Estimation apparatus, estimation method, breeding method, and program

Abstract

To effectively breed organisms to be bred.SOLUTION: An estimation apparatus 100 acquires actual result values including actual result values of candidates of increase / decrease cause factors of organisms to be bred and actual result values of candidates of breeding state factors of the organisms, as actual result values in a breeding farm selected based on at least one of an estimated value and an actual result value of environmental factor in a candidate place from among candidates of artificial breeding farm. The estimation apparatus 100 performs, based on the actual result values, processing including generating a causal relationship model representing a relationship between a plurality of variables including an increase / decrease cause variable indicating a value of increase / decrease cause factor and a breeding state variable indicating a value of breeding state factor, as a process to select at least one increase / decrease cause factor which can be adjusted in the breeding farm.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to an estimation device, an estimation method, a breeding method, and a program, which are particularly suitable for use in breeding specific organisms. [Background technology]

[0002] There is a need to promote the growth of organisms such as algae, seafood, and microorganisms, suppress their decline, and prevent their extinction by propagating them. For example, the area of ​​seaweed beds has been decreasing in various parts of Japan and around the world. This has led to a need to artificially create new seaweed beds. In doing so, it is necessary to investigate the factors that affect the increase or decrease of algae that are the target for propagation in seaweed beds. Non-Patent Document 1 discloses the use of structural equation modeling to derive the factors that affect the increase or decrease of aquatic plants (submerged aquatic vegetation (SAV)) in a specific marine area. [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] Lefcheck, JS, Orth, RJ, Dennison, WC, Wilcox, DJ, Murphy, RR, Keisman, J., Gurbisz, C., Hannam, M., Landry, JB, Moore, KA, Patrick, CJ, Testa, J., Weller, DE, & Batiuk, RA (2018). Long-term nutrient reductions lead to the unprecedented recovery of a temperate coastal region. PNAS, 115(14), 3658-3662. Summary of the Invention [Problem to be solved by the invention]

[0004] However, the technology described in Non-Patent Document 1 targets naturally occurring aquatic plants (submerged aquatic vegetation (SAV)). Therefore, using the technology described in Non-Patent Document 1, it is not easy to select effective breeding sites for the propagation of aquatic organisms. Furthermore, the technology described in Non-Patent Document 1 does not perform structural equation modeling from the perspective of promoting the increase of organisms or suppressing their decline. For example, the path diagram described in Non-Patent Document 1 only shows factors that affect the increase or decrease of aquatic organisms in a specific marine area. Furthermore, the path diagram described in Non-Patent Document 1 collectively represents aquatic plants (SAV) that inhabit a specific marine area. Therefore, using the structural equation modeling described in Non-Patent Document 1, it is not easy to estimate factors that are effective in promoting the increase of target organisms or suppressing their decline.

[0005] The present disclosure has been made in consideration of the above problems, and aims to effectively propagate organisms that are the target of propagation. [Means for solving the problem]

[0006] The estimation device disclosed herein is an estimation device that estimates increase / decrease factor elements including at least one of increase / decrease factor elements and decrease factor elements of a specific organism to be bred, and is equipped with a selected site performance acquisition unit that acquires performance values ​​including performance values ​​of candidate increase / decrease factor elements and actual measured values ​​of candidate reproductive status elements of the organism as actual values ​​at a breeding site selected from candidate sites for an artificial breeding site for the organism based on at least one of estimated values ​​and actual measured values ​​of environmental elements at the candidate site, and an increase / decrease factor selection processing unit that performs processing to select at least one of the increase / decrease factor elements that can be adjusted at the breeding site, including creating a causal relationship model that represents the relationship between multiple variables including increase / decrease factor variables that indicate the values ​​of the increase / decrease factor elements and reproductive status variables that indicate the values ​​of the reproductive status elements, based on the performance values ​​acquired by the selected site performance acquisition unit.

[0007] The estimation method disclosed herein is a method for estimating increase / decrease factor elements, including at least one of an increase factor element and a decrease factor element of a specific organism to be bred, and includes: a selection process for selecting a breeding site from among candidate sites based on at least one of estimated values ​​and actual measured values ​​of environmental elements at candidate sites for artificial breeding sites for the organism; a selected site performance acquisition process for acquiring actual values, including actual values ​​of candidate increase / decrease factor elements and actual measured values ​​of candidate reproductive status elements of the organism, as actual values ​​at the breeding site selected by the selection process; and an increase / decrease factor selection process for performing a process, based on the actual values ​​acquired by the selected site performance acquisition process, to select at least one of the increase / decrease factor elements that can be adjusted at the breeding site, which process includes creating a causal relationship model that represents the relationship between multiple variables, including an increase / decrease factor variable that indicates the value of the increase / decrease factor element and a reproductive status variable that indicates the value of the reproductive status element. The breeding method of the present disclosure breeds the organisms in the breeding ground using the increase / decrease factor elements selected based on the results of processing by the increase / decrease factor selection processing step of the estimation method.

[0008] The program of the present disclosure causes a computer to function as each part of the estimation device. [Effects of the Invention]

[0009] According to the present disclosure, organisms to be propagated can be propagated effectively. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 2 is a diagram illustrating an example of a functional configuration of an estimation apparatus. [Figure 2] 10 is a flowchart illustrating an example of an estimation method. [Figure 3] FIG. 1 is a diagram showing an example of the relationship between the range of values ​​of environmental elements suitable for the growth of Laminaria religiosa and the estimated values ​​of environmental elements in a candidate site for an artificial seaweed bed. [Figure 4]FIG. 10 is a diagram showing an example of candidates for increase / decrease factor elements and candidates for reproductive status elements. [Figure 5A] FIG. 10 is a diagram illustrating an example of a path diagram. [Figure 5B] FIG. 10 is a diagram illustrating an example of a path coefficient. [Figure 6] 10 is a diagram showing an example of the current value (current state) of an increase / decrease factor element and the increase / decrease amount of the increase / decrease factor element. FIG. [Figure 7] FIG. 10 is a diagram showing an example of the results of deriving the distribution area of ​​Laminaria religiosa. DETAILED DESCRIPTION OF THE INVENTION

[0011] Hereinafter, an embodiment of the present disclosure will be described with reference to the drawings. Note that the term "comparison objects" as being the same in terms of length, position, size, spacing, etc., includes not only cases where they are exactly the same, but also cases where they are different within the scope of the present disclosure (for example, differences within the tolerance range determined at the time of design).

[0012] Fig. 1 is a diagram showing an example of the functional configuration of the estimation device 100. Fig. 2 is a flowchart illustrating an example of an estimation method performed using the estimation device 100. The estimation device 100 has, as its hardware, one or more hardware processors such as a CPU (Central Processing Unit), and one or more memories such as a RAM (Random Access Memory) and a ROM (Read Only Memory), and performs various calculations by executing one or more programs stored in the memories using the one or more hardware processors.

[0013] FIG. 1 illustrates a case where the input device 110 and the output device 120 are communicatively connected to the estimating device 100. The communication between the estimating device 100 and the input device 110 and the output device 120 may be wired communication or wireless communication. The communication between the estimating device 100 and the input device 110 and the output device 120 may be communication via a communication network. The estimating device 100 may include the input device 110 and the output device 120. The estimating device 100 may be realized by dedicated hardware such as an ASIC (Application Specific Integrated Circuit) or an FPGA (Field Programmable Gate Array).

[0014] In FIG. 1, this embodiment illustrates an example in which the estimation device 100 includes a candidate site environment estimation unit 101, a selected site performance acquisition unit 102, an increase / decrease factor selection processing unit 103, and an increase / decrease factor adjustment unit 104.

[0015] The estimation device 100 performs processing to estimate the increase / decrease factor factors of the breeding target organisms. The increase / decrease factor factors include at least one of an increase factor factor and a decrease factor factor. The increase factor factor is a factor that causes the breeding target organisms to increase in an artificial breeding site. The decrease factor factor is a factor that causes the breeding target organisms to decrease in an artificial breeding site. The increase / decrease factor factors may be expressed as a physical quantity, a basic statistical quantity, or other quantity. From the perspective of more efficiently increasing the breeding target organisms in an artificial breeding site, it is preferable that the increase / decrease factor factors include an increase factor factor, and it is more preferable that both an increase factor factor and a decrease factor factor are included. However, the increase / decrease factor factors may include a decrease factor factor or not include an increase factor factor. In the following description, the artificial breeding site for the breeding target organisms will be simply referred to as a breeding site as necessary.

