A method for predicting tree species structure of subtropical forest

Abstract

The application discloses a prediction method of tree species structure of arbor forest in subtropical areas, and comprises the following steps: 1) dividing the tree species of arbor forest into three categories of pines, firs and broad-leaved trees, and establishing a prediction model according to volume proportion data of each tree species category obtained from fixed sample plots in previous forest resource surveys; 2) taking two-period survey data as modeling data, and fitting parameters by using STATA software; 3) substituting measured data of each sample plot into the prediction model for iterative calculation, obtaining volume proportion values of the future arbor forest in the two-period interval as prediction values; calculating average prediction values of the prediction values of the tree species proportions of all sample plots; and 4) comparing the prediction value of the volume proportion of the broad-leaved trees or the sum of the prediction values of the volume proportions of the pines and firs in the average prediction values of the tree species proportions with a critical value of coniferous and broad-leaved mixed forest, and determining whether artificial intervention needs to be applied in actual forestry management. The application can predict the tree species structure of arbor forest in future years, and has the function of guiding forestry production and operation activities.

Description

Technical Field

[0001] This invention relates to a method for predicting the tree species structure of arbor forests in subtropical regions. Background Technology

[0002] Forest succession reflects the fundamental ecological processes of terrestrial plants, revealing changes in the overall forest composition. Therefore, it influences ecological functions and timber supply, and is closely related to ecosystem services, carbon sequestration, soil carbon accumulation, and flora and fauna habitats. Thus, revealing the patterns of forest tree species structure and predicting forest succession processes helps us understand the impacts of natural evolution and climate change on current and future forest species composition, and is crucial for predicting forest carbon sink dynamics and implementing sustainable management.

[0003] Most current knowledge about forest succession is based on time series, meaning that forest succession follows a single, largely deterministic trajectory over time. The application of mathematical models to simulate and analyze the process of forest succession has attracted widespread attention, leading to the development of various forest dynamic models based on mathematical formulas, such as the JABOWA model, the FOREST single-tree model, and the SORTIE model.

[0004] Changes in tree species structure are a crucial aspect of forest succession, reflecting the composition of tree species within the forest ecosystem and the rationality of their proportional distribution. Revealing the patterns of these changes can provide a basis for decision-making regarding the rational management of forest resources. Norden et al.'s work, "Successional dynamics in Neotropical forests are asuncertain as they are predictable," highlights this. Proceedings of the National Academy of Sciences of the United States of America 2015, 112 In the paper "8013-8018, doi:10.1073 / pnas.1500403112", the dynamic interactions between stem density, basal area, and species density are incorporated into a forest succession model to quantify how forest interacting components systematically influence succession. In Chen et al.'s paper "Application of meta-population competition mechasism in forest succession simulation of Mount Lushan",... 中文 《生态学报》 2017, 36In [reference 862-868, doi:10.13292 / j.1000-4890.201703.003], a Logistic model was used to simulate the impact of habitat and population competition on the number of tree species in Lushan. In Zhang et al.'s paper, "How evergreen and deciduous trees coexist during secondary forest succession: Insights into forest restoration mechanisms in Chinese subtropical forest,"... 《全球生态与保护》 2021, 25 In “e01418”, the coexistence and competition between deciduous and evergreen broadleaved forests in subtropical secondary forests were studied, and it was found that disturbed secondary forests gradually transformed into regional climax vegetation types dominated by evergreen broadleaved forests. In addition, some scholars have focused on and conducted in-depth research on changes in vertical structure, community structure, phylogenetic turnover, and ecological memory of subtropical tree species.

