Ecological Archives E091-030-A3

Jana S. Petermann, Alexander J. F. Fergus, Christiane Roscher, Lindsay A. Turnbull, Alexandra Weigelt, and Bernhard Schmid. 2010. Biology, chance, or history? The predictable reassembly of temperate grassland communities. Ecology 91:408–421.

Appendix C. One table showing the experimental design and six tables providing the results of statistical analyses.

TABLE C1. Experimental design. Original (sown) species richnesss and functional-group composition of resident communities in the Jena Experiment.

Species richnesss

1

1

1

1

2

2

2

2

2

2

2

2

4

4

4

4

4

4

4

4

4

4

4

4

4


grasses

1

     

2

     

1

 

1

 

4

     

2

 

2

 

2

1

1

 

1

small herbs  

1

     

2

   

1

   

1

 

4

   

2

   

2

1

 

1

2

1

tall herbs    

1

     

2

   

1

1

     

4

   

2

2

 

1

2

 

1

1

legumes      

1

     

2

 

1

 

1

     

4

 

2

 

2

 

1

2

1

1


Replicates

4

4

4

4

2

2

2

2

2

2

2

2

1

1

1

1

1

1

1

1

1

1

1

1

4


 
Species richnesss

8

8

8

8

8

8

8

8

8

8

8

8

8

16

16

16

16

16

16

16

16

16

16

16

60


grasses

8

     

4

 

4

 

2

3

3

 

2

16

 

8

 

8

 

5

5

6

 

4

16

small herbs  

8

   

4

   

4

3

 

3

2

2

   

8

   

8

6

 

5

6

4

12

tall herbs    

8

   

4

4

 

3

2

 

3

2

 

16

 

8

8

 

6

5

 

5

4

20

legumes      

8

 

4

 

4

 

3

2

3

2

     

8

 

8

 

6

5

5

4

12


Replicates

1

1

1

1

1

1

1

1

1

1

1

1

4

1

1

1

1

1

1

1

1

1

1

4

4


TABLE C2. Analysis of variance (ANOVA) of the number of species and the biomass (gm-2) of internal invaders per functional group per harvest quadrat. Data exclude the 60-species level, because it always contains all four functional groups. The inclusion of resident biomass as a covariable did not change the significance of the results and was omitted from the model. "Species richnesss" stands for the log2-transformed sown species richnesss of the resident community. The deviation of the species richnesss effect from log-linearity was not significant and was omitted. The results were relatively robust to the order of the terms species richnesss and presence of particular functional groups, so only the results from the model with species richnesss tested first are shown. The presence of the four functional groups was tested in the order of their explanatory power in the model. Two- and three-way interactions of functional group effects were small and therefore were omitted from the model. The weeding and seed-addition treatments (c-, c+, and w+) were used to form two contrasts. Only the first of them ("Seed addition"), representing seed addition (c+ and w+) vs. no seed-addition treatments (c-), was included in the model (the contrast between w+ and c+ was not significant). The "Home-away contrast", which represents the main contrast within the total "Invader × resident FG" interactions was tested against its deviation ("Other invader-resident FG interactions"). Other error terms are printed in italics. FG = functional group.

