EP0721626A4

EP0721626A4 - Apparatus and method for generating growth alternatives for living tissue

Info

Publication number: EP0721626A4
Application number: EP94931760A
Authority: EP
Inventors: Jinkui Liu; Fazal Wala; Harry E Meek
Original assignee: AMERICAN INSTITUTE OF NUTRITION AND MANAGEMENT DOING BUSINESS AS WALAGROUP; AMERICAN INST OF NUTRITION AND
Current assignee: AMERICAN INSTITUTE OF NUTRITION AND MANAGEMENT DOING BUSINESS AS WALAGROUP; AMERICAN INST OF NUTRITION AND
Priority date: 1993-09-22
Filing date: 1994-09-22
Publication date: 1998-04-15
Also published as: AU688808B2; EP0721626A1; WO1995008812A1; AU8071694A; BR9407646A

Abstract

The present invention is an apparatus and method for generating relationships between factors that effect the growth of an animal. These relationships are then used to determine optimal operation parameters in oder to achieve a maximum return on investment. A computer (figure 4) generates the relationships and then determines the optimal parameters such as the amount and type of feed, growing cycle, population density, etc. The computer (figure 4) consists of a microprocessor (20), random access memory (30), a storage device (32), a keyboard (24), a computer screen (26), and a math co-processor (22).

Description

APPARATUS AND METHOD FOR GENERATING GROWTH ALTERNATIVES FOR LIVING TISSUE

TECHNICAL FIELD The present invention relates generally to an apparatus and method for generating optimized living entity growth alternatives, and more specifically to a method and apparatus for generating an inter-variable relationship between growth factors of a creature in order to optimize output production.

BACKGROUND

The economic optimization and viability of an enterprise depends on the ability to accurately analyze the relationship between the cost of materials, services, and labor that are input into the enterprise and the return that is achieved on the product that is output by the enterprize. In agribusiness industries that raise creatures such as cattle, poultry, fish, etc., the inputs include the creature itself, food, shelter, and services. The output, of course, is the marketed creature. One of the most critical relationships in optimizing the economic gain of such a farming enterprise is the relationship between the controllable and uncontrollable factors that affect the rate at which a creature grows and the final size of the creature at marketing age. Thus, it is important to have a model that describes the relationship between each of these factors and the rate of growth of a population of creatures.

Variables affecting the growth of creatures can be divided into genetic and non-genetic categories. Genetic variables are fixed and are reflected by the growth potential of the individual type of creature of interest. It will be appreciated by those skilled in the art that the growth rate of a creature is never higher and only lower than the maximum potential. During its life, a creature seeks to achieve its genetic potential, but fails due to the impediment of non- genetic variables.

Non-genetic variables that are partially controllable by the commercial operator can be divided urther into living factors and food factors. Living factors encompass environmental conditions such as temperature, humidity, creature density, ventilation, disease conditions, air quality, etc. Food factors encompass the types and amounts of material that are ingested by a creature. One skilled in the art will appreciate that food factors can be controlled in a commercial environment through nutrition. The food factor reflects a major portion of the cost during the growth period.

To maximize an enterprise's net margin, many scientists have used models to simulate the growth of various types of creatures. (see G. C. Emmans, "The Growth of Turkeys," 21 Recent Advances in Turkey Science. 135-166 (C. Nixey and T.C. Grey eds. 1989); H. Talpaz et al., "Dynamic Optimization Model for Feeding of Broilers," Aσaric. Savs, 121-132 (1986); H. Talpaz et al., "Economic Optimization of a Growth Trajectory for Broilers," 70 Amer. J. Aσ. Econ.. 382-390 (1988); P. E. aibel et al. , TURKS Program Agricultural Extension Service (University of Minnesota 1985)). It will be appreciated that the various models represent efforts to take into account the incredibly complex and diverse structure of living entities, as well as the innumerable variables that affect the living entities in their environment. The most popular formula to describe poultry growth is the Gompertz curve (B. Gompertz, "On the Nature of the Function Expressive of the Law of Human Mortality, and on a New Mode of Determining the Value of Life Contingencies," Philos. Trans. Roy. Soc, 513-585 (1825)), which shows the current mass weight as a function of age with known constant parameters. Many Gompertz curves have been developed to describe the growth of poultry in terms of different genetic, living, and food factors (G. C. Emmans, "The Growth of Turkeys," 21 Recent Advances in Turkey Science. 135-166 (C. Nixey and T.C. Grey eds. 1989); R. M. Gous et al., "A Characterization of the Potential Growth Rate of Six Breeds of Commercial Broiler, " 2 Proceedings of XIX World's Poultry Congress, 20-24 (Amsterdam, The Netherlands, Sept. 1992); N. B. Anthony et al., "Comparison of Growth Curves of Weight Selected Populations of Turkeys, Quail and Chickens," 70 Poultry Sci. , 13-19 (1991)). However, because all the parameters are independent from one to another among all the curves, each Gompertz curve can describe growth in terms of only one set of conditions. Because of the complexity of a life form, there is a need for a model that describes growth alternatives in terms of a plurality of different conditions. Such a model would permit an accurate economic analysis that allows a commercial operator to simultaneously optimize the relationship between the conditions and growth. In turn, the production of living creatures would be more easily controlled in order to optimize production and hence maximize economic return.

Additionally, those skilled in the art will appreciate that the efficiency of nutrient utilization (i.e., the proportion of digestible nutrient that can be biologically utilized by an animal or poultry) is one of the most important factors that influences the accuracy of growth models. However, due to variations in results from experiment to experiment, as well as the many approaches in interpreting the results, a single efficiency value of a dietary amino acid often ranges broadly. For example, the efficiency value of dietary amino acids for body tissue protein deposition of poultry ranged from 75 to 85%. These singular, rigid estimates been used by different researchers for model construction. See e.g. C. Fisher, "The Physiological Basis of the Amino Acid Requirements of Poultry", In: Protein Metabolism and Nutrition (editors M. Arnal, R. Pion and D. Bouin) , Proc IV Int. Sy . , Clermont-Ferrand. Vol. 1, pp. 385-404, Les colloques de l'INRA, No. 16 (1983); H. Talpaz et al., "Dynamic Optimization Model for Feeding Boilers," 20 Agricultural Systems 121-132 (1986) .

One of the major reasons for the large differences in efficiency is that animals are fed on a population basis. Thus, each individual has its own growth potential and its own nutrient requirement to meet that potential. When these diverse individuals are put together in a flock or herd, the resultant average efficiency of nutrient utilization depends on the variation of individual nutrient requirements and dietary nutrient level of a tested population. The results of these tests indicated that higher dietary nutrient levels resulted in lower efficiency. The reason for these results was that the nutrient requirement for a larger proportion of animals/birds in the population was met, and only the higher requirement animal/birds use the extra nutrient for production.

The Reading Model is the only one to describe the nutrient requirement for a population. See. C. Fisher et al., "A Model for the Description and Prediction of the Response of Laying Hens to Amino Acid Intake," 14 British Poultry Science 469-484 (1973). It was originally used in the description of egg production for laying hens. The essential feature of this approach is to look at nutrient response of each bird independently and then to derive the population response as an integration of each individual bird response. The "optimum" flock requirement of each nutrient can be calculated through this approach by knowing the unit cost of this individual nutrient and value of unit output product.

Although the Reading Model approach is useful in some applications, it has several drawbacks as follows:

1. The approach calculates optimum requirement of each nutrient independently. Therefore, nutrition balance and interactions among nutrients are ignored.