[0005] Broadleaf forests are the apex communities of subtropical forest vegetation, possessing numerous advantages. However, from the perspective of providing high-quality timber to society, the presence of coniferous species (including pines and firs) is still necessary. Therefore, mixed coniferous and broadleaf forests are currently the most prevalent form. Mixed coniferous and broadleaf forests play a vital ecological role in water conservation, soil and water conservation, increasing growth rates, providing diversified forest products, reducing the incidence of pests and diseases, and preventing the spread of fires. They also hold an important strategic position in constructing new multifunctional forest ecosystems with high ecological, economic, and social benefits (Zhou Qiujing et al. Species Composition and Community Structure of Natural Mixed Coniferous and Broadleaf Forests in Shennongjia. Journal of Ecology, 2019). However, the successional patterns of the structural proportions of various tree species in subtropical forests remain poorly understood. The reasons are as follows: First, most forest succession studies focus on changes in a specific tree species type, rarely considering the dynamic changes in the future structural proportions of tree species in regional forests. Second, the Continuous Forest Inventory (CFI), a survey conducted in many countries worldwide, possesses a sufficient time span and abundant, statistically significant field measurement data; however, there are few reports on using this data source to build mathematical models, quantitatively summarize the long-term evolution patterns of tree species outcomes, and simulate and predict future evolution directions. Finally, difference equations mostly focus on predicting stand variables closely related to harvest forecasting, such as diameter at breast height (DBH) and basal area, and there is a lack of research on forest succession, and an understanding of the relationship between difference equations and forest succession. Therefore, none of the above-mentioned studies can predict the succession trajectory of tree species structure in historical environments, thus failing to provide guidance for forestry production and management activities based on tree species structure. Summary of the Invention

[0006] To address the problem that various studies mentioned above cannot predict the succession trajectory of tree species structure in subtropical forests within historical contexts, thus failing to provide guidance for forestry production and management activities based on tree species structure, this invention provides a method for predicting tree species structure in subtropical forests, offering guidance for the direction and strategies of artificial intervention in forest management.

[0007] The method for predicting the tree species structure of subtropical arbor forests according to the present invention includes the following steps:

[0008] 1) The tree species in the arbor forest are divided into three major categories: pine, fir, and broadleaf. Based on the volume ratio data of each major category of tree species obtained from the sample plots of the two previous surveys, the following prediction model (1) is established:

[0009] (1)

[0010] In equation (1), y 10 , y 20These represent the volume proportions of pine and fir trees in the first phase of two consecutive surveys. y 11 , y 21 , y 31 Let represent the volume ratios of pine, fir, and broadleaf trees in later stages, respectively, and use them as predicted values, satisfying: ,and y 11 , y 21 , y 31 , y 10 , y 20 Both are greater than or equal to 0 and less than or equal to 1; These are the six parameters of the prediction model;

[0011] 2) Using the two survey periods of the sampling population as modeling data, the parameters were obtained by fitting the data using STATA software. ;

[0012] 3) Substitute the measured data of each sample plot into the prediction model of formula (1) for iterative calculation to obtain the volume ratio of arbor forest species in future years with several step lengths of modeling data interval, and use it as the predicted value of the tree species ratio in future years of the sample plot; calculate the average value of the predicted values ​​of the tree species ratio of all sample plots in the sampling population to obtain the predicted value of the tree species ratio of the population.

[0013] 4) Compare the sum of the predicted values ​​of the broadleaf timber volume ratio or the predicted values ​​of the pine and fir timber volume ratios in the overall tree species ratio prediction with the critical value of 0.65 for mixed coniferous and broadleaf forests. If the predicted value is greater than 0.65, human intervention measures need to be implemented in actual forestry management.

[0014] In step 1), the above-mentioned pine species include one or more of Masson pine, black pine, Huangshan pine, and slash pine; the fir species include one or more of Chinese fir, Japanese cedar, dawn redwood, pond cypress, and yew; the broad-leaved species include one or more of Schima superba, Liquidambar formosana, oak, Castanopsis fargesii, and camphor tree; the above-mentioned volume is calculated by measuring each tree of a normally growing arbor forest species with a diameter at breast height ≥ 5.0 cm and using a one-dimensional volume table model.

[0015] In step 2), obtain the parameters. When setting parameters The initial value of each is 1.

[0016] In step 3), the step size mentioned above is the interval between two survey periods.

[0017] In step 4), the aforementioned human intervention measures are afforestation or forest tending. Afforestation includes, for example, cultivating artificial fir or pine timber forests, while forest tending includes, for example, thinning or replanting.