[0016] The organisms to be propagated are not limited, and may be, for example, algae, seafood, microorganisms, or other organisms (protists, plants, animals, etc.). The breeding grounds may be constructed by humans in order to propagate the organisms to be propagated. For example, the organisms to be propagated may be moved (from another location) to the breeding ground, and breeding may be carried out using the organisms. Alternatively, eggs of the organisms to be propagated may be moved (from another location) to the breeding ground, and breeding may be carried out using the eggs. Furthermore, a component (such as an enclosure) that restricts access to the breeding ground may be installed in the location where the organisms to be propagated live (or occur naturally). In this case, it is not necessary to move the organisms to be propagated or their eggs from another location to the breeding ground. Furthermore, the organisms to be propagated may be of one type or multiple types.

[0017] As described above, the organisms to be propagated are not limited, but in this embodiment, the case where the organisms to be propagated are a specific type of algae (seaweed, seaweed, etc. (more specifically, for example, Laminaria religiosa or Laminaria japonica)) is taken as an example. Furthermore, in this embodiment, the case where the estimation device 100 performs processing to estimate factors that cause an increase or decrease in a seaweed bed when the seaweed bed is artificially created is taken as an example. In the following description, an artificially created seaweed bed will be referred to as an artificial seaweed bed as necessary.

[0018] <Candidate site environment estimation unit 101, candidate site environment estimation step S201> The candidate site environment estimation unit 101 acquires estimated values ​​of environmental elements in candidate breeding sites. Environmental elements are elements that make up the environment and, for example, represent the environmental conditions of a (candidate for) breeding site. The environmental conditions of a (candidate for) breeding site include at least one of quantitative conditions and qualitative conditions. Quantitative conditions are conditions expressed numerically, such as temperature, speed, concentration, density, number, weight, mass, area, and volume. Qualitative conditions are conditions that are not expressed numerically, such as type and shape. A numerical value corresponding to the content of the qualitative condition may be expressed as the value of the qualitative condition. Furthermore, environmental elements may represent the conditions of substances other than living organisms, or may represent the conditions of living organisms. The living organisms may be organisms of the same species as the breeding target, or may be organisms of a different species. Furthermore, environmental elements may be represented by physical quantities, basic statistics, or other quantities.

[0019] The environmental elements for which estimated values ​​are obtained may include environmental elements whose values ​​cannot be (substantially) adjusted in the breeding grounds. Furthermore, the environmental elements for which estimated values ​​are obtained may include only environmental elements whose values ​​cannot be (substantially) adjusted in the breeding grounds. However, the environmental elements for which estimated values ​​are obtained may include environmental elements whose values ​​can be adjusted in the breeding grounds. Note that "a value that cannot be (substantially) adjusted in the breeding grounds" means, for example, that the value cannot be controlled in the breeding grounds (the value cannot be set to the target value). Furthermore, if environmental factors that affect the increase or decrease of breeding target organisms are known from publicly known literature, etc., the candidate site environment estimation unit 101 may obtain estimated values ​​of environmental factors including the environmental factors that affect the increase or decrease of breeding target organisms.

[0020] The candidate site environment estimation unit 101 may derive estimated values ​​of environmental elements at candidate breeding sites by, for example, performing processing including numerical simulation. For example, the candidate site environment estimation unit 101 may derive estimated values ​​of environmental elements at candidate breeding sites by performing data assimilation (a method of determining values ​​of unobserved variables (state variables) so that the difference between the measured values ​​of observable variables (observation variables) and the calculated values ​​obtained by numerical simulation is small (minimum)). Alternatively, the candidate site environment estimation unit 101 may derive estimated values ​​of environmental elements at candidate breeding sites by using a machine learning model.

[0021] In order to obtain environmental elements at a higher positional resolution (finer mesh) and over a wider range, the candidate site environment estimation unit 101 may derive estimated values ​​of environmental elements at each position (mesh) of the candidate breeding site using actual values ​​of the environmental elements obtained at the candidate breeding site. Furthermore, the candidate site environment estimation unit 101 may derive estimated values ​​of environmental elements at each position of the candidate breeding site using actual values ​​of elements of a different type from the environmental elements for which estimates are to be obtained. The elements of a different type from the environmental elements for which estimates are to be obtained may include environmental elements of a different type from the environmental elements in question.

[0022] The candidate site environment estimation unit 101 may, for example, acquire the actual values ​​of the environmental elements in the candidate breeding site from the input device 110. The input device 110 may be a storage medium, a user interface, a receiving device, or a device that includes two or three of these.

[0023] The candidate site environment estimation unit 101 may then output information indicating the estimated values ​​of the environmental elements at each location of the candidate breeding sites derived as described above to the output device 120. The output device 120 may be a computer display, a storage medium, a transmission device, or a device that includes two or three of these.

[0024] Specifically, this embodiment illustrates a case where the candidate site environment estimation unit 101 derives estimated values ​​of environmental elements at a candidate site for an artificial seaweed bed. Furthermore, this embodiment illustrates a case where, among elements that represent the ocean conditions at a candidate site for an artificial seaweed bed, at least seawater temperature (e.g., sea surface temperature), salinity, and horizontal current velocity (horizontal current velocity in the ocean) are included in the environmental elements for which estimates are to be obtained. This is because it is known that these environmental elements affect the proliferation of seaweed and seagrass. For example, actual data made publicly available by the Japan Coast Guard or the Japan Meteorological Agency may be used as actual values ​​of seawater temperature, salinity, and horizontal current velocity. The period of the actual data is, for example, about one to three years.

[0025] Specifically, in this embodiment, the candidate site environment estimation unit 101 uses the performance data to estimate the potential site environment. The candidate site environment estimation unit 101 estimates the potential site environment using the performance data ...<URL:https: / / dreams-c1.riam.kyushu-u.ac.jp / vwp / > ) to derive estimated values ​​of water temperature, salinity, and horizontal flow velocity at a candidate site for an artificial seaweed bed in a 1.5 km × 1.5 km mesh. This embodiment also illustrates a case where the candidate site environment estimation unit 101 derives estimated values ​​of water temperature, salinity, and horizontal flow velocity at each of a plurality of candidate sites for an artificial seaweed bed, and outputs them to the output device 120.

[0026] The candidate site environment estimation unit 101 may derive estimated values ​​of environmental elements for one candidate site for a breeding site. In this case, if the candidate site is not selected as a breeding site in the selection step S202 described below, the candidate site environment estimation unit 101 may derive estimated values ​​of environmental elements for another candidate site. The candidate site environment estimation unit 101 may also derive estimated values ​​of environmental elements for multiple candidate sites for breeding sites. In this case, a breeding site is selected from the multiple candidate sites in the selection step S202 described below. If the multiple candidate sites are not selected as a breeding site in the selection step S202 described below, the candidate site environment estimation unit 101 may derive estimated values ​​of environmental elements for another candidate site.

[0027] Furthermore, if estimated values ​​of environmental elements have been obtained by an external organization or the like, the candidate site environment estimation unit 101 may, for example, acquire the estimated values ​​from the input device 110. In this case, the aforementioned derivation of estimated values ​​of environmental elements may not be performed.

[0028] Furthermore, when selecting a breeding site from among candidate breeding sites, for example, if actual measurements of environmental elements are obtained with sufficient position resolution, the estimation device 100 does not necessarily have to include the candidate site environment estimation unit 101. In this case, the candidate site environment estimation step S201 does not necessarily have to be performed.

[0029] Furthermore, the selection of candidate breeding grounds may be based on estimated values ​​of environmental elements obtained in the past. If it has been possible to inhabit (preferably breed) breeding target organisms in a breeding ground selected in the past, a location having values ​​of environmental elements similar to the estimated values ​​of the environmental elements in the breeding ground may be selected as a candidate breeding ground. In this embodiment, if it has been possible to create an artificial seaweed bed in a location selected in the past, a location having values ​​of environmental elements similar to the estimated values ​​of the environmental elements in the location of the artificial seaweed bed may be selected as a candidate artificial seaweed bed.

[0030] <Selection process S202> In the selection step, a breeding site is selected from the breeding site candidate sites based on at least one of estimated values ​​and measured values ​​of environmental elements at the breeding site candidate sites. As described above, the estimated values ​​of environmental elements at the breeding site candidate sites acquired by the candidate site environment estimation unit 101 may be output to the output device 120. In this case, for example, the operator may select a breeding site from the breeding site candidate sites by checking whether the estimated values ​​of the environmental elements at the breeding site candidate sites are within a range known from publicly known literature or the like as a range of values ​​suitable for the growth (including growth; the same applies hereinafter) and breeding of the organism.

