Table 1.

Assemblage similarity between continuous forest and modified habitats (fragments and secondary forest), ~15 years and ~30 years after experimental forest clearance. We defined the similarity between two assemblages as the correlation between model-predicted occurrence probabilities or abundances (both log-transformed). The values in the table show posterior mean similarities between assemblages inhabiting continuous forests and modified habitats for the two study periods, as well as the posterior probability by which the similarities were lower in the first period than in the second.

Generalists	Fragments		Secondary Forest
	Occupancy	Abundance	Occupancy	Abundance
1996–2002	0.64	0.71	0.60	0.68
2011–2013	0.76	0.79	0.51	0.56
Posterior probability	0.8	0.74	0.21	0.31
Specialists
1996–2002	0.84	0.85	0.78	0.78
2011–2013	0.74	0.76	0.61	0.63
Posterior probability	0.16	0.16	0.09	0.09