantly from that elsewhere in European waters, including Ireland, where DSP toxicity is much more common, frequent and persistent than ASP. Elsewhere, ASP is rarely reported. This regional pattern occurs despite the presence of a common Pseudo-nitzschia and Dinophysis flora throughout European coastal waters.

With regard to paralytic shellfish poisoning, the toxigenic PSP species, A. tamarense, is common in European and Scottish coastal waters. However, the UK populations are comprised of two distinct lineages based on molecular evidence (Figure 13; Higman et al., 2001; Medlin et al., 2001). Alexandrium tamarense populations in Scottish coastal waters, including Orkney, have been assigned to the "North American lineage" because of their toxicity in culture and molecular similarity to toxic North American strains. [The Faroe Island population of A. tamarense probably also falls within this lineage given bloom-linked mortality of farmed salmon there (Mortensen, 1985).] Alexandrium tamarense strains from the coastal waters of England, Northern Ireland and the Republic of Ireland are non-toxic in culture, and are related molecularly. These populations have been assigned to the non-toxic "Western European Lineage" (Higman et al., 2001). Within the British Isles, therefore, there is an apparent north-south gradient in PSP toxicity of A. tamarense populations: northern populations (e.g. Scotland; Faroe Islands (possibly) are toxic; southern populations (England, Ireland) are not toxic.

The toxin profile and lineage of A. tamarense populations in the coastal waters from Norway to Portugal are unresolved. PSP events in those waters are often ascribed to A. tamarense, but Fraga and Franco (1994) state the evidence is circumstantial. They recommend that toxin analyses of clonal A. tamarense from those regions be carried out to confirm their purported toxicity. The occurrence of PSP toxicity in southern European waters is irrefutable, however, and reflects another notable HAB difference from Scottish coastal waters. In Scottish coastal waters, A. tamarense is usually considered to be the only indigenous PSP-producer, although recent reports of A. minutum and A. ostenfeldii presence in Orkney waters challenge this (see Section 4.2; Töbe et al., 2001). The first toxic bloom of A. tamarense reported in Spanish coastal waters occurred in 1984 (Blanco et al., 1985). But A. tamarense appears to be a relatively minor species in the upwelling dominated waters along the Portuguese coast and in the mussel culturing rías in Spain. The major source of PSP toxicity in those waters is the chain-forming dinoflagellate Gymnodinium catenatum which is rarely found north of the Iberian peninsula (Hallegraeff and Fraga, 1998). Reports of its pre-historic ("Little Ice Age") blooms in the Kattegat-Skagerrak region based on fossil cyst deposits (Dale et al., 1993) and germination experiments that suggested prolonged retention of cyst viability (Nehring, 1995) have proven incorrect. It is now known that the non-toxic, "look-alike" species, Gymnodinium nolleri, was being dealt with, and not G. catenatum (Ellegaard and Oshima, 1998).

Two primary PSP-species, therefore, characterize European coastal waters: A. tamarense, most prominent in northern European (Scandinavia, UK) waters, and G. catenatum restricted to Iberian waters. Of the two, A. tamarense has the greater regional distribution. Blooms of a third PSP-toxic species, A. minutum, have been increasingly reported from European coastal waters since first recorded in 1985 off the north-western coast of Brittany (see Nehring, 1998). This appearance is consistent with evidence for its global range expansion (see Maestrini et al., 2000) and/or increased bloom frequency, possibly in its emergence from the "hidden flora". Until its recent discovery at the Orkney Isles (Töbe et al., 2001), A. minutum has been conspicuously absent from the HAB dinoflagellate taxonomic checklist for Scottish coastal waters, and has even been cited as "not known" from these waters (MacDonald and Davidson, 1998). A similar situation applies to toxic A. ostenfeldii, also recently reported from the Orkney Isles (Töbe et al., 2001) and along the Scottish east coast (John et al., 2003). If these new reports are correct, then three of the ten known PSP-toxic Alexandrium species occur in Scottish coastal waters, and not only A. tamarense.

The north-south gradient in the distribution of toxic A. tamarense strains within the UK and the different primary PSP species in northern and southern Europe are not the only noteworthy regional differences. There may be a significant west-east difference in HAB components between Scotland and Scandinavian + eastern North Sea coastal systems. Blooms of harmful raphidophytes and other phytoflagellate groups, including haptophytes, are rarely reported for Scottish coastal waters, unlike in the coastal waters extending from Norway to the Dutch Wadden Sea, where toxic blooms of multiple species of Chrysochromulina, Prymnesium, Pyramimonas, Chattonella and Heterosigma akashiwo frequently occur [see Sections 6.0, 7.0].

Collectively, the characteristics of the HAB community in Scottish coastal waters and their divergences from behavior elsewhere in European coastal systems pose an interesting issue. Do the observed differences reflect the presence of ecophysiological barriers, or unique selective factors within Scottish coastal water habitats? The sea lochs and firths of Scotland have many physiographic features in common with the coastal systems of southwest Ireland, Galician (Spain) rías, Norwegian fjords, and their contiguous offshore physical oceanographic features (coastal currents, fronts, upwelling). The following sections will evaluate the extent to which Scottish coastal waters are "closed" versus "open systems", and their resistance or vulnerability to externally induced changes in phytoplankton and HAB dynamics. This is of interest because HAB species have increasingly exploited coastal waters exhibiting altered and novel bloom behavior in an apparent global spreading of their harmful blooms. Four primary causation theories have been advanced to explain this phenomenon (Hallegraeff, 1993): the effects of cultural eutrophication; aquaculture and fish-farm initiatives; unusual climatological conditions, and the spreading of species through dispersal of their vegetative cells and resting cysts in ballast water and in transplanted shellfish stock. The first three of these mechanisms deal with habitat modification - regrouped here as the 'changing environment' theory. The fourth mechanism - the 'emigration' theory - addresses geographic range expansion and the role of ballast water in this. It is considered first.

