1

Download the full official version for free 1

http://onlinelibrary.wiley.com/doi/10.1111/j.1365-2656.2009.01609.x/full 2

3

Classic metapopulations are rare among common beetle species from a naturally fragmented 4

landscape 5

6

Don A. Driscoll1,2 7

Jamie B. Kirkpatrick1 8

Peter B. McQuillan1 9

Kevin J. Bonham1 10

11

1. School of Geography and Environmental Studies, University of Tasmania, Hobart, 12

Australia. 13

2. Fenner School of Environment and Society, The Australian National University, Canberra, 14

Australia 15

16

Corresponding Author: Dr Don Driscoll 17

Fenner School of Environment and Society 18

Australian National University 19

W.K Hancock Building (43) 20

Biology Place Canberra ACT 0200 21

Phone +61 2 6125 8130 22

Fax +61 2 61250757 23

email: don.driscoll@anu.edu.au 24

2

25

26

running headline: Classic metapopulations are rare27

3

Summary 28

1. The general importance of metacommunity and metapopulation theories is poorly 29

understood because few studies have examined responses of the suite of species that occupy 30

the same fragmented landscape. In this study we examined the importance of spatial 31

ecological theories using a large-scale, naturally fragmented landscape. 32

2. We measured the occurrence and abundance of 44 common beetle species in 31 natural 33

rainforest fragments in Tasmania, Australia. We tested for an effect on beetle distribution of 34

geographic variables (patch area, patch isolation, amount of surrounding habitat) and of 35

environmental variables based on plant species, after first accounting for spatial autocorrelation 36

using principal coordinates of neighbour matrices. The environmental variables described a 37

productivity gradient and a post-fire succession from eucalypt-dominated forest to late-38

successional rainforest. 39

3. Few species had distributions consistent with a metapopulation. However, the amount of 40

surrounding habitat and patch isolation influenced the occurrence or abundance of 30% of 41

beetle species, implying that dispersal into or out of patches was an important process. 42

4. Three species showed a distribution that could arise by interactions with dominant 43

competitors or predators with higher occurrence in small patches. 44

5. Environmental effects were more commonly observed than spatial effects. Twenty-three 45

percent of species showed evidence of habitat-driven, deterministic metapopulations. 46

Furthermore, almost half of the species were influenced by the plant succession or productivity 47

gradient, including effects at the within-patch, patch and regional scales. The beetle succession 48

involved an increase in the frequency of many species, and the addition of new species, with 49

little evidence of species turnover. Niche-related ecological theory such as the species-sorting 50

metacommunity theory was therefore the most broadly applicable concept. 51

6. We conclude that classic and source-sink metapopulations are probably rare in this large-52

scale, naturally fragmented system, although dispersal processes like those occurring in 53

4

metapopulations may have a substantial influence on community composition. However, 54

deterministic processes (niche specialisation, species-sorting metacommunities, and 55

deterministic metapopulations) drive the occurrence or frequency of the majority of species. 56

We urge further research into the prevalence of spatial ecological processes in large-scale 57

natural ecosystems to expand our understanding of the processes that may be important in 58

nature. 59

60

Key-words: adaptation, dispersal limitation, edge effect, habitat fragmentation, habitat loss, 61

island biogeography, invertebrate, insect, mass effects, neutral theory 62

63

Introduction 64

Substantial emphasis has been placed on the random nature of extinction, dispersal and 65

colonisation in ecological theory (MacArthur and Wilson, 1967; Hanski, 1998; Hanski, 1999). 66

Stochastic colonisation and extinction are fundamental to the classic metapopulation concept 67

(Levins, 1970; Hanski, 1998) and neutral metacommunity theory (Bell, 2001; Hubbell, 2001). 68

However, other manifestations of metapopulation and metacommunity theory recognize that 69

the distributions of species can be influenced by the spatial and temporal distribution of 70

suitable habitat (Thomas and Hanski, 1997; Leibold et al., 2004; Cottenie, 2005). Despite this 71

very rich theoretical arena, and supporting evidence from a range of empirical studies for most 72

of these concepts (e.g. Thomas, Singer and Boughton, 1996; Elmhagen and Angerbjorn, 2001; 73

Muneepeerakul et al., 2008), the relative importance of any of these concepts in explaining 74

spatial patterns in nature is unknown. There is very little information about the proportion of 75

species that conform to particular spatial ecological theories (Driscoll, 2007; Driscoll and 76

Lindenmayer, 2009). We aim to help address this knowledge gap by examining the occurrence 77

and abundance of beetle species in a large-scale naturally fragmented landscape. 78

79

5

Correlates of a species’ distribution among a set of habitat patches can reveal the most 80

important processes influencing spatial dynamics. For example, extinction and colonisation in 81

a classic metapopulation can lead to lower occupancy in smaller patches due to higher rates of 82

extinction, and lower occupancy in the most isolated sites due to reduced chances of 83

recolonisation (Etienne, Ter Braak and Vos, 2004). Therefore, if the occurrence of a species is 84

negatively related to patch isolation and positively related to patch size, classic metapopulation 85

dynamics would be supported (Hanski and Simberloff, 1997; Etienne et al., 2004; Matthies et 86

al., 2004), although without population turnover data, these cannot be distinguished from 87

source-sink metapopulations (Driscoll, 2008). Contrasting with the emphasis that theoretical 88

ecology places on habitat isolation, there is evidence that the total amount of surrounding 89

habitat can strongly influence species' distributions, regardless of the amount of fragmentation 90

(Fahrig, 1997; Fahrig, 2002; Harrison et al., 2006; Radford and Bennett, 2007). 91

92

However, for many species, the geography of patches may be less important than patch quality 93

(Harrison and Bruna, 1999; Jellinek, Driscoll and Kirkpatrick, 2004). Spatial dynamics within 94

habitat archipelagos can be deterministic, with colonisation and extinction driven by the state 95

of the habitat patch (Thomas, 1994). For example, habitat degradation (Harrison and Bruna, 96

1999), succession after a disturbance (Stelter et al., 1997) or after habitat creation (Sjogren-97

Gulve, 1994) can be major drivers of spatial population dynamics. If occurrence is influenced 98

by patch quality, a deterministic metapopulation is supported. A deterministic metapopulation 99

differs from a classic, or source-sink metapopulation, because the latter assume that patches are 100

always available to a species and that stochastic extinction and colonisation account for the 101

dynamics (Thomas, 1994). 102

103

Niche theory encompasses the occurrence pattern described by the deterministic 104

metapopulation concept, but also describes differences in density that arise from habitat 105

6

specialisation (Hutchinson, 1957; Leibold et al., 2004; Richter-Boix, Llorente and Montori, 106

2007). If abundance but not occurrence of a species is dependent on habitat characteristics, 107

metapopulation dynamics are not supported, though niche partitioning would be inferred. 108

109

Interactions among species can also influence occurrence patterns (Taylor, 1990; Leibold et al., 110

2004). When dominant competitors or predators have lower dispersal ability than subordinate 111

competitors or prey, the subordinate species can find refuge in more isolated sites (Tilman, 112

1994; Yu et al., 2004). Subordinate species may therefore have higher occurrence in the most 113

isolated sites, a pattern opposite to that expected under single-species metapopulation theory 114

(Driscoll, 2008). 115

116

We examined the occurrence and abundance of beetles in 31 natural cool-temperate rainforest 117

patches which ranged in size and isolation. Floristic composition also varied among sites in 118

response to a productivity gradient and succession after fire. Our aim was to determine how 119

many species showed evidence that was consistent with theories describing species and 120

community responses in fragmented landscapes. We took advantage of the information 121

harboured in species’ distributions (Etienne et al., 2004) to test for evidence of: 122

