Each 9 ha plot contains a buffer zone 25 m wide. Trees are monitored inside the core zone, i.e. in an area of 6.25 ha, while silvicultural treatments were applied to the whole plot. The core zone is subdivided into four squares in order to facilitate the numbering of trees. All trees of ≥10 cm DBH are localised by Cartesian coordinates and botanically identified when possible. Three types of silvicultural treatments of increasing intensity were defined and, according to a randomised block design, applied to one plot inside each block, the last plot remaining a control. This left us with three replicates of three treatments plus three control plots. The treatments, applied between 1986 and 1988, were the following :
- T1: selective felling for timber, about 10 trees/ha of DBH ≥ 50–60 cm, belonging to 58 commercial species;
- T2: selective felling for timber plus thinning by poison-girdling of all non-commercial species (NCS) of DBH ≥ 40 cm (30 trees/ha). Poison-girdling of commercial species when trees had major defects;
- T3: selective felling for timber plus logging of NCS of DBH between 40 and 50 cm (for fuelwood: 15 trees/ha), followed by poison-girdling of all NCS of DBH ≥50 cm (20 trees/ha). Poison-girdling of commercial species when trees had major defects.
Logging treatments have induced a very large range of disturbances (from 10 to 60% of aboveground biomass (AGB) lost) that was spatialized with a high confidence level because gaps as well as skidtrails and logging roads were all precisely mapped during logging operations. Damaged trees were recorded with a standard code for each damage type.
|Plots||Timber||Thinning||Fuelwood||% AGB lost|
|Control||1 6 11||0|
|T1||2 7 9||DBH ≥ 50 or 60 cm, according to species, about 10 trees/ha||[12% - 33%]|
|T2||3 5 10||DBH ≥ 50 or 60 cm, according to species, about 10 trees/ha||DBH ≥ 40 cm, all non-valuable trees, about 30 trees/ha||[33% - 56%]|
|T3||4 8 12||DBH ≥ 50 or 60 cm, according to species, about 10 trees/ha||DBH ≥ 50 cm, all non-valuable trees, about 15 trees/ha||40 cm ≤ DBH ≤ 50 cm, all non-valuable trees, about 20 trees/ha||[35 - 56%]|
From 1984 to 1994, an forest inventories have been carried out on the 12 plots each year during the long dry season (from August until the beginning of December). From 1995 on, systematic measurements of girth have been made once every 2 years. Data collected are the following:
- Girth of each living tree with a DBH of ≥10 cm, measured at 1.30 m when possible or above main defaults and/or buttresses by increments of 50 cm. The measurement is made with a steel tape and rounded up or down to the nearest half centimetre. When the point of measurement is too high (well developed buttresses), a Bitterlich relascope is used. Anomalies causing imprecise measurement (non-circular trunks, point of measurement moved up, use of the relascope) are systematically coded in the data base, allowing the user to correct and/or disregard them: about 12% of the trees are concerned.
- Identification of dead trees and of the type of death. Natural death: standing death, primary windthrow (fallen tree), secondary windthrow (fallen and broken tree as a result of a primary windthrow) and human-induced death: felling, death following damage due to logging operations, poisoning;
- Location and botanical identification of new recruits (trees reaching the threshold DBH of 10 cm).