Efficient Method for Assigning Students to Proper Groups

European Journal of Scientific Research
ISSN: 1450-216X Volume 21, No 2 August, 2008
Editor-In-chief or e
Adrian M. Steinberg, Wissenschaftlicher Forscher
Editorial Advisory Board e
Parag Garhyan, Auburn University Teresa Smith, University of South Carolina
Morteza Shahbazi, Edinburgh University Ranjit Biswas, Philadelphia University
Raj Rajagopalan, National University of Chiaku Chukwuogor-Ndu, Eastern Connecticut

Singapore State University
Sang-Eon Park, Inha University John Mylonakis, Hellenic Open University

(Tutor)
Said Elnashaie, Auburn University
M. Femi Ayadi, University of Houston-Clear
Subrata Chowdhury, University of Rhode Lake
Island
Emmanuel Anoruo, Coppin State University
Ghasem-Ali Omrani, Tehran University of
Medical Sciences H. Young Baek, Nova Southeastern University
Ajay K. Ray, National University of Singapore Dimitrios Mavridis, Technological Educational

Institure of West Macedonia
Mutwakil Nafi, China University of
Geosciences Mohand-Said Oukil, Kind Fhad University of
Petroleum & Minerals
Felix Ayadi, Texas Southern University
Jean-Luc Grosso, University of South Carolina
Bansi Sawhney, University of Baltimore
Richard Omotoye, Virginia State University
David Wang, Hsuan Chuang University
Mahdi Hadi, Kuwait University
Cornelis A. Los, Kazakh-British Technical
University Jerry Kolo, Florida Atlantic University
Jatin Pancholi, Middlesex University Leo V. Ryan, DePaul University
As of 2005, European Journal of Scientific Research is indexed in ULRICH, DOAJ and CABELL
academic listings.
http://www.eurojournals.com/ejsr.htm
Editorial Policies:
1) European Journal of Scientific Research is an international official journal publishing high quality
research papers, reviews, and short communications in the fields of biology, chemistry, physics,
environmental sciences, mathematics, geology, engineering, computer science, social sciences,
medicine, industrial, and all other applied and theoretical sciences. The journal welcomes submission
of articles through ejsr@eurojournals.com.
2) The journal realizes the meaning of fast publication to researchers, particularly to those working in
competitive & dynamic fields. Hence, it offers an exceptionally fast publication schedule including
prompt peer-review by the experts in the field and immediate publication upon acceptance. It is the
major editorial policy to review the submitted articles as fast as possible and promptly include them in
the forthcoming issues should they pass the evaluation process.
3) All research and reviews published in the journal have been fully peer-reviewed by two, and in some
cases, three internal or external reviewers. Unless they are out of scope for the journal, or are of an
unacceptably low standard of presentation, submitted articles will be sent to peer reviewers. They will
generally be reviewed by two experts with the aim of reaching a first decision within a three day
period. Reviewers have to sign their reports and are asked to declare any competing interests. Any
suggested external peer reviewers should not have published with any of the authors of the manuscript
within the past five years and should not be members of the same research institution. Suggested
reviewers will be considered alongside potential reviewers identified by their publication record or
recommended by Editorial Board members. Reviewers are asked whether the manuscript is
scientifically sound and coherent, how interesting it is and whether the quality of the writing is
acceptable. Where possible, the final decision is made on the basis that the peer reviewers are in
accordance with one another, or that at least there is no strong dissenting view.
4) In cases where there is strong disagreement either among peer reviewers or between the authors and
peer reviewers, advice is sought from a member of the journal's Editorial Board. The journal allows a
maximum of two revisions of any manuscripts. The ultimate responsibility for any decision lies with
the Editor-in-Chief. Reviewers are also asked to indicate which articles they consider to be especially
interesting or significant. These articles may be given greater prominence and greater external
publicity.
5) Any manuscript submitted to the journals must not already have been published in another journal or
be under consideration by any other journal, although it may have been deposited on a preprint server.
Manuscripts that are derived from papers presented at conferences can be submitted even if they have
been published as part of the conference proceedings in a peer reviewed journal. Authors are required
to ensure that no material submitted as part of a manuscript infringes existing copyrights, or the rights
of a third party. Contributing authors retain copyright to their work.
6) Submission of a manuscript to EuroJournals, Inc. implies that all authors have read and agreed to its
content, and that any experimental research that is reported in the manuscript has been performed with
the approval of an appropriate ethics committee. Research carried out on humans must be in
compliance with the Helsinki Declaration, and any experimental research on animals should follow
internationally recognized guidelines. A statement to this effect must appear in the Methods section of
the manuscript, including the name of the body which gave approval, with a reference number where
appropriate. Manuscripts may be rejected if the editorial office considers that the research has not been
carried out within an ethical framework, e.g. if the severity of the experimental procedure is not
justified by the value of the knowledge gained. Generic drug names should generally be used where
appropriate. When proprietary brands are used in research, include the brand names in parentheses in
the Methods section.
7) Manuscripts must be submitted by one of the authors of the manuscript, and should not be submitted
by anyone on their behalf. The submitting author takes responsibility for the article during submission
and peer review. To facilitate rapid publication and to minimize administrative costs, the journal
accepts only online submissions through ejsr@eurojournals.com. E-mails should clearly state the name
of the article as well as full names and e-mail addresses of all the contributing authors.
8) The journal makes all published original research immediately accessible through
www.EuroJournals.com without subscription charges or registration barriers. European Journal of
Scientific Research indexed in ULRICH, DOAJ and CABELL academic listings. Through its open
access policy, the Journal is committed permanently to maintaining this policy. All research published
in the. Journal is fully peer reviewed. This process is streamlined thanks to a user-friendly, web-based
system for submission and for referees to view manuscripts and return their reviews. The journal does
not have page charges, color figure charges or submission fees. However, there is an article-processing
and publication charge.
Further information is available at: http://www.eurojournals.com/ejsr.htm

© EuroJournals Publishing, Inc. 2005
Volume 21, No 2 August, 2008
Contents
Détermination de L’âge et Modélisation de la Croissance D’un Poisson: L’anchois

Européen (Engraulis Encrasicolus) dans le golfe de Skikda 229-238
L. Mezedjri, A. T. Bensahla and A. Tahar
Efficient Fault Tolerant Mapping of Large Three-Dimensional Simulation onto 3D Tori Graph 239-248
Saleh H. Al-Sharaeh
Efficient Method for Assigning Students to Proper Groups 249-358

Moh'd Belal Al- Zoubi, Imad Salah, Azzam Sleit, Saleh Al-Sharaeh
Ammar Huneiti and Nadim Obeed
Evolution de la Microflore et Caractéristiques Physico-Chimiques D’un Aliment

Traditionnel à Base de Manioc (Manihot Esculenta Crantz) Fermenté: le Bêdêcouman 259-267
Rose Koffi-Nevry, Marina Koussemon, Zinzendorf Y. Nanga
Mobio Marguerite Abro, Celah Kakou and Tano Kablan
Study of Performance of Anthropometric Indices in Screening Children with

Protein-Energy Malnutrition at Risk of Death 268-276
Hassan Barouaca, Ftima Zahra Squali Houssaini, Moustapha Hida
Ahamed Omar Touhami Ahami and Youssef Aboussaleh
Synthesis and Bioactivity of a Novel Cyclic Hexapeptide from Stellaria Delavayi 277-287
Rajiv Dahiya and Ravi Dutt Sharma
On-Line Power Quality Monitoring System Using CompactRIO Device: A Review

on System Development 288-295
M. Z. A Ab Kadir, S. S. Tuan Othman and L. Lombigit
Earnings, Market Value, and Option-Based Equity Valuation 296-304

Mohsen Dastgir and Abdolreza Talaneh
Regime Alimentaire du Poisson-Chat Schilbe Mandibularis (Günther, 1867) Dans Deux

Bassins Fluviaux de Cote D’ivoire (Bia et Agnebi) 305-313
Doumbia Lassina, Ouattara Allassane and Gourene Germain
Weights Adjustment Neural Networks 314-318

Al-Salaimeh Safwan and Hanna Jaber
Shape Preserving Convex Surface Data Visualization Using Rational Bi-Quartic Function 319-327
Malik Zawwar Hussain, Maria Hussain and Tahira Sumbal Shaikh
Dynamics of the Functional Groups of the Meadow Spontaneous Species Around the
Lakes of the Area of El-Tarf (Algeria) 328-345
Slimani. A, Plantureux. S, Brinis. L and Soltane. M
Effects of Job Stress on Health, Personal and Work Behaviour of Nurses in
Public Hospitals in Ibadan Metropolis, Nigeria 346-352
J.K. Mojoyinola
Modified Atmosphere Packaging of Strawberries (Fragaria X Ananassa Duch.)

Stored Under Temperature Fluctuation Conditions 353-364
Kablan Tano, Françoise A. Kouamé, Rose K. Nevry and Mathias K. Oulé
A Comparative Study of Engine Performance by Using Different Bio-Fuels 365-375

M. Sadat Ali, P.S.V. Ramana Rao and C.V. Gopinath
ISSN 1450-216X Vol.21 No.2 (2008), pp.229 -238
Détermination de L’âge et Modélisation de la Croissance D’un

Poisson: L’anchois Européen (Engraulis Encrasicolus)
Dans le Golfe de Skikda
L. Mezedjri
Département de biologie, Faculté des sciences et des sciences de l’ingénieur
Université du 20 août 1955, BP 26 route El Hadeik, 21000 Skikda. Algérie
E-mail: lyamine_mez@yahoo.fr
Tel: +213 660 322 125
A. T. Bensahla
Département de Biologie, Faculté des Sciences, Université Sénia
BP 16 Sénia Oran. Algérie
E-mail: ahmedbensahla@yahoo.fr
A. Tahar
Département de biologie, Faculté des sciences
Université de Badji Mokhtar, BP12, 23000 Annaba. Algérie
E-mail: pr_tahar_ali@hotmail.com
Abstract
During a yearly cycle spreading from July 2003 to June 2004, on a sample of 1605
fishes collected in the gulf of Skikda (Algerian east costs), a study of age and the growth of
the European anchovy (Engraulis encrasicolus) has been achieved.
In population’s dynamics, and for a better knowledge of the growth, the survey of
age is primordial. The determination of age has been done by two indirect statistical
methods: the graphic method of Petersen and the method of the logarithmic differences or
the method of Battacharya.
On the other hand, to explain the absolute linear growth and the absolute ponderal
growth of fish, we used the statistical model of Von Bertalanffy that is the more used in
ichthyology and whose use is very flexible. The relative growth has been studied by mean
of a cubic shape allometric model expressing the weight according to the size.
Key words: European anchovy, growth, age, Statistic methods, model of Von Bertalanffy,
allometric model.
Introduction
L’anchois européen, Engraulis encrasicolus (Linné, 1758) a fait l’objet d’un bon nombre d’études
consacrées à la croissance de cette espèce (Pauly, 1978; Dorel, 1986; Djabali et Hamida, 1989;
Karacam et Düzgunes, 1990; Campillo, 1992; Djabali et al., 1993; Stergiou et Moutopoulos, 2001 et
Abdallah, 2002). Ces études se sont intéressées à la croissance linéaire et pondérale de cette espèce,
Européen (Engraulis Encrasicolus) dans le golfe de Skikda 230
ainsi qu’à la croissance relative. Le présent travail apporte de nouvelles informations sur la croissance
de Engraulis encrasicolus dans le golfe de Skikda (côtes Est algériennes).
L’étude de l’âge, l’étude de la croissance absolue linéaire et pondérale et en fin l’étude de la
croissance relative (taille poids) ont été réalisés et les résultats obtenus sont confrontés à ceux recueillis
dans la littérature dans différentes régions de la méditerranée.
Matériels et méthodes
L’étude de la croissance a été consacrée à des poissons pêchés au moyen des sennes coulissantes au
niveau du golfe de Skikda durant la période s’étalant entre juillet 2003 et juin 2004. Juste au
débarquement au niveau du port de Stora (golfe de Skikda) (figure 1), un échantillon d’un minimum d’
1 kg est prélevé chaque mois. Sur chaque individu on effectue une série de mesures (tableau 1). Les
mesures de longueurs sont faites à l’aide d’un ichtyomètre au millimètre prés, les poids, total et
éviscéré, au moyen d’une balance de précision Sartorius avec une précision de 0,01 g.
Figure 1: Zone d’étude: golfe de Skikda (G S), Côtes Est algériennes.
Tableau 1: Mesures effectuées sur les poissons échantillonnés
N° Code Description
1 Lt Longueur totale
2 Wt Poids total
3 Wev Poids éviscéré
1) Etude de l’âge
En dynamique des populations et pour de meilleures connaissances de la croissance, l’étude de l’âge
est primordiale. La détermination de l’âge se fait en générale soit par des méthodes directes (lecture sur
des pièces anatomiques), soit par des méthodes indirectes. Pour l’étude de l’âge un échantillon de 725
poissons est prélevé instantanément au mois d’avril, puis la longueur totale de tous ces poissons est
mesurée au moyen d’un ichtyomètre. Les poissons sont regroupés dans des classes de tailles d’un
intervalle de 0,2 cm.
Dans notre étude parmi les méthodes indirectes utilisées, on a:
 La méthode graphique de Petersen:
Elle consiste à une observation de l’évolution des modes des histogrammes de fréquences des
tailles et a la recherche des sous populations.
 La méthode des différences logarithmiques ou Battacharya:
On applique cette méthode sur le même échantillon étudié précédemment. Cette méthode est
basée sur une transformation logarithmique des effectifs regroupés en classes de tailles d’égales
amplitudes «h = 0,2 cm» et de centres de classes «xi», puis on établit un graphique ayant en ordonnées
pour chaque centre de classe «xi» log z avec:
231 L. Mezedjri, A. T. Bensahla and A. Tahar
Log z = log z(xi+h) – log z(xi),

où:
z(xi+h): effectif de la classe de taille de centre de classe (x+h)
z(xi): effectif de la classe de taille de centre de classe xi.
Cette méthode est souvent soumise à deux conditions:
 L’histogramme ne doit pas comporter de classes vides, ces dernières peuvent être évitées
par un lissage ou une moyenne entre les deux classes intermédiaires.
 L’intervalle « h » doit être petit par rapport à chacun des écarts types (h/s ≤ 2,2).
La détermination de l’âge est une étape très importante pour le calcul des paramètres de la
formule de Von Bertalanffy au moyen du logiciel FISHPARM 3.0 (Saïla et al., 1988).
2) Croissance linéaire absolue

D’une manière générale deux modèles sont habituellement employés pour expliquer la croissance des
poissons: Le modèle de Compertz et le modèle de Von Bertalanffy.
Généralement on emploie le modèle de Von Bertalanffy qui est le plus utilisé en ichtyologie et
dont l’emploi est très souple. Il faut signaler que ce modèle décrit mal la croissance dans sa fraction pré
recrutée. Ce modèle s’exprime selon la formule suivante:
L  L 1 e  k t
t 

t0 

avec:
Lt: La longueur totale en (cm) à l’instant t,
L∞: La longueur théorique maximale en (cm) ou la taille asymptotique,
k: Taux de croissance,
t0: Temps théorique où Lt = 0 cm,
Les paramètres de ce modèle sont déterminés en utilisant le logiciel FISHPARM 3.0 (Saïla et
al., 1988).
3) Croissance relative
Le suivi mensuel de l’évolution du poids en fonction de la longueur nous renseigne, d’une part, sur
l’estimation du poids d’un individu à partir de sa longueur totale et, d’autre part, nous permet de suivre
et de décrire les changements de forme et de poids durant la période de croissance.
La relation liant la taille au poids est une fonction cubique de la forme:
W a L ,
b
où:
b: proche de 3,
W: c’est le poids en gramme,
L: la taille en cm,
a: une constante,
b: coefficient d’allométrie.
La valeur du coefficient d’allométrie est comparée statistiquement à 3 au moyen du test t de
Student (Dagnélie, 2006). Ce test est réalisé au moyen du logiciel d’analyse et de traitement des
données Minitab (X, 2003).
Le calcul du coefficient d’allométrie est effectué aussi bien en fonction du temps qu’en fonction
du sexe. Pour obtenir les paramètres (b et a) de l’équation précédente on introduit les couples de
données (W, L) dans le logiciel FISHPARM 3.0 (Saïla et al., 1988).
4) Croissance pondérale absolue

Sachant que le poids est en général une fonction cubique de la taille et que le modèle de Von
Bertalanffy est exprimé de la sorte:
L t  L  1  k t  t 0   ,
 e 
nous avons donc

 
W t  W  1 e  k t  t 0 
b
,
avec:
Wt: c’est le poids total du poisson à l’instant t,
W∞: c’est le poids correspondant à L∞,
b # 3: c’est le taux de croissance pondéral.
Les paramètres k et t0 sont identiques à ceux utilisés dans l’équation de croissance absolue
linéaire.
Tous les paramètres de l’équation de Von Bertalanffy étant connus, nous pouvons calculer le
poids total pour chaque classe d’âge.
5) Coefficient K
Le coefficient de condition k est un indice permettant l’appréciation de l’état pondéral relatif des
individus étudiés. Selon Barnabé (1976), à tailles égales, des poissons ayant le même sexe, peuvent
avoir des différences de poids liées à de divers facteurs tels que l’abondance de nourriture ou la période
de ponte. Il est exprimé par la relation suivante:
W
K  3  100 ,
L
où:
W: le poids éviscéré,
L: la longueur totale.
Dans notre cas, nous utiliserons, d’une part, le poids éviscéré pour palier aux fluctuations liées
aux poids des gonades et, d’autre part, le coefficient d’allométrie mensuel calculé, pour observer les
variations mensuelles de k. on utilisera également le coefficient d’allométrie global calculé pour
chaque sexe séparément, afin d’observer les variations de k en fonction de la classe de tailles.
Résultats
1) Etude de l’âge
Pour la détermination de l’âge nous avons opté pour les méthodes indirectes graphiques de Petersen et
la méthode des différences logarithmiques Battacharya.
Méthode de Petersen
Après établissement de l’histogramme de fréquences nous pouvons émettre les hypothèses suivantes:
Âge 1, 1 an: représenté par le pic au niveau de la classe de taille 11,1cm
Âge 2, 2 ans: représenté par le pic au niveau de la classe de taille 12,5cm
Figure 2: Distribution de taille: Histogramme des fréquences de taille chez Engraulis encrasicolus au niveau
du golfe de Skikda (Avril 2003).
Méthode de Battacharya (figure 3)

L’application de la méthode de Battacharya a permis de tracer 6 droites à pente négative pouvant être
considérées comme des sous populations dans l’échantillon global.
Figure 3: représentation graphique des résultats obtenus par l’application de la méthode de Battacharya
Les résultats de l’application de la méthode de Battacharya sont consignés dans le tableau 2

suivant:
Tableau 2: ésultats de l’application de la méthode de Battacharya
âge 1 2 3 4 5 6
m 10,70 12,60 13,70 14,90 16,00 16,90
θ 70° 54° 54° 82° 83° 80°
s 0,264 0,377 0,377 0,157 0,146 0,179
h/s 0,759 0,531 0,531 1,271 1,373 1,119
m =age moyen / θ = angle de la pente négative
s = écart-type / h/s = rapport intervalle écart-type
Il faut signaler que dans notre travail, les conditions d’application de la méthode de Battacharya
ont été respectées.
2) Croissance linéaire absolue

Les résultats obtenus, après l’application de la méthode de Battacharya, nous permettent de dresser des
couples âges longueurs. L’utilisation de ces couples à l’aide du logiciel FISHPARM 3.0 (Saïla et al.,
1988) a permis de déterminer les paramètres du modèle de Von Bertalanffy suivant:
L t

 22 ,11 1  e ,
 0 ,154 t  3 , 341 
où:
L∞= 22,11
k= 0,154
t0= -3,341
Les valeurs théoriques de la longueur totale pour chaque cohorte ou groupe d’âge ainsi que
l’accroissement moyen annuel sont présentés dans le tableau 3. On peut observer une croissance en
longueur importante par rapport à ce genre de poisson, relativement petit, durant la première année de
vie de ces poissons (plus de 10 cm). Ensuite, la tendance est à un allongement annuel qui diminue
progressivement.
Tableau 3: Valeurs théoriques des longueurs Lt aux âges observés chez Engraulis encrasicolus dans le golfe de
Stora.
Facteurs/Age A1 A2 A3 A4 A5 A6
Lobs 10,70 12,60 13,70 14,90 16,00 16,80
Lth 10,77 12,40 13,78 14,97 15,99 16,86
Accroissement 10,77 1,63 1,38 1,19 1,02 0,87
Lobs = Longueur observée
Lth = Longueur théorique
3) Croissance relative
La collecte des données donne des couples taille-poids au nombre de 880 (individus). L’utilisation du
logiciel FISHPARM 3.0 (Saïla et al., 1988) permet d’écrire l’équation suivante modélisant la relation
taille poids:
W  3 , 220  10 L
3 3 , 225
,
avec
Un coefficient R2 = 95,91%,
a = 3,220 x10-3,
b = 3,225.
Le test t de Student a permis de conclure, que le coefficient d’allométrie (b =3,225) indique une
allométrie majorante.
Le suivi mensuel des relations tailles poids nous permet de remarquer une forte corrélation.
Globalement la croissance relative est généralement majorante à l’exception des mois de janvier et
mars où la croissance relative est dans un cas d’isométrie.
L’expression mathématique de la relation taille-poids est calculée séparément pour les mâles et
les femelles et on peut écrire les formules suivantes:
Mâles → W  4,122 10 L 3 ,128 Allométrie majorante.
3
Femelles → W  3,288 10

3 3 , 218
L Allométrie majorante.
Chez les mâles on est en présence d’une isométrie le long de l’année, sauf pour les mois d’août
(b = 3,602) et septembre (b = 3,271) où nous avons le cas d’une croissance relative majorante. Chez les
femelles la croissance est majorante dans 6 mois sur 10. Les mois où on a une isométrie, sont ceux de
septembre (b = 3,301), janvier (b = 3,194), mars (b = 2,722) et le mois de juin (b = 3,661).
4) Croissance pondérale absolue

Après avoir déterminé la relation taille-poids, et connaissant la relation qui détermine la croissance
linéaire absolue, donc tous les paramètres du modèle de Von Bertalanffy, on a alors obtenu le modèle
de croissance pondérale absolue qui suit:
W t

 69,841  e 
0,154t 3,341 3, 225
,
Le tableau suivant donne les couples âges-poids selon le modèle défini précédemment:
Tableau 4: Couples âge-poids selon le modèle de la croissance Pondérale absolue dans le golfe de Stora.
Facteurs/Age A1 A2 A3 A4 A5 A6
Wth 6,89 10,81 15,21 19,86 24,56 29,16
Accroissement 6,89 3,92 4,40 4,65 4,70 4,60
Wth = poids théorique
On constate qu’à un an, nous avons des poissons d’un poids très faible de l’ordre de 6,89 g,
alors que les adultes peuvent avoir trois fois plus du poids moyen (29,16 g). Le gain de poids se fait
progressivement et atteint 4,60 g entre 5 ans et 6 ans.
5) Coefficient de condition k
Les variations de l’indice pondéral chez Engraulis encrasicolus (figure 4) sont assez marquées. Nous
pouvons observer un maximum observé en mars (k = 0,77). Le minimum est observé en été (juillet k =
0,08).
Figure 4: Variations mensuelles de k
0 ,8
0 ,7
0 ,6
0 ,5
vale urs de k
0 ,4
0 ,3
0 ,2
0 ,1
0 ,0
3 3 03 03 3 03 4 4 04 4 4 4
-0 -0 e- e- e-
0
e- r-0 r-0 s- -0 -0 -0
et ût br br br br ie ri e ar ri l ai in
ill ao to nv av m ju Mois
ju em oc e m em ja fé v m
pt ov dé
c
se n
Le coefficient k chez les mâles présente une valeur moyenne (k = 0,38) légèrement supérieure à
la valeur moyenne enregistrée chez les femelles (k = 0,30).
Les variations de k en fonction de la classe de taille (figure 5) chez les mâles et les femelles, ne
présentent aucune allure bien distincte. L’allure est d’aspect presque rectiligne.
Figure 5: Variations de k en fonction du sexe et des classes de tailles
0,42
Variable
k m
0,40 k f
0,38
Valeurs de k
0,36
0,34
0,32
0,30
[ [ [ [ [ [ [ [ [ [ [ [ [ [ [
10 .5 11 .5 12 .5 13 .5 14 .5 15 .5 16 .5 17
.5 10 .5 11 .5 12 .5 13 .5 14 .5 15 .5 16 .5
[9 [1
0
[1
0
[1
1
[1
1
[1
2
[1
2
[1
3
[1
3
[1
4
[1
4
[1
5
[1
5
[1
6
[1
6
Clas s es de taille
Discussion
Dans notre cas l’étude de l’âge est basée sur les méthodes indirectes, à cause de la difficulté rencontrée
lors des essais de lectures sur les écailles (difficultés de collecte) et sur les otolithes (difficultés de
maîtriser la méthode). La méthode de Petersen nous permet de décomposer notre population d’anchois
européen Engraulis encrasicolus des côtes est algériennes, échantillonnée à un instant (t) fixe, en 6
groupes ou cohortes. La méthode de Battacharya nous donne les mêmes résultats avec des longueurs
moyennes relativement proches. Dans la méthode de Battacharya, les valeurs moyennes observées pour
chaque cohorte sont relativement proches et semblables aux valeurs calculées, ce qui montre que
l’équation de Von Bertalanffy, modélise bien la croissance linéaire absolue de notre poisson. Avec une
taille, vers laquelle tend notre poisson L∞= 22,11 cm supérieur à la taille Lmax= 17,1 cm. Le tableau 13
suivant montre les paramètres de l’équation de Von Bertalanffy obtenus par certains auteurs dans
différentes régions de la méditerranée.
Tableau 5: Paramètres du modèle de Von Bertalanffy obtenus dans certaines régions.
L∞ (cm) K t0 Pays Localité Auteur

11,70 1,160 -0,17 Russie Mer d'Azov Pauly (1978)
13,90 1,730 \ Russie Mer noire Pauly (1978)
15,30 2,444 \ Italie Mer Adriatique Pauly (1978)
16,80 0,324 -1,99 Turquie Mer noire 1986-1987 Karacam et Düzgunes (1990)
18,60 0,260 \ Algérie Baie d'Alger Djabali et al. (1993)
20,00 0,420 -1,15 France Golfe du Lion Pauly (1978)
22,11 0,154 -3,34 Algérie Golfe Skikda Mezedjri (2004)
Le modèle qui représente l’évolution du poids en fonction de la longueur totale indique que le
poids de l’anchois européen Engraulis encrasicolus des côtes est algériennes (golfe de Skikda) évolue
plus vite que la taille. Une allométrie majorante est conclue suite à un coefficient calculé (b = 3,225)
qui est significativement supérieur à 3. Ce cas d’allométrie majorante est observé aussi bien chez les
mâles que chez les femelles.
Tableau 6: Paramètres de la relation taille-poids obtenus dans certaines régions de la méditerranée.
a b Sexe n Pays Auteur

0,0080 2,86 \ 156 Egypte Aabdallah (2002)
0,0070 2,95 \ \ Grèce Stergiou et Moutopoulos (2001)
0,0065 2,98 \ 2219 France Dorel (1986)
0,0040 3,00 \ \ Algérie Djabali et al. (1993)
0,0065 3,02 \ \ France Campillo (1992)
0,0028 3,33 femelle 524 Turquie Karacam et Düzgunes (1990)
0,0021 3,45 Mâle 610 Turquie Karacam et Düzgunes (1990)
0,0032 3,225 \ 880 Algérie Mezedjri (2004)
0,0041 3,128 Mâle 316 Algérie Mezedjri (2004)
0,0033 3,218 femelle 487 Algérie Mezedjri (2004)
L’équation de la croissance pondérale absolue obtenue, montre que l’évolution du poids en

fonction de l’âge de l’anchois européen Engraulis encrasicolus des côtes Est algériennes (golfe de
Skikda), est de l’ordre de 6,89 g à 1 an et atteint 29,16 g à l’âge de 6ans, avec un gain maximum en
poids de l’ordre de 4,70 g entre l’âge 4 et l’âge 5. Le poids asymptotique P∞ = 69,84 g est supérieur à
Pmax = 29,16 g.
Les variations saisonnières du coefficient de condition (k) montre que chez l’anchois européen
Engraulis encrasicolus des côtes est algériennes, les valeurs les plus faibles sont durant les mois les
plus chauds et les valeurs les plus élevées durant l’hiver et au début du printemps. Ceci montre que
notre poisson rend disponible ses réserves énergétiques pendant la période de reproduction et il les
emmagasine durant le repos sexuel. Djabali et Hamida (1989), aboutissent aux mêmes conclusions
dans leur étude sur l’anchois dans la baie d’Alger durant la période 1983 à1984.
Conclusion
L’application des méthodes indirectes pour l’étude de l’âge de l’anchois européen, montre l’existence
de 6 sous populations correspondantes aux âges de 1 à 6 ans. L’anchois européen des côtes Est
Algériennes, croit d’environ 10,77 cm la première année. Cependant, le taux de croissance diminue
pour atteindre 0,87 cm au cours de la sixième année.
Par ailleurs, le coefficient d’allométrie résultant de la relation taille-poids, est statistiquement
supérieur à 3 ce qui montre l’existence d’une allométrie majorante entre les deux paramètres. Le
modèle de Von Bertalanffy décrit parfaitement la croissance linéaire de notre poisson. Du point de vue
pondéral notre poisson pèse 6,89 g à un an et atteint 29,16 g à six ans.
References
[1] Abdallah M., 2002. Length-weight relationship of fishes caught by trawl off Alexandria, Egypt.
Naga ICLARM Q. 25(1):19-20.
[2] Barnabé G., 1976. Contribution à la connaissance de la biologie du loup Dicentrarchus labrax
(L.) (Poisson Séraranidé). Thèse d’état, Univ. Sc. Tech. Langueduc, Montpellier. 426p.
[3] Campillo A., 1992. Les pêcheries françaises de Méditerranée: synthèse des connaissances.
Institut Français de Recherche pour l'Exploitation de la Mer, France. 206 p.
[4] Dagnélie P., 2006. Statistique théorique et appliquée. Volume 2: Inférences à une et à deux
dimensions. Bruxelles-université DE BOECK et LARCIER – 659p.
[5] Djabali F., & Hamida F., 1989. Pelagos. Bulletin de l’institut des sciences de la mer et de
l’aménagement du littoral ISMAL. Volume VII fasicule 1(1989): 11-26.
[6] Djabali F., Mehailia A., Koudil M., & Brahmi B., 1993. Empirical equations for the estimation
of natural mortality in Mediterranean teleosts. Naga. ICLARM Q. 16(1): 35-37.
[7] Dorel D., 1986. Poissons de l'Atlantique nord-est relations taille-poids. Institut Français de
Recherche pour l'Exploitation de la Mer.
[8] Karacam H., & Düzgünes E., 1990. Age, growth and meat yield of the European anchovy
(Engraulis encrasicolus, L. 1758) in the Black Sea. Fish. Res. 9(2):181-186.
[9] Mezedjri L., 2004. Biométrie comparée de l’anchois européen Engraulis encrasicolus (Linné,
1758). Thèse de Magistère. Département des sciences de la mer, faculté des sciences, université
Badji mokhtar Annaba. Algérie.
[10] Pauly D., 1978. A preliminary compilation of fish length growth parameters. Ber. Inst.
Meereskd. Christian-Albrechts-Univ. Kiel (55):1-200.
[11] Saïla S.B., Recksiek C.W., & Parger M.H., 1988. Basic fisheries biology Programs. Elsevier
science publisher B.V.
[12] Stergiou K.I., & Moutopoulos D.K., 2001. A review of length-weight relationships of fishes
from Greek marine waters. Naga ICLARM Q. 24(1&2):23-39.
[13] X., 2003. MINITAB version 14.1 for windows.
ISSN 1450-216X Vol.21 No.2 (2008), pp.239-248
Efficient Fault Tolerant Mapping of Large Three-Dimensional

Simulation onto 3D Tori Graph
Saleh H. Al-Sharaeh
Department of Computer Science, King Abuallah II School for Information Technology
Jordan University, Amman 11942n, Jordan
E-mail: ssharaeh@ju.edu.jo
Tel: +96225355000 ext.: 22576
Abstract
In this paper we present an efficient 3D large volume simulation decomposition and

mapping technique onto a 3D torus interconnection network, based on base-b reflected gray
codes with fault tolerant capability. The algorithm was tested on real 3D space plasma
simulation of the Aurora region of the Earth’s Ionosphere, when the fault occurs, the
proposed algorithm preserve the Hamiltonian mapping (i.e. ring topology) onto 3D
interconnection network due to either node or link failure. Hamiltonian mapping is used to
minimize the communication among processors. We also studied it is impact on the speed
as well as the efficiency. The new algorithm, showed a small degradation in the speedup as
well as the efficiency.
Keywords: Parallel Simulation; Computational Modeling; Plasma Physics; Hamiltonian

Cycle; Gray codes
1. Introduction
Plasma simulation is considered a complex scientific application because of the high density of
particles (rang from 1 to 1018 particles per cm3) that exist in the simulation domain. One method to
simulate the plasma is Particle-in-Cell code technique (PIC), where the plasma behavior acts as a
collection of particles which move freely in space in the manner dictated by the interaction forces and
external imposed forces [1-2].
We consider a PIC algorithm to simulate the movement of particles in electromagnetic field.
Plasma simulation of the Earth's Ionosphere is considered High Performance Applications (HPC),
because they are the most demanding of computing power. Due to perceived difficulties with
programming in a multi-computing environment, interconnection topologies play a vital role [3-7]. The
Massively Parallel Processing (MPP) Cray T3D/T3E architecture employs a 3D torus interconnection
topology. Each Processing Element (PE) has local memory. Such memory is physically distributed, but
it is logically shared because each PE can access any other PE memory without involving the
microprocessor in that PE [7-8]. Cray Research Inc. introduced the first phase which has 256
processors followed by the T3E with 1024 processors, and recently Cray XD1[8] a shared memory
system. In this new generation of MPP based on 3D torus interconnection topologies, each node has
16M bytes that are physically distributed, but are logically shared. When a processor needs data at a
certain address, which exists in another processor, that processor number combined with the address
requested will formulate the effective address, and then a very fast search engine handles the
Efficient Fault Tolerant Mapping of Large Three-Dimensional Simulation onto 3D Tori Graph 240
communication without interrupting the computation processor. Latency can be effectively hidden in
this manner by transparently overlapping communication with computation, which makes this new
generation very effective and scalable.
The aforementioned characteristics of the T3D make this kind of architecture suitable to our
application, and with a proper mapping methodology, the algorithm developed was ported to a 3-D
torus (the Cray T3D interconnection topology). The method is based on base-b reflected Gray code to
minimized the inter-processor communication delay and reducing the next node address software
overhead.
2. Task assignment, mapping, and scheduling

The targeted hardware, Cray T3D, is made up of a set of processing nodes that are interconnected by
some form of interconnection topology, in this case 3D Torus. In a homogenous environment, the
processing elements all have the same computing capability, and more often than not, the network is
regular in structure.
In general, the act of allocating the task graph to the target topology can be described in terms
of three basic operations: assignment, scheduling, and mapping [9]. The assignment operation involves
the clustering or partitioning of tasks into processing groups in a manner that ensures a one-to-one
correspondence between the number of groups and the number of processing elements. Scheduling is
the process of determining the order in which the tasks (and possibly communications) are to execute
within each processing group. The mapping problem involves assigning processing groups to
individual processing elements for execution, in a way that match the communication structure present
between groups with the topology of the inter-processor communication network. Each of these three
operations is NP complete and can occur in a static (preprocessing) or dynamic (run time)
environment. For many algorithms, optimal allocations can be made statistically by carefully
examining the structure of the algorithm. One technique that has been used successfully for such
problems employs base-b reflected Gray codes [9-11].
3. Base-b reflected gray codes

A well-known strategy for mapping an algorithm efficiently onto a binary hypercube is based on the
binary reflected Gray codes. A class of generalized Gray codes is called base-b reflected Gray codes.
The generalized reflected Gray codes are then used to port the restructured 3-D PIC code (a method to
simulate space plasma particles) task graph developed in this work, onto a Cray T3D multi-computer
system, which has 3D torus interconnection topology.
To provide the necessary background, we will briefly describe the reflected base-b Gray Codes.
(b)
Let G (n) denote the set of all n-digit code words of the base-b reflected Gray code [8]. It is
assumed that b=2m, m≥1, N=2km, and n=km. When m=1 (that is, b=2), it is called a binary reflected
( 2)
Gray code. The definition of G (n) is as follows:
G
(2)
( n )  {G
(2)
0
(2)
( n ), G 1 ( n ),..., G
(2)
2 n 1
( n )} (1)
(2)
is called the encoding of integer i for i=0,..,2n-1, and G i (n)  2
(2) n
where G i (n)
(2)
The encoding process involved in the definition of G (n ) is as follows. Let i  <N>, the binary
(2)
representation of i is i  in i n 1i n 2 ......i1i0 (note that ). in  0 Let Gi (n)  g n g n 1 g n  2 ...g1 and E(2) be the
corresponding bijection function (one-to-one and onto function), that is,
(2)
E : N  G ( n )
(2)
(2)
where gj  ij ij1,j j {1,2,..,n}
241 Saleh H. Al-Sharaeh
The successive code words differ in exactly one bit by what is called a Hamming distance of
one. Another definition of base-b reflected Gray codes and its relationship to Hamiltonian cycle
embedding onto the 3-D torus will be presented later.
4. The 3D PIC Model

The following represents a basic overview of the PIC method as applied to such plasma problems:
4.1. Initialization
The three-dimensional geometric space associated with the simulated special domain is subdivided into
a number of equal size cubic cells (x y z ) appropriately large. Each cell is then initialized with an
equal number of particles whose initial velocity components (vx,vy,vz) and position (x,y,z) are chosen
randomly from a Maxwellian and Uniform distributions, respectively. As shown in Figure 1, the
simulated volume is of size Lx Ly Lz and total number of cells are Nx Ny Nz. The particles
represent electrons and ions of the plasma.
4.2. Partial Charge Scatter Operation

Each grid point is then assigned a partial charge, based upon the locations of the particles which reside
in the eight adjacent cubic cells.[1].
4.3. Gather Operation

4.3.1. Grid Point Potential Calculation
The electric potential of each grid point is calculated by solving the Poisson equation:
 2 ( x , y , z )    ( x , y , z )  (3)
i j k i j k 0
where  is the electric potential,  is charge density, and 0 is the permittivity of the space.
4.3.2. Electric Field Calculation


The electric field E is determined by solving:

E    ( x i , y j , z k ) (4)
4.4. Particle Advancement Operation

The equations of motion for the particles are then solved using the leap-frog scheme in a uniform

magnetic field, B , to evaluate the new particle position and velocity in the x, y, and z dimension:
 
d v  ( q / m )( E  v  B )
(5)
dt e

d r  v (6)
dt 
Where m and q are the particle mass and charge, and v is the velocity (vx,vy,vz).
The PIC code technique is an efficient, simple method of solving a wide variety of complex
scientific problems such as the nonlinear evolution of the lower hybrid waves in the Earth’s ionosphere
and charge interaction with space vehicles, which involving a large number of particles moving under
the action of internal and external forces. The defining characteristic of PIC simulations is the method
of calculating the force on each particle. By dividing the simulation box into cubic cells in which the
particles fall, we are able to use the resulting grid in the solution of a field equation from which the
force on each particle can then be determined; a periodic spatial domain is assumed. The calculated
forces on each particle determine the new position of the particle. This is much more efficient than
summing N3 coulomb interactions among N particles [1,7,18-19].
5. Cray T3D Interconnection Network

In the Cray T3D system the interconnection network provides paths among the processing elements
and the I/O gateways through bi-directional communication links. It forms a three-dimensional matrix
of paths that connect the nodes in the X, Y, and Z dimensions. Each communication link connects two
nodes in one dimension. In a torus interconnection topology the communication links connect the
smallest numbered node in a dimension directly to the largest numbered node in the same dimension.
5.1. Dimension order routing

Dimension order routing is used by the Cray T3D to determine the path a message takes after it leaves
the sender node. The message travels in the x dimension first, then in the y dimension, and finally in
the z dimension, until it reaches its destination node [8]. In the Cray T3D system each node uses a
look-up table to obtain the routing tag. For example, with a message from a node at (1,3,4) to a node at
(3,2,5) one possible routing tag will be X=+2, Y=-1, and Z=+1. So the message travels two hops
along the positive x direction, then 1 hop along the negative direction of y, and finally one hop along
the positive z direction.
5.2. 3-D Torus Partitioning

The users of the MPP Cray T3D system have the option to use two kinds of partitioning supported by
the host, which is the operating controlled or user controlled. Since our objective to fully utilize the
interconnection characteristics, the second approach will be used. Rather than let the operating system
converts the virtual PE numbers used by the application into logical PE numbers, an algorithm is
developed to get the physical ID of the PE, so that to minimize the number of hops a message traverse
In a hardware partition, when the application transfers data between PEs, the operating system
is not involved in the transfer. Hardware in each PE contains two parts: the virtual node number and
the PE bit. The virtual node number ranges from 1 to 10 bits and indicates in which hardware partition
the PE resides. The PE bit indicates whether the PE is PE0 or PE1 in the node.
The virtual node number has 0 to 3 bits assigned to the x dimension, 0 to 4 bits assigned to the
y dimension, and 0 to 3 bits assigned to the z dimension. By assigning bits of the virtual node number
to the appropriate dimensions, software arranges the virtual nodes into one of several shapes [8-9]. To
determine the kind of partitioning, developer could use a system call. In our application we decided to
use the hardware partitioning, to minimize the software overhead and to reshape the interconnection
topology such that, the number of hops traveled by the messages is at a minimum.
6. Algorithm mapping onto ring topology

In the following sections we will discuss the two cases. The first case when we have a fault-free
mapping. That is the Hamiltonian cycle length will be equal to the number of nodes. The number of
nodes will be equal to 2 d 1  2 d 2  2 d 3 , where d1, d2, and d3, are number of bits in the x, y, and z-
dimension respectively. The second case when we have at least one of the node either not reachable or
it is down. The Hamiltonian cycle length will be equal to (2 d 1  2 d 2  2 d 3 )  n f , where n f is the
number of nodes at fault.
6. 1. Algorithm mapping onto fault free ring topology

In the Cray T3D interconnection topology, the user can overwrite the software partitioning, as
mentioned earlier. The process of mapping a cluster of tasks (slabs) onto a set of processors connected
via a 3-D torus, the Cray T3D interconnection topology graph, should minimize the number of hops
that a message must traverse to reach its destination. In the proposed algorithm each processor
communicates only with two adjacent processors. As we demonstrated earlier, the communication time
of a message is directly proportional to the number of hops a message traverses. Thus, the proper
topology is a ring topology. For that, the adjacent slabs should be assigned to adjacent processors.
The mapping function was designed to assign consecutive slab numbers to consecutive
Hamiltonian cycle vertices. This raises the question of how to obtain an algorithm to automatically find
the Hamiltonian cycle of the 3-D torus. Next we will introduce this algorithm, which is based on, base
b reflected Gray Codes.
Using indirect labeling to the slabs, such that
(2)
 : V ( H )  G (n) (7)
where  is one-to-one and onto function, such that, each successive slabs are at addresses that differ in
( 2)
only one bit. Let H be the encoding graph represented by G (n) =(Ve,Ee), where
Ve  (ed 1 ed 1-1 ed 1- 2 ..e0 , ed 2 ed 2-1 ed 2- 2 ..e0 , ed 3 ed 3-1 ed 3- 2 ..e0 ) that is defined in Section 3. It is clear
that H is a cycle where V ( H )  2 n Let T (t1 , t 2 , t 3 ) represent the interconnection topology, where t1, t2,
t3 are the processor (vertex) coordinates in the x, y, and z dimensions respectively. The number of
processors is powers of two, i.e., t1  2 d 1 , t 2  2 d 2 , t 3  2 d 3 . Thus, our proposed mapping function of
the binary hypercube Hamiltonian cycle onto a 3-D torus interconnection network is given by
f : V ( H )  V (T ) , (8)
where V(H) represent the vertices of the cycle (each vertex represents a chunk of data, slab), and each
vertex of T represents a processing element, where
 t , e e , l  0,......., d  1 
 1
(t , t , t )   l l  1 1 (9)
1 2 3  j , e e ,l  d  1,.., d j 
 l l  1 j  1 
where  is the XOR function. An example of a Hamiltonian cycle in a 3-D torus of size (2,4,2) is
given in Table 1 and Fig.1. Clearly, each slab is assigned to a processor whose coordinates are
calculated by Eq. 9.
Table 1: Mapping 4-Cube onto 3-D Torus of Size (2,4,2).
i Vg G(2)(4) T(t3 t2, t1) Vt

