Study On Seedling Growth Model Of Karst Forest Based On Mixed Effect

Forest natural regeneration is one of the important ways of ecosystem restoration and reconstruction.Seedling regeneration is an important part of the forest,and the number and distribution pattern of regeneration determine the structure and function of the stand in the future.Well-renewed stands play an important role in the stability and sustainable development of the forest ecosystem.In this study,natural regeneration in a karst evergreen deciduous broad-leaved mixed forest in Maolan was studied.Based on the seedling survey data for five consecutive times from 2017 to 2021,the spatial distribution pattern and regeneration limiting factors of regeneration seedlings in this community were analyzed,and the interannual seedling survival was analyzed by using the generalized linear mixed effect model.A counting model of seedling regeneration based on impact factors was constructed in Maolan karst forest.At the same time,the ground diameter tree height structure of regenerated seedlings was analyzed,and the mixed effect model was used to establish the ground diameter tree height model of seedlings of different regenerated species in the karst forest of Maolan,and the accuracy evaluation and independent sample test of the model were carried out.The main results are as follows:(1)Geostatistical analysis showed that the seedling density distribution in the plot was mainly concentrated in 0～10 plants/m~2,and the seedling density above 20 plants/m~2 was distributed in the north of the plot,presenting a patchy distribution.(2)The effects of biological factors and habitat factors on seedling survival were analyzed by generalized linear mixed effect model.It was found that seedling survival was mainly affected by habitat factors.Seedling survival and evergreen seedling survival at the community level were significantly positively correlated with soil water content,altitude,canopy density and tree effect,and negatively correlated with available P and rock exposure rate.The survival of deciduous seedlings was significantly positively correlated with slope and the number of individual trees,and negatively correlated with available P and rock exposure rate.(3)Based on Poisson distribution model and negative binomial distribution model,a generalized linear mixed effect counting model is constructed by adding random effects.It is found that the fitting effect of negative binomial distribution model is better than that of Poisson distribution model.Altitude,slope,herb density,soil water content and available N had a significant positive correlation with the number of seedlings.There was a significant negative correlation between rock exposure rate and available P and the number of seedlings.(4)Based on the form of linear equation,power function and exponential function,the ground diameter tree height models of Clausena dunniana,Mallotus philippensis,Boniodendron minus,Laurocerasus australis,Machilus rehderi and Distylium myricoides in karst forest are established by using mixed effect model.The results showed that the linear equation could better fit the relationship between ground diameter and tree height of seedlings of various tree species.Based on the basic model,a quadrat level mixed effect model is constructed.Comparing the mixed effect model with the basic model,it is found that the fitting result of the mixed effect model is significantly better than that of the basic model.The independent sample test of the model shows that the ME of the mixed effect model is closer to0,and the MAE and the MRAE are smaller than the basic model,indicating that the mixed effect model can not only improve the fitting effect of the model,but also improve the prediction accuracy of the model.