123

1. Classic and source-sink metapopulations (negative relationship between occurrence and 124

patch isolation, positive relationship with patch area, only a proportion of patches occupied, 125

occurrence not influenced by environmental variables) (Levins, 1970; Pulliam, 1988; Hanski, 126

1998); 127

2. Deterministic metapopulations (occurrence related to environmental variables, not 128

geographic variables, only a proportion of patches occupied) (Thomas, 1994); 129

3. Habitat-amount theory (positive relationship of occurrence or abundance with amount of 130

surrounding forest, not with patch isolation or size) (Fahrig, 1997; Fahrig, 2002); 131

7

4. Dominant-species / dispersal tradeoffs (higher occurrence in most isolated patches) (Taylor, 132

1990; Leibold et al., 2004); 133

5. Niche theory (in addition to evidence from point 2 above, abundance related to 134

environmental parameters) (Hutchinson, 1957; Leibold et al., 2004; Richter-Boix et al., 2007). 135

136

We examine a large number of species from the same fragmented landscape to assess the likely 137

general importance of these contrasting ecological theories. 138

139

Materials and Methods 140

STUDY AREA 141

In south-west Tasmania, Australia, cool-temperate rainforest occurs naturally in discrete 142

patches surrounded by sedgeland (Read, 1999). The patches range in size from a few hectares 143

up to several square kilometres (Jarman, Brown and Kantvilas, 1984; Kirkpatrick and 144

Dickinson, 1984). The rainforest varies in plant species composition (Jarman, Kantvilas and 145

Brown, 1999), and forest patches can occur in a range of post-fire successional stages (Jackson, 146

1968). The forest beetle communities are distinct from those in the surrounding sedgeland 147

matrix (Driscoll, 2005). 148

149

We have deliberately eschewed selecting an “ideal” study landscape, with patches of a 150

prescribed size, isolation and with uniform habitat. We want to know how important spatial 151

ecological theory is in a large-scale naturally fragmented landscape where there is substantial 152

variation in patch geography and, as in most landscapes, the plant communities are 153

heterogeneous. 154

155

BEETLE COLLECTION 156

8

Beetles were sampled from 31 rainforest patches in two adjacent regions (Fig. 1). Region one 157

has slightly higher precipitation with infertile soils formed on quartzites and phyllites whereas 158

many sites in region two are underlain by fine-textured sedimentary rocks. Site productivity is 159

driven by rainfall, topography, altitude and soil fertility (Kirkpatrick, 1984) which results in a 160

bias towards more productive sites in region 2. 161

162

Each patch was sampled using three trap grids separated by approximately 100 m (Driscoll in 163

press). Equal sampling effort was used regardless of patch size to ensure that detection rates 164

were proportional to population density and not sampling effort. Grid locations were stratified 165

to sample different successional ages in the patch. In the absence of fire, forest patches expand 166

into the buttongrass matrix so usually have a fringe of early successional habitat. Each grid 167

consisted of 16 pairs (32 cups in total) of 225 ml plastic cups used as pit-fall traps. We used 168

plastic container lids suspended on wooden skewers to prevent the cups from filling with rain, 169

hail or snow. Pairs were spaced at 5 m intervals, and cups within pairs separated by 1 m. Each 170

cup had 50 ml of Gault’s solution as a preservative (Walker and Crosby, 1988). The traps were 171

set in January 2002 and left open for eight weeks. We used this very large sampling effort (96 172

pit-fall traps per site) to ensure that a reasonable number of species would be detected reliably, 173

if they were present (Driscoll in press). Trapped animals were preserved in ethanol, and all 174

identifications were completed by the same person (KB). Vouchers are housed in the 175

Tasmanian Forestry Insect Collection, Hobart, code numbers FT44089 to FT44145. 176

177

PLANT SURVEYS 178

Percentage cover of all plant species within each of the pit-fall grids (three grids of 15 m × 15 179

m) was estimated in three categories: < 20 %, 20-50 %, > 50 %. The size of tree species was 180

estimated in three categories: (1) young (< 15 cm diameter at breast height (DHB)); (2) mid 181

9

(not reaching canopy, 15-50 cm DBH); (3) mature and old-growth (DBH > 50cm, canopy tree). 182

Nomenclature follows Buchanan (2006). 183

184

AMOUNT OF HABITAT AND ISOLATION MEASURES 185

Geographic parameters were the area of forest within 300 m of grid B (usually located between 186

the other two) and the area of forest within 300 m of the patch boundary (isolation). The 187

distance of 300 m was chosen to maximise dispersion of values across the range of possible 188

values. We estimated the area of forest around the points or patches at 100 m intervals (100-189

1000 m) and 250 m intervals (1000-2000 m). We then counted the number of patches or points 190

in 10% interval categories of percent forest cover. The buffer distance that best separated the 191

sites was where the maximum number of the ten categories included sites and in approximately 192

equal numbers. The average proportion of forested area within 300 m of a sample point was 193

0.49 (SD 0.24), and the average proportion of forested area in a 300 m buffer around patches 194

was 0.21 (SD 0.21). The buffer values therefore spanned a broad range of possible values. We 195

examined the influence of using a 1000 m buffer and found this made no difference to the 196

apparent prevalence of contrasting processes. 197

198

ENVIRONMENTAL VARIABLES 199

To summarise plant community differences between patches we used non-metric 200

multidimensional scaling (NMDS) with a Bray-Curtis distance matrix. The distance matrix 201

was based on the plant species percent cover data, with values for each species averaged across 202

grids within sites. NMDS using two dimensions was completed using functions initMDS and 203

isoMDS of R version 2.5.0 (R Development Core Team, 2007). Multiple starts of the NMDS 204

procedure were undertaken to avoid being trapped in local minima. New starts were continued 205

until the stress had not been reduced for 100 successive iterations. The maximum iterations 206

within isoMDS was 200, and convergence tolerance 10-7. Procrustes rotation ensured that most 207

10

of the variation was captured in axis 1 (vegan 1.8 Oksanen et al., 2007). To assist with 208

interpretation of the two NMDS axis, we used ANOVA to examine the relationship of each 209

plant species, region, and forest succession stage with each NMDS axis. 210

211

We also estimated the stage of post-fire succession. Eucalypts dominate the early stages of 212

succession, but a dense canopy of the rainforest species Nothofagus cunninghamii, Eucryphia 213

lucida and Atherosperma moschatum prevents eucalypt regeneration (Jackson, 1968). After 214

the eucalypts die (100-450 years), rainforest remains (Jackson, 1968; Read, 1999). The size 215

and abundance of eucalypts compared with rainforest trees indicates successional status. Tree 216

growth rate is influenced by nutrient availability (Read, 2001), so tree size cannot be used to 217

accurately date the stand (Koch, Driscoll and Kirkpatrick, 2008). We devised a succession 218

scale to represent the gradient from (1) pure rainforest through to (3) eucalypt stands: 219