0 0000 0000 0,00,0 0,0,0
1 0001 0001 0,00,1 1,0,0
2 0010 0011 0,01,1 1,1,0
3 0011 0010 0,01,0 0,1,0
4 0100 0110 0,10,0 0,2,0
5 0101 0111 0,10,1 1,2,0
6 0110 0101 0,11,1 1,3,0
7 0111 0100 0,11,0 0,3,0
8 1000 1100 1,11,0 0,3,1
9 1001 1101 1,11,1 1,3,1
10 1010 1111 1,10,1 1,2,1
11 1011 1110 1,10,0 0,2,1
12 1100 1010 1,01,0 0,1,1
13 1101 1011 1,01,1 1,1,1
14 1110 1001 1,00,1 1,0,1
15 1111 1000 1,00,0 0,0,1
Figure 1: The Cray T3D Partition of size (2,4,2).
031 131
030 130
021 121
020 120
011 111
010 110
001 101 Wraparound

Communication
Link
001 100
Processor Elements with
its Coordinates
6. 2. Algorithm mapping onto faulty node(s)

Due to the existence of the probability of either nodes failure as well as line failure between nodes raise
the demand of a fault tolerant technique. In the following, we present an algorithm to preserve the
Hamiltonian mapping (i.e. ring topology) onto 3D interconnection network due to either node or link
failure. Say, v(t1 , t 2 , t 3 ) is the old destination node, then we have to calculate the new destination node
coordinate v ' (t1 , t 2 , t 3 ) . The algorithm that handles the re-mapping and at the same time still we have a
Hamiltonian cycle except those nodes at fault is as follows.
Fault Tolerant Algorithm
Case 1 : Fault along the x - axis

  d
 t 1
, (t  1) 
 1) mod( 2
 1 1 
'
v ( t1 , t 2 , t 3 )   t ,e  e ,l  d  1,......., d 
 2 l l 1 j 1 j
t ,e  e ,l  d  1,......., d 
3 l l 1 j 1 j 
 
Case 2 : Fault along the y - axis
t ,e  e , l  0 ,1,..., d  1 
 1 l l 1 1 
 d 
2
v ' ( t1 , t 2 , t 3 )   t , ( t  1) mod( 2  1) 
 2 2 
t3 ,e  e ,l  d  1,......., d 
l l 1 j 1 j
 
Case 3 : Fault along the z - axis
 
 
t ,e  e , l  0 ,1,..., d  1 
 1 l l 1 1 
v ' ( t1 , t 2 , t 3 )   t , (e  e ,l  d  1,......., d 
 2 l l  1 j  1 j 
 d 
t 3
, ( t  1) mod( 2  1) 
 3 3 
where v ' ( t1 , t 2 , t 3 ) is the new destinatio n node coordinate .
It is evident that when a node fail according to case 1, case 2, and case 3, will be an extra
overhead of (2 d 1 ) , (2 d 2 ) , and (2 d 3 ) respectively.
6.3. Example
Here we show an example of (4,2,4) 3D torus interconnection network. According to Eq.9, the
Hamiltonian cycle will be: (0,0,0), (1,0,0), (2,0,0), (2,1,0), (1,1,0), (0,1,0), (0,1,1), (1,1,1), (2,1,1),
(2,1,2), (1,1,2), (0,1,2), (0,0,0). Applying the fault tolerant algorithm, assuming a fault at node with
coordinate of (1,0,0) (i.e. case 1 a fault along the x-dimension). Then the new Hamiltonian cycle will
be, (0,0,0),(2,0,0), (2,1,0),(1,1,0),(0,1,0),(0,1,1),(1,1,1),(2,1,1),(2,1,2),(1,1,2),(0,1,2),(0,0,0). Fig. 2
illustrates this scenario.
Figure 2: A (4,2,4) 3D Torus with fault at node (1,0,0)
7. Algorithm Performance
7.1. Fault Free Algorithm
In Table 2, we show the results of running the 3-D PIC model on the Cray T3D (Fault-Free Scenario).
In this experiment, each processor has a slab size of 33 33 9 grid points, and 8 particles per cubic
cell. The total number of particles in the simulation domain is 131072 per processor. Fig. 3 shows both
the speedup and the parallel execution time of the simulation versus the number of processors [18]. As
predicted by the scalability analysis [18], the efficiency is near constant, which reflects the optimum
mapping of the simulated domain onto the 3D torus topology. Furthermore, the execution time has a
slight increase due to the increased size of the communication buffer, which hold the particles that are
needed to be communicated among the adjacent processors, and is considered almost constant. The
speedup is also near the ideal speedup. Because of the utilization of the method proposed in this
research, packet collisions in the routing network were brought to a minimum.
Table 2: Geometric Scaling of the 3-D PIC Parallel Simulation.
Number of Processors Sequential Time (Sec.) Parallel Time (Sec.) Speedup Efficiency
2 399.97 199.99 1.999 0.999
4 800.00 200.00 3.999 0.999
8 1603.13 200.39 7.999 0.999
16 3229.97 201.87 15.998 0.999
32 6479.12 202.48 31.990 0.999
64 13281.55 207.53 63.975 0.999
128 28167.00 220.06 127.906 0.999
256 63671.61 248.74 251.722 0.984
Figure 3: Geometric Space Scaling Speedup and the Efficiency Vs. Number of Processors (Fault-Free
Scenario).
1.20
1.15 Actual Efficiency
1.10 Ideal Efficiency
Efficiency
1.05
1.00
0.95
0.90
0.85
0.80
250
Actual Speedup
200
Speedup
Ideal Speedup
150
100
50
0
0 50 100 150 200 250
Number of Processors
7.2 Faults-Tolerant Algorithm

For the case when we have a fault either at the node or the communication link the proposed algorithm
was implemented. It worth mentioning here, that we have to implant a load balancing technique. The
technique is simple, in the sense that we have re-size the slab size to fit the new number of working
nodes. We impose a single node error, then we measure the speedup and efficiency. The results of our
simulation are depicted in Fig.4. As shown in the figure, both Speedup and Efficiency, as expected,
showed degradation as compared with fault free configuration. This is due to the re-routing mechanism
that we used inside the simulation as well as we have to increase the size of the slabs to fit the new
configuration after the occurrence of the single error (load balancing).
Figure 4: Geometric Space Scaling Speedup and the Efficiency Vs. Number of Processors (Single Fault Node
Scenario).
1
Effieciency
0.5
……. Ideal efficiency
_____ Measured Efficiency
0
50 100 150 200 250
300
…… Ideal Speedup
200 ____ Actual Speedup
Speedup
100
0
0 50 100 150 200 250 300
8. Conclusion
In our methodology, proper restructuring and mapping of the 3-D PIC code was developed, in which
the workload is characterized in terms of the number of particles and grid points and mapped to a
proper interconnection topology ring. The problem is fully parallelized unlike previous efforts on this
problem, where the fixed-time model is used to calculate the parallel speedup and parallel efficiency.
Performance metrics such as speedup and efficiency are two important measures that characterized the
performance of an algorithm. Because of the routing path that follows a Hamiltonian cycle, it reduces
the communication significantly between nodes. Furthermore, it reduces the probability of messages
collisions. For the case when we impose a fault node, an extra number of hops needs to be traveled as
well as we have to redistribute the work load to achieve load balancing. This impacts the speedup as
well as the efficiency as shown in Fig.4. The degradation for the fault case was not significant this due
to proposed fault tolerant technique.
References
[1] C. K. Birdsall, A. B. Langdon, Plasma Physics via Computer Simulation, New York: Adam
Hilger, 1991.
[2] R. W. Hokney, J. W. Easetwood, Computer Simulation Using Particles, New York: Adam
Hilger, 1988.
[3] George Papanicolaou, Knut Solna, Chrysoula Tsogka, Hongkai Zhao, Kai Huang, Efficient
numerical simulation for long range wave propagation, Journal of Computational Physics 215
(2006) 448–464.
[4] Huaxi Gu, Jie Zhang, Kun Wang, Zengji Liu, Guochang Kang, Enhanced fault tolerant routing
algorithms using a concept of ‘‘balanced ring’’, Journal of Systems Architecture, 53 (2007)
pp.902–912
[5] Yu Zhuang, A parallel and efficient algorithm for multicompartment neuronal modeling,
Neurocomputing 69 (2006) 1035–1038.
[6] R. Hatzky, Domain cloning for a particle-in-cell (PIC) code on a cluster of symmetric-
multiprocessor (SMP) computers In Press Parallel Computing xxx (2006) xxx–
xxx,doi:10.1016/j.parco.2006.03.001.
[7] Kwang Jin Oh, Michael L. Klein ,A general purpose parallel molecular dynamics simulation
program, Computer Physics Communications 174 (2006) 560–568.
[8] Cray Research Homepage, http:// www.cray.com.
[9] Cray Research Inc., Closing the gap between peak and achievable performance in high
performance computing, White Paper WP-0020404.
http://www.cray.com/downloads/whitepaper_closing_the_gap.pdf
[10] S. Lakshmivarahan, S. Dhall, Analysis and Design of Parallel Algorithms: Arithmetic and
Matrix Problems, New York, McGraw-Hill, 1990
[11] J. Conway, N. Sloane, and A. Wilks, Gray codes and reflection groups, Graphs and
combinatorics 5(1989)315-325
[12] S. Bokhari, Assignment Problems in Parallel and Distributed Computing, Boston, Kluwer
Academic Pub., 1987.
[13] nCube Corp., nCUBE 2 Programmer’s Guide, r2.0, 1990.
[14] J. L. Gustafson, The Consequences of Fixed-Time Performance Measurement, Proc. 25th
Hawaii Intr. Conf. System Sciences, Vol. III, 1992, pp. 113-124.
[15] J. L. Gustafson, Reevaluating Amdahl’s Law, Comm. ACM, Vol. 31, No. 5, 1988, pp. 532-533.
[16] Sun, X.-H.; Chen, Y.; Wu, M. ,Scalability of Heterogeneous Computing; ICPP 2005,
International Conference on Parallel Processing 14-17 June 2005 Page(s):557 – 564.
[17] Grama, A., Gupta, A., Kumar, V., Isoefficiency: Measuring the Scalability of Parallel
Algorithms and Architectures, IEEE Parallel & Distributed Technology, August 1993, pp. 12-
21.
[18] S. Hosni Al-Sharaeh, B. Earl Wells, Nagendra Singh, A Massively Parallel Particle-In-Cell
Technique for Three-Dimensional Simulation of Plasma Phenomena, Proceedings of the 9th
International Conference on Parallel and Distributed Computing Systems (PDCS96),
Dijon,France, Sep. 25-27, 1996.
[19] Saleh H. Al-Sharaeh, On the Hamiltonian cycle mapping onto 3-D torus interconnection
network based on base-b reflected gray codes, ,Applied Mathematics and Computation,
Volume 186, Issue 2, 15 March 2007, pp. 1311-1321
ISSN 1450-216X Vol.21 No.2 (2008), pp.249-358
Efficient Method for Assigning Students to Proper Groups
Moh'd Belal Al- Zoubi

Department of Computer Information Systems,University of Jordan, Amman, Jordan
E-mail: mba@ju.edu.jo
Tel: +962-6-5355000; Fax: +962-6-5350704
Imad Salah
E-mail: isalah@ju.edu.jo
Tel: +962-6-5355000; Fax: +962-6-5350704
Azzam Sleit
E-mail: azzam.sleit@ju.edu.jo
Tel: +962-6-5355000; Fax: +962-6-5350704
Saleh Al-Sharaeh
E-mail: ssharaeh@ju.edu.jo
Tel: +962-6-5355000; Fax: +962-6-5350704
Ammar Huneiti
E-mail: a.huneiti@ju.edu.jo
Tel: +962-6-5355000; Fax: +962-6-5350704
Nadim Obeed
E-mail: obein@ju.edu.jo
Tel: +962-6-5355000; Fax: +962-6-5350704
Abstract
The problem of partitioning a large number of entities into a fixed number of

groups on the basis of a set of attributes associated with each entity has required an
intensive research work in the literature. We have investigated a previous research related
to this problem and proposed a method that is based on fuzzy clustering techniques. Jordan
University students who are enrolled in the course of Computer Skills 2 were taken as an
example to test the new method. Successful grouping will help academic community group
students based on their qualifications. Distributing students in groups with
comparable/similar backgrounds and skills will help instructors use proper education
methods based on their qualifications and skills. The proposed method was tested on three
real data sets, and good results were achieved.
Efficient Method for Assigning Students to Proper Groups 250
Keywords: Fuzzy Clustering, Silhouette Coefficients, Assigning Students to Groups.
1. Introduction
Computer Skills 2 (CS2) is a compulsory course at all universities in Jordan. It serves as a foundation
for Information Technology (IT) literacy, promoted by the information technology faculties and
schools in the Jordanian universities. Because of the diverse interests, different disciplines and widely
varying IT experience among students, teaching these courses has proved to be challenging. The
problem is more intense for non-technical students (e.g., students who are enrolled in humanities and
other non-technical faculties). This is due to a lack of integration of computer skills within their study
schemes.
We believe that the problem is bound to occur, possibly with less intensity, in any other
institute. Multi-disciplinarily is a fact of life and is reflected (or rather emphasized) in our education
systems a few years before a student leaves school and in universities as well. The need for
interdisciplinary is rapidly increasing because many complex problems cannot be solved using the
expertise/knowledge of single disciplines. Solutions to many complicated scientific and societal issues
require interdisciplinary efforts and capabilities. Interdisciplinary studies involve generally various
theories, methodologies and knowledge from many different disciplines. Such a practice seems to be
the only way to deal with complex problems (Daily, 1999; Kostoff, 2002; Naiman, 1999).
Whether the interdisciplinary work is only cooperative to solve interdisciplinary problems
(Sharma, 1997; Collins, 2002.), or integrative to build a new whole (Kostoff, 2002), communication is
always an essential issue. It involves the ability to fully understand the content of what is being
negotiated from the perspective of the discipline that has proposed it (Biocca, 2002). Furthermore,
education is the activity of communicating insights, assessments, experiences and/or skills. However, it
is important to emphasize that what is important in education is the extent to which the
students/receivers acquire potentially useful knowledge and utilize this knowledge in their own
disciplines. This explains the reason why students taking CS2 have to acquire the content knowledge
and skills as judged necessary by IT experts.
Few researchers would disagree that some of the major problems that face interdisciplinary
communications are:
(D1) Experts are trained to be specialists in their disciplines. Hence, their ability to clarify a
message or to find a weakness in an argument outside their domain of expertise or discipline is rather
limited.
(D2) It is difficult to find a common language because of the discipline specialization (Biocca,
2002). Hence, interdisciplinary communication requires cooperation between the disciplines to agree
on an ontology and on defining the rules of collaboration.
(D3) Inferential and reasoning abilities of the experts who are exchanging knowledge: one
expert could communicate some knowledge to another who may not have the reasoning capability to
make a proper use of it.
Teaching CS2 seems to fall within the scope of interdisciplinarity, and as such, one has to deal
with some of the problems of interdisciplinary communication. However, the situation does not, even
remotely, fit this pattern regarding CS2. While the IT instructor could be classified as an IT expert, the
students do not represent potential experts in their disciplines. A large percentage of the students who
enter universities are not studying the disciplines that reflect their capabilities, cognitive skills and/or
their problem-solving abilities. Thus, discipline, as a criterion for partitioning students into classes, was
not successful, and clearly, randomness is not an option.
The aspiration to maximize the learning output of CS2, which-as mentioned above-is intended
as a foundation for IT literacy, emphasizes the need for proper groupings of students into more
homogenous groups with similar cognitive and problem-solving abilities. Such grouping will help
create a fairer and healthier environment for the students and will make the task of the instructor more
tolerable. The failure to distribute the students in the appropriate groups could be a non-positive impact
251 Moh'd Belal Al- Zoubi, Imad Salah, Azzam Sleit,
Saleh Al-Sharaeh, Ammar Huneiti and Nadim Obeed
on both the students and the instructor. The instructor will be placed under a severe pressure to make
extreme ends meet, and the students’ reactions may range from completely disinterested to non-coping.
Therefore, efficient methods must be followed to assign students of CS2 to proper groups.
Mathematically, the process is called clustering, and the obtained groups are called clusters (Susanto,
2002). We propose a method based on fuzzy clustering techniques. The question is to find
criteria/factors that:
(a) are accessible/available,
(b) can be employed to indicate students' abilities, and
(c) yield proper groupings.
We have decided to consider the following factors:
[1] The High Schools General Exam (HSGE) score.
[2] The student’s performance in the course of Computer Skills 1 (CS1).
[3] The student’s Grade Point Average (GPA) score.
Factors (i) - (iii) satisfy conditions (1) and (2) above. Our aim is to show that they do yield
groups that are similar in their learning abilities.
2. Previous Work
To our knowledge, no similar previous work has been conducted. We believe that the problem at hand
is not very common. Many related existing applications involve assigning students to projects (Hill,
1983; Beheslitian-Ardekani, 1986; Muller, 1989; Weitz, 1992), rather than assigning them to classes.
In these applications, experienced students in the group are allocated to similar classes while the less
experienced ones are allocated to different classes. Most of the previous work on this problem was
based on optimal algorithms including linear and integer programming or multiple criteria models.
Bafail and Moreb (Bafail, 1993) proposed a goal programming model for allocating students to
various departments in an engineering school.
In (Saber, 2001), the authors state the major assignment problem as a linear programming
model through the use of value functions.
As discussed in (Weitz, 1992; Dhar, 1990), mathematical programming models may not
achieve optimal solutions and are extremely complex. This is beside the difficulties of formulating
such models.
In (Weitz, 1992), a Multi-Criteria Allocation Decision Support System (MCADSS) was
designed in order to partition MBA students into sections and groups. The authors noted that the
MCADSS system is not guaranteed to find a solution. As an alternative, the authors suggested to
present the obtained solution to the administrators, allowing them to manually adjust it in order to
achieve an acceptable final solution.
In (Susanto, 2002), the authors applied the fuzzy c-means clustering method (Bezdek, 1981) to
assign students of OR II course into three classes based on the their mastery level of the course’s
prerequisite subjects. Students with a similar achievement are grouped in the same class. Students with
different levels of achievement will be distributed in different classes. However, the authors used a
fixed number of groups (three groups). In practical applications, we would like to know what the most
appropriate number of groups to be formed is. In addition, the c-means clustering algorithm is sensitive
to outliers and is not efficient in recognizing non-spherical clusters (Kaufman, 1990; Trauwaert, 1991).
3. Proposed Algorithm
The main advantage of fuzzy clustering over hard clustering is that it yields much more detailed
information on the structure of the data (Kaufman, 1990). In fact, fuzzy clustering generates cluster
membership weights (degrees) that indicate how tight each object is to its cluster relative to other
clusters.
One of the most robust fuzzy clustering techniques is the Fuzzy Analysis Program (FANNY)
(Kaufman, 1990), which aims at minimizing the following objective function:
n
k u 2
iv u 2jv d (i, j )
J 
i , j 1
n
v 1
2 u 2jv
j 1
(1)
where d(i, j) represents the given distances (or dissimilarities) between objects i and j, and uiv is the
membership of object i to cluster v. The membership functions are subject to the constraints:
1. uiv  0 ,i, v
These constraints imply that memberships must be positive and that each object has a certain
total membership distributed over different clusters. This total membership is normalized to 1. The
objective function is minimized numerically by means of iterative steps, taking into account the
constraints above.
It was shown in (Trauwaert, 1991) that FANNY has the advantage over the c-means method in
that it is less sensitive to outliers and can achieve better recognition of non-spherical clusters.
Therefore, in this study, we propose the FANNY clustering method to assign students to proper groups.
In order to determine the right number of clusters (groups), the silhouette coefficients
(Kaufman, 1990) are used. For a given point i in a cluster, the silhouette of i , s (i ) , is then defined as
follows:
b(i )  a (i )
s (i ) 
max a(i ), b(i )
(2)
where a(i ) is the average dissimilarity between the point i and all other points in A (the cluster to
which i belongs), and b(i) is the average dissimilarity between the point i and the points in the closest
cluster to A , which is B in this case.
The average of all silhouettes in the data set, S ' is called the average silhouettes width for all
points in the data set. The value S ' will be denoted by S ' (k ) which is used for the selection of the right
value of the number of clusters, k, by choosing that k for which S ' (k ) is as high as possible. The
Silhouette Coefficient (SC ) ) is then defined as follows:
SC  max S ' (k )   (3)
where the maximum is taken over all k for which the silhouettes can be constructed.
The value s(i) may be interpreted as follows:
s(i) ≈ 1 ⇒ object i is well classified (in cluster A);
s(i) ≈ 0 ⇒ object i lies intermediate between two clusters (A and B);
s(i) ≈ -1 ⇒ object i is badly classified (closer to B than to A).
The silhouette plot shows which objects lie well within the cluster and which ones are merely
somewhere inbetween clusters. A wide silhouette width indicates large s(i) values, and hence, the
quality of the clusters can be compared: a wide silhouette width is better than a narrow one. The height
of a cluster is simply equal to the number of objects in the cluster. The silhouette plot is useful for
assessing how well the individual objects are clustered.
4. Experimental Results
In order to test the efficiency of our proposed method, three data sets are tested. The first set (data set
1) contains 295 students who were just admitted at the University of Jordan and whose CS1’s scores
(the prerequisite for CS2). Since these students have just been admitted, we do not have their GPA
scores yet, but we have their HSGE scores. So, in this set, only two variables (dimensions) are
available, which are CS1 and HSGE, and the clustering of these students is based only on these two
variables.
Table 1 shows The silhouettes width for all points in the data set, S’, associated with the
number of clusters (groups), k. The silhouette coefficient is the maximum of the S’ values, shown as a
shaded cell in Table 1, and is associated with (k=2). We can conclude that the right number of clusters
for the current data set is 2.
Table 1: The average silhouettes width, S’, for all points in data set 1, associated with different k values.
k S’
2 0.4546
3 0.3022
4 0.3706
5 0.3198
6 0.3299
7 0.2895
8 0.2996
As mentioned above, the silhouette plot can be used to assess how well the individual objects
are clustered. Figure 1 shows the silhouette plots for k = 2, 3, 4 and 5. As illustrated in the figure, wider
silhouette values are obtained when k = 2, while narrower ones are obtained when k ≠ 2.
After applying the FANNY program, we found that each cluster contains a subset of students as
shown in Table 2: 125 students belong to cluster 1, while 170 students belong to cluster 2.
The second set (data set 2) which contains 1336 students was obtained from a previous
semester for students with available GPA scores, in addition to their CS1 and HSGE scores. The
clustering of these students is based on these three variables. The right number of clusters was (k = 2)
after calculating the silhouette coefficient values for a different number of clusters, as shown in Table
3.
Figure 1: Silhouette plot for different values of k in data set 1.
K=2 K=3
K=4 K=5
Table 2: Number of students (NoStd) distributed to each cluster.
Cluster NoStd
1 125
2 170
Figure 2 shows the silhouette plots for k = 2, 3, 4 and 5.
k S’
2 0.5775
3 0.5444
4 0.5063
5 0.4998
6 0.4703
7 0.4880
8 0.4524
K=2 K=3
K=4 K=5
Applying the FANNY program, we have found that each cluster contains a subset of students as
shown in Table 4. The Table shows that 603 students belong to cluster 1, while 733 students belong to
cluster 2.
Table 4: Number of students (NoStd) distributed to each cluster for data set 2.
Cluster NoStd
1 603
2 733
The third data set (data set 3) contains 3553 students from the first semester in the academic
year 2006/2007. The GPA, CS1 and HSGE scores are available. The clustering of these students is
based on these three variables. As shown in Table 5, the right number of clusters was (k = 3).
Figure 3 shows the silhouette plots for k = 2, 3, 4 and 5.
k S’
2 0.5176
3 0.6539
4 0.5035
5 0.4970
6 0.5972
7 0.5940
8 0.4436
K=2 K=3
K=4 K=5
Table 6 shows the distribution of students to clusters after applying the FANNY program. The
table shows that 1031 students belong to cluster 1, 1982 students belong to cluster 2, and 540 students
belong to cluster 3. We have found that the third cluster includes students with very high scores.
Table 6: Number of students (NoStd) distributed to each cluster for data set 3.
Cluster NoStd
1 1031
2 1982
3 540
Having found the right group to which each student belongs, it is the responsibility of the
Registration Office to organize the distribution of students in sections. Furthermore, the students still
have the choice to register in any section within their group.
5. Conclusion
Computer skills 2 is a compulsory course at all universities in Jordan. Assigning students to groups is a
difficult task because of the students’ different backgrounds. In this paper, a proposed approach for
assigning students into proper groups was presented. The approach is based on fuzzy clustering
techniques. As a result, students can be assigned to the right group. This enables lecturers to set
different acceleration levels appropriate to the background level of the students in the group.
This paper is only a first step in a joint effort, of an interdisciplinary nature, that aims to address
emergent teaching problems in many University-wide courses in Jordanian University. We hope to
extend the approach to take into consideration additional criteria that help understanding better
students’ ability and cognitive skills. We also aim to look at other approaches to find out which can
help better in clustering. We shall elaborate in some of these issues in future work.
References
[1] Daily, G. and P. Ehrlich, “Managing Earth’s Ecosystems: An Interdisciplinary Challenge”,
Ecosystems, Vol. 2, 1999, pp. 277-280.
[2] Kostoff, R., 2002. Overcoming Specialization, BioScience, 52(10): 937-941.
[3] Naiman, R., 1999. A Perspective on Interdisciplinary Science, Ecosystems, 2: 292-295.
[4] Sharma, A., 1997. Professional as Agents Knowledge Asymmetry in Agency Exchange,
Academy of Management Review, 22(3): 758-798.
[5] Collins, J., 2002. May You Live in Interesting Times: Using Multidisciplinary and
Interdisciplinary Programs to Cope with Change in the Life Sciences, Bioscience, 52(1): 75-
853.
[6] Biocca, Z. and F. Biocca, 2002. Building Bridges Across Fields, Universities and Countries:
Successful Funding Communication Research Through Interdisciplinary Collaboration, Journal
of Applied Communication Research, 30(4): 350-357.
[7] Susanto, S., P. Suharto and P. Sukapto, 2002. Using Fuzzy Clustering Algorithm for Allocation
of Students, Trans. on Engineering and Technology Education, 1(2): 245-248.
[8] Hill, A., J. Naumann and N. Chervany, 1983. SCAT and SPAT: Large-scale Computer Based
Optimization Systems for the Personnel Assignment Problem, Decision Sciences, 14( ): 207-
220.
[9] Beheslitian-Ardekani, M. and M. Mahmood, 1986. Development and validation of a tool for
assigning students to groups for class projects. Decision Sciences, 17(1): 92-113.
[10] Muller, T., 1989. Assigning Students to Groups for Class Projects: An Exploratory Test of Two
Methods, Decision Sciences, 20(3): 623-634.
[11] Weitz, R. and M. Jelassi,1992. Assigning Students to Groups: A Multi-Criteria Decision
Support System Approach, Decision Sciences, 23(3): 746-757.
[12] Bafail, A, and A. Moreb, 1993. Optimal Allocation of Students to Diferent Departments in an
Engineering College, Computers and Industrial Engineering, 25:295–308.
[13] Saber, H. and J. Ghosh, 2001. Assigning students to academic majors, Omega 29: 513–523.
[14] Dhar, V. and N. Ranganathan, 1990. Integer programming vs. expert systems: An experimental
comparison. Communications of the ACM, 33(3):323-336.
[15] Bezdek, J.C., 1981. Pattern Recognition with Fuzzy Objective Function Algorithms. Plenum
Press, New York.
[16] Kaufman, L. and P. Rousseeuw, 1990. Finding Groups in Data: An Introduction to Cluster
Analysis. John Wiley & Sons.
[17] Trauwaert, E., L. Kaufman and P. Rousseeuw, 1991. Fuzzy Clustering Algorithms Based on the
Maximum Likelihood Principle, Fuzzy Sets and Systems, 42(2): 213-227
ISSN 1450-216X Vol.21 No.2 (2008), pp.259-267
Evolution de la Microflore et Caractéristiques

Physico-Chimiques D’un Aliment Traditionnel à Base de
Manioc (Manihot Esculenta Crantz) Fermenté: Le Bêdêcouman
Rose Koffi-Nevry
UFR des Sciences et Technologies des Aliments
Université d’Abobo-Adjamé, 02 BP 801 Abidjan 02, Côte d’Ivoire
E-mail: rosenevry2002@yahoo.fr
Marina Koussemon
Zinzendorf Y. Nanga
Laboratoire National de Santé Publique (LNSP), Abidjan, Côte d’Ivoire
Mobio Marguerite Abro

Celah Kakou
Tano Kablan
Laboratoire des Sciences des Aliments
Abstract
Physico-chemical and microbiological evolution of Bêdêcouman, a traditional food

made from cassava (Manihot esculenta Crantz), were studied. Bêdêcouman is the product
obtained when cassava mash is fermented, cooked, pounded, cooked again and shaped.
Results of fresh mash, fermented mash and Bêdêcouman showed changing in the
composition of proteins, lipids, ash, moisture, pH and titratable acidity during cassava
processing into Bêdêcouman. This cassava-based product is rich in carbohydrate (45.27%),
poor in protein (1.09%) and lipid (0%) and has a high energy density (185.44 Kcal/100g).
Evolution of the microflora, pH and titratable acidity was assessed during 4 days storage at
ambient temperature. Level of total aerobic count, and yeast and moulds increase
significantly from 3.103 to 2.1.105 CFU/g, and from 1.2.102 to 1.5.105 CFU/g respectively
while pH decrease from 4.16 to 3.82. Level of coliforms is low in bêdêcouman. Total
coliforms and faecal coliform not detected the first days of storage reached level of 1.6 102
CFU/g and 22 CFU/g respectively the 4th day.
Traditionnel à Base de Manioc (Manihot Esculenta Crantz) Fermenté: le Bêdêcouman 260
Keywords: Bêdêcouman, cassava, microflora, fermentation, evolution, storage
Introduction
Le manioc est une denrée alimentaire essentielle dans les pays tropicaux où il est devenu l’aliment de
base des populations (Mentaldo, 1977). En Côte, d’Ivoire, avec une production de 1,9 millions de
tonnes par an, le manioc occupe le 2ème rang au niveau des cultures vivrières, après l’igname (FAO,
2001).
La culture et l’utilisation du manioc sont cependant confrontées à des problèmes, notamment
celui de la toxicité du manioc, et celui de sa haute périssabilité. Les processus de transformation du
manioc permettent à la fois de réduire les dérivés toxiques (glucosides cyanogéniques) contenus dans
ce produit et d’assurer sa conservation. Ces techniques de transformation ont conduit à une variété de
produits propres à chaque région. Dans la région ouest-africaine, où il est particulièrement très
consommé, le manioc est transformé en divers produits parmi lesquels on cite le kokondé (pâte obtenue
à partir de farine de manioc cuite dans de l’eau) au Cameroun, le gari (farine de manioc fermentée,
séchée et légèrement frite) au Bénin et au Nigéria (Chuzel et Griffon, 1987, Zoumenou et al., 1999).
En Côte d’Ivoire, on cite comme aliments dérivés du manioc l’attieké (semoule de manioc),
l’attoukpou (galette cuite de manioc fermenté), le placali (pâte de manioc fermentée et cuite) et le
foutou (pain de manioc) (Aboua, 1992; Kadio, 2004; Zoumenou et al., 1999; Koffi-Nevry et al., 2007).
L’importance accrue de ces aliments dans les régimes alimentaires de populations a entraîné la mise en
place de programmes de recherche en vue de leur valorisation. Ainsi, plusieurs travaux ont été
effectués sur le manioc et ses dérivés dont le foufou, le gari, l’attiéké, le placali et le foutou, le fufu et
le kokondé (Chuzel et Griffon, 1987; Vinck, 1982; Oyewole et Ofunda, 1992 Zoumenou, 1994;
Kouadio et al., 1991; Assanvo et al., 2001).
Parmi les dérivés du manioc, certains sont très peu connus et leur production demeure encore
au plan familial et local. C’est le cas du bêdêcouman qui est très essentiellement produit et prisé par les
populations du sud-est de la Côte d’Ivoire, notamment les ethnies Abouré et N’Zima du Sud-Comoé.
C’est un produit obtenu après fermentation de la pâte de manioc, suivie de cuisson, de pilonnage et de
façonnage. Le bêdêcouman se présente sous la forme d’un pain blanchâtre, de taille variant entre 10 et
15 cm, confectionné dans des feuilles de Tomatococcus danielli communément appelées « feuilles
d’attiéké». Il se consomme accompagné de sauces et peut être conservé et consommé pendant 4 jours à
température ambiante. Compte tenu de l’inexistence de travaux et d’informations scientifiques sur cet
aliment, cette étude a été menée. Ainsi certains paramètres physico-chimiques dont les teneurs en
glucides, lipides, protéines, humidité et cendres, ont été déterminés sur les produits obtenus à différents
niveaux du processus de fabrication du bêdêcouman (pâte de manioc non fermentée, pâte fermentée,
produit final). De plus, l’évolution de la microflore, du pH et du taux d’acide du bêdêcouman, au cours
de sa conservation à température ambiante pendant 4 jours a été suivie. Les germes recherchés sont les
germes aérobies mésophiles, les coliformes et les levures et moisissures.
Materiel et Methodes
Matériel
Le matériel d’étude est constitué de tubercules de manioc de la variété amère IAC obtenus avec des
producteurs de manioc de la ville de Bonoua (Sud-Est de la Côte d’Ivoire).
261 Rose Koffi-Nevry, Marina Koussemon, Zinzendorf Y. Nanga,
Préparation du bêdêcouman
Vingt (20) kg de tubercules de manioc sont épluchés, découpés lavés et broyés pour obtenir la pâte de
manioc fraîche (pâte non fermentée). Cette pâte est ensuite ensemencée avec 8% (p/p) d’un ferment
obtenu après conservation pendant 4 jours, à température ambiante (30 ± 2°C) de racines de manioc
épluchées, lavées, découpées et conservées dans un sac en jute. La pâte de manioc ainsi ensemencée est
additionnée de 0,2‰ d’eau, homogénéisée, puis stockée dans des sacs en fibre synthétiques sur
lesquels sont ensuite déposées des pierres, permettant un égouttage. La pâte est laissée ainsi fermenter
pendant 72 heures, puis tamisée par pressage contre les mailles d’un tamis en acier. La pâte tamisée est
emballée par petites quantités (environ 150g) sous forme de petits pains, dans des feuilles de
Tomatococcus danielli, communément appelées « feuilles d’attiéké », puis cuite dans de l’eau pendant
20 à 25 minutes. Le produit issu de cette 1ère cuisson est alors pilonné chaud dans un mortier, re-
emballé dans des « feuilles d’attiéké » et mis à cuire à nouveau pendant une heure. Le produit final est
encore pilonné dans un mortier, façonné sous forme des petits pains et reconditionné dans des feuilles
d’attiéké préalablement désinfectées par cuisson à l’eau pendant une heure (Figures 1 et 2).
Figure 1: Diagramme de fabrication du Bêdêcouman
TUBERCULES DE MANIOC
(variété IAC)
EPLUCHAGE
DECOUPAGE ET LAVAGE
BROYAGE
FERMENTATION (2 à 3 jours) ET EGOUTTAGE
TAMISAGE
CUISSON (25 min ~ au feu de bois)
PILONNAGE ET CONDITIONNEMENT
(feuilles de Tomatococcus danielli)
CUISSON (60 min ~ au feu de bois)
PILONNAGE ET CONDITIONNEMENT
(dans des feuilles de Tomatococcus danielli)
BEDECOUMAN
NN
Figure 2: Bêdêcouman dans une « feuille d’attiéké»
Echantillonnage
Cinq échantillons de 20 g chacun, de pâte de manioc non fermentée et de pâte fermentée sont prélevés,
ainsi que 5 échantillons de bêdêcouman déjà conditionnés et prêts à être consommés sont prélevés pour
les analyses physico-chimiques.
Les analyses microbiologiques, de même que le pH et le taux d’acide lactique ont été
déterminés quotidiennement, pendant 4 jours, sur 15 échantillons de bêdêcouman conservés à
température ambiante.
Analyses physico-chimiques
Les teneurs en humidité, matières grasses, sucres totaux, cendres et acide lactique sont déterminés pour
chaque échantillon selon la méthode AOAC (1980). Le taux de protéines est déterminé selon la
méthode de Kjeldhal. La valeur énergétique du bêdêcouman est déterminée selon la formule suivante:
Valeur énergétique (kcal/100g d’aliment) = (% Glucides x 4 Kcal) + (% Protéines x 4 Kcal). Pour la
détermination du pH des échantillons, 10 g de chaque échantillon sont additionnés à 20 ml d’eau
distillée dans un sachet stomacher, puis le tout est broyé. L’électrode du pHmètre (pH-Meter 604),
directement plongée dans la solution obtenue permet de lire le pH de chaque échantillon.
Evolution de la microflore du bêdêcouman

Sur les échantillons de bêdêcouman conservés à température ambiante pendant 4 jours, le nombre de
germes aérobies mésophiles (GAM), de coliformes totaux et fécaux, de levures et moisissures est
quotidiennement déterminé selon la méthode AFNOR (1996).
Resultats
Analyse chimique
La teneur en eau des dérivés du manioc diminue progressivement au cours du processus de fabrication
du bêdêcouman. Ainsi, cette teneur qui est en moyenne de 59,80% dans la pâte non fermentée, passe à
57,70% dans la pâte fermentée puis à 52,72% dans le produit final (bêdêcouman). Il est en de même
pour le taux de cendres qui diminue au fur et à mesure de la production du bêdêcouman. Les dérivés de
manioc sont très pauvres en lipides (0,2% pour les pâtes non fermentée et fermentée) et le bêdêcouman
n’en contient pas. Le taux de protéines est le plus élevé dans la pâte fermentée, avec une valeur de
2,04% contre 1,97% et 0,92% respectivement dans le pâte non fermentée et le bêdêcouman. Ce dernier
contient 45,27% de glucides et a une valeur énergétique de 185,44 kcal pour 100g d’aliment. Le taux
en acide lactique augmente dans les dérivés du manioc au fur et à mesure de la production du
bêdêcouman, tandis que le pH diminue de 6,20 dans la pâte non fermentée, à 4,16 dans le produit final
(Tableau 1).
Tableau 1: Composition chimique moyenne des dérivés de manioc au cours de la production du bêdêcouman
Dérivés du manioc
Paramètres Pâte non fermentée Pâte fermentée Bêdêcouman
Humidité (%) 59,80 ± 1,99 57,70 ± 1,81 52,72 ± 1,45
Acide lactique (% MS) 0,03 ± 0,01 0,27 ± 0,04 0,53 ± 0,02
Protéines (% MS) 0,98 ± 0,09 2,04 ± 0,07 1,09 ± 0,08
Lipides (% MS) 0,20 ± 0,10 0,2 ± 0,10 0
Cendres (% MS) 1,97 ± 0,10 1,40 ± 0,06 0,92 ± 0,08
Glucides (% MS) Nd Nd 45,27 ± 0,12
pH 6,20 ± 0,07 4,23 ± 0,05 4,16 ± 0,05
Valeur énergétique (kcal/100g) 185,44 ± 0,09
MS: matière sèche; Nd: non déterminé; Les résultats sont la moyenne de 3 prises d’essai
Variation du pH et du taux d’acide au cours de la conservation du bêdêcouman

Alors que le taux d’acide lactique augmente (de 0,53 à 0,71%) au cours de la conservation du
bêdêcouman à température ambiante, le pH diminue, allant de 4,16 juste après la préparation de
l’aliment à 3,82 après 4 jours (Figure 3).
Figure 3: Evolution du pH et du taux d’acide lactique au cours de la conservation du bêdêcouman
4,2 0,8
Taux d'acide lactique (%)
0,6
pH
pH
4
Acide lactique
0,4
3,8 0,2
0 24 48 72 96
Durée de conservation (heures)
Evolution de la microflore au cours de la conservation du bêdêcouman

