单纯形加速法拟合生态学中的非线性模型

本文以Logistic模型,Taylor幂法则模型,Holling功能反应模型,以及种群内禀增长力Rm等模型的拟合和参数估计为例,探讨单纯形加速法在生态模型优化拟合和参数估计中的应用.结果表明,单纯形加速法拟合生态学中的非线性模型不仅适用广泛,而且拟合过程是直接求原来非线性模型的最优拟合,因而优于生态学中通常使用的将原模型“线性化后再拟合”的方法,而与其它一些最优化方法,如:麦夸方法、枚举选优法等比较,由于单纯形法不需计算目标函数的偏导数,因而计算不受目标函数及其偏导函数复杂程度的限制,而且对于各种模型其求优计算过程十分相似,可以编制统一的计算程序.本研究所编制的计算机程序对于本文未提到的其它一些模型也是完全适用的,在应用时仅需修改定义目标函数的自定义函数语句即可.研究也发现,在求优过程中,只要搜索系数选择适当和实际数据合理,是可以保证寻优成功的.     

麦草栽培姬菇配方的数学模型研究

借助二次通用旋转组合设计法研究了麦草栽培姬菇配方的数学模型,以探讨利用麦草替代棉籽壳的新配方.试验结果得出了配料的取值范围及最佳栽培配方,即麦草79％,麸皮15％,玉米粉4％,过磷酸钙1％,黄豆饼粉1％.在实际生产中很有应用价值.     

Enzymatic hydrolysis of lignocellulosic materials: II. Experimental investigations of theoretical hydrolysis--process models for an increased enzyme recovery

Stream pretreatment of wheat straw solubilized most of the xylan present. Xylose and other sugars were recovered by washing the substrate with water but only a minor part (34%) was monomeric. Treatment of this solutions with celulases and hemicellulases improved the yield of monomeric sugars to 69%, the main product being xylose. Some xylose was also obtained during enzymatic hydrolysis of the solid substrate although the pretreatment step contributed 64% (mean value) of total xylose formed. A reference model, No. 1, and two other models, Nos. 2 and 4, described in the first part of this article series (this issue) have been studied experimentally and results confirm the theoretical conclusions. An uninterrupted hydrolysis over a given time period leads to a lower degree of saccharification than when hydrolysate is withdrawn several times. Saccharification is also favored if the residue is removed at a late stage, i.e., at the end of the 24 h hydrolysis cycle. Extended recirculation of the enzymes during a 4 x 24-h experimental period gave the following average yields of saccharification on a 24-h basis: 65% (Reference), 73% (Model 2), and 79% (Model 4). It is concluded that enzyme recovery with model 4 is 70% or more, while the Reference and Model 2 attain a lower level of recovery. The design of an improved hydrolysis model is also discussed.     

Correlating the effect of pretreatment on the enzymatic hydrolysis of straw

Avicell, Alkali-treated straw cellulose (ATSC), and wheat straw were ball-milled to reduce crystallinity; wheat straw was delignified by hot (120 degrees C) sodium hydroxide solutions of various concentrations. The physically and chemically pretreated cellulosic materials were hydrolyzed by the cellulases of Fusarium oxysporum strain F3. Enzymic hydrolysis data were fitted by the hyperbolic correlation of Holtzapple, which involves two kinetic parameters, the maximum conversion (X(max)), and the enzymic hydrolysis time corresponding to 50% of X(max) (t(1/2)). An empirical correlation between X(max) and cellulose crystallinity, lignin content, and degree of delignification has been found under our experimental conditions. Complete cellulose hydrolysis is shown to be possible at less than 60% crystallinity indices or less than 10% lignin content.     

Estimates of marker-associated QTL effects in Monte Carlo backcross generations using multiple regression

Summary The decision of whether or not to use QTLassociated markers in breeding programs needs further information about the magnitude of the additive and dominance effects that can be estimated. The objectives of this paper are (1) to apply some of the Moreno-Gonzalez (1993) genetic models to backcross simulation data generated by the Monte Carlo method, and (2) to get simulation information about the number of testing progenies and mapping density in relation to the magnitude of gene effect estimates. Results of the Monte Carlo study show that the stepwise regression analysis was able to detect relatively small additive and dominance effects when the QTL are independently segregating. When testing selfed families derived from backcross individuals, dominance effects had a larger error standard deviation and were estimated at a lower frequency. Linked QTL require a higher marker mapping density on the genome and a larger number of progenies to detect small genetic effects. Reduction of the environmental error variance by evaluating selfed backcross families in replicate experiments increased the power of the test. Expressions of the number of progenies for detecting significant additive effects were developed for some genetic situations. The ratio of the within-backcross genetic variance to the square of a gene effect estimate is a function of the number of progenies, the heritability of the trait, the marker map density and the portion of the genetic variance explained by the model. Different values (from 0 to 1) assigned to (relative position of the QTL in the marker segment) did not cause a large shift in the residual mean square of the model.     

Detection of significant demographic differences between subpopulations of prehispanic Maya from Copan, Honduras, by survival analysis

Heterogeneity and small sample size are problems that affect many paleodemographic studies. The former can cause the overall distribution of age at death to be an amalgam that does not accurately reflect the distributions of any of the groups composing the heterogeneous population. The latter can make it difficult to separate significant from nonsignificant demographic differences between groups. Survival analysis, a methodology that involves the survival distribution function and various regression models, can be applied to distributions of age at death in order to reveal statistically significant demographic differences and to control for heterogeneity. Survival analysis was used on demographic data from a heterogeneous sample of skeletons of low status Maya who lived in and around Copan, Honduras, between A.D. 400 and 1200. Results contribute to understanding the collapse of Classic Maya civilization.     

On the generalization of the logistic law of growth

This communication presents a discussion of some extensions of the formalism of Verhulst's simple logistics, which may constitute an autonomous growth model of a more general scope.For that purpose, the basis concept of growth diagram or trajectory is called upon, as it affords the graphic representation of the change in the growth variable y, using two relevant kinetic parameters: the instantaneous rate and the instantaneous acceleration. The two possible kinds of trajectories are in relation to the use of absolute (V = dyldt; = dV/dt) or relative (or specific) values (R = (1/y)(dy/dt); _R = dR/dt).In the case of simple logistics, the trajectory (V, ) allows 4 growth phases or states to be distinguished. The diagram (R, _R ) shows that the deceleration of the specific rate is not monotonous.In the case of Richards - Nelder's generalized logistics, the qualitative variation of the growth trajectory depends on the value of the dissymmetry parameter (occurrence of a critical value which determines the number of growth states).Blumberg's model is characterized by an analogous property and, moreover, can account for a non monotonous variation of the specific growth rate.     

Nutritional criteria in machiguenga food production decisions: A linear-programming analysis

Machiguenga Indians of the Peruvian Amazon, like other low populationdensity, technologically “simple” peoples, produce ample food for a nutritious diet. Assuming that this is an intended outcome of their foodproduction strategy, to what extent is it a labor-efficient solution to the problem of producing a “balanced diet”? A linear-programming model of the “diet problem” is constructed with parameters reflecting the Machiguenga economy, and solutions are computed. These are then compared to observed Machiguenga food production; the degree of fit between model and behavior is examined and reasons for discrepancies are discussed.