1. No eucalypts and either Nothofagus of size category 2 (150-500mm) or 3 (>500mm), or 220

Atherosperma at size 2 or Eucryphia at size 3 or other rainforest understory species 221

predominant (two sites only) 222

2. Eucalypts at size 3 and Nothofagus, Eucryphia, or Phyllocladus at size 2 or 3. 223

3. Eucalypts at size 1 (< 150mm) or 2 and Nothofagus or Eucryphia present, or eucalypts at 224

size 2 and 3 with no Nothofagus, Eucryphia or Phyllocladus. 225

226

The minimum score from the three plant grids was used to represent the latest stage of 227

succession present in the patch. 228

229

STEP-WISE GENERALISED LINEAR MODELS 230

We analysed separately presence/absence data and frequency data (number of trap-pairs in 231

which a species was captured in a patch; a surrogate for abundance). 232

233

11

To account for spatial autocorrelation (Gonzalez-Megias, Gomez and Sanchez-Pinero, 2005) 234

prior to fitting a step-wise model, six principal coordinates of neighbour matrices (PCNM) 235

were calculated (Borcard and Legendre, 2002; Dray, Legendre and Peres-Neto, 2006). For 236

each analysis we fitted the PCNM variables in a binomial generalised linear model with a logit 237

link-function (McCullagh and Nelder, 1989). We permuted the beetle frequency data 1000 238

times to obtain a distribution of r2 (deviance explained by factor/total deviance) and this was 239

used to estimate P values for each of the six variables (proportion of random r2 values ≥ actual 240

value). We adjusted the P values to control the false discovery rate (Benjamini and Hochberg, 241

1995) using p.adjust in the R package stats 2.5.0 (R Development Core Team, 2007). Each 242

species and analysis type (presence/absence, frequency) was considered a separate family. 243

PCNM variables with a q value of ≤ 0.1 were used as the base model for the step-wise 244

procedure. An alternative approach of using analysis type alone to designate family was 245

extremely conservative with all P values above zero having q values > 0.1. This did not 246

capture obvious spatial structure, such as the near absence from one region but not the other, 247

and so the less conservative approach was used. 248

249

Using a simultaneous forwards and backwards stepwise approach (McCullagh and Nelder, 250

1989) with AIC as a stopping rule, we fitted three vegetation variables (MDSaxis1, MDSaxis2, 251

succession score) and three geographic variables (patch-area, patch isolation, forest area) to the 252

base model containing any significant PCNMs. Variable selection procedures produce over-253

optimistic P values (Maindonald and Braun, 2007). We therefore tested the significance of 254

fitted parameters using a permutation method (Arditi, 1989; Good, 1994). We permuted the 255

beetle species data among sites 1000 times and with each permutation re-fitted the parameters 256

selected by the step-wise procedure then calculated r2 as the test statistic. We adjusted P values 257

(Benjamini and Hochberg, 1995) for all variables included in the models, based on all variables 258

12

available for selection. We report results where the adjusted P is < 0.1 to reduce the risk that 259

low power leading to type-2 errors may obscure support for metapopulation processes. 260

261

When using step-wise models, the inclusion of a variable that explains the most variation may 262

obscure slightly weaker relationships with other variables if they are correlated with the first 263

variable. To determine if this was a problem in our analyses, we examined the amount of 264

independent and joint variation explained by the environmental, geographic and spatial PCNM 265

variables, using hierarchical partitioning (Chevan and Sutherland, 1991; MacNally, 1996). We 266

only included PCNM variables that were significantly correlated with a species to simplify the 267

model before analysis. Significance of the independent variation explained was ascertained by 268

permutation. 269

270

Species analysed separately occurred in at least 20% of patches (six) and had a frequency of at 271

least 0.049 within patches in which they were trapped (probability of detection with 48 traps > 272

90% McArdle, 1990). Species’ frequency is a reasonable guide to the probability of detection 273

(Driscoll in press). Six patches was used as an arbitrary cut off below which we did not expect 274

reliable models to arise. Species that occurred in 20-80% of patches were used in the 275

presence/absence analyses. Verheyen et al. (2004) also used a 20-80% occupancy range, 276

beyond which they regarded metapopulation dynamics as unlikely or undetectable. 277

278

We applied a similar step-wise approach, but using an ANOVA model, for the number of 279

species in a patch, and the number of rare species. Rare species were those with a frequency of 280

less than 0.049. 281

282

SPECIES-ENVIRONMENT RELATIONSHIPS WITHIN PATCHES 283

13

Pooling data from the three grids has the potential to obscure relationships between beetle 284

abundance and environmental parameters. We therefore performed similar analyses to those 285

described for patch-level data, but analysed beetle frequency data and plant data from each grid 286

within a patch. We first fitted patch-name to remove patch-level variation. We then fitted 287

three environmental parameters in the step-wise binomial GLMs, including succession score 288

for each grid and two NMDS axes from analyses of plant species data at the grid level. The 289

grid-level NMDS axes had the same biological interpretation as the patch-level axes (see 290

results). 291

292

Examining species traits such as flight capacity can provide valuable additional insight into the 293

response of species to geographic and environmental variables (Driscoll and Weir, 2005), but 294

our analyses in this respect were not very informative (Appendix S1). 295

296

Results 297

Although 168 morphospecies were identified from 26532 individuals, only 44 species satisfied 298

the criteria for species-level analysis (Appendix S2). Eighteen of 44 species occurred on more 299

than 80% of patches. One site (B13) had a unique beetle and plant community compared with 300

the rest of the sites. It was the only site on the floodplain of a large river and has been 301

excluded from further analyses (Appendix S3). 302

303

ENVIRONMENTAL AXIS 304

MDS axis 1 represented the forest succession gradient with early succession eucalypt 305

communities at low values and late succession rainforest at high values (mean axis 1 value for 306

succession scores 1 versus 2+3 were 0.138, -0.2415, P = 0.003). Plant species that were most 307

negatively related to MDS axis 1 included Melaleuca squarrosa (sub-canopy tree, r2 = 0.69, P 308

< 0.0001); Eucalyptus nitida (canopy tree, r2 = 0.60, P < 0.0001);Gleichenia microcarpa (fern, 309

14

r2 = 0.61, P < 0.0001); Bauera rubioides (shrub, r2 = 0.60, P < 0.0001); Nematolepis squamea 310

(shrub, r2 = 0.57, P < 0.0001); and Pteridium esculentum (fern, r2 = 0.69, P < 0.0001). These 311

species are typical of sclerophyllous plant communities in SW Tasmania. Plant species most 312

positively correlated with MDS axis 1 included Grammitis billardierei (fern, r2 = 0.45, P = 313

0.0001); Atherosperma moschatum (canopy tree, r2 = 0.28, P = 0.0026); Eucryphia lucida 314

(canopy tree, r2 = 0.24, P = 0.0051); Nothofagus cunninghamii (canopy tree, r2 = 0.18, P = 315

0.019) and are typical rainforest species 316

317

Mean scores on MDS axis 2 were significantly differentiated by region (region 1 versus 2, -318

0.1229, 0.0668, P = 0.040). The scores on this axis had a positive relationship with 319

Anodopetalum biglandulosum (usually sub-canopy tree, r2 = 0.32, P = 0.0012), a species 320

characteristic of the more tangled rainforests on low nutrient sites. The scores also had 321

negative relationships with Nothofagus cunninghamii (canopy tree, r2 = 0.35, P = 0.0006), 322

Phyllocladus aspleniifolius (canopy tree, r2 = 0.38, P = 0.0003), Blechnum wattsii (fern, r2 = 323

0.43, P = 0.0001), and Trochocarpa gunnii (shrub, r2 = 0.41, P = 0.0001), species that are 324

widespread in tall rainforests with sparse understorey. We therefore interpret low values of 325

MDS axis 2 as an indicator of the more productive sites and callidendrous to thamnic rainforest 326