[0031] Figure 3 shows an example of the relationship between the range of values ​​of environmental elements (seawater temperature, salinity, and horizontal current velocity) suitable for the growth of kelp, and the estimated values ​​of those environmental elements at candidate sites for artificial seaweed beds. In Figure 3, the range of values ​​of environmental elements (seawater temperature, salinity, and horizontal current velocity) suitable for the growth of kelp are shown in the column under "Suitable Value Range." The estimated values ​​of the environmental elements at candidate sites for artificial seaweed beds are also shown in the columns under "Candidate Site A," "Candidate Site B," and "...." The "..." symbol indicates that estimated values ​​of the environmental elements have also been obtained at candidate sites other than candidate sites A and B.

[0032] In Figure 3, the estimated values ​​of environmental elements at candidate site A (seawater temperature, salinity, and horizontal current velocity) are all within the range of values ​​suitable for the growth and reproduction of Religious Laminaria. On the other hand, the seawater temperatures at candidate site B are all outside the range suitable for the growth and reproduction of Religious Laminaria. In this case, candidate site A can be selected as an artificial seaweed bed for Religious Laminaria. On the other hand, candidate site B cannot be selected as an artificial seaweed bed for Religious Laminaria.

[0033] The number of breeding sites selected in the selection step S202 may be one or more. For example, one candidate site with the most suitable estimated value may be selected from among candidate sites where the estimated values ​​of all types of environmental elements are within the range of values ​​of the environmental elements suitable for the growth and reproduction of the breeding target organisms. Alternatively, multiple candidate sites where the estimated values ​​of all types of environmental elements are within the range of values ​​of the environmental elements suitable for the growth and reproduction of the breeding target organisms may be selected. In this case, all candidate sites where the estimated values ​​of all types of environmental elements are within the range of values ​​of the environmental elements suitable for the growth and reproduction of the breeding target organisms may be selected. Alternatively, two or more candidate sites where the most suitable estimated values ​​are obtained may be selected from among candidate sites where the estimated values ​​of all types of environmental elements are within the range of values ​​of the environmental elements suitable for the growth and reproduction of the breeding target organisms.

[0034] Furthermore, in this embodiment, an example is given in which the operator selects the breeding grounds. However, this is not necessarily the case. For example, the estimation device 100 may include a breeding ground selection unit that selects breeding grounds from among candidate breeding grounds based on at least one of estimated values ​​and measured values ​​of environmental elements at the candidate breeding grounds. In this case, the estimation device 100 (breeding ground selection unit) may select one or more breeding grounds based on the results of processing that includes, for example, determining whether the estimated values ​​of the environmental elements at the candidate breeding grounds are within a range that is preset as a range of values ​​suitable for the growth and reproduction of the organism. Furthermore, the estimation device 100 (breeding ground selection unit) may select one or more breeding grounds using a method similar to the selection performed by the operator described above, in which the closer the estimated values ​​of the environmental elements are to the median of the range of values ​​suitable for the growth and reproduction of the organism, the more suitable the breeding ground is for the growth and reproduction of the organism to be bred.

[0035] <Selected location performance acquisition unit 102, selected location performance acquisition step S203> The selected site performance acquisition unit 102 acquires performance values ​​including the performance values ​​of candidate increase / decrease factor elements and the actual measured values ​​of candidate reproductive status elements as performance values ​​for the breeding ground selected as described above. The reproductive status elements are elements that represent the reproductive status of the breeding target organisms, and for example, represent the degree of reproduction of the breeding target organisms. The reproductive status elements may be expressed as physical quantities, basic statistical quantities, or other quantities. Note that the causal relationship model described below includes increase / decrease factor variables that indicate the values ​​of the increase / decrease factor elements and reproductive status variables that indicate the values ​​of the reproductive status elements. The "candidates" for the candidate increase / decrease factor elements and candidate reproductive status elements are candidate increase / decrease factor variables and reproductive status variables to be included in the causal relationship model.

[0036] The performance values ​​of the reproductive status elements of the propagation target organisms can only be obtained after the organisms have taken up residence. Therefore, these performance values ​​are acquired after the propagation target organisms have taken up residence in the propagation site selected as described above. Furthermore, these actual measured values ​​are preferably acquired after the propagation of the propagation target organisms has been confirmed in the propagation site selected as described above. In this embodiment, an example is given of the case where these performance values ​​are acquired after an artificial seaweed bed has been successfully created in the selected location.

[0037] Candidates for reproductive status elements include, for example, at least one of the number, weight, size (area, volume, etc.) and coverage of organisms to be bred. It is preferable to have multiple types of candidates for reproductive status elements, but it is also possible to have only one type (note that weight may be mass. This also applies to the weight of organisms other than organisms to be bred). In addition, candidates for increase / decrease factor elements are selected in advance from elements known in public literature, etc. as factors that may have a positive or negative effect on candidate reproductive status elements. It is preferable to have multiple types of candidates for increase / decrease factor elements, but it is also possible to have only one type.

[0038] In the present embodiment, when the organisms to be propagated are a specific type of algae, the weight and area of ​​the specific type of algae are examples of the propagating condition element. Also, the volume may be used instead of the area.

[0039] Factors (increase / decrease factors) that may have a positive or negative effect on the weight or area of ​​these specific types of algae include, for example, environmental factors in the breeding ground (the place where the artificial seaweed bed is constructed) selected as described above. These environmental factors preferably include environmental factors whose values ​​can be adjusted in the breeding ground. Furthermore, environmental factors included in the increase / decrease factors may include environmental factors whose values ​​cannot be (substantially) adjusted in the breeding ground.

[0040] Environmental factors whose values ​​can be adjusted in an artificial seaweed bed, which is an example of a breeding ground, include the amount of nutrients, the amount of trace elements, and the amount of animals that feed on specific types of algae. These amounts are expressed, for example, as weight per unit volume or number per unit volume. Examples of nutrients include nitrogen and phosphorus. Examples of trace elements include dissolved iron and silica. Examples of animals that feed on specific types of algae include sea urchins, seafood, etc.

[0041] Environmental factors whose values ​​cannot be (practically) adjusted in artificial seaweed beds, which are an example of breeding grounds, include seawater temperature (e.g., sea surface temperature), salinity, horizontal flow velocity, wave height, electrical conductivity, and dissolved oxygen concentration.

[0042] Figure 4 is a diagram showing an example of candidate increase / decrease factor elements and candidate reproductive status elements when the organism to be propagated is kelp. In Figure 4, the population score for kelp is a numerical representation of the coverage measured as qualitative information (e.g., information on a five-level scale: many, somewhat many, normal, somewhat few, and few). Similarly, the population score for sea urchins is a numerical representation of the number of individuals measured as qualitative information (e.g., information on a five-level scale: many, somewhat many, normal, somewhat few, and few). The actual measurement values ​​acquired by the selected site performance acquisition unit 102 may be actual measurement values ​​measured by a third party.

[0043] The selected site performance acquisition unit 102 acquires the performance values ​​described above (performance values ​​of candidate increase / decrease factor elements and actual measured values ​​of candidate reproductive status elements). The selected site performance acquisition unit 102 may input the performance values ​​from, for example, the input device 110. As described above, the input device 110 may be a storage medium, a user interface, a receiving device, or a device that includes two or three of these.

[0044] <Increase / Decrease Factor Selection Processing Unit 103, Increase / Decrease Factor Selection Processing Step S204> The increase / decrease factor selection processing unit 103 performs processing for selecting at least one adjustable increase / decrease factor element in the breeding ground, including creating a causal relationship model that represents the relationship between multiple variables, including increase / decrease factor variables that indicate the values ​​of the increase / decrease factor elements and breeding state variables that indicate the values ​​of the breeding state elements, based on the performance values ​​(performance values ​​of candidate increase / decrease factor elements and actual measured values ​​of candidate breeding state elements) acquired by the selected site performance acquisition unit 102. The processing for selecting at least one adjustable increase / decrease factor element in the breeding ground includes, for example, at least one of selecting at least one adjustable increase / decrease factor element in the breeding ground and deriving information necessary for selecting at least one adjustable increase / decrease factor element in the breeding ground.

[0045] In this embodiment, a case where a causal relationship model is created by Structural Equation Modeling (SEM) is exemplified. However, the causal relationship model is not limited to a model created by SEM.