2. The cost of each calculated nutrient is required to use this model. This can be a severe obstacle due to the fact that nutrient cost is mostly associated with each ingredient (i.e. , each ingredient contains many nutrients) . The final nutrient cost depends on the final ingredient composition of the diet due to nutrient competition among the available ingredients to meet minimum nutrient constraints during an optimization process. 3. The calculation of optimum nutrient requirement of a population is based on the economic break point of nutrient cost and value of product in the Reading Model. This may not be true due to the commercial and financial integration of multiple "divisions" within a modern enterprise. Optimum nutrient level can be higher or lower than the one at the economic break point due to higher or lower overhead costs such as costs of processing, labor, production, etc. Therefore, there is an additional need for a method and apparatus that determines the optimum utilization effectiveness of nutrients for a population. The method and apparatus should also preferably be capable of being used in combination with an apparatus and method for generating animal and/or poultry growth alternatives. The present invention directly addresses and overcomes the shortcomings of the prior art.

SUMMARY One advantage of the present invention is that the apparatus and method are based on the correlation between constant parameters among a multitude of Gompertz curves that describe each of the growth curves for poultry. Thus, a commercial poultry operator can use the apparatus and method to simultaneously optimize growth and yield in a plurality of living and food conditions in order to maximize economic return. The present invention generally relates to a computer and method of operating a computer that calculates the earliest time that a bird can reach its maximum rate of growth. Using this information, the computer can determine an appropriate size for a flock of birds, the type and amount of feed that should be fed to the flock, and the age at which a flock should be sold to a food processor, in order to maximize the profits realized by the farmer who raises the flock. More specifically, the present invention is an apparatus for optimizing the ratio between expenditures and rate of growth for living creatures. This apparatus includes processing means for optimizing the ratio between expenditures and the rate of growth for creatures, wherein the processing means calculates the minimum age at which the creatures can experience their maximum rate of growth according to the equation W = Ae^Λ(-e^Λ(-k(t-t*) ) ) , where W is the current body weight of the creature, A is the weight of the creature at physical maturity, k is a constant, t is the current age of the creature, and t* is the age at which the creature has its maximum rate of growth. The apparatus also includes memory means for storing data corresponding to information about feed, information about the characteristics of the creatures, and information generated by the processing means. The memory means is operationally coupled to the processing means.

The present invention is also in the form of a method for operating the apparatus. The method steps include calculating the minimum age at which the creature can experience its maximum rate of growth according to the equation W = Ae^Λ(-e(-k(t-t*) ) ) where W is the current body weight of the creature, A is the weight of the creature at physical maturity, k is a constant, t is the current age of the creature, and t* is the age at which the creature has its maximum rate of growth. The method includes the additional step of storing data corresponding to information about feed, information about the characteristics of the creatures, and information generated by the processing means.

The present invention further provides a method and apparatus which illustrates that the growth response of nutrients is capable of being described in a population by changing the utilization efficiency of nutrients. This approach overcomes the limitations discussed above and can be used in an automated-machine optimization process or a multi-purpose computer. However, it will be readily apparent to those skilled in the art upon a reading of the present specification that the invention is also applicable to other environments. Therefore, while the computer example will be discussed herein, the present invention is not so limited, and various aspects may be applied to other methods and applications.

The nutrition modeling of land and marine animals, and poultry, is a critical component of the respective enterprise optimization. The optimization accuracy depends on the description of utilization efficiency of nutrients in a population of animals. The present invention provides for a method and apparatus to determine utilization efficiency of nutrients for meat production in a population. The present invention uses a model which describes utilization efficiency on at least three component parts of nutrient utilization— more specifically, maintenance, linear gain, and non¬ linear gain. The efficiencies for maintenance and linear gain are not different among individuals in a population, but it has been found that nutrient efficiency for non-linear gain is a product of population variation and efficiency of linear gain. In a preferred embodiment constructed according to the principles of the present invention, a computer processor acts on program logic to receive input from a user or a stored file on the population and nutrients, uses a plurality of stored simultaneous equations, and solves them simultaneously using an optimizer program. The results are then stored and/or displayed. Additionally, the results are also used when generating poultry growth alternatives, which is described herein. Therefore, according to one aspect of the present invention, there is provided: A method of determining utilization effectiveness of nutrients in a population, comprising the steps of: a) determining the standard deviation for the average period gain of the population and the average period potential gain of the population; and b) comparing the actual gain of the population with the average period potential gain to determine the number of standard deviations of non-linear gain, wherein the nutrients to support the linear and/or non- linear gain may be determined.

According to another aspect there is provided a computerized system for determining the utilization effectiveness of nutrients in a population based on the theoretical average period gain of the population, comprising: a) memory means for storing data on the standard deviation, of period gain of the population and the period gain potential of the population; b) first determining means for determining the number of standard deviations for the average non-linear gain of the population from information stored in said memory means; and c) second determining means for determining the nutrition efficiency ratio (EFR) in accordance with the following equation: EFR = 1.6551e^Λ(-0.31768e^Λ(0.35 X STD)) where EFR is the nutrient efficiency ratio, e=2.71828, and STD is the number of standard deviations (-3 to 3) . While the invention will be described with respect to a preferred embodiment computer based method and apparatus, and with respect to particular computer program operational steps and components used therein, it will be understood that the invention is not to be construed as limited in any manner by either such configuration or components described herein. Further, while the preferred embodiment of the invention will be described in relation to poultry, it will be understood that the scope of the invention is not to be limited in any way by the particular animal or poultry environment in which it is employed. The principles of this invention apply to animal growth, and more specifically for modeling the utilization of effectiveness of nutrients in a population using a computer program. These and other advantages and features, which characterize the present invention, are pointed out with particularity in the claims annexed hereto and forming a further part hereto. However, for a better understanding of the invention, its advantages, and objects obtained by its use, reference should be made to the drawings, which form a further part hereto, and to the accompanying descriptive matter, which illustrates and describes a preferred embodiment of the present invention.

DESCRIPTION OF THE DRAWINGS

Figure 1 is a graph showing the characteristic of a Gompertz curve. Figure 2 is a chart showing the values of and . relationship between the rate factor, k, and the inflection point, t*, for a variety of strains of birds.

Figure 3 is a graph showing the relationship between the rate factor, k, and the inflection point, t*, using the data that is included in the chart of Figure 2.

Figure 4 is a functional block diagram of a multipurpose computer useful for practicing the method of the present invention.

Figures 5-19 are menus, screen displays, and a sample report of a preferred embodiment computer program which implements the present invention.

Figure 20 is a functional block diagram of program logic used to implement the principles of the invention.

Figure 21 is an information flow diagram for the program logic of Figure 20. Figures 22a-22ai is a flow chart showing the detailed operation of the program logic shown in Figures 20 and 21.

Figure 23 is a bell curve illustrating population distribution. Figures 24 and 25 are standard deviation tables.

Figure 26 is the source code that controls the computer system shown in Figure 4.

DETAILED DESCRIPTION A preferred embodiment of the invention will be described in detail with reference to the Drawings, wherein like reference numerals represent like parts and assemblies throughout the several views. Reference to the preferred embodiment does not limit the scope of the invention which is limited only by the scope of the claims attached hereto.

The present invention relates to an apparatus and method for correlating the equations that describe the multitude of Gompertz curves for various variables that describe the growth of living creatures. The results of the correlation allow a farmer to simultaneously optimize the ratio between expenditures and growth and thus optimize profit margins. One skilled in the art will realize that the present invention may be used for any type of creature whose growth can be described by a Gompertz curve. However, for the sake of description, the present invention is described in the context of poultry.