Source

Number of species

Biomass

df

SS

F

P

df

SS

F

P

Spatial variation

9

51.1

3.01

0.005

9

619241

2.70

0.010

Species richnesss

1

151.3

80.23

<0.001

1

815930

32.03

<0.001

Legume presence

1

83.7

44.38

<0.001

1

302853

11.89

0.001

Tall herb presence

1

18.7

9.89

0.003

1

85362

3.35

0.072

Grass presence

1

0.1

0.05

0.827

1

31173

1.22

0.273

Small herb presence

1

0.2

0.09

0.771

1

8594

0.34

0.563

Plot

63

118.8

1.57

0.010

63

1604762

0.75

0.907

Invader FG

3

154.1

42.67

<0.001

3

1602127

15.77

<0.001

Home-away contrast

1

145.8

37.94

<0.001

1

2071215

6.50

0.027

Other invader-resident FG interactions

11

42.3

3.19

<0.001

11

3506087

9.41

<0.001

Species richnesss × Home-away contrast

1

4.5

1.77

0.205

1

15375

0.14

0.710

Species richnesss × Other invader-resident FG interactions

14

35.2

2.09

0.014

14

1493090

3.15

<0.001

Plot × Invader FG

201

242.0

2.88

<0.001

202

6841052

3.03

<0.001

Seed addition

1

75.4

180.62

<0.001

1

93776

8.40

0.004

Species richnesss × Seed addition

1

19.8

47.44

<0.001

1

15019

1.35

0.246

Legume presence × Seed addition

1

19.6

46.91

<0.001

1

3215

0.29

0.591

Tall herb presence × Seed addition

1

1.1

2.66

0.104

1

1568

0.14

0.708

Grass presence × Seed addition

1

0.1

0.20

0.652

1

15890

1.42

0.233

Small herb presence × Seed addition

1

0.0

0.09

0.770

1

62645

5.61

0.018

Invader FG × Seed addition

3

13.9

11.07

<0.001

3

181883

5.43

<0.001

Home-away contrast × Seed addition

1

13.6

20.41

<0.001

1

119186

6.65

0.026

Other invader-resident FG interactions × Seed addition

11

7.3

1.59

0.097

11

197138

1.61

0.093

Plot × Invader FG × Subplot

596

248.8

0.57

1.000

595

6639929

0.75

1.000

Year

1

580.2

791.58

<0.001

1

2289985

154.69

<0.001

Plot × Invader FG × Subplot × Year

901

660.4

2.49

<0.001

 

 

 

 

Residual

944

278.3

 

 

921

13634455

0.65

1.000


TABLE C3. Analysis of variance (ANOVA) of the number of species of external invaders per harvest quadrat. Data exclude the 60-species level. The inclusion of resident biomass as a covariable did not change the significance of the results and was omitted from the model. "Species richnesss" stands for the log2-transformed sown species-richnesss of the resident community. The deviation of the species richnesss effect from log-linearity was not significant and was omitted. The results were relatively robust to the order of the terms species richnesss and presence of particular functional groups, so only the results from the model with species richnesss tested first are shown. The presence of the four functional groups was tested in the order of their explanatory power in the model. Two- and three-way interactions of functional group effects were small and were therefore omitted from the model. All error terms are printed in italics.

 

df

SS

F

P

Spatial variation

9

10.3663

0.88

0.549

Species richnesss

1

28.9082

22.03

<0.001

Legume presence

1

10.9671

8.36

0.005

Small herb presence

1

6.0697

4.63

0.035

Tall herb presence

1

0.0013

0.00

0.975

Grass presence

1

0.2808

0.21

0.645

Plot

67

87.9229

4.12

<0.001

Seed addition

1

0.823

2.58

0.112

Species richnesss × Seed addition

1

0.6031

1.89

0.173

Legume presence × Seed addition

1

0.0001

0.00

0.986

Small herb presence × Seed addition

1

1.2136

3.81

0.055

Tall herb presence × Seed addition

1

0.7482

2.35

0.130

Grass presence × Seed addition

1

0.044

0.14

0.711

Plot × Subplot

76

24.2346

1.18

0.190

Year

1

2.686

9.97

0.002

Species richnesss × Year

1

1.0648

3.95

0.049

Legume presence × Year

1

2.732

10.14

0.002

Small herb presence × Year

1

2.3659

8.78

0.004

Tall herb presence × Year

1

0.0043

0.02

0.900

Grass presence × Year

1

0.3525

1.31

0.255

Seed addition × Year

1

0.1736

0.64

0.423

Plot × Subplot Year

155

41.7673

0.91

0.713

Residual

160

47.1871

 

 


TABLE C4. Analysis of variance (ANOVA) of the biomass (gm-2) of external invaders per harvest quadrat. Data exclude the 60-species level. The inclusion of resident biomass as a covariable did not change the significance of the results and was omitted from the model. "Species richnesss" stands for the log2-transformed sown species richnesss of the resident community. The deviation of the species-richnesss effect from log-linearity was not significant and was omitted. The results were relatively robust to the order of the terms species richnesss and presence of particular functional groups, so only the results from the model with species richnesss tested first are shown. The presence of the four functional groups was tested in the order of their explanatory power in the model. Two- and three-way interactions of functional group effects were small and were therefore omitted from the model. All error terms are printed in italics.

