Wildfires triggered by lightning have caused significant losses of ecology, property and life worldwide. With climate warming and El Nino phenomena, there has been a significant increase in the activity of thunderstorms and lightning around the world, and the area burned by lightning-caused fires has increased significantly. However, at present, there is still a big knowledge gap in the research on the lightning ignition process and mechanism. In this paper, we summarized the rule of lightning, factors affecting lightning density and the pattern of lightning days in China, made a link between the occurrence and prediction of lightning-caused fires, and explored the relationship between the pattern of lightning and lightning-caused fires. The lightning occurrence has the characteristics of randomness and instantaneity in time, and the characteristics of lightning are different in different regions. The lightning density is higher in the east than in the west, in the south than in the north, and on the land than in the sea. Lightning-caused fires in the forest are caused by the thermal effect of cloud-ground lightning current. Ignition or not is closely related to the polarity and current intensity of lightning. Lightning density is also affected by altitude, vegetation distribution, soil type, topography and burned area. Higher elevations cause forced convection, which leads to lightning strikes. Woodland, shrub and grassland have different lightning strike densities due to different surface heating levels. Large burned areas cause lightning strikes through enhanced free convection or mesoscale circulation; and the areas strongly influenced by human activities (such as urban, mining, and industrial) have more lightning activity. Fluvisols, cambic-arenosols and luvisols are associated with increased lightning activity. The cumulus clouds generated by forest fires may also trigger another fire. The mean thunderstorm duration in the region east of 105°E increases with decreasing latitude. The process of lightning-induced lightning fires includes three stages: discharge heating stage, thermal feedback stage and spontaneous flame stage. Moreover, climate change can affect the occurrence and ignition probability of lightning. In the future, the increase of lightning occurrence and ignition probability will largely lead to the increase of fire risk in space and time. At present, lightning prediction models often do not rigorously assess the occurrence of lightning. Measurement systems to detect long-term trends in lightning should be developed in the future. By means of satellite communication, UAV, atmospheric electric field measuring system, combined with artificially triggered lightning technique and field ignition experiment, the intelligent forecast models of lightning-caused fires should be established to improve the accuracy of the probability forecast of lightning-caused fires in China, and realize the effective prevention and control of lightning-caused fires.

Objective: The Huzhong area of the Daxing’anling Mountains is a region with a high incidence of lightning wildfires in China, and lookout tower monitoring is the most important means for fire prevention and control. In this study, we analyzed and evaluated the lookout tower monitoring system in the Huzhong area, in order to improve the local lookout tower layout and construction, and effectively enhance the local forest fire prevention and control capability. Method: In this study, Huzhong area of the Daxing’anling Mountains was targeted. Based on the geographical information data, the visibility of the existing 36 lookout towers in Huzhong area was analyzed by using ArcGIS, Origin and other analysis tools. In addition, the fire monitoring capability of the lookout towers was also evaluated combine with the historical forest fire data from 1990 to 2020 in the study area. Result: More than 96% of the forest fires in the study area were caused by lightning wildfires, most of which occurred from noon to evening in summer. The southern part of Huzhong National Nature Reserve, the central part of Huyuan forest farm, the western, central and northern part of Huzhong forest farm, and the junction of Bisui forest farm and Huzhong forest farm were the areas with high incidence of lightning wildfires. Lightning wildfires were mostly distributed on the gentle slope, southeast slope and south slope of the mountain with an altitude of 700?1 100 m. The overall visibility of the lookout tower system in Huzhong area was good, and the visibility of the lookout tower system between different forest farms was basically the same. However, there were significant differences in visibility between the individual lookout towers, among which the visibility level of each lookout tower in Huzhong National Nature Reserve was relatively high and the most average. Overlay analysis showed that 72.73% of forest fires fell in the visible area of the lookout tower, while the fires that fell in the monitoring blind area were mainly distributed in the southwest of Huzhong Forest Farm and the southeast of Bishui Forest Farm. There was a significant difference in the slope direction of fires that fell in the visible and invisible areas of the lookout tower. The former was mainly distributed in the southeast and south slope, while the latter was mainly distributed in the north, northeast and east slope. Conclusion: The overall visibility level of the lookout tower system in Huzhong area is good, higher than the national average level. However, the visibility of individual lookout towers varies greatly, and there are obvious monitoring blind areas. In particular, the fire monitoring ability at shady and semi-shady slopes is relatively weak, which needs to be strengthened.

