[Objective]Taxus cuspidata is a tertiary relic species and the national level endangered plants. The species composition and community structure of T. cuspidata forestand the spatial correlation between T. cuspidata and other major tree species were studied to explore the formation and maintenance mechanism of community biodiversity. [Method] According to the field protocol of the 50 hm² plot in Panama (Barro Colorado Island, BCI), a 25 hm² plot was established in 2014 in Muling National Nature Reserve of Heilongjiang Province. All woody plants with diameter at breast height (DBH) ≥ 1 cm were mapped, tagged, and identified to species. I (DBH < 10 cm), II (10≤DBH<30 cm), III (DBH≥ 30 cm). [Result] In the 2014 census, we documented 57 woody species with 63 877 individuals, belonging to 38 genera and 22 families. The mean DBH of all trees was 7.83 cm, all species were found when sampling area was set as 21 hm². The DBH distribution of all individuals showed a reversed "J" type. The DBH distributions of the T. cuspidata and Corylus mandshurica were approximate normal and "L" type, respectively, other major species showed reversed "J" type, including Tilia amurensis, Acer mono, Acer barbinerve, Acer tegmentosum, Betula costata, Abies nephrolepis and Pinus koraiensis. Within a scope of less than 50 m, Tilia amurensis, Acer mono, Corylus mandshurica, Acer tegmentosum, Betula costata displayed a gathered distribution, Abies nephrolepis, Acer barbinerve, T. cuspidata first displayed a gathered distribution, followed by random distribution and uniform distribution. The distribution pattern of Pinus koraiensis fluctuated between aggregate distribution and random distribution. T. cuspidata and Acer mono were negatively correlated at small spatial scales (1, 3, 5 m), and with the increase of spatial scale, became independent of each other, and went positively correlated when the spatial was larger than 30 m. T. cuspidata was negatively correlated with five species in different spatial scales, including Acer barbinerve, Acer tegmentosum, Betula costata, Corylus mandshurica and Pinus koraiensis. Generally. T. cuspidata was negatively correlated with Abies nephrolepis, but showing independent of each other at few scales. In contrast, T. cuspidata was independent with Tilia amurensis at most scales, and showing negative correlation at few scales. [Conclusion]T. cuspidata forest had a relatively high species richness. Sapling regeneration of most species was fairly good, but very poor for T. cuspidata.Spatial distribution of species was significantly correlated with habitat, and the species showed different habitat preferences. T. cuspidata may be a niche differentiation with some species, including Pinus koraiensis, Abies nephrolepis, Tilia amurensis, Acer tegmentosum, Corylus mandshurica, Betula costata and Acer barbinerve.

[Objective]Taxus cuspidata is a tertiary relic species and the national level endangered plants. The species composition and community structure of T. cuspidata forestand the spatial correlation between T. cuspidata and other major tree species were studied to explore the formation and maintenance mechanism of community biodiversity. [Method] According to the field protocol of the 50 hm² plot in Panama (Barro Colorado Island, BCI), a 25 hm² plot was established in 2014 in Muling National Nature Reserve of Heilongjiang Province. All woody plants with diameter at breast height (DBH) ≥ 1 cm were mapped, tagged, and identified to species. I (DBH < 10 cm), II (10≤DBH<30 cm), III (DBH≥ 30 cm). [Result] In the 2014 census, we documented 57 woody species with 63 877 individuals, belonging to 38 genera and 22 families. The mean DBH of all trees was 7.83 cm, all species were found when sampling area was set as 21 hm². The DBH distribution of all individuals showed a reversed "J" type. The DBH distributions of the T. cuspidata and Corylus mandshurica were approximate normal and "L" type, respectively, other major species showed reversed "J" type, including Tilia amurensis, Acer mono, Acer barbinerve, Acer tegmentosum, Betula costata, Abies nephrolepis and Pinus koraiensis. Within a scope of less than 50 m, Tilia amurensis, Acer mono, Corylus mandshurica, Acer tegmentosum, Betula costata displayed a gathered distribution, Abies nephrolepis, Acer barbinerve, T. cuspidata first displayed a gathered distribution, followed by random distribution and uniform distribution. The distribution pattern of Pinus koraiensis fluctuated between aggregate distribution and random distribution. T. cuspidata and Acer mono were negatively correlated at small spatial scales (1, 3, 5 m), and with the increase of spatial scale, became independent of each other, and went positively correlated when the spatial was larger than 30 m. T. cuspidata was negatively correlated with five species in different spatial scales, including Acer barbinerve, Acer tegmentosum, Betula costata, Corylus mandshurica and Pinus koraiensis. Generally. T. cuspidata was negatively correlated with Abies nephrolepis, but showing independent of each other at few scales. In contrast, T. cuspidata was independent with Tilia amurensis at most scales, and showing negative correlation at few scales. [Conclusion]T. cuspidata forest had a relatively high species richness. Sapling regeneration of most species was fairly good, but very poor for T. cuspidata.Spatial distribution of species was significantly correlated with habitat, and the species showed different habitat preferences. T. cuspidata may be a niche differentiation with some species, including Pinus koraiensis, Abies nephrolepis, Tilia amurensis, Acer tegmentosum, Corylus mandshurica, Betula costata and Acer barbinerve.

