Geographic location of the 164 25 9 25 m quadrats in Jianfengling, Hainan Island, China.
Species accumulation curves based on (a, b) number of sampling quadrats and (c, d) number of sampling stems using rarefaction method for old-growth forests (OG), older selectively logged forests (SLold), recently selectively logged forests (SLyoung), older clear-cut forests (CLold) and recently clear-cut forests (CLyoung).
Basal areas at the breast height of selectively logged forests (SL) and clearcut forests (CL) compared to old-growth forests (OG) along recovery times, with (a) d.b.h. ≥2 5 cm, (b) d.b.h. ≥40 cm, (c) d.b.h. ≥2 5 cm for eighteen pioneer species and (d) d.b.h. ≥2 5 cm for eighteen shadetolerate species. Average value for each of the old-growth forest category is plotted as a straight line for comparison, respectively. Basal area unit: m2/625 m2.
1. The extent to which tropical rain forests recover tree species richness, composition and community structure after clear-cutting or selective harvest is controversial. Thus, the conservation value of forests regenerated after harvest remains unclear, but critically important for biodiversity.
2. We sampled trees from 164 25 9 25 m quadrats across a 160-km2 tropical rain forest landscape on Hainan Island, China, which had been clear-cut or selectively harvested and left to recover without management for up to 50 years. Species richness and species abundance distributions were compared among five successional categories to investigate changes in species richness and species abundance over time. Basal areas and three different species similarity indices were compared to reveal temporal changes in species composition and community structure.
3. Species richness recovered faster than species composition and structure in both selectively logged and clear-cut forests. Both total number of species and number of rare species tended to increase from younger harvested forests through older harvested forests to old-growth intact forests. Within 20–40 years after harvest species composition of harvested forests tended towards that of old-growth forests, community similarity between harvested and oldgrowth forests decreased subsequently and basal area did not recover, given even a half-century of succession undisturbed by anthropogenic forces.
4. Shortly after harvest, pioneer species increased rapidly, but shade-tolerant species required much more time to recover to former abundances. The shift from pioneer species to shadetolerant species indicates significant recovery of logged forests.
5. Selectively logged forests recovered more quickly and had higher conservation values than clear-cut forests.
6. Synthesis and applications. Our findings indicate that logged tropical forests only partially recovered the characteristics of pre-harvest, primary forest after a half-century of succession. Recovery of the original tree biodiversity on such post-harvest landscapes will be slow at best, if measured by species composition or stand structure. Our study amplifies the importance of conserving tropical forest integrity and developing harvest and management approaches that facilitate full recovery of logged tropical forests.
Key-words: Tropical forest, forest structure, landscape scale, rare species, regenerated
forests, selective harvest, clear-cutting, species composition, species richness, succession