A. Theory

As shown in Figure 1, a Gompertz curve represents mass as a function of time, and is commonly used to represent the growth of poultry. The Gompertz curve that describes a growth pattern is as follows: where W is the current body weight of the a bird, W₀ is the initial body weight of the bird, L is a constant, k is a constant, t is the current age of the bird, and ^Λ represents an exponent. Equation (1) can be rearranged as follows: W=f(t)=W₀e^Λ((L/k)e^Λ((-L/k)e~(-kt))). (2)

The limit of equation (2) as t→∞ is defined as: lim f(t) = A = W₀e^Λ(L/k) (3) where A is the bird's mature body weight. Combining equation (2) and equation (3) results in the following equation: which can be written as follows:

W = Au (5) where u = e^Λ( (-L/k)e^Λ(-kt) ) . (6₎ Equation (4) can be rewritten as:

W = Ae^Λ(-Be^Λ(-kt) ) (7) where B = L/k. From equation (7), the average daily gain is: f' (t) = WkBe^Λ(-kt) . (8) The rate at which the average daily gain changes is defined as: f*^*(t) = k^Λ(2)BWe^Λ((-kt) (Be^Λ(-kt)-l)) (9)

If f ' (t) = 0 at the age of maximum gain, then: 0 = k^Λ(2)BWe^Λ(-kt*) (Be^Λ(-kt*)-1) ) Be^Λ(-kt*) = 1

B = e^Λ(kt*) (10) where t* is defined as the inflection point, which represents the age at which the maximum daily weight gain is achieved. The inflection point and the constant k govern the form of growth curve. If equation (10) is substituted into equation (7) , then

W = Au (11)

Where u = e^Λ(-e^Λ(-k(t-t*) ) ) . Equation (11) shows that current body weight depends on mature weight A and u.

Mature weight A is a genetically inherited value. Given fixed genetic conditions, therefore, the growth trajectory depends on u, i.e., the growth rate factor k and inflection point t*. The growth trajectory represents body weight over age.

Rate factor k and inflection point t* are independent of each other among multiple growth curves even though they are constrained by equation (10) within one curve. Their inter-relationship among different curves has to be established in order to make equation (11) cover multiple curves so that it can be used in an automated computer optimization process, i.e., make constant k a function of the inflection point, (k=f(t*)). The equations that are utilized in the program of the present invention is:

W = Ae^Λ(-e^Λ(f(t*)(t-t*)) . (12)

When mature weight A and age t is known, only one variable t* is left to predict body weight W in equation (12). The difference between equation (11) and equation (12) is that equation (11) represent only one growth curve and t* is a constant, but equation (12) represent multiple curves and t* is a variable, which can be optimized in an optimization process. Therefore, the relationship between k and t* must be defined. This relationship will be in the form of a function k = f(t*).

Experimental growth data for broilers, quails, and turkeys with different genetic and environmental conditions have been obtained from public domain sources and summarized. This information is contained within the program of the present invention and can be used to define the relationship between k and t*. The body weight for male turkeys of age 0 to 18 weeks (Waibel, P.E. et al., "Factorial Study of Protein Level Sequence and Diet Energy/Pelleting on Performance of Large White Tom Turkeys," 67 Poult. Sci.. 170 (1988)) and female turkeys of age 0 to 15 weeks (Waibel, P.E. et al., "Factorial Study of Protein Level Sequence and Diet Energy/Pelleting on Performance of Large White Hen Turkeys," 68 Poult. Sci, 133 (1989)) are each comprised of 24 different protein sequence treatments. The body weights of each treatment at different ages was independently fitted into equation (11) and the corresponding value for k and t* was calculated. These values are shown in Figure 2. More specifically, constant k and t* were experimentally determined by (See Hurwitz, S. et al., "Estimation of the Energy Needs of Young Broiler Chicks," Proceedings of the Meeting,

Arkansas Nutrition Conference 16-21 (Riverfront Hilton, North Little Rock, Arkansas, Sept. 10-12, 1991); Talpaz, H. et al., "Dynamic Optimization Model for Feeding of Broilers," Agaric. Savs , 121-132 (1986); Talpaz, H. et al., "Modeling of Dynamics of Accelerated Growth

Following Feed Restriction in Chicks," 36 Agric . Sys . , 125-135 (1991); Gous, R.M. et al. , "A Characterization of the Potential Growth Rate of Six Breeds of Commercial Broiler," 2 Proceedings of XIX World's Poultry Congress, 20-24 (Amsterdam, The Netherlands, Sept. 1992); Emmans, G.C., "The Growth of Turkeys," 21 Recent Advances in Turkey Science, 135-166 (C. Nixey and T.C. Grey eds. 1989); Anthony, N.B. et al. , "Comparison of Growth Curves of Weight Selected Populations of Turkeys, Quail and Chickens," 70 Poultry Sci.. 13-19 (1991)) and fitted into equation (11) by mathematical methods that are commonly known in the art. Figure 2 also includes the values of B and L, which were calculated using equation (10). Figure 3 is a graph in which k is plotted against t*. The graph of figure 3 demonstrates the relationship of k=f(t*) and that the relationship between k and t* is non-linear. Examining the graph of Figure 3, one skilled in the art will realize that statistical methods demonstrate that k=0.79878t*⁽"°^*83747), where adjusted correlation coefficient r=0.9746. One skilled in the art will appreciate that the non-linear relationship between k and t* also exists in other curves such as the Richards equation (Richards, F.J., "A Flexible Growth Function for Empirical Use, " 10 J. of Experimental Botany. 290-300 (1959)). Equation (12) can be rewritten as W=Ae^Λ(-e^Λ(-( .79878t*(-.83747) (t-t*) ) ) ) . (13) This equation covers a multitude of growth curve possibilities and can be used for different types of poultry including turkey, broiler, duck, quail, etc. Given equation (13), constant t* is the only variable that needs to be optimized.

Equation (12) reveals that the rate at which a bird grows depends on only one variable—t*. As discussed above, t* is the age at which a bird has its maximum rate of gain. The earlier the age, the quicker the bird will grow to the weight at which it may be marketed. The commercial applications of equation (12) will be a very important tool in selecting the most efficiently growing type of bird. One skilled in the art will appreciate that the present invention may also have applications related to the production of other types of creatures as well as vegetation.

Equation (12) can be utilized in optimizing poultry production because it correlates multiple growth curves, which include a genetic potential growth curve of the type shown in Figure 1. A curve of this type is required in order to implement a computer optimization process. As discussed above, the genetic potential growth curve of Figure 1 defines the minimum age at which a bird's maximum growth rate is reached. Given the curve of Figure 3, a computer can calculate optimum weight gain and average body weight for each feeding period of a flock of birds. The weight gain and average body weight is then used to determine the optimal living and food environments.

The following example shows how the potential weight gain can be modified by changing the density of turkeys within a certain living space:

Change of weight gain = 0.71556 + 7.9902 MDNSITY - 57.765 MDNSITY² where r (correlation coefficient) = 0.8846; overall p-value (possibility value) = 0.0006; and MDNSITY-body weight density ranged 0.03 to 0.06 meter²/kg^0*67. Similar predictions can be derived by establishing the effect of temperature, humidity, ventilation, etc. on weight gain.

In addition to predicting physical mass of the entire bird, the inflection point t* can be used to predict the growth of each component part of a bird's body. The following is an example for turkeys:

Breast (% of Eviscerated carcass) = 67.121-2.2824 Sex + 0.37094 Age - 0.00093294 Age² - 93116 In(Age) - 0.14238 t* where r = 0.843; and p - value of coefficient t* = 0.0000. Thigh (% of Eviscerated carcass) = 14.6+ 0.056919 Age - 0.00022113 Age² - 0.026625 t* where r = 0.875; and p - value of coefficient t* = 0.0000.

Wing (% of Eviscerated carcass) = 26.399-2.3552 Sex + 0.10141 Age - 0.0018162 Age² + 0.0000064398 Age³ -

0.10284 t* where r = 0.90; p - value of coefficient t* = 0.0000.