(sensu Jarman et al., 1984) with or without emergent eucalypts, and high values as an indicator 327

of less productive sites and thamnic to implicate rainforests (Jarman et al., 1984) with or 328

without emergent eucalypts (later referred to as low-nutrient patches). 329

330

GENERALISED LINEAR MODELS 331

The hierarchical partitioning analyses usually either supported the GLM results, or were not 332

significant. Only two GLM results were contradicted by the hierarchical partitioning analyses 333

(Roptoperus tasmaniensis, Myrmicholeva ligulata, Appendix S4, S3). These made no 334

difference to our interpretation or conclusions. The key message from the hierarchical 335

15

partitioning analyses is that relationships between patch geography or the environment and 336

beetle species were not obscured in the step-wise models by covariation of the explanatory 337

variables. 338

339

Spatial parameters derived from the PCNM analysis were included in occurrence models of 340

nine species and in frequency models for 13 species. PCNM1 separated patches from the 341

northern and southern regions. Four species had lower occurrence or lower frequency in the 342

southern region while nine species had the opposite pattern (Appendix S5). PCNM2, 3, 4, and 343

5 each influenced occurrence or frequency of six, four, one and three species respectively. 344

Although difficult to interpret, their inclusion in species' models indicates the removal of some 345

spatial autocorrelation effects prior to examining environmental and patch geography 346

relationships. 347

348

Next we focus on the environmental and geographic results of the GLM. The results are 349

presented in order of the five processes listed at the end of the introduction. We draw attention 350

to species that show evidence in support of each process, but also note when those species 351

appear to have other processes acting simultaneously. 352

353

Classic and source-sink metapopulations 354

Of the 26 species occurring on 20-80% of patches, only one species had a pattern consistent 355

with classic or source-sink metapopulation predictions. Catoposchema tasmaniae occurred 356

more frequently in larger patches (Appendices S2, S5). However, three other species had 357

higher frequency on the least isolated patches (Lissotes curvicornis; Palimbolus victoriae; 358

Telura vitticollis Appendices S2, S5). 359

360

Deterministic metapopulations 361

16

Environmental parameters influenced the occurrence of 10 species suggesting they have a 362

deterministic metapopulation, although three of these species also had their occurrence 363

influenced by geographic parameters. Four of these species had higher occurrence on low-364

nutrient sites (Aridius nodifer, Aspidiphorus humeralis, Palimbolus victoriae, Pselaphinae 365

TFIC sp. 29) and occurrence of the latter species was also higher with more surrounding forest 366

(Appendix S2). Six species had higher occurrence in late successional rainforest 367

(Austronemadus TFIC sp. 03, Brycopia hexagona, Dinichus terreus, Pedilophorus gemmatus, 368

Pterocyrtus globosus, Pterocyrtus tasmanicus), although occurrence of Dinichus terreus was 369

also higher in smaller patches, and Pedilophorus gemmatus occurred more often with more 370

surrounding forest (Appendices S2, S5). 371

372

Amount of surrounding habitat 373

The prediction that the amount of surrounding habitat should influence occurrence was 374

supported by two species (Pselaphinae TFIC sp. 29, Pedilophorus gemmatus), which had 375

higher occurrence in sites with more surrounding forest. The latter species also had higher 376

frequency and occurrence on low-nutrient sites while P. gemmatus also had higher occurrence 377

in rainforest. Seven additional species had higher frequency in sites with more surrounding 378

forest (Aspidiphorus humeralis, Austronemadus TFIC sp. 03, Coripera deplanata, Decilaus 379

striatus, Roptoperus tasmaniensis, Sogdini ANIC Gen B. TFIC sp. 01, Zeadolopus TFIC sp. 380

02), though the first two and last species also had frequency and/or occurrence influenced by 381

environmental parameters (Appendices S2, S5). 382

383

Dominant-species / dispersal tradeoffs 384

Three flightless species had higher occurrence in small patches (Sloaneana tasmaniae, 385

Trechimorphus diemenensis, Dinichus terreus, the latter also with occurrence higher in late-386

successional rainforest, Appendices S2, S5). 387

17

388

Niche effects on abundance 389

The frequency but not the occurrence of ten species was only influenced by environmental 390

parameters (though three of these included PCNM variables in their models). Four of those 391

species were more frequent in late-successional rainforest (Myrmicholeva ligulata; 392

Nargomorphus globulus; Pselaphaulax CHANDLER Tasmania 1; Rybaxis parvidens), four on 393

low-nutrient sites (Microchaetes hystricosus; Thalycrodes cylindricum; Scydmaenidae TFIC 394

sp. 04; Thalycrodes pulchrum), one in patches with eucalypts and in low-nutrient sites 395

(Heteronyx tasmanicus), and one in high-nutrient sites (Decilaus TFIC sp. 19, Appendices S2, 396

S5). Two additional species had higher frequency in low-nutrient sites, but these also had 397

frequency influenced by amount of surrounding forest (Zeadolopus TFIC sp. 02) or patch 398

isolation (Telura vitticollis). 399

400

Within patches, at the grid level, environmental parameters explained from 4 to 12% of model 401

variation for 14 species (at P < 0.1, Appendix S6). Two of these reflected the same response as 402

observed at the patch level of analysis, with higher frequency or occurrence towards the 403

rainforest end of the succession (Pterocyrtus tasmanicus, Pterocyrtus globosus). Nine species 404

that had not shown a relationship with environmental parameters at the patch scale, did so at 405

the grid scale, including eight species that were more frequent in rainforest (Agonica simsoni, 406

Choleva TFIC sp 01, Decilaus striatus, Decilaus TFIC sp 04, Pseudomicrocara TFIC sp 02, 407

Roptoperus tasmaniensis, Stichonotus leai, Trechimorphus diemenensis), and one in low 408

nutrient sites (Pseudomicrocara TFIC sp 01). Three species that had higher frequency in low-409

nutrient sites at the patch scale had higher frequency in eucalypts at the grid scale (Aridius 410

nodifer, Thalycrodes cylindricum, Telura vitticollis, Appendix S6). There were 17 species 411

with environmental relationships at the patch level but without a relationship with the 412

environment at the grid level (Appendices 2, 5, 6). 413

18

414

Number of Species 415

The number of beetle species in a patch and the number of rare species were significantly 416

positively correlated with MDSaxis2, implying more species in low-nutrient sites (Appendix 417

S5). The number of species and the number of rare species were also significantly negatively 418

correlated with succession score, with more species in late-successional rainforest (Appendix 419

S5). 420

421

Summary of patterns that were consistent with theoretical expectations 422

Of 44 common species, only 2.3% of species showed evidence to support a classic 423

metapopulation model, whereas 22.7% supported a deterministic metapopulation, with patch 424

occupancy influenced by the stage of forest succession and rainforest type (Fig. 2). The 425

occurrence or frequency of 20.5% of species was strongly influenced by the amount of habitat 426

surrounding the sample site, regardless of patch size or isolation. Three common beetle species 427

(6.8%) had higher frequency or occurrence in small patches which may reflect negative 428

interactions with dominant species in large patches. Twelve species (27.3%) showed evidence 429

of weak niche-partitioning at the patch scale, with habitat type influencing only their 430

frequency. Taking into account within-patch environmental relationships, 21 species (47.7%) 431

showed weak niche-partitioning (Appendix S6). 432

433

Discussion 434

CLASSIC AND SOURCE-SINK METAPOPULATIONS 435

Only one beetle species had the positive relationship between occurrence and patch size that 436

was consistent with classic and source-sink metapopulation predictions (Hanski and 437

Simberloff, 1997; Etienne et al., 2004). Three more species had a lower frequency on isolated 438

patches, consistent with a boost in beetle density by higher immigration into the least isolated 439

