Table 1.

Results of the multiple regression on matrices analysis by spatial scale

All scalesWithin marshRegionalContinental
R2 = 0.46 bR2 = 0.27 bR2 = 0.61 bR2 = 0.17 b
Ln (geographic distance)0.40***0.47***
Ln (plant similarity + 0.005)
Ln (100 – % sediment moisture)0.17**0.63**0.19*0.14
Ln (ammonium +1)
Ln (nitrate +1)0.100.17*
Ln (phosphate +1)0.11
pH
% plant cover0.13*0.34***n.s.
Salinity0.100.16
Water temperature0.09na0.45***
Air temperature0.14**na0.24*0.16*
  • The variation (R2) of ln community distance (distance = 1 – similarity) that is explained by the remaining variables and the partial regression coefficients (b) of the final model is reported (all values are significant at P ≤ 0.0001, n = 62). If a partial regression coefficient is reported, its significance level (one-way tests) is < 0.0500. *P ≤ 0.0100, **P ≤ 0.0010, and ***P ≤ 0.0001. Water and air temperature were measured for each marsh, and therefore are not included in the within-marsh models.