Skip to main content
. 2012 Jun 29;7(6):e39571. doi: 10.1371/journal.pone.0039571

Table 3. Results of direct gradient ordination analyses performed on the data.

Predictors data F-ratio p-value Trace
treatment × sampling time interaction species’ covers (CCA) 1.739 0.052 0.029
treatment × sampling time interaction (removal vs. invaded plots) species’ covers (CCA) 2.417 0.074 0.024
treatment × sampling time interaction (removal vs. adjacent uninvaded plots) species’ covers (CCA) 1,45 0.368 0.017
treatment × sampling time interaction (invaded vs. adjacent uninvaded plots) species’ covers (CCA) 2.208 0.002 0.027
treatment × sampling time interaction binnary data (RDA) 1.523 0.034 0.002
treatment × sampling time interaction (removal vs. invaded plots) binnary data (RDA) 1.452 0.116 0.002
treatment × sampling time interaction (removal vs. adjacent uninvaded plots) binnary data (RDA) 0.772 0.9380 0.001
treatment × sampling time interaction (invaded vs. adjacent uninvaded plots) binnary data (RDA) 2.386 0.004 0.003
treatment (main effect) species’ covers (CCA) 1.375 0.07 0.309
treatment (main effect - removal vs. invaded plots) species’ covers (CCA) 1.077 0.370 0.135
treatment (main effect - removal adjacent uninvaded plots) species’ covers (CCA) 1.740 0.034 0.260
treatment (main effect - invaded vs. adjacent uninvaded plots) species’ covers (CCA) 1.941 0.004 0.304
treatment (main effect) binnary data (RDA) 1.139 0.2280 0.034
treatment (main effect - removal vs. invaded plots) binnary data (RDA) 1.201 0.2360 0.027
treatment (main effect - removal vs. adjacent uninvaded plots) binnary data (RDA) 0.824 0.742 0.019
treatment (main effect - invaded vs. adjacent uninvaded plots) binnary data (RDA) 1.523 0.05 0.032

For each of the analyses, the table shows the tested term (main effect or interaction), type of data (species’ covers estimates or binary presence – absence), results of test statistics (F-ratio and p-value) and Trace. Trace is, in this case, a sum of canonical eigenvalues of the model and represents the ordination model’s explanatory power.