Context- and density-dependent effects of introduced oysters on biodiversity

1

Context- and density-dependent effects of introduced oysters on biodiversity 1

2

Dannielle. S. Green1,2

and Tasman. P. Crowe1 3

1Marine Biodiversity, Ecology and Evolution Group, School of Biology and Environmental 4

Sciences, University College Dublin, Dublin, Ireland. 5

2Urban Institute Ireland, Earth Institute, University College Dublin, Ireland. 6

7

Corresponding author: 8

Dannielle Senga Green 9

Marine Biodiversity Ecology and Evolution 10

School of Biology and Environmental Sciences 11

Science Centre West 12

University College Dublin, Belfield, Dublin 4, Ireland. 13

Email: [email protected] 14

Phone: +353 1716 2687 15

16

Running title: Effects of Crassostrea gigas on biodiversity 17

Key words: Invasive species, biodiversity, ecosystem engineer, environmental context, 18

density-dependent, Crassostrea gigas 19

20

Citation: Green, D.S. and Crowe, T.P. (2013) Context- and density-dependent effects of 21

introduced oysters on biodiversity. Biological Invasions DOI: 10.1007/s10530-013-0569-x 22

23

24

25

26

27

2

Abstract 28

Pacific oysters, Crassostrea gigas, have been introduced throughout much of the world, 29

become invasive in many locations and can alter native assemblage structure, biodiversity 30

and the distribution and abundance of other species. It is not known, however, to what extent 31

their effects on biodiversity change as their cover increases, and how these effects may differ 32

depending on the environmental context. Experimental plots with increasing cover of oysters 33

were established within two estuaries in two different habitats commonly inhabited by C. 34

gigas, (mussel-beds and mud-flats) and were sampled after 4 and 15 months. Within mud-flat 35

habitats, macroscopic species living on or in the substratum increased in richness, Shannon-36

Wiener diversity and number of individuals with oyster cover. In mussel-bed habitats, 37

however, these indices were unaffected by the cover of oysters except at one estuary after 15 38

months when species richness was significantly lower in plots with the greatest cover of 39

oysters. Assemblage structure differed with oyster cover in mud-flats but not in mussel-beds, 40

except at 100% cover in one location and at one time. Within mud-flats at one location and 41

time (of four total tests), assemblages became more homogenous with increasing cover of 42

oysters leading to a significant decrease in β-diversity. These responses were primarily 43

underpinned by the facilitation of several taxa including a grazing gastropod (Littorina 44

littorea), an invasive barnacle (Austrominius modestus) and a primary producer (Fucus 45

vesiculosus) with increasing cover of oysters. Although there were consistent positive effects 46

of C. gigas on mud-flat biodiversity, effects were weak or negative at higher cover on 47

mussel-beds. This highlights the need for the impacts of invasive species to be investigated at 48

a range of invader abundances within different environmental contexts. 49

50

3

1 Introduction 51

Predicting the impacts of invasive species can be complicated by the fact that their effects 52

are often context-dependent, varying in nature and magnitude depending on the properties of 53

the receiving environment (Hewitt et al 2001; Agrawal et al 2007; Padilla 2010; Queirόs et al. 54

2011) and the abundance of the invader (Thomsen et al. 2011). There is increasing 55

recognition of the need to explore the effects of invasive species in different types of habitats 56

in order to understand mechanistic interactions between invaders and abiotic conditions 57

(Sousa et al. 2009; Thomsen et al. 2011). Also, as the abundance of an invasive species 58

increases, the effect it has on the physical, chemical and biological properties of the 59

environment may also change in nature or magnitude (Sousa et al. 2009) and although there 60

has been extensive research into the impacts of invasive species, few studies have compared 61

their impacts at different abundances (Thomsen et al. 2011). This is a significant shortcoming 62

in invasion biology because it precludes the ability to make generalisations about abundance-63

dependent impacts or to predict how impacts may change as the invasion proceeds 64

(Yokomizo et al. 2009; Thomsen et al. 2011). 65

Among the most influential invasive species are those which alter existing habitats and/or 66

create new habitats that persist in the environment for a long time. Such species have been 67

referred to as “ecosystem engineers” (Jones et al. 1994) or “foundation species” (Dayton 68

1975). Invasive ecosystem engineers can have far reaching impacts on the structure of 69

assemblages over space and time (Jones et al. 1994) through multiple direct and indirect 70

mechanisms (Crooks 2002). Many bivalves are ecosystem engineers, physically altering the 71

environment and affecting the availability of resources to other organisms either through their 72

physical structure (shell) or their biological activities (filter-feeding and biodeposition). Due 73

to their potential to alter the structure and functioning of receiving ecosystems, invasive 74

4

bivalves have been identified as a priority for conservation and management initiatives 75

(Sousa et al. 2009) as they can have severe cascading economic impacts (Aldridge et al. 76

2004). 77

C. gigas is a commercially important species (accounting for 98% of global commercial 78

oyster production) and, as a consequence of deliberate introduction, has become invasive 79

throughout much of the world (Ruesink et al. 2005). C. gigas is able to modify habitats in a 80

way that enhances its own further spread (Diederich 2005) and as such, it can form very 81

dense populations (over 500 individuals per m2; Wrange et al. 2010). C. gigas is an 82

ecosystem engineer which modifies the physical properties of the receiving environment and 83

provides a complex three dimensional habitat (Markert et al. 2010). 84

Invasive populations of C. gigas have been found to alter the abundance and distribution of 85

native species in several parts of the world including the United Kingdom, Canada, Australia, 86

New Zealand and Western Europe (Troost 2010). Effects have been documented in a variety 87

of different habitats, including mussel-beds ( Kochmann et al. 2008;), marshes (Escapa et al. 88

2004), seagrass beds (Wagner et al. 2012) and rocky platforms (Krassoi et al. 2008;) and 89

possible effects may occur in polychaete reefs (Dubois et al. 2006) and mud-flats (Trimble et 90

al. 2009). In some cases, C. gigas was associated with an increase in the abundance and 91

diversity of organisms (Dubois et al. 2006; Markert et al. 2010; Lejart and Hily 2011). In 92

other cases, diversity was unchanged, but the structure of assemblages associated with C. 93

gigas differed from those associated with indigenous habitats (Kochmann et al. 2008; 94

Markert et al. 2010). 95

Padilla (2010) predicted how the impacts of C. gigas may vary depending on the receiving 96

habitat. For example, in mussel-beds if the native bivalve provides many of the same 97

engineering functions as C. gigas, such as the facilitation of other species through the 98

5

provision of habitat or refugia, then the impacts of C. gigas on the local diversity are more 99

likely quantitative rather than qualitative. Alternatively, in mud-flats the presence of C. gigas 100

creates hard habitat for sessile algae and invertebrates that are typically not found in soft 101

sediments, while also increasing available habitat and refugia for mobile organisms, thereby 102

the changes to local diversity are likely to be qualitative and quantitative. 103

Therefore, the effects of C. gigas on biodiversity may differ as their abundance, cover or 104

density increases, and their effects may differ depending on the receiving habitat, but these 105

factors have not yet been explicitly examined together. The aims of this study were therefore 106

to determine 1) How C. gigas affects biodiversity, assemblage structure and the abundance of 107

individual taxa, 2) How these impacts vary with increasing cover of C. gigas and 3) whether 108

these impacts vary in different habitats. 109

110

2 Materials and methods 111

2.1 Study sites 112

This study was done at two locations in Ireland, Lough Foyle (Quigley’s Point, County 113

Donegal: 55° 7' 14.87", -7° 11' 53.59) and Lough Swilly (Ballylin Point, County Donegal, 114

Ireland: 55° 2' 36.12", -7° 33' 36.09"). At each location experimental plots were set up on two 115

different types of habitat, both in lower intertidal areas: mussel-beds and mud-flats. Mussel-116

bed habitats consisted of dense populations of blue mussels, Mytilus edulis. Mud-flat habitats 117

were patches of mud, ranging from 4 to 20 m2 in area, interspersed between the mussel-beds 118

which were not dominated by any other biogenic habitat building organism or by hard 119

substratum. Sediment within mussel-beds was a mixture of fine sand, silt with large shell 120

