Ecological characterization of toxic phytoplankton species (Dinophysis ...

This article was originally published in a journal published by Elsevier, and the attached copy is provided by Elsevier for the author’s benefit and for the benefit of the author’s institution, for non-commercial research and educational use including without limitation use in instruction at your institution, sending it to specific colleagues that you know, and providing a copy to your institution’s administrator. All other uses, reproduction and distribution, including without limitation commercial reprints, selling or licensing copies or access, or posting on open internet sites, your personal or institution’s website or repository, are prohibited. For exceptions, permission may be sought for such use through Elsevier’s permissions site at: http://www.elsevier.com/locate/permissionusematerial

Marine Pollution Bulletin 52 (2006) 1504–1516 www.elsevier.com/locate/marpolbul

co

py

Ecological characterization of toxic phytoplankton species (Dinophysis spp., Dinophyceae) in Slovenian mariculture areas (Gulf of Trieste, Adriatic Sea) and the implications for monitoring Janja France *, Patricija Mozeticˇ

al

National Institute of Biology, Marine Biology Station, Fornacˇe 41, 6330 Piran, Slovenia

on

Abstract

pe

rs

Diarrhetic shellfish poisoning (DSP) events are often registered in Slovenian mariculture areas (Gulf of Trieste, Adriatic Sea) and are related to the occurrence of Dinophysis spp. The annual dynamic of this genus and succession of the most important species were studied at two shellfish farms during monitoring fieldwork in the period 1995–2003. Results indicate that the Dinophysis genus maintains a relatively stable inter-annual dynamic at both sites. The Dinophysis community is characterized by two surface maxima in June and September, while in the middle layer only the autumn peak is pronounced (peak median 92 cells l 1). Occasional abundance maxima of around 2000 cells l 1 in the surface layer indicate that potential outbursts of toxic species are less predictable than their seasonal dynamic. On the basis of multivariate analysis, Dinophysis sacculus was characterized as a typical late spring–early summer species, and Dinophysis caudata and Dinophysis fortii as autumn species. Correlation analysis revealed the influence of stratified conditions only on the most abundant species, D. sacculus. Ecological characteristics of the species were combined with shellfish safety requirements towards a more effective monitoring. 2006 Elsevier Ltd. All rights reserved.

r's

Keywords: Dinophysis spp.; Inter-annual dynamic; Environmental factors; DSP; Monitoring; Adriatic Sea

1. Introduction

Au

th o

Species belonging to the Dinophysis genus are found in the phytoplankton communities of coastal waters throughout the world but they rarely reach abundances higher than 100 cells l 1 (Hallegraeff and Lucas, 1988). Their appearance and persistence in the water column is controlled by several factors. Following the general consensus that dinoflagellates prefer warm and stratified waters as well as conditions of low turbulence (Paerl, 1988; Smayda, 1997) several authors confirmed these preferences for the Dinophysis species per se (Peperzak et al., 1996; Palma et al., 1998). More ambiguity can be found in connection with relation-

*

Corresponding author. Tel.: +386 5 671 29 34; fax: +386 5 671 29 02. E-mail address: [email protected] (J. France).

0025-326X/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.marpolbul.2006.05.012

ship between Dinophysis growth and inorganic nutrient availability. Some authors argue that occurrence of Dinophysis is apparently independent of concentrations of major nutrients (e.g., Delmas et al., 1992; Lindahl and Andersson, 1996), while others observed enhanced growth of Dinophysis cells in nutrient rich waters, whether due to increased phosphate (Cabrini et al., 1995; Santhanam and Srinivasan, 1996) or nitrogen (Caroppo et al., 2001). Other factors besides nutrients and water column stratification may play an important role in the dynamics of Dinophysis populations (Caroppo, 2001) such as polymorphic life histories, inter- and intra-specific relationships and prey availability for mixotrophic and heterotrophic species. The mixotrophic nutrition mode can be a significant factor for the periodic predominance of dinoflagellates in the water column as well as for their role in aquatic food webs and ecosystem dynamics (Lessard, 1991; Stoecker, 1999). In addition, heteromorphic life forms (e.g., temporary

