Appendix A. ANOVAs comparing the number of new settlers observed on each treatment during the experiment.
Table A1. ANOVAs comparing the number of new settlers observed on each treatment during the 50-d experiment for (A) all damselfish species excluding Neopomacentrus cyanomos and (B) N. cyanomos (Pomacentridae), (C) butterflyfish (Chaetodontidae), (D) surgeonfish (Acanthuridae), and (E) rabbitfish (Siganidae).
Source
df
SS
MS
F
P
Levene's
F test†P
A) Damselfish
1.30
0.330
Block
4
46.30
11.58
0.33
0.853
Piscivores
1
5379.20
5379.20
153.22
<0.0001
Damselfish
1
561.80
561.80
16.00
0.002
P × D
1
423.20
423.20
12.05
0.005
Error
12
421.30
35.11
B) N. cyanomos‡
2.20
0.110
Block
4
11.38
2.85
2.82
0.073
Piscivores
1
2.03
2.03
2.02
0.181
Damselfish
1
0.78
0.78
0.77
0.396
P × D
1
0.003
0.003
0.003
0.955
Error
12
12.07
1.01
C) Butterflyfish‡
1.00
0.475
Block
4
0.20
0.05
0.56
0.698
Piscivores
1
21.69
21.69
236.02
<0.0001
Damselfish
1
0.05
0.05
0.59
0.457
P × D
1
0.05
0.05
0.59
0.457
Error
12
1.10
0.09
D) Surgeonfish
Block
4
31.50
7.88
1.17
0.373
0.72
0.657
Piscivores
1
168.20
168.20
24.95
0.0003
Damselfish
1
0.20
0.20
0.03
0.866
P × D
1
3.20
3.20
0.47
0.504
Error
12
80.90
6.74
E) Rabbitfish
2.15
0.117
Block
4
7.20
1.80
0.66
0.632
Piscivores
1
76.05
76.05
27.82
0.0002
Damselfish
1
14.45
14.45
5.29
0.040
P × D
1
14.45
14.45
5.29
0.040
Error
12
32.80
2.73
† Levene's F test tests the assumption of equal variance among treatments. P > 0.05 indicates this assumption has been met.
‡ Response (X) was ln(X + 1) transformed to correct unequal variance.