1
“Effects of the ‘Prestige’ oil spill on macroalgal assemblages: large-scale
2
comparison”
3
Carla Lobón1, Consolación Fernández, Julio Arrontes, José Manuel Rico, José Luis Acuña,
4
Ricardo Anadón, Augusto Monteoliva
5
Departamento de Biología de Organismos y Sistemas, Unidad de Ecología, Universidad de Oviedo. C/
6
Catedrático Rodrigo Uría s/n 33071 Oviedo, Spain.
7
1Corresponding
8
ABSTRACT
9
An assessment of the effects of the ‘Prestige’ oil spill on intertidal, macroalgal
10
assemblages was carried out comparing abundance data obtained before and after the
11
spill. Three zones in the North and Northwest coast of Spain were sampled, one of them
12
located at the immediate vicinity of the spill, the zone most heavily oiled. Macroalgal
13
assemblages had similar structure between years. Neither critical decrease in abundance
14
of the dominant macroalgae, nor increase in opportunistic species was found. Some
15
differences in abundance were found, but they did not show any pattern, being more
16
likely the result of the natural variability of the ecosystem. Extensive, but not intense
17
fuel deposition on the shores, and a limited use of aggressive cleanup methods are
18
suggested as possible causes for the lack of the effects in these assemblages after the
19
‘Prestige’ oil spill.
20
Keywords: ‘Prestige’ oil spill; macroalgal assemblages; community structure;
21
species distribution; pollution; North of Spain.
author: e-mail: carlalobon.uo@uniovi.es, fax: (+34) 98 510 4777
22
1
23
1. Introduction
24
In November 2002, the ‘Prestige’ oil tanker sank at ca. 130 nautical miles off the
25
Galician Coast, carrying more than 77000 tons of heavy fuel oil M-100. This fuel is
26
characterized by its low solubility and volatility, which makes it very persistent over
27
time (Markarian et al., 1993). After the wreckage, the ‘Prestige’ released more than
28
10000 tons of fuel oil, which were carried by prevailing winds and ocean currents and
29
reached extensive areas of the Cantabrian Coast, N of Spain (Montero et al., 2003;
30
García-Soto, 2004; Acuña et al., submitted). After sinking, the ‘Prestige’ gradually
31
released the rest of its fuel during ca. 4 months, causing a series of oil waves, which
32
mainly affected the Galician Coast, NW of Spain (Acuña et al., submitted). Except for
33
most oiled areas located at the "Costa da Morte" (Galicia, Fig. 1), oil deposition on the
34
Atlantic and Cantabrian Spanish coasts was extensive, but not very intense, affecting
35
mainly a 100 Km coastal zone E of Cape Peñas (Acuña et al., submitted; Fig. 1).
36
To date, moderate to negligible effects of the ‘Prestige’ oil spill have been documented
37
on benthic (Serrano et al., 2006) and planktonic (Varela et al., 2006; Bode et al., 2006;
38
Salas et al., 2006) communities, and on cell and tissue condition biomarkers in mussel,
39
hake and anchovy (Marigómez et al., 2006). This is consistent with low fuel contents
40
measured
41
http://www.ieo.es/prestige/resultados.htm) and in the water column off the Cantabrian
42
coast (IEO http://www.ieo.es/prestige/resultados.htm; González et al., 2006). Therefore,
43
the most likely target to detect any significant ecological effect should be the shoreline,
44
where much of the spilled fuel was deposited (Acuña et al., submitted).
during
the
winter
2002-2003
in
shelf
sediments
(IEO
2
45
Proper assessment of the ecological effects of disturbances requires baseline time-series
46
studies, documenting the situation before the impact (e.g. for application of a BACI
47
sampling design, Underwood 1992). Information of the natural variability of the
48
ecosystem is essential to differentiate and evaluate the effects of anthropogenic impacts.
49
In spite of the remarkable recurrence of oil spills in certain areas, including the Galician
50
coast, baseline studies with adequate replication for hypothesis testing are rarely
51
available (Teal and Howarth, 1984). During 2000 and 2002, right before the ‘Prestige’
52
oil spill, we conducted a survey to characterize the structure of macroalgal assemblages
53
at the mid and low rocky intertidal in the northern Spanish coast (Galicia, Asturias and
54
Basque Country, Fig. 1). Hierarchical sampling allowed us to partition variance in
55
community structure among different spatial scales (from kilometers to meters). After
56
the ‘Prestige’ disaster we decided to repeat the same sampling program to evaluate the
57
effects of oil spill on these assemblages. Coincidentally, our study included locations
58
from the most affected area in Galicia to the eastern affected coast in the Basque
59
Country. Although these data lacked temporal replication before and after the spill, they
60
allowed us to test of hypothesis concerning differences in macroalgal community
61
structure after the ‘Prestige’ oil spill.