6

fragments throughout. Sediment within mud-flats was a mixture of very fine sand, silt and 121

clay. 122

2.2 Experimental set-up 123

The experiment was set-up during late April 2009 and had 2 factors: Habitat (fixed and 124

orthogonal with 2 levels, mussel-bed or mud-flat) and cover (fixed and orthogonal with 4 125

levels, 0, 5, 50 or 100 % cover of oysters). These four levels of cover were arranged into 50 x 126

50 cm plots in mussel-bed and muddy sediment habitat and equated to the following 127

densities: 0, 16 +/- 0.5, 120 +/- 8 and 240 +/- 12 individuals per m2 which equated to 0, 26.06 128

+/- 2.13, 390.88 +/- 31.95 and 781.77 +/- 63.90 g per m2 ash free dry weight. All oysters used 129

in this experiment were found in situ and were not moved between locations. Oysters were 130

rinsed with seawater and cleaned of any epibionts prior to use and then simply inserted 131

upright into the mud and mussel-bed habitats to mimic the positions in which they are found 132

in natural populations. Oysters were inserted to simulate an overgrowth, rather than a 133

complete replacement of mussels. Because sampling was destructive, 8 replicate plots of each 134

treatment were set up and 4 replicates were sampled at each of 2 times. Plots were spaced at 135

least 5 m apart and were spread across a section of shore that was approximately 25 m in 136

length and 100 m in width. 137

138

2.3 Sampling methods 139

The experiment was sampled after 4 and 15 months, during late July, 2009 and late June, 140

2010 respectively. To measure epifauna and flora, a 25 x 25 cm quadrat was randomly placed 141

onto each plot and a 2 minute search was conducted in which large mobile macrofauna, 142

including gastropods and crabs, were counted and percentage covers of algae were estimated 143

visually. After this, a core with 10 cm diameter was taken down to 10 cm depth in the centre 144

7

of the plot. The top section, which included epifauna and the first 1-2 cm of sediment, was 145

removed and placed into a container and preserved in 5% formalin. Any epifauna or flora on 146

the surface of the bivalves was later identified and counted and the organisms in the sediment 147

were sieved through a 125µm mesh, sorted and identified. Sessile epifauna including 148

bryozoans, barnacles and polychaetes were counted as individuals not as colonies. The 149

remainder of the core (from 2 to 10 cm depth) was sieved through a 500 µm mesh to retain 150

macrofauna, which were then stained with rose bengal and placed into a container topped up 151

with 5% formalin to preserve fauna for later sorting and identification in the laboratory. 152

Organisms in quadrats and cores were scaled up to numbers per m2. 153

154

2.4 Statistical analyses 155

Univariate analyses were done using 2-factor analysis of variance (ANOVA) using the design 156

described above. Total number of individuals (N), species richness (SR) and Shannon-Wiener 157

diversity (H’) were analysed. Differences in the abundance of taxa thought to be susceptible 158

to the impacts of oysters based on previous studies, or found as dominant space occupiers on 159

the shore were also analysed. These included common periwinkles, Littorina littorea, green 160

shore crabs, Carcinus maenas, invasive barnacles, Austrominius modestus (accidently 161

introduced to Ireland in the 1940's and now widespread (Allen et al. 2006)), and 162

Phyllodocidae polychaetes. Separate analyses were done for each variable at each sampling 163

time and within each location. 164

Variation in assemblage structure among treatments at each sampling time and within each 165

location was compared using two-factor PERMANOVA (Anderson 2001) based on the 166

design described above and using Bray-Curtis dissimilarities (Bray and Curtis 1957) of 167

square root transformed data with 9999 permutations under the reduced model. Assemblage 168

8

data were ordinated on a 2-dimensional non-metric multidimensional scaling (nMDS) 169

diagram, with the stress values representing the level of distortion of the actual rank order of 170

dissimilarity among samples (Clarke 1993). Where significant differences in assemblage 171

structure were found, SIMPER (Clarke 1993) analyses were computed on square root 172

transformed data and were used to identify the contribution of different taxa to dissimilarities 173

between treatments. Tests for homogeneity of multivariate dispersions (PERMDISP: 174

Anderson, 2004) were done to identify heterogenous variability among experimental groups 175

and also represented a measure of β diversity (Anderson et al, 2006). 176

Analyses of Variance were computed using the software Win-GMAV (Underwood and 177

Chapman, 1998). Prior to univariate ANOVAs, homogeneity of variance was tested using 178

Cochran’s C-test and when significant, data were square root or log transformed to make 179

variances homogenous. Transformations were consistently applied to separate analyses of the 180

same variable (done for each sampling time and location) to ensure that outcomes were 181

comparable among analyses. When significant differences were detected by ANOVA (P < 182

0.05), post-hoc Student-Newman Keuls (SNK) tests were used to identify patterns of 183

difference. All multivariate analyses were completed using the PRIMER package (PRIMER-184

e, 2009). 185

186

3 Results 187

A total of 61 taxa were found throughout the duration of the experiment (detailed in 188

Appendix 1). Not all taxa were present at both locations in both years, but in aggregate they 189

comprised 4 algal species, 9 gastropod species, 6 bivalve species, 4 amphipod species, 190

harpacticoid copepods, Janiridae isopods, cumaceans, oligochaetes, leeches, bryozoans, 191

thaliaceans, holothurians, chitons, sponges and 16 different families of polychaetes. At Lough 192

9

Foyle after 15 months several plots were missing or damaged so only 3 replicate plots for 193

each treatment could be used in the analyses. 194

195

3.1 Effects of C. gigas on the total number of individuals of non-oyster taxa 196

At each lough after 4 months, N was affected by the interaction between habitat and cover of 197

oysters (Table 1). N did not differ among covers in mussel-beds, but at each lough on mud-198

flats was greater in plots with 5, 50 or 100 % cover than in those with 0 % cover of oysters 199

(Table 1, Fig. 1a). 200

After 15 months N increased by an order of magnitude at Lough Foyle. The increase was 201

mostly attributable to an increase in bryozoans, Conopeum seurati. Regardless, at each lough, 202

N differed between habitats (Table 1), with greater N in mussel-beds than in mud-flats (Fig. 203

1a). In addition, at Lough Swilly N was also affected by the cover of oysters and was greater 204

within plots with 50 or 100 % cover than in those with 0 % cover of oysters (Table 1, Fig. 1a) 205

206

3.2 Effects of C. gigas on biodiversity 207

After 4 months at each lough SR and H' were affected by the interaction between habitat and 208

cover of oysters (Table 1) with no differences among covers within mussel-beds. Within 209

mud-flats, however, plots with 100 % cover of oysters had greater SR than the plots with 0 210

and 5 % cover and plots with 100 or 5 % cover had greater H' than plots with 0 % cover of 211

oysters at Lough Foyle (Fig. 1, Table 1). In mud-flats at Lough Swilly, SR and H’ were 212

greater in plots with 5, 50 and 100 % cover of oysters than in plots with 0 % cover of oysters 213

(Table 1, Fig. 1). 214

10

After 15 months, at Lough Foyle, SR was affected by the cover of oysters and was greater in 215

plots with 50 or 100 % cover than in those with 0 % cover of oysters (Table 1, Fig. 1). At 216

Lough Swilly, SR and H' were affected by the interaction between habitat and cover and 217

within mussel-beds, SR in plots with 0, 5 or 50 % cover of oysters was greater than in those 218

with 100 % cover of oysters and in mud-flats, plots with 50 and 100 % cover of oysters had 219

greater SR and H’ than plots with 0 and 5 % cover of oysters (Table 1, Fig. 1). 220

221

3.3 Effects of C. gigas on assemblage structure 222

At each time and within each lough, assemblage structure was affected by the interaction 223

between habitat and cover of oysters (Fig. 2). Assemblages within mussel-beds were only 224

affected by the cover of oysters at Lough Swilly after 15 months, when plots with 100 % 225

cover of oysters differed from those with no oysters (Fig. 2d). Differences in assemblage 226

structure were driven by greater densities of A. modestus and Tubificoides benedii in plots 227

with 100 % cover of oysters and reduced densities of harpacticoid copepods, Janiridae 228

isopods and polychaetes of the families Phyllodocidae, Polynoidae, Hesionidae and 229