Le nombre en germes aérobies mésophiles (GAM) du bêdêcouman, initialement de 3.103 UFC/g,
augmente au cours de la conservation du produit et atteint après 96 heures la valeur de 2,1.105 UFC/g.
Il en est de même pour les autres germes, à savoir les coliformes totaux, les coliformes thermotolérants
et les levures et moisissures. Après les GAM, les germes majoritaires du bêdêcouman sont les levures
et moisissures, avec des nombres variant de 1,2.102 à 1,5.105 UFG/g. Les coliformes totaux et les
coliformes thermotolérants sont les germes minoritaires, avec des nombres respectifs de 1,6.102 et 22
UFC/g après 4 jours de conservation du bêdêcouman (Tableau 2).
Tableau 2: Evolution de la microflore du bêdêcouman au cours de sa conservation
Durée de Germes recherchés (UFC/g)

conservation (heures) GAM CT CF Levures et Moisissures
0 3.103 ± 2.102 < 15 <1 1,2.102 ± 21
24 2,9.104 ± 103 21 ± 3 < 15 5,2.103 ± 2.102
48 3,1.104 ± 2,9.103 1,3.102 ± 29 < 15 1,1.104 ± 2,3.103
72 1,5.105 ± 7.103 1,8.102 ± 36 24 ± 4 105 ± 8,2.103
96 2,1.105 ± 8,8.103 1,6.102 ± 27 22 ± 5 1,5.105 ± 104
GAM: germes aérobies mésophiles; CT: coliformes totaux; CF: coliformes fécaux; UFC: unité formant colonie
Discussion
La teneur moyenne en eau du bêdêcouman est plus faible que celle des sous-produits obtenus au cours
de sa production. Le processus de fabrication du bêdêcouman comprenant des étapes d’égouttage et de
cuisson justifie ce résultat. Le taux d’humidité obtenu pour le bêdêcouman (52,72%) est relativement
faible en comparaison de ceux obtenus pour d’autres aliments similaires à base de manioc, notamment
le placali (60,5%) et le kokondé (72%) (Aboua, 1988; Zoumenou, 1994). Ce taux d’humidité
relativement bas dans le bêdêcouman constitue un avantage pour sa conservation du fait qu’une faible
activité de l’eau limite considérablement la croissance des microorganismes et donc l’altération des
aliments. Ce taux d’humidité justifie la durée de conservation du bêdêcouman qui est de 4 jours à
température ambiante.
Les produits dérivés du manioc au cours de sa transformation en bêdêcouman sont pauvres en
lipides (0,2% pour les pâtes de manioc et 0% pour le bêdêcouman) à l’instar d’autres dérivés du
manioc dont le foutou, l’attiéké, l’attoukpou et le placali (Zoumenou, 1994; Yao et al., 2006, Koffi-
Nevry et al,. 2007). Ces faibles teneurs en lipides suggèrent et justifient que la consommation du
bêdêcouman s’accompagne de celles d’aliments riches en ces constituants (lipides) tels que les sauces
au poisson ou à la viande.
La teneur en protéines de la pâte fermentée (2,04%) est plus élevée que celle de la pâte non
fermentée (0,98%). Ce résultat s’explique par la richesse du ferment en microorganismes,
microorganismes qui synthétisent de nombreuses enzymes au cours de la fermentation. La teneur en
protéines de la pâte fermentée de manioc diminue ensuite pour atteindre la valeur de 1,09% dans le
bêdêcouman. Cette diminution est due à la dénaturation des protéines par la chaleur, au cours des
différentes étapes de cuisson. La teneur en protéines du bêdêcouman est aussi similaire à celles
d’autres aliments dérivés du manioc, notamment le gari, l’attiéké, le placali, l’attoukpou… (Meuser et
Smolnik, 1980: Ezeala, 1984; Aboua, 1992; Koffi-Nevry et al., 2007).
Le bêdêcouman représente un aliment très énergétique, avec un apport de 185,44 kcal pour 100
g d’aliment. Il peut donc répondre en grande partie aux besoins énergétiques journaliers des
populations.
Le pH des divers dérivés du manioc diminue au fur et à mesure de la production du
bêdêcouman, tandis que le taux d’acide augmente. Cette situation résulte certainement de la croissance
de microorganismes fermentaires contenus en grande quantité dans le ferment rajouté à la pâte de
manioc. D’autres auteurs ont en effet montré que le ferment de manioc est constitué à majorité de
bactéries lactiques et de levures et moisissures (Amoa-Awua et Jacobsen, 1995). Ainsi, le
développement de tels microorganismes dans la pâte de manioc accélérerait le processus de
fermentation, notamment la production d’acide à partir des sucres fermentescibles présents (Meuser,
1978; Regez et al., 1989).
Au cours de la conservation à température ambiante du bêdêcouman, son pH diminue tandis
que son taux d’acide lactique augmente. Ces résultats peuvent s’expliquer par l’utilisation des sucres
contenus dans le produit par divers microorganismes. Ces résultats sont en accord avec ceux de
l’analyse microbiologique qui ont montré que les contenus en germes du bêdêcouman augmentent au
cours de sa conservation. Les GAM, dont le nombre varie de 3 .103 UFC/g après la fabrication du
bêdêcouman, à 2,1.105 UFC/g après 4 jours de conservation, sont probablement à majorité des germes
acidophiles (notamment des bactéries lactiques et des champignons microscopiques) capables de
supporter le pH acide (4,16 à 3,82) du produit. Ce faible pH combiné à la forte teneur en glucides et à
la température de conservation (30 ± 2°C) du bêdêcouman représentent en effet des conditions
favorables au développement des levures et moisissures dont la charge augmente au cours du temps.
Cet important développement de microorganismes dans le bêdêcouman peut avoir une incidence sur la
qualité du produit. Toutefois, les germes acidophiles (bactéries lactiques, levures et moisissures)
n’étant généralement pas des pathogènes, il est peu probable que leur présence affecte la qualité
sanitaire du produit.
De tous les germes recherchés, les coliformes sont ceux présentant les plus faibles nombres,
avec, après 96 heures de fermentation, des valeurs de 1,6.102 UFC/g pour les coliformes totaux et 22
UFC/g pour les coliformes thermotolérants. Le faible développement des coliformes dans le
bêdêcouman est dû au pH du milieu (généré et entretenu par la présence d’acide lactique) qui est
défavorable à la croissance de ces germes. Tomkins et al. (1987) ainsi que Nout et al. (1987) ont
précisé que lorsque le pH des pâtes fermentées baisse au-dessous de 4,5 consécutivement à la
production d’acide lactique, le développement des entérobactéries est inhibé. Le pH acide du
bêdêcouman et la présence d’acide lactique, constituent également une barrière à la croissance des
microorganismes pathogènes (Desmazeaud, 1996), ce qui justifie la faible présence des coliformes
thermotorérants (fréquemment associés à des pathogènes) durant la conservation du bêdêcouman.
Svanger (1991) a aussi indiqué que, dans les produits amylacés, la fermentation lactique conduit à la
diminution de nombreux germes pathogènes dont Salmonella, Staphylococcus aureus, Esherichia coli
et Shigella.
La présence dans le bêdêcouman de nombreux germes, notamment celle des coliformes
thermotolérants, traduit une post-contamination du produit survenue après l’étape finale de cuisson, et
un défaut d’hygiène.
Conclusion
Le bêdêcouman est un aliment riche en glucides, pauvre en protéines et il ne contient pas de lipides.
C’est un aliment très énergétique ayant un taux d’humidité relativement bas. Au cours de sa
conservation, ses teneurs en germes augmentent tandis que son pH diminue. La présence de coliformes
dans le bêdêcouman dénote d’une hygiène défectueuse et d’une post-contamination du produit après sa
cuisson.
References
[1] Aboua F., 1988. “Traditional food preparations of cassava in Ivory Coast”, Cassava Newletter,
12 (1). ISSN-00259/3688 CIAT.
[2] Aboua F., 1992. “Chemical changes in cassava tubers (Manihot esculenta Crantz) during
production of placali”. Trop. Sci, 32 (4): 343-360.
[3] AFNOR, 1996. Analyses microbiologiques Tome II. Contrôle de la qualité des produits
alimentaires. AFNOR, Ed. 545 p.
[4] Amoa-Awua W. K., et Jacobsen, 1995. “The role of Bacillus species in the fermentation of
cassava”, J. Appl. Bacteriol, 79: 250-256.
[5] AOAC, 1980. Official methods of Analysis, 13 th, Association of Official Analytical Chem.
Washingon, DC.
[6] Assanvo J. B, G., Agbo, Z., Farah, P., Coulin et Y. E. N., Behi, 2001. “Etude de la microflore
du ferment de manioc pour la production de l’attiéké Adjaoukrou (Nouvel Ousrou, Orbaff et
Bouboury) ”, Communication écrite. Journées portes ouvertes, Jubilé centre Suisse de
recherches Scientifiques, Abidjan, 2001.
[7] Chuzel G. et D., Griffon, 1987. Le gari: un produit traditionnel fermenté à base de manioc.
Aspect fermentation et détoxification. Montpellier. Centre d’étude et d’expérimentation du
machinisme agricole tropical. 13p. Fr., 23 Ref II.
[8] Desmazeaud M., 1996. “Les bactéries lactiques dans l’alimentation humaine: utilisation et
innocuité”, Cahiers Agriculture. 5 (5): 331-342, 9-10.
[9] Ezeala D. O., 1984. “Changes in the Nutritional quality of fermented cassava tuber”, J. Agric.
Food. Chem, 32: 467-469.
[10] FAO, 2001. Production-Annuaire. Collection FAO Statistique, 55 (170).
[11] Kadio A., 2004. Etude comparative de quelques propriétés physico-chimiques et
organoleptiques d’une variété amère (IAC) et d’une variété douce (Bonoua) du manioc au cours
de la fabrication de l’attoukpou. Mémoire de DEA. Université d’Abobo-Adjamé, Abidjan, Côte
d’Ivoire. UFR-STA. 52p.
[12] Koffi-Nevry R., M., Koussémon, et F., Aboua, 2007. “Chemical and organoleptic properties of
attoukpou made from two cassava varieties (Manihot esculenta Crantz), Bonoua and IAC”, J.
Food Technol, 54 (4): 300-304.
[13] Kouadio N., K. E., Kouakou, S. F., Angbo, et K., Mousso, 1991. “Etude comparative des
méthodes traditionnelles de préparation de l’attiéké dans le sud de la Côte d’Ivoire”, Ind Alim
Et Agric (108): 703-706.
[14] Mentaldo P., 1977. La transformation des produits agricoles en zone tropicale. JJA ed. Karthala
et CTA. 1991p.
[15] Meuser F., 1978. Cassava flour and starches: some considerations. Cassava harvesting and
processing. Procedings of a workshop help at CIAT, Cali, Colombia.
[16] Meuser F. et H. D., Smolnik, 1980. “Processing of cassava to gari and other foodstuffs”,
Starch, 32 (4): 116-122.
[17] Nout M.J.R., J. G. A., Hautvast, J., Van der Haar, W. E. W. E., Marks et F. M., Rombout, 1987.
La composition et la sécurité microbiologique des produits de sevrage à base de céréales. In:
Pour améliorer l’alimentation des jeunes enfants en Afrique orientale et australe. Compte-rendu
d’atelier. Nairobi. UNICEF-SIDA-CRDI. pp 280-297.
[18] Oyewole O. B. et S. A. S. A., Ofunda, 1992. “Effects of processing variables cassava
fermentation for fufu production”, Trop. Sci, 32 (3): 213-240.
[19] Regez P. F., C., Rigamonti, et O., Geiges, 1989. “Evaluation de l’activité amylolytique et de la
formation de biomasse de quelques champignons”, Microbiologie, Aliment, Nutrition (MAN), 7:
7-24.
[20] Svanger U., 1991. Lactic fermentation of cereal based weaning gruels and improved nutritional
quality. In: Traditional African Food. Quality and nutrition. Ed IFS. pp 53-60.
[21] Tomkins A., D., Alnwick, et P., Haggert, 1987. L’emploi des produits fermentés. In: Pour
améliorer l’alimentation des jeunes enfants en Afrique orientale et australe. Compte-rendu
d’atelier. Nairobi. UNICEF-SIDA-CRDI. pp 156-192.
[22] Vinck D., 1982. La transformation du manioc en gari. Mémoire fac. Sci Agron Gembloux-
Belgique. 220 p.
[23] Yao K. J, A. R., Koffi, et F., Aboua, 2006. “Composition of deshydrated attiéké powder”, Trop
Sci, 46 (4): 224-226
[24] Zoumenou V., 1994. Etude physico-chimique et nutritionnelle de quelques préparations
alimentaires à base de manioc (Manihot esculenta Crantz). Thèse de doctorat 3ème cycle.
Université nationale de Côte d’Ivoire, FAST. 101 p.
[25] Zoumenou V., F., Aboua, D., Gnakri et A., Kamenan, 1999. “Etude des caractéristiques
physico-chimiques de certains plats traditionnels dérivés du manioc (foutou, placali et
kokondé)”, Tropicultura, 3: 120-126.
ISSN 1450-216X Vol.21 No.2 (2008), pp.268-276
Study of Performance of Anthropometric Indices in Screening

Children with Protein-Energy Malnutrition at Risk of Death
Hassan Barouaca
Faculty of Sciences, Sidi Mohamed Ben Abdellah University
Biology Department, Laboratory of Biochemistry
Nutritional Biochemistry Unit, PB: 1796 Fez Morocco
E-mail: barouaca@yahoo.fr
Ftima Zahra Squali Houssaini

Faculty of Sciences, Sidi Mohamed Ben Abdellah University
Biology Department, Laboratory of Biochemistry
Nutritional Biochemistry Unit, PB: 1796 Fez, Morocco
Moustapha Hida
Faculty of Medicine, Sidi Mohamed Ben Abdellah University
Paediatrics Ward, University Hospital Centre Hassan II Fez, Morocco
Ahamed Omar Touhami Ahami

Faculty of Sciences, Biology and Health Laboratory, Neurosciences and Nutrition Team
Human Biology and Population, Health Research Unit
Ibn Tofail University PB: 133 Kénitra, Morocco
E-mail: ahami_40@yahoo.fr
Youssef Aboussaleh
Faculty of Sciences, Biology and Health Laboratory, Neurosciences and Nutrition Team
Human Biology and Population Health Research Unit
Ibn Tofail University PB: 133 Kénitra, Morocco
Abstract
The purpose of this investigation was to determine the anthropometrical measures

predicting mortality in children admitted to the hospital. Anthropometric data from 70
children, protein-energy malnutrition (PEM), who survived were compared with those of
13 children who died during nutritional rehabilitation in hospital set up in Morocco,
between 2002-2005. To evaluate the performance of anthropometric indices in screening
children at risk of death, we perform an analysis by receiver operating characteristic curve
(ROC) (sensitivity vs. specificity) for anthropometric indices. The performances, expressed
by area under ROC curve (AUROC), of z-score index (Weight/Age, Weight/Height and
Height/Age) were lowest and equivalent. Also, the performance of mild upper arm
circumference/Head circumference (MUAC/HC), MUAC/Height and MUAC/Age were
lowest and approximately equal to that of z-score indices. We conclude that the MUAC
alone was the best of all anthropometrics indices in screening children at risk of death.
Protein-Energy Malnutrition at Risk of Death 269
Keywords: Screening, Children, Death, Protein-energy malnutrition and Anthropometric.
Introduction
Severe acute malnutrition (SAM), is defined as a weight-for-height measurement of 70% or more
below the median, or three SD or more below the mean National Centre for Health Statistics reference
values, which is called “wasted”; the presence of bilateral pitting oedema of nutritional origin, which is
called “oedematous malnutrition”; or a mid-upper-arm circumference of less than 110 mm in children
age 1–5 years [1]. Case-fatality rates in hospitals treating SAM in developing countries average 20–
30% and have remained unchanged since the 1950s [2]. During the follow-up period, severely
malnourished young children had approximately 6.8 times risk of dying than better nourished children
[3]. Good diagnosed precision in identifying prognostic factors associated with mortality, early
intervention and good subsequent management are important in reducing in-hospital mortality for
children with PEM. However, recent report suggests that there are problems in initially identifying
severely malnourished children at hospital admission [5]. To discriminate those children at high
mortality risk, several measures were investigated. In this context, several studies have reported an
association between poor nutritional status assessed by anthropometrics measurements and child
mortality. Some of these studies reported that anthropometric indicators may be sufficient markers to
predict death [6], while others recommended biochemical markers and clinical signs [7]. We conduct
this study to evaluate predictors of in hospital-mortality in Moroccan children, with protein-energy
malnutrition, received for nutritional rehabilitation.
Objective
The study aims to determine the predictive ability of anthropometrics measures to discriminate
children at high risk of death admitted to the hospital for nutritional rehabilitation.
Materiel and Methods

Subjects
Eighty-three children with protein-energy malnutrition (PEM) were admitted to the paediatric unit
ward of Ibn Khatib hospital, University Hospital Hassan II Fez, Morocco, for nutritional rehabilitation,
between January 2002 and June 2005. Children classified as suffering from PEM, when
anthropometrics measures (Z-score Weight-for-Height<-2 standard deviation of reference National
Centre for Health Statistics and World Health Organization) and/or having symmetric oedema. Ethical
approval for this study was obtained from the Ministry of Health and the proposed studies were
explained in detail to the parents of the children and oral informed consent was obtained.
Weight
Children at follow up were, weighed with minimal clothing on a Balance (RGZ-20), with an accuracy
of 10 grams. The scales were calibrated every morning before used for weighing.
Height and Age

Two nurses were helping to take height measurements. The height was measured to the nearest
millimetre with a stadiometer. Mid-upper arm circumference (MUAC) was measured with a plastic
flexible tape at the mid-point of the upper arm, and the readings recorded to the nearest 0.1cm. Age
was obtained from mothers' interviews. All data on the children, mother and family were collected in a
questionnaire during interviews at admission.
270 Hassan Barouaca, Ftima Zahra Squali Houssaini, Moustapha Hida,
Treatment
The children were admitted to the hospital Ibn Khatib ward of the paediatric unit of the University
Hospital Centre Hassan II Fez. They were managed according to a standard protocol this involved
correction of fluid and electrolyte imbalances and administration of broad-spectrum antibiotics.
Renutrition was starting within the first 24 h after admission, and was continued until full clinical
recovery had been achieved, after 21 days. All children were feed on an infant formula that provided
75 Kcal/100 ml and 1.2 g protein.Kg-1.d-1 until appetite was restored (day 9 post admission). During
the catch-up growth phase, the patients were fed a milk-based formula that provided (100-150kcal/100
ml) and 3g protein per kg per day. The feed was given every 3 h throughout the day or as a smaller
quantity every 2 h when the child was having problems tolerating the feed. Mothers were invited to
stay in hospital with their child during nutritional rehabilitation.
Statistical Methods
Construction of receiver operating characteristics curves
The ability of routine anthropometrics measures to differentiate between children who died and those
who survived was assessed by plotting receiver operating characteristic curves (sensitivity vs. 1 -
specificity) [8]. Theses curves are drawn for each index by systematically varying die value of the cut-
off point. The distance between the receiver operating curve and the diagonal is a measure of the
discriminating ability of an indicator. This measure expresses how are well the anthropometrics
measures distinguish between patients who lived and those who died. The performance of a diagnostic
test in terms of sensitivity and specificity will be better if the area under receiver operating
characteristic (ROC) curve is higher above the diagonal. The area under ROC curve is obtained by
plotting the proportion of true positives or the sensitivity of a diagnostic test (the proportion of positive
tests among the deceased in this case) against the proportion of false positives (the proportion of
positive tests among the survivors) or 1 - specificity, with these two parameters estimated at different
operating points. The diagonal of the graph is called the "chance line" since its points correspond to
equal proportions of false and true positives.
Comparison of Groups
Mean age and the anthropometric indices were compared between children who survived and children
who died by Student’s t-test. The odds ratio at 95% confidence intervals, were used to determine the
association between factors and risk of death. To compare nonparametric variables (categorical age and
sex), the χ2 test was used. Difference was significant when P<0.05.
Software used
The MedCalc 9.02, statistical package, was used to compare the performance of anthropometrics
measures, by plotting receiver operating characteristic curves and to determine the odds ratio. Epi-Info
2000 program, Version 6.04 was used to determine the z-scores of Weight-for-Age, Height-for-Age
and Weight-for-Height.
Result
Table 1: Relative Risk, confidence interval and p values for mortality according to age and sex.
Physical variables Dead group Recovered group Odds ratio (C.I. at 95%) χ2 P
Sex
Femelle 9 29 3.18 (0.89 to 11.33) 3.41 0.12
Mâle 4 41
Age
6 - 24 12 57 2.74 (0.33 to 22.96) 0.93 0.35
24 - 60 1 13
Age (mon) 11.60 ± 5.54 16.2 ± 4.32 - - 0.59
Mon: month, I.C: Confidence interval.
The mean age of the DG and RD were not significantly different. As revealed by adds ratio,
there were no significant association between age and death. Also, the analysis of sex gender (Table 1)
revealed no significant association between sex and mortality.
Table 2: The physical characteristics of children at admission to the Hospital for nutritional rehabilitation.
Physical characteristics Recovered group Dead group P

Number (N) 70 13 -
Weight (Kg) 5.72 ± 2.41 5.51 ± 1.45 0.80
Height (cm) 66.35 ± 12.08 63.11 ± 11.00 0.37
HC (cm) 42.20 ± 2.11 39.94 ± 3.74 0.06
MUAC (cm) 9.56 ± 1.81 8.94 ± 2.17 0.30
MUAC for HC 0.22 ± 0.03 0.20 ± 0.02 0.11
MUAC for Height 0.14 ± 0.02 0.13 ± 0.01 0.51
MUAC for Age 0.92 ± 0.46 0.91 ± 0.51 0.54
Z-HFA -2.72 ± 2.11 -2.41 ± 2.10 0.65
Z-WFA -2.92 ± 0.99 -3.35 ± 0.86 0.83
Z-WFH -2.25 ± 1.59 -2.54 ± 1.42 0.56
HC: Head circumference; MUAC: Mild upper arm circumference; Z-HFA: Z-score height for age; Z-WFA: Z-score weight for age and Z-WFH: Z-
score weight for height.
The physical characteristics of dead and recovered children are given in Table 2. There is no
significant difference in all anthropometrics indices between dead and recovered groups.
Table 3: Diagnostic value of anthropometrics indices to discriminate children at risk of death.
Cut off points Sensibility (95 % C.I.) Specificity (95 % C.I.) AUROC curve
Z-WFA ≤ -3.05 45.5 ( 16.9 - 76.5) 71.9 ( 59.2 - 82.4) 0.566
Z-WFH ≤ -2.1 72.7 ( 39.1- 93.7) 50.0 (37.4 - 62.6) 0.561
Z-HFA < -4.39 91.7 (61.5 - 98.6) 27.1 (17.2 - 39.1) 0.557
MUAC ≤ 8.25 (cm) 63.6 (30.9 to 88.8) 73.9 (61.9 - 83.7) 0.719
MUAC/HC ≤ 0.217 63.6 (30.9 - 88.8) 58.0 (45.5 - 69.8) 0.530
MUAC/Height ≤ 0,121 81.8 (48.2 - 97.2) 40.6 (28.9 - 53.1) 0.561
MUAC/Age ≤ 1.071 63.6 (30.9 - 88.8) 59.1 (46.3 - 71.0) 0.596
CI: Confidence interval; AUROC curve: Area under receiver operating characteristic curve.
The receiver operating characteristic and the diagnostic values of the anthropometrics indices
according to cut-offs to discriminate children who are at risk of death are presented in (Table 3). The
sensitivity of simple MUAC, MUAC for HC and MUAC for Age were identical, but in regard to their
specificity and AUROC curve, simple MUAC was the highest. Among all anthropometrics indices z-
score WFH and MUAC for height give better balance of sensitivity and specificity but they have week
and equal AUROC curve (Table 2). Also, all z-score indices show equal and week AUROC curve than
do simple MUAC.
Figure 1: Receiver operating characteristic curve for simple MUAC and Z-scores indices.
Figure 2: Receiver operating characteristic curve for simple MUAC and MUAC with correction of Age,
Height and Head circumference.
Figure 3: Receiver operating characteristic curve for simple MUAC and others anthropometrics indices based
on Age.
The figure 1, 2 and 3 present a comparison of performance, expected by AUROC, of all

anthropomerics indices. As we can see in figure 1, 2 and 3, the MUAC had the highest AUROC curve
compared to all anthropometrics indices. We can see also that, the correction with Age, Height and
Head circumference did not improve the predictive performance of MUAC (Figure 2). All indices
based upon age shows no improvement in diagnosis of death (Figure 3).
Discussion
In the present study, an overall in-hospital mortality rate of (15.66%) was observed and remains high.
The fatality rate in some African countries increased from 15 to 24 % and can reach up to > 50% in
Mulago Hospital Uganda's (2002) [9]. This high rate of death may be the result of a lack of knowledge
of main factors associated with paediatric mortality to plan the most efficient methods of care at higher
mortality risks. After the analysis, the variable ‘sex’ was not found to have a statistically significant
association with death. This finding agrees with Bachou et al (2006) who also found no difference in
mortality between sexes [10]. The association between age and mortality revealed that patient less than
24 months showed a mortality risk of 2.74 times higher when compared to patients older than 24
months, but this association was found to be not significant. Our results are consistent with other
reports [11] where no clear relationship was found between age and mortality as well as those
performed at paediatric wards of Mulago, Uganda's by Bachou and al (2006); however, this contrasts
with the findings of Mitra et al (2000) [10-12].
The clinical signs of malnutrition, biochemical and anthropometric indicators can be used to
assess the nutritional status, but their efficiency in defining malnutrition depends on their ability to
show a direct relationship with morbidity and mortality [13]. In our study, the mortality discriminating
power of anthropometric indices was assessed by the receiver operating characteristic curve analysis.
Results, however, show that discriminating power of risk of death can be made simple than is usually
suggested; sample MUAC alone, was a slightly better predictor of death than nutritional indices based
on height, weight or age. Also, all of the examined anthropometric indices demonstrated some
discriminatory capacity of risk of death, with the highest performance and specificity for simple
MUAC. Our finding of the high discriminatory capacity of MUAC over other anthropometric indices
in prognosticating subsequent mortality risk was consistent with those of other studies [14, 15]. Our
results also showed that, for given values of nutritional status measured by z-score indices WFA, WFH
and HFA performed equalled and lowest discriminatory capacity. Erinoso et al, (1993), suggested that
Z-WFA, Z-WFH and Z-HFA were associated with poor screening children at risk of death [16]. In
addition, it was confirmed that there are no significant differences in the prediction of mortality across
four indicators (WFA, HFA, WFH, and mild arm circumference-for-age) when applied to young
children [17]. As revealed by Bern et al, (1977) and Van den Broeck et al, (1996), very low WFA was
not associated with an increased risk of mortality [18, 19]. Similarly, Chen et al, (1980) demonstrated
using the best indicator (weight for age) that a sensitivity of only 25% at a specificity of 95%, leading
them to conclude that the nutrition screening of individual by anthropometry does not appear to be
highly efficient [20]. The weakest discrimination of child mortality, obtained by Z-WFA and Z-WFH,
was explained by the observation that the weight body is more closely related to total body water, the
proportion of which is thought to be abnormally high in malnourished children [11]. This could affect
the relevance of the indices based on weight. Prudhon et al, (1996) suggested that anthropometrics
indices seem to have a better predictive value for mortality among children without oedema than for
those with oedema [21]. This is not surprising because the weight recorded upon admission includes
the weight of oedema fluid; thus, oedematous children are likely to have the severity of their tissues
wasting underestimated to an extent that is itself dependent upon the degree of oedema and hence to be
highly variable. Our results also show that age-dependent indices such as arm circumference for age, z-
score weight for age, and height for age do not emerge better as screening measures for mortality than
an age-independent index such as mild upper arm circumference. This is because MUAC measure is
largely determined by muscle mass and is closely related to total body protein and selects
proportionately younger children at an earlier stage of malnutrition compared with WFA [22].
However, MUAC is more practical for use in areas where age is not available or reliable because an
estimate of malnutrition by age-dependent indices is affected by error in age and measurement [23]. In
addition, in many under developing countries accurate ages are mostly unknown; the anthropometrics
indices based on age in this area have a little advantage in screening children at risk. As we have
suggested, the predictive power of simple MUAC did not improve after adding Age, HC or Height
based indicators. This result is consistent with previously published work [13, 24-6], demonstrating
that the inclusion of either height or age provided little improvement in the sensitivity and specificity
of MUAC. A study in Malawian children between the ages of 3 and 48 months revealed that adding the
HC to MUAC offered no advantage in screening for PEM in these children [25]. Thus, the use of
MUAC for Age, MUAC for Height or MUAC for HC indices in screening children at risk of death was
not recommended. These confirmations and the results presented above suggest that simple arm
circumference had a best performance to discriminate children at risk of death.
Conclusion
Our findings, as well as considerations of cost and practicality, suggest that MUAC may be more
appropriate than z-score (WFA, HFA and WFH), MUAC for HC, MUAC for Height and MUAC for
age in identifying groups of children at the greatest risk of dying. Also, the performance of indices
based age was not higher than simple MUAC and the correction of height, age, or HC did not improve
the performance of MUAC.
Acknowledgment
The authors would like to thank the paediatric staff of Ibn Khatib hospital, Hassan II University
hospital. This research was supported by project PROTARS P1T1/05.
References
[1] Sphere Project Team.(2003) The SPHERE humanitarian charter and minimum standards in
disaster response, 2nd edn. Geneva: The SPHERE Project.
[2] Schofield C., Ashworth A. (1996) why have mortality rates for severe malnutrition remained so
high.Geveva, Bulltin World Health Organisation; 74: 223–9.
[3] Schroeder D.G., and Brown K.H. (1994) Nutritional status as a predictor of child survival:
summarizing the association and quantifying its global impact. Bull World Health Organ;
72:569–79.
[4] Barouaca H., Squali F-Z., Iraqui MR, Amor CH., Ahami AOT. (2006) Etude rétrospective et
facteur de complication de la malnutrition protéino-énergétique, in Santé environnement et
éducation, Health environment and education. Digi Edition, Rabat.
[5] Hamer C., Kvatum K., Jeffries D., Allen S. (2004) Detection of severe protein-energy
malnutrition by nurses in The Gambia. Archives of disease in childhood; 89:181–4.
[6] Trowbridge F.L., and Sommer A. (1981) Nutritional anthropometry and mortality risk. The
American journal of clinical nutrition; 34: 2591–2.
[7] Dramaix M., Hennart P., Brasseur D., Bahwere P., Mudjene O., Tonglet R., Donnen P., Smets
R. (1993) Serum albumin concentration, arm circumference, and oedema and subsequent risk
of dying in children in central Africa. British Medical Journal; 307: 710–3.
[8] Metz LE. (1978) Basic principles of ROC analysis. Seminars in nuclear medicine; 8:283-98.
[9] Mulago Hospital: Inpatient Activity Statistics. Records office, Kampala, Mulago Hospital
Complex, 2002.
[10] Bachou H., Tumwine J.K., Mwadime R.K., Tylleskar T. (2006) Risk factors in hospital deaths
in severely malnourished children Kampala, Uganda. BMC pediatrics; 6:7.
[11] McLaren D.S., Burman D. Textbook of paediatrics nutrition. 2nd ed. Edinburgh: Churchill
Livingstone, 1982.
[12] Mitra A.K., Rahman M.M., and Fuchs G.J. (2000) Risk factors and gender differentials for
death among children hospitalized with diarrhoea in Bangladesh. Journal of health population
and nutrition; 18: 151–6.
[13] Trowbridge F.L., Staehling N. (l980) Sensitivity and specificity of arm circumference
indicators in identifying malnourished children. The American journal of clinical nutrition; 33:
687-96.
[14] Alam N., Wojtyniak B., Rahaman M.M. (1989) Anthropometric indicators and risk of death.
The American journal of clinical nutrition; 49(5):884–8.
[15] Briend A., Dykewics C., Graven K., Mazumder N.R., Wojtyniak B., Bennish M. (1986)
Usefulness of nutritional indices and classifications in predicting death of malnourished
children. British medical journal; 293:373–5.
[16] Erinoso H.O., Akinbami F.O., Akinyinka O.O. (1993) Prognostic factors in severely
malnourished hospitalised nigerian children anthropometrics and biochemical factors. Tropical
and geographical medicine; 45: 290–3.
[17] Pelletier D.L., Low J.W., Johnson F.C., Msukwa L.A. (1994) Child anthropometry and
mortality in Malawi: testing for effect modification by age and length of follow-up and
confounding by socioeconomic factors. The Journal of nutrition; 124: 2082–5.
[18] Bern C., Zucker J.R., Perkins B.A., Otieno J., Oloo A.J., Yip R. (1977) Assessment of potential
indicators for protein-energy malnutrition in the algorithm for integrated management of
childhood illness. Geveva, Bulletin World Health Organsation; 75: 87–96.
[19] Van den Broeck J, Eeckels R, Massa G. (1996) Validity of single-weight measurements to
predict current malnutrition and mortality In children. The Journal of nutrition; 126: 113–20.
[20] Chen L.C., Chowdhury A.K.M.A., Huffman S.L. (1980) Anthropometrics assessment of
energy-protein malnutrition and subsequent risk of mortality among preschool aged children.
The American journal of clinical nutrition; 33:1836–1845.
[21] Prudhon C., Briend A., Laurier D., Mary J.Y, Golden M.H. (1996) Comparison of weight- and
height-based indices for assessing the risk of death in severely malnourished children. The
American journal of epidemiology; 144:116–23.
[22] Zeitbin M.F. (1986) Comparison of malnourished children selected by weight-for-age, mid-
upper-arm circumference and maximum thigh circumference. Journal of tropical paediatrics;
32:190-5.
[23] Bairagi R. (1986) Effects of bias and random error in anthropometry and in age on estimation
of malnutrition. American journal of epidemiology; 23(b):185-91.
[24] Sommer A., Loewenstein M.S. (1975) Nutritional status and mortality: a prospective validation
of the Quac stick. The American journal of clinical nutrition; 28:287–92.
[25] Ball T.M., Pust R.E. Arm circumference v. arm circumference/head circumference ratio in the
assessment of malnutrition in rural Malawian children. Journal of tropical paediatrics, 1993,
39(5):298-3.
ISSN 1450-216X Vol.21 No.2 (2008), pp.277-287
Synthesis and Bioactivity of a Novel Cyclic Hexapeptide from

Stellaria Delavayi
Rajiv Dahiya
Department of Pharmaceutical Chemistry, Rajiv Academy for Pharmacy
Mathura - 281 001, Uttar Pradesh, India
E-mail: rajivdahiya04@yahoo.co.in; rajivdahiya77@rediffmail.com
Tel: +91-9897417450; +91-565-6531680
Ravi Dutt Sharma

Department of Pharmaceutical Chemistry, Rajiv Academy for Pharmacy
Mathura - 281 001, Uttar Pradesh, India
Abstract
Synthesis of a novel phenylalanine-rich cyclopeptide – delavayin B (12) was

accomplished by coupling of dipeptide Boc-L-ser-L-ile-OH and tetrapeptide L-phe-L-phe-
L-ala-gly-OMe followed by cyclization of linear hexapeptide fragment. Structure
elucidation of newly synthesized peptide was done on basis of FT-IR, 1H-NMR, 13C-NMR,
FABMS and elemental analysis data. From biological evaluation, it was concluded that
cyclic hexapeptide 12 exhibited high level of cytotoxic activity alongwith good
antimicrobial activity against pathogenic Gram-positive bacterium Staphylococcus aureus
and dermatophytes Trichophyton mentagrophytes and Microsporum audouinii.
Dicyclohexylcarbodiimide (DCC) was found to be a good coupling agent and
pentafluorophenol (PFP) was proved better for activation of acid functionality of linear
hexapeptide segment prior to cyclization.
Keywords: Delavayin B, Cyclic Hexapeptide, Stellaria delavayi, Solution-Phase Peptide

Synthesis, Cytotoxicity, Antibacterial and Antifungal Activities
1. Introduction
Past literature has provided several findings proving the potential of plant roots to produce a wide array
of natural products with interesting bioactivities (Nakamura et al, 2007; Tan and Zhou, 2006; Ireland et
al, 2006). Among these, cyclopolypeptides and related congeners have received special attention due to
their unique structures and wide pharmacological profile and may prove better candidates to overcome
the problem of resistance towards conventional drugs. A novel cyclic hexapeptide, delavayin B has
been isolated from roots of Stellaria delavayi (Caryophyllaceae) and its structure was elucidated from
spectroscopic evidence and by chemical degradation. The stereochemistry of all amino acid residues
was found to be in the L-configuration by Marfey’s derivatization followed by HPLC analysis (Morita
et al, 1997). As part of our continuing synthetic efforts on biologically active cyclic peptides (Dahiya,
2008; 2008a; 2008b; 2008c; 2007; 2007a; 2007b; Dahiya and Kaur, 2008; 2007; Dahiya and Kumar,
2008; 2007; Dahiya and Pathak, 2007; 2007a; 2006; Dahiya et al, 2006), the present investigation was
aimed at the synthesis of a novel cyclic hexapeptide of plant origin. Keeping in view of significant
Synthesis and Bioactivity of a Novel Cyclic Hexapeptide from Stellaria Delavayi 278
bioactivities possessed by natural cyclic peptides (Dahiya and Pathak, 2006a; Pathak and Dahiya,
2003), the above synthetic peptide was further subjected to cytotoxic, antibacterial and antifungal
activity studies.
2. Experimental
Melting points were determined by open capillary method and are uncorrected. L-amino acids, di-tert-
butylpyrocarbonate (Boc2O), DCC, triethylamine (TEA), N-methylmorpholine (NMM), trifluoroacetic
acid, pyridine and p-nitrophenol were obtained from SpectroChem Limited (Mumbai, India).
Diisopropylcarbodiimide (DIPC) and pentafluorophenol were procured from RANKEM RFCL
Limited (New Delhi, India). IR spectra were recorded on Shimadzu 8700 FTIR spectrophotometer
(Shimadzu, Japan) using a thin film supported on KBr pellets for solids and CDCl3 as solvent for
intermediate semisolids. 1H NMR and 13C NMR spectra were recorded on Bruker AC NMR
spectrometer (300 MHz), (Brucker, USA) using CDCl3 as solvent and TMS as internal reference. Mass
spectra was recorded on JMS-DX 303 Mass spectrometer (Jeol, Tokyo, Japan) operating at 70 eV
using fast atom bombardment technique. Elemental analyses of all compounds were performed on
Vario EL III elemental analyzer (Elementar, Germany). Optical rotation of the synthesized peptides
was measured on automatic polarimeter (Optics Tech, Ghaziabad, India) in a 2 dm tube at 25 °C using
sodium D-light and methanol as solvent. Purity of synthesized cyclopeptide as well as intermediates
was checked by TLC on precoated silica gel G plates utilizing CHCl3/MeOH as developing solvent.
2.1. Protection and Deprotection of Amino Terminal of Amino Acid/Peptide Units

The free amino group of L-amino acids was protected by introduction of Boc group. A solution of L-
amino acid (20 mmol), Boc2O (6 mL, 2.6 mmol) in 1N NaOH (40 mL) and isopropanol (30 mL) was
stirred at RT for 2.5 h, washed with light petroleum ether (b.p. 4060ºC) (20 mL), acidified to pH 3.0
with 2N H2SO4 and finally extracted with chloroform (3  20 mL). The organic layer was dried over
anhydrous Na2SO4 and evaporated under reduced pressure to give the crude product which was finally
crystallized from chloroform and petroleum ether (b.p. 4060ºC) to get pure compound. Physical
characterization data of the synthesized boc-amino acids 13 is given in Table 1.
Boc group was removed by treatment with trifluoroacetic acid (TFA). A solution of Boc-
di/tetrapeptide methyl ester (2.6 g (9)/5.55 g (10), 10 mmol), CF3COOH (2.28 g, 20 mmol) in CHCl3
(15 mL) was stirred at RT for 1 h, washed with saturated NaHCO3 solution (25 mL). The organic layer
was dried over anhydrous Na2SO4 and concentrated under reduced pressure to yield crude product
which was finally purified by crystallization from CHCl3 and petroleum ether (b.p. 4060ºC) to get
pure deprotected di/tetrapeptide units 9a and 10a.
2.2. Protection and Deprotection of Carboxyl Terminal of Amino Acid/Peptide Units

The carboxyl group of L-amino acids was protected by esterification with MeOH/SOCl2. A solution of
L-amino acid (20 mmol), methanol (100 mL) containing thionyl chloride (1.4 mL, 20 mmol) was
refluxed for 10 h. After evaporating the solvent, the residue was triturated with ether at 0 ºC. The crude
product was crystallized from methanol/ether to get pure amino acid methyl ester hydrochloride.
Physical characterization data of the synthesized amino acid methyl ester hydrochlorides 46 is given
in Table 1.
Methyl ester group was removed by alkaline hydrolysis with LiOH. A solution of Boc-
dipeptide methyl ester (3.32 g (7)/4.27 g (8), 10 mmol), lithium hydroxide (0.36 g, 15 mmol) in
THF:H2O (1:1, 36 mL) was stirred at RT (1 h) and acidified to pH 3.5 with 1N H2SO4. The aqueous
layer was extracted with Et2O (3 × 35 mL) and combined organic extracts were concentrated under
reduced pressure. The crude product was crystallized from methanol/ether to get pure deprotected
dipeptide units 7a and 8a.
279 Rajiv Dahiya and Ravi Dutt Sharma
2.3. General Procedure for Synthesis of Dipeptide and Tetrapeptide Fragments (710)
Amino acid methyl ester hydrochloride/dipeptide methyl ester (10 mmol) was dissolved in
dichloromethane (DCM, 20 mL). To this, TEA (2.21 mL, 21 mmol) was added at 0ºC and the reaction
mixture was stirred (15 min). Boc-amino acid/dipeptide (10 mmol) dissolved in DCM (20 mL) and
DCC / DIPC (2.1 g/1.26 g, 10 mmol) was added to above reaction mixture with stirring. After 24 h, the
reaction mixture was filtered and the residue was washed with DCM (30 mL) and added to the filtrate.
The filtrate was washed with 5 % sodium bicarbonate and saturated sodium chloride solutions. The
organic layer was dried over anhydrous Na2SO4, filtered and evaporated in vacuum. The crude product
was crystallized from a mixture of chloroform and petroleum ether (b.p. 6080ºC) followed by cooling
at 0ºC.
Table 1: Physical Characterization Data for Synthesized Compounds 1–12
Elemental analysis
Mol. formula a b (Calc.) / (found)
Compd. Physical state M.p. (°C) % Yield Rf [D]
(Mol. weight)
%C %H %N
C8H15NO5 80 46.82 7.37 6.83
1 White crystals 51-52 0.49 –7.61
(205) (3.28 g) 46.83 7.34 6.85
C14H19NO4 89 63.38 7.22 5.28
2 White crystals 86-87 0.68 +24.92
(265) (4.7 g) 66.40 7.23 5.24
C8H15NO4 84 50.78 7.99 7.40
3 White solid 80-81 0.60 –26.23
(189) (3.18 g) 50.79 7.98 7.43
C7H16ClNO2 91 46.28 8.88 7.71
4 Viscous liquid – 0.78 –112.1
(181) (3.3 g) 46.29 8.86 7.74
C10H14ClNO2 88 55.69 6.54 6.49
5 White crystals 155-156 0.85 +36.84
(215) (2.9 g) 55.69 6.57 6.50
C3H8ClNO2 78 28.70 6.42 11.16
6 White solid 175 0.53 –
(125) (1.95 g) 28.69 6.46 11.15
C15H28N2O6 75 54.20 8.49 8.43
7 Semisolid mass – 0.63 –54.7
(332) (2.5 g) 54.18 8.52 8.42
C24H30N2O5 83 67.59 7.09 6.57
8 White solid 91-93 0.51 –93.2
(426) (3.54 g) 67.58 7.12 6.55
C11H20N2O5 81 50.76 7.74 10.76
9 Semisolid mass – 0.76 +19.1
(260) (2.1 g) 50.79 7.75 10.76
C29H38N4O7 79 62.80 6.91 10.10
10 Semisolid mass – 0.81 –71.9
(554) (4.37 g) 62.82 6.89 10.14
C38 H54N6O10 77 60.46 7.21 11.13
11 Semisolid mass – 0.55* +78.3
(754) (5.8 g) 60.45 7.18 11.15
C32H42N6O7 86 61.72 6.80 13.50
12 White solid 184-185 0.72* +5.95 (+6.0)
(622) (2.67 g) 61.75 6.82 13.49
a
(CHCl3:MeOH / 7:3); * (CHCl3:MeOH / 9:1)
b
c, 0.5 in MeOH; 1 c, 2 in H2O; 2 c, 1 in C2H5OH; 3 c, 2 in AcOH; 4 c, 2 in C2H5OH; 5 c, 0.2 in MeOH
2.3.1. tert-Butyloxycarbonyl-phenylalanyl-phenylalanyl-alanyl-glycine methyl ester (10).

