Statistique Theorique et Appliquee, Tome 1: Les Bases Theoriques. by P. DagnelieBiometrics, Vol. 50, No. 3 (Sep., 1994), pp. 895-896Published by: International Biometric SocietyStable URL: http://www.jstor.org/stable/2532818 .

Accessed: 25/06/2014 10:22

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

.

International Biometric Society is collaborating with JSTOR to digitize, preserve and extend access toBiometrics.

http://www.jstor.org

This content downloaded from 185.2.32.58 on Wed, 25 Jun 2014 10:22:14 AMAll use subject to JSTOR Terms and Conditions

Book Reviews 895

The authors set out to provide "a unified theory for the analysis and interpretation of data on resource selection". They succeed in providing a comprehen- sive account of resource selection data together with a less comprehensive account of statistical models for their analysis. The book's major weakness is that it treats statistical modelling too superficially. Despite this, field ecologists and biologists should find it a useful starting point for the analysis of such data.

A program to implement some of the methods for analysis is available from the authors.

REFERENCE

Aebischer, N. J., Robertson, P. D., and Kenwood, R. E. (1993). Compositional analysis of habitat use from annual radio-tagging data. Ecology 75, 1313-1325.

D. BORCHERS Wildlife Population Assessment Research Group

School of Mathematical and Computational Sciences University of St Andrews

St Andrews, Scotland

DEANGELIS, D. L. Dynamics of Nutrient Cycling and Food Webs. Chapman and Hall, London, 1992. xv + 270 pp. ?19.99. ISBN 0-412-29840-6 (paper).

This book is No. 9 in the "Population and Community Biology Series". It is intended for graduate students in ecology with "some acquaintance of elementary dif- ferential equations". However, a substantial back- ground in maths, matrices as well as differential equa- tions, is required if much benefit is to be gained. It is for the reader with a serious interest in ecological modelling. The author's aim is to integrate, through mathematical modelling, the interrelationships be- tween the dynamics of ecological food webs and the cycling of nutrients. My interest and expertise is in nutrient cycling, so I was interested to see if the author succeeded in his aim and was able to introduce me to, and interest me in ecological modelling.

The main text consists of twelve chapters, four appendices giving details of some of the calculations, a very comprehensive list of about 530 references [in- valuable in itself], and a subject index. It covers a broad range of scales in ecology from local pollution events to global issues such as carbon cycling. It starts " small" with nutrient cycles and builds by adding living organisms and interactions to the system.

Chapter 1 introduces energy and nutrient flows or cycles, food webs, and models. I liked the emphasis on

food webs rather than chains, pointing up the interac- tions that are always involved in real ecological sys- tems. Chapter 2 establishes the basis of nutrient fluxes, and introduces the ideas of stability and resil- ience and the first, relatively simple, equations for single-compartment models of linear systems. As a nonspecialist I found the development of models and equations clearly set out and easy to follow.

Chapter 3 introduces living components into the models and thus greatly increases their complexity. Chapters 4-6 develop this complexity, beginning with autotrophs [primary producers], and moving on to herbivores and their interactions with autotrophs and with nutrient cycling. The riole of nutrient limitation is extensively covered and catastrophe introduced. We are now into nonlinear systems and the maths is be- coming progressively more complex but, again, the development and explanation are excellent and the interested reader should have no difficulty in following the arguments.

Chapter 7, Nutrient Interactions of Detritus and Decomposers, was especially interesting to me. Mod- ellers of nitrogen and carbon cycling in agriculture realize the need to make a much better quantification of the retur-n of detritus [organic matter] to the soil and its decomposition. The author goes into the detail of modelling mineralization and immobilization reac- tions, and the effects of N:C ratios on these processes.

Chapters 8 and 9 cover the effects of nutrient limi- tation on food webs and of competition for nutrients, and Chapters 10 and I 1 focus on temporal and spatial variation in ecological systems. Chapter 12 consists of a general discussion of human effects on the global carbon cycle, the effects of increasing CO, on climate, and the effects of climate change on the environment.

