Harmful Algae 39 (2014) 365–373

Changing wind patterns linked to unusually high Dinophysis bloomsaround the Shetland Islands, Scotland

Callum Whyte *, Sarah Swan 1,2, Keith Davidson 1,2

Microbial and Molecular Biology Department, Scottish Association for Marine Science, Scottish Marine Institute, Oban, Argyll PA37 1QA, United Kingdom

A R T I C L E I N F O

Article history:

Received 9 May 2014

Received in revised form 16 September 2014

Accepted 17 September 2014

Available online

Keywords:

Dinophysis

Harmful algal bloom

Dinophysistoxins

Advection

Wind pattern

Shetland Islands

A B S T R A C T

During the summer of 2013, 70 people received Diarrhetic Shellfish Poisoning following consumption of

mussels harvested in the Shetland Islands, Scotland. At this time, large numbers of the biotoxin-

producing phytoplankton genus Dinophysis was observed around the Shetland Islands. Analysis

indicated this increase was not due to in situ growth but coincided with a change in the prevalent wind

direction. A previous large bloom of Dinophysis during 2006 also coincided with a similar change in the

prevalent wind patterns. Wind direction and speed in the North East Atlantic and the North Sea is

strongly influenced by the North Atlantic oscillation (NAO) with a positive relationship between it and

wind direction. It has been noted that a positive trend in the NAO is linked to climate change and

predictions suggest there will be an increasingly westward component to prevalent wind directions in

the North Sea which could lead to an increase in the occurrence of these harmful algal blooms. Analysis of

wind patterns therefore offers a potential method of early warning of future bio-toxicity events.

� 2014 Published by Elsevier B.V.

Contents lists available at ScienceDirect

Harmful Algae

jo u rn al h om epag e: ww w.els evier .c o m/lo cat e/ha l

1. Background

In July 2013 there was an outbreak of Diarrhetic ShellfishPoisoning (DSP) in the South East of England. Although it is notknown how many people were actually affected, 70 people soughtmedical advice after consuming mussels which, they had beenserved in several prominent restaurants in London. They were laterfound to be suffering the symptoms of DSP. The shellfish had beensupplied by a farm in the coastal waters of the Shetland Islands(Fig. 1). Although the numbers of Dinophysis had been slowlyincreasing during the preceding few weeks, analysis did notindicate that the concentration of toxins in the shellfish was high.However three days after the farmer had harvested the shellfish anunusually high level of toxins was detected by the Food StandardsAgency (FSA) weekly monitoring programme (FSA, 2013), accom-

Abbreviations: ASIMUTH, Applied Simulations and Integrated Modelling for the

Understanding of Toxic and Harmful Algal Blooms; BADC, British Atmospheric Data

Centre; DSP, Diarrhetic Shellfish Poisoning; DTX (1–3), dinophysistoxins; GDP,

Gross Domestic Product; FSA, Food Standards Agency; NAO, North Atlantic

Oscillation; OA, okadaic acid; SAMS, Scottish Association for Marine Science.

* Corresponding author. Tel.: +44 01631 559 226.

E-mail addresses: callum.whyte@sams.ac.uk, calwhyte@gmail.com (C. Whyte),

sarah.swan@sams.ac.uk (S. Swan), keith.davidson@sams.ac.uk (K. Davidson).1 These authors contributed equally to this work.2 Tel: +44 01631 559 226.

http://dx.doi.org/10.1016/j.hal.2014.09.006

1568-9883/� 2014 Published by Elsevier B.V.

panied by a very rapid accumulation of Dinophysis cells. The areawas closed and commercial harvesting was suspended, unfortu-nately too late for the luckless consumers of the affected mussels.This paper investigates the possibility that the unusually largeblooms of Dinophysis observed around the coast of Shetland during2013, and another large bloom event during 2006, were not due tohigh in situ growth rates, but rather to changes in the annualpattern of prevalent wind direction. We discuss the possibility thatan analysis of wind patterns could be used to provide an earlywarning of biotoxic events.

Yasumoto et al. (1980) was among the first to suggest thatgastro-intestinal pain, accompanied by diarrhoea, nausea andvomiting, following the consumption of mussels (Mytilus edulis)was linked to the presence of dinoflagellates in the water column,in particular Dinophysis fortii. This link was confirmed afteroutbreaks of severe gastro-intestinal shellfish poisoning in Japanduring 1976 and 1977 and led to the identification of a new set oftoxins associated with the genus Dinophysis (Reguera et al., 2012).

Identified as Diarrhetic Shellfish Poisoning (DSP) causativetoxins, these lipophilic compounds are comprised principally ofOkadaic Acid (OA) and its derivatives DTX-1, DTX-2 and DTX-3(Smayda, 2006; Taylor et al., 2013). They have been found to bindwith protein phosphatase receptors in the body, leading to anaccumulation of phosphorylated proteins with symptoms oflethargy, general weakness, vomiting, cramps and diarrhoea,

C. Whyte et al. / Harmful Algae 39 (2014) 365–373366

symptoms that can appear from 30 min to several hours afteringestion of the contaminated shellfish (Gerssen et al., 2010).