IR (CDCl3), v/cm1: 3134-3127 (m, -NH str, amide), 3065, 3059 (w, -CH str, rings), 2954, 2874 (m, -
CH str, asym and sym, CH3), 2929, 2925 (m, -CH str, asym, CH2), 2852-2848 (m, -CH str, sym, CH2),
1745 (s, -C=O str, ester), 1647-1636 (s, -C=O str, 2º amide), 1557-1554, 1471-1466 (m, skeletal bands,
rings), 1535-1530, 1526-1520 (m, -NH bend, 2º amide), 1465 (m, -CH bend(scissoring), CH2), 1392,
1370 (m, -CH bend, butyl-t), 1271 (s, C−O str, ester), 715-710, 695-692 (m, -CH def, oop, rings). 1H
NMR (CDCl3), : 8.15 (br. s, 1H, NH), 8.10 (br. s, 1H, NH), 7.52 (tt, 2H, J = 7.2, 4.4 Hz, H-m, phe-1),
7.22 (tt, 2H, J = 7.15, 4.35 Hz, H-m, phe-2), 7.05 (t, 1H, J = 6.2 Hz, H-p, phe-2), 6.94 (t, 1H, J = 6.15
Hz, H-p, phe-1), 6.86-6.81 (m, 4H, H-o, phe-1 and phe-2), 6.76 (br. s, 1H, NH), 6.62 (br. s, 1H, NH),
4.78-4.73 (m, 1H, H-, phe-1), 4.39-4.33 (m, 1H, H-, phe-2), 4.27-4.22 (m, 1H, H-, ala), 4.05 (d,
2H, J = 4.8 Hz, H-, gly), 3.64 (s, 3H, OCH3), 3.19-3.13 (m, 4H, H- , phe-1 and phe-2), 1.55 (9H, s,
butyl-t), 1.22 (d, 3H, J = 4.15 Hz, H-, ala). 13C NMR (CDCl3), : 171.7 (C=O, ala), 170.3 (C=O, phe-
2), 169.0 (C=O, gly), 166.2 (C=O, phe-1), 156.5 (C=O, boc), 140.6 (C-, phe-1), 137.4 (C-, phe-2),
131.1 (2C, C-m, phe-1), 129.5 (2C, C-o, phe-2), 129.1 (2C, C-o, phe-1), 128.6 (C-p, phe-1), 128.0 (2C,
C-m, phe-2), 127.3 (C-p, phe-2), 79.7 (C-, butyl-t), 57.5 (C-, phe-1), 52.3 (OCH3), 51.2 (C-, ala),
49.7 (C-, phe-2), 40.1 (C-, gly), 38.5 (C- , phe-2), 36.2 (C-, phe-1), 28.8 (3C, C-, butyl-t), 18.2
(C-, ala).
2.4. Procedure for Synthesis of Linear Hexapeptide Fragment (11)

3.18 g (10 mmol) of 7a was dissolved in 25 mL of dimethylformamide (DMF) and solution was
neutralized with 2.21 mL (0.021 mol) of NMM at 0 C and the resulting mixture was stirred for 15
min. 4.55 g (10 mmol) of 10a was dissolved in 25 mL of DMF and resulting solution with 2.1 g/1.26 g
(0.01 mol) of DCC / DIPC were added to above mixture. Stirring was first done for 1 h at 05 C and
then further for 35 h at RT. After the completion of reaction, the reaction mixture was diluted with
equal amount of water. The precipitated solid was filtered, washed with water and recrystallized from a
mixture of chloroform and petroleum ether (b.p. 40-60 C) followed by cooling at 0 C to get tert-
Butyloxycarbonyl-L-seryl-L-isoleucyl-L-phenylalanyl-L-phenylalanyl-L-alanyl-glycine methyl ester
(11) as pale-yellow semisolid mass. IR (CDCl3), v/cm1: 3334 (m/br, -OH str), 3138-3127 (m, -NH str,
amide), 3068, 3056 (w, -CH str, rings), 2957-2954, 2876-2872 (m, -CH str, asym and sym, CH3),
2928-2923 (m, -CH str, asym, CH2), 2853-2847 (m, -CH str, sym, CH2), 1750 (s, -C=O str, ester),
1648-1632 (s, -C=O str, 2 amide), 1556-1552, 1473-1468 (m, skeletal bands, rings), 1539-1532, 1525-
1522 (m, -NH bend, 2 amide), 1390, 1368 (s, -CH bend, butyl-t), 1268 (s, C−O str, ester), 714-710,
697-693 (m, -CH def, oop, rings), 930 (w, CH3 rocking, butyl-t), 666 (m/br, -OH bend, oop). 1H NMR
(CDCl3), : 8.50 (br. s, 1H, NH, phe-2), 8.15 (br. s, 1H, NH, phe-1), 8.12 (br. s, 1H, NH, ala), 7.42 (br.
s, 1H, NH, ile), 7.25 (br. s, 1H, NH, gly), 7.227.15 (m, 4H, H-m, phe-1 and phe-2), 7.13 (br. s, 1H,
NH, ser), 7.06 (t, 1H, J = 6.15 Hz, H-p, phe-2), 6.96 (t, 1H, J = 6.2 Hz, H-p, phe-1), 6.876.82 (m, 4H,
H-o, phe-1 and phe-2), 5.72 (s, 1H, OH, ser), 4.524.47 (m, 1H, H-, ile), 4.32-4.27 (m, 1H, H-, ala),
4.224.17 (m, 1H, H-, ser), 4.03 (d, 2H, J = 4.75 Hz, H-, gly), 3.953.90 (m, 2H, H-, phe-1 and
phe-2), 3.86 (dd, 2H, J = 6.55, 4.15 Hz, H-, ser), 3.63 (s, 3H, OCH3), 2.99-2.93 (m, 4H, H- , phe-1
and phe-2), 2.051.98 (m, 1H, H-, ile), 1.661.62 (m, 2H, H-, ile), 1.54 (s, 9H, butyl-t), 1.52 (d, 3H,
J = 4.2 Hz, H-, ala), 1.05 (d, 3H, J = 5.85 Hz, H-΄, ile), 0.96 (t, 3H, J = 7.8 Hz, H-, ile). 13C NMR
(CDCl3), : 176.2 (C=O, ser), 173.5 (C=O, ile), 172.0 (C=O, ala), 170.2 (C=O, gly), 169.1 (C=O, phe-
2), 168.2 (C=O, phe-1), 158.5 (C=O, boc), 140.6, 140.2 (2C, C-, phe-1 and phe-2), 130.6 (2C, C-o,
phe-2), 130.2 (2C, C-m, phe-1), 129.4 (2C, C-o, phe-1), 128.1 (2C, C-m, phe-2), 127.7 (C-p, phe-1),
127.2 (C-p, phe-2), 79.7 (C-, butyl-t), 63.5 (C- , ser), 58.9 (C-, ser), 57.4, 56.2 (2C, C-, phe-1 and
phe-2), 53.3 (OCH3), 49.1 (C-, ile), 47.5 (C-, ala), 41.3 (C-, gly), 40.1 (C-, phe-2), 37.9 (C-,
ile), 36.7 (C-, phe-1), 28.9 (3C, C-, butyl-t), 25.4 (C-, ile), 19.5 (C- , ala), 16.2 (C-΄, ile), 10.5
(C-, ile).
2.5. Cyclization of Linear Hexapeptide Fragment

To synthesize 12, linear hexapeptide unit 11 (3.77 g, 5 mmol) was deprotected at carboxyl end using
lithium hydroxide (0.18 g, 7.5 mmol) to get Boc-L-ser-L-ile-L-phe-L-phe-L-ala-gly-OH. The
deprotected hexapeptide unit (3.70 g, 5 mmol) was now dissolved in chloroform (45 mL) at 0 C. To
the above solution, DCC (1.06 g, 5 mmol) and p-nitrophenol / pentafluorophenol (0.94 g/1.23 g, 6.7
mmol) were added and stirred was done at room temperature (12 h). The reaction mixture was filtered
and the filtrate was washed with 10 % NaHCO3 (3  15 mL) and 5 % HCl (2  10 mL) to get the
corresponding p-nitrophenyl/pentafluorophenyl ester Boc-L-ser-L-ile-L-phe-L-phe-L-ala-gly-O-
pnp/pfp. To this compound (3.45 g/3.63 g, 4 mmol) dissolved in chloroform (35 mL), CF3COOH (0.91
g, 8 mmol) was added, stirred at room temperature (1 h) and washed with 10 % NaHCO3 solution (2 
25 mL). The organic layer was dried over anhydrous Na2SO4 to get L-ser-L-ile-L-phe-L-phe-L-ala-gly-
O-pnp/pfp which was dissolved in chloroform (25 mL) and NMM/pyridine (2.8 mL/1.61 mL, 21
mmol) was added. Then, whole contents were kept at 0 C (7 days). The reaction mixture was washed
with sufficient quantity of 10 % NaHCO3 solution and finally washed with 5 % HCl (3  25 mL). The
organic layer was dried over anhydrous Na2SO4. Finally, chloroform was distilled off and crude
product was crystallized from chloroform/n-hexane to get pure Cyclo (-seryl-isoleucyl-phenylalanyl-
phenylalanyl-alanyl-glycyl-) (12). Yield 2.67 g, (86 %, NMM), 2.15 g (69 %, C5H5N). IR (KBr),
v/cm1: 3330 (m/br, -OH str), 3140-3128 (m, -NH str, amide), 3065, 3052 (w, -CH str, rings), 2956-
2953, 2877-2874 (m, -CH str, asym and sym, CH3), 2926-2922 (m, -CH str, asym, CH2), 2852-2848
(m, -CH str, sym, CH2), 1652-1646, 1642-1633 (s, -C=O str, 2 amide), 1555-1550, 1472-1469 (m,
skeletal bands, rings), 1540-1533, 1524-1519 (m, -NH bend, 2 amide), 717-712, 695-692 (m, -CH def,
oop, rings), 669 (m/br, -OH bend, oop). 1H NMR (CDCl3), : 9.68 (br. s, 1H, NH, ile), 8.65 (br. s, 1H,
NH, ser), 8.35 (br. s, 1H, NH, phe-1), 7.78 (br. s, 1H, NH, phe-2), 7.67 (br. s, 1H, NH, ala), 7.43 (s,
1H, OH, ser), 7.35 (br. s, 1H, NH, gly), 7.237.16 (m, 4H, H-m, phe-1 and phe-2), 7.02 (t, 1H, J = 6.2
Hz, H-p, phe-2), 6.98 (t, 1H, J = 6.15 Hz, H-p, phe-1), 6.866.81 (m, 4H, H-o, phe-1 and phe-2), 5.98-
5.92 (m, 1H, H-, ala), 5.825.74 (m, 1H, H-, ser), 5.695.62 (m, 2H, H-, phe-2 and phe-1),
4.264.20 (m, 1H, H-, ile), 5.11 (d, 2H, J = 4.8 Hz, H-, gly), 4.04 (dd, 2H, J = 6.6, 4.2 Hz, H-,
ser), 2.43-2.36 (m, 4H, H-, phe-1 and phe-2), 1.651.54 (m, 3H, H- and H-, ile), 1.45 (d, 3H, J =
4.15 Hz, H-, ala), 0.98 (d, 3H, J = 5.9 Hz, H-΄, ile), 0.95 (t, 3H, J = 7.75 Hz, H-, ile). 13C NMR
(CDCl3), : 174.6 (C=O, phe-2), 173.6 (C=O, ile), 172.2 (C=O, ala), 171.0 (C=O, ser), 170.8 (C=O,
phe-1), 168.4 (C=O, gly), 137.9, 137.4 (2C, C-, phe-1 and phe-2), 130.8 (2C, C-o, phe-2), 130.3 (2C,
C-m, phe-1), 129.5 (2C, C-o, phe-1), 128.7 (2C, C-m, phe-2), 128.0 (C-p, phe-1), 127.5 (C-p, phe-2),
64.1 (C-, ser), 61.4 (C-, ile), 58.7 (C-, phe-1), 57.4 (C-, ser), 56.6 (C-, phe-2), 49.0 (C-, ala),
45.7 (C-, gly), 39.6 (C-, phe-2), 37.2 (C-, phe-1), 34.5 (C-, ile), 24.8 (C-, ile), 17.7 (C-, ala),
15.9 (C-΄, ile), 10.2 (C-, ile). FABMS: (m/z, rel. int.) 623.7 [(M + H)+, 100], 595.7 [(623.7–CO)+,
22], 552.6 [(gly-ser-ile-phe-phe)+, 35], 536.6 [(ile-phe-phe-ala-gly)+, 29], 524.6 [(552.6–CO)+, 17],
508.6 [(536.6–CO)+, 11], 479.6 [(ile-phe-phe-ala)+, 44], 476.6 [(phe-ala-gly-ser-ile)+, 67], 405.5 [(gly-
ser-ile-phe)+, 52], 408.5 [(ile-phe-phe)+, 79], 380.5 [(408.5–CO)+, 14], 377.5 [(405.5–CO)+, 9], 363.4
[(phe-ala-gly-ser)+, 47], 335.4 [(363.4–CO)+, 10], 276.3 [(phe-ala-gly)+, 32], 261.4 [(ile-phe)+, 19],
258.3 [(gly-ser-ile)+, 26], 219.3 [(phe-ala)+, 41], 191.3 [(219.3–CO)+, 14], 148.2 [(phe)+, 11], 145.2
[(gly-ser)+, 24], 120.2 [(C8H10N)+, 34], 117.2 [(145.2–CO)+, 8], 114.2 [(ile)+, 17], 91.1 [(C7H7)+, 17],
86.2 [(C5H12N)+, 22], 65.1 [(C5H5)+, 14], 60.1 [(C2H6NO)+, 19], 58.1 [(gly)+, 11], 57.1 [(C4H9)+, 7],
44.1 [(C2H6N)+, 16], 31.0 [(CH3O)+, 9], 30.1 [(CH4N)+, 12], 29.1 [(C2H5)+, 11], 15.0 [(CH3)+, 10].
2.6. Cytotoxic and Antimicrobial Activities

Synthesized cyclopeptide 12 was subjected to short term in vitro cytotoxicity study against Dalton’s
lymphoma ascites (DLA) and Ehrlich’s ascites carcinoma (EAC) cell lines at 62.5–3.91 μg mL1 using
5-FU as reference compound (Kuttan et al, 1985). Activity was assessed by determining the percentage
inhibition of DLA and EAC cells. CTC50 values were determined by graphical extrapolation method.
The results of cytotoxic activity are tabulated in Table 2. Antimicrobial activity studies were carried out
against four bacterial strains Corynebacterium pyogenes (MUMC 73), Staphylococcus aureus (MUMC
377), Pseudomonas aeruginosa (MUMC 266) and Klebsiella pneumoniae (MUMC 95) and four fungal
strains Microsporum audouinii (MUMC 545), Trichophyton mentagrophytes (MUMC 665), Candida
albicans (MUMC 29) and Aspergillus niger (MUMC 77) at 25–6 μg mL1 using gatifloxacin and
griseofulvin as reference drugs (Bauer et al, 1966). MIC values of test compound were determined by
tube dilution technique using sterile DMF. The petri plates inoculated with bacterial/fungal cultures
were incubated at 37 °C for 18 h and 48 h, respectively. The average diameters of the zones of
inhibition (in mm) of test compound were calculated for triplicate sets and compared with that
produced by the standard drugs. The results of antibacterial and antifungal activity studies are
summarized in Table 3. Experimental details of anthelmintic and antimicrobial activity procedures are
described in our previously published reports (Dahiya and Pathak, 2007b).
Table 2: In vitro Cytotoxic Activity Data for Synthesized Cyclopeptide 12
DLA cells EAC cells

Conc.
Compd. Live cells No. of % growth CTC50‡ Live cells No. of % growth CTC50
(μg/mL)
counted dead cells inhibition† (μM) counted dead cells inhibition (μM)
62.5 0 38 100.0 0 28 100.0
31.25 2 36 94.77 4 24 85.71
12 15.63 11 27 71.05 12.56 12 16 57.14 22.64
7.81 19 19 50.00 22 6 21.43
3.91 25 13 34.21 25 3 10.71
62.5 38 0 – 28 0 –
31.25 38 0 – 28 0 –
Control 15.63 38 0 – – 28 0 – –
7.81 38 0 – 28 0 –
3.91 38 0 – 28 0 –
Standard 62.5 0 38 100.0 0 28 100.0
31.25 0 38 100.0 0 28 100.0
15.63 10 28 73.68 37.36 11 17 60.71 90.55
(5-FU)
7.81 13 25 65.79 19 9 32.14
3.91 22 16 42.11 23 5 17.86
†
% growth inhibition = 100 – [{(Celltotal – Celldead)  100} / Celltotal]
‡
CTC50 = cytotoxic concentration inhibiting 50 % of percentage growth
Table 3: Response of Pathogenic Microorganisms to Synthesized Cyclopeptide 12
Diameter of zone of inhibition (mm)

Bacterial strains Fungal strains
Compd.
C. S. P. K. C. M. A. T.
pyogenes aureus aeruginosa pneumoniae albicans audouinii niger mentagrophytes
12 14(12.5)# 27(6) 9(25) 11(25) 14(6)
16(6) 8(25) 22(6)
Control – – – – – – – –
Gatifloxacin 20(12.5) 28(6) 24(6) 25(6) – – – –
Griseofulvin – – – – 20(6) 17(6) 18(12.5) 20(6)
#
Values in bracket are MIC values (g mL1).
3. Results and Discussion

3.1. Chemistry
In the present work, disconnection strategy was employed to carry out the first total synthesis of
delavayin B (12). The cyclopeptide molecule was split into three dipeptide units Boc-L-ser-L-ile-OMe
(7), Boc-L-phe-L-phe-OMe (8) and Boc-L-ala-gly-OMe (9). To obtain dipeptide 7, Boc-L-ser-OH (1)
was coupled with L-ile-OMe.HCl (4) employing DCC and DIPC as coupling agents and TEA/NMM as
bases. Other dipeptide units 8 and 9 were prepared by coupling of Boc-amino acids viz. Boc-phe-OH
(2) and Boc-L-ala-OH (3) with corresponding amino acid methyl ester hydrochlorides such as L-phe-
OMe.HCl (5) and gly-OMe.HCl (6) (Fig. 1a). Ester group of 8 was removed using lithium hydroxide
(LiOH) and Boc group of 9 was removed using CF3COOH. Deprotected dipeptide units were now
coupled to get the tetrapeptide unit Boc-L-phe-L-phe-L-ala-gly-OMe (10) (Fig. 1b). Dipeptide 7 was
deprotected at carboxyl end and coupled with tetrapeptide 10 after deprotection at amino terminal,
using DCC/DIPC to get the linear hexapeptide unit Boc-L-ser-L-ile-L-phe-L-phe-L-ala-gly-OMe (11)
(Fig. 1c). The methyl ester group of linear peptide fragment was replaced by p-nitrophenyl /
pentafluorophenyl ester group via hydrolysis of 11 followed by coupling of deprotected unit with p-
nitrophenol / pentafluorophenol (PNP/PFP). Boc group of resulting unit was removed using TFA and
deprotected linear fragment was now cyclized by keeping the whole contents at 0 C for 7 days in
presence of catalytic amount of NMM/pyridine (C5H5N) to get final product 12 (Fig. 1d).
Structure of the newly synthesized cyclohexapeptide as well as intermediates linear
di/tetra/hexapeptides were confirmed by spectral as well as elemental analysis. Peptide units were
prepared by solution phase technique (Bodanzsky and Bodanzsky, 1984) employing DCC/DIPC as
coupling agents and TEA/NMM as bases.
Synthesis of cyclohexapeptide 12 was accomplished with 86 % yield utilizing DCC as coupling
agent and NMM was proved to be a yield effective base as compared to pyridine, for cyclization of
linear hexapeptide fragment. Pentafluorophenol (PFP) was found to be better as compared to p-
nitrophenol (PNP), for activation of acid functionality of linear hexapeptide segment prior to
cyclization. Cyclization of linear peptide fragment was indicated by disappearance of absorption bands
at 1750, 1268 and 1390, 1368 cm1 (C-O str of ester and C-H bend of tert-Butyl group) and presence of
additional Amide I and II bands of the -CO-NH- moiety at 1652-1649 and 1521-1519 cm1 in IR
spectra of 12. Formation of cyclopeptide was further confirmed by disappearance of singlets at 3.63
and 1.54 ppm corresponding to three protons of methyl ester group and nine protons of tert-Butyl
group of Boc, in 1H NMR spectrum and disappearance of singlets at 158.5, 79.7, 28.9 and 53.3 ppm
corresponding to carbon atoms of tert-Butyl and ester group, in 13C NMR spectrum of 12.
Figure 1a: Synthesis of dipeptide units 7−9
O O
O O
HO
O DCC/DIPC NH H O
+ O Coupling
O N
O N + - HO
H NH3 Cl
OH O
1 4 7
O O
O O
O
O
O
N
HN OH DCC/DIPC H
+ +
H3N - Coupling
HN O
Cl
O O
2 5 8
O
O O HN
HO
O NH O
DCC/DIPC O
+ O
N + - Coupling O
O NH3 Cl O
H
3 6 9
Figure 1b: Synthesis of tetrapeptide unit 10
O
HO
O
N
H
LiOH
O
8 Deprotection
NH
O
O
NH O
8a O
O
DCC/DIPC
+ Coupling NH
O NH
H
H O N
O O
N
CF3COOH
9 O
O
Deprotection
H2N
10
9a
Figure 1c: Synthesis of linear hexapeptide unit 11
HO
O
O O
H OH
NH
LiOH N
7 Deprotection
O O
N
HO O H NH
O
O
7a NH
O
O
DCC/DIPC
+
Coupling NH
O NH
H
NH2 O N
O O
CF3COOH O
10 NH
Deprotection O NH
H
N
11
O
10a
Appearance of pseudomolecular ion peak (M + 1)+ at m/z 623.7 corresponding to the molecular
formula C32H42N6O7 in mass spectra of 12, along with other fragment ion peaks resulting from
cleavage at ‘ala-gly’, ‘ser-ile’ and ‘phe-phe’ amide bond levels, showed exact sequence of attachment
of all the six amino acid moieties in a chain. In addition, presence of immonium ion peaks at m/z 120.2
(phe), 86.2 (ile), 60.1 (ser), 44.1 (ala), and 30.0 (gly) further confirmed all five amino acid moieties in
cyclopeptide structure. Furthermore, elemental analysis of 12 afforded values ( 0.03) strictly in
accordance to the molecular composition.
Figure 1d: Synthesis of cyclohexapeptide – delavayin B 12
O OH
O
OH O
O N
O H N
N H
O H NH O
O NH OPNP/PFP
O
NH O
O LiOH
PNP/PFP NH
NH O NH
H
O NH
H N
O
N
O
O
O
NMM / pyridine
11 CF3COOH
(cyclization)
HO
O H O
N
NH
NH
O O
NH
NH
NH
O O
12
3.2. Pharmacology
Synthesized cyclopeptide exhibited potent cytotoxic activity against DLA and EAC cell lines with
CTC50 values of 12.56 and 22.64 μM respectively, in comparison to standard drug – 5-FU (CTC50
values – 37.36 and 90.55 μM). Moreover, 12 showed good activity against pathogenic Gram-positive
bacterium S. aureus and dermatophytes. Antifungal activity against T. mentagrophytes was found to be
even better than that possessed by standard drug – griseofulvin, at MIC value of 6 μg mL−1. Moderate
level of activity was seen against pathogenic microbes C. albicans and C. pyogenes, in comparison to
reference drugs – gatifloxacin/griseofulvin. However, 12 displayed no significant activity against A.
niger and Gram-negative bacteria. On passing toxicity tests, synthesized cyclohexapeptide 12 may
prove good candidate for clinical studies and can be new cytotoxic and antidermatophyte drug of
future.
Acknowledgement
Authors are grateful to U.S.I.C., DU, Delhi (India) and R.S.I.C., I.I.T., Delhi (India) for spectral and
elemental analysis. Also great thanks to Mr. Gajender Saini for his valuable contribution for the
completion of the project.
References
[1] Bauer, A.W., W.M., Kirby, J.C., Sherris and M., Turck, 1966. “Antibiotic Susceptibility
Testing by a Standardized Single Disk Method”, American Journal of Clinical Pathology 45,
pp. 493–496.
[2] Bodanzsky, M. and A., Bodanzsky, 1984. “The Practice of Peptide Synthesis”, Springer-
Verlag, New York, pp. 78–143.
[3] Dahiya, R., 2008. “Synthetic Studies on a Cyclic Hexapeptide from Dianthus superbus”,
Chemical Papers 62, in press.
[4] Dahiya, R., 2008a. “Synthesis and in vitro Cytotoxic Activity of a Natural Peptide of Plant
Origin”, Journal of Iranian Chemical Society 5, in press.
[5] Dahiya, R., 2008b. “Total Synthesis and Biological Potential of Psammosilenin A”, Archiv der
Pharmazie - Chemistry in Life Sciences 341, pp. 502–509.
[6] Dahiya, R., 2008c. “Synthesis, Spectroscopic and Biological Investigation of Cyclic
Octapeptide: Cherimolacyclopeptide G”, Turkish Journal of Chemistry 32, pp. 205–215.
[7] Dahiya, R., 2007. “Synthesis of a Phenylalanine-Rich Peptide as Potential Anthelmintic and
Cytotoxic Agent”, Acta Poloniae Pharmaceutica - Drug Research 64, pp. 509–516.
[8] Dahiya, R., 2007a. “Synthesis, Characterization and Biological Evaluation of a Glycine-Rich
Peptide – Cherimolacyclopeptide E”, Journal of Chilean Chemical Society 52, 1224–1229.
[9] Dahiya, R., 2007b. “Synthetic and Pharmacological Studies on Longicalycinin A”, Pakistan
Journal of Pharmaceutical Sciences 20, pp. 317–323.
[10] Dahiya, R. and K., Kaur, 2008. “Synthetic and Pharmacological Investigation of Segetalin C as
a Novel Antifungal and Cytotoxic Agent”, Arzneimittelforschung 58, pp. 29–34.
[11] Dahiya, R. and K., Kaur, 2007. “Synthetic and Biological Studies on Natural Cyclic
Heptapeptide: Segetalin E”, Archives of Pharmacal Research 30, pp. 1380–1386.
[12] Dahiya, R. and A., Kumar, 2008. “Synthetic and Biological Studies on a Cyclopolypeptide of
Plant Origin”, Journal of Zhejiang University Science B 9, pp. 391–400.
[13] Dahiya, R. and A., Kumar, 2007. “Synthesis and Biological Activity of a Potent Analog of
Natural Cyclopeptide”, International Journal of Natural and Applied Sciences 3, pp. 433–440.
[14] Dahiya, R. and D., Pathak, 2007. “First Total Synthesis and Biological Evaluation of
Halolitoralin A”, Journal of Serbian Chemical Society 72, pp. 101–107.
[15] Dahiya, R. and D., Pathak, 2007a. “Synthesis, Characterization and Biological Evaluation of
Halolitoralin B – A Natural Cyclic Peptide”, Asian Journal of Chemistry 19, pp. 1499–1505.
[16] Dahiya, R. and D., Pathak, 2007b. “Synthetic Studies on Novel Benzimidazolopeptides with
Antimicrobial, Cytotoxic and Anthelmintic Potential”, European Journal of Medicinal
Chemistry 42, pp. 772–798.
[17] Dahiya, R. and D., Pathak, 2006. “Synthetic Studies on a Natural Cyclic Tetrapeptide –
Halolitoralin C”, Journal of Pharmaceutical Research 5, pp. 69–73.
[18] Dahiya, R. and D., Pathak, 2006a. “Cyclic Peptides: New Hope for Antifungal Therapy”,
Egyptian Pharmaceutical Journal (NRC) 5, pp. 189–199.
[19] Dahiya, R., D., Pathak, M., Himaja, and S., Bhatt, 2006. “First Total Synthesis and Biological
Screening of Hymenamide E”, Acta Pharmaceutica 56, pp. 399–415.
[20] Ireland, D.C., M.L., Colgrave, and D.J., Craik, 2006. A Novel Suite of Cyclotides from Viola
odorata: Sequence Variation and the Implications for Structure, Function and Stability”,
Biochemistry Journal 400, pp. 1–12.
[21] Kuttan, R., B., Bhanumathy, K., Nirmala, and M.C. George, 1985. “Potential Anticancer
Activity of Turmeric (Curcuma longa)”, Cancer Letters 29, pp. 197–202.
[22] Morita, H., T., Kayashita, A., Uchida, K., Takeya, and H., Itokawa, 1997. “Cyclic Peptides
from Higher Plants. 33. Delavayins A-C, Three New Cyclic Peptides from Stellaria delavayi”,
Journal of Natural Products 60, pp. 212–215.
[23] Nakamura, S., X., Li, H., Matsuda, K., Ninomiya, T., Morikawa, K., Yamaguti, and M.,
Yoshikawa, 2007. “Bioactive Constituents from Chinese Natural Medicines. XXVI. Chemical
Structures and Hepatoprotective Effects of Constituents from Roots of Rhodiola sachalinensis”,

Chemical and Pharmaceutical Bulletin (Tokyo) 55, pp. 1505–1511.
[24] Pathak, D. and R., Dahiya, 2003. “Cyclic Peptides as Novel Antineoplastic Agents: A Review”,
Journal of Science and Pharmacy 4, pp. 125–131.
[25] Tan, N.-H. and J., Zhou, 2006. “Plant Cyclopeptides”, Chemical Reviews 106, pp. 840–895.
ISSN 1450-216X Vol.21 No.2 (2008), pp.288-295
On-Line Power Quality Monitoring System Using CompactRIO

Device: A Review on System Development
M. Z. A Ab Kadir
Department of Electrical and Electronics Engineering, Faculty of Engineering
Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia
E-mail: mzainal@eng.upm.edu.my
Tel: +60-389464362; Fax: +60-389466327
S. S. Tuan Othman
Department of Electrical and Electronics Engineering, Faculty of Engineering
Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia
L. Lombigit
Malaysian Nuclear Agency, Bangi 43000 Kajang, Selangor, Malaysia
Abstract
Power quality is an important issue in the power system. If power quality problem
occurred, the effects such as manufacturing interruptions, loss of production, loss of
revenue, lost of opportunities, product damage, and wasted energy and decreased
equipment life will significantly be affected. One of the ways to detect the power quality
problem is by on-line monitoring. This monitoring can be done by accessed to the point
address in networking and those who involved in this site can go through for a quick
assessment. This paper comprises of modelling the system, which is developed in
LabVIEW environment and integrated using a powerful CompactRIO system. Detail of the
modules and integration process are presented, together with the initial result obtained from
data monitoring.
Keywords: Power quality, LabVIEW software, CompactRIO system, online monitoring
1. Introduction
Power quality (PQ) covers a wide range of issues, from voltage disturbances like sags, swells, outages
and transients, to current harmonics, to performance wiring and grounding. The symptoms of poor PQ
include intermittent lock-ups and resets, corrupted data, premature equipment failure, overheating of
components for no apparent cause, etc (Dugan et al, 1996). The ultimate cost is in downtime, decreased
productivity and frustrated personnel. The term ‘power quality’ refers to the purity of the voltage and
current wave-form, and a power quality disturbance is a deviation from the pure sinusoidal form.
Harmonics superimposed on the fundamental are one cause of such deviations, and this chapter
describes the nature, generation and effects of harmonics on power supply systems, together with the
limitation of such effects and harmonic studies. In recent years power quality issues have captured the
attention of many researchers, energy suppliers and customers. The problem of controlling power
quality is becoming more relevant because of the widespread use of non-linear and time-varying
On-Line Power Quality Monitoring System Using CompactRIO Device:
A Review on System Development 289
single-phase or three-phase loads that increasingly affect the operation of distribution networks in
residential, commercial and industrial areas.
Power quality deterioration is due to transient disturbances (voltage sags, voltage swells,
impulses, and so on) and steady-state disturbances (harmonic distortion, unbalance, and flicker). Some
definitions relevant to non-sinusoidal systems are provided in order to deeply understand the solutions
proposed in the remainder. An outline of IEC and IEEE Standards guidelines, which deals with
harmonic disturbances, is presented. A topic such standards deal with and several researchers are
interested in is that non-linear loads should not be considered the only cause of the detrimental effects
related to power systems in these conditions. The responsibility for the power quality deterioration
should be shared between the supplier and the customer. This kind of information could be provided by
indices in a straightforward and simple way. The traditional indices used by the international standards
cannot satisfy these specifications.
2. Methodology on Monitoring System Development

This project is focused on modelling a monitoring system in Malaysian Nuclear Agency. The purpose
of this project is to monitor power quality problem that happened on the system. A power quality
problem is any problem manifested in voltage, current or frequency deviations that result in failure or
disoperation of customer equipment (Dugan et al, 1996; TNB, 1996). The term of power quality refer
to wide variety of electromagnetic phenomena characterize the voltage and current at a given time and
location on the power system (IEEE Std.1159-199, 1995). The methodology of modelling the modules,
as well as integration with the hardware components used for complete monitoring activities, is shown
in Figure 1. This system consists of the integration between the interfaces developed in LabVIEW and
the hardware of CompactRIO, host and module slot for PQ measurements.
Figure 1: Diagram of complete on-line monitoring systems
Modelling of the modules are done using the LabVIEW software and signal processing tools is
used to develop this system.
290 M. Z. A Ab Kadir, S. S. Tuan Othman and L. Lombigit
3. Modelling the modules in LabVIEW

3.1. Direct monitoring system
This direct monitoring system consists of three modules, namely the FPGA target, the CompactRIO
module and the host module. The system is designed for a single host for direct monitoring and
analysing. The FPGA is a built-in module in CompactRIO for data processing, as shown in Fig. 2.
Whilst the host, that consists of CPU and monitor, is directly connected to the CompactRIO device.
Figures 3, 4 and 5 show the block diagrams modelled in LabVIEW for monitoring system respectively.
Figure 2: CompactRIO input/output module
Figure 3: Block diagram of FPGA target.

Figure 4(a): CompactRIO block diagram I
Figure 4 (b): CompactRIO block diagram II
Figure 5: Host block diagram
3.2. Multiple monitoring systems

The modules for multiple monitoring systems are the same as for direct system. The only different is
the additional of diagram called Transmission Control Protocol/Internet Protocol (TCP/IP) port, which
has to be developed and assigned the IP address for remote monitoring. Figure 6 shows the block
diagram developed for TCP/IP function.
Figure 6: Block diagram for TCP/IP port.
For multiple monitoring, the CompactRIO must be connected to the main server. Before
accessing to the network, the computer should firstly be connected to the setting IP address. Inside the
block diagram of the host, there is a function to open TCP/IP port [4].
4. Hardware and system integration

The hardware systems consist of two main parts, which are the CompactRIO module and the host.
Figure 7 shows the CompactRIO module (with built in FPGA), power supply (240V/50Hz) and slots
for measurement. There is one slot available for temperature detector and another two slots for voltage,
current or frequency.
Figure 7: CompactRIO 9002 system.
Power Supply
CompactRIO
Figure 8 shows the host (for modelling and direct monitoring) that is connected to the
CompactRIO module.
Figure 8: Host for modelling the interfaces and direct monitoring
5. Preliminary results
Figure 9 shows the preliminary result which demonstrates the capability of the system to detect any
power quality problem i.e. undervoltage in this case, which compared the setting value and the
amplitude of the system voltage recorded. Any disturbance will be detected by the CompactRIO and
transformed into analogue waveform. This waveform later will be converted into the graph of
amplitude (V) versus time (t), which is displayed on the host monitor.
Figure 9: Initial result obtained from complete developed system
Analogue input waveform from CompactRIO
Peak detector from the host
Failure
(undervoltage)
upper value: 9.27V
Lower value: -9.26V
Legend
Setting parameters: delay (10ms), upper limit (10V), lower limit (-10V)
The process of transforming the analogue waveform of CompactRIO to the V-t graph is done
within 10 ms (delay time). As far as the monitoring is concerned, whatever the results displayed on the
host can easily be monitored elsewhere simply by accessing the IP assigned to the host.
6. Discussion and Conclusion

There are many ways to monitor power quality issues. The use of the CompactRIO module, integrated
with the interfaces built in LabVIEW environment, offer users the ability to control and monitor the
activities. Among the parameters available, amplitude was chosen for monitoring either in the case of
transient, harmonics, voltage sag. Integration between the interfaces and the hardware has produced
very good approach or method of monitoring the disturbances that occur in power system. Moreover,
with the capability of monitoring elsewhere, there are a lot of advantages in term of time saving,
manpower and overall cost of maintenance.
Acknowledgement
The authors want to express their gratitude to Malaysian Nuclear Agency for allowing the system to be
developed and tested and for the permission to use the lab and other facilities.
References
[1] Dugan, R.C., M.F. McGranaghan and H.W. Beaty, 1996. Electrical Power System Quality.
McGraw Hill, New York. ISBN-10: 0070180318
[2] Tenaga Nasional Berhad (TNB), 1996. A Guidebook on Power Quality-Monitoring, Analysis &
Mitigations Distribution Engineering, TNB Publication.
[3] IEEE, 1995. IEEE Standard.1159-1995: Recommended Practice for Monitoring Electric Power
Quality.
[4] Sokoloff, L, 2004. Application in LabVIEW, Pearson Education, New Jersey. ISBN-10:
0130161942
ISSN 1450-216X Vol.21 No.2 (2008), pp.296-304
Earnings, Market Value, and Option-Based Equity Valuation
Mohsen Dastgir
Accounting Department, College of Economics and Social Sciences
Shahid Chamran University, Ahvaz, I.R.IRAN
E-mail: dastmw@yahoo.com
Tel:+989131130587
Abdolreza Talaneh
Accounting Department, Azad University, Firoozkuh Branch, I.R.IRAN
E-mail: abdolrezatalaneh@yahoo.com
Tel: +989121011382
Abstract
This paper empirically investigates, in an emerging market, the functional form of

the relationship between earnings and stock prices under an option-based equity valuation
framework and uses in the analyses a sample data of 226 Iranian manufacturing firms listed
on Tehran Stock Exchange (TSE) for the period of 1991 to 2004. Employing after-tax
earnings and beginning-of-year equity book value, both on a per share basis and deflated by
the General Consumer Price Index, the paper provides graphical and statistical weak
evidences on the convex form of price-earnings relationship and contributes in the market-
based accounting literature.
Keywords: Equity Valuation, Earnings, Book Value, Convexity, and Adaptation Option
1. Introduction
Many theoretical and empirical researches to date have shown that the basic accounting variables
(earnings and equity book value) play a central role in equity valuation models. Such studies use
current profit or expected earnings as a proxy for dividends in those valuation models, which in turn
assume a linear relationship between accounting and market variables. Ohlson (1995) and Feltham and
Ohlson (1995; 1996) show that, assuming the clean-surplus relation and a specific linear information
dynamics, equity value can be represented as a linear function of earnings and book value.
However, recent empirical studies assume a non-linear role of either earnings or book value in
the valuation function, which depends on the levels of these two variables in the model, Burgstahler
and Dichev (1997), Zhang (2000); Hayn (1995) and Collins et al. (1999), for example. Findings by
Collins et al. (1999) indicate that, “the relative and the incremental importance of earnings versus book
value in explaining equity value vary with the level of earnings”.
The financial options’ functions follow a non-linear relationship and as such the real options’
functions follow the same behaviour. If employing the real option variable into equity valuation model
is right, then the relationships between basic accounting variables and equity values should be non-
linear. Burgstahler and Dichev (1997), Chen and Zhang (2000, a,b) show that the relationship between
earnings and book value with stock prices is non-linear and convex.
Earnings, Market Value, and Option-Based Equity Valuation 297
Holthausen and Watts (2001), argue that empirical tests often rely on valuation models that “do
not allow for growth opportunities or abandonment options”. Studies by Burgstahler and Dichev
(1997), Zhang (2000) and Saqafi and Talaneh (2006), introduce real options in their valuation
functions.
Few studies have been conducted in Iran, which empirically investigate the relationship
between earnings and book value with the stock prices. Therefore, present study is an attempt to follow
the previous studies incorporating real option studies into the valuation model in Iran. We use in our
analyses the empirical models developed by Burgstahler and Dichev (1997) to test empirically the
functional form of the relations between variables. However, as we apply the data of manufacturing
firms listed on Tehran stock Exchange in the analyses, the paper provides new evidences for the equity
valuation studies in the emerging markets.
Using the data of 226 listed firms in Tehran’s stock exchange for the period of 1991-2004 this
paper investigates the non-linear relationship between earnings and stock prices. After controlling for
book value of equity, our findings reveal that the relationship between earnings and stocks’ prices is
non-linear and convex.
The remainder of this paper is organized as follows. Section 2, discusses the relevant literature
and explains research variables with an emphasis on real options. In section 3, research method and
hypotheses are introduced. Fourth section presents descriptive statistics and tests of hypotheses.
Research findings and discussion of the findings, research limitations and suggestions for the future
research are presented in the fifth section.
2. Literature Review
In the fifties and sixties scholars such as Graham and Dod (1951), Walter (1956), Miller and
Modigiliani (1961) and Gordon (1959) developed theoretical models for equity valuation. Such models
were based on discounted cash flows (dividends) to capitalize stock price. In these studies current
profit or expected earnings were used as a proxy for dividends, and the relationship between variables
were assumed to be linear.
Studies relating equity values to balance sheet items include Landsman (1986), Barth (1991)
and Shevlin (1991). Other studies relating both balance sheet and income statements items (book value
of equity and earnings) to valuation of equity can be found in the literature, see Amir and Lev (1994),
Ohlson (1995), Feltham and Ohlson (1995), Collins et al. (1997), and Collins et al. (1999).
Despite the relative success of the previous studies, none of them have incorporated
management flexibility in their valuation models and did not consider the role of the real options in
their valuation functions. Some argue that inconsistent results with the expectations observed in studies
by Jan and Ou (1995), Hayn (1995) and Kothari and Zimmerman (1995), are due to the lack of real
option variables in their models. Burgstahler and Dichev (1997); Hayn, (1995); Wysocki, (1999) have
examined the effect of adaptation (abandonment) options on equity value, but they do not develop
formal valuation models. Kothari (2001) also argues that introducing real options in valuation model is
a right direction to equity valuation.
Real-option-based valuation models assume that if the future earnings are not promising and
the firm suffers losses, mangers have this flexibility to stop the operations and liquidate the firm. Since
owners (through their managers) have the control of abandoning the firm’s resources, selling assets
will change future earnings expectations. In an efficient market and in the absence of arbitrage, the
value of option to liquidate will affect the stock prices. Thus, adaptation option and liquidation option
will limit the downside losses of stock prices. There is a similarity between these real options and put
option.
In the last decade researchers such as Berger, Ofek and Swary (1996); Burgstahler and Dichev
(1997); and recent studies by Chen and Zhang (2000) and Saqafi and Talaneh (2006), have tried to
introduce the real options in their valuation models. The results of these studies are promising, for
example Berger, Ofek and Swary (1996) show that adaptation option (option to abandon) is understood
298 Mohsen Dastgir and Abdolreza Talaneh
by market participants and its value is reflected in the stock prices. Chen and Zhang (2000) indicate
that the Zhang model relaying on two options of expansion and contraction of the operation explains
the unexpected phenomena. Saqafi and Talaneh (2006) show that equity book value, when the level of
earnings is low or negative can play the role of adaptation option. When the option value is significant
this role is appreciated by market participants.
The financial options’ functions (call and put options) follow a non linear relationship. The real
options have analogy with financial ones; therefore, follow the same behaviour. Thus, employing real
option variable into equity valuation model causes the function to become a non linear function.
Burgstahler and Dichev (1997), Chen and Zhang (2000, a, b) show that the relationship between
earnings and book value with stock price is non linear and convex.
Based on the previous discussion, to investigate the shape of the relationship between stock
price and earnings, Burgstahler and Dichev (1997) write the equity valuation model on the weighted
average basis of recursion value and adaptation value. The weight of each variable is the likelihood of
its occurrence. If value of the recursion value is reduced compare to adaptation value, then the
probability of exercising adaptation option by management is increased and the market value is more
affected by adaptation value. Opposite is also true. When the weight of recursion value is increased
compare to adaptation value, the probability of liquidation will decrease and recursion value plays
major role in market value. In the middle situation, the market value will be affected equally by the two
values.
If we substitute the current earnings and equity book value, for recursion value and adaptation
value respectively, as the consequence the shape of the function of stock price based on the current
earnings and equity book value will be uprising and convex. Similar reasoning is also seen in Chen and
Zhang (2000).
3. Regression Model
Following Burgstahler and Dichev (1997), if current earnings and book value of equity are used as
substitute of recursion and adaptation value respectively, then the stock price will be weighted average
of current earnings and book value1 as follow:
Pt   Bt 1   Et   (1)
Where:
P; is the market value of equity (stock),E; is the earning per share, which is used as
proxy for recursion value, B; is book value per share, which is used as proxy for
adaptation value.
When earnings variable, E, compare to the book value of stock, B, is getting too small, then the
book value of stock determines the value function and  get close to zero and in the same time  get
close to the parity coefficient of stock price and equity book value. On the other hand when earnings
variable, E, compare to the book value of stock, B, is getting large,  gets close to zero and  gets
close to capitalizing factor of earnings; when earnings become main determinant of stock value. To
allow the coefficient to vary between the range of earnings and book value in the previous regression
model we divide the two side of the equation 1, by book value variable to obtain the following
equation:
Pt / B t  1     E t / B t  1  e t (2)
Regression model 2, shows the relationship of price and earnings by keeping constant the
equity book value. Although this equation is a linear equation, but we can use the β coefficient for
convexity test. By introducing dummy variable to control different ratios of earning to book value,
equation 2 changes to the following Piece-Wise Regression model:
1
In this model we used beginning of the year book value of stocks. This equation can be driven from Ohlson [1995], see
Collins, Pincus and Xie [1999].
( Pt / Bt 1 )  a1  a2 M  a3 H  b1 ( Et / Bt 1 )  b2 M ( Et / Bt 1 )  b3 H ( Et / Bt 1 )  et
(3)
Where: the dummy variables, H (high) and M (middle), are used to divide the range of earning
to book value into three partitions.
We expect from equations 1 and 2, when the earnings are low the book value coefficient (a) to
be positive and significant. Meaning that in the lower partition of earnings the (a1) coefficient is
positive and significant and gradually by moving toward the upper partition of earnings the amount of
(a) will be reduced. Therefore, we expect that in higher earnings partition the amount of (a3) to get
close or less than zero. When earnings are high and moving higher again we expect the coefficient bs
to get larger and larger. If the relation between stock values and earnings is non linear and convex, the
coefficient of upper bound earnings (b3) should be larger than the coefficient of middle bound earnings
(b2) and the coefficient of middle bound earnings should be larger than the coefficient of lower bound
earnings (b1).
4. Sample Selection
The sample of this study is selected from the companies listed in Tehran Stock exchange with four
criteria in mind:
1. Companies having fiscal year end in accordance with Iranian calendar year (Iranian
calendar year ends in 20th of March).
2. Manufacturing companies
3. Companies with shares actively traded in the market
4. Companies with positive equity values.
After enforcing the above criteria the remaining population reduced to 226 companies divided
between 22 industries. The selected sample included all the 226 companies for the period of 1991-
2004.
5. Descriptive Statistics and Model Estimates

Descriptive statistics of earnings, book value of stocks and stock prices are presented in table 1. The
Maximum amounts of the variables are too much larger than Mean and Median statistics for the same
variables because we have not eliminated outliers. In the period under study, at least 79 companies
reported loss at least in one period and overall the number of losses reported by companies under study
amounted to 121 cases, which led earnings to be negative only in the first quartile.
Table 1: Descriptive statistics
Pt Bt-1 Et
Obs. 1901 1901 1901
Mean 6829.9 4837.3 1694.9
Std. Dev. 8654.3 47648.3 3203.0
Min. 89.5 5.8 -9851.6
Q1 1836.6 915.9 286.4
Median 3989.5 1555.6 793.5
Q3 8223.2 3155.9 1912.1
Max. 94101.0 71243.9 47295.5
P = stock price, B = book value per share, E = net earning per share which is adjusted by retail price index for the effect of inflation.
6. Graphic Analysis
Before running the regression models, first scatter diagram for variables used in model 2 (P/BV and
E/BV) is drawn. This diagram more or less shows the expected convexity of the variables.
The regression line is not linear and its coefficient of determination is slightly more than 31
percent. When the relationship between stock price and earnings, controlling for the opening and
ending book value of assets was drawn the results improved and more convexity became apparent.
Figure 1: Price/Book Value versus Earnings/Book value
70
60
50
40
30
20
10
0
PBV1
-10 Rsq = 0.3117