J. France, P. Mozeticˇ / Marine Pollution Bulletin 52 (2006) 1504–1516

2. Materials and methods

on

al

co

py

Field sampling was carried out in the area of two mussel farms in the south-eastern part of the Gulf of Trieste (northern Adriatic) in the period from 1995 to 2003. The sampling stations are located in two shallow embayments: Strunjan (station A, depth of 14 m) and Secˇovlje (station B, depth of 12 m) (Fig. 1). Samples for cell counts were collected biweekly from May to November, i.e., in the period that turned out to be critical for the appearance of toxic or potentially toxic species (Sidari et al., 1995; Poletti et al., 1998), and monthly in other months. From 1995 to 1999 and again in 2002 and 2003 a quantitative sampling method was used: sea water samples were taken with a 5 l Niskin bottle at the surface and from the 6 m layer. This layer corresponds to the depth at which shellfish stocks are positioned in the sea water. In contrast, in the years 2000 and 2001 a semi-quantitative method of vertical net hauls was adopted with the use of 20 lm plankton net. Since the latter sampling method did not provide us comparable results on cell counts, we did not use 2000 and 2001 data in the analysis. In 1997, measurements of the main inorganic nutrients (ammonium, nitrite, nitrate and phosphate) in the water samples, collected at 0 and 6 m depth, were carried out according to standard colorimetric methods (Grasshoff et al., 1983) with the use of a Perkin Elmer UV/VIS Lambda 14 Spectrophotometer. The temperature, salinity and density profiles of the water column were measured at each sampling from 1997 onward by the using a CTD probe (Centre for Water Research, Western Australia). Water samples for phytoplankton counts were fixed with 2% (final concentration) formaldehyde neutralized with hexamethylenetetramine immediately after being brought to the laboratory and then examined on an inverted microscope ZEISS Axiovert 135 following Utermo¨hl

Au

th o

r's

pe

rs

and resting cysts) allow regulation of dinoflagellates’ seasonal succession through dispersal, seeding and repopulation processes (Pfiester and Anderson, 1987; Montresor, 2001). Several Dinophysis species are found in the Adriatic Sea and they are generally implicated in incidences of diarrhetic shellfish poisoning (DSP) along the northern Adriatic coast (Mozeticˇ and Obal, 1995; Poletti et al., 1998). DSP events are regularly registered in the central (Marasovic´ et al., 1998) and southern Adriatic (Caroppo, 2001), as well. In the northern Adriatic Dinophysis species are present in water samples mainly from May to October with abundance peaks in early summer and early autumn (Sidari et al., 1995; Poletti et al., 1998). During this period a marked species succession can be observed (Della Loggia et al., 1993; Bernardi Aubry et al., 2000). The present study was performed in the shallow waters along the Slovenian coast of the Gulf of Trieste (northern Adriatic). In these waters the first DSP case related to an increase of Dinophysis cells was reported in 1989 (Sedmak and Fanuko, 1991). Since then mussel intoxications have been periodically registered and are the cause of bans on the selling of shellfish (Sedmak et al., 2003). Previous studies have dealt with short-term seasonal occurrences of the Dinophysis species in the northern Adriatic (e.g., Malej et al., 1994; Sidari et al., 1995). The aim of this work was to reveal the seasonal occurrence patterns of the most recurrent Dinophysis species and to determine the predictability of their succession with the help of multivariate statistical analysis on the nine-year monitoring data. An attempt was made to relate Dinophysis appearance to selected environmental parameters. As the knowledge of species ecology is essential for the effective control of harmful algal blooms and the taking of proper measure in case of their emergence, our objective was also to improve the existing mussel watch protocol.

1505

Fig. 1. Map of the study area indicating sampling stations A and B.


py

test the differences in monthly cell abundances between the two stations and between the two sampled depths. The relationship between Dinophysis species and environmental parameters in the surface water layer was tested using correlation analysis after log transformation of the original abundance data. The significance was calculated using the Student’s t-test. 3. Results

co

3.1. Dinophysis diversity and seasonal dynamic

al

During the nine-year study period (from 1995 to 2003) 16 species of Dinophysis were found in samples from stations A and B: Dinophysis acuminata Clapare`de et Lachmann, Dinophysis acuta Ehrenberg, Dinophysis caudata SavilleKent, Dinophysis diegensis Kofoid, Dinophysis fortii Pavillard, Dinophysis hastata Stein, Dinophysis mitra (Schu¨tt) Abe´ vel Balech, Dinophysis norvegica Clapare`de et Lachmann, Dinophysis operculoides (Schu¨tt) Balech, Dinophysis ovum Schu¨tt, Dinophysis parva Schiller, Dinophysis recurva Kofoid and Skogsberg, Dinophysis rotundata Clapare`de et Lachmann, Dinophysis sacculus Stein, Dinophysis cf. similis Kofoid and Skogsberg and Dinophysis tripos Gourret. Temperature and salinity conditions in the surface and 6 m layer of both sampling stations during the study period are illustrated in Fig. 2. Mean surface water temperature