Spionidae compared to plots with no oysters (Table 2). 230

Assemblages within mud-flats were, however, affected by the cover of oysters in both loughs 231

and at each time. After 4 months, at Lough Foyle, plots with 100 % cover of oysters differed 232

from those with 0, 5 and 50 % cover of oysters and plots with 50 % cover of oysters differed 233

from those with no oysters (Fig. 2a). The differences in assemblage structure in plots with 234

100 % compared to plots with 0, 5 or 50 % cover of oysters were driven by variations in the 235

densities of A. modestus, C. seurati and Spionidae polychaetes, which had greater densities in 236

plots with 100 % cover of oysters (Table 2). Similarly, differences in assemblage structure in 237

plots with 50 % cover from plots with no oysters, were also driven by these same taxa which 238

11

occurred in greater densities in plots with 50 % cover of oysters than in plots with no oysters. 239

At Lough Swilly after 4 months, assemblage structure in plots with 0 % cover of oysters 240

differed from those with 5, 50 and 100 % cover in mud-flats (Fig. 2b). Differences in 241

assemblage structure were mostly driven by greater densities of A. modestus, C. seurati, 242

harpacticoid copepods, juvenile Mytilus edulis and Spionidae polychaetes in plots with 5, 50 243

or 100 % than in plots with 0 % cover of oysters (Table 2). 244

After 15 months at Lough Foyle although there was a significant interaction between habitat 245

and cover of oysters, the post-hoc procedure did not detect any patterns of difference 246

(although due to the loss of replication, only 10 unique permutations were possible for 247

comparing levels of cover within habitats so statistical power was very limited). At Lough 248

Swilly after 15 months assemblage structure in plots in mud-flat habitats with 0 % cover of 249

oysters differed from those with 5, 50 or 100 % cover (Fig. 2d). Differences were largely 250

driven by greater densities of T. benedii, A. modestus, Polydora sp., Phyllodocidae 251

polychaetes and a greater cover of F. vesiculosus but a lower density of harpacticoid 252

copepods in plots with either 5, 50 or 100 % cover than plots with no oysters (Table 2). 253

254

3.3.1 Effects on β-diversity 255

β-diversity (multivariate dispersion indices) significantly varied among covers of oysters in 256

mud-flat habitats at Lough Swilly after 15 months, where plots with 100 % cover of oysters 257

were less dispersed than those with 0, 5 or 50 % cover (PERMDISP: F7,24 = 6.83, P = 0.006). 258

Homogeneity of multivariate dispersions did not differ significantly at any other occasion 259

(time, lough or habitat) throughout the experiment. 260

261

12

3.4 Effects of C. gigas on density or cover of individual taxa 262

After 4 months, the density of L. littorea was greater in mussel-beds than in mud-flats and at 263

Lough Swilly was also greater in plots with 100 % cover than in those with 0, 5 or 50 % 264

cover of oysters (Table 3, Fig. 3). After 15 months at Lough Foyle, the density of L. littorea 265

was greater in mussel-beds than in mud-flats and greater in plots with 100 % than with 0 or 5 266

% cover of oysters. At Lough Swilly, the density of L. littorea was greater in plots with 50 or 267

100 % cover of oysters than in those with 0 or 5 % cover in mudflats, but was not affected by 268

C. gigas in mussel-beds (Table 3, Fig. 3). 269

After 4 months, at each lough, the density of A. modestus was affected by the interaction 270

between habitat and cover but in both habitats barnacle numbers generally increased with 271

oyster cover (Table 3, Fig. 3). After 15 months at Lough Foyle the density of A. modestus 272

was greater in mussel-beds than in mud-flats and greater in plots with 50 or 100 % than in 273

those with 0 or 5 % cover of oysters (Table 3, Fig. 3). At Lough Swilly, the density of A. 274

modestus was greater in plots with 50 or 100 % than those with 0 or 5 % cover of oysters 275

(Table 3, Fig. 3). 276

After 4 months, at Lough Foyle the density of C. maenas was affected by the interaction 277

between habitat and cover with a greater density in plots with 100 % than in those with 0, 5, 278

or 50 % cover of oysters in mussel-beds, and a greater density in plots with 50 % cover of 279

oysters than in those with 0 or 5 % cover of oysters in mud-flats (Table 3, Fig. 3). At Lough 280

Swilly at each sampling time and at Lough Foyle after 15 months, the density of C. maenas 281

was greater in mussel-beds than in mud-flats (Table 3, Fig. 3). 282

After 4 months, at Lough Foyle, the density of Phyllodocidae polychaetes was greater in 283

mussel-beds than in mud-flats and, in both habitats was greater in plots with 100 % cover of 284

oysters than in plots with 0 or 5 % cover of oysters (Table 3, Fig. 3). After 4 months at Lough 285

13

Swilly the density of Phyllodocidae polychaetes was greater in plots with 50 % cover of 286

oysters than in plots with 0, 5 or 100 % cover of oysters (Table 3, Fig. 3). After 15 months at 287

Lough Swilly, the density of Phyllodocidae polychaetes was affected by the interaction 288

between habitat and cover and was greater in plots with 0 or 50 % cover of oysters than in 289

plots with 100 % cover of oysters in mussel-beds and greater in plots with 5, 50 or 100 % 290

cover than in those with 0 % cover of oysters in mud-flats (Table 3, Fig. 3). 291

292

4 Discussion 293

Crassostrea gigas affected biodiversity, assemblage structure and the abundance of several 294

organisms, and the nature of these effects depended both on the habitat and on the cover of C. 295

gigas. 296

Overall effects on biota were more consistent in mud-flat than in mussel-bed habitats. This is 297

not surprising since the addition of C. gigas to mud-flats provides hard substratum where it 298

was formerly rare, thus completely changing the habitat from soft unstructured mud-flat to 299

one with hard biogenic reef (Lejart and Hily 2011) with more habitat available for the 300

recruitment of sessile organisms and refugia from desiccation and heat stress for mobile 301

organisms (Padilla 2010). This is in contrast to mussel-beds, where C. gigas are overgrowing 302

an existing hard biogenic habitat formed by M. edulis, which provides similar engineering 303

functions; both species are suspension-feeders that change the characteristics of the sediment 304

and provide shelter for other organisms thereby enabling a different assemblage to develop 305

than would otherwise be present (Padilla 2010). Regardless, the formation of oyster reefs in 306

either habitat results in an increase of habitat complexity and heterogeneity (McCoy and Bell 307

1991), and although not always consistent, there were also strong effects of C. gigas in 308

mussel-beds. In mud-flats the total number of individuals, species richness and Shannon-309

14

Wiener diversity generally increased with increasing cover of C. gigas. The importance of 310

availability of refuges and structural complexity of habitat on intertidal assemblage structure 311

has long been recognised (Barshaw and Lavalli 1988; Gee and Warwick 1994) and the 312

increase in the total number of individuals with increasing cover of oysters may be a result of 313

reduced mortality or increased recruitment (Crooks and Khim 1999; Grabowski 2004). 314

Effects on assemblage structure in mud-flats were also consistent across locations and time, 315

with plots with the greatest cover differing from those with less or no cover of C. gigas. 316

These changes were mostly underpinned by increases of organisms which require a hard 317

surface for colonisation, such as barnacles, bryozoans and fucoid algae, with the oysters’ 318

shell allowing these organisms to exist in a habitat where they previously could not. 319

Assemblage structure in mussel-beds did not differ at Lough Foyle, (perhaps due to reduced 320

statistical power due to the loss of replicates) but at Lough Swilly after 15 months, plots with 321

the greatest cover differed from those with no cover of C. gigas. Since C. gigas is bigger, has 322

overlapping ridges and layers on its shell and is more complex in structure than M. edulis 323

(Lejart and Hily 2011), in both habitats differences in assemblage structure could be due to an 324

increase in surface area for settlement or habitat complexity. 325

Because C. gigas enhances habitat heterogeneity with its shell and forms complex reefs, it is 326

expected to increase biodiversity (McCoy and Bell 1991). Previous studies have confirmed 327

this hypothesis, finding C. gigas to increase biodiversity and the abundance of taxa compared 328

to adjacent habitats (Dubois et al. 2006; Troost 2010; Lejart and Hily 2011; Ferraro and Cole 329

2011). Several studies have found increases in the number of individuals and species richness 330

associated with C. gigas on mud-flats, similar to results of the current study (Hosack 2003; 331

Lejart and Hily 2011). Previous studies of mussel-beds similar to those in the present study, 332

found that C. gigas reefs had greater biodiversity than adjacent mussel-beds (Markert et al. 333