The Dinophysis species most commonly found in Scottishwaters are D. acuminata, D. acuta, and D. norvegica, all of whichhave been confirmed to produce toxins with particular emphasison D. acuminata and D. acuta (Bresnan et al., 2005; Hart et al.,2007). While contamination of both farmed and wild shellfish byOA and DTX’s is sporadic, it is widespread and has been reportedfrom eleven countries across Europe including Denmark, France,Germany, Ireland, Italy, Norway, Portugal, Spain, Sweden and theNetherlands (Smayda, 2006). Perhaps the severest outbreakoccurred in Spain during 1981 when an outbreak of DSP affectedapproximately 5000 people, mostly from Madrid (Fraga et al.,1984, 1988; Gestal-Otero, 2000).

While unrecorded cases are likely to have occurred, the firstreliable record of Diarrhetic Shellfish Poisoning in the UK was in1997 when 49 people became ill after eating shellfish in twoLondon restaurants (Hinder et al., 2011; Scoging and Bahl, 1998).Since then the frequency of closures due to DSP around the UKappears to be increasing. This may simply be due to a betterunderstanding of the underlying causes of shellfish poisoning, agreater awareness of the problem and a more rigorous monitoringprogramme. Unlike some species of harmful algae, Dinophysis doesnot need to be present in large numbers to cause toxicity inshellfish. A few hundred cells/l can be enough to lead to the closureof a harvesting area. In 2000, DSP toxins were found in shellfish inareas on the East coast of Scotland, Orkney, the Outer Hebrides, theShetland Islands and the Firth of Clyde. In all, closures of thevarious shellfish harvesting areas amounted to 24 weeks (Howardet al., 2001). In 2002, DSP toxins were found throughout Scotlandwith the majority of the closures lasting between four to six weeks,although some areas were closed for up to seven months. Between2008 and 2009, thirteen areas were closed in the Shetland Islandsand seven areas in Argyll and Bute (Hinder et al., 2011) and mostrecently, large areas of Shetland were closed for several weeksduring 2013.

In addition to the extremely distressing effects contaminatedshellfish can have on anyone who consumes them, a DSP outbreakcan have far reaching consequences for the shellfish industry.While there is an obvious detrimental effect on farmers, processorsand distributors, the impact can extend well outside the closedharvesting area. In what is sometimes referred to as a ‘‘halo’’ effect(Rural Affairs Committee, 1999) consumer confidence is often lost,not only in the closed fishery, but in shellfish products in general,regardless of the area where they were harvested. This loss ofconfidence can easily outlast the closure itself and, as much ofScotland’s production is destined for overseas markets, can haveserious economic impacts for the country’s GDP (Rural AffairsCommittee, 1999).

DSP events are particularly damaging to aquaculture in theSouth West of Ireland, an area responsible for 80% of the ropegrown blue mussel (Mytilus edulis) and 50% of the pacific oyster(Crassostrea gigas) cultivated in Ireland. Production in this areasuffers from a history of harmful algal events; often closing thearea for months at a time (Raine et al., 2010). The frequency ofshellfish harvesting area closures due to DSP events in SouthernIreland, particularly in the large bays of Bantry, Dunmanus andLong Island, would suggest that these areas are particularly suitedfor the growth of Dinophysis. However the hydro-dynamicprocesses, in particular, the exchange of substantial fractions ofthe bay volume during thermal stratification, present in these bayssuggests that they are unsuited to the development of indigenousphytoplankton populations (Raine et al., 2010). Instead, it isbelieved to be wind driven water exchange between the Irish coastand the continental shelf that is responsible for these events (Raineet al., 2010). Similarly, Escalera et al. (2010) have confirmed that

accumulations of Dinophysis species in Galician rıas during Octoberand November were due not to intrinsic growth but to physicaltransport from other areas.

In Scottish waters Dinophysis are widely distributed, usually inlow numbers (Tett and Edwards, 2002; Davidson and Bresnan,2009) and are generally found in offshore waters. Once advectedonshore, however, they can accumulate in semi-enclosed areaswhere in the right conditions they may thrive. Given theirwidespread distribution, outbreaks of DSP can occur almostanywhere around the Scottish coast.

Responsible for 54% of the mussel production in Scotland with105 businesses involved in shellfish production (Shetland inStatistics, 2011), Shetland plays an important role in the Scottisheconomy. Given its geographical location and its topography,comprised of numerous fjords and embayments, it provides anideal environment for shellfish aquaculture. However, as theclosures due to an exceptionally large bloom of Dinophysis

experienced during the summer of 2013 illustrate, it is alsovulnerable to blooms of these harmful algae. The ScottishGovernment has called for an increase in shellfish productionfrom 6,525 tonnes in 2012 to a proposed 13,000 tonnes by 2020. Itis important therefore to understand the mechanism underlyingthis unusual event and try to estimate its likelihood of recurrence.