To explore successional dynamics of community structure and associated soil response within Pinus massoniana forests, species composition, community structure and habitat properties were determined among three forests differing in restoration stages (i.e., 10, 20-30 and 50 a) in Guniujiang Nature Reserve in Anhui Province, by using a spacefortime substitution approach. The results showed that 1) with the succession, the species richness decreased, but the Shannon diversity index increased. The dominance of deciduous trees declined, but the dominance of evergreen trees progressed in the canopy layer; 2) The distribution of diameter size followed the unimodal type for P. massoniana across the three successional stages, but changed from L to inverse-J type for Castanopsis eyrei through forest succession; evergreen shrubs exhibited good performance in growth; 3) With the forest succession, soil total nitrogen and organic matter contents increased, and soil fertility increased gradually, while the pH value decreased.

To explore successional dynamics of community structure and associated soil response within Pinus massoniana forests, species composition, community structure and habitat properties were determined among three forests differing in restoration stages (i.e., 10, 20-30 and 50 a) in Guniujiang Nature Reserve in Anhui Province, by using a spacefortime substitution approach. The results showed that 1) with the succession, the species richness decreased, but the Shannon diversity index increased. The dominance of deciduous trees declined, but the dominance of evergreen trees progressed in the canopy layer; 2) The distribution of diameter size followed the unimodal type for P. massoniana across the three successional stages, but changed from L to inverse-J type for Castanopsis eyrei through forest succession; evergreen shrubs exhibited good performance in growth; 3) With the forest succession, soil total nitrogen and organic matter contents increased, and soil fertility increased gradually, while the pH value decreased.

Based on the numerical classification of the forest communities in Pangquangou Nature Reserve of Shanxi by  Two-Way Indicator Species Analysis (TWINSPAN), and by using the indices of richness, species diversity, and evenness, this paper studied the species diversity of the forest communities within the Reserve, and by using Spearman’s rank correlation coefficient, analyzed the relationships among the species diversity in the tree, shrub, and herbaceous layers of the communities. The richness, species diversity, and evenness indices of the 15 associations of the communities well reflected the diversity differences of all the associations. Overall, the richness and species diversity indices of most associations presented the sequence of herbaceous layer > shrub layer > tree layer, and the evenness index was the lowest in herb layer but had less difference in tree and shrub layers. There existed significant differences (P<0.01) among the Patrick index, Simpson index, and Shannon index and between the Pielou index and Alatalo index. In Ass. 3 (Larix principisrupprechtii-Spiraea pubescens+Rosabella-Fragaria orientalis), there was a significant negative correlation (r=-0.643, P<0.05) between shrub and herbaceous layers; in Ass. 8 (Betula platyphylla+Populus davidiana-Cotoneaster acutifolius+R.bella-Carex stenophylloides), shrub layer had a significant negative correlation (r=-0.458, P<0.05) with tree layer but a significant positive correlation (r=0.404, P<0.05) with herbaceous layer; in Ass. 11(Picea meyeri-C. stenophylloides+Carpesium cernuum), tree layer was significantly negatively correlated with herbaceous layer (r=-0.949, P<0.05).

Based on the numerical classification of the forest communities in Pangquangou Nature Reserve of Shanxi by  Two-Way Indicator Species Analysis (TWINSPAN), and by using the indices of richness, species diversity, and evenness, this paper studied the species diversity of the forest communities within the Reserve, and by using Spearman’s rank correlation coefficient, analyzed the relationships among the species diversity in the tree, shrub, and herbaceous layers of the communities. The richness, species diversity, and evenness indices of the 15 associations of the communities well reflected the diversity differences of all the associations. Overall, the richness and species diversity indices of most associations presented the sequence of herbaceous layer > shrub layer > tree layer, and the evenness index was the lowest in herb layer but had less difference in tree and shrub layers. There existed significant differences (P<0.01) among the Patrick index, Simpson index, and Shannon index and between the Pielou index and Alatalo index. In Ass. 3 (Larix principisrupprechtii-Spiraea pubescens+Rosabella-Fragaria orientalis), there was a significant negative correlation (r=-0.643, P<0.05) between shrub and herbaceous layers; in Ass. 8 (Betula platyphylla+Populus davidiana-Cotoneaster acutifolius+R.bella-Carex stenophylloides), shrub layer had a significant negative correlation (r=-0.458, P<0.05) with tree layer but a significant positive correlation (r=0.404, P<0.05) with herbaceous layer; in Ass. 11(Picea meyeri-C. stenophylloides+Carpesium cernuum), tree layer was significantly negatively correlated with herbaceous layer (r=-0.949, P<0.05).

Changes in physiognomy, species composition and structure, and dispersal mechanisms of canopy and subcanopy plant assemblages were investigated along a chronosequence of three ages: 12, 20, and 50+ years old (=old-growth), three replications in each, in an Atlantic Forest landscape in Northeastern Brazil. Our objective was to investigate whether there is floristic and structural convergence along secondary succession. There were significant differences between secondary and old-growth forests in density and basal area only for the subcanopy. Differences in density between forest ages were noted when the assemblage was analyzed per diameter and height classes. Richness of canopy species of both secondary ages differed from those of old-growth forests. Some dominant species in the canopy of secondary forests showed a significant decrease in density with increasing age, which indicates an ongoing process of floristic changes. The low level of shared species between secondary and old-growth forests supports the idea that species composition is one of the last components to recover during successional process. Zoochory was the most important dispersal guild in species percentage and number, irrespective of stand age. Although regenerating areas can take alternative pathways, our results indicate that secondary Atlantic Forest sites have a high potential for natural regeneration. This recovery is recorded as a physiognomic convergence of the canopy layer in as little as 12 years, and progressive introductions of later successional species into the plant assemblage that lead to convergence in terms of the diversity and richness of the subcanopy and of dispersal guilds just 20 years after abandonment.