Au

th o

r's

pe

rs

(1958). 1 l of each sample was left to settle in 1 l cylinders for at least 48 h; thereupon 200 ml of the concentrated sample was split into quarters. One quarter of this concentrated sample was then left to settle in the sedimentation chamber for 24 h. The whole chamber bottom was counted at 200· magnification. Identification of Dinophysis species was based on Dodge (1982), Rampi and Bernhard (1980), Schiller (1931–1933), Steidinger and Tangen (1996) and Faust and Gulledge (2002). The dynamics of the Dinophysis genus as a whole is presented at two depths: the surface and the 6 m layer. For the evaluation of the occurrence pattern of the four most frequent and abundant Dinophysis species through selected years, the STATIS multi-table method (Lavit et al., 1994) was used, as it combines within-year and between-years analyses. Furthermore, the STATIS method is suitable for the comparison of years in which the numbers of samples are not equal as was the case in present study. The STATIS analysis was performed using ADE-4 software (Thioulouse et al., 1997). Only surface layer data was taken into account in the STATIS analysis of seasonality since in this layer seasonal peaks of the entire dinoflagellate community were found to be the most pronounced (France and Mozeticˇ, 2003). Details on the procedure and computations in STATIS analysis can be found in Licandro and Ibanez (2000) and Anneville et al. (2002). The Mann–Whitney U-test and the Kolmogorov–Smirnov test were used to

on

1506

Fig. 2. Temperature and salinity (mean ± SD) in the surface and 6 m water layers of stations A and B in the period 1997–2003.


co

py

increased gradually from June until the autumn peak (September median of 62 cells l 1 at station A and October median of 92 cells l 1 at station B) (Fig. 3). Therefore, during the June peak Dinophysis abundances were significantly higher in the surface layer in comparison with the 6 m layer at both stations (Mann–Whitney U = 44.5, p = 0.05 and Kolmogorov–Smirnov D = 0.51, p = 0.05 for station A and U = 39, p = 0.03 and D = 0.59, p = 0.01 for station B). In contrast, the autumn peak was significantly higher in the 6 m layer at station B (Mann–Whitney D = 24, p = 0.05) while there were no differences at station A. During the winter months and up to May, the number of Dinophysis cells in the water column was low or undetectable. Except for a few very high abundances (up to some 103 cells l 1) observed at station A in June, most of the highest monthly abundances were noted at station B.

al

3.2. STATIS analysis of predominant Dinophysis species

on

3.2.1. Inter-structure analysis – similarity between different years Four Dinophysis species were regularly found in the samples with abundances high enough to be considered as predominant species: D. caudata, D. fortii, D. rotundata and D. sacculus. We therefore present their detailed dynamics as revealed by STATIS analysis. Graphical representation (Fig. 4) of the STATIS first step – the interstructure analysis – is based upon the correlation matrix and shows the eigenvectors for individual years on the plane made by the 1st and 2nd axes. For station A, a high

Au

th o

r's

pe

rs

ranged from 8.9C in February to 25.3C in August at station A and from 9.0C in February to 25.6C in August at station B. The highest temperature was about a degree lower in the 6 m layer of both stations, whereas February temperatures were nearly the same in both layers. Mean salinity values displayed larger variations in the surface layer varying from 35.53 to 37.94 and from 35.54 to 37.67 at stations A and B, respectively. The lowest surface salinities were recorded in June and July (32.76) and a second minimum was registered in November (32.21). The highest mean surface salinities were noted during winter months. The mean salinity range was much narrower in the 6 m layer, with the lowest values of 36.21 in June at station A and 36.41 in July at station B. Water column stratification persisted from May to September with the most pronounced stratification in June. From October to April the water column was generally well mixed. The Dinophysis genus as a whole displayed very similar annual dynamics at the two sampling sites with no statistically significant differences. It showed two annual maxima in the surface layer, the first in June (medians of 60 cells l 1 at station A and 73 cells l 1 at station B) and the second in September (medians of 57 and 39 cells l 1 at stations A and B, respectively) (Fig. 3). It can be seen in the Box–Whiskers plots that abundance was particularly variable during these peaks given the largest inter-quartile distance. The interquartile distance was large in November as well, despite the low median, indicating that Dinophysis abundance occasionally peaked during this month. The dynamics was different in the 6 m layer, where cell abundances

1507

Fig. 3. Box–Whiskers Plots representing the annual dynamics of Dinophysis species in the surface and 6 m layers at stations A and B in the period 1995– 2003 (2000 and 2001 excluded). Note the different scale on y-axis.


co

py

1508

on

3.2.2. Intra-structure analysis – the stable part of the annual pattern Additional intra-structure analysis (analysis of the compromise) was carried out to reveal the annual pattern and to better explain the differences/similarities among years. The graphical illustration of the analysis of the compromise (Fig. 5) shows the average position of each species in respect to the 1st and 2nd axes. Each position corresponds to the single species average occurrence during the entire investigation period. Interpretation of these figures is eased by a comparison with Fig. 6 that shows the projections of all samples and months barycentre on the same planes. The first two axes of the analysis of the compromise account for a very high percentage of the cumulative inertia (82% for both sampling sites) providing, hence, a good summary of the species annual dynamics.