[0046] For example, the causal relationship model may be a machine learning model such as a multiple regression equation in which the increase / decrease factor variables are used as explanatory variables and the breeding status variable is used as a target variable. The increase / decrease factor selection processing unit 103 may select the explanatory variables (increase / decrease factor variables) of the machine learning model using, for example, a stepwise method. In this case, the increase / decrease factor selection processing unit 103 can select increase / decrease factor variables that can be adjusted in the breeding field at the time the machine learning model is created.

[0047] Furthermore, the increase / decrease factor selection processing unit 103 may select explanatory variables (increase / decrease factor variables) of the machine learning model based on the coefficient of determination of the machine learning model, etc. For example, the increase / decrease factor selection processing unit 103 may create candidates for multiple regression equations using candidates for explanatory variables (increase / decrease factor variables). Then, the increase / decrease factor selection processing unit 103 may derive a coefficient of determination for the candidate multiple regression equation, and if the coefficient of determination exceeds a reference value, select the explanatory variables (increase / decrease factor variables) included in the candidate as increase / decrease factor variables that can be adjusted in the breeding field. In this case, the coefficient of determination is an example of an increase / decrease factor selection index (an index for selecting at least one increase / decrease factor element that can be adjusted in the breeding field).

[0048] As described above, the causal relationship model may include only the increase / decrease factor elements (increase / decrease factor variables) and the reproductive status elements (reproductive status variables). However, for example, if there are multiple types of at least one of the increase / decrease factor elements and the reproductive status elements, the relationship between the increase / decrease factor elements and the reproductive status elements can be more easily and in more detail expressed by using an element that groups some of them. Therefore, this embodiment illustrates a case in which the causal relationship model includes a parameter indicating the value of an intermediary element, which is an element associated with the increase / decrease factor elements and the reproductive status elements (note that the parameter is not necessarily synonymous with the parameter commonly used in fields such as mathematics). Furthermore, structural equation modeling can take into account latent variables, which are variables that cannot be (directly) observed (or are not observed). Therefore, this embodiment illustrates a case in which the intermediary element is represented as a latent variable in structural equation modeling, and the increase / decrease factor variables and the reproductive status variables are represented as observed variables. The intermediary element may be, for example, an element that directly affects the reproductive status element and is directly affected by the increase / decrease factor element. However, it is preferable to use an intermediary element that affects both the increase / decrease factor elements and the reproductive status elements. In this way, the relationship between the mediating factors can be expressed, and therefore the relationship between the increase / decrease factor factors and the reproductive status factors can be expressed more easily and in more detail.

[0049] Therefore, in this embodiment, an example is given in which the mediator elements include a first mediator element that affects the increase / decrease factor element and a second mediator element that affects the reproductive status element and is affected by the first mediator element.

[0050] In structural equation modeling, at least one of the structural equations and the measurement equations is represented as a path diagram. In this embodiment, an example is shown in which both the structural equations and the measurement equations are included in the causal relationship model. On the other hand, in Non-Patent Document 1, only the structural equations are included in the model. In the path diagram, one-sided arrows (one-directional arrows) represent cause-and-effect relationships (causal relationships). Two-sided arrows (two-directional arrows) represent correlations. Furthermore, rectangular frames represent observed variables. Oval frames represent latent variables. Note that not all of the arrows set in the causal model analysis are necessarily represented in the path diagram; some of the arrows may be omitted in the path diagram. Furthermore, error variables may be omitted in the path diagram (error variables, etc. are omitted in Figures 5A and 5B, which will be described later).

[0051] The path diagram is created by, for example, an operator. In this case, the operator creates a path diagram candidate (performs so-called path setting) using one or more candidate increase / decrease factor elements selected from candidate increase / decrease factor elements, one or more candidate reproductive state elements selected from candidate reproductive state elements, and candidate intermediary elements (first intermediary element and second intermediary element), for example, by operating the input device 110 (user interface). In this case, the increase / decrease factor selection processing unit 103 may output information indicating the contents of the path diagram candidate to the output device 120 based on the operation of the input device 110. Then, the output device 120 may display the information indicating the contents of the path diagram on a computer display. Furthermore, the information indicating the contents of the path diagram candidate may be stored in the input device 110 (storage medium).

[0052] As described above, in this embodiment, the case where the increase / decrease factor selection processing unit 103 acquires information indicating the contents of the path diagram candidates from the input device 110 will be exemplified.

[0053] FIG. 5A is a diagram illustrating an example of a path diagram. 5A illustrates an example in which D-Fe (concentration of dissolved iron), DIN (concentration of dissolved nitrogen), EC (electrical conductivity), sea surface temperature, and sea urchin population score are selected as increase / decrease factor elements 510 from among the candidate increase / decrease factor elements shown in FIG. Here, an example is shown in which increase / decrease factor elements whose values ​​can be adjusted in the artificial seaweed bed (D-Fe concentration 511, DIN concentration 512, sea urchin population score 515) and increase / decrease factor elements whose values ​​cannot (substantially) be adjusted in the artificial seaweed bed (EC 513, sea surface temperature 514) are included in the path diagram as increase / decrease factor elements 510. However, for example, only increase / decrease factor elements whose values ​​can be adjusted in the artificial seaweed bed may be included in the path diagram as increase / decrease factor elements.

[0054] 5A illustrates a case in which a distribution area 521 of Laminaria religiosa and a wet weight 522 of Laminaria religiosa are selected as reproductive condition elements 520 from among the candidates for reproductive condition elements shown in FIG.

[0055] As described above, the first mediator element 530 and the second mediator element 540 are elements that associate the increase / decrease factor element 510 and the reproductive status element 520. The first mediator 530 is a factor that affects the increase / decrease factor element 510. For example, by setting a first mediator 530 for each group of the increase / decrease factor element 510, the relationship between the increase / decrease factor element 510 and the reproductive status element 520 can be easily and in detail expressed. The increase / decrease factor element 510 is grouped, for example, based on the attributes of the increase / decrease factor element 510. The attributes of the increase / decrease factor element 510 include at least one of a quantitative attribute of the increase / decrease factor element 510 and a qualitative attribute of the increase / decrease factor element 510. The quantitative attribute is an attribute expressed by a numerical value, such as number, weight, mass, concentration, density, area, or volume. The qualitative attribute is an attribute that is not expressed by a numerical value, such as type or shape. Note that a numerical value corresponding to the content of the qualitative attribute may be expressed as the value of the qualitative attribute.

[0056] In FIG. 5A , since the D-Fe concentration 511 and the DIN concentration 512 belong to nutrients, a nutrient factor 531 is set as one of the first mediators 530 as a factor (element) that determines the value of the nutrient attribute. Furthermore, since the sea urchin population score 515 relates to sea urchins that feed on narrow-sea kelp, a sea urchin factor 532 is set as one of the first mediators 530 as a factor (element) that determines the value of the sea urchin attribute. Furthermore, since the values ​​of EC 513 and sea surface temperature 514 cannot be adjusted, an other factor 533 is set as one of the first mediators 530 as a factor (element) that determines the values ​​of increase / decrease factors other than sea urchin and nutrients. Since the first mediators 530 (nutrient factor 531, sea urchin factor 532, and other factor 533) are not observed, first mediators that indicate their values ​​are set as latent variables.

[0057] The second intermediary element 540 is an element that influences the reproductive condition element 520 and is influenced by the first intermediary element 530. For example, by setting a second intermediary element 540 for each group of reproductive condition elements 520, the relationship between the increase / decrease factor element 510 and the reproductive condition element 520 can be easily and in detail expressed. The reproductive condition elements 520 are grouped, for example, based on the attributes of the reproductive condition element 520. Like the attributes of the increase / decrease factor element 510, the attributes of the reproductive condition element 520 include at least one of a quantitative attribute of the reproductive condition element 520 and a qualitative attribute of the reproductive condition element 520. Note that a numerical value corresponding to the content of the qualitative attribute may be expressed as the value of the qualitative attribute. In FIG. 5A, the distribution area 521 of Kelp and the wet weight of Kelp belong to Kelp, and therefore a kelp factor 541 is set as one of the second intermediary elements 540 as a factor (element) that determines the value for Kelp.

[0058] FIG. 5A also illustrates an example in which the value of kelp factor 541 is determined by each of nutrient salt factor 531, sea urchin factor 532, and other factor 533 as factors. In FIG. 5A, double-headed arrows are set between all of the observed variables (rectangular frames), but for convenience of notation, the double-headed arrows are omitted from the illustration.

