Quantitative Ecology Of Picea Wilsonii Forest In The Luya Mountain Nature Reserve

The Luya Mountain National Nature Reserve (LMNR) is to protect national-level protectedanimal Crossoptilon mantchuricum and cold temperate coniferous forests, including Picea wilsonii,Picea meyeri and Larix principis-rupprechtii. Picea wilsonii forest is an important vegetation typein north China, and occupies an important position in the forestry production. In this work, we hadsurveyed the flora, population dynamics and spatial distribution pattern of dominant species,numerical classification and ordination of vegetation, species diversity and functional diversity ofPicea wilsonii forest in LMNR. The results showed that：（1）There were125species of spermatophyte belonging to99genera,41families in Piceawilsonii forests. Temperate distribution and Cosmopolitan distribution were the main geographicaltypes of spermatophyte flora. There were no endemic families, but2endemic genera and a certainnumber of endemic species of China were recorded in the studied areas.（2）Picea wilsonii was the constructive species of the surveyed areas, and Larix principis-rupprechtii is the main competitive species. The age structure and survivorship curves indicatedthat the Picea wilsonii population belonged to steady growing population; Larix principis-rupprechtii population would be in a recession due to extremely few young individuals and toomuch old trees.（3）Point pattern method was used to analyze spatial distribution pattern and interspeciesrelation of Picea wilsonii and Larix principis-rupprechtii populations. It turned out that spatialdistribution pattern and interspecies relationship changed with succession proceeding. Piceawilsonii monodominant community in the earlier stage of succession gradually changed to Piceawilsonii and Larix principis-rupprechtii co-dominant forest, in the latter stage of succession andcorrespondingly the distribution of Picea wilsonii was inclined from clumped to random. For Larixprincipis-rupprechtii population, its clumped scale increased firstly and then decreased. Piceawilsonii had no significant correlation with Larix principis-rupprechtii at Picea wilsonii dominantstage; significant positive correlation at small scale and non relevant at large scale duringcompetition stage; negative correlation at small scale and positive correlation at large scale duringco-dominant stage. The change in spatial distribution pattern and interspecies relationship duringsuccession was mainly driven by intraspecific and interspecific competition for residual resources.(4) The results of two-way indicator species analysis (TWINSPAN) and redundancy analysis(RDA) showed that60Picea wilsonii forest quadrats were classified into9types of plant communities.(2) RDA ordination clearly reflected the relationships between the distributionpattern of communities and environmental gradients. The distribution pattern of communities wasobvious along the first axis gradient. Elevation was the decisive factor that affects the communitydistribution.(3) Environmental factors influenced distribution of dominant species and thusaffected the distribution of the corresponding communities.(4) After eliminating collinearityeffects among environmental factors, only elevation, slope position and soil pH had significantinfluences on spatial distribution pattern of communities. Variation partitioning with3environmental variables through partial RDA revealed their independent effects and sharedfraction quantitatively.(5) The results of the self-organizing feature map (SOM) showed that60quadrats wereclassified into8types of plant communities and each type had its boundary and distributionalrange on the ordination map, revealed the relationships among communities. By visualizingenvironmental factors gradient on topology map, indicated that elevation was the mainenvironmental factor affecting the distribution of Picea wilsonii forests. Although SOM andTWINSPAN classification results were different, the information covered by the two methods washighly matching. Results of application suggested that SOM is applicable to describe communitycharacteristics and to analyze relationship between communities and environmental variables.(6) Generalized linear model (GLM) and generalized additive model (GAM) were used to fitresponse rules of single species with environmental variables and predict the distribution of species.Based on the key environmental factors identified by RDA and variation partitioning, GLM andGAM were applied to fit the response of dominant species with elevation, slope position and soilpH. Results indicated that GAM had better fitting effect than GLM and was more suitable fordescribing the response rules of dominant species with environmental factors. Picea wilsoniimonodominant community gradually changed to Picea wilsonii and Larix principis-rupprechtiico-dominant forest along with the increasing number of Larix principis-rupprechtii. In general,The importance value of Picea wilsonii decreased with elevation increasing, slope positionchanged from lower to upper slope and soil pH decreasing; Larix principis-rupprechtii，by contrast，increased with elevation increasing, slope position changed from lower to upper slope and soil pHdecreasing.(7) Picea wilsonii forest was rich in functional diversity. Functional diversity was closelyrelated to species diversity. FAD (Functional attribute diversity), MFAD (Modified FAD), FDp(Plot-based functional diversity), FDc(Community-based functional diversity), Rao (quadraticentropy), FRic (convex hull hyper-volume) and FEve (Functional evenness) were significantlycorrelated with species richness (S), which were increasing functions of S; FEve and FDiv had nosignificant relationship with Shannon-Wiener (H); MFAD, Rao and FEve were significantcorrelate with Pielou index (E); Rao, FEve, FDiv and FDis (Functional dispersion) had nosignificant correlation with Simpson index (D), which indicated that community with higher species richness and larger heterogeneity would have higher functional diversity. Relationshipsbetween functional diversity and environmental factors analyzed by using linear regressionanalysis indicated that functional diversity decreased with the elevation increasing and with slopeposition changed from lower to upper slope. The results of principal component analysis showedthat the9functional indices can be divided into three main categories: FAD, MFAD, FDp, FDcand FRic representing functional richness, FEve and FDiv representing functional eveness, andFDis and Rao representing functional divergence.The conservation and management of Picea wilsonii forests were discussed.