2. Methods

Following the FSA guidelines, seawater samples integrated overa depth of 10 m were collected by Lund tube as part of the FSAphytoplankton monitoring programme for Scottish waters. Thesamples were gently mixed and 500 ml sub-samples transferred tobrown, Nalgene bottles and immediately treated with acidifiedLugol’s solution to obtain a final concentration of 1% by volume.Samples were collected weekly from nine distinct sites aroundShetland between April and October. From November to March thefrequency of sampling was reduced to one sample per month. Aftercollection, these samples were sent to the Scottish Association forMarine Science (SAMS) where 50 ml aliquots were settled usingthe Utermohl sedimentation method as outlined in Lund et al.(1958). Following FSA protocols the phytoplankton was allowed tosettle for a minimum of 20 h before examination. Full chambercounts at 200� magnification were carried out using Carl ZeissAxiovert inverted microscopes.

While regulatory biotoxin phytoplankton surveillance andmonitoring has occurred in Scotland since 1991, early samplingwas spatially and temporally variable. However, since 2006 a morestructured collection regime has been used, with FSA-fundedsampling officers collecting waters samples from designatedrepresentative monitoring points within classified shellfishproduction areas. Hence, to determine whether the number ofDinophysis cells recorded in Shetland during 2013 had changedsignificantly, a plot of all the Dinophysis counts made between2006 and 2013 was created. The median was calculated for eachday of the year between 2006 and 2013 and was drawn onto thisplot (see Fig. 2 solid line) along with the 5th and 95th percentiles.These are represented in the figure by the dashed lines and form anenvelope encompassing 90% of the observations made. Theindividual datum points have been omitted for clarity. Counts ofDinophysis recorded during 2013 were then plotted against thismedian and a Chi square test was performed, making theassumption that values recorded during 2013 would be equallydistributed around the median. Using the same envelope for countsmade between 2006 and 2013 this method was used to investigateDinophysis counts made during 2006 (see Fig. 3).

Active phytoplankton growth in Scottish waters occurs mostlybetween spring and autumn when light, water temperature andnutrients are sufficient (Davidson et al., 2011). As the monitoring

Fig. 1. Map of location showing sampling sites mentioned in the text.

Fig. 2. (A) Dinophysis counts recorded in Shetland during 2013 (circles) plotted

against median values of Dinophysis recorded in Shetland between 2006 and 2013

(solid black line). Dashed lines represent the 5th and 95th percentiles of the data

collected between 2006 and 2013. (B) Dinophysis counts recorded in Shetland

during 2006 (circles) plotted against median values of Dinophysis recorded in

Shetland between 2006 and 2013 (solid black line). Dashed lines represent the 5th

and 95th percentiles of the data collected between 2006 and 2013.

C. Whyte et al. / Harmful Algae 39 (2014) 365–373 367

programme takes place throughout the year, albeit at a reducedfrequency during the winter months, many of the observationsmade in the early and late parts of the year consist of zero counts.This has the tendency to drag the median down and can bias

summer counts, making them appear more significant than theyare. To counter this, counts were transformed using Log10(X + 0.5z),where X = the number of Dinophysis cells counted and z = thedetection limit, in this case 20 cells/l.

To ascertain wind direction, data were downloaded from theBritish Atmospheric Data Centre (BADC) (BADC, 2012) for Lerwick,Shetland (BADC station identifier src-id 9). The BADC providedhourly wind direction and wind speed for each day of the year.These data were then averaged to obtain a mean daily scalar windspeed and direction for each year between 2006 and 2013. A‘‘climatology’’ was then created by calculating the mean values foreach day of the year between 2006 and 2013. To generate themonthly wind roses these mean daily values were binned intosixteen 22.5 degree bins and plotted in Microsoft Excel.

2.1. Growth rates

Any significant changes in the number of Dinophysis recordedcould potentially be due to natural, inter-annual differences ingrowth rates. In situ and cultured growth rates for Dinophysis

species grown at different temperatures vary widely in theliterature. In an attempt to choose a realistic value, only thosegrowth rates found from in situ studies involving those species ofDinophysis growing in Scottish waters and at the temperaturesfound in the waters around Shetland were used. To estimate theexponential growth rates of Dinophysis around Shetland, a meanvalue of 0.32 d�1 calculated from the values shown in Table 1 wasused. In situ growth rates cover a wide range of values and thevalue chosen can have a marked effect on the outcome. To addressthis, growth rates were also calculated for � one standard deviationfrom the mean. As with growth rates, loss rates can also vary widely inthe literature. Zhai et al. (2010) working in the North-west Atlanticcalculated loss rates between 10% and 60%, while Kozlowsky-Suzukiet al., 2006 estimated 25% losses from copepod grazing in the BalticSea. Again taking a conservative approach, the lowest of these lossrates, i.e. 10% was chosen.