Objective: This study aims to understand the spatial and temporal distribution of lightning fires in the Daxing’anling Mountains and the dynamic changes over the past 42 years by analyzing the occurrence history of lightning fires in the region from 1980 to 2021, so as to provide a scientific basis for the prevention and control of lightning fires. Method: Based on the statistical data of lightning fires in the Daxing’anling Mountains from 1980 to 2021, the administrative division of the study area and the DEM digital elevation model, the elevation, aspect and slope distribution maps and corresponding areas of the study area were obtained by processing the DEM data and calculating the corresponding areas. Combined with the study zoning map, the spatial-temporal distribution characteristics and dynamic change trend of lightning fires were analyzed. Result: 1) From 1980 to 2021, a total of 1 651 lightning fires occurred in the Daxing’anling Mountains, with a total area of 473 088.8 hm². Approximately 89.4% of lightning fires burned within 100 hm². The number of lightning fires from 2000 to 2021 was about 3.5 times that of from 1980 to 1999. Around 95.9% of the lightning fires occurred from May to August, and 52% of the lightning fires occurred from 13:00 to 17:00. The number of lightning fires increased rapidly in the days following the end of the spring fire prevention period on June 15. In terms of longitude, lightning fires were mainly concentrated near the two centers of 121.2°E and 122.5°E. In terms of latitude, 80% of lightning fires were concentrated between 51°N and 53.5°N. There were significant differences in the lightning fire density and lightning fire area ratio among different administrative divisions, with Heilongjiang being overall higher than Inner Mongolia. The lightning fires were concentrated on slopes of 4°–12°, accounting for 48.9% of the total. When the slope was greater than 4°, the number of lightning fires decreased with the increase of the slope. The most lightning fires occurred within 600–800 m altitude, accounting for 33.2% of the total, while only 1.3% of lightning fires occurred in areas above 1 200 m. The southwest slope had the largest number of lightning fires, with 219 (13.9%). The northwest slope had the fewest lightning fires, with 177 (11.2%). 2) The result of Pearson correlation analysis between impact factors and lightning fires showed that there was a highly significant negative correlation between the number of lightning fires and longitude (R=–0.155, P≤0.001), slope (R=–0.523, P≤0.001), the corresponding elevation area (R=0.336, P≤0.001), forest bureau area (R=0.559, P≤0.001), the corresponding slope area (R=0.734, P≤0.05). There was no significant correlation between the number of lightning fires with latitude, altitude and the corresponding aspect area. From 1980 to 1985, the number of active days of lightning fires was the least and the most stable (35±16.1) days. 3) From 1986 to 1997, the number of active days of lightning fires began to rise to (77.3±58.5) days. 3) From 1998 to 2011, the number of active days of lightning fires reached the peak of (112.1±47.9) days, and from 2012 to 2021, the number of active days of lightning fires decreased to (68.1±33.2) days. The period 1980?2021 was divided into 8 periods with a 5-year interval. From the third period to the seventh period, the first peak day of lightning fires gradually moved forward. Conclusion: The number of lightning fires has significantly increased since 2000, with the most lightning fires in June and the least in August of the year, and with the most at 13:00?16:00 and least at 23:00?5:00. The spatial distribution of lightning fires has a tendency to gather, and except slope, all other spatial types basically conform to the law that the larger the area, the more the number of lightning fires. There is obvious difference in the distribution of lightning fires in different aspect, mainly manifested as more in the south than in the north, and more in the east than in the west. The active days of lightning fires from 1980 to 2021 can be roughly divided into four periods, namely 1980?1985, 1986?1997, 1998?2011, and 2012?2021, reflecting the dynamic change of “flat, rising, peak and fall”. On the annual scale, there is no obvious regularity on the day of the first peak of lightning fire, but on the scale of 5-years, from the beginning of the third period to the end of the seventh period, the date of the first peak of lightning fires accumulatively advances by 24 days.

Objective: This study aims to investigate the temporal and spatial distribution and occurrence regularity of lightning fires in Sichuan Province, and analyze the main influencing factors of lightning fires, so as to provide a theoretical basis and reference for the prevention and control of lightning fires in Sichuan Province. Method: Based on the historical statistical data of lightning fires in Sichuan Province from 2001 to 2021, the change characteristics of lightning fires in Sichuan at different time scales were analyzed. The spatial distribution status and pattern of lightning fires in Sichuan were analyzed using multi distance spatial clustering analysis. Geographic information technology, SPSS statistical software and other tools were used to process the historical statistics data of lightning fire, meteorological data, digital elevation model (DEM) data and forest resource data and thereby analyze the impact of terrain, meteorology, vegetation and other factors on lightning fire in Sichuan Province. Result: A total of 72 lightning fires occurred in Sichuan from 2001 to 2021, most of which occurred from March to June. Among them, April was a concentrated and high occurrence period. The main occurrence period of time was 14:00―18:00, with 15:00―16:00 being the most prone time period for lightning fires. The spatial aggregation of lightning fires in Sichuan was obvious, and lightning fires mainly concentrated in Liangshan Prefecture and Panzhihua City in southwest Sichuan, and Ganzi Prefecture in west Sichuan, especially in Yalong River basin. Precipitation and relative humidity were the main meteorological factors affecting the frequency of lightning fires in Sichuan, while temperature and wind speed had no significant effect. In addition to meteorological factors, lightning fires in Sichuan were also affected by topography and vegetation type. Lightning fires were mainly concentrated in the area of the steep slope, semi-sunny slope or semi-shady slope with an altitude of 2 500–4 000 m and a slope of 26–45°. The most prone forest type was alpine pine forest, followed by Yunnan pine forest. Conclusion: The distribution pattern of lightning fires in Sichuan is obvious under the two time scales of month and day, and they mostly occur in the afternoon of summer. It has significant spatial aggregation characteristics, mainly concentrated in Ganzi Prefecture, Liangshan Prefecture, especially in the middle and lower reaches of the Yalong River. The occurrence of lightning fires is significantly affected by meteorological factors such as precipitation and relative humidity, and is prone to occur in specific terrain conditions and vegetation types. It is suggested to carry out lightning fire prevention and control on the basis of grasping the spatiotemporal distribution characteristics and influencing factors of lightning fires in Sichuan.