11.2 Ballast water introductions of HAB species into Scottish waters

Dinoflagellate resting stages can survive ballast water transport and germinate under laboratory conditions (Hallegraeff and Bolch, 1992). This finding, combined with the novel blooms of three harmful dinoflagellate species previously unrecorded from Australian waters; the historical absence of their resting stages in local sediment cores at those sites, and their disjunct global distributions led to the theory that ballast water vectoring (i.e. facilitated emigration of harmful species) occurs (Hallegraeff and Bolch, 1992; Hallegraeff, 1998; Hallegraeff et al., 1988). This theory is based primarily on the persistent blooms of PSP toxigenic G. catenatum that now occur at Tasmanian aquacultural sites since this species was first detected there in the late 1980s. There is evidence that G. catenatum may have been ballasted from either Japan and/or Europe coastal waters (Hallegraeff, 1998). [ Gymnodinium catenatum, as mentioned earlier, is the major PSP toxigenic bloom species along the Iberian peninsula.] The ballast water theory has become inflated into the widely held view that such vectoring of species into new growth areas is an important contributor to the global HAB expansion in progress. Subba Rao et al. (1994) recorded about 100 taxa in vessels arriving from overseas ports, and expressed concern over the risk to Canadian mariculture posed by ballast water introductions of undesirable phytoplankton species.

There are four steps in the ballast transfer function: (1) ballast water and suspended sediments taken up at embarkation ports (or en route) capture phytoplankton cells and their resting stages; (2) these stages survive the ship's darkened ballast water chambers during transit; (3) their discharge at the deballasting ports, when accompanied by successful germination of cysts and/or vegetative cell survival, leads to local population growth; and (4) the expatriated population then spreads within local coastal currents. In this process, the ballast tanks can be both death chambers and growth incubators; which effect dominates is influenced by transit time, ballast water temperature, and the life cycle stage and ecophysiology of the captured cells and species (Gollasch et al., 2000). Ballasted species introductions may be sterile, with the species becoming extinct even if they achieve blooms (Smayda, 2002).

Transoceanic and regional expatriations of emigrant species leading to indigenous status and blooms require completion of a three stage colonization process (Smayda, 2002). The introduced (pioneer) species must first grow and complete its life cycle to establish a "founder" population; thereafter, this population must continue to grow at a level that will allow it to survive habitat variability and persist. Since invading species always interact with the resident species, they must establish and maintain themself within the community. Community entry is the third stage of the colonization process and involves competition for niche space. Once becoming a permanent member of the invaded community, subsequent bloom events are influenced by local growth conditions and species competition.

The extent to which these physical and ecological rigors limit successful ballasted-assisted dispersal and seedings of phytoplankton is poorly understood. Despite the increasingly invoked explanation that ballasted assisted introductions are a major factor in the global HAB expansion, there is little species-specific evidence in support of this theory (Smayda, 2002). This is not to deny such occurrences or potential ballasted-assisted bioinvasions. The following survey sketches some of the evidence that Scottish coastal waters are "open" to ballast water / sediment introductions of toxic and benign species of phytoplankton, and both immigrant and indigenous species.

The presence of HAB species in the ballast water and sediments in 127 vessels arriving at Scottish ports (Macdonald and Davidson, 1998) and 76 vessels to English and Welsh ports (Hamer et al., 2001) has been surveyed. The PSP-producing taxa listed as the Alexandrium tamarense / catenella complex and A. minutum were prominent among the 29 (Scotland) and 48 (England/Wales) species recorded. Gymnodinium catenatum was recorded in the Scotland survey, and the toxic species, Protoceratium reticulatum (= Gonyaulax spinifera), a producer of yessotoxin (Satake et al., 1999), in the England/Wales survey. Both surveys reveal the taxonomic problems that ballast water analysts confront in seeking to identify the species present from their abnormal appearance often acquired during the stress of ballast transport. Species identifications of ballasted populations are usually based on examination of resting cysts, as in the two UK surveys, a difficult task given that species complexes of closely related and morphologically similar taxa occur. For example, G. catenatum identified in the Scottish survey is an important PSP-producer along the Iberian peninsula and throughout its global distributional range. And, as mentioned earlier, it may have become indigenous in Tasmanian aquacultural areas through bioinvasion (Hallegraeff and Fraga, 1998). However, in European coastal waters, reliance on cyst morphology can easily lead to the misidentification of the non-toxic, "look-alike" species, Gymnodinium nolleri, as G. catenatum (Bolch et al., 1999). Such misidentification has falsely reported G. catenatum occurrences in the perimeter waters of the North Sea (Ellegaard and Moestrup, 1998; Nehring, 1998). Macdonald and Davidson (1998) acknowledged that the cysts on which they based their identification of G. catenatum may have been produced by another " similar species of Gymnodinium" . Hamer et al. (2001) found G. nolleri, but not G. catenatum during their survey. It seems likely that Macdonald and Davidson actually found non-toxic G. nolleri, and not toxic G. catenatum in the ballast of ships arriving in Scottish ports. However, their report can not be totally dismissed given the southern European occurrences of PSP-toxic G. catenatum.

