WO2018193984A1

WO2018193984A1 - Method for predicting productivity of afforested area

Info

Publication number: WO2018193984A1
Application number: PCT/JP2018/015526
Authority: WO
Inventors: 英治岩田; 悠生久永; 澁澤　栄; 正和小平
Original assignee: 日本製紙株式会社; 国立大学法人東京農工大学
Priority date: 2017-04-17
Filing date: 2018-04-13
Publication date: 2018-10-25
Also published as: JPWO2018193984A1; JP6708789B2

Abstract

The objective of the present invention is to provide a method capable of rapidly, simply and accurately predicting the productivity of an afforestation candidate area, including the productivity of planted trees in an afforested area. The present invention provides a method for predicting the productivity of an afforestation candidate area, including (A) to (D): (A): sort soil component values and soil spectra in reference forest area by means of multivariate analysis to obtain a correlation between soil component values and soil spectrum; (B): sort productivity, topographic information and soil component values in reference forest area by means of multivariate analysis to obtain a correlation between productivity, topographic information and soil component values; (C): from the soil spectrum of the afforestation candidate area, obtain a predicted value of the soil component values on the basis of the correlation obtained in (A); and (D): obtain a predicted value of the productivity of the afforestation candidate area on the basis of the topographic information of the afforestation candidate area and either the soil component values of the afforestation candidate area or the predicted values of the soil component values obtained in (C), on the basis of the correlation obtained in (B).

Description

How to predict plantation productivity

The present invention relates to a method for predicting plantation productivity.

It is important to predict the productivity of plantations in advance in order to efficiently manage afforestation and produce production plans in industrial plantations for the purpose of timber production and environmental plantations such as forest conservation and environmental conservation activities. is there. In industrial afforestation, in order to maximize the growth amount (accumulated amount) of planted trees and reduce costs, in environmental afforestation, in order to select tree species and reduce costs, land and operations suitable for planting Is required.

Patent Document 1 includes ground altitude, soil hardness, moisture content in soil, redox potential, pH, conductivity, C content, N content, ratio of C and N, Ca content, Mg content The height of planted trees in the plantation after a predetermined number of years, based on a plurality of parameters related to the growth of mangroves in the plantation environment, such as the amount, K content, Na content, etc. and the height of the mangrove in this plantation environment A method for creating a judgment formula for estimating the tree height and predicting the tree height using the judgment formula from the measured values of the parameters is described. Non-Patent Document 1 describes that the soil depth is greatly related to the productivity of planted radiata pine in relation to the balance between soil and water and nutrient accumulation, and that the phosphorus atomic weight of leaves is determined by a specific judgment formula. It is described that tree height can be predicted from soil depth, pH, and soil total nitrogen content.

In Patent Document 2, an explanatory variable is obtained from the analysis value of the color, characteristics, and nutrients of the soil near the target area, and a calibration equation with the growth data of the tree near the target area is obtained from the explanatory variable. A method for predicting the growth degree of trees in the target area by applying each analysis result of the target area to the equation is described. In Patent Document 3, a soil spectrum is obtained from reflected light obtained by irradiating light on an analysis target soil, and is compared with a characteristic spectrum obtained from a waveform of a soil spectrum similarly obtained from a plurality of other soils. Describes that the characteristics of soil can be analyzed. Non-Patent Document 2 describes a method of estimating a soil analysis value by measuring a visible light-near infrared light soil diffusion spectrum using a tractor-mounted soil analysis system and performing a PLS regression analysis.

JP 2008-206421 A International Publication No. 2016/43007 JP 2006-38511 A

However, all of the existing technologies focus only on the height of planted trees or the properties of soil as prediction targets, and are not methods for consistently evaluating the productivity of planted land and planted trees. Possible reasons for the lack of progress in developing methods for predicting the productivity of plantations and trees are as follows: (1) In forestry, soil components and topography are closely and complexly related; 2) It is common to use all of the land that is owned, and the selection of the land to be used is not assumed; (3) The logging season is long in forestry (usually 40-50 years in domestic forestry), and current logging Prior to planting the planted trees that had reached the season, there was no means to assess the productivity of the land and the need for such means had not been sought; and (4) There was variation, and the growth could not be accurately predicted after planting.

This invention solves such a subject and aims at providing the method of predicting the productivity of the afforestation candidate site including the productivity of afforestation trees quickly, simply and accurately.