[0059] The path diagram is not limited to the example shown in FIG. 5A . For example, the first intermediary element 530 may be any element that influences the increase / decrease factor element 510. Therefore, for example, another element (at least one of an observed variable and a latent variable) may be set between the first intermediary element 530 and the increase / decrease factor element 510. Furthermore, the second intermediary element 540 may be any element that influences the reproductive state element 520 and is influenced by the first intermediary element 530. Therefore, for example, another element (at least one of an observed variable and a latent variable) may be set between the second intermediary element 540 and the reproductive state element 520. Furthermore, another element (at least one of an observed variable and a latent variable) may be set between the first intermediary element 530 and the second intermediary element 540. Furthermore, arrows (double-sided arrows or single-sided arrows) may be set between the increase / decrease factor elements 510 (the same applies to the reproductive state element 520, the first intermediary element 530, and the second intermediary element 540).

[0060] Furthermore, for example, when multiple types of organisms are propagated in one propagation area, a second intermediary factor may be set for each type of organism. In this case, a single-sided arrow or a double-sided arrow may be set between the second intermediary factors. For example, when both narrow-skinned kelp and Laminaria japonica are propagated in the same artificial seaweed bed, the kelp factor 541 may include two factors, a narrow-skinned kelp factor and a Laminaria japonica factor, as second intermediary factors (latent variables).

[0061] The creation of a path diagram itself can be realized by known techniques. Therefore, a detailed description thereof will be omitted here. Note that if latent variables are not used, the observed variables (in this embodiment, the increase / decrease factor variables and the reproductive status variables) may be converted into latent variables.

[0062] When a candidate path diagram is created, the increase / decrease factor selection processing unit 103 derives the suitability of the candidate path diagram using the actual values ​​of the increase / decrease factor elements and breeding status elements included in the candidate path diagram from the actual values ​​(actual values ​​of candidate increase / decrease factor elements and actual measured values ​​of candidate breeding status elements) acquired by the selected site actual results acquisition unit 102.

[0063] The goodness of fit may be a known goodness of fit used in structural equation modeling. For example, the increase / decrease factor selection processor 103 derives at least one of the Goodness of Fit Index (GFI), Adjusted GFI (AGFI), Comparative Fit Index (CFI), and Root Mean Square Error of Approximation (RMSEA) as the goodness of fit for the candidate path diagram. In this case, if the goodness of fit derived by the increase / decrease factor selection processor 103 satisfies a predetermined criterion, the candidate path diagram is adopted; otherwise, the candidate path diagram is not adopted. For example, if the values ​​(all values) of the GFI, AGFI, CFI, and RMSEA for the candidate path diagram satisfy the predetermined criterion, the candidate path diagram is adopted; otherwise, the candidate path diagram is not adopted.

[0064] The increase / decrease factor selection processing unit 103 may output information indicating the result of the derivation of the degree of compatibility to the output device 120. The increase / decrease factor selection processing unit 103 may also output information indicating whether the degree of compatibility satisfies a predetermined criterion to the output device 120. The output device 120 may display this information on a computer display. In this case, the operator may input information indicating whether to accept or reject the path diagram candidate to the input device 110 based on the information displayed on the computer display. Then, the increase / decrease factor selection processing unit 103 may decide whether to accept or reject the path diagram candidate based on the information indicating the acceptance or rejection of the path diagram candidate.

[0065] Furthermore, if a path diagram candidate is not adopted based on information indicating whether the path diagram candidate is adopted, the increase / decrease factor selection processing unit 103 may output information for recreating the path diagram candidate to the output device 120. The output device 120 may display information for recreating the path diagram candidate on a computer display. Based on the information, an operator performs an operation on the input device 110 (user interface) to re-create the path diagram candidate. When re-creating the path diagram candidate, at least one of the variables included as the increase / decrease factor element 510 (increase / decrease factor variable), the reproductive state element 520 (reproductive state variable), the first mediator element 530 (first parameter), and the second mediator element 540 (second parameter) may be added or deleted. Furthermore, a variable other than these (at least one of a latent variable and an observed variable) may be added. Furthermore, the arrangement of each variable in the path diagram may be changed. In this case, the position of the arrow line may be changed (at least one of the position of the tip and base of the arrow line may be changed). The form of the arrow line may also be changed (for example, the direction of the arrow line may be changed, or a single-sided arrow line may be changed to a double-sided arrow line, or a double-sided arrow line may be changed to a single-sided arrow line). Arrow lines may also be added or deleted. Note that a path diagram may include variables that are not connected to any variables by arrow lines.

[0066] Furthermore, the increase / decrease factor selection processing unit 103 may determine whether to adopt the path diagram candidate based on the degree of suitability, and output information indicating the determination result to the output device 120. Furthermore, when it is determined that the path diagram candidate is to be adopted, the increase / decrease factor selection processing unit 103 may derive the following path coefficients without outputting information to the output device 120.

[0067] After determining the path diagram to be adopted from the candidate path diagrams as described above, the increase / decrease factor selection processing unit 103 derives path coefficients for the path diagram using the actual values ​​of the increase / decrease factor elements and reproductive status elements included in the path diagram from the actual values ​​acquired by the selected site actual value acquisition unit 102 (actual values ​​of candidate increase / decrease factor elements and actual measured values ​​of candidate reproductive status elements).

[0068] Figure 5B shows an example of path coefficients in the path diagram shown in Figure 5A. Note that in structural equation modeling, information such as variance is also derived, but this is not shown in Figure 5B. In Figure 5B, the numbers shown next to the arrows are path coefficients. If the value of the path coefficient is positive, it indicates that the variable connected to the base end of the arrow has a positive influence (a positive influence) on the variable connected to the tip of the arrow, and the larger the absolute value, the greater the degree of influence. On the other hand, if the value of the path coefficient is negative, it indicates that the variable connected to the base end of the arrow has a negative influence (an unfavorable influence) on the variable connected to the tip of the arrow, and the larger the absolute value, the greater the degree of influence.

[0069] In the example shown in Figure 5B, the path coefficient between nutrient factor 531 and kelp factor 541 is a positive value (=0.05), indicating that nutrient concentration has a positive effect on the distribution area 521 and wet weight 522 of kelp. Therefore, D-Fe concentration 511 and DIN concentration 512, whose values ​​are determined using nutrient factor 531 as a factor, are factors that contribute to the increase of kelp in the selected artificial seaweed bed. Therefore, it is expected that the proliferation of kelp will be promoted by increasing the concentrations of D-Fe and DIN through fertilization, etc.

[0070] Furthermore, because the path coefficient between the sea urchin factor 532 and the kelp factor 541 is a negative value (=-0.83), the number of sea urchins (sea urchin population score 515), whose value is determined by the sea urchin factor 532, is a factor that is causing a decline in narrow-skinned kelp in the selected artificial seaweed bed. Therefore, it is expected that the proliferation of narrow-skinned kelp will be promoted by reducing the number of sea urchins through the extermination or removal of sea urchins, etc.

[0071] Additionally, the path coefficient between other factors 533 and kelp factors 541 is a positive value (= 0.42). EC 513 and sea surface temperature 514, which are determined by other factors 533, cannot be (substantially) adjusted in the selected artificial seaweed bed. Therefore, the fact that EC 513 and sea surface temperature 514 have a positive effect on the distribution area 521 and wet weight 522 of narrow-sea kelp is used as supplementary information.

[0072] As described above, the path coefficient is an example of an increase / decrease factor selection index (in this embodiment, an index for selecting at least one increase / decrease factor that can be adjusted in the artificial seaweed bed). As described above, for example, of the path coefficients, the path coefficients that indicate the relationship between the first mediator 530 and the second mediator 540 ("0.05" and "-0.83" in FIG. 5B) may be used as the increase / decrease factor selection index. Furthermore, of the path coefficients, only the path coefficients that indicate the relationship between the first mediator 530 and the second mediator 540 ("0.05" and "-0.83" in FIG. 5B) may be used as the increase / decrease factor selection index.

[0073] The increase / decrease factor selection processing unit 103 may output information indicating the contents of the path coefficients to the output device 120. Then, the output device 120 may display the information indicating the contents of the path coefficients on a computer display. The causal model analysis using structural equation modeling is realized by using, for example, R (a statistical analysis program).