2010) or a greater number of organisms (Kochmann et al. 2008). The current study is, 334

15

however, the first to find a decrease in macro- invertebrate and algal diversity associated with 335

C. gigas, where species richness decreased at the greatest cover of C. gigas on mussel-beds at 336

Lough Swilly after 15 months. Although not statistically significant, this trend was visible for 337

both diversity metrics (species richness and Shannon-Wiener) in both years at Lough Swilly. 338

Also at Lough Swilly after 15 months, β-diversity (assessed using multivariate dispersion 339

indices) was reduced with the greatest cover of C. gigas in mud-flat habitats. Tests for 340

homogeneity of multivariate dispersions have been identified as a way of measuring β-341

diversity (Anderson et al. 2006). The exact meaning of this measure, however, is largely 342

dependent on the dissimilarity index and any transformations used and therefore should be 343

carefully interpreted (Anderson et al. 2011). β-diversity, as measured here, based on Bray-344

Curtis, square-root transformed data, captures variation in assemblage structure with 345

emphasis on both species composition and relative abundance (Anderson et al. 2011). 346

Therefore, the reduction in β-diversity with increasing cover of C. gigas means that 347

assemblages were more homogeneous in terms of their composition (the types of taxa 348

present) and the relative abundance of these taxa. The potential for homogenisation (Olden 349

and Rooney 2006), in terms of decreased β-diversity (Olden et al. 2008; McKinney and 350

Lockwood 1999) is an important, but often overlooked, consequence of invasive species 351

(Wright 2011).The issue of homogenisation may become magnified if invasive species also 352

cause homogenisation of the habitat (Rahel 2002). 353

Experimental manipulations, such as those in the current study, are a more powerful method 354

than purely mensurative studies of investigating the impacts of invasive species because they 355

allow inferences concerning the cause and effect of ecological processes to be assigned 356

(Underwood 1997). However, in the current study it was necessary for the experimental plots 357

to be relatively small, only 25cm2 (to allow for replication and manipulation), and scaling up 358

from the scale of centimetres to that of an estuary (kilometres) can be very uncertain and has 359

16

long been recognised as a challenging and pervasive issue in ecology (Levin 1992). C. gigas 360

is, however, an ecosystem engineer which can dominate large areas (for example, a 361

continuous reef covering 26.5 ha was reported in The Netherlands (Fey et al. 2010)) thereby 362

causing extensive habitat homogenisation. 363

It is not uncommon for invasive species to have “positive” effects, such as increasing the 364

diversity and abundance of native organisms (Simberloff and Von Holle 1999; Altieri et al. 365

2010). In fact, some previous studies have found positive effects to increase with invader 366

abundance (Thomsen 2010), but it is possible that there is a threshold after a critical invader 367

abundance is reached whereupon the positive effects on an invasive species become negative, 368

or the negative effects increase substantially (White and Shurin, 2011). Despite this, very few 369

studies have explicitly tested the effects of invasive species at a range of abundances 370

(Thomsen et al. 2011), but this is paramount if we are to avoid the detrimental consequences 371

of under-estimating their effects at greater abundances (Yokomizo et al. 2009). For example 372

in the current study, at lower cover, C. gigas might be interpreted as having positive effects 373

by increasing biodiversity. At greater cover, however, there was evidence of a threshold 374

being reached, and in mussel-beds, the effects on biodiversity became negative. 375

Whilst the effects of any invasive species on biodiversity can be difficult to predict, some 376

single species, such as L. littorea and A. modestus, exhibited very consistent responses, 377

increasing with increasing cover of C. gigas within both habitats, locations and at both times. 378

Similar results have been found by others for both of these species (Kochmann et al. 2008; 379

Markert et al. 2010) and are probably due to an increase in habitat complexity provided by 380

oyster shells providing more refuges from desiccation or predation, or in the case of the 381

periwinkles, an increase in available microalgae for grazing. Although C. maenas initially 382

increased at Lough Foyle with increasing cover of oysters, it was mostly affected by the type 383

of habitat, with more crabs found in mussel-beds than on mud-flats, probably due to greater 384

17

source of food from the mussels themselves (Mascaro and Seed, 2001). This was in contrast 385

to past studies which have found C. maenas to increase in association with C. gigas 386

(Kochmann et al. 2008; Markert et al. 2010). The density of Phyllodocidae polychaetes at 387

Lough Swilly increased with increasing cover of C. gigas on mud-flats, but decreased with 388

the greatest cover of C. gigas in mussel-beds. In addition, Polynoidae, Hesionidae and 389

Spionidae polychaetes were also less abundant in plots with the greatest cover of oysters 390

compared to control plots with no oysters. 391

Declines in polychaetes and biodiversity at the greatest cover of oysters in mussel-beds may 392

have been caused by an increase in the concentration of nutrients in the pore-water, such as 393

ammonium, resulting either directly from oyster excretion (Dame et al. 1984, 1985) or 394

indirectly due to the breakdown of organic matter in oyster “biodeposits” which may be 395

composed of faeces or pseudofaeces. In a concurrent study, abiotic measurements were also 396

taken from the experimental plots at Lough Swilly after 15 months, and the concentration of 397

porewater ammonium was indeed greater beneath a greater cover of C. gigas (Green et al. 398

2013). Oysters produce large quantities of biodeposits (Hayakawa et al. 2001) which 399

organically enrich sediments and supply food for other organisms (Castel et al. 1989). But the 400

decomposition of biodeposits increases the uptake of oxygen in the sediment (Christensen et 401

al. 2003) and if the rate of biodeposition is great, this can lead to sediment anoxia, thus 402

making the habitat unsuitable, and decreasing the diversity and abundance of infaunal 403

organisms (Callier et al. 2009). At Lough Swilly, community respiration (CO2 flux) from 404

mussel-beds and mud-flats was greatest from plots with 100 % cover of C. gigas (Green et al. 405

2013), most likely indicating an increase in microbially mediated decomposition rates (Green 406

et al. 2012). 407

Alternatively, some of the differences in assemblage structure or biodiversity could also be 408

due to structural effects, such as competition for space or alterations to hydrodynamics, 409

18

caused by the oysters’ shells (Moulin et al. 2007). Of course, in order to understand the 410

mechanisms underpinning these changes, it would be necessary to separate the effects of the 411

physical structure and the biological activities of C. gigas. This was not explored in the 412

current study, but Wagner et al (2012) found that the shells of C. gigas can decrease eelgrass 413

shoot density and size and Green and Crowe (2013) found that the structure of C. gigas was 414

at least as important as their biological activity in determining patterns of difference in the 415

abundance and diversity of organisms in boulder-fields. 416

It is also possible that indirect effects of organisms that were facilitated by increasing cover 417

of C. gigas may have excluded the establishment of others, thereby decreasing biodiversity. 418

For example, some species which consistently increased in density with increasing cover of 419

C. gigas, at one or both loughs, are known to be able to reduce the establishment of other 420

invertebrates and algae, for example, A. modestus through competition for space and food 421

(Little et al. 1992), L. littorea via bull-dozing effects from grazing (Buschbaum 2000) and F. 422

vesiculosus via mechanical abrasion caused by the wave-induced movement of fronds (Grant 423

1977; Kiirikki 1996). The ecological importance of facilitation of native organisms by 424

invaders, and the cascading indirect effects on biodiversity, have recently been highlighted 425

(White et al. 2006; Rodriguez 2006; Pope et al. 2008) and warrant further investigation. 426

Estuarine habitats such as those examined here are inherently patchy (Morrisey et al. 1992) 427

characterised by a high level of spatial and temporal variability not only in the abundance and 428

distribution of invertebrates (Underwood and Chapman 1996) but also geochemical 429

properties of sediments (Tolhurst and Chapman 2007). The current study found that some 430

impacts of C. gigas on individual taxa were consistent across habitats, but the majority were 431

context dependent varying in direction and magnitude with regards to the type of habitat 432

being invaded and on the cover of C. gigas. Generally in mud-flats the number of individuals 433

and their diversity increased at 5 and 50% cover of C. gigas and then plateaued, with no 434

