Applying Maxent And Arcgis To Predict Mosses Geographic Distirbution Range

Predicting the potential geographical distribution of bryophytes is of great value forbryophytes specimen collection, ecological surveys, introduction and cultivation, aswell as their germplasm conservation. In the present work, an investigatin of mossflora in Huaping Nature Reserve was conducted firstly, and then according to19bioclimatic variables and the geographical occurrences of12moss species withdifferent climatic adaptation patterns, the method how to apply MaxEnt combinedArcGis to predict the geographical distribution of plants was introduced, by theirprediction in China, especially in Huaping Nature Reserve, and the comparison oftheir prediction with their factual distribution in Huaping Nature Reserve. The mainresults are as follows:1. The moss flora of Huaping National Nature Reserve consists of232species(including infraspecific taxa) belonging to103genera in30femilies. Among them,31species and3genera are new record to Guangxi.2. There are9dominant families, including66genera and157speciesaccounting for64.08%of the total genera and67.4%of the total species. Hypnaceaeis the richest among these9dominant femilies with29species in it, which accountfor12.45%of the total species and followed by Brachytheciaceae andSematophyllaceae with22species and18species of mosses. The families with sixspecies or more as dominant genera was considered. There are11dominant generawith80species of mosses in Huaping National Nature Reserve, which account for34.48%of the total species. Their are Plagiothecium, Brachythecium, Entodon,Ctenidium, Brotherella, Dicranum, Thuidium, Hypnum, Cyrto-hypnum, Eurhynchiumand Campylopus. 3. There exist nine geographical distribution patterns of mosses in HuapingNature Reserve. EastAsian element accounting for37.07%of the entire flora is themost dominant type followed by Cosmopolitan element (17.67%) and the tropicalAsia element (12.93%).4. Using a maximum entropy algorithm modelling program (MaxEnt3.3.2)and a GIS software system (ArcGIS9.3) to study and model the potentialgeographical distribution of12moss species in China was applied, especial inHuaping, Guangxi, in relation to19bioclimatic variables and percentage of treecover. Their occurrence data from the field work was got, relevant literature andherbarium specimens examination. The prediction clearly showed their suitabledistribution range in China, and the main bioclimatic factors influencing theirdistribution.5. The paper discusses the principles and ways to improve the predictionaccuracy of the geographical distribution of plants. By different niche models, thenumber of different samples, different combinations of environment variables, wereexpected to improve the prediction accuracy of plant according to AUC values. Thegeographical elements of the mosses in Huaping Nature Reserve was studied, and itsspecies abundance, the flora similarity coefficient, with those in other tengeographical units was compared. The geographical elements of mosses among theHuaping and other ten geographical units by using a cluster ananysis and aordination of Detrended Correspondence Analysis were compared. The results showthat Huaping Nature Reserve has closer moss flora with those of Mt. Jiuwanshan, Mt.Maoershan and Mt. Yinzhulaoshan, which results from their closer geographicallocation, their similar climatic conditions and their environments includingvegetation types etc.