Objective: This study aims to explore the impact of land-use/land-cover (LULC) changes on soil carbon sink and its stability in the conversion of Pinus massoniana forests to tea gardens, clarify the role of forest edge formation on forest soil carbon content, and provide a scientific foundation for accurate estimation of forest soil carbon storage and carbon sequestration. Method: The study was conducted in a P. massoniana forest with the area partially converted to tea plantations in Wuyishan National Park. The soil carbon components in the topsoil (0·20 cm) collected from the forest at the locations of 1, 20 and 50 m distances from the edge shared by the two land use types, including the total organic carbon (TOC), dissolved organic carbon (DOC), and microbial biomass carbon (MBC) were analyzed. Result: After 16 to 28 years of converting P. massoniana forest into tea plantation, the TOC, DOC and MBC contents reduced by 57.1%, 48.8% and 16.9%, respectively, compared to the original P. massoniana forest. The TOC and DOC gradually decreased from the interior of P. massoniana forest to the forest-tea edge. However, no significant difference was found for the MBC at those locations (P>0.1). The TOC content was 31.08, 30.67, 22.26 and 13.25 g·kg^?1 at the locations of 50, 20 and 1 m distance from the forest-tea edge and adjacent tea plantation, respectively. The DOC content was 269.7, 259.8, 226.5 and 135.5 mg·kg^?1 at the same location gradient above. The TOC and DOC were able to be well-fitted with explanatory variables (topsoil physical and chemical properties, LULC, and boundary age) by the stepwise regression equation, with independent variables explaining 87.5% and 76.6% of their respective variations, respectively. Nonetheless, MBC provided the poorest fitting. Specifically, TOC and DOC were affected by variables of soil nitrogen compositions and LULC. Boundary age only had effect on the DOC content. Conclusion: The LULC conversion of P. massoniana forest to tea plantation leads to the decrease of topsoil carbon content (particularly TOC and DOC) , while the carbon loss magnitude reduces with the increasing distance to forest edge. Thus, the forest edge effects of soil carbon content need to be considered to reduce uncertainty in the estimation of forest soil carbon sinks on a landscape scale.

Objective: An activation function based on theoretical growth equation (Richards formula) is proposed to solve the problem that the output of neural network algorithm does not accord with biological laws in forest growth modeling, in order to provide a new idea and method for the application of neural network in forest growth modeling. Method: Using the stem analysis data of 96 Abies nephrolepis trees in Wangqing Forestry Bureau of Jilin Province, the traditional nonlinear regression model, the deep neural network tree height-DBH model based on ordinary activation function and the new deep neural network tree height DBH model based on Richards activation function were established respectively. Result: Compared with the traditional nonlinear regression model, R² of the deep neural network model increased by 0.175%, root mean squared error (RMSE) and mean absolute error (MAE) decreased by 2.282% and 4.011%, respectively. Compared with the deep neural network model of ordinary activation function, the new model has obvious tree height asymptote, which is more fitted with the biological law. Conclusion: A neural network activation function based on Richards formula is proposed, which has the following advantages: 1) There must be a maximum value for the output. 2) The output must be greater than 1.3 with a reasonable ANN structure. 3) The new method can take the parameters fitted by the traditional regression method as the input of the neural network model, so that the neural network training can obtain a priori knowledge.