19

additional increase at 100% cover. In mussel-beds species richness was reduced at 100% 435

cover at Lough Swilly after 15 months. If the objective of management is to maintain or 436

enhance biodiversity, then results from this study suggest that action should be taken at an 437

early stage to restrict the spread of C. gigas before dense reefs (exceeding 50% cover) are 438

formed. At present in Ireland, wild populations of intertidal C. gigas are being harvested by 439

hand (F. O’Beirn, pers. comm.). This may be a cost effective solution to help control their 440

population size. 441

Typically studies on the impacts of invasive species have compared the presence or absence 442

of invaders (Thomsen et al. 2011). In order to improve predictions of their impacts, it is 443

advantageous to relate invader effects directly to their abundance, which may vary through 444

time, as populations expand or decline, or differ spatially through patchy distribution. 445

Although impacts may form continuous gradients with abundance (Parker et al. 1999), it is 446

more common for the relationship between effect and invader abundance to be non-linear 447

(Yokomizo et al. 2009). If responses are non-linear, then small abundances of an invader may 448

have different or disproportionate effects on native ecosystems than larger abundances. For 449

instance, if small populations of an invader have positive or neutral impacts on native species, 450

then a threshold value may exist, above which effects become negative (Groffman et al., 451

2006). The current study suggests that this may be the case for C. gigas and confirms the 452

recommendations of others (Sousa et al. 2009; Padilla 2010; Thomsen et al. 2011) that more 453

experimental studies spanning a range of habitats and a range of invader abundances are 454

needed in order to assess the context-dependency of invasive species and to aid in the 455

development of cost-effective mitigation strategies. 456

457

458

20

Acknowledgements 459

This research was funded by the project SIMBIOSYS (2007-B-CD-1-S1) as part of the 460

Science, Technology, Research and Innovation for the Environment (STRIVE) Programme, 461

financed by the Irish Government under the National Development Plan 2007–2013, 462

administered on behalf of the Department of the Environment, Heritage and Local 463

Government by the Irish Environmental Protection Agency (EPA). We would also like to 464

thank Bas Boots, the editor Jennifer Ruesink and two anonymous reviewers for helpful 465

comments that helped to improve this manuscript. 466

467

References 468

Agrawal AA, Ackerly DD, Adler F, Arnold E, Cáceres C, Doak DF, Post E, Hudson PJ, 469

Maron J, Mooney KA, Power M, Schemske D, Stachowicz J, Strauss S, Turner MG, Werner 470

E (2007) Filling key gaps in population and community ecology. Frontiers in Ecology and the 471

Environment 5: 145–152 472

Aldridge DC, Elliott P, Moggridge GD (2004) The recent and rapid spread of the zebra 473

mussel (Dreissena polymorpha) in Great Britain. Biological Conservation 119:253-261 474

Allen BM, Power AM, O’Riordan RM, Myers AA, McGrath D (2006) Increases in the 475

abundance of the invasive barnacle Austrominius modestus Darwin in Ireland. Biology and 476

Environment: Proceedings of the Royal Irish Academy 106: 155-161 477

Altieri AH, van Wesenbeeck BK, Bertness MD, Silliman BR (2010) Facilitation cascade 478

drives positive relationship between native biodiversity and invasion success. Ecology 479

91:1269-1275 480

Anderson MJ (2001) A new method for non-parametric multivariate analysis of variance. 481

Austral Ecology 26:32-46 482

Anderson MJ (2006) Distance-based tests for homogeneity of multivariate dispersions. 483

Biometrics 62:245-253 484

Anderson MJ, Crist TO, Chase JM, Vellend M, Inouye BD, Freestone AL, Sanders NJ, 485

Cornell HV, Comita LS, Davies KF, Harrison SP, Kraft NJB, Stegen JC, Swenson NG 486

(2011) Navigating the multiple meanings of beta diversity: a roadmap for the practicing 487

ecologist. Ecology Letters 14:19-28 488

21

Anderson MJ, Ellingsen KE, McArdle BH (2006) Multivariate dispersion as a measure of 489

beta diversity. Ecology Letters 9:683-693 490

Barshaw DE, Lavalli KL (1988) Predation upon postlarval lobsters Homarus americanus by 491

Cunners Tautogolabrus adspersus and mud crabs Neopanope sayi on 3 different substrates – 492

eelgrass, mud and rocks. Marine Ecology-Progress Series 48:119-123 493

Bray JR, Curtis JT (1957) An ordination of the upland forest communities of Southern 494

Wisconsin. Ecological Monographs 27:326-349 495

Buschbaum C (2000) Direct and indirect effects of Littorina littorea (L.) on barnacles 496

growing on mussel-beds in the Wadden Sea. Hydrobiologia 440:119-128 497

Buttger H, Asmus H, Asmus R, Buschbaum C, Dittmann S, Nehls G (2008) Community 498

dynamics of intertidal soft-bottom mussel-beds over two decades. Helgoland Marine 499

Research 62:23-36 500

Callier MD, Richard M, McKindsey CW, Archambault P, Desrosiers G (2009) Responses of 501

benthic macrofauna and biogeochemical fluxes to various levels of mussel biodeposition: An 502

in situ "benthocosm" experiment. Marine Pollution Bulletin 58:1544-1553 503

Castel J, Labourg PJ, Escaravage V, Auby I, Garcia ME (1989) Influence of seagrass beds 504

and oyster parks on the abundance and biomass patterns of meiobenthos and macrobenthos in 505

tidal flats. Estuarine Coastal and Shelf Science 28:71-85 506

Christensen PB, Glud RN, Dalsgaard T, Gillespie P (2003) Impacts of longline mussel 507

farming on oxygen and nitrogen dynamics and biological communities of coastal sediments. 508

Aquaculture 218:567-588 509

Clarke KR (1993) Non-parametric multivariate analyses of changes in community structure. 510

Australian Journal of Ecology 18:117-143 511

Crooks JA (2002) Characterizing ecosystem-level consequences of biological invasions: the 512

role of ecosystem engineers. Oikos 97:153-166 513

Crooks JA, Khim HS (1999)Architectural vs. biological effects of a habitat-altering, exotic 514

mussel, Musculista senhousia. Journal of Experimental Marine Biology and Ecology 240:53-515

75 516

Dame RF, Wolaver TG, Libes SM (1985) The summer uptake and release of nitrogen by an 517

intertidal oyster reef. Netherlands Journal of Sea Research 19:265-268 518

Dame RF, Zingmark RG, Haskin E (1984) Oyster reefs as processors of estuarine materials. 519

Journal of Experimental Marine Biology and Ecology 83:239-247 520

Dayton PK (1975) Experimental evaluation of ecological dominance in a rocky intertidal 521

algal community. Ecological Monographs 45:137-159 522

22

Diederich S (2005) Differential recruitment of introduced Pacific oysters and native mussels 523

at the North Sea coast: coexistence possible? Journal of Sea Research 53:269-281 524

Dubois S, Commito JA, Olivier F, Retiere C (2006) Effects of epibionts on Sabellaria 525

alveolata (L.) biogenic reefs and their associated fauna in the Bay of Mont Saint-Michel. 526

Estuarine Coastal and Shelf Science 68:635-646 527

Escapa M, Isacch JP, Daleo P, Alberti J, Iribarne O, Borges M, Dos Santos EP, Gagliardini 528

DA, Lasta M (2004) The distribution and ecological effects of the introduced Pacific oyster 529

Crassostrea gigas (Thunberg, 1793) in northern Patagonia. Journal of Shellfish Research 530

23:765-772 531

Fey F, Dankers N, Steenbergen J, Goudswaard K (2010). Development and distribution of 532

the non-indigenous Pacific oyster (Crassostrea gigas) in the Dutch Wadden Sea. Aquaculture 533

International 18:45-59 534

Ferraro SP, Cole FA (2011) Ecological periodic tables for benthic macrofaunal usage of 535

estuarine habitats in the US Pacific Northwest. Estuarine Coastal and Shelf Science 94:36-47 536

Gee JM, Warwick RM (1994) Metazoan community structure in relation to the fractal 537

dimensions of marine macroalgae. Marine Ecology-Progress Series 103:141-150 538

Grabowski JH (2004) Habitat complexity disrupts predator-prey interactions but not the 539

tropic cascade on oyster reefs. Ecology 85:995-1004 540

Grant WS (1977) High intertidal community organization on a rocky headland in Maine, 541

