anmarine evolutionary perspective ecology ^ »
EDITORIAL Mar'ne Ecology-ISSN °173-9565 Marine biology in time and space
Indeed for the study of long-lived organisms such as Introduction cetaceans, long-term and extensive data sets are necessary This volume comprises a number of the papers presented to derive even the most fundamental life-history traits at the 44th European Marine Biology Symposium (EMBS) (A rrigoni et al. 2011). hosted by the University of Liverpool in September 2009. It is clear that marine systems may be influenced by The theme of the science programme was ‘Marine Biology large scale environmental phenomena such as climatic in Time and Space’. The papers focused on describing pat variations and human activities, especially in heavily terns across a variety of spatial and temporal scales but exploited areas such as the Mediterranean Sea (Ligas et al. with the emphasis on seeking understanding and explana 2011). It is also becoming increasingly clear that while we tions for those patterns. Time and space define the four strive to understand the mechanisms controlling the dimensions in which scientific observations are grounded. dynamics of marine communities, the communities them Indeed Vito Volterra’s first model of coupled temporal selves, such as those around the UK are changing over interactions was developed by Umberto D’Ancona to time (Spencer et al. 2011). In contrast, surveys of the rel study the interaction between fishery stocks and fishing atively unmodified White Sea indicate an absence of sub effort, a moment considered by some to be the starting stantial change in the structure of benthic communities point for modern ecology (Boero 2009; Gatto 2009). during the past 50 years (Solyanko et al. 2011). It is prob In the 21st century, new observational techniques, from ably most important to assess the impacts of long-term DNA genetic profiling to data storage tags and remote change on species composition (Spencer et al. 2011) or sensing, have been developed to document these patterns, ecosystem functioning (Neumann & Kröncke 2011). while experimental and modelling approaches are being However, at a finer scale, species-specific studies indicated applied to develop understanding of the factors responsi differential variability to different sources of anthropenic- ble for them. The 44th EMBS therefore took as its theme induced change (Ligas et al. 2011), an important consid ‘Marine Biology in Time and Space’ with the aim of con eration in the management of commercially important sidering recent advances in our understanding of the driv species. Assessment of ecosystem functioning is an ers of long term change in marine organism communities increasingly important tool for a number of management and ecosystems, the causes of spatial patterns in ecology purposes but, for benthic systems at least, it is essential and the consequences of catastrophic phenomena in mar that methodologies be consistent and consider biological ine systems. Papers were presented under these three main traits as well as simple count and biomass metrics themes, though in reflection of the naturally inter-disci (Aarnio et al. 2011). plinary nature of marine biology, many of them contained elements from more than one theme. Theme 2: Spatial Patterns
An understanding of patterns of species distribution and Theme 1: Long-Term Dynamics community composition can also be gained by studies of The University of Liverpool had previously hosted the spatial variability, and the presented papers also high European Marine Biology Symposium (EMBS13) in 1978. lighted the importance of considering a range of scales. At that meeting the theme was ‘Cyclic Phenomena in Designation of Marine Protected Areas (MPAs) to effec Marine Plants and Animals’ (Naylor & Hartnoll 1979) tively protect vulnerable habitats from exploitative activi and temporal dynamics reappeared as one of the themes ties and preserve biodiversity should be based on for the 44th meeting. Temporal change in marine systems knowledge of spatial factors such as distribution and dis occurs on long, multi-decadal, time scales. The impor persal (Kinlan & Gaines 2003). However MPAs will not tance of collecting and maintaining long-term data sets of be able to provide protection from extreme climatic the marine environment is well recognised (Ducklow events (Huete-Stauffer et al. 2011). et al. 2009), and several of the presented papers high At a regional scale, such as in the English Channel, the lighted this (e.g. Ligas et al. 2011; Spencer et al. 2011). trophic structure of benthic ecosystems appears to be
Marine Ecology 32 (Suppl. 1) (2011) v-vii © 2011 Blackwell Verlag GmbH V Editorial Green, Paramor, Robinson, Spencer, W atts & Frid determined by sedimentary conditions, whatever the geo J. A. Green1, O. A. L. Paramor1’3, L. A. Robinson1, graphic area (Garcia et al. 2011), showing the importance M. Spencer1, P. C. Watts2 & C. L. J. Frid1 of abiotic factors. It is possible to consider how species 1 School of Environmental Sciences, University of Liverpool, distributions are controlled by spatial factors at micro to Liverpool L69 3GP, UK; 2Institute of Integrative Biology, regional scales, and their interactions, within single stud Biosciences Building, University of Liverpool, Liverpool L69 ies. For example, both specific frond segments and envi 7BX, UK; 3University of Nottingham Ningbo, China, ronmental factors of salinity and wave exposure are Environmental Sciences, 199 Taikang East Road, important in determining the composition of epiphyte Ningbo 315100, Zhejiang, P.R. China and mobile fauna communities on habitat forming macro algae (Kersen et al. 2011). At local scales such as estuarine R eferences habitats, the coexistence of sympatric and seemingly com petitor species can be explained by partitioning of Aarnio K., Mattila J., Tornroos A., Bonsdorff E. (2011) resources. In the case of juvenile plaice and flounder, both Zoobenthos as an environmental quality element: the flexibility and heterogeneity of diets appears to reduce ecological significance of sampling design and functional niche overlap and bring order to an apparently chaotic traits. Marine Ecology, 32(Suppl. 1), 58-71. habitat (Mariani et a í 2011). Finally, a key requirement in Arrigoni M., Manfredi P., Panigada S., Bramanti L., examining spatial effects is the establishment of discrete Santangelo G. (2011) Life-history tables of the Mediterranean populations, a process which is rapidly being facilitated by fin whale from stranding data. Marine Ecology, 32(Suppl. 1), molecular techniques (Luis et a í 2011). 1-9. Boero F. (2009) Recent innovations in marine biology. Marine Ecology, 30, 1-12. Theme 3: Consequences of Catastrophic Events Cerrano C., Bavestrello G. (2009) Mass mortalities and extinc tions. In: Wahl M. (Ed.), Marine Hard Bottom Communities. Finally, both temporal and spatial factors collide when Springer, Berlin: 295-307. considering the impacts and influence of catastrophic Ducklow H.W., Doney S.C., Steinberg D.K. (2009) Contribu events. This will become increasing important if the inci tions of long-term research and time-series observations to dence of these events continues to increase in frequency marine ecology and biogeochemistry. Annual Review of Mar and/or magnitude (Cerrano & Bavestrello 2009). Mass ine Science, 1(1), 279-302. mortality events can lead to change in ecosystem structure Garcia C., Chardy P., Dewarumez J.M., Dauvin J.C. (2011) As- and function, particularly if the subjects affected are eco sessement of benthic ecosystem functioning through trophic system engineers (Huete-Stauffer et a í 2011). In the case web modelling: the example of the eastern basin of the Eng of Mediterranean corals, high temperature appeared to lish Channel and the Southern Bight of the North Sea. Mar precipitate a mass-extinction event which was exacerbated ine Ecology, 32(Suppl. 1), 72-86. by opportunistic bacterial infection (Huete-Stauffer et a í Gatto M. (2009) On Volterra 8c D’Ancona’s footsteps: the tem 2011). Catastrophic events may of course also be predict poral and spatial complexity of ecological interactions and able and avoidable. On tidal mudflats, commercial dredg networks. Italian Journal of Zoology, 76(1), 3-15. ing for cockles can cause disruption of benthic Huete-Stauffer C., Vielmini I., Palma M., Navone A., Panzalis communities. However these systems can recover to their P., Vezulli L., Misic C., Cerrano C. (2011) Paramuricea clav original state if dredging occurs at an appropriate inten ata (Anthozoa, Octoralia) loss in the Marine Protected Area sity and frequency (Wijnhoven et al. 2011). of Tavolara (Sardinia, Italy) due to a mass mortality event. Marine Ecology, 32(Suppl. 1), 107-116. Kersen P., Kotta J., Martynas B., Kolesova N., Dekere Z. Perspectives (2011) Epiphytes and associated fauna on the brown alga Fucus vesiculosus in the Baltic and the North Seas in relation As ever in the study of biology, it is worth considering the to different abiotic and biotic variables. Marine Ecology, work of Charles Darwin. As well as his more widely-publi 32(Suppl. 1), 87-95. cised work, Darwin’s studies of barnacles remain authorita Kinlan B.P., Gaines S.D. (2003) Propagule dispersal in marine tive (Rainbow 2011). In the bicentennial of his birth, a and terrestrial environments: a community perspective. Ecol consideration of the influence of marine biology on the ogy, 84(8), 2007-2020. work of Darwin reveals the importance of detailed observa Ligas A., Sartor P., Colloca F. (2011) Trends in population tion, critical thinking and experiments to test ideas and dynamics and fishery of Parapenaeus longirostris and Nephr ultimately communicate the results (Rainbow 2011). These ops norvegicus in the Tyrrehenian Sea (NW Mediterranean): principles clearly continue to underpin the work of marine the relative importance of fishery and environmental biologists and the work presented at the 44th European variables. Marine Ecology, 32(Suppl. 1), 25-35. Marine Biology Symposium was a testament to this.
vi Marine Ecology32 (Suppl. 1 ) (2011 ) v-vii © 2011 Blackwell Verlag GmbH Green, Paramor, Robinson, Spencer, W atts & Frid Editorial
Luis J.R., Comesafia A.S., Sanjuan A. (2011) mtDNA Solyanko K., Spiridonov V., Naumov A. (2011) Biomass, com differentiation in the mussel Mytilus galloprovincalis Lmk. monly occurring and dominant species of macrobenthos in on the Iberian Peninsula coast: first results. Marine Ecology, Onega Bay (White Sea, Russia): data from three different 32(Suppl. 1), 102-106. decades. Marine Ecology, 32(Suppl. 1), 36-48. Mariani S., Boggan C., Balata D. (2011) Food resource use in Spencer M., Birchenough S.N.R., Mieszkowska N., Robinson sympatric juvenile plaice and flounder in estuarine habitats. L.A., Simpson S.D., Burrows M.T., Capasso E., Cleall-Har- Marine Ecology, 32(Suppl. 1), 96-101. ding P., Crummy J., Duck C., Eloire D., Frost M., Haii A.J., Naylor E., Hartnoll R.G. (1979) Cyclic phenomena in marine Hawkins S.J., Johns D.G., Sims D.W., Smyth T.J., Frid L.J. animals and plants. Proceedings of the 13th European Marine (2011) Temporal change in UK marine communities: trends Biology Symposium Isle of Man, 27 September-4 October or regime shifts? Marine Ecology, 32(Suppl. 1), 10-24. 1978. Pergamon Press, Oxford: p. 477. Wijnhoven S., Escaravage V., Herman P.M.J., Smaal A.C., Neumann H., Kröncke I. (2011) The effect of temperature var Hummel H. (2011) Short and mid-long term effects of iability on ecological functioning of epifauna in the German cockle-dredging on non-target macrobenthic species: a Bight. Marine Ecology, 32(Suppl. 1), 49-57. before-after-control-impact experiment on a tidal mudflat in Rainbow P.S. (2011) Charles Darwin and marine biology. the Oosterschelde (The Netherlands). Marine Ecology, Marine Ecology, 32(Suppl. 1), 130-134. 32 (Suppl. 1), 117-129.
Marine Ecology32 (Suppl. 1 ) (2011 ) v-vll © 2011 Blackwell Verlag GmbH anmarine evolutionary perspective ecology ^ »
Marine Ecology. ISSN 0173-9565
ORIGINAL ARTICLE Life-history tables of the Mediterranean fin whale from stranding data Massimo Arrigoni1,2, Piero Manfredi3, Simone Panigada2, Lorenzo Bramanti1,4 & Giovanni Santangelo1
1 Department of Biology, University of Pisa, Pisa, Italy 2 Tethys Research Institute, Milan, Italy 3 Department of Statistics and Mathematics Applied to Economics, University of Pisa, Pisa, Italy 4 Institut de Ciencias del Mar, CSIC, Barcelona, Spain
Keywords A bstract Conservation; demography; Mediterranean Sea; Mysticeta. The conservation of long-lived species requires extensive, in-depth knowledge of their population structure and vital rates. In this paper we examine the Correspondence structure of the Mediterranean fin whale (Balaenoptera physalus) population Giovanni Santangelo, Department of Biology, based on the available mortality figures from European stranding network data University of Pisa, Via Volta 6, 1-56126 Italy, bases compiled over the past 22 years. Such data has enabled us to lay out a Pisa. E-mail: [email protected] first life-history (mortality) table of the population using a simple age-struc
Accepted: 15 December 2010 tured demographic model with three life-tables: calf, immature and mature. Our results reveal a high mortality rate in the first stage of life (77% per year), doi: 10.1111/j. 1439-0485.2011,00437.x which decreases during the immature stage and falls further during the mature adult stage. In addition, we have calculated the corresponding life expectancies at birth (e0), at entry in the immature stage (ej and at maturity (e2) under different hypotheses on survival at the maximum age of 90 years (s90) ranging between 0.1 and 3% of newborns still alive. The life expectancy at birth (e0) at the lower bound of the chosen range (s90 = 0.001) is about 6 years, entry in the immature stage (ej is 8.2 years, and entry in the mature stage (e2) is about 15.6 years. This large increase is the consequence of the higher mortality in the first two stages compared with the mature one. The life expectancies are 10.1, 14.3, and 37.8 years for s90 at the upper bound of the chosen range (s90 = 0.03). The resulting population intrinsic growth rates (r) ranged between -1.3. and +1.7 per year. High juvenile mortality patterns imply that the sta tionary reproductive value (the number of female offspring produced by each female after a given age x) at the start of maturity reaches a value about seven times higher than at birth. Only optimistically high survival patterns of older individuals would allow positive intrinsic growth rates, thereby enhancing the chances of the population survival.
While some demographic studies have been conducted Introduction using industrial whaling data on Northeast Atlantic popu Surprisingly, the fin whale (Balaenoptera physalus ), which lations (Aguilar & Lockyer 1987), little is known about is the world’s second largest cetacean and one of its the demography of their counterparts in the Mediterra longest-lived mammals (Lockyer et a í 1977), is also one nean, where industrial whaling has never been practised of the least-known Mysticetes in demographic terms (Notarbartolo di Sciara et a í 2003). Although the data (Notarbartolo di Sciara et a í 2003). from Aguilar & Lockyer (1987) are a fundamental
Marine Ecology 32 (Suppl. 1) (2011) 1-9 © 2011 Blackwell Verlag GmbH 1 Demography of fin whales from strandings Arrigoni, Manfredi, Panigada, Bramanti & Santangelo contribution to the understanding of the demography of strandings along the Italian coasts has been drawn from the fin whale, no population dynamics model has ever the CSC (Cetacean Study Center) database, available been developed for this species. Moreover, as the Medi online at CIBRA (2010). The Spanish and French coast terranean fin whale population is genetically distinct from stranding data have been collected respectively from the its Northeast Atlantic counterpart (the nearest population MEDACES database (2009) and the French National in geographic terms; Bérubé et al. 1998; Palsboll et al. Stranding Network RNE (2008). Further data from Medi 2004), it therefore represents a separate unit of conserva terranean countries without dedicated stranding databases tion, requiring ad hoc studies. were found in the scientific literature (Notarbartolo di According to the IUCN Red data book criteria (Reeves Sciara et al. 2003). In our analyses we used the stranded and Notarbartolo di Sciara, 2006), the conservation status animals’ sex and length at death. Unfortunately, the infor of this Mediterranean species has been judged data defi mation is not uniform, as in many cases sex was not cient due to the lack of demographic information. How determined and exact size measurements are possible only ever, a more recent assessment, still under review by the for recently dead animals due to their rapid decomposi Red List Authority, has classified the Mediterranean pop tion. ulation as vulnerable (Panigada, pers. comm.) The survival of this population is threatened by many Basic life-history data sources of mortality and environmental stress (Notarbar tolo di Sciara & Gordon 1997; Notarbartolo di Sciara et a í Fin whales are characterized by fast growth in the first 2002), the most important of which are ship collisions part of their life, which then slows as they reach full phys (Panigada et a í, 2006), fishing gear entanglement, human- ical maturity at about 25 years of age (Aguilar & Lockyer induced natural habitat degradation, unregulated whale- 1987). As a first step, we transformed the size distribution watching (Airoldi et al. 1999), and acoustic disturbance of the stranded whales into a size-stage distribution, and (Notarbartolo di Sciara et al. 2003; Abdulla et al. 2008). then into an age distribution by stage. This was carried Although some ecological features, such as seasonal abun out using the growth and reproductive parameters mea dance (Forcada et al. 1996), habitat use (Panigada et a í sured in the Northeast Atlantic population (Lockyer 1984; 2005, Panigada et al. 2008; Monestiez et al. 2006; Laran & Aguilar & Lockyer 1987; Aguilar et al. 1988), criteria Gannier 2008), site fidelity, diving profiles (Panigada et a l which yielded the following three age-stages: Calf (0-0.5) 1999) and contamination by pollution (Fossi et al. 2003) years, Immature (0.5-7.5) years, and Mature (7.5-90) have been investigated, no population dynamics study has years. The value of 90 years represents the maximum life been performed on fin whale populations to date. span for fin whales estimated by Lockyer et al. (1977). Only recently have the demographic models widely As a preliminary assumption we hypothesized that used in studying other animal and plant populations stranding data represent a faithful description of the real (Ebert 1998; Caswell 2001; Santangelo & Bramanti 2006) mortality by stage. This, however, holds only if the proba been applied to the study of cetaceans (Buckland 1990; bility of stranding is equal in all life-tables. Indeed, only Fujiwara & Caswell 2001). Two different approaches are under such circumstances would we expect the relative commonly applied in demographic studies; these are distribution of stranding by stage to be the same as the based on static or cohort life-tables. A third approach is to true underlying distribution of deaths by stage. As precise compile mortality tables (Caughley 1966; Caughley & Sin information in this regard is lacking, such an assumption clair 1994; Ebert 1999), which provide precise informa is therefore necessary to compute the mortality table. tion about size/age and sex of dead individuals. Herein we have adopted this latter approach, which to date has The mortality table never been applied to cetaceans, by using stranding data. Our aim is to develop a demographic model for the Med To build up a complete mortality table for the population iterranean fin whale population based on a life-history we used a simple demographic model based on the three table (mortality table sensu Bergher 1990; Ricklefs and above-defined life-tables, with continuous age distribution Miller 2001) built on Mediterranean stranding records. and constant mortality rates within each stage, under the assumption of population stationarity [i.e. the population Material and methods is assumed to be constant in number and age structure over time). Stranding data As we assumed that no animals survive beyond the age Our demographic model has been based on all available of 90, to apply the model with constant mortality rates, it data on fin whale strandings recorded on Mediterranean is necessary to know the fraction (s90) of newborn indi coasts between 1986 and 2007. The information on viduals that survive up to the maximum age cu = 90.
2 Marine Ecology32 (Suppl. 1 ) (2011 ) 1-9 © 2011 Blackwell Verlag GmbH Arrigoni, Manfredi, Panigada, Bramanti & Santangelo Demography of fin whales from strandings
Given the lack of information on this quantity, we per on average one offspring every 22-24 months (Lockyer formed a detailed analysis of the sensitivity of the life- 1984). This corresponds to an age-specific female fertility table to different assumptions on survival rates up to the rate of about 0.25-0.28 females per year, assuming a sex maximum age (ranging between 0.1 and 3%). This has ratio at birth of 1:1 (Zanardelli et a í 1999). By combining enabled us to compute the mortality rates (or mortality this assumption with our mortality table, we have calcu risks) /( for each age-stage (a¡_1, a¡) via the equation: lated the standard reproduction measures: the net repro ductive number (Ro), the mean age of mothers at reproduction in the corresponding stationary population (T), the population intrinsic growth rate (r, Keyfitz & Caswell 2007) and the reproductive value at each age in a In equation 1, the age interval (a;-!, a;) denotes the i- stationary population (SRV). The net reproductive rate th age-stage. Thus, as per the definitions in the previous R0 represents the average number of female offspring a section, the calf stage is defined by (a0 = 0, a: = female expects to have during her entire life under the 0.5 years), the immature stage by (a: = 0.5 years, mortality described by the given life-table (so that the a2 = 7.5 years), and the difference h¡ = a; - a;_! is the value R0 = 1 represents the threshold between population corresponding class length. Finally, s(a¡)denotes the frac growth and decline). The stationary reproductive value tion of newborn individuals still alive at precisely age a; (SRV) represents the number of female offspring remain [i.e. at the moment of transition from stage i to stage ing to be born to a female mother after any given age x. (i + 1)]. Further details are reported in Appendix 1. The assumption of constant mortality rates within each age group implies that the corresponding survival curve Results has the following exponential form: Stranded population structure s(a) = s(ai_i)e-"-(a-a' ^ a¡_i < a Reproduction parameters J l 11111 lui 1 1 1 1 y 11 n Ini 1985 1990 1995 2000 2005 2010 The fertility rates for females have been drawn from the Year literature using the standard assumption that a mature Fig. 1. Mediterranean fin whale: distribution of total strandings over female that has not been subjected to mortality produces the period 1986-2007. Marine Ecology32 (Suppl. 1) (2011) 1-9 © 2011 Blackwell Verlag GmbH 3 Demography of fin whales from strandings Arrigoni, Manfredi, Panigada, Bramanti & Santangelo eo I 11988-2007 50 50 S01 30 30 30 20 20 10 Fig. 2. Mediterranean fin whale: distribution of strandings by life-stage over two different periods: (A) 1986-96, (B) 1997-2007, and (C) 0 C a R Immatura Matura CaH Immatura Matura CaR Im m atura M atura the entire measurement period: 1986-2007. entire period. There was a statistically significant differ cohort of 1000 (instead of 134) recruits, as shown in ence in the population mortality structure between the Table 2, column 3: of 1000 recruits, only 679 (67.9%) two 11-year periods (omnibus-type likelihood ratio test enter the immature stage and only 186 enter the mature on the multinomial distribution significant at 1%), sug stage. Under the stationarity hypothesis, this represents gesting that some change in mortality could have the living population structure. occurred in the more recent period (1997-2007). How Clearly, the distribution of individuals by life-tables is ever, even though the data suggest the possibility of a biased by the different durations of the stages. To correct change in the mortality structure, we calculated a single for this, we normalized the distribution to the duration mortality table for the entire period to avoid the exces of the first stage (6 months), which is the shortest sively small sample size that would result by splitting the (Table 1). data. This was supported by the lack of any evident Using equation 1, we computed the continuous mortal trends in the distribution of total strandings over time ity rates /í¡ inside each stage. For calves, /q was 0.774 per (Fig. 1), as well as by some evidence of stability suggested year, while for immature individuals, ß2 was 0.184 per by the monotonically decreasing structure of the normal year, less than a quarter of the calf rate. For mature ized data (Table 1, Fig. 3). whales the mortality rate ranged between 0.063 per year for P90 = 0.001 and 0.022 per year for P90 = 0.03. Figure 3 shows the corresponding survival curves com Mortality table puted according to equation 2 using the values As a second step we built a new mortality table (Table 1) P90 = 0.001 and P90 = 0.03. The plotted survival curves based on the strandings according to occurrence in the differ from each other only in the third age-stage (7.5— three discrete life-tables into which the species life cycle 90 years), a consequence of the different assumptions on has been divided (calf, immature and mature). In this the fraction surviving at 90 years. The curves furthermore perspective, the total number of strandings (134) can be show first an exponentially decreasing pattern within each interpreted as the number of newborns in a hypothetical age-stage, a result of the assumption of constant-rate birth cohort, of which 43 die during the calf stage, 66 mortality in each stage, and secondly a marked difference during the immature stage, and the remaining 25 individ in the rate of decline in the different stages, which is con uals during the mature stage (Table 1 column 5). Inter sistent with the different mortality rates in the various preting the data is simplified by using a hypothetical stages. Table 1. Mediterranean fin whale: mortality table by life-stages. No. of deaths No. of survivors No. of survivors Stage duration normalized by stage at onset of each at onset of each Life-table No. of deaths (years) duration (year1) stage stage (per 1000) Calf 43 0.5 86 134 1000 Immature 66 7.0 9.42 91 679 Mature 25 82.5 0.30 25 186 4 Marine Ecology32 (Suppl. 1 ) (2011 ) 1-9 © 2011 Blackwell Verlag GmbH Arrigoni, Manfredi, Panigada, Bramanti & Santangelo Demography of fin whales from strandings B 1 0.9 0.9 0.8 0.8 p90 = 0.0°1 0.7 0.7 0.6 0.6 cl 0.5 0.5 0.4 0.4 0.3 0.3 0.2 0.2 Fig. 3. Mediterranean fin whale: age-specific survival curves under tw o different 0.1 0.1 assumptions on the probability of survival to 0 0 maximum age: (A): p90 = 0.001; (B) 0 20 40 60 80 0 20 40 60 80 P9 0 = 0.03. Age (years) Age (years) Reproduction parameters The life expectancies at the age of onset of the various stages (Fig. 4) show a marked increasing trend as a func By combining our set of mortality tables with fertility tion of both the age of entry and the fraction surviving at data we can develop scenarios of long-term population age 90. In particular, for s90 at the lower bound of the trends. The net reproductive rate Ro ranges from a value chosen range (s90 = 0.001), the life expectancy at birth well below one (0.73) for the s90 = 0.001 hypothesis, to a (e0) is about 6 years, whereas the life expectancy upon value considerably above unity (1.77) for s90 = 0.03. The onset of the immature stage (e j is 8.2 years, and the life mean age of mothers at reproduction correspondingly expectancy entering into the mature stage (e2) is about ranges between 22.8 and 36.8 years (Fig. 5). Finally, the 15.6 years. This large increase follows from the very high corresponding intrinsic population growth rate (r) ranges mortality in the first two stages as compared with the between -1.3% and +1.7% year-1 (Fig. 6). Figure 7 shows mature stage. the trend of the SRV with the age of the mother. In par At the upper bound of the chosen range (s90 = 0.03) ticular, the value at sexual maturity is about seven times the respective life expectancies are 10.1, 14.3, and higher than the value at birth, due to the huge mortality 37.8 years. during pre-reproductive ages. 2 9_ 0 0.005 0.01 0.015 0.02 0.025 0.03 Fraction alive at maximal age © (© = 90 years) 0 0 0.005 0.01 3.015 0.02 0.025 0.03 Fig. 4. Mediterranean fin whale: life expectancies at the age of entry Fraction alive at maximal age © (© = 90 years) into the various stages as functions of the fraction surviving at age 90 (ranging between 0.001 and 0.03); e0 = life expectancy at birth, Fig. 5. Mediterranean fin whale: net reproduction rate R0 (left verti e-i = life expectancy at onset of the immature stage, e2 = life expec cal axis) and the corresponding mean age of mothers at reproduction tancy at onset of maturity. T (right vertical axis) as functions of the fraction surviving at age 90. Marine Ecology32 (Suppl. 1) (2011) 1-9 © 2011 Blackwell Verlag GmbH 5 Demography of fin whales from strandings Arrigoni, Manfredi, Panigada, Bramanti & Santangelo 0.02 towards female mortality previously suggested for this species (Clark 1982; De La Mare 1985). 0.015 Our results show that the first stage of the life cycle is the most life-threatening, with a yearly risk of death of 0.01 about 77%, whereas in the immature stage, death is 5 0.005 nearly four times less likely (18%). This indicates a strong impact, natural and/or anthropogenic, on calves and immature animals, which prevents their reaching sexual maturity. On the other hand, the risk of death of mature -0.005 individuals is much lower: under the most pessimistic scenario it is still only about 6.3% per year. These results - 0.01 confirm a pattern common to several mammals: high mortality in the youngest age classes and low ones in -0.015 0 0.005 0.01 0.015 0.02 mature stages (Caughley 1966; Emelen 1970). Nonethe Fraction alive at maximal age ® (® = 90 years) less, a very low proportion of newborns reach sexual maturity, which may represent a serious threat for the Fig. 6. Mediterranean fin whale: the population Intrinsic growth rate survival of this population. Indeed, even under very opti as a function of the fraction surviving at age 90. mistic hypotheses on the length of the maximum lifespan and prolonged fertility, the intrinsic growth rate r is likely to be positive only if the percentage of offspring surviving SRV up to maximum age is quite high {i.e. well above 0.005). The SRV clearly shows that the contribution to the popu lation in terms of survival is biased toward adults: most calves and young whales do not contribute to reproduc tion because they will never reach sexual maturity. In conducting this study we examined all the available data on strandings. However, these are far from represen tative of all Mediterranean strandings. In addition, only some of the available data were suitable for analysis, as Age (years) indications on sex are lacking in about half the cases and the reported size is often approximate or even missing. Fig. 7. Mediterranean fin whale: change in the stationary reproduc To make up for this lack of data uniformity, we resorted tive value (SRV) with age; a2 Is the age at sexual maturity. to some necessary assumptions (population stationarity and identical stranding probabilities in each stage). Although the proposed model is rather simple, the study nevertheless suggests that stranding data and the use of D iscussion demographic models may well allow enhancing our cur This study seeks to describe the structure of the Mediter rently limited knowledge of the demographics of this ranean fin whale population by analyzing stranding important cetacean. Future work to improve our ongoing records from the period 1986-2007. As the ecological study of fin whale populations will focus on comparing characteristics of this species make data collection at sea the approach applied herein with analyses of photo-iden- particularly difficult (fin whales usually live far from the tification data recorded by the Tethys Research Institute coast and are difficult to observe due to weather con on the live population inhabiting the waters of the Pela- straints and the high costs of dedicated research vessels), gos Sanctuary. We also plan to investigate population strandings may prove to be an alternative source of dynamics under different conservation scenarios, and demographic data (Orsi Relini et al. 2004). This study is thereby assess its current status and risk of extinction. the first to analyze the demographic features of the Medi At present, in spite of the existence of the Pelagos terranean fin whale population and, to our knowledge, Sanctuary (Notarbartolo di Sciara et a í 2007), an MPA the first to set out a mortality table based on cetacean specifically designated to protect cetaceans and which stranding data. represents the most important feeding grounds for the The dataset examined does not reveal any significant Mediterranean fin whale (Notarbartolo di Sciara et a í divergence from a balanced sex ratio in the strandings. 2003), no specific regulation is currently in force for We are therefore unable to confirm the natural bias protection of this species. Regulation of naval traffic and 6 Marine Ecology32 (Suppl. 1 ) (2011 ) 1-9 © 2011 Blackwell Verlag GmbH Arrigoni, Manfredi, Panigada, Bramanti & Santangelo Demography of fin whales from strandings whale-watching activities could enhance this population’s Bergher J. (1990) Persistence of different-sized populations: an chances of survival (Panigada et a í, 2006). empirical assessment of rapid extinction in Bighorn sheep. Our findings suggest that mitigation measures targeted Conservation Biology, 4, 91-98. to reproductive adults, particularly addressed to increase Bérubé M., Aguilar A., Dendanto D., Larsen F., Notarbartol- the mean age of mothers at reproduction, are likely to be o di Sciara G., Sears R., Sigurjonsson J., Urban-Ramirez the most effective and need to be taken into account in J., Palsboll P.J. (1998) Population genetic structure of designing proper conservation plans. The discovery of North Atlantic, Mediterranean Sea and Sea of Cortez fin breeding grounds where calves may enjoy greater protec whales, Balaenoptera physalus (Linnaeus, 1758): analysis of mitochondrial and nuclear loci. Molecolar Ecology, 7, 585- tion, could further increase survival rates. On another 600. track, special naval traffic regulations, aimed at reducing Buckland S.T. ( 1990) Estimation of survival rates from sight mortality rates from ship collisions, could enhance the ings of individually identifiable whales. Report of the Interna survival of mature females and calves. Mitigating other tional Whaling Commission, Special Issue 12, 149-154. sources of mortality and stress, such as chemical and Caswell H. (2001) Matrix Population Models: Construction, acoustic pollution, whale-watching activities and natural Analysis and Interpretation, 2nd edn. Sinauer Associates, habitat degradation, could further improve the popula Sunderland, MA: 722 pp. tion’s chances of survival. Caughley C. (1966) Mortality patterns in mammals. Ecology, 47, 906-918. Acknowledgements Caughley G., Sinclair A.R.E. (1994) Wildlife Ecology and M an agement. Blackwell Scientific Publishing, Oxford. We wish to thank the Italian National Strandings Data CIBRA Centro Interdisciplinare di Bioacustica e Ricerche Ambi- base, CIBRA, University of Pavia and Italian Ministry of entali. (2010) University of Pavia http://www.cibra.unipv.it/ the Environment. We are also grateful to the Mediterra spiaggiamenti.html (last modified lanuary 2010). nean Database of Cetacean Strandings (MEDACES) at the Clark W.G. (1982) Historical rates of recruitment to Southern University of Valencia (Spain), which is supported by the Hemisphere fin whale stocks. Report of the International Spanish Ministry of the Environment, and the French Whaling Commission, 32, 305-324. National Stranding Network (RNE), Centre de Recherche De La Mare W.K. ( 1985) On the estimation of mortality rates sur les Mammifères Marins - Université de La Rochelle. from whale age data, with particular reference to mink We also thank C. Carota of the Statistics Dept, of Turin whales ( Balaenoptera acutorostrata ) in the Southern University for statistical references and A. Cafazzo for his Hemisphere. Report of the International Whaling Commis revision of the English text. This research was supported sion, 35, 239-250. Ebert T.A. (1998) Plant and Animal Populations: Methods in by the Italian Ministry of Education and Scientific Demography. Academic Press, New York. Research PRIN 2007 project 200777BWEP. Mathematical Entelen J.M. (1970) Age specificity and ecological theory. Ecol population theory and L.B. were supported by a Marie- ogy, 51, 588-601. Curie IEF (CORGARD, project no. 221072). We also Forcada J., Aguilar A., Hammond P., Pastor X., Aguilar R. thank two referees and the Editor, whose useful sugges (1996) Distribution and abundance of fin whales (Balaenop tions improved the manuscript. tera physalus) in the western Mediterranean Sea during the summer. Journal of Zoology London, 238, 23-34. Fossi M.C., Marsili L., Neri G., Natoli A., Politi E., Panigada S. R eferences (2003) The use of a non-lethal tool for evaluating toxicolog- Abdulla A., Gornei M., Maison E., Piante C. (2008) Status of ical hazard of organochlorine contaminants in Mediterra Marine Protected Areas in the Mediterranean Sea. IUCN, nean cetaceans: new data 10 years after the first paper Malaga and WWF France, Paris: 152 pp. published in MPB. Marine Pollution Bulletin, 46, 972-982. Aguilar A., Lockyer C.H. (1987) Growth, physical maturity, Fujiwara M., Caswell H. (2001) Demography of the endan and mortality of fin whales ( Balaenoptera physalus ) inhabit gered North Atlantic right whale. Nature, 414, 537-541. ing the temperate waters of the northeast Atlantic. Canadian Keyfitz N., Caswell H. (2007) Applied Mathematical Demogra Journal of Zoology, 65, 253-264. phy. Springer-Verlag, New York: 505 pp. Aguilar A., Oimos M., Lockyer C.H. (1988) Sexual maturity of Laran S., Gannier A. (2008) Spatial and temporal prediction of fin whales ( Balaenoptera physalus ) caught off Spain. Report fin whale distribution in the northwestern Mediterranean of the International Whaling Commission, 38, 317-322. Sea. ICES (International Council for the Exploration of the Airoldi S., Azzellino A., Nani B., Ballardini M., Bastoni C., Seas) Journal of Marine Science, 65, 1260-1269. Notarbartolo di Sciara G., Sturlese A. (1999) Whale-watching Lockyer C.H. (1984) Review of Baleen Whale (Mysticeti) Repro in Italy: results of the first three years of activity. European duction and Implications for Management. Report of the Research on Cetaceans, 13, 153-156. International Whaling Commission, Special Issue 12. Marine Ecology32 (Suppl. 1) (2011) 1-9 © 2011 Blackwell Verlag GmbH 7 Demography of fin whales from strandings Arrigoni, Manfredi, Panigada, Bramanti & Santangelo Lockyer C.H., Gambell R., Brown S.G. (1977) Notes on age (Western Mediterranean Sea) with physiographic and data of fin whale taken off Iceland, 1967-74. Report of the remote sensing variables. Remote Sensing of the Environment, International Whaling Commission, 27, 427-450. 112, 3400-3412. MEDACES Mediterranean Database of Cetacean Strandings Reeves R., Notarbartolo di Sciara G. (eds) (2006) The Status (2009) University of Valencia, http://www.medaces.uv.es and Distribution of Cetaceans in the Black Sea and Mediterra (last modified September 2009). nean Sea. Iucn Centre for Mediterranean Cooperation, Monestiez P., Dubroca L., Bonnin E., Durbec J.P., Guinet C. Malaga, Spain, 137 pp. (2006) Geostatistical modelling of spatial distribution of Bal Ricklefs R.E., Miller L.G. ( 1999) Ecology VI. Freeman and Co., aenoptera physalus in the Northwestern Mediterranean Sea New York. 822 pp. from sparse count data and heterogeneous observation RNE Réseau National d’échouage (2008) Université de La efforts. Ecological Modelling, 193, 615-628. Rochelle et le Groupe d’Etudes des Cétacés de Méditerranée Notarbartolo di Sciara G., Gordon J. (1997) Bioacoustics: a http://crmm.univ-lr.fr/index.php/ff/echouages/carte (last tool for the conservation of cetaceans in the Mediterranean modified 2010). Sea. Marine and Freshwater Behaviour and Physiology, 30, Santangelo C., Bramanti L. (2006) Ecology through time; an 125-146. overview. Biology Forum, 99, 395-424. Notarbartolo di Sciara G., Aguilar A., Bearzi G., Birkun A., Zanarddelli M., Panigada S., Airoldi S., Borsani J.F., Jahoda Frantzis A. (2002) Overview of known or presumed impact M., Notarbartolo di Sciara G. (1999) Site fidelity, seasonal on the different species of cetaceans in the Mediterranean residence and sex ratio of fin whales ( Balaenoptera physalus ) and Black Seas. In Notarbartolo di Sciara G. (ed.), Cetaceans in the Ligurian Sea feeding grdouns. European Research on in the Mediterranean and Black Seas: State of Knowledge and Cetaceans, 12, 24. Conservation Strategies. Report to the ACCOBAMS Secretar iat, Monaco, February 2002, pp. 194-196, 219pp. A ppendix Notarbartolo di Sciara G., Zanardelli M., Jahoda M., Panigada 5., Airoldi S. (2003) The fin whale Balaenoptera physalus (L. We report some technical details on the various demo 1758) in the Mediterranean Sea. Mammal Review, 33, 105- graphic measures used in the paper. 150. Notarbartolo di Sciara G., Agardy T., Hyrenbach D., Scovazzi T., Van Klaveren P. (2007) The Pelagos Sanctuary for Medi The mathematical model of the mortality table terranean marine mammals. Aquatic Conservation: Marine The mortality table is based on the following piecewise Freshwater Ecosystem, 18, 367-391. constant mortality rate over the various (continuous) age- Orsi Relini L., Palandri G., Garibaldi F., Lantieri L. (2004) stages [ai_!,ai) Note about some population parameters of the fin whale Balaenoptera phisalns (Linneo 1758). Biol. Marine Med., 11, /((a) = ( - a 8 Marine Ecology32 (Suppl. 1 ) (2011 ) 1-9 © 2011 Blackwell Verlag GmbH Arrigoni, Manfredi, Panigada, Bramanti & Santangelo Demography of fin whales from strandings surviving until the mature stage times the life expectancy in the mature stage. The corresponding mean age of e0 = / s(a)da = ^ 1 ( 1 (A3) £ e - mothers at the birth of their female offspring (computed i=i A with reference to the stationary population of birth den Similar equations can be derived for the life expectan sity R0"1m(a)p(a)) is given by; cies at the age of entry in the immature and mature stages, denoted by e!, e2 in the main text. For example / 1 \ e-P3(ffl-32) T — I a2 H I — (oj — a2 ) ------(^3V) the life expectancy at the age a2 of entry in the mature stage is given by: a^=co The intrinsic growth rate r of the population, which represents the speed of growth or decay that the popula (', 4 ) tion would achieve in the long-term on the assumption es-/32 By combining the life-table with suitable assumptions that the vital rates are maintained constant over time, is on age-specific fertility rates of female whales (taken as given as: given) we can compute a variety of reproduction indices. r = mFp(a2)e-ra2 (l - e- Marine Ecology 32 (Suppl. 1) (2011) 1-9 © 2011 Blackwell Verlag GmbH 9 anmarine evolutionary perspective ecology » Marine Ecology. ISSN 0173-9565 ORIGINAL ARTICLE Temporal change in UK marine communities: trends or regime shifts? M. Spencer1*, S. N. R. Birchenough2*, N. Mieszkowska3*, L. A. Robinson1*, S. D. Simpson4*, M. T. Burrows5, E. Capasso3,6, P. Cleall-Harding6, J. Crummy7, C. Duck8, D. Eloire9,10, M. Frost3, A. J. Haii8, S. J. H aw kins3'6, D. G. Joh n s11, D. W . Sim s3'12, T. J. Sm yth9 & C. L. J. Frid1* 1 School of Environmental Sciences, University of Liverpool, Liverpool, UK 2 Cefas Laboratory, Pakefleld Road, Lowestoft, Suffolk, UK 3 Marine Biological Association of the United Kingdom, The Laboratory, Citadel HUI, Plymouth, UK 4 School of Biological Sciences, University of Bristol, Bristol, UK 5 The Scottish Association for Marine Science, Dunstaffnage Marine Laboratory, Oban, Argyll, UK 6 School of Ocean Sciences, Bangor University, Menai Bridge, Ynys Mon, UK 7 British Geological Survey, Murchison House, Edinburgh, UK 8 Sea Mammal Research Unit, Scottish Oceans Institute, University of St Andrews, St Andrews, UK 9 Plymouth Marine Laboratory, Prospect Place, The Hoe, Plymouth, Devon, UK 10 Laboratoire Ecosystème Lagunaire, UMR B119, CNRS - Université Montpellier II - IRD - IFREMER, CC093, Montpellier, France 11 Sir Allster Hardy Foundation for Ocean Science, Plymouth, UK 12 Marine Biology and Ecology Research Centre, Marine Institute, School of Marine Sciences and Engineering, University of Plymouth, Drake Circus, Plymouth, UK Keywords A bstract Abundance; population trends; principal components; regime shift detection; regime A regime shift is a large, sudden, and long-lasting change in the dynamics of shifts; time series; UK marine ecosystems. an ecosystem, affecting multiple trophic levels. There are a growing number of papers that report regime shifts in marine ecosystems. However, the evidence Correspondence for regime shifts is equivocal, because the methods used to detect them are not Matthew Spencer, School of Environmental yet well developed. We have collated over 300 biological time series from seven Sciences, University of Liverpool, Liverpool L69 marine regions around the UK, covering the ecosystem from phytoplankton to 7ZB, UK. E-mail: [email protected] marine mammals. Each time series consists of annual measures of abundance for a single group of organisms over several decades. We summarised the data *These authors wish to be considered as joint for each region using the first principal component, weighting either each time first authors. series or each biological component (e.g. plankton, fish, benthos) equally. We then searched for regime shifts using Rodionov’s regime shift detection (RSD) Accepted: 16 November 2010 method, which found regime shifts in the first principal component for all seven marine regions. However, there are consistent temporal trends in the doi : 10.1111/j.1439-0485.2010.00422.x data for six of the seven regions. Such trends violate the assumptions of RSD. Thus, the regime shifts detected by RSD in six of the seven regions are likely to be artefacts caused by temporal trends. We are therefore developing more appropriate time series models for both single populations and whole commu nities that will explicitly model temporal trends and should increase our ability to detect true regime shift events. Lathrop 2008; Greene et al. 2008; Hagerthey et al. 2008; Introduction Heath & Beare 2008; Hemery et al. 2008; Petersen et al. A growing number of studies report major changes in 2008). These, often high profile, observations have con biological systems (Reid et a í 2001; Rudnick & Davis tributed to a move towards more holistic and integrated 2003; Lees et a í 2006; Beaugrand et al. 2008; Carpenter & ‘ecosystem-based’ environmental management (United 10 Marine Ecology 32 (Suppl. 1) (2011) 10-24 © 2010 Blackwell Verlag GmbH Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et al. Temporal change in UK marine communities Nations 1992). Several such studies have highlighted rela However, it seems important to evaluate whether the tively large-scale change in aspects of the system over a assumption of stationarity in mean except at discrete shift short period of time (e.g. Hare 8c Mantua 2000; Reid points is appropriate for a broad range of biological time et al. 2001; Chavez et al. 2003; Weijerman et al. 2005; series. Daskalov et al. 2007). Such phenomena have been Here, we apply RSD to a large collection of marine referred to as ecological regime shifts and have been seen biological time series from seven regions around the Brit as evidence of non-linear interactions and feedbacks in ish Isles. We argue that many of the apparent shifts found the ecological system, with the potential for hysteresis by RSD are in fact the consequence of gradual, rather (Beisner et al. 2003; Potts et al. 2006). A large number of than sudden, changes over time. We think that such grad such events have been documented (Folke et al. 2004). ual changes are biologically interesting, and are important For example, the Thresholds database (accessed 27 Febru because they form a dynamic baseline of genuine and ary 2010) contains 102 instances of regime shifts in sometimes large changes in marine ecosystems (Hard- terrestrial, freshwater and marine ecosystems (Resilience man-Mountford et al. 2005). However, they do not fall Alliance, Santa Fe Institute 2004). Lees et al. (2006) within the usual definition of regime shifts. provide a review of many of the ‘regime shift’ papers published prior to 2005 and emphasise that for changes Material and Methods actually to constitute a regime shift, the change must propagate across multiple physical and biological compo Data nents of the ecosystem. The UK’s Marine Environmental Change Network In the majority of cases where a regime shift was pos (MECN) links research units and universities that hold tulated it was attributed to changes in the climatic system long-term or historic data on aspects of UK marine eco (Lees et al. 2006). There is little doubt that planetary systems. The MECN dataset compiled for this study warming is occurring and the temperature records suggest includes 324 biological time series of annual observations that this has been most rapid in the last three decades, from seven marine regions (Fig. 1). The series cover five with a warming trend apparent in most atmospheric and biological components (plankton, infaunal benthos, rocky sea surface temperature datasets (IPCC 2007, section 1.1). shore invertebrates, fish, and marine mammals), although Many published analyses of biological data focus on not all components were represented in every region. All regime shifts that may have been caused by environmen series within a region were reduced to the length of the tal change (e.g. Reid et al. 2001; Weijerman et al. 2005; shortest series from that region (from 19 to 30 years, fin Beaugrand et al. 2008). This raises the fundamental ques ishing in 2006: Table 1). Some components were sampled tion: do biological interactions generally result in discon tinuous dynamics at the system level? The answer has profound implications for understanding and predicting the impacts of global climate change. A number of studies have examined time series of bio logical and physical variables simultaneously (Hare 8c Mantua 2000; Weijerman et al. 2005), often by summariz ing many time series using principal components. Statisti cal methods can then be used to look for sudden changes in the levels of the principal components. The regime shift detection (RSD) algorithm is one such method (Ro dionov 2004, 2005, 2006). RSD was initially developed for climatic data such as the Pacific Decadal Oscillation (Ro dionov 2004, 2005), but has subsequently been applied to biological data such as the abundances of organisms in several trophic levels in the Black Sea ecosystem (Daska lov et al. 2007), fish stocks in four marine ecosystems (Link et al. 2009) and plankton in the Northern Adriatic -10° W -5° W 0° 5° E Sea (Kamburska 8c Fonda-Umani 2009). Fig. 1. UK marine regions used in this study. The regions are the RSD assumes that the univariate time series of interest standard continuous plankton recorder regions (Richardson et al. (such as the first principal component of a multivariate 2006): B2 (Northern North Sea); C2 (Central North Sea); C3 (Irish dataset) is stationary in mean except at an unknown Sea); C4 (West Scotland); D2 (Southern North Sea); D3 (English Chan number of discrete regime shifts (Rodionov 2004). nel); and D4 (Celtic Sea). Marine Ecology32 (Suppl. 1) (2011) 10-24 © 2010 Blackwell Verlag GmbH 11 Temporal change in UK marine communities Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et al. Table 1. Summary of the number and length of time series, and the types of biological component covered by those series In each of the seven marine regions (the continuous plankton recorder areas - see Fig. 1). The last year of data used for any of the series was 2006. region number of time series length (years) start year biological components B2 35 30 1977 Marine mammals, fish, Zooplankton, phytoplankton C2 65 28 1979 Marine mammals, fish, Infaunal benthos, Zooplankton, phytoplankton C3 37 19 1988 Fish, Zooplankton, phytoplankton C4 38 19 1988 Marine mammals, fish, Zooplankton, phytoplankton D2 34 30 1977 Marine mammals, fish, Zooplankton, phytoplankton D3 68 19 1988 Fish, rocky shore Invertebrates, Zooplankton, phytoplankton D4 47 25 1982 Fish, rocky shore Invertebrates, Zooplankton, phytoplankton more often than annually. For these, we calculated annual influenced by the Tamar Estuary, with increased nutrients means. More detailed information on the variables for and periodic incursions of fresher surface water following each component is given below. heavy rain (Rees et al. 2009; Smyth et al. in press). Weekly Zooplankton samples have been collected at L4 since 1988, using vertical net hauls from the sea floor to Plankton the surface of a WP2 net with a mesh-size of 200 p m and Continuous plankton recorder data a 0.5-m diameter aperture corresponding to a mouth area For every region, we obtained eight time series derived of 0.25 m2 (UNESCO 1968; Southward et al. 2005). The from monthly continuous plankton recorder (CPR) sam 22 species and groups used cover more than 99% of ples collected by the Sir Alister Hardy Foundation for the total Zooplankton abundance at this site. We used the Ocean Sciences (SAHFOS). These were the phytoplankton lowest taxonomic level available for each group, and no colour index (PCI), total abundances of diatoms and species contributed to more than one variable. Originally, dinoflagellates, echinoderm and decapod larvae, and eup- the Zooplankton time series contained three missing data hausids, and abundances of the important copepods (Eloire et al. in press). The missing data for January and Calanus finmarchicus (boreal) and Calanus helgolandicus February 1988 were replaced by the average value of the (temperate). The PCI is based on the ‘greenness’ of each month over the entire time series. For August 2000, sample, as referenced to standard colour charts, giving one the missing data were replaced by the average value of of four category values per sample. These values are based the monthly averages of the previous and following on a ratio scale of acetone extracts using spectrophotomet- months, and the average value for August over the entire ric methods, and give an indication of phytoplankton bio time series. Finally, annual averages were calculated from mass (Richardson et al. 2006; section 5.1). For all other the monthly averages. planktonic samples, the units are the number of organisms per sample where each sample represents approximately Infaunal benthos 10 nautical miles (18.5 km) of tow, which equates to 3 m3 of filtered seawater (Batten et al. 2003). For all these vari For region C2 (Central North Sea), we obtained infaunal ables, we calculated annual means from the monthly con macrobenthic data from the Dove M l time series (Bucha tinuous plankton recorder samples. For some variables in nan & Moore 1986a). These data are based on five 0.1 -m2 some years, there were no individuals observed in the grabs collected in September each year. The Dove station majority of months (Richardson et al. 2006). Ml (55°07' N, 01°20' W) is 10.5 km off the NE English coast. It has predominantly sandy sediment, with a 20% English Channel Zooplankton data silt-clay content and lies in 55 m of water (Frid et al. In region D3 (English Channel), we also obtained 22 Zoo 1996, 2009). Sampling commenced in September 1972 plankton variables from Station L4, which is situated in and the dataset analysed here covers samples taken in the Western English Channel (50° 15.00' N, 4° 13.02' W) September of each year between September 1979 and and forms part of the Western Channel Observatory run 2006. No samples were taken, due to weather or opera by the Plymouth Marine Laboratory working with the tional constraints, in September 1987, 1991 and 2002. Marine Biological Association of the UK. The water is Buchanan & Warwick (1974) and Buchanan & Moore 50 m deep and is tidally influenced, with a 1.1-knot sur (1986b) describe the methods of sampling in detail. face stream at mean spring tide. Typically stratification The data used in this study are total genera abundance starts in early April, persists throughout the summer and per square metre based on at least five replicate samples is eroded by the end of October. Station L4 is strongly (Frid et al. 2009). Analysis at the genus level avoided any 12 Marine Ecology32 (Suppl. 1) (2011) 10-24 © 2010 Blackwell Verlag GmbH Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et al. Temporal change in UK marine communities problems in identification at the species level, or changes and Aquaculture Science (CEFAS) (Ellis et al. 2005) dur in taxonomy leading to problems with homonyms. The ing five separate long-term surveys. Four surveys used full dataset included 327 genera. To extract a shorter otter trawls [B2, C2 and D2 (Northern, Central and number of macrobenthic variables, genera were ranked Southern North Sea) and D4 (Celtic Sea)], and three used separately based on total abundance across all years and beam trawls [D3 (English Channel), D4 (Celtic Sea), and persistence (frequency of occurrence). The ranks were C3 (Irish Sea)]. All surveys were conducted in autumn summed and the combined score used to select the top 30 (August-December), fish were identified to species level, species. For years where there were missing data, an inter and catches were standardised to catch per hour. For each polated value was obtained by averaging the densities of of the seven SAHFOS CPR Standard Areas, we calculated the 2 years before and the 2 years after the missing year. the mean catch per trawl per year for all 179 species, and then identified the dominant 30 species in each area in a way that combines persistence and abundance. Species Rocky shore invertebrates were ranked by abundance from highest (1) to lowest (0), Quantitative, replicated counts of abundance were made the ranks were multiplied by the proportion of times the annually using replicate 50-cm2 quadrats for the boreal species appeared in hauls, and the species with the 30 limpet species Patella vulgata (Linneaus) and the lusita- highest scores were selected. Similar methods were used nian Patella depressa (Pennant) in the midshore region of by Genner et al. (2004, 2010). Pelagic species [i.e. those semi- to exposed rocky shores in regions D3 (Western with an adult pelagic phase) were subsequently removed English Channel: seven locations) and D4 (Celtic Sea: because of concerns that the gear used in these surveys eight locations for P. vulgata, but only seven for P. de sampled them incidentally. Thus, some important aspects pressa, whose numbers were not recorded at the eighth of ecosystem change such as the ratio of pelagic to site in some years). The cosmopolitan Patella ulyssiponen demersal fish (de Leiva Moreno et al. 2000) cannot be sis was also counted, but was excluded from our analyses detected by our analyses. because most counts in most years were zero. In region D4 (Celtic Sea) these data were supplemented These sites were part of a wider UK survey (Mies by the ‘Standard Haul’ series from the Marine Biological zkowska et a l 2006). Three surveyors (S. Hawkins, Association’s Laboratory, Plymouth, as part of the Wes M. Burrows and N. Mieszkowska) were involved in data tern Channel Observatory. Otter trawls were undertaken collection, and have undertaken multiple cross-calibration at 30-50 m depth over a spatial scale of 51 X 22 km off exercises to ensure continuity and standardisation in Plymouth (50°08'-50°20' N, 03°55'-04°39' W) during collection methodology across the time series. Not all 1911, 1913-1914, 1919-1922, 1950-1958, 1967-1979, sites were surveyed in every year and there are gaps in the 1983-1994 and 2001-2007. The abundance of demersal time series. A least-squares fit general linear regression fish taxa was recorded. Over the series, seven vessels have model with sites and years as fixed factors was fitted to been used for sampling, ranging in overall length from the log10(x + 10)-transformed survey data to generate 18.3 to 39.0 m. Where data are available, trawls were predicted values for combinations of sites and years for undertaken at the same speeds and were comparable in which data was missing. The complete data matrix had dimensions: headline length range, 16.2-19.8 m; ground- 435 elements for each species. Of these, 227 elements rope length range, 19.8-27.4 m; main net stretched mesh (52%) contained real data, and 208 missing elements diameter, 75-270 mm, and all vessels used a fine-mesh (48%) were filled using model data. Calculated values of cod end or a cover (Genner et a l 2010). Data included in R 2 were P. depressa 0.84 and P. vulgata 0.77. We subse this analysis were standardised mean catch per hour, of quently excluded P. depressa at Lynmouth because as a all hauls during the year, for 30 species, as an annual range edge location it was not recorded in all years, lead average of all trawls collected, as previous analyses of this ing to several zero values. Although we are concerned data suggested interannual variation is much stronger about the number of missing data in these time series, we than seasonal variation (Genner et a l 2010). think it important to consider their inclusion. Without them, the rocky shore habitat, on which many studies of Marine mammals the effects of climate change have focused (Hawkins et al. 2009), would not be represented in our analyses. Estimates of total grey seal ( Halichoerus grypus) pup p ro duction were obtained at individual breeding colonies in regions B2 (Northern North Sea: Orkney), C2 (Central Fish North Sea: Isle of May, Fast Castle, Farne Islands), D2 The most extensive fish survey data used in this analysis (Southern North Sea: Lincolnshire, Norfolk), and C4 were collected by the Centre for Environment Fisheries (West Scotland: Inner and Outer Hebrides). All seal data Marine Ecology32 (Suppl. 1 ) (2011 ) 10-24 © 2010 Blackwell Verlag GmbH 13 Temporal change in UK marine communities Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et al. were derived from surveys either conducted or reported components with the maximum possible autocorrelation, by the Sea Mammal Research Unit (Duck & Mackey rather than variance. This is useful for identifying smooth 2008; Duck et al. 2008). There are no reliable data on trends in time series. However, sudden changes such as total population size. Pup production at a fixed set of regime shifts do not necessarily result in strong autocor sites can be measured reliably, although its relationship to relations and might not be extracted by MAFA. As we population size is complicated by density dependence in want to determine whether changes are sudden or some cases. Pup production in B2, C2 (apart from the gradual, we think that principal components are more Farne Islands), and C4 was estimated from repeated aerial appropriate than MAFA for our purposes. surveys during the breeding season using maximum likeli The time series in each region were grouped into bio hood methods (Duck & Mackey 2008). C4 data were logical components (Table 1). We are interested in regime combined estimates of production from 11 colonies in shifts that affect multiple categories. However, the the Inner Hebrides and 15 in the Outer Hebrides from unweighted principal component analysis gives equal 1984 onwards. Prior to 1983, data are for colonies in the weight to each time series. Categories for which many Outer Hebrides only. B2 data are from up to 26 colonies time series are available may therefore dominate the first in Orkney. C2 data were from two colonies in the Firth principal component. This may be undesirable because of Forth, of which only one (a new colony) was included the number of time series in a category reflects only the in 1997. No aerial survey data were available in 1983 for availability and taxonomic resolution of data. We there most colonies. Where necessary, we used the mean of the fore also calculated weighted principal components, in values from 1982 and 1984 as an estimate of 1983 pup which we gave equal weight to each category rather than production for B2 and C2 (we did not need to interpolate each time series. To do this, we found the principal com for C4 because the starting year for this region was 1988). ponents of the transformed variables Pup production for the Farne Islands (C2) is the cumula x ij(k) — fij(k) tive total from repeated ground counts carried out by zm = — ¡— r ~ National Trust staff. Data for the southern North Sea \J nkGm (D2) are from three relatively recently established colonies that are similarly ground-counted by staff from Lincoln where x ^ is the value of the ith observation from vari shire Wildlife Trust, the National Trust and Natural able j, which is in category k, f t ^ i s the sample mean for England. the jth variable, oj^is the sample variance for the jth variable, and nk is the number of variables in the /rth cat egory (Deville & Malinvaud 1983; Jolliffe 2002, section Statistical methods 14.2.1). The RSD algorithm is designed to detect changes in uni The RSD algorithm is designed to detect changes at variate time series (Rodionov 2004). We therefore used discrete times in an otherwise-stationary time series (Ro the first principal component of the data for each region dionov 2004). If there are temporal trends in populations, as a univariate summary of the major patterns of com the original variables will not be stationary in mean, and munity change. Data for each time series were natural log it is unlikely that the first principal component will be (x + 1)-transformed because we expect a positive rela stationary in mean. However, if the population is growing tionship between level and variability (we added 1 to each at a constant average rate, the first differences (the differ observation because some time series contained zeros). ence between the value in a given year and the value in We then centred and scaled the log-transformed data so the previous year) of the log-transformed observations that each series had mean 0 and standard deviation 1. We will be stationary in mean. We therefore also analysed the calculated principal components of the centred and scaled first principal component of the first differences of log- data for each region, treating each year as an observation transformed data, using both the unweighted and and each time series as a variable. Because the data are weighted approaches described above. When this is done, time series, observations in successive years are not inde the events that RSD is searching for will be changes in pendent. However, such dependencies are not a serious the average rate of change of the first principal compo problem when the principal components are used as nent, rather than changes in the level of the first principal descriptions of data (Jolliffe 2002; section 12.1). Maxi component. Such changes are of biological interest, and mum autocorrelation factor analysis (MAFA: Solow 1994) correspond to changes in a weighted sum of population is an alternative technique for extracting components that growth rates, but they are not always included in defini describe changes in multivariate time series which explic tions of regime shift (reviewed by Lees et al. 2006). itly deals with autocorrelation. It differs from principal We used Version 2.1 of the MATLAB implementa components analysis in that it extracts orthogonal tion of the RSD software (downloaded 21 December 2009 14 Marine Ecology32 (Suppl. 1) (2011) 10-24 © 2010 Blackwell Verlag GmbH Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et al. Temporal change in UK marine communities fro m http://www.climatelogic.com/stars.html ). RSD of concerns about gear changes and other sampling arte (Rodionov 2004) searches for shifts in the level of a sta facts in the fish data, we repeated all the analyses without tionary time series by performing t-tests on individual the fish. Similarly, we repeated all the analyses without observations, with the null hypothesis that the nth obser the limpets, for which many observations were interpo vation is drawn from the same population as the preced lated using a linear model. ing sequence of observations. If the null hypothesis of no shift is initially rejected, follow-up tests are performed on Results a specified number of subsequent observations. The null hypothesis may not be finally rejected if these subsequent A fairly large number of principal components are needed observations do not appear consistent with the proposed to account for most of the variability in the data (Fig. 2: shift. All analyses were done with default parameters weighted principal components gave similar results, not [I (cutoff length) = 10, a (nominal size of test) = 0.05, shown). Over all transformations, principal component h (Huber weight parameter) = 1]. We have not used any methods, and regions, the first principal component correction for multiple testing among regions because we explained between 16 and 47% of the variability in the see RSD mainly as an exploratory tool, and its statistical data (Table 2). Thus, although the first principal compo properties are not well enough known to undertake an nent is capturing a substantial amount of variability, there appropriate correction. Within any single time series, tests are important features of the data that are not readily at each time point have nominal size a. However, the sta summarised in one dimension. tistical properties of the follow-up tests have not been When applied to the first principal component examined in detail. The null hypothesis is finally not (whether unweighted or weighted) of the log-trans rejected if any of these follow-up tests does not provide formed data, regime shift detection (Figs 3-9) found strong enough evidence against it. The overall size of the one regime shift in each of the regions C3 (Irish Sea, test at any time point is therefore not well defined but is Fig. 5A,C), C4 (West Scotland, Fig. 6A,C), and D3 certainly less than a. There is no correction in regime (English Channel, Fig. 8A,C), and two regime shifts in shift detection for multiple testing within a single time each of the regions B2 (Northern North Sea, series, which further complicates the issue. Fig. 3A,C), C2 (Central North Sea, Fig. 4A,C) (the last All analyses were done using MATLAB R2009b for regime shift here for the unweighted analysis was in Linux (The Mathworks, Inc., Natick, MA, USA). Because 2006, the final year of the series), D2 (Southern North Eigen o o L o 10 20 30 20 40 60 Component number Component number a> 15 10 20 30 10 20 30 Component number Component number a> 15 Fig. 2. Scree plots for unweighted principal 10 20 20 40 components analysis of the natural log Component number Component number (x + 1)-transformed data for each region. a> 20 Components are arranged on the horizontal axis in descending order of the am ount of variance they account for. The vertical axis (eigenvalue) is the variance of each 10 20 30 40 component. Component number Marine Ecology32 (Suppl. 