Objective: Suitable edge correction methods were selected to eliminate the influence of edge effects on the spatial structure parameters of the sample plots and to provide a theoretical basis for forest spatial structure analysis. Method: The edge trees were removed according to different edge correction methods such as the Voronoi diagram-based method, the internal buffer method and the nearest-neighbor (NN) method in the sample plots of different sizes, then the structural parameters i.e. the uniform angle index, the size differentiation index, the mingling index and the tree species diversity mingling were calculated in terms of those remained trees within the different-size plots, and we further analyzed how these structural parameters varied against the changes of the sample plots' sizes so as to recognize the applicability of the three different correction methods on different plots. Result: For those plots with each side length of no more than 40 m, the internal buffer method removed the least amount of edge wood. As the scale of plots increased furthermore, the internal buffer method started to removed the most amount of edge wood while the NN method removed more edge trees than the Voronoi diagram-based method, but the differences were not significant. The spatial distribution patterns of small sample plots mostly appeared as clumped or random, and the spatial distribution pattern of large sample plots tends to be clumped but close to random. The mean value of size differentiation index of each sample plot gradually transitioned to intermediate wood status when the side length of the sample plot increased. The mean curves of the simple mingling index showed a trend of strong mixed degree, and the tree species diversity minglings were between medium and strong mixed degree. Conclusion: When the side length of the sample plot was no more than 40 m, the structural parameters varied greatly with the size of the sample plots changing. When the side length of the sample plot was greater than 60m, the edge correction or not had little effect on the calculation of the size differentiation index, the mingling index, and the tree species diversity mingling. The internal buffer method has applicability limitations in calculating the uniform angle index since it did not induce convergence of the index like the other two methods. In general, the NN method is less dependent on the plot scale than the internal buffer method, and is the best performing one among the three methods. The three methods used in this paper all belong to minus-sampling method, the calculation of the structural parameters only utilized the retained trees in the sample plots and wasted part of the survey information. As an alternative, the mirror replication or eight-neighborhood translation correction method may form a larger sample plot (i.e. plus-sampling method) to offset the edge effects, but whether they have scale effects is a question that needs to be further investigated in the future.

Objective: To study the characteristics of individual differentiation and its impact on productivity in the growth process of different densities of Pinus massoniana, it provides a basis for determining suitable management densities, cultivating large-diameter woods and improving stand productivity in P. massoniana forests. Method: Based on 15 continuous years observation data of four kinds of P.massoniana plantations with planting density of 2 500, 3 333, 4 500 and 6 000 plants·hm^?2 (denoted as: A, B, C, D), and random forests and structural equations were used to explore the influence of forest age, planting density and self-thinning process on individual differentiation and productivity of P. massoniana forests. Result: Gini coefficient (individual size differentiation) decreased significantly with increasing stand age and increased significantly with increasing density. In the process of stand self-thinning, the number of standing trees decreased significantly with the increase of stand age, and the self-thinning rate fluctuated and increased significantly with the increase of stand density. The A density (2 500 tree·hm^?2) is the first to reach 25% of the large diameter wood (21 years old) and 50% of the medium diameter wood (17 years old) compared with other densities. The stand volume increased logarithmically with the increase of stand age of P. massoniana, and the increase order in 15 years was A (420.03 m³·hm^?2) > C (202.51 m ³·hm^?2) > B (186.76 m ³·hm^?2) > D (148.95 m ³·hm^?2), the stand productivity increased first and then decreased and tended to be stable. The stand age, planting density and the number of living standing trees have significant effects on Gini coefficient and stand productivity, among which stand age has a significant negative effect on Gini coefficient, and planting density has a significant positive effect. The effect of stand age and planting density on stand productivity was mainly through the indirect effect of “stand age-number of living standing trees-stand productivity, planting density-Gini coefficient-stand productivity”. Conclusion: At density A (2 500 tree·hm^?2), the individual differentiation degree is small and the high productivity level of the P. massoniana stand, which is suitable for the cultivation of medium and large diameter wood in the south subtropical region of P. massoniana plantations.