[0074] <Increase / Decrease Factor Selection Step S205> As described above, the increase / decrease factor selection processing unit 103 (increase / decrease factor selection processing step S204) may perform a process of selecting at least one increase / decrease factor element that can be adjusted in the breeding ground, which process itself performs the selection. In this case, the increase / decrease factor selection step S205 is unnecessary. Furthermore, the increase / decrease factor selection processing unit 103 (increase / decrease factor selection processing step S204) may perform a process of acquiring an increase / decrease factor selection index such as a path coefficient, which process is used to select at least one increase / decrease factor element that can be adjusted in the breeding ground. In such a case, the increase / decrease factor selection step S205 may select at least one increase / decrease factor element that can be adjusted in the breeding ground based on the results of the processing in the increase / decrease factor selection processing unit 103 (increase / decrease factor selection processing step S204).

[0075] For example, as described above, when information indicating the contents of the path coefficients is displayed on a computer display by the increase / decrease factor selection processing unit 103, the operator may select at least one increase / decrease factor that can be adjusted in the breeding ground based on that information. In the example shown in Figure 5B as described above, the operator may select, for example, at least one of the D-Fe concentration 511, the DIN concentration 512, and the sea urchin population score 515 as an increase / decrease factor that can be adjusted in the selected artificial seaweed bed.

[0076] Furthermore, for example, the estimation device 100 may include an increase / decrease factor selection unit that selects at least one increase / decrease factor that can be adjusted in the breeding bed based on the results of processing in the increase / decrease factor selection processing unit 103 (increase / decrease factor selection processing step S204). In this case, the estimation device 100 (increase / decrease factor selection unit) may select, for example, an increase / decrease factor 510 associated with the first intermediary element 530, for which a path coefficient whose absolute value exceeds a threshold is set, from among the path coefficients between the first intermediary element 530 and the second intermediary element 540, an element whose value can be adjusted in the artificial seaweed bed.

[0077] <Increase / Decrease Factor Adjustment Unit 104, Increase / Decrease Factor Adjustment Step S206> The increase / decrease factor adjustment unit 104 performs processing to derive adjustment amounts for the increase / decrease factor elements selected based on the results of processing by the increase / decrease factor selection processing unit 103. The processing to derive adjustment amounts for the increase / decrease factor elements includes, for example, at least one of deriving adjustment amounts for the increase / decrease factor elements and deriving information necessary for deriving adjustment amounts for the increase / decrease factor elements. For example, the increase / decrease factor adjustment unit 104 may derive a predetermined constant amount for each type of increase / decrease factor element as the adjustment amount for the increase / decrease factor element. However, in this embodiment, in order to more effectively breed the target organisms, the increase / decrease factor adjustment unit 104 performs processing to derive adjustment amounts for the increase / decrease factor elements using an evaluation model that has, as explanatory variables, increase / decrease factor variables indicating the values ​​of the increase / decrease factor elements selected based on the results of processing by the increase / decrease factor selection processing unit 103, and has, as objective variables, reproductive status variables associated with the increase / decrease factor variables in the employed causal relationship model. In this case, the increase / decrease factor adjustment unit 104 derives the pre-adjustment value and post-adjustment value of the increase / decrease factor element as a process for deriving the adjustment amount of the increase / decrease factor element. Note that the reproductive state variable associated with the increase / decrease factor variable may be a reproductive state variable that is directly associated with the increase / decrease factor variable, or may be a reproductive state variable that is associated with the increase / decrease factor variable via another variable. For example, even if the increase / decrease factor variable and the reproductive state variable are connected by an arrow in a path diagram, or even if the increase / decrease factor variable and the reproductive state variable are connected by an arrow with another variable in between, the increase / decrease factor variable and the reproductive state variable may be considered to be associated with each other.

[0078] The increase / decrease factor adjustment unit 104 may acquire, from the input device 110, information specifying an increase / decrease factor variable indicating the value of the increase / decrease factor element selected based on the processing result by the increase / decrease factor selection processing unit 103 and a reproductive status variable associated with the increase / decrease factor variable in the adopted causal relationship model. In this case, an operator may input the information to the estimation device 100 by operating the input device 110 (user interface). The increase / decrease factor adjustment unit 104 may also acquire the information from the increase / decrease factor selection processing unit 103.

[0079] The evaluation model may be a machine learning model such as a neural network or a multiple regression equation. The machine learning model may be a model created by supervised learning or unsupervised learning.

[0080] For example, when performing supervised learning, the increase / decrease factor adjustment unit 104 may create learning data using the actual values ​​of the increase / decrease factor elements and reproductive status elements included in the path diagram adopted by the increase / decrease factor selection processing unit 103, out of the actual values ​​(actual values ​​of the candidate increase / decrease factor elements and the actual measured values ​​of the candidate reproductive status elements) acquired by the selected site actual value acquisition unit 102. In this case, the actual value of the reproductive status element becomes the correct label. The increase / decrease factor adjustment unit 104 may create the above-mentioned machine learning model using, for example, the learning data. More specifically, for example, the increase / decrease factor adjustment unit 104 may derive the multiple regression equation (partial regression coefficients a1 to a5 and constant b) shown in the following equation (1) by performing multiple regression analysis, which is a multiple regression equation that includes a variable indicating the value of the distribution area 521 of Kelp as the objective variable, and variables indicating the values ​​of the D-Fe concentration 511, the DIN concentration 512, the EC 513, the sea surface temperature 514, and the sea urchin population score 515 as the explanatory variables.

[0081] [konb]=a1×[D-Fe]+a2×[DIN]+a3×[EC]+a4×[SST]+a5×[UNI]+b ···(1) Here, [konb] is (a variable indicating the value of) the distribution area of ​​narrow kelp (units are, for example, hectares). [D-Fe] is (a variable indicating the value of) the concentration of dissolved iron (units are, for example, μg / L). [DIN] is (a variable indicating the value of) the concentration of dissolved nitrogen (units are, for example, μg / L). [EC] is (a variable indicating the value of) the electrical conductivity (units are, for example, mS / m = 10 -3 S / m). [SST] is the sea surface temperature (variable indicating the value of) (units are, for example, °C). [UNI] is the sea urchin population score (variable indicating the value of) (units are, for example, dimensionless (pieces)).

[0082] The signs of the partial regression coefficients a1 to a5 are positive if, for example, the distribution area of ​​Kelp increases as the value of the explanatory variable (increase / decrease factor) multiplied by the partial regression coefficient increases, and negative if the distribution area of ​​Kelp increases as the value of the explanatory variable (increase / decrease factor) multiplied by the partial regression coefficient decreases. The method of creating a machine learning model, including multiple regression analysis, can be realized using known technology, and detailed explanations thereof will be omitted here.

[0083] After the evaluation model is created as described above, for example, an operator sets the values ​​of the explanatory variables (increase / decrease factor elements) of the evaluation model by operating the input device 110 (user interface). The increase / decrease factor adjustment unit 104 derives the value of the objective variable (increase / decrease factor) in the evaluation model that corresponds to the value set by the operator.

[0084] Specifically, for example, the operator may set the actual measured value of the explanatory variable (increase / decrease factor element) as the current value for the explanatory variable (increase / decrease factor element) of the evaluation model. Then, the increase / decrease factor adjustment unit 104 may derive the value of the objective variable (reproduction state variable) corresponding to the value set by the operator in the evaluation model (e.g., equation (1)) as the current estimated value of the reproductive state element. Thereafter, the operator sets a desired value for at least one of the explanatory variables (increase / decrease factor elements) that is different from the current value. Then, the increase / decrease factor adjustment unit 104 may derive the objective variable (reproduction state variable) corresponding to the value set by the operator in the evaluation model as the future estimated value (predicted value) of the reproductive state element. The increase / decrease factor adjustment unit 104 may output information indicating the derivation result of the evaluation model (the current estimated value and future estimated value of the reproductive state element) to the output device 120. The output device 120 may display the information on a computer display. In this way, the operator can know how much the objective variable (reproduction status element) will change when an explanatory variable (increase / decrease factor element) is changed from its current value, based on the derived results of the evaluation model (current and future estimated values ​​of the reproductive status element) and the explanatory variables (current values ​​and changed values ​​of the increase / decrease factor element) set for the evaluation model. Furthermore, the increase / decrease factor adjustment unit 104 may derive an adjustment amount for the explanatory variable (increase / decrease factor element) based on the value of the explanatory variable (increase / decrease factor element) used when deriving the current estimated value of the reproductive status element and the value of the explanatory variable (increase / decrease factor element) used when deriving the future estimated value of the reproductive status element. In this case, the increase / decrease factor adjustment unit 104 may output information indicating the adjustment amount for the explanatory variable (increase / decrease factor element) to the output device 120. The output device 120 may display the information on a computer display.

