Forest productivity is defined by the environment the population is inserted, by
the implementation of genetically superior material and by the correct silvicultural
management applied at this material. The correct prescription of technical
recommendations and the execution of those operations according to acceptable
quality standard are intended to eliminate or minimized the growth constrains.
Despite the silvicultural improvements that have occurred in recent decades, it is
known that it is still possible to obtain gains in productivity through monitoring, early
detection and correction of deviations in the silviculture. The objective of this study
was to investigate an index for monitoring the silvicultural quality using forest
inventory networks. For this, there were three stages of work: i) Definition of
uniformity indexes of dendrometric variables; ii) Standardization of so-called optimal
range of uniformity (ORU) and validation of the methodology through its application in
a network of forest inventory for a single commercial clone. In the first stage we used
three tests of the network BEPP (Brazil Eucalyptus Potential Productivity) with
different levels of productivity to establish the appropriate indexes to characterize the
silvicultural uniformity. In the second stage we used five clonal test of Eucalyptus in
Sao Paulo state to validate the concept of Optimum Range of Uniformity. In the last
step, the indexes and the concept of IOU were applied on a commercial scale in a
network of forest inventory plots installed in 12.000 hectares of Eucalyptus clonal
plantations in northeastern of São Paulo state, with about 2 years old, and planted
within 1995 to 2009. In the definition phase, the index that represents the percentage
of the total volume of 50% of smaller trees planted (which includes the planting holes)
(PV50) was shown as the best adapted to the purpose of this study because it has
finite limits (50% to 0%), includes the holes of planting and indirectly represents the
distribution of classes of growth. The uniformity at 5, 9, 12 and 24 months were
strongly correlated to uniformity to 6 years (r2 > 0.74) showing the possibility of early
monitoring for detection of quality deviations in forestry. Furthermore, the initial rate
PV50 was highly correlated with the final yield (p <0.001). At the stage of
standardization, there was no statistical difference (Tukey, 5%) of PV50 among
clonal tests despite their different yields, showing that the uniformity index can be
generalized, irrespective of the productivity of the site. The ORU of PV50 was from
37 to 50%, i.e., sample plots which have the PV50 within this range can be
considered satisfactory "uniform". In the validation phase, when the concept was
applied on a commercial scale there was a strong temporal evolution of PV50. In the
plantations made in 1995 the average of PV50 was 29% and increased to 42% in
2009. The percentage of plots within the optimal range of uniformity clearly tended to
rise over this period and could be related to major improvements in forestry
operations and their monitoring through quality control.
