The Relationships Between Biodiversity And Ecosystem Functioning In Natural Tropical Forests Of Hainan Island, China

The ecosystems in the globe are experiencing unexpected change. Biodiversity is thebackbone of ecosystems.The relationships between biodiversity and ecosystem functioning(BEF) is a central issue in ecology. Many experiments on BEF showed that the loss ofbiodiversity will significantly affect ecosystem functioning. Until recently, most of the controlexperiments on BEF have beenconducted in grassland by manipulating combinations ofherbaceous species richness. Due to the limited numbers of species and the controlledenvironment,the consistency of these experiental results with the actual field situations stillneed to be proven.Actual field ecosystems are far more complex than the ones designed bycontrolled experiments.Up to now, few studies on BEF have been done in the actual naturalecosystems, especially in complex ecosystems such as the tropical rainforests. Observationsbased on natural ecosystems may reveal the actual relationships between biodiversity andecosystem functioning. Furthermore, those controlled experiments have been mainly designedto assess how species richness affected ecosystem functioning and ignored other facets ofbiodiversity such as functional diversity and phylogenetic diversity. In this thesis,based on fieldinvestigations of communities and measurements of functional traits and environmental factorsin the tropical lowland rainforests of different successional stages after shifting cultivation inthe Bawangling Nature Reserve,in the old growth forests distributed in different elevationalranges in the Jianfengling Nature Reserve,as well asin the tropical montane rianforests locatedin different habitat types within a30haforest dynamics plot(FDP)in the Jianfengling NatureReserve,we explored the relationships between biodiversity and ecosystem functioning in thenatural tropical forests on Hainan Island,China.We did this research by the four followingsteps:Firstly,the variations of functional trait, biodiversity (including species diversity,functional diversity, phylogenetic diversity) and ecosystem function (stand abovegroundbiomass,AGB) with successional stages,elevations,habitat types and their influences by environmental factors were assessed;Secondly,the correlations among different facets ofbiodiversity(species/functional/phylogenetic diversity) were probed;Thirdly, the relationshipsbetween functional traits versusAGBand different facets of bidiversity versus AGB undervaried conditions (i.e.different successional stages, elevations and habitat types) wereexamined. Finally, the influence of environmental factors on the relationships betweenbiodiversity and ecosystem functioning through functional traits were comprehensivelyanalyzed by the structural equation modeling (SEM) approach. This thesis was one of the fewstudies on BEF based on actual field observationsin forest ecosystems,which is helpful infurther exploring the theory of biodiversity and for evaluating ecosystem functioning andservices. The main results are as follows:1. For tropical lowland rainforests with different successional stages (15-,30-and60-year-old secondary forests and old growth forest) in Bawangling Nature Reserve,(1)specific leaf area, leaf nitrogen content, leaf phosphorus content, leaf total organic carboncontent decreased and leaf dry matter content, wood density, potential maximum heightincreased with the process of succession. Leaf potassium content changed non-significantlywith the processof succession;(2) Species richness and Shannon-Weaver index increased withthe process of succession. There was a no significant change for Pielou’s index in the secondaryforests, but the value of Pielou’s index in secondary forests was lower than that in old growthforest;(3) Functional richness, functional evenness, functional dispersion and Rao’s entropydecreased first and then increased with the process of succession. Functional divergencereached peak in the60-year-old secondary forest;(4) Mean phylogenetic distance andphylogenetic diversity increased while net relatedness index and mean nearest phylogenetictaxon distance decreased with the process of succession. Nearest Taxon Index in the secondaryforests changed no significantly but were higher than that in old growth forest. Phylogeneticstructure transferred from clustering to overdispersion with the process of succession;(5) AGBincreased with the process of succession;(6) Functional traits and biodiversity (speciesdiversity, functional diversity and phylogenetic diversity) in different successional stages affected by different environmental factors. The key environmental factors to drive functionaltraits and biodiversity in15-,30-,60-year-old secondary forest and old-growth forest were pHand soil organic matter, canopy openness, soil total potassium content and availablephosphorus content, soil phosphorus content successively.2. For old growth forests distributed in different elevations in Jianfengling Nature Reserve,(1) specific leaf area and potential maximum height decreased while leaf dry matter contentand wood density increased with the increase of elevation;(2) Species richness andShannon-Weaver index increased with the increase of elevation;(3) Functional evenness,functional divergence, functional dispersion and Rao’s entropy decreased whereas functionalrichness increased with the increase of elevation;(4) Phylogenetic diversity enhanced andphylogenetic structure transferred from overdispersion to clustering with the increase ofelevation;(5) AGBincreased significantlywith the increase of elevation.3. For the tropical