[0085] Figure 6 shows an example of the current values ​​(current status) of the increase / decrease factors and the increase / decrease amounts of the increase / decrease factors. Here, we assume three cases (Cases A1-A3) in which the dissolved iron concentration ([D-Fe]) is increased relative to the current value, two cases (Cases B1-B2) in which the sea urchin population score ([UNI]) is decreased relative to the current value, and two cases (Cases C1-C2) in which the sea surface temperature ([SST]) is increased relative to the current value. The distribution area of ​​kelp [konb] is calculated using Equation (1) for each case. In Equation (1), a1 = 0.013, a2 = 0.19, a3 = -0.014, a4 ​​= -0.49, a5 = -0.063, and b = 7.0.

[0086] Figure 7 shows an example of the results of deriving the distribution area of ​​narrow-skinned kelp [konb] for each of the cases shown in Figure 6. As shown in Figure 7, for example, if the sea urchin population were halved, it was calculated that the distribution area of ​​narrow-skinned kelp would increase by 22.3% compared to its current value (see Case B1). Furthermore, for example, if the concentration of dissolved iron were doubled, it was calculated that the distribution area of ​​narrow-skinned kelp would increase by 0.7% compared to its current value (see Case A3). Note that while artificial seaweed beds cannot (effectively) adjust sea surface temperatures, it is clear that the distribution area of ​​narrow-skinned kelp will decrease due to future climate change and other factors (see Cases C1 and C2).

[0087] Performing the process for deriving the adjustment amounts of the increase / decrease factor elements as described above is preferable because it makes it possible to quantify which increase / decrease factor elements should be adjusted and how in the breeding grounds. However, it is not necessary to perform the process for deriving the adjustment amounts of the increase / decrease factor elements. For example, the increase / decrease factor elements selected as described above may be changed little by little while confirming changes in the reproductive status elements by actual measurement in the selected breeding grounds. In this case, the estimation device 100 does not need to be equipped with the increase / decrease factor adjustment unit 104.

[0088] <Breeding method> In a breeding ground selected from among the candidate sites as described above, the increase / decrease factor elements selected as described above are used to breed specific organisms to be bred. For example, if the D-Fe concentration 511, the DIN concentration 512, and the sea urchin population score 515 are selected, fertilization may be performed in the breeding ground so that the D-Fe concentration 511 and the DIN concentration 512 increase. Alternatively, the sea urchins may be exterminated. Furthermore, if the adjustment amount of the increase / decrease factor elements is obtained, the value of the increase / decrease factor elements in the breeding ground may be adjusted based on the adjustment amount.

[0089] <Summary> As described above, in this embodiment, the estimation device 100 acquires actual values ​​for a breeding site selected from candidate artificial breeding sites based on at least one of the estimated values ​​and actual measured values ​​of environmental elements at the candidate site, including actual values ​​of candidate increase / decrease factor elements for the breeding target organisms and actual measured values ​​of candidate reproductive status elements for the organisms. Based on the actual values, the estimation device 100 performs a process to select at least one adjustable increase / decrease factor element for the breeding site, including creating a causal relationship model representing the relationship between multiple variables, including an increase / decrease factor variable indicating the value of the increase / decrease factor element and a reproductive status variable indicating the value of the reproductive status element. Therefore, after selecting a breeding site effective for breeding organisms, it is possible to select increase / decrease factor elements for each breeding target organism from among the candidates in the breeding site. Therefore, by carrying out breeding while taking the increase / decrease factor elements into consideration, the breeding target organisms can be effectively bred.

[0090] In this embodiment, the estimation device 100 further includes, in the multiple variables included in the causal relationship model, a parameter variable indicating the value of a parameter that is an element associated with the increase / decrease factor element and the reproductive status element. Therefore, the causal relationship model can more easily and accurately represent the relationship between the increase / decrease factor element and the reproductive status element.

[0091] In this embodiment, the estimation device 100 correlates, in the causal relationship model, a first parameter indicating the value of a first mediator that influences the increase / decrease factor element and a second parameter indicating the value of a second mediator that influences the reproductive status element and is influenced by the first mediator. Therefore, the causal relationship model can more easily and in more detail represent the relationship between the increase / decrease factor element and the reproductive status element.

[0092] Furthermore, in this embodiment, the estimation device 100 creates a causal relationship model using structural equation modeling, and in this causal relationship model, the mediator variables are latent variables, and the increase / decrease factor variables and reproductive status variables are observed variables. In Non-Patent Document 1, only observed variables are included in the model. Therefore, only factors that actually exist in the area where aquatic organisms exist are taken into consideration. In contrast, in this embodiment, the causal relationship model can express the relationship between the increase / decrease factor factors and the reproductive status factors using variables that cannot be observed (or are not observed). Therefore, the causal relationship model can express the relationship between the increase / decrease factor factors and the reproductive status elements more easily and in more detail.

[0093] Furthermore, in this embodiment, when creating a causal relationship model using structural equation modeling, the estimation device 100 acquires a path coefficient as an increase / decrease factor selection index (an index for selecting at least one increase / decrease factor element that can be adjusted in the breeding field). Therefore, a quantitative index can be acquired as the increase / decrease factor selection index. Therefore, an increase / decrease factor element that is effective for breeding can be easily selected.

[0094] Furthermore, in this embodiment, the estimation device 100 acquires estimated values ​​of environmental elements in candidate breeding grounds. Therefore, for example, information on environmental elements that cannot (or is difficult to) obtain through actual measurements can be obtained. Therefore, breeding grounds that are more suitable for breeding can be selected.

[0095] In this embodiment, the estimation device 100 performs processing to derive the adjustment amount for the increase / decrease factor selected as described above. Therefore, information on how to adjust the increase / decrease factor can be obtained. This makes it possible to carry out breeding in the breeding grounds effectively and efficiently.

[0096] Furthermore, in this embodiment, the estimation device 100 derives the adjustment amount of the increase / decrease factor variable using an evaluation model that has, as an explanatory variable, an increase / decrease factor variable indicating the value of the increase / decrease factor factor selected as described above, and has, as a target variable, a reproductive status variable associated with the increase / decrease factor variable in the adopted causal relationship model. This improves the accuracy of information on how to adjust the increase / decrease factor. This enables more effective and efficient breeding in the breeding grounds.

[0097] (Other embodiments) The above-described embodiments of the present disclosure can be realized by a computer executing a program. A computer-readable recording medium having the program recorded thereon and a computer program product such as the program can also be applied as embodiments of the present disclosure. Examples of recording media that can be used include flexible disks, hard disks, optical disks, magneto-optical disks, CD-ROMs, magnetic tapes, non-volatile memory cards, and ROMs. The embodiments of the present disclosure can also be realized by a programmable logic controller (PLC) or dedicated hardware such as an application-specific integrated circuit (ASIC). Furthermore, the above-described embodiments of the present disclosure are merely examples of specific embodiments for carrying out the present disclosure, and the technical scope of the present disclosure should not be interpreted as being limited by these. In other words, the present disclosure can be embodied in various forms without departing from its technical concept or main features.