62
2. Materials and methods
63
2.1. Sampling
64
We sampled 12 localities along Galicia (Muxía, Lobeiras, and Lobadiz), West Asturias
65
(Novellana, Artedo, and Aramar), East Asturias (Rodiles, La Griega, and Vidiago), and
66
the Basque Country (Sakoneta, Zumaia, and Igeldo) coasts (Fig. 1). West and East
67
Asturias localities were sampled in August 2000 and 2003, and Galicia and Basque
3
68
Country localities in September 2002 and 2003. Six sites were randomly chosen at each
69
locality, three at "lower" (between 0.4 and 0.7 m above the Lowest Astronomical Tide)
70
and three at "upper" (between 0.9 and 1.3 m) intertidal zone. One 15 m transect parallel
71
to the coastline was sampled at each site. Five, 50x50 cm quadrats were randomly
72
placed in each transect and photographed. Algae were identified to species, or assigned
73
to higher taxonomic categories when species identification was not possible (e.g. Order
74
Ceramiales). Abundance (as percentage cover) of each taxon was estimated in the
75
laboratory by the point-contact method (Hawkins and Jones, 1992). A grid of 100
76
regularly spaced points was superimposed over the digitized pictures of the quadrats
77
and interceptions for each taxon were counted. This technique reduces sampling time at
78
the low intertidal level, but overestimates canopy species in multilayered assemblages
79
(Meese and Tomich, 1992; Dethier et al., 1993).
80
2.2 Data analysis
81
Percentage cover data were analysed using both univariate and multivariate techniques.
82
Spatio-temporal differences of most abundant taxa from each intertidal level
83
(comprising more than 75% of total cover) were analyzed. These taxa were Fucus spp.
84
(including Fucus spiralis L. and Fucus vesiculosus L., difficult to distinguish on
85
photographs), Fucus serratus L., Mastocarpus stellatus (Stack.) Guiry, Corallina
86
elongata J. Ellis & Sol., Ralfsia verrucosa Aresch., and Ceramiales (mainly Ceramium
87
spp. and Callithamnium spp.) for the "upper" level; and Bifurcaria bifurcata R. Ross,
88
Himanthalia elongata (L.) S.F. Gray, C. elongata, Ceramiales, Stypocaulon scoparium
89
(L.) Kütz., and Cladostephus spongiosus (Hudson) C. Agardh for the "lower" level. In
90
addition, species of Ulva spp. were included due to their importance as colonizers on
91
open substrata. Hypothesis of absence of spatio-temporal differences was tested using
4
92
four-way univariate analysis of variance (ANOVA). Separate analyses were done for
93
high and low tidal levels. “Time” (fixed, with 2 levels, Before and After spillage),
94
“Zone” (fixed, with 4 levels, GAL (Galicia), ASTw (West Asturias), ASTe (East
95
Asturias), and BC (Basque Country) coasts), “Locality” (random and nested in Zone,
96
with 3 levels), and “Site” (random and nested in the “Time x Locality (Zone)”
97
interaction, with 3 levels) were the factors considered in the Corallina elongata,
98
Ceramiales, Ralfsia verrucosa, and Ulva spp. analyses. A different analysis, using three
99
levels for factor Zone was applied for Fucus spp., and Mastocarpus stellatus, because
100
they did not appear in the Basque Country, and for Stypocaulon scoparium, and
101
Cladostephus spongiosus, since they were not present in samples from Galicia. For
102
Himanthalia elongata, only present in Galicia and West Asturias, factor Zone had two
103
levels. Variances were homogeneous (Cochran’s test, P > 0.05) for Bifurcaria bifurcata
104
and, after arcsin transformation for Stypocaulon scoparium, and ln (x + 1)
105
transformation for the rest of the taxa. Despite heterogeneity of variances for Fucus
106
spp., Himanthalia elongata, and Corallina elongata, ANOVA was used because of its
107
robustness and its validity in case of non-significant results in large and balanced
108
designs (Underwood, 1997). When significant Time x Zone interactions were found,
109
Student-Newman-Keuls (SNK) a posteriori test was applied.
110
Multivariate analyses were used to examine changes in the community structure before
111
and after the oil spill. Percentage cover estimates of all taxa (30 and 29 in the "upper"
112
and "lower" levels respectively) were included in the analyses. A matrix of similarities
113
between each pair of samples was calculated using the Bray-Curtis similarity coefficient
114
(Bray and Curtis, 1957). Several non-metric multivariate approaches were utilized. Null
115
hypothesis of no differences in macroalgal assemblages among Times and Zones were
5
116
tested using two-way crossed analysis of similarities (ANOSIM, Clarke 1993). A
117
maximum of two factors can be included in ANOSIM, and they were Time (2 levels)
118
and Zone (4 levels). The R-statistic generated by ANOSIM ranges from –1 to 1; the
119
higher the absolute value is, the greater the dissimilarity between macroalgal
120
communities. In addition, multiple ANOSIMs for one-factor (Time) were done within
121
each level of the other factor (Zone).
122
Hierarchical clustering analysis (CLUSTER) and multidimensional scaling (MDS) were
123
used to define similarities between the macroalgal assemblages of each Zone and Time.
124
A similarity percentage analysis (SIMPER) was used to determine the contribution of
125
each taxa to Bray-Curtis dissimilarities between groups obtained from CLUSTER and
126
MDS. Multivariate analyses were performed with PRIMER statistical software package
127
(Clarke and Warwick, 1994), and univariate analyses with GMAV5 for Windows
128
(Underwood et al., 1998).