-4 -2 0 2 4 6
EBV1
7. Estimating Regression Models

The results of regressing model 2 are reported in table 3. As can be seen in table 2, the constant
coefficients in 12 years are positive and significant. The average and pooled results are also significant.
The earnings coefficient except for 1991 in all other years in this study for average and pooled is
significant. The coefficient of determination has a value of 0 in 1991 and 86 percent in 1996. The R-
square in the average row is 46 percent and in the pooled row is 33.5 percent. These findings indicate
that predicting the model for analyzing the convexity was right. The only problem is that the constant
coefficient and the slope of the regression model is less than what was expected. The constant
coefficient of the model represents the parity of book value and market value, and the earnings
coefficient represents capitalizing earnings. Normally we expected that these coefficients to be higher
than actually calculated.
Table 2: Results for the regression of prices on earnings for a fixed amount of book value
Regression Model Pt / Bt 1   1   2 E t / Bt 1   t
Year Obs. 1 Prob. 2 Prob. Adj. R Sq.
1 2 3 4 5
1991 46 F 2.839 .000 .166 .576 -.008
1992 57 1.890 .000 .810 .054 .071
1993 80 1.411 .000 1.092 .004 .143
1994 105 .817 .060 2.407 .000 .566
1995 130 1.432 .000 2.802 .000 .399
1996 163 1.117 .000 2.585 .000 .857
1997 174 1.078 .000 1.932 .000 .571
1998 172 .531 .001 2.322 .000 .824
1999 188 1.217 .012 2.487 .000 .571
2000 187 1.014 .000 3.665 .000 .707
2001 181 .736 .119 5.697 .000 .563
2002 183 1.054 .261 8.883 .000 .481
2003 160 1.825 .013 6.015 .000 .406
2004 75 1.969 .033 3.960 .040 .372
Average 14 1.352 .000 3.202 .000 .466
Pooled 1901 1.457 .000 2.841 .000 .335
a. All regression results are based on a White's (1980) heteroskedasticity-consistent covariance matrix.
b. F shows that the regression is not significant even at 90% in terms of F statistics for goodness of fit. Other regressions are significant at 95%.
c. Probability values adjacent to coefficients are used for significance test of coefficients (two-tailed Student’s t-test). The values less than 0.05 indicate
the significance of coefficients. All significant probabilities at 95% are in bold face.
d. The “Average” row shows the average coefficients for 14 years. Probability values adjacent to coefficients are calculated based on two-tailed
Student’s t-test.
Results of regressing model 3 are presented in table 3. This table has 17 columns. The last
column shows the explanatory power of each yearly regression models, which compared to previous
model on average, is improved by 11 percent. This improvement indicates that partitioning earnings is
right thing to do, and the relationship between earning and stock value is not the same in different level
of earnings.
Modeling and Temperature Controller Design for Yazd Solar Power Plant 302
Table 3: Results for the piece-wise regression of prices on earnings for a fixed amount of book value
Adj. R
Yr Obs. 1 Prob. 1  2 2 Prob. 1 3 3 Prob. 1 Prob. 1 2 2 Prob. 1 3 3 Prob.
Sq.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
91 46 2.354 .000 2.498 .144 .892 3.825 1.471 .094 -.164 .000 .907 3.071 .149 -.081 2.083 .002 .110
92 57 2.012 .024 1.990 -.022 .987 2.029 .017 .991 .731 .799 .315 -.416 .899 .769 .038 .990 .007
93 80 1.346 .000 .828 -.518 .249 2.793 1.447 .164 2.211 .050 1.682 -.529 .710 .287 -1.924 .139 .153
94 105 1.207 .000 4.880 3.673 .347 -.203 -1.410 .023 -.651 .498 -4.169 -3.518 .561 2.593 3.244 .003 .578
95 130 2.286 .000 -1.740 -4.026 .013 1.133 -1.153 .180 -3.183 .351 9.170 12.353 .014 3.092 6.275 .072 .449
96 163 1.858 .000 1.741 -.117 .857 .599 -1.259 .000 -1.719 .000 1.342 3.061 .011 2.755 4.474 .000 .891
97 174 1.664 .000 .629 -1.035 .049 .095 -1.569 .000 -1.098 .002 2.254 3.352 .000 2.579 3.677 .000 .685
98 172 .997 .000 .490 -.507 .130 -.217 -1.214 .000 .000 .999 2.117 2.117 .006 2.800 2.800 .000 .875
99 188 1.237 .000 .038 -1.199 .000 .218 -1.019 .264 -1.055 .104 3.807 4.862 .000 3.517 4.572 .000 .803
00 187 1.381 .000 -.893 -2.274 .000 -.180 -1.561 .087 -.356 .384 6.382 6.738 .000 4.843 5.199 .000 .840
01 181 1.222 .000 -1.384 -2.606 .015 -.074 -1.296 .282 -1.532 .214 8.433 9.965 .001 6.947 8.479 .000 .675
02 183 1.085 .000 .446 -.639 .544 -6.869 -7.954 .000 -.643 .090 6.765 7.408 .045 18.156 18.799 .000 .765
03 159 1.253 .000 .071 -1.182 .086 -.607 -1.860 .498 -1.381 .004 6.041 7.422 .001 9.321 10.702 .000 .590
04 75 1.765 .000 -.460 -2.225 .007 .043 -1.722 .373 -1.479 .000 7.946 9.425 .000 6.736 8.215 .001 .619
Ave 14 1.548 .000 .652 -.895 .079 .185 -1.363 .036 -.880 .025 3.785 4.665 .002 4.594 5.474 .001 .574
Pooled 1900 1.466 .000 .524 -.942 .004 1.893 .427 .174 -.989 .000 3.970 4.959 .000 2.894 3.883 .000 .383
a. All regression results are based on a White's (1980) heteroskedasticity-consistent covariance matrix.
b. F shows that the regression is not significant even at 90% in terms of F statistics for goodness of fit. F1 shows that the regression is significant at 10% and other regressions are significant at 95%.
c. Probability values adjacent to coefficients are used for significance test of coefficients (two-tailed Student’s t-test). The values less than 0.05 indicate the significance of coefficients. All significant probabilities at
95% are in bold face.
d. The “Average” row shows the average coefficients for 14 years. Probability values adjacent to coefficients are calculated based on two-tailed Student’s t-test.
Modeling and Temperature Controller Design for Yazd Solar Power Plant 303
As was expected the coefficients a1 to a3 gradually is reduced. The average for a1 is 1.548, for
a2 is 0.652 and for a3 is 0.185. These coefficients show that, when level of earnings is low, based on
our expectation, share book value has higher weight than stock value.
Findings for the earnings coefficients also more or less are according to expectations. In lower
levels of earnings the earning coefficient (calculated -0.880) has lower weight in determining stock
value. With increasing earnings levels, the average of earnings coefficients are calculated 3.785 and
4.594 respectively for middle and higher level of earnings.
8. Conclusion and Discussion

Present study only focuses on the shape of the relationship between earnings and share value and does
not investigate the relationship between book value and share value and is highly based on the
methodology presented by Burgstahler and Dichev (1997).
The paper studies the relationship between earnings and stock value after controlling for stock
book value. Analysis shows that the shape of the relationship between earnings and stock value after
controlling for stock book value, can be non linear and convex. This study provides evidence, although
not very strong, that the above relationship is non linear and convex.
Future studies may shed more lights on the issue and provide supporting evidence for present
study. Findings of this paper has this message that linear equity valuation model may not be a good
equity valuation model. Zhang (1999) argues that, in general, equity value is convex in both earnings
and book value. Thus, non linear equity valuation models based on real options are new avenue for
future research.
References
[1] Amir, E., and B., Lev, 1996, Value-relevance of non-financial information: the wireless
communications industry, Journal of Accounting and Economics 22, 3-30.
[2] Barth, M., 1991, Relative measurement errors among alternative pension asset and liability
measures, The Accounting Review 66, 433-463.
[3] Berger, P, Ofek E. and Swary I. 1996. Investor Valuation of the Abandonment Option. Journal
of Financial Economics 42: 257-287.
[4] Bergestahler, D, and Dichev. I.1997. Earnings, Adaptation and Equity Value. The Accounting
Review 72: 187-215.
[5] Chen, P. and Zhang G. 2002a. Profitability, Earnings and book Value in Equity Valuation: A
Geometric View and Empirical Evidence. Working Paper, Hong Kong University of Science &
Technology.
[6] Chen, P. and Zhang G. 2002b. The Role of Earnings and Book Value in Equity Valuation: A
Real options Based Analysis. Working Paper, Hong Kong University of Science & Technology.
[7] Collins, D., E. Maydew and I. Weiss. Changes in the Value-Relevance of Earnings and Book
Values over the Past Forty Years." Journal of Accounting and Economics 24 (1997): 39-67.
[8] Collins D, Pincus M.and Xie H. 1999. Equity Valuation and Negative Earnings. The
Accounting Review (January): 29-61.
[9] Feltham, G., and J. Ohlson. 1995. Valuation and Clean Surplus Accounting for Operating and
Financial Activities. Contemproray Accounting Research, 11: 689-731
[10] Gordon, Merton J. 1959. Dividends, Earnings, and Stock Prices, Review of Economic and
Statistics, (May): 99-105
[11] Hayn, C. 1995. The Information Content of Losses. Journal of Accounting and Economics 20:
125-153.
[12] Jan, C. L. and Ou J. 1995. The Role of Negative Earnings in the Valuation of Equity Stocks.
Working paper, New York University and Santa Clara University.
[13] Kothari, S. P. 2001. Capital Market Research in Accounting. Journal of Accounting and
Economics 31: 105-231.
[14] Kothari, S. P., and Zimmerman. J. 1995. Price and Return Models. Journal of Accounting and
Economics 20:155-192.
[15] Landsman, W., 1986, An Empirical Investigation of Pension Fund Property Rights, The
Accounting Review 61, 662-692.
[16] Miller Merton H., and Franco Modigliani. 1961. Dividend Policy, growth, and Valuation of
Shares, journal of Business,Vol.34 (Oct.), 411-433.
[17] Ohlson, J. A. 1995. Earnings, Book values, and Dividends in Security Valuation. Contemporary
Accounting Research 11: 661-687.
[18] Saqafi, Ali, and Abdolreza Talaneh. 2006. The Role of Earnings, Book Value, and the
Abandonment Option in Equity Valuation. The Iranian Accounting and Auditing Review (Farsi
Version), No. 34. pp 3-38.
[19] Schwartz, E. S. and Trigeogis L. 2001. Real Options and Investments under Uncertainty. The
MIT Press, London, England.
[20] Shevlin, T., 1991, The valuation of R & D firms with R & D limited partnership, The
Accounting Review 66, 1-21.
[21] Wysocki, P., 1999, Real options and informativeness of segment disclosures, Working paper,
University of Michigan.
[22] White, H. 1980. A Heteroscedasticity Consistent Covariance Matrix Estimator and a Direct
Test for Heteroscedasticity. Econometrica 48: 817-838.
[23] Wright, F. K. 1967. Adaptation and the Assets Measurement Problem. Abacus (August): 74-79.
[24] Zhang, G. 2000. Accounting Information, Capital Investment Decisions, and Equity Valuation:
Theory and Empirical Implications. Journal of accounting Research 38: 271-295
ISSN 1450-216X Vol.21 No.2 (2008), pp.305-313
Regime Alimentaire du Poisson-Chat Schilbe Mandibularis

(Günther, 1867) Dans Deux Bassins Fluviaux de
Cote D’ivoire (Bia et Agnebi)
Doumbia Lassina
Université d’Abobo-Adjamé (UFR-SGE)
Laboratoire d’Environnement et de Biologie Aquatique
02 B.P. 801 Abidjan 02, République de Côte d’Ivoire
E-mail: doumbia@uabobo.ci
Tel: (225) 20 30 42 62, (225) 20 30 42 59, (225) 07 92 80 58
Ouattara Allassane
E-mail: ou dflassina@yahoo.fr
Tel: (225) 20 30 42 62, (225) 20 30 42 59, (225) 07 92 80 58
Gourene Germain
Tel: (225) 20 30 42 62, (225) 20 30 42 59, (225) 07 92 80 58
Abstract
This survey carried on the diet of the catfish Schilbe mandibularis in the Bia river
(with two hydroelectric dams) and the Agnébi river (without hydroelectric dam). 726
specimens coming from stations localized in the upper, middle and lower streams of the
two rivers of sampling have been fished with the help of nets maillants during two
consecutives years. The diet has been studied according to a gradient longitudinal uphill-
downstream, the climatic season and the size of individuals. The characterization of the
food habits of this fish reveals that the diet is opportuniste omnivorous type. The global
results show that whatever is the river considered the food regime of Schilbe mandibularis
is composed mainly of four trophiques categories (insects, fishes, macrophytes, animals
and plants remnants). The different groups of preys don't vary following meaningfully the
season, the size of fish and stations of sampling.
Keywords: Feeding, Schilbe mandibularis, Rivers, Bia, Agnébi, Ivory Coast.
Introduction
Les Schilbeidae constituent l’un des groupes de poissons des eaux douces et saumâtres les plus
importants économiquement au monde (Teugels, 1996). En effet, le Schilbeidae Schilbe mandibularis,
Bassins Fluviaux de Cote D’ivoire (Bia et Agnebi) 306
couramment rencontré dans toute la région Ouest africaine (Lévêque et al., 2003), est une des espèces
les plus prisées et appréciées par les populations de cette région (Vanga, 1994).
En Côte d’Ivoire, Schilbe mandibularis est fréquemment pêchée dans les rivières et les lacs de
barrages hydroélectriques. Malgré l’importance économique de cette espèce, son écologie alimentaire
reste encore peu connue. Quelques observations sur le comportement alimentaire de S. mandibularis
ont été faites par Vidy (1976) et Sangaré (1977). Cependant, ces études ont abordé de façon très
parcellaire l’alimentation de l’espèce.
Ce travail a donc pour but d’étudier le régime alimentaire de S. mandibularis en tenant compte
de facteurs de variabilité (saison, taille des poissons), et le comportement trophique de l’espèce le long
d’une rivière modifiée par l’édification de deux barrages hydroélectrique (Bia) et d’une rivière sans
aménagement (Agnébi). Notons que la présence de ces barrages sur la Bia a engendré une répartition
en trois zones écologiques: le cours supérieur, le lac d’Ayamé et le cours inférieur.
Materiel et Methodes
Sites d’étude
L’étude du régime alimentaire est basée sur des échantillonnages ichtyologiques effectués en trois sites
caractéristiques (cours supérieur, moyen et inférieur) des rivières Bia et Agnébi (Figure 1). Il s’agit de
l’amont à l’aval de la rivière Bia des stations de Bianouan, du lac d’Ayamé I et d’Aboisso et sur la
rivière Agnébi des stations de Gbessé, du Pont-autoroute et d’Armébé.
Les deux rivières échantillonnées développent leurs cours dans la partie sud de la Côte d’Ivoire.
Le climat de cette région est caractérisé par deux saisons sèches (décembre à mars et août à septembre)
et par deux saisons des pluies (avril à juillet et octobre à novembre).
Figure 1: Situation géographique des stations d’échantillonnage (●) sur les rivières Bia et Agnébi (d’après
Gourène et al., 1999).
307 Doumbia Lassina, Ouattara Allassane and Gourene Germain
Echantillonnage ichtyologique et analyse des contenus stomacaux

Les pêches mensuelles ont été faites pendant 24 mois aux stations définies sur les deux rivières. Les
poissons ont été collectés avec deux batteries de 10 filets maillants (longueur: 30 m, hauteur: 1,5 m) de
mailles croissantes entre 10 et 50 mm de vide de maille. Chaque poisson a été identifié d’après
Lévêque et al. (1992), mesuré, pesé et disséqué. L’estomac a été prélevé et conservé dans du formol à
5 %. Au laboratoire, l’observation des contenus stomacaux s’est effectuée sous une loupe binoculaire.
Le matériel ingéré a été dénombré et quantifié.
La caractérisation du régime a été faite à partir des pourcentages suivants:
 Pourcentage d’occurrence (Hyslop, 1980): F = Nie/Net x 100 avec
Nei: nombre d’estomacs contenant l’item i et Net: nombre d’estomacs pleins examinés;
 Pourcentage numérique (Hureau, 1970): N = ni/Nt x 100 avec
ni: le nombre total d’une catégorie de proies i et Nt; le nombre total des proies;
 Pourcentage pondéral (Gherbi-Barré, 1983): W = Wi/Wt x 100 avec
Wi: poids total de l’item i et Wt: poids total des proies.
Les classes de tailles des spécimens analysés ont été définies sur la base de la taille de première
maturité sexuelle (123 mm de longueur standard) déterminée par Ouattara (2000). Ainsi, les spécimens
de Schilbe mandibularis ont été classés en deux groupes: les juvéniles (longueur standard inférieure à
123 mm) et les adultes (longueur standard supérieure ou égale à 123 mm).
La comparaison des régimes alimentaires suivants les saisons climatiques, les stations
d’échantillonnage et la taille des individus a été faite à partir de l’analyse de variances (ANOVA)
réalisée en tenant compte du pourcentage pondéral des proies. Selon Macdonald et Green (1983), ce
paramètre permet de mieux évaluer l’apport nutritionnel des proies. Le seuil de significativité
statistique est de 5 %.
Resultats
Caractérisation générale du régime
Au total, 726 estomacs de Schilbe mandibularis contenant de la nourriture ont été analysés. Parmi
ceux-ci, 617 proviennent de la rivière Bia et 109 sont issus de la rivière Agnébi. Le contenu stomacal
est composé de fractions animal et végétal. La fraction animale renferme des invertébrés, des vertébrés
et les débris animaux. La fraction végétale est constituée par les fruits et les débris végétaux. Les
différents pourcentages calculés (occurrence, numérique, pondéral) figurent dans le tableau I.
L’analyse des contenus stomacaux a révélé que les débris animaux (plus de 45 %), les insectes
(environ 41 %) et les macrophytes (environ 23 %) constituent les aliments les plus fréquents dans le
régime de S. mandibularis dans les rivières Bia et Agnébi. Parmi les insectes, les coléoptères et les
hyménoptères (plus de 16 % chacun) sont les ordres les plus fréquemment observés. Les autres ordres
d’insectes ainsi que les autres catégories de proies (annélides, myriapodes, arachnides, crustacés,
mollusques, amphibiens, poissons) ont un faible pourcentage d’occurrence (entre 0,10 et 3,34 %).
En ce qui concerne les pourcentages numériques, les insectes (60,80 %) et les macrophytes
(20,28 %) sont les aliments les plus nombreux dans le régime de S. mandibularis, dans la rivière Bia.
Les hyménoptères (30,46 %) et les diptères (12,68 %) sont les insectes les mieux représentés. Les
autres items alimentaires (annélides, myriapodes, arachnides, crustacés, mollusques, amphibiens,
poissons) sont peu nombreux dans les contenus stomacaux (moins de 1 %). Dans la rivière Agnébi, en
revanche, les macrophytes sont les aliments les plus consommés (71,73 %) suivis des insectes (14,59
%).
L’estimation du poids des différents aliments consommés montre que les macrophytes, les
poissons et les insectes constituent une part importante de l’alimentation de S. mandibularis dans la
rivière Bia (respectivement de 32,81 %, 20,14 % et 19,53 %) et dans la rivière Agnébi (respectivement
42,69 %; 30,14 % et 10,28 %).
L’ANOVA ne montre pas de différence significative de régime alimentaire entre les deux
rivières (p > 0,05).
Tableau I: Pourcentages d’occurrence (F), numérique (N) et pondéral (P) des aliments consommés par les
spécimens de Schilbe mandibularis dans les rivières Bia et Agnébi. n=nombre d’estomacs contenant
des proies.
Aliments Rivière Bia n = 617 Rivière Agnébi n = 109

Invertébrés F N P F N P
Insectes
Coléoptères 17,99 8,63 4,21 16,51 2,94 1,77
Diptères 8,43 12,68 6,77 3,67 0,51 0,32
Ephéméroptères 4,70 1,25 0,65 3,67 0,41 1,55
Hémiptères 0,97 0,28 0,07 0,92 0,10 0,21
Hyménoptères 16,86 30,46 1,74 18,35 8,00 1,88
Odonates 1,62 0,34 0,19 2,75 0,30 0,98
Orthoptères 5,19 6,15 5,15 10,09 1,42 2,09
Trichoptères 1,62 0,31 0,04 3,67 0,51 0,09
Autres insects 1,78 0,68 0,71 3,67 0,41 1,38
Annélides 1,78 0,40 6,87 0,92 0,10 0,0001
Myriapodes 2,11 0,63 0,67 3,67 0,51 0,19
Arachnides 0,81 0,14 0,10 3,67 0,41 0,42
Crustacés 0,97 0,28 0,29 3,67 3,34 1,50
Mollusques 1,30 0,23 1,11 0,92 0,10 3,27
Vertébrés
Amphibiens 0,16 0,03 0,37 1,83 0,20 1,18
Poissons 5,35 0,97 20,14 13,76 2,43 30,14
Macrophytes
Fruits 10,53 17,78 30,55 12,84 70,42 42,30
Débris végétaux 13,13 2,51 2,26 11,93 1,32 0,39
Débris animaux 55,59 16,24 18,12 46,79 6,59 10,34
Variation saisonnière du régime alimentaire

Le régime alimentaire en fonction de la saison climatique est étudié au moyen de valeurs du
pourcentage pondéral des items alimentaires spécifiques en excluant les débris animaux dont les
origines sont diverses et souvent difficiles à apprécier.
Le tableau II présente le pourcentage pondéral (P) des différentes proies consommées par S.
mandibularis dans les stations d’échantillonnage des rivières Bia et Agnébi.
Dans le cours supérieur de la rivière Bia, la majorité des aliments consommés en terme de
biomasse est constituée en saison sèche de macrophytes (39,92 %) et de poissons (33,44 %). Dans le
lac d’Ayamé, à cette même période, S. mandibularis se nourrit principalement de poissons (72,23 %).
En revanche, dans le cours inférieur la part des macrophytes (76,91 %) est très importante dans
l’alimentation de l’espèce.
En saison des pluies, S. mandibularis se nourrit, pour une grande part, de macrophytes (31,83
%) dans le cours supérieur, de poissons (45,46 %) dans le lac d’Ayamé, d’insectes (43,24 %) et de
macrophytes (33,27 %) dans le cours inférieur.
Comparativement à la rivière Bia, on observe également des variations dans les proportions des
différentes proies consommées le long de la rivière Agnébi aux deux saisons climatiques (Tableau II).
En saison sèche, S. mandibularis consomme beaucoup plus d’insectes (85,17 %) dans le cours
supérieur. En revanche, dans le cours moyen, les poissons (46,61 %) et les macrophytes (41,29 %)
couvrent une part importante du régime alimentaire. Dans le cours inférieur, ce sont les macrophytes
qui sont consommés majoritairement (jusqu’à 87,68 %).
En saison des pluies, dans le cours supérieur de l’Agnébi, les poissons (35,45 %) et les insectes
(28,89 %) représentent la majorité des proies consommées. Dans le cours moyen, on retrouve comme
en saison sèche, les macrophytes (49,76 %) et les poissons (20,96 %) comme ressources alimentaires
les plus importantes. Dans le cours inférieur, S. mandibularis se nourrit majoritairement de
macrophytes (29,90 %).
L’ANOVA ne montre pas de différence significative de régime alimentaire saisonnier (p >
0,05) entre les stations aux deux rivières.
Tableau II:Variations saisonnières du pourcentage pondéral des contenus stomacaux de Schilbe mandibularis
selon les stations d’échantillonnage des rivières Bia et Agnébi: nSs: nombre d’estomacs de
spécimens pêchés en saison sèche; nSp: nombre d’estomacs de spécimens pêchés en saison des
pluies.
Rivières Stations Saisons sèches Saisons des pluies

Cours supérieur Macrophytes 39,92 % Macrophytes 31,83 %
nSs = 197 Poissons 33,44 % Insectes 17,54 %
nSp = 119
Lac d’Ayamé Poissons 72,23 % Poissons 45,46 %
Bia nSs = 66 Insectes 19,20 %
nSp = 72
Cours inférieur Macrophytes 76,91 % Insectes 43,24 %
nSs = 66 Macrophytes 33,27 %
nSp = 97
Cours supérieur Insectes 85,17 % Poissons 35,45 %
nSs = 13 Insectes 28,89 %
nSp = 9
Cours moyen Poissons 46,61 % Macrophytes 49,76 %
Agnébi nSs = 24 Macrophytes 41,29 % Poissons 20,96 %
nSp = 48
Cours inférieur Macrophytes 87,68 % Macrophytes 29,90 %
nSs = 7
nSp = 8
Variation du régime alimentaire selon la taille des spécimens

Tout comme l’étude saisonnière, le régime alimentaire suivant la taille des spécimens s’est fait sur la
base du pourcentage pondéral des items alimentaires spécifiques (Tableau III).
Les juvéniles de S. mandibularis dans la rivière Bia se nourrissent principalement d’annélides
(24,27 %) et de poissons (22,46 %) dans le cours supérieur. Dans le lac d’Ayamé les aliments
préférentiels sont les poissons (25,59 %) et les insectes (25,59 %). Dans le cours inférieur, les aliments
consommés majoritairement sont les insectes (32,98 %).
Quant aux adultes, leur alimentation est dominée par les macrophytes (39,39 %) et les poissons
(16,13 %) dans le cours supérieur. Les poissons (64,04 %) constituent l’essentiel des proies au lac
d’Ayamé. Les macrophytes (50,76 %) et les insectes (29,48 %) sont abondants dans les contenus
stomacaux des spécimens capturés dans le cours inférieur.
Dans la rivière Agnébi, les juvéniles de S. mandibularis exploitent plus ou moins les mêmes
ressources trophiques que dans la rivière Bia. Dans le cours supérieur, le seul spécimen juvénile pêché
s’est nourrit de crustacés (100 %). En revanche, les poissons (19,90 %) dominent les contenus
stomacaux des juvéniles capturés dans le cours moyen. Au cours inférieur, les insectes (11,63 %) sont
abondants. Dans le cours supérieur de cette même rivière, les individus adultes ont un régime
alimentaire basé sur les insectes (49,99 %) et les poissons (29,24 %). Ce sont les macrophytes (47,73
%) et les poissons (32,62 %) qui dominent dans le cours moyen. Les macrophytes (67,17 %) sont
prépondérants dans le cours inférieur.
L’ANOVA réalisée ne fait pas état de différences significatives de régime alimentaire entre les
juvéniles et les adultes aux différentes stations de l’Agnébi (p > 0,05).
Tableau III: Variations du pourcentage pondéral des contenus stomacaux des juvéniles et des adultes de
Schilbe mandibularis dans les rivières Bia et Agnébi. nj: nombre d’estomacs de spécimens
juvéniles; nad: nombre d’estomacs de spécimens adultes.
Rivières Stations Juvéniles (LS<123 mm) Adultes (LS>123 mm)

Cours supérieur Annélides 24,27 % Macrophytes 39,39 %
nj =126 Poissons 22,46 % Poissons 16,13 %
nad = 190 Insectes 15,25 %
Lac d’Ayamé Poissons 25,59 % Poissons 64,04 %
Bia nj =62 Insectes 25,59 %
nad =76
Cours inférieur Insectes 32,98 % Macrophytes 50,76 %
nj = 22 Insectes 29,48 %
nad = 141
Cours supérieur Crustacés 100 % Insectes 49,99 %
nj = 1 Poissons 29,24 %
nad = 21
Cours moyen Poissons 19,90 % Macrophytes 47,73 %
Agnébi nj = 17 Poissons 32,62 %
nad = 55
Cours inférieur Insectes 11,63 % Macrophytes 67,17 %
nj = 4
nad = 11
Variation du régime alimentaire en fonction des stations

La variation du régime alimentaire de Schilbe mandibularis en fonction des stations s’est faite sur la
base de la biomasse (% pondéral) des aliments spécifiés. Le régime alimentaire varie peu d’une station
à l’autre avec une forte prédominance des macrophytes à quatre stations sur l’ensemble des six milieux
d’étude (Tableau IV). En effet, les macrophytes prédominent dans les cours supérieur et inférieur de la
Bia avec respectivement 34,30 % et 49,76 %. Il en est respectivement de même dans les cours moyen
et inférieur de l’Agnébi où l’on enregistre 46,06 % et 39,63 %. Les poissons (58,14 %) dominent le
régime alimentaire S. mandibularis au lac d’Ayamé. En revanche, au cours supérieur de l’Agnébi,
l’espèce est insectivore (44,16 %). L’ANOVA n’indique toutefois pas de différences significatives
entre les différentes stations d’échantillonnage (p > 0,05).
Tableau IV: Variations du pourcentage pondéral des contenus stomacaux de Schilbe mandibularis selon le
gradient longitudinal amont-aval; P: Pourcentage pondéral; n: nombre d’estomacs contenant de la
nourriture.
Rivières Stations Aliments (% pondéral)

Cours supérieur Macrophytes 34,30 %
n = 316 Poissons 17,05 %
Lac d’Ayamé Poissons 58,14 %
Bia
n = 138 Insectes 13%
Cours inférieur Macrophytes 49,76 %
n = 163 Insectes 29,58 %
Cours supérieur Insectes 44,16 %
n = 22 Poissons 25,83 %
Cours moyen Macrophytes 46,06 %
Agnébi
n = 72 Poissons 32,16 %
Cours inférieur Macrophytes 39,63 %
n = 15
Discussion
L’analyse du régime alimentaire de Schilbe mandibularis dans les rivières Bia et Agnébi a révélé que
les insectes, les macrophytes et les poissons constituent les aliments préférentiels. Cette diversité de
proies montre que l’espèce étudiée a un régime éclectique comme déjà signalé par Vidy (1976) et
Lévêque et al. (1988) qui ont travaillé sur Eutropius mentalis (synonyme de Schilbe mandibularis). Les
proportions des différentes proies consommées varient d’une station à l’autre et d’une rivière à l’autre.
Dans le lac d’Ayamé, Schilbe mandibularis consomme plus de poissons que dans les autres stations
des deux rivières. Cette tendance à l’ichtyophagie s’expliquerait par le fait que dans le lac d’Ayamé, S.
mandibularis y trouve une forte densité de poissons de petite taille notamment le CHARACIDAE
Bricinus sp. En effet, cette ressource alimentaire est plus importante en saison sèche (environ 72 %)
qu’en saison de pluies (environ 45 %). En revanche, dans les parties fluviatiles des rivières Bia et
Agnébi, le régime est à base d’insectes terrestres, de macrophytes et parfois de poissons. Les deux
premières proies proviennent du couvert végétal qui surplombe les cours d’eau. Elles sont très
importantes dans les contenus stomacaux de S. mandibularis en saison des pluies. Dans ces milieux, les
insectes aquatiques sont très peu représentés dans la nourriture de S. mandibularis en saison sèche.
C’est plutôt en saison de pluies que ces proies contribuent le plus à l’alimentation de ce poisson. Selon
Dietoa (2003), la reproduction de nombreux insectes à lieu pendant cette dernière saison, ce qui
expliquerait leur prépondérance dans le milieu et aussi dans les contenus stomacaux de S.
mandibularis. De telles observations ont été faites par N’Guessan et al. (2006) sur le régime
alimentaire d’Heterobranchus isopterus dans la rivière Bia. De même Dietoa et al. (2007) ont obtenu
des résultats similaires chez Brycinus imberi dans le même bassin. Contrairement aux milieux lotiques,
les spécimens de S. mandibularis du lac d’Ayamé ne changent pas véritablement de régime d’une
saison à l’autre. L’absence de canopée au lac est une situation qui distingue en partie l’alimentation des
spécimens de la retenue de ceux des autres stations. Les résultats font apparaître une faible variation du
régime entre les populations juvéniles et adultes. Par ailleurs, les spécimens juvéniles et adultes ne
consomment pas les mêmes proportions de proies aux trois stations sur l’Agnébi dans l’ensemble et au
cours inférieur sur la Bia. Les juvéniles se nourrissent d’insectes et parfois de poissons. Quant aux
adultes, ils consomment majoritairement les macrophytes et les poissons. Toutefois, l’on observe une
similarité de régime entre les juvéniles et les adultes dans le lac d’Ayamé et dans le cours supérieur de
la Bia. Une observation similaire a été faite par Lévêque et al. (1988). Ces auteurs indiquent que S.
mandibularis est un poisson opportuniste. De plus, ceci pourrait être dû au fait que les ressources
trophiques consommées sont assez disponibles.
Conclusion
Cette étude a permis de montrer que le poisson chat Schilbe mandibularis se nourrit d’une diversité de
proies disponibles dans le milieu. Son régime alimentaire est de ce fait de type omnivore opportuniste.
Concernant les variations du régime en fonction des stations d’échantillonnage, des saisons climatiques
et de la taille des populations étudiées, des différences n’ont pu être notées. Une approche plus
approfondie dans un système sémi-contrôlé pourrait sans doute permettre de préciser ces résultats.
Remerciements
Cette étude fait partie du projet VL.I.R. / K.U.L. (Vlaamse Interuniversitaire Raad) « Evolution de la
Biodiversité des poissons après la construction d’un barrage hydroélectrique: cas de la rivière Bia en
Côte d’Ivoire », financé par l’Agence Générale pour la Coopération au Développement de la Belgique.
Nous tenons à remercier le promoteur Thys Van Den Audenaerde D.F.E. Nous saluons la mémoire du
co-promoteur Teugels G. G.
References
[1] Dietoa, Y. M., 2003. Entomofaune et stratégies alimentaires des poissons du genre Brycinus
(CHARACIDAE) en milieux fluviatiles et lacustres (Bassins Bia et Agnébi; Côte d’Ivoire).
Thèse de doctorat de l’Université d’Abobo-Adjamé, Côte d’Ivoire, 261 p.
[2] Dietoa, Y., M., G. Gourène, G., & Ouattara, A., 2007 Habitudes alimentaires de Brycinus
longipinnis dans le complexe fluvio-lacustre de la Bia, Côte d’Ivoire. Belgian journal of
zoology, 137 (1): 3-9.
[3] Gherbi-Barré, A., 1983. Biologie de Trisopterus luscus –Linné, 1758) Gadidae de la Baie des
Douarnemez (reproduction, croissance, régime alimentaire). Thèse de 3ème cycle, Université de
Bretagne Occidental.
[4] Hureau, J. C., 1970. Biologie comparée de quelques poissons antartiques (Nototheniidae).
Bull. Inst. Océanogr.; Monaco, 68, 139: 1-250.
[5] Hyslop, E. J., 1980. Stomach contents analysis, a review of methods and their application.
Journal of Fish Biology, 17: 411-429.
[6] Lévêque, C., Bruton, L. M. N. & Ssetongo, G. W., 1988. Biologie et écologie des poissons
d’eau douce africains. Edition de l’ORSTOM, Paris, 508 p.
[7] Lévêque, C., Paugy, D. & Teugels, G. G., (Eds.), 1992. Faune des poissons d'eau douce et
saumâtre de l'Afrique de l'ouest. Edition ORSTOM / MRAC, Tome II: 385-902.
[8] Macdonald, J. S. & Green, R. H., 1983. Redundancy of variables used to describe importance
of prey species in fish diets. Can. J. Fish. Aquat. Sci. 40: 635-637.
[9] N’guessan, Y. S., Ouattara, A. & Gourène, G., 2006. Régime alimentaire du poisson-chat
Heterobranchus isopterus Bleeker, 1863 (Clariidae) dans un hydrosystème fluvio-lacustre ouest
africain (rivière Bia, Côte d’Ivoire) Sciences & Nature Vol. 3 N° 1: 83-90
[10] Ouattara, M., 2000. Stratégie de reproduction et réactions aux pressions de l’environnement
chez le poisson africain Mormyrops anguilloides et Marcusenius ussheri (Mormyridae), Schilbe
mandibularis et S. intermedius (Schilbeidae) (Bassins Bia et Agnébi; Côte d’Ivoire). Thèse de
doctorat de l’Université d’Abobo-Adjamé, Côte d’Ivoire, 266 p.
[11] Sangaré, Y., 1977. Données biologiques et écologiques de Eutropius mentalis. Institut
Universitaire d'Ecologie Tropicale. Université Nationale de Côte d'Ivoire, 85 p.
[12] Teugels, G. G., 1996. Taxonomy, phylogeny and biogeography of catfishes (Ostariophysi,
Siluroidei): an overview. LEGENDRE, M. et PROTEAU, J. P., (Eds.) ORSTOM, CIRAD,
Aquat. living Resour., 9, suppl.: 9 – 34.
[13] Vanga, A. F., 1994. La pêche au barrage d'Ayamé. Mémoire de D.E.A. Département de
Sociologie, Université Nationale de Côte d’Ivoire, 50 p.
[14] Vidy, G., 1976. Etude du régime alimentaire de quelques poissons insectivores dans les rivières
de Côte d’Ivoire. Rapport ORSTOM 2, 30 p.
ISSN 1450-216X Vol.21 No.2 (2008), pp.314-318
Weights Adjustment Neural Networks
Al-Salaimeh Safwan
Irbid National University, Faculty of Information Technology, Irbid
The Hashemiet Kingdom of Jordan
E-mail: safwan_71@yahoo.com
Hanna Jaber
Computer and Automatic Control Department, Teshreen University, Latakia, Syria
E-mail: jabrhanna@yahoo.com
Abstract
The study aims at knowing the neural networks, their components, neuron cells and
their mechanism. One of these networks types is the most common in digital systems,
Perception that is Perceptron network or as called touch networks. An equation of how
these networks work was studied in order to design new system.
Keywords: Weights adjustment, neural networks, training category, Perceptron network,

digital systems, neuron cells, transformation.
Introduction
One million neural cells spread in human body and divide themselves into million free nerve endings.
These cells bring touch and reactions from and to human brain through spinal cord. Through the cells,
information bout outside world has been saved in the human brain by weights adjustment inside the
cells.
Scientists were thinking of a way to be able to imitate the process that happens in the human
brain. They found that the science of neural networks is under Artificial Intelligence Science in which
computers become intelligent machines. These machines acquire information as same as human beings
in terms of the way of weights adjustment during learning [1].
The Purpose of the Study

The study aims at knowing the neural networks, their components, neuron cells and their mechanism.
One of these networks types is the most common in digital systems, Perception that is Perceptron
network or as called touch networks. An equation of how these networks work was studied in order to
design new system.
Artificial Neural Networks

A machine designed to imitate the way in which the human brain performs certain function. A huge
processor distributes side by side. It consists of simple processed units that save the practical
information to be available for the user through weights adjustment.
Weights Adjustment Neural Networks 315
Imitation Ways
There are many ways of imitation regarded the way in which human learns that are:
1. Through pantered as input then certain mathematical processes were made in which weights
were adjusted to be desired result. This way is unpractical and always used in order to clarify
the concept of neural network's work only.
2. Through many persons, each of who has simple calculator in which each one represents a
neural cell doing weights. This way is not efficient for the above same reasons.
3. Through numerous computers that connected to each other in which each one represents neural
cell doing simple mathematical operations to adjust the weights. This way is not efficient due to
a very large number of computers that are often very expensivper and pen in which the
characteristics were ee [1,2].
The Applications for the Neural Networks' Use

The neural networks gave the high-qualified solutions for many applications in many fields. Some of
which are mentioned:
1. Recognizing patterns and realizing pictures.
2. Ability to realize the disfigured pictures.
3. Completing the pictures whose part is lost as sending pictures through satellites.
4. Classification processes of a number of groups, for example, classifying animals into tame and
predatory animals.
For instance the process of realizing the creatures, if the program is prepared in a traditional
way to realize these animals, it will be very difficult in addition to determined capacities. The
traditional applications have many stages that mostly require the existence of human being and giant
program to realize each animal independently.
While in the neural networks, the network follows the same way by which human learns. That
is by showing animals' pictures, weights adjustment in which the information has been saved in a
correct way in the computer's memory. The network learns through the repetition of pictures and their
variation to be able to give correct answers. This requires preparing a giant program as in the
traditional program [2,3].
The Components of Neural Networks

Figure (1) illustrates these components.
Figure 1: Components of Neural Networks

316 Al-Salaimeh Safwan and Hanna Jaber
The human being has input units connected to outside world that are five senses. These as same
as the neural networks need input units and processors in which mathematical operations are madder to
adjust the weights and so a suitable reaction for each one of the networks input.
The input units contain a layer called input layer and processing units constituting a processing
layer that result the network products. Within each layer, there is a layer of interjoint that link each
layer and follow one in which weights are adjusted regarded each interjoint. The network includes one
layer of input units but it includes more than one processing layer.
The Ways of Neural Networks' Learning

The network learns through giving a group of examples that should be selected carefully because this
will contribute in the speed of the network's learning. This group of examples is called training
category. The ways of neural network learning divide into two parts according to training category
shown on the network.
Both are:
 Supervised Learning:
In this way, the training category shown on the network consisting of two pairs of sides, one is
input. The input consists of entered values of network. The output consists of values that the
network should make.
Input: (01010001)
Output :( 011)
 Unsupervised Learning
In this way, the training category contains input only without showing the output on the
network.
 (Perceptron Networks)OR (Perceptron)
After knowing the nature of perceptron network and their aim, one of ways to form the neural
network will explain below.
The perceptron network is used in the applications that need pattern classification into two
categories only.
This means that the patterns are isolated figurally in straight line. This kind of pattern is Linear
Separability Patterns.
This network consists of input layer and one or two layers of processing in which do not exceed
two. Moreover, the weights are adjusted of one only of interjoint layers that link between the previous
layers to be the other layer (if available) has stable weights. This network learns through supervised
learning. There are advantages and disadvantages of this network, one of advantages is preparing very
easy program. Some disadvantages are:
1. It cannot classify the pattern for more two categories.
2. The number of processing layers is determined in one or two layers only.
3. The weights are adjusted through one of inter-joints layers.
The stages of this learning for these neural networks are two that: A. Learning Stage & B.
Examination Stage
A. Learning Stage
This stage includes that inter-joints' weights are adjusted to the extent of weights which are able to give
correct answers. This has done through processing units doing three main processes:
Weights Adjustment Neural Networks 317
Weighted Sum Process

Each processing unit does the sum process for each input weight and added to inter-joint that link
between it and the existing unit in the previous layer multiplicand the resultant value of this unit as in
the formula:
SJ=∑ai wji
where wj is a weight added to inter-joint that links the processing unit (j) and the unit (i) that is existing
in the layer (ai). That is output value from the unit (i). SJ is the result of the sum for each processing
unit.
Transformation
This process has done in the last one of processing layers in which the result of the mentioned sum
process is transformed in the previous process to one of values that is proposed within the result of
desired network. For example, if the network was learning how to classify the numbers into odd and
double. Each odd number has a value of (0) and each double number has a value of (1).
The value of SJ is a result of sum often not the value of (0) or(1). So, this result should be
transformed to one of two values through transformation rule that is determined by programmer. For
instance, the rule is to be as followed:
If SJ>0 then XJ=1
If SJ<=0 then XJ=0
XJ is the result value from processing unit (J).
Weights Adjustment
After transformation process, the result that is from the network is compared with the correct result that
is proposed from the network. This has been done by subtracting the aimed result (correct) from the
result of network. If the result is equal to zero, this means that the network achieved correct result.
However, if the result is not zero, the network needs its weights adjustment through learning rule.
Wjinew=wjiold+c (Tj –Xj) ax
where
 wjinew is the value of new weight added to inter-joint between unit (J) and unit (i).
 In addition, wjiold is the value of old weight added to inter-joint between unit (j) and (i).
 c is the learning rate that is stable value is always value less than 1.
 Tj is the aim value of network.
 Xj is the value from the network.
 ai is the output from the unit (i).
B. Examination Stage
The network Exam is exactly as same as learning process. Nevertheless, the network in this stage is not
adjusted-weights only doing two operations; addition and transformation, then comparing the result
from the network with the aim result. Where it has been shown the category of exam on the network.
This category contains a group pf input and output accompanied with each input. The exam category is
preferred to be different from the training category. If the network passed the exam and gave correct
answers, the result would be the network's learning and the network would be used.
The reasons for the network's failure

The design and training of network have done but finally failed.
The reasons for its failure are numerous. Some are:
1. The number of units is not suitable.
2. The weights from which the network starts are not sufficient.
3. The learning rate is not satisfactory.
318 Al-Salaimeh Safwan and Hanna Jaber
4. The training category has not carefully chosen.