Neck (% of Eviscerated carcass) = 18.056 - 2.1653 Sex - 0.0095747 Age - 0.085037 t* where r = 0.6367; and p - value of coefficient t* = 0.0000 where Sex - 1 for male, 2 for female; age = age in days; t* = inflection point (days); r = correlation coefficient; and p-value = possibility value.

All the above regression equations show that the inflection point t* has a significant effect on dependent variables as indicated by the small number of p-values.

Optimizing the ratio between expenditures and growth of a population is also impacted by the population's effectiveness of utilizing nutrients. The reason is that a discrepancy between individual and population response to nutrients on growth results when the overall nutrient requirement of a small proportion of individuals in the population are met. In this setting, even with increased levels of nutrition, more and more individuals in the population will not respond via higher growth. Therefore, the per unit nutrient input results in a smaller overall production gain — which reflects the icroeconomic concept of diminishing return.

It is logical to assume that at a starting point of diminishing return, the nutrient requirement for maintenance of all individuals and partial gain (or linear gain) are met for growing poultry. The population variation has an effect only on the remainder gain (i.e., the non-linear gain). Thus, if the nutrient efficiency is known at this point, then the point will also be known at which efficiency will decrease with each increase in the nutrient level.

Many experiments have been conducted to study the poultry nutrient requirements for maintenance and body weight gain. The most popular model for poultry is the nutrient maintenance and gain model, i.e. maintenance is proportional to the 0.67 power of body weight (BW^0*67) and the requirements for gain are linearly related to the gain itself. Therefore, the total dietary nutrient requirement for maintenance and gain can be calculated using total body weight, weight gains, and their respective efficiencies.

Several questions follow from this model. First, what is the average linear gain. Second, what is the average non-linear gain in a population. Finally, what is the efficiency for the foregoing.

It is commonly accepted that the average daily gain in a population follows the Normal distribution (See Fig. 23). Two standard deviations on each side of the mean should cover about 95% of individual daily gains in a population and three standard deviations should cover 99% of them. Thus, if three standard deviations are used, then over 99% of the individual total non-linear gains will be in the range of six (i.e., 3 x 2) standard deviations (e.g., from point "a" to "d") around the mean. It can then be theorized that the point of diminishing return start at point "a". When all the birds are fed to meet their potential, then the non¬ linear gain distribution in the population should be close to point "b", i.e. three standard deviations. However, due to changes in economic conditions, many times the nutrient levels are not supplied to meet the full growth potential (under feeding) . If the nutrition level is supplied to meet the bird proportion to point "c" instead of "d", then the number of standard deviations for non-linear gain, which is less than three, can be obtained from a Normal distribution table based on the fraction area of higher than point "c".

Still referring to Fig. 23, the nutrient efficiency of non-linear gain is affected by the fraction of the area higher than point "c". With decreases of this area, nutrient utilization will become less efficient. Therefore, calculation of an efficiency ratio can be done as follows.

Assuming that there are one hundred birds in a population, when the nutrient level is low (point "a" of Fig. 23, e.g., a -3 standard deviation) 99.87% of birds in the population would gain body weight with each increase in nutrient level. At that instant, the efficiency ratio would be 99.87%. However, with increased nutrition to -2 standard deviations, then 97.72% of birds in the population would gain (i.e. 2.28% of the birds have reached their genetic potential based on Normal Distribution) . Similarly the efficiency ratios at standard deviations 0, 2 and 3 would be 50, 2.28 and 0.13%, respectively. The efficiency is not instant for calculating the nutrient requirement of non-linear gain. Instead, the average efficiency for the total non-liner gain is calculated. This can be done by averaging all the instant efficiency ratios and decreasing the distance between each instant point. For the above example, the approximate efficiency ratio for three standard deviations of non-linear gain would be (99.87 + 97.72 + 50 + 2.28 + 0.13)/5=50%. Supposing the efficiency for linear gain is 90%, the one for non-linear gain would be 90% X 50% = 45%.

A more accurate description of efficiency ratios for non-linear gains are listed in Figs. 24 and 25 with a distance of 0.01 standard deviation between two instant points. This table ranges from -3 to 3 standard deviations and demonstrates the decreasing efficiency with increasing the standard deviation. This data is fitted into a regression equation as follows: EFR = 1.6551e^Λ(-0.31768e^Λ(0.35 X STD)) (1) r = 0.9940 Where: EFR - Nutrient efficiency ratio e - 2.71828 STD - number of standard deviations (-3 to 3) r - Correlation coefficient The efficiency of non-linear gain is calculated as: EFNL = EFL X EFR

(2) Where: EFNL - nutrient efficiency of non-linear gain.

EFL - nutrient efficiency of linear gain.

Nutrient efficiency ratios (Fig.s 2 and 3) can be used not only for calculating nutrient efficiency of non-linear gain but also for determining the average amount of non-linear gain in the population. For example, if the nutrition level is supplied from point "a" to "d", then the individual distribution for non¬ linear gain increased six standard deviations (e.g., -3 to +3), but the average non-linear gain of population only increased three standard deviations (e.g., -3 to 0). This is because at standard deviation +3, only 50% (0.5001, Fig. 3) of gains in the population for six standard deviations is achieved (50% X 6 = 3). Therefore, the average non-linear gain for the population is three standard deviations when a nutrition level is supplied close to their maximum growth potential.

For conditions of underfeeding, if birds are fed to standard deviation two (i.e. five standard deviations: -3 to 2) for individual distribution, then the average non-linear gain of the population would be 2.991 (5 x 0.5982) standard deviations (Fig. 3). Similarly, for standard deviations 0, -2, and -3 their population average would be 2.6004 (-3 to 0 is 3 standard deviations, 3 X 0.8668 = 2.6004), 0.9931 (-3 to -2 is 1 standard deviation, 1 X 0.9931 = 0.9931), and 0 (-3 to - 3 is 0 standard deviation, 0 X 0.9987 = 0), respectively.

With known weight gain potential and three standard deviations as nonlinear gain the standard deviation difference can be calculated by comparing current average gain and potential gain. Subtracting the difference from 3 would be the number of standard deviations for nonlinear gain.

The result of the number of standard deviations of nonlinear gain and the standard deviation of gain is the total average non-linear gain of the population. The average linear gain is the difference of total average gain and average non-linear gain.

The current invention demonstrates that nutrient utilization efficiency is not a constant number due to the population variation. Instead, the nutrient requirement for gain can be separated into linear gain and non-linear gain. The changed efficiency is only associated with non-linear gain which can be calculated with efficiency of linear gain and the efficiency ratio of non-linear gain.

This idea can be applied in a robust non-linear optimization process with a known standard deviation and potential of weight gain. The standard deviation is utilized for deciding proportion of non-linear gain among the total gain and its nutrient efficiency.

B. Commercial Embodiment

As one skilled in the art will realize and as shown in Figure 4, the present invention is preferably utilized with a personal computer (hereinafter PC) that is based on Intel's 80486 microprocessor 20 with a 66 MHz clock or Intel's PENTIUM™ microprocessor. The computer also preferably has a math co-processor 22 for completing mathematical computations. The computer also includes a keyboard 24, screen 26, printer 28, random access memory 30, and a storage device 32. The storage device 32 may include magnetic means (i.e., floppy disk drive, hard drive, or tape drive), optical disk means, firm ware, or any other appropriate storage means. The storage device 32 is used to store the execution program and data generated by the execution program. The computer may also include means such as a modem 34 and communications software for loading input data or the execution program from a remote location. As one skilled in the art will further appreciate, other types of computers might be used such as a main frame, portable computer, note-book computer, or mini-computer.

