GA-iForest: An Efficient Isolated Forest Framework Based on Genetic Algorithm for Numerical Data Outlier Detection

With the development of data age,data quality has become one of the problems that people pay much attention to.As a field of data mining,outlier detection is related to the quality of data.The isolated forest algorithm is one of the more prominent numerical data outlier detection algorithms in recent years.In the process of constructing the isolation tree by the isolated forest algorithm,as the isolation tree is continuously generated,the difference of isolation trees will gradually decrease or even no difference,which will result in the waste of memory and reduced efficiency of outlier detection.And in the constructed isolation trees,some isolation trees cannot detect outlier.In this paper,an improved iForest-based method GA-iForest is proposed.This method optimizes the isolated forest by selecting some better isolation trees according to the detection accuracy and the difference of isolation trees,thereby reducing some duplicate,similar and poor detection isolation trees and improving the accuracy and stability of outlier detection.In the experiment,Ubuntu system and Spark platform are used to build the experiment environment.The outlier datasets provided by ODDS are used as test.According to indicators such as the accuracy,recall rate,ROC curves,AUC and execution time,the performance of the proposed method is evaluated.Experimental results show that the proposed method can not only improve the accuracy and stability of outlier detection,but also reduce the number of isolation trees by 20%-40%compared with the original iForest method.

Life forms affect beta-diversity patterns of larch forests in China

Beta-diversity reflects the spatial changes in community species composition which helps to understand how communities are assembled and biodiversity is formed and maintained. Larch(Larix) forests, which are coniferous forests widely distributed in the mountainous and plateau areas in North and Southwest China, are critical for maintaining the environmental conditions and species diversity. Few studies of larch forests have examined the beta-diversity and its constituent components(species turnover and nestedness-resultant components). Here, we used 483 larch forest plots to determine the total betadiversity and its components in different life forms(i.e., tree, shrub, and herb) of larch forests in China and to evaluate the main drivers that underlie this beta-diversity. We found that total betadiversity of larch forests was mainly dependent on the species turnover component. In all life forms,total beta-diversity and the species turnover component increased with increasing geographic, elevational, current climatic, and paleoclimatic distances. In contrast, the nestedness-resultant component decreased across these same distances. Geographic and environmental factors explained 20%-25% of total beta-diversity, 18%-27% of species turnover component, and 4%-16% of nestedness-resultant component. Larch forest types significantly affected total beta-diversity and species turnover component. Taken together, our results indicate that life forms affect beta-diversity patterns of larch forests in China, and that beta-diversity is driven by both niche differentiation and dispersal limitation. Our findings help to greatly understand the mechanisms of community assemblies of larch forests in China.

PLFA Analysis of Soil Microbial Community Structure in Different Forest Types

Soil soluble organic matter is an important component in the study of carbon and nitrogen cycling in terrestrial ecosystems. Soil microorganisms, as soil decomposers, participate in soil biogeochemical processes and play an important role in maintaining the balance of soil ecosystems. As a typical subtropical regional unit, Queensland, Australia, is a relatively concentrated distribution area of forests in Australia. It is very sensitive to climate change and plays an important role in Australian climate and even global climate change. Its unique natural environment and ecosystem occupy a special position in the world. However, the knowledge of available carbon and nitrogen pool and microbial activity in forest soil is still very limited. Pinus elliottii, Araucaria cunninghamii and Agathis australis are the three most important forest types in southern Queensland, Australia. In our research, the function and structural diversity of soil microbial communities of these three forest types were studied using biochemical and molecular biological methods, and the effective carbon and nitrogen pools of soil of different forest types and related microbial processes were discussed, which has important theoretical guiding significance for further research on the structure and function of soil ecosystem. The number of PLFAs in the soil of P. elliottii was 45, the number of PLFAs in the soil of Araucaria cunninghamii and Agathis australis was 39 and 35, respectively. The number and content of PLFAs monomer in P. elliottii were higher than those in the other two kinds of forest soil.