5. The transformation rule is not adequate.
Results and Discussion

The perceptron or perceptron network, the analysis of equation regarding the mechanism of network
and its training to form AND system with two entrances were studied. The network should be
examined to compare the existing results with the desired results were noticed. The speed of exam
depends on selecting the weights that are near the correct weight in addition to taking into account the
suitable equation of exam.
The process has done in the last one of processing layers in which the result of the mentioned
sum process is transformed in the previous process to one of values that is proposed within the result of
desired network. After transformation process, the result that is from the network is compared with the
correct result that is proposed from the network
References
[1] Neural Networks: A Comprehensive Foundation, Simon Haykin, Macmillan Coll Div, New
York, 2006.
[2] Digital Neural Networks, S.Y. Kung, Prentice Hall, 1997.
[3] Digital Neurocomputing From Theory to Implementation, Kung, S.Y., Pearson professional
education country, USA, 2000.
ISSN 1450-216X Vol.21 No.2 (2008), pp.319-327
Shape Preserving Convex Surface Data Visualization Using

Rational Bi-Quartic Function
Malik Zawwar Hussain

Department of Mathematics, University of the Punjab, Lahore-Pakistan
E-mail: malikzawwar@math.pu.edu.pk
Maria Hussain
Tahira Sumbal Shaikh

Abstract
The rational bi-quartic partially blended function [8], an extension of rational

quartic function [14] is used to visualize the convex data. The rational bi-quartic partially
blended function has eight free parameters in each rectangular patch. Simple sufficient data
dependent constraints are derived on these free parameters to preserve the shape of convex
data. The presented method applies equally well to data or data with derivatives, is local in
its representation, computationally economical and verified numerically. The high degree
of the rational function used in this paper, also make it suitable candidate for interactive
CAD.
Keywords: Shape preservation; Rational bi-quartic partially blended function; Free

parameters; Convex surfaces.
AMS Subject Classification Codes: 65D05, 65D10, 65D07, 65D18, 68U07
1. Introduction
In recent years, a good amount of work has been published [1-14] that focuses on shape preserving
curves and surfaces. Convexity is very important shape property and plays a vital role in non-linear
programming that arises in engineering and scientific applications such as design, optimal control,
parameter estimation and approximation of the functions. Some of the work on convexity has already
been discussed. A brief review of some of the past work is as follows:
Asaturyan [1] scheme divides each rectangle of the grid into nine sub-rectangles to preserve the
shape of convex data. The scheme is not local i.e. the change in the data along one edge of a sub-
rectangle affects the edges of all sub-rectangles located in that vary direction.
Constantini [4] proposed a semi global scheme to preserve the shape of 3D convex data. The
author extended some of the univariate results developed in [3] for bivariate interpolation. The
drawback of this scheme was that in some rectangular patches degree of interpolant became too high
and polynomial patches tend to be linear in x and/or y , visually non pleasing surfaces. Dodd et al. [6]
used quadratic spline to define partial derivatives on the boundaries of the grid rectangles. The scheme
developed in [6] preserved the convexity of the data along the edges of grid but failed to preserve the
Shape Preserving Convex Surface Data Visualization Using Rational Bi-Quartic Function 320
convexity in the interior of the grids and produced the undesirable flat spots due to vanishing of second
order mixed partial derivatives.
The paper is organized as follows. In Section 2, rational quartic function [14] is rewritten. In
Section 3, rational quartic function developed in [8] is rewritten. In Section 4, a method is developed to
preserve the shape of convex data by making constraints on free parameters in the description of
rational bi-quartic partially blended function. The Section 5, demonstrates the method developed in
Section 4. Section 6 concludes the paper.
2. Rational Quartic Function

Let  x , f  , i  0,1, 2,, n
i i be given set of data points, x0  x1  x2    xn . A piecewise rational
quartic function developed by Wang and Tan [14] is defined in each interval I i  [ xi , xi 1 ] as:
pi  
Si    , (1)
qi  
with
pi     i fi 1     { 3 i  i  fi   i hi d i } 1     {3 i f i 1  3i f i } 2 1   
4 3 2
 { i  3i  f i 1  i hi di 1} 3 1      i f i 1 4 ,

x  xi
qi     i 1      i , and hi  xi 1  xi ,  .
hi
The rational quartic function (1) has the following properties:
S ( xi )  f i , S ( xi 1 )  f i 1 , S 1 ( xi )  di , S 1 ( xi 1 )  di 1.
Here S 1 ( x) denotes the derivative with respect to x and di denotes the derivative value
(given or estimated by some method) at the knots xi . S  x   C1  x0 , xn  has freedom of  i and i as
parameters in the interval [ xi , xi 1 ]. It is noted that in each interval [ xi , xi 1 ] , when  i  1 and  i  1,
the piecewise rational quartic function reduces to non-rational quartic polynomial.
Samreen [13] developed the following result:
Theorem 1
The piecewise rational quartic function defined in (1) preserves convexity in each subinterval  xi , xi 1  ,
if the parameters  i and  i satisfy the following conditions:
i  0
and
     3 i  di   i  3 i  di   i  9 i  4di 1  3di   i di 1 
 i  mi  Max 0, i i , i , , ,  , mi  0.
 di (5 i  3di )  3 i  di 1   3i  di   i 
3. Rational Bi-quartic Partially Blended Function

The rational bi-quartic partially blended function [8] used in this paper is defined as:
Let  xi , y j , Fi , j  , i  0,1, 2,..., n ; j  0,1, 2,..., m be given set of data points defined over the
rectangular domain D   a, b    c, d  ,  : a  x0  x1    xn  b be partition of  a, b  and
 : c  y0  y1    ym  d be the partition of  c, d  .
The rational bi-quartic partially blended function S  x, y  is defined over each rectangular
patch  xi , xi 1    y j , y j 1  , i  0,1, 2, , n  1 ; j  0,1, 2, , m  1 as:
321 Malik Zawwar Hussain, Maria Hussain and Tahira Sumbal Shaikh
S  x, y    AFBT , (2)
where
A   1 a0   a1    , B   1 b0   b1    ,
 0 S  x, y j  S  x, y j 1  
 
F   S  xi , y  S  xi , y j  S  xi , y j 1   ,
 
 S  xi 1 , y  S  xi 1 , y j  S  xi 1 , y j 1  
 
with
a0  1    1  2  , a1   3  3  2  , b0  1    1  2  , b1   2  3  2  ,
2 2
y  yj
 .
hˆ j
S  x, y j  , S  x, y j 1  , S  xi , y  and S  xi 1 , y  are rational quartic functions, defined on the

boundary of rectangular patch  xi , xi 1    y j , y j 1  as:
4
 1   
4 i
 i Ai
S  x, y j   i 0
, (3)
q1  
with
A0   i , j Fi , j , A1   3 i , j   i , j  Fi , j   i , j hi Fi ,xj , A2  3 i , j Fi 1, j  3 i , j Fi , j ,
A3   i , j  3 i , j  Fi 1, j  i , j hi Fi x1, j , A4  i , j Fi 1, j ,
q1     i , j 1     i , j ,
4
 1   
4 i
 i Bi
S  x, y j 1   i 0
, (4)
q2  
with
B0   i , j 1Fi , j 1 , B1   3 i , j 1   i , j 1  Fi , j 1   i , j 1hi Fi ,xj 1 , B2  3 i , j 1 Fi 1, j 1  3i , j 1 Fi , j 1 ,
B3   i , j 1  3i , j 1  Fi 1, j 1  i , j 1hi Fi x1, j 1 , B4  i , j 1 Fi 1, j 1 ,
q2     i , j 1 1     i , j 1 ,
4
 1   
4 i
 i Ci
S  xi , y   i 0
, (5)
q3  
with
 
C0  î , j Fi , j , C1  3î , j  î , j Fi , j  î , j hˆ j Fi ,yj , C2  3î , j Fi , j 1  3î , j Fi , j ,
 
C3  î , j  3î , j Fi , j 1  î , j hˆ j Fi ,yj 1 , C4  î , j Fi , j 1 ,
q3    î , j 1     î , j ,
4
 1   
4 i
 i Di
S  xi 1 , y   i 0
, (6)
q4  
with
 
D0  ˆ i 1, j Fi 1, j , D1  3î 1, j  î 1, j Fi 1, j  î 1, j hˆ j Fi y1, j , D2  3ˆ i 1, j Fi 1, j 1  3î 1, j Fi 1, j ,
 
D3  î 1, j  3î 1, j Fi 1, j 1  î 1, j hˆ j Fi y1, j 1 , D4  î 1, j Fi 1, j 1 ,
q4    ˆ i 1, j 1     î 1, j .
4. Convex Surface Interpolation

Let  xi , y j , Fi , j  be convex data defined over rectangular grid Ii, j  xi , xi1 yj , yj1 , i  0, 1, 2,..., n  1 ;
j  0,1, 2,..., m  1 , satisfying the following conditions:
Fi ,xj  i , j  Fi x1, j , Fi ,yj  ˆ i , j  Fi ,yj 1 , Fi ,xj  Fi x1, j , Fi ,yj  Fi ,yj 1 ,  i , j   i 1, j , ˆ i , j  ˆ i , j1 ,
where
F  Fi , j ˆ F F
 i , j  i 1, j ,  i , j  i , j 1 i , j .
hi hˆ j
The following result was developed in [2]:
Bi-quartic partially blended surface patch inherits all the properties of network of boundary
curves.
According to above result bi-quartic partially blended surface patch (2) is convex if boundary curves
defined in (3)–(6) are convex.
4
 1     L
4 i i
i
S  x, y j  is convex if S 
2
 x, y  i 0
0,
h  q   
j 2 3
i 1
L0  6  i , j  4 F  2 i , j F ,
3
i, j
3
i, j
x
i, j
2
i, j
x
i, j
L1  2 i3, j Fi ,xj  6 i3, j  i , j  18 i2, j i , j  i , j  6 i2, j i , j Fi x1, j  8 i2, j  i , j Fi ,xj ,

L2  18 i , j i2, j  i , j  18 i2, j  i , j  i , j  6 i2, j  i , j Fi ,xj  6 i , j i2, j Fi x1, j ,
L3  18 i , j  i2, j  i , j  8 i , j i2, j Fi x1, j  6 i , j  i2, j Fi ,xj  6 i3, j  i , j  2 i3, j Fi x1, j ,
L4  6 i3, j  i , j  4 i3, j Fi x1, j  2 i , j i2, j Fi x1, j .
4
 1   
4 i
 i Li  0 if Li  0 , i  0,1, 2,3, 4 .
i 0
Li  0 , i  0,1, 2,3, 4 if  i , j  0 and
i, j
  F x  
 Max 0, i, j i, j
, i, j i, j
, ,

i , j 3Fi x1, j  3 i , j i , j 3Fi x1, j  3 i , j     .

 F x
F x
 3  3Fi ,xj  5i , j
 i , j i 1, j i , j i , j

4
 1     M
4 i i
i
S  x, y j 1  is convex if S  2
 x, y  i 0
 0,
h  q   
j 1 2 3
i 2
M 0  6 3
i , j 1  i , j 1  4 3
F x
i , j 1 i , j 1  2 2
i , j 1  F x
i , j 1 i , j 1 ,
M 1  2 3 x
F
i , j 1 i , j 1  6 3
 i , j 1  18
i , j 1
2
i , j 1  i , j 1 i , j 1  6 i2, j 1i , j 1Fi x1, j 1 
8 i2, j 1i , j 1 Fi ,xj 1
M 2  18 i , j 1i2, j 1i , j 1  18 i2, j 1i , j 1 i , j 1  6 i2, j 1i , j 1 Fi ,xj 1  6 i , j 1 i2, j 1Fi x1, j 1 ,
M 3  18 i , j 1i2, j 1 i , j 1  8 i , j 1i2, j 1 Fi x1, j 1  6 i , j 1i2, j 1 Fi ,xj 1  6 i3, j 1 i , j 1 

2 i3, j 1 Fi x1, j 1
M 4  6 i3, j 1i , j 1  4 i3, j 1Fi x1, j 1  2 i , j 1i2, j 1 Fi x1, j 1 .
4
 1   
4 i
 i M i  0 if M i  0 , i  0,1, 2,3, 4 .
i 0
M i  0 , i  0,1, 2,3, 4 if  i , j 1  0 and

  F x
 i , j 1  Max 0, i , j 1 i , j 1
, ,
x
,
x

i , j 1 i , j 1 i , j 1 3Fi 1, j 1  3i , j 1 i , j 1 3Fi 1, j 1  3 i , j 1     .

  i , j 1 Fi x1, j 1 Fi ,xj 1  3 i , j 1 3Fi ,xj 1  5 i , j 1 
4
 1     N
4i i
i
S  x , y  is convex if S    x , y  
2 i 0
0,
hˆ  q   
i i 2 3
j 3
N 0  6ˆ ˆ i , j  4ˆ F  2ˆ î , j F ,

3
i, j
3
i, j
y
i, j
2
i, j
y
i, j
N1  2ˆ i3, j Fi ,yj  6î3, j ˆ i , j  18î2, j î , j ˆ i , j  6ˆ i2, j î , j Fi ,yj 1  8ˆ i2, j î , j Fi ,yj ,
N  18ˆ ˆ 2 ˆ  18ˆ 2 ˆ ˆ  6ˆ 2 ˆ F y  6ˆ ˆ 2 F y ,
2 i, j i, j i, j i, j i, j i, j i, j i, j i, j i, j i, j i , j 1
N 3  18ˆ i , j î2, j ˆ i , j  8î , j î2, j Fi ,yj 1  6ˆ i , j î2, j Fi ,yj  6 î3, j ˆ i , j  2 î3, j Fi ,yj 1 ,
N  6ˆ 3 ˆ  4ˆ 3 F y  2ˆ ˆ 2 F y .
4 i, j i, j i, j i , j 1 i, j i, j i , j 1
4
 1   
4i
 i N i  0 if N i  0 , i  0,1, 2,3, 4 .
i 0
N i  0 , i  0,1, 2,3, 4 if î , j  0 and

 ˆ F y ˆ ˆ

 i , j  Max 0,
ˆ i, j i, j
,
i, j i, j
, ,

î , j 3Fi ,yj 1  3ˆ i , j î , j 3Fi ,yj 1  3ˆ i , j     .

 ˆ i , j Fi ,yj 1 Fi ,yj  3ˆ i , j 3Fi ,yj  5ˆ i , j 
4
 1     P
4 i i
i
S  x , y  is convex if S    x , y    0,
2 i 0
hˆ  q   
i 1 i 1 3
2
j 4
P0  6î31, j ˆ i 1, j  4î31, j Fi y1, j  2ˆ i21, j î 1, j Fi y1, j ,

P1  2ˆ i31, j Fi y1, j  6ˆ i31, j ˆ i 1, j  18ˆ i21, j î 1, j ˆ i 1, j  6ˆ i21, j î 1, j Fi y1, j 1
8ˆ i21, j î 1, j Fi y1, j
P2  18î 1, j î21, j ˆ i 1, j  18î21, j î 1, j ˆ i 1, j  6î21, j î 1, j Fi y1, j  6î 1, j î21, j Fi y1, j 1 ,
P3  18ˆ i 1, j î21, j ˆ i 1, j  8ˆ i 1, j î21, j Fi y1, j 1  6ˆ i 1, j î21, j Fi y1, j  6 î31, j ˆ i 1, j
2 î31, j Fi y1, j 1
P4  6 î31, j ˆ i 1, j  4 î31, j Fi y1, j 1  2î 1, j î21, j Fi y1, j 1 .
4
 1   
4i
 i Pi  0 if Pi  0 , i  0,1, 2,3, 4 .
i 0
Pi  0 , i  0,1, 2,3, 4 if î 1, j  0 and

 ˆ F y

î 1, j  Max 0, i 1, j i 1, j
, ,
 ,
 y

î 1, j ˆ i 1, j î 1, j 3Fi 1, j 1  3ˆ i 1, j î 1, j 3Fi 1, j 1  3ˆ i 1, j 
y

.
 ˆ i 1, j Fi y1, j 1 Fi y1, j  3ˆ i 1, j 3Fi y1, j  5ˆ i 1, j 
The above can be summarized as:
Theorem 2
The piecewise rational bi-quartic interpolant S  x, y  , defined over the rectangular mesh
D   x0 , xn    y0 , ym  , in (2), is convex if the following sufficient conditions are satisfied.
  0, 
i, j  0 , ˆ  0 , ˆ
i , j 1 i, j  0, i 1, j
 i , j  Max 0, Dk , 1  k  4, k  Z   ,
 i , j 1  Max 0, Dk , 5  k  8, k  Z   ,
î , j  Max 0, Dk , 9  k  12, k  Z   ,
ˆ i 1, j  Max 0, Dk , 13  k  16, k  Z   ,
with
 i , j Fi ,xj i , j  i , j  i , j  3Fi x1, j  3 i , j  i , j  3Fi x1, j  3 i , j 
D1  , D2  , D3  , D4  ,
i, j Fi x1, j Fi ,xj  3 i , j 3Fi ,xj  5 i , j
i , j 1 Fi ,xj 1 i , j 1i , j 1 i , j 1  3Fi x1, j 1  3i , j 1 
D5  , D6  , D7  ,
 i , j 1 Fi x1, j 1 Fi ,xj 1  3 i , j 1
 i , j 1  3Fi x1, j 1  3i , j 1 
D8  ,
3Fi ,xj 1  5 i , j 1
D9 
î , j Fi ,yj
, D10 
î , j ˆ i , j
, D11 
 
î , j 3Fi ,yj 1  3ˆ i , j
, D12 
 
î , j 3Fi ,yj 1  3ˆ i , j
,
ˆ i , j Fi ,yj 1 Fi ,yj  3ˆ i , j 3Fi ,yj  5ˆ i , j
D13 
î 1, j Fi y1, j
, D14 
î 1, j ˆ i 1, j
, D15 
 
î 1, j 3Fi y1, j 1  3ˆ i 1, j
,
ˆ i 1, j Fi y1, j 1 Fi y1, j  3ˆ i 1, j
D16 

î 1, j 3Fi y1, j 1  3ˆ i 1, j
.

3F y  5ˆ
i 1, j i 1, j
The above constraints can be rearranged as:

 i , j  ti , j  Max 0, Dk , 1  k  4, k  Z   , ti , j  0.
 i , j 1  ui , j  Max 0, Dk , 5  k  8, k  Z   , ui , j  0.
î , j  vi , j  Max 0, Dk , 9  k  12, k  Z   , vi , j  0.
ˆ i 1, j  wi , j  Max 0, Dk , 13  k  16, k  Z   , wi , j  0.
5. Numerical Examples
First example: The data in Table 1 is generated from the following function:
F1  x, y   e x2  y 2
; 1  x, y  6 .
Table 1: A convex data:
y/x 1 2 3 4 5 6
1 4.1000 9.4000 23.600 61.800 163.90 438.20
2 9.4000 16.900 36.800 87.500 218.10 558.10
3 23.600 36.800 69.600 148.40 340.70 819.10
4 61.800 87.500 148.40 286.20 603.70 1354.4
5 163.90 218.10 340.70 603.70 1177.4 2465.7
6 438.20 558.10 819.10 1354.4 2465.7 4843.0
Figure 1: Convex rational bi-quartic partially blended function.
Figure 1 is generated by the convexity preserving scheme developed in Section 4 with

i , j   i , j 1  î , j  î 1, j  0.5 and ti , j  ui , j  vi , j  wi , j  0.47 . It is clear from Figure 1, shape of the
data in Table 1 has been preserved smoothly.
Second example: The data set in Table 2 is generated by the following function:
F2  x, y   x 2  y 2 ;  3  x, y  3 .
Table 2: A convex data:
y/x -3 -2 -1 0 1 2 3
-3 18.000 13.000 10.000 9.0000 10.000 13.000 18.000
-2 13.000 8.0000 5.0000 4.0000 5.0000 8.0000 13.000
-1 10.000 5.0000 2.0000 1.0000 2.0000 5.0000 10.000
0 9.0000 4.0000 1.0000 0.0000 1.0000 4.0000 9.0000
1 10.000 5.0000 2.0000 1.0000 2.0000 5.0000 10.000
2 13.000 8.0000 5.0000 4.0000 5.0000 8.0000 13.000
3 18.000 13.000 10.000 9.0000 10.000 13.000 18.000
Figure 2: Convex rational bi-quartic partially blended function.
Figure 2 is generated with the convex surface scheme developed in Section 4 with  i , j 
 i , j 1  î , j  î 1, j  0.05 and ti , j  ui , j  vi , j  wi , j  0.4. It is clear from the Figure 2, shape of the
data in Table 2 is smoothly preserved.
6. Conclusion
In this paper, the problem of shape preserving convex surface has been discussed. Data dependent
sufficient constraints are derived on free parameters in the description of rational bi-quartic function to
preserve the shape of convex data.
The scheme developed in this paper has number of advantageous features over already
developed schemes. The scheme developed in [1, 4] is global whereas the scheme presented in this
paper is local. The scheme in [6] preserved convexity only along grid lines and failed to preserve
convexity in the interior of the grid whereas in our scheme convexity is preserved along grid lines as
well as in the interior of rectangle. The scheme developed in [9] with cubic function, and constraints on
two parameters whereas the scheme presented in this paper with quartic function, and constraints on
eight parameters so it provides more degree of freedom to the user. Thus the presented method is more
flexible and more suitable for interactive CAD and applies equally well to data or data with
derivatives.
References
[1] Asaturyan, S., Shape preserving surface interpolation; Ph.D. Thesis, (1990), Department of
Mathematics and Computer Science, University of Dundee, Scotland, UK.
[2] Casciola, G. and Romani, L., Rational interpolants with tension parameters, Curve and Surface
Design, Tom Lyche, Marie-Laurence Mazure and Larry L. Schumaker(eds.), (2003), 41-50.
[3] Costantini, P., On monotone and convex spline interpolation, Mathematics of Computation,
46(173), (1986), 203-214.
[4] Costantini, P. and Fontanella, F., Shape preserving bivariate interpolation, SIAM Journal of
Numerical Analysis, 27(2), (1990), 488-506.
[5] Delbourgo, R. and Gregory, J. A., Shape preserving piecewise rational interpolation, SIAM
Journal of Statistics and Computing, 6, (1985), 967-976.
[6] Dodd, S. L., Mc Allister, D. F. and Roulier, J. A., Shape preserving spline interpolation for
specifying bivariate functions on grids, IEEE Computer Graphics and Applications, 3(6),
(1983), 70-79.
[7] Floater, M. S., Total positivity and convexity preservation, Journal of Approximation Theory,
96, (1999), 46-66.
[8] Hussain, M. Z., Hussain, M. and Tahira, S. S., Shape preserving positive surfaces using rational
quartic spline, submitted in Applied Mathematical Modelling, (2007).
[9] Hussain, M. Z. and Hussain, M., Visualization of 3D data preserving convexity, Journal of
Applied Mathematics and Computing, 23(1), (2007), 397-410.
[10] Renka, R. J., Interpolation of scattered data with a convexity-preserving surface, ACM
Transactions on Mathematical Software, 30(2), (2004), 200-211.
[11] Roulier, J. A., A convexity preserving grid refinement algorithm for interpolation of bivariate
functions, IEEE Computer Graphics and Applications, (1987), 57-62.
[12] Sarfaraz, M. and Hussain, M. Z., Data visualization using rational spline interpolation, Journal
of Computational and Applied Mathematics, 189(1-2), (2006), 513-525.
[13] Samreen, S., Shape preserving data visualization with rational quartic splines; M.Phil Thesis,
(2006), University of the Punjab Lahore, Pakistan.
[14] Wang, Q. and Tan, J., Rational quartic spline involving shape parameters, Journal of
Information and Computational Science, 1(1), (2004), 127-130.
ISSN 1450-216X Vol.21 No.2 (2008), pp.328-345
Dynamics of the Functional Groups of the Meadow Spontaneous

Species Around the Lakes of the Area of El-Tarf (Algeria)
Slimani. A
Institut des sciences Agronomiques, Centre universitaire El-Tarf 36 000 (Algérie)
E-mail: ali.slimani@ensaia.inpl-nancy.fr
Tel: 21338600942; Fax: 21338600942
Plantureux. S
Laboratoire d’Agronomie et Environnement
2, Avenue Forêt de Haye 54505 Vandœuvre lès nancy (France)
Brinis. L
Département de Biologie Faculté des sciences
Université Badji Mokhtar Annaba (Algérie)
Soltane. M
Institut des Sciences Agronomiques
Centre Universitaire d’El-Tarf 36 000 (Algérie)
Abstract
The area of El-Tarf constitutes has environment remarkable natural, for its
exceptional biodiversity in the Mediterranean basin. It is ecosystem which proves more to
be weakened. Its lakes complex undergoes since years year increasing environmental
pressure. These banks of the lakes are overgrazing and plowed frequently, which affect
certainly the floristic composition and their habitat. Vegetation, natural species and abiotic
factors, soil characteristic physicochemical, were evaluated in 66 stations distributed over
the lakes, along moisture gradient. 25 native species were founded: Poaceae 48%, trifolium
20%, médicago 16% and 16% other Fabaceae. A long term of disturbance causing species
to response to environmental variables. 3 groups of species were formed sharing the same
strategies adaptive and have similar ecological attributes. These functional groups are
abundant around the lakes. The results indicate that the floristic composition is
homogeneous in the 3 lakes, the soil fertility is low but, did not affect the floristic
composition, owing to the fact that the species are tolerate. The moisture gradient did not
effect the species distribution. In return, human activity (practices) plays a major role in the
3 lakes since 19th century. The last 20 years, human activity has increased dramatically in
this area, which explains the loss of plant diversity and has direct effect reducing the
species richness chiefly due to degradation of habitats and reducing seed production and
the stock seed in the soil.
Keywords: Drilling, native, species, lake, functional groups, floristic composition, abiotic
factors, richness specific,
Lakes of the Area of El-Tarf (Algeria) 329
1. Introduction
L’activité agricole a joué depuis des siècles un rôle critique dans la formation du paysage
Méditerranéen et sa biodiversité (27). Le bassin méditerranéen figure parmi les régions les plus riches
du monde, il possède de plus de 4,3% des plantes de toute la planète (29) sur une superficie
relativement modeste (10 000 km2). Face à la densité humaine, il devient urgent d’identifier les sites
irremplaçables en terme de conservation des écosystèmes et de la diversité végétale et animale. Devant
une demande sociale et à la pression anthropique croissante, certains biologistes préconisent
aujourd’hui d’évaluer le coût économique de la protection des sites encore naturels (26); (21).
Cependant, les changements socio-économiques qui ont eu lieu durant ces dernières années en Algérie
et l’introduction des réformes dans le secteur agricole, depuis 1970, ont conduit une baisse des
productions agricoles dans les régions arables. Tel est notamment le cas dans les zones humides, qui
sont les régions les mieux arrosées du pays (600 à 800 mm/an), mais aussi considérées comme les plus
vulnérables sur le plan écologique, (19). Ces changements sont dus principalement à l’augmentation de
la pression exercée par la croissance rapide de la population dans ces zones, cette situation étant plus
sérieuse autour des lacs. Les sites autour de ces lacs subissent une fragmentation (morcellement), ainsi
que la détérioration des écosystèmes et, qui entraîne une très forte diminution de la biodiversité.
La région d’El-tarf (Algérie) composée en grande partie par la zone des lacs du Parc National
d’El-Kala, (P.N.E.K) qui constitue des milieux remarquables, et exceptionnels, tant par les fonctions
écologiques qu’ils remplissent que, pour la biodiversité qu’ils abritent (15). Ce sont ces derniers
milieux qui s’avèrent être les plus fragilisés. Les diverses menaces ont depuis longtemps détruit les
zones humides et le littoral aux environs d’Alger au cours du XXe siècle (développement de
l’agriculture, l’urbanisation et du tourisme). Cette région, qui fait partie des zones humides du bassin
méditerranéen, sont connues pour leurs multiples valeurs pour l’homme et la nature (32). Une de ces
valeurs est de constituer une source d’alimentation inestimable pour des animaux de ferme et aussi la
faune sauvage. Le plus précieux sous type des zones humides pour le pâturage est la prairie humide i.e,
zone avec une table d’eau dans le sol assez élevée qui permet la croissance des plantes herbacées à
travers la plupart des saisons (34). Cependant, une nouvelle menace, très active et irréversible, est en
train de détruire la flore et l’écosystème de tout le secteur.
Des ressources fourragères sont détruites à un taux alarmant en raison d’une combinaison de
facteurs. Un des problèmes majeurs auquel est confrontée l’agriculture pastorale dans cette région est
la régression rapide des espèces fourragères natives (1);(2);(3) et ce malgré le statut juridique national
(loi 2003) et international (31) permettant la protection de ces écosystèmes extrêmement fragiles, et de
ce biotope menacé d’extinction. C’est parce qu’ils sont très vulnérables à la surexploitation et au
changement du climat que les écosystèmes de cette région doivent faire l’objet d’une attention
particulière et immédiate, (8). Sur le plan économique ces zones sont vitales pour un développement
agricole durable puisque l’essentiel des ressources fourragères des élevages de la région provient de la
végétation des rives des lacs. Afin d’utiliser au mieux les potentialités écologiques et économiques de
ces écosystèmes, il est nécessaire de comprendre leur fonctionnement, pour mettre au point des outils
de gestion conciliant le maintien de la biodiversité, de la gestion de l’espace, l’alimentation des
animaux et la protection des sols. D’une manière générale, les effets des pratiques agricoles sur la
végétale prairiale sont de mieux en mieux documentés. Cependant, tel n’est pas le cas dans les zones
méditerranéennes humides, ce qui ne permet pas de bien comprendre ces écosystèmes et de faire des
améliorations sur le plan agronomique et écologique.
Le Maghreb méditerranéen (Afrique du nord), majoritairement représenté par l’Algérie, est
donc au cœur des préoccupations mondiales en matière de conservation de la biodiversité menacées par
une anthropisations croissantes (10). Dans ce contexte l’étude de la biodiversité des pâturages
(parcours) est au vaste point chaud méditerranéen par le fait que la menace est liée à deux tendances:
l’une par les pratiques (la surcharge ou le surpâturage) et l’autre par la mise en culture des pâturages.
330 Slimani. A, Plantureux. S, Brinis. L and Soltane. M
Figure 1: On the left card location of the study area in North East of Algeria, on the right: location study area
showing the three lakes (Oiseaux lake, Oubeïra, Tonga).
En effet, les terres fertiles situées en bordure des lacs (13), ou alluvionales (16), se trouvent
constamment côtoyées par la pression humaine (animal) pour leur rentabilité quantitative (28) et aussi
qualitative (17). Finalement les conséquences des interventions humaines ou naturelles fournissent
l’occasion d’approfondir nos connaissances des processus de la dynamique de la végétation, d’autre
part cette approche d’analyse de ces changements de la composition floristique. Durant les dernières
années décennies, les berges des lacs du P.N.E.K ont été sérieusement dégradées par des perturbations
intensives liées à l’activité humaine (9), qui se traduit par la détérioration du couvert végétal, de son
habitat, et une perte de la biodiversité.
Les objectifs de cette étude sont: de comprendre, la diversité et la dynamique du couvert de
la végétation herbacée fourragères des berges des lacs Tonga, Oubeïra, et le lac des oiseaux.
Pour atteindre ces objectifs, nous formulons l’hypothèse qu’il faut prendre en compte le
comportement des espèces (stratégies fonctionnelles biologiques). Les stratégies observées par
l’intermédiaire des traits biologiques représentent les éléments pertinents à prendre en compte pour les
réponses à cette problématique. Au delà de la démarche de classification, l’objectif est d’expliquer les
réponses des communautés végétales aux changements environnementaux concernant notamment les
modalités d’usage des terres et le climat. La démarche se donne également comme objectif de pouvoir
prédire les effets de changements environnementaux sur la structure et la composition des
communautés et donc sur leur dynamique. Plusieurs travaux ont été menés dans ce sens: Botkin (4) a
émis l’hypothèse qu’il est possible de grouper les espèces végétales ayant les profils biologiques
similaires et de définir ainsi le groupe fonctionnel. Plusieurs possibilités de classification dites
fonctionnelles peuvent être conçues selon Lavorel et al., (22). Bouzillé (5), définit que l’objectif majeur
et l’établissement d’une liste synthétique de traits pertinents qui permettent une généralisation des
informations, mais un aspect clairement affiché de ces démarches et aussi de comprendre et prédire les
réponses des communautés aux facteurs environnementaux. Mac Intyre et al., (24), soulignent que les
classifications fonctionnelles constituent un cadre de travail important pour décrire les mécanismes de
réponse de la végétation, en d’autres termes, il s’agit de comprendre les variations de la composition
floristique des communautés végétales soumises à des contraintes ou des perturbations (18).
2. Etude du Site
La région d’El-tarf est connue pour sa grande richesse biologique et d’exceptionnelle originalité
écologique, ses lacs présentent une grande diversité sur plan floristique, et faunistique (8). Le climat est
méditerranéen, située dans l’étage subhumide chaud de l’échelle d’EMBERGER. La région est située à
l’extrême Est (8° 25’ ‘’ E) de l’Algérie limitée au nord (36°, 49’ ‘’ N) par la mer méditerranée et à
l’Est par la Tunisie. (Fig. 1).Elle s’étale sur un relief plat, sa superficie est de 305 000 ha, de nombreux
lacs de taille variable s’inscrivent dans un contexte géomorphologique sublittoral et possèdent des eaux
lagunaires salées (Mellah), saumâtres (Mafragh) ou d’eau douce à tendance plus au moins
mésoeutrophe: le lac Tonga, Oubeïra et le lac des Oiseaux qui font l’objet de notre étude, forment une
grande partie de complexe humide et la plus importante zone lacustre du pays et troisième position
après le delta de l’Ebre, en Espagne et la Camargue en France dans les zones humides du bassin
méditerranéen. L’altitude des plans d’eau s’étage de quelques mètres à une trentaine de mètres leur
surface varie de quelques hectares pour les plus petits à plus de 3160 hectares pour les plus grands.
2.1. Le lac Tonga

Ce lac est naturellement du type endoréique et le plan d’eau douce eutrophe est situé nettement au
dessus du niveau de la mer (2,2 mètres en moyenne). Il est caractérisé par des berges faiblement
inclinées ce qui facilite leur mise en exploitation par les riverains, et des boisements tourbeux
inondables, (25). Les zones de balancement des eaux, permettent la mise en pâturage en automne et au
printemps et la mise en cultures spéculatives (arachide et melon), pendant la période d’étiage (en été)
Les terres agricoles occupent 12.5 % de la surface totale du bassin versant et s’étendent sur une
superficie de 2059 ha. Elles sont constituées par les parcelles de terres travaillées par les riverains, mais
de plus en plus dégradées et abandonnées.
2.2. Le lac Oubeïra

Le lac est endoréique à eaux mésoeutrophes, situé à 25 mètres d’altitude, le marnage des eaux dégage
de larges ceintures de végétation très fortement pâturées, le bétail pouvant pénétrer profondément le
plan d’eau, ce qui explique la rareté de l’herbe sur les prairies des berges qu’ils ont l’aspect de
paillasson. C’est le lac qui subit le plus de pression de part de l’homme et de l’animal tout au long des
années. Il est désormais le seul site en Algérie abritant la châtaigne d’eau (Trapa natans) et l’unique
station en Afrique du nord ou l’on trouve le nénuphar jaune (Nuphar luteum). En période hivernale, il
abrite plus de 20.000 Oiseaux d’eaux toute espèces confondues.
2.3. Le lac des Oiseaux

Ce lac est relié aux larges plaines de la Mafragh par des exutoires par lesquels s’évacuent les hautes
eaux d’hiver, il est couramment asséché et mis en culture, l’eutrophisation de ses berges se traduit par
la destruction de l’habitat des espèces prairiales et du couvert végétal natif (14). Malgré sa faible
superficie, il constitue un habitat idéal pour une grande diversité floristique et faunistique.
Comparée à la plupart de la région méditerranéenne, la précipitation annuelle dans le bassin des
lacs est relativement élevée (960 ± 45mm (N.E.) pour les années 1970 – 1995 tombant principalement
pendant l’hiver et le printemps.
Les sols des lacs sont alluvionales originaires des collines environnantes, ou la roche mère est
principalement des grés numidiens sur lequel sont parfois plaqués des bourrelets dunaires. La plupart
des sables, et alluvions (matière organique), sont transportés profondément dans les couches du sol, par
les fortes précipitations dans la région, par conséquent nous obtenons des valeurs de pH légèrement
acide dans les stations des trois lacs. C’est un site de 2200 ha, profond de 4 mètres au maximum. Il
s’inscrit dans un quadrilatère de 5 x 4 km et développe 19 km de rives.
L’hétérogénéité des caractéristiques des sols a été "améliorée" par les pratiques agricoles
précédents, i.e. les terres des berges, sont souvent mises en culture (arachide). La mise en culture s’est
intensifiée ces dernières années, depuis ce temps le pâturage a probablement été influencée par les
pratiques agricoles lesquelles ont provoqué du changement dans la composition des espèces des berges,
e.i. labourer les terres chaque 2 ou 3 ans en avril-mai pour les cultures vivrières, et pâturées par des
animaux (bovins, ovins et caprins) en deux saisons, l’une au printemps et l’autre en automne. (Fig. 2)
Figure 2: Intensity and grazing period used annually around the shore lacs of Tanga, Oubeira, and Oiseaux.
La densité charge est de 10 animaux ha ֿ◌1, bien que la densité des espèces des plantes soit
relativement faible. La communauté des espèces fourragères, familles des Poaceae (graminées) et
Fabaceae (Légumineuses) est la plus menacée par la pression humaine qui ne cesse de grimpée, et le
surpâturage accru sur les berges où se situe précisément le maximum de diversité.
3. Matériel et Méthodes
3.1. Méthodes d’Etude
Une analyse floristique et une analyse de sol ont été effectuées sur un ensemble de 66 stations entre le
début du mois de mars et du mois d’avril 1998. Pour chaque lac, des stations ont été choisies dans les
parcelles en cours de pâturage en excluant les zones cultivées (arachide, melon, maraîchage) et les
zones non agricoles (forêts et falaises). Dans chaque parcelle un échantillonnage aléatoire a été
pratiqué sur les berges, allant de 30 à 200 mètres du bord de l’eau. Le nombre de stations varie en
fonction de la taille des berges des lacs, 20, 15, 31 pour les lacs, Tonga, le lac des Oiseaux, Oubeïra
respectivement. Le nombre de station par parcelle est globalement proportionnel à la taille de la
parcelle. Pour chaque station les relevés, ont été faits à l’intérieur d’un carré permanent de dimension,
1m x 1m (1m2), au sein de chaque carré, toutes les espèces fourragères présentes sont notées et un
coefficient de présence / absence leur est attribué. A l’exclusion des espèces à rosettes non fourragères
et des espèces arbustives dont leurs contributions est minime. Les espèces a rosettes et les arbustes, et
aussi une communauté d’hydrophytes et d’hélophyte qui apparaissent au printemps regroupant:
Nymphaea alba, Potamogeton nodosus, et Ranunculus aquatilis et de roselière flottante tel, que: la
Typhaie à feuilles larges
L’ensemble des sites d’étude a une faible pente, ≈ 1%, déterminée à l’aide de Sunnto
clinométer. Les berges sont généralement inondées par les eaux de crues des différents bassins versants
des lacs durant les fortes pluies d’hiver, depuis décembre à avril.
L’étude de la végétation de chaque station des trois lacs a consisté à déterminer sa composition
floristique et la structure de sa végétation des espèces herbacées collectées, suivant la méthode
d’évaluation, de Braun-Blanquet, (1964). Des coefficients de fréquence (F%) et d’abondance-
dominance (B%)ont été calculées pour chaque espèce. Les spécimens des plantes collectés ont été
identifiés au Parc National d’El-kala (P.N.E.K), en se referant à la flore d’Algérie tomes I et II de
Quezel et Santa (1962, 1963).
D’autres informations, le mode d’utilisation principal (pâturage ou mise en culture, la durée de
fréquentation de pâturage, (deux saisons l’une en automne et l’autre au printemps, Fig. 2) ont été
collectées pour chaque lac. Il en fut de même pour le nombre les catégories des animaux qui
fréquentent les berges des lacs, l’estimation de la population résident autour des lacs,
Ont été collectées pour chaque station en 1998: des échantillons sol étaient prélevés à une
profondeur de 35 cm. Ainsi que la distance des stations par rapport au bord de l’eau, a été notée.
 Le type de sol et les caractéristiques chimiques des échantillons de
Des analyses (physique et chimique) du sol était réalisée pour déterminée la granulométrie 5
fractions sans décarbonatation. L’analyse a porté sur: pH eau (NF ISO 10390), était mesuré dans
0,01M cacl2 (1:1 w/v). la teneur en Phosphore, (P2O5) selon la méthode de Joret – Hébert, le Potassium
(K2O) échangeable à l’acétate d’ammonium, le Calcium (CaCO3) total (NF ISO 10693), le magnésium
(Mg0), échangeable à l’acétate d’ammonium, la Matière Organique Totale (M.O.T) par correction
calcaire, et enfin l’ Azote total (N) (méthode de Kjeldahl, 2300 Kjeltec Analyser, Foss Tetator,
Höganäs, Sweden).
Ces analyses des échantillons de terre ont été analysées dans le laboratoire des analyses des sols
d’Arras (I.N.R.A. France), accrédité par la Cofrac (Essais).
3.2. Analyse des données

3.2.1. Traitement des données et analyses statistiques
L’analyse statistique des données a été effectuée à l’aide d’un programme informatique appelé SPAD.
Des analyses statistiques types: “factorielles de correspondances (A.F.C.)” et “en composantes
principales (A.C.P.)”, des présences et des dominances des espèces ont été performées par rapport aux
facteurs du milieu étudié. Les données acquises au cours de l’année 1998 ont permis dans un premier
temps, d’évaluer la présence et la fréquence des espèces dans chaque station et pour chaque lac, puis
dans un deuxième temps, de calculer la contribution spécifique (C.S %) pour chaque espèce présente.
La C.S. correspond au rapport de la fréquence de l’espèce à la somme des fréquences de toutes les
espèces.
Les données des variables explicatives environnementales collectées sont synthétisées dans le
(Tableau 1).
Tableau 1: Les variables environnementales prises en compte dans cette étude.
Variable Abréviation Mensuration

Sol
Sable SABLE % %
Limon LIMON % %
Argile ARGILE % %
pH pH Numérique
Matière organique totale MOT g/kg
Phosphore P g/kg
Potassium K g/kg
Afin d’exécuter ces analyses statistiques, nous avons identifié les variables à expliquer (les
espèces), des variables nominales qualitatives (lac, le mode d’utilisation, distance de l’eau, les analyses
du sol) et des variables quantitatives (texture, pH, P, K, MOT, Mg), de manière à pouvoir traiter l’effet
de chaque variable.
Dans un premier temps, une A.F.C. a été utilisée pour étudier les interactions significatives
entre les principaux facteurs. L’analyse n’est pas arbitraire, elle est basée sur les hypothèses
bibliographiques actualisées, permettant de tester et d’expliquer les effets de chaque variable étudiée
dans notre étude
4. Résultats
Au total 25 espèces (dont certaines sont natives), représentant trois grandes familles fourragères
annuelles ont été collectées et identifiées dans cette étude (tableau 4). Elles sont constituées de 12
graminées, et 13 légumineuses dont 5 du genre trifolium, 4 du genre médicago, et 4 représentant
d’autres espèces légumineuses fourragères considérées en 1992 par Gehu et al., comme des espèces
rares de la flore de cette région, et qui ont des valeurs intéressantes du point de vue fourrager.
4.1. L’AFC Espèces

L’ordination des 25 espèces sur un plan de projection d’une AFC (Fig. 3.), explique respectivement 12,3 et
10,4 % de l’inertie sur les axes 1 et 2. La dispersion des points variables (espèces) fournit une explication
écologique des espèces recensées dans les 66 stations. Les informations tirées de la figure 3, se résument
dans les points suivants: formation d’un premier groupe d’espèces dont les points sont rapprochés et situés
proche de l’origine, leurs valeurs propres sont élevées, et représentent les espèces suivantes (appréciation
tableau 4): Cynodon dactylon (L.) Pers (cda), Lolium perenne L (fov), Trifolium repens L.(tre), Lolium
multiflorum Lam. (lmu), Trifolium pratense L. (tpr), Medicago ciliaris (mci), Vicia sativa L. subsp sativa
(vst), Trifolium arvense L. (tan). Certaines sont même superposées, exemple, cda et fov. Ce groupe
d’espèces semble répondre à une caractéristique d’habitat similaire. En revanche, un deuxième groupe
d’espèces, représenté par: Poa trivialis L.(ptr), Mililotus indica (mhi), Lotus uliginosus Schkuhr (lul),
Sécale céréale (sce), Alopecurus pratensis L. (apr), Trifolium pratense L. (tpr), Poa annua L.(pan),
Medicado murex (mmu), Mililotus officinalis (mei), et Alopecurus geniculatus L. (age), dont les valeurs
propres sont faibles, et les points situés sur l’axe 2, sont éloignés les uns des autres, ce qui caractérise une
dissemblance de leur habitat.
Figure 3: Analyse factorielles des correspondances de présences de 25 espèces étudiées sur les berges des
trois lacs (Plan de projection des espèces)
L’AFC des espèces réalisée sur les 66 relevées représenté ici par des stations des berges des
lacs permet de projeter sur un plan à deux dimensions un nuage multidimensionnel afin de définir la
répartition des variables actives (présence, absences des espèces). Les plans factoriels 1, et 2 l’A.F.C
stations (figure 7), permettent d’expliquer les informations des stations. Sur la figure 7, la projection
des stations dans le système axes 1 et 2, semble être très dispersée. Les stations d’un même lac se
trouvent éloignées les unes des autres, exemples: Le lac Tonga: les stations 8 et 18, le lac des Oiseaux:
les stations 1 et 10, le lac Oubeïra: les stations 10 et 12.
Ce qui signifie que leur composition floristique est caractérisée par une certaine dissemblance.
En conséquences, sur la figure 7, les points de stations des trois lacs sont situés de part et d’autres du
plan de projection. Sur les axes1 et 2, on constate un premier groupe de nuage de points (stations) qui
se forme, ce nuage de points est constitués de relevés de différentes stations des trois lacs, et sont
proches les uns des autres. Par ailleurs, le deuxième groupe de nuages de points est situés dans la partie
droite du plan de projection et sont constitués des relevées des trois lacs, et qui sont dispersés sur l’axe
2
4.2. La Classification Hiérarchique Ascendante (CAH)