montane rainforests distributedin different habitat types (valley, downhill,middle slope, uphill and mountaintop successively) in the30ha.FDP on Jianfengling NatureReserve,(1) specific leaf area decreased while leaf dry matter content, wood density andpotential maximum height increased from valley to mountaintop;(2) There was a reducingtrend for species richness and Shannon-Weaver index from valley to mountaintop;(3)Functional richness, functional evenness, functional dispersion and Rao’s entropy decreasedwhile functional divergence decreased at first and then increased from valley to mountaintop;(4) Phylogenetic diversity decreased from valley to mountaintop. Phylogenetic structuretransferred from clustering to overdispersion from valley to middle slope whereas phylogeneticstructure transferred from overdispersion to random from middle slope to mountaintop. Thewhole phylogenetic structure of the30ha FDP was close to random;(5) AGB increased fromvalley to mountaintop.4. The correlations among species richness, functional richness and phylogenetic diversitywere positive in tropical forests under different conditions. The correlation between speciesrichness and phylogenetic diversity was higher than the other correlationsin all study area. For tropical lowland rainforests with different successional stages in Bawangling Nature Reserve,the correlation between species richness and phylogenetic diversity increased with the processof succession; The correlations between species richness and functional diversity as well as thecorrelation between phylogenetic diversity and functional diversity were higher in15-and60-year-old secondary forests. For old growth forests in different elevational ranges inJianfengling Nature Reserve, species richness explained more variation of phylogeneticdiversity in the low elevations(<800m) than that in the high elevations(>800m). For the oldgrowth tropical montane rainforest in the30ha.FDP in Jianfengling Nature Reserve, thecorrelations between species richness and functional diversity increased from valley tomountaintop. The correlation between species richness and phylogenetic diversity was lower indownhill than that in valley or mountaintop, while the correlation between phylogeneticdiversity and functional diversity was higher in downhill than that in valley ormountaintop.5. For tropical lowland rainforests with different successional stages in Bawangling NatureReserve, AGB in secondary forest (15-,30-and60-year-old forest) was not significantlyinfluenced by specific leaf area or leaf dry matter content; AGB in old growth forest decreasedmonotonously with the increase of species leaf area or the decrease of leaf dry matter content;AGB was positively related to wood density in all successional stages; Contrasting with30-year-old secondary forest, AGB in15-year-old secondary forest or old growth forest waspositively related to potential maximum height, but this relationship was not significant in60-year-old secondary forest. For old growth forests with different elevations in JianfenglingNature Reserve, AGB was not significantly related to functional traits in low elevation; specificleaf area was negatively related to AGB,leaf dry matter content and wood density werepositively related to AGB and potential maximum height was not significantly related toAGBin high elevation. For the old growth tropical montane rainforest in the30ha.FDP inJianfengling Nature Reserve, AGB decreased with the increase specific leaf area and increasedwith the increase of leaf dry matter content and wood density in valley and downhill; AGBonly increased with the increase of wood density from middle slope to mountaintop; Potential maximum height were not significantly related to AGB in all habitat types and the correlationsbetween other functional traits and AGB declined from valley to mountaintop.6. For tropical lowland rainforests with different successional stages in Bawangling NatureReserve, there was a significant and positive effect of biodiversity on AGB; Biodiversityexplained more variation of AGB in the15-year-old secondary forest and old growth forestthan those in the30-and60-year-old secondary forests. For old growth forests with differentelevations in Jianfengling Nature Reserve, biodiversity was not related to AGB both in low orhigh elevations. For the old growth tropical montane rainforest in the30ha.FDP in JianfenglingNature Reserve, species richness and phylogenetic diversity were positively to related to AGBin valley while the effect of species richness on AGB was greater than that of phylogeneticdiversity; Biodiversity was not significantly related to AGB from downhill to mountaintop.7.Initial SEM hypothesized that: the community characteristic (functional traits orbiodiversity) responded to environmental factors and affected ecosystem functioning (AGB)and environmental factors/community characteristic/ecosystem functioning were influencedby successional stage. Final SEM showed that specific leaf area was negatively related to wooddensity or AGB whereas wood density was positively related to AGB; Biodiversity waspositively related to AGB and the order for the effect of biodiversity on AGB was speciesrichness, phylogenetic diversity, functional diversity successively; The successional stage hadthe maximum effect on AGB and functional traits/biodiversity had a larger effect on AGB thanenvironmental factors. Overall, our results suggest functional traits or biodiversity are themajor drivers of ecosystem functioning and environmental factors regulate ecosystemfunctioning indirectly through functional traits or biodiversity in the tropical natural forests ofHainan Island.