Seedling dynamics differ between canopy species and understory species in a tropical seasonal rainforest,SW China

We used 11 years of census data from 450 seedling quadrats established in a 20-ha forest dynamics plot to study seedling dynamics in tree species of a tropical seasonal rainforest in Xishuangbanna,southwestern China.We found that overall seedling recruitment rate and relative growth rate were higher in the rainy season than in the dry season.Both the recruitment rate of seedlings from canopy tree species(two species)and the relative growth rate of seedlings from understory species(nine species)were higher in the rainy season than in the dry season.However,in the rainy season,the recruitment rate of seedlings was higher for canopy tree species than for understory tree species.In addition,relative growth rate of seedlings was higher in the canopy species than in understory seedlings in the dry season.We also observed that,in both rainy and dry seasons,mortality rate of seedlings was higher for canopy species than for understory species.Overall,canopy tree species appear to have evolved a flexible strategy to adapt to the seasonal changes of a monsoon climate.In contrast,understory tree species seem to have adopted a conservative strategy.Specifically,these species mainly release seedlings in the rainy season and maintain relatively stable populations with a lower mortality rate and recruitment rate in both dry and rainy seasons.Our study suggests that canopy and understory seedling populations growing in forest understory may respond to future climate change scenarios with distinct regeneration strategies.

A bibliometric analysis using machine learning to track paradigm shifts and analytical advances in forest ecology and forestry journal publications from 2010 to 2022

Forest habitats are critical for biodiversity,ecosystem services,human livelihoods,and well-being.Capacity to conduct theoretical and applied forest ecology research addressing direct(e.g.,deforestation)and indirect(e.g.,climate change)anthropogenic pressures has benefited considerably from new field-and statistical-techniques.We used machine learning and bibliometric structural topic modelling to identify 20 latent topics comprising four principal fields from a corpus of 16,952 forest ecology/forestry articles published in eight ecology and five forestry journals between 2010 and 2022.Articles published per year increased from 820 in 2010 to 2,354 in 2021,shifting toward more applied topics.Publications from China and some countries in North America and Europe dominated,with relatively fewer articles from some countries in West and Central Africa and West Asia,despite globally important forest resources.Most study sites were in some countries in North America,Central Asia,and South America,and Australia.Articles utilizing R statistical software predominated,increasing from 29.5%in 2010 to 71.4%in 2022.The most frequently used packages included lme4,vegan,nlme,MuMIn,ggplot2,car,MASS,mgcv,multcomp and raster.R was more often used in forest ecology than applied forestry articles.R software offers advantages in script and workflow-sharing compared to other statistical packages.Our findings demonstrate that the disciplines of forest ecology/forestry are expanding both in number and scope,aided by more sophisticated statistical tools,to tackle the challenges of redressing forest habitat loss and the socio-economic impacts of deforestation.

A real-time intelligent lithology identification method based on a dynamic felling strategy weighted random forest algorithm

Real-time intelligent lithology identification while drilling is vital to realizing downhole closed-loop drilling. The complex and changeable geological environment in the drilling makes lithology identification face many challenges. This paper studies the problems of difficult feature information extraction,low precision of thin-layer identification and limited applicability of the model in intelligent lithologic identification. The author tries to improve the comprehensive performance of the lithology identification model from three aspects: data feature extraction, class balance, and model design. A new real-time intelligent lithology identification model of dynamic felling strategy weighted random forest algorithm(DFW-RF) is proposed. According to the feature selection results, gamma ray and 2 MHz phase resistivity are the logging while drilling(LWD) parameters that significantly influence lithology identification. The comprehensive performance of the DFW-RF lithology identification model has been verified in the application of 3 wells in different areas. By comparing the prediction results of five typical lithology identification algorithms, the DFW-RF model has a higher lithology identification accuracy rate and F1 score. This model improves the identification accuracy of thin-layer lithology and is effective and feasible in different geological environments. The DFW-RF model plays a truly efficient role in the realtime intelligent identification of lithologic information in closed-loop drilling and has greater applicability, which is worthy of being widely used in logging interpretation.

Physico-Chemical and Structural Assessment of Akono Riparian Forest Watershed, Tributary of the Nyong Basin (Centre-Cameroon) at Different Stages of Degradation

Freshwater bodies are natural resources that should be exploited to the fullest, while maintaining the sustainability of ecosystems and ecosystem services which they support. Riparian forests are more important as they contain rivers which are vital sources of fresh water for local populations. However, the quality and quantity of water issued from the watershed depend on the structural state of these forests. The aim of this work was to assess the physico-chemical and structural state of the Akono gallery forest. To achieve this, fieldwork consisted of selecting six major streams of the watershed including Ndjolong, Menyeng adzap, Emomodo, Mvila, Negbe and Ossoé kobok. On each of these, two stations, one intact and one degraded, were marked by transects. The method involved measuring Hydrometric parameters (depth, length, width) of the stream and Physico-chemical parameters of water in the streams while dendrometric parameters were measured along 100 m-transects laid using the point-centred quarter method modified for water bodies to collect tree, shrub and palm variables such as trunk diameter, crown diameter and height. Macrophytes and species identification were carried out using standard botanical procedures. Results showed that, the majority of physico-chemical parameters measured differed significantly between intact and degraded stations (P Pentachletra mancrophylla, whereas on degraded sites, this index was low and characterized by the relative dominance of species Piptadeniastrum africanum. Sorensen’s index (0.56) and CFA showed that the different stands were homogeneous. We can affirm that the riparian forests of Akono watershed are towards a state of stability notwithstanding the perpetuation of anthropological actions.