Objective: This study aims to clarify the long-term water consumption pattern of short-rotation Populus tomentosa plantation during the rotation period and construct a (basic) crop coefficient curve, so as to provide a basis for water-saving irrigation and management of P. tomentosa plantation. Method: We continuously monitored the transpiration (Tr), soil evaporation (E_s), and stand evapotranspiration (ET) of 2–6-year-old P. tomentosa plantations under sufficient water supply. At the same time, we also collected data of meteorological factors, daily trunk growth (DGR), leaf area index (LAI), and groundwater level, and further calculated reference crop potential evapotranspiration (ET₀), basic crop coefficient (K_cb), and crop coefficient (K_c). Result: 1）The main growth stages of the stand were different each year, but the growth rate was the fastest from May to July, during which the diameter growth of trees in this period accounted for 69%~88% of the total growth in the whole growth season, and the cumulative ET accounted for 47%~61% of the total ET in the whole growth season. According to the average DGR and LAI seasonal dynamics of stand during the rotation period, the growth stages of P. tomentosa plantation were divided into the early growth stage, development stage, middle growth stage, and late growth stage, and the corresponding periods were from early April to mid-April, from mid-April to mid-June, from mid-June to mid-August, and from mid-August to the end of October, respectively. 2）The results showed that there were significant seasonal changes in Tr, E_s, and ET in the stand, but the change patterns were quite different in different stand ages. In addition, the seasonal variation of stand Tr was significantly positively correlated with ET₀ only in 3–6-year-old stands (P<0.05), and there was positive correlation betweenE_s and ET₀ (P<0.05) only in 2-year-old and 6-year-old stands, but there was a positive correlation between ET and ET₀ in all stand ages (P<0.05). The annual total ET and Tr increased exponentially (P<0.001) and linearly (P=0.004) with the increase of stand age, respectively, while the stand E_s gradually decreased with the plant age, but suddenly increased significantly in the 6-year-old stand. The proportions of Tr and E_s in ET increased and decreased yearly, respectively and tended to be stable in 5–6-year-old stands. 3）There were significant differences in seasonal variation characteristics of K_cb and K_c in different stand ages. The seasonal variation of K_cb and K_c was not affected by groundwater level, however, it was controlled by LAI in 3–6-year-old stands (P<0.05). Moreover, bothK_cb (R² = 0.44–0.87) and K_c (R² = 0.42–0.77) could establish an excellent quantitative relationship model with LAI. In addition, we constructed the (basic) crop coefficient curve and list of P. tomentosa plantations according to the change patterns of K_cb and K_c and the divided tree growth stages. Conclusion: The seasonal dynamic characteristics of evapotranspiration, transpiration, and evaporation of P. tomentosa stands have interannual changes. The fluctuation of ET₀ is an essential factor in controlling the seasonal dynamics of water consumption, but the control mode varies with water consumption components and stand ages. The growth stage of P. tomentosa can be divided into four periods, and the key period of water management in stands is from May to July every year. The established (basic) crop coefficient curve, list, and prediction model of P. tomentosa plantation can be used to estimate the stand water consumption and thereby formulate and optimize the irrigation schedule. In addition, the results can not only provide a reference for the efficient water management of other tree species, but also help deepen the understanding of the water relationship of the plantation.

Objective: Both leaf hydraulic and economic traits affect biomass accumulation as they can quantify and reflect the ability of plants to acquire water-carbon resources and their trade-off strategies. This study aims to explore the relationship between leaf hydraulic and economic traits and biomass of poplar clones, so as to find a reliable index for predicting clonal biomass, which could have guiding significance for early breeding of high-yield and fast-growing clones. Method: Eight 4-year-old white poplar clones with different growth rates and derived from the same parents (Populus alba 'I-101' × (P. Alba × P. Glandulosa) '84K') were used as the study subjects. Hydraulic traits (leaf vascular hydraulic conductance K_leavesHPFM, leaf hydraulic conductance K_leafEFM, vein density VD, vessel diameter D_v, hydraulic weighted vessel diameter D_h, stomatal density SD and stomatal length SL) and the economic traits (specific leaf area SLA, leaf total carbon content C, leaf total nitrogen content N, and leaf total carbon/nitrogen ratio C/N) were measured in the growing season (July to September). Above-ground biomass (AGB) was calculated at the end of the growing season (October). Then the variations of biomass, leaf hydraulic and economic traits among the clones were analyzed to investigate the relationships among them. Result: There were significant differences in above-ground biomass, leaf hydraulic and economic traits among the clones. The AGB overall ranked as follows: K1 > K2 > K3 > Z2 > Z1 > M3 > Z3 > M1, suggesting that the growth trend of most clones was consistent with the parent tree (8-year-old). Compared with that of leaf economic traits, the variation coefficient of hydraulic traits was larger (2.0%-12.0% VS 4.5%-13.2%). Leaf hydraulic traits and economic traits decoupled with each other, however, they were both related to above-ground biomass. In terms of hydraulic traits, K_leavesHPFM, VD, D_v and D_h increased with AGB, all showing a significant positive correlation. However, SD, SL and K_leafEFM were not correlated with AGB. As for economic traits, SLA was significantly negatively correlated with AGB, while C, N and C/N were not correlated with AGB. Further multiple linear regression analysis showed that VD was the main index affecting biomass (t = 2.957). Conclusion: The white poplar clones show a steady growth trend, and there are differences in leaf hydraulic and economic traits among the clones, which would provide a possibility for finding predictive biomass indicators at early growth stage. Compared with leaf economic traits, hydraulic traits can be the better predictors of the aboveground biomass of those clones, with vein density (VD) being the key factor.

