Font Size: a A A

Study On Recovery Ecology Of The Degraded Tropical Forest Vegetation In Hainan Island, South China

Posted on:2007-09-19Degree:DoctorType:Dissertation
Country:ChinaCandidate:Y DingFull Text:PDF
GTID:1103360182997915Subject:Ecology
Abstract/Summary:PDF Full Text Request
The tropical forest of Hainan Island lies on the northern edge of Asian tropical rain forest,and is recognized as one of the forest types with richest biodiversity in China. Due toextensive deforestation and long-term intensive agricultural land use, the area of primaryforests has been reduced markedly, and resulted in the degraded ecosystems of the variedrecovery stages. Therefore, the secondary forests that developed from those disturbed forestshave become the most important natural resources in this region. However, the past ecologicalstudies were mainly conducted in primary or old-growth forests, and we still know little aboutthose degraded vegetation after anthropogenic disturbance. In this study, field investigationswere conducted in totally 15.15 ha sample plots of recovery vegetation and old-growth in theBawangling forest region of Hainan Island, South China. The recovery vegetation includeddifferent stages of fallows on the abandoned shifting cultivation lands and logged forestsunder different logging systems. On the basis of those field investigations, we analyzed therecovery patterns and successional processes of the degraded tropical forest vegetations bycomparing their community composition and structure. The main results are as follows:1. Only a few families, genera or species apparently dominated in the 5-year-old recoverytropical lowland rain forest on the abandoned land of shifting cultivation. The small-sizedindividuals dominated the woody community. The abundance and basal area of sproutingstems were greater than those of seeder stems in the tree functional group. By contrast,the abundance and basal area of resprouters and seeders did not differ significantly in theshrub functional group. The total abundance of stems for the community, stemabundances for species in tree or shrub functional groups, and for seeder or resprouterstems were all negatively correlated with coverage of the grass functional group. Themean sprouting ability in the tree functional group was greater than that in the shrubfunctional group, while the sprouting ability of different species was also significantlydifferent in the same functional group.2. In the four recovery phases of the tropical lowland rainforests after shifting cultivation,the richness of the deciduous species decreased gradually with the stands age, and thesame change tendency of the deciduous species proportion also occurred in each size classof stems. In addition, the deciduous species proportion of stems ≥5cm diameter at breastheight (dbh) was higher than that of the small stems (dbh<5cm) in all recovery stage. Thechanges of the density and basal area proportion of deciduous species in varied sizeclasses and in total with the succession showed the unimodal pattern, in which theyusually reached the maximum values at the early stage of succession. In contrast, thedensity and basal area proportion of those deciduous species of the stems of dbh<5cmshowed a different change tendency. The density and basal area proportion of deciduousspecies approached 50% in initial stage of succession, but they decreased to the lowest inthe early stage of succession before increasing in the middle stage.3. The species richness, abundance, and basal area of lianas in old-growth tropical lowlandrain forests were all significantly greater than fallows in different succession stages on theabandoned shifting cultivation lands. They all increased with the succession, andapproached about half of those for the old-growth forest in the middle and late phases ofsuccession. In contrast to the younger forests, large lianas were more abundant incommunities of late recovery stage communities and old-growth forests. Four climbingmechanisms were categorized in this study, including twiner, tendril climber, scrambler,and root climber. Significant differences existed among the proportions of liana stemswith different climbing mechanisms. However the proportions of liana species withdifferent types of climbing mechanisms didn't show significant differences. Twiner wasthe main climbing mechanism in all kinds of recovery communities. With the successionof fallows, the proportion in species richness and stem abundance of twiners increased,while tendril climber and scramblers decreased. Root climbers usually appeared in thelate stage, and mainly distributed in the old-growth forest. The proportion of stemsinfested by lianas increased with age of the stands, and the larger stems had higherprobability of infestation by lianas.4. After different anthropogenic disturbances, the species richness, community height, andproportion of late species in recovery tropical lowland rainforest on abandoned shiftingcultivation lands were lower than those in logged forest. The faster recovery rate ofcommunities usually existed in recovery series of logged forest, for example, thecommunity structure of the 40-year-old selective logged forest was almost similar withthat of the old-growth tropical lowland rain forest. The results of NMS (non-metricmultidimensional scaling) and DCCA (detrended canonical correspondence analysis)revealed that the community composition of recovery vegetation on abandoned shiftingcultivation lands was significantly different from those of the logged forest andold-growth forest. Among environmental and disturbance factors, the disturbance type,distance to remnant old-growth forest, elevation, and successional time mainly influencedthe community composition of recovery vegetation. In recovery vegetation on abandonedshifting cultivation lands, density and basal area proportion of sprouting stems decreasedgradually with the stands age. Based on their successional status and regeneration niche,the species were classed into three functional groups (short-lived pioneer species group,long-lived pioneer species group, and late species group). The short-lived pioneer speciesgroup and long-lived pioneer species group were dominant in recovery vegetation onabandoned shifting cultivation lands, and the late species group dominated in loggedforest and old-growth forest. The proportion of late species group in selective loggedforest was more similar to the old-growth forest than the clear-logged forest.5. Commercial logging, including high-intensity logging, moderate-intensity logging, andlow-intensity logging, enhanced the species richness of recovery communities in thetropical montane rainforest, which was mainly caused by the increase of species richnessin small stems. The stems in logged forest were significantly more abundant than primarymontane rain forest, but the basal area of logged forest was apparently less than primarymontane rain forests. After ca. 25 years of succession, the community structure of thelow-intensity logged forest was much more similar to that of the primary montane rainforest. The proportion of long-lived pioneer species increased significantly in the loggedforest.6. The species composition and community structure of recovery tropical montane rainforestdiffered significantly on different positions (up-road, road, and down-road) of anabandoned timber transportation road. Although with the same time of succession, thespecies richness on the road was less than those on the two road-sides (up-road anddown-road), but the species richness was similar on the two road-sides. The speciesrichness of trees and lianas on the road-side position were 5 and 1.5 times greater thanthose on road, respectively. However the shrub species richness had no significantdifferences between the different road positions. The tree species number per stem on thedown-road was higher than the up-road position, while shrub and liana species numbersper stem on the two road-sides showed no differences. The community composition onvaried road position was different significantly by MRPP (multiple response permutationprocedures). In contrast to the differences of tree and shrub compositions on the threeroad positions, the lianas were more similar. By ISA (indicator species analysis), therewas significant difference of the number of indicator species between different roadpositions. Most species exhibited a significant affinity to up-road position. The densityand basal area of lianas, shrubs, trees, and total stems on road were all significantly lessthan those on the road-sides.
Keywords/Search Tags:Species diversity, Degraded tropical forest, Functional groups, Hainan Island, Vegetation recovery
PDF Full Text Request
Related items