The present invention provides the following.
[1] Method for predicting productivity of candidate plantations including (A) to (D):
(A): Obtaining a correlation between the soil component value and the soil spectrum by arranging the soil component value and the soil spectrum in the reference forest by multivariate analysis;
(B): Organizing productivity, terrain information, and soil component values of the reference forest by multivariate analysis, and obtaining correlation between productivity, terrain information, and soil component values;
(C): Obtaining predicted values of soil component values based on the correlation obtained in (A) from the soil spectrum of the proposed plantation site; and (D): Topographic information of the proposed plantation site and the proposed plantation site From the soil analysis value or the predicted value of the soil component value obtained in (C), a predicted value of the productivity of the proposed plantation site is obtained based on the correlation obtained in (B).
[2] The soil component value is the water content ratio, organic matter content, pH, phosphorus atom content, potassium atom content, calcium atom content, magnesium atom content, sodium atom content of the soil of the reference forest. Aluminum atom content, hydrogen atom content, aluminum saturation, acidity, total exchangeable base, base substitution capacity, base saturation, sulfur atom content, boron atom content, copper atom content, iron atom content, manganese Includes at least one selected from the group consisting of atomic content, zinc atomic content, total nitrogen content, ammonia nitrogen content, chlorine atom content, clay content, silt content, sand content, and total carbon content The prediction method according to [1].
[3] The prediction method according to [1] or [2], wherein the productivity is an annual growth amount of the tree.
[4] The prediction method according to any one of [1] to [3], wherein the topographic information includes at least one selected from an inclination angle and an altitude.
[5] The prediction method according to any one of [1] to [4], wherein the multivariate analysis is a PLS regression analysis.
[6] The correlation between the soil component value and the soil spectrum in (A) is an equation for estimating the soil component value from the soil spectrum, with the soil component value as the objective variable and the soil spectrum as the explanatory variable. The prediction method described.
[7] Correlation between productivity, topographic information and soil component value in (B) is a candidate for plantation from soil component value and topographic information, with productivity as objective variable and soil component value and topographic information as explanatory variables The prediction method according to [5] or [6], wherein the prediction method is a ground productivity estimation formula.
[8] (C) is to obtain the predicted value of the soil component value by applying the soil spectrum of the candidate plantation site to the soil component value estimation formula obtained in (A), in [6] or [7] The prediction method described.
[9] (D) is to obtain the predicted value of productivity of the candidate plantation site by applying the topographical information and the soil component value of the candidate plantation site to the productivity estimation formula obtained in (B), [6] ] The prediction method according to any one of [8] to [8].
[10] The prediction method according to any one of [1] to [9], wherein the afforestation candidate site is a candidate plantation site for a Eucalyptus plant.
[11] Obtain a predicted productivity value of at least one afforestation candidate site by the prediction method according to any one of [1] to [10], and select an afforestation candidate site with a high productivity predictive value. A method of selecting plantations, including selecting as.
[12] An afforestation method including afforesting in an afforestation area selected by the selection method according to [11].
[13] A method for producing a tree, comprising producing a tree in a plantation selected by the selection method according to [11].
[14] A candidate plantation site prediction system that can implement the method according to any one of [1] to [10], including a soil spectrum measurement device, a soil component value measurement device, and data processing means.

According to the present invention, it is possible to quickly and easily evaluate the productivity of candidate plantations from soil component values, soil spectra, topographic information, etc., and select plantations. In particular, it is possible to evaluate the productivity of the afforestation candidate site based not only on the tree height but also on the growth amount of the afforestation tree including the breast height diameter.

FIG. 1 is a diagram showing a test procedure of the example. FIG. 2 is a diagram showing a soil spectrum (before pretreatment) obtained in the examples. FIG. 3 is a diagram showing a soil spectrum (after pretreatment) obtained in the examples. FIG. 4 is a diagram showing the estimation accuracy of productivity from the actually measured value of the soil component value in the example. FIG. 5 is a diagram showing the estimation accuracy of the potassium atom content in the examples. FIG. 6 is a diagram showing the estimation accuracy of the sulfur atom content in the examples. FIG. 7 is a diagram illustrating the estimation accuracy of the clay content in the example. FIG. 8 is a diagram showing the estimation accuracy of productivity from the estimated value of the soil component value in the example.