19

sites (Brown and Kodric-Brown, 1977; Pulliam, 1988; Leibold et al., 2004). If this pattern was 440

generated by dispersal limitation, metapopulation dynamics may become possible for those 441

species in landscapes with more isolated patches. The rarity of evidence for classic or source-442

sink metapopulation dynamics is not consistent with the results of most other studies (Table 1), 443

but is consistent with the results of Driscoll (2008), despite that work taking place at a much 444

smaller spatial scale. 445

446

Eighteen species occurred on >86% of patches and so probably do not form metapopulations. 447

We suggest that high rates of migration between patches may prevent populations of these 448

species from ever becoming extinct (Harrison, 1991). Alternatively, the populations may be 449

isolated, but are so large that the time to extinction exceeds the time over which the patches 450

change configuration in response to changing climatic conditions (on the order of thousands of 451

years, Leigh, 1981; Jackson, 1999). 452

453

A key assumption of our interpretation is that dispersal ability limits access to patches (Hanski 454

and Simberloff, 1997). However, if establishment was limiting, isolation effects would not be 455

observed (Eber and Brandl, 1996; Verbeylen, De Bruyn and Matthysen, 2003). If 456

establishment limitations was stochastic, the process would be consistent with Levins' simple 457

metapopulation model (Levins, 1970). However, deterministic processes could also limit 458

colonisation (such as natural enemies: Ryall and Fahrig, 2006). Conceivably, seven species 459

that occurred on 20-80% of sites and that did not have significant model effects at the patch 460

scale, could have establishment-limited metapopulations. However, we can find few examples 461

of this kind of metapopulation in the literature (Werth et al., 2006) and dispersal limitation is 462

commonly assumed to be more realistic (Hanski, 1998). Population turnover data, genetic data 463

and knowledge of competitor or predator distributions could help distinguish dispersal from 464

establishment-limited processes (Driscoll, 2007). 465

20

466

DETERMINISTIC METAPOPULATIONS 467

Deterministic processes influenced the occurrence of more species (ten) than did isolation and 468

size (one). Deterministic metapopulations may be more common in nature than the traditional 469

ideas of a metapopulation that is driven by stochastic processes (Thomas, 1994). We note also 470

that niche concepts (Hutchinson, 1957), including the species-sorting metacommunity concept 471

(Leibold et al., 2004) can provide an equally valid construct for describing (presumed) spatial 472

dynamics that are driven by differences in the environment. 473

474

HABITAT AMOUNT 475

The amount of surrounding habitat can have a strong influence on species’ distributions 476

(Fahrig, 1997; Fahrig, 2003; MacDonald and Kirkpatrick, 2003; Driscoll, 2004; Radford and 477

Bennett, 2007). Habitat amount influenced the occurrence of two species, implying they may 478

have spatial dynamics that are similar to the classic metapopulation. However, direct 479

connectivity, which defines the patches of a metapopulation, appears to be less important than 480

the area of near-by habitat. Occupancy may be lower with less surrounding habitat through 481

lower immigration (Brown and Kodric-Brown, 1977), higher emigration (Hill, Thomas and 482

Lewis, 1996; Fahrig, 2001) or increased immigration of dominant matrix species (Turner, 483

1996; Harrison and Bruna, 1999). We think the latter explanation is less likely because there 484

was very little overlap of matrix and forest beetle fauna (Driscoll, 2005; Driscoll, 2008). 485

Similar processes may account for the influence of amount of surrounding habitat on the 486

frequency of seven additional species. These findings emphasise that the patch-based model of 487

a landscape, with the assumption that a separate panmictic population resides within each 488

patch, does not always apply because spatial population dynamics occurs at a different, in this 489

case smaller, scale (Lindenmayer, McIntyre and Fischer, 2003). 490

491

21

DOMINANT SPECIES / DISPERSAL TRADEOFFS 492

Three species had higher occurrence on small patches. They were all flightless, and therefore 493

had a distribution inconsistent with the dispersal-limited species-sorting metacommunity 494

model. The model suggested that species that would otherwise be outcompeted or devoured 495

may survive in remote patches because of their superior dispersal ability (Tilman, 1994; 496

Driscoll, 2008). These three species may be finding refuge on patches too small to support 497

persistent populations of their predators and competitors. However, we have only examined 498

spatial patterns and more direct approaches to identifying competitive or predatory 499

relationships are now needed (e.g. Yu et al., 2004). Small or isolated patches may be important 500

refuges for a suite of presumably subordinate, forest specialist beetles (Tscharntke et al., 2002; 501

Driscoll, 2008). 502

503

NICHE THEORY 504

Environmental parameters influenced the frequency of 21 species but not their occurrence and 505

so we did not discuss them in the context of deterministic metapopulations. The influence of 506

environmental parameters on frequency could reflect differential reproduction and survival in 507

different habitats (Barker and Mayhill, 1999) or habitat selection (Fletcher, 2007). Our results 508

are consistent with recent evidence that species-sorting metacommunity processes have the 509

strongest influence on community composition (Cottenie, 2005; Parris, 2006; Brooks et al., 510

2008). 511

512

Species with an environmental relationship at the patch scale, but not within patches at the grid 513

scale imply that habitat type within a patch is important for reproduction, but that animals can 514

disperse throughout the patch, eliminating any differences between habitats within patches 515

(mass effects, Shmida and Ellner, 1984; Leibold et al., 2004). On the other hand, eight species 516

with higher frequency in rainforest at the grid but not the patch scale probably reflect 517

22

methodological limitations. Patch-level environmental measures may not reflect habitat 518

availability within patches, preventing us from detecting an effect at the patch scale 519

(Whittingham et al., 2005). 520

521

In support of niche theory, post-fire succession was characterised by increased frequency and 522

addition of species, rather than species turnover. This additive effect contrasts with beetle 523

succession in boreal forests, which has more distinct communities (Paquin, 2008). However, 524

evidence from eucalypt forests younger than those that we sampled suggests there is a suite of 525

early successional species in Tasmania (Baker, 2006), similar to that observed in the boreal 526

forest (Paquin, 2008). A clear implication of our results for management is that burning or 527

logging old-growth forests and rainforest will reduce local beetle diversity. 528

529

Although most of the PCNM spatial variables could not be interpreted, the main spatial pattern 530

was the difference between the northern and southern regions (PCNM1). This pattern could 531

partly be driven by environmental differences because region was correlated with MDSaxis2, 532

with a tendency for more low nutrient sites to occur in the north. Environmental filters 533

therefore seem to act at multiple spatial scales, with species distributions influenced by factors 534

that vary within patches at the grid level, between patches and between regions. 535

536

CONCLUSIONS 537

Within a large-scale, naturally fragmented ecosystem, we have shown that beetles have a broad 538

range of responses, exhibiting patterns that are consistent with a diverse range of theory (Fig. 539

2). Metapopulation theory widely permeates the conservation and ecological literature (Hanski 540

and Gaggiotti, 2004), but in our study landscape, evidence for classic or source-sink 541

metapopulations was rare. Part of the reason for this was that a proportion of species 542

responded to the amount of surrounding habitat at a scale smaller than the patches. Habitat 543