USA. Marine Biology 44:15-25 542

Green DS, Boots B, Crowe TP (2012) Effects of non-indigenous oysters on microbial 543

diversity and ecosystem functioning. PLoS ONE 7(10): e48410. 544

doi:10.1371/journal.pone.0048410 545

Green DS, Rocha C, Crowe TP (2013) Effects of non-indigenous oysters on ecosystem 546

processes vary with abundance and context. Ecosystems: DOI: 10.1007/s10021-013-9659-y 547

Green DS, Crowe TP (2013) Physical and biological effects of introduced oysters on 548

biodiversity in an intertidal boulder field. Marine Ecology Progress Series 482:119-132 549

Gutierrez JL, Jones CG, Strayer DL, Iribarne OO (2003) Mollusks as ecosystem engineers: 550

the role of shell production in aquatic habitats. Oikos 101:79-90 551

Hayakawa Y, Kobayashi M, Izawa M (2001) Sedimentation flux from mariculture of oyster 552

(Crassostrea gigas) in Ofunato estuary, Japan. Ices Journal of Marine Science 58:435-444 553

Hosack G (2003) Effects of Zostera marina and Crassostrea gigas culture on the intertidal 554

communities of Willapa. Bay, Washington. Msc Thesis, University of Washington 555

Jones CG, Lawton JH, Shachak M (1994) Organisms as ecosystem engineers. Oikos 69:373-556

386 557

23

Kater BJ, Baars JMDD (2003) Reconstructie van oppervlakten van litorale Japanse 558

oesterbanken in de Oosterschelde in het verleden en een schatting van het huidig oppervlak. 559

(RIVO rapport C017/03) Nederlands Instituut voor Visserij Onderzoek (RIVO), Wageningen, 560

The Netherlands 561

Kiirikki M (1996) Experimental evidence that Fucus vesiculosus (Phaeophyta) controls 562

filamentous algae by means of the whiplash effect. European Journal of Phycology 31:61-66 563

Kochmann J, Buschbaum C, Volkenborn N, Reise K (2008) Shift from native mussels to 564

alien oysters: Differential effects of ecosystem engineers. Journal of Experimental Marine 565

Biology and Ecology 364:1-10 566

Krassoi FR, Brown KR, Bishop MJ, Kelaher BP, Summerhayes S (2008) Condition-specific 567

competition allows coexistence of competitively superior exotic oysters with native oysters. 568

Journal of Animal Ecology 77:5-15 569

Lang AC, Buschbaum C (2010) Facilitative effects of introduced Pacific oysters on native 570

macroalgae are limited by a secondary invader, the seaweed Sargassum muticum. Journal of 571

Sea Research 63 119-128 572

Lejart M, Hily C (2011) Differential response of benthic macrofauna to the formation of 573

novel oyster reefs (Crassostrea gigas, Thunberg) on soft and rocky substrate in the intertidal 574

of the Bay of Brest, France. Journal of Sea Research 65:84-93 575

Levin SA (1992) The problem of pattern and scale in ecology. Ecology 73: 1943-1967 576

Little C, Morritt D, and Stirling P (1992) Changes in the shore fauna and flora of Lough 577

Hyne. The Irish Naturalists’ Journal 24: 87-95 578

Markert A, Wehrmann A, Kroncke I (2010) Recently established Crassostrea-reefs versus 579

native Mytilus-beds: differences in ecosystem engineering affects the macrofaunal 580

communities (Wadden Sea of Lower Saxony, southern German Bight). Biol. Invasions 12:15-581

32 582

Mascaro M, Seed R (2001) Choice of prey size and species in Carcinus maenas (L.) feeding 583

on four bivalves of contrasting shell morphology. Hydrobiologia 449: 159-170 584

McCoy ED, Bell SS (1991)Habitat structure: the evolution and diversification of a complex 585

topic. In: Bell SS, McCoy ED, Mushinsky HR (eds). Habitat Structure: The Physical 586

Arrangement of Objects in Space, Chapman and Hall, New York, pp. 3–27 587

McKinney ML, Lockwood JL (1999) Biotic homogenization: a few winners replacing many 588

losers in the next mass extinction. Trends in Ecology & Evolution 14:450-453 589

Morrisey DJ, Howitt L, Underwood AJ, Stark JS (1992) Spatial Variation in Soft-Sediment 590

Benthos. Marine Ecology Progress Series 81:197-204 591

24

Moulin FY, Guizien K, Thouzeau G, Chapalain G, Mulleners K, Bourg C (2007) Impact of 592

an invasive species, Crepidula fornicata, on the hydrodynamics and transport properties of 593

the benthic boundary layer. Aquatic Living Resources 20:15-31 594

Muschenheim DK (1987) The dynamics of near-bed seston flux and suspension-feeding 595

benthos. Journal of Marine Research 45:473-496 596

Olden JD, Kennard MJ, Pusey BJ (2008) Species invasions and the changing biogeography 597

of Australian freshwater fishes. Global Ecology and Biogeography 17:25-37 598

Olden JD, Rooney TP (2006) On defining and quantifying biotic homogenization. Global 599

Ecology and Biogeography 15:113-120 600

Padilla DK (2010) Context-dependent Impacts of a Non-native Ecosystem Engineer, the 601

Pacific Oyster Crassostrea gigas. Integrative and Comparative Biology 50:213-225 602

Peterson CH, Grabowski JH, Powers SP (2003) Estimated enhancement of fish production 603

resulting from restoring oyster reef habitat: quantitative valuation. Marine Ecology Progress 604

Series 264:249-264 605

Pope KL, Garwood JM, Welsh HH, Lawler SP (2008) Evidence of indirect impacts of 606

introduced trout on native amphibians via facilitation of a shared predator. Biological 607

Conservation 141:1321-1331 608

Queiros AD, Hiddink JG, Johnson G, Cabral HN, Kaiser MJ (2011) Context dependence of 609

marine ecosystem engineer invasion impacts on benthic ecosystem functioning. Biological 610

Invasions 13:1059-1075 611

Rahel FJ (2002) Homogenization of freshwater faunas. Annual Review of Ecology and 612

Systematics 33:291-315 613

Reise K, Olenin S, Thieltges DW (2006) Are aliens threatening European aquatic coastal 614

ecosystems? Helgoland Marine Research 60:77-83 615

Rodriguez LF (2006) Can invasive species facilitate native species? Evidence of how, when, 616

and why these impacts occur. Biological Invasions 8:927-939 617

Ruesink JL (2007) Biotic resistance and facilitation of a non-native oyster on rocky shores. 618

Marine Ecology Progress Series 331:1-9 619

Ruesink JL, Lenihan HS, Trimble AC, Heiman KW, Micheli F, Byers JE, Kay MC (2005) 620

Introduction of non-native oysters: Ecosystem effects and restoration implications. Annual 621

Review of Ecology Evolution and Systematics 36:643-689 622

Simberloff D, Von Holle B (1999) Positive interactions of nonindigenous species: Invasional 623

meltdown? Biological Invasions 1:21-32 624

Sousa R, Gutierrez JL, Aldridge DC (2009) Non-indigenous invasive bivalves as ecosystem 625

engineers. Biological Invasions 11:2367-2385 626

25

Thomsen MS (2010) Experimental evidence for positive effects of invasive seaweed on 627

native invertebrates via habitat-formation in a seagrass bed. Aquatic Invasions 5:341-346 628

Thomsen MS, Wernberg T, Olden JD, Griffin JN, Silliman BR (2011) A framework to study 629

the context-dependent impacts of marine invasions. Journal of Experimental Marine Biology 630

and Ecology 400:322-327 631

Tolhurst TJ, Chapman MG (2007) Patterns in biogeochemical properties of sediments and 632

benthic animals among different habitats in mangrove forests. Austral Ecology 32:775-788 633

Trimble AC, Ruesink JL, Dumbauld BR (2009) Factors preventing the recovery of a 634

historically overexploited shellfish species, Ostrea lurida Carpenter 1864. Journal of 635

Shellfish Research 28:97-106 636

Troost K (2010) Causes and effects of a highly successful marine invasion: Case-study of the 637

introduced Pacific oyster Crassostrea gigas in continental NW European estuaries. Journal of 638