Cysts identified as those produced by the A. tamarense / catenella species complex were common in the UK ballast water surveys. These species are also virtually indistinguishable based on cyst morphology. Alexandrium tamarense is widely distributed in European coastal waters, while chain-forming A. catenella has not been reported, even though it is considered to be a cold water species well known for its blooms along the west coast of North America, from Alaska to Chile, and in South Africa, Tasmania and Japan. A. catenella has recently appeared in the Mediterranean Sea where it is spreading along the north-west coast and producing annually recurrent blooms at ca. 20°C (Vila et al., 2001). Germination experiments with ballasted cysts produced motile, vegetative cells identified as A. tamarense (Hamer et al., 2001) . While it appears likely, then, that A. tamarense is the most likely source of the cysts assigned to the A. tamarense / catenella complex, the potential de-ballasting of A. catenella cysts into Scottish coastal waters can not be discounted given its Mediterranean Sea occurrences.

The common occurrence of toxic A. minutum in the ballast tanks prompted investigators to point out that it had not been reported previously from UK waters. Macdonald and Davidson (1998) stated that the colder waters along Scotland prevented A. minutum cysts from germinating into vegetative cells. Hamer et al. (2001) induced ballasted A. minutum cysts to germinate at 16°C. Töbe et al. (2001) subsequently claimed to have found motile, vegetative cells of A. minutum in Orkney coastal waters which they attributed to advection from oceanic provinces based on drift buoy observations. [Note: Töbe et al. (see p. 253 in 2001) attributed this source to Hummert et al. (2001), but the latter do not claim this in their publication]. It is uncertain whether an active A. minutum population occurs in Scottish coastal waters and, if so, whether it is seeded through germination from local resting stage seed beds, through advection, or from both sources. Ballast water introductions of A. minutum founder populations are viable initial sources, should this species have expanded its range into Scottish coastal waters.

Dinophysis spp., the genus associated with DSP, were also found in ballast samples, including Dinophysis caudata and Dinophysis tripos, species that Macdonald and Davidson (1998) characterized as uncommon in Scottish waters. Hamer et al. (2001) do not mention having found Dinophysis spp.

Three other potentially toxic dinoflagellate species were recorded in the ballast samples: Alexandrium margalefi, an obscure species known principally from Spain (Balech, 1995; Macdonald and Davidson, 1998); Cochlodinium sp. and Gonyaulax verior, also an obscure species (Hamer et al., 2001). Of these, Cochlodinium spp. have begun to attract attention because of their harmful bloom effects at shellfish culture and fish farm sites, notably Cochlodinium polykrikoides (Kim, 1998). Cysts of non-toxic Scrippsiella hangoei were abundant in a crude oil tanker arriving from Poland. This species forms intense winter - spring blooms in the northern Baltic Sea (Larson et al., 1995). Its presence in Scottish sea lochs has been debated. Larson et al. are skeptical that the cells which germinated from cysts collected in sea lochs, and provisionally identified by Lewis et al. (1984) as Protoperidinium hangoei, are conspecific with Scrippsiella hangoei.

Relatively large ballast water populations (up to 50,000 cells L ^-1) of two Pseudo-nitzschia species were also found: P. seriata and P. delicatissima. Pseudo-nitzschia species, responsible for Amnesic Shellfish Poisoning ( ASP), are common in Scottish coastal waters. They are also common components in ballast water exchanges between Temperate North American and Canadian coastal waters, where ASP has caused problems (Macdonald and Davidson, 1998).

Resting stages resembling those produced by the ichthyotoxic raphidophyte Chattonella antiqua were found in the ballast of two vessels entering English and Welsh ports (Hamer et al., 2001). The unprecedented series of toxic Chattonella blooms that have occurred in the Skagerrak since 1998 has been summarized in Section 7.1. The short transit time between that region and Scottish ports could facilitate ballasted introductions of toxic raphidophytes and haptophytes into Scottish coastal waters

A diverse assemblage of dinoflagellates indigenous to UK coastal waters was also found in ballast water, including Heterocapsa (= Katodinium) rotundatum, Heterocapsa triquetra, Lingulodinium polyedrum, Scrippsiella trochoidea, and numerous heterotrophic Protoperidinium spp. Both cyst and motile stages were found; the cyst stages of several species were induced to germinate in culture (Hamer et al., 2001). In cases where the introduced species are indigenous, whether toxic or benign, hybridization between the local and expatriated populations might result in a genetically modified cohort leading to altered bloom behavior. However, there is little data available to evaluate the influence of species' hybridizations in altering their bloom behavior and toxicity.

Clearly, Scottish coastal waters are "open" to ballast water and sediment introductions of toxic and benign species of phytoplankton, both immigrant and indigenous species. This exposure is of concern, given the significant increase in harmful and novel species blooms recorded in European waters the past three decades, including of species seemingly absent or insignificant in Scottish coastal waters. Scotland is in close proximity to continental Europe, and much of the ballast water potentially discharged at Scottish ports originates from European ports (Macdonald and Davidson, 1998). The relatively short transit times between these regional ports favor survival of ballasted vegetative cells and cysts and their deballasted seedings as bloom-starter populations. Vessels also arrive into UK ports from worldwide originations (see Table 2 in Hamer et al., 2001). There is concern over whether current International Maritime Organization ( IMO) guidelines are adequate to protect against harmful species introductions into UK waters. Macdonald and Davidson (1998) and Hamer et al. (2001) specifically expressed concern over recommendations that ballast water exchanges by shipping traffic within continental Europe be carried out in the North Sea, Irish Sea or English Channel. The retention of sediments (and cysts) in the ballast tanks during exchange with waters harboring bloom taxa, they believed, may compromise IMO guidelines.