The method for predicting productivity of a candidate plantation site of the present invention includes the following steps:
Step (A): Obtaining a correlation between soil component values and soil spectra by arranging soil component values and soil spectra in the reference forest by multivariate analysis;
Process (B): Organizing productivity, topographic information, and soil component values of the standard forest by multivariate analysis to obtain a correlation between productivity, topographic information, and soil component values;
Step (C): Obtaining the predicted value of the soil component value from the soil spectrum of the candidate plantation site based on the correlation obtained in (A); and Step (D): Topographical information of the candidate plantation site, and planting Obtaining the predicted value of productivity of the candidate plantation site from the soil analysis value of the candidate site or the predicted value of the soil component value obtained in (C) based on the correlation obtained in (B).

[Process (A)]
In the step (A), the soil component value and the soil spectrum in the reference forest are arranged by multivariate analysis, and the correlation between the soil component value and the soil spectrum is obtained.

The reference forest land may be an area where soil component values, topographic information, and productivity can be obtained, and may be at least a part of the forest land. The forest land may be either a natural forest or an artificial forest, but an artificial forest (planting area) is preferred. The area of the reference forest is not particularly limited, but is usually 50 m ² or more. Further, the reference forest is preferably a place where position information can be managed by GPS (Global Positioning System) or DGPS (Differential Global Positioning System). The reference forest preferably meets one of the following requirements: the climate is the same or similar to the proposed plantation; the topography is the same or similar to the proposed plantation; the distance from the proposed plantation is close ; Trees of the same or related species are growing in the proposed plantation site.

The tree growing in the reference forest is not particularly limited as long as it is a woody plant. For example, Eucalyptus plant, Cryptomeria plant, Pineus plant, and Prunus plant. Plants (such as cherry (Prunus spp.), Ume (Prunus mume), Prunus tomentosa), Mango genus (Mangifera indica), Acacia (Acacia) plant, Yamamo genus, Yamamo genus Plants, Quercus plants (such as Quercus acutissima), grape (Vitis) plants, apple (Malus) plants, rose (Rosa) plants, camellia (C melia plant, Jacaranda plant (Jacaranda mimosifolia etc.), crocodile (Persea) plant (Avocado (Persea americana) etc.), Hinoki genus (Chamaecyparis) plant, L (Abies) plant and Quercus plant are mentioned, and among these, Eucalyptus plant, Cedar plant and Pine plant are preferable.

The soil component value may be a numerical value indicating the properties of the soil, for example, water content ratio, organic matter content, pH, phosphorus atom content, potassium atom content, calcium atom content, magnesium atom content, sodium atom content Amount, aluminum atom content, hydrogen atom content, aluminum saturation, acidity, total exchangeable base, base substitution capacity, base saturation, sulfur atom content, boron atom content, copper atom content, iron atom content Manganese atom content, zinc atom content, total nitrogen content, ammonia nitrogen content, chlorine atom content, clay content, silt content, sand content, and total carbon content, potassium atom content, Sulfur atom content and clay content are preferred. The soil component value may be at least one of the above, but a combination of 2 or more is preferable, and a combination of 3 or more is more preferable. There is no particular upper limit, and it can be appropriately determined according to the analysis accuracy.

The soil component value in the reference forest can be obtained by measuring in the reference forest. A measuring method is not specifically limited, What is necessary is just to utilize the method used in the case of the measurement of each soil component value.

The soil component value may be a measurement value obtained at one measurement point selected from the reference forest, but is preferably an average value of measurement values obtained at two or more measurement points. The measurement point is preferably 10 or more, more preferably 15 or more, still more preferably 20 or more, and even more preferably 25 or more.

The soil spectrum is a spectrum obtained when the soil is irradiated with light. The soil spectrum is usually obtained by irradiating light and measuring reflected light. The wavelength region of the soil spectrum is usually a visible light or near infrared light region, preferably 350 to 2500 nm, more preferably 500 nm to 2500 nm, and still more preferably 500 to 1600 nm. For the measurement of the soil spectrum, any device that can irradiate light such as near-infrared light and analyze the transmitted light or diffuse reflected light spectrum may be used. (For example, FieldSpec4 (ASD), SAS (Shibuya Seiki), etc.) can be used.

The measurement point in the reference forest of the soil spectrum is preferably 2 or more, more preferably 10 or more, still more preferably 15 or more, still more preferably 20 or more, and particularly preferably 25 or more.

The soil spectrum is preferably pretreated prior to multivariate analysis. As a result, variations between samples can be corrected and the influence of noise, outliers, and the like can be excluded, so that data quality can be improved. Examples of preprocessing include smoothing processing, differentiation processing (for example, primary differentiation processing, secondary differentiation processing), correction processing (for example, multiple scattering correction processing (MSC), standard normal variable processing (SNV), extended multiple scattering correction). Processing (EMSC), baseline correction processing), averaging processing, auto scale processing, range scale processing, dispersion scale processing, normalization processing, common logarithm processing (log 10), multiplication processing, and subtraction processing.