23

patches did not determine the structural basis of dynamics for many species, contrary to 544

metapopulation assumptions. Nevertheless, one fifth of common beetles may have been 545

subject to the same dispersal limitation that can lead to metapopulation dynamics. Dispersal 546

limitation may have influenced community development, even though metapopulations were 547

rare. 548

549

In contrast to the limited evidence supporting spatial processes, niche concepts were widely 550

supported. Vegetation type more frequently influenced occurrence than habitat size, amount or 551

isolation, suggesting that deterministic spatial dynamics are common relative to stochastic 552

dynamics (Hanski and Simberloff, 1997). If this is generally true, spatial dynamic models may 553

need to incorporate environmentally driven population turnover rather than just stochastic 554

processes (Verheyen et al., 2004; Wilcox, Cairns and Possingham, 2006). 555

556

We hope that more studies take the approach of examining multiple species in large-scale 557

fragmented landscapes. Turnover data and genetic analyses would be valuable additions to our 558

approach to enable recognition of establishment-limited rather than just dispersal-limited 559

metapopulations. Although individual species studies will continue to be important for 560

exploring mechanisms (e.g. Lindenmayer, 2000), taking a multi-species approach in a natural 561

ecosystem is essential for assessing the importance of competing theories, and thereby for 562

understanding how communities work. 563

564

Acknowledgements 565

We thank Madelaine Hanson for assistance in the field, Amelia Koch for her feedback on an 566

earlier draft and Simon Grove for access to Forestry Tasmania’s beetle collections. Jeff Wood 567

and Hwan-Jin Yoon provided statistics advice. This research was completed under Tasmanian 568

Parks and Wildlife Service permit number FA 01030, complying with relevant State and 569

24

Federal laws. The field work was funded through an Australian Research Council Post 570

Doctoral Fellowship, which DD held in the School of Geography and Environmental Studies at 571

the University of Tasmania. Beetle collections were sorted by KB with funding to DD from 572

Flinders University of South Australia. The project was written up at The Australian National 573

University. 574

575

References 576

Arditi, R. (1989) Avoiding Fallacious Significance Tests in Stepwise Regression - a Monte-577

Carlo Method Applied to a Meteorological Theory for the Canadian Lynx Cycle. 578

International Journal of Biometeorology, 33, 24-26. 579

Baker, S. C. (2006) A comparison of litter beetle assemblages (Coleoptera) in mature and 580

recently clearfelled Eucalyptus obliqua forest. Australian Journal of Entomology, 45, 581

130-136. 582

Barker, G. M. & Mayhill, P. C. (1999) Patterns of diversity and habitat relationships in 583

terrestrial mollusc communities of the Pukeamaru Ecological District, northeastern 584

New Zealand. Journal of Biogeography, 26, 215-238. 585

Bell, G. (2001) Ecology - Neutral macroecology. Science, 293, 2413-2418. 586

Benjamini, Y. & Hochberg, Y. (1995) Controlling the False Discovery Rate - a Practical and 587

Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society Series 588

B-Methodological, 57, 289-300. 589

Borcard, D. & Legendre, P. (2002) All-scale spatial analysis of ecological data by means of 590

principal coordinates of neighbour matrices. Ecological Modelling, 153, 51-68. 591

Brooks, D. R., Perry, J. N., Clark, S. J., Heard, M. S., Firbank, L. G., Holdgate, R., Mason, N. 592

S., Shortall, C. R., Skellern, M. P. & Woiwod, I. P. (2008) National-scale 593

metacommunity dynamics of carabid beetles in UK farmland. Journal of Animal 594

Ecology, 77, 265-274. 595

25

Brown, J. H. & Kodric-Brown, A. (1977) Turnover rates in insular biogeography: effect of 596

immigration on extinction. Ecology, 58, 445-449. 597

Buchanan, A. (2006) A census of the vascular plants of Tasmania, Tasmanian Herbarium, 598

Hobart, Tasmania. 599

Chevan, A. & Sutherland, M. (1991) Hierarchical Partitioning. American Statistician, 45, 90-600

96. 601

Cottenie, K. (2005) Integrating environmental and spatial processes in ecological community 602

dynamics. Ecology Letters, 8, 1175-1182. 603

Dahlgren, J. P. & Ehrlen, J. (2005) Distribution patterns of vascular plants in lakes - the role of 604

metapopulation dynamics. Ecography, 28, 49-58. 605

Dray, S., Legendre, P. & Peres-Neto, P. R. (2006) Spatial modelling: a comprehensive 606

framework for principal coordinate analysis of neighbour matrices (PCNM). Ecological 607

Modelling, 196, 483-493. 608

Driscoll, D. A. (2004) Extinction and outbreaks accompany fragmentation of a reptile 609

community. Ecological Applications, 14, 220-240. 610

Driscoll, D. A. (2005) Is the matrix a sea? Habitat specificity in a naturally fragmented 611

landscape. Ecological Entomology, 30, 8-16. 612

Driscoll, D. A. (2007) How to find a metapopulation. Canadian Journal of Zoology-Revue 613

Canadienne De Zoologie, 85, 1031-1048. 614

Driscoll, D. A. (2008) The frequency of metapopulations, metacommunities and nestedness in 615

a fragmented landscape. Oikos, 117, 297-309. 616

Driscoll, D. A. & Lindenmayer, D. B. (2009) An empirical test of metacommunity theory 617

using an isolation gradient. Ecological Monographs, 79, 485-501. 618

Driscoll, D. A. & Weir, T. (2005) Beetle responses to habitat fragmentation depend on 619

ecological traits, remnant condition and shape. Conservation Biology, 19, 182-194. 620

26

Eber, S. & Brandl, R. (1996) Metapopulation dynamics of the tephritid fly Urophora cardui: 621

An evaluation of incidence-function model assumptions with field data. Journal of 622

Animal Ecology, 65, 621-630. 623

Elmhagen, B. & Angerbjorn, A. (2001) The applicability of metapopulation theory to large 624

mammals. Oikos, 94, 89-100. 625

Etienne, R. S., Ter Braak, C. J. F. & Vos, C. C. (2004) Application of stochastic patch 626

occupancy models to real metapopulations. Ecology, genetics, and evolution of 627

metapopulations (eds I. Hanski & O. E. Gaggiotti), pp. 105-132. Elsevier, Burlington 628

USA. 629

Fahrig, L. (1997) Relative effects of habitat loss and fragmentation on population extinction. 630

Journal of Wildlife Management, 61, 603-610. 631

Fahrig, L. (2001) How much habitat is enough? Biological Conservation, 100, 65-74. 632

Fahrig, L. (2002) Effect of habitat fragmentation on the extinction threshold: A synthesis. 633

Ecological Applications, 12, 346-353. 634

Fahrig, L. (2003) Effects of habitat fragmentation on biodiversity. Annual Review of Ecology, 635

Evolution and Systematics, 34, 487-515. 636

Fletcher, R. J. (2007) Species interactions and population density mediate the use of social cues 637

for habitat selection. Journal of Animal Ecology, 76, 598-606. 638

Gonzalez-Megias, A., Gomez, J. M. & Sanchez-Pinero, F. (2005) Consequences of spatial 639

autocorrelation for the analysis of metapopulation dynamics. Ecology, 86, 3264-3271. 640

Good, P. (1994) Permutation tests : a practical guide to resampling methods for testing 641

hypotheses, Springer-Verlag, New York. 642

Hanski, I. (1998) Metapopulation dynamics. Nature, 396, 41-49. 643

Hanski, I. (1999) Metapopulation Ecology, Oxford University Press, Oxford. 644

27

Hanski, I. & Gaggiotti, O. E. (2004) Metapopulation biology: past, present, and future. 645