[0018] Beneficial Effects: This invention constructs a predictive model of the tree species structure in subtropical forests, enabling the prediction of changes in the tree species structure in future years. This allows for a deeper understanding of the volume and stock of each tree species. By comparing the volume ratio of broadleaf trees or the sum of the volume ratios of pine and fir trees with the critical value for mixed coniferous and broadleaf forests, it determines whether human intervention measures (such as selecting afforestation tree species or carrying out targeted forest management and tending) are necessary in actual forestry management. This ensures a sustainable and stable supply of regional ecological products and timber products, providing a basis for decision-making in the rational management of forest resources. It is of great significance for protecting biodiversity and formulating sustainable macro-forestry management strategies. Attached Figure Description

[0019] Figure 1 The graph shows the results of iterative prediction based on data from the 2014-2019 review period using the method of the present invention, with the measured data of 2019 as the initial value. Detailed Implementation

[0020] The technical solution of the present invention will be described in detail below through embodiments, but the scope of protection of the present invention is not limited to the embodiments described.

[0021] This invention is a method for predicting the tree species structure of subtropical forests. The data used in this invention were all collected from forest lands in Zhejiang Province, China, located in the subtropical zone. Zhejiang Province is situated on the southeastern coast of China, in a subtropical monsoon humid climate zone (118°01′E ~123°10′E, 27°02′N ~31°11′N), with an average annual temperature of 16.1℃~18.6℃ and an average annual precipitation of 1109.1~2132.3 mm. In 2020, the province's forest area was 60,800 square kilometers, with a forest coverage rate of 61.17% and a forest stock volume of 378 million cubic meters. The main tree species in the forests include oaks (…). 栎属 ), camphor ( 樟树 ), wood lotus ( 木荷 Broadleaf tree species such as Masson pine; 马尾松 ), slash pine ( 湿地松 ), Huangshan pine ( 台湾松 ),Chinese fir( 杉木属 杉木 Coniferous tree species such as )

[0022] The data used in this invention are fixed plots of dominant tree species (arbor forests) from seven CFI (Centralized Tree Index) sampling plots in Zhejiang Province. The CFI interval (step size) is 5 years, and fixed plot data of arbor forests from 1989, 1994, 1999, 2004, 2009, 2014, and 2019 were selected, with a total time span of 30 years. The fixed plots were systematically laid out according to a 4km × 6km grid, and each plot was 800 m². 2 The sample plots are square. There are a total of 4252 sample plots in Zhejiang Province (including non-forest sample plots), and this invention only uses the arbor forest sample plots.

[0023] Tree species in arbor forests are classified into three major categories: pines, firs, and broad-leaved trees. The pines category includes Masson pine. 马尾松 Black Pine 黑松 Huangshan Pine 台湾松 早田 , slash pine 湿地松 etc.; Cedar species include cedar trees 杉木 (拉姆.)胡克 Japanese cedar 柳杉 Metasequoia 水杉 郑 , Chishan 池杉 Yew 南方红豆杉 ( 皮尔格 .) 弗洛林 etc.; broad-leaved trees include oak. 栎属 Castanopsis 栲属 Liquidambar formosana 枫香属 , wood lotus 木荷 camphor tree 樟科 The reason for dividing tree species into three major categories is that these three categories are representative, basically meet the requirements, and are also ecologically significant.

[0024] The trees in each sample plot were categorized by major species, and the volume ratio of each major species category in each sample plot was calculated. Within each sample plot, the sum of the volume ratios of the three major species categories was equal to 1. Specifically, for all normally growing arbor trees with a diameter at breast height (DBH) ≥ 5.0 cm in the sample plot, each tree was measured, and relevant factors were recorded, including tree number, standing type, measurement type, species name, and DBH. Based on the DBH of each tree species, the volume of a single tree (i.e., standing volume) was calculated using a one-dimensional volume table model applicable to Zhejiang Province. The total volume of the sample plot was then calculated, and the volume ratio of each major species category was calculated.

[0025] The data from the seven inventory surveys were divided into six review periods, named in the format of "Review Period Start Year - Review Period End Year," namely 1989-1994, 1994-1999, 1999-2004, 2004-2009, 2009-2014, and 2014-2019. Each review period consisted of two consecutive periods; for example, the 1989-1994 review period included data from both the 1989 and 1994 reviews. Then, adjacent review plots were paired. The land type at the start of each review period had to be arbor forest, while the land type at the end of the review period was not critical, but the plots had to contain trees with a diameter at breast height (DBH) of ≥5cm.

