Ecological Study Of Cyclobalanopsis Glauca Community On The Karst Hills Of Guilin

Cyclobalanopsis glauca is one of the constructive species of climax community in karst areas, and plays a very important role in forest ecosystem. In this paper, based on the data collected from 15 plots, the population and community characteristics of C. glauca were systematically and comprehensively studied by applying various methods of quantitative ecology. The main study contents included community structure and composition, classification and ordination, species diversity, population structure and spatial distribution pattern, intraspecific and interspecific competition, interspecific relationship and niche breadth and niche overlapping of dominant population, etc. The aims of this study were to reveal the ecological relationships between plant community and environment, and to provide some theory bases that could be applied in the ecological restoration and reestablishment of degraded ecosystem in karst areas.(1)In the karst hills of Guilin, Cyclobalanopsis glauca community mainly distributs fragmently in some well preserved holy hills. 15 plots were set up and investigated in the field. The results show that the height of C. glauca community was from 12m to 16m; the vertical structure was simpler, and obviously divided into three layers: arborous, shrub and herb; arborous layer could be subdivided into two layers and interstratum plant was developed. There were totally 128 species belonging to 104 genera of 63 families in C. glauca community, Euphorbiaceae was the family with the richest species, next was Rubiaceae,Rosaceae and Rutaceae. The species dominated in the arborous layer were 36, in which evergreen species were 19, deciduous species were 17. Obviously, the community had the character of mixed evergreen- deciduous broad-leaved forests. There were 76 species of plants in shrub layer, and the species composition was richer than arborous layer. There were 24 species of plants in herb layer, and Cyperaceae and Liliaceae were the predominant family, in which Carex sp. was dominant species. The flora of 96 seed plants founded in the community was classified into 11 areal types and 5 subtypes, of which tropical elements had 59 genera, accounting for 65.56%, temperate elements had 31 genera, accounting for 34.44%, and tropical floristic elements dominated in the community. Community physiognomy was characterized mainly by evergreen broad-leaved mesophaenerophytes with small-sized, single, leathery and non-entire leaves.(2)Six diversity index and spacies-abundance distribution model were applied to analyze the species diversity characteristics of C. glauca community. The results show that all these indexes could effectively reflect the character of communities, such as the structure and habitat. Meanwhile, they also reflect the particularity of composition and structure of C. glauca community. Species richness and species diversity index of woody plants were obviously higher than those of herbs, but there was no significant difference in the evenness. The variety of diversity in different plots was significantly related with habitat conditions. The relationship between vertical structure and species diversity of the community shows that species richness and species diversity index of both tree layer and shrub layer were higher than those of herb layer. The order of the magnitude of species diversity was: shrub layer>herb layer >tree layer or shrub layer>tree layer>herb layer. The diversity indexes of tree layer, shrub layer and herb layer was varied in the communities. In terms of the specific-abundance distribution, plants in tree layer and herb layer conformed to the logarithmic series distribution, indicating that there are fewer dense species but more sparse species, and that community evenness is relatively low. However, plants of shrub layer nearly conformed to the logarithmic normal distribution, indicating that more enriched species dominated in shrub layer, while sparse species was fewer. It may be concluded that the quantitative analysis of species diversity index was significant to make a through study on the forest community.(3)In terms of diameter structure, height structure and survival curve, population structure of C. glauca population was analyzed. The results show that C. glauca population was a stable population with better natural regeneration, and had a high mortality rate of seedling. Survival curves changed from Deevey typeⅢto DeeveyⅡtype, and population structure transited from growing type to stationary.The spatial distribution pattern of C. glauca population was analyzed by setting up a 50m×50m plot on Yanpo hill and using the spatial point pattern analysis (Ripley's K). Based on the DBHs of individuals of C. glauca population, six age classes (I,II,III,IV,V andⅥ) were distinguished, and pattern dynamic mechanism of the population were discussed. The results showed that the individuals of age classⅠhad a clumped distribution because of patch distribution of microhabitats, but the individuals of age class II,III andⅥrandomly distributed.In terms of spatial scales, C. glauca population was significantly clumped at the small spatial scales, but randomly distributed at the larger spatial scales. These differences of spatial distribution pattern were significantly related to reproduction character of C. glauca and habitat heterogeneity, indicating the adaptive strategy of C. glauca population to environmental condition of Karst hill. By using box-counting dimension, the fractal characters of different age classes of C. glauca population were discussed. The result shows that the box-counting dimension values of different age classes ranged from 0.7 to 1.1. The order of spatial occupation degree was age classⅣ> age classⅤ> age classⅡ> age classⅢ> age classⅥ. In fact, box-counting dimension reflects the spatial occupation degree of an object. The box-counting dimension of population pattern may disclose the spatial occupation degree of population and its scale variation degree.