Dinophysis is a mixotroph, requiring light, nutrients and liveprey to thrive. Unfortunately, suitable data on nutrient levels,irradiance and the availability of prey that would allow a detailedgrowth model to be constructed for the sites in question were notavailable. Nevertheless taking a conservative approach, andassuming idyllic growth conditions existed at the sites, a simple

Fig. 3. Dinophysis numbers recorded at (A) Braewick Voe, (B) Ronas Voe, (C) East of Linga and (D) Busta Voe, Shetland during July and August 2013 (solid line) plotted alongside

the numbers of Dinophysis that would be expected from a simple exponential growth model with 10% losses (dashed line). The exponential growth rates for � one standard

deviation are represented by the lighter shaded lines. Diagrams (E) Basta Voe Cove and (F) Scarvar Ayre, both located on the East coast of Shetland, are shown as a comparison.

Table 1Dinophysis growth rates used in calculations (see text).

Growth rate (1/d) Species Authors

0.11 Dinophysis acuta Stolte and Garces (2006)

0.22 Dinophysis acuminata

0.63 Dinophysis norvegica

0.40 Dinophysis acuminata Velo-Suarez et al., 2009

0.45 Dinophysis caudata

0.11 Dinophysis acuminata Tong et al. (2010)

0.23 Dinophysis acuminata

0.47 Dinophysis acuta Farrell et al. (2013)

0.57 Dinophysis acuta

0.10 Dinophysis norvegica Gisselson et al. (2002)

0.17 Dinophysis norvegica

0.40 Dinophysis norvegica

C. Whyte et al. / Harmful Algae 39 (2014) 365–373368

model of exponential growth with 10% losses was used to predictthe numbers of Dinophysis (N) that could be expected to be presentduring the bloom, i.e. dN/dt = rmN (0.9), where rm = 0.32 d�1.

3. Results

3.1. Increase in Dinophysis spp.

Fig. 2A shows the distribution of Dinophysis counts madebetween 2006 and 2013. Counts of Dinophysis made in 2013 havebeen plotted over this envelope providing a graphic illustration ofthe year’s distribution. It is obvious from Fig. 2A that 2013 hadexceptionally high levels of Dinophysis. Indeed a Chi squared testcarried out, making the assumption that the 2013 observationsshould be equally distributed along the median, showed asignificant change, x2 = 42.13, p < 0.001 in Dinophysis numbers.

Fig. 2B was constructed in a similar way to Fig. 2A. In this casethe situation is not quite as visually clear as that in 2013.Nevertheless carrying out a Chi square test, again making theassumption that the 2006 observations should be equally

Table 2Chi square values for Dinophysis numbers recorded between 2006 and 2013. When calculating the Chi square value for the whole year 2006, 2007, 2008 and 2010 and 2013 all

showed a significant difference in Dinophysis numbers. However when the growth period between May and August was considered only 2006 and 2013 showed a significant

change (see text for further explanation).

Year Significant Higher Lower Chi square Month Significant Chi square

2006 x x 29.57 p < 0.001 May–August x 8.17 p < 0.01

2007 2.22 May–August 1.31

2008 x x 33.94 p < 0.001 May–August 1.33

2009 0.12 May–August 2.27

2010 x x 5.23 p < 0.05 May–August 0.55

2011 2.49 May–August 1.36

2012 1.66 May–August 0.17

2013 x x 42.13 p < 0.001 May–August x 36.5 p < 0.001

C. Whyte et al. / Harmful Algae 39 (2014) 365–373 369

distributed along the median, still shows that there was asignificant change in the numbers of Dinophysis observed in2006 relative to the 2006–2013 median, x2 = 29.57, p < 0.001.

The results of calculating Chi square tests for each of the yearsbetween 2006 and 2013 can be found in Table 2. Although not aslarge as that found during 2006 and 2013, a significant increase inannual Dinophysis numbers was also observed in 2010. In Shetlandthe highest numbers of Dinophysis occur in the summer betweenMay and August. A Chi square test carried out on each year for thisperiod reveals that only 2006 and 2013 show a significant increasein Dinophysis numbers relative to the 2006–2013 median.

3.2. Net growth rates

Fig. 3 shows the result of plotting the actual numbers ofDinophysis cells counted per litre (solid line) in (A) Braewick Voe,(B) Ronas Voe, (C) Vaila Sound and (D) Busta Voe, Shetland, duringJuly 2013, against the numbers that would be expected, assumingexponential growth of Dinophysis, at these sites (represented bydashed line, exponential growth � 1 standard deviation representedby lighter shaded lines). It can be seen that in each case the numbersof Dinophysis observed exceeded the numbers that would beexpected. As a comparison the actual numbers of Dinophysis cellscounted per litre (solid line) from two additional sites (E) Basta Voeand (F) Dales Voe, both located on the east coast of the ShetlandIslands, are also shown. In these two locations the accumulation ofDinophysis is less than that expected from exponential growth.