Au

th o

r's

pe

rs

first eigenvalue deriving from the correlation matrix was calculated, which explains 82% of the total inertia. Considering this value, the uniform distribution of the years’ eigenvectors on the 1st axis (all eigenvectors have high values on the 1st axis, Fig. 4(a)) indicates a relatively strong common annual structure. On the 2nd axis of the interstructure analysis (10% of the total inertia) two years appear to be separated from the rest: 1995, which had a shorter sampling period, and 1997, both with negative values on the 2nd axis. The results of the inter-structure analysis for station B (Fig. 4(b)) do not differ a lot from that of station A. The years’ eigenvectors all have high values on the 1st axis, indicating again a large similarity among years. The only isolated year appeared to be 2002 with a high negative value on the 2nd axis in which the sampling was interrupted at the end of June.

al

Fig. 4. STATIS inter-structure analysis: Position of each single year on the plane defined by 1st (I) and 2nd (II) axes for stations A (a) and B (b). Proximity of the arrows shows the similarity between different annual structures.

Fig. 5. STATIS intra-structure analysis: Projections of each single species on the plane defined by 1st (I) and 2nd (II) axes of the analysis of the compromise for stations A (a) and B (b). Projections correspond to the average position of each species during the study period and show the global relationship between species.

1509

co

py


on

(Fig. 6(a)) despite its quite uniform distribution. The typical autumnal species is D. fortii (Figs. 5(a) and 6(a)). It is hardly found in water samples until September; afterwards it persists till November but in general in low numbers (Me < 10 cells l 1) although abundances of approximately 100 cells l 1 are occasionally found (September 1996). At station B the annual occurrence pattern of the four Dinophysis species is similar, even though with some differences. Barycentre for December, January and February are placed in the origin of the compromise plane (0, 0) (Fig. 6(b)) since no cells of Dinophysis were found in these months. The 1st axis of the analysis divides spring and early summer months on the positive side from summer and autumn months on the negative side. In addition, the 2nd axis separates autumn months (October, November) on the positive half from summer (August, September) on the negative half. D. sacculus displays essentially the same annual pattern as at station A, confirming it as a typical June species. The maximum abundance of D. sacculus (398 cells l 1 in June 1999; Fig. 7) was, however, lower than at station A. D. caudata is present in the water samples from early summer onwards (Fig. 7). It reaches its highest but relatively modest median values in the period July–September which is reflected in the position of D. caudata on the compromise plane (Figs. 5(b) and 6(b)). Nevertheless, its maximum abundances were found later on in autumn (145 cells l 1 in October 1997 and 224 cells l 1 in November 2003). D. rotundata has a distribution similar to that of station A (Fig. 7) but with slightly higher abundances (maximum 55 cells l 1 in September 1999). In contrast to station A, D. rotundata is positioned on the positive side of the 1st axis (Figs. 5(b) and 6(b)) as it was, on average, more abundant during the early summer months. The autumn species D. fortii has a similar occurrence pattern (Figs. 5(b) and 6(b)) with comparable median abundances (Fig. 7) to those at station A.

Au

th o

r's

pe

rs

The 1st axis of this analysis separates spring months (May and June) on the positive side from summer and autumn months on the negative side for station A (Fig. 6(a)). In addition, the 2nd axis separates summer months (July, August and September) on the positive half from autumn months (October and November) on the negative half. The barycentre of February, March, and April are positioned in the origin of the plane (0, 0); in none of these months were any of the four Dinophysis species detected in seawater samples. The barycentre of December and January probably does not reflect the proper position due to low number of findings (one and two findings in December and January, respectively). Other barycentre fit reasonably well into the annual structure (see Fig. 3). Thus, the first Dinophysis species becoming abundant is D. sacculus, a typical late spring–early summer species (Figs. 5(a) and 6(a)). It starts to appear in May and reaches its maximum in June. During this peak the abundance is mostly below 100 cells l 1 (Fig. 7), although we occasionally observed much higher abundances of more than 1000 cells l 1 (2876 cells l 1 in 1996 and 1543 cells l 1 in 1997). D. sacculus usually disappears by the end of summer and is succeeded by D. caudata (Figs. 5(a) and 6(a)). D. caudata can be found in samples from May onward but remains rather rare until July. On average, it reaches the highest abundance in September and persists in the water column until December. However it has generally low abundances with the median not exceeding 25 cells l 1 and occasional maxima of approximately 150 cells l 1 (Fig. 7). Another species typical for the second half of the year is D. rotundata that shows the lowest abundance among all four species. Its abundance never exceeds 25 cells l 1 and the highest median abundances are