The potential transfer of HAB species accompanying aquacultural transplantation and importation of shellfish stocks should not be overlooked. In an importation of Pacific oyster, Crassostrea gigas, to Ireland from France, 67 phytoplankton species were recorded from the gut contents and sediments of the consignment, including 15 dinoflagellate cyst producing species, three of which were harmful (O'Mahoney, 1993). Dijkema (1992) calculated that 2.5 million viable dinoflagellate cysts can be transferred per tonne of mussels imported into The Netherlands from "red tide" areas, yielding an annual introduction of ca. 10 ¹⁰ dinoflagellates into Dutch coastal waters.

It is recommended that Scottish authorities reevaluate the local procedures followed to minimize potential ballast water / sediment assisted bioinvasions of species that pose threats to fish farms, shellfish culture sites, and natural shellfish stocks.

11.3 Advection of HAB species in coastal currents

Ballast water vectoring is a human-assisted mechanism of dispersal that supplements a more natural, omni-present delivery system of species introductions: advective transport and dispersal of cells (species) during their entrainment within coastal currents. While ballast vectoring favors transoceanic introductions of novel species potentially expanding their ranges and leading to unusual or anomalous blooms, the advected regional and local dispersals of species from a common species pool functions as a bloom-seeding mechanism. The vagaries associated with such redistributions contribute to the unpredictability that generally characterizes bloom events of harmful species. This "natural" mechanism also potentially allows for the step-wise regional dispersal of species into new habitats where they may produce harmful blooms and achieve indigenous status. [This has been the case for the regional spreading of toxic Pyrodinium bahamense var. compressum in Indo-Pacific waters (see Hallegraeff and Maclean, 1989).] The purported ballast water delivery of PSP toxic G. catenatum into Scottish ports was discussed earlier. Under appropriate physical oceanographic and receptive habitat conditions, this species might also enter Scottish coastal waters via coastal currents. Thus, Scottish coastal waters are "open" to unpredictable seedings leading to blooms of species entrained within current systems advecting into, and away from local waters.

Harmful blooms developing at coastal sites after physical advection of seed species (communities) that had developed elsewhere, both nearfield and farfield, are well known. This bloom-seeding mechanism has been demonstrated for K. mikimotoi (= Gyrodinium aureolum) advected into Swedish (Lindahl, (1986), Norwegian (Dahl and Tangen, 1993) and Irish (O'Boyle et al, 2001) coastal waters; for A. tamarense dispersed along the north-east coast of the U.S. while entrained within a buoyant surface current (Franks and Anderson, 1992); Gulf Stream dispersion of ichthyotoxic Karenia brevis from its "natural habitat" in the Gulf of Mexico (Tester et al., 1991); the Indo-Pacific regional spreading of highly toxic ( PSP) Pyrodinium bahamense var. compressum (see Hallegraeff and Maclean, 1989).

Unlike for Scandinavia and Irish waters, the role of advective transport in harmful bloom dynamics in Scottish coastal waters is poorly known. This relationship in Scandinavian and Irish coastal waters is of special interest because the physical oceanographic conditions, the phytoplankton flora, and the intense aquaculture in those regions are similar to those found in Scottish coastal waters. Accordingly, the following summary of the influence of advection on harmful blooms in hydrographically complex Scandinavian and Irish [Section 3.4] coastal waters is presented as an analogue of behavior expected to take place in Scottish waters.

The spectacular, 1988 novel bloom of the haptophyte, Chrysochromulina polylepis, propagated along the Swedish and Norwegian coast accompanied by widespread mortality of the benthos and natural and farmed fish stocks, has been discussed in Section 7.1.2. The advective dispersion and serial seeding of Chrysochromulina cells was a major driver of this bloom. Cells from this bloom, initiated in the Kattegat in water of low salinity that entered from the Baltic Sea, became entrained within the Baltic Current and progressively spread to the north and west at a rate of 25 km d ^-1 into the Skagerrak (Figure 14). This dispersal serially seeded Norwegian coastal waters with cells entrained within the Norwegian Coastal Current which moved at a rate of 5 to 30 km d ^-1 (Dahl et al., 1989). The advected population eventually reached the Ryfylke fjord area in south-west Norway, where it dispersed. During this mesoscale event, a subsurface concentration of cells (at 15-20 m depth) became entrained within the Jutland Current, which underlies the Baltic Current, and moved southwards into the Kattegat.

In another advection event (Figure 15), serial seeding of Ceratium furca resulted in a widespread bloom that developed over a six week period extendig from the Skagerrak to mid-Norway (63°N) along its western coast. This bloom developed after a major outflow of low salinity water from the Skagerrak entrained cells from a bloom in the outer Oslofjord (59°N) and dispersed this population along the Norwegian coast where it continued to bloom (Johnsen et al., 1997). [Blooms of Ceratium spp. have caused anoxia in the Kattegat and German Bight (Granéli et al., 1989; Hickel, 1982)]. The duration and distance over which advective dispersion can seed bloom sites can be extensive. A bloom of the coccolithophore, Emiliania huxleyi, that began along the Norwegian west coast (60°N) in April passed the Arctic Circle (70°N) in July (Figure 14); Johnsen et al., 1997).

Advection of DSP-producing Dinophysis spp. into mussel farm sites and other embayments from offshore Skagerrak waters occurs (Dahl and Aune, 1996; Godhe et al., 2002), consistent with reports that the high Dinophysis population densities when found in French and Spanish coastal waters are the result of advection (see Godhe et al., 2002). Advection of Dinophysis acuta in Irish coastal waters leads to DSP toxin accumulation in farmed mussels (McMahon et al., 1998). Norwegian coastal blooms of K. mikimotoi are thought to originate from pycnocline populations in the open North Sea and Skagerrak (see Kristiansen et al., 1995).