129
3. Results
130
3.1. Abundance of dominant taxa
131
Abundance of main taxa in the two tidal levels did not change in the studied zones after
132
the oil spill (Figs. 2 and 3). A west-east gradual variation of dominant taxa was
133
observed. The "upper" level was dominated by Fucaceae, mainly Fucus spiralis and
134
Fucus vesiculosus, in GAL and AST, and by Corallina elongata and Ceramiales in BC
135
(Fig. 2). Mastocarpus stellatus, abundant in GAL, gradually decreased to the east.
136
Ralfsia verrucosa, and Ulva spp. (Fig. 4) showed very low percentages in all of the
137
samples and, in most cases, Ulva spp. were epiphytes. The Time x Zone interaction was
138
significant for Ulva spp. only (Table 1A), being significantly more abundant in pre-spill
6
139
Basque Country samples than in the other Time x Zone combinations (SNK, Table 1B).
140
In agreement with the alongshore-spatial patterns mentioned above, Ceramiales, and
141
Corallina elongata were more abundant in the Basque Country (significant Zone effect
142
in ANOVA). Although significant Zone differences in Fucus spp., and Mastocarpus
143
stellatus abundance were expected (Fig. 2), ANOVA failed to detect them; possibly due
144
to the large locality effect. Abundance of Ralfsia verrucosa did not vary spatially but it
145
was large before the oil spill (significant Time effect). Fucus serratus, appeared just in
146
one West Asturias locality, where it was dominant, and decreased significantly from
147
2000 to 2003 (ANOVA, F = 17.33, p < 0.01). The Time x Locality interaction was
148
significant for C. elongata and Ceramiales, indicating differences in some localities
149
without a clear pattern and not reflected at Zone level.
150
Similar results were found at the “lower” level, where dominant taxa were Himanthalia
151
elongata in GAL, Bifurcaria bifurcata in AST, and Corallina elongata and Stypocaulon
152
scoparium in BC (Fig. 3). No significant Time x Zone interactions were detected (Table
153
2). Only Bifurcaria bifurcata, and Corallina elongata abundances showed significant
154
differences between Zones. Ceramiales increased slightly but significantly after the
155
‘Prestige’ oil spill. Ulva spp. did not vary significantly in Zones between Times of
156
sampling (Fig. 4, Table 2). Time x Locality interaction was significant for Ceramiales
157
and Ulva spp., which increased or decreased significantly in some localities but any
158
pattern could be observed.
159
3.2. Assemblage structure
160
Significant differences in the structure of macroalgal assemblages structure in the
161
“upper” intertidal level between Times and Zones were detected by ANOSIM (Table
7
162
3A). However, minimal differences can be detected by ANOSIM when a large number
163
of replicates are included. In this case (n = 45), despite the statistical significance, we
164
examined the R statistic values to determine the degree of dissimilarity (R values near
165
to unity) or similarity (R values close to cero) between the communities compared
166
(Clarke and Warwick, 1994). The R value for Time comparison was close to zero
167
(0.081), in contrast with higher R values for Zone comparisons (Table 3A). This
168
indicates that assemblages after the ‘Prestige’ oil spill were very similar to those before,
169
but there was a great dissimilarity between zones. Analyzing each Zone separately
170
(Table 3B), differences between years were detected for ASTw, ASTe, and BC.
171
Statistical significance was again affected by the number of replicates (n = 15), and the
172
R values very close to zero indicated a similar macroalgal structure after ‘Prestige’ oil
173
spill.
174
Samples were clearly grouped according to their spatial distribution along the west-east
175
gradient, not by sampling date, in both classification analyses. The CLUSTER grouped
176
pre-spill and post-spill samples of the same zone, resulting in four groups coincident
177
with the levels of factor Zone (Fig. 5A). Same results showed MDS ordination, being
178
samples of the same zone (Before and After Prestige) closer to each other than samples
179
of different zones (Fig. 5B). This indicates that communities at each Zone were similar
180
between years. Main taxa responsible of the dissimilarity between Zones were Fucus
181
spp., Mastocarpus stellatus, and Corallina elongata (SIMPER, Table 4). The GAL
182
samples were characterized by greater abundance of Fucus spp., and M. stellatus,
183
whereas C. elongata was more abundant in BC samples. ASTw differed from ASTe by
184
the presence of Fucus serratus, and the lower abundance of Fucus spp.
8
185
Similar results were found in the “lower” level, significant differences in the macroalgal
186
assemblage structure between Zones and Times were detected by ANOSIM (Table 5B).
187
However, the R value (0.081) for Time comparison indicates that communities were
188
similar between years. Likewise, although we found significant differences between
189
years for GAL, ASTw, and ASTe, the R values indicated a similar macroalgal
190
assemblage structure (Table 5B).
191
Both, CLUSTER and MDS ordination grouped samples according to the levels of factor
192
Zone. Pre-spill and post-spill samples of the same Zone were not separate, indicating a
193
similar macroalgal structure (Fig. 6). The SIMPER analysis (Table 6) revealed that
194
Bifurcaria bifurcata, Himanthalia elongata, and Corallina elongata were the main
195
contributors to Zone differences. Himanthalia elongata explained most of the
196
dissimilarity between GAL and the other Zones. Corallina elongata and, secondly,
197
Stypocaulon scoparium, were the most important macroalgae in BC. ASTw
198
communities were characterized by the presence of Himanthalia elongata, and a higher
199
abundance of Bifurcaria bifurcata than in ASTe communities, where Stypocaulon
200
scoparium was very abundant.