A method for determining the spatial pattern of forest trees based on the uniformity theory

The spatial pattern of trees is an important feature of forests,and different spatial patterns of trees exhibit different ecological stability.Research has confirmed that natural forests with random patterns have higher biodiversity and stronger resistance to unstable factors such as pests and diseases.Even if they are disturbed or destroyed by unstable factors such as pests and diseases,they can still recover and rescue themselves;while artificial forests with uniform and clustered patterns have lower biodiversity and are susceptible to unstable factors such as pests and diseases.And once pests and diseases occur,it’s more difficult for them to recover.In order to promote the healthy and stable develop-ment of the forestry industry and protect the diversity of the biological environment,it is necessary to protect the random pattern of natural forests from being destroyed in the process of forest management,while effectively transforming the spatial pattern of artificial forests into a random pattern.Therefore,in order to ensure the convenient and accurate determination of the type of forest spatial pattern,research on methods for determining forest spatial pattern has become particularly important.Based on the theory of uniformity,this study proposes definitions and related theories of included exclusive sphere,included exclusive body,included random pattern,and included uniformity.Under the guidance of the definition of inclusion uniformity and related theories,and by using mathematical method,it is proved that the uniformity of inclusion(CL)is asymptotically subject to the Eq.18,Therefore,the relationship between the included uniformity(CL)and the number of trees in the sample plot was established,and the corresponding relationship formula was obtained,and then the determination of the spatial pattern type of trees was completed by using the corresponding relationship formula.Through rigorous reasoning and case verification,the determination method of forest spatial pattern is effective.

Ecosystem service multifunctionality of mixed conifer-broad-leaved forests under climate change and forest management based on matrix growth modelling

Climate change and forest management are recognized as pivotal factors influencing forest ecosystem services and thus multifunctionality.However,the magnitude and the relative importance of climate change and forest management effects on the multifunctionality remain unclear,especially for natural mixed forests.In this study,our objective is to address this gap by utilizing simulations of climate-sensitive transition matrix growth models based on national forest inventory plot data.We evaluated the effects of seven management scenarios(combinations of various cutting methods and intensities)on the future provision of ecosystem services and multifunctionality in mixed conifer-broad-leaved forests in northeastern China,under four climate scenarios(SSP1-2.6,SSP2-4.5,SSP5-8.5,and constant climate).Provisioning,regulating,cultural,and supporting services were described by timber production,carbon storage,carbon sequestration,tree species diversity,deadwood volume,and the number of large living trees.Our findings indicated that timber production was significantly influenced by management scenarios,while tree species diversity,deadwood volume,and large living trees were impacted by both climate and management separately.Carbon storage and sequestration were notably influenced by both management and the interaction of climate and management.These findings emphasized the profound impact of forest management on ecosystem services,outweighing that of climate scenarios alone.We found no single management scenario maximized all six ecosystem service indicators.The upper story thinning by 5%intensity with 5-year interval(UST5)management strategy emerged with the highest multifunctionality,surpassing the lowest values by more than 20%across all climate scenarios.In conclusion,our results underlined the potential of climate-sensitive transition matrix growth models as a decision support tool and provided recommendations for long-term strategies for multifunctional forest management under future climate change context.Ecosystem services and multifunctionality of forests could be enhanced by implementing appropriate management measures amidst a changing climate.