 

df

SS

F

P

Spatial variation

9

42165

1.66

0.115

Species richnesss

1

38314

13.61

<0.001

Grass presence

1

4248

1.51

0.224

Small herb presence

1

3561

1.26

0.265

Legume presence

1

4541

1.61

0.209

Tall herb presence

1

0

0.00

1.000

Plot

67

188639

1.58

0.027

Seed addition

1

1522

0.85

0.359

Species richnesss × Seed addition

1

1535

0.86

0.357

Grass presence × Seed addition

1

31

0.02

0.896

Small herb presence × Seed addition

1

570

0.32

0.574

Legume presence × Seed addition

1

763

0.43

0.515

Tall herb presence × Seed addition

1

4455

2.49

0.118

Plot × Subplot

76

135726

2.61

<0.001

Year

1

5827

8.53

0.004

Species richnesss × Year

1

3751

5.49

0.020

Grass presence × Year

1

230

0.34

0.563

Small herb presence × Year

1

634

0.93

0.337

Legume presence × Year

1

39

0.06

0.811

Tall herb presence × Year

1

2636

3.86

0.051

Seed addition × Year

1

778

1.14

0.288

Plot × Subplot Year

155

105942

0.31

1

Residual

160

353360

 

 


TABLE C5. Analysis of variance (ANOVA) of the total number of species per harvest quadrat and of community biomass (gm-2). Target species were residents in weeded controls (w-), but residents, external and internal invaders in non-weeded subplots (c- and c+). "Species richnesss" is the sown species richnesss of the resident community. The weeding and seed-addition treatments (w-, c- and c+) were used to form two contrasts. The first of them ("Invasion") represents weeded controls (w-) vs. invasion treatments (c- and c+), the second ("Seed addition") represents non-weeded treatments without seed addition (c-) vs. with seed addition (c+). Error terms are printed in italics.

Source

Number of species

Biomass

df

SS

F

P

df

SS

F

P

Spatial variation

9

1100.7

9.05

<0.001

9

4440201

1.16

0.333

Species richnesss (log2)

1

3655.3

270.36

<0.001

1

12139477

28.63

<0.001

Species richnesss (deviation from log-linear)