201
4. Discussion
202
Along the coast of northern Spain, from West to East, there is a gradual replacement of
203
cold-temperate species (Fucus spp., Himanthalia elongata, Mastocarpus stellatus, and
204
Bifurcaria bifurcata) by warm-temperate ones (Corallina elongata, Stypocaulon
205
scoparium, and Cladostephus spongiosus). This transition has been previously
206
documented (Fischer-Piette, 1957; Anadón and Niell, 1981; Anadón, 1983; Arrontes
207
1993), and is probably related to a summer upwelling centred on the westernmost Spain,
9
208
Galician coast (Fraga et al., 1982; Botas et al., 1992), which influence decreases
209
towards the East.
210
(Topinka and Tucker, 1981; Hawkins and Southward, 1992; Peterson, 2001; Peterson
211
et al., 2003). But none of these effects were observed. No relevant changes were found
212
neither in the structure of the assemblage nor in the biomass of dominant macroalgal
213
species.
214
The structure of the studied assemblages did not change noticeably (the same dominant
215
macroalgae and a similar number of species). Some differences in macroalgal
216
abundance were found in various localities; however they did not exhibit a clear spatial
217
trend. Some taxa increase in ones and decrease in others. Although lacking a time series
218
prior to the oil spill, the magnitude of the changes in abundance of some macroalgae
219
found in some localities could be considered within the range of natural variability, and
220
these changes would render statistical significant differences between years (see
221
Mathieson et al. 1976 for an example in Fucaceae).
222
When localities were grouped by zones, no significant reduction of the main macroalgae
223
after the ‘Prestige’ oil spill was found. Only Ulva spp. in the “upper” intertidal level of
224
the Basque Country decreased significantly after the disaster. However, this trend is
225
opposite to the expected result, because these opportunistic algae rapidly colonize the
226
newly available surface after the removal of the canopy-forming algae by intense oil
227
deposition (Houghton et al., 1996; Southward and Southward, 1978; Floc’h and Diouris,
228
1980; and Kingston et al., 1997, respectively) but no destruction of the algal canopy was
229
observed in any of the localities of this study.
10
230
What may be the reason for the lack of significant effects of the ‘Prestige’ oil spill on
231
macroalgal assemblages? The degree and persistence of damage from oil spills depends
232
on several factors like type of fuel, quantity and duration of the spill, oceanographic and
233
meteorological conditions, and the type of clean-up treatments used (Clark and Finley,
234
1977). In this case, the tanker sank carrying most of its cargo far from the coast,
235
releasing the fuel in several pulses which impacted a large area of the coast. This
236
generated extensive but not intense fuel deposition (Acuña et al., submitted) but the
237
most likely causes for the absence of severe impacts were fuel dilution due to intense
238
winter mixing and advection during the wreckage period (García-Soto, 2004; Acuña et
239
al., submitted) and limited use of aggressive cleanup methods, that sometimes cause
240
more damage to organisms than fuel itself, delaying recovery of the ecosystem for
241
several years (Southward and Southward, 1978; Houghton et al., 1996).
242
Acknowledgements
243
We would like to thank D. Álvarez, I. Martínez, J.L. Menéndez, J. Oliveros, I. Sánchez,
244
and L. S. Pato, for their assistance during sampling. We also thank M. López-Álvarez
245
and I. Sánchez for their comments on an early version of this paper. Financial support
246
during the first stage of the study came from the Spanish Ministry of Education and
247
Science (CICYT Project MAR1999-1162). Carla Lobón was financially supported by a
248
fellowship from the Ministry of Education for the improvement of higher education
249
(MEC-03-EA2003-0066).
250
References
11
251
Acuña, J.L., Anadón, R., Puente, A., Fernández, C., Vera, M.L., Rico, J.L., Arrontes, J.,
252
& Juanes, J, submitted. Large scale fuel deposition patterns in northern Spanish shores
253
following the ‘Prestige's’ oil spill.
254
Anadón, R. 1983. Zonación en la costa asturiana: variación longitudinal de las
255
comunidades de macrófitos en diferentes niveles de marea. Investigación Pesquera 47,
256
125-141.
257
258
259
260
Anadón, R., & Niell, F.X. 1981. Distribución longitudinal de macrófitos en la costa
asturiana (N de España). Investigación Pesquera 45, 143-156.
Arrontes, J. 1993. Nature of the distributional boundary of Fucus serratus on the north
shore of Spain. Marine Ecology Progress Series 93, 183-193.
261
Bode, A., González, N., Lorenzo, J., Valencia, J., Varela, M., & Varela, M. 2006.
262
Enhanced bacterioplankton activity after the ‘Prestige’ oil spill off Galicia, NW Spain.
263
Aquatic Microbial Ecology 43, 33-41.
264
265
266
267
Botas, J.A., Fernández, E., Bode, A., & Anadón, R. 1990. A persistent upwelling off the
central cantabrian coast. Estuarine Coastal and Shelf Science 30, 185-199.
Bray, J.R., & Curtis, J.T. 1957. An ordination of upland forest communities of southern
Wisconsin. Ecological Monographs 27, 325-349.