The book reminded me again of the complexity of even the simplest ecosystem and the simplification that the most complex model must accept if it is to be practicable. However, the book also exemplifies one of the values of modelling: the ability of a model to enable the researcher to focus on a system or process, consider its interactions with other systems or pro- cesses, and see where the gaps in understanding lie.

This is a well-written book; I noted only a few errors. It is very good value at ?17.95 for 270 pages in paperback, an excellent introduction to ecological modelling.

K. W. T. GOULDING Soil Science Department

Rothamsted Experimental Station, Harpenden England

BRIEF REPORTS BY THE EDITORS

DAGNELIE, P. Statistique Theorique et Appliquee, Tome 1: Les Bases Theoriques. Les Presses Agronomiques de Gembloux, Gembloux, Belgique, 1992. x + 492 pp. FF187.00. ISBN 2-87016-039-9.

Professor Dagnelie's series of books on statistical the- ory and methodology has been appreciated widely for the clarity with which he has made the mathematical foundations accessible to students and researchers in biometry. Volumes 1 and 2 of Theorie et Methodes Statistiques: Applications Agronomiques first ap- peared in 1969 and 1970. They were joined by a third

volume, Analyse Statistique a Plusieurs Variables, in 1975. Pr incipes d'Expe&rimentation and ThMeroie et Methodes Statistiques: Exercices were added in 1981 and 1975.

A second edition of Volume 1 was published in 1969 and was reprinted many times. Its replacement begins with two new chapters on the objectives and nature of data collection and data analysis. Here, as generally throughout the book, there are plentiful French and English references which enable the reader to pursue topics in much greater depth; in all there are now more than 400 references, many from the research litera-

This content downloaded from 185.2.32.58 on Wed, 25 Jun 2014 10:22:14 AMAll use subject to JSTOR Terms and Conditions

896 Biometrics, September 1994 ture. Material on one- and two-dimensional descrip- tive statistics is now in Chapters 3 and 4, whilst the material on three-dimensional descriptive statistics has been dispersed. There is greater attention to prop- erties of distributions and to generating functions and the characteristic function (Chapter 5), and more ma- terial on particular distributions (Chapter 6). The last three chapters (8, 9, and 10), on sampling distribu- tions, estimation, and tests of hypotheses, have also been rewritten. The text is well presented with very good typography.

BOFINGER, E., DUDEWICZ, E. J., LEWIS, G. J., and MENGERSEN, K. (eds). The Frontiers of Modern Statistical Inference Procedures, II. American Sciences Press, Columbus, Ohio, U.S.A., 1992. xii + 498 pp. $98.75. ISBN 0-935950-30-3.

This volume contains the proceedings of the Second International Advanced Seminar/Workshop on Infer- ence Procedures Associated with Statistical Ranking and Selection, IPASRAS II, held at the University of Sydney, Australia, in 1978. The intention of the con- ference was to honour the eminent Australian statisti- cian Professor P. A. P. Moran as a specially invited participant. Sadly, he died one week before the meet- ing. The proceedings are dedicated to his memory.

The meeting adhered to the format established at IPASRAS I, i.e., 30-60 minutes for each paper, with 15-30 minutes for discussion and preplanned discus- sants. There are 18 papers in the volume, with topics rianging firom "Ranking and Selection from Paired- Comparison Data" (H. A. David) and "Subset Selec- tion in Crop Variety Trials" (K. L. Butler) to "Select- ing the Best Candidate for the Job" (K. Mengersen). Every paper is accompanied by an edited version of the taped discussion.

The volume also includes reports fiom the follow- ing five discussion groups: Journals and RSM (ranking and selection methods); Notation and RSM (a topic that arose repeatedly throughout the meeting); Bayes- ian and Conditional Inference Approaches to RSM; Applications, Software, and Computing in RSM; and Teaching RSM at Undergraduate and Graduate Lev- els.

HARTER, H. L. The Chronological Annotated Bibli- ography of Order Statistics, Volume VII: 1968-1969; Volunme VIII: Indices, with a Supplenment on 1970-1992. American Sciences Press, Syracuse, New York, 1992- 1993. Volume VII: vi + 292 pp., $95.00, ISBN 0-935950-25-7; Volume VIII: vii + 267 pp., $110.00, ISBN 0-935950-26-5.