Objective: This study aims to compare and analyze the differences in physiological indicators among different combinations of rootstocks and scions of Catalpa bungei, so as to provide a reference basis for the selection of rootstock and scion combinations and the evaluation of their affinity and growth status. Method: The grafting was carried out with Catalpa bungei ‘Nanlin No.1’, Catalpa bungei ‘Luoqiu’ and Catalpa bungei ‘Suqiu No.1’ as scions, and Catalpa ovata and Catalpa fargesii as rootstocks, and the survival rate, osmoregulatory substance content, antioxidant enzyme activity, endogenous hormone content of each grafting combination were determined. Correlation analysis was conducted on various physiological indicators, and the principal component analysis and orthogonal partial least squares discriminant analysis were used to comprehensively evaluate each grafting combination. Result: 1) The survival rate of the C. fargesii/C. bungei ‘Nanlin No.1’ combination was the greatest among the six rootstock and scion combinations, reaching up to 76.91%, and the survival rate of the C. fargesii/C. bungei ‘Suqiu No.1’ combination was the lowest, as low as 32.05%. 2) The soluble sugar and starch contents of each grafting combination showed a general trend of decreasing, then increasing and then decreasing, and the antioxidant enzyme activities showed a trend of increasing and then decreasing, and all of them reached the maximum after 5 months of grafting. However, the endogenous hormone contents reached the maximum after 4 months of grafting in general. 3) Throughout the measurement period, the soluble sugar content, starch content, superoxide dismutase (SOD) activity, polyphenol oxidase (PPO) activity, and abscisic acid (ABA) content of C. fargesii rootstock grafting combinations were all to varying degrees greater than those of C. ovata rootstock grafting combinations, except for the grafting combination with C. bungei ‘Suqiu No.1’ as the scion. 4) The overall evaluation scores of the six combinations were as follows: C. fargesii/C. bungei ‘Nanlin No.1’ > C. ovata/C. bungei ‘Nanlin No.1’ > C. ovata/C. bungei ‘Suqiu No.1’ > C. fargesii/C. bungei ‘Luoqiu’ > C. ovata/C. bungei ‘Luoqiu’ > C. fargesii/C. bungei ‘Suqiu No.1’. Conclusion: C. fargesii is suitable as rootstock for C. bungei ‘Nanlin No.1’ and C. bungei ‘Luoqiu’, and C. ovata is suitable as rootstock for C. bungei ‘Suqiu No.1’. Among the six combinations, C. fargesii/C. bungei ‘Nanlin No.1’ has the best-combined performance and C. fargesii/C. bungei ‘Suqiu No.1’ has the worst-combined performance. Superoxide dismutase (SOD) activity, polyphenol oxidase (PPO) activity, auxin (IAA) content, and abscisic acid (ABA) content can be used as the main physiological indicators to reflect the affinity and growth status of grafted plants of C. bungei.

Objective: Genome-wide microsatellite sequence and distribution characteristics of Magnolia biondii were analyzed, and highly polymorphic and stable simple sequence repeat (SSR) primer pairs were developed. Furthermore, DNA fingerprinting profiles were constructed for important commercial Magnolia cultivars. Results of this study would provide molecular marker resources for population genetic structure and diversity analyses, and also provide technological support and scientific basis for identification of elite Magnolia cultivars, and also for the protection of intellectual property rights of breeders. Method: Based on the published whole genome sequence of M. biondii, the MISA program was used to characterize the microsatellites. The Primer Premier 5.0 software was used to design SSR primers, and 203 pairs of primers were randomly synthesized and screened using six M. biondii individual plants. The selected polymorphic primers were further used to construct the DNA fingerprinting profiles of 26 commercial Magnolia cultivars. Result: A total of 2 820 303 microsatellites were identified in the genome of M. biondii, among which the hexanucleotide repeats were most abundant (62.13%), followed by dinucleotide repeats (12.02%) and mononucleotide repeats (9.98%). Sequence analysis revealed 501 different repeat motifs, among which AAACCT/AGGTTT (16.46%) repeats were the most frequent, followed by A/T (8.63%) and AG/CT (5.9%). Among the 203 randomly synthesized primers, 94 (46.31%) primers were able to effectively amplify the genomic DNA of six M. biondii individuals, and 25 (12.32%) primers could generate a clear, easily scored and highly polymorphic banding pattern, with the polymorphism information content (PIC) values ranging from 0.24 to 0.72. The top 10 primers ranked by the PIC value were used to construct DNA fingerprinting profiles of 26 commercial varieties, and a total of 85 genotypes were obtained. The number of genotypes produced by each primer varied from 4 to 18. The discrimination power of each primer ranged from 0 to 13. The 26 commercial Magnolia cultivars could be distinguished by a subset of three primers. Conclusion: On the basis of microsatellite sequence statistics and analysis of the entire genome of M. biondii, in this study 10 SSR primer pairs have been developed with clear bands, high polymorphism and good reproducibility. These primers have been used to construct DNA fingerprint profiles for 26 commercial Magnolia cultivars. Comparison of the fingerprint profiles has revealed that there are no synonyms or homonyms within the 26 commercial Magnolia cultivars.