Examples of multivariate analysis used to organize soil component values and soil spectra in the reference forest include, for example, partial least squares (PLS) regression analysis, multiple regression analysis, principal component analysis (PCS), principal component regression analysis (PCR), And Fourier transform analysis. Among these, the problem of multicollinearity that can occur in other methods such as multiple regression analysis can be avoided, estimation with a large number of explanatory variables such as spectra, estimation with a small number of samples, prediction PLS regression analysis is preferable because of its high performance.

An example of the step (A) when the multivariate analysis is PLS regression analysis is shown below. According to the regression model in the PLS regression analysis, a soil spectrum is used as an explanatory variable, a matrix is created using the soil component value as an objective variable, the explanatory variable is converted based on the PLS algorithm, and is substituted into the formula of the PLS regression model. A soil component value estimation formula (calibration curve, calibration formula model) is obtained from the spectrum. The PLS algorithm and regression model in PLS regression analysis are, for example, “Model selection in PLS regression” (Yuki Hashimoto, Yutaka Tanaka, Academia. Bulletin of Information Science and Engineering: Bulletin of Nanzan University, 2010, http: //www.seto. nanzan-u.ac.jp/st/nas/academia/vol_010pdf/10-039-049.pdf).

[Process (B)]
In the step (B), the productivity of the reference forest, the topographic information, and the soil component value are arranged by multivariate analysis, and the correlation between the productivity, the topographic information, and the soil component value is obtained.

The reference forest land is as described in the process (A), and the reference forest land in the process (A) is the same as the reference forest land in the process (B).

The productivity of the standard forest is preferably expressed numerically, more preferably the annual growth of trees growing in the standard forest, the monthly growth, etc., and the productivity generally used in forestry Since it is an index, annual growth is more preferable. For example, the annual growth rate can be calculated by measuring the tree height and breast height diameter of the tree, estimating the volume of the desired tree using a growth curve, and the accumulated amount ( There is a method for estimating the volume (volume), and the method shown in the examples is preferable because the felling period and the age of the tree can be omitted.

The productivity may be a measurement value obtained at one measurement point selected from the reference forest, but is preferably an average value of measurement values obtained at two or more measurement points. The measurement point is preferably 10 or more, more preferably 15 or more, still more preferably 20 or more, and even more preferably 25 or more.

The topographic information of the reference forest is preferably a numerical value representing the topography. The terrain information includes, for example, the inclination angle, altitude, direction, soil layer depth, and soil layer hardness. The inclination angle and altitude are preferable, and the altitude is more preferable. The topographic information may be obtained as a measured value, or may be obtained using a geographic information analysis system provided by a public institution or the like. The method for measuring the terrain information is not particularly limited, and can be a known method.

The terrain information may be one terrain information selected from the reference forest, but is preferably an average value of two or more terrain information. The terrain information preferably has an average value of 10 or more locations, more preferably an average value of 15 or more locations, more preferably an average value of 20 or more locations, and even more preferably an average value of 25 or more locations.

Specific examples and preferred examples similar to the multivariate analysis in the step (A) are given as the multivariate analysis for organizing the productivity of the reference forest, topographic information, and soil component values. An example of the step (B) when the multivariate analysis is a PLS regression analysis is shown below. Using soil component values and topographic information as explanatory variables, creating a matrix with productivity as the target variable, converting the explanatory variables based on the PLS algorithm, and substituting them into the PLS regression model formula, soil component values and topographic information The productivity estimation formula of the candidate plantation site is obtained.

[Process (C)]
In step (C), a predicted value of the soil component value is obtained from the soil spectrum of the candidate plantation site based on the correlation obtained in (A).

The soil spectrum of the proposed plantation site and the measurement method thereof are the same as the definition and specific example described in the step (A). The soil spectrum of the afforestation candidate site is preferably an average value of measured values obtained at two or more measurement points of the afforestation candidate site. The measurement point is preferably 10 or more, more preferably 15 or more, still more preferably 20 or more, and even more preferably 25 or more. When the estimation formula of the soil component value is obtained in (A), the predicted value of the soil component value can be obtained by applying the measured value of the soil spectrum of the candidate plantation site to the formula.

[Process (D)]
In step (D), from the topographic information of the candidate plantation site and the soil analysis value of the candidate plantation site or the predicted value of the soil component value obtained in (C), based on the correlation obtained in (B), A predicted value of productivity of the candidate plantation site is obtained.