These two volumes complete Dr Harter's definitive bibliography of order statistics for the period up to and including 1969. Volume VIII (for 1968-1969) upholds the same high standards and has the same format as the previous volumes. Volumes I and II (pre-1950 and 1950-1959, respectively) were reviewed in Biotnetrics 40, p. 873, Volume III (1960-1961) was reviewed in Biometrics 47, p. 1638, and Volume IV (1962-1963) was reviewed in Biomnetrics 48, p. 660. A review of Volumes V and VI (1964-1965 and 1966-1967) ap- peared in Bionmetrics 48, p. 977.

Volume VIII is different. It contains author and subject indexes for Volumes III-VII, citation lists for Volumes I-VIII, and a nine-page list of errata for Volumes III-VII. Volume VIII ends with an article, with references, by Dr N. Balakrishnan commenting on developments in order statistics from 1970 to the date of publication.

MEIER, P. C. and ZUND, R. E. Statistical Methods in Analytical Chemistry. Wiley, New York, 1993. xiv + 321 pp. $64.95/?54.00. ISBN 0-471-58454-1.

This is a book by two analytical chemists for workers in GMP/GLP (Good Manufacturing Practice/Good Laboratory Practice) environments. Its subject matter comprises the calibration of measuring equipment and the collection and interpretation of data obtained fiom automated instrumentation. The book comes with a floppy disk containing 32 BASIC programs and some 30 sample data sets on quality control and reliability. Chapters 1 and 2 are on Univariate and Bi- and Mul- tivariate Data-the material here includes confidence limits, t-tests, ANOVA, CUSUM charts, linear and non- linear regression, and full factorial designs. Chapter 3 gives an overview of, initer ailiai, EDA, error propaga- tion, smoothing, filtering, and computer simulation. The authors find "no room for proving theory, how- ever trivial". Instead they provide a wealth of practi- cal examples in Chapter 4; many have a pharmaceuti- cal origin. There are several appendices, including ones on numerical approximation for some fiequently used sampling distributions and instructions for the use of the software.

BAKER, F. B. Item Response Theory: Parameter Es- timation Techniiques. Marcel Dekker, New York, 1992. xiii + 440 pp. $125.00/?83.99. ISBN 0-8247-8636-X.

This is a book for psychometric researchers, testing and evaluation specialists, and those who are inter- ested in the r6le of the item characteristic curve in psychometric test theory and in the estimation of its parameters. Item response theory has developed rap- idly over the last few decades, with applications in a wide field; concomitants are fiagmentation of the lit- erature and notational inconsistency.

The author has endeavoured to unify the coverage of item response theory by concentrating on the un- derlying logic and by presenting the mathematics in a coherent notation. A feature of the book is the inclu- sion of seven appendices with BASIC computer pro- grams that help the reader to understand the imple- mentation of the various estimation procedures (MLE of item parameters, MLE of examninee's ability, joint MLE for the Rasch model, the Bayesian approach, item and parameter estimation under a graded re- sponse model, and MLE of item and ability parame- ters under norninal response scoring).

CLEMENT, P. and LUMER, G. (eds). Evolution Equations, Control Theory, and Bioimathematics. Mar- cel Dekker, New York, 1994. xxvi + 580 pp. $165.00/ ?f109.55. ISBN 0-8247-8885-0.

This collection of nearly 50 papers, arranged alphabet- ically by author, is the outcome of the Third (1991) International Workshop-Confer-ence on1 Evolution Equations, Control Theory, and Biomathematics. Many of the papers are purely mathematical and some of the others are based on problems in physics and engineering. Very few relate directly to biological ap- plications; these include "On the Balance between Mutation and Frequency-Dependent Selection in Ev- olutionary Game Dynamics" by Bomze and Biiger, "The Cumulative Formulation of (Physiologically) Structured Population Models" by Diekmann, Gyllen- berg, Metz, and Thieme, and "Can We Understand What Life Is?" by Thom.

This content downloaded from 185.2.32.58 on Wed, 25 Jun 2014 10:22:14 AMAll use subject to JSTOR Terms and Conditions