La Classification Hiérarchique Ascendante, CHA des 66 stations fait apparaître une classification des
stations relativement similaires sur un dendrogramme de performance. L’émergence de 2 grands
groupes de stations divisibles en 8 sous groupes, qui s’identifient par la présence de certaines espèces
et l’absence d’autres dans les relevés (Fig. 4). Les stations les plus ressemblants, ayant donc des
valeurs de similitude élevées sont définies dans les sous groupes: A, B, C, D. Ensuite, quatre autres
sous groupes dont les valeurs de similitude sont faibles, se désignent sur le dendrogramme: E, F, G, H.
La méthode de classification a conduit à l’existence de plusieurs assemblages d’espèces qui
représente des variations de composition floristique. Donc dans l’assemblage toutes les espèces ne
présentent pas la même distribution, certain sont très présentes ou constantes (espèces ubiquistes), tel
est le cas, de fov, cda, et tre, d’autres moins et certains peu présentes, exemple ptr, mhi, mei.
L’élaboration d’un tableau en tenant compte des informations obtenues à partir de l’analyse des
résultats, permet de visualiser précisément la distribution des espèces dans l’ensemble des stations des
lacs et de mieux mettre en évidence les assemblages d’espèces
En effet, sur la base de la Classification Ascendante Hiérarchique (CAH), les résultats obtenus
à partir des calculs de fréquences des espèces des stations qui forment les deux groupes hétérogène et
leurs sous groupes permettent de classer les espèces en fonction de leurs fréquences relatives. Plutôt
que la fréquence relative. Il est proposé en phytosociologie, un indice de présence en chiffres romains
(5) attribué en fonction des classes de fréquences des espèces (tableau 2). Cet indice correspond à la
probabilité de rassembler les espèces fidèles à un cortége floristique.
Tableau 2: Présentation d’un tableau synoptique définitif avec les différentes catégories d’espèces et leur
appartenance aux groupes.
Classe I Classe II Classe III Classe IV Classe V

Classes
Espèces < 20% 21 à 40% 41 à 60% 61à 80% > 81%
G.1 G.2 G.1 G.2 G.1 G.2 G.1 G.2 G.1 G.2
Age 4* 2 0* 0 0 1 0 0 0 1
Lul 4 0 0 1 0 1 0 1 0 1
Mei 4 2 0 0 0 1 0 0 0 1
Sce 4 0 0 1 0 2 0 0 0 1
Mhi 3 2 0 1 0 1 0 0 1 0
Apr 3 0 1 0 0 3 0 0 0 1
Ptr 2 2 1 1 2 0 0 0 0 1
Pan 2 1 2 2 0 0 0 1 0 0
Mmu 2 1 2 0 0 3 0 0 0 0
Ppr 2 1 0 1 2 0 1 1 1 1
Tfr 2 0 1 2 0 2 0 0 1 0
Dgl 1 2 1 1 2 0 0 1 0 0
Mci 0 0 0 0 2 1 1 3 1 0
Tan 1 1 0 0 0 0 3 2 0 1
Lpe 0 0 0 0 0 1 1 1 3 2
Cda 0 0 0 0 0 0 0 0 4 4
Fov 0 0 0 0 0 0 0 0 4 4
Tre 0 0 0 1 0 0 0 3 4 0
Lmu 0 0 0 1 0 0 2 2 2 1
Min 0 0 2 1 1 2 1 1 0 0
Tsu 1 1 0 0 2 2 1 1 0 0
Ast 0 1 0 1 2 1 1 1 1 0
Mlu 1 0 1 2 1 1 0 0 1 0
Tpr 1 1 1 0 0 1 1 0 0 2
Vst 1 1 1 1 1 1 1 0 0 1
* espèce présente dans les sous groupes ** espèce absente dans les sous groupes
Le traitement des stations par groupe et sous groupe à permis de mettre en évidence une
combinaison floristique de base constituée de cda, fov, tre, et au moins deux variations, ayant peut être
le rang de sous association, l’une caractérisée par lpe, tan, et l’autre par lmu, et mci. Il est toutefois,
nécessaire de préciser que la valeur diagnostique de la combinaison sous-groupe floristique est une
dimension statistique. Dans nos résultats, on a tenu compte que de dix stations (relevées), qui
constituent un minimum à réunir pour mettre en évidence une combinaison digne d’intérêt.
Figure 4: Dendrogramme des stations, Classification Hiérarchique Ascendante (CAH) SPAD. Regroupement
des stations similaires des trois lacs
Classification hiérarchique directe
O0_12
O0_11
T0_4
X0_9
O0_24
T0_3
O0_4
T0_9
X0_10
T0_7
T0_8
T0_11
T0_5
O0_20
O0_5
O0_19
O0_18
O0_21
T0_16
O0_23
O0_15
O0_14
O0_16
T0_6
T0_14
T0_20
O0_28
X0_3
O0_9
O0_22
X0_5
T0_15
O0_29
O0_27
O0_30
X0_14
0%
T0_17
O0_26
X0_6
T0_12
T0_13
X0_7
X0_12
T0_10
O0_31
X0_4
X0_8
T0_18
X0_15
O0_17
X0_2
O0_3
O0_7
O0_8
X0_1
O0_1
T0_19
O0_6
T0_2
O0_25
O0_10
O0_2
O0_13
X0_11
X0_13
T0_1
1
Le traitement des stations par groupe et sous groupe à permis de mettre en évidence une
combinaison floristique de base constituée de cda, fov, tre, et au moins deux variations, ayant peut être
le rang de sous association, l’une caractérisée par lpe, tan, et l’autre par lmu, et mci. Il est toutefois,
nécessaire de préciser que la valeur diagnostique de la combinaison sous-groupe floristique est une
dimension statistique. Dans nos résultats, on a tenu compte que de dix stations (relevées), qui
constituent un minimum à réunir pour mettre en évidence une combinaison digne d’intérêt.
4.3. Richesse spécifique (R.S en %) des sous groupes

La variabilité des cortèges floristiques peut être traduite par le nombre d’espèces par station, c’est la
richesse spécifique. Le tableau 3 et la figure 5 montrent, que la richesse spécifique est plus élevée dans
le groupe 2, (E,F,G,H) que dans le groupe 1(A,B,C,D); elle est homogène dans les sous groupes du
groupe 2 (12,9 % en H à 13,8% en F), en revanche elle est hétérogène dans ceux du groupe1 (9,1% en
C et 12,3% en B).
La structure d’une communauté, dépend des adaptations et des réponses des organismes
(Espèces) à l’environnement, ainsi qu’à la cohabitation entre les espèces.
Figure 5: Diagrammes montrant les résultats de la richesse spécifique en espèces par sous groupe de stations
(A,B,C,D et E,F,G,H)
16
Richesse spécifique (%)
14
12
10
8
6
4
2
0
A B C D E F G H
Sous groupes
Rappelons que le critère de biodiversité peut s’exprimer par la richesse spécifique ou l’indice
de Shannon qui représentent des indicateurs de l’état de référence de l’habitat.
4.4. Les résultats analyses physicochimique du sol

Une différence significative des résultats des analyses physiques caractérise les deux groupes de
stations. En effet, les sols des sous groupes A, B,C,D, sont riches en sable (≈ 60%), moyens en limon
(25%) et pauvres en argile (15%), et correspondent aux faibles richesses spécifiques (Tableau 3), en
revanche, les sols des sous groupes E,F,G,H, sont riches en limon (40%) et moyens en sable (41%) et
argile (19%), la richesse spécifique est relativement élevée dans le groupe 2 (13) par rapport au groupe
1 (11,4).
Table 3: Pourcentages moyens des différents Constituants du sol par rapport au coefficient de la richesse
spécifique par groupe, 1 et 2.
Constituants sol Groupe1 Groupe2

Sable (%) 60% 41,5
Argile (%) 15 % 19
Limon (%) 25% 39,7
R.spécifique 11, 4% 13 %
Figure 6: Diagrammes montrant les différents proportions (en %) physiques du sol (Sable, Argiles et limon)
respectivement dans les sous groupes de stations (A,B,C,D, et E,F,G,H)
16
Richesse spécifique (%)

14
12
10
8
6
4
2
0
A B C D E F G H
Sous groupes
La courbe représentant la matière organique totale (Fig.8. (b)) est négativement corrélée aux
différents niveaux de la richesse spécifique. Les figures (c) et (d) représentent les courbes de tendances
des éléments extractibles P et K, ne semble corréler avec la richesse spécifique des sous groupes.
Certaine étude ont démontré que le P et K peuvent réduire la richesse des espèces même dans le cas ou
ces éléments sont présents à un niveau élevé dans le sol (20). Cependant, le meilleur ajustement
linéaire en relation avec la richesse spécifique est l’azote (e), plus il est abondant dans le sol plus la
richesse spécifique augmente. Les mêmes résultats sont obtenus dans différentes régions du bassin
méditerranéen.
Les caractéristiques du sol varient en fonction du niveau de fertilité. La figure Fig.8. (a) montre
que les valeurs du pH des sous groupes sont stables, elles sont légèrement acide en C (pH= 6,62) à
neutre en E (pH=7.24), en conséquences les valeurs du pH n’influent pas sur la richesse spécifique des
sous groupes.
Figure 7: Ordination de l’AFC stations, montrant les différents groupes (1 et 2) ainsi que leurs sous groupes
formés à partir de la sortie de la CAH (A,B,C,D et E,F,G,H). Homogénéisation de la composition
floristique dans les groupes 1 et 2.
Figure 8: Relations de cinq caractéristiques physicochimiques du sol (pH, Matière organique totale,
Extractible P, Extractible K, N total) par rapport la richesse spécifique des berges des lacs ligne ou
courbe sont les meilleurs ajustements des données *P<0,05
La richesse spécifique (%)
5. Discussion
5.1. L’homogénéité de la végétation au niveau des trois lacs
Cette étude a permis de dégager des résultats intéressants, sur la situation des espèces naturelles
fourragères existantes sur les berges des trois lacs. Il apparaît clairement, que les résultats obtenus sont
relativement conforment aux résultats attendus.
Les analyses effectuées montrent, qu’il n’y a pas de discrimination entre les trois lacs, que les
stations ont des compositions végétatives similaires. Les sous groupes perfectionnés par la CAH, puis
reportés sur l’AFC stations (Fig. 7) confirment l’hypothèse d’une homogénéisation de la composition
végétative des berges. En conséquence, cette homogénéité trouve son explication dans deux types de
facteurs, le premier est attribué à la proximité des lacs: deux (2) km séparent le lac Tonga du lac
Oubeïra, et une dizaine de km sépare le lac des Oiseaux des deux autres lacs (Fig.1.). Cette proximité
des lacs, permet de justifier le second facteur, qui est le déplacement (va-et-vient) des animaux autour
des trois lacs, qu’ils soient domestiques, bovins, ovins et caprins, ou sauvages, sangliers, oiseaux
(d’eau principalement); cette situation correbore avec celle est qui expliquée par Timothy, (33), qui
qualifie les animaux comme étant responsables de l’augmentation et de l’efficacité de la dispersion des
semences, même pour les longues distances (11). Les animaux transportent les semences soit dans leur
fèces ou leur toison.
Le fonctionnement de l’écosystème prairial relève de deux grandes fonctions: la biodiversité
fonctionnelle et les cycles biochimiques. Ces deux fonctions regroupent un ensemble de processus, qui
peuvent expliquer, les effets de la gestion ou les variables du milieu, sur les flux et les états de
l’écosystème (35). La combinaison des facteurs du milieu (sol, climat) et des facteurs de gestion
(intensité du pâturage, ou autres interventions) déterminent, la composition floristique et la
performance du peuplement végétal. Dans notre étude, trois groupes fonctionnels de réponse ont été
identifiés, rendant compte de la dynamique de la composition botanique des berges.
Le premier groupe, constitué des espèces suivantes: cda Cynodon dactylon, fov Lolium
perenne, tre, Trifolium repens, qui ont des stratégies développées type " Guérilla ", (23), explorent le
milieu et colonisent les espaces sur la totalité des stations. Ces espèces peuvent contribuer à un
rétablissement rapide des écosystèmes des prairies après un pâturage, puisque le pâturage affect leur
abondance relative (12) Leurs abondances sont relativement élevées, un indice de 4, leurs est attribué
sur l’échelle de Braun-Planquet (7), ce qui correspond à un recouvrement avoisinant 70% de la
végétation totale. Ce groupe d’espèces doit être considéré dans n’importe quel projet de réhabilitation
ou d’ensemencement des pâturages détériorés
Les espèces Lpe Lolium perenne et tan Trifolium arvense constituent, le second groupe, leur
stratégie développée est du type " phalange " où les méristèmes auxiliaires se développent en touffes
denses isolées, ce sont des espèces caractérisées par des courts entre-nœuds et un système racinaire
développé, efficient dans l’absorption, abondant, et qui joue un rôle important lors des exsudats
racinaires et avec les mycorhizes. Leur abondance est variable, entre 30 et 50 % de la végétation totale
et, qui correspond à 3 sur l’échelle de Braun-Planquet. Le denier groupe d’espèces est composé de
lmu, Lolium multiflorum et mci, Medicago ciliaris ce sont des espèces non compétitives et non
résistantes aux piétinements et autres perturbations (labour), leur stratégie de développement n’est pas
adaptée, pour intercepter et valoriser les ressources du milieu. Ce groupe reste sensible au pâturage, il
est positivement affecté par la protection et négativement effectué par le pâturage Ils constituent 25%
de la végétation totale soit l’indice 1 sur la même échelle.
5.1.1. La richesse spécifique et les facteurs du milieu

Les résultats de l’AFC milieu et richesse en espèces, permet de conclure dans un premier point que la
richesse est à l’opposé de la structure sableuse des stations, cela signifie que les sols a tendance
sableuse ont une perte d’espèces significative; ceux-ci est le cas de la majorité des stations du lac
Oubeïra et en partie le lac Tonga.
Figure 9: Ordination montrant les premiers deux axes de l’AFC. Les facteurs sont la richesse en espèces,
distance de l’eau et facteurs abiotiques. La longueur du vecteur richesse est opposée à l’habitat sable
(%), le vecteur distance eau n’a pas d’effet sur la richesse en espèces, de même que les autres
facteurs
Dans le deuxième point concernant les éléments physico-chimiques étudiés, nous constatons
qu’il n’ y a pas d’effet effet direct sur la richesse spécifique, de même que pour la distance des stations
du niveau d’eau des lacs.
Pour en conclure, en général, 2 facteurs majeurs influencent donc fortement le processus de
cœxistence des espèces et les processus dynamique de la végétation et le régime de perturbation
auxquels sont soumis les habitats et les capacités de dispersion des espèces dans l’environnement, un
effet non évalué mais visible, c’est que le stock semence (seed bank) dans le sol est en baisse chaque
année
Enfin l’activité humaine reste l’élément majeur dans la destruction de la biodiversité végétale:
(1) altération de l’habitat, fragmentations et pertes, (2) surexploitation des espèces, (3) détérioration
des ressources non renouvelables (terre et eau)
Table 4: liste des espèces (abréviation) fourragères rencontrées et étudies sur les berges des lacs
Nom Vernaculaire Nom Scientifique Abréviation Famille Botanique V.F.

1 Agrostide stolonifère Agrostis stolonifera L. Ast Gramineae (=Poaceae) 6
2 Vulpin genouillé Alopecurus geniculatus L. Age Gramineae (=Poaceae) 4
3 Vulpin des prés Alopecurus pratensis L. Apr Gramineae (=Poaceae) 7
4 Cynodon dactyle Cynodon dactylon (L.) Pers. Cda Gramineae (=Poaceae) 3
5 Dactyle aggloméré Dactylis glomerata L. Dgl Gramineae (=Poaceae) 10
6 Fétuque ovine Lolium perenne L. Fov Gramineae (=Poaceae) 1
7 Ray-grass italien Lolium multiflorum Lam. Lmu Gramineae (=Poaceae) 6
8 Ray-grass anglais Lolium perenne L. Lpe Gramineae (=Poaceae) 10
9 Fléole des prés Phleum pratense L. Ppr Gramineae (=Poaceae) 9
10 Pâturin annuel Poa annua L. Pan Gramineae (=Poaceae) 2
11 Pâturin commun Poa trivialis L. Ptr Gramineae (=Poaceae) 7
12 Seigle Sécale céréale Sce Gramineae (=Poaceae) 7
13 Lotier des marais Lotus uliginosus Schkuhr Lul Fabaceae (=Leguminosae) 7
14 Luzerne ciliée Medicago ciliaris Mci Fabaceae (=Leguminosae) 7
15 Luzerne enchevêtrée Medicago intertextal Min Fabaceae (=Leguminosae) 7
16 Minette Medicago lupulina L. Mlu Fabaceae (=Leguminosae) 7
17 Trèfle des champs Trifolium arvense L. Tan Fabaceae (=Leguminosae) 4
18 Trèfle porte-fraise Trifolium fragiferum L. Tfr Fabaceae (=Leguminosae) 7
19 Trèfle des prés Trifolium pratense L. Tpr Fabaceae (=Leguminosae) 8
20 Trèfle blanc Trifolium repens L. Tre Fabaceae (=Leguminosae) 8
21 Trèfle souterrain Trifolium subterraneum L. Tsu Fabaceae (=Leguminosae) 6
22 Vesce cultivée Vicia sativa L. subsp sativa Vst Fabaceae (=Leguminosae) 6
23 Luzerne murex Medicado murex Mmu Fabaceae (=Leguminosae) 7
24 Mélilot des indes Mililotus indica Mhi Fabaceae (=Leguminosae) 3
25 Mélilot jaune Mililotus officinalis Mei Fabaceae (=Leguminosae 4
Remerciements
Nous tenons à remercier Pr. S. Plantureux ainsi que l’équipe du laboratoire d’Agronomie et
environnement de Vandœuvre lès nancy (France) pour leur aide et assistance
References
[1] Abdelguerfi A., Chapot J.Y., Conesa A.P, 1990. Contribution à l’étude de la répartition des
luzernes spontanées en Algérie, selon certains facteurs du milieu, Fourrages 113:89-106
[2] Abdelguerfi, A.1991. Les espèces spontanées du genre Medicago L. en Algérie.
Caractéristiquement des gousses et des graines de six espèces. IVe Congrès International des
terres de parcours, Montpellier, France, 1991
[3] Bensalem, K., Abdelguerfi, A., Abdelguerfi-Berrekia, R., 1990. Relations du genre Scorpulus
L. avec certains facteurs du milieu en Algérie. Fourrages 124:407-419.
[4] Botkin, BD. 1975. Fonctionnel groups of organisms in model ecosystems. In: Levin SA.
Ecosystem analysis and prediction. Society for Industial and applied Mathematics,
Philadelphia, 98-102.
[5] Bouzillé J.B. 2007. Gestion des habitats naturels et biodiversité: concepts, méthodes et
démarches. Edition Tec & Doc. Lavoisier. Paris. 325 p
[6] Braun-Blanquet, J.1964. Pflanzensoziologie. Grundzüge der vegetationskunde.spingered.,
New-York, 865p.
[7] Braun-Planquet, J. 1928. Pflanzensoziologie Springer-Verlag, Berlin
[8] De Belair, G. 1990. Structure, fonctionnement et perspectives de gestion de quatre
écocomplexes lacustres et marécageux (El-Kala, Est Algérien). Manuscrit, 193P., 36tab.,
38fig., 8 cartes. Thése Uni. Montpellier.
[9] Donjó G., Pardini, G., Gispet, M., 2004. the role of land use-land cover on runoff generation
and sediment yiels at a microplot scale, in a small Mediterranean catchment. Journal of
environment, 57: 99-116.
[10] Errol, V., et Benhouhou S., 2007. Evaluation s’un nouveau point chaud de biodiversité
végétale dans le Bassin méditerranéen (Afrique du Nord). C.R. Biologies 330:589-605.
[11] Fischer SF, Poschlod P, Beinlich B (1996) Experimental studies on the dispersal of plants and
animals on sheep in calcareous grasslands. J. Appl Ecol 33:1206–1222
[12] Fernandez-Ales, R., Laffarga, J.M. & Ortega, F. 1993. Strategies in Mediterranean grassland
annuals in relation to stress and disturbance. J. Veg. Sci. 4: 313-322.
[13] Gehu, J.M., Kaabache, M., et Gharzouli, R., 1992. L’aulnaie glutineuse de la région d’El-Kala
(Annaba, Algérie). Fitosociologia (sous press). Pavia. (Texte distribute au Congresso della
Italia di Fitosociologia à pavia les 26-28 nov-1992).
[14] Gehu, J.M., Kaabache, M., et Gharzouli, R., 1992. Observations phytosociologiques sur le
littoral Kabylie de Bejaia à Jijel. Doc. Phytosoc. N.S. 14: 305-322. Camerino.
[15] Gehu, J.M., Kaabache, M., et Gharzouli, R., 1993. Phytosociologie et typologie des rives des
lacs de la région de El-kala (Algérie). Colloques phytosociologiques XXII Syntaxonomie
typologique des habitats Bailleul (France) 298-309p
[16] Grévillot F.1992. Les écosystèmes prairiaux de la plaine alluviale de la Meuse lorraine:
phytosociologie, dynamique et fonctionnement, en relation avec les gradients hydriques et les
modifications des pratiques agricoles. Thèse, université de Metz, 217 p et annexes
[17] Grevillot, F. et Muller, S. 1996. Etude de l’impact des changements des pratiques agricoles sur
la biodiversité végétale dans les prairies inondables du Val de Meuse: présentation
méthodologie et premiers résultats. Acta botanica Gallica. 143:317-338.
[18] Grime, J.P.1979. Plant strategies and vegetation processes. John Wiley and sons.
Chichester.186 PP.
[19] IUCN, 1990. Biodiversity in-Saharan Africa and its Islands: conservation, management and
sustainable utilisation. Occasional Papers of the IUCN species survival commission
N°.6.IUCN, Gland Switzerland.
[20] Janssens, F. 1998. Restauration des couverts herbacés riches en espèces. Thèse, Université
Catholique de Louvain, 111 p.
[21] Kalpana A., Syed Ainul H., et Ruchi B. 2007. Social and economic considerations in
conserving wetlands of indo-gangetic plains: A case study of Kabartal wetland, India.
Environmentalist 27:261–273
[22] Lavorel, S., McIntyre, S., Landsberg, J., & Forbes, T.D.A., 1997. Plant functional
classifications: from general groups to specific groups on response to disturbance. Trends Ecol.
Evol. 12:474-478.
[23] Lovett-Doust, 1981. Population dynamics and local specialization in a clonal perennial
(ranunculus). I. the dynamics of ramets in contrastings habitats. J. Ecol, 69: 743-755
[24] Mac Intyre, S., Lavorel, S., Landsberg, J., et Forbes TDA. 1999. Disturbance response in
vegetation- towards a global perspective on functional traits. J. Veg. Sci, 10: 621-630.
[25] Maire, 1961, 1962. Flore de l’Afrique du Nord. Vol. 7 et 8. Paris
[26] Matthew, J.C., Mark, T.B., Keith D.S., 2006. Estimating the environmental costs of soil
erosion at multiple scales in Kenya using emergy synthesis. Agriculture, ecosystems and
Environment. 114: 249-269.
[27] McCraken, D.I.,Bibnal, E.M., Wenlock, S.E., (Eds.), 1995. Farming on the edge: the nature of
traditional farmland. Proceeding of the Fourth European Forum on Nature Conservation and
Pastoralism. Trujilo, November 1994.
[28] Muller, S., 2002. Appropriate agricultural management practices required to ensure
conservation and biodiversity of environmentally sensitive grassland sites designated under
Natura 2000. Agriculture Ecosystems and Environment, 89: 261-266.
[29] Myers, N., Mittermeier, R. A., Mittermeier, C. G. et al. 2000. Biodiversity hotspots for
conservation priorities. Nature 403: 853/858.
[30] Quezel et Santa 1962, 1963. Nouvelle flore de l’Algérie et des régions désertiques méridional.
Tome I et II, Ed. C.N.R.S. Paris. 1170 p.
[31] RAMSAR (1990). Conservation on wetlands, http://www.ramsar.org
[32] Skinner, J., Zalewski, S., 1995. Functions and values of Mediterrenean wetland,
stationbiologique de la Tour du Valet, Arles. A Med-Wet Publication, France.
[33] Timothy, J. K., 2007. The roles of seed mass and persistent seed banks in gap colonisation in
grassland. Plant Ecol 193:233–239
[34] Tzialla, C.E., Veresoglou, D.S., Papakosta, D., Mamolos, A.P., 2006. Changes in soil
characteristics and plant species composition along a moisture gradient in a Mediterranean
pasture. J. of envir manag 80: 90-98.
[35] Vertés, F., Loiseau, P. et Soussana, J.F., 2002. Fourrages, 171, 265-276.
ISSN 1450-216X Vol.21 No.2 (2008), pp.346-352
Effects of Job Stress on Health, Personal and Work Behaviour

of Nurses in Public Hospitals in Ibadan Metropolis, Nigeria
J.K. Mojoyinola
Department of Social Work, Faculty of Education, University of Ibadan
E-mail: jkmojoyinola@yahoo.com
Abstract
The study investigated the effects of job stress on the physical health, mental health
personal and work behaviours of nurses in public hospitals in Ibadan Metropolis, Nigeria. It
aimed at addressing the issue of how stress at work can be effectively managed, reduced, or
prevented by the government and hospital management boards in order to enhance the
health of the nurses, as well as improving their personal and work behaviours.
The study was carried out among 153 nurses working in two public hospitals in
Ibadan Metropolis, Nigeria.
Expost-factor research design was adopted for the study. A single questionnaire
tagged “Stress Assessment Questionnaire for Hospital Nurses (SAQFHN) was developed
and used for the study. It contains 72 items, measuring demographic variables, job stress,
physical and mental symptoms, personal and work behaviour.
Two hypotheses were formulated and tested in the study, using analysis of variance
and independent t-test. The study established that job stress has significant effect on
physical and mental health of the nurses. (F = 2.376, df = 10/143, P > .05). It also
established that there was a significant difference in personal and work behaviour of highly
stressed nurses and less stressed nurses. (t = 2.178, d = 152, P > .05).
Based on these findings, it was recommended that the government (Federal or
State) and Hospital Management Boards should improve the welfare of the nurses. It was
also recommended that their morale should be boosted by involving them in policy or
decision-making concerning their welfare or care of their patients. Their salary should be
reviewed and that they should be promoted as at when due.
Keywords: Job stress, Health, Personal and Work Behaviour, Nurse, Public Hospitals,
Metropolis.
Introduction
Stress is an important psychological concept that can affect health, well-being and job performance in
negative dimensions, (Mojoyinola, 1984; and Olaleye, 2002).
Stress according to Arnold et al (1995), is a word derived from Latin word “Stingere” meaning
to draw tight. It is regarded as a force that pushes a physical or psychological factor beyond its range of
stability, producing a strain within the individual.
Stress is the process by which environmental events (stressors or challenges) threaten us, how
these threats are interpreted, and how they make us feel (Baum et al, 1997). Lazarus (1966), conceived
stress to be a threat of anticipation of future harm, either physical or psychological events that lower an
Public Hospitals in Ibadan Metropolis, Nigeria 347
individual self-esteem. It is an affective behaviour and physical response to aversive stimuli in the
environment.
According to Selye (1976), stress is a state within the organism characterized by general
adaptation syndrome. In other word, it is the non-specific response of the body to the demand made
upon it. It suggest excessive demands that produce disturbance of physiological, sociological and
psychological systems.
Stress may be acute or chronic in nature (Akinboye et al. 2002). It exists in different forms. It
may be psychological, emotional, social, occupation or job related.
Stress experienced by workers at work is called job stress. It may be due to a number of factors
such as poor working condition, excessive work load, shift work, long hours of work, role ambiguity,
role conflicts, poor relationships with the boss, colleagues or subordinate officers, risk and danger, to
mention a few.
Certain responses indicate the presence of job stress in an individual, or group. It may manifest
by the presence of headache, sleep disturbances, difficulty in concentration, short temper, upset
stomach, job dissatisfaction and low morale (NIOSH, 1998). Other manifestations or indications of
presence of job stress include muscular tensions and ache, tightness in the chest, high blood pressure,
heat problems, snapping and arguing with others, aggressive or hostile behaviour, blaming others or
administration for tension, absenteeism and high staff on job turnover.
The above manifestations can be clearly observed in hospital nursing staff, which may have
negative effects on their health, personal and work behaviours.
The problem of this study therefore, is to investigate how job related stress affect the physical
health, mental health, personal and work behaviour of nurses in public hospitals. The study also aims at
addressing the issue of how stress at work can be effectively managed, reduced, or prevented by the
government and hospital managements in order to enhance the physical and mental health of the nurses
or improve their personal and work behaviour.
Furthermore, it attempts to find out how the quality of care and treatment the nurses give to
their patients under stressful working conditions can be improved.
Review of Relevant Study

Several studies have confirmed that the combination of high demands and low control produces job
stress and is also related to heart disease. For instance, Vitaliano et al (1990) found that physicians
whose jobs include a very high level of demands but also a high degree of control suffer less from
stress than medical students, who are burdened with undesirable combination of high demands and low
control.
Alterman et al (1994) in their study, found that the more latitude men had in making decisions
on their job, the lower was their death rate from coronary heart disease. In addition, they found that
workers in high demand and low decision jobs had an elevated risk of heart disease mortality and the
risk was greater for white-collar workers than for the blue-collar workers.
Control at work has been found to play significant role in physical and mental health of
workers.
Very low level of personal control have been found to be psychologically harmful whereas
greater control has been associated with better mental health (Evans and Carrere, 1991).
The actual use of the control the individuals have in their job to cope with stressful working
situation is a determinant of their health and well-being. For instance in a study of Dutch nurses, de
Rijik, et al (1998) found that overall job control was positively related to employees, well-being, but
for nurses who used active or control copping, high job control reduced the increase in emotional
exhaustion due to job demands. In contrast, for nurses with low active coping, and high job control
overtaxed such individuals when faced with high job demands, resulting in a lowered well-being;
having high levels of control acted as a stressor for these individuals.
348 J.K. Mojoyinola
Increased managerial pressure can impact on employee well-being. By virtue of their superior
position in an organization, managers and supervisors, intentionally or unintentionally can cause stress
for their subordinates. For example, Gauster et al (1990) found that Type A behaviour patterns
exhibited by supervisors were positively related to subordinates’ physical health symptoms. When
under pressure, many managers may react by exhibiting a negative managerial style. Managerial
support such as effective communication and feedback are important factor for employee well-being.
Poor supervisor support has been linked with increased stress level and symptoms of depression.
Bullying management style is detrimental to workers’ health. Hoel, Rayner and Cooper (1999)
found that bullying at work is linked with employee ill-health including psychosomatic stress
symptoms muscles-skeletal symptoms, anxiety and depression.
Repetti (1993) found strong evidence that jobs with the combination of high demand and low
control constitute a risk factor for hypertension and heart disease. He observed that social environment
at work is an important factor contributing to stress on the job, which play a role in both physical and
mental health.
Repetti (1993) also found that poor relationship between the superior and the workers
contribute to the level of stress experienced by the workers. He found that the workers experienced
more negative moods on days when they had distressing interactions with superiors and co-workers.
Holt (1993) found that shift work can lead to a variety of physical complaints, including sleep
and gastro-intestinal problems and can also interfere with the family life.
AlbarMarin and Garcia-Ramirez (2005) in their study examined the effect of social support on
job stress and emotional exhaustion among hospital nursing staff in Servile, South of Spain. They
found that social support had significant buttering effect on the level of stress and emotional
exhaustion experienced by the nurses at work. Nurses that received high kin support, and high levels of
co-workers and supervisors support experienced low level of job stress and emotional exhaustion than
those who did not.
Olaleye (2002) in her study among nurses working in government (state-owned) hospitals
found that job stress and burnout syndrome had greater effect on their health and coping ability at
work.
Cheng and Kawachi (2002) in their study among female registered nurses in America,
examined the association between psychosocial characteristics and health functioning. They found that
examined separately, low job control, high job demands and low work related social support were
associated with poor health status at baseline as well as greater functional decline over the four year
follow up period. When examined jointly, they found that those with low job control, high job
demands and low work related social support had the greatest functional declines. They concluded that
adverse psychosocial work conditions are important predictors of poor functional status and its
declines over time.
Occupational or job stress has been found to be negatively related to job satisfaction in nursing.
For instance, Achalu (1995) found that nurses that experienced high level of stress were less satisfied
with their career, had higher absenteeism rate and significantly less career commitment.
In a study carried out by Kennedy et al (1997), there were some evidences that creating
supportive and enabling work environment for nursing staff is a way of finding solution to the problem
of stress and burnout associated with their duties. He found that nurses who perceived their work as
supportive were more satisfied with their jobs and in their ability to provide high quality patient care.
Piko (2003) investigated the relative effects of socio-demographics, socio-economic factors,
and some variables of psychosocial work environment on the psychosomatic health of female
Hungarian nurses.
The participants were registered and students nurses drawn from public hospitals (n = 420).
Multiple regression models revealed that the frequency of stressful situations and emotionally
provoking problems, and the lack of social support from peers (i.e. measure of psycho-social work
environment) proved to be the only significant contributors to psychosomatic health complaints after
controlling for other variables in both registered and student nurses.
The various studies reviewed, point to the fact that job stress has adverse effects on physical
health, mental health, personal and work behaviour of nurses.
It is imperative therefore, that the government, hospital management boards, or employers of
nurses be committed to the reduction or prevention of high job stress experienced by the nurses. This
will enhance their welfare, efficiency and quality of care give to the patients. Hence, there is need for
further research in this direction.
Research Hypotheses
1. There will be a significant effect of job stress on physical and mental health of nurses in public
hospital.
2. There will be a significant difference in personal and work behaviour of highly stressed nurses
and less stressed nurses in public hospital.
Methodology
The study was carried out among nurses working in two public hospitals in Ibadan, Nigeria.
Expost-facto research design was adopted for the study. One hundred and fifty three nurses
working in the State Hospital, Ring Road, Ibadan, and University College Hospital, Ibadan, constituted
the sample used for the study.
The major instrument used for the study was a single questionnaire tagged “Stress Assessment
Questionnaire for Hospital Nurses (SAQFHN)”. It contains items measuring job stress, state of health,
personal and work behaviours. The items were drawn from Stress Less INC (2005). Job Stress
Assessment Scale.
The scale contains items measuring symptoms of stress, physical and mental symptoms, and
signs of personal and work behaviours. The instrument was reliably validated, yielding coefficient
alpha of 0.80 at 0.05 level of significance.
The nurses were randomly selected from five units or wards in which their tasks were highly
demanding (peadiatrics, accident and emergency, surgical, outpatients, and orthopeadic wards or
units). They were asked to assess the level of stress they experienced at work in the last six months.
They were also asked to assess the effect of such stress on their physical and mental health, as well on
their personal and work behaviour.
The data collected after two weeks were analysed, using one way analysis of variance
(ANOVA) and independent t-test. The stated hypotheses were tested at 0.05 level of significance.
Result
Hypothesis I
There will be a significant effect of job stress on physical and mental health of nurses in public
hospital. The hypothesis was put to test, using analysis of variance. This was based on items measuring
signs and symptoms of job stress and items measuring physical and mental symptoms.
The results obtained from the test are summarized in table 1 below.
Table 1: Anova Showing the Effect of Job Stress on Physical and Mental Health of Public Hospital Nurses
Source of variance Sum of squares df Mean square F-critical F-cal P Remark

Within group variance 23656.318 143 165.429 1.63 2.376 .05 Significant
Between group variance 3931.247 10 393.125
Total variance 27587.565 153 -
F = 2.376, df = 10/143, P >.05
350 J.K. Mojoyinola
Table 1 above reveals the effect of job stress on state of health of nurses in public hospital. The
table shows that there was a significant effect of job stress on physical and mental health of nurses in
public hospital (F = 2.736, df = 10/143, P >.05).
The result gives support to the first hypothesis. Hence, the hypothesis was accepted.
Hypothesis II
There will be a significant difference in personal and work behaviour of highly stressed nurses and less
stressed nurses in public hospital. The hypothesis was put to test using independent t-test. This was
based on items measuring symptoms of stress and signs of personal and work behaviour of public
hospital nurses.
The results obtained from the test are summarized in table 2 below.
Table 2: Independent T-Test Showing the Difference in Personal and Work Behaviour of Highly Stressed
Nurses and Less Stressed Nurses in Pubkic Hospitals
Standard Standard t-
Category of nurses N Mean df t-cal P Remark
Deviation Error mean critical
Highly stressed public
85 60.2471 14.36174 1.55775
hospital nurses
152 2.198 1.96 .05 Significant
Less stressed public
69 55.5662 11.77420 1.41745
hospital nurses
t = 2.178, df = 152, P >.05
The difference observed in personal and work behaviour of the public hospital nurses is
presented in table 2 above. The table revealed that there was a significant difference in personal and
work behaviour of the highly stressed nurses and less stressed nurses in public hospital (t = 2.178, df =
152, P >.05).
The result gives support to the hypothesis. Therefore, the second hypothesis was accepted.
Discussion of Findings
The results obtained from testing the first hypothesis revealed that there was a significant effect of job
stress on physical and mental health of nurses in public hospital (F = 2.376, df = 10/143, P >.05).
The result is consistent with the findings of Mojoyinola (1984) that job stress had adverse
effects on mental health and physical well-being of student nurses.
The above result is also supported by the findings of Cheng and Kawachi (2002) that women
(female registered nurses) who reported low job control, high job demand, and low work related social
support had greater declines in the physical health subscales and less improvement in mental health
subscales.
The result is consistent with the findings of Olaleye (2002) that job stress and burnout
syndrome had significant interactive effect on state of health and coping ability of nurses. It is also in
line with the finding of Wong et al (2001) that one-third of Chinese nurses had poor mental health.
Findings from the study revealed that 85 (or 55.5%) of the public health nurses experienced high level
of stress at work. This is due to work overload, lack of promotion, inadequate staffing, poor working
and salary conditions, job dissatisfaction and frustration of all kinds. The effects of the stress on their
health were manifested in form of headache, back or neck pain, muscular aches, worry, high blood
pressure, lack of concentration or attention, mental chatter, and difficulty in making decision. This
implies that both their physical and mental health were adversely affected by job stress.
In situations where nurses experienced high level of stress and their health being greatly
injured, their level of functioning at work will become greatly reduced. It is imperative therefore, that
the workload of nurses be reduced to the level they can cope with in order to become efficient at work
and give effective nursing care to their patients. Adequate nursing staff, medical equipment, dressings
and drugs should be provided in public hospitals to enable the nurses perform their duties effectively.
Their welfare should be addressed, and psycho-social needs adequately met.
The result of the second hypothesis indicated a significant difference in personal and work
behaviour of highly stressed nurses and less stressed nurses (t = 2.178, df = 152, P >.05). It is not
gainsaying the fact that the highly stressed nurses will engage in behaviours quite different from the
less stressed nurses at work.
Findings from the study revealed that the highly stressed nurses (85 or 55.5%) exhibited
personal and work behavioural problems like bullying, absenteeism, resignation or turnover. Due to
being stressed or frustrated, some of the nurses engage in aggressive hostile or vindictive behaviour on
their wards. The result was supported by the finding of Mojoyinola (1984) that junior student nurses
who experienced greater stress at work behaved differently than the senior student nurses. There were
cases of bullying, absenteeism, and turnover from them than their seniors. The result was also in line
with the finding of Wong et al (2001) that Chinese nurses adopted behavioural and cognitive coping
behaviours like direct actions, positive thinking, avoidance, resignation, alcohol use, positive coping
and negative coping when they experienced stress at work.
It should be noted that negative personal and work behaviour may not bring about positive
treatment outcomes. The patients and their families may become afraid of negative attitude and
behaviours of the nurses. For this reasons, the patients may not receive best treatment from them.
In situations where the nurses in public hospitals adopt negative personal and work behaviours
like absenteeism, apathy, dissatisfactions, tardiness irresponsibility, irritability demoralization and
withdrawal from colleagues, adequate treatment and nursing care cannot be given to the patients. They
may also experience delayed recovery.
Conclusion and Recommendations

Conclusion
Job stress is negatively associated with increased symptoms of ill-health. The complaints of physical
and mental symptoms indicated the presence of stress in the nurses, and this suggests that their state of
health is greatly injured.
As revealed in this study, the presence of stress among the nurses also make them to engage in
withdrawal, displaced or hostile aggressive behaviour to their patients and other people alike. Hence,
under stressful working conditions, they could not give humane treatment to their patients.
Recommendations
To ensure that efficient nursing care is given to the patients, the (Federal, State) the Ministries of
Health or the hospital management boards should help in reducing sources of stress in the nurses. Their
working conditions need to be quickly improved by giving them adequate salary that commensurate
with the demands of their jobs. Their promotion should be done as at when due to boost their morale.
They should also be involved in vital decisions concerning their jobs and their patients. In-service
training, workshops and seminars should be organized for nurses to update their knowledge and skills.
They should be sent for courses on human behaviour, resource management, interpersonal relation,
stress management and crisis interventions.
It is hoped that when nurses are given adequate support by their employers or when their needs
are adequately met many of them will experience less tension or stress at work. They will become less
aggressive or hostile to the patients or their families. The patients will also receive better and adequate
nursing care from them.
352 J.K. Mojoyinola
References
[1] Achulu, E. 1995. “Occupational Stress: Your Job can be dangerous to your health”. Nigeria
School Health Journal. 8(1): 46-49.
[2] Akinboye, J. O., Akinboye, D. O. and Adeyemo, D. A. 2002. Coping with Stress in Life and
Work place. Ibadan: Stirling-Horden Publishers (Nig) Ltd.
[3] Albar Marin, M. J. and Garcia-Ramirez. 2005. “Social Support and Emotional Exhaustion
among Hospital Nursing Staff”. European Journal of Psychiatry. 19, (2).
[4] Alterman, T.; Shekelle, R. B.; Vernon, S. W. and Burau, K. D. 1994. “Decision Latitude,
Psychologic demand, job strain, and coronary disease in the Western Electric Study”. American
Journal of Epidemology 139: 620-627.
[5] Arnold, J.; Cooper, C. L. and Robertson, I. T. 1995. Work Psychology: Understanding Human
Behaviour in the Work Place London: Pitman Publishing.
[6] Baum, A.; Gatchel, R. J. and Krantz, D. S. 1997. An Introduction to Health Psychology. New
York: The McGraw-Hill Companies, Inc.
[7] Cheng Y. and Kawachi, T. 2000. “Association between Psychosocial Work Characteristics and
health functioning in American Women: Prospective Study”. BMJ 320: 1432-1436.
[8] de Rijik, A. E.; Le Blanc, P. M.; Schaufeli, W. B. and de Jonge, J. 1998. “Active coping and
need for control as moderator of job demand-control model: Effects of burnout.” Journal of
Occupational and Organizational Psychology 71: 1-18.
[9] Evans, G. W. and Carrere, S. 1991. “Traffic Congestion, Perceived control and Psychological
stress among urban bus drivers”. Journal of Applied Psychology 76: 658-663.
[10] Gauster, D. C.; Schaubrock, J. and Sume, W. E. 1990. Unhealthy Leader dispositions, work
group strain and performance. Best papers, Proceedings of the Academy of Management 191-
195.
[11] Hoel, H.; Rayner, C. and Cooper, C. L. 1999. “Work Place Bullying”. In C. L. Cooper and I. T.
Robertson (eds), International review of Industrial and Organizational Psychology Chichester
U.K.: Willey.
[12] Holt, R. R. 1993. “Occupational Stress”. In L. Goldberger and S. Breznitz (eds) Handbook of
Stress: Theoretical and Clinical Aspects New York: Free Press.
[13] Kennedy, P. and Erey, N. 1997. “High pressure areas”. Nursing Times. 93 (29) 26-32.
[14] Lazarus, R. S. 1966. Psychological Stress and the coping process New York: McGraw-Hill.
[15] Mojoyinola, J. K. 1984. Occupational Stress Among The Student Nurses of University College
Hospital (U.C.H), Ibadan: It’s Effects on Their Mental Health and Well-being. Unpublished
B.Sc. Project, University of Ibadan.
[16] NIOSH. 1998. Stress at Work: Job Stress and Health. Retrieved January 9, 2006, from
http://www.cdc/gov/noish/jobstress.html.
[17] Olaleye, B. A. 2002. “Psycho-Social Effects of Job Stress and Burnout Syndrome Among
Nurses in State-Owned Hospitals in Oyo State”. M.S.W. Project, (Unpublished) University of
Ibadan, Nigeria.
[18] Piko, B. F. 2003. “Psychosocial work environment and Psychosomatic health of nurses in
Hungary”. Work and tress 17 (1): 93-100.
[19] Repetti, R. L. 1993. “The effects of work load and Social Environment on Health”. In L.
Goldberger and S. Breznitz (eds). Handbook of Stress: Theoretical and Clinical Aspects. New
York: Free Press.
[20] Selye, H. 1976. The Stress of Life New York: McGraw-Hill.
[21] Wong, D.; Leung, S.; So, C. and Lam, D. 2001. “Mental Health of Chinese Nurses in Hong-
Kong: The Roles of Nursing Stress and Coping Strategies”. Online Journal of Issues in Nursing
5 (2).
[22] Vitaliano, P. P.; Maiuro, R. D., Russo, J. and Katon, W. 1990. “Coping profiles associated with
psychiatric, physical, work and family problems”. Health Psychology 9, (3): 348-376.
ISSN 1450-216X Vol.21 No.2 (2008), pp.353-364
Modified Atmosphere Packaging of Strawberries