C. In Operation

In operation, the user loads the execution program from the program memory storage location into the random access memory 30. Those skilled in the art will appreciate that the program might be stored on magnetic media, (i.e., floppy disk drive, hard drive, or tape), read only memory (i.e., optical disk), firm ware, or any other appropriate storage medium 32. The program might also be transmitted from a remote location such as from a file server, a main frame, or other PC that has a communication link with the user's terminal. Referring to Figure 5, a menu is displayed on the computer screen after the program is loaded. The menu has the following options: Setup 36, Products 38, Time Value 40, Management Spec's 42, Grow Out Spec's 44, Fixed/Variable Costs 46, Raw Materials 48, Choose Data Sets 50, Solve/Optimize 52, Management Report 54, Review/Predictions/Diets 56, Model Creation 58, Change Database 60, Use DOS Commands 62, and Exit to DOS (Quit)

64.

The first menu option is Setup 36. On invoking this option, a user with basic industry knowledge can define a new flock of birds or edit information concerning an existing flock. As shown in Figure 6, the Flock Data computer screen 68 is displayed when the Setup menu item is chosen. From this screen, the user has four options. The user can highlight an existing flock and press enter at which time the Flock Data Maintenance screen 70, Figure 7, will appear on the display. At this time the user can edit the displayed information, which includes the name of the farm 72 where the flock is kept, the name of the particular flock 74, the entity from which the flock was purchased 76, a reference code 78 that identifies the flock, the strain of bird 80 that comprises the flock, and whether the user wishes to have automatic age calculation 82. Automatic age calculation calculates the age based on the batch date. The user can also choose to delete the listing of a particular flock, or enter escape in order to return to the main menu.

The second item on the main menu is Products 38. Upon choosing this menu item, the Electronic Data table (EDT) entitled "TABLE: PRODUCT.T" 84 is displayed. See Figure 8. The information entered into this EDT includes the price per pound for a whole bird, a gutted carcass, and each of the individual body parts. The information entered also includes the amount of poultry product that the user wants to have available for market. More specifically, the user enters the range of acceptable tonnage that he/she plans to sell. If the user plans to market the poultry in parts, an acceptable range of tonnage for each type of part is entered. The price is entered into column 86, the minimum acceptable tonnage is entered into column 88, and the maximum acceptable tonnage is entered into column 90.

The third item on the main menu is Time Value 40. Upon choosing this menu item, a screen entitled "TABLE: TIME.T" 92 is displayed on the computer screen. See Figure 9. The data that is entered into the EDT displayed in this screen includes, the age that the poultry will be sold 94, the amount of time that a barn will be empty between flocks 96, the length of the brooding period if the particular strain of birds has a brooding period 98, and the square foot the user wants to provide for each bird within the barn 100. The unit of measurement for all time periods is days. The desired values are entered into the first column 102 of the table if the user knows the precise time period or allowable square foot per bird. Otherwise the user can enter an acceptable range of time or square footage in the second and third columns 104 and 106. If the user enters a range, the program will calculate the optimum value in order to maximize the user's return on investment.

The fourth item on the main menu is Management Spec's 42. Upon choosing this menu item, the EDT entitled "INFORMTN.T" 108 is displayed on the screen. See Figure 10. Information in this EDT is broken down into a plurality of time intervals during the life of the poultry. Each interval is called a series 110 and corresponds to a production period. In the column entitled "Age, Days" 112 the user can enter the age of the flock at the end of each interval. In the column entitled "TEMP (F) " 114 the user can enter the ambient temperature of the flocks environment. In the column entitled "HUMIDITY, %" 116 the user can enter the humidity of the flock's environment. One skilled in the art will realize that data concerning other environmental factors may also be included in the INFORMTN.T table 108.

The fifth item on the main menu is Grow Out Spec 's 44. Upon choosing this item, an EDT entitled "RECOMEND.T" 110 is displayed. See Figure 11. Information in this table is broken down into a plurality of time intervals 112 during the life of the poultry. Each interval is called a series and corresponds to a production period.

The sixth item on the main menu is Fixed/Variable Cost 46. Upon choosing this item, the EDT entitled "COST.T" 126 is displayed. See Figure 12. Data listed in this table includes "FIX, $/YR" 128, which is fixed costs per year; "PRCSS, $/YR" 130, which is the cost of processing per year; "CHICK, $/BD" 132, which is the cost of purchasing each chick; "MARKT, $/YR" 134, which is the cost of marketing per year; "PRPNE, $/YR" 136, which is the cost of building heat per year; "BROOD, $/FL" 138, which is the cost of brooding each flock of birds if the flock is of the type that requires brooding; and "GRWER, $/LB" 140, which is the cost of live weight per pound for contract grower.

The seventh item on the main menu is Raw Materials 48. Upon selection of this item, a sub-menu entitled "Raw Materials" 142 is displayed. See Figure 13. The first item on the sub-menu is Select and Price

Ingredients 144. Upon selecting this first sub-menu item, the EDT entitled "INGREDIENT UPDATE" 146 is displayed. See Figure 14.

The table includes columns entitled AVAIL. 147, GROUP 148, SHORT NAME 150, MIN 152, MAX 154, CTRL 156,

COST 157/CWT, NO 158, and HA 160. The AVAIL. 140 column lists whether that particular ingredient is available to be included in the feed. As shown in Figure 15, the possible listings in this column include Avail 162, which means that the ingredient is available to the user; Maybe 164, which means that the ingredient has a high price and the computer will try to use an alternative ingredient; No 166, which means that the ingredient is not available to the user; and Cost 168, which means ingredient will not be used in formulation but the computer will give a price at which the ingredient could be used. The GROUP column 148 lists the classification of ingredients. The SHORT NAME 150 column lists the common name of the ingredient. The MIN column 152 lists the minimum amount of that ingredient that the user wants to include in the feed. The units of measurement for this data is percentage. The MAX column 154 lists the maximum amount of the ingredient that the user wants to include in the feed. The CTRL column 156 marks those settings that cannot be changed by user in this screen. The COST/CWT column 157 lists the cost of each ingredient per 100 pounds. The NU column 158 lists the choice of predicting nutrient level. The HA column 160 lists hand add value. As will be discussed in more detail below, an ingredients database lists the types and amounts of the nutrients that are included in each ingredient. The amount of each ingredient listed in the database corresponds to the amount of ingredient that is found in a typical crop that has a standard weight per bushel. One skilled in the art will further realize that the amount of each nutrient can vary with the weight of the crop per bushel. Thus, the program of the present invention has the capability of recalculating the amount of nutrients in each ingredient if the weight per bushel is entered into the computer. The eighth item on the main menu is Choose Data Sets 50. When this item is chosen, the "TO BE FORMULATED" 170 screen is displayed. See Figure 16. This menu option allows a user to select the particular flock that is to be optimized. The ninth item on the main menu is Solve/Optimize 52. When this item is chosen the computer of the present invention will calculate the optimum rate of growth. The computer will make these calculations for each designated time interval during the life of the flock. The computer will simultaneously calculate the optimal diet, living environment, and age at which the flock should be sold. The diet consists of the amount of ingredients that should be included in the feed. The living environment includes the number of birds that are included in each flock and the density of the birds

(e.g., the square feet per bird within the barn). The age of the bird is number of days between the birth of the birds and the date at which the bird should be sold to a processing plant. One skilled in the art will realize that the computer also calculates data concerning the volume of poultry that each flock will generate and financial data concerning the amount of revenue, costs, and return on investment. One skilled in the art will further realize that other financial data may be calculated by the computer.