The topographical information of the proposed plantation site is the same as the definition and specific example described in the process (B). The soil analysis value of the candidate plantation site may be a measured value or a predicted value. That is, it may be a measurement value obtained by actually measuring from the soil analysis value of the candidate plantation site, or obtained by applying the topography information of the candidate plantation site to the estimation formula of the soil component value obtained in (C). It may be a predicted value. The soil spectrum of the candidate plantation can be measured at the candidate plantation. The method for measuring the soil spectrum is the same as that described in the step (A). When the formula for estimating the productivity is obtained in (B), the predicted value of the productivity of the plantation is obtained by applying the soil analysis value of the plantation candidate site and the soil spectrum of the plantation candidate site to the formula. Can do.

According to the present invention, since the productivity of various plantation candidate sites can be predicted quickly, simply and accurately, afforestation and tree production can be promoted systematically. Specifically, the productivity of the afforestation candidate site can be predicted based not only on the tree height but also on the amount of planted tree growth including the breast height diameter. For example, a predictive value of productivity of at least one afforestation candidate site can be obtained by the above prediction method, and afforestation candidate sites with high productivity predictive values can be selected as afforestation sites. When there are two or more candidate plantations, the predicted productivity value of each candidate plantation is obtained, and the candidate plantation with the higher predicted value may be selected as the plantation site. Thereby, it is possible to carry out afforestation systematically in the selected plantation, that is, to produce trees.

According to the present invention, there is also provided a prediction system including a soil spectrum measurement device, a soil component value measurement device, and data processing means for carrying out the above method for predicting a candidate plantation site. The soil component value measuring device may be any device that can measure the soil component value obtained by the prediction method. The prediction system may further include a data storage device. The data storage device may store each measured value (including average value, highest value, lowest value, and deviation), each predicted value, and each correlation (for example, an estimation formula) in steps (A) to (D). it can. The prediction stem may further include a computing device. The arithmetic device is a device that can calculate each correlation and predicted value in steps (A) to (D).

Example 1 [Prediction of Eucalyptus Plantation Productivity]
A productivity estimation test was conducted in 5.5 years (total 30 points) of Eucalyptus plantation in Amapa State, Brazil. The test was performed according to the flowchart (FIG. 1).

[Calculation of annual growth of trees at the test site]
Annual growth (m ³ / ha / year) was used as the productivity of plantations. The annual growth was calculated according to the following procedure. First, a single eucalyptus tree species inhabiting each of 30 test sites, tree heights of 40 lines (Eucalyptus urograndis), and breast height diameter were measured. The annual growth was calculated according to the following procedure. The single wood volume at the time of the survey was calculated based on the following formula, and the average value of each test site was obtained. Next, multiply the number of trees planted per ha (1000-1500 depending on the land), calculate the eucalyptus volume growth per 1 ha of the test area, divide by the test period (year), and the growth volume (annual volume growth volume) (M ³ / ha / year)) was calculated.

[formula]
Single wood volume = (chest height diameter (m)) ² x tree height (m) x coefficient Coefficient: 0.3 to 0.4

[Collecting topographic information of the test site]
As topographic information, elevation and inclination angles were collected at 30 points on the test site. Both elevation and inclination angle were obtained from geographic information analysis software (ArcGIS ESRI).

[Measurement of soil component value of test site]
At 30 points of the test site, soil from the ground to a depth of 20 cm was sampled to obtain a soil sample. About each soil sample, each soil component value of 28 items shown in Table 1 was obtained by the conventional method.

[Measurement of soil spectrum of standard forest land]
About each soil sample, the spectrum measurement was implemented by FieldSpec4 (ASD company) (FIG. 2). The measured values were subjected to preprocessing (smoothing processing, differentiation processing, etc.) using “The Unscrambler X” (Camo Software Japan Co., Ltd.), and the obtained correction values were used in the following multivariate analysis (FIG. 3). ).

[Soil component value estimation formula]
Multivariate analysis (PLS regression) using “The Unscrambler X” (Camo Software Japan) with soil component values (K content, S content and clay) as objective variables and corrected soil spectrum as explanatory variables Analysis) and a soil component value estimation formula was created.

[Productivity estimation formula]
Using productivity as an objective variable, soil component values (K content, S content and clay) and topographic information (tilt angle) as explanatory variables, multivariate analysis (PLS analysis) was performed to create a productivity estimation formula.