1 ) (2011 ) 10-24 © 2010 Blackwell Verlag GmbH 15 Temporal change in UK marine communities Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et al. Table 2. Percentage variation explained by two different forms of between two apparently stationary regimes (Fig. 4A). first principal component (unweighted and weighted) applied to two Superficially, this is the pattern expected from a genuine different transformations [natural log (x+ 1) and first differences of regime shift. However, the step is absent from the natural log (x + 1)] of the data for each region. weighted first principal component (Fig. 4C). Separate unweighted unweighted weighted weighted analyses of the unweighted first principal component of region log first difference log first difference the log-transformed data for each category (Fig. 10) show that the step is present only in the infaunal benthos B2 27 16 44 28 C2 39 19 36 17 (Fig. 10B). Because the infaunal benthos make up 30 of C3 32 24 23 30 65 time series for C2, they dominate the overall pattern C4 31 17 47 38 when equal weight is given to each series, but not when D2 37 16 45 33 equal weight is given to each category. The absence of the D3 21 20 24 26 step in other categories suggests that whatever change D4 36 18 29 18 occurred in the seabed community was not transmitted to the other categories included in this study, and is Sea, Fig. 7A,C), and D4 (Celtic Sea, Fig. 9A,C). How therefore not a regime shift in the usual sense of the ever, in all regions other than C2 (Central North Sea), term. the log-transformed data do not appear stationary in Excluding either fish or limpets did not change the mean except at the shift points. Thus, except in C2, overall qualitative pattern, except that without fish there the apparent regime shifts found by the regime shift was no evidence of either trends or step changes in C3 detection algorithm may be more appropriately (Irish Sea, neither unweighted nor weighted: results not described as trends. shown). Thus, the trend in C3 (Fig. 5A,C) is largely The unweighted first principal component of the log- driven by changes in the fish data, whether real or arte- transformed data for C2 shows a strong step in 1995 factual. B2 unweighted B2 unweighted first difference 1980 1985 1990 1995 2000 2005 1980 1985 1990 1995 2000 2005 Year Year weighted B2 weighted first difference O CL 1980 1985 1990 1995 2000 2005 1980 1985 1990 1995 2000 2005 Year Year Fig. 3. Regime shift detection applied to region B2 (Northern North Sea), using two different forms of principal component (unweighted and weighted) and two different transformations [natural log (x + 1) and first differences of natural log (x + 1)]. (A) Unweighted first principal compo nent of the natural log (x + 1 [-transformed data. (B) Unweighted first principal component of the first differences of natural log (x + 1 [-trans formed data. (C) Weighted first principal component of the natural log (x + 1 [-transformed data. (D) Weighted first principal component of the first differences of natural log (x + 1 [-transformed data. In each panel, time in years is on the horizontal axis, the vertical axis is the value of the first principal component, the solid lines are the observed values, and the dashed lines are the regime means found by regime shift detection with default parameters [/ (cut-off length) = 10, a (nominal size of test) = 0.05, h (Huber weight parameter) = 1], 16 Marine Ecology32 (Suppl. 1) (2011) 10-24 © 2010 Blackwell Verlag GmbH Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et al. Temporal change in UK marine communities A C2 unweighted C2 unweighted first difference o CL "O0 £O) 0 'o c5 => -5 -10 -10 -20 1980 1985 1990 1995 2000 2005 1980 1985 1990 1995 2000 2005 Year Year C D C2 weighted C2 weighted first difference o o Q_ Q_ T3 T3 0 0 Fig. 4. Regime shift detection applied to -2 -2 regime C2 (Central North Sea). See Fig. 3 '1980 1985 1990 1995 2000 2005 '1980 1985 1990 1995 2000 2005 legend for explanation. Year Year C3 unweighted C3 unweighted first difference 1990 1995 2000 2005 1990 1995 2000 2005 Year Year C3 weighted C3 weighted first difference O Q_ g> '0 £ Fig. 5. Regime shift detection applied to regime C3 (Irish Sea). See Fig. 3 legend for 1990 1995 2000 2005 1990 1995 2000 2005 explanation. Year Year The first principal components of the first differences principal component explained more of the variability in of the log-transformed data (Figs 3-9B,D) are approxi the log-transformed data than in the first differences of mately stationary in mean. In all but three cases (C2 log-transformed data. This is consistent with the idea that unweighted, Fig. 4B; D3 unweighted, Fig. 8B; D4 the dominant pattern in the log data is a temporal trend, weighted, Fig. 9D), no regime shifts were detected in the which is relatively easy to capture in one dimension. principal components of first differences. In these three exceptional cases, the detected shifts were trivial (very D iscussion small and in the final observation point). Thus, overall, there is little evidence for consistent changes in popula There have been many reports of responses to climate tion growth rates. In all but two cases (Table 2), the first change in marine species from the seas around the UK. Marine Ecology 32 (Suppl. 1) (2011) 10-24 © 2010 Blackwell Verlag GmbH 17 Temporal change in UK marine communities Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et al. A B C4 unweighted C4 unweighted first difference o o Q_ Q_ T3 T3 0 0 JZg> '0 '0 1 -5 cs Z> Z> -10 -5 1990 1995 2000 2005 1990 1995 2000 2005 Year Year C D C4 weighted C4 weighted first difference o o Q_ Q_ T3 T3 0 0 -2 Fig. 6. Regime shift detection applied -3 -2 1990 1995 2000 2005 1990 1995 2000 2005 to regime C4 (West Scotland). See Fig. 3 Year Year legend for explanation. A B D2 unweighted D2 unweighted first difference o O Q_ Q_ ■a ■a 0 0 .g> c .g> c 0 0 s S c C Z> Z) -5 -5 1980 1985 1990 1995 2000 2005 1980 1985 1990 1995 2000 2005 Year Year C o Q_ ■a 0 .ro - 1 -2 D2 weighted D2 weighted first difference Fig. 7. Regime shift detection applied -2 -3 1980 1985 1990 1995 2000 2005 1980 1985 1990 1995 2000 2005 to regime D2 (Southern North Sea). See Year Year Fig. 3 legend for explanation. Over the past few decades poleward shifts in biogeo primary producers or primary consumers, and this mis graphic boundaries have been recorded for plankton match has the potential to disrupt ecosystem function (Beaugrand & Reid 2003), intertidal rocky benthos (Mies (Thackeray et a l 2010). Trends in abundance have also zkowska et a l 2006), subtidal benthos (Hinz et a l in been documented, including increases in population review) and fish (Perry et a l 2005). Changes in phenology abundances of benthic and pelagic species with warm are also widespread and important in marine systems. For water affinities close to northern range limits and cold example, the advancement of spring and summer events water species close to southern range limits (Beaugrand has occurred more slowly in secondary consumers than 2003; Beaugrand & Ibanez 2004; Mieszkowska et a l 2005, 18 Marine Ecology 32 (Suppl. 1) (2011) 10-24 © 2010 Blackwell Verlag GmbH Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et al. Temporal change in UK marine communities A B D3 unweighted o o CL Q_ T3 T3 0 '0I '0I =§ -10 =§ -10 D3 unweighted first difference -15 -15 1990 1995 2000 2005 1990 1995 2000 2005 Year Year C D D3 weighted o o Q_ Q_ T3 T3 0 0 -2 -2 Fig. 8. Regime shift detection applied to D3 weighted first difference -3 -3 regime D3 (English Channel). See Fig. 3 1990 1995 2000 2005 1990 1995 2000 2005 legend for explanation. Year Year D4 unweighted o O Q_ Q_ T3 T3 0 0 -C g> '0s '0 c I -5 3 -5 =) D4 unweighted first difference -10 -10 1985 1990 1995 2000 2005 1985 1990 1995 2000 2005 Year Year D4 weighted D4 weighted first difference o o Q_ Q_ T3 T3 0 0 -2 —2 Fig. 9. Regime shift detection applied to —3 -4 regime D4 (Celtic Sea). See Fig. 3 legend for 1985 1990 1995 2000 2005 1985 1990 1995 2000 2005 explanation. Year Year 2007; Vance 2005; Rees et al. 2006; Callaway et al. 2007; decades, rather than sudden shifts that affect many compo Hiddink & ter Hofstede 2008; Genner et al. 2010; Eloire nents of the community at the same time. Trends instead et al. in press; Widdicombe et al. in press). However, of sudden shifts seem plausible ecologically for this area of long-term assessments that cover a combination of time the world given that physical conditions are largely being series observations representing a wider range of ecosys driven by the gradual response to climatic trends such as tems components are limited in UK waters (Southward changing sea temperature (Southward et al. 2005). There is et al. 2005). evidence of multiple equilibria in a wide range of mathe Overall, changes in UK marine communities appear to matical models for ecosystems including North African be dominated by gradual trends over the last two to three vegetation (Higgins et al. 2002), coral reefs (Mumby et al. Marine Ecology32 (Suppl. 1 ) (2011 ) 10-24 © 2010 Blackwell Verlag GmbH 19 Temporal change in UK marine communities Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et al. A marine mammals Infaunal benthos o o CL CL T3 T3 Q) Ql - 2 -5 1980 1985 1990 1995 2000 2005 1980 1985 1990 1995 2000 2005 Year Year C D fish phytoplankton o o CL CL T3 T3 Fig. 10. Unweighted first principal Q) Ql component of natural log (x + 1)-transformed data for region C2 (Central North Sea), separated Into biological categories [A: marine -5 -5 1980 1985 1990 1995 2000 2005 1980 1985 1990 1995 2000 2005 mammals (two time serles), B: Infaunal Year Year benthos (30 time series), C: fish (25 time E series), D: phytoplankton (three time series), Zooplankton o E: Zooplankton (five time series)]. In each CL panel, time In years Is on the horizontal axis, T3ai the vertical axis Is the value of the first principal component, the solid lines are the observed values, and the dashed lines are the -5 1980 1985 1990 1995 2000 2005 regime means found by regime shift Year detection. 2007) and lakes (Carpenter 2005). In such models, gradual was little evidence for a discrete regime shift between two changes in physical conditions, or large enough distur locally stable states. In contrast, Beaugrand & Reid (2003) bances to state variables such as the abundances of organ report multiple temporal discontinuities in North Sea isms, can lead to sudden changes in ecosystem state biological time series. However, the events they are (Beisner et al. 2003). There is no doubt that regime shifts detecting are of a different type, because they analysed can occur, and have important socioeconomic conse one taxon (e.g. euphausiids, copepods) at a time, and quences (Folke et al. 2004). However, we do not see com looked at a small number of taxa over a longer time per pelling evidence for such sudden changes in our data. iod than that of our data. They calculated the principal The same may be true of other ecosystems to which components of spatiotemporal data for each taxon, and regime shift detection has been applied. For example, the used a method which will detect adjacent 6-year blocks changes in six time series from the Black Sea interpreted with significantly different mean values. Thus, the events as regime shifts by Daskalov et al. (2007, their Figure 1) they detect are relatively short-term changes in the abun could be interpreted as noisy trends rather than discrete dances of individual taxa. However, they noted that long changes in level. The same is true of the changes in phy term trends were the dominant pattern in their data, toplankton and Zooplankton in the Northern Adriatic which is consistent with our results. (Kamburska & Fonda-Umani 2009; their Figure 13) and Our analyses do suggest that there was a potential in the first principal component of a set of 114 physico regime shift in region C2 (Central North Sea) around chemical and biological variables from the North Sea 1995 driven primarily by the infaunal benthos. The infau (Kenny et al. 2009; their Figure 11). However, distin nal benthic time series has been previously analysed in guishing between these alternatives statistically requires isolation in a more extensive form. Among the 89 domi more sophisticated time series models that are currently nant genera, there are a series of sudden changes in in development. In the meantime, we should not too species composition at 5-10-year intervals between 1972 readily accept the idea of step-like regime shifts when and 2005 (Frid et al. 2009). Despite these changes, trends appear plausible. This conclusion is consistent with higher-level properties such as total abundance and gen some other, smaller-scale statistical analyses of UK marine era richness remained roughly constant. The most marked time series. For example, Solow & Beet (2005) found that changes in species composition identified in the 33-year although there were substantial changes in the abun series were in the early 1980s and early 1990s. In the dances of phytoplankton, copepods, cod, haddock and shorter time series analysed using RSD here, the 1980s herring in the North Sea between 1963 and 1997, there shift may not have been detected because it was too close 20 Marine Ecology32 (Suppl. 1) (2011) 10-24 © 2010 Blackwell Verlag GmbH Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et al. Temporal change in UK marine communities to the start of the dataset. The 1995 shift we detected ment strategies need to consider the entire ecosystem, may correspond to the early 1990s shift found by Frid including abiotic components, rather than focusing only on et al. (2009). Thus, the RSD analysis is consistent with one ecosystem component alone. Furthermore, the integra previous reports of roughly decadal shifts in species com tion of direct human effects on exploited species with direct position in the Central North Sea. However, these shifts or indirect climatic effects is needed for a realistic perspec did not propagate to the other components of the ecosys tive on ecosystem responses. Our analyses have highlighted tem analysed here for this region (plankton, fish, and multi-decadal trends across a broad range of ecosystem marine mammals). There are three possible (and not components, but further analyses will be required to exam mutually exclusive) explanations. First, the replacement of ine the likely drivers of the trends observed. one polychaete-dominated assemblage by another may Although across a range of ecosystem components, our have little impact on other trophic levels. This could analysis has concentrated on changes in species abun occur if many of the polychaete species are similarly suit dances. Ecosystem approaches to management also con able food items for consumers such as fish, and consume sider ecosystem functions and services. Biological traits resources at similar rates. Secondly, if stocks of many fish analysis (Bremner et al. 2006) assumes that ecosystem species in the region were reduced as a result of exploita functions are linearly related to species abundances. If this tion (FAO 2009, p. 196), there may have been little scope is true, then our results imply gradual changes in ecosys for a response to changes in prey abundance. Thirdly, the tem function for most biological components of most UK benthic data were from a single location, and we do not marine ecosystems. However, at least some ecosystem ser know the geographical extent of the changes in species vices (such as coastal protection provided by marshes, composition. Local changes may have had little effect on mangroves, seagrasses, and coral reefs) are nonlinearly wide-ranging consumers, or on samples taken from other related to species abundances (Koch et al. 2009). In such locations. cases, there may be utility thresholds (Samhouri et a l Distinguishing gradual from sudden change is important 2010), such that management actions produce much for ecosystem-based management. Thrush & Dayton greater responses for some ecosystem states than others, (2010) discuss several examples of ‘ecological ratchets’, in even in the absence of regime shifts. This is another area which a sustained fishing impact pushes an ecosystem into in which models with broader scope are needed. a new state from which recovery to the original state is dif ficult. In such cases, removing the impact will not restore Conclusions the system. In contrast, the lack of strong evidence for widespread regime shifts in our data suggests that most Overall, change in UK marine communities may be better components of UK marine ecosystems might return to described as temporal trends than as abrupt regime shifts, states observed in previous decades if both abiotic variables although abrupt shifts may have occurred in some and human impacts were returned to their 1970s levels. regions. Future analyses of change in UK marine commu This does not imply that irreversible changes did not occur nities should therefore be based on statistical models that before the start of our data, and may tell us little about the explicitly include trends, but also allow the possibility of risk of future regime shifts (Thrush et al. 2009). abrupt shifts. One promising approach is to use state- Information on the responses of multiple trophic levels space models (Durbin & Koopman 2001), which are to climate change is required to feed into national and much more general and flexible than many other international legislation (e.g. UK Marine & Coastal Access approaches to time-series analysis, and which can be Bill, UK Climate Change Act, EU Habitats Directive and applied easily to non-stationary time series. EU Marine Strategy Framework Directive). Furthermore, Looking at large numbers of time series across an the Common Fisheries Policy review recognises the need entire region complements the analysis of individual time for anthropogenic impacts on marine ecosystems and spe series, and broadens our ability to describe and under cies to be viewed in the context of pervasive climate change stand regime shifts and other ecosystem changes. We have (Commission of the European Communities 2008a,b, shown that getting meaningful results from such an 2009). The effects of climate change on marine ecosystems approach requires not only collaboration between large are important both economically and ecologically, whether numbers of data providers, but also the critical evaluation they are gradual or sudden. For example, climate-driven of statistical methods. changes in plankton may be an important determinant of cod recruitment in the North Atlantic (Beaugrand & Kirby Acknowledgements 2010), and declines in the abundance and distribution of kelps are likely to be decreasing available nursery grounds The authors would like to thank the Marine Environ for juvenile fish (Mieszkowskaet al. 2005). Thus, manage mental Change Network ( http://www.mba.ac.uk/mecn) Marine Ecology 32 (Suppl. 1) (2011) 10-24 © 2010 Blackwell Verlag GmbH 21 Temporal change in UK marine communities Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et al. and Defra for funding and organising the data analysis using biological traits analysis (BTA). Ecological Indicators, workshop. Defra has also provided funding through the 6, 609-622. MECN to support a number of the time series used in Buchanan J.B., M oore J.J. (1986a) A broad review o f variability this analysis including the Dove, PML, Liverpool Bay, and persistence in the Northumberland benthic fauna - MBA fish time series and MarClim benthic datasets, 1971-85. Journal of the Marine Biological Association of the and the Cefas fish data. The Natural Environment United Kingdom, 66, 641-657. Research Council (NERC) Oceans 2025 Strategic Buchanan J.B., M oore J.J. (1986b) Long-term studies at a Research Programme funded data collection by the Wes Benthic Station off the coast of Northumberland. Hydrobio- logia, 142, 121-127. tern Channel Observatory, a research collaboration Buchanan J.B., Warwick R.M. (1974) An estimate of benthic between PML and the MBA. Additional funders macrofaunal production in the offshore mud of the North acknowledged for their contributions to various time umberland coast. Journal of the Marine Biological Association series are Defra, Countryside Council for Wales, Envi of the United Kingdom, 54, 197-222. ronment Agency, JNCC, English Nature/Natural Eng Callaway R., Engelhard G.H., D ann J., Cotter J., R um ohr H. land, NERC, Scottish Government, Scottish Natural (2007) A century of North Sea epibenthos and trawling: Eleritage, States of Guernsey, The Crown Estates, and comparison between 1902-1912, 1982-1985 and 2000. WWF. We are grateful to Sergei Rodionov for discus M arine Ecology Progress Series, 346, 27-43. sion on the regime shift detection algorithm, Martin Carpenter S.R. (2005) Eutrophication o f aquatic ecosystems: Genner for his contributions to the 2009 MECN work bistability and soil phosphorus. Proceedings of the National shop on which this paper was based, and Claire Widdi- Academy of Sciences of the United States of America, 102, combe for assistance with plankton data. We would like 10002-10005. to thank two anonymous reviewers for thorough and Carpenter S.R., Lathrop R.C. (2008) Probabilistic estimate of a constructive criticism. Historical data retrieval at the threshold for eutrophication. Ecosystems, 11, 601-613. MBA was also assisted by funding from the NAGISA Chavez F.P., Ryan J., Lluch-Cota S.E., Niquen M. (2003) From History of the Nearshore Programme. anchovies to sardines and back: multidecadal change in the Pacific Ocean. Science, 299, 217-221. Commission of the European Communities (2008a) Directive R eferences 2008/56/EC of the European Parliament and of the Council Batten S.D., Clark R., Flinkman J., Hays G.C., John E., John of 17 June 2008 - establishing a framework for community A.W.G., Jonas T., Lindley J.A., Stevens D.P., Walne A. action in the field of marine environmental policy (Marine (2003) CPR sampling: the technical background, materials Strategy Framework Directive). Official Journal of the and methods, consistency and comparability. Progress in European Union, L164/19, 22. Oceanography, 58, 193-215. Commission of the European Communities (2008b) A Euro Beaugrand G. (2003) Long-term changes in copepod abun pean Strategy for Marine and Maritime Research. Brussels, dance and diversity in the north-east Atlantic in relation to COM: 534 pp. fluctuations in the hydroclimatic environment. Fisheries Commission of the European Communities (2009) Green Oceanography, 12, 270-283. Paper, Reform of the Common Fisheries Policy. Brussels, Beaugrand G., Ibanez F. (2004) Monitoring marine plankton COM: 28 pp. ecosystems. II: Long-term changes in North Sea calanoid Daskalov G.M., Grishin A.G., Rodionov S., Mihneva V. (2007) copepods in relation to hydro-climatic variability. Marine Trophic cascades triggered by overfishing reveal possible Ecology Progress Series, 284, 35-47. mechanisms of ecosystem regime shifts. Proceedings o f the Beaugrand G., Kirby R.R. (2010) Climate, plankton and cod. National Academy of Sciences of the United States of America, Global Change Biology, 16, 1268-1280. 104, 10518-10523. Beaugrand G., Reid P.C. (2003) Long-term changes in phyto Deville J.-C., Malinvaud E. (1983) Data analysis in official plankton, Zooplankton and salmon related to climate. Global socio-economic statistics. Journal of the Royal Statistical Change Biology, 9, 801-817. Society Series A, 146, 335-361. Beaugrand G., Edwards M., Brander K., Luczak C., Ibanez F. D uck C.D., Mackey B.L. (2008) Grey Seal Pup Production in (2008) Causes and projections of abrupt climate-driven Britain in 2007. Sea Mammal Research Unit, University of ecosystem shifts in the North Atlantic. Ecology Letters, 11, St Andrews, Available at: http://www.smru.st-and.ac.uk/ 1157-1168. documents/SCOS_2008_vl.pdf. Beisner B.E., Haydon D.T., Cuddington K. (2003) Alternative Duck C.D., Thompson D., Mackey B.L. (2008) The Status of stable states in ecology. Frontiers in Ecology and the Environ British Common Seal Popidations in 2007. Sea Mammal ment, 1, 376-382. Research Unit, University of St Andrews, Available at: Bremner J., Rogers S.I., Frid C.L.J. (2006) M ethods for describ http://www.smru.st-and.ac.uk/documents/SCOS_2008_ ing ecological functioning of marine benthic assemblages vl.pdf. 22 Marine Ecology32 (Suppl. 1) (2011) 10-24 © 2010 Blackwell Verlag GmbH Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et ai Temporal change in UK marine communities D urbin J., Koopm an S.J. (2001) Time Series Analysis by State Heath M.R., Beare D.J. (2008) New primary production in Space Methods. Oxford University Press, Oxford. northwest European shelf seas, 1960-2003. Marine Ecology Ellis J.R., Dulvy N.K., Jennings S., Parker-H um phreys M., Progress Series, 363, 183-203. Rogers S.I. (2005) Assessing the status of demersal elasmo- Hemery G., D’Amico F., Castege I., Dupont B., D’Elbee J., branchs in UK waters: a review. Journal of the Marine Lalanne Y., Mouches C. (2008) Detecting the impact of Biological Association of the United Kingdom, 85, 1025-1047. oceano-climatic changes on marine ecosystems using a Eloire D., Somerfield P.J., Conway D.V.P., Halsband-Lenk C., multivariate index: the case of the Bay of Biscay (North Harris R., Bonnet D. (in press) Temporal variability and Atlantic-European Ocean). Global Change Biology, 14, community composition of Zooplankton at station L4 in the 27-38. Western Channel: 20 years of sampling. Journal of Plankton Hiddink J.G., ter Hofstede R. (2008) Climate induced increases Research. in species richness of marine fishes. Global Change Biology, FAO (2009) The State o f World Fisheries and Aquaculture 2008. 14, 453-460. FAO, Rome: 196. Higgins P.A.T., Mastrandrea M.D., Schneider S.H. (2002) Folke C., Carpenter S., Walker B., Scheffer M., Elmqvist T., Dynamics of climate and ecosystem coupling: abrupt Gunderson L., Holling C.S. (2004) Regime shifts, resilience, changes and multiple equilibria. Philosophical Transactions of and biodiversity in ecosystem management. Annual Review the Royal Society o f London Series B Biological Sciences, 357, of Ecology Evolution and Systematics, 35, 557-581. 647-655. Frid C.L.J., Buchanan J.B., Garwood P.R. ( 1996) Variability and Hinz H., Capasso E., Lilley M., Frost M., Jenkins S.R. (in stability in benthos: twenty-two years of monitoring off review) Investigating temporal changes of sub-tidal benthic Northumberland. ICES Journal of Marine Science, 53, 978-980. communities across a bio-geographic boundary. Global Frid C.L.J., Garwood P.R., Robinson L.A. (2009) Observing Change Biology. change in a North Sea benthic system: a 33 year time series. IPCC (2007) Climate change 2007: Synthesis report. Contribu Journal of Marine Systems, 77, 227-236. tion of working groups I, II and III to the Fourth Assess Genner M.J., Sims D.W., Wearmouth V.J., Southall E.J., ment Report of the Intergovernmental Panel on Climate Southward A.J., Henderson P.A., Hawkins S.J. (2004) Regio Change. In: Pachauri R.K., Reisinger A. (Eds), Core Writing nal climatic warming drives long-term community changes Team. IPCC, Geneva. of British marine fish. Proceedings of the Royal Society of Jolliffe I.T. (2002) Principal Component Analysis. 2nd edn. London Series B Biological Sciences, 271, 655-661. Springer, New York. Genner M.J., Sims D.W., Budd G.C., Masterson P., Mchugh Kamburska L., Fonda-Umani S. (2009) From seasonal to deca M., Rendle E., Southall E.J., W earm outh V., Hawkins S.J. dal inter-annual variability of mesozooplankton biomass in (2010) Body size-dependent responses of a marine fish the Northern Adriatic Sea (Gulf of Trieste). Journal of Mar assemblage to climate change and fishing over a century- ine Systems, 78, 490-504. long scale. Global Change Biology, 16, 517-527. Kenny A.J., Skjoldal H.R., Engelhard G.H., Kershaw P.J., Reid Greene C.H., Pershing A.J., Cronin T.M., Ceci N. (2008) Arctic J.B. (2009) An integrated approach for assessing the relative climate change and its impacts on the ecology of the North significance of human pressures and environmental forcing Atlantic. Ecology, 89, S24-S38. on the status of Large Marine Ecosystems. Progress in Hagerthey S.E., Newman S., Rutchey K., Smith E.P., Godin J. Oceanography, 81, 132-148. (2008) Multiple regime shifts in a subtropical peatland: Koch E.W., Barbier E.B., Silliman B.R., Reed D.J., Perillo community-specific thresholds to eutrophication. Ecological G.M.E., Hacker S.D., Granek E.F., Primavera J.H., Muthiga Monographs, 78, 547-565. N., Polasky S., H alpern B.S., Kennedy C.J., Kappel C.V., Hardman-Mountford N.J., Allen J.I., Frost M.T., Hawkins S.J., Wolanski E. (2009) Non-linearity in ecosystem services: Kendall M.A., Mieszkowska N., Richardson K.A., Somerfield temporal and spatial variability in coastal protection. Fron P.J. (2005) Diagnostic monitoring of a changing environ tiers in Ecology and the Environment, 7, 29-37. ment: an alternative UK perspective. Marine Pollution Bulle Lees K., Pitois S., Scott C., Frid C., Mackinson S. (2006) Char tin, 50, 1463-1471. acterizing regime shifts in the marine environment. Fish and Hare S.R., Mantua N.J. (2000) Empirical evidence for North Fisheries, 7, 104-127. Pacific [climatic] regime shifts in 1977 and 1989. Progress in de Leiva M oreno J.I., Agostini V.N., Caddy J.F., Carocci F. Oceanography, 47, 103-145. (2000) Is the pelagic-demersal ratio from fishery landings a Hawkins S.J., Sugden H.E., Mieszkowska N., M oore P.J., useful proxy for nutrient availability? A preliminary data Poloczanska E., Leaper R., Herbert R.J.H., Genner M.J., exploration for the semi-enclosed seas around Europe. ICES Moschella P.S., Thompson R.C., Jenkins S.R., Southward Journal of Marine Science, 57, 1091-1102. A.J., Burrows M.T. (2009) Consequences of climate-driven Link J.S., Stockhausen W.T., Skaret G., Overholtz W., Megrey biodiversity changes for ecosystem functioning of North B.A., Gjosæter H., Gaichas S., D om m asnes A., Falk-Petersen European rocky shores. Marine Ecology Progress Series, 396, J., Kane J., M ueter F.J., Friedland K.D., Hare J.A. (2009) 245-259. A comparison of biological trends from four marine Marine Ecology32 (Suppl. 1 ) (2011 ) 10-24 © 2010 Blackwell Verlag GmbH 23 Temporal change in UK marine communities Spencer, Birchenough, Mieszkowska, Robinson, Simpson, Frid et a!. ecosystems: synchronies, differences, and commonalities. Rodionov S.N. (2005) A Brief Overview of the Regime Shift Detec Progress in Oceanography, 81, 29-46. tion Methods. UNESCO-ROSTE/BAS Workshop on Regime Mieszkowska N., Leaper R., Moore P., Kendall M.A., Burrows Shifts, Paris, 14-16 June 2005, Varna, Bulgaria: 17-24. M.T., Lear D., Poloczanska E., Hiscock K., Moschella P.S., Rodionov S.N. (2006) Use of prewhitening in climate regime Thom pson R.C., H erbert R.J., Laffoley D., Baxter J., South shift detection. Geophysical Research Letters, 33, L12707. ward A.J., Hawkins S.J. (2005) Marine biodiversity and climate Rudnick D.L., Davis R.E. (2003) Red noise and regime shifts. change: assessing and predicting the influence of climatic change Deep-Sea Research Part I, 50, 691-699. using intertidal Rocky Shore Biota. Occasional Publications, no. Samhouri J.F., Levin P.S., Ainsworth C.H. (2010) Identifying 20. Marine Biological Association of the United Kingdom, 53 pp. thresholds for ecosystem-based management. PLoS ONE, 5, Mieszkowska N., Kendall M.A., Hawkins S.J., Leaper R., e8907. Williamson P., Hardman-Mountford N.J., Southward A.J. Smyth T.J., Fishwick J.R., Al Moosawi L., Cum mings D.G., H ar (2006) Changes in the range of some common rocky shore ris C., Kitidis V., Rees A., Martinez Vincente V., Woodward species in Britain - a response to climate change? Hydrobio- E.M.S. (in press) A broad spatio-temporal view of the western logia, 555, 241-251. English Channel observatory. Journal of Plankton Research. Mieszkowska N., Hawkins S.J., Burrows M.T., Kendall M.A. Solow A.R. (1994) Detecting change in the composition of a (2007) Long-term changes in the geographic distribution multispecies community. Biometrics, 50, 556-565. and population structures of Osilinus lineatus (Gastropoda: Solow A.R., Beet A.R. (2005) A test for a regime shift. Fisheries Trochidae) in Britain and Ireland. Journal of the Marine Oceanography, 14, 236-240. Biological Association of the United Kingdom, 87, 537-545. Southward A.J., Langmead O., Hardman-Mountford N.J., Mumby P.J., Hastings A., Edwards H.J. (2007) Thresholds and Aiken J., Boalch G.T., Dando P.R., Genner M.J., Joint I., the resilience of Caribbean coral reefs. Nature, 450, 98-101. Kendall M.A., Halliday N.C., Harris R.P., Leaper R., Mies Perry A., Low P.J., Ellis J.R., Reynolds J.D. (2005) Climate zkowska N., Pingree R.D., Richardson A.J., Sims D.W., Smith change and distribution shifts in marine fishes. Science, 308, T., Walne A.W., Hawkins S.J. (2005) Long-term oceano 1912-1915. graphic and ecological research in the western English Petersen J.K., Hansen J.W., Laursen M.B., Clausen P., Carsten- Channel. Advances in Marine Biology, 47, 1-105. sen J., Conley D.J. (2008) Regime shift in a coastal marine Thackeray S.J., Sparks T.H., Frederiksen M., Burthe S., Bacon ecosystem. Ecological Applications, 18, 497-510. P.J., Bell J.R., Botham M.S., Brereton T.M., Bright P.W., Potts D.L., H uxnian T.E., Enquist B.J., W eltzin J.F., Williams Carvalho L., Clutton-Brock T., Dawson A., Edwards M., D.G. (2006) Resilience and resistance of ecosystem func Elliot J.M., Harrington R., Johns D., Jones I.D., Jones J.T., tional response to a precipitation pulse in a semi-arid Leech D.I., Roy D.B., Scott W.A., Smith M., Smithers R.J., grassland. Journal o f Ecology, 94, 23-30. Winfield I.J., Wanless S. (2010) Trophic level asynchrony in Rees H.L., Pendle M.A., Limpenny D.S., Mason C.E., Boyd rates of phenological change for marine, freshwater and ter S.E., Birchenough S., Vivian C.M.G. (2006) Benthic restrial environments. Global Change Biology, in press. responses to organic enrichment and climatic events in the Thrush S.F., Dayton P.K. (2010) What can ecology contribute western N orth Sea. Journal of the Marine Biological Associa to ecosystem-based management? Annual Review of Marine tion of the United Kingdom, 86, 1-18. Science, 2, 419-441. Rees A.P., Hope S.B., W iddicom be C.E., Dixon J.L., W ood Thrush S.F., Hewitt J.E., Dayton P.K., Coco G., Lohrer A.M., ward E.M.S., Fitzsimons M.F. (2009) Alkaline phosphatase Norkko A., Norkko J., Chiantore M. (2009) Forecasting the activity in the western English Channel: elevations induced limits of resilience: integrating empirical research with the by high summertime rainfall. Estuariae Coastal and Shelf ory. Proceedings o f the Royal Society B Biological Sciences, Science, 81, 569-574. 276, 3209-3217. Reid P.C., Borges M.F., Svendsen E. (2001) A regime shift in UNESCO (1968) Monographs on Oceanographic Methodology: the North Sea circa 1988 linked to changes in the North Sea Zooplankton Sampling. United Nations, Paris. horse mackerel fishery. Fisheries Research, 50, 163-171. United Nations (1992) Convention on Biological Diversity. UN, Resilience Alliance, Santa Fe Institute (2004) Thresholds and New York: 31 pp. Alternate States in Ecological and Social-Ecological Systems. Vance T. (2005) Loss o f the Northern Species Alaria Escidenta Resilience Alliance, Available at: http://www.resalliance.org/ from Southwest Britain and Implications for Macroalgal index.php?id=183 (accessed 27 February 2010). Succession. University of Plymouth, Plymouth: 31. Richardson A.J., Walne A.W., John A.W.G., Jonas T.D., Lind- Weijerman M., Lindeboom H., Zuur A.F. (2005) Regime shifts ley J.A., Sims D.W., Stevens D., Witt M. (2006) Using con in marine ecosystems of the North Sea and Wadden Sea. tinuous plankton recorder data. Progress in Oceanography, M arine Ecology Progress Series, 298, 21-39. 68, 27-74. Widdicombe C.E., Eloire D., Harbour D., Harris R.P., Somer Rodionov S.N. (2004) A sequential algorithm for testing cli field P.J. (in press) Long-term phytoplankton community mate regime shifts. Geophysical Research Letters, 31, L09204, dynamics in the western English Channel. Journal of Plank doi: 10.1029/2004GL019448. ton Research. 24 Marine Ecology32 (Suppl. 1) (2011) 10-24 © 2010 Blackwell Verlag GmbH anmarine evolutionary perspective ecology ' » Marine Ecology. ISSN 0173-9565 ORIGINAL ARTICLE Trends in population dynamics and fishery of Parapenaeus longirostris and Nephrops norvegicus in the Tyrrhenian Sea (NW Mediterranean): the relative importance of fishery and environmental variables Alessandro Ligas1, Paolo Sartor1 & Francesco Colloca2 1 Centro Interuniversitario di Biología Marina ed Ecología Applicata 'G. Bacci', viale N. Sauro 4, Livorno, Italy 2 Dlpartlmento di Biología Animale e dell'Uomo, University of Rome 'La Saplenza', viale dell'UnlversItà 32, Rome, Italy Keywords A bstract Environmental variables; fishing effort; landings; series analysis; trawl survey. Temporal variation in the population abundance of the deep-water rose shrim p, Parapenaeus longirostris (Lucas, 1846) (Decapoda, Penaeidae), and the Correspondence Norway lobster, Nephrops norvegicus (Linnaeus, 1758) (Decapoda, Nephropi A. Ligas, Centro Interuniversitario di Biología dae), in the Tyrrhenian Sea (NW Mediterranean), were evaluated using Marina ed Ecología Applicata 'G. Bacci', viale time-series data (1994—2008) from experimental trawl surveys and commercial N. Sauro 4, 1-57128 Livorno, Italy. landings. The influence of several environmental variables (sea surface tempera E-mail: [email protected] ture, wind-mixing index and NAO index) and fishing effort indices (number Accepted: 1 February 2011 of days at sea per month and mean engine power of the trawl fleet) were inves tigated. The time series were analysed by means of min/max auto-correlation doi: 10.1111/j. 1439-0485.2011,00440.x factor analysis (MAFA) and dynamic factor analysis (DFA). The abundance of P. longirostris showed a clear increasing trend, significantly correlated with the fishing effort index (number of days at sea per month), the sea surface temper ature and the wind-mixing index. The temporal variations in the stock of P. longirostris, which has a preference for warm waters, were positively corre lated with the rise of the sea surface temperature and the decrease of wind cir culation. For N. norvegicus, an increasing trend of landings per unit of effort and recruitment index contrasted with a decreasing trend of relative population abundance (biomass and density indices). (e.g. higher growth rate, earlier age-at-maturity) (Fromen Introduction tin & Fonteneau 2001), effects on the populations of Understanding the causes and mechanisms of change in non-target species (e.g. cetaceans, sea birds, sea turtles) the abundance of species over time is a crucial issue in resulting from by-catch (Kaiser & De Groot 2000), sus marine ecology. Fishing exploitation is considered to be pension of superficial sediments (Smith et a í 2003), and one of the main factors determining the dynamics of a reduction of habitat complexity and alteration of ben marine populations and ecosystems (Baum et al. 2003; thic community structure (Kaiser et a í 2000). M orato et al. 2006; Ciannelli et al. 2008; Pauly 2009). Large-scale changes in climate and oceanographic condi These effects include: changes in predator-prey relation tions are also known to influence the dynamics of marine ships that lead to shifts in food-web structure (Cartes populations (Gislason et a í 2000; Lloret et a í 2001; et al. 2001), effects on abundance and body-size distribu Rothschild et a í 2005). For example, the influence of tions that can result in fauna dominated by small-size global warming in the 20th century on long-term changes individuals (Jennings et a í 2001), genetic selection for in phytoplankton concentration in the North Atlantic different physical characteristics and reproductive traits has been demonstrated (Reid et a í 2001). In the Marine Ecology32 (Suppl. 1 ) (2011 ) 25-35 © 2011 Blackwell Verlag GmbH 25 Trends in P. congirostris an d N. norvegicus Ligas, Sartor & Colloca Mediterranean Sea, enrichment of nutrients in surface ing effort, mainly due to the EU Common Fishery Policy, waters has been shown to affect pelagic food-web dynamics which promotes the reduction of fishing effort through and fishery productivity (Molinero et al. 2008). This in incentives to decommission. turn can affect deep-sea necto-benthic communities which To analyse the effects of environmental and anthropo are known to depend on the downward flux of organic genic factors on demersal communities, two species which matter from the surface layers (Company et a í 2008). display different life cycles and behavioural strategies were Additionally, changes in river discharge and surface selected: the deep-water rose shrimp, Parapenaeus longi production can alter trophic webs and assemblage compo rostris (Lucas, 1846) and the Norway lobster Nephrops sitions in deep Mediterranean waters (Cartes et a í 2009). norvegicus (Linnaeus, 1758). Parapenaeus longirostris is a C om pany et a í (2008) described how climate-driven fast-growing, short-lived species with thermophilic prefer cascading dense shelf water influences the ecology of deep- ence (Abellö et a l 2002), inhabiting the water column sea populations on a decadal timescale. Bartolino et a í layers close to the seabed. Nephrops norvegicus is a long- (2008) linked the wind circulation to the recruitment of lived decapod, typical of temperate and cold waters, the European hake, Merluccius merluccius, one of the most which dwells in burrows and exerts territorial behaviour important demersal species in Mediterranean waters. Many (Aguzzi et al. 2003). It was hypothesised that, due to other studies have highlighted significant relationships these contrasting characteristics, the two species would between large-scale atmospheric variables (such as the show different behaviours in relation to changes in envi North Atlantic Oscillation index, NAO, which is tradition ronmental and anthropogenic factors. From the results, ally defined as the normalised pressure difference between we suggest a mechanism and cause-effect relationships Azores and Iceland) or local scale (surface temperature, linking the atmospheric and environmental variables with wind circulation, etc.) atmospheric variables and demersal changes in the abundance of both species. populations (Lloret et a í 2001; Fariña & González Herraiz 2003; Zuuret al. 2003a,b; Zuur & Pierce 2004; Erzini 2005; Study area Maynou 2008; Cartes et al. 2009; González Herraiz et a l 2009). The study area covered part of the continental shelf and the Since the 1950s, a warming process has occurred in the upper and middle slope off the western coasts of Italy (Cen Western Mediterranean basin. This is demonstrated by tral-northern Tyrrhenian Sea, Fig. 1). The Tyrrhenian Sea both environmental changes (e.g. surface temperature is semi-enclosed between islands (Corsica, Sardinia and increase; see Vargas-Yánez et al. 2009) and biological Elba) and the mainland (Italy), and is separated from the changes (e.g. northward advance of thermophilic species; rest of the western basin by a channel of moderate depth. It see CIESM 2008). In addition to environmental changes, can therefore be considered a distinct entity within the Italian fishing grounds, similar to those of other Mediter Central-western Mediterranean basin (Artale et al. 1994; ranean countries, have been affected by a decrease of fish Gasparini et al. 2005). The circulation in the Tyrrhenian H ‘p O 117c r ? f c 137er VJ ------f ( Y , y. ví * Porto S.Stefano R o m a 500 k vi V- a «ta o Fig. 1. Study area; the main isobaths are shown, as well as the sampling stations V' investigated during the experimental trawl survey Medits 2008. The black triangles show . i the three points at which satellite data were ■T, “ J (TO collected (42°30' N, 11°00' E; 42°00' N, 12°00' E; and 41°00' N, 13°00' E). 26 Marine Ecology32 (Suppl. 1) (2011) 25-35 © 2011 Blackwell Verlag GmbH Ligas, Sartor & Colloca T rends in P. congirostrís an d N. norvegicus Sea is organised in a series of cyclonic (anti-clockwise) and recruits per square kilometre, no. recruits-km-2) was also anticyclonic (clockwise) gyres determined by the wind (Ar- computed. Following Mori et al. (2000) and Orsi Relini tale et a í 1994). Three main cold water gyres, two cyclonic et al. (1998), specimens under the size of 20 mm CL (car and one anticyclonic, have been detected. They undergo apace length) were considered recruits. The indices of significant seasonal change, particularly the central anticy N. norvegicus were computed taking into account only clonic gyre that spreads over most of the basin in spring the hauls carried out in the 200-800 m depth stratum. and summer and almost disappears in autumn and winter. To investigate the effect of hydrological conditions, The intermediate (LIW) and deep waters have a constant mean monthly values of satellite-derived (1994-2008) sea temperature (12.8-13.0 °C). Mixing of surface and deep surface temperature (SST, °C) and wind speed (W, m-s-1) layers by wind-driven turbulence enriches the upper layer were gathered from the Physical Oceanography Distrib with nutrients (Nezlin et a í 2004), giving the Tyrrhenian uted Active Archive Centre (PO.DAAC: http://pod- Sea a relatively high concentration of nutrients within the aac.jpl.nasa.gov/index.html). Data taken from three Mediterranean basin. locations (42°30/ N -11°00' E; 42°00/ N -12°00/ E and 41°00' N -13°00/ E; see Fig. 1) in the T yrrhenian Sea were used to compute a mean monthly value. A wind-mixing Material and methods index was calculated as the cube of the wind speed From 1994 to 2008, landing data were collected monthly according to Bartolino et al. (2008). Monthly data of the over 3-5 days of observation at the auction of Porto NAO from 1994 to 2008 were obtained from the Pacific Santo Stefano, one of the most important fishing har Fisheries Environmental Laboratory (PFEL: http://las . bours of the area. The exploitation of Parapenaeus longi pfeg.noaa.gov/). rostris takes place in the fishing grounds between 200 and The time series were explored by means of auto- and 400 m depth, while catches of Nephrops norvegicus are cross-correlation functions. The auto-correlation function obtained from a greater depth range (200-600 m) (Sbrana gives an indication of the amount of association between et al. 2003). The number of trawlers habitually targeting variables Yt and Yt_k, where the time lag k takes the val the two species decreased during the investigated period: ues 1, 2, 3, etc. (Zuuret al. 2007). Thus it is used to from 30 vessels in 1994 to 12 in 2008 (Sbrana et al. highlight the presence of cyclic patterns in time series. 2006). Data on specific composition of the landing (total Formulated differently, the auto-correlation with a time weight by species or commercial category) and fishing lag of k years represents the overall association between effort (number of fishing days) were collected for each values that are separated by k time points. vessel. The landing rates (landing per unit of effort, The cross-correlation function shows the relationship LPUE) were calculated by taking into account the fishing between Yt and Xt_k. Therefore this tool can be used to day as a unit of effort (kg per day per vessel). In addition, explore whether there is a (linear) relationship between two indices of fishing activity and capacity were com two variables (Zuur et al. 2007, 2009). In time series anal puted: (i) the total number of days at sea performed by ysis, the use of significantly cross-correlated variables the fleet per month, and (ii) the mean engine power should be avoided. The confidence intervals of the auto (kW) of the fleet per month. correlation were obtained from ±2/Vn, where n is the During the investigated period (1994-2008), two exper length of the time series. imental trawl surveys per year were carried out under the To analyse the long-term changes of the variables, framework of the International bottom trawl survey in cyclical patterns were removed from the data obtained by the Mediterranean (Medits) and the Italian demersal the seasonal decomposition by Loess smoothing (Zuur resources program (Grund). According to the sampling et al. 2007). The data were then analysed by means of protocols (see Relini 1998 and Bertrand et al. 2002), the multivariate time series analysis techniques: min/max Medits survey was performed in spring and the Grund auto-correlation factor analysis (MAFA) and dynamic fac survey in autumn. The two surveys were carried out tor analysis (DFA) to estimate common trends. These according to a depth-stratified sampling design with ran tools were used to estimate common underlying trends domly allocated hauls within each stratum. In addition, from the multiple time series dataset, and to evaluate the the number of hauls in each stratum was proportional to correlations with species abundance and environmental the surface of the stratum itself. The haul position of the and fishery factors. For this purpose, the time series of Medits trawl survey 2008 is shown in Fig. 1. LPUE and of biomass and density indices were used as Mean biomass (kg-km-2) and abundance (N-km-2) response variables and the environmental and fishing indices were calculated for both species to obtain time effort factors as explanatory variables. All analyses were series composed of two observations per year, for a total performed using the software BRODGAR 2.6.6 (http:// of 30 observations. A recruitment index (number of www.brodgar.com ). Marine Ecology32 (Suppl. 1 ) (2011 ) 25-35 © 2011 Blackwell Verlag GmbH 27 Trends in P. congirostris an d N. norvegicus Ligas, Sartor & Colloca MAFA can be described in various ways: a type of were significantly correlated. In the case of Parapenaeus principal component analysis especially for (short) time longirostris, LPUE time series were positively correlated to series; a method for extracting trends from multiple experimental trawl survey time series (biomass and den time series; a method for estimating index functions from sity indices) (Table la), suggesting a good match between time series; a smoothing method; or a signal extraction fishery-dependent and fishery-independent data. The procedure. The underlying idea is that a trend is associ recruitment index was significantly correlated to the ated with high auto-correlation at time lag 1. Therefore, biomass and density indices, and the maximum cross the first MAFA axis represents the trend, or the main correlations were at time lags 0 and 2, respectively underlying pattern in the data. This axis can also be seen (Table lb). This means that a peak of recruitment was as an index function or smoothing curve. Cross-correla directly reflected in a peak of density, while the peak of tions (canonical correlations) between the variables (both biomass followed with a time lag 2, which corresponds to response and explanatory variables) and the trends were 1 year. The correlations between the landing and survey computed to evaluate the significance of the relationship time series of Nephrops norvegicus were significant, but between the variables and the trends (Erzini et a í 2005; negative, suggesting an inverse relationship between the Z uur et al. 2007). The mathematics behind MAFA are two variables. However, the recruitment index was posi described in Solow (1994). The underlying formula is tively correlated to the LPUE, with the maximum correla similar to principal component analysis. The MAFA cal tion corresponding to a time lag 3. This suggests that a culations involve a principal component analysis on cen peak in recruitment was followed by a peak in LPUE after tred data, followed by a first-differencing on the principal a time lag of more than 1 year. components, and a second principal component analysis The analysis of cross-correlations among the explana on these differenced components. As a result, the MAFA tory variables also provided significant results (Table 2). axes are mutually uncorrelated with unit variance, and the MAFA axes have decreasing auto-correlation with time lag 1 (Zuur et al. 2007). Table 1. Response variables, (a) Cross-correlations. L = landing per The DFA is a method to estimate common trends, unit of effort (kg per day per vessel): B = biomass Index (kg-km-2); D effects of explanatory variables and interactions between = density Index (n-krrT2); R = recruitment Index (no. recruits-kirT2). (b) the response variables in a multivariate time series data Maximum cross-correlations: the time lags corresponding to the maxi set. Statistical details and applications of DFA are given mum cross-correlations are shown In the grey part of the table. in Z uur et al. (2003a,b) and Zuur & Pierce (2004). DFA Parapenaeus longirostris Nephrops norvegicus applies a dimension reduction to the N time series. The dynamic factor model, in words, is given by N Time L B D R L B D R series = linear combination of M common trends + (a) explanatory variables + noise. P. longirostris DFA models with one common trend and a symmetric, L 1.00 non-diagonal covariance matrix were used to analyse the B 0.54 1.00 datasets. A series of models were fitted, ranging from the D 0.49 0.95 1.00 R 0.12 0.46 0.69 1.00 simplest, with only one explanatory variable, to the most N. norvegicus complex, with all the explanatory variables. Akaike’s L 1.00 information criterion (AIC) was used as a measure of B -0 .5 5 1.00 goodness-of-fit and to compare models (Zuur et al. D -0.54 0.94 1.00 2003b), with the best model having the smallest AIC. Fac R 0.42 0.38 0.52 1.00 tor loadings were used to make inferences regarding the (b) importance of particular trends, representing underlying P. longirostris L 0.79 0.82 0.54 common patterns over time, both to specific response B 0.95 0.53 variables and to different groups of response variables D 0 0.69 (Erzini 2005; Erzini et a í 2005; Zuuret al. 2007). R 1 2 0 N. norvegicus L 0.66 0.66 0.54 Results B 0.94 -0 .4 7 The analysis of the time series dataset by means of D 0 -0 .5 5 R 3 1 -1 cross-correlation functions allows us to identify significant relationships between the response variables. For both Significance level for correlations: ± 0.37: significant correlations are species, time series of landings and trawl surveys data highlighted In bold. 28 Marine Ecology32 (Suppl. 1) (2011) 25-35 © 2011 Blackwell Verlag GmbH Ligas, Sartor & Colloca Trends in P. congirostris and N. norvegicus T ab le 2. Explanatory variables: cross-correlations. 0 .8 - SST W3 NAO Days kW c o « 0.4- SST 1.00 0 0 .2 - W3 -0 .4 5 1.00 o O NAO 0.19 -0.12 1.00 - 0 .2 - Days 0.30 -0.02 -0.02 1.00 -0.4 kW 0.18 0.14 0.00 0.41 1.00 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 Time lag SST = Sea Surface Temperature (°C); W3 = wind-mixing index (m3-s~3); NAO = North Atlantic Oscillation index; Days = days at sea Fig. 3. Auto-correlation plot of the landing per unit of effort (LPUE) per month; kW = mean engine power (kW). Significance level for time series of Parapenaeus longirostris and Nephrops norvegicus (thick correlations: ± 0.37. Significant correlations are highlighted in bold. line). Dotted lines: confidence Interval limits (± 0.15). Sea surface temperature (SST) and wind-mixing index 600-, (W3) were negatively correlated. SST showed an increas £ 500- ing trend, whereas the W3 series followed a decreasing ! 400- pattern. For the fishing efforts parameters, a positive cor o 300 - relation was found between days at sea and mean engine « 2 0 0 - power. nj Q 1 0 0 - According to the results obtained by means of cross correlations, it was decided to use the LPUE of P. longi 30-, B rostris and N. norvegicus only as response variables for the 26- analyses of time series. Among the explanatory variables, ? 2 2 - the wind-mixing index, the NAO index and the days at sea per month were used. The LPUE time series of P. longirostris and N. norvegi 14- cus were characterized by wide fluctuations, making it impossible to identify any clear trend (Fig. 2). Both time 12 C series showed peaks in late spring (April-June), when it is 10 known that the catch of the two species is higher. The 8 presence of a seasonal pattern was confirmed by the auto 0 6 £ correlation function: significant correlations were identi 4 fied at time lags of 12 and 24 months (Fig. 3). Some 2 examples of the time series of the explanatory variables 0 -I------1------1------1------1------1------1------1----- are shown in Fig. 4. The time series of days at sea per Jan-94 Jan-96 Jan-98 Jan-00 Jan-02 Jan-04 Jan-06 Jan-08 Months month was characterised by a clear decreasing trend (Fig. 4A). Figure 4B and C show fluctuations over time of Fig. 4. Time series plot of monthly data of explanatory variables from sea surface temperature and wind speed. Sea surface tem January 1994 to December 2008. (A) Number of days at sea per perature peaked in summer, when the wind speed was month (computed from the Porto Santo Stefano trawl fleet). (B) Sea surface temperature (SST, °C). (C) Wind speed (m-s~1). 7°-i lower. These trends explain the significant, but negative, ® 60- cross-correlation between the two variables. To remove 0 0 _Ä O 50- the seasonal patterns, the explanatory variables were also o 40- smoothed by means of the seasonal decomposition by Loess sm oothing (see Fig. 5). ■g 2°- The results obtained by means of MAFA described a io- clear scenario in the case of P. longirostris. The estimated Jan-94 Jan-96 Jan-98 Jan-00 Jan-02 Jan-04 Jan-06 Jan-08 trend showed an increasing pattern, although character Months ized by fluctuations (Fig. 6). High and positive correla Fig. 2. Landing per unit of effort (LPUE) time series of Parapenaeus tions between the trend and the LPUE time series of the longirostris and Nephrops norvegicus (thick line) from the Porto Santo two species were identified (Table 3). While for P. longi Stefano trawl fleet. rostris all the response variables considered (LPUE, Marine Ecology32 (Suppl. 1 ) (2011 ) 25-35 © 2011 Blackwell Verlag GmbH 29 Trends in P. congirostris and N. norvegicus Ligas, Sartor & Coi loca 0.3 According to the factor loadings, only P. longirostris was correlated to the trend computed by means of DFA (0.223 for P. longirostris, and 0.006 for N. norvegicus). The estimated trend (Fig. 7) was similar to that obtained o.o using MAFA, with a general increasing pattern and two main peaks in 2001 and 2006. Flowever, this model, cn -0 .1 which was characterized by the lowest AIC value - 0.2 (Table 4), suggested no significant relationship with fish 1.0 ing effort, as it was correlated only to the monthly time d)X 7? series of wind-mixing index (W3) and the NAO index. ■d Q) 0.5 .E w CD CO In fact, the estimated t-values for the regressions for individual species with W3 and NAO were relatively large - 1.0 ,------,------,------,- , , — ,------Jan-94 Jan-96 Jan-98 Jan-00 Jan-02 Jan-04 Jan-06 Jan-08 Months 1 0 - Fig. 5. Sea surface temperature (SST) and win d-m ixin g index time series obtained by means of seasonal decomposition by Loess smooth ing. o - 0.3-, 0 .2 - Û) o o —0.2 -I------1------.------1------.------1------.------1----- Fig. 7. Common trend computed by means of DFA from the LPUE Jan-94 Jan-96 Jan-98 Jan-00 Jan-02 Jan-04 Jan-06 Jan-08 time series of Parapenaeus longirostris and Nephrops norvegicus. Months Fig. 6. Common trend computed by means of MAFA from the LPUE Table 4. Values of Akalke's Information criterion (AIC) for DFA mod time series of Parapenaeus longirostris and Nephrops norvegicus. els with one common trend and different sets of explanatory variables (W3 = wind-mixing Index; NAO = North Atlantic Oscillation Index; Days = number of days at sea per month), based on diagonal and Table 3. Canonical correlations between the common trend obtained symmetric matrices. through MAFA and, respectively, the response variables (landing per unit of effort time series of Parapenaeus longirostris and Nephrops Model Matrix Explanatory variables AIC norvegicus), and the explanatory variables (W3 = wind-mixing Index; NAO = North Atlantic Oscillation index; Days = number of days at sea 1 Diagonal W3, NAO, Days 319.7 per month, days). 2 Diagonal W3, NAO 298.4 3 Diagonal W3, Days 307.5 Response variables Explanatory variables 4 Diagonal NAO, Days 178.2 5 Diagonal W3 291.5 P. longirostris 0.97 W3 -0 .45 6 Diagonal NAO 388.3 N. norvegicus 0.35 NAO -0 .09 7 Diagonal Days 248.5 Days -0.31 8 Diagonal - 258.0 Significance level for correlations: ± 0.1 S. Significant values are high 9 Symmetric W3, NAO, Days 296.3 lighted in bold. 10 Symmetric W3, NAO 159.0 11 Symmetric W3, Days 348.5 12 Symmetric NAO, Days 279.8 biomass, density and recruitment indices) were correlated 13 Symmetric W3 379.9 with the estimated trend, in the case of N. norvegicus, two 14 Symmetric NAO 406.9 15 Symmetric Days 369.7 contrasting scenarios emerged from the results: an 16 Symmetric - 372.1 increase in terms of LPUE and recruitment indices, and a decrease in terms of density and biomass indices. The lowest AIC value Is highlighted In bold. 30 Marine Ecology32 (Suppl. 1) (2011) 25-35 © 2011 Blackwell Verlag GmbH Ligas, Sartor & Colloca Trends in P. congirostrís an d N. norvegicus for N. norvegicus, indicating strong relationships involved in linking upper layers and benthic habitats have (Table 5). However, a large diagonal element of the error been made (Company et al. 2008; Cartes et al. 2009). covariance matrix (R > 0.74) was obtained for N. norvegi Many other environmental and oceanographic variables, cus, confirming that these variables did not fit well to the such as primary production, chlorophyll and nutrient model. concentrations, salinity, upwelling, currents, and river dis charge, have been shown to be influential in the life cycles and dynamics of marine ecosystems (Bahamón & Cru Discussion zado 2003; Erzini 2005; Erzini et al. 2005; Rothschild The present study aimed to understand the change over et al. 2005; Company et al. 2008; Cury et al. 2008; Cartes time of two demersal stocks in relation to environmental et al. 2009; Sardà et al. 2009). In the Tyrrhenian Sea, and anthropogenic factors, using analysis of a relatively oceanographic parameters are poorly and irregularly con long and complete time series of stock abundance data. sidered. The value of using SST, wind circulation and the In the Atlantic Ocean and North Sea, data have been col NAO index was the availability of an extensive and com lected since at least the 1950s (e.g. for many fish stocks in plete dataset covering the time span of the available fish the North Atlantic; Rijnsdorp et a í 2006). In contrast, in eries data. This enabled a thorough investigation to be the Mediterranean Sea, time series of fisheries data usually made into their influence over the course of 15 years. only cover the last few decades. Therefore, the availability Among the environmental variables used, only SST and of 15 years of fisheries data from two of the most abun the wind-mixing index (W3) were clearly related to the dant decapods of the demersal communities of Mediterra trend showed by P. longirostris and N. norvegicus, whereas nean waters, both important target species (Aguzzi et a í the NAO index was not significantly associated with 2004; Sobrino et al. 2005; Guijarro et al. 2009; Morello either. Parapenaeus longirostris is considered to be a spe et al. 2009), should be regarded as an unique opportu cies with a preference for warm waters, being more abun nity. dant in the Southeastern Mediterranean than in the The results clearly showed an increasing trend in the Northwestern basin (Abellô et al. 2002). The current abundance of the stock of P. longirostris and a more het warming of the upper and intermediate water layers of erogeneous scenario for N. norvegicus in the Tyrrhenian the Western Mediterranean (Vargas-Yánez et al. 2009), Sea during the investigated period. In addition, our reflected in the observed increase in SST and decrease in results strongly suggest that temporal variations in the wind circulation (W3), could have had a positive effect abundance of the two species were correlated with both on the life cycle and abundance of this species in the Tyr environmental and fishing activity variables. rhenian Sea. A possible explanation for this phenomenon is provided by Cartes et al. (2009). They hypothesised an association between high temperatures, low rainfall Environmental effects regimes and river discharges and a reduction in the flux Atmospheric and surface water environmental variables of organic matter that maintain deep-water benthic com were investigated in this study. Although these variables munities off the Catalonian coasts. These environmental have been used in several studies which highlighted signi conditions resulted in a higher abundance of Zooplankton ficant correlations with demersal and deep-sea communi and increased production of suprabenthos (Cartes et a í ties (Lloret et al. 2001; Fariña & González Herraiz 2003; 2009). Although P. longirostris displays a wide range of Z uur et al. 2003a,b; Zuur & Pierce 2004; Erzini 2005; prey items, its diet is mainly based on suprabenthic crus Maynou 2008; Cartes et al. 2009; González Herraiz et al. taceans, such as mysids (especially Lophogaster typicus) 2009), very few attempts to explain the mechanisms (Sobrino et al. 2005). Therefore, the warming phase observed in recent years could have favoured the P. longi rostris population in the Tyrrhenian Sea. In addition to Table 5. Estimated t-values for regressions between the response this, Bartolino et al. (2008) found a positive correlation variables (LPUE time series of Parapenaeus longirostris and Nephrops norvegicus) and the explanatory variables (W3 = wind-mixing Index; between the recruitment of the European hake, Merluccius NAO = North Atlantic Oscillation index). merluccius, and wind circulation in the Tyrrhenian Sea: high recruitment rates were associated with strong water t-value and wind circulation. The recruitment of the two species, W3 NAO M. merluccius and P. longirostris, takes place in the same area, at a bottom depth of 100-150 m (Colloca et a í P. longirostris 1.26 1.23 2004). European hake juveniles are known to prey on N. norvegicus -4 .8 2 -3.11 crustaceans, such as the juveniles of P. longirostris (Carp- Significant values are highlighted in bold. entieri et al. 2005b). The environmental conditions which Marine Ecology32 (Suppl. 1 ) (2011 ) 25-35 © 2011 Blackwell Verlag GmbH 31 Trends in P. congirostris an d N. norvegicus Ligas, Sartor & Colloca positively affect P. longirostris, such as high temperatures graphic evaluation by bottom trawl sampling. Commer and low wind circulation, are the same as those that neg cial fishing often operates on a 24-h basis, whereas atively affect the recruitment of M. merluccius (Bartolino experimental trawl surveys are usually carried out in the et a í 2008). The resulting lower predation pressure could daytime. Trawl fleets exploiting the Tyrrhenian Sea often have further enhanced the recruitment success of the perform 2-3 days of fishing operations, carrying out hauls deep-water rose shrimp. during both the day and night (Sbrana et a í 2003). In contrast, N. norvegicus, which showed a negative Although diurnal versus nocturnal bias in sampling has trend in terms of density and biomass indices, was nega been found to be moderate when trawl catches are per tively correlated with the NAO and the wind-mixing formed on fishing grounds on the continental slope (Agu index. Again this may be related to mechanisms linking zzi & Bahamón 2009; Bahamón et al. 2009), the fact that the productivity in the upper layers with the structure of N. norvegicus is a predominantly nocturnal species (Agu demersal communities, as proposed by Cartes et al. zzi & Sardá 2008) may help explain the differences (2009). While phases of warmer and dryer atmospheric observed between commercial and experimental survey conditions favour planktonic/suprabenthic feeders, ben data. Nephrops norvegicus spend most of the time within thic feeders and predators such as N. norvegicus (Aguzzi or at the entrance of their burrows and are caught by et a í 2009) are disadvantaged by the reduction of the trawling only when they emerge. Emergence varies with organic matter flux resulting from the decreased rainfall time of day, season, animal size, food presence, hunger and river discharge. Similar responses to atmospheric state, sex and reproductive status. Thus the fisheries warming processes by N. norvegicus have been observed exploit the population selectively and in a different man in other areas. In the context of only minor changes in ner with respect to males and females (Aguzzi et a í 2003; fishing pressure, Fariña & González Herraiz (2003) and Aguzzi & Sardá 2008; Aguzzi & Bahamón 2009). In par González Herraiz et al. (2009) showed a decline in the ticular, egg-bearing females spend most of their time in population abundance of N. norvegicus in the Atlantic. their burrows during the entire egg-incubation period, This decline was associated with the positive phase of the which lasts for 4-6 months in the Mediterranean Sea (Ag NAO index, which determines warmer temperatures in uzzi et a í 2003). Furthermore, juveniles rarely leave their Northern Europe (Halliday & Pinhorn 2009). burrows. These factors, related to the biology of the spe cies, therefore, strongly contribute to protection of the N. norvegicus life-stages that are perceived as sensitive to Fishing activity effects trawling exploitation. The fishing effort is a complex variable that is difficult to Parapenaeus longirostris also shows variations in density quantify because it is influenced by many different fac and depth distribution according to daytime rhythms and tors, such as increasing catch efficiency and changes in photoperiod length. Carpentieri et a í (2005a) observed fleet characteristics. Increasing catch efficiency of the fleet higher catch rates of P. longirostris at night in the shelf- (also known as ‘technological creep’) is usually related break in the Tyrrhenian Sea. In addition, they found the positively to an increase in skipper skills, investments in highest density index during late winter-spring, which auxiliary equipment, more efficient gear and materials, corresponds to the spawning peak (Ardizzone et a í replacement of old vessels with new ones and, to a lesser 1990). As most larvae occur around the 100-m isobath, extent, upgraded engines (Rijnsdorp et a í 2006). During adults could displace during the spawning period to shal the investigated period, a decrease in the number of ves lower depths (Sobrino et al. 2005). sels occurred: the fleets of Porto Santo Stefano and adja It is worth highlighting that the trend of P. longirostris cent ports decreased by about 50%, producing an almost was characterized by huge interannual fluctuations. Apart proportional decrease in fishing effort. from environmental conditions and fishing activity, this In the case of N. norvegicus the data showed two con interannual variability was probably related to the short trasting trends: an increase of landings per unit of effort life-span and fast growth rates of this species (Abellô and recruitment index, and a decrease of relative popula et al. 2002). A similar pattern, characterized by a biomass tion abundance. The daily activity of N. norvegicus could peak in 2001, was observed in other areas of the Western help explain these divergent trends. Light intensity influ Mediterranean, such as in the Balearic sub-basin (Guijar ences how organisms perceive their environment, modu ro et al. 2009). lating their inter- and intra-specific interactions. Demersal communities exposed to light intensity variations are Conclusions expected to react to them, producing changes in species composition and density. Therefore, the diurnal activity Long-term changes in the abundance of two important cycles of demersal species may consistently bias demo demersal species in the Tyrrhenian Sea, the deep-water 32 Marine Ecology32 (Suppl. 