Ecology, genetics, and evolution of metapopulations (eds I. Hanski & O. E. Gaggiotti), 646

pp. 3-22. Elsevier Academic Press, Burlington USA. 647

Hanski, I. & Simberloff, D. (1997) The metapopulation approach, its history, conceptual 648

domain and application to conservation. Metapopulation biology. Ecology, genetics and 649

evolution (eds I. Hanski & M. E. Gilpin), pp. 5-26. Academic Press, San Diego. 650

Harrison, S. (1991) Local extinction in a metapopulation context: an empirical evaluation. 651

Biological Journal of the Linnean Society, 42, 73-88. 652

Harrison, S. & Bruna, E. (1999) Habitat fragmentation and large-scale conservation: what do 653

we know for sure? Ecography, 22, 225-232. 654

Harrison, S., Safford, H. D., Grace, J. B., Viers, J. H. & Davies, K. F. (2006) Regional and 655

local species richness in an insular environment: Serpentine plants in California. 656

Ecological Monographs, 76, 41-56. 657

Hill, J. K., Thomas, C. D. & Lewis, O. T. (1996) Effects of habitat patch size and isolation on 658

dispersal by Hesperia comma butterflies: implications for metapopulation structure. 659

Journal of Animal Ecology, 65, 725-735. 660

Hubbell, S. P. (2001) The unified neutral theory of biodiversity and biogeography, Princeton 661

University Press, Princeton. 662

Hutchinson, G. E. (1957) Concluding Remarks. Cold Spring Harbor Symposium of 663

Quantitative Biology, 22, 415-427. 664

Jackson, W. D. (1968) Fire, air, water and earth - an elemental ecology of Tasmania. 665

Proceedings of the Ecological Society of Australia, 3, 9-16. 666

Jackson, W. D. (1999) Palaeohistory of vegetation change: The last 2 million years. Vegetation 667

of Tasmania. Flora of Australia Supplementary Series No. 8 (eds J. B. Read, R. S. Hill, 668

M. J. Brown & M. J. Hovenden), pp. 64-88. Australian Biological Resources Study, 669

Environment Australia, Canberra. 670

28

Jarman, S., Kantvilas, G. & Brown, M. (1999) Floristic composition of cool temperate 671

rainforest. Vegetation of Tasmania. Flora of Australia Supplementary Series No. 8 (eds 672

J. B. Read, R. S. Hill, M. J. Brown & M. J. Hovenden), pp. 145-159. Australian 673

Biological Resources Study, Environment Australia, Canberra. 674

Jarman, S. J., Brown, M. J. & Kantvilas, G. (1984) Rainforest In Tasmania, National Parks and 675

Wildlife Service, Tasmania. 676

Jellinek, S., Driscoll, D. A. & Kirkpatrick, J. B. (2004) Environmental and vegetation variables 677

have a greater influence than habitat fragmentation in structuring lizard communities in 678

remnant urban bushland. Austral Ecology, 29, 294-304. 679

Kirkpatrick, J. B. (1984) Altitudinal and Successional Variation in the Vegetation of the 680

Northern Part of the West-Coast Range, Tasmania. Australian Journal of Ecology, 9, 681

81-91. 682

Kirkpatrick, J. B. & Dickinson, K. J. M. (1984) Vegetation Map of Tasmania, 1:50 000. 683

Tasmanian Forestry Commission, Hobart. 684

Koch, A. J., Driscoll, D. A. & Kirkpatrick, J. B. (2008) Estimating the error of tree ageing 685

methods in mature Eucalyptus obliqua forest Tasmania. Australian Forestry, 71, 147-686

159. 687

Leibold, M. A., Holyoak, M., Mouquet, N., Amarasekare, P., Chase, J. M., Hoopes, M. F., 688

Holt, R. D., Shurin, J. B., Law, R., Tilman, D., Loreau, M. & Gonzalez, A. (2004) The 689

metacommunity concept: a framework for multi-scale community ecology. Ecology 690

Letters, 7, 601-613. 691

Leigh, E. G. (1981) THE AVERAGE LIFETIME OF A POPULATION IN A VARYING 692

ENVIRONMENT. Journal of Theoretical Biology, 90, 213-239. 693

Levins, R. (1970) Extinction. Some mathematical questions in biology. Lectures on 694

mathematics in life sciences 2. (ed M. Gerstenhaber), pp. 77-107. American 695

Mathematical Society, Providence, R I. 696

29

Lindenmayer, D. B. (2000) Factors at multiple scales affecting distribution patterns and their 697

implications for animal conservation - Leadbeater's Possum as a case study. 698

Biodiversity and Conservation, 9, 15-35. 699

Lindenmayer, D. B., McIntyre, S. & Fischer, J. (2003) Birds in eucalypt and pine forests: 700

landscape alteration and its implications for research models of faunal habitat use. 701

Biological Conservation, 110, 45-53. 702

MacArthur, R. H. & Wilson, E. O. (1967) The theory of island biogeography, Princeton 703

University Press, Princeton. 704

MacDonald, M. A. & Kirkpatrick, J. B. (2003) Explaining bird species composition and 705

richness in eucalypt-dominated remnants in subhumid Tasmania. Journal of 706

Biogeography, 30, 1415-1426. 707

MacNally, R. (1996) Hierarchical partitioning as an interpretative tool in multivariate 708

inference. Australian Journal of Ecology, 21, 224-228. 709

Maindonald, J. & Braun, J. (2007) Data analysis and graphics using R. An example-based 710

approach. Second Edition, Cambridge University Press, Cambridge. 711

Matthies, D., Brauer, I., Maibom, W. & Tscharntke, T. (2004) Population size and the risk of 712

local extinction: empirical evidence from rare plants. Oikos, 105, 481-488. 713

McArdle, B. H. (1990) When are rare species not there? Oikos, 57, 276-277. 714

McCullagh, P. & Nelder, J. A. (1989) Generalized Linear Models (Second Edition), Chapman 715

and Hall, London. 716

Muneepeerakul, R., Bertuzzo, E., Lynch, H., Fagan, W., Rinaldo, A. & Rodriguez-Iturbe, I. 717

(2008) Neutral metacommunity models predict fish diversity patterns in Mississippi–718

Missouri basin. Nature, 453, 220-223. 719

Nieminen, M. & Hanski, I. (1998) Metapopulations of moths on islands: a test of two 720

contrasting models. Journal of Animal Ecology, 67, 149-160. 721

30

Oksanen, J., Kindt, R., Legendre, P. & O'Hara, R. B. (2007) vegan: Community Ecology 722

Package version 1.8-6, http://cran.r-project.org/. 723

Paquin, P. (2008) Carabid beetle (Coleoptera : Carabidae) diversity in the black spruce 724

succession of eastern Canada. Biological Conservation, 141, 261-275. 725

Parris, K. M. (2006) Urban amphibian assemblages as metacommunities. Journal of Animal 726

Ecology, 75, 757-764. 727

Pulliam, H. R. (1988) Sources, sinks and population regulation. American Naturalist, 132, 652-728

661. 729

Quintana-Ascencio, P. F. & Menges, E. S. (1996) Inferring metapopulation dynamics from 730

patch-level incidence of Florida scrub plants. Conservation Biology, 10, 1210-1219. 731

R Development Core Team. (2007) R: A language and environment for statistical computing., 732

R Foundation for Statistical Computing, Vienna, Austria. URL http://www.R-733

project.org. 734

Radford, J. Q. & Bennett, A. F. (2007) The relative importance of landscape properties for 735

woodland birds in agricultural environments. Journal of Applied Ecology, 44, 737-747. 736

Read, J. (1999) Rainforest Ecology. Vegetation of Tasmania. Flora of Australia 737

