The Spatial Pattern And Height Prediction Models Of Natural Regenerated Seedlings And Saplings Under Barod-leaved Forest In Maoershan

Regeneration seedlings and saplings under canopy are an important part of forests.The number and spatial pattern of sampings and samplings determine the future structure and function of forests.Good natural regeneration under forest canopy plays an important role in the stability and sustainable development of forest ecosystems.Therefore,it is of great significance to study the structure,spatial distribution and growth of seedlings and samplings under the canopy.Based on the investigation data of seedlings and saplings in natural broad-leaved forest in Maoershan Experimental Forest Farm of Heilongjiang Province,the ground diameter structure,tree height structure and spatial distribution of seedlings and saplings were analyzed.The models of ground diameter and tree height and height growth of seedlings and saplings were developed by using mixed effect model for broad-leaved forest in this area.The evaluation and valiadation test of the model was conducted.The results are as follows:Slope is the main topographic factor affecting natural regeneration.In a certain range,the larger the slope is,the more regenerated plants are under the forest.Tree species composition is the main stand factor affecting natural regeneration.Generally speaking,the larger the proportion of breast basal area in the total basal area of the upper stand,the better the regeneration of young trees under the forest.The distribution of ground diameter and tree height of young seedlings can reflect the horizontal and vertical structure of regenerated seedlings under forest conopy.The seedling saplings were classified according to a certain size class and the number of regenerated seedlings at different levels was analyzed.Generally speaking,the number of young seedlings decreased with the increase of ground diameter and tree height,showing a reverse "J"distribution.The number of young trees less than 2cm in diameter,seedlings less than 0.3m in height,and the number of renewal plants of Fraxirtus manshuric in initial stage are the most,but they die quickly.The best species for regeneration are Acer mono and Acer mandshuricaum,followed by Ulmus propinqua,Ulmus laciniata,Acer tegmentosum,Tilia,Juglans mandshurica,Fraxinus manshuric,Quercus mongolica and Populus davidiana.Seven kinds of distance-independent indices,including mean variance ratio(Co),negative binomial index(K),aggregation index(PAI)and Morisita index(I),were used to judge the spatial distribution pattern of young trees.The aggregation intensity of saplings was determined by the values of aggregation index {PAI),negative binomial index(K)and Morisita index(I)at different sampling scales.The aggregation intensity of saplings was the highest at the scale of 5 m×5 m and the lowest at the scale of 30 m×20 m.That is to say,with the increase of sampling scale,the distribution of saplings gradually showed a diffusive trend.At the same sampling scale,saplings of different tree species and different ground diameter classes show different distribution patterns.Generally speaking,aggregation distribution generally occurs in small ground diameter class,and large diameter class is uniform or random distribution,which indicates that the spatial distribution pattern of trees will change with individual growth,and gradually diffuse from aggregation distribution to random distribution.Using mixed deeect model,the ground diameter-tree height and tree height growth models were developed for ten tree species,such as Acer mono,Tilia amurensis and Fraxinus mandshurica,which are mainly regenerated under broad-leaved forest.The basic models are mostly polynomial and power or exponential function.The fitting result of ground diameter tree height model is very good,Ra2 is above 0.85.Besides the proper selection of model,the close relationship between ground diameter and tree height is also a reason for good fitting effect.Compared with Ra2,the fitting effect of tree height grolth model is slightly worse in the range of 0.3-0.7.One reason is that tree height varies greatly with age,and the other is that the amount of data is relatively small.Based on the basic model,the mixed effect model at sample level is constructed.The AIC of the mixed model of most tree species is smaller than the AIC value of the basic model,which improves the fitting effect.The model is tested by independent samples.The average absolute error(MAE)and the average prediction error percentage(MPSE)of the mixed effect model are smaller than the basic model,which shows that the mixed effect model is not only effective.It can improve the fitting effect of the model and the prediction accuracy of the model.