3.3. Wind Patterns

Fig. 5 was produced after constructing a climatology for theperiod between 2006 and 2013 then extracting the mean prevalentwind directions for June, July and August, shown in diagrams (A),(C) and (E). It can be seen that the mean wind direction for theperiod is mostly from the South (S) or South, South East (SSE).

Any surface water movement induced by wind from thisdirection would tend to move water northwards towards Shetlandfrom the North Sea. In contrast during 2013 the prevalent winddirection changed dramatically to the West (Fig. 4B, D, F)particularly during June and July (Fig. 4B, D) when the unusuallyhigh numbers of Dinophysis were observed. Interestingly, as thewind patterns began to move back towards a more Southerlydirection in August (Fig. 4F) the numbers of Dinophysis began toreturn to more usual densities.

Fig. 5 shows the same mean prevalent wind direction observedbetween 2006 and 2013 at Lerwick, Shetland but in this case it iscompared with the mean wind direction observed during June, Julyand August, 2006 (Fig. 5B, D, F). While the bloom during 2006 didnot reach the same densities as that observed during 2013 theincrease in numbers of Dinophysis observed was still significant.Again it can be seen that there was a major change in winddirection during June, July and August. It is interesting to note that

while the prevalent wind directions observed in August 2006 weresimilar to those in 2013, the southerly wind speeds were lowerthan those in 2013 and represented a smaller percentage of thedays during August when the wind blew from the south. This mayaccount for the difference in the Dinophysis densities in the twoyears (Fig. 2). While the Dinophysis bloom observed during 2013ended rather abruptly at the end of July, corresponding to thechange in wind direction, the bloom observed during 2006 peakedagain in August of that year before gradually declining duringSeptember and October.

4. Discussion

It is apparent from Fig. 2 that significantly more Dinophysis wasobserved in Shetland during 2006 and 2013. Although thereappeared to be significant increases in annual Dinophysis numbersduring 2007, 2008 and 2010 (see Table 2), these are not reflected inthe changes observed during the summer months when numbersof Dinophysis are expected to be at their highest. This appears to bean artefact of the large amount of zero values in the dataset. Thesetend to accrue towards the beginning and end of the year whenDinophysis is scarce. Occasional observations of Dinophysis cells,albeit in low densities at these times of the year, strongly affect theresults of a Chi square test when median values are inevitably zero.Analysis of the data for the period between May and August of eachyear reveals that only 2006 and 2013 show a significant increase inDinophysis numbers.

The information presented in Fig. 3 suggests that, rather thanideal growing conditions accounting for the rapid increase inDinophysis numbers, the increase was due in part to somemechanical means. Diaz et al. (2013) found that changes to localwind patterns affected upwelling events offshore and speculatethat this increased the availability of suitable prey, sustaining largeDinophysis populations which were then advected into the GalicianRıas Baixas. The experience from the bays of the west coast ofIreland (Raine et al., 2010) is that rapid increases in Dinophysis

numbers are again associated with wind driven water advection.Our analyses suggest a similar process may occur in the voes ofShetland. During 2013 unusual wind patterns were observedduring June, July and August that, in contrast to the normallyprevalent Southerly winds, blew from the West. Edwards et al.(2006) note that over the last four decades the centre fordistribution of Dinophysis in the North Sea has moved eastwardstowards the coast of Norway. They also observe that there appearsto have been a Northwards shift in the distribution of Dinophysis.

In this case it could be expected that the usual southerly windswould advect water from the Eastern side of the North Sea which asEdwards et al. (2006) have observed, now contains relatively fewercells of Dinophysis. Instead, Westerly winds would advect wateroriginating on the edges of the Faroe-Shetland channel. Smaydaand Reynolds (2001), classify Dinophysis as a type VII dinoflagellatelife-form, adapted to survive in coastal currents and to grow in

Fig. 4. A series of wind roses illustrating the mean prevalent wind direction observed during (A) June, (C) July and (E) August between 2006 and 2013 compared with the

prevalent wind direction observed during (B) June, (D) July and (F) August 2013 at Lerwick, Shetland. Higher wind speeds are shown by darker shades. The percentage of days

during the month when the wind blew from a given direction is represented by the length of the diamond.

C. Whyte et al. / Harmful Algae 39 (2014) 365–373370

weak frontal systems. The conditions to the west of Shetland alongthe edges of this channel which, experience regular upwellingevents, followed by periods of relaxation, could be expected toprovide ideal conditions for the growth of Dinophysis.