The regional pattern of the massive 1998 Chattonella verruculosa bloom that developed in the Skagerrak and the North Sea off western Denmark was highly influenced by advective transport (Backe-Hansen et al., 2001). An unusual current system advected the bloom northward along the Swedish coast, then westward along the south-west coast of Norway where it looped to the south from Skagen along the west coast of Denmark and continued towards the German Bight.

11.4 Scottish waters as advective sources of HAB species

Lewis et al. (1995) discovered a major "seed bank" of dinoflagellate cysts in the Firth of Forth during a regional survey in the region between Aberdeen (57°03'N) and Flamborough Head (54°05'N) (Figure 16). This led them to endorse the theory that North Sea populations of A. tamarense are derived from cells excysted at that site and then advected southwards in currents. There is physical oceanographic evidence for this advection (Brown et al., 2001): a density-driven current (associated with strong bottom fronts) continuously flows southwards along the ca. 300 km north-east coast, from the Firth of Forth to Flamborough Head, before turning offshore to Dogger Bank (Figure 16). Brown and co-workers were able to match up the seasonal and regional occurrences and progression of A. tamarense and PSP monitored along this coastal stretch, and advective maintenance, dispersal and bloom-seedings in the southerly current flow, with an origination of A. tamarense in the vicinity of Firth of Forth and Farne Island. The evidence favoured a diffuse source of seed stock, rather than exclusive reliance on the Firth of Forth seed bank. At typical flow rates, the transit time from Farne Islands (south of Firth of Forth) to Flamborough Head (210 km) is 33 days, a delivery rate adequate to allow local seedings and blooms of A. tamarense en route. Similar advective behavior has been reported for A. tamarense entrained within buoyant surface currents along the north-east coast of the U.S. (Franks and Anderson, 1982). The jet-like coastal current passes eastward into the North Sea near the latitude of Whitby, skirting the north-west corner of the Dogger Bank. As a result, there is no significant transport A. tamarense to the south of Flamborough Head. Brown et al. (2001) linked the absence of PSP outbreaks south of Flamborough Head, despite a extensive shellfishery there, to this dispersion pattern.

Joint et al. (1997), however, rejected the notion that coastal current diffusion was the mechanism transporting A. tamarense cells or cysts >100 km down the north-east UK coast. They concluded that its blooms there develop from a diffuse supply, rather than a single source such as from the cyst "seed bank" in the Firth of Forth / Farne Island region. They state (p. 953 in 1997) " it is unlikely that there could be any significant transport of dinoflagellate cells or cysts in a coastal current". It is not clear from their presentation whether they offer this conclusion as a general dictum, or is restricted to local waters. If the former, then it is contradicted by the examples of advective dispersal reported from other regions, summarized above and to be considered further in another section. The dispute over the sources and advection of A. tamarense along the north-east UK coast is relevant to another problem posed by the proposed advection of A. tamarense south from the Firth of Forth and nearby seed stocks. As discussed earlier, Scottish strains of A. tamarense, including those from the Firth of Forth, are toxic in culture unlike English strains - there is a north-south gradient in PSP toxicity in UK waters (Higman et al., 2001). The unresolved question is whether A. tamarense cells advected south from north-east Scotland, and which grow during their transport towards Flamborough Head, i.e. in English coastal waters, retain or lose their toxicity. Determination of the toxic profiles of A. tamarense strains from along that gradient would help to resolve the conflicting views of Brown et al. (2001) and Joint et al. (1997).

11.5 Scottish waters as advective recipients of HAB species

11.5.1 Offshore advections

HAB events are highly unpredictable and stochastic, which suggests that blooms of a given species are often the result of " being in the right place at the right time" (Smayda and Reynolds, 2001). Incursions of seed stocks of regionally indigenous species resulting from growth elsewhere, and sustained and dispersed in the prevailing circulation patterns, are of interest. Seedings of species from a common, regional species pool would not confront the same colonization and growth hurdles that face ballast water introductions of novel and allochthonus (= non-indigenous, visiting) species (Smayda, 2002).

The circulation pattern and hydrographic complexity of the waters off western Scotland (Craig,1959; Milne,1972; Simpson and Hill, 1986; Rydberg et al., 2002) particularly favor intra-regional and farfield recruitments of phytoplankton cells advected as passive particles. The Scottish Coastal Current ( SCC), flowing northwards into the Minch through the North Channel, carries a mixture of Irish and Clyde Sea waters that is slightly diluted by the outflow of less saline water from fjordic sea lochs along the Scottish west coast (see Hill et al., 1997). The mean velocity (ca. 5 km d ^-1) of the SCC, probably buoyancy-driven and guided by bottom topography as it flows northwards into the Minch, contrasts with the sluggish flow rates found generally in UK shelf seas (Simpson and Hill, 1986). A striking feature of the SCC is its behaviour upon entering into the Minch where it encounters the Hebrides island chain. Caesium tracer (Cs ₁₃₇) distributions suggest that some of this flow turns westward, crosses the Minch, and then moves southward along the Hebrides before veering northwards again to flow along the western Hebrides (Simpson and Hill, 1986). Drifter deployments revealed current velocities during this flow ranged from 0.2 to 0.4 m s ^-1, rates considerable greater than the mean SCC velocity (0.06 m s ^-1; Hill et al., 1997). This circulation is accompanied by the formation of frontal systems, localized upwelling zones and stratified and stagnant water masses in the Sound of Jura (Jones et al., 1984) and (probably) contiguous waters. This complex of phytoplankton growth zones and currents, while spatially heterogeneous and temporally variable, represents the physical environmental template upon which bloom events originate, disperse and terminate. An analogue of the expected advective behavior of phytoplankton is provided by the larval drift of the commercially important Norway lobster ( Nephrops norvegicus) in the Minch (Hill et al., 1997). Larvae serially released annually from March to July are either retained locally or lost from the regionally depending upon the circulation pattern. Larvae from South Minch populations can be expatriated by currents to far distant sites in the North Minch and, even, Noup grounds. Thus, both nearfield and farfield advection of larval Nephrops occurs, even though it is capable of migratory and related "zooplanktonic" behaviour that would mute such transport. Such anti-drift countermeasures are expected to be considerable weaker among phytoplankton.