[0026] The number of re-measured plots in each review period is shown in Table 1. The number of plots in the first three periods was relatively smaller than that in the latter three periods. This is mainly because in the three surveys conducted in 1994, 1999, and 2004, one-third of the plots were relocated each time, meaning one-third of the plots were abandoned and no longer investigated, while the same number of new plots were established. In the latter three review periods, the location of the plots remained unchanged, and the number of re-measured arbor forest plots increased in each of the latter three review periods. This is due to the increase in arbor forest area and forest coverage rate caused by afforestation and greening projects and forest closure for natural regeneration throughout the province.

[0027] Table 1. Data of arbor forest plots during each review period

[0028]

[0029] Note: The number of dominant plots for each tree species category refers to the number of plots with the largest proportion of timber volume for each tree species category.

[0030] Table 2 shows the changes in the actual average volume ratio of the three major tree species in the seven CFI surveys. It is not difficult to find that over the past 30 years, the tree species structure of the arbor forest has shown a trend of decreasing average volume ratio of pine and fir, and increasing proportion of broad-leaved trees. Specifically, (1) Over the past 30 years, the volume of pine in the province reached its highest value in 1994 (0.581533), and then continued to decrease, gradually declining from an early dominant position to a disadvantage. (2) The volume of fir in the province fluctuated, showing a fluctuating increase from 1989 to 2004, and reaching its highest peak in 2004 (0.320324), and then continued to decline. (3) The volume of broad-leaved trees in the province maintained a stable upward trend, and in the 2009 survey, it surpassed pine and fir to become the largest category. From 1989 to 2004, pine and fir were dominant, while from 2009 to 2019, broad-leaved trees were dominant, and their dominance continued to expand.

[0031] Table 2. Changes in the actual average volume ratio of the three major tree species categories in the seven CFI surveys.

[0032]

[0033] Based on the volume ratio data of various tree species obtained from previous survey plots, the following basic model (2) is established:

[0034] (2)

[0035] In equation (2), y 10 , y 20 , y 30 These represent the volume proportions of pine, fir, and broadleaf timber in the two consecutive surveys. y 11 , y 21 , y 31 These represent the volume ratios of pine, fir, and broadleaf timber in later stages, satisfying the following conditions: , And 0≤ y 11 , y 21 , y 31 , y 10 , y 20 , y 30 ≤1. , , These are the 12 parameters of the model;

[0036] according to ,Right now:

[0037] The basic model of equation (2) is adjusted to the proportional model of equation (3):

[0038] (3)

[0039] Then simplify equation (3) to equation (4):

[0040] (4)

[0041] Let the relevant parameters of equation (4) be:

[0042] (5)

[0043] Equation (6) can be obtained from equation (5):

[0044] (6)

[0045] Based on equation (6), equation (4) can be rewritten as equation (7):

[0046] (7)

[0047] according to , eliminate Equation (8) is obtained:

[0048] (8)

[0049] make: Then the prediction model of equation (1) is finally obtained:

[0050] (1)

[0051] In equation (1), t 1- t 6 is a parameter.

[0052] Using STATA software (Stata 16.0, StataCorp LLC, College Station, TX, USA), the survey data from two adjacent periods of the systematically distributed sampling population were used as modeling data to fit and calculate the parameters. ; In obtaining parameters When setting parameters The initial values ​​of all parameters are 1; the parameters obtained by fitting are shown in Table 3.

[0053] Table 3. Parameters obtained from the fitting

[0054]

[0055] Since the data collection interval of the sample plot survey is 5 years, the measured data of each survey sample plot is substituted into the prediction model of formula (1) for iterative calculation, and the data of the next 5 years or several years later with a step size of 5 years can be obtained, which is the predicted value; the average value of the predicted value of the tree species ratio of all sample plots in the sampling population is calculated to obtain the predicted value of the tree species ratio of the population.