Supplementary Series No. 8 (eds J. B. Read, R. S. Hill, M. J. Brown & M. J. 738

Hovenden), pp. 160-197. Australian Biological Resources Study, Environment 739

Australia, Canberra. 740

Read, J. (2001) Soil and rainforest composition in Tasmania: correlations of soil characteristics 741

with canopy composition and growth rates in Nothofagus cunninghamii associations. 742

Australian Journal of Botany, 49, 121-135. 743

Richter-Boix, A., Llorente, G. A. & Montori, A. (2007) Structure and dynamics of an 744

amphibian metacommunity in two regions. Journal of Animal Ecology, 76, 607-618. 745

Ryall, K. L. & Fahrig, L. (2006) Response of predators to loss and fragmentation of prey 746

habitat: A review of theory. Ecology, 87, 1086-1093. 747

31

Shmida, A. & Ellner, S. (1984) Coexistence of Plant-Species with Similar Niches. Vegetatio, 748

58, 29-55. 749

Sjogren-Gulve, P. (1994) Distribution and extinction patterns within a northern metapopulation 750

of the pool frog, Rana lessonae. Ecology, 75, 1357-1367. 751

Stelter, C., Reich, M., Grimm, V. & Wissel, C. (1997) Modelling persistence in dynamic 752

landscapes: lessons from a metapopulation of the grasshopper Bryodema tuberculata. 753

Journal of Animal Ecology, 66, 508-518. 754

Taylor, A. D. (1990) Metapopulations, Dispersal, and Predator Prey Dynamics - an Overview. 755

Ecology, 71, 429-433. 756

Thomas, C. D. (1994) Extinction, colonization and metapopulations: environmental tracking 757

by rare species. Conservation Biology, 8, 373-378. 758

Thomas, C. D. & Hanski, I. (1997) Butterfly metapopulations. Metapopulation biology. 759

Ecology, genetics and evolution (eds I. Hanski & M. E. Gilpin), pp. 359-386. Academic 760

Press, San Diego. 761

Thomas, C. D., Singer, M. C. & Boughton, D. A. (1996) Catastrophic extinction of population 762

sources in a butterfly metapopulation. American Naturalist, 148, 957-975. 763

Tilman, D. (1994) Competition and Biodiversity in Spatially Structured Habitats. Ecology, 75, 764

2-16. 765

Tscharntke, T., Steffan-Dewenter, I., Kruess, A. & Thies, C. (2002) Contribution of small 766

habitat fragments to conservation of insect communities of grassland-cropland 767

landscapes. Ecological Applications, 12, 354-363. 768

Turner, I. M. (1996) Species loss in fragments of tropical rain forest: A review of the evidence. 769

Journal of Applied Ecology, 33, 200-209. 770

Verbeylen, G., De Bruyn, L. & Matthysen, E. (2003) Patch occupancy, population density and 771

dynamics in a fragmented red squirrel Sciurus vulgaris population. Ecography, 26, 118-772

128. 773

32

Verheyen, K., Vellend, M., Van Calster, H., Peterken, G. & Hermy, M. (2004) Metapopulation 774

dynamics in changing landscapes: A new spatially realistic model for forest plants. 775

Ecology, 85, 3302-3312. 776

Walker, A. K. & Crosby, T. K. (1988) The preparation and curation of insects, Science 777

Information Publishing Centre (DSIR), Wellington. 778

Werth, S., Wagner, H. H., Gugerli, F., Holderegger, R., Csencsics, D., Kalwij, J. M. & 779

Scheidegger, C. (2006) Quantifying dispersal and establishment limitation in a 780

population of an epiphytic lichen. Ecology, 87, 2037-2046. 781

Whittingham, M. J., Swetnam, R. D., Wilson, J. D., Chamberlain, D. E. & Freckleton, R. P. 782

(2005) Habitat selection by yellowhammers Emberiza citrinella on lowland farmland at 783

two spatial scales: implications for conservation management. Journal of Applied 784

Ecology, 42, 270-280. 785

Wilcox, C., Cairns, B. J. & Possingham, H. P. (2006) The role of habitat disturbance and 786

recovery in metapopulation persistence. Ecology, 87, 855-863. 787

Yu, D. W., Wilson, H. B., Frederickson, M. E., Palomino, W., De la Colina, R., Edwards, D. P. 788

& Balareso, A. A. (2004) Experimental demonstration of species coexistence enabled 789

by dispersal limitation. Journal of Animal Ecology, 73, 1102-1114. 790

791

792

33

Supporting Information 793

794

The following Supporting Information is available for this article online: 795

Appendix S1. Trophic Level, Size and Flight analyses. MS Word 97-2003 796

Appendix S2. Beetle species occurrence and summary of responses to geographic and 797

environmental variables for 44 common beetle species. Tab delimited text. 798

Appendix S3. Justification for excluding site B13 as an ecological outlier. MS Word 97-2003 799

Appendix S4. Hierarchical partitioning results. Tab delimited text. 800

Appendix S5. Generalised linear model results at the patch level. Tab delimited text. 801

Appendix S6. Generalised linear model results at the grid level. Tab delimited text. 802

803

34

804

Table 1. Examples of studies that have assessed the prevalence of metapopulation dynamics. 805

Proportion of community that

may have a metapopulation

Study System Reference

15% of plant species

75% of specialist plants

Rosemary scrub in Florida USA (Quintana-Ascencio and

Menges, 1996)

37% of forest-specialist plant

species

regenerating English forests (Verheyen et al., 2004)

58-66% of moths

rocky islands in Finland (Nieminen and Hanski, 1998)

54% of common water plants

Swedish lakes (Dahlgren and Ehrlen, 2005)

7.5% of beetle species Eucalyptus-dominated fragments

in Tasmania, Australia

(Driscoll, 2008)

806

35

Figure 1. Each of thirty one rainforest patches (10 in region 1 from the north west, 21 in region 807

2 from the south east) were sampled using 48 pairs of pit-fall traps in South-West Tasmania, 808

Australia. Hashed area = hydro-electric dams, dark grey = forest, light grey = sedgeland, 809

dashed line = road, white circles = sampled patch. The white arrow in the top centre of the 810

right hand panel indicates the outlier population (B13). 811

812

Figure 2. Summary of the number of species showing support for each of the theories. The 813

boxes for each row indicate the evidence supporting each theory and the number of species 814

showing that evidence. Species with higher frequency on the least isolated sites in the 815

metapopulation row do not support a metapopulation, but imply that a similar process of 816

dispersal limitation without population turnover. Double headed arrows indicate where one 817

species had more than one response. Species with occurrence and frequency showing the same 818

response are only counted under occurrence and the arrow links are omitted. 819

820

36

821

822

Location

HOBART

TASMANIA

37

823

Classic and source-sink metapopulations

Occurrence higher on least isolated patches

0

Occurrence higher on largest patches

1

Frequency higher on least isolated patches

3

Deterministic metapopulations

Higher occurrence in Rainforest

6

Higher occurrence on low nutrients

4

Habitat-amount Higher frequency with more forest

7

Higher occurrence with more forest

2

Dominant-species / dispersal tradeoffs

Higher occurrence in small patches

2

Higher occurrence in more isolated patches

1

Niche theory, patch level

Highest frequency in Rainforest

4

Highest frequency on low nutrients

7 on high nutrients

1

Highest frequency in eucalypts

1

Niche theory, grid level

Highest frequency in Rainforest

10

Highest frequency on low nutrients

1

Highest frequency in eucalypts

3

No significant response

At patch level

14 (6 without PCNMs)

At patch and grid level

8 (2 without PCNMs)

824