268
Clark, R.C., & Finley. J.S. 1977. Effects of oil spills in Arctic and Subarctic
269
environments. Chapter 9 In: Malins, D.C. (ed.) Effects of Petroleum on Arctic and
270
Subarctic Marine Organisms. Vol II: Biological Effects. New York: Academic Press,
271
Inc., pp 411-476.
12
272
273
Clarke, K.R. 1993. Non-parametric multivariate analyses of changes in community
structure. Australian Journal of Ecology 18, 117-143.
274
Clarke, K.R., & Warwick, R.M. 1994. Change in marine communities: an approach to
275
statistical analysis and interpretation. Natural Environmental Research Council,
276
United Kingdom, p. 144.
277
Dethier, M.N., Graham, E.S., Cohem, S., & Tear, L.M. 1993. Visual versus random-
278
point percent cover estimations: 'objective' is not always better. Marine Ecology
279
Progress Series 96, 93-100.
280
Fischer-Piette, E. 1957. Sur des déplacements de frontières biogéographiques observés
281
au long des côtes ibériques dans le domaine intercotidale. Publ. Inst. Biol. Apl.
282
Barcelona 26, 35-40.
283
284
285
286
287
288
Floc'h, J.Y., & Diouris, M. 1980. Initial effects of ‘Amoco Cadiz’ oil on intertidal algae.
Ambio 9, 284-286.
Fraga, F., Mouriño, C., & Manríquez, M. 1982. Las masas de agua en la costa de
Galicia: junio-octubre. Resultados Expediciones Científicas 10, 51-77.
García-Soto, C. 2004. ‘Prestige’ oil spill and Navidad flow. Journal of the Marine
Biological Association of the United Kingdom 84, 297-300.
289
González, J.J., Viñas, L., Franco, M.A., Fumega, J., Soriano, J.A., Grueiro, G.,
290
Muniategui, S., López-Mahía, P., Prada, D., Bayona, J.M., et al. 2006. Spatial and
291
temporal distribution of dissolved/dispersed aromatic hydrocarbons in seawater in the
292
area affected by the ‘Prestige’ oil spill. Marine Pollution Bulletin 53, 250-259.
13
293
294
Hawkins, S.J., & Jones, H.D. 1992. Marine field course guide 1. Rocky Shores. Immel
Publishing, London.
295
Hawkins, S.J., & Southward, A.J. 1992. The ‘Torrey Canyon’ oil spill: recovery of
296
rocky shore communities. In: Thayer, G.W. (Ed.), Restoring the Nation's
297
Environment. Maryland, USA pp. 583-631.
298
Houghton, J.P., Lees, D.C., Driskell, W.B., Lindstrom, S.C., & Mearns, A.J. 1996.
299
Recovery of Prince William Sound epibiota from ‘Exxon Valdez’ oiling and shoreline
300
treatments, 1989 through 1992. American Fisheries Society Symposium 18, 379-411.
301
Kingston, P.F., Dixon, I.M.T., Hamilton, S., Moore, C.G., & Moore, D.C. 1997. Studies
302
on the response of the intertidal and subtidal marine benthic communities to the
303
‘Braer’ oil spill. In The Impact of an Oil Spill in Turbulent Waters: the ‘Braer’,
304
Davies, J.M., and Topping, G., eds, pp. 209-233. The Stationery Office Limited,
305
Edinburgh.
306
Marigómez, I., Soto, M., Cancio, I., Orbea, A., Garmendia, L., & Cajaraville, M.P.
307
2006. Cell and tissue biomarkers in mussel, and histopathology in hake and anchovy
308
from Bay of Biscay after the ‘Prestige’ oil spill (Monitoring Campaign 2003). Marine
309
Pollution Bulletin 53, 287-304.
310
Markarian, R.K., Nicolette J.P., Barber T.R., & Giese L.H. 1993. A critical review of
311
toxicity values and an evaluation of the persistence of petroleum products for use in
312
natural resource damage assessments. American Petroleum Institute, Publication
313
number 4594, Washington, DC.
14
314
Mathieson, A.C., Shipman, J.W., O'Shea, J.R., & Hasevlat, R.C. 1976. Seasonal growth
315
and reproduction of estuarine Fucoid algein New England. Journal of Experimental
316
Marine Biology and Ecology 25, 273-284.
317
Meese, R.J. & Tomich, P.A. 1992. Dots on the rocks, a comparison of percent cover
318
estimation methods. Journal of Experimental Marine Biology and Ecology 165, 59-73
319
Montero, P., Blanco, J., Cabanas, J.M., Maneiro, J., Pazos, Y., Moroño, A., Balseiro,
320
C.F., Carracedo, P., Gomez, B., Penabad, E., Pérez-Muñuzuri, V., Braunschweig, F.,
321
Fernandes, R., Leitao, P.C., & Neves, R. 2003. Oil spill monitoring and forecasting on
322
the Prestige-Nassau accident. Proceedings of the 26th Arctic and Marine Oil spill
323
Program (AMOP) Technical seminar 2, pp. 1013-1029.
324
325
Peterson, C.H. 2001. The ‘Exxon Valdez’ Oil Spill in Alaska: Acute, Indirect and
Chronic Effects on the Ecosystem. Advances in Marine Biology 39, 1-103.
326
Peterson, C.H., Rice, S.D., Short, J.W., Esler, D., Bodkin, J.L., Ballachey, B.E., &
327
Irons, D.B. 2003. Long-term ecosystem response to the ‘Exxon Valdez’ oil spill.