[0098] The disclosure of the above embodiment can be implemented as follows, for example. [Disclosure 1] An estimation device for estimating increase / decrease factor elements including at least one of an increase factor element and a decrease factor element of a specific organism to be propagated, a selected site performance acquisition unit that acquires performance values ​​including the performance values ​​of the candidate increase / decrease factor elements and the actual measurement values ​​of the candidate reproductive status elements of the organism as performance values ​​at a breeding site selected from candidate sites for the artificial breeding site of the organism based on at least one of estimated values ​​and actual measurement values ​​of environmental elements at the candidate site; an increase / decrease factor selection processing unit that performs processing including creating a causal relationship model that represents the relationship between multiple variables, including an increase / decrease factor variable that indicates the value of the increase / decrease factor element and a breeding status variable that indicates the value of the breeding status element, based on the performance value acquired by the selected site performance acquisition unit, as processing for selecting at least one of the increase / decrease factor elements that can be adjusted in the breeding site; An estimation device comprising: [Disclosure 2] The estimation device according to Disclosure 1, wherein the plurality of variables further includes a parameter variable indicating a value of a mediator parameter that is an element associated with the increase / decrease factor element and the reproductive status element. [Disclosure 3] The mediating factors include a first mediating factor that affects the increase / decrease factor element, and a second mediating factor that affects the reproductive status element and is affected by the first mediating factor, The estimation device according to Disclosure 2, wherein in the causal relationship model, a first parameter indicating a value of the first mediator and a second parameter indicating a value of the second mediator are mutually associated. [Disclosure 4] the causal relationship model includes a model created by structural equation modeling; the parameter is a latent variable, The estimation device according to Disclosure 2 or 3, wherein the increase / decrease factor variable and the reproductive status variable are observed variables. [Disclosure 5] The relationship is expressed using a path coefficient, The estimation device described in Disclosure 4, wherein the increase / decrease factor selection processing unit acquires the path coefficient as an increase / decrease factor selection index, which is an index for selecting at least one of the increase / decrease factor elements that can be adjusted in the breeding ground. [Disclosure 6] The estimation device described in any one of Disclosures 1 to 5, wherein the increase / decrease factor selection processing unit acquires an increase / decrease factor selection index, which is an index for selecting at least one of the increase / decrease factor elements that can be adjusted in the breeding field, based on the causal relationship model. [Disclosure 7] a candidate site environment estimation unit that acquires estimated values ​​of the environmental elements in the candidate site for the breeding ground; The estimation device according to any one of Disclosures 1 to 6, wherein the selection of the breeding grounds is performed based on the estimated value acquired by the candidate site environment estimation unit. [Disclosure 8] The estimation device according to any one of Disclosures 1 to 7, further comprising an increase / decrease factor adjustment unit that performs processing to derive an adjustment amount of the increase / decrease factor element selected based on the result of the processing by the increase / decrease factor selection processing unit. [Disclosure 9] The estimation device described in Disclosure 8, wherein the increase / decrease factor adjustment unit has, as an explanatory variable, an increase / decrease factor variable indicating the value of the increase / decrease factor element selected based on the result of the processing by the increase / decrease factor selection processing unit, and performs processing to derive an adjustment amount of the increase / decrease factor variable using an evaluation model having, as a target variable, the reproductive status variable associated with the increase / decrease factor variable in the causal relationship model. [Disclosure 10] The estimation device according to any one of Disclosures 1 to 9, wherein the increase / decrease factor elements include the increase factor element and the decrease factor element. [Disclosure 11] The estimation device according to any one of Disclosures 1 to 10, wherein the breeding ground is an artificially created seaweed bed. [Disclosure 12] An estimation method for estimating increase / decrease factor elements including at least one of an increase factor element and a decrease factor element of a specific organism to be propagated, a selection step of selecting a breeding site from among the candidate sites based on at least one of estimated values ​​and actual measured values ​​of environmental factors at the candidate sites for artificial breeding sites of the organism; A selection site performance acquisition process for acquiring performance values ​​including the performance values ​​of the candidate increase / decrease factor elements and the actual measured values ​​of the candidate reproductive status elements of the organism as performance values ​​at the breeding site selected by the selection process; an increase / decrease factor selection process for selecting at least one of the increase / decrease factor elements that can be adjusted in the breeding site, the process including creating a causal relationship model that represents the relationship between multiple variables, including an increase / decrease factor variable that indicates the value of the increase / decrease factor element and a breeding status variable that indicates the value of the breeding status element, based on the performance values ​​acquired in the selected site performance acquisition process; The estimation method comprises: [Disclosure 13] A breeding method in which the organisms are bred in the breeding ground using the increase / decrease factor elements selected based on the results of processing by the increase / decrease factor selection processing step of the estimation method described in Disclosure 12. [Disclosure 14] A program for causing a computer to function as each part of the estimation device according to any one of Disclosures 1 to 11. [Explanation of symbols]

[0099] 100 Estimator 101 Candidate Site Environment Estimation Department 102 Selection Site Performance Acquisition Department 103 Increase / decrease factor selection processing unit 104 Increase / Decrease Factor Adjustment Department 110 Input Device 120 Output Device 510 Increase / decrease factors 511 D-Fe concentration 512 DIN concentration 513 EC 514 Sea surface temperature 515 Sea Urchin Population Score 520 Reproductive Status Element 521 Distribution area of ​​narrow-skinned kelp 522 Wet weight of narrow-skinned kelp 530 First mediating element 531 Nutritional Factors 532 Sea Urchin Factor 533 Other factors 540 Second mediating element 541 Kelp Factor

Claims

1. An estimation device for estimating increase / decrease factor elements including at least one of an increase factor element and a decrease factor element of a specific organism to be propagated, a selected site performance acquisition unit that acquires performance values ​​including the performance values ​​of the candidate increase / decrease factor elements and the actual measurement values ​​of the candidate reproductive status elements of the organism as performance values ​​at a breeding site selected from candidate sites for the artificial breeding site of the organism based on at least one of estimated values ​​and actual measurement values ​​of environmental elements at the candidate site; an increase / decrease factor selection processing unit that performs processing including creating a causal relationship model that represents the relationship between multiple variables, including an increase / decrease factor variable that indicates the value of the increase / decrease factor element and a breeding status variable that indicates the value of the breeding status element, based on the performance value acquired by the selected site performance acquisition unit, as processing for selecting at least one of the increase / decrease factor elements that can be adjusted in the breeding site; An estimation device comprising:

2. The estimation device according to claim 1 , wherein the plurality of variables further includes a parameter variable indicating a value of a parameter element that is an element associated with the increase / decrease factor element and the reproductive status element.

3. The mediating factors include a first mediating factor that affects the increase / decrease factor element, and a second mediating factor that affects the reproductive status element and is affected by the first mediating factor, 3. The estimation device according to claim 2, wherein in the causal relationship model, a first parameter indicating a value of the first mediator and a second parameter indicating a value of the second mediator are mutually associated.

4. the causal relationship model includes a model created by structural equation modeling; the parameter is a latent variable, The estimation device according to claim 2 or 3, wherein the increase / decrease factor variable and the reproductive status variable are observed variables.

5. The relationship is expressed using a path coefficient, The estimation device according to claim 4 , wherein the increase / decrease factor selection processing unit acquires the path coefficient as an increase / decrease factor selection index that is an index for selecting at least one of the increase / decrease factor elements that can be adjusted in the breeding ground.

6. The increase / decrease factor selection processing unit acquires an increase / decrease factor selection index, which is an index for selecting at least one of the increase / decrease factor elements that can be adjusted in the breeding field, based on the causal relationship model. The estimation device according to any one of claims 1 to 3.

7. a candidate site environment estimation unit that acquires estimated values ​​of the environmental elements in the candidate site for the breeding ground; The estimation device according to any one of claims 1 to 3, wherein the selection of the breeding ground is performed based on the estimated value acquired by the candidate site environment estimation unit.

8. The estimation device according to any one of claims 1 to 3, further comprising an increase / decrease factor adjustment unit that performs processing to derive an adjustment amount of the increase / decrease factor element selected based on a result of the processing by the increase / decrease factor selection processing unit.

9. The increase / decrease factor adjustment unit has an increase / decrease factor variable indicating the value of the increase / decrease factor element selected based on the result of the processing by the increase / decrease factor selection processing unit as an explanatory variable, and the reproductive status variable associated with the increase / decrease factor variable in the causal relationship model is used as a target variable. The estimation device according to claim 8, wherein processing is performed to derive the adjustment amount of the increase / decrease factor variable.

10. The estimation device according to claim 1 , wherein the increase / decrease factor elements include the increase factor element and the decrease factor element.

11. The estimation device according to any one of claims 1 to 3, wherein the breeding ground is an artificially created seaweed bed.

12. An estimation method for estimating increase / decrease factor elements including at least one of an increase factor element and a decrease factor element of a specific organism to be propagated, a selection step of selecting a breeding site from among the candidate sites based on at least one of estimated values ​​and actual measured values ​​of environmental factors at the candidate sites for artificial breeding sites of the organism; A selection site performance acquisition process for acquiring performance values ​​including the performance values ​​of the candidate increase / decrease factor elements and the actual measured values ​​of the candidate reproductive status elements of the organism as performance values ​​at the breeding site selected by the selection process; an increase / decrease factor selection process for selecting at least one of the increase / decrease factor elements that can be adjusted in the breeding site, the process including creating a causal relationship model that represents the relationship between multiple variables, including an increase / decrease factor variable that indicates the value of the increase / decrease factor element and a breeding status variable that indicates the value of the breeding status element, based on the performance values ​​acquired in the selected site performance acquisition process; An estimation method comprising:

13. A breeding method, comprising breeding the organisms in the breeding ground using the increase / decrease factor elements selected based on the results of processing by the increase / decrease factor selection processing step of the estimation method according to claim 12.

14. A program for causing a computer to function as each part of the estimation device according to any one of claims 1 to 3.

Images