Sea Research 64:145-165 639

Underwood AJ, Chapman MG (1996) Scales of spatial patterns of distribution of intertidal 640

invertebrates. Oecologia 107:212-224 641

Underwood AJ, Chapman MG (1998) Spatial analyses of intertidal assemblages on sheltered 642

rocky shores. Australian Journal of Ecology 23:138-157 643

Wagner EL, Ruesink JL, Dumbauld BR, Hacker SD, Wisehart LM (2012) Density-dependent 644

effects of an introduced oyster, Crassostrea gigas, on native intertidal eelgrass, Zostera 645

marina. Mar Ecol Prog Ser 468:149-160 646

White EM, Wilson JC, Clarke AR (2006) Biotic indirect effects: a neglected concept in 647

invasion biology. Diversity and Distributions 12: 443-455 648

White LF, Shurin JB (2011) Density dependent effects of an exotic marine macroalga on 649

native community diversity. Journal of Experimental Marine Biology and Ecology 405: 111-650

119 651

Wrange AL, Valero J, Harkestad LS, Strand O, Lindegarth S, Christensen HT, Dolmer P, 652

Kristensen PS, Mortensen S (2010) Massive settlements of the Pacific oyster, Crassostrea 653

gigas, in Scandinavia. Biological Invasions 12:1145-1152 654

Wright S (2011) Invasive Species and the Loss of Beta Diversity. Ethics and the Environment 655

16:75-97 656

Yokomizo H, Possingham HP, Thomas MB, Buckley YM (2009) Managing the impact of 657

invasive species: the value of knowing the density-impact curve. Ecological Applications 658

19:376-386 659

26

Tables and Figures 660

Table 1. ANOVA for the total number of individuals (N), species richness (number of taxa) and Shannon-Weiner (H') diversity on plots with 661

increasing cover of oysters on mussel-beds and mud-flats after 4 and 15 months. N was ln(x+1) transformed to correct heterogeneity of variance. 662

Significant P values are indicated in italics. Post-hoc (SNK) results are detailed below each analysis. 663

Source of variation 4 mths: L. Foyle L. Swilly 15 mths: L. Foyle L. Swilly

Number of individuals d.f. MS F P d.f. MS F P d.f. MS F P d.f. MS F P

Habitat (=H) 1 5.35 23.51 0.0001 1 3.89 16.46 0.0005 1 11.56 7.68 0.0136 1 1.40 7.72 0.0104

Cover of oysters (=C) 3 1.46 6.42 0.0024 3 0.63 2.68 0.0698 3 1.15 0.8 0.5137 3 0.62 3.44 0.0329

H x C 3 0.84 3.68 0.0260 3 0.97 4.08 0.0178 3 1.15 0.79 0.5148 3 0.28 1.55 0.2264

Residual 24 0.23

24 0.24

16 1.45

24 0.18

SNK (H)

Mus > Mud

Mus > Mud SNK (C)

100 = 50 > 0

SNK (H x C)

Mussel-bed NS

NS

Mud-flat 100 =50 = 5 > 0

100 = 50 = 5 > 0

Species richness

Habitat (=H) 1 101.53 26.13 0.0000 1 26.28 4.81 0.0383 1 7.04 1.28 0.2745 1 81.28 8.64 0.0072


H x C 3 14.61 3.79 0.0241 3 22.03 4.03 0.0187 3 11.15 2.03 0.1506 3 126.53 13.45 0.0000

Residual 24 3.88

24 5.47

16 5.5

24 9.41

SNK (C)

100 = 50 > 0 SNK (H x C)

Mussel-bed NS

NS

0 = 5 = 50 > 100 Mud-flat 100 > 0 = 5

100 = 50 = 5 > 0

100 = 50 > 5 = 0

Shannon-Weiner (H’)

Habitat (=H) 1 0.78 10.21 0.0039 1 0 0.03 0.8727 1 0.03 0.08 0.7854 1 0 0 0.9761


H x C 3 0.28 3.73 0.0249 3 0.5 4.75 0.0097 3 0.07 0.18 0.9107 3 0.71 4.67 0.0104

Residual 24 0.08

24 0.1

16 0.4

24 0.15

SNK (C)

SNK (H x C)

Mussel-bed NS

NS

NS Mud-flat 100 > 0, 5 > 0 100 = 50 = 5 > 0 100 = 50 > 5 = 0

27

Table 2. SIMPER analyses based on square-root transformed data corresponding to 664

significant PERMANOVA results (Table 2). Listed are the 5 taxa which contributed most to 665

the dissimilarity between pair-wise comparisons of different covers of oysters. Av. 666

abundance: average abundance at each density of oysters indicated (square root transformed); 667

Av. diss: average dissimilarity among pairs of samples in terms of the species in question; 668

Diss/SD: a measure of variation in the contribution of the species to dissimilarities between 669

pairs of samples; Contrib%: percentage contribution of the species to the average overall 670

dissimilarity between groups of treatments. 671

672 Habitat Lough Time Taxon Av. abundance Av. diss Diss/SD Contrib%

Mussel-bed Swilly 15 mo 0% 100%

A. modestus 80.08 19.61 8.57 2.14 18.26

T. benedii 66.53 58.82 3.58 1.57 7.63

Harpacticoida 9.77 21.72 2.38 1.28 5.08

Janiridae 0.00 16.44 2.32 3.65 4.94

Phyllodocidae 3.99 17.44 2.19 1.44 4.66

Mud-flat Foyle 4 mo 0% 50%

C. seurati 0.00 125.95 23.5 5.05 41.68

Spionidae 15.62 49.86 6.88 1.73 12.2

A. modestus 0.00 24.63 4.13 0.72 7.33

T. benedii 3.99 22.48 3.57 1.96 6.34

Nematoda 80.11 95.86 3.19 1.59 5.65

0% 100%

C. seurati 0.00 128.62 19.54 4.01 28.06

A. modestus 0.00 121.87 18.53 6.85 26.60

Spionidae 15.62 66.42 7.59 2.00 10.89

Nematoda 80.11 61.73 4.97 0.81 7.14

T. benedii 3.99 23.42 3.15 1.88 4.52

5% 100%

C. seurati 45.25 128.62 12.04 1.78 28.5

A. modestus 52.14 121.87 8.68 1.89 20.56

Nematoda 87.82 61.73 4.61 0.88 10.92

Spionidae 46.67 66.42 3.84 1.62 9.09

T. benedii 13.76 23.42 1.68 1.25 3.99

50% 100%

A. modestus 24.63 121.87 11.3 2.19 32.5

C. seurati 125.95 128.62 4.74 1.43 13.62

Nematoda 95.86 61.73 4.73 0.96 13.61

Spionidae 49.86 66.42 3.18 1.33 9.14

Harpacticoida 14.66 17.34 1.88 1.32 5.4

Swilly 4 mo 0% 5%

A. modestus 0.00 71.57 10.39 3.76 19.93

C. seurati 0.00 41.08 5.23 0.73 10.03

Nematoda 101.51 103.51 3.76 1.26 7.21

Spionidae 14.01 36.48 3.34 1.67 6.41

L. cinerea 0.00 22.39 3.22 7.68 6.17

0% 50%

C. seurati 0.00 136.27 18.16 3.82 31.23

A. modestus 0.00 62.39 8.31 4.93 14.29

Spionidae 14.01 40.28 3.49 1.38 6.01

Nematoda 101.51 88.65 3.33 2.87 5.73

28

Harpacticoida 31.04 53.13 3.22 1.16 5.54

0% 100%

A. modestus 0.00 110.53 15.37 7.58 25.25

C. seurati 0.00 75.1 10.31 1.49 16.94

Nematoda 101.51 79.0 6.48 1.08 10.65

Harpacticoida 31.04 58.35 3.86 1.85 6.34

M. edulis (juv) 0.00 22.75 3.18 2.75 5.22

Swilly 15 mo 0% 5%

Harpacticoida 57.45 33.68 8.01 1.3 11.43

Polydora sp. 39.87 6.31 7.37 1.3 10.51

A. modestus 0.00 26.46 6.13 0.83 8.74

T. benedii 0.00 24.41 5.01 1.35 7.14

Oligochaete 25.26 5.64 4.74 1.76 6.76

0% 50%

Harpacticoida 57.45 40.43 6.54 1.16 9.41

A. modestus 0.00 31.14 5.29 1.11 7.62

T. benedii 0.00 28.54 5.2 1.53 7.49

Polydora sp. 39.87 12.45 4.85 1.26 6.98

Phyllodocidae 2.82 29.46 4.37 1.78 6.29

0% 100%

A. modestus 0.00 72.93 10.57 5.34 17.1

Phyllodocidae 2.82 41.41 5.6 4.58 9.06

T. benedii 0.00 32.41 4.81 1.46 7.77

Harpacticoida 57.45 53.84 4.02 1.75 6.5

L. littorea 2.24 25.36 2.24 3.64 5.4

673

29

Table 3. Results of analyses of densities (No. per m2) of L. littorea, E. modestus, C. maenas and polychaetes on plots with increasing cover of 674

oysters on mussel-beds and mud-flats at Lough Foyle and Lough Swilly after 4 and 15 months. All data were sqrt(x+1) transformed to correct 675

heterogeneity of variance. Significant P values are indicated in italics. Post-hoc (SNK) results are detailed below each analysis. 676

