Background/Question/Methods
I examined effects of disparate non-mechanized highly selective logging histories on forest structure and tree species composition, among 4 contiguous afro-tropical montane forests in the West Usambara Mountains, Tanzania. All trees ³ 10cm dbh were measured and identified from 181 0.04-ha forest plots, stratified by elevation and disturbance, and canopy openness measured from 136 50m hemispherical photo transects. Forests consisted of an unlogged old-growth montane forest used as an ecological reference, two forests continuously selectively logged over the past 50 years, primarily of the medium hardwood Ocotea usambarensis (Lauraceae), and one forests logged for 9 years but now 17-years post-logging. I investigated the impact of structural and composition differences on aboveground biomass density among the four forests. In the absence of empirically derived biomass equations for afro-montane forest, I used four pan-tropical equations for moist tropical forest, incorporating wood specific gravity, and evaluated how large tree density and species composition, hence wood density, contributed to estimate variability among the equations.

Results/Conclusions

Results suggest negative but non-significant changes in some broad measures of forest structure (e.g., basal area), but significant differences in canopy openness, mean tree diameter, stem density, diameter distributions of timber species, and relative species abundance, which fundamentally alter the composition of the 50-yr logged forest. Although biomass density was significantly lower (27%) in the most pervasively 50-yr logged forest, compared to the unlogged ecological reference forest, it was not significantly different between the 9-yr logged forest and the reference. Unexpectedly, altered canopy dominance in the 50-yr logged forest included a greater proportion of high wood density species, partially mitigating biomass loss due to logging.