The tenth item of the main menu is Management Report 54. Upon selection of this menu item, a list of the possible reports 172 is displayed on the screen. See Figure 17. There are seven reports that the user can choose. The first report is entitled OPTIMUM RESULTS 174 and lists the optimal performance and environmental constraint to which the user must conform in order to realize the maximum possible Return On Investment. Figure 18. One skilled in the art will realize that such data includes the optimal flock size, the optimal age at which the flock should be sold, the optimal bird density in units of bird per square foot, the weight of the bird at sale, etc. The second report is entitled OPTIMUM PERFORMANCE 176 and includes data that relates to the length of each feeding period, the amount of feed given to the flock, the amount of feed that is consumed by the flock, etc. Figure 18a. The third report is entitled OPTIMUM YIELD 178 and includes data that relates to the total weight of the flock that is available for sale, the costs of raising the flock, and the price received for the flock. Figure 18b. The fourth report is entitled OPTIMUM FD/FACTORS 180 and includes information that relates to the amount and cost of the feed that a flock will consume. Figure 18c. The fifth report is entitled OPTIMUM NUT/ALLNCE 182 and includes information that relates to the optimal nutrient amounts that need to be consumed and that can be metered to a flock. Figure 18d. The sixth report is entitle RESOURCES RAW/MATRLS 184 and includes information related to the amount of ingredients that are consumed and inventoried for use by a flock. Figure 18e. The seventh report is entitled OPTIMUM INDIV-BIRD 186 and includes information related to the characteristics of the birds in each flock, its yield characteristics, the environmental conditions in which the flock will live, the average size of each bird within the flock, and the average amount of feed consumed by each bird within the flock. Figure 18f. Samples of the reports that are generated are shown in Figure 18 and labeled 174', 176', 178', 180', 182', 184' , and 186' .

The eleventh item on the main menu is Review/Predictions/Diets 56. When this item is selected, the computer of the present invention will display the predicted value of data concerning the weight of the flock, the amount of feed consumed, the weight of the various part of a bird, and other miscellaneous data concerning the environment of the flock. See Figure 19. This information may also be updated to reflect actual data during the life of a flock. Upon entering the actual values, the SOLVE/OPTIMIZE 52 menu item may be re-selected in order to update the optimal diet, living environment, and age at which the flock should be sold.

Preferably, the computer of the present invention is programed using the Clarion database software. Clarion is published by Clarion Software Corporation, which is located in Pompano Beach, Florida. One skilled in the art will realize that other database software packages such as Paradox, DB2, Access, etc., may be used. One skilled in the art will further realize that the computer may also be programed using the C++, Fortran, and Pascal programming languages. During execution of the program, the microprocessor sequentially executes each individual instruction. However, as described herein, the operation of the microprocessor implementing the program will be defined in terms of major functional steps.

Referring to Figure 20, the program that controls the computer of the present invention begins at block 200. The user may input information into the databases at Block 202. The information inputted may enter either the Journal database 204, Ingredient database 206, or Model database 208. The Journal database, block 204, stores information that relates to the characteristics of the flock such as sex, weight, number, strain, etc. This database also stores the information that is generated by the model and the optimizer. Such information relates to the optimal diet, environmental conditions, flock size, predicted mortality rate, predicted yield, financial figures, etc. The Ingredient database, block 206, stores information that relates to the potential ingredients that may be included within the feed and the nutritional values of the various ingredients. One skilled in the art will realize that the Ingredient database also includes equations that the user can execute to recalculate the value of the amino acid nutrients. These equations are based on the weight per bushel of the ingredients. The Model database, block 208, includes information that relates to the actual code of the execution files. The model database also includes information that relates to the variables that are used within the execution files.

One skilled in the art will realize that the blocks 210, 212, 214, 216, 218, and 220 represent the various execution programs that are required to control the computer of the present invention. One skilled in the art will further realize that any one of these block may contain a plurality of execution files in order to fulfill its function. As described above, the execution files and the databases are preferably written utilizing the Clarion database software.

At block 212, the user may execute the model that forms the equations that are described in the section above titled A. Theory. The source code for the model is shown in Figure 26 and explained in greater detail below. This section also forms equations that calculate the predicted mortality rate and other effects of living conditions, predicted yield for various economic body parts, and nutrient calculations. More specifically, the model will create a plurality of simultaneous equations that it will pass through the interface, block 222, to the Optimizer, block 224.

The interface, Block 222, reconfigures the information generated by the Model, Block 212, into a form that is acceptable by the Optimizer. The interface is preferably written in C++. The source code for the interface is attached as Figure 22. The Optimizer, Block 224, will solve the simultaneous equations in order to create the optimal values for each of the variable that describe the predicted mortality rate and other effects of living conditions, predicted yield for various economic body parts, and nutrient calculations. This information is then passed to the Journal database. Block 204, where it is stored. The preferred optimizer is GINO or Mines, which is published by The Scientific Press of San Francisco, California. At block 210, the user may execute the files that generate and print reports. These reports are described in detail above. At block 214, the user may edit the tables that store information that about the various ingredients that may be included in feed. More specifically, the user may delete or add ingredients, and edit the nutritional values associated with each ingredient. Additionally, the user may execute amino acid equation that recalculates the values of the amino acid nutrients based on the weight per bushel of each ingredient. The information manipulated by block 214, is stored in the Ingredient database, block 206.

At block 216, the user may create variables that are used in the various execution files. One skilled in the art will realize that at block 218 the user may create and edit the various tables that are used to organize and store information within the databases. Finally, the user may create and edit the execution files and databases at block 220.

Figure 21 describes the information flow of the program execution. The information flow is shown generally at 300. Block 301 illustrates the various inputs into the logical program flow in order to calculate and solve the various equations. Block 302 includes information on nutrient composition and digestibility which may be stored in the form of a look¬ up table or some other known database structure. This data is provided to block 307 where data and/or equations on the nutrient efficiency is stored. Additional information is provided to block 307 from the growth model block 303 and the nutrient to support growth block 306. Each of the various blocks 302, 307, 306, 305, and 304 provide data and equations to optimizer block 308 which solves the equations in an optimized manner. The outputs of optimizer block 308 are provided to output block 309 which provides the results to the journal data base 204 (best seen in Fig. 20). This information is illustrated as including: optimum marketing age block 310, optimum raw material tonnage & mixes block 311, optimum growth & yield of creatures block 312, optimum nutrient level/period feeding block 313, and optimum creature space density & number 314.

D. Operation of the Model Referring now to Figures 22a-22ai, the program begins at block 300. The mathematical constant e (e=2.71828) and a space factor (SPACE_FACTOR = 1000) are established at block 302. If this is the first time the model is executed, set the age, temperature, humidity values from a table for the current conditions, and set the mortality correction to zero, block 304. The sequence value is then set to the current period, block 306. If the sum of the percentage of males and females does not total 100% (plus or minus 1%) indicate a failure in the program at block 308. The next step is to give the optimizer an impossible condition at block 310 and indicate that this is the last of the series of passes, thus the user does not see the incorrect values. If the starting feed period begins when the creature is born or hatched, (P_FEED_START=1) , block 312, set the field body weight to zero (FDBWT=0), block 314. Then skip to block 342. If the starting feed period begins at some point other than the birth of the creature ( _FEED_START > 1), block 316, update the (model) sequence number and field body weight, blocks 318 and 320, respectively. Current condition information such as age, temperature and humidity is then entered, block 322.

If this is not the first pass of the program skip to block 340. Otherwise, compare optimized body weights to the field body weights at blocks 324-334. More specifically, find the current age of the creatures, look up the values in the age database and compare it to the real body weights (P_AVE_BWT) . Also compute the standard deviation (REUSE1) in the weighed creatures and the gross rate factor (REUSE3), block 324. Check for errors at block 326 in order to eliminate faulty values for body weights, its standard deviation, and the number of creatures weighed. Upon finding an error, set the standard deviation to a very large number, block 328. If the standard deviation is more than two or less than negative two, add the equation for weight at the beginning of the current period (WTB) with the equations in block 332. Otherwise add the equation for weight at the beginning of the current period (WTB) with the equations in block 334.