Insight into the potential dispersal distances of phytoplankton cells in Scottish waters has been sought assuming their entrainment within currents flowing at 0.2 and 0.4 m s ^-1 reported by Hill et al. (1997), and that the population doubling rate is one and two days. Table 9 summarizes the results of four combinations of dispersion and growth rates. Assuming a constant velocity, a single cell entrained in a current flowing at 0.4 m s ^-1 would be transported 175 km over a five day period; the population, assuming a constant growth rate of µ = 1.0 d ^-1 and no loss, would increase 32-fold during that period. In contrast, the slower growing population (µ = 0.5 d ^-1) would increase only 6-fold over that time and distance. The strength and potential impact of advective dispersal are functions of the current speed and the net population growth rate of the entrained cells. Farfield dispersal of slow growing cells can occur when entrained in high velocity current systems. When growth rates are high and the cells are entrained in high velocity currents, higher seed populations are available on both local and farfield scales. When cells are entrained in low velocity currents, nearfield seedings are favored, with the strength of this seeding also a function of prior growth rates.

Table 9. Advected distances traversed and relative population increases of entrained HAB species estimated for assumed combinations of current velocities and growth rates.

The advective velocities chosen for the calculations are among the higher rates for local waters. Extending the view of Joint et al. (1997), it might be argued that low velocity rates are accompanied by poor growth conditions that thwart advective seedings. The equivalent 5-day displacement distance for A. tamarense reported to have been advected ca. 210 km in 33 days is 30 km d ^-1 (Brown et al., 2001). At such low transport rates, the problems of population maintenance and survival become paramount, with periodic access to growth refugia and/or local recruitment en route needed to prevent extinction. If physico-chemical conditions favourable to growth are accessed during advective drift, the augmented population can continue as potential seed stock and initiate blooms upon periodic entry into suitable habitats found along the drift route. The hydrographic conditions reported for the Minch (as elsewhere) provide such recruitment zones: at fronts, upwelling loci and during exposure of the drifting cells to sites where vertical mixing fluxes nutrients upwards from depth. Frontal zones can serve as "pelagic seed banks" that facilitate the dispersal and regional seedings of harmful bloom species (Smayda, 2002). These depots of vegetative cells are the pelagic counterpart of cyst "seed beds". Stratified watermasses also aide flagellate survival and growth. The degree of stratification influences the irradiance field and nutrient supply and, hence, the population growth rate (Jones and Gowen, 1990). In the Sound of Jura and Firth of Lorne, the degree of mixing vs. stratification, scaled to depth and the irradiance field, determined the shift from a diatom to a dinoflagellate dominated community ( Alexandrium spp., Gymnodinium spp, Heterocapsa triquetra and Scrippsiella trochoidea). The switch to dinoflagellate dominance at those sites can take place within four days depending on current velocities.

11.5.2 Onshore - offshore advections

Onshore-offshore advections make aquacultural regions particularly vulnerable to HAB events, both as sources and sinks of seed populations. Most Scottish sea lochs are fjordic estuaries, characterized physically by freshwater-driven circulation and tidal exchange with adjacent coastal waters (Jones and Gowen, 1985). Tidal exchange influences onshore-offshore advections in both directions. Since many sea lochs are sites of fish farms, the outflow from these fjords is nutrient-conditioned (Tett and Edwards, 2002). Excreted nutrients in excess of local use, together with the phytoplankton species and biomass developing in response to this enrichment, are available for transfer to adjacent offshore areas, while tidal inflow advects offshore populations into the sea lochs. The capacity for hydrographic exchange and species advections varies among sea lochs. Jones and Gowen (1985) analyzed the summer flushing rates of 11 sea lochs and distinguished among three types. Type A sea lochs (Ardbhair, Eishort, Moldart, Slapin) are rapidly flushed, with flushing times <2.5 days. Type B sea lochs (Craignish, Creran, Seen, West Loch Tarbert) had flushing time scales ranging from 6 to 11 days. In Type C sea lochs (Etive, Nevis, Striven), flushing is very slow, on the order of several weeks. There is also considerable variability in tidal current velocities, which tend to be higher than geostrophic and baroclinic current velocities. Tidal velocities in Scottish coastal waters on the order of ~0.5 m s ^-1 are not unusual (Joint et al., 1997; Wyatt and Saborido-Rey, 1993).

Onshore-offshore advections occur in both directions; hence, sea lochs are both sources and sinks of nutrients and phytoplankton. Two primary onshore-offshore dispersions occur: alongshore advections and advection to offshore sites. In rapidly flushed sea lochs, the phytoplankton community, bloom species and biomass usually will be similar to that in offshore waters that are in tidal exchange with the sea loch, particularly where tidal velocities are high. The contributions of nutrients and phytoplankton from these sea lochs are blended into recipient waters, and their advective signatures become obscure. In regions of slowly flushed sea lochs, the phytoplankton will usually diverge from that in adjacent waters, partly because of the relatively more prolonged exposure of the sea loch phytoplankton to the altered water quality accompanying fish farming. These outflowing sea loch populations can then become entrained in alongshore currents and seed adjoining sea lochs during alongshore dispersion in a process similar to that reported for A. tamarense during its widespread regional bloom in 1990 [Section 4.2.1; Figs. 7, 17].