1) (2011) 25-35 © 2011 Blackwell Verlag GmbH Ligas, Sartor & Colloca T rends in P. congirostrís an d N. norvegicus rose shrimp ( Parapenaeus longirostris ) and the Norway cean Nephrops norvegicus on the activity rhythms of continen lobster (Nephrops norvegicus), were found to correlate sig tal margin prey decapods. Marine Ecology, 30, 366-375. nificantly with identified environmental and anthropo Aguzzi J., Sarda F. (2008) A history of recent advancements on genic factors. While the increasing abundance of Nephrops norvegicus behavioral and physiological rhythms. P. longirostris was correlated to a rise of sea surface tem Reviews in Fish Biology and Fisheries, 18, 235-248. perature, a corresponding decrease of wind circulation Aguzzi J., Sarda F., Abello P., Com pany J.B., Rotlant G. (2003) and to the reduction of fishing effort, a corresponding Diel and seasonal patterns of Nephrops norvegicus (Deca trend for N. norvegicus was not evident. On one hand, poda: Nephropidae) catchability in the western Mediterra nean. Marine Ecology Progress Series, 258, 201-211. the population abundance of N. norvegicus was negatively Aguzzi J., Sarda F., Allué R. (2004) Seasonal dynamics in correlated with environmental variations, while on the Nephrops norvegicus (Decapoda: Nephropidae) catches off other hand, it did not show any association with the gen the Catalan coast (Western Mediterranean). Fisheries eral decrease of fishing effort in the area. However, the Research, 69, 293-300. recruitment index, as calculated in the study, could be Ardizzone G.D., Gravina M.F., Belluscio A., Schintu P. ( 1990) used as a proxy for change in stock abundance. Depth-size distribution pattern of Parapenaeus longirostris Some mechanisms have been proposed to link atmo (Lucas, 1846) (Decapoda) in the central Mediterranean Sea. spheric conditions (sea surface temperature, wind circula Journal of Crustacean Biology, 10, 139-147. tion and the NAO) to the trophic webs and community Artale V., Astraldi M., Buffoni C., Gasparini G.P. (1994) Sea structure in the deep-water benthic habitats. However, sonal variability of gyre-scale circulation in the northern these models need to be improved to achieve a deeper Tyrrhenian Sea. Journal of Geophysical Research C, 99, and more accurate understanding of the mechanisms 14127-14137. linking these ecosystems. Further analyses are required to Bahamón N., Cruzado A. (2003) Modelling nitrogen fluxes in better understand the relationships between variations in oligotrophic environments: NW Mediterranean and NE the abundance of demersal species and environmental Atlantic. Ecological Modelling, 163, 223-244. and anthropogenic factors. The availability of suitable Bahamón N., Sarda F., Aguzzi J. (2009) Fuzzy diel patterns in information on environmental characteristics, such as catchability of deep-water species on the continental margin. appropriate sea-floor topography, sediment composition, ICES Journal of Marine Science, 66, 2211-2218. hydrographical characteristics, and trophic webs (prey Bartolino V., Colloca F., Sartor P., Ardizzone G.D. (2008) availability, presence and abundance of predator species) Modelling recruitment dynamics of hake Merluccius merluc is necessary to better understand the temporal change in cius, in the central Mediterranean in relation to key environ species abundance, distribution and biology. mental variables. Fisheries Research, 93, 277-288. Baum J.K., Myers R.A., Kehler D.G., W orm B., Harley S.J., Doherty P.A. (2003) Collapse and conservation of shark Acknowledgements populations in the Northwest Atlantic. Science, 299, 389- 392. The authors would like to thank the Editor and the Ref B ertrand J.A., Gil de Sola L., Papaconstantinou C., Relini G., erees for useful comments and revisions that significantly Souplet A. (2002) The general specifications of the MEDITS improved the structure and the quality of this paper. We survey. Scientia Marina, 66, 9-17. acknowledge also Dr Jessica Craig, University of Aber Carpentieri P., Colloca F., Ardizzone G.D. (2005a) Day-night deen, Anna Davies, Andrew Silva and Letizia Simeni for variations in the demersal nekton assemblage on the Medi proof-reading and English revision. terranean shelf-break. Estuariae and Coastal Shelf Science, 63, 577-588. Carpentieri P., Colloca F., Cardinale M., Belluscio A., Ardizz R eferences one G.D. (2005b) Feeding habits of European hake ( Merluc Abello P., Abella A., Adamidou A., Jukic-Peladic S., Maiorano cius merluccius ) in the central Mediterranean Sea. Fishery P., Spedicato M.T. (2002) Geographical patterns in abundance Bulletin, 103, 411-416. and population structure ofNephrops norvegicus and Parape Cartes J.E., Elizalde M., Sorbe J.C. (2001) Contrasting life his naeus longirostris (Crustacea: Decapoda) along the European tories, secondary production, and trophic structure of Perac- Mediterranean coasts. Scientia Marina, 66, 125-141. arid assemblages of the bathyal suprabenthos from the Bay Aguzzi J., Bahamón N. (2009) Modeled day-night biases in of Biscay (NE Atlantic) and the Catalan Sea (NW Mediter decapod assessment by bottom trawling survey. Fisheries ranean). Deep-Sea Research I, 48, 2209-2232. Research, 100, 274-280. Cartes J.E., M aynou F., Fanelli E., Papiol V., Lloris D. (2009) Aguzzi J., Bahamón N., Marotta L. (2009) The influence of light Long-term changes in the composition and diversity of availability and predatory behaviour of the decapod crusta deep-slope megabenthos and trophic webs off Catalonia Marine Ecology32 (Suppl. 1 ) (2011 ) 25-35 © 2011 Blackwell Verlag GmbH 33 Trends in P. congirostris an d N. norvegicus Ligas, Sartor & Colloca (western Mediterranean): are trends related to climatic oscil groundfish populations in the early 1990s. Fisheries Research, lations? Progress in Oceanography, 82, 32-46. 97, 163-182. Ciannelli L., Fauchald P., Chan K.S., Agostini V.N., Dingsor Jennings S., Dim ore T.A., Duplisea D.D., W arr K.J., Lancas G.E. (2008) Spatial fisheries ecology: recent progress and ter J.E. (2001) Trawling disturbance can modify benthic future prospects. Journal of Marine Systems, 71, 223-236. production processes. Journal o f A nim al Ecology, 70, 459- CIESM (2008) Climate warming and related changes in Medi 475. terranean marine biota. CIESM Workshop Monographs, 35, Kaiser M.J., De Groot S.J. (2000) Effects of Fishing on Non- 152. Target Species and Habitats: Biological, Conservation, Socio- Colloca F., Carpentieri P., Balestri E., Ardizzone G.D. (2004) A Economic Issues. Blackwell Science, Oxford: 399 pp. critical habitat for Mediterranean fish resources: shelf-break Kaiser M.J., Ramsay K., Richardson C.A., Spence F.E., Brand areas with Leptometra phalangium (Echinodermata: Crinoi A.R. (2000) Chronic fishing disturbance has changed shelf dea). Marine Biology, 145, 1129-1142. sea benthic community structure.Journal o f Anim al Ecololo- Company J.B., Puig P., Sarda F., Palanques A., Latasa M., gy, 69, 494-503. Scharek R. (2008) Climate influence on deep sea popula Lloret J., Lleonart J., Sole I., From entin J.M. (2001) Fluctua tions. PLoS ONE, 3, el431. doi:10.1371/journal.pone. tions of landings and environmental conditions in the 0001431 north-w estern M editerranean Sea. Fisheries Oceanography, Cury P.M., Shin Y.J., Planque B., D urant J.M., From entin 10, 33-50. J.M., Kramer-Schadt S., Stenseth N.C., Travers M., Grimm Maynou F. (2008) Environmental causes of the fluctuations of V. (2008) Ecosystem oceanography for global change in fish red shrimp ( Aristeus antennatus ) landings in the Catalan eries. Trends in Ecology and Evolution, 23, 338-346. Sea. Journal of Marine Systems, 71, 294-302. Erzini K. (2005) Trends in NE Atlantic landings (southern Molinero J.C., Casini M., Buecher E. (2008) The influence of Portugal): identifying the relative importance of fisheries the Atlantic and regional climate variability on the long and environmental variables. Fisheries Oceanography, 14, term changes in gelatinous carnivore populations in the 195-209. northwestern Mediterranean. Limnology and Oceanography, Erzini K., Inejih C.A.O., Stobberup K.A. (2005) An application 53, 1456-1467. of two techniques for the analysis of short, multivariate Morato T., Watson R., Pitcher T.J., Pauly D. (2006) Fishing non-stationary time-series of Mauritanian trawl survey data. down the deep. Fish and Fisheries, 7, 24-34. ICES Journal of Marine Science, 62, 353-359. Morello E.B., Antolini B., G ram itto M.E., Atkinson R.J.A., Fariña A.C., González Herraiz I. (2003) Trends in catch-per- Froglia C. (2009) The fishery for Nephrops norvegicus unit-effort, stock biomass and recruitment in the North and (Linnaeus, 1758) in the central Adriatic Sea (Italy): preli Northwest Iberian Atlantic Nephrops stocks. Fisheries minary observations comparing bottom trawl and baited Research, 65, 351-360. creels. Fisheries Research, 95, 325-331. Fromentin J.M., Fonteneau A. (2001) Fishing effects and life Mori M., Sbrana M., De Ranieri S. (2000) Reproductive biol history traits: a case study comparing tropical versus tem ogy of female Parapenaeus longirostris (Crustacea, Decapoda, perate tunas. Fisheries Research, 53, 133-150. Penaeidae) of the north Tyrrhenian Sea (western Mediterra Gasparini G.P., Ortona A., Budillon G., Astraldi M., Sansone nean). Atti della Societä Toscana di Scienze. Naturali Memo E. (2005) The effect of the Eastern Mediterranean Transient rie Serie B, 107, 1-6. on the hydrographic characteristics in the Strait of Sicily Nezlin N.P., Lacroix G., Kostianoy A.G., Djenidi S. (2004) and in the Tyrrhenian Sea. Deep-Sea Research Part I, 52, Remotely sensed seasonal dynamic of phytoplankton in the 915-935. Ligurian Sea in 1997-1999. Journal of Geophysical Research, Gislason H., Sinclair M., Sainsbury K., O’Boyle R. (2000) Sym 109, C07013, doi:10.1029/2000JC000628. posium overview: incorporating ecosystem objectives within Orsi Relini L., Zamboni A., Fiorentino F., Massi D. fisheries management. ICES Journal of Marine Science, 57, (1998) Reproductive patterns in Norway lobster Nephrops 468-475. norvegicus (L.) (Crustacea Decapoda Nephropidae) of Gonzalez Herraiz I., Torres M.A., Fariña A.C., Freire J., Cance different Mediterranean areas. Scientia Marina, 62, lo J.R. (2009) The NAO index and the long-term variability 25-41. o f Nephrops norvegicus population and fishery off West of Pauly D. (2009) Beyond duplicity and ignorance in global Ireland. Fisheries Research, 98, 1-7. fisheries. Scientia Marina, 73, 215-224. Guijarro B., M assuti E., M oranta J., Cartes J.E. (2009) Short Reid P.C., Borges M., Svendsen E. (2001) A regime shift in the spatio-temporal variations in the population dynamics and North Sea circa 1988 linked to changes in the North Sea biology of the deep-water rose shrimp Parapenaeus longiros horse mackerel fishery. Fisheries Research, 50, 163-171. tris (Decapoda: Crustacea) in the western Mediterranean. Relini G. ( 1998) Demersal trawl surveys in Italian seas: a short Scientia Marina, 73, 183-197. review. IFREMER, Actes de Colloques, 26, 46-75. Halliday R.G., Pinhorn A.T. (2009) The roles of fishing and Rijnsdorp A.D., Daan N., Dekker W. (2006) Partial fishing environmental change in the decline of Northwest Atlantic mortality per fishing trip: a useful indicator of effective 34 Marine Ecology32 (Suppl. 1) (2011) 25-35 © 2011 Blackwell Verlag GmbH Ligas, Sartor & Colloca T rends in P. congirostrís an d N. norvegicus fishing effort in mixed demersal fisheries. ICES Journal of Parapenaeus longirostris, in European Atlantic and Mediter Marine Science, 63, 556-566. ranean waters (Decapoda, Dendrobranchiata, Penaeidae). Rothschild B.J., Chen C., Lough R.G. (2005) Managing fish Crustaceana, 78, 1153-1184. stocks under climate uncertainty.ICES Journal of Marine Solow A.R. ( 1994) Detecting changes in the composition of a Science, 62, 1531-1541. multispecies community. Biometrics, 50, 556-565. Sarda F., Com pany J.B., Baham ón N., Rotllant G., Flexas M.M., Vargas-Yánez M., Moya F., Tel E., García-Martínez M.C., Sánchez J.D., Zúfiiga D., Coenjaerts J., Orellana D., Jordà G., Guerber E., Bourgeon M. (2009) Warming and salting in Puigdefábregas J., Sánchez-Vidal A., Calafat A., Martin D., the western Mediterranean during the second half of the Espino M. (2009) Relationship between environment and the 20th century: inconsistencies, unknowns and the effect of occurrence of the deep-water rose shrimp Aristeus antennatus data processing. Scientia Marina, 73, 7-28. (Risso, 1816) in the Blanes submarine canyon (NW Mediter Zuur A.F., Pierce G.J. (2004) Common trends in Northeast ranean). Progress in Oceanography, 82, 227-238. Atlantic squid time series. Journal of Sea Research, 52, Sbrana M., Sartor P., Belcari P. (2003) Analysis of the factors 57-72. affecting catch rates of crustacean trawl fishery of the north Zuur A.F., Fryer R., Jolliffe I., Dekker R., Beukema J. (2003a) ern Tyrrhenian Sea (western Mediterranean). Fisheries Estimating common trends in multivariate time series using Research, 65, 271-284. dynamic factor analysis. Environmetrics, 14, 665-685. Sbrana M., Viva C., Belcari P. (2006) Observation on the fish Zuur A.F., Tuck I.D., Bailey N. (2003b) Dynamic factor analy ery of the deep water rose shrimp Parapenaeus longirostris sis to estimate common trends in fisheries time series. Cana (Lucas, 1846) (Crustacea: Decapoda) in the northern Tyr dian Journal o f Fisheries and Aquatic Sciences, 60, 542-552. rhenian Sea (western Mediterranean). Hydrobiologia, 557, Zuur A.F., leño E.N., Smith G.M. (2007)Analysing Ecological 135-144. Data. Springer, London: 680 pp. Smith C.J., Rumohr LL, Karakassis I., Papadopoulou K.N. Zuur A.F., leño E.N., Elphick C.S. (2009) A protocol for data (2003) Analysing the impact of bottom trawls on sedimen exploration to avoid common statistical problems. Methods tary seabeds with sediment profile imagery. Journal of Exper in Ecology and Evolution, 1, 3-14. doi: 10.111 l/j.2041- imental M arine Biology and Ecology, 285-286, 479-496. 210X.2009.00001 .x Sobrino I., Silva C., Sbrana M., Kapiris K. (2005) A review of the biology and fisheries of the deep water rose shrimp, Marine Ecology32 (Suppl. 1 ) (2011 ) 25-35 © 2011 Blackwell Verlag GmbH 35 anmarine evolutionary perspective ecology » Marine Ecology. ISSN 0173-9565 ORIGINAL ARTICLE Biomass, commonly occurring and dominant species of macrobenthos in Onega Bay (White Sea, Russia): data from three different decades Katya Solyanko1, Vassily Spiridonov2 & Andrew Naumov3 1 Institute for Estuarine and Coastal Studies, University of Hull, Hull, UK 2 P.P. Shirshov Institute o f Oceanology of the Russian Academy of Sciences, Moscow, Russia 3 Zoological Institute o f the Russian Academy of Sciences, Unlversltetskaya Naberezhnaya, St. Petersburg, Russia Keywords A bstract Anthropogenic impact; biomass; decadal variation; European seas; macrobenthos. Onega Bay is the largest bay in the White Sea, characterised by shallow depth, a range of sediment types and strong tidal currents. All these factors provide Correspondence conditions for high species richness and biomass. This study reviews data from Katya Solyanko, Institute for Estuarine and three surveys of sublittoral macrobenthos undertaken by Russian institutes: the Coastal Studies, University of Hull, HU6 7RX benthic survey covering the entire Onega Bay in 1952; the survey performed in Hull, UK.E-mail: [email protected] the northern part of the area in 1981/90, and a study carried out in 2006 in Accepted: 1 February 2011 the eastern part of the bay. In total, data from 107 stations were analysed. The data in different surveys were collected by different grab types. The datasets of doi : 10.1111/j.1439-0485.2011,00438.x both 1981/90 and 2006 overlap the 1952 survey area. The pattern of biomass distribution was consistent between the years of survey and was characterised by the low biomass at the northern periphery of the bay and the highest bio mass observed in the coastal waters of the Solovetsky Islands. Bivalves and cir- ripeds (mostly Modiolus modiolus, Arctica islandica, Balanus balanus and Verucca stroemia) dominated in biomass. Neither the biomass share of domi nant species nor the frequency of occurrence of several common species in these groups changed markedly between 1952 and 1981/90. Although the results of the 2006 survey appear somewhat different from the patterns of pre vious years, this does not indicate major changes in the benthic communities, because the survey in 2006 was designed in a different way and its overlap with the 1952 survey was minimal. However, the dominant species (by biomass) - A. islandica, M. modiolus and V. stroemia - held their leading positions. Results of the multidimensional scaling analysis based on the biomass data for all taxa encountered in the 1952 survey indicate considerable mixing of the samples from all surveys. This may be interpreted as the absence of major shifts in the sublittoral communities of the macrobenthos of Onega Bay at decadal scale. This kind of stability may be explained by an oceanographical regime resilient to climate variation and a relatively low anthropogenic environmental impact when compared to other shallow European seas. species frequency of occurrence, population density and Introduction biomass often show considerable changes over time. Studies conducted in most European seas have shown Sometimes sudden inter-annual changes are detected, but that the composition of macrobenthic communities, changes are more likely across decades. Drastic shifts in 36 Marine Ecology 32 (Suppl. 1) (2011) 3 6-48 © 2011 Blackwell Verlag GmbH Solyanko, Spiridonov & Naumov Macrobenthos of Onega Bay (White Sea, Russia) species composition and structure of communities biological stations (Kudersky 1966; Fokin et al. 2006; detected in the Black Sea were found to be due to eutro Naumov 2006). The main objective of the present study phication, fishing and the introduction of alien species was to analyse data from three different decades with (Chikina & Kucheruk 2005). Reduced biomass and num regard to the composition, occurrence and biomass of ber of species was found in the Kattegat and Skagerrak dominant and common macrobenthic species and discuss due to direct effects of trawling, long-term temperature if any temporal pattern is revealed by these historical fluctuations and eutrophication of the area (Pearson et a í datasets. 1985; Rosenberg et a í 1987). Considerable changes in benthic communities were detected in the North Sea and Study area the Irish Sea due to eutrophication, bottom trawling, dredging, oil drilling operations and climate variation Onega Bay is the largest bay in the White Sea, with an (Frid et al. 1999, 2009; Bradshaw et al. 2002; W ieking & area of 12,800 km2. The depth of the bay is generally Krönke 2003; Krönke et a í 2004). In the Barents Sea, <50 m, with the exception of northern parts, where changes were associated with bottom trawling pressure depths can reach 87 m. The bottom relief is uneven, espe and climate variation (Galkin 1998; Brown et al. 2005; cially along the coastline. Particularly complex bathymetry Denisenko 2008; Carroll et al. 2008). These studies indi is observed along the bay’s western coast, where numer cate the role of anthropogenic effects on the composition ous islands are concentrated. A broad range of sediment of macrobenthic communities. types characterises Onega Bay, but coarse and hard sedi In the White Sea, the ‘youngest’ sea of Europe (existing ments with a small percentage of silt are the dominant only since the beginning of the Holocene), belonging to substrata (Berger & Naumov 2001). Onega Bay is con both the Northeast Atlantic and Arctic realms and charac nected to the central part of the sea by the Western and terised by a very peculiar oceanographical regime (Berger the Eastern Solovetsky Salma, or strait (Fig. 1). Deep & Naumov 2001; Filatov et al. 2005a,b), there has been waters of the Salmas enable large volumes of water to no attempt to analyse historical datasets on subtidal ben enter the bay, generating strong tidal currents exacerbated thic communities. The emphasis of previous studies has by the shallow depths in the Bay. The maximum speed of been on identifying spatial patterns (Derjugin 1928; Ku- a spring tide is 1.5-2.0 m-s_1 in the Eastern Salma, and dersky 1966; Beklemishev et al. 1980; Golikov et al. 1985; 1.5-1.7 m-s_1 in the Western Salma (Babkov 1998; Filatov Lukanin et al. 1995; Berger & Naumov 2001); long-term et al. 2005a). Strong tidal currents increase the turbidity temporal trends in benthic communities have received of the water, leading to vertical homothermy and homo- relatively little attention. Nevertheless, there has been a halinity in many parts of the bay. A developed thermo- long tradition of benthic research associated with marine cline is largely absent in most areas of the Northern and Solovetsky Islands ▲ R/V "Professor M esy atsev", 1952 (0.1 m2 Petersen grab) • R/V "Kartesh", 1981/1990 (0.25 m2 "Ocean" grab) ♦ R/V "Professor V. Kuznetsov", 2006 (0.1 m2 van Veen grab) Boundary of overlapping stations 30 60 kilometres Fig. 1. Location of benthic sampling stations of surveys of Onega Bay In 1952, 1981/90 and 2006 showing overlapping station boundaries. Marine Ecology32 (Suppl. 1) (2011) 36-48 © 2011 Blackwell Verlag GmbH 37 Macrobenthos of Onega Bay (White Sea, Russia) Solyanko, Spiridonov & Naumov Central Onega Bay (personal observations in July 2006 White Sea Biological Station of the Karelian-Finnish and June 2010). Onega Bay is the most species-rich area Branch of the Academy of Sciences of USSR (WSBS KFB) of the entire White Sea, with around 500 species of inver are deposited in the Archive of the Karelian Science Cen tebrates and a high benthic biomass (Golikov et a í 1985; tre of the Russian Academy of Sciences (KSC RAS) in Lukanin et a í 1995). The area may be regarded as being Petrozavodsk (Anonymous 1952a,b). They were digitised exposed to lower anthropogenic impacts than many other in Microsoft EXCEL format suitable for further use in Northeast Atlantic seas, as the industrial activity in the electronic databases. Material on Porifera, Hydrozoa, area has never been particularly high and has decreased Polychaeta, Pantopoda and Bryozoa from this survey was recently (Terzhevik et al. 2005). transferred to the Zoological Institute of the then Acad emy of Sciences of USSR (now Russian Academy of Sci ences) in Leningrad, now St. Petersburg (ZIN RAN) M ethods (Ivanova 1957). The fate of the material on other groups This study is based on the data from three benthic surveys remains unknown. conducted respectively in 1952, 1981/90 and 2006 (Fig. 1, Material from the 1981/90 surveys was identified Table 1). The data in the different surveys were collected mostly to species level, with the exception of Nemertini, using a Petersen (0.1 m2), a Petersen Ocean-50 (0.25 m2) Oligochaeta and some families of Porifera, Hydrozoa and and a Van Veen (0.1 m2) grab. Table 1 shows the dates of Bryozoa, which were identified by A. D. Naumov, V. V. the surveys, vessels, number of stations, samples at a sta Fedyakov and V. V. Lukanin in consultation with special tion, depth of sampling, and on-deck processing protocol. ists at ZISP on some faunal groups. The data are main The 1952 survey data pooled with other material col tained in the information system ‘Benthos of the White lected in Onega Bay were described by Ivanova (1957) Sea’ implemented in CLIPPER 5.0 algorithmic language and Kudersky (1966), but our re-analysis of these data is (Naumov 2006). based on the original protocols of sample examination. Benthic collections from the 2006 survey were pro The 1952 survey was processed incompletely: Mollusca, cessed with methods and taxonomic resolution similar to Cirripedia, Brachiopoda, Echinodermata and other taxa the one used in 1981/90. Most of the identification was were identified to species level by S. S. Ivanova and L. A. done by A. Rogacheva and K. Solyanko in consultation Kudersky. Other groups were recorded as higher taxa and with other specialists. The material is stored in the Zoo the total abundance and biomass of Porifera, Hydrozoa, logical Museum of the Moscow University. Due to the Polychaeta, Pantopoda, Bryozoa, Tunicata and several unclear status of the taxon usually identified as Hiatella orders of Crustacea, i.e. Amphipoda, Cumacea and My arctica (L., 1867), namely, the possible presence of sidacea, were calculated. The original station data and the another, yet unidentified species of the genus (Naumov protocols for processing the benthic collections by the 2006), the bivalve was listed as Hiatella sp. for all surveys. Table 1. Basic data for surveys in Onega Bay used in the study. Survey Characteristics KFB ZIN RAS IO RAS Notes Dates 10 August - 10 26 September 1981 - 2 15 July -17 Stations 113 and 114 September 19S2 September 1981; 2 July 1990 July 2006 were sampled in 1990 Vessel Professor Mesyatsev Kartesh Professor Vladimir Kuznetsov Gear Petersen grab - 0.1-m2 Petersen grab Ocean van Veen grab - 0.1 -m2 sampling area SO - 0.25-m2 sampling area sampling area No. of stations 70 28 10 No. of casts per station 2 1 3-5 Total no. of grab samples 134 27 38 Finest mesh size in process 0.7S 1 1 of rinsing samples, mm Depth range, m 7-S3 5-70 6-36 Mean depth, m 26 24 19 KFB = Karelian-Finnish Branch of the Academy of Sciences of USSR; ZIN RAS = Zoological Institute of the Russian Academy of Sciences, St. Peters burg; IO RAS = Institute of Oceanology of the Russian Academy of Sciences. 38 Marine Ecology32 (Suppl. 1 ) (2011 ) 36-48 © 2011 Blackwell Verlag GmbH Solyanko, Spiridonov & Naumov Macrobenthos of Onega Bay (White Sea, Russia) To test for differences between surveys in the biomass Results of 11 biomass-predominant and common bivalve species in 1952 and 2006, univariate techniques were applied Biomass and abundance of macrobenthos such as ANO VA, the Mann-Whitney (7-test, median test Biomass distribution in the Northern Onega Bay showed and Kolmogorov-Smirnov test (Hammer et a í 2001). a consistent pattern in 1952 and 1981/90 (Fig. 2). This Species composition and biomass data for these areas consistency was also found in the eastern part of the bay were compared using multivariate techniques (Clarke & when the 1952 and 2006 data were compared. In the Warwick 2001). The one-way ANOSIM test (PRIMER northern periphery of the bay, and generally in the Wes v6) was used to determine the differences in species com tern and the Eastern Salma, the biomass was relatively position and biomass between the studied years (overlap low and this zone of low biomass extended to the coastal ping stations), using 17 species of bivalves and seven areas in the northwestern part (Fig. 2). The lowest bio species of echinoderms. Only species from the 1952 mass (5.5 g'tn-2) was recorded at Station 4 near River (mean values) and the 1981/90 surveys were compared. Zolotitsa in 1952. In the coastal zone of the Solovetsky ANOSIM is analogous to analysis of variance (ANOVA) Islands, biomass varied greatly; however, most stations in univariate statistics. The 1952 data were not compared with biomass exceeding 1000 g'm-2 were concentrated in to the 2006 data because of the small number of overlap this area. The highest biomass recorded was 9200 g'm-2 ping stations. In the ANOSIM procedure, the probability at Station 237 in 1981/90 south of Bolshoi Solovetsky o f a priori groupings of samples was estimated by Island. In the central part of Onega Bay and off the repeated permutations of data {i.e. repeated random rela Onega Peninsula coast the biomass was generally lower belling of samples in the matrix). Initially, a global R sta than around the islands (in most cases <500 g'm-2) but tistic was calculated to determine whether significant greater than in the north of the bay (Fig. 2). In general, differences exist between all groups (analogous to the glo the macrobenthic biomass in Onega Bay can be consid bal F test in ANOVA). If differences were significant at a ered significant: it exceeded 100 g'm-2 at more than 60% global level, then pairwise comparisons between sample of all stations. groups were conducted to test for differences between Among large taxonomic groups, bivalves made a major pairs. In global tests, the null hypothesis {i.e. ‘no differ contribution to total benthic biomass, constituting at least ence between groups’) was rejected at a significance level 40% of the biomass of each survey (Fig. 2). Horse mussel of P < 0.05. Modiolus modiolus and quahog Arctica islandica together Possible changes in the community structure in terms with barnacles Balanus crenatus and Verruca stroemia con of abundance and biomass were measured by the ABC stituted the greatest biomass within all surveys. Cirripeds (abundance/biomass comparison, statistics W) curves were the next most important contributors to the total method. This method was applied only for stations where benthic biomass (above 20%) in 1952 and 1981/90, fol the biomass and abundance had been recorded ade lowed by polychaetes. However, this was not the case in quately. The abundance-biomass comparison (ABC) curves were conducted using the PRIMER v.6.0 software package. Non-metric multidimensional scaling (nMDS) based Biomass, g-m on Bray-Curtis similarity was carried out using loga o 0-50 O 50-150 rithm-transformed biomass data (all replicates included). O 150-500 The data of different years were pooled into one dataset. Q 500-1000 Q 1000-2000 The list of taxa contained species from the 1952 survey: ( J2000-10.000 20 species of bivalves were included (other species of biv alves appearing in later surveys were pooled into group ‘Other bivalves’), 13 species of gastropods (plus ‘Other 1981 and I990#urvey# gastropods’ group), seven species of echinoderms (plus rtiivey ‘Other echinoderms’ group), three species of cirripeds and one species of brachiopod. The rest of the taxa were entered as higher taxonomic groups (Porifera, Cnidaria, Polychaeta, Amphipoda, Cumacea, Decapoda, Pantopoda, Bryozoa and Ascidiacea). Although the list of taxa did not include information about all species, the taxa identi fied to species level were the most important in terms of Fig. 2. Distribution of macrozoobenthos biomass (g-m 2) In Onega biomass. Bay In the years 1952, 1981/90 and 2006. Marine Ecology32 (Suppl. 1 ) (2011 ) 3 6 -4 8 © 2011 Blackwell Verlag GmbH 39 Macrobenthos of Onega Bay (White Sea, Russia) Solyanko, Spiridonov & Naumov the Southeastern Onega Bay in 2006, where the positions there was no significant difference between the 1952 sur of these two groups were reversed (Fig. 3). Sponges, hy- vey (taking all stations or overlapped stations) and droids, brachiopods, bryozoans and echinoderms contrib 1981/90 surveys. uted to similar fractions of the total macrobenthic biomass (averaging 3-9%) in 1952 and 1981/90 (Fig. 3). Frequency of occurrence and biomass of particular taxa Median biomass values in the surveys were in the range 114-151 g'm-2 and rather similar (Table 2). However, Most of the bivalve species which were listed as dominant comparison of biomass values (using a non-parametric and subdominant in benthic communities of Onega Bay Mann-Whitney U-test) showed statistically significant dif in 1952 (Ivanova 1957; Kudersky 1966) and the 1980s ferences between all stations in 1952 and in 1981/90 (Golikov et al. 1985; Lukanin et a l 1995; N aum ov 2001) (P < 0.05) and between the 1981/90 and the 2006 sta occurred with similar frequency in 1952 and 1981/90 tions (P < 0.01). This was due to some exceptionally high (Table 3). Furthermore, Heteranomia spp., Nicania values (>2000 g'm-2) in 1981/90 (several stations around montagui , Nuculana sp., Modiolus modiolus and M ytilus the Solovetsky Islands). No statistically significant differ edulis showed nearly the same values. Only Leionucula ences in biomass were found between the 1952 and 2006 bellotii, Clinocardium ciliatum , Macoma calcarea occurred data (Table 2). No significant difference was detected 1.6-2.1 times more frequently in 1981/90 compared to between the 1952 stations and the overlapping 1981/90 1952, whereas Thyasira gouldi was about five times more stations. common in this year (Table 3). Correlation between fre Benthic abundance varied considerably between surveys quencies of occurrence of the bivalve species listed in from 10 to 43,604 ind-m-2 (Table 2). Abundance in Table 3 (w ithout T. gouldi , which was the most dissimilar 1981/90 was notably higher (mean of 5182 ind-m-2) in this respect) in the 1952 and the 1981/90 surveys was compared to the 1952 survey (mean of 2029 ind-m-2) high and statistically significant (r = 0.73, P < 0.005, and 2006 survey (mean of 2407 ind-m-2). The bio n - 13). mass/abundances ratio (B/A), or a mean mass of a speci In the 2006 survey area some of the bivalve species men, was remarkably similar (Table 2) and was not were found at a higher frequency than at the overlapping significantly different between the 1952 and the 1981/90 stations in 1952 (Table 3). In contrast, Heteranomia spp. surveys. The B/A ratio of the 2006 survey was lower than was much rarer in 2006 than in 1952. Mytilus edulis and in the other two surveys (Mann-Whitney U-test, P < 0.01 Chlamys islandica were found only in 1952. Furthermore, for the 1952 and 2006 comparison and P < 0.05 for the both absolute biomass and the biomass shares of particu 1981/90 and 2006 comparison). However, there was no lar species in 1952 and 1981/90 were also similar in many significant difference in the B/A values for the 1952 sta cases (Tables 3 and 4). Non-parametric tests indicate sta tions overlapping with the 2006 survey (Table 2). In tistically significant differences in absolute biomass only terms of abundance-biomass comparison (ABC curves) for Elliptica elliptica , Heteranomia squamula , M. calcarea A 100 1952 H 1981/1990 H2006 U M i i l U . Bivalvia Cirripedia Hydrozoa Polychaeta Echinodermata Gastropoda 40 11952 H 1981/1990 H2006 30- * 20 V c«VI Fig. 3. Percentage of contribution to the ¡ io total biomass (mean for stations + SD): for S taxonomic groups with a high relative o contribution (A) and other groups with lower Ascidia Brachiopoda Bryozoa Porifera Apyrae relative contribution (B). 40 Marine Ecology32 (Suppl. 1) (2011) 36-48 © 2011 Blackwell Verlag GmbH Solyanko, Spiridonov & Naumov Macrobenthos of Onega Bay (White Sea, Russia) Table 2. Comparison of the macrobenthic biomass and abundance. For mean biomass, mean abundance and mean biomass ratio, the standard deviation Is presented In brackets. 1952 - stations 1952 - stations In the area In the area 19S2 - all 1981/90 -all 2006 -a ll overlapping with overlapping with Parameters stations stations stations 1981/90 survey 2006 survey No. of stations 70 27 10 41 6 Benthic biomass (B) range, g-m-2 6 - 2188 11-9210 5-1195 2-2188 14-1706 Mean B, g-mT2 (SD) 273 (371) 959 (2008) 190 (254) 332 (408) 374 (504) Median B, g-mT2 1S1 142 114 188 68 Benthic abundance (A) range, Ind-m-2 10-22,310 60-43,604 250-19,020 10-22,310 230-8630 Mean A, Ind-m-2 (SD) 2029 (368S) B182 (9S28) 2407 (3540) 2452 (4182) 2928 (2956) Median A, Ind-m-2 S9S 1332 1170 890 1985 B/A, range, g-m-2 0.02-1.62 0.01-1.24 0.01-0.58 0.02-1.62 0.02-0.53 Mean B/A, g-m-2 (SD) 0.28 (0.32) 0.26 (0.3S) 0.1 (0.12) 0.26 (0.32) 0.13 (0.15) Median B/A, g-m-2 0.14 0.11 0.06 0.13 0.07 Table 3. Frequency of occurrence and mean contribution to the total biomass of common bivalves, cirripeds, echinoderms, gastropods, and brachlopods (In descending order of frequency of occurrence for the 1952 survey). 1952 - stations 1952 - stations 1981/90 - all overlapping with overlapping with 1952 - all stations stations 2006 - all stations 1981/90 survey 2006 survey Species FO ± SE BS ± SE FO ± SE BS ± SE FO ± SE BS ± SE FO ± SE BS ± SE FO ± SE BS ± SE Bivalvia Arctica islandica 18 ± 3 ■t* l+ 17 26 ± 8 30 ± 13 26 ± 7 62 ± 10 15 ± 74 38 ± 11 17 ± 11 35 Chlamys Islandica 10 ± 3 16 ± 6 22 ± 8 11 ± 5 15 ± 4 11 ± 4 17 ± 11 30 ± 3 Clinocardium ciliatum 16 ± 3 24 ± 6 52 ± 10 11 ± 4 24 ± 7 25 ± 6 22 ± 5 24 ± 6 8 1 Elliptica elliptica 43 ± 4 13 ± 2 52 ± 10 5 ± 3 37 ± 8 1 49 ± 6 12 ± 3 58 ± 14 14 ± 7 Heteranomia spp. 47 ± 4 4 ± 1 56 ± 10 1 ± 1 8 ± 4 1 ± 1 59 ± 5 4 ± 1 83 ± 11 9 ± 5 Hiatella sp. 27 ± 4 1 ± 1 56 ± 10 1 ± 1 13 ± 5 1 ± 1 33 ± 5 1 ± 1 33 ± 14 3 ± 1 Leionucula bellotii 18 ± 3 1 ± 1 52 ± 10 1 ± 1 58 ± 8 2 ± 1 20 ± 4 2 ± 1 33 ± 14 1 ± 1 Macoma calcarea 9 ± 2 8 ± 4 33 ± 9 7 ± 4 55 ± 8 20 ± 6 12 ± 4 5 ± 2 25 ± 13 1 ± 1 Modiolus modiolus 24 ± 4 46 ± 5 26 ± 8 46 ± 10 53 ± 8 4 ± 2 28 ± 5 51 ± 6 50 ± 14 35 ± 11 Mytilus edulis 8 ± 2 18 ± 4 15 ± 7 12 ± 6 6 ± 3 18 ± 3 8 1 Nicania montagui 31 ± 4 5 ± 2 44 ± 10 1 ± 1 50 ± 8 4 ± 2 30 ± 5 3 ± 1 8 1 Nuculana sp. 52 ± 4 6 ± 1 67 ± 9 2 ± 1 42 ± 8 9 ± 4 56 ± 5 3 ± 1 58 ± 14 6 ± 2 Thyasira gouldi 10 ± 3 2 ± 1 63 ± 9 1 ± 1 37 ± 8 1 ± 1 7 ± 3 1 ± 1 8 1 Cirripedia Balanus balanus 37 ± 4 22 ± 4 4 5 3 2 44 ± 5 21 ± 4 33 ± 14 10 ± 2 Balanus crenatus 13 ± 3 31 ± 6 56 ± 10 21 ± 6 37 ± 8 7 ± 3 17 ± 4 36 ± 7 8 6 Verruca stroemia 55 ± 4 15 ± 2 56 ± 10 8 ± 2 21 ± 7 19 ± 8 72 ± 5 15 ± 3 67 ± 14 19 ± 6 Echinodermata Henricia sp. 27 ± 4 1 ± 1 26 ± 8 1 ± 1 3 1 ± 1 Ophiopholis aculeata 4 ± 2 2 ± 1 22 ± 8 2 ± 2 5 ± 2 2 ± 1 Ophiura robusta 27 ± 4 2 ± 1 59 ± 9 1 ± 1 5 1 23 ± 5 1 ± 1 17 ± 11 1 ± 1 Stegophiura nodosa 12 ± 3 4 ± 3 26 ± 8 3 ± 2 18 ± 6 1 ± 1 13 ± 4 6 ± 5 Gastropoda Margarites g. groenlandicus 4 ± 2 1 ± 1 15 ± 7 1 ± 1 7 ± 3 2 ± 1 Puncturella noachina 4 ± 2 1 ± 1 7 ± 5 1 ± 1 3 1 6 ± 3 1 ± 1 8 1 Buccinum undatum 4 ± 2 6 ± 2 19 ± 7 1 ± 1 16 ± 6 9 ± 4 5 ± 2 6 ± 2 Brachiopoda Hemithiris psittacea 33 ± 4 8 ± 2 37 ± 9 10 ± 3 5 1 ± 1 43 ± 5 6 ± 1 58 ± 14 11 ± 5 FO = frequency of occurrence (%); BS = biomass share (%); SE = standard error. Marine Ecology32 (Suppl. 1 ) (2011 ) 3 6 -4 8 © 2011 Blackwell Verlag GmbH 41 Macrobenthos of Onega Bay (White Sea, Russia) Solyanko, Spiridonov & Naumov Table 4. Differences in biomass and statistical comparison of biomass data for dominant bivalve species at the stations in the overlapping area between the surveys in 19S2 and 1981/90 in Onega Bay. Biomass, g-m 2 1952 (n = 36) 1981/90 (n = 27) Kolmogorov-Smirnov Species Mean (SE) Mean (SE) Mann-Whitney (J-test (P) Test of median (P) test K (P) Species which may be dominant in the benthic communities (Kudersky 1966; Golikov eta/. 1985; Lukanin eta/. 1995) Arctica islandica 14.41 ± 7.98 39.34 ± 27.58 468.00 (0.56) 0.92 0.70 (0.70) Chlamys Islandica 25.54 ± 14.09 95.91 ± 70.33 461.50 (0.51) 0.54 0.55 (0.92) Clinocardium ciliatum 15.97 ± 4.48 12.62 ± 5.49 470.50 (0.66) 0.39 0.63 (0.82) Elliptica elliptica 30.08 ± 7.20 12.71 ± 7.13 242.50 (0.001)** 0.01* 1.91 (0.001)** Modiolus modiolus 203.48 ± 70.89 466.97 ± 253.78 453.00 (0.45) 0.46 0.62 (0.84) Nicania montagui 1.85 ± 0.55 1.43 ± 0.44 419.50 (0.25) 0.61 0.76 (0.60) Nuculana spp. 1.89 ± 0.83 1.43 ± 0.44 474.50 (0.73) 0.91 0.50 (0.97) Other common species Heteranomia squamula 13.87 ± 3.18 3.42 ± 1.48 233.50 (0.001)** 0.001** 1.97 (0.001)** Hiatella arctica 1.94 ± 0.52 4.93 ± 2.32 469.00 (0.86) 0.51 0.61 (0.85) Macoma calcarea 0.19 ± 0.15 2.29 ± 1.4 381.00 (0.02)* 0.06 0.92 (0.37) Tyasira gouldi 0.06 ± 0.03 0.45 ± 0.14 251.50 (0.001)** 0.001** 1.95 (0.001)** *Different levels of statistical significance of differences. SE = standard error; P = probability of belonging to the same general set of variables. and T. gouldi (Table 4). Biomass data for these and other B. crenatus, was not common in 1981/90 or 2006. Ver common bivalves (17 species) for 49 overlapping stations ucca stroemia and B. crenatus occurred much more fre were also tested for differences using a one-way ANOSIM quently than B. balanus in the 1981/90 and 2006 survey test. No significant difference between the studied years areas (Table 3). was found (Table 5). The most common echinoderms in the 1952 and the The 2006 survey indicated a greater contribution (aver 1981/90 surveys were (in descending order) Ophiura aged to nearly 67%) of A. islandica. In 1952 the contribu robusta, Stegophiura nodosa, Ophiopholis aculeata and tion was lower but the species still made the greatest Henricia sp. The frequencies of occurrence and average contribution to total benthic biomass (Table 3). Nonethe contributions to biomass were higher in 1981/90 for all less, one should bear in mind that the 1952 and 2006 species, although S. nodosa showed higher occurrence in data allow little direct comparison due to the small num 1952 in the area which overlapped with the 1981/90 sur ber of widely scattered stations in the earlier survey versus vey. Again, the one-way ANOSIM test for seven species of much more closely set stations along the shoreward tran relatively common echinoderms for 34 overlapping sta sects in 2006. tions did not show a significant difference in biomass Amongst common cirripeds Verucca stroemia occurred between the studied years (Table 4). In the overlapping at a very similar rate and made similar contributions to area of the 2006 and 1952 surveys, O. aculeata did not biomass in 1952 and 1981/90, whereas Balanus balanus, occur in either year and other three species were not which occurred twice as frequently in 1952 compared to found in 1952 (Table 3). Table 5. Comparison of the biomasses of common bivalves (17 species) and echinoderms (seven species) for all overlapping stations between the 1952 and the 1981/90 surveys using a one-way ANOSIM test. Taxa Species No. of species No. of stations Global R P-value Bivalvia Arctica islandica ; Chlamys islandicus ; Clinocardium ciliatum ; 17 49 0.039 0.08 Elliptica elliptica; Heteranomia squamula ; Hiatella arctica; Leionucula bellotii ; Macoma calcarea; Modiolus modiolus ; Mya truncata ; Mytilus edulis ; Nicania montagui ; Nuculana minuta ; Nuculana pernula ; Pandora glacialis ; Serripes groenlandicus ; Thyasira gouldi Echinodermata Asterias rubens ; Henricia sanguinolenta ; Ophiacantha 7 34 0.08 0.06 bidentata ; Ophiopholis aculeata ; Ophiura robusta ; Stegophiura nodosa; Strongylocentrotus pallidus 42 Marine Ecology32 (Suppl. 1 ) (2011 ) 36-48 © 2011 Blackwell Verlag GmbH Solyanko, Spiridonov & Naumov Macrobenthos of Onega Bay (White Sea, Russia) Gastropods were not commonly found in the 1952 sur Stations 16-19 of the 2006 survey. However, the biomass vey; only three species occurred with a frequency above o f Modiolus modiolus in 1952 was somewhat higher than 5%: Margarites groenlandicus groenlandicus, Puncturella in 2006 (Fig. 5, sub-area F). In Southeastern Onega Bay noachina and Buccinum undatum. In 1981/90 these three (Fig. 5, sub-area G) the community was also dominated were also the most frequently occurring species, with by Arctica islandica in both 1952 and 2006. P. noachina occurring with a similar rate, whereas in 1952 the two other species were found more frequently Discussion (Table 3). Hemithyris psittacea, the only brachiopod species living The benthic surveys considered in the present study were in the White Sea, showed a very similar occurrence rate not designed to study inter-annual variation in benthic and average contribution to total benthic biomass in 1952 communities. When planning the 2006 survey the stations and 1981/90. In 2006 the species was not as common as were intentionally set in the area which was covered the in the 1952 survey area overlapping with the 2006 survey least by the surveys in earlier years. Furthermore, the (Table 3). methods of sampling and gears differed between surveys. Bearing this in mind, we expected to find greater differ ences between the surveys from three different decades. Comparison at the assemblage level Median benthic biomass was very similar in all years of Results of the MDS analysis based on the biomass data investigation and clearly different from other areas of the for all taxa accounted for in the 1952 survey indicate con White Sea with similar depth and bottom topography. In siderable mixing of the samples of all surveys: variation particular, in the Gorlo (the shallow strait separating the between samples of the 1981/90 and the 2006 surveys is outer part of the White Sea from its deep basin) and in largely inside the variation of the 1952 survey performed the Dvina Bay the median biomass was one order of mag at a wider spatial scale (Fig. 4). A pairwise ANOSIM test nitude lower (Naumov 2001). revealed no statistically significant differences between the Neither the biomass of dominant bivalves and cirri 1952 and the 1981/90 data. However, differences at a sta peds nor the frequency of occurrence of the most com tistically significant level (P < 0.05) were found between mon species showed any considerable changes between these surveys and the 2006 survey. 1952 and 1981/90. The contribution to the total bio To compare the communities at a smaller scale, the mass of some bivalves and cirripeds ( Modiolus modiolus, study area was divided into sub-areas (Fig. 5). Compari Arctica islandica, Chlamys islandica, Mytilus edulis, Ellip son of the dominant pattern in particular sub-areas tica elliptica, Balanus balanus, Verucca stroemia and, to between the stations of the 1952 survey and the 1981/91 lesser extent, Clinocardium ciliatum) did not change survey also did not indicate major shifts (Table 6). Sub- between 1952 and 1981/90. All these species were area F covered the northern stations of the 2006 survey. described as dominant in various benthic communities These stations were located near Station 65 of the 1952 identified using different methods in the 1950s and the survey, which had a similar species composition with 1980s (Kudersky 1966; Golikov et a í 1985; Lukanin et al. 1995). Nuculana pernula and Nuculana minuta may be added to this list but it is possible that these morphologically similar species were poorly distinguished S17 8my CWtu m nùm ùy (Naumov 2006) in earlier surveys and so their presence cannot be confirmed with certainty. Furthermore, the frequency of occurrence and biomass of other common species ( Heteranomia squamula, Hiatella sp., Nicania montagui, Hemithyris psittacea and common echino derms) did not show much variation. At the assemblage * level, few differences were revealed using multivariate A +J statistics and direct comparison of the closely located AA stations from different surveys. The stability of the ABC curves and the average mass of a specimen also indicate the absence of shifts in benthic communities similar to those observed in some areas under the influence of Fig. 4. Results of multidimensional scaling (MDS) analysis of the sam eutrophication (Rosenberg 1987). Although the results of ples of surveys In Onega Bay conducted In 19S2, 1981/90 and 2006. the 2006 survey appear somewhat different from the Explanations are given In the text. patterns of previous years this does not indicate major Marine Ecology 32 (Suppl. 1) (2011) 3 6 -4 8 © 2011 Blackwell Verlag GmbH 43 Macrobenthos of Onega Bay (White Sea, Russia) Solyanko, Spiridonov & Naumov i l'A2 survey • 1 'IS 1 1 '»'in su rveys Fig. 5. Sub-areas of detailed comparison at ♦ 2()(in survey the assemblage level. Explanations are given In the text. Table 6. Benthic taxa composition (In terms of biomass) at stations of the 19S2 and the 1981/90 surveys performed for sub-areas In Onega Bay. Subarea No. of Survey (see Flg. S) samples Dominant taxa One-way ANOSIM test resul 19S2 A 9 Hemithyris psittacea, Cirripedia, Nuculana minuta, Elliptica elliptica R = 0.176; P > 0.0S and other bivalves 1981/90 7 Hemithyris psittacea, Cirripedia, Elliptica elliptica and other bivalves No significant differences 19S2 B 7 Modiolus modiolus, Cirripedia, Heteranomia squamula, Hydrozoa R = 0.38S; P > 0.0S 1981/90 3 Modiolus modiolus, Cirripedia, Heteranomia squamula, Ascidia No significant differences 19S2 C 10 Clinocardium ciliatum, Elliptica elliptica, Cirripedia, Ascidia R = 0.634; P < 0.0S 1981/90 3 Arctica islandica, Clinocardium ciliatum, Serripes groenlandicus, Significantly different Ascidia, Cirripedia 19S2 D 7 Cirripedia, Modiolus modiolus, Chlamys islandicus, Hiatella sp., R = 0.138; P > 0.0S Bryozoa, Hydrozoa 1981/90 4 Cirripedia, Hemithyris psittacea, Modiolus modiolus, Chlamys islandicus, No significant differences Hiatella sp., Bryozoa, Hydrozoa 19S2 E 8 Modiolus modiolus, Cirripedia, Ascidia, Hiatella sp., Hemithyris psittacea R = 0.323; P < 0.0S 1981/90 4 Ascidia, Cirripedia, Hemithyris psittacea, Nuculana minuta and other bivalves Significantly different 19S2 G 9 Arctica islandica, Nicania montagui, Nuculana spp., Elliptica elliptica ANOSIM test failed 1981/90 2 Arctica islandica, Elliptica elliptica, Clinocardium ciliatum, Nuculana minuta changes in the benthic communities because the survey tury (Stephen 1938; Parsons et a í 1977; Lukanin et a í in 2006 was designed in a different way and its overlap 1989); these are not necessarily related to environmental with the 1952 survey was minimal. However, even in variation (Naumov 2006). In the White Sea, a patch of this case, the dominant species A. islandica, M. modiolus benthic assemblages with a strong dominance of A . islan and V. stroemia held their positions. dica (population density of about 15,000 ind-m-2) has In spite of a similarity overall, there are apparent dif been monitored in Chupa Inlet for more than 25 years. ferences between the surveys which need to be discussed. The structure and quantitative characteristics of this clam First, the maximum and the average biomass of keystone population at depths >10 m remained stable for 23 years, species such as M. modiolus, C. islandica and A. islandica before the fraction of large (30-40 mm) specimens were higher in 1981/90 than in 1952. This may reflect declined owing to a drastic natural elimination. In subse population dynamics related to cohort growth and turn quent years, restoration of the population structure was over. In clams and mussels, long-term population cycles observed, probably as a result of the re-distribution of the have been known since the second half of the 20th cen clams (Guerassimova et a í 2008). 44 Marine Ecology32 (Suppl. 1 ) (2011 ) 36-48 © 2011 Blackwell Verlag GmbH Solyanko, Spiridonov & Naumov Macrobenthos of Onega Bay (White Sea, Russia) A lower biomass of dominant bivalves in 1952 may shore part of the White Sea (Babkov 1998; Howland et a í also be explained by the use of different sampling meth 1999), shows neither strong positive nor negative anoma ods. The Ocean-50 grab used in 1981/90 has a slightly lies since the late 1950s. The average temperature for the larger sampling area (0.25 m2) than two casts of a Peter 50-65 m layer indicates particularly little inter-annual sen grab (0.2 m2). It is possible that large sessile species variation; the anomalies do not significantly exceed with aggregated distribution were underestimated by tak 0.5 °C and show a weak correlation with the anomalies in ing two replicate samples of smaller size versus the one of the upper 15-m layer (Berger et al. 2003). As Onega Bay larger size. Furthermore, an Ocean-50 grab is much hea is open to the influence of the deep part of the White vier than a Petersen grab because their mass is propor Sea, owing to tidal wave propagation and an anti-clock- tional to L3, where L is a linear dimension of the open wise system of permanent currents (Babkov 1998; Filatov grab. A heavy grab is probably more effective in penetrat et al. 2005a), the pattern of inter-annual variation of ther ing the dense coverage of large bivalves than a lighter mal regime is not expected to be very different from that one. Further studies using both theoretical models and in the entrance of Chupa Inlet. River discharge, which field experiments are needed to check these hypotheses. can potentially strongly affect benthic communities in the Some species which were not dominant in their bio coastal zone, also shows no well expressed trends or mass but are relatively common in Onega Bay showed an major changes (Filatov et a í 2005b). apparent increase in the frequency of occurrence and bio Modelling of yearly average primary production based mass between 1952 and 1981/90. These species include on satellite chlorophyll data indicates that Onega Bay is small clam Thyasira gouldi, echinoderms Ophiura robusta one of the most productive areas in the White Sea (Ro- and Ophiopholis aculeata, and whelk Buccinum undatum. mankevich & Vetrov 2004). A considerable part of the In 2006, T. gouldi was also common and B. undatum phytoplankton production and allochthonous organic occurred with much higher frequency than found in pre matter supplied by river run-off is consumed by seston- vious surveys. In this case the differences in sampling feeding bivalves (Aí. modiolus, A. islandica, C. islandica, methodology may also have biased estimates for these and M . edulis) and cirripeds, which constitute the major species. For example, one may suppose that such mobile ity of the biomass in Onega Bay. These bivalve species are and probably aggregating species such as ophiuran and long-living (Naumov 2006) and have few consumers - whelks are underestimated by taking only two replicate mostly flatfish, which do not predate on older age groups samples of the Petersen grab. However, it is questionable o f M. modiolus, large clams and scallops (Ivanova 1957), whether this explanation also holds for T. gouldi. A lterna and eiders, which are highly abundant in the area. Eiders tive explanations would be trends for extension and/or use the area for breeding, moulting and wintering in the increasing abundance in the aforementioned species. polynyas but mostly concentrate for feeding close to the Regardless of whether these changes or trends are real or shore, in particular on blue mussels, M . edulis (Bianki artefacts of sampling design, they are not essential in 1991; Galaktionov 2001; Makarevich & Krasnov 2005). It comparison with the apparent absence of shift in the is therefore unlikely that predators have a strong impact dominance pattern in benthic communities and the rela on the population dynamics of dominant sessile benthic tive stability of biomass characteristics of most common species at the scale of the entire Onega Bay. Due to their species at a decadal scale. Taking into account high spa role in filtration of organic particles, influencing near-bot tial variability and methodological constraints of surveys, tom hydrodynamics and producing shell material as sub we may also speculate that such shifts could be potentially strate for epibenthos (Naumov 2006), the dominant overlooked. However, the consistency of the dominance bivalves and cirripeds may be considered keystone spe- pattern in benthic assemblages in small sub-areas (Fig. 5) cies-modifiers (Mills et al. 1993) in seabed biotopes. Thus over decades suggests that this is not the case. Indeed, in the stability or quasi-periodic changes in their popula the dynamics of the environmental conditions in the tions contribute to the relative stability of the subtidal White Sea region we see hardly any major changes that macrobenthic communities in Onega Bay. could drive shifts in the dominance pattern in benthic The characterisation of Onega Bay would be incom communities. plete without mentioning that the anthropogenic influ The period from the early 1940s until the first half of ence on its marine ecosystem was low to moderate in the the 1980s was characterised by general cooling, but from 20th century. Although the trend for eutrophication of the mid-1980s onwards, temperatures increased (Tolsti- the marine waters was seen in the White Sea in the 1980s kov et al. 2004; Filatov et a í 2005b). For water tempera compared to the 1950s (Maksimova 1991), the White Sea ture, the data from a permanent station at the entrance watershed area was not an area of intensive agriculture of the low-shore fjord in the Kandalaksha Bay, Chupa and pulp production in the second half of the 20th cen Inlet, which has unrestricted water exchange with the off tury (Terzhevik et al. 2005), and natural organic matter Marine Ecology32 (Suppl. 1 ) (2011 ) 3 6 -4 8 © 2011 Blackwell Verlag GmbH 45 Macrobenthos of Onega Bay (White Sea, Russia) Solyanko, Spiridonov & Naumov input from river run-off was always considerable (Ro- modified by human impact. The specific oceanographical mankevich & Vetrov 2004). Although there is pollutant regime of the White Sea has possibly also made it resilient transport with river run-off, much of the pollution is to climate variation in the past decades. Taking this into entrapped by so-called marginal filters in estuaries (Iva account, Onega Bay with its largely boreal and in several nov & Brizgalo 2005). In Onega Bay, background pollu respects similar characteristics to the North Sea and the tion with hydrocarbons and organochlorides is low; the Western Baltic biota (Zenkevich 1963; Naumov 2001) is a trace metal concentrations in bivalve tissues may be prospective area for studies of natural variation in benthic somewhat higher than in the neighbouring Kandalaksha communities and possible future climate-forcing in this Bay but they are still not high compared with seas sur yet undisturbed ecosystem belonging to the Northeast rounded by areas of high population and industrial den Atlantic realm. sity (Savinov et al. 2001). Onega Bay has always been an important area for herring and navaga fishing for local Acknowledgements and regional markets, but fishermen mostly used passive gears and there was practically no impact of bottom We thank Captain Jan Stelmakh and the crews of R.V. trawling and dredging on seabed habitats. Apart from Kartesh and Professor Vladimir Kuznetsov in 1981/90 and pink salmon, which was introduced in the late 1950s- 2006. We are very appreciative of the support of the 1960s, there are no alien species established in the region director of the Institute of Biology of KSC RAS, Prof. (Berger 2001). Nina N. Nemova, in giving us access to the archive mate It is of interest to compare the presumed stability of rial of the 1952 survey, and the help of Luydmila B. Cal benthic communities in Onega Bay to examples known anina and the staff of the Archive of the KSC RAS, from the other shelf areas of similar scale. In the neigh Petrozavodsk in working with the archive material. Valu bouring Southwestern Barents Sea the response of zoo- able critics and comments of the anonymous referees benthos to long-term fluctuations of temperature and the helped to improve the manuscript and made our state inflow of Atlantic waters is relatively rapid and manifests ments clearer. The field study in 2006 was supported by itself in changes of occurrence of the arctic and the boreal the INTAS project 51-5458 (Population and trophic func species (Galkin 1998; Denisenko 2008). However, the tion in shrimp species Crangon allmanni), and the visits principal factor influencing variation of quantitative char to the Archive of KSC RAS became possible due to the acteristics of infaunal benthic communities has been the travel funds provided by the Lighthouse Foundation bottom trawl fishery (Denisenko 2001; Carroll et a í (Hamburg and Kiel) and WWF Russia to V. Spiridonov. 2009). In the 20th century, Skagerrak, Kattegat, the North This study is part of the Russian Foundation for Basic and the Irish Seas show examples of significant changes Research Project 10-05-00813a. in benthic (mostly infaunal) communities that are likely caused by eutrophication (Pearson et al. 1985; Rosenberg R eferences et al. 1987), bottom trawling, background pollution (Krönke 1990; Wieking & Krönke 2003; Türkay & Krönke Anonymous (1952a) Log Book of on Deck Works in the 1st 2004) and scallop dredging (Bradshaw et a í 2002). Begin Cruise of R.V. ‘Professor Mesyatsev’, 17.08-11.09.1952. ning from the 1980s an increasing impact of climatic Archive of the Karelian Science Centre of the Russian Acad trends on benthic communities of the southern part of emy of Sciences, Petrozavodsk (In Russian). the North Sea can be traced (Beukema & Dekker 2003; Anonymous (1952b) Protocols of Processing Benthic Samples Sonnewald 2008). The studies on the Black Sea benthos Collected in the 1st Cruise of R.V. ‘Professor Mesyatsev’, indicated that changes in the structure of bottom com 17.08-11.09.1952. Archive of the Karelian Science Centre munities manifesting in the change of dominant spe of the Russian Academy of Sciences, Petrozavodsk (In Russian). cies and the high magnitude of variation in abundance Babkov A.V. (1998). Hydrology of the White Sea. Belomorskaya and biomass of common species may happen within a biostantsiya, St. Petersburg: 94 pp (In Russian). few years under the cumulative influence of the conse Beklemishev K.V., Pantyulin A.N., Semenova N.L. ( 1980) Bio quences of eutrophication and introduction of alien logical composition of the White Sea. II. New findings rela species. The misbalanced benthic communities are contin ting to vertical zoning of the White Sea. Transactions of the uing to experience rapid changes in their quantitative spe White Sea Biological Station of the Moscow University, 5, 20- cies composition (Chikina & Kucheruk 2005; Kucheruk 28 pp (In Russian). et al. 2009). Berger V.Y. (2001) Fishes. In: Berger V., Dahle S. (Eds), White Amongst the seas around Europe, the White Sea and Sea. Ecology and Environment. Derzhavets Publisher, St. particularly Onega Bay may represent a rare case of a Petersburg and Trornso: 55-76. shallow-water benthic ecosystem which has not yet been 46 Marine Ecology32 (Suppl. 1 ) (2011 ) 36-48 © 2011 Blackwell Verlag GmbH Solyanko, Spiridonov & Naumov Macrobenthos of Onega Bay (White Sea, Russia) Berger V.Y., Naumov A.D. (2001) General features. In: Berger Global Change. Springer-Praxis Publishing, Chichester: V., Dahle S. (Eds), White Sea. Ecology and Environment. 73-154. Derzhavets Publisher, St. Petersburg and Tromso: 9-22. Filatov N.N., Nazarova L.E., Salo Ju A., Tolstikov A.V. (2005b) Berger V.J., Naumov A.D., Usov N., Zubaha V., Smolyar M.A., Climate of the White Sea catchment scenarios of climate Tatusko R., Levitus S. (2003).36-Years Time-Series (1963- and river runoff changes. In: Filatov N.N., Pozdnyakov 1998) of Zooplankton, Temperature and Salinity in the White D.V., Johannessen O.M. (Eds), White Sea. Its M arine Envi Sea. NOAA, St. Petersburg: 362 pp. ronment and Ecosystem Dynamics Influenced by Global Beukema J.J., Dekker R. (2003) Redistribution o f spat-sized Change. Springer-Praxis Publishing, Chichester: 53-72. Macoma balthica in the Wadden Sea in cold and mild win Fokin S.I., Smirnov A.V., Layus Yu.A. (2006).Marine Biologi ters. Marine Ecology Progress Series, 265, 117-122. cal Stations in the Russian North (1881-1938). KMK, M os Bianki V.V. (1991) The birds. Oceanographic Conditions and cow: 130 pp. Biological Productivity of the White Sea. Annotated atlas. Frid C.L.J., Clark R.A., Haii J.A. ( 1999) Long-term changes in PINRO Publication, Murmansk: 191-201. the benthos on a heavily fished ground off the NE coast of Bradshaw C., Veale L.O., Brand A.R. (2005) The role of scal- England. Marine Ecology Progress Series, 188, 13-20. lop-dredge disturbance in long-term changes in Irish Sea Frid C.L.J., Garwood P.R., Robinson L.A. (2009) The North benthic communities: a re-analysis of an historical dataset. Sea benthic system: a 36 year time-series. Journal of the Journal of Sea Research, 47, 161-184. Marine Biological Association of the United Kingdom, 89, Brown E.J., Finney B., Hills S., Dommisse M. (2005) Effects of 1- 10. commercial otter trawling on benthic communities in the Galaktionov K.V. (2001) Marine and coastal birds. In: Berger south-eastern Bering Sea. In: Barnes B.W., Thomas J.P. V., Dahle S. (Eds), W hite Sea. Ecology and Environment. (Eds), Symposium on Effects of Fishing Activities on Benthic Derzhavets Publisher, St. Petersburg and Tromso: 77-88. Habitats. American Fisheries Society, Bethesda, Maryland: Galkin Y.I. ( 1998) Long term changes in the distribution of 439-460. molluscs in the Barents Sea related to the climate. Berichte Carroll M.L., Denisenko S.G., Renaud P.E., Ambrose W.G. zur Polaiforschung, 287, 100-143. (2008) Benthic infauna of the seasonally ice-covered western Golikov A.N., Babkov A.I., Golikov A.A., Novikov O.K., Shere- Barents Sea: patterns and relationships to environmental metevskiy A.M. ( 1985) Ecosystems of the Onega Bay and forcing. Deep-Sea Research Part I, 55, 2340-2351. neighbouring areas of the basin of the White Sea. In: Scarla Carroll M.L., Johnson B.J., Henkes G.A., McMahon K.W., ta O.A. (Ed.), Ecosystems of the Onega Bay of the White Sea. Voronkov A., Ambrose W.C., Denisenko S.G. (2009) Explorations of the Fauna of Seas. Zoological Institute of the Bivalves as indicators of environmental variation and USSR, Leningrad: 20-87 (In Russian). potential anthropogenic impacts in the southern Barents Guerassimova A.V., Kuznetsova E.K., Maximovich N.V. (2008) Sea. Marine Pollution Bulletin, 59, 193-206. On multi-annual dynamics of a local population of Arctica Chikina M.V., Kucheruk N.V. (2005) Long-term changes in islandica (Mollusca, Bivalvia) and peculiarities of macroben the structure of coastal benthic communities in the north thos distribution in the area of Keretsky Archipelago (White eastern part o f the Black Sea: influence o f alien species. Sea). Presentations of the Scientific Conference dedicated to Oceanology, 45, 176-182. the 70 years anniversary of the White Sea Biological Station of Clarke K.R., Warwick R.M. (2001).Change in Marine Commu the Moscow University. Grif 8c Co. Publishers, Moscow: 55- nities: An Approach to Statistical Analysis and Interpretation, 58 (In Russian). 2nd edn. Primer-E, Plym outh, UK: 176 pp. Hammer 0., Harper D.A.T., Ryan P.D. (2001) Paleontological Denisenko S.G. (2001) Long-term changes of zoobenthos bio statistical software package for education and analysis. Pale mass in the Barents Sea. Zoological sessions (Annual reports ontología Electrónica, 4, 1-9. 2000). Proceedings of the Zoological Institute RAS: 59-66 (In Howland R.J.M., Pantyulin A.N., Millward G.E., Prego R. Russian). (1999) The hydrography of the Chupa Estuary, White Sea, Denisenko S.G. (2008) Macrozoobenthos of the Barents Sea in Russia. Estuarine Coastal and Shelf Science, 48, 1-12. the conditions of changing climate and anthropogenic Ivanov V.V., Brizgalo V.A. (2005) White Sea watershed hydrol impact. Dissertation Dr. Biol. Sei. Zoological Institute of ogy and anthropogenic impact. In: Filatov N.N., Pozdnya Russian Academy of Sciences, St. Petersburg (In Russian). kov D.V., Johannessen O.M. (Eds), W hite Sea. Its Marine Derjugin K.M. (1928) Fauna des Weissen Meeres und ihre Environment and Ecosystem Dynamics Influenced by Global Existenzbedignungen. Exploration des mers d’URSS, 7-8, Change. Springer-Praxis Publishing, Chichester: 15-47. 1-511 (In Russian with extended German summary). Ivanova S.S. (1957) Qualitative and Quantitative Characteristics Filatov N.N., Pozdnyakov D.V., Ingebeikin Yu.I., Zdorovenov of Benthos in Onega Bay of the White Sea. Materials on com R.E., Melentyev V.V., Tolstikov A.V., Pettersson L.H. plex studies of the White Sea (Moscow): 355-380 (In Russian). (2005a) Oceanographical regime. In: Filatov N.N., Pozdnya Krönke I. ( 1990) Macrofauna standing stock of the Dogger kov D.V., Johannessen O.M. (Eds), White Sea. Its Marine Bank. A comparison: II. 1951-1952 versus 1985-1987. Are Environment and Ecosystem Dynamics Influenced by changes in the community of the north-eastern part of the Marine Ecology32 (Suppl. 1 ) (2011 ) 3 6 -4 8 © 2011 Blackwell Verlag GmbH 47 Macrobenthos of Onega Bay (White Sea, Russia) Solyanko, Spiridonov & Naumov Dogger Bank due to environmental changes? Netherlands Pearson T.H., Josefson A.B., Rosenberg R. (1985) Petersen Journal of Sea Research, 25, 189-198. benthic stations revisited. 1. Is the Kattegatt becoming Krönke I., Stoeck T., Wieking G., Palojarvi A. (2004) Relation eutrophic? Journal of Experimental Marine Biology and Ecol ship between structural and functional aspects of microbial ogy, 92, 157-206. and macrofaunal communities in different areas of the Romankevich E.A., Vetrov A.A. (2004). Carbon Cycle in Rus N orth Sea. Marine Ecology Progress Series, 282, 13-31. sian Arctic Seas. Springer, Berlin: 331 pp. Kucheruk N.V., Flint M.V., Maksimova O.V., Chikina M.V., Rosenberg R., Gray J.S., Josefson A.B., Pearson T.H. (1987) Siniakova U.V. (2009) Contemporary Dynamics of Benthic Petersen benthic stations revisited. 2. Is the Oslofjord and Communities in the North-East Shelf of the Black Sea Natural eastern Skagerrak enriched? Journal of Experimental Marine and Social Economic Factors of Environment Changes in Rus Biology and Ecology, 105, 219-251. sia. Nauka, Moscow: 1-22 (In Russian). Savinov V., Savinova T., Dahle S. (2001) Contaminants. In: Kudersky L.A. (1966). Bottom Fauna of the Onega Bay of the Berger V., Dahle S. (Eds), White Sea. Ecology and Environ W hite Sea. Karelian Book Publisher, Petrozavodsk: 204-371 ment. Derzhavets Publisher, St. Petersburg and Tromso: (In Russian). 120-136. Lukanin V.V., Naumov A.D., Fedyakov V.V. ( 1989) Multiyear Sonnewald M. (2008) Langzeituntersuchung zur Biodiversität structural changes in an estuarine population of blue mussel der decapoden Crustaceen und deren Begleitfauna in der De in the W hite Sea. Journal of General Biology, 50, 366-371 (In utschen Bucht: 2007 versus 1986, 1987 und 1990. D iplom ar Russian). beit im Fachbereich Biologie Johann Wolfgang Goethe, Lukanin V.V., Naumov A.D., Fedyakov V.V. ( 1995) Peculiari Universität, Frankfurt am Main. ties of benthos distribution in the Onega Bay. In: Berger Stephen A.C. (1938) Production of Large Broods in Certain N.J. (Eds), White Sea. Biological Resources and Problems of Marine Lamellibranchs with a Possible Relation to Weather Their Rational Exploitations. St. Petersburg: 232-236 (In Conditions. Journal o f Anim al Ecology, 7, 130-143. Russian). Terzhevik A.Y., Litvinenko A.V., Druzhinin P.V., Filatov N.N. Makarevich P.R., Krasnov J.V. (2005) Aquatic ecosystem pro (2005) Economy of the White Sea watershed. In: Filatov file. In: Filatov N.N., Pozdnyakov D.V., Johannessen O.M. N.N., Pozdnyakov D.V., Johannessen O.M. (Eds), W hite Sea. (Eds), White Sea. Its Marine Environment and Ecosystem Its Marine Environment and Ecosystem Dynamics Influenced Dynamics Influenced by Global Change. Springer-Praxis Pub by Global Change. Springer-Praxis Publishing, Chichester: lishing, Chichester: 155-177. 241-301. Maksimova M.P. (1991) Hydrochemistry of the White Sea. Tolstikov A.V., Petrov M.P., Filatov N.N. (2004) Influence of Hydrometeorology and Hydrochemistiy of Seas of the USSR. climate change on oceanographic characteristics. In: Filatov Hydrometeoizdat, Leningrad: 8-112 (In Russian). N.N. (Ed.), The Climate of Karelia: Variability and Influence Mills L.S., Soulé M.E., Doak D.F. (1993) The Keystone-Species on Water Objects and Catchment Areas. Karelian Research Concept in Ecology and Conservation. BioScience, 43, 219- Centre, RAS, Petrozavodsk: 130-134. 224. Türkay M., Krönke I. (2004) Eine Insel unter Wasser: die Naumov A.D. (2001) Benthos. Ch. 4. In: Berger V., Dahle S. Doggerbank. Natur und Museum, 134, 261-276. (Eds), W hite Sea. Ecology and Environment. Derzhavets, St. Wieking G., Krönke I. (2003) Macrofauna communities of the Petersburg and Tromso: 41-53. Dogger Bank (central North Sea) in the late 1990s: spatial Naumov A.D. (2006) Clams o f the White Sea. Ecological and distribution, species composition and trophic structure. Hel Fannistic Analysis. Explorations of the fauna of the seas. goland Marine Research, 57, 34-46. Zoological Institute of the Russian Academy of Sciences, St. Zenkevich L.A. (1963) Biology of the Seas of the USSR. Acad Petersburg: 1-367 (In Russian). emy of Sciences of USSR Publishing House, Moscow: 740 Parsons T.R., Takahashi M., Hargrave B. (1977). Biological pp (In Russian). Oceanographic Processes. Pergamon Press, Oxford: 323 pp. 48 Marine Ecology32 (Suppl. 1 ) (2011 ) 36-48 © 2011 Blackwell Verlag GmbH anmarine evolutionary perspective ecology ^ / Marine Ecology. ISSN 0173-9565 ORIGINAL ARTICLE The effect of temperature variability on ecological functioning of epifauna in the German Bight Hermann Neumann & Ingrid Kröncke Department for Marine Research, Senckenberg Institute, Wilhelmshaven, Germany Keywords A bstract Benthos; biological traits analysis; cold winter; ecosystem functioning; functional diversity; Benthic epifauna was sampled in an area of 10 X 10 nautical miles in the Ger North Sea; temperature anomalies. man Bight. Samples were taken in January and July/August from 1998 to 2009. The ecological functioning of the epifaunal community was assessed using bio Correspondence logical traits analysis (BTA). Twelve ecological traits of 26 epifaunal species were Hermann Neumann, Department for Marine considered and analysed using non-metric multidimensional scaling (nmMDS). Research, Senckenberg Institute, Südstrand Anomalies in the sea surface temperature (SST) close to the study area were 40, 26382 Wilhelmshaven, Germany. E-mail: [email protected] mainly above the long-term mean during the study period. SST was exception ally high during the autumn months between 2002 and 2006. Additionally, the Accepted: 22 November 2010 cold winter of 1995-96 was clearly reflected in strong negative SST anomalies. Trait composition changed in 2002, mainly due to a decreasing trend of traits doi: 10.1111/j. 1439-0485.2010.00420.x related to an opportunistic life mode from 1998-2002. Traits related to repro duction showed a much clearer response to the high autumn SST anomalies from 2002 to 2006 than other traits. We concluded that the cold winter resulted in an increase in opportunistic species in the study area followed by characteris tic post-disturbance succession stages to the point of an established community in 2002. This indicates a recovery time of epifaunal communities in the German Bight of 7-8 years. Additionally, the results give evidence that climate-induced variability of SST in the German Bight affects mainly the reproduction of epi faunal species rather than other traits such as feeding type. of warm-water species (Edwards & Richardson 2004; Introduction Edwards et al. 2008; Kirby et al. 2008; Martens & van In the North Sea ecosystem, a regime shift occurred in Beusekom 2008). Biogeographical shifts of fish species the late 1980s from a ‘cold dynamic equilibrium’ to a have been identified and interpreted as reflecting a ‘warm dynamic equilibrium’ (Beaugrand 2004). This shift response to increasing water temperature (Ehrich & was linked to pronounced modifications in large-scale Stransky 2001; Brander et al. 2003; Perry et al. 2005). hydro-metrological forcing and ecosystem parameters, Migration patterns of species have changed (Sims et al. including a marked increase in oceanic inflow and sea 2001). Benthic communities in the Southern and North surface temperature (Beaugrand 2004). The warm tem ern North Sea have been affected by temperature changes perature period has continued to the present day (Hughes (Kröncke et al. 1998; Neumann et al. 2009a,b). However, & Holliday 2006) and there is strong evidence to suggest the trend of increasing temperature was interrupted by that many different species and communities in the North extreme cold winter conditions in the North Sea region Sea ecosystem are responding to these temperature during 1995-1996. Cold winters influence benthic fauna changes. For example, the phenology of phyto- and Zoo greatly, through direct (enhanced mortality) and indirect plankton in the North Sea has changed and plankton (reduced reproduction and production) effects on the communities have shifted due to an increasing prevalence species, especially in shallow areas (Reiss et a í 2006; Marine Ecology 32 (Suppl. 