Appendix A. Supplementary Figure.

1

Figure A1. Relationship between algal volume gently packed into a graduated cylinder and algal wet biomass used to estimate algal addition biomass in the field. Shown here for two species, A) Ulva sp and B) Ectocarpus. For pooled algae including Ulva lactuca, Ectocarpus, Gracilaria verrucosa and Bryopsis sp., the relationship between biomass and volume is log(wt) = 1.07*log(Vol)- 0.87 (r2=0.87, p < 0.001), and the relationship between biomass and percent cover of 0.5 m2 is log(wt) = 1.07*log(cover) + 0.82 (r2 = 0.94, p < 0.001).

Appendix B. Supplementary Statistical Results.

Table A1. Statistical results for tests of cage control treatment versus open treatment (one-way anova).

Response / MS / df / F / p
Spartina density / 28.125 / 1 / 0.096 / 0.768
Error / 294.625 / 6
Proportion flowering / 0.007 / 1 / 0.184 / 0.682
Error / 0.036 / 6
Spartina shoot height / 2.679 / 1 / 0.020 / 0.895
Error / 138.218 / 5
Above ground biomass / 4163.74 / 1 / 0.604 / 0.467
Error / 6892.31 / 6
Below ground biomass / 2.565 / 1 / 0.136 / 0.725
Error / 18.912 / 6
Shoot weight / 2.639 / 1 / 1.672 / 0.244
Error / 1.578 / 6

Tables A2-A6. Statistical results from model comparisons for surveys of benthic primary producer abundance at restored and natural sites through time. Models were compared using Akaike Information Criteria adjusted for small sample sizes (AICC) and evidence ratios calculated as in Burnham and Anderson (2002). Best models have the lowest AICc value, and the weight of evidence in support of this conclusion is indicated by the evidence ratio with 1 indicating complete support and 0 indicating no support. Models are presented relative to the full model (Eqn 1), and for each comparison a factor is removed in a logical order.

Table A2.Model results for mean algal abundance (percent cover) from models fit with a Guassian distribution on arcsine square-root transformed percent cover data. The row in bold indicates the best model (highest evidence ratio).Factors included in the model are indicated in the Model column, and RE: indicates random effects.

Model / LogLikelihood / K / AICc / δ / ωi / Ev ratio
Full model (Eqn 1): Sill, Time and RE: Sill | Transect within Site. / 369.5 / 6 / -726 / 35 / 0 / 0
Sill, Time and RE: Transect within Site. / 366.3 / 6 / -720 / 35 / 0 / 0
Sill, Time and RE: Transect / 365.4 / 5 / -720 / 35 / 0 / 0
Sill and RE: Transect / 381.6 / 4 / -755 / 0 / 1 / 1
RE: Transect / 355.4 / 3 / -705 / 50 / 0 / 0

Table A3.Macroalgal taxonomic richness model comparison results. Evidence ratios for two models are small (unconvincing) and similar, so these models are both considered plausible.

Model / LogLikelihood / K / AICc / δ / ωi / Ev ratio
Full model (Eqn 1): Sill, Time and RE: Sill | Transect within Site. / -109.0 / 6 / 231 / 0 / 1 / 0.77
Sill, Time and RE: Transect within Site. / -115.5 / 6 / 244 / 13.00 / 0.00 / 1.15 x 10-3
Sill, Time and RE: Sill | Transect. / -111.7 / 5 / 234 / 3.25 / 0.20 / 0.15
Sill and RE: Sill | Transect within Site. / -112.3 / 5 / 235 / 4.45 / 0.11 / 0.08
RE: Transect / -141.6 / 3 / 289 / 45.84 / 0.00 / 8.52 x 10-11

Table A4. Microalgal abundance model comparison results.

Model / LogLikelihood / K / AICc / δ / ωi / Ev ratio
Full model (Eqn 1): Sill, Time and RE: Sill | Transect within Site. / -899.4 / 6 / 1811 / 8.75 / 0.01 / 0.01
Sill, Time and RE: Transect within Site. / -901.1 / 6 / 1815 / 12.15 / 2.23 x 10-3 / 2.22 x 10-3
Sill, Time and RE: Sill | Transect. / -910.1 / 5 / 1831 / 28.00 / 8.32 x 10-7 / 8.07 x 10-7
Sill and RE: Sill | Transect within Site. / -896.1 / 5 / 1803 / 0 / 1 / 0.97
RE: Transect within Site / -901.4 / 4 / 1811 / 8.48 / 0.01 / 8.52 x 10-11

Table A5.Benthic algal primary production (macroalgae + microalgae abundance).

Model / LogLikelihood / K / AICc / δ / ωi / Ev ratio
Full model (Eqn 1): Sill, Time and RE: Sill | Transect within Site. / -975.0 / 6 / 1963 / 0 / 1 / 0.89
Sill, Time and RE: Transect within Site. / -979.6 / 6 / 1972 / 9.2 / 0.01 / 8.98 x 10-3
Sill, Time and RE: Sill | Transect. / -979.2 / 5 / 1969 / 6.25 / 0.04 / 0.04
Sill and RE: Sill | Transect within Site. / -978.8 / 5 / 1968 / 5.44 / 0.7 / 0.06
RE: Transect within Site / -999.9 / 4 / 2008 / 45.5 / 1.30 x 10-10 / 1.16 x 10-10

Table A6. Marsh cordgrass Spartina alterniflora abundance. Evidence ratios for two models are small (unconvincing) and similar, so these models are both considered plausible.

Model / LogLikelihood / K / AICc / δ / ωi / Ev ratio
Full model (Eqn 1): Sill, Time and RE: Sill | Transect within Site. / -15.97 / 6 / 44.44 / 24.4 / 4.96 x 10-6 / 2.65 x 10-6
Sill, Time and RE: Transect within Site. / -17.04 / 6 / 46.58 / 26.57 / 1.70 x 10-6 / 9.12 x 10-7
Sill, Time and RE: Sill | Transect. / -13.73 / 5 / 37.82 / 17.8 / 1.36 x 10-4 / 7.30 x 10-5
Sill and RE: Sill |Transect. / -7.47 / 4 / 23.18 / 3.16 / 0.21 / 0.11
RE: Sill | Transect within Site. / -5.89 / 4 / 20.02 / 0 / 1 / 0.54
RE: Transect within Site / -7.35 / 3 / 20.82 / 0.82 / 0.66 / 0.35

1