Offshore advections of sea loch nutrients and phytoplankton are of special interest, given concerns that nutrient wastes from fish farms may have stimulated the ASP blooms of Pseudo-nitzschia spp. leading to closure of commercial scallop harvesting grounds (Figure 2). In addition to the reasons presented in Section 2.11, i.e. that this stimulation is unlikely, the dilution and utilization of nutrients advected offshore from fish farms are expected to be attenuated and greatly diminish such fertilization effects. Moreover, as evident from the discussion above, Pseudo-nitzschia spp. are adapted to grow and bloom in shelf seas, behavior uncoupled from a dependency on delivery of nutrients from nearshore sources (Horner et al., 2000).

The primary onshore-offshore event of concern to aquaculture is the seeding of HAB species from offshore sites where the farmed fish and cultured shellfish can become exposed to introduced populations, or from local blooms of seeded populations. Such events, extensively documented in earlier Sections [3.4, 4.2, 5.1, 7.1], are further considered in the following Section.

In summary, inshore, nearshore and offshore waters are advective sources and sinks and bloom sites of species potentially harmful to fish farming, shellfish cultivation and natural stocks. Seed populations are advected along onshore-offshore gradients, from offshore in shelf waters, in coastal currents along latitudinal gradients, and alongshore. Both local and far-field sources provide the advected populations. These multiple population sources and advective pathways contribute significantly to the unpredictability that characterizes HAB events and making coastal waters open systems, both generally and, specifically, Scottish waters.

11.6 Advections of HAB species in Irish coastal waters: an analogue of Scottish coastal waters

Among the potential far-field sources of species advected into western Scottish waters are those originating in blooms along the coast of Ireland. It is not suggested that this is a primary, or even important source of phytoplankton populations advected into Scottish coastal waters. Rather, the relevant events and mechanisms described for Irish coastal waters are viewed as analogues of processes that take place in western Scotland waters given the similar phytoplankton communities, dynamics, hydrography and importance of fish farming and shellfish culture. An excellent series of satellite and synoptic field surveys has documented the frequent advective seedings and redistribution of harmful species detrimental to aquaculture in Irish coastal waters (O'Boyle et al., 2001; McMahon et al., 1998; Raine et al., 2001; among others). Blooms of K. mikimotoi, harmful to fish and shellfish, and the DSP-producing species, Dinophysis acuta and Dinophysis acuta, are advected from offshore growth regions into bays through wind-driven exchanges. Bloom populations within coastal embayments and adjacent shelf regions are also advected offshore and northward in coastal currents. The Irish Coastal Current flowing northwards along the western Irish coast dispersed K. mikimotoi into aquaculturally sensitive regions during unfavorable wind-forcing events (O'Boyle et al., 2001). Raine et al. (2001) have described the complex oceanography of southwestern Ireland coastal waters and its system of shifting, wind-induced upwelling zones, frontal systems and vertical thermocline structure that impact phytoplankton abundance and species composition. They describe a particularly large scale advection event involving K. mikimotoi and un-named species of Pseudo-nitzschia. An extensive surface "red tide" bloom formed a plume 150 km long and 40 km wide extending from the mouth of Bantry Bay southeastwards and across the northern Celtic Sea towards Cork. Dynamics such as this might lead to entrainment of cells within currents moving into western Scotland waters via passage through the North Channel (Figure 17). Other examples of advection of harmful species into Irish coastal waters are presented in Section 3.4.

Advective diffusion of HAB species into, and away from growth zones are commonplace in European waters. In addition to the advective seedings cited for Scandinavian and Irish coastal waters, advective seedings led to PSP blooms of G. catenatum in Spanish rías (Fraga et al., 1988; Figueras et al., 1995) and Dinophysis induced DSP outbreaks along the French and Spanish coasts (Lassus et al., 1993; Pazos et al., 1995; Reguera et al., 1995). In reality, HAB events in European coastal waters often are advected population responses. This first-order property of HAB dynamics is expected to occur in Scottish coastal waters.

11.7 Ecophysiological capacity of HAB species for advective seedings

The intra-regional and far-field expatriations of advected populations of the type expected in the Minch [Section 11.5.1] are primarily latitudinal (north-south) excursions, characteristic of shelf waters generally. There is also considerable habitat structure to sustain the growth and regional dispersion of phytoplankton. It is exactly in this ecological zone that many harmful species thrive, including species that have been, or are potential problems to Scottish aquaculture: Pseudo-nitzschia ( ASP) , Dinophysis ( DSP) , Alexandrium tamarense ( PSP) and K. mikimotoi (ichthyotoxic). The ability of these species to grow, bloom and disperse in open waters indicates that their dynamics are not dependent on aquacultural stimulation. There is greater need for the entrained cells to tolerate biophysically the shear and turbulent fields associated with advective transport, and for their vertical migration requirements to be accommodated by the mixing regime. The view that dinoflagellates (unlike diatoms) are intolerant of turbulence and shear, and require stratified conditions for nutrient stimulated growth is prevalent among phytoplankton ecologists. Margalef's classical Mandala codifies this putative requirement in its depiction of harmful dinoflagellate blooms as clustering into a single, basic red tide ecological zone of high irradiance, low turbulence and elevated nutrients (Margalef, 1978; Margalef et al., 1979). The Mandala also groups all red tides as similar phenomena independent of the bloom-species or bloom-habitat. Both paradigms are incorrect.