[0056] Specifically, for example, for the data from two adjacent periods during the 1989-1994 review period, namely the data from 1989 and 1994, the volume ratio values ​​of the three major tree species in all arbor forest plots were first used as modeling sample data. The STATA software was then used to fit the above prediction model to obtain the parameters. t 1- t6. Using all data from 1994 as the independent variable (early stage), the volume ratio value for 1999 was calculated as the predicted value (later stage). The average of the predicted values ​​for the proportion of each tree species in all sample plots of the overall sample was calculated to obtain the predicted value for the proportion of each tree species in the overall sample in 1999. Then, using the predicted value for the volume ratio of all sample plots in 1999 as the independent variable (early stage), the calculation for 2004 was iteratively performed. This process was repeated iteratively to calculate the predicted values ​​for the volume ratio in 2009, 2014, and 2019, and the predicted average value for the proportion of each tree species in the overall sample was calculated. Similarly, the above operations were performed on the data from each of the survey periods: 1994-1999, 1999-2004, 2004-2009, and 2009-2014. The parameters were calculated to obtain the prediction model and the predicted values ​​were obtained. By comparing the predicted values ​​with the actual measured values ​​of the year, the predictive function of the model can be evaluated. The results are shown in Table 4. In Table 4, the proportion data are predicted values, and the error is the relative error between the predicted values ​​and the actual values ​​in Table 2. Positive numbers indicate that the predicted values ​​are too high, and negative numbers indicate that they are too low. This method of estimating the next period's data based on the current period's data is similar to the method of succession from one period to the next in Markov theory. It is only related to the previous state and not to the previous state.

[0057] Table 4 Comparison of Model Predictions and Actual Values

[0058]

[0059] Note: All volume ratios mentioned above are averages.

[0060] Table 4 shows that the prediction errors of the models in the first two review periods (1989-1994, 1994-1999) were larger than those of other models, especially in predicting 2009 and beyond, where the errors deviated significantly from reality. This indicates that the trends in the first two review periods differed considerably from those in the subsequent review periods. In the third to fifth review periods (1999-2004, 2004-2009, 2009-2014), the models showed smaller errors in predicting 2019, particularly in the latter two review periods. This suggests that the trends in the latter four review periods were similar.

[0061] As shown in Table 4, the model has achieved good prediction accuracy from the 1999-2004 review period to 2019, especially from the 2004-2009 review period model, where the relative error of the prediction is below 2%. Therefore, for relatively stable forest ecosystems, the prediction model of this invention has high prediction accuracy for at least 15 years.

[0062] The predicted values ​​for the broadleaf timber volume ratio or the sum of the predicted pine and fir timber volume ratios in the overall tree species proportion predictions are compared with the critical value of 0.65 for mixed coniferous and broadleaf forests (see National Forestry and Grassland Administration, "Technical Regulations for Continuous Forest Resources Inventory," GB / T 38590-2020, 2020-03-06). When the predicted value is greater than 0.65, human intervention measures need to be implemented in actual forestry management. For example, when the predicted broadleaf timber volume ratio is greater than the critical value of 0.65 for mixed coniferous and broadleaf forests, that is, when the sum of the predicted pine and fir timber volume ratios is less than 0.35, it indicates that the overall forest stand is transforming from a mixed coniferous and broadleaf forest with a large biomass and rich biodiversity to a broadleaf forest. At the same time, the region's timber supply capacity will decrease. Therefore, intervention measures such as artificial afforestation or forest tending are needed, such as artificial selection of afforestation tree species, management of target tree species forests, and cultivation of artificial fir and pine timber forests, to achieve continuous improvement in the quality of the forest ecosystem and a continuous and stable supply of ecological products and timber.

[0063] The following example illustrates the use of the prediction method of this invention for prediction. For instance, using a model fitted with data from the 2014-2019 review period for iterative prediction, the volume ratio of major categories of arbor forest tree species is predicted to 2119, i.e., a prediction of 100 years (2019-2119). The predicted data are shown in Table 5, and a curve graph is generated as follows. Figure 1 As shown, based on the sample plot prediction results, the average sample plot volume ratio for the predicted year is then calculated. From Figure 1 It is not difficult to observe that the volume ratio of pine trees decreases monotonically, the volume ratio of broad-leaved trees increases monotonically, and the volume ratio of fir trees initially increases slightly before decreasing monotonically. Figure 1 The data shows that the tree species structure will become increasingly stable over the next 20 years. Furthermore, starting in 2034, the predicted volume ratio of broadleaf trees will exceed the critical value of 0.65 for mixed coniferous and broadleaf forests, meaning the sum of the predicted volume ratios of pines and firs will be less than 0.35. Therefore, attention needs to be paid to coniferous forests, and preparations should be made for the artificial cultivation of firs and pines, as well as for forest management and tending with firs and pines as the target species, and for strengthening the cultivation of artificial timber forests of firs and pines.