4

1206.8

22.32

<0.001

4

1651207

0.97

0.428

Plot

67

905.8

3.28

<0.001

67

28408418

5.88

<0.001

Invasion

1

2915.7

706.83

<0.001

1

5018898

69.59

<0.001

Seed addition

1

203.5

49.34

<0.001

1

251564

3.49

0.064

Species richnesss (log2) × Invasion

1

289.2

70.10

<0.001

1

1080842

14.99

<0.001

Species richnesss (deviation from log-linear) × Invasion

4

0.8

0.05

0.995

4

445103

1.54

0.193

Species richnesss (log2) × Seed addition

1

40.2

9.75

0.002

1

65241

0.90

0.343

Species richnesss (deviation from log-linear) × Seed addition

4

20.2

1.22

0.303

4

256648

0.89

0.472

Plot × Subplot

150

618.8

0.99

0.522

150

10818315

0.75

0.970

Year

1

663.5

159.32

<0.001

1

8354184

87.12

<0.001

Species richnesss (log2) × Year

1

246.6

59.20

<0.001

1

38361

0.40

0.528

Species richnesss (deviation from log-linear) × Year

4

36.3

2.18

0.072

4

1745162

4.55

0.001

 Invasion × Year

1

357.2

85.75

<0.001

1

1586364

16.54

<0.001

Seed addition × Year

1

114.4

27.46

<0.001

1

61793

0.64

0.423

Species richnesss (log2) × Invasion × Year

1

87.4

20.98

<0.001

1

59669

0.62

0.431

Species richnesss (deviation from log-linear) × Invasion × Year

4

7.2

0.43

0.787

4

203245

0.53

0.714

Species richnesss (log2) × Seed addition × Year

1

12.6

3.02

0.084

1

87566

0.91

0.340

Species richnesss (deviation from log-linear) × Seed addition × Year

4

5.2

0.31

0.868

4

291002

0.76

0.553

Plot × Subplot × Year

224

932.9

2.13

<0.001

224

21480355

1.41

0.005

Residual

242

472.5

 

 

242

16501297

 

 


TABLE C6. Analysis of variance (ANOVA) of community biomass, including the effect of observed (=realized) species richnesss. Target species were residents in weeded controls (w-), but residents, external and internal invaders in non-weeded subplots (c- and c+). The influence of "realized species richnesss (log2-transformed)" on community biomass was tested against the interaction "Plot × Subplot × Year". The weeding and seed-addition treatments (w-, c-, and c+) were used to form two contrasts. The first of them ("Invasion") represents weeded controls (w-) vs. non-weeded treatments (c- and c+), the second ("Seed addition") represents non-weeded treatments without seed addition (c-) vs. with seed addition (c+). Error terms are printed in italics.

Source

df

SS

F

P

Spatial variation

9

4085251

6.00

0.307

realized richness (log2)

1

15619722

160.67

<0.001

Plot

72

36869246

6.83

<0.001

Invasion

1

75710

1.01

0.316

Seed addition

1

27015

0.36

0.549

realized richness (log2) × Invasion

1

557075

5.73

0.017

realized richness (log2) × Seed addition

1

658371

6.77

0.010

Plot × Subplot

160

11987260

0.77

0.962

Year

1

3874550

39.85

<0.001

realized richnesss (log2) × Year

1

1850417

19.03

<0.001

Invasion × Year

1

1347

0.01

0.906

Seed addition × Year

1

15993

0.16

0.685

realized richness (log2) × Invasion × Year

1

563858

5.80

0.017

realized richness (log2) × Seed addition × Year

1

85679

0.88

0.349

Plot × Subplot × Year

239

23234778

1.55

<0.001

Residual

233

14571877

 

 


TABLE C7. Analysis of variance (ANOVA) of the community biomass, including the effect of observed (=realized) functional-group proportions. In this analysis, external invaders were excluded because they could not be grouped into the respective functional groups, so target species in this case were residents in weeded controls (w-), but residents and internal invaders in non-weeded subplots (c- and c+). Three outliers with very high biomass were excluded. The realized functional-group proportions were included in the model in the order of their explanatory power and their influence was tested against the "Residual". The weeding and seed-addition treatments (w-, c-, and c+) were used to form two contrasts. The first of them ("Invasion") represents weeded controls (w-) vs. non-weeded treatments (c- and c+), the second ("Seed addition") represents non-weeded treatments without seed addition (c-) vs. with seed addition (c+). Error terms are printed in italics.

Source

df

SS

F

P

Spatial variation

9

2609852

2.10

0.041

Proportion of legumes

1

4847743

80.23

<0.001

Proportion of small herbs

1

1161391

19.22

<0.001

Proportion of grasses

1

1483

0.02

0.876

Plot

72

9961269

2.29

<0.001

Invasion

1

3875505

64.14

<0.001

Seed addition

1

142731

2.36

0.126

Proportion of legumes × Invasion

1

924491

15.30

<0.001

Proportion of small herbs × Invasion

1

41616

0.69

0.408

Proportion of grasses × Invasion

1

256598

4.25

0.041

Proportion of legumes × Seed addition

1

39499

0.65

0.420

Proportion of small herbs × Seed addition

1

9209

0.15

0.697

Proportion of grasses × Seed addition

1

410

0.01

0.934

Residual

147

8882268

 

 


[Back to E091-030]