Source of variation 4 mths: L. Foyle L. Swilly 15 mths: L. Foyle L. Swilly

Littorina littorea d.f. MS F P d.f. MS F P d.f. MS F P d.f. MS F P

Habitat (=H) 1 1.8x103 107.60 0.0000 1 247.20 51.87 0.0000 1 1.4x10

3 67.25 0.0000 1 1.0x10

3 63.10 0.0000


H x C 3 22.45 0.28 0.2882 3 14.44 3.03 0.0489 3 128.96 6.21 0.0053 3 156.03 9.46 0.0003

Residual 24 16.89

24 4.76

16 20.77

24 16.49

SNK (H) Mus > Mud

SNK (C) 100 = 50 > 5 = 0 SNK (H x C)

Mussel-bed

100 > 0

NS

NS

Mud-flat

100 > 50 = 5 > 0

100 = 50 > 5 = 0

100 = 50 > 5 = 0 Austrominius modestus

Habitat (=H) 1 6.7x104 52.84 0.0000 1 7.4x10

3 25.34 0.0000 1 5.2x10

4 77.3 0.0000 1 1.8x10

3 3.40 0.0777

Cover of oysters (=C) 3 1.2x104 9.6 0.0002 3 7.2x10

3 24.58 0.0000 3 4.8x10

3 7.19 0.0029 3 6.6x10

3 12.48 0.0000

H x C 3 7.8x103 6.13 0.0030 3 2.3x10

3 7.87 0.0008 3 859.9 1.28 0.3144 3 484.08 0.92 0.4468

Residual 24 1.3x103

24 294.22

16 670.75

24 526.99

SNK (H)

Mus > Mud SNK (C)

100 = 50 > 5 = 0

100 > 50, 5, 0 and 50 > 0

SNK (H x C) Mussel-bed 100 = 50 > 5 = 0

100 = 50 > 5 = 0

Mud-flat 100 > 50 = 5 = 0

100 > 50 = 5 > 0 Carcinus maenas

Habitat (=H) 1 159.99 44.81 0.0000 1 146.07 36.20 0.0000 1 94.13 11.14 0.0042 1 110.44 35.29 0.0000


H x C 3 12.71 3.56 0.0292 3 1.40 0.35 0.7877 3 10.7 1.27 0.3194 3 1.47 0.47 0.7067

Residual 24 3.57

24 4.04

16 8.45

24 3.13

SNK (H)

Mus > Mud

Mus > Mud

Mus > Mud SNK (C)

SNK (H x C)

Mussel-bed 100 > 0 = 5 = 50 Mud-flat 50 > 0 = 5 Phyllodocidae

30

Habitat (=H) 1 95.55 5.72 0.0250 1 4.84 0.15 0.7018 1 12.94 0.15 0.7022 1 852.40 0.70 0.0270


H x C 3 32.30 1.93 0.1511 3 54.53 1.69 0.1952 3 19.74 0.23 0.8733 3 1036.18 0.19 0.0018

Residual 24 16.71

24 32.21

16 85.36

24 153.52

SNK (H) Mus > Mud

SNK (C) 100 > 5 = 0

50 > 100 = 5 = 0 SNK (H x C)

Mussel-bed

100 < 0

Mud-flat 100 = 50 = 5 > 0

677

31

678

Figure 1. Mean (+/- S.E.) numbers of individuals per m2

after (a) 4 and (b) 15 months (note 679

there is a break between 3x105 and 8x10

5), species richness after (c) 4 and (d) 15 months and 680

Shannon-Wiener diversity after (e) 4 and (f) 15 months in mussel-beds (hatched bars) and 681

mud-flats (white bars) with increasing cover of C. gigas at Lough Foyle and Lough Swilly. 682

683

32

684

Figure 2. nMDS plots with stress values on square root transformed data of assemblages on 685

mussel-beds (black shapes) and mud-flats (white shapes) with 0, 5, 50 or 100% cover (as 686

indicated in the legend) of C. gigas at Lough Foyle after 4 (a), and 15 (c) months and Lough 687

Swilly after 4 (b) and 15 (d) months. Pseudo-F and P values from PERMANOVA are 688

indicated for the main effects, Habitat (Ha) and Cover of oysters (Co) and their interaction 689

(Ha*Co). Significant P values are indicated in italics. 690

691

33

692

Figure 3. Mean (+/- S.E.) density (number per m2) of L. littorea after (a) 4 and (b) 15 693

months, A. modestus after (c) 4 and (d) 15 months, C. maenas after (e) 4 and (f) 15 months 694

and Phyllodocidae polychaetes after (g) 4 and (h) 15 months in mussel-beds and mud-flats 695

with increasing cover of C. gigas at Lough Foyle and Lough Swilly. 696

34

Appendix 1. Species list detailing all of the taxa found throughout the experiment. Classifications were obtained from the World Register of 697 Marine Species (WoRMS) and the European Register of Marine Species (ERMS). 698

Phylum Class Order Family Genus Species

Heterokontophyta Phaeophyceae Fucales Fucaceae Fucus vesiculosus

Rhodophyta Florideophyceae Ceramiales Ceramiaceae Ceranium rubrum

Rhodophyceae Ceramiales Rhodomelaceae Laurencia

Florideophyceae Gigartinales Petrocelidaceae Mastocarpus

Annelida Hirudinea

Oligochaeta

Haplotaxida Tubificidae Tubificoides benedii

Polychaeta Aciculata Amphinomidae

Hesionidae

Nereidae Nereis

Nereis diversicolor

Canalipalpata Ampharetidae

Serpulidae Pomatoceros triqueter

Eunicida Eunicidae

Phyllodocida Aphroditidae

Glyceridae Glycera

Glycera tridactyla

Nephtyidae Nephtys

Phyllodocidae

Phyllodoce lamelligera

Polynoidae

Syllidae

Spionida Spionidae

Polydora ciliata

Prionospio fallax

Terebellida Cirratulidae

Capitellidae Capitella capitata

Orbiniidae

Scoloplos armiger

Bryozoa Gymnolaemata Cheilostomata Membraniporidae Conopeum seurati

Chordata Thaliacea

Crustacea Cirripedia Sessilia Balanidae Austrominius modestus

Copepoda Harpacticoida

Eumalacostraca Amphipoda Corophiidae Corophium volutator

Ischyroceridae Siphonoecetes striatus

Melitidae Melita palmata

35

Malacostraca Amphipoda Caprelloidea Pariambus typicus

Cumacea

Decapoda Crangonidae

Paguridae Pagurus bernhurdas

Portunidae Carcinus maenas

Isopoda Janiridae

Echinodermata Holothuroidea

Mollusca Bivalvia Myoida Myidae Mya arenaria

Mytiloida Mytilidae Mytilus edulis

Veneroida Cardiidae Cerastoderma edule

Pectinidae Chlamys varia

Tellinidae Macoma balthica

Veneridae Tapes rhomboides

Gastropoda Caenogastropoda Cerithiidae Bittium reticulatum

Hydrobiidae Hydrobia ulvae

Littorinidae Littorina littorea

Littorina obtusata

Littorina saxatalis

Muricidae Nucella lapillus

Patellidae Patella vulgata

Trochidae Gibbula cineraria

Gibbula umbilicalis

Polyplacophora Chitonida Lepidochitonidae Lepidochitona cinerea

Nematoda

Porifera Demospongiae Halichondrida Halichondriidae Halichondria panicea

699

Documents

Context- and density-dependent effects of introduced oysters on biodiversity