328
Science 302, 2082-2086.
329
Salas, N., Ortiz, L., Gilcoto, M., Varela, M., Bayona, J.M., Groom, S., Alvarez-Salgado,
330
X.A., & Albaiges, J. 2006. Fingerprinting petroleum hydrocarbons in plankton and
331
surface sediments during the spring and early summer blooms in the Galician coast
332
(NW Spain) after the Prestige oil spill. Marine Environmental Research 62, 388-413.
333
Serrano, A., Sánchez, F., Preciado, I., Parra, S., & Frutos, I. 2006. Spatial and temporal
334
changes in benthic communities of the Galician continental shelf after the ‘Prestige’
335
oil spill. Marine Pollution Bulletin 53, 315-331.
15
336
Southward, A.J., & Southward, E.C. 1978. Recolonization of rocky shores in Cornwall
337
after use of toxic dispersants to clean up the ‘Torrey Canyon’ spill. Journal of the
338
Fisheries Research Board of Canada 35, 682-706.
339
340
Teal, J.M., & Howarth, R.W. 1984. Oil spill studies: a review of ecological effects.
Environmental Management 8, 27-44.
341
Topinka J.A., & Tucker L.R. 1981. Long-term oil contamination of fucoid marcroalgae
342
following the ‘Amoco Cadiz’ oil spill. In: ‘Amoco Cadiz’. Consequences d’une
343
pollution accidentelle par les hydrocarbures. Fates and Effects of the oil spill.
344
CNEXO, Paris, France pp. 393-403.
345
Underwood, A.J. 1992. Beyond BACI: the detection of environmental impact on
346
populations in the real, but variable, world. Journal of Experimental Marine Biology
347
and Ecology 161, 145-178.
348
Underwood, A.J. 1997. Experiments in ecology: their logical design and interpretation
349
using analysis of variance. Cambridge University Press, Cambridge, United Kingdom.
350
Underwood, A.J., Chapman, M.G., & Richards, S.A. 1998. GMAV5 for Windows.
351
Institute of Marine Ecology, University of Sydney, Australia.
352
Varela, M., Bode, A., Lorenzo, J., Álvarez-Ossorio, M.T., Miranda, , A., Patrocinio, T.,
353
Anadón, R., Viesca, L., Rodríguez, N., Valdés, L., et al. 2006. The effect of the
354
‘Prestige’ oil spill on the plankton of the N–NW Spanish coast. Marine Pollution
355
Bulletin 53, 272-286.
356
16
357
FIGURES
358
Figure 1. Geographic situation of sampling localities in the study area (N and NW
359
Spanish coast). The line shows the ‘Prestige’ course since it started to leak oil on
360
November 13th () until the sinking on 19th (). 1 = Muxía; 2 = Lobeiras; 3 = Lobadiz;
361
4 = Novellana; 5 = Artedo; 6 = Aramar; 7 = Rodiles; 8 = La Griega; 9 = Vidiago; 10 =
362
Sakoneta; 11 = Zumaia; 12 = Igeldo.
363
Figure 2. Abundance (% cover, mean  SE) of the most abundant taxa in the high
364
intertidal level in the four zones sampled before (light grey bar) and after (dark grey
365
bar) ‘Prestige’ oil spill. GAL = Galicia; ASTw = West Asturias; ASTe = East Asturias;
366
BC = Basque Country.
367
Figure 3. Abundance (% cover, mean  SE) of the most abundant taxa in the low
368
intertidal level in the four zones sampled before (light grey bar) and after (dark grey
369
bar) ‘Prestige’ oil spill (abbreviations as in Fig. 2).
370
Figure 4. Abundance (% cover, mean  SE) of Ulva spp. in the four zones sampled
371
before (light grey bar) and after (dark grey bar) ‘Prestige’ oil spill (abbreviations as in
372
Fig. 2).
373
Figure 5. (A) MDS ordination plot and (B) dendrogram of CLUSTER analysis of the
374
macroalgal assemblages of the “upper” intertidal level in each Zone and Time
375
(abbreviations as in Fig. 2; * samples taken after Prestige).
376
Figure 6. (A) MDS ordination plot and (B) dendrogram of CLUSTER analysis of the
377
macroalgal assemblages of the “lower” intertidal level in each Zone and Time
378
(abbreviations as in Fig. 2; * samples taken after Prestige).
17
379
Figure 1. Lobón et al.
380
381
18
382
Figure 2. Lobón et al.
383
384
19
385
Figure 3. Lobón et al.
386
387
20
388
Figure 4. Lobón et al.
389
390
21
391
Figure 5. Lobón et al.
392
393
22
394
Figure 6. Lobón et al.
395
396
23
397
Table 1. Summary of the four-way ANOVA (A) results of taxa abundance in the
398
“upper” intertidal level (n = 5) showing F-values. In order to obtain homogeneity of
399
variances, variables were ln(X+1) transformed. Variances were still heterogeneous for
400
Fucus spp. (Cochran’s test, P < 0.05). SNK tests (B) of the significant Time x Zone
401
interaction (GAL = Galicia; ASTw = West Asturias; ASTe = East Asturias; BC =
402
Basque Country). *P < 0.05; **P < 0.01; ***P < 0.001; ns not significant.