Objective: This study takes the disturbed Pinus armandii forest in the Shennongjia forest region as the research subject, and measures the enzyme activity and soil chemical properties of the P. armandii forest in the growing season and soil respiration after the insect disturbance, aims to explore the changes of soil enzyme activity and soil chemical properties of P. armandii forest and their effect on soil respiration which provides a theoretical basis for the soil carbon emission of P. armandii forest after insect disturbance. Method: In two types of samples (disturbed P. armandii forest-DPAF, healthy P. armandii forest-HPAF), soil samples of 0-10 cm are selected near the soil respiration ring (randomly laid out in the sample) to determine the soil chemical properties and soil enzyme activity, respectively. Result: The soil organic carbon(SOC), total nitrogen(TN), available nitrogen(AN), available potassium(AK) contents and C/N, C/P, N/P of the disturbed P. armandii forest are lower than the healthy P. armandii forest, but the soil total phosphorus(TP), total potassium(TK) and available phosphorus(AP) content are higher than that of the Healthy P. armandii forest. Soil catalase(CAT), acid protease(ACPT) and sucrase(SC) activities of healthy P. armandii forest are significantly higher than those of healthy P. armandii forest(P<0.05), and soil acid protease(ACP) activity is significantly lower than that of healthyP. armandii forest(P<0.05). N/P is the dominant factor affecting ACP activity, and TK is the dominant factor affecting ACPT and CAT activity. Redundant analysis shows that the effects of TP and AP on soil enzyme activities reach a highly significant level (P<0.01), with an interpretation amount of 58.9% and 50.5%, respectively. RDA analysis shows that the effects of N/P and C/P on soil enzyme activities reach a significant level (P<0.05), with an explanation amount of 34.0% and 26.5%, respectively. The soil respiration(R_s) of the disturbed P. armandii forest and the healthy P. armandii forest are basically the same as the growth season, and the monthly change of R_s in the growing season has the characteristics of a unipolar curve. Analysis of variance shows that R_s reaches a highly significant difference between different months (P<0.01).R_s also varies at different times of the growing season, with R_s significantly higher in summer than in Spring, Autumn and Winter, and significant differences in Summer R_s. The R_s of the disturbed P. armandii forest are lower than those of the healthy P. armandii forest during the growing season, and the concentration differences are in the summer growth period. There are no significant differences between the R_s and the chemical factors in the disturbed P. armandii forest, and the R_s of healthy P. armandii forest is significantly inversely correlated with pH. Conclusion: After the disturbance of insect disturbance, the soil chemical properties and enzyme activities of P. armandii forest have significant variations. The effects of insect disturbance on soil enzyme activities in P. armandii forest are mainly achieved through the control of the soil phosphorus element (TP, AP) cycle and the regulation of related ecological measurement ratios (C/P, N/P), and there is no soil phosphorus element limitation. After the disturbance, the soil respiration rate of P. armandii plantation forest in the growing season is significantly reduced, and the difference is related to the soil enzyme activity and chemical properties, but there is no significant relationship.

Objective: To select an optimal way of the semi-arid coal mining restoration in China, we illustrated the effect of Hippophae rhamnoides planting density, arbuscular mycorrhizal fungi (AMF) inoculation and their interaction on the understory vegetation and soil properties. Method: A two-factor experimental design with three planting densities (L: 836 tree ? hm^?2; M: 1 111 tree ? hm^?2; H:1 667 tree ? hm^?2), two inoculation treatments (AMF, CK) and the natural recovery area selected as the control was conducted in Daliuta mine in Shenmu county, Shanxi province, northwest China. The shrubs were planted in April 2012 and the AMF inoculation was conducted in July 2012 in small hole near shurbs. In October 2020, a vegetation survey and mixed sampling in 0-20 cm topsoil were carried out. All plots were totally 8 400 m². The indices of understory vegetation and soil properties were selected for principal component analysis, and five principal components with characteristic values greater than 1 were selected to obtain the weight by the factor load matrix and calculate the comprehensive score. Result: 1) Improving planting density and AMF inoculation both improved the canopy coverage and existing density of Hippophae rhamnoides. 2) The planting density significantly increased the Pielou evenness of understory vegetation and decreased the Simpson dominance of understory vegetation, and the inoculation of AMF significantly increased the Pielou evenness index and decreased the Simpson dominance index of understory vegetation. Planting density significantly influenced the community composition, biomass and coverage of understory vegetation, and the biomass of understory vegetation was the largest at medium planting density. AMF inoculation significantly changed the community composition and biomass of understory vegetation. 3) The planting density and AMF inoculation significantly influenced the biomass of AMF and soil properties. The AMF biomass, microbial biomass and soil physical and chemical properties were at a high level under medium planting density. 4) There was significant interaction in planting density and mycorrhizal inoculation. The influence of planting density on understory vegetation and soil properties was less in inoculation treatments, compared to the control. 5) The comprehensive score results showed that the ecological restoration effect of medium planting density was the best in our study. Conclusion: In this experiment, the ecological restoration effect of medium planting density was the best in our study. Inoculation with AMF provided more ecological niches for planting, and relieved the limitation of high planting density on the growth and positive successions of understory vegetation.