Compute the field mortality correction (M0RT_FLD__C0RRCT) at block 336. If it is more than four or less than negative four, set it to zero at block 338. Add equations to the model for the number of birds placed, block 340.

At block 342, set the beginning age (AGEB) to be the current age (AGE) and set the feeding period ₍FEEDING_PERIOD) to be 0. At block 344, set the ending age (AGEE) of the current sequence to be the beginning age of the next sequence as found in the database provided. The feeding period is now the difference between the ending and beginning age. Retrieve the temperature for the end of the sequence (TEMPFE, which is the same as the temperature for the next sequence) from the database, block 346.

Next set the minimum and maximum market age. If the market lock age is zero, block 348, and the ending age of the sequence is greater than the minimum allowed market age, block 350, indicate that this will be the end of the series of passes through the program (ESERIES = SERIES), block 352. Also at block 352, add equations that set the range, for the market age, setup values for the market range (MKTRGE) and the beginning market ages (MKTB, step 32). If the market lock age is not zero. block 354, and the ending age of the sequence is greater than the maximum allowed market age, block 356, indicate that this will be the end of the series of passes through the program (ESERIES = SERIES), block 358. Also at block 358, add equations that set the range for the market age, set the market range and market beginning day equal to one.

If SERIES = ESERIES add an equation for the period as shown in block 360. If SERIES does not equal ESERIES, add an equation for the period as shown in block 362.

At block 364, set the age for the diet formulation (FORMULATING_AGE) equal to the age at the middle of the current period. Then compute the beginning mortality (for current conditions) for the males and females depending on the respective percentages and add in the correction factor. At blocks 366-372, compute the mortality for each day from the beginning of the period until the end of the period and add the results to get the cumulative mortality. At block 374, divide the cumulative value by the number of days in order to obtain the incremental mortality (MORTINC) .

Also compute the effective temperature at block 374. However, if the age of the current diet formulation is less than 21 days, block 376, retrieve the value of the effective temperature from the reference temperature table for the current conditions, block 378. Next compute the adjustment period for temperature effects on body weight at block 380-384. At block 386, add that correction factor to the total correction factor for body weight affected by temperature (BW_TEMP_TOTAL) , and setup the period over which the correction factor is applied (BW_TEMP_PERIOD) .

If this is not the end of the series skip to block 400. If the feed starting period is not at the beginning, skip to block 398. Otherwise, for each period in which temperature effects on body weight ^• (REUSE) is used, find the number of days from a table for the current conditions and sum them

(BW_TEMP_PERIOD) , blocks 392 and 394. At block 398, add an equation to the model for the body weight temperature correction (TEMPBW) .

At block 400, compute the weight at maturity (MATUWT), the lower age at maximum gain (AGEMG_LB) , and upper ages at maximum gain (AGEMG_UB) . Additionally, add an equation to the model for the creature density (DNSITY) . If the feed starting period is not the first, block 402, add an equation to the model that indicates mass density (MDNSTY whose units are sq. meter/kg^Λ0.67) is greater than a very small number, block 404. Otherwise calculate the lower limit of the mass density at block 406. At block 408, set an upper limit on the mass density and then add an equation to the model for the correction factor for body weight as a function of bird density (DNBW) . If the ending age for the period is less than 35 days, block 410, set the mortality as a function of body weight (BWMORT) and mortality due to density (DNMORT) to zero, block 412. Otherwise they are calculated in either block 414 or 416. More specifically, if this is the end of the series use the equations for BWMORT and DNMORT in block 414, which are based on the market age (MKTAGE) . Otherwise the use the equations in block 416, which are based on the ending age (AGEE) for the current period.

At block 418, add an equation to the model to take into effect the blistering on the breast of the bird. If this is the end of the series, block 420, add equations to the model for number of birds at processing time (FINUMB) and the average number of birds in the period (DBIRD) based on the market age (MKTAGE), block 422. Otherwise, only add an equation for DBIRD based on the ending age of the current period, block 424.

If this is the last of the series, block 426, and the first period for the feed (P_FEED_START = 1), block 428, then add equations for creature density as provided in tables, blocks 430 and 432. If the user has supplied a creature density, use it (block 430) otherwise set the constraints as found in a table (block 432). At step 434, set constraints in the model for the maximum and minimum weight at market time from a table. If the objective is the weight of the carcass without giblets (W.O.G.), add equations for eviscerated carcass yield at block 436. If the objective is cut up parts, set constraints on breast yield at block 438.

If this is not the last of the series skip to block 448. Otherwise, add equations to the model that effect the body weight loss from fasting (FASTLS) during the time that it is being taken to market, block 440. Additionally, find the percentage of skin on breast (BRESKN) and neck (NECKSKIN) from tables for the current conditions . Use these corrections in equations that are added to the model for breast bone (BREBON), market weight (MKTWT) , and yield (YIELD).

If the objective is cut up parts, add an equation for waste (WASTE), either block 442 or 444. Next, add the equations for breast, neck, drumsticks, thighs, wings, and back to the model, block 446. Then add the equations to the model for the Gompertz rate factor. (RATEF) and the rate factor for potential growth (PRATEF) , block 448. Depending on whether this is the first period in the sequence, add the equations from either block 450 or 452 to the model. These blocks included different variations for the equations for the weight at the beginning (WTB), weight at beginning of the period (WTPB) and age at the beginning for maturity (AGEMTB) . Similarly add an equation for the weight at the end (WTE), block 454 or 456. At block 458, add the equations to the model for the weight at period end (WTPE) and set the constraint weight at the beginning period equal to the weight at the beginning.

If the feed type is not zero, block 460, assume the availability of amino acids is 90%, block 462. Otherwise the availability of amino acids is 100%, block 458. At block 464, add equations for the standard deviation for body weight (STD), the number of standard deviations for average gain in body weight (STDNO), fraction of normal curve (FRAC), and efficiency of non¬ linear gain (UTILG) .

Depending on whether the creature's age is less than ten days, add equations for protein gain and feather gain set forth in block 468 or 470.

At block 472, add the equations for the total gain (TOTALG), extra gain (EXTRAG) and the average metabolic weight (METAWT) . Then obtain the amino acid content for maintenance (AA_M) , weight gain (AA_G) and feather gain (AA_F) from the database for Arginine, Histidine, Isoleucine, Leucine, Methionine and Cystine combination, Methionine, Phen. and Tryptophan combination, Phen., Threonine, Tryptophan, and Valine. Add constraints for each of these nutrients to the model at blocks 472-490, respectively. Lysine and the Methionine and Cystine combination have lower as well as upper bounds.

The next step is to add equations for fat gain (FATG), feed intake (FI) and nutrient Metabolizable Energy (N002) to the model, block 492. Then add constraints for the effect of metabolizable energy on body weight taking into account nutritional density and feed form at this point (MEBW and MEBWT) , block 494.

If the user is reoptimizing, block 496, compute the feed cost to the present (FD_COST_TO_NOW) , block 498. For each sequence from the beginning of the period, look up the effect of metabolizable energy on gains in the sequences, the feed costs in the database, and sum them together. The user is reoptimizing if the feed start period is not the first period.

Depending on the current pass in the series, compute the effect of metabolizable energy on weight gain (MEBWT) by adding up the metabolizable energy weights from the previous passes and one in the current pass and dividing the sum by the lengths of the periods of the previous passes, blocks 500-546. At block 548, add the equations to the model for body weight correction that is dependent on metabolizable energy and the feed form (MEFFBW) .