A. ANOVA
Source of variation
Time = T
Zone = Z
TxZ
Locality = L(Z)
T x L(Z)
Site = S(T x L(Z))
Residual
Transformation
Cochran's Test
F
df
F versus
1
T x L(Z)
2
L(Z)
2
T x L(Z)
6 S(T x L(Z))
6 S(T x L(Z))
36 Residual
216
Fucus spp.
M. stellatus
ns
2,51
0,96ns
0,72ns
12,27***
2,67ns
3,14ns
0,97ns
10,66***
1,73ns
4,38***
1,21ns
2,97***
none
0,128*
ln(x+1)
0,095ns
F
Source of variation
Time = T
Zone = Z
TxZ
Locality = L(Z)
T x L(Z)
Site = S(T x L(Z))
Residual
Transformation
Cochran's Test
403
df
F versus C. elongata Ceramiales R. verrucosa Ulvaceae
8,2*
1
T x L(Z)
0,05ns
0,13ns
0,0ns
ns
ns
3
L(Z)
4,56*
2,12
3,97
3,2ns
3
T x L(Z)
4,4*
2,26ns
0,35ns
0,14ns
7,45***
8 S(T x L(Z)) 29,8***
10,08***
1,72ns
8 S(T x L(Z))
3,83**
4,26***
2,35*
2,08ns
2,14***
48 Residual
4,39***
4,78***
4,45***
288
ln(x+1)
none
ln(x+1)
ln(x+1)
0,065ns
0,076ns
0,069ns
0,059ns
B. SNK of "Time x Zone" interaction in Ulvaceae
Time
Before GAL = ASTw = ASTe < BC
Zone
GAL
Before = After
After
GAL = ASTw = ASTe = BC
ATSw Before = After
ASTe Before = After
BC
Before > After
404
24
405
Table 2. Summary of the four-way ANOVA results of taxa cover in the “lower”
406
intertidal level (n = 5) showing F-values. In order to obtain homogeneity of variances,
407
C. spongiosus, Ceramiales, and Ulva spp. were ln(x + 1), whereas S. scoparium was
408
arcsin(%) transformed. Variances were still heterogeneous for H. elongata, and C.
409
elongata (Cochran’s test, P < 0.01). *P < 0.05; **P < 0.01; ***P < 0.001;
410
significant.
ANOVA
F
Source of variation df F versus H.elongata
Time = T
T x L(Z)
2,87ns
Zone = Z
L(Z)
0,52ns
TxZ
T x L(Z)
0,01ns
Locality = L(Z)
S(T x L(Z)) 20,67***
T x L(Z)
S(T x L(Z))
1,46ns
Site = S(T x L(Z))
Residual
3,83***
Residual
144
Transformation
none
Cochran's Test
0,122**
411
Source of variation df F versus Ceramiales
Time = T
T x L(Z)
7,31*
Zone = Z
L(Z)
2,53ns
TxZ
T x L(Z)
0,8ns
Locality = L(Z)
S(T x L(Z)) 6,48***
T x L(Z)
S(T x L(Z))
3,04**
Site = S(T x L(Z))
Residual
3,7***
Residual
288
Transformation
ln(x+1)
Cochran's Test
0,05ns
ns
not
F
df S.scoparium C.spongiosus
1
0,95ns
4,37ns
ns
2
1,49
5,13ns
2
3,11ns
0,21ns
9,37***
3,28*
6
ns
6
1,28
0,41ns
11,24***
4,6***
36
216
arcsin(%)
ln(x+1)
ns
0,088
0,072ns
F
B.bifurcata C.elongata Ulvaceae
0,02ns
5,14*
0,13ns
4,35*
0,64ns
4,38***
1,17ns
21,6***
1,57ns
4,68***
1,85ns
1,68**
none
none
0,208**
0,07
ns
1,92ns
3,33ns
3,06ns
2,7*
2,76*
4,59***
ln(x+1)
0,074ns
412
25
413
Table 3. Two-way ANOSIM (A) testing differences in the “upper” intertidal level
414
macroalgal assemblages among Times and Zones. The Bonferroni adjusted probability
415
of type I error for multiple comparisons in the pairwise test was /6 = 0.008. ANOSIM
416
(B) for one factor, Time, was done within each level of the other factor, Zone
417
(abbreviations as in Table 1). In all analyses, 999 permutations, a random sample from a
418
large number, were used.
A. Two-way crossed ANOSIM
Global Test
Time
R
0,081
P
0,001
Global Test
Pairwise Test
ASTw, ASTe
ASTw, BC
ASTw, GAL
ASTe, BC
ASTe, GAL
BC, GAL
B. One-way ANOSIM
Time Global Test
- GAL
- ASTw Time Global Test
- ASTe Time Global Test
Time Global Test
- BC
0,454
0,001
0,221
0,354
0,26
0,746
0,267
0,899
R
0,014
0,073
0,156
0,072
0,001
0,001
0,001
0,001
0,001
0,001
P
0,139
0,007
0,001
0,005
Zone
419
420
26
421
Table 4. Summary of SIMPER analysis results of the “upper” intertidal level macroalgal
422
contributions to dissimilarity between the four Zones (abbreviations as in Table 1)
423
obtained by CLUSTER and MDS techniques. X = average abundance,  = average
424
dissimilarity between groups, SD = standard deviation of the , % = percentage
425
contribution of each taxon to the overall dissimilarity between two groups. Only main
426
contributors to the overall dissimilarity between groups are shown.