It is interesting to note that the two sites shown that werelocated on the East coast of Shetland did not show any greatincrease in Dinophysis numbers during the same period nor did therates of accumulation of Dinophysis exceed that predicted by

Fig. 5. [A series of wind roses illustrating the mean prevalent wind direction observed during 2006]. A series of wind roses illustrating the mean prevalent wind direction

observed during (A) June, (C) July and (E) August between 2006 and 2013 compared with the prevalent wind direction observed during (B) June, (D) July and (F) August 2006

at Lerwick, Shetland. Higher wind speeds are shown by darker shades. The percentage of days during the month when the wind blew from a given direction is represented by

the length of the diamond.

C. Whyte et al. / Harmful Algae 39 (2014) 365–373 371

C. Whyte et al. / Harmful Algae 39 (2014) 365–373372

exponential growth. This conclusion is given strength by theobservations made during the summer of 2006, when unusuallyhigh numbers of Dinophysis were again accompanied by a changein the prevailing winds from South to West.

The changes in Dinophysis distribution reported by Edwardset al. (2006) appear to be related to higher temperatures duringwinter. A similar change has been found by Wasmund et al. (1998).Working in the Baltic Sea, they observed a change from diatom todinoflagellate dominance and also connected this to milderwinters. Given that dinoflagellates make up the biggest proportionof toxic algae, rising sea temperatures could have seriousconsequences for outbreaks of harmful algal blooms around ourcoasts.

These impacts will only be acerbated by changes in theprevalent wind patterns. Evidence that winds in the North Sea arebecoming stronger and that they increasingly have a west or south-westerly component is growing (Siegismund and Schrum, 2001;Brown, 2009; Christensen et al., 2007). Large scale bipolar patternssuch as the NAO are known to influence both the speed and thedirection of winds in the North East Atlantic and the North Sea(Phillips et al., 2013) and as these are positively linked with climatechange (Gillett et al., 2003) the incidence of changing windpatterns in Shetland may be set to increase and with them thepossibility of more prevalent bio-toxic events connected to thosespecies such as Karenia mikimotoi (Davidson et al., 2009), andDinophysis spp. that are influenced by Westerly, wind advectedcurrents.

Predicting harmful algal blooms is fraught with difficulties.Rigorous monitoring is still the safest way to ensure thatconsumers can eat shellfish with confidence and without worryingabout food poisoning. However, as the recent event in Londonillustrates, the possibility for contaminated shellfish to enter thefood supply still exists. Models, such as those developed in the EUproject ASIMUTH (ASIMUTH, 2014), that predict harmful algalblooms are gaining in accuracy and sophistication, although thereis no room for complacency. As the frequency of harmful algalblooms around the globe is perceived to be on the increase, and asthe levels of investment in aquaculture rise, an understanding oftheir underlying causes, accompanied by tools to help mitigatetheir impact is more important than ever.

Competing interests

None.

Authors contributions

CW has contributed in generation of data, analysis and paperwriting. SS has contributed in generation of data and paper writing.KD has contributed in analysis and paper writing.

Author’s information

For the past nine years SS has been the manager of themonitoring programme for the presence of toxin producingplankton in shellfish production areas in Scotland. Her interestslie in the distribution and development of harmful algal bloomsand the taxonomy of potentially toxic phytoplankton.

KD is the head of the Microbial and Molecular BiologyDepartment at the Scottish Association for Marine Science. Hiscurrent interests lie in the processes governing the initiation ofharmful algal blooms and the influences that bathymetry andphysical drivers play in phytoplankton distribution and produc-tivity.

CW is a member of the Department of Microbial and MolecularBiology and deputy manager of the monitoring programme for the

presence of toxin producing plankton in Scottish shellfishproduction areas. His research interests lie in phytoplanktonecology; specifically those few species that produce bio-toxins. Heis trying to understand the drivers that lead to blooms of harmfulalgae, with the eventual goal of developing tools and methods thathelp to identify and predict their occurrence.

Acknowledgements

The authors would like to acknowledge the NERC PURE researchprogramme and the EU FP7 project ASIMUTH who supplied thefunding for this study. They would like to thank the following fortheir assistance in identifying and enumerating the phytoplanktonsamples: Deborah Brennan, Christine Campbell, Eilidh Cole, JoanneFielding, Sharon McNeill, Elaine Mitchell and Rachel Saxon. Theywould also like to thank the British Atmospheric Data Centre andthe Food Standards Agency for the generous use of their data.[TS]

References

ASIMUTH, 2014. Applied Simulations and Integrated Modelling for the Understand-ing of Toxic and Harmful Algal Blooms. EC FP7 Space Theme, http://www.asi-muth.eu/en-ie/Pages/default.aspx.

BADC: UK meteorological Office, 2012. Met Office Integrated Data Archive System(MIDAS) Land and Marine Surface Stations Data (1853-Current). NCAS BritishAtmospheric Data Centre, http://badc.nerc.ac.uk/view/badc.nerc.ac.uk__A-TOM__dataent_ukmo-midas.

Bresnan, E., Fryer, R., Hart, M., Percy, L., 2005. Correlation Between Algal Presence inWater and Toxin Presence in Shellfish. Fisheries Research Services ContractReport 04/05 27 pp.

