Natural populations of Tabebuia cassinoides Lam (DC) have been intensively
harvested in the last 70 years. The pressure for continuum exploitation of remaining
populations remains, although few populations remain in conditions that allow
commercial exploitation and recent studies indicated that intensive harvest causes strong
loss of genetic diversity. Thus, the objective of this work was to study the impacts of
harvesting on genetic diversity levels of T. cassinoides natural populations from Vale do
Ribeira-SP, Brazil, using isozymes markers. We collected leaf tissues of adult trees from
seven populations, four natural and three harvested. Sixty trees were sampled per
population, with the exception of one population, where we sampled 100 trees and also
measured their diameter at breast height (DBH). The effects of harvest were evaluated
by simulating different logging intensity in this one population, considering several
DBH classes and retaining different populational sizes per hectare (20, 30, 50, 75 and
100 remaining trees). The different resulting pictures were evaluated comparing the indexes of the percentage of polymorphic loci (P95%), the mean numbers of alleles per locus (A), observed (H0) and expected in Hardy-Weinberg heterozigosity (He) and the fixation index (f). The intrapopulational spatial genetic structure was studied in five of the seven populations, sampling 12 to 20 random groups, formed by the five nearest
trees. In order to study the spatial genetic structure, we registered, with the use of GPS,
the geographic coordinates of the sampled trees in five populations. The results showed that T. cassinoides has high levels of genetic diversity (A = 3.1; He = 0.455 e H0 = 0.445) when compared with other tropical tree species. By comparing the mean values found
for the natural and for the harvested populations, we detected higher levels of genetic diversity in natural populations (A = 2.64; He = 0.491 e H0 = 0.504), relatively to the harvested ones (A = 2,29; He = 0,406 e H0 = 0,353). The largest part of the genetic diversity was found within population (minimum 27.6%). In natural populations, 12.8%
of genetic diversity was found among populations and 28.4% on the harvested ones.
Within natural populations, 12.3% of genetic diversity was found among groups and
9.7% in the harvested ones, suggesting that the mean coancestry within population is
close to the expected in half-sibs (0.125) and, thus, that there is spatial genetic structure
in T. cassinoides populations. The analysis of genotypes spatial genetic structure by
spatial autocorrelation detected a significant indication for spatial genetic structure at
approximately 50 meters distance radius. The genetic diversity index estimation for five diametric classes did not detect significant correlations among diametric classes and A and He indexes. However, significant associations were detected among the diametric classes and H0 (r = 0,813) and f (r = -0,910) indexes, suggesting that trees of the higher diametrical classes have higher heterozigosity and, thus, preferential selection for
heterozygous. When simulating the retention of different sample sizes per hectare, we
observed that all genetic parameters were affected and that, in order to maintain low
negative effects, it is necessary to retain approximately 75 trees per hectare.