[0064] Table 5. Tree species structure ratio predicted for the next 100 years based on data from the 2014-2019 review period (2019-2119)

[0065]

[0066] The predictive model of this invention is based on CFI plot data deployed according to the principle of systematic sampling, thus the prediction results have sampling statistical properties. Based on CFI data, this invention establishes a nonlinear difference equation to explore the structural evolution patterns of major arbor forest species in subtropical regions and predicts the changing trends of arbor forest species volume ratios up to 2119. CFI data covers a large environmental gradient, which is valuable for using models to analyze the evolution patterns of tree species structure. It is reliable at the overall level and suitable for macro-regional applications. Furthermore, CFI fixed plot data from China and other countries around the world can create excellent conditions for the widespread application of the predictive model.

[0067] Unless otherwise specified, all technologies mentioned above refer to existing technologies.

[0068] As described above, although the invention has been shown and described with reference to specific preferred embodiments, it should not be construed as limiting the invention itself. Various changes in form and detail may be made without departing from the spirit and scope of the invention as defined in the appended claims.

Claims

1. A method for predicting the tree species structure of subtropical forests, characterized in that, Includes the following steps: 1) The tree species in the arbor forest are divided into three major categories: pine, fir, and broadleaf. Based on the volume ratio data of each major category of tree species obtained from the sample plots of the two previous surveys, the following prediction model (1) is established: In equation (1), y 10 y 20 These represent the volume ratios of pine and fir trees in the two consecutive surveys, respectively. 11 y 21 y 31 Let y represent the volume ratios of pine, fir, and broadleaf trees in the later stages, respectively, and use them as predicted values, satisfying: 11 +y 21 +y 31 =1, and y 11 y 21 y 31 y 10 y 20 All are greater than or equal to 0 and less than or equal to 1; t1, t2, t3, t4, t5, and t6 are the six parameters of the prediction model; 2) Using the two-period survey data of the sampling population system as modeling data, the parameters t1, t2, t3, t4, t5, t6 were obtained by fitting the data using STATA software; 3) Substitute the measured data of each sample plot into the prediction model of formula (1) for iterative calculation to obtain the volume ratio of arbor forest species in future years with several step lengths of modeling data interval, and use it as the predicted value of the tree species ratio in future years of the sample plot; calculate the average value of the predicted values ​​of the tree species ratio of all sample plots in the sampling population to obtain the predicted value of the tree species ratio of the population. 4) Compare the sum of the predicted values ​​of the broadleaf forest volume ratio or the predicted values ​​of the pine and fir forest volume ratio in the overall tree species ratio prediction values ​​with the critical value of 0.65 for mixed coniferous and broadleaf forests. When the predicted value is greater than 0.65, human intervention measures need to be implemented in actual forestry management.

2. The method for predicting the tree species structure of subtropical arbor forests according to claim 1, characterized in that, In step 1), the pine species include one or more of Masson pine, black pine, Huangshan pine, and slash pine; the fir species include one or more of Chinese fir, Japanese cedar, dawn redwood, pond cypress, and yew; the broad-leaved species include one or more of Schima superba, Liquidambar formosana, oak, Castanopsis fargesii, and camphor tree; the volume is calculated by measuring each tree of a normally growing arbor forest species with a diameter at breast height ≥ 5.0 cm using a one-dimensional volume table model.

3. The method for predicting the tree species structure of subtropical arbor forests according to claim 1, characterized in that, In step 2), when obtaining parameters t1, t2, t3, t4, t5, t6, the initial values ​​of parameters t1, t2, t3, t4, t5, t6 are all set to 1.

4. The method for predicting the tree species structure of subtropical arbor forests according to claim 1, characterized in that, In step 3), the step size is 5 years.

5. The method for predicting the tree species structure of subtropical arbor forests according to claim 1, characterized in that, In step 4), the human intervention measures are artificial afforestation or forest tending.

Images