1) (2011) 4 9-57 © 2010 Blackwell Verlag GmbH 49 Effect of temperature variability on epifauna N e u m a n n & K röncke N eum ann et al. 2008a,b, 2009a,b). These effects are showed an increase in diversity and secondary production observed as a reduced number of species, diversity and in conjunction with increasing sea surface temperature biomass (Ziegelmeier 1970; Buchanan & Moore 1986; (N eum ann et a í 2008a; Neumann et al. 2009b). The Beukema 1992; Kröncke et al. 1998). As the effects of cold objectives of our study were (i) to assess the seasonal and winters might influence the ecosystem for several years, it annual effects of SST variability on the functional compo is essential to understand them more precisely in order to sition of epifauna and (ii) to compare these results with interpret long-term dynamics in the North Sea ecosystem. the outcome of the species composition analysis in the Limited attention has been paid to the question of how same area. these climate-induced changes affect the functioning of the North Sea ecosystem despite a growing demand for Material and Methods the functional aspects of systems to be incorporated into conservation and management efforts (Bremner 2008; Study site Frid et al. 2008). According to Jax (2005), the term ‘func The area o f investigation (Box A; 10 X 10 nautical miles) tion’ in ecology refers to (i) processes and the causal rela was situated about 25 nautical miles northwest of the tions that give rise to them, (ii) the role of organisms Island of Helgoland, in close proximity to the old Elbe within an ecological system, (iii) the overall processes that glacial valley 30 m depth contour (54° 17' N-54°27/ N sustain an ecological system and finally (iv) to the services and 006°58/ E-007°15/ E) (Fig. 1). The m ean depth of a system provides for human or other organisms. Studies this area was 40 m and the water column was generally on species composition were often inadequate to address well mixed throughout the year. Sediments in the south these issues as ecosystem processes are determined by the west corner of Box A were more than 20% mud (<63 p m functional characteristics of the organisms involved, fraction). This percentage gradually decreased towards the rather than by taxonomic identity (Odum 1969; Grime northeast (0-5%). The time series started in 1998 and 1997; H ooper et a í 2002). The same conclusion was were part of the German small scale bottom trawl survey drawn by Diaz & Cabido (2001), who stated that ecosys (GSBTS) (Ehrich et al. 2007 for further information). tem stability is strongly attributed to the functional traits Epifauna was sampled twice a year in January (first quar of species and their interactions rather than to species ter) and in July/August (third quarter) on board the FRV composition per se. Walther Herwig. Sampling did not take place in winter Biological traits analysis (BTA), which was developed 1998 and 1999 due to ship time constraints. in terrestrial and freshwater ecology, is a useful analytical approach to describe different aspects of functioning (Brem ner et a í 2003b). BTA has been applied successfully Epifauna data to assess fishing effects on benthic fauna (Bremner et a í Epifauna was sampled with a standardized 2 m beam 2003a, 2005; Tillin et al. 2006), to assess the functional trawl made of galvanized steel with a chain matt attached. diversity in different species assemblages (Bell 2007; M ouillot et a í 2007; Schratzberger et al. 2007) as well as for management and conservation purposes (Bremner 2008; Frid et al. 2008). BTA uses a comprehensive set of functional traits (e.g. mobility, feeding type, size, longev ity, and reproductive technique), which can be used as indicators for ecosystem functioning. The wide range of traits used by BTA, the strong link between them and ecosystem processes (Diaz & Cabido 2001), as well as the sound theoretical framework (see Bremner 2008) are a considerable advance over traditional methods dealing with ecosystem functioning. To understand the impact of temperature variability on ecosystem functioning in the German Bight, this study focused on a single component of the ecosystem, the mobile epifauna. We applied BTA on an epifaunal time series in the German Bight covering a period of 12 years re re re with summer and winter sampling. Previous studies showed that the epifaunal communities in this area were Fig. 1. Location and depth of Box A and temperature station Helgo severely affected by the cold winter in 1995-96, but also land (Hel) In the North Sea. 50 Marine Ecology32 (Suppl. 1) (2011) 49-57 © 2010 Blackwell Verlag GmbH Neumann & Kröncke Effect of temperature variability on epifauna The beam trawl was fitted with a 20-mm net and a cod 7° 18.0' E) close to Box A. Monthly standardized tempera end of 4 mm mesh size. A Scanmar depth-finding sonar ture anomalies from 1998 to 2008 were calculated from was attached to the top of the net just behind the steel the Helgoland station based on the 1968-2008 mean. beam to determine the exact time and position of contact with the seabed. From the moment of contact with the Biological trait analysis (BTA) ground, the beam was towed at a speed of about 1.5-2 knots for 5 min. Altogether, 197 beam trawls were taken The epifaunal dataset was reduced to the 26 most domi between 1998 and 2009. In general, nine replicates were nant species in terms of abundance and occurrence in taken in each sampling season but the replicate number Box A, previously described by Neumann et al. (2008a,b). varied between 3 (winter 2000) and 13 (winter 2004) These species were coded in a ‘species by trait table’ according to weather conditions. On average, 8.9 repli (Fig. 2) to the extent they displayed the categories of 12 cates were taken. Samples were sieved through 5-mm traits. Coding was done using a ‘fuzzy coding’ approach mesh and the epibenthic fauna was separated from the which uses positive scores to describe the affinity of spe remains. Species were identified onboard to the lowest cies to trait categories (Chevenet et al. 1994). In this possible taxonomic level and abundance data were stan study, a scoring range from 0 (no affinity) to 4 (total dardized to a tow length of 250 m (area sam affinity) was applied. For example, the shrimp Crangon pled = 500 m 2). allmanni was coded 0 (permanent attached), 0 (tempo rally attached), 1 (Burrower), 2 (Crawler) and 1 (Swim mer) for the trait ‘adult mobility’. Ten of the traits used Temperature data (including the categories) were adopted from Tillin et a í The Federal Maritime and Hydrographic Agency of Ger (2006) reflecting a wide range of ecological function and many (BSH) provided weekly sea surface temperature life-history modalities of species. The traits ‘fertilization (SST) data at the Helgoland Station (Hel; 54°9.6/ N, type’ (internal, external) and ‘reproductive season’ Species by trait table Species by station table Mobility 1998 1999 Perm Temp. Burrower Craw|er Swimmer attach attach Aporrhais pespelecani 34 345 Aporrhais pespelecani 0 0 3 1 0 Asterias rubens 0 0 0 4 0 Asterias rubens 12 123 Trait Categories Perm, attached; Temp, attached; Burrower; Crawler; Mobility 1. Multiply category scores with Swimmer abundance data Habitat Infauna; Epifauna; Epizoic 2. Sum category scores over taxa Feeding type Deposit; Filter/suspens.; Grazer; Scavenger; Predator Algae; Invertebr.A/ertebr.; Carrion; Detritus; Plankton; Food type Suspend, org. matter; Microorg. small (1-2); small-medium (3-10); medium (11-20); Size (cm) medium-large (11-20); large (> 50) Station by trait table Adult longevity (yr) <2; 2-5; 5-10; 10+ 1998 1999 Age sexual maturity (yr) <2; 2-5; 5-10; 10+ Perm. 23 0 Asexual; Sexual (spawner); Sexual (egg lay/brood - attach Reprod. technique mini adults); Sexual (egg lay/brood - plankt. larvae) Annual once; Annual (2 or more); Biennial; Temp. 0 587 Reprod. frequency Semelparous attach Reprod. season Winter; Spring; Summer; Autumn Burrower 211 265 No pelagic life stage; Pelagic life stage; Low mobility Dissemination Crawler 2395 115 adult; Highly mobile adult; Migratory Swimmer 0 135 Fertilization type Internal; External Fig. 2. Stages of the biological trait analysis (BTA) including the trait variables with the corresponding number of categories used to describe eco logical functioning of the epifaunal community in Box A. Marine Ecology32 (Suppl. 1 ) (2011 ) 49-57 © 2010 Blackwell Verlag GmbH 51 Effect of temperature variability on epifauna N e u m a n n & K röncke 2 .5 - Fig. 3. Anomalies in SST (°) at the station Helgoland (Hel) close to Box A, based on the 1968-2008 mean. (autumn, winter, spring, summer) were added to focus and October in 2005 and 2006 (1.6 and 1.9). SST anoma on the reproduction of species, as temperature variability lies were exceptionally high at the first half of 2007 (Janu has a large impact on the reproduction cycles of species ary to June) ranging from 1.4 (June) to 2.1 (April), which and on benthic-pelagic coupling in the North Sea (e.g. (together with September 2002) is the highest recorded Edwards & Richardson 2004; Kirby et al. 2007, 2008). anomaly at that station in 1995-2008. Trait category scores for each species present in a year/season were then weighted by their abundance in Biological trait analysis (BTA) that year/season by multiplying the scores with abun dance data and then summing the resulting values over The nmMDS analysis based on fourth root-transformed all species (Fig. 2). The result is a ‘station by trait table’ trait data revealed distinct changes in the trait composi which contains the frequencies of occurrence of biological tion of the epifauna in Box A in 2002 (Fig. 4). The ANO traits in each year and season (Fig. 2). SIM randomization test confirmed significant differences It is important to mention that the BTA used here in trait composition between the years 1998-2002 (win describes only a single aspect of functioning as it does ter) and 2002 (summer) to 2009 (R = 0.625, P < 0.001). not include other components of the ecosystem, nor does The dissimilarity between these two periods was 21%. it quantify processes or properties. The most important trait categories contributing to this Non-metric multidimensional scaling (nmMDS) in the dissimilarity were categories which belonged to an oppor PRIMER v 6 package (Plymouth Marine Eaboratory) was tunistic life mode, such as small size, early onset of sexual applied to the station-by-trait table based on fourth root- maturity and a short life span. Thus, the relative abun transformed data. Similarities were calculated using the dance of small species ( 1 - 2 cm) which have an age of Bray-Curtis coefficient. This method describes the simi maturity <2 years and an adult longevity of 2-5 years larities between the years and seasons in terms of their decreased continuously in the first years and seasons of trait composition and is appropriate for providing a gen the study period (Fig. 4). Shifts in trait composition eral picture of functioning in marine assemblages (Brem around 2 0 0 2 were also evident in the traits: adult mobil ner et a í 2006). An ANOSIM randomization test was ity, feeding type, dissemination, as well as reproductive performed to test the differences in trait composition type, -frequency and -season (Fig. 5). For example, the between the years (H0: no differences in trait composi proportion of deposit and filter-/suspension feeders was tion). much higher before 2003 (52-80%) whereas grazers, scav engers and particularly predators were the dominant feed ing mode after 2003 (73-98%). Additionally, the Results abundance of migratory species was higher after 2003 SST temperature anomalies from 1995 to 2008 (11-21%) than in the period before 2003 (1-7%). Traits such as adult mobility, reproductive type, reproductive In general, the temperature anomalies were mainly above frequency and reproductive season not only shifted the long-term mean (1968-2008) at Helgoland Station between 2002 and 2003 but showed the highest percent from 1995 to 2008 (Fig. 3). However, the cold winter of ages or even a seasonality in the period of the warm 1995/96 was clearly reflected in strong negative SST autumn months (2002-2007). The proportion of sexual anomalies from the start of 1996, which persisted until egg layers with a planktonic larvae peaked in winter 2005 May 1997. Since 2002, the positive SST anomalies have (75%) and 2006 (75%) and decreased slightly afterwards. often persisted throughout the year, and were exception The percentages of species which had their reproductive ally high during the autumn months in 2002-2006. For season in autumn and reproduce twice (or more) per year example, the highest yearly anomalies were found in Sep showed a clear seasonality. The highest abundance of tember in 2002-2004 (2.1, 1.3 and 0.9) and in November 52 Marine Ecology32 (Suppl. 1) (2011) 49-57 © 2010 Blackwell Verlag GmbH Neumann & Kröncke Effect of temperature variability on epifauna Small size (1-2 cm) Age at sexual maturity <2 yrs Adult longevity 2-5 yrs Fig. 4. nmMDS plot of biological trait composition in Box A from 1998 to 2009 including summer and winter sampling (top left). nmMDS plot was overlaid with the relative occurrence of the trait categories 'small size (1-2 cm)' (top right), 'age at sexual maturity <2 years' (bottom left) and adult longevity 2-5 years (bottom right). Data were fourth root-transformed. Adult mobility Feeding type ■ Temp, attach. «Burrower «Crawler «Swimmer ■ Deposit feeder ■ Filter/susp. feeder «Grazer ¡Scavenger «Predator 100% 80% 80% 60% 60% 40% 40% 20% 20% 0% 0% Cq Cq ^ <0 <0 <0 ^ Cq ^ Cq ^ C o ^ Co ^ Co ^ Co ^ Co Cq Cq £.Gq £.CO'£.G q £.GQ'£.Gq £.OQ'£.C q £.Gq £.Cq £.Gq s ' ?/# /e /$ /'S-/ Reproductive type Dissemination ■ Asexual budding «Sexual spawner No pelagic life stage ■ Pelagic life stage Low mobility ■ Sexual egg layer-planktonic larvae «Sexual egg layer-mini adults High mobility ■ Migratory 100% 100% 80% 80% MIII 60% 40% 40% 20% 20% 0% I I 0% co Co ^ Co ^ co ^ co £ co ^ c q ^ co Co $ co o Cq Cq Cq ^ Cq Gq « 0 ^ Cq £ Cq Cq ^ Cq •'s'g's's's'& '& 's's'g'g'ÿ'& 'ys'& 'è'g's'g's' Reproductive frequency Reproductive season 3Annual once «Annual (2 or more) «Biennial ■ Semel parous ■Winter «Spring «Summer «Autumn 80% 1111111111111 60% 40%- 20 %- CO C q ^ C o ^ Cq ^ Cq ^ C o ^ Cq ^ O q ^ C o ^ Cq ^ Cq ^ Cq # ' # ' s /S‘' s /S /3'/3'' Fig. 5. Percentages of trait categories of the trait variables adult mobility, feeding type, reproductive type, dissemination as well as reproductive frequency and -season in Box A from 1998 to 2009. Marine Ecology32 (Suppl. 1 ) (2011 ) 49-57 © 2010 Blackwell Verlag GmbH 53 Effect of temperature variability on epifauna N e u m a n n & K röncke these species was found in winter 2003 to winter 2007, effect of cold winters on ecosystems was less clear in the correlating to the winters following exceptional warm later years due to the clear response of ecosystem compo autumn months. nents to the warming of the North Sea. However, cold winters are also part of climate variability and their impacts have been observed in long-term records of D iscussion plankton (Martens & van Beusekom 2008) and benthic The shift in trait composition in 2002 largely coincided infauna (Krönckeet al. 1998; Schröder 2005). The current with shifts in the epifaunal species composition in Box A study revealed that the recovery time of epifaunal com (N eum ann et al. 2008a,b), which was also observed in the munities after cold winters was 7-8 years, compared to a shallow West and North Frisian coasts (Neumann et a í much faster recovery in plankton communities ( < 1 year) 2009a,b). In all these areas, exceptionally high abundances and benthic infauna communities (2-5 years) in the Ger of the brittle star Ophiura albida were found in 1998, man Bight (Schröder 2003; Martens & van Beusekom which decreased in subsequent years parallel to an 2008). The prolonged recovery time of epifauna in shal increase in diversity, secondary production as well as low, well-mixed areas has important implications for the abundance and biomass of other epifaunal species. The assessment of ecosystem health and consequently for BTA revealed a decrease in traits related to an opportu management and conservation strategies. nistic life mode after summer 1998 (Fig. 4). The patterns Succession of the epifaunal community in Box A was found follow the theory on the effects of disturbance on mainly driven by biotic factors such as trophic interac communities going back to the models of Odum (Odum tions (Neumann et al. 2008a,b). During our study, SST 1969) and Pearson-Rosenberg (Pearson & Rosenberg anomalies at the station close to Box A were mainly 1978). In the Pearson-Rosenberg model, the second stage above the long-term mean of 1968-2008, with exception in the faunal succession was characterized by the appear ally high anomalies in the autumn months of 2002-2006. ance of opportunistic species, which were able to recolo- It is obvious that traits related to the reproduction of spe nize a disturbed habitat faster than K-selective species. cies such as reproductive type, -frequency and -season The opportunistic life mode (r-strategy) involves have shown a clear response to these high autumn water increased reproductive effort through early onset of matu temperatures. Thus, a high proportion of sexual egg layers rity, short life span and small body size (Heip 1995) pro with planktonic larvae were found during the period of viding a selective advantage in disturbed environments by high autumn anomalies. Additionally, higher abundances utilizing free resources faster than others. Thus, the of species which reproduce in autumn and are able to decrease in small species ( 1 - 2 cm) with an age of matu reproduce twice a year were found in the winters after rity <2 years and an adult longevity of 2-5 years con the warm autumn months. Direct positive effects of tem firmed (according to the BTA) that the community in perature on key stages of reproduction are well known Box A was at a succession stage between the opportunistic for many species which were found in Box A. For exam dominance and the established community between 1998 ple, the larval development of the swimming crab Liocar and 2002. The occurrence of seasonal variation formed a cinus holsatus (from hatching to metamorphosis) is faster decisive part of ‘persistence stability’ in benthic commu with increasing temperature (Choy 1991). Furthermore, nities following catastrophic events in temperate areas the duration of the larval stage as well as the frequency of with a highly dynamic physical regime (Arntz & Rumohr breeding of Pandalus spp. is positively linked to tempera 1982). Seasonality was found for the traits reproductive ture (Bergström 2000), as has also been observed in other frequency and reproductive season since 2003, but other caridean species (Wear 1974). Henderson et al. (2006) traits also show conspicuous changes between 2 0 0 2 and found that the recruitment of the shrimp Crangon cran 2003 (Fig. 5). For instance, the proportions of predators, gon in autumn was correlated to SST and the winter scavengers and grazers were much higher after winter index of the North Atlantic Oscillation. Similar changes 2003, whereas those of deposit and filter/suspension feed were also observed in plankton communities where, in ers decreased. In conjunction with the shift in overall trait particular, echinoderm and decapod larvae were found in composition in 2 0 0 2 , this indicated a shift in the commu higher abundances and/or earlier due to temperature- nity from the disturbance stage after the cold winter of induced shifts in the reproduction cycle of benthic species 1995/96 to the recovery stage between 2002 and 2003. (Lindley et al. 1993; Greve et al. 2001; Edwards & Rich Thus, the variability of trait composition underpins the ardson 2004; Kirby et al. 2007; Richardson 2008). hypothesis that the cold winter resulted in the outbreak In contrast to traits related to reproduction, traits such of opportunists, followed by characteristic post-distur- as feeding type and dissemination showed no obvious bance succession stages to the point of an established response to the exceptional warm SST in autumn but a community in 2002/2003 (Neumann et al. 2008a,b). The clear shift in winter 2003 (when there was a higher 54 Marine Ecology 32 (Suppl. 1) (2011) 49-57 © 2010 Blackwell Verlag GmbH Neumann & Kröncke Effect of temperature variability on epifauna proportion of predators, scavengers, grazers and migra functional aspects of one single component (the benthic tory species) in relation to overall positive anomalies of epifauna), it provides useful information on how the SST. However, it was difficult to determine whether these functional composition of epifauna was linked to envi changes could be attributed to the climate-induced vari ronmental changes. But there is an obvious need for ability of SST or whether the trait compositions just holistic studies that include more biotic components as reflect the stage of the common community following the well as ecological processes. succession after the cold winter. As Neumann et a í (2008a,b, 2009a,b) argued that the epifauna in the shallow Acknowledgements Southeastern North Sea were influenced by increased food supply due to a much longer period of primary produc We thank the captains and crews of RV Walther Herwig tion in conjunction with higher SSTs (Hughes & Holliday III for their assistance during sampling. We gratefully 2007), we expected higher proportions of organisms such acknowledge the Federal Research Institute for Rural as deposit feeders in this study, but that was not the case. Areas, Forestry and Fisheries for providing ship time and However, the increased abundance of epifaunal predators data. In particular, we thank Achim Schulz (BSH) for might be attributed to higher abundances of, for example, providing temperature data. The present study was pre deposit feeders of lower trophic levels {i.e. benthic pared at the Biodiversity and Climate Research Centre infauna, meiofauna) which in turn benefited from (BiK-F), Frankfurt a.M., and financially supported by the increased food supply. Thus, a higher proportion of pre research funding programme ‘LOEWE -Landes-Offensive dators since 2003 in Box A might also be attributed to zur Entwicklung Wissenschaftlich-ökonomischer Exzel climate-induced temperature increase, especially as Hen lenz’ of Hesse’s Ministry of Higher Education, Research, derson et a í (2006) also suggested that there must also be and the Arts. an increase in food availability if an increase of larval development due to higher temperature is to be translated R eferences into higher recruitment. The results of the BTA primarily revealed a greater Arntz W.E., Rumohr H. ( 1982) An experimental study of mac- impact of temperature variability on traits related to robenthic colonization and succession, and the importance reproduction than on feeding traits in Box A. Therefore, of seasonal variation in temperate latitudes. Journal of Exper we assumed that the climate-induced temperature vari imental Marine Biology and Ecology, 64, 17-45. ability indeed had a greater influence on the recruitment Beaugrand G. (2004) The North Sea regime shift: evidence, of epifauna than on the food availability in the study causes, mechanisms and consequences. Progress in Oceanog area. In contrast, food supply was found to be an impor raphy, 60, 245-262. tant factor influencing epifaunal species in the northern, Bell J.J. (2007) Contrasting patterns of species and functional composition of coral reef sponge assemblages. Marine Ecol stratified North Sea (Neumann et al. 2009a,b), which ogy Progress Series, 339, 73-81. were regarded to be food limited (Davies & Payne 1984). Bergström B.I. (2000) The biology of Pandalus. Advances in Further analysis, including our data from the Northern Marine Biology, 38, 55-245. North Sea, should be used to test these hypothesized Beukema J.J. ( 1992) Expected changes in the Wadden Sea ben functional differences between the Northern and Southern thos in a warmer world: lessons from periods with mild N orth Sea. winters. Netherlands Journal of Sea Research, 30, 73-79. We are aware that we will have missed some aspects of Brander K., Blom G., Borges M.F., Erzini K., Henderson G., functioning as we excluded rare species in the BTA and MacKenzie B.R., Mendes H., Ribeiro J., Santos A.M.P., Tore- thus their contribution to functionality. Additionally, we sen R. (2003) Changes in fish distribution in the eastern do not cover all aspects of functioning due to the choice North Atlantic: are we seeing a coherent response to changing of only 12 functional traits in the BTA. Two expert temperature? ICES Marine Science Symposia, 219, 261-270. workshops in Plymouth and London have identified 10 Bremner J. (2008) Species’ traits and ecological functioning in key aspects of marine system functioning and 24 corre marine conservation and management. Journal of Experi sponding functional traits for the BTA (Frid et a í 2008). mental Marine Biology and Ecology, 366, 37-47. We excluded traits such as ‘Energy transfer efficiency’ or Bremner J., Frid C.L.J., Rogers S.I. (2003a) Assessing m arine ‘Intra-specific sociability’ to reduce the demand for func ecosystem health: the long-term effects of fishing on func tional data, which were mostly impossible to get. How tional biodiversity in North Sea benthos. Aquatic Ecosystem ever, this does not greatly affect our conclusions. The Health & Management, 6, 131-137. definition of functioning by Jax (2005) given in the Bremner J., Rogers S.I., Frid C.L.J. (2003b) Assessing func introduction emphasizes both processes and the role of tional diversity in marine benthic ecosystems: a comparison single components. Although our study only incorporates o f approaches. Marine Ecology Progress Series, 254, 11-25. Marine Ecology32 (Suppl. 1 ) (2011 ) 49-57 © 2010 Blackwell Verlag GmbH 55 Effect of temperature variability on epifauna N e u m a n n & K röncke Bremner J., Frid C.L.J., Rogers S.I. (2005) Biological traits of Hooper D.U., Solan M., Symstad A., Diaz S., Gessner M.O., the North Sea benthos: does fishing affect benthic ecosystem Buchmann N., Degrange V., Grime P., Hulot F., Mermillod- function? American Fisheries Society Symposium, 41, 477- Blondin F., Roy J., Spehn E., van Peer L. (2002) Species 489. diversity, functional diversity, and ecosystem functioning. Bremner J., Rogers S.I., Frid C.L.J. (2006) M ethods for describ In: Loreau M., Naeem S., Inchausti P. (Eds), Biodiversity ing ecological functioning of marine benthic assemblages and Ecosystem Functioning: Synthesis and Perspectives. Oxford using biological traits analysis (BTA). Ecological Indicators, University Press, New York: 195-208. 6, 609-622. Hughes S.L., Holliday N.P. (2006) ICES Report on Ocean Cli Buchanan J.B., Moore D.C. (1986) Long-term studies at a ben mate 2005. ICES Cooperative Research Report. Report No. thic station off the coast of Northumberland. Hydrobiologia, 280, International Council for the Exploration of the Sea, 142, 121-127. Copenhagen. Chevenet F., Doledec S., Chessel D. (1994) A fuzzy coding Hughes S.L., Holliday N.P. (2007) ICES Report on Ocean Cli approach for the analysis of long-term ecological data. mate 2006. ICES Cooperative Research Report. Report No. Freshwater Biology, 31, 295-309. 289, International Council for the Exploration of the Sea, Choy S.C. (1991) Embryonic and larval biology of Liocarcinus Copenhagen. holsatus and Necora puber (Crustacea: Brachyura: Portuni Jax K.. (2005) Function and “functioning” in ecology: what dae). Journal of Experimental Marine Biology and Ecology, does it mean? Oikos, 111, 641-648. 148, 77-92. Kirby R.R., Beaugrand C., Lindley J.A., Richardson A.J., Davies J.M., Payne R. (1984) Supply of organic matter to the Edwards M., Reid P.C. (2007) Climate effects and benthic- sediment in the northern North Sea during a spring phyto pelagic coupling in the North Sea. Marine Ecology Progress plankton bloom. Marine Biology, 78, 315-324. Series, 330, 31-38. Diaz S., Cabido M. (2001) Vive la difference: plant functional Kirby R.R., Beaugrand C., Lindley J.A. (2008) Climate-induced diversity matters to ecosystem processes. Trends in Ecology effects on the meroplankton and the benthic-pelagic cou & Evolution, 16, 646-655. pling ecology of the North Sea. Limnology and Oceanogra Edwards M., Richardson A.J. (2004) Impact of climate change phy, 53, 1805-1815. on marine pelagic phenology and trophic mismatch. Nature, Kröncke I., Dippner J.W., Heyen H., Zeiss B. (1998) Long 430, 881-884. term changes in macrofaunal communities off Norderney Edwards M., Johns D.G., Beaugrand C., Licandro P., John (East Frisia, Germany) in relation to climate variability. A.W.G., Stevens D.P. (2008) Ecological Status Report: Results Marine Ecology Progress Series, 167, 25-36. from the CPR Survey 2006/2007., SAHFOS, Plymouth. Lindley J.A., Williams R., Hunt H.G. (1993) Anomalous sea Ehrich S., Stransky C. (2001) Spatial and temporal changes in sonal cycles of decapod crustacean larvae in the North Sea the southern species component of North Sea fish assem plankton in an abnormally warm year. Journal of Experimen blages. Senckenbergiana Maritima, 31, 143-150. tal Marine Biology and Ecology, 172, 47-65. Ehrich S., Adlerstein S., Brockmann U., Floeter J., Garthe S., Martens P., van Beusekom J.E.E. (2008) Zooplankton response Hinz H., Kröncke I., N eum ann H., Reiss H., Sell A.F., Stein to a w arm er northern W adden Sea. Helgoland Marine M., Stelzenmüller V., Stransky C., Temming A., Wegner G., Research, 62, 67-75. Zauke G.P. (2007) 20 years of the German Small-Scale Bot Mouillot D., Dumay O., Tomasini J.A. (2007) Limiting simi tom Trawl Survey (GSBTS): a review. Senckenbergiana Mari larity, niche filtering and functional diversity in coastal tima, 37, 13-82. lagoon fish communities. Estuariae Coastal and Shelf Science, Frid C.L.J., Paramor O.A.L., Brockington S., Bremner J. (2008) 4, 443-456. Incorporating ecological functioning into the designation Neumann H., Ehrich S., Kröncke I. (2008a) The effects of cold and management of marine protected areas. Hydrobiologia, winters and climate on the temporal variability of an epi- 606, 69-79. benthic community in the German Bight. Climate Research, Greve W., Lange U., Reiners F., Nast J. (2001) Predicting the 37, 241-251. seasonality of North Sea Zooplankton. Senckenbergiana Neumann H., Ehrich S., Kröncke I. (2008b) Spatial variability Maritima, 31, 263-268. of epifaunal communities in the North Sea in relation to Grime J.P. (1997) Biodiversity and ecosystem functioning: the sampling effort. Helgoland Marine Research, 62, 215-225. debate deepens. Science, 277, 1260-1261. Neumann H., Ehrich S., Kröncke I. (2009a) The effect of Heip C. (1995) Eutrophication and zoobenthos dynamics. increasing temperature on epibenthic communities and spe Ophelia, 41, 113-136. cies in the northern North Sea. Marine Biology, 156, 1817— Henderson P.A., Seaby R.M., Somes J.R. (2006) A 25-year 1826. study of climatic and density-dependent population regula Neumann H., Reiss H., Rakers S., Ehrich S., Kröncke I. tion of common shrimp Crangon crangon (Crustacea: Cari (2009b) Temporal variability of southern North Sea epi dea) in the Bristol Channel. Journal of the Marine Biological fauna communities after the cold winter 1995/1996. ICES Association of the United Kingdom, 86, 287-298. Journal of Marine Science, 66, 2233-2243. 56 Marine Ecology32 (Suppl. 1) (2011) 49-57 © 2010 Blackwell Verlag GmbH Neumann & Kröncke Effect of temperature variability on epifauna Odum E.P. ( 1969) The strategy of ecosystem development. Sci Schröder A. (2005) Community dynamics and development of ence, 164, 262-270. soft bottom macrozoobenthos in the German Bight (North Pearson T.H., Rosenberg R. ( 1978) Macrobenthic succession in Sea) 1969-2000. Report of Polar Research, 494, 1-181. relation to organic enrichment and pollution of the marine Sims D.W., Genner M.J., Southward A.J., Hawkins S.J. (2001) environment. Oceanography and Marine Biology. An Annual Timing of squid migration reflects North Atlantic climate Review, 16, 229-311. variability. Proceedings. Biological Sciences/The Royal Society, Perry A.L., Low P.J., Ellis J.R., Reynolds J.D. (2005) Climate 268, 2607-2611. change and distribution shifts in marine fishes. Science, 308, Tillin H.M., Hiddink J.G., Jennings S., Kaiser M.J. (2006) 1912-1915. Chronic bottom trawling alters the functional composition Reiss H., Meybohm K., Kröncke I. (2006) Cold winter effects of benthic invertebrate communities on a sea-basin scale. on benthic macrofauna communities in near- and offshore Marine Ecology Progress Series, 318, 31-45. regions of the North Sea. Helgoland Marine Research, 60, Wear R.G. (1974) Incubation in British decapod crustacea, 224-238. and the effects of temperature on the rate and success of Richardson A.J. (2008) In hot water: Zooplankton and climate embryonic development. Journal of the Marine Biological change. ICES Journal of Marine Science, 65, 279-295. Association of the United Kingdom, 54, 745-762. Schratzberger M., Warr K., Rogers S.I. (2007) Functional Ziegelmeier E. (1970) Uber Massenvorkommen verschiedener diversity of nematode communities in the southwestern makrobenthaler Wirbelloser während der Wiederbesied N orth Sea. Marine Environmental Research, 63, 368-389. lungsphase nach Schädigungen durch ‘katastrophale1 Schröder A.. (2003) Community dynamics and development of Umwelteinflüsse. Helgoländer wissenschaftliche Meeresunter soft bottom benthos in the German Bight (North Sea) suchungen, 21, 9-20. 1969-2000. Ph.D. thesis, University Bremen (Germany). Marine Ecology32 (Suppl. 1 ) (2011 ) 49-57 © 2010 Blackwell Verlag GmbH 57 anmarine evolutionary perspective ecology < Marine Ecology. ISSN 0173-9565 ORIGINAL ARTICLE Zoobenthos as an environmental quality element: the ecological significance of sampling design and functional traits Katri Aarnio, Johanna Mattila, Anna Törnroos & Erik Bonsdorff Department of Biosciences, Environmental and Marine Biology & Huso Biological Station, Abo Akademl University, Turku, Finland Keywords A bstract Brackish water benthic Index; biological traits analysis; ecological status; sampling The EC Water Frame Directive (WFD) states that all coastal water bodies must methodology; water frame directive; achieve ‘good ecological status’ by the year 2015. A range of different classifica zoobenthos. tion methods have been developed and used to define ecological status to sup port the WFD. The aim of this study was to compare the effects of using two Correspondence different mesh sizes of sieve (1.0 and 0.5 mm) on zoobenthic assemblages and Katri Aarnio, Department of Biosciences, on the ecological status of benthic macrofauna (using the Brackish water Environmental and Marine Biology & Huso biological station, Abo Akademi University, benthic index, BBI) in three ecologically distinct archipelago areas (Inner, Mid FI-20S00 Turku, Finland. dle and Outer) in the Aland Islands, Northern Baltic Sea. We performed a bio E-mail: [email protected] logical trait analysis (BTA) to evaluate differences in the functional (trait) diversity of macrofauna collected using different mesh sizes and estimate the Accepted: 25 November 2010 ecological relevance of mesh size. The results showed that sieve mesh size had significant effects on the recorded number of species, abundance, and total bio doi:10.1111/j. 1439-0485.2010.00417.x mass of the zoobenthos. Small-bodied species and juveniles (e.g. Macoma balth ica) were not observed when using a 1.0-mm mesh. The ecological status (sensu WFD) was only slightly affected by the mesh size, and all areas had good or high ecological status. BTA showed a difference in trait composition when using 0.5- or 1.0-mm mesh, particularly in the Outer area, where the propor tion of small-sized species was high. Our results highlight how biological traits, in addition to species number and biomass, can play a key role when analyzing ecosystem structure for assessment and classification of coastal ecosystems. We show that combining traditional monitoring for the EU WFD with a functional analysis strengthens our ability to interpret environmental quality, and thus increases the precision of our advice for management purposes. disturbance of coastal areas leads to changes in species Introduction number and the composition of assemblages, as well as in Eutrophication is one of the most severe environmental their abundance and biomass (Pearson & Rosenberg issues in the Northern Baltic Sea, affecting both pelagic 1978; Cederwall & Elmgren 1980; Diaz & Rosenberg 1995; and benthic environments, in shallow and deep areas Norkko & Bonsdorff 1996; Bonsdorff & Pearson 1999; (Elmgren 1989; Bonsdorff et al. 1997; HELCOM 2009). Perus & Bonsdorff 2004). Zoobenthos is widely used as an indicator of change in In the Northern Baltic Sea, the number of benthic spe environmental conditions, as the organisms are relatively cies is low primarily due to brackish water conditions stationary and several species live for many years. Thus (salinities vary between 3 and 7 psu). Species are either of changes in environmental conditions are reflected in marine or limnic origin and they live at the limits of their zoobenthos as altered community parameters. Increased physiological tolerance (Bonsdorff 2006). Due to the low 58 Marine Ecology 32 (Suppl. 1) (2011) 58-71 © 2010 Blackwell Verlag GmbH Aarnio, Mattila, Törnroos & Bonsdorff Zoobenthos as an environmental quality element salinity, the organisms are small in size compared with poor and bad (2000/60/EG WFD; Borja et a í 2000; their relatives in fully marine areas. Hence, the size-spec- Rosenberg et a í 2004; Perus et al. 2007; Josefsson et a l trum of benthic organisms is narrow, and most organisms 2009). As the EU WFD classifies the entire Baltic Sea as have an adult size smaller than a few centimeters one eco-region, this implies that the same quality ele (although certain polychaetes may reach 1 0 cm in length). ments and indicators should be valid across the entire To fully illustrate and understand community dynamics, sea. As has been shown in numerous recent studies, this individuals’ size must be considered a priori when choos is not the case (Perus & Bonsdorff 2004; Leonardsson ing methods for the study. Sieve mesh sizes of 1 mm or et al. 2009; Rosenberg et al. 2009). Not only must more are widely used for benthic surveys in marine areas species-sensitivity values be set according to regional (Borja et al. 2009 and references therein), while a 0.5-mm ecological baselines, but also indicators other than just sieve mesh is more often used in local studies in the presence/absence, abundance and biomass must be Northern Baltic Sea (Perus et al. 2007). included. One potential way forward is to include Assessments of anthropogenic effects on benthic sys functioning (cf. Bonsdorff & Pearson 1999) and specific tems have mostly been based on taxonomic composition biological traits (Bremner 2008). In the case of the North and relative abundance of taxa, which are still valuable ern Baltic Sea species, size must also be considered. and easy to comprehend, serving as a base for further The aim of this study was to investigate how different assessments (Blomqvist & Bonsdorff 1986; Bonsdorff & mesh sizes of sieve affect the results of benthic studies Blomqvist 1993; Perus & Bonsdorff 2004). However, both structurally and functionally. We studied the effects recent studies have questioned the use of only species of macrofauna assemblage observations, using two sieve number and other basic parameters as measures of eco mesh sizes: 1.0 or 0.5 mm. The effects were measured on system health and functioning, especially as marine ben basic community parameters, as well as on the ecological thic systems harbour great numbers of phyla for which status of the environment, along an environmental gradi the taxonomic divisions are still uncertain (Warwick & ent from inner to outer archipelago using a specially Somerfield 2008). The increasing need for broadening of developed index, the brackish water benthic index (BBI; the quality concept for habitats or coastal areas and Perus et al. 2007). In addition we performed a biological accepting the importance of ecosystem functioning, has trait analysis (BTA; Bremner et a í 2003) to evaluate dif promoted a scientific and applied discussion about pres ferences in functional (trait) diversity of different mesh ent quality indices, what they describe and how they fit sizes and estimated the ecological relevance of mesh size modern management approaches (Frid et al. 2008; Tillin choice in relation to the selected macro-habitats. et al. 2008; Borja et al. 2009). The ability of management directives (e.g. sampling methods), ecological indices and Study Areas other measures of anthropogenic stress to encompass dif ferent scales of functioning and biodiversity during eco The field sampling for this study was conducted in three system change has been questioned. For example, the areas (macro-habitats) of different exposure and degree of contribution of some species to the functioning of the organic input to the sediments in the Aland archipelago ecosystem may decrease and that of other species increase (N Baltic Sea): Inner, Middle and Outer archipelago areas during environmental change, and to include this requires (Fig. 1). The Inner area was sheltered from the open sea an appropriate time and spatial sampling scale (Thrush and had limited water exchange. It was directly affected et al. 1997, 2000; Yachi & Loreau 1999; Stachowicz et al. by human activities, mainly agriculture and food industry. 2002). Further, organisms that appear to perform similar The Middle area was semi-exposed and moderately roles may not always respond to stress in the same way affected by local sources of eutrophication. The Outer (Ramsay et a í 1998). Hence, the next step is to connect area was exposed to the open sea with high water the changes in abiotic characteristics of the habitat or exchange rates. It was only slightly affected by local environment, the number and abundance (or biomass) of human activities, and had low nutrient levels in compari species (species diversity) and the functioning of these son with the other areas studied. The sediment quality species (functional diversity) in order to best identify and ranged from sand/gravel in the Outer exposed areas, with direct management efforts (Bremner et a í 2003; Jax 2005; low organic content (mean 1 .8 %) to clay/mud with high Bremner 2008). According to the EU Water Frame Direc organic content (mean 5.6%) in the Inner sheltered areas tive (2000/60/EG; WFD) all coastal waters should achieve (Table 1). The entire region (Fig. 1) has been thoroughly a good ecological status by 2015. For this purpose, water studied for benthic infauna since the early 1970s and areas are being classified using biological quality elements there were comprehensive records of the benthic assem (macrophytes, phytoplankton, zoobenthos) and divided blages in these areas (e.g. Helminen 1975; Bonsdorff et a í into five classes of ecological status: high, good, moderate, 1991, 2003; Perus et al. 2001; Perus & Bonsdorff 2004). Marine Ecology 32 (Suppl. 1) (2011) 58-71 © 2010 Blackwell Verlag GmbH 59 Zoobenthos as an environmental quality element Aarnio, Mattila, Törnroos & Bonsdorff Riiiand Fig. 1. Map of the study areas in the Aland archipelago, Northern Baltic Sea. I, Inner area; M, Middle area; 0, Outer area. Table 1. Site characteristics of the three archipelago areas. The four variables, Depth, Salinity, Oxygen (02%) in bottom w ater and Organic con tent in sediment are presented as: mean (min-max), the variable Dominating sediment type with the following abbreviations; C, clay; M, mud; S, sand; G, gravel. Archipelago area Site characteristics Inner Middle Outer Depth m 10.7 (5-16) 11.2 (6-27) 14 (4-26) Salinity, %0 5.6 (5.6-5.7) 5.6 (5.3-5.8) 6.0 (5.4-6.2) 0 2% in bottom water 86 (80-91) 93 (87-99) 94 (86-113) Dominating sediment type C, MC C, MC, CG CS, S, SG Organdie content in sediment (loi), % 5.6 (2.7-8.6) 5.5 (0.9-9.2) 1.8 (0.4-10.7) 1.0- and 0.5-mm screens (reported as 1.0 and 0.5 mm Material and Methods pooled). Fauna were identified to the lowest possible taxo Sampling was conducted between 14 August 2007 and nomic level, counted and weighed (wet weight). The length 7 September 2007 when the animal summer-recruitment o f individual Macoma balthica was measured to the nearest had occurred. In all areas, zoobenthic samples were taken mm for estimates of population structure and recruitment using an Ekman-Birge grab sampler (289 cm2) from two success. A paired i-test was used to compare the number of depth zones: < 1 0 m (hereafter called ‘shallow’) and > 1 0 m species, abundance and biomass between the 0.5- and 1.0- (hereafter called ‘deep’). Five replicate samples were taken mm mesh results in the different depth zones and areas. An from six stations in each area; three on shallow and three overall analysis of mesh size effects was done using two- on deep bottoms. The samples were sieved through both way ANOVA, with sieves (0.5 and 1.0 mm) and areas 60 Marine Ecology32 (Suppl. 1) (2011) 58-71 © 2010 Blackwell Verlag GmbH Aarnio, Mattila, Törnroos & Bonsdorff Zoobenthos as an environmental quality element (Inner, Middle, Outer) as factors. Prior to the analyses, Individual taxa were then coded for the extent to which data were log (x + 1 ) transformed if they did not meet the they display the modalities in a scoring range between 0 assumptions of normality and homogeneity of variances. and 3, with 0 being no affinity to a modality and 3 being All statistical analyses were performed using the statistical total affinity. The category scores were then standardized and graphical software GRAPH PAD PRISM, version 5.0. to 1 within a trait. This ‘fuzzy coding’ procedure (Cheve- net et a í 1994) allows taxa to exhibit the modalities of a variable (trait) to different degrees. The procedure was Ecological status developed and first applied to terrestrial plants and fresh The ecological status sensu WFD was measured using the water invertebrates (Olff et a í 1994; Townsend & Hildrew BBI equation 1 (Perus et al. 2007). It was adopted from 1994) but has now also been introduced for marine sys the benthic quality index (BQI; Rosenberg et al. 2004) tems (Bremner et a í 2003, 2006a). For our purposes, data and adjusted for low-saline coastal areas, with low species on traits were obtained from the primary and secondary numbers, and the sensitivity values for the species have literature and by consulting expert advice (A. Törnroos been adjusted to their actual environment (cf. Rumohr unpubl. data). Thus, relevant and reliable information on et al. 1996). BBI follows the assumption that biodiversity all traits was obtained for all taxa sampled. This proce increases with increasing distance from a pollution source dure resulted in a taxa by trait matrix, one of two tables along a gradient of disturbance, and can take values included in the BT analysis. To link the abundance of between 0 and roughly 1 (Pearson & Rosenberg 1978; taxa at each station and mesh size with the traits dis Perus et al. 2007). played by the taxa, we conducted a co-inertia analysis (Coi; Dolédec & Chessel 1994). This analysis assesses the co-structure between two data tables, in this case a ‘taxa by station’ table and a ‘taxa by trait’ table, and simulta neously ordinates the two tables, maximizing both the variance from the individual tables and the correlation between them (Dolédec & Chessel 1994; Dray et a í where BQI = benthic quality index (sensu Rosenberg et al. 2003). As a first step, two separate ordinations were con 2004); BQImax = maximum BQI-value recorded within ducted: a centered PCA (principle component analysis) each type after calculating all available data within the on the “taxa by station” table and a FCA (fuzzy corre national Finnish zoobenthos database ‘Hertta’ (http:// spondence analysis) on the “taxa by trait” table (Cheve- www.ymparisto.fi ); H’ = Shannon-Wiener diversity index net et a í 1994; Charvet et a í 1998). These were then used (log 2 -base); H’max = maximum H’-value recorded within in the Coi analysis, and the significance of the resulting type after calculating all available data within national co-structure was examined with the RV coefficient (a zoobenthos database; AB = total abundance at each measure of similarity between squared symmetric station; and S = number of species/taxa at each station. matrices). To evaluate if the value of RV significantly dif fered from zero, a Monte-Carlo random permulation test was performed (n repetitions = 999) (Dolédec & Chessel Biological trait analysis (BTA) 1994). To investigate trait patterns in relation to mesh For the BTA we selected 10 specific traits representing the size along the environmental gradient, we conducted two principal aspects of morphology, life-history, living-, separate Coi analyses; one on the Inner area stations, and feeding habit, and movement of the studied taxa one on the Outer area stations. This was done to best (Table 2). These were chosen to maximize the differences elucidate patterns in the two areas, and to prevent the among species or taxa, and thus to elucidate representative uneven abundance distribution in the inner archipelago trait patterns in sampling design and differences between to mask the more even pattern (lower variability) of areas. The 10 traits were separated into sub-categories, i.e. the outer archipelago. BTA was thus not conducted modalities. Body design, for example, was divided into on the Middle area, as we wanted to highlight the verminiform unsegmented, verminiform segmented, opposite ends of the range of environmental and faunal bivalved, turbinate and articulate. This resulted in a abundance. Prior to the analysis, down weighting of total of 45 different trait modalities for the species-pool abundant taxa was done using square-root transforma selected (Table 2). The division of traits and categories tion. The scores were plotted on ordination maps, with was deduced from Bonsdorff & Pearson (1999), Pearson each point representing the abundance-weighted biologi (2001), and Bremner et a í (2003, 2006a,b), and revised cal trait composition of each station. All analyses were and applied to fit the Baltic Sea benthic species (A. Törn done in the R environment (R Development Core Team roos unpubl. data). 2009). Marine Ecology32 (Suppl. 1 ) (2011 ) 58-71 © 2010 Blackwell Verlag GmbH 61 Zoobenthos as an environmental quality element Aarnio, Mattlla, Törnroos & Bonsdorff Table 2. Biological traits (10) and modalities (45) ascribed to species and used In the BTA. Labels listed correspond to trait modalities In Fig. 5. Biological traits Trait modalities Labels Explanations Mean size 0.1-1 mm VS Very small 1-5 mm S Small 5 mm-1 cm SM Small-medium 1-3 cm M Medium Body design Vermiform unsegmented VermLuns Wormlike, lacking true segments Vermiform segmented Verml-seg Wormlike, seml-lndependent units Blvalved Blvalved Shell with two valves jointed by a ligament Turbinate Turbinate Whorled shell Articulate Articulate Jointed, arthrous Larval type Planktotrophlc Planktotrophlc Feeding on materials captured from the plankton Lecltotrophlc Lecltotrophlc Nourished on Internal resources, yolk Direct development Dlrect_dev Direct development of mini adults Living habit Attached Attached Adherent to substratum (>95% of adult time) Tube-dweller Tube_dweller Burrow dweller Burrow_dweller Free-living Free In or on sediment, In water column Environmental Infauna (>5 cm) lnf_deep Living within substrate, deeper than 5 cm position Infauna middle (2-5 cm) lnf_mlddle Living within substrate, between 2 and 5 cm Infauna (top 2 cm) lnf_top Living within top 2 cm of substrate Eplbenthlc Eplbenthlc Living on the surface of substrate Bentho-pelaglc Bent_pel Living In the water column but (primarily/occasionally) feeds on the bottom Feeding habit Detrltlvore Detrltlvore Feeds on detritus Omnivore Omnivore Feeds on mixed diet of plant and animal material Herbivore Herbivore Feeds on plants Carnivore Carnivore Feeds on animals (predator) Scavenger Scavenger Feeds on dead organic material Resource capture Jawed Jawed Jaws, mandibles method Siphon Siphon Tentaculate Tentaculate Pharynx Pharynx Both with and without jaws Radula Radula Rasping Mobility Sessile Sessile Temporary (e.g. Mytilus edulis) Seml-moblle Seml_mobll Mobile Mobil Movement Byssus Byssus Occasional movement with byssus threads method Swimmer Swimmer Fins, legs, appendages via undulatory movement Rafter/dri fter Raft_drlft Rafting on e.g. algal mats, drifting Crawler Crawler On substrate via muscles, legs or appendages Burrower Burrower Lives and or moves In a burrow Tube-bullder Tube_bullder Lives and moves In a tube Sediment No transport No_trans No transport transportation Diffusive mixing Dlff_mlxlng Random diffusive transport (e.g. reworking, excavatlor Surface deposition Surf_deposltlon Surface deposition of particles, 'regeneration' (e.g. excavation, egestlon) Conveyer belt transport Conv_belt_trans Translocation of sediment, depth to surface (e.g. egestlon, excavation, defecations) Reverse conveyer Rev_conv_belt_trans Subduction of particles from surface to depth belt transport (e.g. egestlon, excavation) Results nated by gastropods and oligochaetes, typically occurring on muddy bottoms. Macoma balthica was also abundant Basic community parameters and dominated by biomass. In the Middle area, the assem Benthic faunal assemblages in the studied areas of the blage was similar but M. balthica dominated both numeri Aland archipelago consisted of 30 species/taxa (Table 3). cally and by biomass. The invasive polychaete Marenzelleria In the Inner area, the zoobenthic assemblage was domi sp. was relatively abundant in this area. The Outer area was 62 Marine Ecology32 (Suppl. 1) (2011) 58-71 © 2010 Blackwell Verlag GmbH Aarnio, Mattila, Törnroos & Bonsdorff Zoobenthos as an environmental quality element Table 3. Species list of the soft-bottom communities in the three abundance and biomass (P > 0.05, unpaired t-test), the archipelago areas (Inner, Middle and Outer) from both shallow and two depth zones were pooled in the statistical analysis. deep samples. X indicates if a species was present. Total number of The number of species found in each area was signifi species on shallow versus deep stations are given at the bottom of cantly lower using the 1.0-mm mesh than the 0.5-mm the table. Total # spp, total number of species in each area. mesh. In the Inner area, 18 species were found using the Inner area Middle area Outer area 0.5-mm mesh and 14 using the 1.0-mm mesh. In the Middle area, 22 species were recorded (19 with the larger Species/taxa Shallow Deep Shallow Deep Shallow Deep mesh alone). The Outer area had the highest total species Nemertea number: 28 with the 0.5-mm mesh and 27 with the Cyanophthalma X X X X X 1.0-mm mesh (Table 3). The mean number of species obscura was significantly higher using the 0.5-mm mesh compared Priapulida with the 1.0-mm mesh (P < 0.0001; paired f-test) in all Halicryptus X X spinulosus archipelago areas (Table 4). In the overall analysis using a Annelida 2-way ANOVA there was a significant difference in spe Oligochaeta X X X X X X cies number both between the sieves (P < 0.0001) and Harmothoe sarsi X X betw een the areas (P < 0.0001) (Table 5; Fig. 2A). Manayunkia X X X The total abundance values were significantly higher aestuarina with the 0.5-mm mesh than with the 1.0-mm mesh in all Marenzelleria spp. X X X X X areas (P < 0.0001; paired f-test) (Table 4). For both mesh Hediste diversicolor X X X Pygospio elegans X X sizes the abundance was lowest in the Inner area and Mollusca highest in the Outer area. In the overall analysis using a Cerastoderma X X X X 2-way ANOVA there was a significant difference in abun glaucum dance both between sieves (P < 0.0001) and among areas Macoma balthica X X X X X X (P = 0.0162) (Table 5; Fig. 2B). Mya arenaria X X X X X X Biomass values were also lowest in the Inner area and Mytilus edulis X X X X highest in the Outer area, and they were significantly Hydrobia spp. X X X X X X Radix spp. X higher in the 0.5-mm mesh compared with the 1.0-mm Limapontia capitata X mesh fraction in all areas (P < 0.0001; paired f-test) Potamopyrgus X X X X X (Table 4). In the overall analysis using a 2-way ANOVA antipodarum there was a significant difference in abundance both Theodoxus fluviatilis X between sieves (P < 0.0001) and among areas (P = Arthropoda 0.0043). There was also a significant interaction Acarina X (P = 0.0043) between mesh and area (Table 5; Fig. 2C). Ostracoda X X X X X X Semibalanus X X In all areas, both the abundance and the biomass esti Improvisus mates of the numerically dominant species were reduced Idotea chelipes X significantly when using only the 1.0-mm mesh (Fig. 3). Jaera albifrons X Gammarus sp. X X X Table 4. Results of paired f-test analysis of mesh size effects on Leptocheirus pilosus X X X number of species, abundance and biomass In the Inner, Middle and Monoporeia affinis X X X Outer areas. Corophium volutator X X X X X Saduria entomon X X X df f P Neomysis Integer X X X X X Chironomidae X X X X X Species Chironomus plumosus X X X X Inner archipelago 29 9.497 <0.0001 Total shallow versus deep 16 11 19 17 22 19 Middle archipelago 29 1S.S6 <0.0001 Total # spp 18 22 28 Outer archipelago 34 11.29 <0.0001 Abundance Inner archipelago 29 6.S9S <0.0001 Middle archipelago 29 7.236 <0.0001 characterized by M. balthica and other bivalves ( Mytilus ed Outer archipelago 34 8.2S3 <0.0001 ulis and Cerastoderma glaucum) together with Marenzelleria Biomass sp. The amphipod Monoporeia affinis was also common in Inner archipelago 29 S.373 <0.0001 Middle archipelago 29 8.142 <0.0001 this area. As both the shallow and the deep stations in all Outer archipelago 34 9.106 <0.0001 areas showed identical patterns regarding species number, Marine Ecology32 (Suppl. 1 ) (2011 ) 58-71 © 2010 Blackwell Verlag GmbH 63 Zoobenthos as an environmental quality element Aarnio, Mattlla, Törnroos & Bonsdorff Table 5. Results of 2-way ANOVA analysis of mesh size effects on A 14 I I 0.5 mm number of species, abundance and biomass In the Inner, Middle and m 1 2 ■ 1.0 mm Outer areas. 92 T ' 8 1 0 C l Source of « 8 0 variation df SS MS F P o3 6 -Q No. of species 1 4 Mesh 1 83.44 83.44 149.2 <0.0001 Z 2 Area 2 172.9 86.45 25.50 <0.0001 0 Interaction 2 0.8229 0.4115 0.7358 0.4947 B Subjects 16 54.25 3.390 6.063 0.0004 10 000 (matching) T" Residual 16 8.948 0.5592 I 7500- Abundance C Mesh 1 60,830,000 60,830,000 35.34 <0.0001 § 5000- Area 2 55,890,000 27,950,000 5.392 0.0162 O(Ü Interaction 2 6,332,000 3,166,000 1.840 0.1909 C o 2500 Subjects 16 82,930,000 5,183,000 3.012 0.0170 (matching) J 1 Residual 16 27,540,000 1,721,000 i_Ui Biomass C 2 0 0 -, Mesh 1 18.00 18.00 33.67 <0.0001 Area 2 67,730 33,860 7.827 0.0043 Interaction 2 8.331 4.165 7.793 0.0043 - 1 5 0 - Subjects 16 69,230 4327 8094 <0.0001 "o) (matching) s ’ 1 » - CD Residual 16 8.553 0.5345 O£ ûû 50 Q Oligochaetes, which were abundant in the Inner and Mid Inner Middle Outer dle area, were reduced by 96-99% when using only a 1.0- Archipelago area mm mesh size. Ostracods were reduced by 97% in the Inner area and by 100% in the Middle and Outer areas, Fig. 2. Number of species (A), abundance (B) and biomass (C) In 0.5-mm and 1.0-mm mesh In the three archipelago areas (Inner, and polychaetes, namely Marenzelleria sp, were reduced Middle, Outer). Values Indicate mean ± SD. by 75% in the Outer area when using the 1.0-mm mesh. The Baltic clam, M. balthica, dominated in all areas by biomass (58-78%), but abundance estimates were reduced overall status (Fig. 4). All but one sampled station and all significantly, when the 1 .0 -mm mesh was used. three areas were classified as ‘good’ ( 1 .0 -mm mesh) or The size distribution of M. balthica was affected by ‘high’ (0.5-mm mesh) ecological status. The BBI itself was mesh size: w hen the 1 .0 -mm mesh was used, the number highest in the Outer area, but as the class boundaries are of individuals measuring 1 and 2 m m in size (i.e. the different for the different areas (and depths), the ecologi annual spat) was reduced markedly. In the Inner area, cal status was similar to the other areas. 96% of 1-mm-sized M. balthica and 32% of 2-mm-sized individuals were lost using the 1.0-mm mesh. In the Mid Biological traits dle area, the corresponding values were 1 0 0 % ( 1 m m ) and 45% (2 mm), and in the Outer area 99% (1 mm) The Coi analyses for the Inner and Outer area stations and 47% (2 mm). Both sieves captured equal numbers of illustrated the relationship between taxon composition clams of size classes 3 mm and larger. Thus the main dif and abundance at shallow and deep stations with the two ference was the loss (or underestimation) of recruiting mesh sizes, and biological traits. A clear difference in the individuals, which may affect the evaluation of the eco trait composition was identified between the two archi logical status of a benthic habitat. pelago areas in terms of number of modalities found. In the Inner area, no scavengers (feeding type) or tube builders (movement type) were registered, resulting in 43 Ecological status modalities compared with 45 in the Outer area. This survey showed that zoobenthos as an ecological The significance of the resulting correlation (noted R quality element resulted in evaluation of a relatively good value in Table 6 ) between the two sets of coordinates was 64 Marine Ecology32 (Suppl. 1) (2011) 58-71 © 2010 Blackwell Verlag GmbH Aarnio, Mattila, Törnroos & Bonsdorff Zoobenthos as an environmental quality element A Abundance Biomass Inner area Abundance (Ind/m5) Biomass (giri2) 500 1 000 1 500 2000 2500 10 15 2 0 25 00 35 40 45 M. üattflts . fA. ansiTanö C. plbncsius - HyrHoOia spp. - P. antifxxSaru-m . N. On/srsicobr - B Abunda one (ind/m5) Biomass (g/m2) Middle area 500 1000 1500 2000 0 1 0 2 0 0 0 40 5 0 00 7 0 80 00 100 M. úasfita - Oligochaeta - Hydrobia spp, P. aniipoaaruni - MsranzelOnö spp Ostracoda Marenzelleria spp. - Abundance (indVm2) Biom ass (gVm2) Outer anea 500 1 000 1500 2000 2500 3000 20 40 EO 80 100 120 140 160 fA. ¿aivica Msrenzeti&riB spp. Ostracoda C. stáucom - Hydratö spp - Ifarsnzflferfispp. - Chironemidae - Hydrobia spp. C. voyator - Fig. 3. Abundance and biomass of the dominant species In 0.5-mm and 1.0-mm mesh In the Inner (A), Middle (B) and Outer (C) areas. Values Indicate mean ± SD. Note the different scales on x-axes. examined with the Monte-Carlo test. The test showed the first axis of the ordination, the two mesh sizes were that the Outer area had a borderline significance for a separated for both shallow and deep stations, the larger non-random pattern (RV = 0.254, P = 0.071). However, size being grouped more towards the centre and the smal the opposite was true for the Inner area (RV = 0.268, ler mesh size samples more spread out, i.e. separated from P = 0.297). Still, we chose to present both results to illus the others (Figs 5 and 6 ). This means that the mesh size trate the trend and more evident pattern in the Outer of the sieve influences the functional analysis, and implies area compared with the Inner one, considering the higher that for a reliable analysis of biological traits in these rela species diversity and more equal abundance of species tively species-poor assemblages, information is needed for (Figs 5 and 6 ). all species that can be sampled (Fig. 6 ). The shallow areas In the BTA of the Outer area, axes 1 and 2 of the co were characterized by small-sized detritivores (species inertia analysis accounted for 90% (65% and 25%, obtaining food through suspension, surface and/or sub respectively) of the variability in biological trait composi surface feeding) with a diffusive sediment transport mode tion between the stations. A clear separation between (taxa such as Macoma balthica, Potamopyrgus antipoda shallow and deep stations along the second axis could be rum and Chironomidae). The deeper areas showed a seen, especially for the outer archipelago (Fig. 6 ). Along compilation of detritivores performing no sediment Marine Ecology32 (Suppl. 1 ) (2011 ) 58-71 © 2010 Blackwell Verlag GmbH 65 Zoobenthos as an environmental quality element Aarnio, Mattlla, Törnroos & Bonsdorff 0 0.5 mm A 1 . 0 mm 0.8 Fig. 4. Ecological status, measured using BBI, 0.6 In shallow (S) and deep (D) bottoms In the three archipelago areas. Sampling stations are 0.4 Indicated on the x-axls. Background colours: blue = high, green = good, 0.2 yellow = moderate, orange = poor and red = bad ecological status. Note that the 0.0 r — r~- T ' class borders for different status Ile on F18F15 F6 F17 F9 F3 M13M20M7 M5 M15M18 E23 E26 E24 E18 E11 E15 different BBI-levels In different depths and different archipelago areas. transport (e.g. M ytilus edulis) and omnivorous tube- widespread and most abundant species are studied (El builders performing reversed conveyer belt transport (taxa lingsen et al. 2007). such as Marenzelleria sp. and Pygospio elegans). The mesh size of sieve also had a significant effect on the registered population structure of the bivalve Macoma balthica, which is a key species in the Northern Baltic Sea D iscussion (Segersträle 1962; Olafsson 1989). Population structure The results from this study showed that the choice of and recruitment success of the species are often used as mesh size (1.0 or 0.5 mm) in the sieve affected all basic indicators of the environmental conditions. Adult community parameters, in all areas and at both depths M. balthica are quite tolerant and can withstand stressed (shallow and deep). The number of species, abundance environmental conditions for some time, whereas juve and biomass were all significantly reduced when using the niles are sensitive even to small disturbances in the envi larger mesh size alone. The number of species dropped by ronment (Bonsdorff et al. 1995). When the 1.0-mm mesh 42% in the Inner and Middle area, and by 25% in the was used, most juveniles (1 - 2 mm in size) were lost, and Outer area. Also, some species, such as oligochaetes and thus the whole annual recruitment could be missed from polychaetes, were sampled in significantly reduced num any analysis. It is thus impossible to make estimates on bers with the 1.0-mm mesh. Small-sized species, such as recruitment success and population structure if the larger ostracods and small polychaetes (e.g. M anayunkia aestua mesh size is used. rina), as well as juveniles of many species, were lost com The estimate of the ecological status according to the pletely when the larger mesh size was used. In other EU-WFD is done using several different indices, and words, the benthic community may look very different almost every country has developed an index of their when using different mesh sizes. When many species are own, suitable for their particular environment (Diaz et al. seemingly lost (due to large mesh size) from an area with 2004; Zettler et a l 2007; Borja et al. 2009; Josefsson et al. naturally low biodiversity, the ecological assessment and 2009; Leonardsson et al. 2009). For the Northern Baltic representation of that assemblage will be wrong. It is Sea, BBI has been developed to account for the low salin impossible to obtain a representative picture of the overall ity and low species numbers in the area (Perus et al. biodiversity and ecosystem functioning if only the most 2007). Due to the complex topography and the gradients Table 6. Main characteristics of co-lnertla analyses. Axis Covar Vari Var2 Ineri Iner2 R value (a) Coi Inner archipelago F1 2.009 4.439 0.575 19.924 0.548 0.787 F2 0.709 2.137 0.485 24.637 0.972 0.684 (b) Coi Outer archipelago F1 1.803 4.986 0.621 25.552 0.490 0.583 F2 1.127 2.802 0.514 33.653 0.878 0.783 Covar, covariance between both sets of coordinates of co-lnertla analysis; Vari, Inertia of the abundance data projected onto co-lnertla axes; Var2, Inertia of the trait data projected onto co-lnertla axes; R value, correlation between both sets of coordinates resulting from the co-lnertla analysis; Ineri, maximum Inertia projected onto axes of the simple analysis of abundance data (eigenvalues of centered PCA); Iner2, maximum Inertia projected onto axes of the simple analysis of trait data (eigenvalues of FCA). 66 Marine Ecology32 (Suppl. 1) (2011) 58-71 © 2010 Blackwell Verlag GmbH Aarnio, Mattila, Törnroos & Bonsdorff Zoobenthos as an environmental quality element Inner area F2 ------d=0i d = 1 .0 Conv belt trans lnf_middle Rev con belt trans ri'SL Shallow 1 Siphon T en tacu lis |n;_dç; p V . t i J Lecitotrophic Bivalvei M Herbivore Vermi_uns Plankta Articulate Swimmer *F* D eeP \ Uent_pel I 1 Byssus \ F3L /tplbenthic J Radula • Turbinate No trans —y i Omnivore Burrow dweller Carnivore ills free —------ Jawed Tube_dweller Verml_seg Pharynx Fig. 5. Co-lnertla ordination of stations (both mesh sizes) and of biological traits for the Inner area. Position of the trait variables (43 modalities) and some representative taxa on the F1 x F2 co-inertia plain is presented In the large plot (cf. Table 2 for variable labels). Inset ordination plot at top right: position of sites, both 1.0- and 0.5-mm mesh sizes (bold) on the F1 x F2 plain. of salinity and exposure in the archipelago areas, the BBI (Fig. 4). For the Aland archipelago, similar estimates of has different class boundaries for different areas and ecological status have been obtained using macrophytes as depth-zones (Perus et al. 2007). In the Inner and Outer biological parameters (Söderström 2008). archipelago areas, the status was somewhat better in the BTA was found to be a useful approach in the coastal shallow than in the deeper areas. In our study the ecolog areas of the Northern Baltic Sea. Our analysis - one of ical status was good or even high in all areas and depths, the first in this region (but see Boström et al. 2010) - except for one deep station in the Inner area, where the serves to highlight the issue of time consumption versus status was moderate (0.5 mm) or bad (1.0 mm). Identify sampling effort (e.g. mesh size choice) and reliability of ing the border between moderate and good is critical, as comprehensible results in management. When choosing a all water areas should have a good ecological status by less time-consuming and thereby more economical sam 2015. In our study, the proportion of sites with a good pling method (e.g. 1 .0 -mm mesh size), the ecological per ecological status was the same irrespective of mesh size. ception of a system is adversely affected. Although the However, the status was generally somewhat better when larger mesh size did not overestimate the ecological sta a 0.5-mm mesh rather than a 1.0-mm one was used. tus, the ecological functionality of the system in question Using a 1.0-mm mesh would certainly not overestimate could be wrongly interpreted if all species and their func the ecological status, and sometimes a conservative esti tional traits are not covered. However, the application of mate may be desired. As the estimated ecological status functional multi-trait analyses, such as BTA, to species using the BBI was uniformly and significantly higher level or even more properly individual level, is still time- using the 0.5-mm mesh than the 1.0-mm mesh, impor consuming, taxonomically difficult and needs further tant ecological information is lost using only the larger refinement (Albert et a í 2010). The difference in trait mesh size. On the other hand, the more conservative esti composition when using 0.5- or 1.0-mm mesh showed a mate may be valid from a management point of view significant trend, particularly in the Outer archipelago, Marine Ecology32 (Suppl. 