(4)Based on the data collected from 60 quadrats, the classification and ordination of C. glauca community, were studied using two-way indicator species analysis (TWINSPAN) and detrended correspondence analysis (DCA). According to the results of TWINSPAN classification, 60 quadrats of C. glauca community were classified into 8 association types:Ⅰ.ASS. Cyclobalanopsis glauca+Cinnamomum saxatile-Cladrastis wilsonii+Loropetalum chinense-Teucrium pernyi+ Carex sp.;Ⅱ.ASS.Cyclobalanopsis glauca+Cinnamomum saxatile-Loropetalum chinense+ Decaspermum esquirolii-Carex sp.;Ⅲ.ASS. Cyclobalanopsis glauca-Alchornea trewioides+Cladrastis wilsonii-Carex sp.;Ⅳ.ASS. Cyclobalanopsis glauca- Alchornea trewioides+Decaspermum esquirolii-Carex sp.;Ⅴ.ASS. Cyclobalanopsis glauca+ Sapium rotundifolium-Alchornea trewioides-Ophiopogon platyphyllus+ Arthraxon hispidus;Ⅵ.ASS. Cyclobalanopsis glauca+Acer lucidum-Cladrastis wilsonii+ Decaspermum esquirolii-Adiantum malesiana;Ⅶ.ASS. Cyclobalanopsis glauca-Tarenna depauperata-Carex sp.;Ⅷ.ASS. Cyclobalanopsis glauca- Alchornea trewioides -Arthraxon hispidus. DCA ordination analysis not only exhibited the same classification results as the TWINSPAN analysis, but also reflected the relationship between association types and environmental gradient, that is, the first axis of DCA mainly reflected the gradient change of slope, and the diagonal line of the two-dimensional ordination diagram of DCA indicated the gradient change of aspect. All these results showed that the species composition and distribution of C. glauca community were affected mainly by slope factor.(5)Jaccard indices, Ochiai indices, Dice indices, andχ~2 test with 2×2 contingency table were used to analyze the interspecific association among 231 species-pairs of 22 dominant species of C. glauca community. Pearson's correlation coefficient and Spearman's rank correlation coefficient were also applied to calculate correlations between species. The results showed that 100, 105 and 114 species-pairs were positively correlated, while 110, 126 and 117 species-pairs had negative correlations, based onχ~2 test for a 2×2 contingency table, Pearson's correlation coefficient test, and Spearman's rank correlation coefficient test, respectively, and the numbers of positive correlated species-pairs were nearly equal to the negatives. Compared withχ~2 test and Pearson's correlation coefficient test, Spearman's rank correlation coefficient test was an effective method to study the relationships between species. The fluctuation of interspecific relationships in C.glauca community was significantly affected with the degree of habitat heterogeneity.(6)In this paper, the niche breadth and niche overlay of 22 dominant species of C. glauca community were studied using the methods of Levins's indexes of niche breadth, niche proportional similarity formulae and Pianka's index of niche overlap. The result shows that niche breadth values of C. glauca(0.886), Carex sp.(0.476), and Alchornea trewioides(0.413) were larger than those of other species because these species were the constructive or dominant species in the community, and they had wider ecological amplitude and extensive adaptation to environment.Generally, the values of niche overlap is relevant to niche breadth, and the populations with wider niche breadths usually have higher overlap values. These phenomena are often founded among the populations with the same or similar environment requirements and biological properties. In our study, Decaspermum esquirolii and Cladrastis wilsonii had the same similar environment requirements, and they had higher overlap values. The more the niche breadth of populations was, the wider the adaptive range of species to ecological factors, and the more extensive the distribution range of species. The more the niche overlapping of populations were, the more the ecological similarities between populations. The reason for these was that the populations with wider breadth niche might have higher niche overlap, and the populations with narrower breadth niche might have smaller niche overlap. Niche overlap values of most species-pairs were between 0-0.4, indicating that the competition between species was not intensive, and the similarity degree of species in resource utilization was lower.(7)According to the field investigation data, the intraspecific and interspecific competition intensity of C. glauca was quantitatively analyzed by using Hegyi single-tree competition index model. The results show that in C. glauca community, competition intensity decreased gradually with the increase of tree diameter, and intraspecific competition was more intensive than interspecific competition. The degressive order of the intraspecific and interspecific competition intensity was as follows: intraspecific> Cinnamomum saxatile> Acer lucidum > Sapium rotundifolium > Pittosporum glabratum > Cladrastis platycarpa > Firmiana simplex > Pistacia chinensis > Canthium dicoccum> Mallotus philippinensis. The relationships of C. glauca and the competition intensity of the whole forest and C. glauca population nearly conformed to power function ( CI = AD?B), while companion species and C. glauca nearly conformed to logarithm function ( CI = A + B lnD). The change in competition intensity is small when the DBH of the objective tree reaches 25cm. The intraspecific and interspecific competition were stimulated efficiently by the model.