A recent evaluation of the experimentally derived motility rates for 71 flagellate species and the in situ distributions of dinoflagellates has revealed their relatively high tolerance to turbulence and vertical mixing and ability to carry out vertical migrations at high mixing rates (Smayda, 2002). And, Smayda and Reynolds (2001) have shown that dinoflagellates have diverse habitat preferences and adaptive strategies, rather than the uniform behavior depicted in Margalef's Mandala. Nine distinct life-form habitats were recognized, each habitat type selecting for a specific dinoflagellate species assemblage and adaptive strategy. The three harmful dinoflagellate species that are ubiquitous and pose threats to aquaculture in Scottish coastal waters - Alexandrium tamarense ( PSP), Dinophysis ( DSP) , and K. mikimotoi (ichthyotoxic) - all share a major and common life-form feature. They cluster in the three "mixing - drift" habitats recognized by Smayda and Reynolds (Types IV, V, VI), and in closely related habitat Type VII. They are adapted to the increased velocities associated with frontal zones and during entrainment within coastal currents: the frontal zone life-form (Type IV = K. mikimotoi) and the coastal current entrained life-form (Type VI = A. tamarense). Alexandrium tamarense and K. mikimotoi readily survive dispersal, and can grow when entrained within coastal currents; they also bloom in frontal zones. Mixing-drift species are disturbance-tolerant (= ruderal) R-species that tolerate, or depend upon entrainment within mixed or circulating water layers. Pre-adapted to withstand shear/stress effects, R-species have a light harvesting pigment system that allows growth at low irradiance levels (Smayda and Reynolds, 2001). R-species generally are also characterized by strong phototaxic capability, chain formation, a resting stage in their life cycle, marked behavioral and auto-aggregative ability, and produce toxins and/or induce mortality.

While Dinophysis species are also "drift" forms (Type VII), and can achieve modest blooms during upwelling relaxations (see Reguera et al., 1995), they appear to be better adapted to advection within the streamlines accompanying small-scale, Langmuir convection cells and other current systems in which shear/stress is less intense than in Type IV-VI habitats. Dinophysis species are primarily nutrient stress tolerant species ( S-strategists), and able to deal with low nutrient availability. Typically large and ornamented, they are capable of maintaining depth through motility alone (= vertical migration) (Smayda and Reynolds, 2001). They achieve modest, but persistent abundance, often possess endosymbionts, and supplement their photo-autotrophy by mixotrophy (see Hansen, 1991; Jacobsen and Andersen, 1994).

A similar diatom life-form classification is not available, as yet, to typify the harmful ( ASP) Pseudo-nitzschia species found in Scottish coastal waters. The autecology of Pseudo-nitzschia, discussed in Section 2.8, suggests a "mixing-drift" life-mode. This is supported by observations that blooms of advected P. delicatissima and P. pungens were enhanced by the patterns of coastal cross-shelf currents and upwelling/downwelling in north-west Pacific coastal waters (Horner et al., 2000).

Phytoflagellates, other than dinoflagellates, collectively have much lower swimming velocities (Smayda, 2000), suggesting a very diminished capacity for advective transport, with the possible exclusion of species from the raphidophyte genus Chattonella. Phytoflagellate blooms are usually restricted to nearshore habitats and occur during stratified conditions, a preference that may be influenced by their diminished vertical migration speeds.

11.8 Summary

Scottish coastal waters [and aquacultural sites] are open systems: open to HAB species introduced in ballast water and advected by currents. The invasion of K. mikimotoi into Scottish waters during its successful bioinvasion in northern European waters is an example of this openness. There is evidence that Scottish waters historically have been open systems. Molecular analyses of Alexandrium spp. led Scholin et al. (1995) to conclude that ancestral populations from the Pacific dispersed into the North Atlantic when the passage between Canada and Greenland opened. Medlin et al. (1998) believe that the opening between Greenland and Svalbard was the point of entry since toxic strains from Scottish waters can be assigned to the 'North American' lineage. Medlin et al. also suggest that toxic alexandrians in Scottish waters may be introductions, whereas non-toxic strains could be native to UK waters. This is in line with the evidence that the ability of A. tamarense to produce toxins may correlate with lineage: 'North American' and 'Temperate Asian' lineages consist exclusively of toxic strains, while non-toxic strains fall into the 'Western European" lineage (Higman et al., 2001).

Scottish coastal waters are both sources and recipients of advected populations (cells). Evidence for the latter is primarily circumstantial, but rooted in first principles and bloom species behavior observed in Scandinavian, Irish and Iberian waters, and at aquacultural sites in those regions. These have been described in earlier Sections of this report. Sordo et al. (2001) have described the sudden onsets of toxic blooms that can result from advection of PSP-toxic G. catenatum into Spanish rías.

Three basic types of blooms associated with advection are recognizable: blooms that develop offshore; onshore blooms seeded by populations advected from offshore waters, and blooms that develop inshore as a result of local, alongshore advections. Since turbulent mixing and current systems influence the drift of phytoplankton cells as passive particles, the combinations of these primary advective mechanisms contributing to a given bloom event vary.

Scottish coastal waters have some distinct HAB features, and seem remarkably closed to other HAB occurrences relative to European waters despite their openness and phytoplankton features shared in common with Scandinavian waters and the North Sea. The absence of harmful phytoflagellate blooms, and the preeminence of ASP and minimal DSP problems in Scottish waters are the converse of observations in France, Spain and Scandinavia. The diversity of HAB events in Irish coastal waters is seemingly much greater than in oceanographically similar Scottish waters, despite the similar species pool. The DSP compromised shellfish cultivation in Ireland is a particularly notable distinction from Scotland.

The HAB distinctions between Scotland and elsewhere in Europe can not be related to differences in aquaculture intensity or, within this, whether fish farming or shellfish cultivation is the more prominent enterprise.

« Previous | Contents | Next »