1 ) (2011 ) 58-71 © 2010 Blackwell Verlag GmbH 67 Zoobenthos as an environmental quality element Aarnio, Mattlla, Törnroos & Bonsdorff Outer area d = 2 .0 Dm_mixing lnf_top Jawed Semi mobi Burrower Tube_dweller Burrow_dweller rbmate SM V Crawler Herbivore Raft drift Planktotrophi ired.dev Aftieulate Bivalved Detrltlvore Radula free Mqfrile Carnivore Ep benthic Surt_depositfon inf middle Vermr.se . Scavenger Bent pel Conv belt trans Swimmer No_t Siphon Pharynx vermi uns Omnivore Lecitotrophic d=0.2 Rev_con_belt_trans in# ¿ee Tube builder T e n tjc u b te : LSI Byssus Attached Sessile Fig. 6. Co-lnertla of ordination of stations (both mesh sizes) and of biological traits for the outer archipelago area. The position of the trait vari ables (45 modalities) and some representative taxa F1 x F2 co-lnertla plain are presented In the large plot (cf. Table 2 for variable labels). Inset ordination plot at bottom right: position of sites, both 1.0- and 0.5-mm mesh sizes (bold) on the F1 x F2 plain. where the proportion of small-sized species is high. A markedly affect the measurement of benthic functioning functional approach to classifying and assessing habitat of coastal areas. We also show that it is important to dif and ecosystem quality is generally agreed upon today and ferentiate between the shallow ( < 1 0 m; photic zone) and may be the most relevant for delivering ecosystem-based deeper ( > 1 0 m; euphotic zone) areas in complex archipel targets (Bremner 2008; Tillin et al. 2008). Discussion ago areas, where the shoreline is long and topography is about drawbacks and problems still concerned with the complex (Granö et al. 1999). approach have focused mainly on the operational mea sures of functioning, methods to best elucidate function Concluding Remarks ing and the extensive species-specific information required in analysis (Bremner et a í 2003, 2006a,b; Tillin Our results highlight how biological traits, in addition to et a í 2008). However, the scale on which functioning is species number and biomass, can play a key role in ana studied is also significant (Hewitt et a í 2008). As shown lyzing ecosystem structure and in assessment and classifi in this study, it is particularly essential to consider the cation of coastal systems, and for our understanding of scale at which one samples, not only the spatial scale that the complexity of ecological functioning of these systems concerns the design (e.g. between or within habitats and (Bremner 2008; Hewitt et al. 2008). within landscapes). The choice of mesh size, and thereby In conclusion, the ecological implications of using lar inclusion or exclusion of both rare and common species ger mesh sizes of the sieve ( 1 . 0 mm is recommended for in analysis, is essential to the conclusions of habitat qual the Baltic Sea in the WFD) instead of smaller is that, in ity, ecosystem stability and functioning. shallow areas, individuals of small species and their The functional consequences of sampling method on particular traits are not sufficiently sampled, i.e. they may the scale of mesh size of sieve, have not to our knowledge be lost in a functional perspective. In deeper areas, espe been thoroughly evaluated. Our findings suggest this can cially trait modalities linked to the essential process of 68 Marine Ecology 32 (Suppl. 1) (2011) 58-71 © 2010 Blackwell Verlag GmbH Aarnio, Mattila, Törnroos & Bonsdorff Zoobenthos as an environmental quality element bioturbation are lost. From an ecological point of view, a Bonsdorff E., Laine A.O., H änninen J., V uorinen I., Norkko A. broader trait composition is obtained with 0.5-mm mesh (2003) Zoobenthos of the outer archipelago waters (N Baltic sizes, generating a more reliable and complete picture of Sea) - the importance of local conditions for spatial distri ecosystem functioning. We also show that combining tra bution patterns. Boreal Environment Research, 8, 135-145. ditional monitoring for the EU WFD with a functional Borja A., Franco J., Pérez V. (2000) A marine biotic index to analysis of the benthic assemblages strengthens our ability establish the ecological quality of soft-bottom benthos to interpret environmental quality, and thus increases the within European estuarine and coastal environments. Marine precision of our advice for management purposes. Pollution Bulletin, 40, 1100-1114. Borja A., Miles A., Occhipinti-Ambrogi A., Berg T. (2009) Cur rent status of macroinvertebrate methods used for assessing Acknowledgements the quality of European marine waters: implementing the Water Framework Directive. Hydrobiologia, 633, 181-196. This study was funded by Abo Akademi University, the Boström C., Törnroos A., Bonsdorff E. (2010) Invertebrate dis Government of Aland and the Maj and Tor Nessling persal and habitat heterogeneity: expression of biological foundation. We thank Camilla Heilman for field sampling traits in a seagrass landscape. Journal of Experimental Marine and sorting of the benthic samples. Biology and Ecology, 390, 106-117. Bremner J. (2008) Species’ traits and ecological functioning in marine conservation and management. Journal of Experi R eferences mental Marine Biology and Ecology, 366, 37-47. Albert H.C., Thuiller W., Yoccoz N.G., Douzet R., Auberi S., Bremner J., Rogers S.I., Frid C.L.J. (2003) Assessing functional Lavorel S. (2010) A multi-trait approach reveals the struc diversity in marine benthic ecosystems: a comparison of ture and the relative importance of intra- vs. interspecific approaches. Marine Ecology Progress Series, 254, 11-25. variability in plant traits. Functional Ecology, 24, 1192-1201. Bremner J., Rogers S.I., Frid C.L.J. (2006a) M ethods for Blomqvist E.M., Bonsdorff E. (1986) Spatial and temporal describing ecological functioning of marine benthic variations of benthic macrofauna in a sand bottom area on assemblages using biological traits analysis (BTA). Ecological Aland, northern Baltic Sea. Ophelia, 4 (Suppl), 27-36. Indicators, 6, 609-622. Bonsdorff E. (2006) Zoobenthic diversity-gradients in the Bal Bremner J., Paramor O.A.L., Frid C.L.J. (2006b) Developing a tic Sea: continuous post-glacial succession in a stressed eco Methodology for Incorporating Ecological Structure and Func system. Journal of Experimental Marine Biology and Ecology, tioning into Designation of Special Areas of Conservation 333, 383-391. (SAC) in the 0-12 Nautical Mile Zone. School of Biological Bonsdorff E., Blomqvist E.M. (1993) Biotic couplings on shal Sciences, University of Liverpool, Liverpool: 168 pp low water soft bottoms - examples from the northern Baltic Cederwall H., Elmgren R. (1980) Biomass increase of benthic Sea. Oceanography and Marine Biology Annual Review, 31, macrofauna demonstrates eutrophication of the Baltic Sea. 153-176. Ophelia, 1 (Suppl.), 31-48. Bonsdorff E., Pearson T.H. (1999) Variation in the sublittoral Charvet S., Kosmala A., Statzner B. ( 1998) Biomonitoring macrozoobenthos of the Baltic Sea along environmental gra through biological traits of benthic macroinvertebrates: per dients: a functional group analysis. Australian Journal of spectives for a general tool in stream management. Archiv Ecology, 24, 312-326. für Hydrobiologie, 142, 415-432. Bonsdorff E., Aarnio K., Sandberg E. (1991) Temporal and Chevenet F., Doledec S., Chessel D. (1994) A fuzzy coding spatial variability of zoobenthic communities in archipelago approach for the analysis of long-term ecological data. waters of the northern Baltic Sea - consequences of eutro Freshwater Biology, 31, 295-309. phication? Internationale Revue des Gesamten Hydrobiologie, Diaz R.J., Rosenberg R. (1995) Marine benthic hypoxia: a 74, 433-449. review of its ecological effects and the behavioural responses Bonsdorff E., Norkko A., Boström C. (1995) Recruitment of benthic macrofauna, Oceanography and Marine Biology and population maintenance of the bivalve Macoma balth Annual Review, 33, 245-303. ica (L.) - factors affecting settling success and early sur Diaz R.J., Solan M., Valente R.M. (2004) A review of vival on shallow sandy bottoms. In: Eleftheriou A., Ansell approaches for classifying benthic habitat and evaluating A.D., Smith CJ. (Eds), Biology and Ecology of Shallow habitat quality. Journal of Environmental Management, 73, Coastal Waters. Proceedings of the 28th European Marine 165-181. Biology Symposium. Olsen 8c Olsen, Fredensborg: 253- Dolédec S., Chessel D. ( 1994) Co-inertia analysis: an alterna 260. tive method for studying species-environment relationships. Bonsdorff E., Blomqvist E.M., M attila J., Norkko A. (1997) Freshwater Biology, 31, 277-294. Long-term changes and coastal Eutrophication. Examples Dray S., Chessel D., Thioulouse J. (2003) Co-inertia analysis from the Aland Islands and the Archipelago Sea, northern and the linking of ecological data tables. Ecology, 84, 3078- Baltic Sea. Oceanologica Acta, 20, 319-329. 3089. Marine Ecology32 (Suppl. 1 ) (2011 ) 58-71 © 2010 Blackwell Verlag GmbH 69 Zoobenthos as an environmental quality element Aarnio, Mattlla, Törnroos & Bonsdorff Ellingsen K.E., Hewitt J.E., Thrush S.F. (2007) Rare species, Perus J., Liljeqvist J., Bonsdorff E. (2001) A long-term study of habitat diversity and functional redundancy in marine changes in the zoobenthos in the Aland archipelago - benthos. Journal of Sea Research, 58, 291-301. a comparison between 1973, 1989 and 2000. Forskningsrap- Elmgren R. ( 1989) Man’s impact on the ecosystem of the porter frân Huso Biologiska Station, 103, 1-58 (In Swedish Baltic Sea: energy flows today and at the turn of the century. with English summary). Ambio, 18, 326-332. Perus J., Bonsdorff E., Bäck S., Lax H.-G., Villnäs A., W estberg Frid C.L.J., Paramor O.A.L., Brockington S., Bremner J. (2008) V. (2007) Zoobenthos as indicators of ecological status in Incorporating ecological functioning into the designation coastal brackish waters: a comparative study from the Baltic and management of marine protected areas. Hydrobiologia, Sea. Ambio, 36, 250-256. 606, 69-79. R Development Core Team (2009) R: A Language and Environ Granö O., Roto M., Laurila L. (1999) Environment and land ment for Statistical Computing. R Foundation for Statistical use in the shore zone of the coast of Finland. Pnblicationes Computing, Vienna. ISBN 3-900051-07-0, URL http:// Institnti Geographici Universitatis Tnrknensis, 160, 1-76. www.R-project.org . HELCOM (2009) Eutrophication in the Baltic Sea - an inte Ramsay K., Kaiser M.J., Hughes R.N. ( 1998) Responses of ben grated thematic assessment of the effects of nutrient enrich thic scavengers to fishing disturbance by towed gears in dif ment in the Baltic Sea region. Baltic Sea Environment ferent habitats. Journal of Experimental Marine Biology and Proceedings, 115B, 1-148. Ecology, 224, 73-89. Helminen O. (1975) Bottenfaunan i den äländska skärgarden. Rosenberg R., Blomqvist M., Nilsson H.C., Cederwall H., Meddelanden fràn Huso Biologiska Station, 17, 43-71. Dommig A. (2004) Marine quality assessment by use Hewitt J.E., Thrush S.F., Dayton P.D. (2008) H abitat variation, of benthic species-abundance distributions: a proposed species diversity and ecological functioning in a marine new protocol within the European Union Water system. Journal of Experimental Marine Biology and Ecology, Framework Directive. Marine Pollution Bulletin, 49, 366, 116-122. 728-739. Jax K.. (2005) Function and ‘functioning’ in ecology: what Rosenberg R., Magnusson M., Nilsson H.C. (2009) Temporal does it mean? Oikos, 111, 641-648. and spatial changes in marine benthic habitats in relation to Josefsson A.B., Blomqvist M., Hansen J.L.S., Rosenberg R., the EU Water Framework Directive: The use of sediment Rygg B. (2009) Assessment of marine benthic quality change profile imagery. Marine Pollution Bulletin, 58, 565-572. in gradients of disturbance: comparison of different Scandi Rumohr H., Bonsdorff E., Pearson T.H. (1996) Zoobenthic navian multi-metric indices. Marine Pollution Bulletin, 58, succession in Baltic sedimentary habitats. Archive of Fishery 1263-1277. and Marine Research, 44, 179-214. Leonardsson K., Blomqvist M., Rosenberg R. (2009) Theoreti Segerstrâle S.G. (1962) Investigations on Baltic populations of cal and practical aspects on benthic quality assessment the bivalve Macoma balthica (L.) II. What are the reasons according to the EU-Water Framework Directive - examples for periodic failure of recruitment and the scarcity of Maco from Swedish waters. Marine Pollution Bulletin, 58, 1286- ma in deeper waters of the inner Baltic?. Societas Scientia- 1296. rum Fennicae Commentationes Biologicae, 24, 1-26. Norkko A., Bonsdorff E. (1996) Population responses of Söderström S. (2008) Testing of classification methods for coastal zoobenthos to stress induced by drifting algal mats. coastal waters at Aland Islands according to the EU Water M arine Ecology Progress Series, 140, 141-151. Framework Directive - Chlorophyll a and soft-bottom vege Olafsson E.B. (1989) Contrasting influences of suspension- tation. Forskningsrapporter frân Huso Biologiska Station, 121, feeding and deposit-feeding populations of Macoma balthica 1-85 (In Swedish with English summary). on infaunal recruitment. Marine Ecology Progress Series, 55, Stachowicz J.J., Fried H., Osman R.W., Whitlatch R.B. (2002) 171-179. Biodiversity, invasion resistance and marine ecosystem func Olff H., Pegtel D.M., van Groenendael J.M., Bakker J.P. (1994) tion: reconciling pattern and process. Ecology, 83, 2575- Germination strategies during grassland succession. Journal 2590. o f Ecology, 82, 69-77. Thrush S.F., Schneider D.C., Legendre P., W hitlatch R.B., Pearson T.H. (2001) Functional group ecology in soft-sedi- Dayton P.K., Hewitt J.E., Hines A.H., Cum mings V.J., ment marine benthos: the role of bioturbation. Oceanogra Lawrie S.M., G rant J., Pridm ore R.D., Turner S.J., McArdle phy and Marine Biology, 39, 233-267. B.H. (1997) Scaling-up from experiments to complex Pearson T.H., Rosenberg R. (1978) Macrobenthic succession in ecological systems: where to next? Journal of Experimental relation to organic enrichment and pollution of the marine Marine Biology and Ecology, 216, 243-254. environment. Oceanography and Marine Biology, 16, 229- Thrush S.F., Hewitt J.E., Cum mings V.J., Green M.O., Funnell 311. G.A., Wilkinson M.R. (2000) Improving the generality of Perus J., Bonsdorff E. (2004) Long-term changes in macrozoo- field experiments: the interaction of processes operating over benthos in the Aland archipelago, northern Baltic Sea. Jour different spatial scales on intertidal sandflats. Ecology, 81, nal of Sea Research, 52, 45-56. 399-415. 70 Marine Ecology32 (Suppl. 1) (2011) 58-71 © 2010 Blackwell Verlag GmbH Aarnio, Mattila, Törnroos & Bonsdorff Zoobenthos as an environmental quality element Tillin H.M ., Rogers S.I., Frid C.L.J. (2008) Approaches to clas Yachi S., Loreau M. ( 1999) Biodiversity and ecosystem produc sifying benthic habitat quality. Marine Policy, 32, 455-464. tivity in a fluctuating environment: the insurance hypothe Townsend C.R., Hildrew A.G. (1994) Species traits in relation sis. Ecology, 96, 1463-1468. to a habitat templet for river systems. Freshwater Biology, 31, Zettler M.L., Schiedek D., Bobertz B. (2007) Benthic biodiver 265-275. sity indices versus salinity gradient in the southern Baltic Warwick R.M., Somerfield P.J. (2008) All animals are equal Sea. Marine Pollution Bulletin, 55, 258-270. but some are more equal than others. Journal of Experimen tal Marine Biology and Ecology, 366, 184-186. Marine Ecology32 (Suppl. 1 ) (2011 ) 58-71 © 2010 Blackwell Verlag GmbH 71 anmarine evolutionary perspective ecology O » Marine Ecology. ISSN 0173-9565 ORIGINAL ARTICLE Assessment of benthic ecosystem functioning through trophic web modelling: the example of the eastern basin of the English Channel and the Southern Bight of the North Sea Clément Garcia1,2,3, Pierre Chardy4, Jean-Marie Dewarumez1,2,3 & Jean-Claude Dauvin5 1 Université de Lille Nord de France, Lille, France 2 Université Lillei, LOG, Wimereux, France 3 CNRS, UMR8187, Wimereux, France 4 Station Marine d'Arcachon, Université de Bordeauxl, UMR S80S EPOC-OASU, Arcachon, France 5 Université de Caen Basse Normandie, Laboratoire Morphodynamique Continentale et Côtière, UMR CNRS 6143 M2C, Caen, France Keywords A bstract English Channel; functional unit; North Sea; soft-bottom communities; trophic web. Benthic organisms appear to be accurate proxies for assessing coastal ecosystem structures and changes due to climatic and anthropogenic stresses. Functional Correspondence studies of benthic systems are relatively recent, mainly because of the difficul Clément Garcia, centre for Environment, ties in obtaining the basic parameters for each benthic compartment {i.e. detri Fisheries and Aquaculture Science, Lowestoft tus, bacteria, meiofauna and macrofauna). Our study focuses on the eastern Laboratory, Pakefleld Road, Lowestoft, basin of the English Channel and the Southern Bight of the North Sea. Trophic Suffolk, NR33 OHT, UK. E-mail: [email protected] web modelling was used to assess the functioning of the three main benthic community assemblages. To test and assess the relative importance of factors Accepted: 15 December 2010 assumed to influence trophic structure (geographical environment and sedi mentary particle size distribution), the study area was subdivided into divisions dot 10.1111/j. 1439-0485.2011,00428.x defined a priori according to the two main structural factors of community dis tribution; geographic distribution and sedimentary patterns. Then, a steady state trophic model utilising the inverse method was applied to a diagram composed of eight compartments, including detritus, bacteria, meiofauna, mac robenthos and fish. For each compartment, six physiological parameters were assessed, based on our own data, empirical relationships and literature data. This method allowed estimation of the flux of matter and energy within and between the units of the benthic system and assessment of the amount of tro phic energy stored in these units (available mostly to fish). Our results showed that suspension-feeders control most of the matter transfer through the macro- benthic food-web, except in the fine sand community, where deposit-feeders play a dominant role. The results also showed that, whatever the geographic area, trophic structure is strongly linked to the sedimentary conditions. As ben thic communities are connected through hydrodynamics, a model of the entire eastern basin of the English Channel would appear to be acceptable. Elowever, the main sediment types must be taken into account when establishing rela tionships between the functional units. 72 Marine Ecology32 (Suppl. 1) (2011) 72-86 © 2011 Blackwell Verlag GmbH Garcia, Chardy, Dewarumez & Dauvin Assessment of benthic ecosystem functioning Introduction was done using the quantitative dataset for a large spatial area in the eastern basin of the English Channel. Ecosystem functioning has become one of the main fields of interest in marine ecology (see Hooper et al. 2005 for a review). According to Christensen et al. (1996), ecosystem Material and methods functioning includes three main phenomena: ecosystem Study site properties {i.e. the different functional compartments of an ecosystem and the rates of the processes that link the com The area studied is the eastern basin of the English Chan partments together), ecosystem goods {i.e. the direct mar nel and the southern part of the North Sea, called the ket values of an ecosystem) and ecosystem services {i.e. the Southern Bight. This epicontinental sea is a shallow water direct or indirect benefits that ecosystems provide to zone (maximum 50 m) that is subjected to a variety of humans). Marine benthic communities were initially hydrodynamic forces. The tidal range in the area is high, assessed using qualitative methods that identify the taxo reaching about 9 m on the French coast of the Bay of nomic composition of the community. Although such Somme (Salomon & Breton 1991). The tidal current methods highlight environmental stress (Bilyard 1987) such velocities are highly variable, usually being stronger near as resistance to anthropogenic disturbances (Pearson & the French coast than the English coast (Salomon & Bre Rosenberg 1978), it is quite difficult to obtain information ton 1991). The water generally moves from the English about the functioning of the ecosystem from these methods Channel to the North Sea, although a long period of (W arwick et al. 2002). However, other methods based on strong easterly winds can reverse this trend (Salomon & quantitative examination have been developed to better Breton 1991). These patterns are also modified by coastal investigate the function of benthic invertebrates in the geography and the presence of three estuaries (the Seine ecosystem. These new tools include work on mesocosms estuary, the Somme estuary and the Scheldt-Rhine-Meuse (Solan et al. 2003), biological traits analysis (Bremner et al. estuary complex). Local hydrodynamics creates particular 2006) or trophic web modelling (Chardy & Dauvin 1992). structures, such as the gyres that retain the water masses The trophic web modelling tool appears to be essential for in a restricted zone near the Barfleur Cape in the north synthesising data, developing theories and discriminating western part of the Bay of Seine and the Isle of Wright. between alternative competing explanations of how The particle size of the sediment is strongly correlated ecosystems function (Underwood 1990, 1996). Among the with the hydrodynamics described above, with a sedimen different trophic web models already developed, steady- tary gradient (Fig. 1) ranging from coarse sediments in state/dynamic-process models provide the most explicit the middle of the Dover Strait to fine sediments in the representation of trophic interaction (Whipple et a í 2000), area’s bays and estuaries (Larsonneur et al. 1982). This particularly the steady-state inverse model (Chardy 1987; differential sedimentation leads to a bio-sedimentary gra Vézina & Platt 1988; Vézina 1989). dient from pebbles and gravel to fine sand in the places The benthic communities in the eastern English Chan where five main communities have been identified previ nel have been widely studied using both qualitative ously (Cabioch & Glaçon 1975, 1977; Cabioch et a í descriptive methods (Dauvin 1997a) and quantitative 1978): (i) the pebble and gravel community with sessile analyses (Kaiser et al. 1998; Ellien et al. 2000; Newell epifauna and Ophiothrix fragilis (Echinodermata); (ii) the et al. 2001). However, few studies have addressed trophic coarse sand community with Branchiostoma lanceolatum relationships in this area, with those that have, focusing (Cephalochordata); (iii) the fine-to-medium clean sand on local scales only (the Bay of Somme, Rybarczyk et a í community with Ophelia borealis (Polychaeta); (iv) the 2003, and the Bay of Seine, Rybarczyk & Elkaim 2003). In muddy fine sand community with Abra alba (Mollusca); this context, there is a need for large-scale assessment of and (v) the ‘muddy heterogeneous’ community with a energy flow and trophic structure of the system, which mix of species from the pebble and gravel community, can be well approximated using trophic web modelling. the coarse sand community and the muddy fine sand The main objective of this study was to test whether the community. Additionally, the offshore English Channel is trophic structure of benthic communities, through the rela mainly composed of coarse sand to pebble substrates, tionships among benthic invertebrates, depends mainly on whereas the fine sand is confined to bays and the littoral geography or on sediment type. Both of these factors are fringe. considered to have a strong correlation with hydrodynamic patterns, which influence the organisation of the benthic Sampling strategy communities in the area (Dauvin 1997a). We utilised the inverse model simulation technique for this purpose, which Macrofaunal material was collected between 2006 and estimates carbon flows through the benthic ecosystem; this 2008 in an area ranging from the eastern basin of the Marine Ecology32 (Suppl. 1 ) (2011 ) 72-86 © 2011 Blackwell Verlag GmbH 73 Assessment of benthic ecosystem functioning Garcia, Chardy, Dewarumez & Dauvin 5 0 1 0 0 Km Pebbles ESI Gravels (23 Medium sand £•£] Gravely sand I® Muddy fine sand Fig. 1. Distribution of superficial sediment in English Channel the English Channel (from Larsonneur et al. 1982). Representation of the six geographic France and sediment divisions of the eastern part of ■ Gravei and Pebbles the English Channel and the Southern Bight A Coarse sand of the North Sea. NS, North Sea; DS, Dover O Fine sand Strait; BS, Bay of Seine. English Channel to the southern bight of the North Sea of Seine (see Ghertsos 2002 and Dauvin 8 c Ruellet 2008 (from 0° longitude to the Franco-Belgian border). The for details). The second covers the French coast from the main objective was to update the benthic invertebrate Pointe d’Ailly to the Belgium border (see Desroy et a l knowledge 30 years after the first extensive benthic sam 2003 for details). Thus, a total of 403 quantitatively sam pling in the English Channel (during the RCP M anche pled sites from the Bay of Seine to the Southern Bight survey, Cabioch 8 c Glaçon 1975, 1977; Cabioch et a l were available for trophic web analysis (Fig. 1). All mac- 1978). The secondary objective was to supply the first rofaunal/sediment material was gathered and processed quantitative description of the benthic communities over using the same basic methods. the entire area, to allow assessment of the trophic struc ture of the benthic communities and aid the complex Modelling strategy planning and decision-making required for managing anthropogenic pressure in this dynamic area (Martin Many trophic web models have been applied to very large et a l 2009). spatial areas, such as the North Sea (Mackinson 8 c Daska- Quantitative samples were taken with a 0.25-m2 lov 2007), the Irish Sea (Fees 8 c Mackinson 2007) and the Hamon grab (two samples for the macrobenthic fauna Baltic Sea (Harvey et a l 2003). To test whether such a and one sample for sediment). As our focus was on the large-scale model would be appropriate for the eastern macrobenthic component of the total benthic biomass, basin of the English Channel, we examined trophic struc samples were sieved through a 2 -mm mesh (which allows ture at different spatial scales. To accomplish this, an a more than 95% of macrobenthic biomass to be retained, priori division of the area was made based on the two see Ghertsos 2 0 0 2 ). The samples were then sorted, and main factors assumed to influence the benthic communi the organisms were identified to species level where feasi ties and their organisation (taking into account that the ble. Biomass was determined with the ash-free-dry-mass sediment factor is influenced by hydrodynamics which, in method to reduce the variation within and between spe turn, is influenced by the geographic factor). The area cies due to gut content (van der Meer et a l 2005). was split into three geographic divisions (Bay of Seine, To extend and improve the spatial resolution of the Dover Strait and North Sea) and three main sediment study area, two other quantitative databases were also divisions following the dominant sediment types in the included in this study. The first one covers the entire Bay area (i.e. gravel and pebbles, coarse sand and fine sand) 74 Marine Ecology32 (Suppl. 1) (2011) 72-86 © 2011 Blackwell Verlag GmbH Garcia, Chardy, Dewarumez & Dauvin Assessment of benthic ecosystem functioning (Fig. 1). Following Kröncke et al. (2004) and Kröncke unknown parameters are the physiological parameters (2006), the expression ‘geography here refers to all envi (e.g. ingestion rate, egestion rate). These parameter values ronmental variables acting over a spatial area (e.g. hydro are taken from data reported in the literature and/or dynamism, freshwater inputs, food supply and quality). from empirical relationships. From these data, a mean Simulations were then run for the three geographic divi value is calculated for each parameter, and the model sions and the three sediment divisions, and the outputs allows this value to move between an upper and a lower from these simulations were compared. Simulations were bound, which are determined by the confidence interval also run for each sediment type within the three geo of the mean. The relationships linking these bounds are graphic divisions. Thus, there were three sub-divisions for mainly the trophic preferences. This method yields the each division, except for the Bay of Seine, which lacks annual average of the carbon flows connecting the differ fine sand communities, resulting in a total of eight sub ent functional compartments. divisions in the study area. The functional diagram (Fig. 2) of the benthic com partment needs to be simple enough to fit the different benthic communities but also accurate enough to express Model formulation and principles the knowledge about the benthic compartments in the Trophic web structure was assessed using a trophic eastern basin of the English Channel and the Southern inverse model originally introduced by Vézina & Platt Bight of the North Sea. We define the functional com (1988) for the pelagic food-web of the English Channel partments based on the available knowledge about the and the Celtic Sea, which since then has been used by a trophic compartment and on the size of the benthic number of authors in various marine ecosystems (Chardy organism (this criterion is mainly used to divide the ben et al. 1993a,b; Niquil et al. 2001; Leguerrier et al. 2003). thic organisms between macrofauna and meiofauna com The inverse method-based steady state model is a diag partments). This is a steady-state model, an annual nostic method that, although it does not incorporate a average representation of biomasses and flows. Temporal temporal dimension, can provide a comprehensive biomass variations are not considered (d X /d t) = 0. The description of the general trophic structure of an ecosys steady-state hypothesis is expressed by the general equa tem. Our model is composed of eight biotic and abiotic tion: compartments and is used to estimate the carbon flows d X i \ \ resulting from secondary benthic production (Chardy & - ¿ f = J 2 Fp - J 2 Ftp = ° Dauvin 1992). The information required for direct esti mation of most of the flows in such a food-web model is w here Fjú sum of the flows going from 'ƒ to ‘ 1’ (sum either not available or is very difficult to obtain. Thus, it of the inputs, consumption of Y); Fip> sum °f the seems more appropriate to use the flow balance principle, flows going from ‘ 1’ to the other compartment ‘p’ and to as the inputs are equal to the sum of the outputs and the the general outputs of the system (sum of outputs: predi- rate of the biomass standing stock is under steady-state tory mortality, non-preditory mortality, egestion and conditions (Vézina & Platt 1988). respiration). This kind of inverse problem can be encountered in all At the scale of compartments, the steady-state is research fields where the number of observations is less expressed by the balance of the processes: than the number of parameters that need to be known in order to describe the system (Vézina & Platt 1988). To cLXi m — = ^2(Ii- Xi. Cji. (1 - Ei)) - (M í + R i)X i solve this kind of inverse problem, Tarantola & Valette d t i = i (1982) have proposed three fundamental conditions: n • Having a given state of information about the values - x p■ °p ï = 0 of the observed parameters. p= i • Having information about the unknown parameters (we assume that an a priori decision is made about where: j = 1 ...m , is tn number of prey available for ‘1’; the unknown parameter values associated with an p = 1 ...n , is n number of predators of ‘ 1’; Ii is annual interval of confidence). ingestion rate of ‘1’; X i is the biomass of the compartment • Having the necessary information about the theoreti ‘1’; Cji is the feeding preference of the compartment ‘1’ cal relationships between known data and unknown for the resource 'ƒ; Ei is the annual egestion rate of ‘1’; parameters. M i is the annual non-predatory mortality rate of ‘1’; and In trophic web studies, the known data are estimations Ri is the annual respiration rate of ‘i\ of the biomass standing stock in each benthic compart The annual production/biomass (P/B) ratios for all m ent (i.e. the biomass values of each compartment); the seven compartments, from which the physiological param- Marlne Ecology32 (Suppl. 1 ) (2011 ) 72-86 © 2011 Blackwell Verlag GmbH 75 Assessment of benthic ecosystem functioning Garcia, Chardy, Dewarumez & Dauvin P.O.M. Flow (pelagic inputs, phytobenthos...) Detritus XI Suspension-feeder X4 Deposit-feeder Bacteria & mixed X2 X5 Meiofauna X5 Omnivore X7 Carnivore X6 Fig. 2. Functional diagram of the benthic Fish ecosystem in the eastern part of the English X8 Channel and the southern Bight of the North Sea. Trophic fluxes: faeces + non-predatory mortality: Respiration. eters were derived, were found in the literature on either that occurs in this group. The carnivore P/B value (0.65) existing trophic web models (Améziane et al. 1996; Leguer- was the calculated mean of the polychaete predator P/B rier et al. 2003; Mackinson & Daskalov 2007) or field stud value given by George & Warwick (1985) for a hard-bot ies (Warwick et al. 1978; Warwick 1980; Warwick & tom community. The omnivore compartment has a value George 1980; Vranken & Heip 1986; Vranken et al. 1986). close to carnivores (0.7). This value was also chosen based None of the P/B values was obtained with empirical data. on the mean omnivore values calculated from those given As only the largest benthic organisms (2-mm sieve) were by George & Warwick (1985) (between 0.2 and 0.4) and considered in this study, juveniles and small species were those used by Mackinson & Daskalov (2007) (0.55 for the ignored for the biomass estimation. All the large species ‘crab’ compartment and 3 for the ‘shrimp’ compartment). were assumed to be adults and we therefore decided to take The respiration rate was derived from the allometric the smallest P/B found for each compartment. For bacteria, equation developed by Schwinghamer et al. (1986): log 10 we chose a value of 3700, which is an intermediate value Ra = 0.367 + 0.993 log 1 0 Pa, where Ra is the annual res between that of 9470 proposed by Mackinson & Daskalov piration rate (in kCal-year-1) and Pa is the annual pro (2007) and 167 proposed by Améziane et al. (1996). The duction rate (in kCal year-1). The egestion parameter was P/B of 0.95 for the deposit-feeders and 0.8 for the determined from values found in the literature. Ingestion suspension-feeders are the means of the P/B for all the rates were deduced from P/B values, assuming that eges deposit-feeder species and the P/B of bivalves, as calculated tion and respiration are known. The biomass of the four by several authors working in highly different environ macrobenthic compartments - deposit-feeder & mixed, ments (Warwick & Price 1975; Warwick et al. 1978; suspension-feeder, carnivore and omnivore - was taken Warwick 1980; Warwick & George 1980; George & War from the present work. The bacteria biomass was consid wick 1985). The deposit-feeder value agrees with the P/B of ered to be similar to values provided by Améziane et al. the compartment ‘small infauna (polychaetes)’ proposed (1996) for the Bay of Morlaix in the Western English by Mackinson & Daskalov (2007). The lowest meiofauna Channel and the meiofauna and fish biomasses were con P/B value (9) proposed in the literature by Gerlach (1971), sidered to be similar to those in the North Sea (Mackin was chosen to illustrate the auto-predation phenomenon son & Daskalov 2007). These three compartments 76 Marine Ecology32 (Suppl. 1) (2011) 72-86 © 2011 Blackwell Verlag GmbH aie clg 2 Spl 1 (01 728 21 BakelVra GmbH Verlag Blackwell 2011 © 2-86 7 (2011) 1) (Suppl. 32 Ecology Marine Dauvin & Dewarumez Chardy, Garcia, Table 1. Main characteristics of the functional compartments of the different divisions and sub-divisions in the eastern part of the English Channel and the Southern Bight of the North Sea and values of biotic rates used in the simulations. OOOOOOCTiOOOOOOCOOOOOOOCTiOOOOOOCriOOO O O i r C o O o O o O o O o O o O i o T o C o O o o O o O o O o O o O o O o C o O o O o O o O o O o O i o T o C o O o O o O o O o O o o iO o T o C o O O O ooooooooooooooooooooooooooooooooooo C M C M C M C M 0 0 0 ooooooooooooooooooooooooooooooooooo 0 - ^ 0 0 0 0 O O O O 0 l O O 0 O O O 0 O O O 0 O O O O 0 O O C 0 O Oo^-^-oo^-mrvOmLnommoo O 0 O O O 0 O O O O 0 O O O 0 O O C 0 T i O O 0 O 0 0 ooooooooooooooooooooooooooooooooooo 0 0 0 0 0 0 0 0 0 0 0 0 O O O C T i O O O O O O C O O O O O O O C O O O O O O O Co T i O O O O O O C r i O O ö O ö ö ö ö ^ -o ö ^ ö -o ö ^ ro ^ o o coco\r^mcooh OOOOOOOOOOCMOOOOOOOOOOOOOOOOOOOOOOOO CTiCTiO^oor^'xtcOLnLnLnocornLnLncx.criOcOLOO^LnLno C T iCT O O i r s l o L N n o Lfl C o o M o o fÛ x m 'xt •xt m ö ^ m•xt r--. o o o o o o o m o o o o o o o o o lo n co Ln cm c o o m m mcm om O ï om m o c m c o Lm n - x t mm m CM CM O O m r - i — i —o oo - . o <— o r-x. o seseto eti csse functioning ecosystem benthic of Assessment r - x m LO r O - x O o O ' x O t O r - x o o o o o o o o o o o o o o o cm lo lo lo lo cm o r-x. o co CM CM CM CM CM CM CM m r-x. m o o o r-x. o m C MC MC MC M m r--. r--. o ÖÖ Ö Ö Ö Ö Ö Ö Ö in o o o o o o o o o o o o t ri rn N rsi tú ^ lo CM 'xt Ln O ^ ^ cm lo o \ ÛÉ CD II í t II c I—Q- Cû C CT >•> c CL. -C cu CU u_ cr> o> Q i/i o cu Il ^ 3 dö Û L u Ci B ^ cu /1 1/1 1/1 m 11 I I II II i' yi ty U - U o o II II 77 Assessment of benthic ecosystem functioning Garcia, Chardy, Dewarumez & Dauvin (bacteria, meiofauna and fish) had the same biomass in • X3. Deposit-feeders and Mixed. This group includes each of the models applied for each division and sub strict deposit-feeders that feed only or predominantly division of our study area. The physiological parameters on detritus at the sediment layer, but also organisms are presented in Table 1. that are able to feed as either deposit-feeders or sus Diet preference values were determined according to pension feeders. No distinction has been made three different sources of information. Diet data were between sub-surface and surface deposit-feeders, as preferentially taken from experimental studies of gut con they all feed on detritus and bacteria. tents for particular species (e.g. Fauchald & Jumars 1979; • X4. Suspension-feeders. This group is mainly com Langdon & Newell 1990) or groups of species (Lopez posed of filter-feeding bivalves that feed in the et al. 1989). Data from stable isotope studies were also water-sediment interface. They feed more on fresh used, where available (e.g. Le Loc’h & Hily 2005; Carlier matter than the deposit-feeder group does (e.g. phy et al. 2007). Finally, some remaining missing data were toplankton, microphytobenthos, fresh detritus). They also taken from previous published trophic web models can also feed on bacteria bound to particles. Langdon (e.g. Chardy & Dauvin 1992; Améziane et al. 1996; & Newell (1990) have estimated that the bacteria Leguerrier et al. 2003). could represent 3.5 and 25.8% of carbon needs in To assess and compare the functioning of each sector oysters and mussels, respectively. As phytoplankton is of the eastern basin of the English Channel and the not represented in our model, the trophic preference Southern Bight of the North Sea, simulations were based of suspension-feeders was integrated by including a on the same initial mean values for ingestion, egestion, preference for benthic bacteria and a lesser egestion. P/B, non-predatory mortality and initial matter input. • X5. Meiofauna. Nematodes are often the most abun However, the model was allowed to select each parame dant organisms in the permanent meiofauna (Boaden ter’s value within the confidence interval of the mean for 2005). Most of the meiofaunal organisms feed on each simulation. The values of biomass and respiration detritus and bacteria, but some of them are also car were fixed for each simulation (i.e. the model was not nivores, including cannibalism. allowed to change them). The simulation outputs for each • X 6 . Carnivores. This group is composed of predators division and sub-division were the average annual carbon and carnivores. They only feed on living or almost flows per square metre that link all functional compart living organisms; motile nemertean and polychaete ments. predators are the most representative organisms in this group. • X7. Omnivores. This group consists of species that Compartment status have an opportunistic feeding mode. These species We sub-divided the benthic ecosystem into eight com will always prefer to feed on fresh material, but they partments defined by feeding mode and size. The general can also feed as scavengers on dead bodies and detri structure of the diagram (Fig. 2) uses the main compo tus. This group is mainly composed of decapods and nents previously proposed by Chardy et al. (1993a,b), to some gastropods. which we added the omnivore and the demersal fish • X 8 . Fish. The only vertebrate compartment of this compartments. Omnivores have been separated from car trophic web, this group is composed of carnivorous nivores because, as scavengers that recycle organic matter, demersal fish that feed on every macrobenthic com they have a different function in the ecosystem. The fish partment in the model. compartment was added to ‘close’ the benthic trophic web. Results • Fi. The initial flow. This is the necessary amount of organic matter for the whole ecosystem to function. The sum of the average biomass values of benthic inverte It translates as the net sedimentation of pelagic detri brates in the whole study area is 5.253 gOm-2, composed tus (dead phytoplankton cells, faeces) that can be mostly of suspension-feeders (77%), then deposit-feeders used by benthic organisms. and mixed ( 8 %), with omnivores and carnivores having • XL Detritus. This is an inactive compartment. It similar proportions (7.5 and 6 %, respectively) and meio appears to be a cross-road from which the carbon is fauna having the smallest proportion (1.5%). passed to higher levels, receiving matter from outside To assess the trophic structure of each division and sub the system (Fi) as well as from inside the trophic division of our study area, the first step was to identify the web itself (i.e. egestion, non-predation mortality). preferred trophic pathway (i.e. the main compartments • X2. Bacteria. Benthic bacteria are associated with through which most of the carbon will transit). The fish particles of detritus. compartment (X 8 ) is always the most important predator 78 Marine Ecology32 (Suppl. 1) (2011) 72-86 © 2011 Blackwell Verlag GmbH Garcia, Chardy, Dewarumez & Dauvin Assessment of benthic ecosystem functioning P.O.M. flow Fig. 3. Results from the simulation of annual carbon flows (gGrrr2-y~1) showing a representation of (A) TP1, the trophic path way In the total area; (B) TP2, the trophic pathway in the Coarse sand Suspension-feeder division; and (C) TP3, the trophic pathway In the Fine sand division. Trophic fluxes: faeces + non-predatory mortality; main pathway in the carbon transfer between a prey compartment and all of its predator compartments; R, respiration. Deposit-feeder in each of the macrobenthic compartments. To obtain a more accurate picture of the carbon flow through the macrobenthic compartments, we did not take the fish compartment into account when determining the trophic pathways; fish were only considered as the top predators, which came into play following one of the two last macro benthic compartments {i.e. carnivore and omnivore). The preferential trophic pathway for the whole study area begins in the detritus compartment. Most of the car bon in this first compartment is absorbed by the bacteria P.O.M. flow in the second compartment (87.1% of the uptake); the suspension-feeders feed mainly on bacteria (72.4%). The main predators of the suspension-feeder compartment are Suspension-feeder the fish (59.4%), although the main benthic invertebrate’s predators are the omnivores (21.8%). Thus, the total area was considered to have a ‘suspension-feeder/omnivore’ Deposit-feeder trophic pathway (Fig. 3A). Comparison of the divisions The sum of the average biomass values of benthic inverte brates from each trophic compartment in the different divisions of our study area reached values of 0.462, 0.794 and 0.508 gC-m - 2 for the three geographical divisions (North Sea, the Dover Strait and the Bay of Seine, respec tively) and 0.419, 0.917 and 0.761 gC-m - 2 for the three main sediment divisions (gravel and pebbles, coarse sand and fine sand, respectively). P.O.M. flow In the geographic divisions, suspension-feeders were always dominant in terms of biomass proportions, rang ing from 33.3% in the North Sea to 73.7% in the Dover Suspension-feeder Strait. This pattern was also observed in the sediment divisions, where the biomass proportions of the suspen sion-feeders were always between 59.2% in the gravel and Deposit teedei pebbles and 69.1% in the fine sand. The highest biomass proportion for the deposit-feeders and mixed was found in the North Sea geographic division, with 18.6%, and in the fine sand sediment division, with 10.9%. Carnivores and omnivores have quite similar proportions in the Dover Strait (4.4 and 5%, respectively) and in the North Sea (16.5 and 13.6%, respectively), but the carnivores clearly dominate (20.6%) the omnivores (9.8%) in the Bay of Seine. However, all the sediment divisions have quite similar proportions for these two compartments ( 8 . 1 % for the carnivores and 8 .6 % for the omnivores in Marine Ecology32 (Suppl. 1 ) (2011 ) 72-86 © 2011 Blackwell Verlag GmbH 79 Assessment of benthic ecosystem functioning Garcia, Chardy, Dewarumez & Dauvin Table 2. (a) Compartments for each of the three preferential trophic biomass in North Sea is much higher (32.2%) than in the pathways identified, (b) Preferential trophic pathway for each division Dover Strait and the Bay of Seine (5.2 and 4.5%, respec and sub-division of the Eastern English Channel and the Southern tively). However, the carnivore and omnivore compart Bight of the North Sea. ments don’t seem to follow a particular pattern, except (a) N° preferential trophic pathway for the three North Sea sub-divisions, where the biomass proportions in both compartments are systematically TP1 Detritus-Bacteria-Suspension-feeder-Omnivore- higher than in the other five sub-divisions. Fish TP2 Detritus-Bacteria-Suspension-feeder-Carnivore- The dominant trophic pathway for all eight sub-divi Fish sions is one of the three previously identified trophic TP3 Detritus-Bacteria-Deposit-feeder and Mixed-Car- pathways for the divisions. The trophic pathway identified nivore-Fish for the geographic sub-divisions of the gravel and pebbles sediment type changes from one area to another. In the (b) sedimenftsite Bay of Seine Dover Strait North Sea whole Bay of Seine, the trophic pathway is TP2, which is the gravel and pebbles TP1 TP2 TP3 TP2 same pathway as the whole Bay of Seine; in the Dover coarse sand TP1 TP1 TP1 TP1 Strait, the trophic pathway is TP1, which is the same fine sand - TP3 TP3 TP3 pathway as the whole Dover Strait; and in the North Sea, whole TP1 TP2 TP2 TP1 the trophic pathway is TP3, which is the same pathway as the whole fine sand (Table 2b). The trophic pathways in the geographic sub-divisions the gravel and pebbles, 1.9 and 4.4% in the coarse sand of coarse sand sediment and fine sand sediment are much and 4.7 and 4.7% in the fine sand). more stable. The geographic sub-divisions of the coarse Like the trophic pathway for the whole study area, the sand follow the TP2 pathway, whatever the geographic Bay of Seine had a ‘suspension-feeder/omnivore’ path site considered; in other words, the trophic pathway is way. However, the Dover Strait and the North Sea sites the same for the entire coarse sand division. Similarly, the had a different type of trophic pathway, with both of trophic pathways in all the geographic sub-divisions of them having a ‘suspension-feeder/carnivore’ pathway. the fine sand sediment are the same - TP3 - which is also Each of the sediment divisions had a different type of the trophic pathway for the entire fine sand division trophic pathway (Table 2a). The gravel and pebbles divi (Table 2b). sion had the same ‘suspension-feeder/carnivore’ trophic pathway as the whole study area, the Dover Strait and the D iscussion North Sea, denoted in this paper as TP1. The coarse sand division had the same ‘suspension-feeder/omni- Food-web studies and trophic network analysis provide vore’(Fig. 3B) trophic pathway as the Bay of Seine, powerful tools for identifying the global functional prop denoted TP2. Finally, the fine sand division had the path erties of benthic communities (Chardy et al. 1993a,b). way that differed the most, mainly because the principal New techniques for describing and quantifying the flows primary consumers switch from being suspension-feeders of organic matter between compartments have been to being deposit-feeders and mixed. This switch gives this developed at the same time as numerical methods such as division a ‘deposit-feeder and mixed/carnivore’ (Fig. 3C) trophic models. One of these numerical methods for flow trophic pathway, denoted TP3. network assessment in trophic web is the trophic inverse model used in this study. This model is based on an underlying inverse method (Chardy 1987), which states Comparison of the sub-divisions that the sum of inputs is equal to the sum of outputs This pattern of suspension-feeder biomass dominance was (Vézina & Platt 1988). observed in almost all the sub-divisions, except in the Since this model arrived in the late 1980s (Vézina & gravel and pebbles sub-division of the North Sea, where Platt 1988; Vézina 1989; Chardy et al. 1993a,b) it has deposit-feeders and mixed dominate with 32.2%. The been widely used in many different environments and deposit-feeder and mixed biomass in one sediment type biotic compartments, such as the French coast of Brittany for one geographic area has a similar value in the same in the Western English Channel (Chardy 1987; Chardy & sediment type for the other two areas, 13.2% (Bay of Dauvin 1992; Chardy et al. 1993a,b; Améziane et al. Seine), 13% (Dover Strait) and 13.9% (North Sea) in fine 1996), the intertidal mudflat on the French Atlantic coast sand, and 7.1% (Bay of Seine), 4.9% (Dover Strait) and (Leguerrier et al. 2003, 2004), Arcachon Bay (Blanchet 7.3% (North Sea) in coarse sand. The exception is for the 2004), the Baltic Sea (Harvey et al. 2003), the coast of gravel and pebbles, where the deposit-feeder and mixed Norway (Salvanes et al. 1992), the Mediterranean (Coli 80 Marine Ecology32 (Suppl. 1) (2011) 72-86 © 2011 Blackwell Verlag GmbH Garcia, Chardy, Dewarumez & Dauvin Assessment of benthic ecosystem functioning et a l 2006), lagoons on the coast of the Pacific Ocean lish Channel. Thanks to a large tidal range, the hydrog (N iquil et al. 2001) and the coast of the USA (Eldridge & raphical influence of large rivers and the morphology of Jackson 1993; Breed et al. 2004). the Eastern English Channel coast, hydrodynamics in the Inverse methods (i.e. ones in which the values of the Channel vary greatly and have a complex pattern (Salo unknown parameters are deduced from a set of observa mon & Breton 1991). These hydrodynamics lead to differ tions and a system model) (Tarantola & Valette 1982) ential particle-size sedimentation, with a gradient ranging seem to be totally appropriate for food-web research, as from pebbles and gravel to fine sand (Larsonneur et a l researchers in this field face the fundamental limitation 1982). that the number of independent observations of physio Depending on the sediment type, different benthic spe logical rates that can be made is far less than the number cies are able to settle and to undergo a successful meta of parameters needed to describe the whole system (Véz morphosis (Gray 1974). Previous authors have identified ina & Platt 1988). The main limitation of these methods five main bio-sedimentary structures, which have been is that they all assume the system is mass-balanced, which studied since the late 1970s (see Dauvin 1997a). For this permits only a static representation of the food-web for a study, we chose three sediment divisions; coarse sediment particular temporal scale (Pasquaud et al. 2007). In this - gravel and pebbles; intermediate sediment - coarse study, we chose an annual representation of the food- sand; and fine sediment - fine sand. These sediment divi web, mainly because field surveys were conducted over sions were associated with three geographic divisions; the 2Vi years in different seasons, making it impossible to Bay of Seine, the Dover Strait and the North Sea (Fig. 1). take the effect of recruitment into account. This first quantitative approach at such a large spatial The ecological credibility of such inverse methods scale as the Eastern English Channel allowed us to com depends on the functional unit, which needs to be homoge pare the trophic structure in the various sediment and neous in terms of the trophic types, physiological rates and geographic divisions through trophic web modelling. life cycles of its component parts (Chardy et a l 1993a,b). Flowever, inverse methods are based on the parsimony Whatever the precision level reached, determining the principle (Vézina & Platt 1988). Thus, many flows can be boundaries of the functional unit is based on varied crite underestimated or overestimated (Leguerrier et a l 2003). ria, which are always somewhat arbitrary (Warwick & Rad We used similar input parameters to ensure an identical ford 1989). An alternative approach would be to base the estimation error, so that comparisons of the divisions and models on the individual criteria, which would lead to an sub-divisions would remain possible. We also tested each increase in the number of unknown parameters. It is thus of the model outputs by randomly selecting different val necessary to accept that the food-web must be aggregated ues of particular organic matter (P.O.M.). flow entering in functional units, which are inevitably heterogeneous the system inside a confidence interval. We then (C hardy et al. 1993a,d). compared the different outputs of each division and sub Further field studies are necessary to investigate division, to make sure that they were consistent. whether a compartment can be described more accurately. Moreover, sensitivity analyses showed that outputs from For example, the deposit-feeders and mixed compartment inverse methods were robust (Marquis et a l 2007). Using seems to be the most active in the fine sand community; the preferred trophic pathway identified for each division sub-dividing this compartment into sub-surface deposit- (Table 2a), we found only slight differences among geo feeder, surface deposit-feeder sensus stricto and half graphic divisions; the trophic pathways in the Dover deposit-feeder/half suspension-feeder would provide more Strait and the North Sea were similar and were them information on the trophic structure. There is also a need selves similar to the preferred pathway for whole area for more information about the meiofauna compartment, (TP1) (Table 2b). Only the Bay of Seine was found to be which is the least well known compartment in benthic different (TP2). trophic web studies. Flowever, even if more detailed and Flowever, the trophic pathways in the sediment divi numerous compartments would increase the accuracy of sions were very different from each other (Table 2b). For our knowledge about the trophic system, the difficulties example, in the gravel and pebbles and coarse sand sedi of finding physiological parameters would multiply with ment divisions, the suspension-feeders were dominant. Of the increase in the numbers of the compartments. the 1 0 suspension-feeder species that contribute most to Benthic community structure is the result of the the biomass, five species were observed in both of the complex integration of many different factors, including communities. The presence of these five species could be abiotic, biotic and anthropogenic factors (Dauvin 1993, seen as enhancing the efficiency of the suspension-feeder 1997b). Among these factors, hydrodynamics seems to be compartment in terms of trophic web matter transfer. In the most important factor in the organization of the the higher levels of the trophic web of these divisions, the benthic invertebrates in a megatidal sea such as the Eng switch from carnivore in the gravel and pebbles sediment Marine Ecology32 (Suppl. 1 ) (2011 ) 72-86 © 2011 Blackwell Verlag GmbH 81 Assessment of benthic ecosystem functioning Garcia, Chardy, Dewarumez & Dauvin Table 3. Ten main contributive species to the mean biomass of sediment divisions for each trophic compartment. deposit-feeder and mixed B suspension-feeder B carnivore B omnivore B GP Arcopagia crassa 3.725 Laevicardium crassum 6.316 Psammechinus miliaris 0.145 Buccinum undatum 1.881 Cirriformia tentaculata 1.250 Glycymeris glycymeris 5.963 Cerebratulus spp. 0.136 Necora puber 1.394 Chaetopterus variopedatus 0.443 Lutraria spp. 5.093 Pelogenia arenosa 0.077 Sagartia troglodytes 1.263 Golfingia (Golfingia) elongata 0.273 Mytilus edulis 3.926 Asterias rubens 0.059 Urticina felina 0.837 Upogebia deltaura 0.271 Ophiothrix fragilis 3.129 Glycinde nordmanni 0.045 Atelecyclus rotundatus 0.635 Echinocardium cordatum 0.195 Paphia rhomboides 2.611 Tubulanus polymorphus 0.011 Nassarius reticulatus 0.325 Callianassa tyrrhena 0.191 Aequipecten opercularis 2.286 Lepidonotus squamatus 0.008 Ophiura ophiura 0.229 Euclymene lumbricoides 0.179 Mya arenaria 2.080 Notophyllum foliosum 0.007 Liocarcinus depurator 0.173 Golfingia (Golfingia) margaritacea 0.159 Crepidula fornicata 2.052 Harmothoe impar 0.006 Ophiura spp. 0.135 Cs Echinocardium cordatum 6.848 Ensis directus 42.358 Marphysa sanguinea 1.037 Buccinum undatum 2.153 Arcopagia crassa 6.197 Laevicardium crassum 20.484 Nephtys caeca 0.660 Urticina felina 1.025 Amphitrite johnstoni 4.110 Spisula solida 3.116 Pelogenia arenosa 0.352 Nassarius reticulatus 1.019 Cirriformia tentaculata 1.250 Solecurtus scopula 2.912 Nephtys assimilis 0.339 Sagartia troglodytes 0.753 Tellina fabula 1.014 Glycymeris glycymeris 2.580 Lumbrineris latreilli 0.283 Liocarcinus holsatus 0.634 Callianassa tyrrhena 0.703 Ensis arcuatus 2.306 Nephtys hombergii 0.245 Diogenes pugilator 0.584 Tellina tenuis 0.673 Mya arenaria 2.080 Sigalion mathildae 0.193 Nassarius reticulatus 0.533 Callianassa tyrrhena 0.703 Aequipecten opercularis 1.793 Lumbrineris fragilis 0.188 Ophiura ophiura 0.515 Echiurus echiurus 0.395 Ophiothrix fragilis 1.622 Nephtys longosetosa 0.174 Cereus pedunculatus 0.464 Callianassa subterranea 0.382 Solen marginatus 1.493 Gattyana cirrhosa 0.114 Atelecyclus rotundatus 0.346 Fs Echinocardium cordatum 12.955 Ensis directus 11.750 Nemertina 1.017 Nassarius reticulatus 0.986 Cirriformia tentaculata 1.250 Spisula solida 4.446 Nephtys hombergii 0.400 Diogenes pugilator 0.642 Corbula gibba 0.675 Acanthocardia 3.897 Sthenelais boa 0.353 Sagartia troglodytes 0.573 (.Rudicardium) tuberculata Tellina fabula 0.352 Venerupis senegalensis 2.225 Lumbrineris fragilis 0.322 Ophiura ophiura 0.456 Owenia fusiformis 0.284 Mya arenaria 2.080 Nephtys assimilis 0.318 Pagurus bernhardus 0.263 Tellina tenuis 0.262 Donax vittatus 1.927 Nephtys caeca 0.288 Thia scutellata 0.151 Macoma balthica 0.248 Ensis arcuatus 1.921 Sigalion mathildae 0.266 Liocarcinus holsatus 0.145 Callianassa subterranea 0.244 Lutraria angustior 1.827 Nephtys longosetosa 0.107 Trivia monacha 0.074 Acrocnida brachiata 0.179 Spisula elliptica 1.245 Cerebratulus spp. 0.094 Liocarcinus marmoreus 0.065 Upogebia deltaura 0.174 Lutraria lutraria 1.168 Nephtys cirrosa 0.068 Liocarcinus depurator 0.054 B = mean annual biomass of the species in g-m 2-year 1 (ash-free dry weight); GP = Gravel and pebbles; Cs = Coarse sand; Fs = Fine sand. division to omnivore in the coarse sand sediment is more In his trophic model of the benthic trophic web in difficult to explain. This could be because the omnivores Arcachon Bay, Blanchet (2004) found the same deposit- in coarse sand have a more regular biomass distribution feeder dominance in a similar fine sediment type, but am ong the 1 0 species that dominate the biomass with a large difference in the mean deposit-feeder (Table 3). biomass (0.08 gOm - 2 in our study compared with The trophic pathway also switches from suspension-fee 0.52 g O m - 2 in Arcachon Bay), probably due to the der dominance to deposit-feeder and mixed dominance presence of the seagrass Zostera noltii. Unfortunately, fur between coarse sand and fine sand, despite the fact that ther comparisons of our results with those of adjacent both these divisions have quite similar species. This marine areas are quite difficult to carry out. The models switch can be explained by the increase in biomass of one of Mackinson & Daskalov (2007) for the North Sea, and deposit-feeder species, the sea urchin Echinocardium cord Lees & Mackinson (2007) for the Irish Sea, mainly deal atum , associated with the decrease in biomass of the with fisheries management. In addition, these authors did dominant suspension-feeder species, Ensis directus, which not investigate trophic structure at the benthic level and has its biomass value divided by four in the fine sand sed the functional compartments utilised were very different. iment division. Finally, the switch between carnivore In the Marennes-Olérons, Leguerrier et a l (2003) sought dominance in fine sand to omnivore dominance in coarse to highlight the differences between the carbon flows of sand, seems to be due mostly to the absence of a signifi cultivated oysters and those of non-cultivated benthos, cant decrease in the large omnivorous cnidarian Urticina and the models made for the Western English Channel felina and molluscs Buccinum undatum and Nassarius (Chardy & Dauvin 1992; Améziane et al. 1996) were used reticulatus, which contribute most to the biomass of mainly to investigate benthic-pelagic relationships, and the coarse sand sediment division. used different compartments. 82 Marine Ecology32 (Suppl. 1) (2011) 72-86 © 2011 Blackwell Verlag GmbH Garcia, Chardy, Dewarumez & Dauvin Assessment of benthic ecosystem functioning Comparison among sub-divisions tends to show that and determine the spatial limits of the individual func the importance of the geographic factor is low. In fact, tional units, as well as their intrinsic properties. for the geographic coarse sand and the fine sand sub-divi sions, no matter what geographic location is considered, Limitations and Perspectives the trophic structure always follows the trophic pathway of the sediment type to which it belongs (Table 2b). A model is a conceptual representation of a particular However, no clear pattern appeared for the gravel and ecosystem, which has the primary advantage of gathering pebbles sediment. This lack of a clear pattern can be and summarising the current knowledge of ecosystem explained by the sampling method; all the sites were sam functioning (Barnsley 2007). This study shows that pled with the quantitative Hamon grab sampling gear inverse models are extremely useful for investigating the which reaches its functional limits in pebbles and stony trophic structure of benthic ecosystems. Our trophic bottoms. Thus, it appears that the gravel and pebbles sed inverse model allowed us to determine the specific condi iment was under-sampled. In addition, among the three tions under which the detritic compartment is utilised sediment types, gravel and pebbles appear to be the least and to identify which flows are the most important for accurate in terms of representing benthic organisation overall functioning of the system. The comparisons of the and functioning, which could explain the lack of a clear different sediment and geographic divisions were also use pattern observed in this division. It is unclear even for ful for understanding the variations in ecosystem func the coarse sand and fine sand sediment types whether the tions, for identifying general information about benthic observed features are natural or directly dependent on the functioning, and thus for providing a basis for compari sampling strategy. In their study of the effects of sampling son with other benthic ecosystems. effort on food-web structure, Martinez et al. (1999) However, using this kind of model is only possible when showed that major trophic web properties like food chain quantitative data are available, which leads to the main length appear to be robust to variation in sampling reso problem with a study such as ours; as far as the benthic lution. However, further investigations are required to compartment is concerned, inverse models can only be confirm this conclusion, particularly assessment of the used for sediment types in which a quantitative sampling other bio-sedimentary communities of the Eastern English gear can operate. Pebbles and stony sediment are excluded, Channel and the Southern North Sea {i.e. muddy-fine unless SCUBA divers are employed, which involves much sand and mud). The use of a 2-mm sieve is appropriate more work compared with the soft-bottom communities. for sampling large-sized species, but can underestimate Another inconvenience of inverse models is that they small species, including some deposit-feeders and mixed. require access to many physiological parameters that are However, more than 95% of macrobenthic biomass is hard to obtain. One solution to overcome this difficulty retained on 2-mm sieve mesh (Ghertsos 2002) and the is to derive all necessary physiological parameters {e.g. predominantly pebbles to coarse sand communities of the respiration, ingestion) from the P/B value for each com Eastern English Channel are dominated by large-sized partment. The P/B values for many species and/or com suspension-feeder species (Dauvin & Ruellet 2008), so this partments are widely available in the literature, but they bias is mainly restricted to the less common fine sedi vary greatly depending on the authors. This variation can ments. be explained by the different methods used for the assess According to Vézina & Platt (1988), the inverse meth ment, but also by the intrinsic properties of P/B. The ods provide a strong foundation for an effective compara annual P/B of a cohort usually decreases with age; it fol tive analysis of food-web dynamics. This study allowed us lows that populations dominated by older year classes will to determine that there is variation in trophic organisa have a lower P/B than those composed of younger indi tion in the Eastern English Channel and the Southern viduals (Warwick 1980). Another solution would be the North Sea, depending on the scale of observation. In this coupling of this kind of work with other modelling meth respect, the sediment division (thus the bio-sedimentary ods such as the model of size spectra developed by Jen division) appears to be the most important factor con nings et al. (2002) for the benthic system. This method trolling benthic ecosystem functioning in the area, with assumes that organisms with higher body mass feed at the possible exception of the Bay of Seine due to its very higher trophic levels, meaning that it requires fewer particular features {i.e. an enclosed bay in close proximity parameters to assess the main energy flow through the to a river). The trophic organisation of the overall area is food-web. an integration of the specific properties of each individual To the problem of variation in this well known functional unit. Thus, views of the trophic structure of compartment, it is necessary to add the problem of the the system can differ depending on the scales and factors compartments about which little is known - the black considered. Further investigations are needed to identify boxes such as benthic bacteria or meiofauna. As few studies Marine Ecology32 (Suppl. 1 ) (2011 ) 72-86 © 2011 Blackwell Verlag GmbH 83 Assessment of benthic ecosystem functioning Garcia, Chardy, Dewarumez & Dauvin of these compartments are available, it is extremely difficult Boaden P.J.S. (2005) Irish intertidal meiofauna: a modicum of to gather the necessary parameters which, by virtue of their progress. In: Wilson J.G. (Ed.), The Intertidal Ecosystem: The lack of attention, are plagued by incertitude. Value of Ireland’s Shores. Royal Irish Academy, Dublin: Another problem is that macrobenthic communities 81-99. are well known for having large fluctuations in biomass Breed G.A., Jackson G.A., Richardson T.L. (2004) Sedimenta and abundance from year to year, mainly due to preda tion, carbon export and food web structure in the Missis tion on larvae in the plankton (Thorson 1946) or newly sippi River plume described by inverse analysis. Marine settled juveniles. In addition, high adult mortality can Ecology Progress Series, 278, 35-51. Bremner J., Rogers S.I., Frid C.L.J. (2006) M ethods for describ also occur (Warwick 1980). For these reasons, the trophic ing ecological functioning of marine benthic assemblages representation for 1 year may not be valid the following using biological traits analysis (BTA). Ecological Indicators, year. This is particularly evident in populations that are 6, 609-622. dominated by species with long-lived planktonic larvae, Cabioch L., Glaçon R. (1975) Distribution des peuplements which are considered to have highly unstable dynamics benthiques en Manche orientale, de la baie de Somme au (Thorson 1946). Pas-de-Calais. Comptes Rendus de YAcademie des Sciences Our study has shown that, as far as the benthic com (Paris), 280, 491-494. partment is concerned, a large-scale spatial model seems Cabioch L., Glaçon R. (1977) Distribution des peuplements to be acceptable for our study area, as the benthic com benthiques en Manche orientale du cap d’Antifer à la baie munities are all influenced by hydrodynamics. Nonethe de Somme. Comptes Rendus de YAcademie des Sciences less, as benthic trophic functioning is variable and (Paris), 285, 209-212. strongly dependent on sediment type, this model would Cabioch L., Gentil F., Glaçon R., Retières C. (1978) Le bassin have to take into account at least the three main sediment oriental de la Manche, modèle de distribution de peuple types {i.e. gravel and pebbles, coarse sand and fine sand). ments benthiques dans une mer à fortes marées. Journal de Recherche Océanographique, 3, 249-254. Carlier A., Riera P., Amouroux J.-M., Bodiou J.-Y., Grémare Acknowledgements A. (2007) Benthic trophic network in the Bay of Banyuls- The authors wish to thank Dr Aurélie Foveau, Aurore sur-Mer (northwest Mediterranean, France): an assessment Savina, Emilie Houillez and Amélie Charnoz for their help based on stable carbon and nitrogen isotopes analysis. in the sorting and identification of the species, Eric Lecu- Estuariae, Coastal and Shelf Science, 72, 1-15. yer, Dr Nicolas Desroy, Dr Nicolas Spilmont and the offi Chardy P. (1987) Modèle de simulation du système benthique cers and crew of the RV Cotes de la M anche for their effort des sédiments grossiers du golfe normand-breton (Manche). at sea, Lisa Spencer and Dr Julie Bremner for correcting Oceanologica Acta, 10, 421-434. Chardy P., Dauvin J.-C. (1992) Carbon flows in a subtidal the English syntax and grammar, and two anonymous ref fine sand community from the western English Channel: erees for their helpful suggestions and comments. This a simulation analysis. Marine Ecology Progress Series, 81, work was carried out with financial contribution of the 147-161. European Union’s Interreg Illa scheme (European Regio Chardy P., Dauvin J.-C., Lasserre P. (1993a) Biodiver nal Development Funds) CE1ARM 2 project coordinated sity and functional units in a coastal benthic food by A. Carpentier, Ifremer, Boulogne-sur-mer. web. Workshop Biodiversity, Point-á-Pitre, Guadeloupe: 1- 8 . Chardy P., Gros P., Mercier H., Monbet Y. (1993b) Benthic R eferences carbon budget for the Bay of Saint Brieuc (Western Chan Améziane N., Chardy P., Dauvin J.