Appendix D. Analyses based on Gower and Jaccard dissimilarity measures.
To assess the relative importance of compositional changes vs. differences in relative abundance of taxa in driving the response of assemblages to the addition of sediment, data were reanalyzed on the basis of a modified version of the Gower’s dissimilarity measure, as described in Anderson et al. (2006). The Gower measure of dissimilarity was obtained after transforming the data to x' = log2(x) + 1 for x > 0 and x' = 0 for x = 0, and excluding double zeros. This index weights a compositional change equal to a doubling in abundance of species. Analyses were also done using the Jaccard index (Jaccard 1900), which emphasizes compositional change (Anderson et al. 2006).
TABLE D1. Permutational multivariate analysis of variance (PERMANOVA) based on Gower and Jaccard measures of dissimilarity examining effects of sediment on mature assemblages.
Gower |
Jaccard |
||||||||||||||||
Source |
df |
MS |
Pseudo F |
P |
MS |
Pseudo F |
P |
Denominator |
No. permutable |
||||||||
Habitat = Ha |
1 |
4469.70 |
4.02 |
0.004 |
4639.70 |
2.84 |
0.062 |
Ha × Si(Co) |
12 |
||||||||
Condition = Co |
2 |
6792.00 |
8.96 |
0.000 |
6031.70 |
7.86 |
0.001 |
Si(Co) |
6 |
||||||||
C vs. T |
1 |
8471.00 |
9.63 |
0.001 |
4362.32 |
5.05 |
0.031 |
Si(Co) |
6 |
||||||||
T vs. R |
1 |
1530.60 |
2.79 |
0.052 |
3872.81 |
7.60 |
0.008 |
Si(Co) |
6 |
||||||||
Site(Co) = Si(Co) |
3 |
758.14 |
1.56 |
0.063 |
767.11 |
0.96 |
0.498 |
Residual |
36 |
||||||||
Ha × Co |
2 |
2091.50 |
1.88 |
0.077 |
2097.01 |
1.28 |
0.324 |
Ha × Si(Co) |
12 |
||||||||
Ha × C vs. T |
1 |
2448.00 |
2.45 |
0.079 |
3096.12 |
1.87 |
0.217 |
Ha × Si(Co) |
12 |
||||||||
Ha × T vs. R |
1 |
1099.40 |
0.76 |
0.622 |
724.10 |
0.35 |
0.839 |
Ha × Si(Co) |
12 |
||||||||
Ha × Si(Co) |
3 |
1111.00 |
2.29 |
0.002 |
1631.70 |
2.04 |
0.029 |
Residual |
36 |
||||||||
Residual |
24 |
485.36 |
797.44 |
||||||||||||||
Total |
35 |
||||||||||||||||
TABLE D2. Permutational multivariate analysis of variance (PERMANOVA) based on Gower and Jaccard measures of dissimilarity examining effects of sediment on recovering assemblages.
Gower |
Jaccard |
||||||||||||||||||||||
Source |
df |
MS |
Pseudo F |
P |
MS |
Pseudo F |
P |
Denominator |
No. permutable |
||||||||||||||
Habitat = Ha |
1 |
3870.50 |
6.19 |
0.004 |
9422.60 |
7.71 |
0.007 |
Ha x Si(Co) |
12 |
||||||||||||||
Condition = Co |
2 |
5927.80 |
5.25 |
0.003 |
4771.80 |
5.88 |
0.010 |
Si(Co) |
6 |
||||||||||||||
C vs. T |
1 |
6385.90 |
6.15 |
0.008 |
3148.70 |
4.42 |
0.048 |
Si(Co) |
6 |
||||||||||||||
T vs. R |
1 |
1084.40 |
1.21 |
0.359 |
2117.00 |
2.42 |
0.164 |
Si(Co) |
6 |
||||||||||||||
Site(Co) = Si(Co) |
3 |
1129.20 |
2.84 |
0.001 |
811.08 |
1.26 |
0.274 |
Residual |
36 |
||||||||||||||
Ha × Co |
2 |
1191.00 |
1.90 |
0.115 |
1342.70 |
1.10 |
0.440 |
Ha x Si(Co) |
12 |
||||||||||||||
Ha × C vs. T |
1 |
1217.00 |
1.62 |
0.240 |
1575.50 |
0.99 |
0.454 |
Ha × Si(Co) |
12 |
||||||||||||||
Ha × T vs. R |
1 |
422.27 |
0.99 |
0.479 |
374.42 |
0.28 |
0.869 |
Ha × Si(Co) |
12 |
||||||||||||||
Ha × Si(Co) |
3 |
625.50 |
1.57 |
0.095 |
1221.40 |
1.90 |
0.061 |
Residual |
36 |
||||||||||||||
Residual |
24 |
398.20 |
642.47 |
||||||||||||||||||||
Total |
35 |
||||||||||||||||||||||
|
|
|
|
| FIG. D1. Two-factor nonmetric MDS plots of centroids obtained from Gower (A) and Jaccard (B) dissimilarities showing differences in mature assemblages on vertical (triangles) and horizontal (squares) habitats among control (hollow symbols), treated (gray symbols), and reference (black symbols) conditions. |
|
|
| FIG. D2. Two-factor nonmetric MDS plots of centroids obtained from Gower (A) and Jaccard (B) dissimilarities showing differences in recovering assemblages on vertical (triangles) and horizontal (squares) habitats among control (hollow symbols), treated (gray symbols), and reference (black symbols) conditions. |
LITERATURE CITED
Anderson, M. J., K. E. Ellingsen, and B. H. McArdle. 2006. Multivariate dispersion as a measure of beta diversity. Ecology Letters 9:683–693.
Jaccard, P. 1900. Contribution au problème de l’immigration post-glaciare de la flore alpine. Bulletin de la Société Vaudoise des Sciences Naturelles 36: 87–130.