Brown, S., 2009. Future Changes in Storms and Anticyclones Affecting the UK,Climate CHANGE Projections Report, UKCP09, Annex 6, pp. 181–186.

Christensen, J.H., Hewitson, B., Busuioc, A., Chen, A., Gao, X., Held, I., Jones, R., Kolli,R.K., Kwon, W.T., Laprise, R., Magana Rueda, V., Mearns, L., Menendez, C.G.,Raisanen, J., Rinjke, A., Sarr, A., Whetton, P., 2007. Regional climate projections.In: Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor,M., Miller, H.L. (Eds.), Climate Change 2007: The Physical Science Basis. Contri-bution of Working Group 1 to the Fourth Assessment Report of the Intergov-ernmental Panel on Climate Change. Cambridge University Press, Cambridge,UK/New York, NY, USA.

Davidson, K., Bresnan, E., 2009. Shellfish toxicity in UK waters: a threat to humanhealth? Environ. Health 8, 4.

Davidson, K., Miller, P., Wilding, T.A., Shutler, J., Bresnan, E., Kennington, K., Swan, S.,2009. A large and prolonged bloom of Karenia mikimotoi in Scottish waters in2006. Harmful Algae 8 (2) 349–361.

Davidson, K., Tett, P., Gowen, R.J., 2011. Harmful algal blooms. In: Harrison, Hester,(Eds.), Marine Pollution and Human Health. RSC Publishing, pp. 95–127.

Diaz, P.A., Reguera, B., Ruiz-Villarreal, M., Pazos, Y., Velo-Suarez, L., Berger, H.,Sourisseau, M., 2013. Climate variability and oceanographic settings associatedwith interannual variability in the initiation of Dinophysis acuminata blooms.Mar. drugs 11, 2964–2981.

Edwards, M., Johns, D., Leterme, S., Svendsen, E., Richardson, A., 2006. Regionalclimate change and harmful algal blooms in the northeast Atlantic. Limnol.Oceanogr. 51 (2) 820–829.

Escalera, L., Reguera, B., Moita, T., Pazos, Y., Cerejo, M., Cabanas, J.M., Ruiz-Villarreal,M., 2010. Bloom dynamics of Dinophysis acuta in an upwelling system: in situgrowth versus transport. Harmful Algae 9 (3) 312–322.

Farrell, H., Velo-Suarez, L., Reguera, B., Raine, R., 2013. Phased cell division, specificdivision rates and other biological observations of Dinophysis populations insub-surface layers off the south coast of Ireland. Deep Sea Res. Part II: TopicalStud. Oceanogr. 101, 249–254.

Fraga, S., Marino, J., Bravo, I., Miranda, A., Campos, M.J., Sanchez, F.J., Costas, E.,Cabanas, J.M., Blanco, J., 1984. Red tides and shellfish poisoning in Galicia (NWSpain). In: ICES Special Meeting on the Causes, Dynamics and Effects ofExceptional Marine Blooms and Related Events, Copenhagen, 4–5 October1984, p. 5.

Fraga, S., Anderson, D.M., Bravo, I., Reguera, B., Steidinger, K.A., Yentsch, C.M., 1988.Influence of upwelling relaxation on dinoflagellates and shellfish toxicity in Riade Vigo, Spain. Estuar. Coast. Shelf Sci. 27, 349–361.

FSA, 2013. Shellfish Poisoning Outbreak, http://www.food.gov.uk/news-updates/news/jul/shellfish.

Gerssen, A., Pol-Hofstad, I.E., Poelman, M., Mulder, P.P.J., van den Top, H.J., de Boer, J.,2010. Marine toxins: chemistry, toxicity, occurrence and detection, with specialreference to the Dutch situation. Toxins (Basel) 2 (2010 (4)) 878–904, http://dx.doi.org/10.3390/toxins2040878 (published online 23.04.10).

Gestal-Otero, J.J., 2000. Nonneurotic toxins, University of Santiago de Compostelaand University Hospital of Santiago de Compostela, Santiago de Compostela,Spain. Seafood Freshw. Toxins: Pharmacol. Physiol. Detect. 51.

C. Whyte et al. / Harmful Algae 39 (2014) 365–373 373

Gillett, N.P., Graf, H.F., Osborn, T.J., 2003. Climate Change and the North AtlanticOscillation. The North Atlantic Oscillation: Climatic Significance and Environ-mental Impact AGU, Washington, DC, pp. 193–209.

Gisselson, L.-A., Carlsson, P., Graneli, E., Pallon, J., 2002. Dinophysis blooms in thedeep euphotic zone of the Baltic Sea: do they grow in the dark? Harmful Algae 1(4) 401–418.