-C. (1996) Modelling carbon nel). Application of the inverse method. Oceanologica Acta, flows in soft-bottom communities from the bay of Morlaix, 16, 687-694. western English Channel. In: Eleftheriou A.d. (Ed.), Biology Christensen N.L., Bartuska A.M., Brown J.H., Carpenter and Ecology of Shallow Coastal Waters, 28 EMBS Symposium. S.R., D ’Antonio C., Francis R., Franklin J.F., M acM ahon Olsen & Olsen, Fredenstourg, Denmark: 215-224. J.A., Noss R.F., Parsons D.J., Peterson C.H., Turner Barnsley M.J. (2007) Environmental Modeling: A Practical Intro M.G., Woodmansee R.G. (1996) The report of the duction. Taylor & Francis Group, Boca Raton, FL: 406 pp. Ecological Society of America Committee on the Scien Bilyard G.R. (1987) The value of benthic infauna in marine tific Basis for Ecosystem Management. Ecological Applica pollution monitoring studies. Marine Pollution Bulletin, 18, tion, 6, 665-691. 581-585. Coli M., Palomera I., Tudela S., Sarda F. (2006) Trophic flows, Blanchet H. (2004) Structure et fonctionnement des ecosystem structure and fishing impacts in the South peuplements benthiques du Bassin d’Arcachon. Ph.D. thesis, Catalan Sea, Northwestern Mediterranean. Journal of Marine Université de Bordeauxl, Bordeaux: 331 pp. Systems, 59, 63-96. 84 Marine Ecology32 (Suppl. 1) (2011) 72-86 © 2011 Blackwell Verlag GmbH Garcia, Chardy, Dewarumez & Dauvin Assessment of benthic ecosystem functioning Dauvin J.C. (1993) Le benthos: témoin des variations de l’envi Kröncke I. (2006) Structure and function of macrofaunal com ronnement. Oceanis, 19, 25-53. munities influenced by hydrodynamically controlled food Dauvin J.C. (1997a) Les biocénoses marines et littorales français availability in the Wadden Sea, the open North Sea and the es des côtes Atlantique, Manche et Mer dn Nord, syntheses, dees-sea. Senckenbergiana maritima, 36, 1-18. menaces et perspectives. MNHN, Paris: 359 pp. Kröncke I., Stoeck T., Wieking G., Palojaervi A. (2004) Rela Dauvin J.C. (1997b) Evolution à long terme des peuplements tionship between structure and function of microbial and de sédiments fins sablo-vaseux de la Manche et de la mer macrofaunal communities in different areas of the North deu N ord. Oceanis, 23, 113-144. Sea. Marine Ecology Progress Series, 282, 13-31. Dauvin J.-C., Ruellet T. (2008) Macrozoobenthic biomass in Langdon C.J., Newell R.I.E. ( 1990) Comparative utilization of the Bay of Seine (eastern English Channel). Journal o f Sea detritus and bacteria as food sources by two bivalve suspen Research, 59, 320-326. sion-feeders, the oyster Crassostrea virginica and the mussel, Desroy N., Warembourg C., Dewarumez J.-M., Dauvin J.-C. Geukensia demissa. Marine Ecology Progress Series, 58, 299- (2003) Macrobenthic resources of the shallow soft-bottom 310. sediments in the eastern English Channel and southern Larsonneur C., Bouysse P., Auffret J.P. (1982) The superficial N orth Sea. ICES Journal of Marine Science, 60, 120-131. sediments of the English Channel and its Western Eldridge P.M., Jackson G.A. (1993) Benthic trophic dynamics Approaches. Sedimentology, 29, 851-864. in California coastal basin and continental slope communi Le Loc’h F., Hily C. (2005) Stable carbon and nitrogen isotope ties inferred using inverse analysis. Marine Ecology Progress analysis of Nephrops norvegicus/Merluccius merluccius fishing Series, 99, 115-135. grounds in the Bay of Biscay (NE Atlantic). Canadian Jour Ellien C., Thiébaut E., Barnay A.-S., Dauvin J.-C., Gentil F., nal o f Fisheries and Aquatic Sciences, 62, 123-132. Salomon J.-C. (2000) The influence of variability in larval Lees K., Mackinson S. (2007) An Ecopath model of the Irish dispersal on the dynamics of a marine metapopulation in Sea: ecosystems properties and sensitivity analysis. Science the eastern Channel. Oceanologica Acta, 23, 423-442. Series Technical Report. Cefas, Lowestoft: 49 pp. Fauchald K., Jumars P.A. (1979) The diet of worms: a study of Leguerrier D., Niquil N., Boileau N., Rzeznik J., Sauriau P.-G., polychaete feeding guilds. Oceanography and Marine Biology Le Moine O., Bacher C. (2003) Numerical analysis of the an Annual Review, 17, 193-284. food web of an intertidal mudflat ecosystem on the Atlantic George C.L., Warwick R.M. (1985) Annual macrofauna pro coast of France. M arine Ecology Progress Series, 246, 17-37. duction in a hard-bottom reef community. Journal of the Leguerrier D., Niquil N., Petiau A., Bodoy A. (2004) Modeling Marine Biological Association of the United Kingdom, 65, the impact of oyster culture on a mudflat food web in 713-735. Marennes-Oléron Bay (France). Marine Ecology Progress Gerlach S.A. (1971) On the importance of marine meiofauna Series, 273, 147-162. for benthos communities. Oecologia, 6, 176-190. Lopez G., Taghon G.L., Levinton J. (1989) Ecology o f Marine Ghertsos K. (2002) Structure spatio-temporelle des peuple Deposit Feeders. Springer-Verlag, New York. ments macrobenthiques de la baie de Seine à plusieurs Mackinson S., Daskalov G. (2007) An ecosystem model of the échelles d’observation. Ph.D. thesis, Université des Sciences North Sea to support an ecosystem approach to fisheries et Technologies de Lille: 181 pp. management: description and parameterisation. Science Gray J.S. ( 1974) A nim al-sediment relationship. Oceanography Series Technical Report. Cefas, Lowestoft: 142 pp. and Marine Biology: An Annual Review, 12, 223-261. Marquis E., Niquil N., Delmas D., Hartmann H.J., Bonnet D., Harvey C.J., Cox S.P., Essington E., H ansson S., Kitchell J.F. Carlotti F., H erbland A., Labry C., Sautour B., Laborde P., (2003) An ecosystem model of food web and fisheries inter Dupuy C. (2007) Planktonic food web dynamics related to actions in the Baltic Sea. ICES Journal of Marine Science, 60, phytoplankton bloom development on the continental shelf 939-950. of the Bay of Biscay, French coast. Estuariae Coastal and Hooper D.U., Chapin F.S., Ewel J.J., Hector A., Inchausti P., Shelf Science, 73, 223-225. Lavorel S., Lawton J.H., Lodge D.M., Loreau M., Naeem S., M artin C.S., Carpentier A., Vaz S., Coppin F., Curet L., Dauvin Schmid B., Setälä H., Symstad A.J., Vanderm eer J., W ardle J.-C., Delavenne J., Dewarumez J.-M., D upuis L., Engelhard D.A. (2005) Effects of biodiversity on ecosystem functioning: G., Ernande B., Foveau A., Garcia C., Gardei L., H arrop S., a consensus of current knowledge. Ecological Monographs, Just R., Koubbi P., Lauria V., M eaden G.J., M orin J., O ta Y., 75, 3-35. Rostiaux E., Smith R., Spilmont N., Verin Y., Villanueva C., Jennings S., Pinnegar J.K., Polunin N.V.C., W arr K.J. (2002) Warembourg C. (2009) The Channel habitat atlas for marine Linking size-based and trophic analyses of benthic commu resource management (CHARM): an aid for planning and nity structure.Marine Ecology Progress Series, 226, 77-85. decision-making in an area under strong anthropogenic pres Kaiser M.J., Armstrong P.J., Dare P.J., Flatt R.P. (1998) Ben sure. Aquatic Living Resources, 22, 499-508. thic communiies associated with heavily fished scallop Martinez N.D., Hawkins B.A., Dawah H.A., Feifarek B.P. ground in the English Channel. Journal of the Marine Biolog (1999) Effects of sampling effort on characterization of ical Association of the United Kingdom, 78, 1045-1059. food-web structure. Ecology, 80, 1044-1055. Marine Ecology32 (Suppl. 1 ) (2011 ) 72-86 © 2011 Blackwell Verlag GmbH 85 Assessment of benthic ecosystem functioning Garcia, Chardy, Dewarumez & Dauvin van der Meer J., Heip C.H., Herman Moens T., van to the planktonic larvae in the sound (Oresund). Meddelelser Oevelen D. (2005) Measuring the flow of energy and matter Fra Kommissionen for Danmarks fisheri- Og Havundersogelser, in marine benthic animal population. In: Eleftheriou A., Serie: Plankton, 4, 1-523. McIntyre A. (Eds), Methods for the Study of Marine Benthos. Underwood A.J. (1990) Experiments in ecology and manage Blackwell Science, Oxford: 326-407. ment: thei logics, functions and interpretations. Australian Newell R.C., Seiderer L.J., Robinson J.E. (2001) Animal: sedi Journal o f Ecology, 15, 365-389. ment relationships in coastal deposits of the eastern English Underwood A.J. (1996) Detection, interpretation, prediction Channel. Journal of the Marine Biological Association of the and management of environmental disturbances: some roles United Kingdom, 81, 1-9. for experimental marine ecology. Journal of Experimental Niquil N., Pouvreau S., Sakka A., Legendre L., Addessi L., Le M arine Biology and Ecology, 200, 1-27. Borgne R., Charpy L., Delesalle B. (2001) Trophic web Vézina A.F. ( 1989) Construction of flow networks using and carrying capacity in a pearl oyster farming lagoon inverse methods. In: Wulff F., Field J.G., Mann K.H. (Eds), (Takapoto, French Polynesia). Aquatic Living Resources, 14, Network Analysis in M arine Ecology: Methods and Applica 165-174. tions. Springer-Verlag, Berlin: 62-81. Pasquaud S., Lobry J., Elie P. (2007) Facing the necessity of Vézina A.F., Platt T. (1988) Food web dynamics in the ocean. describing estuarine ecosystems: a review of food web ecol I. Best-estimates of flow networks using inverse methods. ogy study techniques. Hydrobiologia, 588, 159-172. M arine Ecology Progress Series, 42, 269-287. Pearson T.H., Rosenberg R. (1978) Macrobenthique succession Vranken G., Heip C. (1986) The productivity of marine nema in relation to organic enrichment and pollution of the todes. Ophelia, 26, 429-442. marine environment. Oceanography and Marine Biology: An Vranken G., Herman P.M.J., Vincx M., Heip C. ( 1986) A re- Annual Review, 16, 229-311. evaluation of marine nematode productivity. Hydrobiologia, Rybarczyk H., Elkaim B. (2003) An analysis of the trophic 135, 193-196. network of a macrotidal estuary: the Seine Estuary (Eastern Warwick R.M. (1980) Population dynamics and secondary Channel, Normandy, France). Estuarine, Coastal and Shelf production of benthos. In: Tenore K.R., Coull B.C. (Eds), Science, 58, 775-791. Marine Benthic Dynamics. The Belle W. Baruch Library in Rybarczyk H., Elkaim B., Ochs L., Loquet N. (2003) Analysis Marine Science, Columbia: 1-24. of the trophic network of a macrotidal ecosystem: the Bay Warwick R.M., George C.L. (1980) Annual macrofauna of Somme (Eastern Channel). Estuarine, Coastal and Shelf production in an Abra community. In: Collins M.B. (Ed.), Science, 58, 405-421. Industrialized Embayments and Their Environmental Prob Salomon J.-C., Breton M. (1991) Courants résiduels de marée lems. Pergamon, Oxford: 517-538. dans la Manche. Oceanologica Acta, 11, 47-53. Warwick R.M., Price R. (1975) Macrofauna production in an Salvanes A.G.V., Aksnes D.L., Giske J. ( 1992) Ecosystem model estuarine mud-flat. Journal of the Marine Biological Associa for evaluating potential cod production in a west Norwegian tion of the United Kingdom, 55, 1-8. fjord. Marine Ecology Progress Series, 90, 9-22. Warwick R.M., Radford P.J. ( 1989) Analysis of the flow network Schwinghamer P., Hargrave B., Peer D., Hawkins C.M. (1986) in an estuarine benthic community. In: Wulff F., Field J.G., Partitioning of production and respiration among size Mann K.H. (Eds), Network Analysis in Marine Ecology: M eth groups of organisms in an intertidal benthic community. ods and Applications. Springer-Verlag, New York: 220-231. M arine Ecology Progress Series, 31, 131-142. Warwick R.M., George C.L., Davies J.R. (1978) Annual macro Solan M., Germano J.D., Rhoads D.C., Smith C., Michaud E., fauna production in a Venus community.Estuarine and Parry D., Wenzhöfer F., Kennedy B., Henriques C., Battle Coastal Marine Science, 7, 215-241. E., Carey D., Iocco L., Valente R., W atson J., Rosenberg R. Warwick R.M., Ashman C.M., Brown A.R., Clarke K.R., Dowell (2003) Towards a greater understanding of pattern, scale B., H art B., Lewis R.E., Shillabeer N., Somerfield P.J., Tapp and process in marine benthic systems: a picture is worth J.F. (2002) Inter-annual changes in the biodiversity and com than a thousand worms. Journal of Experimental Marine munity structure of the macrobenthos in Tees Bay and the Biology and Ecology, 285-286, 313-338. Tees estuary, UK, associated with local and regional environ Tarantola A., Valette B. (1982) Generalized nonlinear inverse mental events. Marine Ecology Progress Series, 234, 1-13. problems solved using the least squares criterion. Reviews of W hipple S.J., Link J.S., Garrison L.P., Fogarty M.J. (2000) Geophysics and Space Physics, 20, 219-232. Models of predation and fishing mortality in aquatic ecosys Thorson G. (1946) Reproduction and larval development of tems. Fish Fisher, 1, 22-40. danish marine bottom invertebrates, with special reference 86 Marine Ecology32 (Suppl. 1) (2011) 72-86 © 2011 Blackwell Verlag GmbH anmarine evolutionary perspective ecology » Marine Ecology. ISSN 0173-9565 ORIGINAL ARTICLE Epiphytes and associated fauna on the brown alga Fucus vesiculosus in the Baltic and the North Seas in relation to different abiotic and biotic variables Priit Kersen1,2, Jonne Kotta1, Martynas Bucas3, Natalja Kolesova4 & Zane D éféré5 1 Estonian Marine Institute, University of Tartu, Tallinn, Estonia 2 Institute of Mathematics and Natural Sciences, Tallinn University, Tallinn, Estonia 3 Coastal Research and Planning Institute, University of Klaipeda, Klaipeda, Lithuania 4 Marine Systems Institute, Tallinn University of Technology, Tallinn, Estonia 5 Laboratory of Marine Ecology, Institute of Biology, University of Latvia, Salasplls, Latvia Keywords A bstract Community composition; dominance structure; eplblonts; fucolds; host frond; Fucus vesiculosus L. is an important habitat-forming macroalga both in the marine benthos; mobile invertebrates; saline and high diverse North Sea and the diluted and low diversity Baltic Sea. seaweeds; spatial variability; wave exposure. Despite its importance, comparisons of the spatial patterns of its epiphytes have rarely been reported. In this study we examined the species composition Correspondence and density of macro-epiphytes and mobile fauna on the canopy-forming mac Priit Kersen, Estonian Marine Institute, roalga F. vesiculosus inhabiting different regimes of wave exposure in the North University of Tartu, Mäealuse 14, 12618 Tallinn, Estonia. and Baltic Seas. The North and Baltic Seas had distinct epiphyte and mobile E-mail: [email protected] faunal communities. Wave exposure and segments of host fronds significantly contributed to the variability in species composition and dominance structure Accepted: 16 November 2010 of epiphytes on F. vesiculosus in the North Sea and Baltic Sea. The study indi cated that there is no clear spatial scale where environmental variables best pre doi: 10.1111/j. 1439-0485.2010.00418.x dicted epiphytic and mobile faunal communities, and the formation of epiphytic and faunal communities is an interplay of factors operating through micro- to regional scales. tidal Baltic Sea (Kiirikki 1996a; Berger et al. 2004; Torn Introduction et al. 2006; Rohde et al. 2008). Fucus vesiculosus hosts a Epiphytism and competition for a substrate is a wide large variety of macroalgal species (Rindi & Guiry 2004), spread phenomenon in marine communities, especially in which provide suitable habitat for sessile invertebrates the rocky intertidal zone (Paine 1990; Kraberg & Norton (Johnson & Scheibling 1987) and associated invertebrates, 2007). Many algal species can grow on some host species mainly grazers (Orav-Kotta & Kotta 2004; Räberg & Ka- or even are obligatory epiphytes (Pavia & Aberg 1999) utsky 2007). Despite its importance, comparisons of the providing potential for mutualistic interspecific associa spatial patterns of its epiphytes have rarely been reported tions (Stachowicz & Whitlatch 2005). (Rindi & Guiry 2004; Fraschetti et al. 2005). Fucoids are widely distributed perennial brown macro Epiphytic organisms, such as micro- and macroalgae, algae in the intertidal Northeastern Atlantic with an invertebrates and bacteria, are often present on the thallus important role in structuring intertidal communities of perennial macroalgae. Their abundance is largely deter (Lüning 1990). They can also extend to brackish non- mined by abiotic factors, e.g. water motion and nutrient tidal Baltic waters. Fucus vesiculosus L. is an im portant availability. The ability of epiphytes to tolerate regular habitat-forming macroalga both in the saline and high desiccation during low tides determines their spatial diverse North Sea and the diluted and low diversity non- distribution (Molina-Montenegro et al. 2005). Elevated Marine Ecology 32 (Suppl. 1) (2011) 8 7-95 © 2011 Blackwell Verlag GmbH 87 Epiphytes and associated fauna on brown alga Kersen, Kotta, Bucas, Kolesova & Deljere nutrient loading is expected to increase both the number We tested the following hypotheses: of epiphytic algae and invertebrates. However, the rela 1 The occurrence and cover of epiphytes are specific to tionship varies among regions and is modulated by a the frond segment of host macroalgae. number of other environmental variables, e.g. wave expo 2 The occurrence and cover of epiphytes are related to sure, regional species pool, characteristics of the host the wave exposure of the site. plant and herbivory (Kotta et al. 2000; Worm et al. 2002; 3 The relationship between abiotic (exposure), biotic fac Kotta & Witman 2009). Among biotic interactions, intra tors (frond segment) and epiphytes (species composi specific competition for space and light and the presence tion, cover) varies among different marine regions of adequate food resources for invertebrates are important (N orth versus Baltic Seas). (Lobban & Harrison 2000). The ability of host algae to resist and avoid epibionts Material and Methods has great importance (Honkanen & Jormalainen 2005), affecting the photosynthetic rate and growth of host algae Study area (K orpinen et al. 2007). The position within the thallus of The study was performed in the North Sea and the Baltic seaweeds is important factor structuring epiphyte com Sea. In each region we selected three sites differing in munities (Lobban & Baxter 1983; Cardinal & Lesage exposure level. The used exposure levels according to the 1992; Longtin et al. 2009). Large epiphytes are associated EUNIS classification were as follows: sheltered, moder with the basal disk; ephemeral epiphytes appearing on the ately exposed and exposed. In the North Sea the sampling tips of the host fucoid fronds (Arrontes 1990). The cover was done on the southwest coast of Norway in Raunefj- of epiphytes often increases with the age of algae. ord and Korsfjord in summer 2007 (Fig. 1). The North The objective of this study was to examine the species Sea study area contains numerous small and large islands composition and density of epiphytes and mobile fauna separated by wide or narrow sounds. The bottom relief of on the canopy-forming macroalga F. vesiculosus inhabit the area is steep and very uneven. Espegrend Marine Bio ing at different regimes of wave exposure in the North logical Station (N l; 60.273°N, 5.218°E) represented shel and Baltic Seas. We considered epiphytes hereafter as tered, Loholmen (N2; 60.266°N, 5.211°E) moderately organism-on-a-plant concept (sensu W ahl 2009; Steel & exposed and Store Kalsoy (M3; 60.113°N, 5.069°E) Wilson 2003, references therein). Thus, epiphytes in exposed areas. During sampling, salinity ranged between this report comprise any macroalgae and sessile inverte 30 and 33 psu and tidal range was c. 1 m in the sampling brates on the host algae. Mobile fauna is defined here as area. motile macroinvertebrates associated with the host mac In the Baltic Sea, samples were collected from the Gulf roalgae. of Riga and the Baltic Proper in summer 2008 (Fig. 1). Fig. 1. Map of the sampling regions and sites. Different stations are Indicated by N1, N2, N3 In the North Sea and B1, B2, B3 In the Baltic Sea. 88 Marine Ecology32 (Suppl. 1) (2011) 87-95 © 2011 Blackwell Verlag GmbH Kersen, Kotta, Bucas, Kolesova & Deljere Epiphytes and associated fauna on brown alga The Gulf of Riga is a wide, shallow, semi-enclosed brack ish water ecosystem. In general, the bottom relief of the area is quite flat, with gentle slopes towards deeps. The northern part of the Gulf is characterized by a wide coastal zone with diverse bottom topography and exten Distal segment sive reaches of boulders. The coasts of the Baltic Proper are very exposed, hydrodynamically active and character Middle segment ized by a steep coastline. The inner part of Köigute Bay, the Gulf of Riga (BÍ; 58.374°N, 22.972°E) represented a sheltered area, the outer part of Köiguste Bay ( B2; 58.370°N, 22.982°E) a moderately exposed area, and Basal segment Kiidema Bay, the Baltic Proper ( B3; 58.568°N, 22.302°E) an exposed area. During sampling, salinity ranged between 4 and 7 psu. The Baltic Sea is nearly tideless, with an average daily tidal component of 15 cm (Schiewer 2008). The water level fluctuations in the area are mainly caused by the meteorological forcing with a seasonal sea Fig. 2. Examined frond segments of the host Fucus vesiculosus. level mean range of 30 cm (e.g. Suursaar & Sooäär 2007). considered an important factor for mobile or associated Sampling fauna. Therefore, 2-factor PERMANOVA was used to In the North Sea, samples were taken at low tide on the determine the abundances of mobile invertebrates. upper littoral zone in the middle of F. vesiculosus belt. In Prior to analysis, a Bray-Curtis similarity matrix was the Baltic Sea samples were collected by a diver from a calculated using raw data (untransformed) and pres Fucus vesiculosus belt at 0.5-1 m depth. Four replicates of ence/absence transformation to detect whether the poten host algae were collected randomly with the aid of a rope tial differences between the assemblages of the epibiota that was placed along the shore and marked at every were due to differences in relative abundances or species metre. Four marks were randomly selected and host composition (Clarke & Warwick 2001). plants were collected nearest to the respective marks on When a factor with more than two levels (i.e. wave the rope. Samples were transported in plastic bags to the exposure and host frond segment) was identified as sig laboratory for the further analyses within 24 h. Mobile nificant (P < 0.05), post-hoc PERMANOVA pair-wise tests fauna were removed from host thalli, counted and identi were conducted to detect which levels were responsible fied to the lowest taxonomic level possible using a dissect for significant interactions. Taxa responsible for observed ing microscope (magnification 4-10x). The cover of differences were identified by similarity percentages (SIM epiphytes were estimated on a six-grade scale ( 0 - PER), where the cut-off percentage was set to 90. Non absence, 1 - from 1 to 20%, 2 - from 21 to 40%, 3 - metric multidimensional scaling (nMDS) was used to from 41 to 60%, 4 - from 61 to 80% and 5 - from 81 to present visual images of the differences in composition of 1 0 0 % of cover on host plant) separately on basal, middle epiphytic and mobile faunal assemblages in distinct mar and distal segments of each host frond (Rindi & Guiry ine regions, exposure levels and frond segment of the 2004) (Fig. 2). host. Data analysis Results All multivariate analyses were conducted using PRIMER 6 A total of 27 epiphytic and mobile faunal taxa were software (Clarke & Gorley 2006). Hierarchical permuta- recorded on the fronds of Fucus vesiculosus in the studied tional multivariate analysis of variance (PERMANOVA) areas: 8 taxa of macroalgae, 5 taxa of sessile invertebrates (A nderson et al. 2008) was used separately for epiphytes (i.e. suspension-feeders) and 14 taxa of mobile inverte and mobile fauna to examine differences in the patterns brates (mainly herbivores) (Table 1). of variation in composition, cover and abundance All investigated factors significantly contributed to the between regions (fixed factor), wave exposure (nested in variability in species composition and coverage of epi region, fixed factor) and frond segment of host macroal phytic and mobile faunal communities on F. vesiculosus gae (nested in region and wave exposure, fixed factor). (PERMANOVA, Table 2). Epiphytic and mobile fauna Due to the mobility of organisms, frond segment was not communities were clearly differentiated in species Marine Ecology 32 (Suppl. 1) (2011) 8 7-95 © 2011 Blackwell Verlag GmbH 89 Epiphytes and associated fauna on brown alga Kersen, Kotta, Bucas, Kolesova & Delyere composition and dominance structure between the North Similarly, the dominance structure of epiphytic com and Baltic Seas (Table 2, PERMANOVA: P = 0.001). The munity (untransformed data) on their host was signifi epiphytes and mobile fauna taxa contributing most to the cantly different between every level of wave exposure in regional variability were the brown algae Pylaiella littoral both the North and Baltic Seas (P < 0.05), except for is, Elachista fucicola, the tube-building polychaete Spiror moderately exposed versus sheltered sites in the North Sea bis spirorbis, and the herbivorous snail Theodoxus (PERMANOVA pair-wise test: P = 0.078). Dissimilarities fluviatilis (Supporting Information Tables S1-S4). between different exposure levels were mostly due to The species composition of epiphytic community (pres E. fucicola and P. littoralis (Supporting Information ence/absence transformed data) on their host was signifi Table S6 ). cantly different between every level of wave exposure in The species composition (presence/absence data) and both the North and Baltic Seas (P < 0.05), except for dominance structure (untransformed data) of mobile fau moderately exposed versus sheltered sites in the North Sea nal community differed significantly for every exposure (PERMANOVA pair-wise test: P = 0.092). Dissimilarities level in the North and Baltic Seas (PERMANOVA between different exposure levels were mostly due to Elac pair-wise test: P < 0.05). Dissimilarities between different hista fucicola and Pylaiella littoralis (Supporting Informa exposure levels were mainly due to Gammarus spp., tion Table S5). Table 1. Recorded epiphytes and mobile fauna on the hostFucus vesiculosus at three study sites in the North (NS) and Baltic Sea (BS). Wave Frond segment NS: Mean BS: Mean No. Taxon Region exposure of host cover/abun ± SE cover/abun ± SE Bryozoa 1 Electra crustulenta (Pallas, 1766) BS S, ME B, M, D 0 0.25 ± 0.07 2 Electra pilosa (L.) NS S, ME B, M, D 0.47 ± 0.14 0 Polychaetae 3 Spirorbis spirorbis (Linnaeus, 1758) NS S, ME B, M, D 1.03 ± 0.18 0 Crustaceans 4 Balanus Improvisus Darwin, 1854 BS E M 0 0.03 ± 0.03 S Gammarus spp. BS, NS S, ME, E - 0.33 ± 0.26 6.5 ± 2.21 6 Idotea balthica (Pallas, 1772) BS, NS S, ME, E - 1.5 ± 0.4 6.17 ± 1.60 7 Idotea chelipes (Pallas, 1766) BS, NS S, ME, E - 8 Jaera albifrons Leach, 1814 BS ME - 0 0.25 ± 0.25 Nemertina 9 Cyanophthalma obscura (Schultzei BS E - 0 0.08 ± 0.08 Mollusca 10 Gibbula cineraria (L.) NS ME - 0.08 ± 0.08 0 11 Elydrobia spp. BS S, ME, E - 0 14.92 ± 6.24 12 Lacuna vincta (Montagu) NS E - 0.17 ± 0.11 0 13 Littorina littorea (L.) NS S, ME, E - 7.17 ± 1.62 0 14 Littorina obtusata (L.) NS S, ME, E - 15 Lymnaea peregra (Müller) BS ME, E - 0 0.42 ± 0.26 16 Mytilus edulis L. NS S - 0.17 ± 0.17 0 17 Mytilus trossulus Gould BS E - 0 0.42 ± 0.29 18 Theodoxus fluviatilis (L.) BS S, ME, E - 0 39.17 ± 6.47 Macroalgae 19 Ceramium tenuicorne (Kützing) Waern BS, NS ME, E B, M, D 0.08 ± 0.06 0.28 ± 0.09 20 Ceramium virgatum Roth BS, NS ME, E B, M, D 21 Chaetomorpha sp. NS S M, D 0.06 ± 0.04 0 22 Chordaria flagelliformis (Müller) Agarth NS ME M 0.06 ± 0.06 0 23 Cladophora glomerata (L.) Kützing BS, NS S, ME B, M, D 0.03 ± 0.03 0.36 ± 0.09 24 Elachista fucicola (Velley) Arenschoug NS S, ME, E B, M, D 0.72 ± 0.17 0.19 ± 0.08 25 Pylaiella littoralis (L.) Kjellman BS, NS S, ME B, M, D 0.06 ± 0.04 1.14 ± 0.26 26 Ulva intestinalis L. NS S, ME B, M, D 0.14 ± 0.06 0 Hydrozoa 27 Dynamena pumila (L.) NS ME B, M, D 0.11 ± 0.07 0 S, Sheltered: ME, moderately exposed; E, exposed site. B, basal; M, middle; D, distal segment of host alga. Means and standard errors were calculated from untransformed coverage and abundance data. 90 Marine Ecology32 (Suppl. 1) (2011) 87-95 © 2011 Blackwell Verlag GmbH Kersen, Kotta, Bucas, Kolesova & Deljere Epiphytes and associated fauna on brown alga Littorina spp. and Theodoxus fluviatilis (Supporting Infor regions than for exposure levels. Nevertheless, in some mation Tables S7 and S 8 ). instances the differences were very clear, e.g. epiphytes Different segments of host fronds had significantly dif were totally absent at the exposed site of the North Sea ferent species composition of epiphytic communities at (Fig. 3). the moderately exposed site in the North Sea (between middle and distal segments; P = 0.033) and at exposed Discussion (between middle and distal segments; P = 0.03) and shel tered sites in the Baltic Sea (between basal and distal seg We predicted that the occurrence and cover of epiphytes ments; P = 0.031). Different segments of host fronds had would be specific to the frond segments of host macro significantly different dominance structure of epiphytic algae. The results agreed with the hypothesis as different communities (untransformed data) at the sheltered site in parts of host seaweeds had different epiphytic and mobile the North Sea (between basal and middle segments; fauna species composition and dominance structure. The P = 0.026) and at exposed (between middle and distal distal segment of host frond had the lowest coverage and segments; P = 0.023) and sheltered sites in the Baltic Sea the lowest species richness of epiphytes. This pattern is (between basal and distal segments; P = 0.03). The differ due to the high metabolic activity of apical (new) parts of ences in species composition and dominance structure Fucus vesiculosus thallus where the host alga produces allel- were mainly caused by E. fucicola and P. littoralis (Sup opathic compounds such as phlorotannins (Wikström & porting Information Tables S9 and SIO). Pavia 2003). Also, the topmost parts of fucoid algae have According to nMDS ordination, the epiphytic commu an anti-fouling strategy of periodically shedding surface nity composition and structure on host macroalgae clearly cell layers (Kiirikki 1996b), which reduces the probability differed between sea regions in epiphyte coverage and also of epibionts settling and becoming established on the host in mobile fauna abundance. It is also possible to detect a plant. A similar strategy of mechanical defence has been separate effect of wave exposure on epiphytic and mobile observed in red algal hosts (Nylund & Pavia 2005). faunal community composition and structure within the We also predicted that the occurrence and cover of epi two sea regions. However, the distinctions are larger for phytes would be related to the wave exposure of the site. Table 2. Main results of PERMANOVA analyses on the effect of region, wave exposure, and frond segment to species composition and domi nance structure of epiphytic and mobile faunal assemblages on Fucus vesiculosus. Source d.f. SS MS Pseudo-F P(perm) Unique perms Epiphytic coverage Presence/absence transformed Region 1 15,498 15,498 41.016 0.001 998 Wave exposure (Region) 4 36,967 9241.6 24.459 0.001 997 Frond segment (Wave exposure (Region) 12 13,869 1155.7 3.0588 0.001 999 Res 54 20,403 377.84 Total 71 86,736 Untransformed Region 1 18,220 18,220 33.454 0.001 998 Wave exposure (Region) 4 50,430 12,607 23.149 0.001 997 Frond segment (Wave exposure (Region)) 12 19,143 1595.2 2.929 0.001 996 Res 54 29,410 544.63 Total 71 117,200 Mobile faunal abundance Presence/absence transformed Region 1 42,429 42,429 123.9 0.001 999 Wave exposure (Region) 4 9522.1 2380.5 6.9515 0.001 999 Res 66 22,601 342.45 Total 71 74,553 Untransformed Region 1 109,390 109,390 12.33 0.001 997 Wave exposure (Region) 4 35,212 8803 10.408 0.001 996 Res 66 55,825 845.83 Total 71 200,430 Res, Residual. Marine Ecology32 (Suppl. 1 ) (2011 ) 87-95 © 2011 Blackwell Verlag GmbH 91 Epiphytes and associated fauna on brown alga Kersen, Kotta, Bucas, Kolesova & Deljere Some epiphytic (e.g. Spirorbis spirorbis, Bryozoa, Clado them to thrive in the low salinity environment. This sug phora glomerata, Pylaiella littoralis) and mobile faunal gests that solinity determines interregional differences in species (e.g. Jaera albifrons) were not observed at exposed epiphyte communities between North and Baltic Seas study sites, suggesting that occurrence and cover/abun- (Kangas & Skoog 1978; Russell 1994; Snoeijs 1999). The dance are related to the exposure level of a site. This pat odoxus fluviatilis was consistently absent in the North Sea tern is explained by the high hydrodynamic pressure on study sites, causing high dissimilarities in abundance the thallus of F. vesiculosus at highly exposed sites, whi among regions. ch removes epiphytic algae and prevents benthic suspen This study also showed a higher variability of epiphytes sion feeders from settling on the algae. and mobile faunal community structure in the Baltic Sea We also predicted that the relationship between abiotic than in the North Sea. It has been proposed that pro (exposure), biotic factors (host frond segment) and epi cesses affect ecosystems simultaneously at various spatial phytes (species composition, cover) would vary among and temporal scales (Denny et a l 2004; Fraschetti et al. different marine regions (North versus Baltic Seas). 2005; Kotta et al. 2008). The relative importance of small- Indeed, our study showed that the North and Baltic Seas and large-scale processes on the formation of marine had different epiphytic and mobile faunal species compo communities is little known and it is likely the patterns sitions and dominance structures. The epiphytic and vary among regions (e.g. H ew itt et al. 2007; Kotta & Wit mobile faunal taxa contributing most to the dissimilarity man 2009). Our study indicates that large-scale factors between North and Baltic Sea communities were the mostly determine the distribution patterns of epiphytes in brown alga P. littoralis and the herbivore Theodoxus flu the North Sea and within these patterns, processes operat viatilis, respectively. Both species have a strong degree of ing at microscale (e.g. due to frond segment) further tolerance to lowered salinity which consequently enables modify the epiphyte communities. On the other hand, Species composition Dominance structure .11 R e g i o n R e g i o n ▲ North sea ▲ North sea • Baltic sea • Baltic sea • • ▲▲ Cfl o M A S % xz Q. Q. LU M0 .06 CC C * E• fi cu ■Q O MAM R e g i o n R e g i o n ▲ North sea ▲ North sea • Baltic sea • Baltic sea Fig 3. Non-metric multidimensional scaling (nMDS) plots showing the effect of sea region and wave exposure (E = exposed; M = moderately exposed and S = sheltered sites) on epiphytic (coverage) and mobile faunal (abundance) community composition and structure. Lefthand plots are composed from presence/absence transformed data, righthand plots from untransformed data. Some marks are overlapped on the figure because epiphytes and mobile fauna were absent in these samples, thus representing 100% similarity. 92 Marine Ecology32 (Suppl. 1) (2011) 87-95 © 2011 Blackwell Verlag GmbH Kersen, Kotta, Bucas, Kolesova & Deljere Epiphytes and associated fauna on brown alga large-, meso- and microscale processes are all equally communities. The formation of epiphytic and mobile important in determining the distribution patterns of epi faunal communities is an interplay of factors operating phytes in the Baltic Sea. through micro- to regional scales. In general, associated faunal community composition was different between all levels of wave exposure in both Acknowledgements marine regions, whereas epiphytic composition and struc ture did not significantly differ between moderately We are especially thankful to Lena Kautsky for her valu exposed and sheltered sites in the North Sea. This able advice on sampling and species identification. We indicates that epiphytic algae inhabiting the North Sea thank Stein Fredriksen, Kjersti Sjotun and Jan Rueness tolerate a larger range of exposure than those inhabiting for their kind support with laboratory species identifica the Baltic Sea. tion. This research was supported by the European Social The effect of host frond segments on the patterns of Fund’s Doctoral Studies and Internationalisation Pro epiphytes varied among North and Baltic Seas, supporting gramme DoRa. Funding for this research was provided by the hypothesis that there are different factors (levels) target-financed projects SF0180013s08 of the Estonian forming different epiphytic communities on F. vesiculosus Ministry of Education and by the Estonian Science Foun in the Baltic and North Sea. It seems likely that at smaller dation grants 7813 and 8254. The study was partly carried spatial scales, biotic factors [i.e. frond segment) play a out in the frame of an advanced course by the Nordic more important role in epiphytic communities in the Marine Academy ‘Biodiversity of northeast Atlantic mac Baltic Sea, whereas abiotic factors [i.e. wave exposure) are roalgae’ at the Espeland Marine Biological Station. more important in the North Sea. Our understanding of the causes of local species diver R eferences sity in marine habitats mostly originates from observa tions performed at small spatial scales. Comparing local Anderson M.J., Gorley R.N., Clarke K.R. (2008) PERMANOVA and regional variability of epiphyte communities, our + for PRIMER: Guide to Software and Statistical Methods. study clearly demonstrated that regional differences define PRIMER-E, Plymouth: 214 pp. broad patterns of species diversity and are among the Arrontes J. (1990) Composition, distribution on host, and most significant factors explaining population variability seasonality of epiphytes on tree intertidal algae. Botanica in these marine environments. However, our study was Marina, 33, 205-211. limited to two-region cases and further studies from mul Berger R., Bergström L., Graneli E., Kautsky L. (2004) How tiple regions may provide us with a generic knowledge of does the eutrophication affect different life stages of Fucus vesicidosus in the Baltic Sea? - a conceptual model. Hydro the processes shaping epiphyte communities. Besides the biologia, 514, 243-248. spatial aspect, epiphytes are known to have a strong com Borum J. ( 1985) Development of epiphytic communities on ponent of seasonal variability (Borum 1985; Vairappan eelgrass ( Zostera marina ) along a nutrient gradient in a 2007; Torn et al. 2010). This was not considered in the Danish estuary. Marine Biology, 87, 211-218. current study but may be highly relevant, as different Cardinal A., Lesage V. (1992) Distribution of the epiphytes regions are characterized by different types and intensity Pilayella littoralis and Polysiphonia lanosa on Ascophyllum of seasonality. nodosum in the Bay of Fundy. Cahiers de Biologie Marine, To conclude, all investigated factors contributed sig 33, 125-135. nificantly to the variability in species composition and Clarke K.R., Gorley R.N. (2006) PRIMER v6: User Man coverage of epiphytes and mobile faunal communities ual/Tutorial. PRIMER-E, Plymouth: 192 pp. on F. vesiculosus. The North and Baltic Seas each had Clarke K.R., Warwick R.M. (2001)Change in Marine Commu distinct epiphyte and mobile faunal communities. nities: An Approach to Statistical Analysis and Interpretation, Within the studied regions, wave exposure and frond 2nd edn. PRIMER-E, Plymouth: 172 pp. segment contributed significantly to the variability in Denny M.W., Helmuth B., Leonard G.H., Harley C.D.G., Hunt species composition and dominance structure of epi L.J.H., Nelson E.K. (2004) Quantifying scale in ecology: phytes on F. vesiculosus in the N orth Sea and Baltic Sea. lessons from a wave-swept shore. Ecological Monographs, 74, Large-scale factors greatly determine the distribution 513-532. patterns of epiphytes in the North Sea, whereas large-, Fraschetti S., Terlizzi A., Benetti-Cecchi L. (2005) Pattern of meso- and microscale processes were all equally impor distribution of marine assemblages from rocky shore: tant in determining the distribution patterns of evidence of relevant scales of variation. Marine Ecology epiphytes in the Baltic Sea. The study indicated that Progress Series, 296, 13-29. there is no clear spatial scale where environmental H ewitt J.E., Thrush S.F., Dayton P.K., Bonsdorff E. (2007) The variables best predicted epiphyte and mobile faunal effect of spatial and temporal heterogeneity on the design Marine Ecology32 (Suppl. 1 ) (2011 ) 87-95 © 2011 Blackwell Verlag GmbH 93 Epiphytes and associated fauna on brown alga Kersen, Kotta, Bucas, Kolesova & Deljere and analysis of empirical studies of scale-dependent systems. Nylund G.M., Pavia H. (2005) Chemical versus mechanical American Naturalist, 169, 398-408. inhibition of fouling in the red alga Dilsea carnosa. Marine Honkanen T., Jormalainen V. (2005) Genotypic variation in Ecology Progress Series, 299, 111-121. tolerance and resistance to fouling in the brown alga Fucus Orav-Kotta H., Kotta J. (2004) Food and habitat choice of the vesiculosus. Oecologia, 144, 196-205. isopod Idotea baltica in the northeastern Baltic Sea. Hydro Johnson S.C., Scheibling R.E. (1987) Structure and dynamics biologia, 514, 79-85. of epifaunal assemblages on intertidal macroalgae Ascophyl Paine R.T. (1990) Benthic macroalgal competition: complica lum nodosum and Fucus vesiculosus in Nova Scotia, Canada. tions and consequences. Journal of Phycology, 26, 12-17. Marine Ecology Progress Series, 37, 209-227. Pavia H., Carr H., Äberg H.C. ( 1999) Habitat and feeding Kangas P., Skoog G. (1978) Salinity tolerance of Theodoxus preferences of crustacean mesoherbivores inhabiting the fluviatilis (Mollusca, Gastropoda) from freshwater and from brown seaweed Ascophyllum nodosum (L.) Le Joi. and its different salinity regimes in the Baltic Sea. Estuarine, Coastal epiphytic macroalgae. Journal of Experimental Marine Biology and Shelf Science, 79, 533-540. and Ecology, 236, 15-32. Kiirikki M. ( 1996a) Mechanisms affecting macroalgal zonation Ráberg S., Kautsky L. (2007) A comparative biodiversity study in the northern Baltic Sea. European Journal of Phycology, of associated fauna of perennial fucoids and filamentous 31, 225-232. algae. Estuarine, Coastal and Shelf Science, 73, 249-258. Kiirikki M. ( 1996b) Experimental evidence that Fucus vesicidosns Rindi F., Guiry M.D. (2004) Composition and spatio temporal (Phaeophyta) controls filamentous algae by the means of the variability of the epiphytic macroalgal assemblage of Fucus whiplash effect. European Journal of Phycology, 31, 61-66. vesiculosus Linnaeus at Clare Island, Mayo, western Ireland. Korpinen S., Jormalainen V., Honkanen T. (2007) Effects of Journal of Experimental Marine Biology and Ecology, 311, nutrients, herbivory, and depth on the macroalgal commu 233-252. nity in the rocky sublittoral. Ecology, 88, 839-852. Rohde S., Hiebenthal C., Wahl M., Karez R., Bischof K. (2008) Kotta J., Witman J. (2009) Regional-scale patterns. In: Wahl Decreased depth distribution of Fucus vesiculosus in the M.. (ed), Marine Hard Bottom Communities. Ecological Western Baltic: effects of light deficiency and epibionts on Studies, 206. Springer, Heidelberg: 89-99. growth and photosynthesis. European Journal Phycology, 43, Kotta J., Paalme T., Martin C., Makinen A. (2000) Major 143-150. changes in macroalgae community composition affect the Russell G. (1994) A Baltic variant of Pilayella littoralis (Algae, food and habitat preference of Idotea baltica. International Fucophyceae). Annales Botanici Fennici, 31, 127-138. Review of Hydrobiologia, 85, 693-701. Schiewer U. (2008) The Baltic Sea. In: Schiewer U.. (Ed). Ecol- Kotta J., Paalme T., Püss T., H erkül K., Kotta I. (2008) C ontri ogy of Baltic Coastal Waters. Ecological Studies 197. Springer, bution of scale-dependent environmental variability on the Berlin: 1-22. biomass patterns of drift algae and associated invertebrates Snoeijs P. ( 1999) Marine and brackish waters. Acta Phytogeo- in the Gulf of Riga, northern Baltic Sea. Journal of Marine graphica Suecica, 84, 187-212. Systems, 74(Suppl. 1), S116-S123. Stachowicz J.J., Whitlatch R.B. (2005) Multiple mutualists pro Kraberg A.C., Norton T.A. (2007) Effect of epiphytism on vide complementary benefits to their seaweed host. Ecology, reproductive and vegetative lateral formation in the brown, 86, 2418-2427. intertidal seaweed Ascophyllum nodosum (Phaephyceae). Steel J.B., W ilson J.B. (2003) W hich is the phyte in epiphyte? Phycological Research, 55, 17-24. Folia Geobotánica, 38, 97-99. Lobban C.S., Baxter D.M. (1983) Distribution of the red algal Suursaar Ü., Sooäär J. (2007) Decadal variations in mean and epiphyte Polysiphonia lanosa on its brown algal host Asco extreme sea level values along the Estonian coast of the Bal phyllum nodosum in the Bay of Fundy, Canada. Botanica tic Sea. Tellus, 59A, 249-260. Marina, 26, 533-538. Torn K., Krause-Jensen D., Martin G. (2006) Present and past Lobban C.S., Harrison P.J. (2000) Seaweed Ecology and Physiol depth distribution of bladderwrack ( Fucus vesicidosus) in the ogy. Cambridge University Press, Cambridge: 366 pp. Baltic Sea. Aquatic Botany, 84, 53-62. Longtin C.M., Scrosati R.A., Whalen G.B., Garbary D.J. (2009) Torn K., Martin G., Kotta J., Kupp M. (2010) Effects of differ Distribution of algal epiphytes across environmental gradi ent types of mechanical disturbances on a charophyte domi ents at different scales: intertidal elevation, host canopies, nated macrophyte community. Estuarine Coastal and Shelf and host fronds. Journal of Phycology, 45, 820-827. Science, 87, 27-32. Lüning K.. (1990) Seaweeds: Their Environment, Biogeography, Vairappan C.S. (2007) Seasonal occurrences of epiphytic algae and Ecophysiology. Wiley-Interscience, New York: 527 pp. on the commercially cultivated red alga Kappaphycus al Molina-Montenegro M.A., Munoz A.A., Badano E.I., Morales varezii (Solieriaceae, Gigartinales, Rhodophyta). Develop B.W., Fuentes K.M., Cavieres L.A. (2005) Positive interac ments in Applied Phycology, 1, 385-391. tions between macroalgal species in a rocky intertidal zone Wahl M. (2009) Epibiosis: Ecology, Effects and Defences. In: and their effects on the physiological performance of Ulva Wahl M. (ed), Marine Hard Bottom Communities. Ecological lactuca. Marine Ecology Progress Series, 292, 173-180. Studies, 206. Springer, Heidelberg: 61-72. 94 Marine Ecology32 (Suppl. 1) (2011) 87-95 © 2011 Blackwell Verlag GmbH Kersen, Kotta, Bucas, Kolesova & Deljere Epiphytes and associated fauna on brown alga Wikström S.A., Pavia H. (2003) Chemical settlement inhibition levels (presence/absence transformed data, coverage of versus post-settlement mortality as an explanation for differ epiphytes on host). ential fouling of two congeneric seaweeds. Oecologia, 138, Table S6. Results of SIMPER analyses testing for differ 223-230. ences in the dominance structure of epiphytic communi Worm B., Lotze H.K., Hillebard H., Sommer U. (2002) ties on Fucus vesiculosus between different wave exposure Consumer versus resource control of species diversity and levels (untransformed data, coverage of epiphytes on ecosystem functioning. Nature, 417, 848-851. host). Table S7. Results of SIMPER analyses testing for differ Supporting Information ences in the species composition of mobile faunal com munities on Fucus vesiculosus between different wave Additional Supporting Information may be found in the exposure levels (presence/absence transformed data, online version of this article: abundance of fauna on host). Table SI. Results of SIMPER analyses testing for differ Table S8. Results of SIMPER analyses testing for differ ences in the species composition of epiphytic communi ences in the dominance structure of mobile faunal com ties on Fucus vesiculosus between the North and Baltic munities on Fucus vesiculosus between different wave Seas (presence/absence transformed data, coverage of epi exposure levels (untransformed, abundance of fauna on phytes on host). host). Table S2. Results of SIMPER analyses testing for differ Table S9. Results of SIMPER analyses testing for differ ences in the dominance structure of epiphytic communi ences in the species composition of epiphytic communi ties on Fucus vesiculosus between the North and Baltic ties on Fucus vesiculosus between different frond segments Seas (untransformed data, coverage of epiphytes on host). (presence/absence transformed data, coverage of epi Table S3. Results of SIMPER analyses testing for differ phytes on host). ences in the species composition of mobile faunal com Table SIO. Results of SIMPER analyses testing for dif munities on Fucus vesiculosus between the North and ferences in the dominance structure of epiphytic commu Baltic Seas (presence/absence transformed data, abun nities on Fucus vesiculosus between different frond dance of fauna on host). segments (untransformed data, coverage of epiphytes on Table S4. Results of SIMPER analyses testing for differ host). ences in the dominance structure of mobile faunal commu Please note: Wiley-Blackwell are not responsible for the nities on Fucus vesiculosus between the North and Baltic content or functionality of any supporting materials sup Seas (untransformed data, abundance of fauna on host). plied by the authors. Any queries (other than missing Table S5. Results of SIMPER analyses testing for differ material) should be directed to the corresponding author ences in the species composition of epiphytic communi for the article. ties on F. vesiculosus between different wave exposure Marine Ecology32 (Suppl. 1 ) (2011 ) 87-95 © 2011 Blackwell Verlag GmbH 95 anmarine evolutionary perspective ecology » Marine Ecology. ISSN 0173-9565 ORIGINAL ARTICLE Food resource use in sympatric juvenile plaice and flounder in estuarine habitats Stefano Mariani, Ciara Boggan & David Balata Marine Biodiversity, Ecology & Evolution, UCD School of Biology & Environmental Science, University College Dublin, Belfield, Dublin, Ireland Keywords A bstract Adaptation; coastal habitats; flatfish; Irish Sea; Platichthys flesus; Pleuronectes platessa ; Environmental conditions in estuarine habitats can vary greatly and influence trophic niche. the composition of fish assemblages and their trophic interrelationships. We investigated feeding habits and patterns of diet overlap in juvenile plaice ( Pleu Correspondence ronectes platessa) and flounder ( Platichthys flesus) from two estuarine habitats Stefano Mariani, Marine Biodiversity, Ecology in the Irish Sea. Plaice was found to vary its diet significantly across environ & Evolution, UCD School of Biology & ments, whereas flounder exhibited a more consistent and homogeneous feeding Environmental Science, University College Dublin, Belfield, Dublin 4, Ireland. pattern. Importantly, sympatric fish sampled at the same station were shown to E-mail: [email protected] reduce diet overlap. The results support the view that environmental hetero geneity in estuaries maintains a wide range of selective forces, the net outcome Accepted: 17 November 2010 of which can produce a diverse array of feeding adaptations among interacting species. doi:10.1111/j.1439-0485.2010.00419.x inhabiting coastal ecosystems (Beyst et al. 1999; D arna- Introduction ude et al. 2001; Mariani et al. 2002; Platell et al. 2006; Despite extensive habitat destruction, pollution and long Russo et al. 2008). Juveniles of flatfish species (Order term irresponsible exploitation regimes (Lotze et al. Pleuronectiformes) are found in abundance in estuarine 2006), estuarine and coastal areas still retain a pivotal role and coastal fish assemblages worldwide, making them in aquatic production processes, conservation, resource good candidates for studies of resource partitioning. In use, economy and commerce. Most of such ‘transitional’ particular, European flounder (Platichthys flesus L.) and habitats represent vital ‘nurseries’ for the juvenile stages plaice ( Pleuronectes platessa L.) represent key species in of several commercially important fish (Beck et al. 2001; cold temperate areas in the Northeast Atlantic and the Kraus & Secor 2005), providing abundant food resources, latter especially sustains a very valuable commercial fish shelter and favourable conditions for rapid growth ery. After hatching, juvenile flounder and plaice use (Haedrich 1983). Yet, estuaries, lagoons and tidal flats are shallow nursery grounds during the first months of life, also characterised by an unparalleled spatial and temporal between March and October (Russell 1976; Gibson 1994; environmental heterogeneity (Elliott & Quintino 2007), Raffaelli & Hawkins 1996), exhibiting at this time a high which necessarily poses severe adaptive challenges to the degree of spatial overlap. organisms that spend at least part of their life cycle Despite the reported similarity in diet (Gibson 1994; therein. Piet et al. 1998), most studies suggest that sympatric pla One main ecological challenge is to be able to parti ice and flounder segregate trophic niches (Aarnio et al. tion resources in a densely populated and variable envi 1996; Beyst et al. 1999; Amezcua et al. 2003; Andersen ronment. Similar species that occupy the same habitat at et al. 2005; Russo et al. 2008). Most comparative studies the same time will likely consume slightly different prey emphasise inter-specific variation and competition in a to minimise niche overlap (Schoener 1974), and a num specific spatial context - sometimes going as far as pro ber of studies have shown this to be the case for fish viding mechanistic explanations for the observed variation 96 Marine Ecology32 (Suppl. 1 ) (2011 ) 96-101 © 2011 Blackwell Verlag GmbH Mariani, Boggan & Balata Food resource use in plaice and flounder in diet (Bels & Davenport 1996; Gronkjaer et al. 2007; Sampling Russo et al. 2008) - but tend to overlook the overall role of coastal environmental heterogeneity in influencing the Juveniles of plaice and flounder were collected using a adaptive responses of species. 10-m long, 4-m high, hand-dragged seine (2-mm mesh), Here we compare the feeding habits of sympatric between June and July 2008, which corresponds to the flounder and plaice in two different estuarine environ period of maximum abundance and activity of 0 -group ments in the Irish Sea and we test the hypothesis that flatfishes in inshore tidal areas in the Northeast Atlantic local conditions will affect the dietary patterns in these (Gibson 1994; Raffaelli 8 c Hawkins 1996). In each locality, species, resulting in spatial variations in their trophic specimens were collected once a week, over 4 weeks, dur inter-relationships. ing daytime, at high and low tide, to maximize the repre sentation of the sample for both species. Specimens were counted each time, and a random subsample was sacri Material and Methods ficed using an overdose of phenoxyethanol and later Study area placed in a 5% formalin solution for preservation and identification. The study was carried out in two inshore tidal inlets in the Irish Sea: North Bull Island (53°22' N, 6°07' W), in Dublin Bay, and Wexford Harbour (52°20' N, 6°27' W), Data processing and analysis at the southernmost end of the Irish Sea basin (Fig. 1). All preserved individuals were measured with a calliper North Bull Island has previously been shown to have (fork length, LF, to the nearest mm) and their stomach higher average salinity (37 psu) and lower average tem contents emptied into a Petri dish and observed under a perature (17 °C) than Wexford Harbour (27 psu and stereomicroscope. Prey items were classified to the lowest 19 °C) throughout the year (Craig et al. 2008). Both taxonomic level possible and recorded as present or localities are characterized by sandy/muddy bottoms, absent. which remain completely exposed during low tide. Within The multivariate dataset was explored and represented each location, two stations ~500 m apart were chosen for using a two-dimensional unconstrained principal coordi sampling. North Bull ¿*£7 r / C Island IRELAND Dublin N. Bull Island Plaice S1 Plaice S2 Flounder S1 Flounder S2 Wexford H Wexford Fig. 1. Map of study areas with bar plots Harbour representing the recorded numbers of plaice and flounder (annotations S1 and S2 refer to Plaice S1 Plaice S2 Flounder S1 Flounder S2 the two stations within each location). Marine Ecology32 (Suppl. 1) (2011) 96-101 © 2011 Blackwell Verlag GmbH 97 Food resource use in plaice and flounder Mariani, Boggan & Balata nate ordination (PCO) based on Gower’s dissimilarity Results measure (Gower 1966). Permutational multivariate analy sis of variance (PERMANOVA; Anderson 2001) was then A total of 1144 flounder and 810 plaice were collected. used to analyse the variation in feeding habits of plaice Llounder were dominant in Wexford Harbour (975 versus and flounder between and within the two different estua- 184) and plaice in North Bull Island (626 versus 169) rine habitats. The experiment consisted of a three-way (Fig- !)• design with ‘Species’ (‘Spe’, two levels) as a fixed factor, A total of 202 fish were examined for stomach con ‘Locality’ (‘Loc’, two levels) as a random and crossed fac tents: 50 flounder (31 + 19) and 50 plaice (30 + 20) from tor, and ‘Station’ (‘Sta’, two levels) as a random factor two stations in North Bull Island, 50 flounder (30 + 20) nested in ‘Locality’. Pairwise tests were also conducted to and 52 plaice (36 + 16) from the two stations in Wexford pinpoint the levels responsible for significant interactions. Harbour. Pish size ranged between 22 and 127 mm and Leeding was also examined by comparing the variation no differences were observed between species and loca of most abundant taxa, which were ‘pooled’ into taxo- tions (ANOVA: F = 1.42, df = 3, P = 0.24) and none of nomical/ecological guilds as follows: ‘Copepods’ (com them had an empty stomach. Twenty-one different prey prising Harpacticoida, Cyclopoida and Calanoida), items were identified, with mysids being generally the pri ‘Worms’ (comprising Polychaeta, Oligochaeta and Nema mary resource for flounder and polychaetes being the toda), ‘Amphipoda + Isopoda’, and ‘Other Malacostraca’ most abundant prey in plaice (Table 1). (comprising Mysidacea, Brachyura and Caridea). These PCO ordination of sample centroids showed a more data were analysed by univariate ANO VA, applying the clumped distribution for flounder and a greater scattering same design described for PERMANOVA, and after test of the plaice data points (Pig. 2). PERMANOVA detected ing for homogeneity of variance using the Cochran’s significant differences for both the interactions: Spe*Loc C-test (Underwood 1997). Student-Newman-Keuls and Spe>fSta(Loc), as well as for the effect ‘Station(Loc)’, (SNK) tests were used for post hoc multiple pairwise revealing the dependence of locality-specific variation in comparisons. governing feeding interactions between these species Table 1. Frequency of occurrence of all prey Items found In the stomach contents of plaice (P) and flounder (F) from Bull Island (B) and Wexford Harbour (W) at stations 1 and 2. FB1 (19) FB2 (31) FW1 (30) FW2 (20) PB1 (30) PB2 (20) PW1 (16) PW2 (36) Harpacticoida 0.11 0.19 0.2 0.65 0.9 0.2 0 0.06 Cyclopoida 0 0.1 0 0 01 0.2 0 0 Calanoida 0 0 0 0 0.03 0 0 0 Polychaeta Err. 0.05 0.1 0 0.15 0.47 0.55 0.75 0.56 Polychaeta Sed. 0.21 0.03 0.2 0.25 0.03 0.35 0.18 0 Amphipoda 0.53 0.48 0.07 0.2 0.27 0.25 0.19 0.31 Isopoda 0.1 0.1 0.03 0 0.03 0.1 0 0.08 Bivalvia 0 0.6 0.03 0 0.2 0.4 0 0 Ostracoda 0 0.3 0.13 0 0.03 0 0 0.17 eggs 0 0.3 0.03 0.5 0.07 0.1 0 0.06 Nematoda 0 0 0 0 0 0 0 0.06 Brachyura 0.16 0.03 0.03 0.25 0 0 0 0 Caridea 0 0 0.03 0 0.50 0 0 0 Mysidacea 0.63 0.23 0.57 0.6 0.1 0 0 0 Oligochaeta 0 0 0 0 0.03 0.05 0 0 Holoturldae 0 0 0 0 0 0 0.06 0 algae 0.1 0.13 0.03 0 0.27 0.2 0.12 0 Hydrozoa 0.05 0 0 0 0 0 0 0.06 Cirripedia 0 0 0 0 0 0 0.06 0 fish 0 0.03 0 0 0 0 0 0 Tunicata 0 0 0.03 0 0 0 0 0 unidentified 0 0 0.03 0 0 0 0 0 Numbers In brackets In the column headers refer to sample size. Values In bold represent 'Important' (>25%) or 'dominant' (>50%) prey Items, according to Albertlne-Berhaut (1973). 98 Marine Ecology32 (Suppl. 1 ) (2011 ) 96-101 © 2011 Blackwell Verlag GmbH Mariani, Boggan & Balata Food resource use in plaice and flounder 40 1.2 Copepods Amphipoda + Isopoda 1.0 T 0.8 A FB1 0.6 PW1 0.4 O O PW2 • F W 1 FB2 0.2 0 FW2 1 0.0 i B-1 B-2 W-1 W-2 B-1 B-2 W-1 W-2 PB2 1.2 - 1,21 Other malacostraca “W orm s” û - - 2 0 1.0 1.0 - 0.8 0.8 - A PB1 0.6 0 .6 - -40 -40 -20 0 20 40 0.4 0.4- PC01 (51.3% of total variation) 0.2 0 .2 - Fig. 2. Ordination plot of sample centroids Inferred with Principal I 0.0 0 .0 - i i i coordinate analysis. F is for flounder (black), P is for plaice (grey), B is B-1 B-2 W-1 W-2 B-1 B-2 W-1 W-2 for 'North Bull Island' (triangles) and W is for 'Wexford Harbour' Flounder = i Plaice (circles). Fig. 3. Frequencies of four main food categories In plaice and floun der from two estuaries. Results of univariate ANOVA tests are In the Table 2. Results summary table for PERMANOVA procedure con text. ducted on presence/absence stomach content data. Values in bold are significant with a = 0.05. source df MS pseudo-F P-value Discussion Spe 1 68,768 2.215 0.114 The study of the feeding habits and resource partitioning Loc 1 20,513 2.073 0.081 in closely related fish can be very useful to understand Sta (Loc) 2 9171.8 3.400 0.001 Spe*Loc 1 31,029 2.612 0.041 the flows of energy across the food web (Darnaude, 2005) Spe*Sta(Loc) 2 11,756 4.358 0.001 and provide important insights into the trophic flexibility residual 191 2697.5 of interacting species (Mariani et a í 2002; Platell et a í total 198 2006; Russo et al. 2008). Yet, the quantification of diet overlap or niche segregation - even when supported by (Table 2). Pairwise tests conducted on both interactions robust explanations of the processes underlying the pat showed consistent significant differences between species terns - may still only provide a very narrow picture of within stations but only plaice was shown to vary its diet the true flexibility of a species’ trophic ecology. significantly between locations. This study expands the analysis of resource partitioning Univariate analyses were also informative in describing between two common flatfish, plaice and flounder, by diet variation between and within species (Fig. 3). Signifi taking into consideration the naturally high environmen cant differences between species were detected in the cate tal heterogeneity of estuarine habitats, and performing gory ‘Worms’ (F = 32.5, P = 0.01). The interactions diet comparisons in two different estuarine habitats. Spe*Loc for ‘Other Malacostraca’ (F = 42.9, P = 0.02) and Wexford Harbour was found to be dominated by Spe*Sta (Loc) for ‘Copepods’ (F = 4.3, P = 0.01) were also flounder, whereas North Bull Island showed a predomi significant, and the Spe*Loc interaction for both ‘Cope nance of plaice. This is probably linked to the lower salin pods’ (F = 10.7, P = 0.08) and ‘Amphipoda + Isopoda ities observed year-round in Wexford (Craig et a í 2008) (F = 14.4, P = 0.06) were only marginally above the proba and the well-known pronounced preference of flounder bility threshold. SNK tests confirmed significant differences for brackish environments. This species appeared to exhi between plaice and flounder across all stations in terms of bit a more spatially homogeneous food spectrum ‘Copepods’, but only in Wexford for ‘Other Malacostraca’. (clumped scatter of data in Fig. 2) relative to plaice, Marine Ecology32 (Suppl. 1 ) (2011 ) 96-101 © 2011 Blackwell Verlag GmbH 99 Food resource use in plaice and flounder Mariani, Boggan & Balata which exhibits remarkable diet changes between the two Conclusions study locations (wider distribution of data in Fig. 2). Univariate tests on the most abundant prey items We have shown that the patterns of niche overlap and (Fig. 3) illustrate well the extent and nature of diet varia resource partitioning between plaice and flounder in estu tion between and within plaice and flounder. Plaice seems arine habitats vary depending on the specific system stud to be the most efficient forager of polychaetes and other ied. This strengthens the view that estuaries, lagoons and worm-like prey, whereas flounder consistently consumes a other transitional habitats may represent a heterogeneous greater amount of malacostraca, particularly mysids ‘mosaic’ of selective forces (Weinig & Schmit 2004), (Table 1). The frequency of different preys across floun which is crucial to the evolution of the life-histories of der samples is much more even than in plaice: for coastal marine fish (Mariani 2006) and perhaps influences instance, considerable frequencies of copepods, decapods the functioning of more ‘extended’ phenotypic traits and bivalves are found in the stomachs of plaice from (sensu Dawkins 1982), such as trophic interactions and North Bull Island, whereas virtually only polychaetes and community structure. amphipods are consumed in Wexford. One explanation Future studies should attempt to evaluate whether the may reside in the fact that Wexford is a flounder-domi observed trophic flexibility results entirely from pheno nated habitat, which might drive the less numerous plaice typic plasticity (Gronkjaer et a í 2007) or also to some to select only preys, such as polychaetes and amphipods, extent from genomic adaptation. that are not ‘preferred’ by flounder, as indeed appears to be the case in Wexford Harbour. Intra-specific competi Acknowledgements tion may also play a role, prompting plaice to select a wider range of prey in the location where it is more fre The study was based on data collected as part of CB’s quent and dominant {i.e. North Bull Island). final year project, internally funded by the UCD School On the other hand, flounder does not seem to go of Biology and Environmental Science. We would like to through the same ‘diet switch’, when the two estuaries thank Debbi Pedreschi for help during sampling and the are compared. The only prey categories affected are am editors and two anonymous reviewers for their valuable phipods and isopods, which show a low frequency in suggestions. Wexford (Fig. 3), although this is counterbalanced by the intense consumption of other malacostraca (espe R eferences cially mysids). Without a parallel investigation on the abundance of potential prey (Wouters & Cabral 2009), Aarnio K., Bonsodorff E., Rosenback N. (1996) Food and feed it remains difficult to assess to what extent the results ing habits of juvenile flounder Platichthys flesus (L.) and may be influenced by differential availabilities of turbot Scophthalmus maximus (L.) in the Aland archipelago, resources; however, the present data unambiguously northern Baltic Sea. Journal of Sea Research, 36, 311— show that estuarine environmental heterogeneity deter 320. mines changes in resource use in plaice and flounder Albertine-Berhaut J. (1973) Biologie des stades juveniles de and that the responses are different for each species, teleosteens, Mugilidae, Mugil auratus Risso 1810, Mugil capito resulting in the variation of patterns of trophic inter Cuvier 1829 et Mugil saliens Risso 1810. Regime alimentaire.. Station Marine d’Endoume, Marseille: 2 pp 251-266. relationships between species. Amezcua F., Nash R.D.M., Veale L. (2003) Feeding habits of The feeding habits of plaice and flounder vary the Order Pleuronectiformes and its relation to the sediment between and within localities (Table 2), resulting in very type in the north Irish Sea. Journal of the Marine Biological interesting patterns at finer scales: plaice in Wexford Association of the United Kingdom, 83, 593-601. consumed significantly different prey than plaice in Bull Andersen B.S., Cari I.D., Gronkjær P., Stottrup J.C. (2005) Island, and in each station within estuary, plaice and Feeding ecology and growth of age 0 year Platichthys flesus flounder consistently preyed upon different items, effec (L.) in a vegetated and a bare sand habitat in a nutrient rich tively reducing niche overlap. This view of trophic fjord. Journal of Fish Biology, 66, 531-552. niches and resource use makes the fluctuating, unstable, Anderson M.J. (2001) A new method for non-parametric mul seemingly chaotic estuarine habitat a much more fine- tivariate analysis of variance. Austral Ecology, 26, 32-46. tuned, sophisticated, elegantly functioning system than Beck M.V., Kenneth L., Heck J.R., Kenneth W.A., Childers generally believed. Further studies including a greater D.L., Eggleston D.B., Gillanders B.M., Halpern B., Hays number of estuaries, and possibly a greater number of C.G., H oshino K., Minello T.J., O rth R.J., Sheridan P.F., species, may add operational complexity, but would cer Weinstein M.P. (2001) The identification, conservation, and tainly help to characterise these mechanisms with greater management of estuarine and marine nurseries for fish and confidence. invertebrates. BioScience, 51, 633-641. 100 Marine Ecology32 (Suppl. 1 ) (2011 ) 96-101 © 2011 Blackwell Verlag GmbH Mariani, Boggan & Balata Food resource use in plaice and flounder Bels V.L., Davenport J. ( 1996) A comparison of food capture Lotze H.K., Lenihan H.S., Bourque B.J., Bradbury R.H., Cooke and ingestion in juveniles of two flatfish species, Pleuronectes R.G., Kay M.C., Kidwell S.M., Kirby M.X., Peterson C.H., platessa and Limanda limanda (Teleostei: Pleuronectifor Jackson J.B.C. (2006) Depletion, degradation and recovery mes). Journal of Fish Biology, 49, 390-401. potential of estuaries and coastal seas. Science, 312, 1806- Beyst C., Cattrijsse A., Mees J. (1999) Feeding ecology of juve 1809. nile flatfishes of the surf zone of a sandy beach. Journal of Mariani S. (2006) Life-history- and ecosystem-driven variation Fish Biology, 55, 1171-1186. in composition and residence pattern of seabream species Craig G., Paynter D., Coscia I., Mariani S. (2008) Settlement (Perciformes: Sparidae) in two Mediterranean coastal of gilthead sea bream Sparus aurata L. in a southern Irish lagoons. Marine Pollution Bidletin, 53, 121-127. Sea coastal habitat. Journal o f Fish Biology, 72, 287-291. Mariani S., Maccaroni A., Massa F., Rampacci M., Tancioni L. Darnaude A.M. (2005) Fish ecology and terrestrial carbon use (2002) Lack of consistency between the trophic interrela in coastal areas: implications for marine fish production. tionships of five sparid species in two adjacent central Journal o f Anim al Ecology, 74, 864-876. Mediterranean coastal lagoons. Journal o f Fish Biology, 61, Darnaude A.M., Harmelin-Vivien M.L., Salen-Picard C. (2001) 138-147. Food partitioning among flatfish (Pisces: Pleuronectiforms) Piet G.J., Pisterer A.B., Rijnsdorp A.D. ( 1998) On factors juveniles in a Mediterranean coastal shallow sandy area. structuring the flatfish assemblage in the southern North Journal of the Marine Biological Association of the United Sea. Journal of Sea Research, 40, 143-152. Kingdom, 81, 119-127. Platell M.E., Orr P.A., Potter I.C. (2006) Inter- and intraspe Dawkins R. ( 1982) The Extended Phenotype. Oxford University cific partitioning of food resources by six large and abun Press, Oxford, 307 pp. dant fish species in a seasonally open estuary. Journal o f Fish Elliott M., Quintino V.M. (2007) The estuarine quality para Biology, 69, 243-262. dox, environmental homeostasis and the difficulty of detect Raffaelli D.G., Hawkins S. (1996) Intertidal Ecology. Chapm an ing anthropogenic stress in naturally stressed areas. Marine 8c Haii, London: 356 pp. Pollution Bidletin, 54, 640-645. Russell S.F. ( 1976) The Eggs and Planktonic Stages of British Gibson R.N. (1994) Impact of habitat quality and quantity on Marine Fishes. Academic Press , London: 524 pp. the recruitment of juvenile flatfishes. Journal of Sea Research, Russo T., Pulcini D., O’Leary A., Cataudella S., Mariani S. 32, 191-206. (2008) Relationship between body shape and trophic niche Gower J.C. ( 1966) Some distance properties of latent root and segregation in two closely related sympatric fishes. Journal of vector methods used in multivariate analysis. Biometrika, 53, Fish Biology, 73, 809-828. 325-338. Schoener T.W. (1974) Resource partitioning in natural com Gronkjaer P., Cari J.D., Rasmussen T.H., Hansen K.W. (2007) munities. Science, 185, 27-39. Effect of habitat shifts on feeding behaviour and growth of U nderw ood A.J. (1997) Experiments in Ecology. Their Logical 0 year-group flounder Platichthys flesus (L.) transferred Design and Interpretation Using Analysis of Variance. Cam between macroalgae and bare sand habitats. Journal o f Fish bridge University Press, Cambridge: 504 pp. Biology, 70, 1587-1605. Weinig C., Schmit J. (2004) Environmental effects on the Haedrich R.L. (1983) Estuarine fishes. In: Ketchum B.H. (Ed), expression of quantitative trait loci and implications for Estuaries and Enclosed Seas. Elsevier Publishing Company, phenotypic evolution. BioScience, 54, 627-635. Amsterdam: pp. 183-207. Wouters N., Cabral H.N. (2009) Are flatfish nursery grounds Kraus R.T., Secor D.H. (2005) Application of the nursery-role richer in benthic prey? Estuarine Coastal and Shelf Science, hypothesis to an estuarine fish. Marine Ecology Progress Ser 83, 613-620. ies, 291, 301-305. Marine Ecology32 (Suppl. 