Effects of site preparation methods on soil physical properties and outplanting success of coniferous seedlings in boreal forests

This study assessed the effect of patch scarification and mounding on the physical properties of the root layer and the success of tree planting in various types of forests.This study was conducted on 12 forest sites in taiga forests of the European part of Russia.A total of 54 plots were set up to assess seedling survival;root collar diameter,height,and heigh increment were measured for 240 seedlings to assess growth.In the rooting layer,240 soil samples were taken to determine physical properties.The study showed that soil treatment methods had no effect on bulk density and total porosity in Cladina sites.However,reduced soil moisture was noted,particularly in mounds,resulting in increased aeration.In Myrtillus sites,there were increased bulk density,reduced soil moisture,and total porosity in the mounds.Mounding treatment in Polytrichum sites resulted in reduced soil moisture and increased aeration porosity.In the Myrtillus and Polytrichum sites,patch scarification had no effects on physical properties.In Polytrichum sites,survival rates,heights,and heigh increments of bareroot Norway spruce seedlings in mounds were higher than in patches;however,the same did not apply to diameter.In Cladina and Myrtillus sites,there was no difference in growth for bareroot and containerised seedlings with different soil treatments.Growing conditions and soil types should be considered when applying different soil treatment methods to ensure high survival rates and successful seedling growth.

Fracture identification of carbonate reservoirs by deep forest model:An example from the D oilfield in Zagros Basin

Identifying fractures along a well trajectory is of immense significance in determining the subsurface fracture network distribution.Typically,conventional logs exhibit responses in fracture zones,and almost all wells have such logs.However,detecting fractures through logging responses can be challenging since the log response intensity is weak and complex.To address this problem,we propose a deep learning model for fracture identification using deep forest,which is based on a cascade structure comprising multi-layer random forests.Deep forest can extract complex nonlinear features of fractures in conventional logs through ensemble learning and deep learning.The proposed approach is tested using a dataset from the Oligocene to Miocene tight carbonate reservoirs in D oilfield,Zagros Basin,Middle East,and eight logs are selected to construct the fracture identification model based on sensitivity analysis of logging curves against fractures.The log package includes the gamma-ray,caliper,density,compensated neutron,acoustic transit time,and shallow,deep,and flushed zone resistivity logs.Experiments have shown that the deep forest obtains high recall and accuracy(>92%).In a blind well test,results from the deep forest learning model have a good correlation with fracture observation from cores.Compared to the random forest method,a widely used ensemble learning method,the proposed deep forest model improves accuracy by approximately 4.6%.

Spatiotemporal characteristics of seed rain and soil seed bank of artificial Caragana korshinskii Kom. forest in the Tengger Desert, China

Vegetation restoration and reconstruction are effective approaches to desertification control and achieving social and economic sustainability in desert areas.However,the self-succession ability of native plants during the later periods of vegetation restoration remains unclear.Therefore,this study was conducted to bridge the knowledge gap by investigating the regeneration dynamics of artificial forest under natural conditions.The information of seed rain and soil seed bank was collected and quantified from an artificial Caragana korshinskii Kom.forest in the Tengger Desert,China.The germination tests were conducted in a laboratory setting.The analysis of species quantity and diversity in seed rain and soil seed bank was conducted to assess the impact of different durations of sand fixation(60,40,and 20 a)on the progress of vegetation restoration and ecological conditions in artificial C.korshinskii forest.The results showed that the top three dominant plant species in seed rain were Echinops gmelinii Turcz.,Eragrostis minor Host.,and Agropyron mongolicum Keng.,and the top three dominant plant species in soil seed bank were E.minor,Chloris virgata Sw.,and E.gmelinii.As restoration period increased,the density of seed rain and soil seed bank increased first and then decreased.While for species richness,as restoration period increased,it gradually increased in seed rain but decreased in soil seed bank.There was a positive correlation between seed rain density and soil seed bank density among all the three restoration periods.The species similarity between seed rain or soil seed bank and aboveground vegetation decreased with the extension of restoration period.The shape of the seeds,specifically those with external appendages such as spines and crown hair,clearly had an effect on their dispersal,then resulting in lower seed density in soil seed bank.In addition,precipitation was a crucial factor in promoting rapid germination,also resulting in lower seed density in soil seed bank.Our findings provide valuable insights for guiding future interventions during the later periods of artificial C.korshinskii forest,such as sowing and restoration efforts using unmanned aerial vehicles.

Effect of thinning intensity on the carbon sequestration of natural mixed coniferous and broadleaf forests in Xiaoxing'an Mountains, China

To study the effect of thinning intensity on the carbon sequestration by natural mixed coniferous and broad-leaf forests in Xiaoxing’an Mountains,China,we established six 100 m×100 m experimental plots in Dongfanghong For-est that varied in thinning intensity:plot A(10%),B(15%),C(20%),D(25%),E(30%),F(35%),and the control sample area(0%).A principal component analysis was performed using 50 different variables,including species diversity,soil fertility,litter characteristics,canopy structure param-eters,and seedling regeneration parameters.The effects of thinning intensity on carbon sequestration were strongest in plot E(0.75),followed by D(0.63),F(0.50),C(0.48),B(0.22),A(0.11),and the control(0.06).The composite score of plot E was the highest,indicating that the carbon sequestration effect was strongest at a thinning intensity of 30%.These findings provide useful insights that could aid the management of natural mixed coniferous and broadleaf forests in Xiaoxing’an Mountains,China.This information has implications for future studies of these forests,and the methods used could aid future ecological assessments of the natural forests in Xiaoxing’an Mountains,China.