FIG. 4 shows the correlation between the predicted productivity value obtained by the productivity estimation formula using the actual measured soil analysis value and the actual measured productivity value. As is clear from FIG. 4, the actual productivity and the productivity obtained from the estimation formula showed a high correlation.

[Measurement of soil spectrum of candidate plantations, collection of topographical information, estimation of soil component values, and prediction of productivity]
Eucalyptus plantation 3.5-year-old (total 9 points) is selected as a candidate planting site, and the soil spectrum is measured in the same manner as described for the test site, and the soil component value is estimated from the soil component value estimation formula. Got the value. In addition, terrain information was collected, and productivity was estimated from the obtained terrain information and estimated values of soil components based on the productivity estimation formula. 5 to 7 show the correlation between the estimated value of potassium content, the estimated value of sulfur content, the estimated value of clay content, and the estimated value of clay content obtained by the soil component value estimation formula. FIG. 8 shows the correlation between the estimated productivity value obtained from the terrain information and the estimated values of the respective soil components using the productivity estimation formula and the measured productivity value. As is apparent from FIGS. 5 to 7, the actual soil component value and the soil component value obtained from the estimation formula showed a high correlation. As is apparent from FIG. 8, the estimated productivity value obtained by the productivity estimation formula and the actual measured productivity value showed a high correlation.

Claims

Method for predicting productivity of candidate plantations including (A) to (D):
(A): Obtaining a correlation between the soil component value and the soil spectrum by arranging the soil component value and the soil spectrum in the reference forest by multivariate analysis;
(B): Organizing productivity, terrain information, and soil component values of the reference forest by multivariate analysis, and obtaining correlation between productivity, terrain information, and soil component values;
(C): Obtaining predicted values of soil component values based on the correlation obtained in (A) from the soil spectrum of the proposed plantation site; and (D): Topographic information of the proposed plantation site and the proposed plantation site From the soil analysis value or the predicted value of the soil component value obtained in (C), a predicted value of the productivity of the proposed plantation site is obtained based on the correlation obtained in (B).
The soil component value is the water content ratio, organic matter content, pH, phosphorus atom content, potassium atom content, calcium atom content, magnesium atom content, sodium atom content, aluminum atom content of the soil of the reference forest. Amount, hydrogen atom content, aluminum saturation, acidity, total exchangeable base, base substitution capacity, base saturation, sulfur atom content, boron atom content, copper atom content, iron atom content, manganese atom content And at least one selected from the group consisting of zinc atom content, total nitrogen content, ammonia nitrogen content, chlorine atom content, clay content, silt content, sand content, and total carbon content. The prediction method according to 1.
The prediction method according to claim 1 or 2, wherein the productivity is an annual growth amount of the tree.
The prediction method according to any one of claims 1 to 3, wherein the topographic information includes at least one selected from an inclination angle and an altitude.
The prediction method according to any one of claims 1 to 4, wherein the multivariate analysis is a PLS regression analysis.
6. The prediction according to claim 5, wherein the correlation between the soil component value and the soil spectrum in (A) is a soil component value estimation formula from the soil spectrum, where the soil component value is an objective variable and the soil spectrum is an explanatory variable. Method.
The correlation between productivity, terrain information, and soil component values in (B) is the production of candidate plantations from soil component values and terrain information, with productivity as the objective variable and soil component values and terrain information as explanatory variables. The prediction method according to claim 5 or 6, which is a sex estimation formula.
(C) is the prediction method of Claim 6 or 7 which is applying the soil spectrum of a plantation candidate site to the soil component value estimation formula obtained by (A), and obtaining the predicted value of a soil component value.
(D) is to apply the topographical information and the soil component value of the candidate plantation site to the productivity estimation formula obtained in (B) to obtain a predicted value of the productivity of the plantation candidate site. The prediction method according to any one of the above.
The prediction method according to any one of claims 1 to 9, wherein the plantation candidate site is a plantation candidate site of a Eucalyptus plant.
A predictive value of productivity of at least one plantation candidate site is obtained by the prediction method according to any one of claims 1 to 10, and a candidate plantation site having a high productivity predictive value is selected as a plantation site. Including methods for selecting plantations.
A tree planting method, comprising planting trees in a plantation area selected by the selection method according to claim 11.
A method for producing a tree, comprising producing the tree in a plantation selected by the selection method according to claim 11.
A plantation site prediction system capable of performing the method according to any one of claims 1 to 10, comprising a soil spectrum measurement device, a soil component value measurement device, and a data processing means.