GAL-ASTw = 77,4
Taxa
Fucus spp.
M. stellatus
C. elongata
F. serratus
Ceramiales
XGAL XASTw /SD
51,5 19,9 1,41
27,3 10,3 1,09
3,9
16,6 0,76
0,0
19,2 0,61
1,4
8,9 0,63
%
29,8
17,7
11,0
12,4
6,0
ASTw-ASTe  = 75,55
%
30,8
20,5
4,9
6,1
ASTw-BC =75,3
GAL-BC = 91,2
XGAL
XBC
51,5
27,3
3,9
1,4
0,0
0,0
44,9
22,1
/SD
1,46
1,03
1,87
1,57
%
28,2
15,0
22,7
11,5
ASTe-BC = 79,3
XASTw XASTe /SD % XASTw XBC /SD % XASTe
Fucus spp.
19,9 44,7 1,33 27,5 19,9
0,0 0,65 13,2 44,7
M. stellatus
10,3
1,7 0,56 7,2
10,3
0,0 0,51 6,8
C. elongata
16,6
5,5
0,8 11,4 16,6 44,9 1,66 21,5
5,5
F. serratus
19,2
0,0 0,61 12,7 19,2
0,0 0,61 12,8
Ceramiales
8,9
8,0
0,9 8,0
8,9
22,1 0,88 6,1
8,0
Bold text indicates best discriminating taxon between groups (major /SD ratio).
Taxa
427
GAL-ASTe  = 65,2
XGAL XASTe /SD
51,5 44,7 1,39
27,3
1,7 1,04
3,9
5,5 0,96
1,4
8,0 0,88
XBC
0,0
44,9
22,1
/SD
1,32
1,81
1,4
%
28,2
25,1
11,1
428
27
429
Table 5. Two-way ANOSIM (A) testing differences in “lower” intertidal level
430
macroalgal assemblages among Times and Zones. The Bonferroni adjusted probability
431
of type I error for multiple comparisons in the pairwise test was /6 = 0.008. ANOSIM
432
(B) for one factor, Time, was done within each level of the other factor, Zone
433
(abbreviations as in Table 1). In all analyses, 999 permutations, a random sample from a
434
large number, were used.
A. Two-way crossed ANOSIM
Global Test
Time
R
0,081
P
0,001
Global Test
Pairwise Test
ASTw, ASTe
ASTw, BC
ASTw, GAL
ASTe, BC
ASTe, GAL
BC, GAL
B. One-way ANOSIM
Time Global Test
- GAL
- ASTw Time Global Test
Time Global Test
- ASTe
Time Global Test
- BC
0,503
0,001
0,243
0,613
0,374
0,472
0,682
0,696
R
0,06
0,08
0,171
0,018
0,001
0,001
0,001
0,001
0,001
0,001
P
0,011
0,007
0,001
0,11
Zone
435
436
28
437
Table 6. Summary of SIMPER analysis results of “lower” intertidal level macroalgal
438
contributions to dissimilarity between the four zones (abbreviations as in Table 1)
439
obtained by CLUSTER and MDS techniques. X = average abundance,  = average
440
dissimilarity between groups, SD = standard deviation of the , % = percentage
441
contribution of each taxon to the overall dissimilarity between two groups. Only main
442
contributors to the overall dissimilarity between groups are shown.
Taxa
GAL-ASTw = 76,9
XGAL XASTw /SD %
GAL-ASTe  = 85,9
XGAL XASTe /SD %
GAL-BC = 86,2
XGAL XBC /SD %
B. bifurcata
H. elongata
C. elongata
Ceramiales
S. scoparium
11,7
38,9
9,6
2,3
-
11,7
38,9
9,6
2,3
0,3
11,7
38,9
9,6
2,3
0,3
Taxa
443
48,3
20,4
4,0
4,6
-
1,47 28,2
1,25 26,1
1,08 5,9
0,81 3,3
-
ASTw-ASTe  = 68,9
31,1
0,0
5,3
7,1
19,1
1,19 17,6
1,16 22,6
1,07 5,0
1,04 4,0
1,05 11,0
ASTw-BC =84,9
7,3
0,0
34,2
9,6
26,7
0,71 9,2
1,16 22,6
1,17 16,1
1,17 5,0
0,92 15,4
ASTe-BC = 73,5
XASTw XASTe /SD % XASTw XBC /SD % XASTe
B. bifurcata
48,3 31,1 1,42 26,7 48,3
7,3 1,49 26,5 31,1
H. elongata
20,4
0,0 0,59 14,8 20,4
0,0 0,59 12,0
C. elongata
4,0
5,3 0,83 4,2
4,0
34,2 1,25 18,4
5,3
Ceramiales
4,6
7,1 1,09 5,5
4,6
9,6 1,19 5,1
7,1
S. scoparium 3,7
19,1 1,02 12,8
3,7
26,7 0,96 15,3 19,1
Bold text indicates best discriminating taxa between groups (major /SD ratio).
XBC
7,3
34,2
9,6
26,7
/SD
1,18
1,21
0,66
1,19
%
20,7
20,5
6,3
18,2
444
29