Evaluation of Soil Fertility Quality and Environmental Driving Factors in Different Soil Types of Artificial Forests

The aim was to clarify the environmental driving factors of soil fertility indicators in artificial forests of Guangxi and comprehensively evaluate the soil fertility level.By collecting data on the current status of soil in artificial forests,the spatial distribution of major soil fertility indicators was analyzed,and the distribution map of the fertility index of artificial forests in the entire region and the comprehensive fertility index of artificial forests of different soil types were obtained.Canonical correspondence analysis method was used to analyze soil fertility indicators and environmental factors,and the environmental driving factors of soil fertility indicators for artificial forests of the main soil types in Guangxi were obtained.The results showed that over 90%of the soil fertility index of artificial forests in the entire region was between 0.20 and 0.50.The order of soil fertility index of different soil types of artificial forests from high to low was yellow brown soil>yellow red soil>yellow soil>red soil>limestone soil>latosolic red soil>laterite.In artificial forests of latosolic red soil,the correlation between soil alkaline nitrogen and organic matter,annual average temperature was high,while the correlation between soil available phosphorus and organic matter,pH was high,and the correlation between soil available potassium and environmental factors such as slope,altitude,rainfall,accumulated temperature,and slope aspect was high.In artificial forests of red soil,the correlation between soil alkaline nitrogen and slope,altitude was high,while the correlation between soil available phosphorus and accumulated temperature,rainfall was high,and the correlation between soil available potassium and pH was high.In artificial forests of limestone soil,there was a high correlation between soil alkaline nitrogen and slope,organic matter,a high correlation between soil available phosphorus and accumulated temperature,rainfall,and a high correlation between soil available potassium and pH.

National Soil Organic Carbon Stocks Inventories under Different Mangrove Forest Types in Gabon

Gabonese’s estuary is an important coastal mangrove setting and soil plays a key role in mangrove carbon storage in mangrove forests. However, the spatial variation in soil organic carbon (SOC) storage remain unclear. To address this gap, determining the SOC spatial variation in Gabonese’s estuarine is essential for better understanding the global carbon cycle. The present study compared soil organic carbon between northern and southern sites in different mangrove forest, Rhizophora racemosa and Avicennia germinans. The results showed that the mean SOC stocks at 1 m depth were 256.28 ± 127.29 MgC ha−1. Among the different regions, SOC in northern zone was significantly (p p < 0.001). The deeper layers contained higher SOC stocks (254.62 ± 128.09 MgC ha−1) than upper layers (55.42 ± 25.37 MgC ha−1). The study highlights that low deforestation rate have led to less CO2 (705.3 Mg CO2e ha−1 - 922.62 Mg CO2e ha−1) emissions than most sediment carbon-rich mangroves in the world. These results highlight the influence of soil texture and mangrove forest types on the mangrove SOC stocks. The first national comparison of soil organic carbon stocks between mangroves and upland tropical forests indicated SOC stocks were two times more in mangroves soils (51.21 ± 45.00 MgC ha−1) than primary (20.33 ± 12.7 MgC ha−1), savanna and cropland (21.71 ± 15.10 MgC ha−1). We find that mangroves in this study emit lower dioxide-carbon equivalent emissions. This study highlights the importance of national inventories of soil organic carbon and can be used as a baseline on the role of mangroves in carbon sequestration and climate change mitigation but the variation in SOC stocks indicates the need for further national data.