Hart, M.C., Green, D.H., Bresnan, E., Bolch, C.J., 2007. Large subunit ribosomal RNAgene variation and sequence heterogeneity of Dinophysis (Dinophyceae) speciesfrom Scottish coastal waters. Harmful Algae 6 (2) 271–287.

Hinder, S., Hays, G., Brooks, C., Davies, A., Edwards, M., Walne, A., Gravenor, M.,2011. Toxic marine microalgae and shellfish poisoning in the British isles:history, review of epidemiology, and future implications. Environ. Health 10 (1)54.

Howard, G., Bresnan, E., Petrie, J., 2001. The biotoxin monitoring programmes forScotland 2000–2001. Shellfish News, CEFAS 11.

Kozlowsky-Suzuki, B., Carlsson, P., Ruhl, A., Graneli, E., 2006. Food selectivity andgrazing impact on toxic Dinophysis spp. by copepods feeding on natural plank-ton assemblages. Harmful Algae 5 (1) 57–68.

Lund, J.W.G., Kipling, C., Cren, E.D., 1958. The inverted microscope method ofestimating algal numbers and the statistical basis of estimations by counting.Hydrobiologia 11 (2) 143–170.

Phillips, M.R., Rees, E.F., Thomas, T., 2013. Winds, sea levels and North AtlanticOscillation (NAO) influences: an evaluation. Global Planet. Change 100, 145–152.

Raine, R., McDermott, G., Silke, J., Lyons, K., Nolan, G., Cusack, C., 2010. A simpleshort range model for the prediction of harmful algal events in the bays ofsouthwestern Ireland. J. Mar. Syst. 83 (3–4) 150–157.

Reguera, B., Velo-Suarez, L., Raine, R., Park, M.G., 2012. Harmful Dinophysis species: areview. Harmful Algae 14 (0) 87–106.

Rural Affairs Committee, 1999. 2nd Report on Subordinate Legislation, Impact ofAmnesic Shellfish Poisoning on the Fisheries Sector, http://archive.scottish.-parliament.uk/business/committees/historic/x-rural/reports-99/rar0201b.

Scoging, A., Bahl, M., 1998. Diarrhetic shellfish poisoning in the UK. Lancet 352(9122) 117.

Shetland in Statistics, 2011. Shetland Islands Council: Shetland Litho. Gremista,Lerwick.

Siegismund, F., Schrum, C., 2001. Decadal changes in the wind forcing over theNorth Sea. Clim. Res. 18 (1/2) 39–45.

Smayda, T.J., Reynolds, C.S., 2001. Community assembly in marine phytoplankton:application of recent models to harmful dinoflagellate blooms. J. Plankton Res.23 (5) 447–461.

Smayda, T.J., 2006. Harmful Algal Bloom Communities in Scottish Coastal Waters:Relationship to Fish Farming and Regional Comparisons – A Review. NaturalScotland Scottish Executive.

Stolte, W., Garces, E., 2006. Ecological aspects of harmful algal in situ populationgrowth rates. In: Graneli, E., Turner, J. (Eds.), Ecology of Harmful Algae. Springer,Heidelberg, Berlin, pp. 139–152.

Taylor, M., McIntyre, L., Ritson, M., Stone, J., Bronson, R., Bitzikos, O., Rourke, W.,Galanis, E., Team, O., 2013. Outbreak of diarrhetic shellfish poisoning associatedwith mussels, British Columbia, Canada. Mar. Drugs 11 (5) 1669–1676.

Tett, P., Edwards, V., 2002. Review of Harmful Algal Blooms in Scottish CoastalWaters. Scottish Environmental Protection Agency, Stirling.

Tong, M., Zhou, Q., David, K., Jiang, T., Qi, Y., Donald, A., 2010. Culture techniques andgrowth characteristics of Dinophysis acuminata and its prey. Chin. J. Oceanol.Limnol. 28 (6) 1230–1239.

Velo-Suarez, L., Garces, E., Reguera, B., Wyatt, T., 2009. Vertical distribution ofdivision rates in coastal dinoflagellate Dinophysis spp. populations: implicationsfor modelling. Mar. Ecol. Prog. Ser. 385, 87–96.

Wasmund, N., Nausch, G., Matthaus, W., 1998. Phytoplankton spring blooms in thesouthern Baltic Sea spatio-temporal development and long-term trends. J.Plankton Res. 20, 1099–1117.

Yasumoto, T., Oshima, Y., Sugawara, W., Fukuyo, Y., Oguri, H., Igarashi, T., Fujita, N.,1980. Identification of Dinophysis fortii as the causative organism of Dhiar-rhetic Shellfish Poisoning. B. Jpn. Soc. Sci. Fish. 46 (11) 1405–1411.

Zhai, L., Platt, T., Tang, C., Sathyendranath, S., Fuentes-Yaco, C., Devred, E., Wu, Y.,2010. Seasonal and geographic variations in phytoplankton losses from themixed layer on the Northwest Atlantic Shelf. J. Mar. Syst. 80 (1–2) 36–46.