Objective: China is the largest country in the production and consumption of paper and paperboard in the world, and the paper industry is an important pillar industry of the national economy. As the basic raw material of the paper industry, wood pulp has become the largest import category of China’s wood forest products since 1995. It has the characteristics of large import volume, rapid growth and high prices. It is necessary to pay attention to the import risk of China’s wood pulp to ensure the supply chain security of the paper industry chain. Method: After standardizing and determining the algorithm of the import market concentration ratio, the import dependence in consumption, and the import international market share of China’s wood pulp, and scientifically setting the risk zone of China’s wood pulp import risk caused by them, the multiplication index and additive index of China’s wood pulp import risk were constructed with their risk zone set scientifically, the reasonable evaluation method of China’s wood pulp import risk index was determined through comparison. Result: The import dependence in consumption, the import market concentration ratio, and the import international market share of China’s wood pulp, are the core factors that lead to the import risk of China’s wood pulp. The import dependence in consumption of China’s wood pulp appeared a slow decline trend in reducing fluctuation and at high level from 2000 to 2021, and it was 62.34% in 2021, the import risk caused by it was in the high risk zone, close to the extremely high risk zone. The import market concentration ratio CR₈ of China’s wood pulp presented a slow decline trend in fluctuations and at a high level from 2000 to 2021, and it was 89.69% in 2021, the import risk caused by it was in the extremely high risk zone. The import international market share of China’s wood pulp was rising rapidly from 2000 to 2021, and it reached 43.28% in 2021, the import risk caused by it was in the high-risk zone. The overall evaluation method of China’s wood pulp import risk based on multi-dimension should adopt the factor multiplication method, but no matter the multiplicative index or additive index constructed, it shows that China’s wood pulp import risk is in the extremely high risk zone. Conclusion: In order to reduce the import risk of China’s wood pulp, it is proposed to promote the construction of industrial pulp forest to expand the supply of domestic wood pulp, guide the domestic demand of pulp and paper from the production, consumption and export of paper and paperboard, adjust the product structure of wood pulp and paper trade by replacing pulp with paper, and optimize the import source market structure of wood pulp by increasing overseas investment.

Objective: To clarify the key factors that affect farmers' choice of measures to deal with human-elephant conflicts, provide decision-making reference for the government to effectively deal with human-elephant conflicts, and promote the harmonious coexistence of human and elephant in the region. Method: This study selected the distribution area of Asian elephants in Xishuangbanna Prefecture, Yunnan Province as the research area, and adopted the selection experiment method to design a selection consisting of five control measures, the improvement of compensation standards, the construction of fences, the adjustment of planting structure, the transfer of problem elephants, and the relocation of immigrants. The data of 472 farmers collected in the field survey were estimated by using the mixed regression method, and the results of farmers' choice of measures to controlling human-elephant conflict were obtained. Result: The result show: 1) The farmers prefer 80% compensation for Asian elephants (0.99), which followed by relocation of problematic elephants (0.97), planting re-organization (0.96), and fence building (0.45); 2) Farmers with different income levels have a higher choice tendency to improve compensation standards, transfer problems and adjust planting structure; 3) Compared with the control measures consisting of improving compensation standards, setting fences, adjusting planting structure, transferring problems and relocating immigrants, the variable coefficient of maintaining the status quo (ASC) is significantly positive, indicating that farmers tend to maintain the existing control measures as a whole; 4) Compared with the farmers who have a positive preference for governance, the number of labor force, the number of people who don't believe in religion and farmers who perceive the conflict between human and elephant tend to choose improved governance measures. Conclusion: Farmers show a high tendency to choose the governance measures formulated by the government, but resettlement and relocation are the most unwilling choice for farmers. Therefore, it is necessary to pay attention to the preference of farmers for the governance measures of human-elephant conflict in order to optimize the governance work and build a compensation system based on compensation.Policy makers need to prioritize rural households with numerous labor forces who suffered most from human-elephant conflict, and create a new governance model consisting of raising standards, transferring problematic elephants, building fences.