1 ) (2011 ) 96-101 © 2011 Blackwell Verlag GmbH 101 anmarine evolutionary perspective ecology » Marine Ecology. ISSN 0173-9565 ORIGINAL ARTICLE mtDNA differentiation in the mussel M ytilus galloprovincialis Lmk. on the Iberian Peninsula coast: first results Javier R. Luis, Angel S. Comesaña & Andrés Sanjuan Xenética Evolutiva Molecular, Facultade de Bloloxia, Unlversldade de Vigo, Vigo, Spain Keywords A bstract Atlantic Ocean; genetic differentiation; Mediterranean Sea; mitochondrial DNA; The nucleotide sequence of the VD1 domain of the Long Unassigned Region Mytilus galloprovincialis. in the F mitochondrial DNA genome was studied for 135 Iberian mussels, 78 from the Atlantic and 57 from the Mediterranean. Significant genetic differenti Correspondence ation between Atlantic and Mediterranean Mytilus galloprovincialis samples was Javier Rodríguez Luis, Xenética Evolutiva found (FSt = 0.262, P < 0 .0 0 0 0 1 ). The four main clusters observed in the Molecular, Facultade de Bloloxia, Edit, de neighbor-joining tree of haplotypes were not exclusive for a specific region, but Ciencias, Unlversldade de Vigo, E-36310Vlgo, Spain. a clear geographic pattern could be observed. One of the clusters contained E-mail: [email protected] 8 6 % of the Atlantic individuals and the remaining three clusters were predomi nantly Mediterranean. The Atlantic-Mediterranean differentiation of the mito Accepted: 2 February 2011 chondrial DNA haplotypes was in agreement with previous data describing the same partitioning in M. galloprovincialis and in many other marine organisms, doi : 10.1111/j.1439-0485.2011,00430.x using different kinds of genetic markers. In all cases the Almeria Oran Oceano graphic front has been associated to this genetic discontinuity. Currently, analysis of mitochondrial DNA (mtDNA) Introduction sequences is one of the most widely used tools in molecu In Europe, the mussel Mytilus galloprovincialis Lmk. is lar phylogenetics studies due to uniparental inheritance, distributed in the Black Sea, the Mediterranean and on abundance of selectively neutral mutations, low rate of the Atlantic coast from the Iberian Peninsula as far north recombination and technical simplicity. Skibinski et a í as the British Isles (Gosling 1992). The Iberian Peninsula (1994) and Zouros et a í (1994) described the existence of occupies an intermediate location in this distribution, doubly uniparental inheritance (DUI) of the mtDNA in between the Atlantic and Mediterranean regions. the Mytilidae family. DUI involves the existence of het- Extensive genetic studies of M. galloprovincialis popula eroplasmic males carrying a maternal (F) and a paternal tions in the Iberian Peninsula have been carried out using (M) mitochondrial genome, and homoplasmic females allozyme polymorphisms (Sanjuan et a í 1994, 1997; bearing only the F genome. Thus, the F genome is mater Q uesada et al. 1995a), mtDNA RFLPs (Quesada et a í nally transmitted to offspring, whereas the M genome is 1995b; Sanjuan et a í 1996) and microsatellites (Diz & paternally transmitted to male progeny only. In species Presa 2008). In all cases, results agree on the existence of with DUI, maternal lineages provide more reliable infor a strong genetic discontinuity between Atlantic and Medi mation for population and phylogenetic studies, as M lin terranean populations associated to the Almeria Oran eages are not appropriate for phylogeographic studies due Oceanographic front (AOOF). However, the direct analy to their very fast evolution and liability to invasion from sis of nucleotide variation would provide a deeper insight the F lineage (Ladoukakis et a í 2002). The mitochondrial into the population dynamics of the species, on both genom e of M ytilus is divided into two parts: the core that sides of the Gibraltar Strait, than is offered by allozymes, contains all protein, rRNA and tRNA coding genes and a restriction enzyme analyses or microsatellites. few noncoding regions of <500 bp, and the LUR (large 102 Marine Ecology32 (Suppl. 1 ) (2011 ) 102-106 © 2011 Blackwell Verlag GmbH Luis, Comesaña & Sanjuan mtDNA differentiation in Iberian M. galloprovincialis Lmk. unassigned region). Cao et al. (2004) have divided the primer (UNFOR1 and UNREV1; Cao et al. 2004); 1.5 U LUR into three parts: the first variable domain (VD1), of Taq DNA polymerase (EcoTaq, Ecogen) and 3 pL of the conserved domain (CD), and the second variable the extracted DNA. An initial dénaturation at 94 °C for domain (VD2). These domains seem to be equivalent to 3 m in, 40 PCR cycles (94 °C: 1 m in, 55 °C: 1.5 m in, those found in human mtDNA and contain sequence 72 °C: 1 min) and a final extension at 72 °C for 6 m in motifs which are known to be involved in the replication were carried out in a GeneAmp PCR System 9700 ther and transcription of the molecule. These observations mal cycler (Applied Biosystems). suggest that the LUR is the main control region of the The PCR products were visualized by UV transillumi mitochondrial genome (Cao et a l 2004). The high degree nation after electrophoresis in 1.5% SeaKem LE agarose of nucleotide variability makes VD1 domain an excellent gels (FMC BioProducts) containing ethidium bromide. tool in order to analyse genetic variation in the mussel Double-stranded DNA from PCR reactions was cleaned M. galloprovincialis. using the Qiaquick PCR purification kit (Qiagen GmbH), Thus, the aim of this work was to assess the degree of following the manufacturer’s instructions. variation and differentiation, and to infer the population From purified PCR products, both strands of a frag dynamics processes leading to the current patterns of ment of 575 bp, belonging to the VD1 domain of the genetic variation in the M. galloprovincialis populations female mtDNA LUR, was sequenced in two separated living on the Iberian coast using the nucleotide sequence sequencing reactions using the CEQ Dye Terminator of the VD1 domain of the large unassigned region (LUR) Cycle Sequencing Kit (Beckman Instruments) and two of the F mtDNA genomes. different primers specific for the female mtDNA molecule. Finally, fragments were migrated in a Beckman CEQ2000 DNA sequencer (Beckman Instruments) at the Facultade Material and Methods de Bioloxia, Universidade de Vigo. Sequences were read A total o f 135 individuals o f M. galloprovincialis were col and edited using the software provided with the sequen lected in two Atlantic and two Mediterranean locations cer and aligned with CLUSTAL X version 2.0.9 (Larkin on the Iberian Peninsula coast (Fig. 1). After dissection, et al. 2007). total DNA was extracted from gonadal tissue using the Nucleotide diversity (Tajima 1983; Nei 1987) for each procedure of DeSalle et al. (1993). sample was estimated in the MEGA software v4.0.2. Amplification of the large unassigned region (LUR) of Gene diversity estimated from haplotype frequencies the mitochondrial genome was carried out in a final vol (Nei 1987) as well as population pairwise FSt distances um e o f 50 pL containing 10 mM Tris-HCl, pH 9.0, 2 mM (Reynolds et al. 1983; Slatkin 1995) between all pairs of M gCl2, 50 mM KC1, 0 . 1 % Triton X- 1 0 0 ; 0 . 2 mM of each populations, were computed using the ARLEQUIN soft dNTP (Amersham Pharmacia Biotech); 0.06 pM o f each ware package v3.1 (Excoffier et al. 2005). In the case of pairwise FST distances we have applied the Bonferroni method, a multiple-comparison correction used when T45 several statistical tests are being performed simulta GSU neously. To avoid spurious positives, the alpha value was lowered to account for the number of comparisons GBA GCQ being performed. Haplotype phylogenetic relationships were estimated with a neighbor-joining (NJ) method (Saitou 8 c Nei GCA ^ --40' 1987) conducted by the MEGA software v4.0.2 (Tamura et al. 2007), using the gamma-corrected Tamura-Nei dis tance (Tamura 8 c Nei 1993). Support for the tree was obtained by 1000 bootstrap replicates (Felsenstein 1985). 150 km Results J- 35 A total of 135 individuals were sequenced for 575 bp of - 10° -5° 0 ° the VD1 domain of the female mtDNA. Seventy-six vari able sites were observed, defining 40 different haplotypes Fig. 1. Sampling locations of M. galloprovincialis on the Iberian Pen insula coast: Suances (GSU), Baiona (GBA), Cambrils (GCA) and Cad- whose frequencies in each of the four samples are detailed aque (GCQ). Also shown is the Almería Oran Oceanographic front in Table 1 . The sequence of haplotype h i 2 was deposited (AOOF). in GenBank (accession no. HQ675011). Four of these Marine Ecology32 (Suppl. 1) (2011) 102-106 © 2011 Blackwell Verlag GmbH 103 mtDNA differentiation in Iberian M. galloprovincialis Lmk. Luis, Comesaña & Sanjuan Table 1. Relative frequencies of the 40 haplotypes found in this respectively), whereas haplotypes h08, h09 and hlO were study. Numbers in parentheses are sample sizes. Sample codes are the most abundant in the Mediterranean, where they indicated in Fig. 1. reach frequencies of 14—19%. It is noteworthy that Atlantic Mediterranean the percentage of singleton haplotypes, calculated over the total number of individuals, was much higher in the haplotypes GSLK33) GBA(45) total(78) GCA(32) GCQ(25) total(57) Atlantic (30%) than in the Mediterranean (5%). When h01 0.36 0.31 0.34 0.09 0 0.05 calculated over the number of haplotypes, the value was h02 0.06 0.02 0.04 0 0 0 also distinctly higher in the Atlantic ( 6 6 %) than in the h03 0 0 0 0.06 0.12 0.09 Mediterranean (21%). All the populations showed similar h04 0.03 0.02 0.03 0 0.04 0.02 degrees of diversity at the haplotype level. Haplotype h05 0.09 0.18 0.13 0 0.04 0.02 diversity (h) averaged 0.87 in the Atlantic samples and h06 0 0.04 0.02 0 0 0 0.9 in the Mediterranean samples. Conversely, the average h07 0.03 0.02 0.03 0 0 0 h08 0 0.02 0.01 0.25 0.08 0.17 of nucleotide diversity (n) in the Mediterranean (0.025) h09 0 0.02 0.01 0.19 0.20 0.19 was almost double the value in the Atlantic (0.013), h 10 0.03 0.02 0.03 0.12 0.16 0.14 although an unpaired t-test resulted in a probability of hi 1 0 0.04 0.02 0.03 0.12 0.08 0.0815. More extensive sampling will be necessary to clar h 12 0 0.04 0.02 0.06 0.04 0.05 ify this apparent difference in nucleotide diversity. h 13 0 0.02 0.01 0.06 0.04 0.05 The population pairwise FST distances (data not h 14 0 0 0 0.12 0.04 0.08 shown) yielded highly significant values for the Atlantic- GSU02 0.03 0 0.01 0 0 0 GSU03 0.03 0 0.01 0 0 0 Mediterranean pairs (P < 0.00001) and nonsignificant GSU04 0.03 0 0.01 0 0 0 values for the Mediterranean pairs (P = 0.748). When the GSU05 0.03 0 0.01 0 0 0 Atlantic populations were compared, the FST distance was GSU07 0.03 0 0.01 0 0 0 found significant at the 5% level, although it became GSU14 0.03 0 0.01 0 0 0 nonsignificant after application of Bonferroni correction GSU17 0.03 0 0.01 0 0 0 for the six pairwise comparisons (alpha level was lowered GSU20 0.03 0 0.01 0 0 0 from 0.05 to 0.0083). The FSt value between the Atlantic GSU21 0.03 0 0.01 0 0 0 GSU25 0.03 0 0.01 0 0 0 and the Mediterranean pooled samples (0.262) was highly GSU28 0.03 0 0.01 0 0 0 significant (P < 0.00001). GSU29 0.03 0 0.01 0 0 0 The NJ tree (Fig. 2) showed four main clusters sup GSU34 0.03 0 0.01 0 0 0 ported for bootstrap values higher than 70%. The largest GBA03 0 0.02 0.01 0 0 0 cluster, supported by a bootstrap value of 97%, included GBA04 0 0.02 0.01 0 0 0 8 6 % of the Atlantic individuals, and 29 different haplo GBA17 0 0.02 0.01 0 0 0 types (24 exclusively Atlantic). The remaining three clus GBA18 0 0.02 0.01 0 0 0 GBA22 0 0.02 0.01 0 0 0 ters were predominantly Mediterranean, encompassing GBA27 0 0.02 0.01 0 0 0 81% of the individuals of this geographic region. GBA29 0 0.02 0.01 0 0 0 GBA37 0 0.02 0.01 0 0 0 GBA40 0 0.02 0.01 0 0 0 D iscussion GBA46 0 0.02 0.01 0 0 0 The results of the analyses applied in this study were GCQ104 0 0 0 0 0.04 0.02 consistent with the existence of a significant genetic dis GCQ122 0 0 0 0 0.04 0.02 GCQ126 0 0 0 0 0.04 0.02 continuity between the Atlantic and Mediterranean popu lations of Mytilus galloprovincialis on the Iberian coasts. On the one hand, the population pairwise F St distances haplotypes (hll, hl2, GBA04 and GCQ126) showed a showed high differentiation between the Atlantic and the previously described duplication (Cao et al. 2004; Smiet- Mediterranean populations, but nonsignificant differences anka et al. 2004) of 36 bp, spanning from nucleotides 262 within each group. On the other hand, the NJ tree dis to 297 of the sequence of haplotype hl2. Haplotypes played four highly differentiated clusters representing four appearing more than once are named with numbers groups of haplotypes separated, one from each other, by (h01-hl4) and the ‘singletons’ are named by the code of a minimum of 10 mutational steps. Although no cluster the individual carrying the haplotype. is exclusive for a specific region, a clear geographic pat There was only one haplotype (hlO) present in all the tern differentiates the largest cluster, which includes 8 6 % samples. Haplotypes hoi and h05 were the commonest of the Atlantic individuals, and the remaining three clus among the Atlantic samples (frequencies of 34 and 14%, ters, predominantly Mediterranean. 104 Marine Ecology32 (Suppl. 1 ) (2011 ) 102-106 © 2011 Blackwell Verlag GmbH Luis, Comesaña & Sanjuan mtDNA differentiation in Iberian M. galloprovincialis Lmk. h02 (3A) gletons in the Atlantic cluster, typical of populations that — GSU07 have gone through episodes of recent expansion. The low h01 (26A/3M) — GCQ104 number of haplotypes in the Mediterranean clusters could 1-GSU25 be explained by recent bottleneck effects that reduced ÍjG B A 03 T— GBA40 genetic diversity. .GSU04 These results were in agreement with previous findings I—GSU20 I— GSU29 describing an Atlantic-Mediterranean partitioning in — h03 (5M) M. galloprovincialis (Sanjuan et al. 1994, 1996, 1997; GSU03 GBA22 Q uesada et al. 1995a,b, 1998; Ladoukakis et al. 2002; GSU17 Smietanka et al. 2004; Diz & Presa 2008) using different — GSU21 kinds of genetic markers (allozyme polymorphisms, GBA46 46 — GSU34 mtDNA restriction fragment length polymorphisms, ------GBA29 mtDNA sequences and nuclear microsatellites). In all 97 ■ h04 (2A/1M) ^ GSU02 cases this genetic discontinuity has been associated with — GBA27 the Almeria Oran Oceanographic front (Fig. 1; Tintore h05 (1JA/1M) 75 — GSU14 et a í 1988), a strong marine surface current between Al 71 h06 (2A) meria (Spain) and Oran (Algeria). The strong current and I— GBA18 — GSU05 a sharp change in oceanographic conditions (higher salin — GBA17 ity and temperature in the eastern part of the front) h07 (2A) GBA37 could represent an important obstacle to the dispersal of ------h08 (1A/10M) mussel larvae. |h09 (1 A/11M) A process of vicariance due to isolation of both popu 9 6 T------111h 10 ^ (2A/8M) . h 11 (2A/4M) lations is consistent with the clear genetic differentiation lGCQ126 100 observed between the Atlantic and Mediterranean. The ■h 12 (2A/3M) GBA04" narrowing of the Gibraltar Strait in the Pleistocene (Pie- GCQ 122 61 -h13 (1A/3M) lou 1979) could account for the isolation of the Mediter 99 - GSU28 ranean basin. Subsequently, after the opening of the — h 14 (5M) Gibraltar Strait, the AOOF would have acted as a partial barrier for genetic flow, maintaining some degree of the Fig. 2. Evolutionary relationships of the 40 haplotypes Inferred using differentiation previously established. the neighbor-joining method. Evolutionary distances were computed using the Tamura-Nei method and the rate variation among nucleo In spite of this barrier to gene flow there are significant tide sites was modeled with a gamma distribution (shape parame proportions of typically Atlantic haplotypes present in the ter = 0.6673). Bootstrap percentages from 1000 replications are Mediterranean and vice versa. Specifically, we have indicated on branches. Numbers In brackets correspond to haplotype observed that 14% of the Atlantic individuals carry Medi absolute frequencies in the Atlantic (A) and the Mediterranean (M). terranean haplotypes and 19% of the Mediterranean mus sels carry Atlantic haplotypes. This asymmetric gene flow explains the mixing of Atlantic and Mediterranean indi We have defined as Atlantic haplotypes those belonging viduals in all the main clusters of the tree. It is congruent to the first cluster as their frequency in the Atlantic is with the predominantly eastward direction of the surface much higher than in the Mediterranean, and the muta water circulation through the Straits of Gibraltar (Perkins tional connection with the rest of the haplotypes requires et al. 1990) and has been described previously in M . gal at least 10 nucleotide changes. Using the same reasoning, loprovincialis using mtDNA RFLPs (Quesada et a í 1998). the remaining haplotypes were defined as Mediterranean. However, this difference is not statistically significant and At the nucleotide diversity level, the Atlantic haplotypes more samples should be analyzed to provide more sup showed a higher degree of homogeneity, as all of them port for the asymmetry of the gene flow. belonged to the same cluster, and the largest difference between two haplotypes was three mutational steps. In Summary contrast, Mediterranean haplotypes are much more diverse, belonging to three clearly differentiated clusters. All the results in this study are consistent with a process This scenario explains the high n value in the Mediterra of isolation, followed by restricted secondary contact, gen nean. The apparent contradiction between the difference erating the current patterns of genetic differentiation in the n value and the similarity of haplotype diversities between the Atlantic and Mediterranean populations of (0.87 and 0.90) is explained by the high frequency of sin Mytilus galloprovincialis in the Iberian Peninsula. The Marine Ecology32 (Suppl. 1) (2011) 102-106 © 2011 Blackwell Verlag GmbH 105 mtDNA differentiation in Iberian M. galloprovincialis Lmk. Luis, C om esaña & Sanjuan temporary closing of the Mediterranean Sea, followed by the interaction of ecological and life-history factors. Marine the opening of the Strait of Gibraltar and subsequent Ecology Progress Series, 116, 99-115. gene flow restricted by the AOOF, would be a plausible Quesada H., Beynon C.M., Skibinski D.O.F. (1995b) A mito geological scenario explaining this situation. Nevertheless, chondrial DNA discontinuity in the mussel Mytilus gallopro a more exhaustive sampling would be necessary to pro vincialis Lmk: pleistocene vicariance biogeography and vide better evidence of the recent dynamics of Atlantic secondary intergradation. Molecular Biology and Evolution, and Mediterranean populations, and the interactions 12, 521-524. between them. Quesada H., Gallagher C., Skibinski D.A.G., Skibinski D.O.F. ( 1998) Patterns of polymorphism and gene flow of gender- associated mitochondrial DNA lineages in European mussel populations. Molecular Ecology, 1, 1041-1051. References Reynolds J., Weir B.S., Cockerham C.C. (1983) Estimation for Cao L., Kenchington E., Zouros E., Rodakis G.C. (2004) Evi the coancestry coefficient: basis for a short-term genetic dence that the large noncoding sequence is the main control distance. Genetics, 105, 767-779. region of maternally and paternally transmitted mitochon Saitou N., Nei M. (1987) The neighbor-joining method: a new drial genomes of the marine mussel ( Mytilus spp.). Genetics, method for reconstructing phylogenetic trees. Molecular 167, 835-850. Biology and Evolution, 4, 406-425. DeSalle R., Williams A.K., George M. (1993) Isolation and Sanjuan A., Zapata C., Alvarez G. (1994) Mytilus galloprovin characterization of animal mitochondrial DNA. Methods in cialis and M. edidis on the coasts of the Iberian Peninsula. Enzymology, 224, 176-234. Marine Ecology Progress Series, 113, 131-146. Diz A.P., Presa P. (2008) Regional patterns of microsatellite Sanjuan A., Comesaña A.S., De Carlos A. (1996) Macrogeo variation in Mytilus galloprovincialis from the Iberian graphic differentiation by mtDNA restriction site analysis in Peninsula. Marine Biology, 154, 277-286. the S.W. European Mytilus galloprovincialis Lmk. Journal of Excoffier L., Laval G., Schneider S. (2005) Arlequin ver. Experimental Marine Biology and Ecology, 198, 89-100. 3.0: an integrated software package for population genet Sanjuan A., Zapata C., Alvarez G. (1997) Genetic differentia ics data analysis. Evolutionary Bioinformatics Online, 1, tion in Mytilus galloprovincialis Lmk. throughout the world. 47-50. Ophelia, 47, 13-31. Felsenstein J. (1985) Confidence limits on phylogenies: an Skibinski D.O., Gallagher C., Beynon C.M. (1994) Sex-limited approach using the bootstrap. Evolution, 39, 783-791. mitochondrial DNA transmission in the marine mussel Gosling E.M. ( 1992) Systematics and geographic distribution Mytilus edulis. Genetics, 138, 801-809. of Mytilus. In: Gosling E. (Eds), The Mussel Mytilus: Ecology, Slatkin M. ( 1995) A measure of population subdivision based Physiology, Generics and Culture. Elsevier, Amsterdam: on microsatellite allele frequencies. Genetics, 139, 457-462. 1- 20 . Smietanka B., Zbawicka M., Wolowicz M., Wenne R. (2004) Ladoukakis E., Saavedra C., Magoulas A., Zouros E. (2002) Mitochondrial DNA lineages in the European populations Mitochondrial DNA variation in a species with two mito of mussels Mytilus. Marine Biology, 146, 79-92. chondrial genomes: the case of Mytilus galloprovincialis from Tajima F. (1983) Evolutionary relationship of DNA sequences the Atlantic, the Mediterranean and the Black Sea. Molecular in finite populations. Genetics, 105, 437-460. Ecology, 11, 755-769. Tamura K., Nei M. (1993) Estimation of the number of nucle Larkin M.A., Blackshields G., Brown N.P., Chenna R., otide substitutions in the control region of mitochondrial McGettigan P.A., McWilliam H., Valentin F., Wallace I.M., DNA in humans and chimpanzees. Molecular Biology and Wilm A., Lopez R., Thompson J.D., Gibson T.J., Higgins Evolution, 10, 512-526. D.G. (2007) Clustal W and Clustal X version 2.0. Bioinfor Tamura K., Dudley J., Nei M., Kumar S. (2007)MEGA4: matics, 23, 2947-2948. Molecular Evolutionary Genetics Analysis (MEGA) software Nei M. (1987) Molecular Evolutionary Genetics. Columbia version 4.0. Molecular Biology and Evolution, 24, 1596-1599. University Press, New York: 512 pp. Tintore J., La Violette P.E., Blade I., Cruzado A. (1988) A Perkins H., Kinder T.H., La Violette P.E. (1990) The Atlantic study of an intense density front in the eastern Alboran Sea: inflow in the western Alboran Sea. Journal of Physical the Almeria-Oran front. Journal of Physical Oceanography, Oceanography, 20, 242-263. 18, 1384-1397. Pielou E.C. (1979) Biogeography. John Wiley 8 c Sons, Inc., Zouros E., Ball A.O., Saavedra C., Freeman K.R. (1994) An London: 351 pp. unusual type of mitochondrial DNA inheritance in the blue Quesada H., Zapata C., Alvarez G. ( 1995a) A multilocus allo- mussel Mytilus. Proceedings of the National Academy of zyme discontinuity in the mussel Mytilus galloprovincialis: Sciences of the United States of America, 91, 7463-7467. 106 Marine Ecology32 (Suppl. 1 ) (2011 ) 102-106 © 2011 Blackwell Verlag GmbH anmarine evolutionary perspective ecology O / Marine Ecology. ISSN 0173-9565 ORIGINAL ARTICLE Paramuricea clavata (Anthozoa, Octocorallia) loss in the Marine Protected Area of Tavolara (Sardinia, Italy) due to a mass mortality event Carla Huete-Stauffer1, Maria Vielmini2, Marco Palma1, Augusto Navone3, Pier Panzalis3, Luigi Vezzulli4, Cristina Misic1 & Carlo Cerrano1 1 Department for the Study of the Territory and its Resources (Dip.Te.Ris), University of Genoa, Genoa, Italy 2 Department of Biology, University of Pisa, Pisa, Italy 3 MPA and SPAMI Tavolara Punta Coda Cavallo, Olbla, Italy 4 Department of Biology (Di.Bio), University of Genoa, Genoa, Italy Keywords Abstract Conservation; global warming; Octocorals; Vibrio. Recent studies highlight an increase in the frequency and intensity of marine mass mortalities of several species over the past 30-40 years, mainly in tropical Correspondence and temperate areas. In the Mediterranean Sea these episodes particularly affect Carla Huete-Stauffer, Department for the benthic suspension feeders, such as sponges and cnidarians. The main objective study of the Territory and its Resources, Corso of this work was to document the loss of one of the main Mediterranean sea Europa 26, 16132, Genoa, Italy. scapes, Paramuricea clavata forests at the Marine Protected Area of Tavolara E-mail: [email protected] Punta Coda Cavallo, Sardinia (Italy), during the summer of 2008. Data regard Accepted: 15 February 2011 ing colony height, density, level of damage, and microbiological community were collected at two sites. Such parameters help us understand how mass doi: 10.1111/j. 1439-0485.2011,00429.x mortality mechanisms act on this ecosystem engineer. We identified a change in size class distribution following a mass mortality that leaves mainly small colonies with a decrease in habitat complexity. Several tests on water chemistry demonstrate that the mortality event was not caused by local contamination. Moreover, microbiological tests on potential pathogenic agents suggest that bacteria belonging to the genus Vibrio are present as an opportunistic and not an etiological cause of P. clavata mortality events. Possible restoration approaches are discussed. ity events coincide with thermal anomalies generally Introduction caused by unusual water warming during prolonged peri Marine communities appear to be facing one of the worst ods of water column stability (CIESM, 2008). periods in their recent history. The direct negative effects Affected species often show modifications in their of several human activities (e.g. oil spills, coastal habitat physiology (Previati et a í 2010), distributions, and some modification, overfishing) are now amplified by climate times phenology (Bavestrello et a í 2006), which can have change, which is compromising both the resistance and unpredictable consequences on species’ interactions the resilience of many marine organisms. During the last (Hughes 2000). Often the affected species are ecosystem decades, the Northwestern Mediterranean Sea has been engineers and their rarefaction and/or disappearance has hit by a series of mass mortality events, which impact profound consequences on the habitat architecture, such benthic suspensivore organisms, such as sponges, cnidari as reducing spatial complexity and decreasing biodiversity ans, bivalves, bryozoans and tunicates, and associated richness (Mafias et al. 2010). In the Mediterranean assemblages (Cerrano & Bavestrello 2009). These mortal Sea, several causes of these mortality events have been Marine Ecology32 (Suppl. 1 ) (2011 ) 107-116 © 2011 Blackwell Verlag GmbH 107 P. clavata loss in the MPA of Tavolara Huete-Stauffer, Vielmini, Palma, Navone, Panzalis, Vezzulli, Misic & Cerrano identified, typically associated with environmental factors Population structure and mortality dynamics: field surveys (Cerrano et a í 2000, 2005; Pérez et al. 2000; Calvisi et al. 2003; Linares et al. 2005, 2008b; Cigliano & Gambi 2007; To study population dynamics, a minimum of six quad Previati et al. 2010) but a number of pathogens have also rats (50 X 50 cm) were randomly sampled from 40 to been implicated (Martin et al. 2002; Gay et al. 2004; Bally 20 m depth every 5 m from the bottom to the top of & Garrabou 2007; Vezzulliet a í, 2010). Among the each shoal, following standard methodology set out by affected areas in the Mediterranean, perhaps the Ligurian similar studies (Cerrano et al. 2005; Coma et al. 2006; Sea can be considered the most severely affected (Garra Linares et al. 2008a) and in relation to the geomorphol b ou et al. 2009). However, in the late summer of 2008, ogy of the sites (Papa 1 from -35 to -20 m; Papa 2 from one of these particular thermal anomalies was registered -40 to -25 m). Within each quadrat the number of colo in the Marine Protected Area of Tavolara Punta Coda nies (converted to colony density), colony heights, colony Cavallo (Central Western Tyrrhenian Sea) and had a dra health (defined as the percentage of colony with damaged matic effect on sea-fan (Octocorallia: Gorgonacea) popu coenenchyme: 0% was considered healthy, <25, <50, <75, lations, particularly on two rocky shoals adjacent to <99 and 100%), and the number of fishing lines and/or Tavolara Island, where presence of large numbers of the nets wrapped around the colonies were recorded. Further gorgonian Paramuricea clavata characterizes one of the more, for each colony the epibiosis level was recorded most famous dive spots in the area. (according to methods provided by Bavestrello et al. Paramuricea clavata is known to be sensitive to high 1997) and the epibiontic organisms were identified. On temperatures (Cigliano & Gambi 2007; Coma et al. 2009; the basis of organisms that had settled on the scleraxis we Fava et al. 2010; Previati et a l 2010) and can show an defined different temporal phases: (i) denuded (when the immediate response to these events. Crucially, this spe scleraxis is visible, with tissue on scleraxis), (ii) new cies is considered an ecosystem engineer and facilitator (covered with filamentous green and/or red algae, and/or species (Bruno & Bertness 2001; Scinto et al. 2009) hydrozoans), (iii) medium (possessing a thick coat of within coralligenous assemblages. The aims of this work algae, and/or sponges) and (iv) old (mainly colonized by were: (i) to describe the mass mortality event that calcareous organisms such as bryozoans). To test the dif affected populations of P. clavata and (ii) to determine ferences in population composition, within sites and whether a range of environmental parameters are corre among times, ANOVA tests were performed after verify lated with P. clavata mass mortality. Possible pathways ing that data were normally distributed and there was for restoration and management of this species are also equality of variances. discussed. We assumed that the pre-mortality population struc ture was very close to that observed in October (i.e. d u r ing/just after mortality) for three reasons: (i) most Material and Methods colonies had only just died, with naked scleraxis and with necrotic coenenchyme portions still present on the colo Study area nies, (ii) no fragments or whole colonies were found Tavolara Island (40°54, 19' N; 9°42, 28' E) is formed (implying that colonies had not detached from the base) from limestone-dolomite rock. Together with the granitic and (iii) colonies were always measured - including the Molara, Molarotto and other minor islets, Tavolara forms denuded and epibionted parts of the colonies. The post a small archipelago with an epibenthic community that mortality structure of colonies was considered to be typi has been well described by Navone & Trainito (2008) and cal of that observed from December 2008. Navone et a í (1992). In the study area, the effects of the mass mortality event on the pre-coralligenous and coral Environmental features - laboratory analyses ligenous assemblages of two nearby sites (termed Papa 1 and Papa 2) were quantified. Papa 1 ranges in depth from To evaluate whether chemical features of seawater and/or 15 to 39 m, and Papa 2 from 24 to 43 m. The current bacterial infections could be involved in the mortality, direction flows in a NE-SW direction, from Papa 2 both seawater samples and portions of colonies were col towards Papa 1. At both sites there is a dense forest of lected. Paramuricea clavata and also Eunicella cavolinii (Calvisi et al. 2003; Bianchi et al. 2007). SCUBA diving surveys Seawater temperature and water chemistry were undertaken during October 2008, December 2008 and June 2009, which corresponded to the period when Temperature was measured during survey dives using two mortality was first noted for P. clavata and 3 and types of underwater computers: (i) UWATEC (± 0.5 °C), 9 months subsequently. with a system that records and memorizes automatically 108 Marine Ecology32 (Suppl. 1 ) (2011 ) 107-116 © 2011 Blackwell Verlag GmbH Huete-Stauffer, Vielmini, Palma, Navone, Panzalis, Vezzulli, Mlslc & Cerrano P. clavata loss in the MPA of Tavolara water temperature during the dives every 4 s, creating a bile salt sucrose (TCBS) agar to isolate the main morpho- profile, and (ii) underwater computers with a punctual types of bacteria belonging to the genus Vibrio. O ther temperature measurement (variation of + 1 °C) that pieces of tissue were frozen (at -20 °C) until total geno divers recorded manually every 5 m during the dive mic DNA could be extracted. After DNA extraction, bac ascent. Water temperature data were taken in both sites. terial samples were identified on the basis of their 16S Average values of water temperature were accomplished rRNA gene sequences (see Vezzulli et a í, 2010 for full with both manually recoded and automatic UWATEC details of methods). To identify isolates, PCR amplifica data, separated by depth ranges: SST (sea surface temper tion of a 798-bp region was performed using the univer ature) or 0, 5, 10, 15, 20, 25, 30, 35 and 40 m depth. sal primers BRI (5'-AGAGTTTGATCCTGGCT-3') and To determine dissolved oxygen concentrations (DOC) BR2 (5'-GGACTACCAGGGTATCTAAT-3'), amplifying during the mortality event, separate seawater samples positions 8-806 of the Escherichia coli numbering of the were carefully collected, avoiding air bubbles, and imme 16S rRNA gene that include hyper-variable regions. 16S diately fixed following Carpenter’s (1965) protocol. Inor rRNA gene sequence similarity was determined with SEQ- ganic nutrient concentrations were determined according MATCH (version 2) analysis of Ribosomal Database Pro to Hansen & Grasshoff (1983). Marine water was pre-fil ject (RDP-II, Release 9) of the Center for Microbial tered with cellulose acetate filters (0.45 jLim pore diame Ecology, Michigan State University ( http://rdp.cme . ter) and maintained at -20 °C until laboratory analysis. msu.edu/seqmatch). To assess the pathogenic potential of Nitrates, nitrites, ammonia and phosphates were analyzed isolated strains towards P. clavata colonies, infection with SYSTEA (nutrient analyzer) and silica concentrations experiments were performed in aquaria at different tem were quantified using a Jasco V-500 spectrophotometer. peratures and environmental conditions simulating those observed in the environment during the occurrence of mortality events (see Vezzulli et a í, 2010 for full details Microbiological analysis of experiments). Top colony pieces (n = 45 in total) of about 5 cm of healthy and damaged P. clavata colonies were taken (20 Results healthy and 25 damaged, all samples from different colo Population structure and mortality dynamics: field surveys nies). Samples were maintained in cold seawater (4 °C) until laboratory manipulation, then washed in sterile sea In total, 476 colonies were observed in 158 quadrats. water to remove other bacteria or fauna that were not Records of colony densities, height and size class distribu strictly related to P. clavata damage and incubated in an tion show how the two sites (shoals) have a different enriched APW (alkaline peptone water) culture broth. population structure, and this appears to have elicited a After 10 h, samples were plated onto thiosulfate citrate different response to the mortality event. October 2008 December 2008 June 2009 Papa 1 I II.n h n »9 is » n j* illn m n ao Papa 2 ¡ I »MH » » » Depth (m) Fig. 1. Density averages for Papa 1 and Papa 2 at different depths In time with ± SE. Marine Ecology32 (Suppl. 1 ) (2011 ) 107-116 © 2011 Blackwell Verlag GmbH 109 P. clavata loss in the MPA of Tavolara Huete-Stauffer, Vielmini, Palma, Navone, Panzalis, Vezzulli, Misic & Cerrano Papa 1. In this site the mean density (+ SE) was healthy colonies (compared with previous surveys) were 9.12 + 2.18 colonies per m2. Significant differences in found and the average size of the population had shifted densities were detected among depths (P < 0.05, Fig. 1), towards smaller sized classes (i.e. 0-30 cm) (see Figs 1, 2 predominantly due to the low number of colonies in the and 4). upper limits of this species’ distribution. The mean height Papa 2. In this site m ean density (+ SE) o f P. clavata of Paramuricea clavata (+ SE) was 21.03 + 4.24 cm dur was 17.33 + 2.13 colonies per m2. No differences in den ing the mortality event and 27.72 + 3.12 cm during the sity were detected both among depths and the investi last survey, with no significant differences detected both gated periods (P > 0.05) (Fig. 1). Mean colony height in mean height of the colonies on the investigated tempo (+ SE) during the mortality event was 29.72 + 6 . 8 6 cm ral scale (P > 0.05, see also Fig. 2) and between the differ and 28.54 + 2.01 cm during the last survey. The smallest ent depths (P > 0.05). Damaged colonies of P. clavata colonies were found at 25 m depth (Fig. 2), while colony were found at all depths, but at 35 m depth the percent height elsewhere was significantly greater (P < 0.001), age of damage was mainly constant and generally lower although there was no obvious correlation between col than at the other (mostly shallower) depths (Fig. 3). ony height and the depth of the substrate they occupied. Between 30 and 20 m depth, mortality occurred at a high Damaged colonies were found at all depths during the percentage of colonies, and was especially prevalent in the entire period of study. At 40 m depth the percentage of larger sized classes. At 25 m depth, all size classes of colo damaged colonies was generally low and varied between nies were 100% damaged, both in October 2008 and in about 20 and 40%. By contrast, at 35 and 25 m depths December 2008. In the last survey, in June 2009, more the frequency of mortality was higher between October October 2008 December 2008 June 2009 40 - j 40 ■ Papa 1 ui 14 - M ii I i i lii 35 30 25 20 CJ, '53 43 Papa 2 20 IO - UL h l l i Depth (m) Fig. 2. Height averages for Papa 1 and Papa 2 at different depths In time with ± SE. '120 r 120 c 100 : loo 'Sm I 60 ô 40 S. 0 October 2008 December 2008 June 2009 October 2008 December 2008 June 2009 Date Date 1 85 m depth ■ 3 0 m depth ■ 25 m depth ■ 20 m depth 140 m depth ■ 3 5 m d e p th ■ 30 m depth ■ 25 m depth Fig. 3. Percentage of damaged colonies In time at all depths for the two study areas: Papa 1 (left) and Papa 2 (right). At deeper depths there are fewer damaged colonies, whereas at the superficial parts of the shoals, damage Is greater and has affected more colonies. There Is a general reduction In the last studied period, probably due to natural recovery and the loss of the larger and more damaged colonies. 110 Marine Ecology32 (Suppl. 1 ) (2011 ) 107-116 © 2011 Blackwell Verlag GmbH Huete-Stauffer, Vielmini, Palma, Navone, Panzalis, Vezzulli, Misic & Cerrano P. clavata loss in th e MPA o f Tavolara October 2008 December 2008 June 2009 35 m U i U ■ I■ I l i i . j •P 'P Q jS> $ p p p P & $ * ? » <& ^ ^ ❖ ^ ^ ^ ^ ^ N = 17 N t = 21 * 40 1 1 I 30 m « i » j l . i - - : 11111. I i l l .P -i> & S i iP iÿ «//W’ ®A V ^ A í A- ❖ * ^ V S* i-' o N = 13 c 25 m I i l u , f i ,& jS> «■ V V V3¡r< r1 ^ V V V «> V «■ y No P. clavata found in this 20 m depth range ...I, AAVAA Size class (cm) Fig. 4. Size class percentage frequency. Frequencies in damaged percentage per size class (black) and healthy percentage per size class (light grey). Papa 1. Frequencies are the percentage of damaged/healthy colonies found in a determinate size class. and December, especially for the 21-30 cm size class indi October) along the water column and the absence (in the viduals. In June, the size class distribution shifted towards investigated depth range of between 20 and 43 m) of the smaller size classes {i.e. there were more colonies in thermocline in October, confirming a prolonged water the range 0-30 cm) (Figs 1, 2 and 5). Overall, the mean stability over it (Fig. 6 ). Temperature is depth-dependent; percentage of damage of colonies indicates that damage is at all depths considered, water temperature was at least a generally greater in the shallower parts of the shoals, in degree higher during the months of September and Octo some cases up to 1 0 0 % throughout the whole size class of ber 2008 than in the previous year. Data for December individuals (Fig. 3). 2007 were not available. At both sites, epibiosis on the denuded parts of the col Although the dissolved oxygen concentrations were onies followed a pattern of four temporally successive always high in both the sampling sites (close to healthy steps: (i) denuded branches, here recorded in October, and damaged colonies), water results have highlighted (ii) newly settled organisms such as filamentous algae and higher levels of nitrite, nitrate and ammonia during the hydroids, (iii) a medium stage of epizootic colonization mortality event, especially around the damaged colonies with algae and small sponges, which was noted in Decem where tissue degradation was taking part (Table 2). In ber, and finally (iv) an old stage with colonization by contrast, phosphate concentrations were higher next to algae, sponges, bryozoans and other calcareous organisms, the healthy colonies. These concentrations led to P-limita- which was found during the last surveys in June tion conditions (higher N/P ratio where N is the sum of (Table 1). nitrite, nitrate and ammonia concentrations) next to the compromised colonies, especially during October. Microbiological tests revealed that Vibrio bacteria were Environmental features: laboratory analyses consistently more abundant in diseased organisms with Seawater temperature measurements confirmed the high up to a twofold higher concentration compared with temperature (22 + 1 °C in September and 21 +0.5 °C in those found on the healthy corals (Fig. 7). The 16S rRNA Marine Ecology32 (Suppl. 1 ) (2011 ) 107-116 © 2011 Blackwell Verlag GmbH 111 P. clavata loss in th e MPA of Tavolara Huete-Stauffer, Vielmini, Palma, Navone, Panzalis, Vezzulli, Misic & Cerrano October 2008 December 2008 June 2009 40 m : l 11. 11 A A A A A A ,*$> ,■$> * •> A A A V / / 1 30 304 35 m i J j L i l 4>.ii « A ll a A i * : 111. . I «f y -?■ •i- í> ’ Cu » 30 m J j ' .lili h i l l i i $ fi ,A A 25 m 1 1 1 A ,A A & & $ "P * A A y Size class (cm) Fig. 5. Size class percentage frequency. Frequencies of damaged percentage per size class (black) and healthy percentage per size class (light grey). Papa 2. Frequencies are percentage of dam aged/healthy colonies found in a determinate size class. Table 1. Number of colonies affected and the type of epibiosis Discussion recorded on the colonies of the studied sites during and after the mass mortality event. The main aims of MPAs, as identified in the IUCN Guidelines for Establishing Marine Protected Areas (Kelle- epibiosis type October 2008 December 2008 June 2009 her & Kenchington 1992), are (i) to maintain essential Papa 1 ecological and life-support systems, (ii) to ensure the sus denuded 60 26 0 tainable utilization of species and ecosystems and (iii) to new 25 52 40 preserve biotic diversity. Monitoring, defined as continu medium 10 15 55 ous observation of conditions over time, is a crucial tool old 5 6 5 for the conservation of marine biological diversity and Papa 2 provides managers with important data from which they denuded 70 43 1 can make informed decisions about patterns and pro new 11 27 36 medium 11 22 53 cesses that affect biodiversity, and thus the functioning old 8 9 10 (or not) of an MPA. Here, we present our monitoring data to describe some of the factors associated with a mass mortality of the gorgonian Paramuricea clavata at gene sequencing of 61 Vibrio isolates associated to dis the MPA at Tavolara Island, Italy. eased and healthy Paramuricea clavata colonies showed a The mortality event described here affected a popula close homology of the majority of the strains with Vib tion of octocoral that is well known and utilized by the rio harveyi (n = 24), Vibrio splendidus (n = 22) and Vib diving tourism industry in a ‘Specially Protected Area of rio coralliilyticus (n = 15), the latter only being identified Mediterranean Importance’ (SPAMI). Our results repre in diseased organisms. sent an important baseline for future monitoring 112 Marine Ecology32 (Suppl. 1 ) (2011 ) 107-116 © 2011 Blackwell Verlag GmbH Huete-Stauffer, Vielmini, Palma, Navone, Panzalis, Vezzulli, Misic & Cerrano P. clavata loss in the MPA of Tavolara T e m p t Temp*C 15 20 25 30 15 5 — 5*2008 10 — Oct20Q8 l o — Dtc2008 15 is £ 30 g 20 I» 30 30- M 40 « 4 5 J Fig. 6. Temperature profile from 2008 of the months of Interest In the Investigated area, recorded by SCUBA operators and SST (sea surface temperature) from http://www.poseldon.ogs.lt . Data from December 2007 were not available. Table 2. Environmental features recorded for the sea-water collected next the decaying (damaged) and the control (healthy) colonies. Dissolved oxygen Silicate Nltrite+nltrate Ammonia Phosphate ml r 1 SD /(M SD /(M SD /(M SD /(M SD N/P ratio October Healthy 7.49 0.13 1.15 0.05 2.40 0.18 1.63 0.08 0.18 0.01 22.9 Damaged 7.23 0.14 1.54 0.21 3.11 0.41 2.10 0.42 0.15 0.01 35.4 December Healthy nd nd 2.42 0.28 0.92 0.10 0.96 0.19 0.12 0.01 15.2 Damaged 7.77 nd 2.12 0.36 0.96 0.06 1.23 0.37 0.10 0.02 22.0 N, nitrogen; nd, not detected; P, phosphate; SD, standard deviation. Vibrio spp. Broadly speaking, Papa 1 is less dense and has smaller colonies than Papa 2. Mean colony height did not differ over a relatively short period of 1-3 months after the 1ô*3 T mortality episode, but there was a clear shift towards ë < 3 small size classes by 9 months after the mortality event. m ie + 2 This trend was reported also in other monitoring studies ID u. on gorgonian mortality (Cerrano et a í 2005; Cupido Ü 1e*1 et a í 2008, 2009; Linares et al. 2008a,b,c). This phenome non is mainly due to the loss of the colonies from the lar ger size classes and also the fragmentation and/or damage Healthy Diseased to branches. Moreover, the presence of recruits caused a shift towards smaller colonies, leading to a general loss of Fig. 7. Concentration of Vibrio found In the colonies (healthy and habitat complexity. Furthermore, there was evidence that damaged) of Papa 1 and Papa 2. CFU = unity of bacteria colony for most of the colony branches with epibionts that were mation with ± SE. This graph considers all the Vibrio species found on counted in December 2008 had either fallen off or were the Paramuricea colonies: of these V. harveyi and V. splendidus were broken in June 2009. the main components on the healthy colonies and V. coralHlyticus was From 35 m (for Papa 2) and 30 m depth (for Papa 1), up also present on the damaged or diseased colonies. to the surface of the shoals, the larger colonies were more programs on this long-lived, sessile species with slow pop affected (i.e. had a higher percentage of damage) by the ulation dynamics (Mistri & Ceccherelli 1994). Even mortality event: deeper colonies were less affected in both though the mean colony density of the studied areas was sites. Colonies living at greater depths could hence consti lower than in other Mediterranean areas (Cerrano & tute a reservoir for the production of planulae for future Bavestrello 2008; Cupido et al. 2009; Garrabou et al. population recoveries (bottom-up and lateral supply). 2009), the seascape before the mortality episode was In coralligenous assemblages, perhaps the most impor dominated by large colonies. tant habitats in the Mediterranean Sea (Ballesteros 2006), Marine Ecology32 (Suppl. 1 ) (2011 ) 107-116 © 2011 Blackwell Verlag GmbH 113 P. clavata loss in the MPA of Tavolara Huete-Stauffer, Vielmini, Palma, Navone, Panzalis, Vezzulli, Misic & Cerrano gorgonians and particularly P. clavata, are considered key Bavestrello 2009). For this reason, monitoring is the only species, being important engineering and/or foundation effective approach to plan adequate programs of interven species ( sensu Dayton, 1972; Mistri & Ceccherelli 1994; tion. Hypotheses to limit damages and/or improve recov Cupido et a í 2009). Mass mortality events have wide ery damaged colonies include: (i) development of an consequences for gorgonian populations (Linares & Doak ‘early warning system’ to measure water stratification and 2 0 1 0 ) and for the community that depends on them, as predict mass mortalities and (ii) utilization of pruning the loss of these species alters sedimentation, turbidity and transplant technique protocols. For example, con and water movement, which negatively affect the complex trolled miniaturization of larger colonies could lead to structure of the hard-bottom benthic communities and more resistant and resilient specimens that will also fur the local biodiversity richness (Scinto et a í 2009). In both nish a number of fragments for transplants to highly September and October 2008, the position of the thermo- damaged areas to be used for restoration. Certainly, cline could not be detected down to 40 m depth during manipulation and transplant experiments need to be the mortality event and temperature was ~2 °C higher designed and tested before this strategy can be used rou than during September 2007. The verified thermal anom tinely (see for example the pilot study of Linares et al. alies and the constant warming of the Mediterranean may 2008c). Sites within MPAs have greater possibilities for have important consequences for the natural biocenosis recovery and adequate management, as the anthropogenic (Bianchi 2007; Coma et a í 2009) and may be the cause impacts are reduced and controlled; but as shown in this of the mass mortality events occurring in the last few and other studies, MPA protection certainly does not pre years (Pérez et al. 2000; Pérez 2008; Coma et al. 2009). vent mass mortalities related to environmental causes. In October, seawater analyses indicated altered values of ammonia, nitrite and nitrate close to damaged colo Conclusions nies, which were higher than usual and likely due to tis sue degradation. These anomalies were not recorded in Short-term effects of mortality events include habitat sim December, after the mortality episode had finished. plification and reducing the economic value due to a Changes in phosphate values could have been related to decrease of touristic appreciation (Trainito 2007; Navone the summer increase of urban sewage outflow due to high & Trainito 2008). Evaluation of long-term consequences tourist density. An alteration of the nutrient concentra needs adequate monitoring programs. On the basis of pre tions was highlighted by the change of the N/P ratio val vious mortality episodes described in different areas (Lig- ues (Table 2), which may potentially favor unusual uro-Provençal Basin and Tyrrhenian Sea), long-term phytoplanktonic and/or bacterial species. The P-limited effects may vary in relation to the presence of healthy situation could have added an energetic constraint on the ‘pocket reservoir’ populations living below the water sta weak P. clavata population. These general conditions bility area delimiting the thermal anomalies. Their pres (high seawater temperature, altered N/P ratio) could have ence may represent an important larval supply to facilitate facilitated the increase of bacteria on damaged colonies. the recovery of shallow populations. High water tempera We uncovered three main groups of Vibrio, o f which Vib ture and prolonged summer conditions are among the rio coralliilyticus has been implicated as an important most relevant causes of these mass mortality events cause of mortality for Mediterranean P. clavata (M artin (including the one reported here), reducing natural et al. 2002; Bally & Garrabou 2007) and Caribbean corals defences of colonies. In these compromised conditions, (Cervino et al. 2004). Vibrio bacteria are normally found colonies are more vulnerable to infections and energetic in seawater and are thermodependent: at high tempera constraints. tures (22-24 °C) Vibrio grows rapidly. Vibrio coralliilyti cus showed the highest virulence toward P. clavata Acknowledgements colonies and satisfied Koch postulates for pathogenicity. Authors are indebted to the Tavolara MPA staff for logis This bacterium appears to act as an opportunistic agent, tic support. This work was partially financed by 2008 infecting weak, thermally stressed colonies and compro Genoa University funds to C.C. mising colony recovery (Vezzulli et a í, 2010). Until now there have been no standardized actions to References mitigate the impact of mass mortalities of P. clavata. Intervention strategies remain to be validated and it is Ballesteros E. (2006) Mediterranean coralligenous assemblages: important, especially where diving activity is frequent, to a synthesis of present knowledge. Oceanography and Marine avoid population fragmentation that would lead to a pro Biology: An Annual Review, 44, 123-195. gressive reduction in population density (Linares & Doak Bally M., Garrabou J. (2007) Thermodependent bacterial pathogens and mass mortalities in temperate benthic 2010) and a general loss of biodiversity (Cerrano & 114 Marine Ecology32 (Suppl. 1 ) (2011 ) 107-116 © 2011 Blackwell Verlag GmbH Huete-Stauffer, Vielmini, Palma, Navone, Panzalis, Vezzulli, Mlslc & Cerrano P. clavata loss in the MPA of Tavolara communities: a new case of emerging disease linked to cli Islands (Tyrrhenian Sea). Biología Marina del Mediterráneo, mate change. Global Change Biology, 13, 2078-2088. 14, 292-293. Bavestrello G., Cerrano C., Zanzi D., Cattaneo-Vietti R. (1997) Coma R., Linares C., Ribes M., Diaz D., Garrabou J., Ballester Damage by fishing activities to the gorgonian coral Paramu os E. (2006) Consequences of a mass mortality in popula ricea clavata in the Ligurian Sea. Aquatic Conservation: tions of Eunicella singularis (Cnidaria: Octocorallia) in Marine and Freshwater Ecosystems, 7, 253-262. Menorca (NW Mediterranean). Marine Ecology Progress Bavestrello C., Puce S., Cerrano C., Zocchi E., Boero N. (2006) Series, 327, 51-60. The problem of seasonality of benthic hydroids in temperate Coma R., Ribes M., Serrano E., Jiménez E., Salat J., Pascuals J. waters. Chemistry and Ecology, 22, 197-205. (2009) Global warming - enhanced stratification and mass Bianchi C.N. (2007) Biodiversity issues for the forthcoming mortality events in the Mediterranean. Proceedings of the tropical Mediterranean Sea. Eiydrobiologia, 580, 7-21. National Academy of Sciences of the United States of America, Bianchi C.N., Cattaneo-Vietti R., Morri C., Navone A., Panzal 106, 6176-6181. is P., Orrù P. (2007) II coralligeno dell’Area Marina Protetta Cupido R., Cocito S., Sgorbini S., Bordone A., Santangelo G. di Tavolara Punta Coda Cavallo (Sardegna Nord.Orientale). (2008) Response of a gorgonian ( Paramuricea clavata) pop Biología Marina del Mediterráneo, 14, 148-149. ulation to mortality events: recovery or loss? Aquatic Conser Bruno J.F., Bertness M.D. (2001) Habitat modification and vation: Marine and Freshwater Ecosystems, 18, 984-992. facilitation in benthic marine communities. In: Bertness Cupido R., Cocito S., Barsanti M., Sgorbini S., Peirano A., M.D., Gaines S.D., Hay M.E. (Eds), Marine Community Santangelo G. (2009) Unexpected long-term population Ecology. Sinauer, Sunderland: 8 , 201-218. dynamics in a canopy-forming gorgonian coral following Calvisi G., Trainito E., Pais M., Franci S., Schiaparelli S. mass mortality. Marine Ecology Progress Series, 394, 195— (2003) Prima segnalazione di un episodio di mortalità di 200. gorgonacei lungo la costa dell’Isola di Tavolara (Sardegna Dayton P.K. (1972) Toward an understanding of community Settentrionale). Biología Marina Mediterranea, 10, 506-508. resilience and potential effects of enrichments to the benthos Carpenter J.H. (1965) The Chesapeake Bay Institute technique at McMurdo Sound, Antartica. In: Parker B.C. (ed), Pro for the Winkler dissolved oxygen method. Limnology and ceedings of the colloquium on conservation problems in Antar Oceanography, 10, 141-143. tica. Allen press, Lawrence, Kansas: 81-95. Cerrano C., Bavestrello G. (2008) Medium-term effects of Fava F., Bavestrello G., Valisano L., Cerrano C. (2010) Sur die-off of rocky benthos in the Ligurian Sea. What can we vival, growth and regeneration in explants of four temperate learn from gorgonians? Chemistry and Ecology, 24, 73-82. gorgonian species. Italian Journal of Zoology, 77, 44-52. Cerrano C., Bavestrello G. (2009) Massive mortalities and Garrabou J., Coma R., Bensoussan N., Bally M., Chelvaldonné extinctions. In: Wahl M. (Ed.), Marine Hard Bottom Commu P., Cigliano M., Diaz D., Harmelin J.G., Gambi M.C., Ker nities. Patterns, Dynamics, Diversity, and Change. Springer- sting D.K., Ledoux J.B., Lejeusne C., Linares C., Marschal Verlag, Berlin: Ch. 21. Ecological Studies. 206, 295-307. C., Perez T., Ribes M., Romano J.C., Serrano E., Torrents Cerrano C., Bavestrello G., Bianchi C.N., Cattaneo-Vietti R., O., Zabala M., Zuberer F., Cerrano C. (2009) A new large Bava S., Morganii C., Morri C., Picco P., Sara G., Schiapar scale mass mortality event in the NW Mediterranean rocky elli S., Siccardi A., Sponga F. (2000) A catastrophic mass- benthic communities: effects of the 2003 heat wave. Global mortality episode of gorgonians and other organisms in the Change Biology, 15, 1090-1103. Ligurian Sea (NW Mediterranean), summer 1999. Ecology Gay M., Renault T., Pons A.M., Le Roux F. (2004) Two Vibrio Letters, 3, 284-293. splendidus related strains collaborate to kill Crassostrea gigas: Cerrano C., Arillo A., Azzini F., Calcinai B., Castellano L., taxonomy and host alterations. Diseases of Aquatic Organ Muti C., Valisano L., Zega G., Bavestrello G. (2005) Gorgo isms, 62, 65-74. nian population recovery after a mass mortality event. Hansen P., Grasshoff K. (1983) Automated chemical analysis, Aquatic Conservation: Marine and Freshwater Ecosystems, 15, methods of seawater analysis (2nd edn). Verlag Chemie, 147-157. Weinheim: 347-379. Cervino J.M., Hayes R.L., Poison S.W., Poison S.C., Goreau Hughes L. (2000) Biological consequences of global warming: T.J., Martinez R.J., Smith G.W. (2004) Relationship of is the signal already apparent? Trends in Ecology and Evolu Vibrio species infection and elevated temperatures to yellow tion, 15, 56-63. Blotch/band disease in Caribbean corals. Applied and Kelleher G., Kenchington R. (1992) Guidelines for establishing Environmental Microbiology, 70, 6855-6864. marine protected areas. A Marine Conservation and Develop CIESM. (2008) Climate warming and related changes in Medi ment Report. IUCN - The World Conservation Union, terranean marine biota. In: Briand F. (Ed.), Marine Science Gland, Switzerland: vii + 79 pp. Commission (CIESM) Workshop Monographs, No.35 Linares C., Doak D.F. (2010) Forecasting the combined effects Monaco: 152. of disparate disturbances on the persistence of long-lived Cigliano M., Gambi M.C. (2007) The long hot summer: a fur gorgonians: a case study of Paramuricea clavata. Marine ther mortality event of gorgonians along the Phlaegrean Ecology Progress Series, 402, 59-68. Marine Ecology32 (Suppl. 1 ) (2011 ) 107-116 © 2011 Blackwell Verlag GmbH 115 P. clavata loss in the MPA of Tavolara Huete-Stauffer, Vielmini, Palma, Navone, Panzalis, Vezzulli, Misic & Cerrano Linares C., Coma R., Diaz D., Zabala M., Hereu B., Dantart L. Navone A., Bianchi C.N., Orrù P., Ulzega A. (1992) Saggio di (2005) Immediate and delayed effects of a mass mortality cartografía geomorfologica nei parco marino di Tavolara- event on gorgonian population dynamics and benthic com Capo Coda cavallo. Oehalia, 17(Suppl), 469-478. munity structure in the NW Mediterranean Sea. Marine Pérez T. (2008) UNEP-MAP-RAC/SPA. Impact of climate Ecology Progress Series, 305, 127-137. change on biodiversity in the Mediterranean Sea. The Regio Linares C., Coma R., Garrabou J., Diaz D., Zabala M. (2008a) nal Activity Centre for Specially Protected Areas. RAC/SPA, Size distribution, density and disturbance in two Mediterra Tunisi-Cedex, Tunisia: 1-90. nean gorgonians: Paramuricea clavata and Eunicella singular Pérez T., Garrabou J., Sartoretto S., Harmelin J.G., Francour P., is. Journal of Applied Ecology, 45, 688-699. Vacelet J. (2000) Mass mortality of marine invertebrates: an Linares C., Coma R., Zabala M. (2008b) Effects of a mass unprecedent event in the Northwestern Mediterranean. mortality event on gorgonian reproduction. Coral Reefs, 27, Compte Rendus de l’Académie des Sciences III, 323, 853-865. 27-34. Previati M., Scinto A., Cerrano C., Osinga R. (2010) Oxygen Linares C., Coma R., Zabala M. (2008c) Restoration of consumption in four Mediterranean octocorals during tem threatened red gorgonian populations: an experimental perature increase. Journal of Experimental Marine Biology and modelling approach. Biological Conservation, 141, and Ecology, 390, 39-48. 427-437. Scinto A., Bertolino M., Calcinai B., Huete-Stauffer C., Previati Martin Y., Bonnefont J.L., Chancerelle L. (2002) Gorgonians M., Romagnoli T., Cerrano C. (2009) Role of a Paramuricea mass mortality during the 1999 late summer in French Med clavata forest in modifying the coralligenous assemblages. iterranean coastal waters: the bacterial hypothesis. Water Proceedings of the 1st Mediterranean Symposium on the Research, 36, 779-782. conservation of the coralligenous and other calcareous bio- Matías M.G., Underwood A.J., Hochuli D.F., Coleman R.A. constructors. The Regional Activity Centre for Specially (2010) Independent effects of patch size and structural com Protected Areas (RAC/SPA). Tabarka, 136-140 pp. plexity on diversity of benthic macroinvertebrates. Ecology, Trainito E. (2007) Progetto mappaggio, valutazione di risorse 91, 1908-1915. sottomarine per lo sviluppo sostenibile del turismo subac- Mistri M., Ceccherelli V.U. (1994) Growth and secondary pro queo nell’AMP Tavolara Punta Coda Cavallo. MPA Tavolar- duction of the Mediterranean gorgonian Paramuricea clavat- a. 140 pp. a. Marine Ecology Progress Series, 103, 291-296. Vezzulli L., Previati M., Pruzzo C., Marchese A., Bourne D.G., Navone A., Trainito E. (2008) Tavolara: Nature at Work...Work Cerrano C.; Vibrio Consortium. (2010) Vibrio infections ing in Nature. Delfino Carlo Editore 8 c Co., Olbia, Rome: 285 triggering mass mortality events in a warming Mediterra p p . nean Sea. Environmental Microbiology, 12, 2007-2019. 116 Marine Ecology32 (Suppl. 1 ) (2011 ) 107-116 © 2011 Blackwell Verlag GmbH anmarine evolutionary perspective ecology ' » Marine Ecology. ISSN 0173-9565 ORIGINAL ARTICLE Short and mid-long term effects of cockle-dredging on non-target macrobenthic species: a before-after-control- impact experiment on a tidal mudflat in the Oosterschelde (The Netherlands) Sander Wijnhoven1, Vincent Escaravage1, Peter M. J. Herman2, Aad C. Smaal3 & Herman Hummel1 1 Monitor Taskforce, Netherlands Institute of Ecology, Centre for Estuarine and Marine Ecology (NIOO-CEME), Yerseke, The Netherlands 2 Department of Spatial Ecology, Netherlands Institute of Ecology, Centre for Estuarine and Marine Ecology (NIOO-CEME), Yerseke, The Netherlands 3 Aquaculture Department, Wageningen IMARES (Institute for Marine Resources & Ecosystem Studies) (WUR), Yerseke, The Netherlands Keywords Abstract Before-and-after-impact design; benthic macrofauna; Cerastoderma edule ; cockle To study the possible environmental impact of hydraulic cockle-dredging on fishery effects; short and mid-long term; macrobenthic communities and the environment, a fishing experiment was species composition. executed on a tidal mudflat in the Oosterschelde (SW Netherlands) according to a BACI (before-after-control-impact) design. Following the characterization of Correspondence the initial situation, a part of the mudflat was commercially fished, after which Sander Wijnhoven, Monitor Taskforce, dredged and undredged areas were compared on the basis of macrofauna Netherlands Institute of Ecology, Centre for Estuarine and Marine Ecology (NIOO-CEME), descriptors and sediment constitution approximately 2 months (short term) and Korringaweg 7, PO Box 140, NL-4401 NT, 1 year (mid-long term) after fishing. Whereas a clear reduction of the larger Yerseke, The Netherlands. Cerastoderma edule cockles (>23 mm) in the fished areas was found, no effect of E-mail: [email protected] dredging on total macrofauna densities or median grain size was observed. No negative effect of fishing on total macrofauna biomass was found; in contrast, an Accepted: 16 November 2010 increase of the biomass of the non-target species almost compensated for the loss in weight due to the extraction of the larger cockles. No significant effect of doi: 10.1111/j. 1439-0485.2010.00423.x dredging on species diversity, richness or evenness was found in the short or mid-long term, these descriptors tending to have increased rather than decreased in the dredged plots after 1 year. The selective fishing for larger cockles reduced the average cockle size, but 1 year after fishing the average size had returned to the initial values in the dredged area. However, compared to the control area, the average size might still be reduced, as the size of the cockles in the control area also increased during the year. Local environmental conditions, with their spe cific macrobenthic communities, seem to be crucial for the type of effects and the impact of dredging. It is therefore of eminent importance to follow a research design with pre-defined environmental conditions, rather than a comparison of different areas that are open or closed to fisheries. The present study based on a BACI approach indicates that mechanical cockle fisheries had no overall negative impact in our study area. ular. Some of these show strong effects (e.g. Beukema Introduction 1995; Piersma et a í 2001; Leopold et a í 2004), but other Several studies have investigated the potential impact of studies show minor effects or none at all (e.g. Cra- dredging or sediment-disturbing activities on macroben eymeersch & Hummel 2004; Ens et al. 2004; Beukema & thic communities and on the non-target species in partic Dekker 2005). These studies differ in the severity of the Marine Ecology32 (Suppl. 1) (2011) 117-129 © 2011 Blackwell Verlag GmbH 117 Short and mid-long term effects of cockle-dredging on non-target macrobenthic species Wijnhoven, Escaravage, Herman, Smaal & Hummel disturbances, especially the disturbance depth (Haii & (Hiddink 2003). Comparisons between observations at Harding 1997; Kaiser et al. 2001), the season of the dis short and mid-long term can show whether the effects turbance (Haii & Harding 1997), the frequency of distur on the species assemblages are transitory, whereas mid- bance (e.g. Kaiser et al. 2001), the possible selectivity of long term observations are required to detect effects on the different fishing techniques used (Ferns et al. 2000) species recruitment and larval settlement (Piersma et a í and the methodological approach involving comparison 2001). of different areas that were fished or unfished for an In this study we specifically investigated whether the extended period (Piersma et al. 2001) versus an experi sediment characteristics (grain size) and the macrobenthic mental approach with exclusion of fishery in cockle beds communities (including non-target species) are negatively (Craeymeersch & Hummel 2004). affected by hydraulic dredging for cockles on a soft-sedi- The many studies on this topic differ also with ment tidal flat. To account for the various sources of var respect to their research questions. They ascertain nega iation besides the direct effect of dredging, a BACI tive effects of fishing disturbances on (i) the environ (before-after-control-impact) design was used (Smith ment as a whole (e.g. Leopold et a í 2004; Zwarts 2004), 2002). A substantial part of a mudflat was commercially (ii) the structure of communities (Leitäo & Gaspar fished, and other parts were left undisturbed. In the 2007), (iii) the abundance of target species (Piersma dredged and undredged areas, 100 X 100 m plots were et a í 2 0 0 1 ), or (iv) processes as settlement, population delimited and randomly sampled before and shortly dynamics and recolonization of selected species (Cotter (short term) and 1 year (mid-long term) after fishing. et al. 1997; Hiddink 2003; Beukema & Dekker 2005). This study is the first to investigate the impact of com These studies mostly consist of inventories after large mercial cockle fisheries with suction dredgers on non-tar impacts, and the next step consists of their integration get benthic macrofauna species and communities using a in policies aiming at a mitigation of the risk that takes BACI approach in which the sensitivity of the experimen into account the opportunities for sustainable fisheries tal design to detect quantitative changes is also taken into (Beukema & Cadée 1999). Thus an impact is expected account. beforehand, where the study investigates the rehabilita tion potential or duration of rehabilitation of the target or non-target populations or the environmental condi Material and Methods tions (Haii & Harding 1997). Study area and experimental design Given that different environments have their specific communities and species assemblages, various impacts of The experimental research on the effects of cockle dredg fisheries might be expected depending on the substrate ing was carried out on the Slikken van de Dortsman tidal specifications (Ferns et a í 2000), tidal range and elevation flats in the Oosterschelde, a semi-open tidal bay in the or depth (Leitäo & Gaspar 2007). Furthermore, different Southwest Netherlands (Fig. 1) with a salinity of above evaluations of effects can be expected according to the 30%o (Coosen et al. 1994). Next to the blue mussel M yti sampling design, ranging from ad hoc inventories in lus edulis, the cockle Cerastoderma edule is the dominant extensive areas that are either fished or protected over suspension feeder in the Oosterschelde. However, nowa various time spans (e.g. Piersm a et a í 2001; Beukema & days, total cockle biomass is lower than it used to be dur Dekker 2005) toa priori elaborated experimental designs ing the 1980s and before. Besides the intensive fishing on on local sites with an exact knowledge of the fishing the cockle populations over several years, the construction intensity and timing (Ferns et a í 2000). of a storm surge barrier in the mouth of the Oostersc We are interested in whether dredging has a destruc helde (from 1976 to 1986) had a great impact on the tive effect on non-target species, which might be dam suitability of the area for cockles. The construction of the aged by being lifted up from the sediment or processed storm surge barrier led to a 30-70% reduction in current through the fishing device. These effects might be visible velocities, and a 1 2 % reduction in the tidal range, pro in the short term through increased mortality resulting ducing clearer waters, crumbling away of the elevated from injuries or from exacerbated predation by other areas, and sedimentation at the brims of the tidal flats macrobenthic species or vertebrates (Ferns et a í 2000; (Geurts van Kessel et al. 2003). The tidal range in the Hiddink 2003). This implies that the predators of the research area varies between 1.7 and 3.8 m. highly dredging-impacted species might profit from this Since the early 1990s, cockle fishery in the Ooster disturbance. Dredging can also induce shifts in the spe schelde has been subject to authorization, which is only cies composition as a result of the alteration of the granted in years with abundant cockle biomass. No environmental conditions such as the sediment compo authorization had been given since 2 0 0 1 in our research sition in the short and especially the mid-long term area (Geurts van Kessel et al. 2003). As a consequence, 118 Marine Ecology32 (Suppl. 1 ) (2011 ) 117-129 © 2011 Blackwell Verlag GmbH Wijnhoven, Escaravage, Herman, Smaal & Hummel Short and mid-long term effects of cockle-dredging on non-target macrobenthic species Oosterschelde The Netherlands ‘Slikken van de Dortsman' i j Dredged plot Control plot '»Sample sites — Route of cockle boats as recorded by satélite tracking systems (STS) Fig. 1. Positioning of the experimental plots (plot numbers Indicated) on the Slikken van de Dortsman tidal flats in the Oosterschelde area (SW Netherlands). the cockle banks in the present study had been free of macrofauna and sediment to detect possible short-term any dredging activity for a 5-year period before the t0 effects, and on 1 and 2 October 2007 (/?)