In the first pass in the series, set the values of the period body weight and the number of the period to zero, block 550. Then look up the effect that body weight has on period in the database at block 552. If the feeding period is greater than 7 take into account the effect of the length of the feeding period on growth, block 554. If this is the last of the series of passes, compute the average effect of body weight for the entire cycle (PRDBWT), block 556. The next step is to obtain the standard metabolizable energy values (MESTD) and calculate the standard metabolizable energy maintenance coefficient for the current conditions (dependent on creature age) . This task is accomplished by looking up the value for males and females in the database and multiplying by the respective percentages, blocks 558-568. Then add an equation to the model for the metabolizable energy at 65 degrees Fahrenheit (ME65F), block 570.

Then change the standard intake of calcium, phosphorus, sodium and chlorine by adjusting it to the 65 degree fahrenheit energy levels. Next add the constraints for each nutrient (N014 and N016) blocks 570 and 572. An equation for the number of cycles per year (CYCLE) is then added to the model, block 572. In order to speed up computations, some initial values are provided at blocks 574 and 576 for AGEMG, RATEF, BWTB, BWTE, FAT, ME, which allow the system to make some initial guesses. If this is the end of the series, add different guesses for BWTE, ME, FINUMB, STNUMB, MKTWT, RATEF, YIELD, and BREAST at block 578. Then add guesses for N002, METAWT, FI, STDNO, WTB, WTE, WTPB, WTPE, AGEMTB, and MTRUE at block 580. If this is the last pass in the series add guesses for MEFFBW and MEBWT at block 582. If market lock age was not set by the user, guess the market age to be half way between the maximum and minimum market age, block 584. Otherwise set it be the lock age, block 586. At block 588, set guess values for DNSITY, MKNSTY and MKTAGE. Finally, the system sets the parameters for GINO, the optimizing software. The optimization package has tunable parameters that are set at block 590 to provide better performance. Equations for feed size (FSIZE) are then added to the model at block 592.

At block 592, price information is the retrieved from the database market weight of creature, and prices for various parts are set. Also look up the fixed costs and the sub-objective to be optimized. If the objective is cut up parts, look up the price of wasted product at block 594. If the objective selected by the user is to maximize the live bird weight, add the equations of block 596 to the model in order to constrain the sub- objectives and the maximum return on investment. If the objective selected by the user is to maximize the eviscerated carcass weight, add the equations of block 598 to the model in order to constrain the sub- objectives and the maximum return on investment. If the objective selected by the user is to maximize the price of the body parts, add the equations of block 600 to the model in order to constrain the sub-objectives and the maximum return on investment. Then obtain the minimum and maximum requirements from the database for the available feed ingredients, blocks 602-608.

While the invention has been described in conjunction with a specific embodiment thereof, it is evident that different alternatives, modifications, and variations will be apparent to those skilled in the art in view of the foregoing description. Accordingly, the invention is not limited to these embodiments or the use of elements having specific configurations as presented herein.

Claims

The invention that we claim is:

1. An apparatus for optimizing the ratio between expenditures and rate of growth for living creatures, the apparatus comprising: a) processing means for optimizing the ratio between expenditures and rate of growth for creatures, wherein the processing means calculates the minimum age at which the creatures can experience its maximum rate of growth according to W=Ae^Λ(-e^Λ(-k(t-t*))) where W is the current body weight of a creature, A is the weight of the creature at physical maturity, t is the current age of the creature, and t* is the age at which the creature has its maximum rate of growth, k is a statistical function of t*; and b) memory means for storing data corresponding to information about feed, information about the characteristics of the creatures, and information generated by the processing means, the memory means being operational coupled to the processing means.

2. The apparatus of claim 1 wherein the life of a creature is divided into a plurality of intervals, further wherein the processing means optimizes the ratio between expenditures and rate of growth for each of the intervals according to

W=Ae^Λ(-e^Λ(-k(t.-t*) ) ) , and W=Ae^Λ(-e^Λ(-k(t₂-t*) ) ) where t₁ is the age of the creature at the beginning of ' the interval and t₂ is the age of the creature at the end of the interval.

3. The apparatus of claim 2 further comprising optimization means for solving a series of simultaneous equations in order to calculate the minimum age at which the creatures can experience its maximum rate of growth. 4. The apparatus, of claim 3 wherein the processing means determines the raw materials and living conditions required in order to maximize the amount of profit realized in raising creatures for sale as a source of food.

5. The apparatus of claim 4 further comprising input means for entering information into the memory means;

6. The apparatus of claim 5 further comprising display means for displaying the raw material and living con_ditions required in order to maximize the amount of profit realized in raising creatures for sale as a source of food.

7. The apparatus of claim 5 further comprising printing means for printing the raw material and living conditions required in order to maximize the amount of profit realized in raising creatures for sale as a source of food.

8. _A method for operating a computer consisting of a processor and memory, the method comprising the steps of: a) calculating the minimum age at which the creatures can experience its maximum rate of growth according to

W=Ae^Λ(-e^Λ(-k(t-t*))) where W is the current body weight of a creature, A is the weight of the creature at physical maturity, t is the current age of the creature, and t* is the age at which the creature has its maximum rate of growth, k is a statistical function of t*; and b) storing data corresponding to information a_bout feed, information about the characteristics of the creatures, and information generated by the processing means. 9. The method of claim 8 wherein the life of a creature is divided into a plurality of intervals, further wherein the step of calculating is accomplished by solving a set of simultaneous equations including

W=Ae^Λ(-e^Λ(-k(t₁-t*))), and

W=Ae^Λ(-e^Λ(-k(t₂-t*) )) where t, is the age of the creature at the beginning of the interval and t₂ is the age of the creature at the end of the interval.

10. The method of claim 9 comprising the additional step of determining the raw materials and living conditions required in order to maximize the amount of profit realized in raising creatures for sale as a source of food.

11. The method of claim 10 comprising the additional step of displaying the raw material and living conditions required in order to maximize the amount of profit realized in raising creatures for sale as a source of food.

12. The method of claim 11 comprising the additional step of printing the raw material and living conditions required in order to maximize the amount of profit realized in raising creatures for sale as a source of food.

13. A method of determining utilization effectiveness of nutrients in a population, comprising the steps of: a) determining the standard deviation for the average period gain of the population and the average period potential gain of the population; and b) comparing the actual gain of the population with the average period potential gain to determine the number of standard deviations of non-linear gain, wherein the nutrients to support the linear and/or non-linear gain may be determined.

14. The method of claim 13, further comprising the steps of determining the efficiency of non-linear and linear gain by calculating the efficiency at predetermined steps of standard deviations and taking the average of each calculated efficiency, wherein the linear gain and the non-linear gain can be used in combination with their respective efficiency to create a diet for the population.

15. A computerized system for determining the utilization effectiveness of nutrients in a population based on the theoretical average period gain of the population, comprising: a) memory means for storing data on the standard deviation of period gain of the population and the period gain potential of the population; b) first determining means for determining the number of standard deviations for the average non¬ linear gain of the population from information stored in said memory means; and c) second determining means for determining the nutrition efficiency ratio (EFR) in accordance with the following equation: EFR = 1.6551e^Λ(- 0.31768e^Λ(0.35 X STD)) where EFR is the nutrient efficiency ratio, e=2.71828, and STD is the number of standard deviations (-3 to 3). 16. The computerized system of claim 3, wherein said second determining means determines the efficiency of non-linear gain according to the following equation: EFNL = EFL X EFR, where EFNL is the nutrient efficiency of non-linear gain and EFL is the nutrient efficiency of linear gain.