| Abstrak |
: |
A permanent sample plot (PSP) dataset was enumerated from primary forest in Jambi
province, Indonesia. No logging treatment was applied to the plot during the period in
which measurements were taken. The dataset including individual tree information,
such as tree identity, diameter and position within the plot, was put into the Sustainable
Yield Management for Tropical Forests (SYMFOR) computer framework. The model has
been calibrated using the Berau PSP dataset (a region in East Kalimantan).
Prior to the simulation, the Jambi dataset was compared to the Berau dataset in terms
of diameter class distribution, the dominant tree species and tree family distribution. It
was concluded that both datasets are not signifi cantly different. Moreover, both datasets
are lowland tropical primary forest types. Therefore, the software recalibration process
is not necessary.
Using the SYMFOR computer model several silvicultural methods were simulated.
These were the conventional methods of TPTI, RIL (Reduced Impact Logging) and a set
of silvicultural methods derived from the RIL namely RIL8, RIL50 and RIL60. The RIL8
indicates that 8 stems ha-1 is the maximum number of trees for the allowable cut, while
the RIL50 and RIL60 are the maximum volume of the allowable cut are 50 m3 ha-1 and
60 m3 ha-1 respectively.
The simulation was run over a long time in order to cover multiple harvest cycles,
and was repeated several times to capture the variability among runs. Some macro
commands in a spreadsheet computer package were developed to obtain the evolution
of timber extracted and the stand quantity follows logging as simulation time went on.
Under 35 years cutting cycle, both RIL50 and RIL60 performed better than other
methods - the amount of timber extracted per hectare in the fi rst harvest was successfully
attained again.
The simulation was also conducted by altering the length of the cutting cycle within
the TPTI, RIL and RIL8. The results showed that timber production increased with the
cutting cycle. In particular, under the RIL8 on 45 years cycle, the quantity of timber
extracted reached its fi rst harvest level. This might be due to the fact that the maximum
allowable cut assigned for the RIL8 was less severe compared to both methods.
The effect of logging on residual stands was also simulated. It was found that both
the RIL50 and RIL60 were consistently better than the other silvicultural methods; the
forest system could be revived almost to the condition of pre-fi rst harvest level.
Upon the cutting cycle extension, it was found that under RIL50 and RIL60 methods,
the level of residual stand beyond the 35 years cycle was successfully returned to its pre-
fi rst harvest level. In contrast, both the TPTI and the RIL methods failed to reach their
respective pre-fi rst harvest levels despite substantial extension of the cutting cycle
applied. Notably, the RIL8 which is considered to be less severe logging compared to
the RIL, still failed to reach the pre-fi rst harvest level if the cutting cycle was less than
45 years. This might suggest that careful logging operations as assigned within the RIL
methods should be in conjunction with the reduction on logging severity.
This study consistently suggested that the current silvicultural guidelines in Indonesia
(the TPTI) would not use our natural tropical forests sustainably. Both RIL50 and RIL60
silvicultural methods on 35 year cycles could be good alternatives for the TPTI.
Keywords: Conventional TPTI, reduced-impact logging (RIL), initial level, maximum
allowable cut, extracted timber, residual stand |