Power of SAR Imagery and Machine Learning in Monitoring Ulva prolifera:A Case Study of Sentinel-1 and Random Forest

Automatically detecting Ulva prolifera(U.prolifera)in rainy and cloudy weather using remote sensing imagery has been a long-standing problem.Here,we address this challenge by combining high-resolution Synthetic Aperture Radar(SAR)imagery with the machine learning,and detect the U.prolifera of the South Yellow Sea of China(SYS)in 2021.The findings indicate that the Random Forest model can accurately and robustly detect U.prolifera,even in the presence of complex ocean backgrounds and speckle noise.Visual inspection confirmed that the method successfully identified the majority of pixels containing U.prolifera without misidentifying noise pixels or seawater pixels as U.prolifera.Additionally,the method demonstrated consistent performance across different im-ages,with an average Area Under Curve(AUC)of 0.930(+0.028).The analysis yielded an overall accuracy of over 96%,with an average Kappa coefficient of 0.941(+0.038).Compared to the traditional thresholding method,Random Forest model has a lower estimation error of 14.81%.Practical application indicates that this method can be used in the detection of unprecedented U.prolifera in 2021 to derive continuous spatiotemporal changes.This study provides a potential new method to detect U.prolifera and enhances our under-standing of macroalgal outbreaks in the marine environment.

基于SemlikiForest病毒RNA复制子的猪瘟RNA疫苗的初步研究

将猪瘟病毒E2基因克隆于我们此前构建的衍生于Semliki Forest病毒(semliki forest virus,SFV)RNA复制子的新型真核表达载体pSFV1CS中,获得重组质粒pSFV1CS-E2。用纯化的pSFV1CS-E2分别转染BHK-21细胞和293T细胞,经间接免疫荧光试验检测显示,CSFV E2基因在转染细胞中得到表达。小鼠接种试验结果表明,10μg或100μg pSFV1CS-E2可诱导小鼠产生猪瘟特异性抗体。

Prediction of rockburst classification using Random Forest

The method of Random Forest （RF） was used to classify whether rockburst will happen and the intensity of rockburst in the underground rock projects. Some main control factors of rockburst, such as the values of in-situ stresses, uniaxial compressive strength and tensile strength of rock, and the elastic energy index of rock, were selected in the analysis. The traditional indicators were summarized and divided into indexes I and 1I. Random Forest model and criterion were obtained through training 36 sets of rockburst samples which come from underground rock projects in domestic and abroad. Another 10 samples were tested and evaluated with the model. The evaluated results agree well with the practical records. Comparing the results of support vector machine （SVM） method, and artificial neural network （ANN） method with random forest method, the corresponding misjudgment ratios are 10%, 20%, and 0, respectively. The misjudgment ratio using index I is smaller than that using index II. It is suggested that using the index I and RF model can accurately classify rockburst grade.

Variation of SPAD values in uneven-aged leaves of different dominant species in Castanopsis carlessi forest in Lingshishan National Forest Park

The greenness （SPAD） of uneven-aged leaves of dominant species in the Castanopsis carlessi forest at different altitude gradients in Lingshishan National Forest Park, Fujian Province, China were measured by using portable chlorophyll meter SPAD-502. In addition, the correlation between SPAD value and the concentration of chlorophyll and foliar nitrogen was also investigated. Significant variations in SPAD values were found between the uneven-aged leaves of different dominant species and different altitude gradients. Regression analysis showed that SPAD value was significantly correlated with the concentration of chlorophyll and the content of foliar nitrogen, indicating that SPAD value could be indicators for foliar chlorophyll and nitrogen. It is suggested that SPAD meter is a useful tool for forest assessments in decision-making and operational nutrient management programs.

Cultivation Techniques of Panax notoginseng F.H. Chen under Different Forests

To overcome the issues of high cost and continuous cropping obstacles in facility cultivation of Panax notoginseng_ F. H. Chen, satisfy the market demand, save the production cost, improve the utilization rate of forest land, increase the in-come of forest farmers and protect the ecological environment, the cultivation tech-niques of high-quality P. notoginseng seedlings from Wenshan, Yunnan under four kinds of forests （walnut forest, China fir forest, grape forest and kiwi forest） were in-vestigated in this study. The results showed that the height growth, crown diameter, survival rate and 3-year-old tuber weight of P. notoginseng_under walnut forest were higher than those under the other three kinds of forests; the height growth, crown diameter, survival rate and 3-year-old tuber weight of P. notoginseng under China fir forest were higher than those under grape forest and kiwi forest; and the crown di-ameter and survival rate under grape forest were higher, and the height growth and tuber weight under grape forest were lower than those under kiwi forest. Walnut is a broad-leaved deciduous tree species, so large-scale cultivation of P. notoginseng_should be conducted under broadleaf deciduous forest with canopy density around 0.8, taking advantage of the cool environment and rich humus layer under forest. This cultivation technology could save labor, shade, fertilizer and other costs, and accord with the ecological habit and the growth rules of P. notoginseng, thus im-proving yield and achieving high economic benefit.