> the sample sites sampling (September 2006). Nine plots of 100 X 100 m were sampled again to detect possible mid-long term were randomly selected within the research area. Because effects. of the expected spatial heterogeneity of habitat and living The current experiment is a standard BACI design communities on the tidal flat, which is largely on a (Smith 2002), which enables the changes observed in North-South gradient, it was decided to separate the nine experimental plots to be compared with those occurring plots into three groups that were spatially clustered in control plots, taking into account autonomic develop (north, middle and south part of the study area). The ments during the study period. positioning and depth of the plots are indicated in Fig. 1. Within each group, two plots were dredged and the last Sampling and measurements plot was used as an undredged/control reference. Within each of the nine plots, five sample sites were randomly At each sample site, five macrofauna and five sediment selected. On 6 September 2006 (io), 45 macrofauna and samples were taken at each sample time. Macrofauna sediment samples were taken, after which the whole area samples consisted of three cores (3 X 0.005 m2) pushed was dredged, with the exception of the three control 30 cm into the sediment within a 1-m radius of the sam plots. The fishing operation was performed for commer ple site, located with a GPS. The macrofauna samples cial purposes by three cockle boats equipped with hydrau were sieved over a 1-mm mesh, fixed with 4% buffered lic dredges. The dredging activity of the ships was formalin and stained with Rose Bengal, after which speci recorded with a satellite tracking system (STS) that mens were determined to the species level, with the revealed, after interpolation of the 1 -min interval signals, exception of the Oligochaeta, Actinaria and Nemertea. dredging tracks all over the experimental plots but not in The numbers per species were counted and densities the control plots (Fig. 1). After dredging, the tracks in the determined. To establish the density of species that are field clearly visible in the sediment were also checked, frequently fragmented, such as polychaetes, the number and were indeed found all over the dredged plots and not of heads was counted. When only body parts were found in the control plots. Fishing activity took place from 5 and no head, the number of specimens was counted as September until 9 November and was restricted to un one. Small or fragmented specimens that could not be sampled areas during the first day of fishing. On 9 classified to species level were classified to genus level November 2006 (iy), all sample sites were sampled for (e.g. Cerastoderma sp., Spio sp. and Arenicola sp.). The Marine Ecology32 (Suppl. 1) (2011) 117-129 © 2011 Blackwell Verlag GmbH 119 Short and mid-long term effects of cockle-dredging on non-target macrobenthic species Wijnhoven, Escaravage, Herman, Smaal & Hummel length of the cockles was also measured as the maximum Data analysis measurable shell length to the nearest millimeter. The total biomass (g ADW, ash-free dry-weight) of The comparisons between treatments were performed each species was determined either directly from the dried according to a standard BACI-ANOVA design where the specimens (2 days at 80 °C) as the decrease in weight effects of treatment, time and time-treatment interaction after 2 h scorching at 560-580 °C, or indirectly by were tested at P < 0.05. As a result of a rather strong length-weight regressions (W = aLb, where W is weight ‘plot’ effect, the individual samples can not be considered in g ADW and L is length in mm). The length-weight to be taken randomly (without consideration of plot ori regressions used were based on (i) specimens scorched gin) within the treatments. Therefore a nested design during this study, (ii) existing data in our BIS (benthos according to: ‘Change in Parameter’ = ‘Parameter aver information system) database from the same area/season, age’ + ‘Treatment effect’ + ‘Plot effect within each treat and (iii) the fresh-weight of the specimens and taxon-spe- ment’ + ‘Unexplained variation’ where ‘Unexplained cific conversion factors from other monitoring campaigns variation’, which is the Error term, equals the variation in BIS. among samples within plots (Sokal & Rohlf 1995). Sediment samples were taken with a 1-cm diameter Because of the decrease in the degree of freedom due to tube pushed 3 cm into the sediment. The median grain the nesting of plots within each treatment, the present size (/im) of the samples was determined by laser-diffrac ANO VA design is a relatively conservative test, which tion methodology using a Mastersizer 2000 (Malvern could fail to detect slight responses to the dredging. We Instruments). are aware that data per treatment should not be gathered for testing when there are differences between plots within treatments. However, we wanted to make sure that Descriptors possible negative effects of cockle-dredging, when present, Plots, treatments and sample times were compared for do not pass unnoticed. We therefore used the more sensi total macrofauna and species densities and biomasses, tive [i.e. without distinction of plot origin) Student-/- test species composition, and frequencies and diversity. The (P < 0.05) for plain comparisons between the treatments length distribution of the cockles (Cerastoderma sp. and or sampling times. Even the robustness of these more Cerastoderma edule combined) was also compared sensitive tests may be relatively low, due to the large vari between treatments. Diversity was measured as species ance among sample sites already at the start of the experi diversity according to the Shannon index, species rich ments (f0), or due to the non-normal distributions. As we ness as the number of individual species and according are especially interested in the developments over time to Margalef, and evenness according to Pielou, calculated for different treatments, independent of autonomous with the software PRIMER 5.2.8 for Windows (Clarke & developments, we calculated the differences between tx Warwick 2001). All total macrofauna and diversity indi and f 0 or f2 and f0, and compared those per treatment cators were calculated with and without taking Cerasto using the Student-f test (P < 0.05). derma sp. and C. edule into account, because it is Effects on individual species were investigated with expected that these cockles are affected by dredging as redundancy analysis (RDA), which is a linear method of the (larger) cockle is the target species. Further, top-10 canonical ordination where environmental variables are lists of the most abundant individual species in chance combined to build the ordination axes, locating the sam of occurrence in samples, in densities and in biomasses ples within the multivariate space defined by the species were put together for each of the sample dates and data, i.e. the log-transformed density, biomass or presence treatments; 18 lists in total. Species mentioned in at frequency data (Ter Braak & Smilauer 1998). The analyses least one of the lists for one of the descriptors were were restricted to the most abundant and dominant spe selected to be related in a multivariate way to time and cies determined as belonging to the top - 1 0 species with treatm ent. respect to the descriptor to be analyzed (density, biomass Differences in the sediment grain size between the or presence frequency) in at least one of the plots at one treatments were also analyzed. To conform to the require of the sample dates. The suitability of the linear response ments of the parametric statistical testing regarding nor model was tested based on the value of the gradient mality in the distribution of the data, the density and length estimated with a detrended correspondence analy biomass data were log-transformed before analyses. The sis (DCA); for a gradient length between 1.5 and 3 SD, diversity indicators (Shannon, Margalef and Pielou), the both linear and unimodal models could be applied. median grain size and cockle length data appeared to be High species score (density, biomass or presence normally distributed in all cases (Kolmogorov-Smirnov frequency) at a given location might be driven by out- test at P < 0.05). of-scope factors coincidental with the treatment, which 120 Marine Ecology32 (Suppl. 1 ) (2011 ) 117-129 © 2011 Blackwell Verlag GmbH Wijnhoven, Escaravage, Herman, Smaal & Hummel Short and mid-long term effects of cockle-dredging on non-target macrobenthic species could lead to misinterpretation based on the RDA plots. □ Control Therefore, the effect of the treatments on densities and 200 : I I I I I : a a a a a a c1 id L b L :’ biomasses of individual species were tested using Student * 0 . : E /■-tests at P < 0.1. Again, this is a rather sensitive test to zi 10,000 : ? 1 avoid missing possible negative impacts. To cope with the ft 180 E 05 bias introduced with the multiple /--testing, a Bonferroni C -S' O)2 co 1000 1 correction according to P < a /n (Sokal & Rohlf 1995) CD will* I was also applied to identify the ‘real’ significant effects on J 160 SCD species. All statistics were executed in SYSTAT for Win dows 1 1 ...... i i i 140 1 2 3 4 5 Power analyses were performed in cases of absence of Plot Plot significant differences to determine the robustness of the 2.5 tests. The minimum difference that could possibly be : ab ab a a a a ab b at? : detected with P < 0.05 was determined taking the varia * • * 5 . : 2.0 tion between samples and the number of samples into 1 1 ■ " i 05 T CD account (Sokal & Rohlf 1995). The number of available | | i : "O SZ observations and their standard deviations were tested at o csz the level of 80% probability, assuming that the data ro belong to one normal distributed population of observa 0.5 tions. The Kolmogorov-Smirnov test indicated that the j ! 0 ! 0.0 values for each of the parameter X treatm ent X tim e 3 4 5 6 intervals can indeed be considered to be normally distrib Plot Plot uted, except for the difference in biomass between f 0 and Fig. 2. Between plots variation before fishing (to). Variation In (A) t2 at the dredged sites. median grain size (um), (B) total density (n-rrT2), (C) total biomass (g ADW-rrT2), and (D) species diversity according to Shannon. Significant differences (P < 0.0S) are Indicated with different letters, whereas the Results same letter In common means no significant differences. Initial situation (f0) Data collected at the start of the experiments (f0) show richness, 24% in species diversity, 48% in total densities distribution patterns over the research area with a clear and 55% in total biomass, with a probability of 80% at differentiation of the three northern plots from the south P < 0.05, as calculated using power analyses. ern plots (Fig. 2). The three northern plots are character ized by smaller median grain sizes and lower total Median grain size macrofauna densities and biomasses (and thereby higher Shannon, Margalef and Pielou diversity indices) than the It was expected that the median grain size would be southern plots (ANOVA, P < 0.05). The intermediate directly influenced by dredging because of the sediment plots showed intermediate values or resembled the resuspension that occurs during fishing activity. However, southern or northern plots. In all cases, as shown also in no difference in median grain size could be detected Fig. 2, the experimental plots clearly resembled the between the control and the dredged areas at any time reference plots (and thus showed the same geographic (f0, or t2; t-test, P < 0.05) or between sampling occa N-S gradient). Therefore, at f 0 the averages of the sions (fo-fi, t0- t 2) (Fig. 3; Table 1). The average median reference plots for median grain size (Fig. 3A), total grain size over all samples equals 175 /zm, varying locally density (Fig 6 A), total biomass (Fig. 6 C), species diver and independently of treatment or time between 150 and sity (Fig. 7E), species richness (Fig. 7A) and evenness 190 /zm. (Fig. 7C) are the same as for the experimental plots. The initial large variance between plots was dealt with by nest Cockles ing the variance within the treatments, which then serves as the error term for the treatment-time interaction to be The present dataset allows an estimation to be made of tested. the impact of the fisheries on the cockle populations. Sig The smallest detectable differences with the used nificantly lower cockle numbers were found in the design, taking initial variability into account, equals 1 % dredged area compared to the control area at A (f-test, in median grain size, 1 0 % in evenness, 2 2 % in species P < 0.05), and lower cockle biomass was found in the Marine Ecology32 (Suppl. 1) (2011) 117-129 © 2011 Blackwell Verlag GmbH 121 Short and mid-long term effects of cockle-dredging on non-target macrobenthic species Wijnhoven, Escaravage, Herman, Smaal & Hummel A □ t0 ED ^ H t2 B D Control b Dredged used by the cockle ships (Hiddink 2003). However, due 200 to the low numbers of observations and their high vari 190 ance, only differences of 82% (larger size classes) to 85% (smaller size classes) can be detected (power analysis; 180 P = 80%, P < 0.05). At A, on average, 38.3% of the cock 'I 170 0 les in the dredged area are large, whereas in the control co 160 area, this figure is 72.2% (significant at P < 0.1). We also found a significant (P < 0.1) decrease in larger cockles 150 between t0 and A in the dredged area compared to the 140 control area. Control Dredged - - 2 0 Fig. 3. Median grain size of the top layer of the experimental plots. Total macrofauna indicators (A) Median grain size fitm) in control and dredged plots before fi0), shortly after fid and 1 year after fi2) fishing. (B) Increase or decrease The total density of macrofauna excluding the cockles of median grain size between t0 and U and between t0 and f2. was relatively stable over time and no significant differ ences were found between the dredged and control areas dredged area than in the control area at both A and t2 (Fig. 6 A; Table 1). A lthough differences in developm ent (P < 0.05) (Fig. 5). The effect of fishing was more evident of the densities might have been present between the two in the biomass changes, as it was primarily the larger treatments, densities do not appear to have been cockles that were fished (Fig. 4). decreased over time in the dredged plots, whereas the The size distribution of the cockles measured in this autonomous trend (visible in the control) shows a slight study can be used to estimate the size selectivity of the decrease (Fig. 6 B). As no significant differences were dredging with respect to the cockles. As indicated by the observed, any differences must have been smaller than average cockle length and length distribution (Fig. 5A), 48% at A and 54% at t2, as calculated by a power analy clear shifts are found towards small-sized individuals sis. between t0 and A and back to the original size distribu In the total biomass, the autonomous effect of decrease tion at t2 in the dredged areas. In the reference area, size over time seems to be even stronger than indicated by the distributions at t0 and A are quite similar, whereas there numbers, and here also such a trend is absent in the is a slight increase in size distribution mode between t0 dredged area (Fig. 6 C,D). However, conducting BACI- and t2 (Figs 4 and 5). Distinguishing different size classes ANOVA, differences in trends do not appear to be signifi (small <23 mm and large >23 mm) of cockles could pro cant (P = 0.444 from t0 to A and P = 0.099 from t0 to t2; vide information about the differential effect of dredging Table 1 ). as a function of shell size. Indeed, a 23-mm shell length is When the total biomass is calculated including the approximately the size of separation of a 15-mm grid as cockles, the possible difference in trends is almost I—I A m u 20 X) 10in E I i ! 20 E 10 Fig. 4. Numbers of cockles distributed over size classes within the control and dredged J in ] tL □ rjfrn 0 10 20 30 40 50 0 10 20 30 40 50 0 10 20 30 40 50 plots (rows) before fi0), shortly after fid and Cockle length (mm) Cockle length (mm) Cockle length (mm) 1 year after fi2) fishing (columns). 122 Marine Ecology32 (Suppl. 1) (2011) 117-129 © 2011 Blackwell Verlag GmbH Wijnhoven, Escaravage, Herman, Smaal & Hummel Short and mid-long term effects of cockle-dredging on non-target macrobenthic species □ Control □ Control ® Dredged ra Dredged B 2 CO CO cCD 40 -C o ^ 1 0 0 0 co 1 CD 'o 0 C l w 100 CD 1.0 CO 0 CO CD 0 0 o "O0 c ■1 Fig. 5. Cockle length in mm (A) and biomass in g ADW-m 2 (B) in control and dredged plots before (f0), shortly after (ffi and 1 year after (f2) fishing. Significant differences in lengths between treatments are indicated with *P < 0.05. 0.0 n Control Dredged 100,000 £-0 .5 10,000 to-ti to-t2 W 1 0 0 0 t0 h t0 t2 CL o CO u 100.00 D io CD° - 1i 10.00 -2 Fig. 7. Macrobenthic assemblage biodiversity descriptors. (A) Species & 0.10 richness according to Margalef in control and dredged plots before (t0), shortly after (ffi and 1 year after (f2) fishing. (B) Increase or decrease of species richness between t0 and U and between t0 and - 1 0 to h t0 t2 t2. (C) Pielou's evenness in control and dredged plots before (f0), shortly after (id and 1 year after (f2) fishing. (D) Increase or decrease Fig. 6. Total macrobenthic density and biomass of the experimental of evenness between t0 and t-\ and between t0 and f2. (E) Shannon's plots. (A) Total density (n-m-2) in control and dredged plots before species diversity in control and dredged plots before (t0), shortly after (i0), shortly after (ffi and 1 year after (f2) fishing. (B) Relative increase (fd and 1 year after (f2) fishing. (F) Increase or decrease of species or decrease of total density between t0 and t-\ and between t0 and f2 diversity between t0 and U and between t0 and f2. calculated as the difference between the natural logarithms. (C) Total biomass (g ADW-m-2) in control and dredged plots before (f0), shortly after (ffi and 1 year after (f2) fishing. (D) Relative increase or decrease Species richness and species diversity show similar pat of total biomass between t0 and U and between t0 and t2 calculated terns when comparing the two treatments over time as the difference between the natural logarithms. (Fig. 7A,E). Initially the two treatments did not differ and this was still the case just after dredging. However, 1 year compensated by the cockle biomass. However, it is clear after dredging the indicators tended to be increased for that on both the short and the mid-long term, with or the dredged area, whereas the control area remained without the cockles included, no decrease in densities or unchanged. The observed trends can not be considered biomass is seen as a result of dredging. significantly different as shown by BACI-ANOVA Marine Ecology32 (Suppl. 1) (2011) 117-129 © 2011 Blackwell Verlag GmbH 123 Short and mid-long term effects of cockle-dredging on non-target macrobenthic species Wijnhoven, Escaravage, Herman, Smaal & Hummel Table 1. Test results of BACI-ANOVA for tlme-treatment Interactions taking 'plot nested within treatment x time' [tlme*plot (treatment)] as Arenicola me 4^ Lanice conchilega error; df((time*treatment)= 1; dfprrnr = 7; cockle data excluded. t Crangon crangon j¡ a Dredged Short-term effects mid-long term OLIGOCHA 'TA Spio martinensis (fo-fi) effects (f0-f 2) Streblespio shrubsolii Mya arenaria ■ammarus locusta mûmçiasp F-ratlo P F-ratlo P Spio sp üTumun Gammarus sp Hydrobia ulvae Capitel 'a capitata median grain size 1.007 0.444 0.000 1.000 ispip'^legtfí^^^:-;::j^ Platynereis dumerilli log (density) 1.285 0.3S2 0.157 0.922 P'OÎÿdçra ligTri*^^ , coloplos armiger t ^ t log (biomass) 0.061 0.979 3.627 0.099 urothoe poseidonis l 2 . Nephtys horni ergii Margalef richness 0.642 0.612 3.S60 0.101 Cerastoderma edule f C b ntrol F 7 Plelou's evenness 0.2S1 0.8S8 0.866 0.S02 Tharyx marioi Shannon diversity 0.494 0.698 3.271 0.113 -1.5 1.0 (P = 0.101 for the trends in richness and P = 0.113 for the trends in diversity between t0 and t2) (Table 1). How Nephtys sp :oloplos armiger ever, there was definitely no negative effect of dredging t 5A . Gammarus locusta on species richness and diversity. Either no effect was Gammarus sp2 k A i f 4 Platynereis dumerilli I ƒ ƒ 4 Nereis longissima found on the evenness, or it must have been smaller than Urothoe poseidonis I ¡krenicola marina Capitella capitata r ! il k Scrobicularia pia 1 0 % according to a power analysis, but then, the impact Carcinus maen\ e lunulata Cerastoderma edule mchilega of dredging seems to be positive instead of negative althica ------Arenicola sp (Fig. 7C). W •WiSfftm. ""“**■ Nephtys a rrosa Control „ M V £í¿:~*f£¡ephtys hombe rgii b pw.-martaken siAA Cerastoderma sp m \ Crangon crangon BRACHYURA Effects on singular non-target species ^ Mya arenaria Hydrobia ^ulvae Dredging could be species-selective in its impact on the -1.5 1.5 macrofauna, either through direct effects on recruitment and mortality rates or, indirectly, through changes in hab Fig. 8. Results of redundancy analyses (RDA) showing relations between species, treatments and sample dates. (A) RDA plot based itat-induced shifts in species composition. Figure 8 shows on log-transformed species densities. (B) RDA plot based on log-trans the results of RDAs based on densities (Fig. 8 A) and bio formed species biomass. mass (Fig. 8 B) for the most abundant species. Results of the presence frequency analyses are not shown, as they are very similar to the density analyses. The graphs show appearance of larger numbers or larger specimens (higher the projection of the gradient axes of the species descrip biomass) in certain plots at certain dates, especially for tor (from low to high values) together with that of the low density species, can not be discriminated from real two environmental treatments (control and dredged) and treatment effects in RDAs. Therefore, for further detailed sampling times (t0, h and t2). The closeness of both pro statistical analyses, per species /--testing would be needed. jections of species and factor gradients points to a direct The series of t-tests on individual species (Table 2) or indirect relationship between the species descriptors showed many species potentially affected by dredging, and either the dredging and/or time (autonomous trend). especially showing increases in densities or biomass. How Species like Cerastoderma edule, Tharyx marioni and ever, in m ultiple tests as perform ed w ith these /--tests, a Hydrobia ulvae seem to be particularly numerous at t0, (conservative) Bonferroni correction should be per indicating an autonomous trend, although the first two formed, after which none of the observed differences are are also related to the control, which indicates a negative really significant. Besides the negative effects on the cockle impact of dredging. Species found in higher numbers in populations, as show n earlier (Figs 4 and 5), the /--tests the control plots at t2, such as Nephtys hombergii and on individual species (Table 2) indicated possible negative Urothoe poseidonis, might also be impacted by dredging. impacts of dredging compared to the autonomous trend On the other hand, for many more species, the highest for three other taxonomic groups. These negative trends numbers were found in the dredged area at /y ( Arenicola are only detected in the short term. Whereas an autono marina, Lanice conchilega) and, especially, at t2. A similar mous increasing trend in densities is observed 6 weeks trend can be found for the analyzed biomass data after the onset of fishing for H. ulvae and the sub-class of (Fig. 8 B), although other species appear in certain cor the Oligochaeta, both were decreased in the dredged area. ners of the graph. It should be noted that coincidental Arenicola sp., which is a special group as it very likely 124 Marine Ecology32 (Suppl. 1 ) (2011 ) 117-129 © 2011 Blackwell Verlag GmbH Wijnhoven, Escaravage, Herman, Smaal & Hummel Short and mid-long term effects of cockle-dredging on non-target macrobenthic species Table 2. Possible differences in density and/or biomass developments between control (C) and dredged (D) plots for the non-target macroben thic species, from f0 to f, and from f0 to f2, as indicated from paired f-tests without Bonferronl correction. negative effects of dredging positive effects of dredging species class development P-level species class development P-level density t0-t, Hydrobia ulvae Gastropoda ?D, Tc 0.042 Carcinus maenas a Malacostraca =D, Tc 0.054 Oligochaeta Clitellata ?D, ÎC 0.006 biomass f0-fr Arenicola sp Polychaeta ÎD, ÎÎC 0.069 Arenicola marina Polychaeta ÎD, Tc 0.059 Capitella capitata Polychaeta ÎD, Tc 0.016 Pygospio elegansa Polychaeta ÎD, Tc 0.016 Streblospio shrubsoliia Polychaeta ÎD, =C 0.030 Urothoe sp.a Malacostraca ÎD, Tc 0.098 density t0-t2 Anaitides mucosa a Polychaeta ÎD, Tc 0.064 Carcinus maenas a Malacostraca =D, Tc 0.054 Harmothoe lunulata a Polychaeta Td , TTc 0.066 Spio sp. Polychaeta Td , Tc 0.036 biomass t0-t2 Crangon crangon Malacostraca Td , Tc 0.014 Gammarus sp. Malacostraca TTd , Tc 0.054 Gammarus locusta Malacostraca TTd , Tc 0.009 Mya arenaria Bivalvia Td , TTc 0.086 Nephtys hombergii Polychaeta Td , Tc 0.097 Platynereis dumerili? Polychaeta TTd , Tc 0.049 Polydora HgnP Polychaeta TTd , Tc 0.027 Scoloplos armiger Polychaeta TTd , Tc 0.051 T Decrease; TT stronger decrease than for the other treatment of the same species;? Increase; TT stronger Increase than for the other treatment of the same species; (=) unchanged. P-levels for significant differences after Bonferroni correction are P < 0.0001 for the dominant species only and P < 0.00002 for all observed species, which were achieved by none of the species. aSpecles not belonging to the 10 most dominant species In densities or biomass in one of the treatment x time combinations. represents small individuals and body parts of A. marina, term, and the decreasing trend in Mya arenaria biom ass tended to increase less in the dredged area than in the seems to be less strong in the dredged area than in the control area (P < 0.1). In contrast to the three groups control area. Whereas most species that tended to that might be affected negatively by dredging in the short increase after dredging belong to the polychaetes, there term, there are six species showing an increase after are also several malacostracans as well as M. arenaria, dredging in the short term. The biomasses of A. marina, which is a bivalve. Capitella capitata, Pygospio elegans, Streblospio shrubsolii and Urothoe sp. appear to increase during the first 6 weeks after the onset of dredging, whereas the opposite Discussion or at least no increase was found in the control area. For Selective fisheries on large cockles Carcinus maenas, the autonomous decreasing trend in density was not observed in the dredged area. The differ Our study demonstrates the efficiency of the fishing pro ence (P < 0.1 in the t-test) between the treatments for cess with a clear reduction of the larger sized (>23 mm) C. maenas was still observable after 1 year. cockles in the dredged areas. This observation confirms, Many more species show an increase after dredging in apart from the visual observation of dredging tracks on the mid-long term. Anaitides mucosa, Harmothoe lunulata the experimental areas, that fishing effectively took place and Spio sp. did not show a decreasing trend or such a on the dredged experimental plots. Partial recovery of strong decreasing trend in density in the dredged area as the cockle population was observed after the fishing. in the control area after a year. Further, seven species Recruitment and growth of the cockles occurred in the appear to show a stronger increase in biomass in the dredged areas, but the average size of the cockles was dredged area than in the control area in the mid-long still smaller after 1 year when compared to the control Marine Ecology32 (Suppl. 1) (2011) 117-129 © 2011 Blackwell Verlag GmbH 125 Short and mid-long term effects of cockle-dredging on non-target macrobenthic species Wijnhoven, Escaravage, Herman, Smaal & Hummel areas. No effect of dredging was detected on smaller Effects on communities and non-target species sized cockles. The failure to detect any effect on the small-sized cockles should be considered, taking into In the present case, no severe environmental impact (on account the low power of the test, which is insensitive either density or biomass) of dredging was detected in the to differences of less than 85%. A thorough analysis short-term observations. Moreover, mid-long term sam addressing the effects on smaller sized cockles would pling showed a slight increase for both densities and bio require larger sampling surfaces to reduce the variance mass in the dredged area, leading to larger biomass in the data and thereby increase the discriminatory compared to the control area after 1 year. This difference power of the test. in biomass mostly resulted from a decrease in the control Whether dredging influences cockle recruitment, as area (autonomous development), which seemingly was suggested by Piersma et al. (2001), could not be deter compensated by an increase in the dredged area. The mined, as no large settlement occurred during the exper higher biomass in the dredged areas was not due to a few iment. Beukema & Dekker (2005) suggest that negative dominant species that might benefit from the disturbed effects of dredging on cockle recruitment mostly occur conditions, but to many species, as shown by the parallel in sediments with very low mud content, where dredg increase in species richness and species diversity under ing might induce a further reduction of the fine material steady levels of evenness. in the sediment below values required for the cockles. Even in such a situation, with increasing biomass, spe In the case of the present study area, where sediment is cies richness and species diversity, one can argue about rather muddy, this effect is therefore not expected to be whether this is a positive or negative development, as cer of great importance. Moreover, the present study did tain species might be favored above others, and some spe not find any effect of dredging on median grain size, cies might be reduced. Therefore we also focused on despite the power of the tests used to detect differences impacts on the individual species. With respect to the indi being large. Natural temporal variation in median grain vidual species, only three species/groups showed a reduc size might have a bigger impact than dredging in the tion on the short term, i.e. with recovery within a year. investigated area. One of the negatively impacted species was the gastropod Hydrobia ulvae, in accordance with Ferns et al. (2000), who showed a depletion of the H. ulvae populations under Effects on the environment the influence of mechanical cockle harvesting. In the pres As indicated above, dredging might affect the composi ent study. H. ulvae was the most numerous species, which tion of the environment, and the top layer of the sedi could positively affect the diversity indices in the dredged ment in particular. On the other hand, sediment areas. This species is also an important food source for sev disturbance can also lead to an increased availability of eral other species (Mendonça et a l 2007). nutrients in the top sediment layer or water layer (Kaiser Previous studies have shown possible negative effects of et al. 2002; Warnken et al. 2003; Nayar et a í 2007), espe dredging on smaller worms (Craeymeersch & Hummel cially in areas with relatively low water turbulence and 2004; Ens et al. 2004), in concordance with the reduction current velocities. Sediment disturbance can undo sedi of oligochaetes observed in the present study under the ment compaction and increase sediment aeration or pore- influence of dredging. Other studies suggested the oppo water renewal (Falcâo et al. 2006). Increased nutrient site response, i.e. an increase of dominance by worms as a availability can also lead to an oxygen decrease due to result of sediment disturbances or increased nutrient increased microbial activities (Riemann & Hoffmann availability in the environment (Reise 1982; Kaiser et al. 1991). The form and depth of disturbance and the type 2002). The positive response of several worm species [i.e. of environment are crucial for whether dredging will have Arenicola marina, Capitella capitata, Pygospio elegans, a negative effect on certain macrobenthic species and Streblospio shrubsolii, Nephtys hombergii, Platynereis whether certain macrobenthic species might profit. dumerilii, Polydora ligni, Scoloplos armiger) as observed in Irrespective of these potential impacts, our study the present experiment might indeed point to a shift failed to detect an effect of dredging on median grain towards worm dominance in the disturbed conditions size, although the power to detect differences was large. after the dredging. In this respect, it is surprising that oli This is in contrast to findings in the Wadden Sea, where gochaetes, which are known to be the first to colonize changes in median grain size and silt content were and dominate in deteriorated conditions (Ysebaert et al. reported in areas that were fished for cockles, although 2003; Wijnhoven et al. 2008), showed decreased the role of winter storms and the vicinity of mussel beds abundances in our study. were also considered relevant factors (Piersma et a í Several Malacostraca species [i.e. Carcinus maenas, 2001). Urothoe sp., Crangon crangon, Gammarus sp., Gammarus 126 Marine Ecology32 (Suppl. 1 ) (2011 ) 117-129 © 2011 Blackwell Verlag GmbH Wijnhoven, Escaravage, Herman, Smaal & Hummel Short and mid-long term effects of cockle-dredging on non-target macrobenthic species locusta) also seem to profit in terms of numbers or bio Consequences of the current design mass from the new conditions in the dredged area. These are mobile species and therefore fast colonizers of Although the current study was able to show, or rule out, disturbed areas. Their increase might result from the fol significant differences in the development of certain lowing two non-exclusive processes. First, they might parameters with a reasonable statistical power, the unbal profit from an increased food availability, i.e. the pres anced design is not ideal. A pairwise experimental design ence of damaged and dead organisms (large macrofa with as many undredged as dredged plots would increase una) in the dredged environment on which they can the power of the tests and would make detection of smal scavenge. Secondly, the increase in space availability due ler differences in development between the treatments to the dredging (with removal of cockles) might sustain possible without increasing the total number of the observed increase of the malacostracans. The hypoth plots/samples. As there is a large variation between plots esis of increased space availability as a result of the and less variation within plots, the power can be dredging might also explain the attenuation in the increased the most by increasing the number of plots, decrease (autonomous trend) of the bivalve M ya arena rather than increasing the number of samples within ria that is observed in the dredged area when compared plots. with the control area. Our hypothesis of decreased com petition and/or increased space availability for non-tar Cockle fisheries in a broader context get species seems to be supported by the partial compensation of the fished-away cockle biomass by bio Under the conditions in the present experiment, it can be mass of non-target species and the apparent slight accel concluded that the impact of dredging on non-target spe eration in growth of the cockles remaining in the cies and the sediment did not appear to be overly destruc dredged area. tive in the mid-long term and therefore likely also not in The present study did not detect any negative effect on the longer term. The negative effects observed for a few bivalve species other than cockles, which are, however, species in the short term were not detected in the mid- potentially influenced by dredging, as shown by other long term, even with the most sensitive tests. The results studies. This result indicates that either dredging has no thus indicate that in the longer term, effects on non-target impact on those bivalves, as they mainly inhabit deeper species of cockle fisheries such as carried out in this study parts of the sediment (e.g. M . arenaria), or that the bival are not to be expected. On the other hand, a longer term ves are not significantly damaged after the processing effect on the cockle populations can not be excluded. through the dredge when they are returned to the sedi Although the cockle stock itself recovered after the dredg m ent (e.g. Macoma balthica). The equivocal effects of the ing and individual growth was observed, a lag in the aver dredging on Arenicola sp. and A. marina should be inter age size was still detected 1 year later in the dredged area preted as a methodological artifact, because Arenicola sp. compared with the control area. An overview of some mostly consists of juveniles and incomplete parts of relevant elements that can help place the present obser A. marina, no other Arenicola species being observed in vations in a larger context is presented below. the research area. The distance between dredged sample sites and The use of the t-test with improved sensitivity, com undredged areas varied in this study between 1 0 and pared with the nested ANO VA, by an increase of the 300 m. In previous studies, showing negative effects of degrees of freedom (no distinction between the plots) cockle-dredging on non-target species, this distance was made it possible to reject the hypothesis that large-scale larger (Hiddink 2003), and thus it might be argued that negative effects on macrofauna density, biomass and recolonization from neighboring areas in our study was diversity would result from the dredging in the investi easier. The extent of recolonization from undredged gated area. Actually, only two groups (H. ulvae and the neighboring areas for both the recovery of the cockles oligochaetes) showed a negative effect of dredging in the and the increase in non-target species was not addressed short term in this experiment, and the differences were directly in this study because the undredged experimental not significant (ANOVA, P > 0.05). This also accounts areas were very small compared to the total area dredged. for the range of species that showed positive effects of We believe therefore that the undredged areas can only dredging on the short and mid-long term. From this, it have contributed in a minor way to the recolonization in can be concluded that there was no negative impact on the much larger dredged area. Recolonization from within the whole range of species, with the exception of the the dredged area might also have been possible, as the two groups just mentioned. Whether some species might dredging intensity was not uniformly distributed over have benefitted from the dredging remains unsubstanti the dredged area. However, satellite tracking system ated. registrations did indicate that undisturbed parts were Marine Ecology32 (Suppl. 1) (2011) 117-129 © 2011 Blackwell Verlag GmbH 127 Short and mid-long term effects of cockle-dredging on non-target macrobenthic species Wijnhoven, Escaravage, Herman, Smaal & Hummel scarce. Therefore, we conclude that recolonization only Acknowledgements contributed in a minor way to the absence of significant results of cockle fisheries on the zoomacrobenthic com We would like to thank the research assistants of the m unity. Monitor Taskforce (NIOO-CEME) for sampling and The present study deals with the impact of a single macrofauna determination. Thanks to the crew of the dredging event, whereas more disruptive effects on com ships YE172, YE98 and YE42, and to Joke Kesteloo- munities can be reasonably postulated by recurrent Hendrikse and Douwe van de Ende (IMARES) for their (yearly or even more frequent) dredging activities. At the assistance during the fieldwork. The research has been more disruptive intensity level, it can be reasonably executed in the frame of the ‘Project Research Sustain expected that dredging activities may prevent the estab able Shellfish Fisheries’ (PRODUS), and we would like lishment of long-living ecostructures such as mussel/oys- to thank the ‘Cooperative Producers Organization of the ter banks and seagrass fields, together with their Dutch Cockle Fisheries’ and Jaap Holstein in particular associated flora and fauna (Dittmann 1990; Boström & for their cooperation. Thanks to the organizing commit Bonsdorff 2000; Jaramillo et al. 2007). The negative tee of the 44th EMBS 2009, who gave us the opportu impact found by many other studies on a range of non nity to present and discuss this study. This is target species could be the result of other dredging tech publication 4892 of the Netherlands Institute of Ecology niques that are more destructive (Haii & Harding 1997; (NIOO-KNAW), and Monitor Taskforce Publication Ser Ferns et a í 2000) than the hydraulic dredging used in the ies 2 0 1 0 - 1 0 . present case. Local conditions of the fishing area should also be References taken into account when considering the effects of fisher ies on benthos. Queirós et al. (2006) and Hiddink et a l Beukema J.J. ( 1995) Long-term effects of mechanical harvest (2007) clearly point at the strong effect of habitat charac ing of lugworms Arenicola marina on the zoobenthic teristics such as sediment and productivity in relation to community of a tidal flat in the Wadden Sea. Netherlands the sensitivity to dredging disturbances. A common con Journal of Sea Research, 33, 219-227. clusion by both papers was that the degree of natural dis Beukema J.J., Cadée G.C. (1999) An estimate of the sustainable turbance determines the degree of sensitivity to fishing rate of shell extraction from the Dutch Wadden Sea. Journal activities. It is thus possible that communities of sandy of Applied Ecology, 36, 49-58. substrates as in the Wadden Sea are differentially sensitive Beukema J.J., Dekker R. (2005) Decline o f recruitm ent success in cockles and other bivalves in the Wadden Sea: possible to disturbances compared with communities found on role of climate change, predation on postlarvae and fisheries. muddy sediments, as in our study. This could explain the Marine Ecology Progress Series, 287, 149-167. differences between the negative impacts of cockle fisher Boström C., Bonsdorff E. (2000) Zoobenthic community ies found in the Wadden Sea (Piersma et a í 2001) and establishment and habitat complexity - the importance of the absence of significant results in our area. However, seagrass shoot-density, morphology and physical disturbance the relationship with the mud content is not consistent, for faunal recruitment.Marine Ecology Progress Series, 205, as effects of fisheries, as indicated by Queirós et a l 123-138. (2006), were more negative on the muddier areas, Clarke K.R., Warwick R.M. (2001)Change in Marine Commu whereas in our muddy area no significant effects could be nities: an Approach to Statistical Analysis and Interpretation, found. Further studies should focus more on these 2nd edn. PRIMER-E, Plym outh: 178 pp. aspects. Several studies also show negative effects of Coosen J., Twisk F., Van der Tol M.W.M., Lambeck R.H.D., dredging on the bivalve recruitment (Piersma et al. 2001; Van Stralen M.R., Meire P.M. (1994) Variability in stock Hiddink 2003). In the absence of massive bivalve recruit assessment of cockles (Cerastoderma edide L.) in the ment in the area during the research period, no conclu Oosterschelde (in 1980-1990), in relation to environmental sion can be drawn relative to the effect of dredging on factors. Hydrobiologia, 282/283, 381-395. the recruitment from the present study. Cotter A.J.R., Walker P., Coates P., Cook W., Dare P.J. (1997) From the results of our study we conclude that sustain Trial of a tractor dredger for cockles in Burry Inlet, South able cockle fisheries may be possible, taking into consider Wales. ICES Journal of Marine Science, 54, 72-83. ation the above-mentioned aspects. The sensitivity and Craeymeersch J.A., Hummel H. (2004) Effectonderzoek kokkel- the recovery potential of a dredged area, amongst other visserij Voordelta. RIVO-rapport C012/04, RIVO, Yerseke, factors related to the type of habitat and probably also to The Netherlands: 41 pp (in Dutch). the period of non-disturbance, should be specified on the Dittmann S. (1990) Mussel beds - amensalism or ameliora basis of adequate field-experiments, preferably using a tion for intertidal fauna? Helgoländer Meeresunters, 44, 335-352. BACI approach. 128 Marine Ecology32 (Suppl. 1 ) (2011 ) 117-129 © 2011 Blackwell Verlag GmbH Wijnhoven, Escaravage, Herman, Smaal & Hummel Short and mid-long term effects of cockle-dredging on non-target macrobenthic species Ens B.J., Smaal A.C., De Vlas J. (2004) The Effects of Shellfish C l/3. Alterra-rapport 955, Wageningen, The Netherlands: Fishery on the Ecosystems of the Dutch Wadden Sea and 146 pp. (in Dutch). Oosterschelde; Final report on the Second Phase of the Mendonça V.M., Raffaelli D.G., Boyle P.R. (2007) Interactions Scientific Evaluation of the Dutch Shellfish Fishery Policy between shorebirds and benthic invertebrates at Culbin (EVA II). Alterra-rapport 1011, RIVO-rapport C056/04, Sands lagoon, NE Scotland: effects of avian predation on RIKZ-rapport RKZ/2004.031, Wageningen, The Nether their prey community density and structure. Scientia lands: 2 1 2 pp. Marina, 71, 579-591. Falcâo M., Caetano M., Serpa D., Gaspar M., Vale C. (2006) Nayar S., Miller D.J., H unt A., Goh B.P.L., Chou L.M. (2007) Effects of infauna harvesting on tidal flats of a coastal Environmental effects of dredging on sediment nutrients, lagoon (Ria Formosa, Portugal): implications on phosphorus carbon and granulometry in a tropical estuary. Environmen dynamics. Marine Environmental Research, 61, 136-148. tal Monitoring and Assessment, 127, 1-13. Ferns P.N., Rostron D.M., Siman H.Y. (2000) Effects of Piersma T., Koolhaas A., Dekinga A., Beukema J.J., Dekker R., mechanical cockle harvesting on intertidal communities. Essink K. (2001) Long-term indirect effects of mechanical Journal of Applied Ecology, 37, 464-474. cockle-dredging on intertidal bivalve stocks in the Wadden Geurts van Kessel A.J.M., Kater B.J., Prins T.C. (2003) Verand Sea. Journal of Applied Ecology, 38, 976-990. erende draagkracht van de Oosterschelde voor kokkels. Rap Queirós A.M., Hiddink J.G., Kaiser M.J., Hinz H. (2006) Effects portage van Thema’s 2 en 3 uit het ‘Lange Termijn of chronic bottom trawling disturbance on benthic biomass, Onderzoeksprogramma Voedselreservering Oosterschelde’, in production and size spectra in different habitats. Journal of het kader van de Tweede Evaluatie van het Nederlands Schel- Experimental Marine Biology and Ecology, 335, 91-103. pdiervisserijbeleid, EVA II. Rapport RIKZ/2003.043, RIVO Reise K. (1982) Long-term changes in the macrobentric inver rapport C062/03, Middelburg, The Netherlands: 128 pp. (in tebrate fauna of the Wadden Sea: Are polychaetes about to Dutch). takeover?. Netherlands Journal of Sea Research, 61, 29-36. Hall S.J., Harding M.J.C. (1997) Physical disturbance and mar Riemann B., Hoffmann E. (1991) Ecological consequences of ine benthic communities: the effects of mechanical harvest dredging and bottom trawling in the Limfjord, Denmark. ing of cockles on non-target benthic infauna. Journal of Marine Ecology Progress Series, 69, 171-178. Applied Ecology, 34, 497-517. Smith E.P. (2002) BACI design. In: El-Shaarawi A.H., Hiddink J.G. (2003) Effects of suction-dredging for cockles on Piegorsch W.W. (Eds), Encyclopedia of Environmetrics, non-target fauna in the Wadden Sea. Journal of Sea Vol. 1. John Wiley 8 c Sons, Ltd, Chichester: pp. 141-148. Research, 50, 315-323. Sokal R.S., Rohlf F.J. (1995) Biometry: the Principles and Hiddink J.G., Jennings S., Kaiser M.J. (2007) Assessing and Practice of Statistics in Biological Research. 3rd edn. W.H. predicting the relative ecological impacts of disturbance on Freeman and Company, New York: 887pp. habitats with different sensitivities. Journal of Applied Ecol Ter Braak C.J.F., Smilauer P. (1998) CANOCO Reference Man ogy, 44, 405-413. ual and User’s Guide to Canoco for Windows. Software for Jaramillo E., Contreras H., Duarte C. (2007) Community Canonical Community Ordination (Version 4). Centre for structure of the macroinfauna inhabiting tidal flats charac Biometry Wageningen, The Netherlands: 351pp. terized by the presence of different species of burrowing Warnken K.W., Gili G.A., Dellapenna T.M., Lehman R.D., bivalves in Southern Chile. Hydrobiologia, 580, 85-96. Harper D.E., Allison M.A. (2003) The effect of shrimp Kaiser M.J., Broad G., Hall S.J. (2001) Disturbance of intertidal trawling on sediment oxygen consumption and the fluxes of soft-sediment benthic communities by cockle hand raking. trace metals and nutrients from estuarine sediments. Journal of Sea Research, 45, 119-130. Estuariae, Coastal and Shelf Science, 57, 25-42. Kaiser M.J., Collie J.S., Hall S.J., Jennings S., Poiner I.R. Wijnhoven S., Sistermans W., Hummel H. (2008) Historic (2002) Modification of marine habitats by trawling activi developments in macrozoobenthos of the Rhine-Meuse ties: prognosis and solutions. Fish and Fisheries, 3, 114— estuary: from a tidal inlet to a freshwater lake. Estuarine, 136. Coastal and Shelf Science, 76, 95-110. Leitâo F.M.S., Gaspar M.B. (2007) Immediate effect of inter Ysebaert T., Herman P.M.J., Meire P., Craeymeersch J., tidal non-mechanised cockle harvesting on macrobenthic Verbeek H., Heip C.H.R. (2003) Large-scale spatial patterns communities: a comparative study. Scientia Marina, 71, in estuaries: estuarine macrobenthic communities in the 723-733. Schelde estuary, NW Europe. Estuarine, Coastal and Shelf Leopold M.F., Dijkman E.M., Cremer J.S.M., Meijboom A., Science, 57, 335-355. Goedhart P.W. (2004) De effecten van mechanische kokkelvis- Zwarts L. (2004) Bodemgesteldheid en mechanische kokkelvisserij serij op de benthische macrofauna en hun habitat. Eindverslag in de Waddenzee. Rapport RIZA/2004.028, Lelystad, The EVA II (Evaluatie Schelpdiervisserij tweede fase), Deelproject Netherlands: 129 pp. (in Dutch). Marine Ecology32 (Suppl. 1) (2011) 117-129 © 2011 Blackwell Verlag GmbH 129 anmarine evolutionary perspective ecology » Marine Ecology. ISSN 0173-9565 ORIGINAL ARTICLE Charles Darwin and marine biology Philip S. Rainbow Department of Zoology, Natural History Museum, London, UK Keywords Abstract Barnacle; Beagle; coral reefs; Darwin; Grant; Lyell; transmutation; unlformltarlanlsm. In a celebration of the 200th anniversary of his birth in 1809, this short essay explores the influence of marine biology on Charles Darwin, and vice versa. Correspondence Darwin made his first forays into the world of marine biology as a medical stu Philip S. Rainbow, Department of Zoology, dent in Edinburgh from 1825 to 1827. He came under the influence there of Natural History Museum, London SW7 5BD, the Lamarckian Robert Grant, and developed an understanding of the simple UK. organisation of the early developmental stages of marine invertebrates. Yet Dar E-mail: [email protected] win balked at Lamarckian transmutation. The voyage of the Beagle led to Dar Accepted: 17 November 2010 win’s perceptive theory of the origin of coral reefs, an origin still mainly accepted today. This theory was steeped in the uniformitarianism of the geolo doi : 10.1111/j. 1439-0485.2010.00421 .x gist Lyell, depending on the slow, gradual growth of billions of coral polyps keeping pace with slow sinking of land to produce an atoll. Prom 1846 to 1854 Darwin revolutionised the understanding of barnacles, producing monographs still relevant today. His barnacle studies gave him the credibility to pronounce on the origin of species; he found great variation in morphology, and a series of related species with remarkable reproductive adaptation culminating in the presence of dwarf males. Barnacles showed him an evolutionary narrative laid out before him, and contributed greatly to his qualification and confidence to write with authority on the origin of species. Born in Shrewsbury on 12 Pebruary 1809, Charles Robert to carry out his external hospital study in Edinburgh Darwin could hardly have been said to have had the sea (Desmond & Moore 1991). Charles, however, became in his blood, but as a child he was an inveterate collector disillusioned with his medical studies as he experienced of objects such as shells and birds eggs and developed an the drudgery of his lectures and poor quality of his lec early interest in natural history. By his teens, hunting had turers, and then the distress of clinical studies with become Charles’ passion, and in 1825 his exasperated associated blood, gore and suffering. Initial diligence father, the well respected local doctor Robert Darwin, gave way to the sampling of student life. On his own came to utter the oft-quoted prediction ‘You care for in his second year, Darwin found a diversion in marine nothing but shooting, dogs, and rat-catching, and you biology. will be a disgrace to yourself and all your family’ (Des In November 1826, Charles Darwin joined the Plinian mond & Moore 1991). Strong action was needed to halt Society, an undergraduate group discussing natural his Charles’ aimless way of life and so, following his father tory and antiquarian researches, and occasionally going and elder brother Erasmus, medicine was chosen as the on collecting expeditions together. This proved a safety required remedy. valve from medicine, and, over the academic year, Darwin In October 1825, Charles Darwin, aged 16, found accompanied his Plinian friends on zoological walks on himself enrolled in the University of Edinburgh to study the shores of the Firth of Forth, and ventured out on medicine, accompanied for the first year by Erasmus trawlers fishing at sea. He became familiar with a host of who, although a medical student at Cambridge, was able marine invertebrates previously strange to him, including 130 Marine Ecology32 (Suppl. 1 ) (2011 ) 130-134 © 2011 Blackwell Verlag GmbH Rainbow Charles Darwin and marine biology sponges, soft corals ( Alcyonium digitatum), sea slugs founded as London University in 1826, admitting stu (Tritonia hombergi), polychaetes (the sea mouse Aphrodite dents regardless of religion and gender, a secular alterna aculeata) and bryozoans (Flustra foliacea). tive to Oxford and Cambridge. London University was At this time Darwin came under the influence and clearly a more suitable venue for Grant’s seditious views mentorship of a man who would be key to the later than God-fearing Edinburgh. development of Darwin’s ideas on evolution. Robert The 2 years at Edinburgh convinced Charles Darwin Edward Grant was a marine invertebrate zoologist and a that medicine was not for him, and in 1827 he left Edin fellow Plinian, living in Edinburgh on a decreasing legacy burgh a disappointed man: he did not like medicine, nor from his late father (Desmond & Moore 1991). Grant the men who pursued it; he had found no qualities in became Darwin’s unofficial tutor on marine invertebrates, professors to generate long-lasting respect (even Grant teaching him to make observations and to dissect speci had disappointed him); and he was not ready to be a mens. Through Grant, Darwin developed an understand transmutationist or be labelled a radical like his grandfa ing of animal development and the simple organisation of ther, Erasmus Darwin (Browne 1995). Medicine was not the early life-history stages of particular invertebrates. for him, but he now had little choice - the family fell Grant was an expert on sponges recognised by his peers, back on the typical safety net for second sons, the Church as exemplified by the naming of the newly erected sponge of England. So from 1827 to 1831, Charles Darwin found genus Grantia in his honour by John Fleming in 1828; himself at Christ’s College at the University of Cambridge the common local sponge Spongia compressa, the purse on the first stage of his journey to Holy Orders. There sponge, became Grantia compressa. It was Grant who was still room for natural history, now under the influ coined the name ‘Porifera’ for the sponges. ence of John Henslow, the Professor of Botany, and for After working with Grant, Darwin (aged 18) gave a talk geology under the influence of Adam Sedgwick, Professor to the Plinian Society on 17 March 1827, showing that of Geology. the larvae of the bryozoan Flustra foliacea use cilia for However, there was no more immediate access to mar locomotion and that the black markings (sea pepper ine biology for Darwin until the portentous year of 1831. corns) on the shells of oysters are the eggs of the marine Then Henslow introduced Darwin to Robert FitzRoy, leech Pontobdella muricata. A triumph for a budding captain of HMS Beagle, who subsequently invited to Dar marine biologist but, according to Darwin’s daughter win to join a voyage around the world as a self-financing Elenrietta, Darwin had been scooped 3 days earlier by gentleman naturalist. The voyage lasted from December Grant in a talk to the more formal (graduate) student 1831 to October 1836. Darwin regularly sent back natural society, the Wernerian Natural History Society - an intro history and geology collections which gained him a scien duction for Darwin to ‘the jealousy of scientific men’ tific reputation in his absence, and the voyage changed (Browne 1995). his life for ever. Grant, however, represented something more - sedition In January 1835, Darwin collected many specimens of a personified (Desmond & Moore 1991). Grant was a fran large intertidal gastropod mollusc on a shore in the Cho cophile who had studied anatomy and embryology in nos archipelago, Chile, a collection not considered even France with Geoffroy Saint-Hilaire. Correspondingly, worthy of reference in several editions of Darwin’s journal Grant was a Lamarckian, more openly so later, for his of the voyage (Darwin 1839). The mollusc concerned was views were still forming at this time. Lamarck (1809) used a muricid gastropod, Concholepas concholepas (as Conc the term ‘transmutation’ for his theory that described the holepas peruviana), the shells of which were riddled with altering of one species into another. Lamarck did not cavities containing minute animals, no bigger than pin propose common ancestry but considered that complex heads. These were to be later identified as boring barna forms transmutated from simple forms of life created cles, and these were also destined to affect Darwin’s continuously by spontaneous generation. As a Lamarck future life enormously. ian, Grant arranged life into chains, considering that the Before leaving on the Beagle, Darwin had been greatly origins of animals and plants lay in the simplest forms; influenced by Sir Charles Lyell, a leading geologist of the and that the natural ordering, simple to complex, of time, who had published the first edition of his Principles sponges represented the historical order of appearance of o f Geology in three volumes (1830, 1832, 1833). Lyell was sponges. Thus Grant directly exposed Darwin to evolu a believer in uniformitarianism, a philosophy claiming tionary theory, with the associated concepts of structural that geological and biological forces have always been homology and unity of plan with similar organs present working in the same way and at the same intensity over in different animals. Grant went on in 1827 to become ages. This view of uniformitarianism was in conflict with Professor of Comparative Anatomy for life (1874) at Uni the then-prevailing theory of catastrophism, which con versity College, London. University College had been sidered that the earth experienced major changes only as Marine Ecology32 (Suppl. 1) (2011) 130-134 © 2011 Blackwell Verlag GmbH 131 Charles Darwin and marine biology Rainbow a result of large catastrophic events. A key consideration unstudied for 6 years. Robert Grant, now in London, was whether the earth was old enough to experience volunteered to help, particularly with ‘lower animals’, but large-scale changes in any other way, but Lyell thought it was turned down by Darwin who had become a compe necessary to create a vast time scale for Earth’s history to tent (and competing) coral expert, given his interest in vouch for fossil remains of extinct species, excluding sud coral reef formation. In fact, ironically, the corals were den geological catastrophes. Charles Darwin shared a sup not monographed. Nor did Darwin and Grant have any port for uniformitarianism. thing more to do with each other (Desmond & Moore The formation of coral reefs was a lively topic for 1991). debate at the time, and in the second volume (1832) of By October 1846, Darwin thought that he had Principles of Geology, Lyell had explained the origin of described all his Beagle specimens, and turned his atten coral atolls as coral reefs growing up from the crater rims tion to a single remaining barnacle species, the boring of underwater volcanoes. Volume 2 was sent out to Dar barnacles in the gastropod shells collected in Southern win by his mentor and faithful correspondent Professor Chile in January 1835. The barnacle was clearly ‘quite Henslow, to reach him in Montevideo in November 1832. new and curious’ - he called it that ‘ill-formed little Darwin’s observations on the Beagle, however, had con monster’ - certainly aberrant and the world’s smallest vinced him otherwise. Darwin drafted an alternative the barnacle. Darwin did not know how to classify it, refer ory for the origin of coral reefs that essentially stands ring to it as ‘Mr Arthrobalanus’. A new microscope was today. Lyell accepted it immediately. necessary, but so was a comparison with ‘more normal’ An essential point in coral reef formation is that reef- barnacles. So Darwin started borrowing the necessary building corals only grow in well-lit shallow waters, sup specimens, and the project grew and grew as he soon plying light to their symbiotic zooxanthellae, and could appreciated the state of chaos of the knowledge of bar not grow up from deeper, dark depths. In Darwin’s nacles. explanation, corals form fringing reefs just below low tide Why did Darwin embark on a project that was to along tropical coastlines, and eventually the coral reef will occupy the next 8 years? Darwin had been formulating grow out to become a barrier reef. Darwin had seen the his ideas on variation and natural selection, resulting effects of earthquakes in Chile and knew that land could eventually in The Origin of Species (Darwin 1859), and he rise or fall. If the coast is sinking slowly, then the growth agreed with the view of his botanist friend Joseph Hooker of coral could keep pace. If the land sinks beneath the that no-one had a right to examine the question of spe waves - an atoll is formed. His first hand observations in cies who had not described many. So Darwin would earn Keeling Atoll helped convince him of his views. Darwin that right. Barnacles would establish his credentials, and a could see that, although coral reefs were huge geological thorough examination of all barnacle varieties could put structures on a world scale, they were created by the slow, him in a commanding position when discussing natural gradual growth of billions of tiny creatures over vast selection. In fact, as he proceeded, he began to uncover reaches of time. This was an example of Lyell’s ‘uniformi- the most extraordinary proofs of his notebook specula tarian’ principle in action, the cumulation of small tions (Desmond & Moore 1991). changes over a long period, to be repeated in Darwin’s So Darwin studied barnacles (‘my beloved barnacles’) studies on earthworms years later. Later observations of in his study at Down House in Kent from 1846 to 1854. cores of coral limestone to great depths (culminating in He would go on to deliver a thorough reappraisal of both studies in Bikini Atoll in the 1950s) provided supporting living and fossil barnacles, monographs that still com evidence for land sinking at an appropriate rate - coral mand the field today (Darwin 1851a,b, 1854a,b). could simply not have grown up from these depths in the So what are barnacles? Zoological folklore has it that absence of light. Louis Agassiz described a barnacle as ‘nothing more than Darwin published Coral Reefs in 1842 (Darwin 1842). a little shrimp-like animal, standing on its head in a lime It showed the tight logical structure to become evident stone house and kicking food into its mouth’. again in the Origin o f Species to be published in 1859. Barnacles are indeed crustaceans, crustaceans that lack On his return from the Beagle in 1836, Darwin sought an abdomen and with a head enormously developed as a to place his collections with experts to identify and stalk or, as in the sessile barnacles, that limestone house describe them - not altogether an easy task. The mam of overcalcified cuticle. Darwin (1851a) drew a diagram mals were placed with Richard Owen - later to found the (Fig. 1) of the relationship between the anatomy of a Natural History Museum in South Kensington and vehe stalked barnacle (Lepas) and a decapod crustacean ( Luci mently oppose Darwin’s views on evolution. He avoided fer) discussing similarities or differences in terms of Robert Brown (Keeper of Botany) at the British Museum the homologies of the body parts, a conceptual way of who had been sitting on a collection of Galapagos plants, thinking with which he was clearly happy. Typically we 132 Marine Ecology32 (Suppl. 1 ) (2011 ) 130-134 © 2011 Blackwell Verlag GmbH Rainbow Charles Darwin and marine biology species of the stalked barnacle genus Scalpellum showed a gradient from complete hermaphroditism to hermaphro dites with small complemental ‘dwarf males with the lar ger ‘hermaphrodite’ acting as a female (e.g. Scalpellum scalpellum). Was this an example of what transmutation might look like? Darwin (1851a) also discovered that the stalked barnacle Ibla cumingii also has large females with small complemental males. The dwarf male shows great reduction in form as it becomes specialised for reproduc tion only. Transmutation? The study of barnacles indeed provided Charles Dar win with many of the facts that he needed to support Fig. 1. The diagram drawn by Darwin (1851a) of the relationship his ideas on evolution - both through comparative between the anatomy of a stalked barnacle (Lepas) and a planktonic anatomy and through the study of fossils. What Darwin decapod crustacean (Lucifer) to show the homologies of the external found in his barnacles (Stott 2003) was variation parts (m, mouth). beyond his wildest imaginings, and reproductive modes that took his breath away, with the development of Table 1. Barnacles and their relatives are members of the class Max complemental males living parasitically on the female, illopoda, together with the likes of tantulocarids, branchiurans, pent- no more than reproductive sacs of sperm, with no astomids, mystacocarids and copepods (Martin & Davis 2001). heads, stomachs or digestive systems. Barnacles had Subclass Thecostraca adapted to their environments and an evolutionary nar Infraclass Facetotecta rative branched out before his eyes (Stott 2003). Dar Infraclass Ascothoracida win’s barnacles showed him what transmutation could Infraclass Cirripedia look like. Bit by bit, each apparently trivial adaptation Superorder Acrothoracica in living structure accumulated, one after another, until Superorder Rhicocephala animals became so distinct from their parents and cous Superorder Thoracica Order Pedunculata ins that they could be called a different species (Browne Order Sessilia 1995). While Darwin made fundamental contributions to the study of coral reefs and barnacles, clearly in return, mar recognize three groups of cirripede barnacles today (Acro ine biology contributed much to Darwin’s development thoracica, Rhizocephala, Thoracica), closely related to two of evolutionary thinking. The interaction of marine biol other taxa, the Ascothoracida and Facetotecta (Table 1, ogy and the intellect of Darwin was key to the develop after Martin & Davis 2001). ment of his supreme contribution to biology - the So what was that ‘ill-formed little monster’ ‘Mr Arthro- mechanism of natural selection acting on natural varia balanus’ found boring into the shells of Concholepas conc tion to explain the origin of species and the evolution of holepas, collected in southern Chile in January 1835? In organisms. 1849, Hancock described a boring barnacle from the colu mella of whelk shells, Buccinum undatum, occupied by Acknowledgements the hermit crab Pagurus bernhardus - Alcippe (now Try petesa) lampas (Hancock 1849). ‘Mr Arthrobalanus’ This essay results from an invited lecture given at the turned out to be a close relative. Darwin (1854a) 44th European Marine Biology Symposium in Liverpool described it as Cryptophialus minutus, and placed both in 2009 in acknowledgement of the 200th anniversary of species in a new order, the Abdominalia. The order was the birth of Charles Darwin. The lecture attempted to renamed the Acrothoracica by Gruvel (1905) because bring Darwin’s specific contributions in marine biology these burrowing barnacles lack an abdomen, the taxon to a wide audience of marine biologists, perhaps under featuring as a superorder in Table 1. Acrothoracicans live standably only aware of Darwin’s hugely significant con in burrows in calcareous rocks and shells, and have a tribution of natural selection as a mechanism body and cirri not dissimilar to that of the more com underpinning evolution. I am indebted to Professor Chris mon thoracicans. Frid for the invitation. It will come as no surprise to Darwin discovered reproductive modes in some barna readers that this is not a paper of original scholarship, cles that astounded him. Most barnacles are hermaphro but simply a secondary compilation from excellent works dite, but not all. Darwin (1851a) discovered that different of real biographical scholarship published on Charles Marine Ecology32 (Suppl. 1) (2011) 130-134 © 2011 Blackwell Verlag GmbH 133 Charles Darwin and marine biology Rainbow Darwin. I have leant heavily on the outstanding biogra Darwin C.R. (1854a) A Monograph on the Subclass Cirripedia. phies by Janet Browne (1995, 2002) and by Adrian Des Vol.2 The Balanidae, The Verrucidae, etc.. Ray Society, Lon mond and James Moore (1991), to whom interested don: 684 pp. readers are referred. Rebecca Stott (2003) has produced a Darwin C.R. (1854b) A Monograph on the Fossil Balanidae and fascinating account of the barnacle years which domi Verrucidae of Great Britain. Palaeontographical Society, Lon nated the life of Darwin and his family at Down House don: 44 pp. from 1846 to 1854. Darwin C.R. (1859) On the Origin of Species by Means of Natu ral Selection, or the Preservation of Favoured Races in the Struggle for Life. John Murray, London: 502 pp. References Desmond A.J., Moore J.R. (1991) Darwin. Michael Joseph, London: 808 pp. Browne J. ( 1995) Charles Darwin: Voyaging. Jonathan Cape, Gruvel A. ( 1905) Monographie des Cirrhipèdes ou Thécosracés. London: 605 pp. Masson et Cie, Paris: 472 pp. Browne J. (2002) Charles Darwin: The Power of Place. Jonathan Hancock A. (1849) Notice of the occurrence on the British Cape, London: 591 pp. coast of a burrowing barnacle belonging to a new order of Darwin C.R.. (1839) Journal of Researches into the Geology and the class Cirripedia. Annals and Magazine of Natural History, Natural Histoiy of the Various Countries Visited by H.M.S. 4, 305-314. Beagle etc.. Henry Colburn: London: 629 pp. Lamarck J.-B. (1809) Philosophie Zoologique, ou Exposition des Darwin C.R. (1842) The Structure and Distribution of Coral Considérations Relatives à l’Histoire Naturelle des Animaux. Reefs. Part 1 of The Geology of the Voyage of the Beagle etc. Dentu et l’Auteur, Paris: 422 pp. Smith Elder and Co., London: 214 pp. Martin J.W., Davis G.E. (2001) An updated classification of Darwin C.R. (1851a) A Monograph on the Subclass Cirripedia. the recent Crustacea. Natural History Museum of Los Angeles Vol.l The Lepadidae. Ray Society, London: 400 pp. County Contributions in Science. 39, 1-124. Darwin C.R. (1851b) A Monograph on the Fossil Lepadidae, or, Stott R. (2003) Darwin and the Barnacle. Faber and Faber, Pedunculated Cirripedes of Great Britain. Palaeontographical London: 309 pp. Society, London: 88 pp. 134 Marine Ecology32 (Suppl. 1 ) (2011 ) 130-134 © 2011 Blackwell Verlag GmbH/<3’'<3-'S/3 /g ' s /<3’'£ 'S /ê>'Sl'g /S '