SEARCH

SEARCH BY CITATION

FilenameFormatSizeDescription
gcb12627-sup-0001-Suppmat.docxWord document3268K

Data S1. Carbon sampling and biomass estimates.

Figure S1. Proportion of (a) coarse and (b) fine sand in the soil of primary forest plots in Paragominas. Letters indicate forest classes with significantly different means following Tukey post hoc tests (P < 0.05). Dots represent outliers. UF = Undisturbed forests, LF = Logged forests, LBF = Logged-and-burned forests. The methodology for estimating sand content is described elsewhere (See Gardner et al., 2013).

Figure S2. Mean percentage contribution of all the 11 different components of the aboveground carbon pool. Results are separated by forest class in (a) Paragominas and (b) Santarém. Error bars indicate SE. Black bars represent significant differences (P < 0.05) from secondary forests within the same region.

Figure S3. Relationship between the aboveground carbon pool and the time-since the last (a) logging event, and (b) fire event in primary forests of both study regions.

Figure S4. Changes in tropical rainforest structure from closed-canopy undisturbed primary forest to open forests with a dense understory dominated by lianas and fast-growing pioneers. All photos taken 10 m away from a 2 × 2 m tarpaulin by E.B. in Paragominas.

Table S1. Mean carbon content (Mg C ha−1) and SE of the four carbon pools assessed in undisturbed, logged, burned, logged-and-burned, and secondary forests across Paragominas and Santarém.

Table S2. Top ranked models of factors driving aboveground carbon stocks in primary forests in Paragominas and Santarém. Generalized mixed-effects models were used, with Catchment set as a random factor and Percentage area disturbed (Area), Mean soil clay content (Clay), Distance to edge (Edge), Mean plot elevation (Elevation), Mean plot slope (Slope), Time-since the last fire event (Fire), and Time-since the last logging event (Logging) as fixed factors. ∆ - AICc differences from Model 1 (e.g. Model 2 AICc – Model 1 AICc). Weight – Akaike weights.

Table S3. Results of generalized mixed-effects models using the dead wood carbon pool in primary forests as the response variable. Time-since the most recent disturbance (either selective logging or understory fire) was set as the only fixed factor and Catchment as a random factor. Results are separated by study region. ∆ - AICc differences from Model 1 (e.g. Model 2 AICc – Model 1 AICc). Weight – Akaike weights.

Table S4. Results of generalized mixed-effects models using the litter carbon pool in primary forests as the response variable. Time-since the most recent disturbance (either selective logging or understory fire) was set as the only fixed factor and Catchment as a random factor. Results are separated by study region. ∆ - AICc differences from Model 1 (e.g. Model 2 AICc – Model 1 AICc). Weight – Akaike weights.

Table S5. Results of generalized mixed-effects models using the soil carbon pool in primary forests as the response variable. Time-since the most recent disturbance (either selective logging or understory fire) was set as the only fixed factor and Catchment as a random factor. Results are separated by study region. ∆ - AICc differences from Model 1 (e.g. Model 2 AICc – Model 1 AICc). Weight – Akaike weights.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.