(Fragaria X Ananassa Duch.) Stored Under Temperature
Fluctuation Conditions
Kablan Tano
UFR of Food Science and Technology
Abobo Adjamé University, Abidjan, Cote d’Ivoire, 02 BP 801 Abidjan 02
Department of Food Science and Nutrition and Horticulture Research Center
Laval University, Sainte-Foy (Québec), Canada G1K 7P4
E-mail: pasqual_kab@hotmail.com
Tel: (225) 20 30 42 79 / 01 16 85 92; Fax: (225) 20 37 81 18
Françoise A. Kouamé
E-mail: akissifrancoise@yahoo.fr
Tel: (225) 20 30 42 79; Fax: (225) 20 37 81 18
Rose K. Nevry
Tel: (225) 20 30 42 79; Fax: (225) 20 37 81 18
Mathias K. Oulé
Faculty of Sciences, University College of Saint-Boniface
200, avenue de la Cathédrale, Winnipeg, MB, Canada R2H 0H7
Abstract
The effect of temperature fluctuation conditions on gas composition in Modified

Atmosphere Packaging (MAP) of fresh strawberry (Fragaria x ananassa Duch. cv. Kent)
and on its quality during the storage period was studied. Fresh strawberries fruits were
packaged in modified atmosphere containers and steady-state of atmosphere of 6% O2 and
15% CO2 were maintained at 4°C. Temperature was fluctuated from 4°C to 14°C during 11
days storage period in cycles: 3 days at 4°C followed by 2 days at 14°C and the cycle was
repeated twice. Temperature, atmospheric composition and quality attributes were followed
throughout the storage. Temperature increase during fluctuations caused anoxic
atmospheres inside the packages both in O2 (1.7%) and CO2 (30% to 15%). The quality of
strawberry stored under the temperature fluctuating regime was severely affected as
indicated by loss of firmness and weight, by level of ethanol in the plant tissue increase,
and infection due to physiological damage, compared to strawberry stored at constant
temperature. It was clear that temperature fluctuation, even if it should occur only once, can
seriously compromise the benefits of modified atmosphere packaging and safety of the
packaged produce. Major problems caused by temperature fluctuation must therefore be
Modified Atmosphere Packaging of Strawberries (Fragaria X Ananassa Duch.) Stored
Under Temperature Fluctuation Conditions 354
addressed to improve the usefulness and reliability of modified atmosphere packaging

technology.
Keywords: Modified atmosphere, packaging, temperature fluctuation; respiration,

permeability, strawberry.
1. Introduction
Fresh strawberries are highly perishable fruits characterized by short postharvest life at ambient
temperature because of their high respiration rates (Li and Kader, 1989) and their high susceptibility to
fungal pathogen deterioration (Woodward, 1972). They also exhibit rapid tissue degradation
accompanied by loss of firmness due to high respiration and metabolic rate. The postharvest
conservation of strawberries is very complex. A very fast metabolism leads to rapid senescence, which
may be controlled by the use of low temperature. Refrigeration is the most useful method to slow down
respiration and fungal development in fresh strawberry. Most of the physical, biochemical,
microbiological and physiological reactions contributing to deterioration of produce quality are largely
dependent on temperature. Metabolic processes including respiration, transpiration and ripening are
particularly temperature dependent. Generally, rates of biological reactions increase by a factor of 2 or
3 for each 10°C increase in temperature (Beaudry et al., 1992; Exama et al., 1993). Modified
atmosphere packaging with combination with low temperature storage extends the shelf-life of many
fruits and vegetables (Kader et al., 1989).
Modified atmosphere packaging is potentially applicable for extending self-life and delaying
postharvest decay of strawberry fruit (Chambroy et al., 1993). At 10°C, high CO2 concentration
(>10%) markedly reduces the development of fungi and the fruit has a better appearance and a firmer
texture, as noted by other researchers (Smith and Skog, 1992). However, off-odors can be produced
when strawberries are kept under higher CO2 atmosphere for more than 4 days as a result of anaerobic
respiration (El-Kazzaz et al., 1983; Li and Kader, 1989; Larsen and Watkins, 1995a). Elevated CO2
(10% to 40%) was found to degrade internal color and caused a decrease in anthocyanin content of the
internal tissue of strawberry (Gill et al., 1997). Modified atmosphere packaging can help to extend the
self-life of strawberries (Picon et al., 1993), but the results can be variable (Shamaila et al., 1992;
Larsen, 1994; Larsen and Watkins, 1995b). The changes in O2 and CO2 levels must not exceed a
certain critical threshold (Beaudry et al., 1992; Almenar et al., 2006).
The reliability of modified atmosphere packaging depends on rigorous temperature control.
Packaged fruits and vegetables are usually exposed to varying surrounding temperatures during
handling, transportation, storage and marketing. During marketing, the surrounding temperature is
usually higher than during shipping or storage. The changes in surrounding temperature create a
problem in MAP design because the rate of change of respiration is different from that of the
permeabilities of MAP films. Due to this fact, it is difficult to maintain an optimum atmosphere inside
the package when the surrounding temperature is not constant. Because of the difference in the rates of
change in permeability and respiration with temperature, a film that produces a favourable atmosphere
at the optimum storage temperature may cause excessive accumulation of CO2/or depletion of O2
inside the package at higher temperatures. This situation could lead to metabolic disorders (Beaudry et
al., 1992; Cameron et al., 1994; Joles et al., 1994).
Several studies on the modified atmosphere packaging focus on the modelling of oxygen and
carbon dioxide responses to temperature without experimental verification. The aim of the present
work was to validate the claims made by these previous researchers by evaluating the effects of
temperature fluctuation on atmospheric composition inside packages, and their impact on the quality of
fresh strawberry. The storage temperature varied between the package design temperature (4ºC) and
damaging temperature under ambient condition, at which produce are often displayed at the retail level.
355 Kablan Tano, Françoise A. Kouamé, Rose K. Nevry and Mathias K. Oulé
2. Materials and Methods

2.1. Materials and Storage Conditions
Mature strawberry (Fragaria x ananassa Duch. Cv. Kent), one-fourth to one-half red on the
surface, were picked from commercial field in “Ile d’orléans”, Quebec, Canada. The fruits were
transported within an hour to the laboratory and held at the optimum temperature of 4oC. The
fruits of uniform size, color and free of external blemishes or infections were then selected. After
12 hours, the pre-cooled fruits were randomly packed in 4-L plastic containers. Each package
containing 1.0 kg of strawberry was fitted with diffusion window for gas exchange. The window
provided an O2 flux rate of 5.58x10-12 mol s-1 pa-1 and CO2 flux rate of 13.55x10-12 mol s-1 pa-1 at
4C. The flux rates for O2 and CO2 at 14C were 5.75x10-12 mol s-1 pa-1 and 14.48x10-12 mol s-1
pa-1, respectively. With the O2 and CO2 flux rates, a steady - state modified atmosphere of 6% O2
and 15% CO2 was obtained in the packages at 4C. The strawberry packages were divided up in
to 3 groups and the first group was stored at a constant temperature of 4°C (package design
temperature) for 11 days, while the second group of packages was stored at temperatures of 4°C
for three days and 14°C for two days alternatively, and the sequence was repeated 2 times during
the 11-day storage period. A third group of treatments was stored in open containers (4-L) and
exposed to regular atmosphere at the product optimum storage temperature, with a relative
humidity (RH) of 95%, and served as the control group. Each group consisted of 18 containers,
of which 3 containers were used for monitoring the atmospheric composition. For strawberry
quality analysis, triplicate containers were removed at the beginning and the end of each
temperature change. Two experiments were carried out. The following determinations were made
in each experiment: gas determination, weight loss, color, texture, pH and titratable acidity and
the infection percent.
2.2. Package Atmosphere

The temperature of the atmosphere inside the packages was monitored using type T thermocouple
probes (POD-237/236, Omega Engineering, Stamford, CT, USA) and a datalogger (Model RR2-1200-
2, Rustrak Ranger II, Automatic RP Inc., Quebec City, Canada).
Carbon dioxide and oxygen concentrations inside the packages were monitored as a function of
time by gas chromatography (Perkin Elmer, Model 8500) using a thermal conductivity detector. Gas
samples of 1 ml were drawn using polypropylene syringes through a septum from each package. The
measurements were carried out in triplicate.
2.3. Fermentation Products

Acetaldehyde and ethanol production in strawberry tissue was measured by head-space gas
chromatographic analysis. A sample of 15 strawberries fruits per replicate were ground in a mortar, and
2 g of ground sample was placed in a sealed tube which was placed in a head space sampler (Hewlett
Packard, Model 19395A), maintained at 90C. After 20 min, the vapour accumulated in the headspace
was immediately analysed by gas chromatography (GC) (Hewlett Packard, Model 5890A), using a
capillary column (DB-225, 30 m; 0.25 mm diameter; film thickness, 0.50 µm). The volume of the
headspace gas injected in the GC was 1 ml. The temperatures of the injector and the detector were
155C and 250C, respectively. The temperature programming used was: initial temperature, 35C for
4 min; final temperature, 75C and heating rate, 20C/min. The measurements were carried out in
triplicate for each treatment at the beginning and the end of each fluctuation.
2.4. Titratable Acidity

A sample of 15 fruits per replicate was homogenate, and the juice was analyzed for titratable acidity
(TA). A sample of 4 g juice was diluted with 20 ml of distilled water; pH and TA were measured
(Titration to pH 8.1) using an automatic titrator fitted with pH meter and ABU80 autoburette;
Radiometer, Copenhagen, Danmark). TA was expressed as percent citric acid.
2.5. Quality Attributes

2.5.1. Color
Fruits samples were analysed initially at the beginning and at the end of each temperature change
during the 11-day storage period. Strawberries exhibiting any sign of decay were discarded. Then,
external color was measured on opposite sides of 15 fruits for each treatment using a Minolta
chromameter (model CR-200; Minolta, Ramsey, NJ) which provided CIE L*, a* and b* values. The
values were used to calculate chroma (C=[a*2+b*2]1/2, which indicate the intensity or color saturation,
and hue angle (ho=arctangent[b*/a*]), where 0o = red-purple; 90o = yellow; 180o = bluish-green; and
270o = blue (McGuire, 1992). A mean from determinations on 15 randomly selected strawberries was
obtained for each treatment.
2.5.2. Firmness, weight loss and infection

Firmness was determined on 15 fruits per replicate sample, as described by Ahmed et al.(1972) with an
Instron Universal Testing Instrument (Model TMS, Instron Canada, Burlington, Ontario, Canada). A
75 mm diameter plunger was used to puncture the fruit to a depth of 4 mm with a load cell of 20 N.
The speed of cross head was 2 cm/min. The fruits were sliced off on side so they could be placed
firmly on the flat base plate. Firmness was measured at the bio-yield point and was expressed in
Newtons (N/mm).
Weight loss was also determined during the storage at days 0, 3, 5, 7, 9 and 11, by monitoring
the weight of the package content before and after storage. Weight loss was expressed as the
percentage of weight loss with respect to the initial weight and was determined in triplicate.
Strawberries were inspected at the beginning and the end of temperature change to determine
the extent of decay until the end of the 11-day storage period. Diseases incidence was expressed as the
percentage of infected fruits per container. Agents causing decay were identified by visual
characteristics (Barnett and Hunter, 1972).
2.6. Statistical Analysis

The experiments were repeated twice. Since there was no significant difference between the 2
experiments, the results were pooled and averaged. Experiments were laid out in a completely
randomized block design with three replications. Data on lightness (L), chroma, angle hue, weight loss,
ethanol, acetaldehyde levels and infection were submitted to an analysis of variance, followed by
Neuwman-Keul’s multiple comparison test (alpha = 0.05)
3. Results and Discussion

3.1. Package atmosphere
The temperature inside strawberry packages subjected to both constant temperature and fluctuating
temperature conditions are shown in Fig. 1a. When the ambient temperature was increased from 4°C to
14°C, the package temperature equilibrated to the ambient temperature over a period of 12 h. The
temperature change inside the package was depended on the package void volume.
The CO2 concentration increased from 0% to 12.8%, while the O2 level fell from 21% to 6.9%,
for both the constant temperature (4°C) group and the temperature fluctuating (4°C - 14°C) group
within 72 h (Fig. 1b,c). Throughout the 11-day storage period, the O2 concentration inside the packages
stored at constant temperature of 4°C, remained stable at steady O2 levels of 7% (Fig. 1b,c). The CO2
concentration inside the package kept at constant temperature of 4°C reached the level of 15.2% after 5
days storage. After the 5 day storage at 4°C, the CO2 concentration remained stable at steady CO2 of
16%. Although the O2 and CO2 levels achieved for strawberries stored at constant temperature were
close to the recommended (6%O2 and 15% CO2), slightly higher levels were obtained for O2 and CO2
(7%O2 and 16%CO2) at the end the 11 days storage period. The similar results have been obtained by
Silva et al. (1999).
In the packages subjected to temperature fluctuation, during a temperature increase from 4°C to
14°C in the first fluctuation cycle, the CO2 concentration increased rapidly, reaching 18.5%, 24.9% and
26.1% after 14, 40 and 48 h, respectively (Fig. 1b), while the O2 concentration fell from 7.0 to 4.1%
and 1.9% after 14 and 40 h, respectively (Fig. 1c). The O2 concentration was down to 1.5% after 24 h
(Fig. 1c). When the temperature was lowered to 4°C after the first 2-day period at 14°C, the CO2
concentration decreased to 18.5%, whereas the O2 concentration increased to 8.3%. During the
temperature increase of the next fluctuation cycle, the CO2 level rose to 30%, higher than the previous
level of 26%, while the O2 level fell again to 2.1%. Carbon dioxide and oxygen levels in the strawberry
packages subjected to temperature fluctuations followed the temperature changes, but the maximum
CO2 accumulation rose with each subsequent cycle. A temperature change may induce a considerable
modification of equilibrium gas composition inside the modified atmosphere packages of fresh
strawberries (Renault et al., 1994). This is due to the small variations of gas diffusion proprieties in the
package films compared to strawberry respiration variation as function of temperature change. To fully
appreciate the impact of temperature on the respiration of strawberry and on the permeability of films,
the QR10 and QP10 values can be compared. Value of QR10 for strawberry was 3.0 while the QP10 of the
package was lower than 1.2 (Exama et al., 1993). This disparity resulted in an accumulation of CO2
and a decrease in O2 inside the packages subjected to temperature fluctuations.
Figure 1: Changes in temperature (A), oxygen (B) and carbon dioxide (C) levels inside strawberries modified
atmosphere packages stored at constant temperature (, 4C) and under temperature fluctuating
conditions (, 4C to 14C). LSDs at 5% are presented.
16
(a ) L S D = 1.0
14
(oC)
T em pe ra tu re
12
10
6
4
2
(b ) L S D = 0.3 1
30
25
C O2 (% )
20
15
10
5
0
(c ) L SD = 0.2 8
18
15
O2 (% )
12
6
3
0
0 2 4 6 8 10 12
S tor ag e tim e (day s)
3.2. Fermentation Products

The acetaldehyde concentrations in the tissue of strawberries for the three treatments increased slightly
during the first three days of storage but there was no significant difference between theses treatments
(Fig. 2a). The difference appears from the fifth day of storage between the packaged strawberries
submitted to temperature fluctuations regime and those stored at constant temperature with and without
packaging. The concentrations of acetaldehyde in the tissue of strawberries stored at constant
temperature with and without packaging became significantly different from the seven day until the
end of the storage period (Fig. 2a). At the end of the 11 days of storage period, the levels of
acetaldehyde in the tissue of strawberries packaged and stored under temperature fluctuation conditions
and at constant temperature and unwrapped strawberries stored at constant temperature were 5.4
mg.kg-1, 4.8 mg.kg-1 and 4.5 mg.kg-1, respectively (Fig. 2a).
Acetaldehyde is a very volatile intermediate product of anaerobic respiration in fruits and
vegetables. Most fruits and vegetables respire aerobically in the presence of sufficient O2. At low O2
concentrations, however, fermentation is to be expected, but there will be a brief time during which
there is insufficient NADH to reduce all of the acetaldehyde to ethanol, resulting in the accumulation
of acetaldehyde. With time, however, ethanol predominated in the tissues of the strawberry. Hertog et
al. (1999) reported that at O2 levels below 5% fermentation occurs, evident from an increase in
respiration quotient (RCO2/RO2). However, the slight accumulation of acetaldehyde in air stores may
be a result of physiological breakdown caused by over – ripeness as reported by Smagula and
Bramlage (1977).
The levels of ethanol in the tissue of strawberries stored under temperature fluctuation
conditions reached 120 mgkg-1 and 195mg.kg-1 at the end of the first two fluctuations, respectively
(Fig. 2b). However, when the temperature was lowered to the optimum storage temperature of 4oC
after each fluctuation, the level of ethanol decreased but did not reach the previous levels. After the
first and the second fluctuations regime, and then the optimum temperature of 4oC were restored, the
concentrations dropped from 120 mg.kg-1 to 95 mg.kg-1 and from 195 mg.kg-1 to 150 mg.kg-1,
respectively (Fig. 2b). Ethanol and acetaldehyde concentrations in strawberries stored at constant
temperature of 4oC in air and in MA package stored at constant temperature became significantly
different (p<0.05) since the fifth day until the 11 days of storage. At the end of the 11 days storage,
ethanol concentrations in the tissue of strawberries stored in air and in MA at constant temperature and
in MA package submitted to temperature fluctuations were 62.2 mg.kg-1, 70.2 mg.kg-1 and 153.8
mg.kg-1, respectively (Fig. 2b).
High CO2 stress has a role in the induction of anaerobic respiration and ethanol production
(Kimmerer and Kozlowski, 1982). Kader (1986) emphasized that the effects of both low O2 and high
CO2 concentrations in the induction of anaerobic respiration are additive. Almenar et al. (2006)
reported that the induction of anaerobic respiration was assigned to the effects of both low O2 and high
CO2 concentrations. Thus, the concentrations of ethanol increased during storage, with higher
increments as the CO2 partial pressure increased and the O2 partial pressure decreased. This behavior
was in agreement with most reports in the literature on cultivated strawberries and can be explained as
a result of anaerobic respiration induced by these gases (Larsen and Watkins, 1995a; Larsen and
Watkins, 1995b)
Ethanol in the tissue stored under modified atmosphere at constant temperature suggests partial
induction of anaerobic respiration when the O2 concentration dropped below 10% and the CO2
concentration rises above 5% (Kader, 1987). This degree of anaerobic metabolism did not have adverse
effects on product quality.
3.3. Titratable Acidity

Modified atmosphere both at constant temperature and under temperature fluctuating regime resulted
in the maintenance of higher titratable acidity of the fruits compared to control where acidity fell
sharply during the 11 days storage. From the fifth day onward, differences between the three
treatments were significant (p<0.05) (Fig. 2c). After 11 days of storage at 4ºC, the titratable acidity
passed from 0.88% to 0.76%. Under temperature abuse conditions (4ºC to 14ºC), the titratable acidity
fell to 0.71%. The titratable acidity of the control strawberries reached 0.61% after the 11 days storage
(Fig. 2c). As would be expected from the above-mentioned acidity values, strawberry pH values
presented a slight increase.
Although, Li and Kader (1989) did not measure any change in TA in “Selva” strawberries
stored in CA treatments. On the other hand, Ke et al. (1991) measured a slight increase in juice pH of
“Selva” strawberries stored in elevated CO2 atmospheres. Holcroft (1998) reported that the effect of
high CO2 concentrations could affect organic acid metabolism and consequently the pH, an explanation
that is in agreement with our results. The acidity reduction during the low temperature storage was in
agreement with the results showed for cultivated strawberries (Holcroft and Kader, 1999; Pérez and
Sanz, 2001;)
Figure 2: Changes in acetaldehyde (A), ethanol (B) concentrations and the titratable acidity (c) of strawberries
stored in air at 4C (▲), in modified atmosphere packages stored at constant temperature (, 4C)
and under temperature fluctuating conditions (, 4C to 14C). LSDs at 5% are presented.
5,5
(a)
Acetaldehyde (mg.kg-1)
4,5
4 LSD=0.15
3,5
180 (b)
Ethanol (mg.kg-1)
140
LSD=3.78
100
60
20
Titratable acidity (% citric acid)
LSD=0.02
0,85
(c)
0,8
0,75
0,7
0,65
0,6
0 2 4 6 8 10 12
Storage time (days)
3.4. Quality attributes

3.4.1. Color
The skin color of air-stored strawberries darkened (L* value decreased), became less intensely red
(Chroma decreased) and was more scarlet than orange-red (hue angle decreased) in comparison to
fruits stored under MA packages at constant temperature or submitted to temperature fluctuation
conditions (Figs. 3a, b, c). Similar changes, although less extreme, were observed in fruits stored in
MA packages at constant temperature and under temperature abuse conditions. The changes in L*, and
hue angle were significantly different (p<0.05) between the three treatments after 3 days storage (Fig.
3a).
For chroma, there was no significant different (p>0.05) between MA storage with fluctuating
temperature conditions and MA storage at constant temperature (Fig. 3b). Nunes et al. (1995) found
that strawberries stored in 5 kPa O2 + 15 kPa CO2, or 10 kPa O2 + 20 kPa CO2 at 4ºC were lighter,
redder (higher a*) and a more intense color (higher chroma) than fruits stored in air. The hue angle was
also higher indicating a more orange – red color. The difference between the color of the fruits stored
MA at constant temperature and those submitted to temperature fluctuations was explained the by
specific effect of high temperature and its effect on the modification of the atmosphere surrounding the
packages.
Figure 3: Changes in Lightness (L) (a), Chroma (b) and hue angle (c) of strawberries stored in air at 4C (▲),
in modified atmosphere packages stored at constant temperature (, 4C) and under temperature
fluctuating conditions (, 4C to 14C). LSDs at 5% are presented.
42
LSD=0.54
(a)
40
lightness (L)
38
36
34
(b)
LSD=0.49
48
Chroma (C*)
46
44
42
40
38
(c) LSD=0.47
38
Hue angle
36
34
32
30
0 2 4 6 8 10 12
Storage time (days)
3.4.2. Firmness, weight loss and infection

The firmness of strawberries stored at constant temperature inside modified atmosphere packages
decreased from 4.40 N to 4.05 N after 5 days of storage. From the fifth day onward, the firmness
decreased sharply to reach the value of 3.40 N (Fig. 4a). This value of 3.40 N was reached after 5 days
when the packages were submitted to temperature fluctuations regime. Strawberry firmness showed a
significant difference between the two modified atmosphere packages treatments from the third day of
storage. Air stored Strawberries at 4ºC softened considerably after 11 days, to 2.50 N (Fig. 4a).
Holcroft and Kader (1999) reported that after 10 days at 5ºC, strawberries stored in air softened
considerably. This observation was in agreement with our results.
Several studies showed that the increase of CO2 concentrations favour the increase of
strawberries firmness. However, with the ripening, the firmness can decrease. Otherwise, the excessive
accumulation of CO2 and high temperature (what is the case when the packages were submitted to
temperature fluctuation conditions) caused the physiological damages with the loss of firmness. Harker
et al. (2000) reported that the storage of strawberries at high temperature decreases their firmness, what
was in agreement with our results.
Unpackaged strawberries lost about 5% of their initial weight after only 11 days of storage (Fig.
4b). This water loss had a negative effect the strawberry fruit appearance, leading to shrivelling and a
dull appearance of the epidermis. From the third day onward, the results showed, however, a
significant (p<0.05) difference between weight loss under MA at constant temperature of 4ºC and that
subjected to temperature fluctuation (Fig. 4b). Strawberry stored in modified atmosphere packages at
constant temperature of 4ºC exhibited only a weight loss of 0.25% at the end the 11 days of storage.
Strawberry subjected to temperature fluctuating regime exhibited a slight increase of weight loss
during the first 7 days of storage. From the seventh day onward, weight loss increased significantly and
reached the value of 1.8% at the end of the 11 days storage (Fig. 4b).
Increasing the temperature decreased the relative humidity inside the packages and increased
the water vapor deficit. The transpiration rate of fresh produce increases with water vapor deficit (Ben -
Yehoshua, 1987; Roy et al., 1995). Otherwise, the high rate of transpiration due to temperature
increased, accelerated produce weight loss. The atmosphere in MAP maintained at constant
temperature was saturated with moisture, which may explain the difference between weight losses in
packages kept at constant temperature and those subjected to temperature fluctuations.
Fungal decay occurred rapidly in unpackaged strawberries stored at 4ºC, with 72.5% of fruits
showing signs of infection after 11 days of storage (Fig. 4c). The decay observed was identified as gray
molds rot, caused by botrytis cinerea 36 (Dennis, 1983). Botritis infection is a major factor limiting
keeping quality of strawberries, with the potential to cause up 50% loss (Hertog et al., 1999; Wszelaki
and Mitcham, 2000). Modified atmosphere packages stored at a constant temperature of 4ºC slowed
down the appearance of gray moulds rot in fruits. Loss due to infection increased slightly but not
reached the limit for acceptance of 10% loss of marketable fruits in a package, after 11 days of storage
(Fig. 4c). Optimal modified atmospheres (6%O2 and 15%CO2) at the optimal storage temperature of
4ºC reduced decay to some extend delayed softening more than air storage. Significant differences
(p<0.05) appeared between the packaged strawberries stored at constant temperature and those
submitted to temperature fluctuations. Excessive CO2 accumulation up 30%, alone or in combination
with very O2 concentrations affected strawberry shelf-life (Holcroft and Kader, 1999).
Figure 4: Changes in firmness (a), weight loss (b) and pourcentage of infection (c) of strawberries stored in air
at 4C (▲), in modified atmosphere packages stored at constant temperature (, 4C) and under
temperature fluctuating conditions (, 4C to 14C). LSDs at 5% are presented.
1,1
(a) LSD=0.03
firmness (N/mm)
0,9
0,8
0,7
LSD=0.12
(b)
5
Weight loss (%)
0
(c)
60
infection (%)
LSD=5.40
40
20
0
0 2 4 6 8 10 12
Storage time (days)
4. Conclusions
The temperature fluctuation encountered in the cold chain during transportation and storage of fresh
strawberries has a considerable impact on both the atmosphere inside the packages and on fruits
quality. Rapid decreases in O2 concentrations accompanied by increases in CO2 concentrations lead to
anaerobic respiration characterized by production of ethanol. The study confirmed that the storage of
the strawberries under optimum modified atmosphere (6 to 7% O2 and 15% CO2) and at the optimal
temperature of 4ºC permitted to reduce considerably the infection and prolong the self life. The
decrease of O2 and the increase of CO2 concentrations at these resulted in the maintaining of the
strawberries firmness at an acceptable level. However, the excessive increase of CO2 concentrations up
to the acceptable level during temperature fluctuating period resulted to the deterioration of the quality
of the strawberries (odour, firmness, infection and color). This loss of quality is assigned to the CO2
that causes physiological damages at high concentrations. The changes in the physiology of
strawberries observed in this study may be interesting as they help to retain some quality attributes of
the fruits. In particular, the lower titratable acidity, the loss of the firmness, suggest that temperature
fluctuation and its impact on the atmosphere inside modified atmosphere package of strawberry had an
irreversible effect on the quality of the strawberries.
References
[1] Ahmed, E.M., Dennison, R.A. and Fluck, R.C. 1972. Textural properties of stored and
irradiated Tioga strawberries. J. Text. Studies 3: 80-88.
[2] Almenar, E., Hermandez-Manoz P., Lagaron J.M., Catala, R. and Gavara, R. 2006. Controlled
atmosphere storage of wild strawberry fruit (Fragaria vesca L.). J. Agric. Food Chem. 54: 86-
91.
[3] Barnett, H.L. and Hunter, B.B. 1972. Illustrated Genera of imperfect Fungi. Burgess
Publishing, Minneapolis, Minn., U.S.A., pp.1-241.
[4] Beaudry, R. M., Cameron, A.C., Shirazi, A. and Lange, D.D. 1992. Modified atmosphere
packaging of blueberry fruit: Effect of temperature on package oxygen and carbon dioxide. J.
Amer. Soc. Hortic. 117: 436-431.
[5] Ben-Yehoshua, A., 1987. Transpiartion, water stress, and gas exchange. In: Postharvest
physiology of vegetables, ed. by J. Weichman. Marcel Dekker, New York, NY, pp.113-170.
[6] Cameron, A. C., Beaudry, R. M., Banks, N. H. and Yelanich, M. V. 1994. Modified atmosphere
packaging of blueberry fruit: Modelling respiration and package oxygen partial pressures as a
function of temperature. J. Amer. Soc. Hort. Sci. 119: 534-539.
[7] Chambroy, Y., Guinebretiere, M.H., Jacquemin, G., Reich, M., Breuils, L. and Souty, M. 1993.
Effects of carbon dioxide on shelf-life and post-harvest decay of strawberry fruit. Science des
Aliments, 13: 409-423.
[8] Dennis, C. 1983. Soft fruits. In: Postharvest pathology of fruits and vegetables. C. Dennis (Ed.)
Academic Press, London, pp. 23-42.
[9] El-Kazzaz, M.K., Sommer, N.F. and Fortlage, R.J. 1983. Effect of different atmospheres on
post-harvest decay and quality of fresh strawberries. Phytopathology. 73: 282-285.
[10] Exama, A., Arul, J., Lencki, R.W., Lee, L.Z. and Toupin, C. 1993. Suitability of plastic films
for modified atmosphere packaging of fruits and vegetables. J. Food Sci. 58: 1365-1370.
[11] Gil, M.I., Holcroft, D.M. and Kader, A.A. 1997. Changes in strawberry anthocyanins and other
polyphenols in response to carbon dioxide treatments. J. Agric. Food Chem. 45: 1662-1667.
[12] Harker, F.R., Elgar H.J., Watkins, C.B, Jackson’ P.J. and Hallett, I.C. 2000. Physical and
mechanical changes in strawberry fruit after high carbon dioxide treatments. Postharvest Biol.
and Technol. 19: 139-146.
[13] Hertog M.L.A.T.M., Boerrigter H.A.M, Van Der Boogaard G.J.P.M, Tijskens L.M.M. and Van
Schaik A.C.R. 1999. Predicting keeping quality of strawberries (cv. Elsanta) packed under
modified atmospheres: an integrated model approach. Postharvest. Biol. Technol. 15: 1-12.
[14] Holcroft, D.M. and Kader, A.A., 1999. Controlled atmosphere-induced changes in pH and
organic acid metabolism may affect color of stored strawberry fruit. Postharvest Biol. Technol.
17: 19–32.
[15] Holcroft, D.M., 1998. Biochemical basis of changes in the colour of strawberry fruit stored in
controlled atmospheres. Ph.D. Dissertation, University of California, Davis.
[16] Joles, D.W., Cameron, A.C., Shirazi, A., Petracek, P.D. and Beaudry, R.M., 1994. Modifed-
atmosphere packaging of “Heritage” red rasberry fruit: respiratory response to reduced oxygen,
enhanced carbon dioxide, and temperature. J. Am. Soc. Hortic. Sci. 119: 540-545.
[17] Kader, A.A. 1986. Biochemical and physiological basis for effects of controlled and modified
atmospheres on fruits and vegetables. Food Technol. 40: 99-104.
[18] Kader, A.A., 1987. Respiration and gas exchanges of vegetables. In: Postharvest physiology of
vegetables. J. Weichman (Ed.). Marcel Dekker, New York. pp. 25-43.
[19] Kader, A.A., Zagory, D., Kerbel, E.L., 1989. Modified atmosphere packaging of fruits and
vegetables. CRC Rev. Food Sci. Nutr. 28: 1-30.
[20] Ke, D., Goldstein L., O’Mahony, K. and Kader A.A. 1991. Effects of short-term exposure to
low O2 and high CO2 atmospheres on quality attributes of strawberries. J. Food Sci. 56: 50-54.
[21] Kimmerer, T.W. and Kozlowski, T.T. 1982. Ethylene, ethane, acetaldehyde and ethanol
production by plants under stress. Plant Physiol. 69: 840-847.
[22] Larsen M. 1994. Flavour changes in strawberries packed in modified atmospheres. Acta Hortic.
368: 78-82.
[23] Larsen M. and Watkins, C.B. 1995a. Firmness and concentrations of acetaldehyde, ethyl acetate
and ethanol in strawberries stored in controlled and modified atmospheres. Postharvest. Biol.
Technol. 5: 39-50.
[24] Larsen, M. and Watkins, C.B. 1995b. Firmess and aroma composition following short-term
high carbon dioxide treatments. HortScience. 30: 303-305.
[25] Li, C. and Kader, A.A. 1989. Residual effects of controlled atmospheres on post-harvest
physiology and quality of strawberry. J. Am. Hort. Sci. 114: 629-634.
[26] McGuire, R.G. 1992. Reporting of objective color measurements. HortScience. 27:1254–1255.
[27] Nunes, M.C. N., Brecht, J.K., Morais, A.M. M.B. and Sargent, S.A. Physical and chemical
quality characteistics of strwberries after storage are reduced by a short delay to cooling. Posth.
Biol. Technol. 1995. 6: 17-28.
[28] Perez A.G. and Sanz C. 2001. Effects of high-oxygen and high-carbon-dioxide atmospheres on
strawberry flavour and other quality traits. J. Agric. Food Chem. 49: 2370-2375.
[29] Picon A., Martinez-Javega J.M., Cuquerella J., Del Rio, M.A. and Navarro P.1993. Effects of
precooling, packaging film, modified atmosphere and ethylene absorber on the quality of
refrigerated Chandler and Douglas strawberry. Food Chem. 48: 189-193.
[30] Renault, Houal L., Jacquemin G., and Chambroy Y., 1994. Gas exchange in modified
atmosphere packaging. 2: Experimental results with strawberries. Inter. J. of Food Sci. and
Technol. 29: 379-394.
[31] Roy, S., Anantheswarn, R.C. and Beelman, R.B., 1995. Fresh mushroom quality as affected by
modified atmosphere packaging. J. Food Sci. 60, 334-340.
[32] Shamaila, M., Powrie, W.D.and Skura, B.J. 1992. Sensory evaluation of strawberry fruit stored
under modified atmosphere packaging (MAP) by quantitative descriptive analysis. J. of food
Sci. 57: 1168-1172.
[33] Silva F.M., Chau K.V., Brecht J.K. and Sargent S.A. 1999. Modified atmosphere packaging for
mixed loads of horticultural commodities exposed to two postharvest temperatures. Postharvest
Biolology and Technology. 17: 1-9.
[34] Smagula, J.M. and Bramlage, W.J. 1977. Acetaldehyde accumulation: it is a cause of
physiological deterioration of fruits,. HortScience. 12: 200-203.
[35] Smith, R.B. and Skog, L.J. (1992). Post-harvest carbon dioxide treatment enhances firmess of
several cultivars of strawberry. HortScience. 27: 420-421.
[36] Woodward, J.R. and Topping, A.J. 1992. The influence of controlled atmospheres on the
respiration rates and storage behaviour of strawberry fruits. Journal of horticultural Science.
47:547-553.
[37] Wszelaki A.L. and Mitcham E.J. 2000. Effects of superatmospheric oxygen on strawberry fruit
quality and decay. Postharvest Biol. Technol. 20: 125-133.
ISSN 1450-216X Vol.21 No.2 (2008), pp.365-375
A Comparative Study of Engine Performance by Using

Different Bio-Fuels
M. Sadat Ali
Asst. Professor, Department of Mechanical Engineering
Jagannath Institute For Technology & Management- Paralakhemundi-761 211, INDIA
E-mail: alisadat@rediffmail.com, mir.sali@jitm.ac.in
P.S.V. Ramana Rao

Asst. Professor, Department of Mechanical Engineering
C.V. Gopinath
Professor, Department of Mechanical Engineering
Abstract
BIODIESEL is natural oil that is taken out of several non-edible oil-bearing plants.
It undergoes organic and chemical processes to yield fatty acid esters, or a diesel like
product. It can be blended with normal diesel. Like ethanol, bio-diesel can be used in
mixing it with anything. Because of invent of different bio-diesel preparing methods it is
tried by many researchers to make bio-diesel from different type of vegetable oils. In this
paper an attempt has been made to study the utility of some of the oils like castor, Mohua
& Karanj in making bio-diesel and compare the properties of the fuel. Also an attempt has
been made to check the performance characteristics of the engines by using these fuels. For
the same engine the power developed by different oils were compared. It was found that
power developed by Castor oil is more than all other oils. The fuel consumed increases
with the increase in load. But the increase was more in case of castor oil where as it follows
a linear trend for all other oils. The increase in brake power was also linear up to a limit for
increasing load. After attaining a maximum value the brake power decreases.
The brake thermal efficiency also shows a similar trend for increasing load
conditions. It increases with increase in load up to a certain limit but after attaining the
maximum value it decreases. All the properties of the fuels and the performance of engine
varies linearly from one trend to other as the blend changes proportionately. The optimum
blend for getting maximum power output and best overall efficiency is yet to be
experimentally decided for future use of bio-diesel fuels in the conventional diesel engines.
Introduction
The traditional definition of fuel is: Fuels are substances which can be burned to obtain heat. With
recent advancements in energy technology, the fuels also include substances used in nuclear reactions,
fuel cells,thermionic generators, biogas plants, fuel processes etc. Fuels are primary, intermediate and
secondary substances having high energy density. Fuels are used for heating water, room heating,
cooking. Fuels are essential for processes like heating, melting, desalination of water, refined chemical
A Comparative Study of Engine Performance by Using Different Bio-Fuels 366
reactions, industrial processes etc., metallurgical and chemical processes etc. Fuels are also used in
internal combustion engines for transportation, mechanical drives, electrical power generation, rockets,
space travel etc. Wood, animal-dung, agricultural waste were the principal fuels used during
agricultural age. These, fuels are being used today even in rural areas and in developing countries.
Energy agricultural farms will provide renewable energy resources during coming decades and
centuries. Fossil fuels are major Source of energy. Since about 1850,the consumption of fossil fuels:
coal, Petroleum oil and natural gas has been increasing at an annual growth rate of 3.5% to 4.5%.The
major supply of world energy is through fossil fuels. Presently we are passing through the peak period
of fossil age. The coal reserves are likely to last for about one thousand years and crude oil/natural gas
reserves for about four hundred years. Thus the fossil age will cover only a span of about 1000years of
human history. Probably the children born today May witness the severe depletion of petroleum oils
and natural gas during their lifetime. Alternate fuel technologies are being developed. Biofuel is one of
the most alternatives.
Biodiesel
BIODIESEL is natural oil that is taken out of several non-edible oil-bearing plants. It undergoes
organic and chemical processes to yield fatty acid esters, or a diesel like product. It can be blended
with normal diesel. Like ethanol, biodiesel can be used in mixing it with anything.
Biodiesel is the name for a variety of ester -based oxygenated fuels made from vegetable oils or
animal fats. Biodiesel is the only alternative fuel that can be used directly in existing, unmodified
diesel engine. Because it has similar properties to petroleum diesel fuel.
Biodiesel usage
A biodiesel blend is a pure biodiesel blended with petrodiesel. It is referred to as Bxx. The “xx”
indicates the amount of biodiesel in the blend. For example a B20 is 20% biodiesel & 80% petrodiesel.
The specification of biodiesel is designed to ensure that consumers will not experience
operational problem from the fuel’s use. Make sure that biodiesel meets this specification and that the
fuel supplier will warrant this fact. Quality fuel will provide the customer with improved air quality
and enhanced operability. Poor quality fuels create problems and increases maintenance activity.
Consumers must look into the following factors while using biodiesel. These are described
below.
 Check fuel filters on the vehicle & in the delivery system frequently upon initial use &
change them as necessary.
 Be aware of biodiesel’s freezing properties & take precautions in cold weather.
 Wipe painted surface immediately after using biodiesel as the fuel is a good solvent.
 Store biodiesel or its blend in a safety can to avoid spontaneous combustion.
 Use the biodiesel within one year since they have shelf life. It should not be stored for a
long time.
Emission
Biodiesel is the first & only alternative fuel to have a complete evaluation of emission results &
potential health effect submitted to the U.S. Environment Protection Unit. The data gathered through
these test complete the most thorough inventory of the environmental and human health effects
attributes that current technology will allow.
The exhaust emissions of particular matter from biodiesel are found to be 30% lower than
overall particulate matter emission from diesel.
 Hydrocarbons -- The exhaust emission of total hydrocarbons were 93% lower for biodiesel
than diesel.
 Nitrogen Oxide -- NOx emissions from biodiesel increases or decreases depending on the
engine family & testing procedures. NOx plays a major role in the localized formation of
367 M. Sadat Ali, P.S.V. Ramana Rao and C.V. Gopinath
smog and ozone. Lack of sulphur allows the use of NOx control technologies that cannot be
used with conventional diesel.
 Health Risk -- Biodiesel emission showed decreased level of PAH and nitrated PAH
compound which have been identified as potential cancer causing compound.
Environmental & Safety Information

Acute Orals : Biodiesel is nontoxic. By comparison,
Toxicity : table salt is 10 times more toxic.
Skin irritation : A 24 hr. human patch test indicated that undiluted biodesel produced
mild irritation less than produced by 4% soap solution.
Aquatic Toxicity : A 96 hr.lethal concentration for bluegill of biodiesel grade methyl ester
was greater than 1000mg/L which is generally deemed insignificant.
Biodegradability : Biodiesel degrade about 4 times faster than petroleum diesel. Within 28
days pure biodiesel degrades 85-88% in water.
Flash Point : Biodiesel flash point is over125ºC, well above petroleum based fuel’s
flash point of around 58ºC.Thus it is safer to store, handle & use than
conventional fuel.
Important advantages of using biodiesel:
1. The higher cetane number of biodiesel compared to petro-diesel indicates potential for
higher engine performance. Tests have shown that biodiesel has similar or better fuel
consumption, horsepower, and torque and haulage rates as conventional diesel
2. The superior lubricating properties of biodiesel increases functional engine efficiency
3. Their higher flash point makes them safer to store
4. The biodiesel molecules are simple hydrocarbon chains, containing no sulfur, or
aromatic substances associated with fossil fuels
5. They contain higher amount oxygen (up to 10%) that ensures more complete
combustion of hydrocarbons
6. Biodiesel almost completely eliminates lifecycle carbon dioxide emissions. When
compared to petro-diesel it reduces emission of particulate matter by 40%, unburned
hydrocarbons by 68%, carbon monoxide by 44%, sulphates by 100%, polycyclic
aromatic hydrocarbons (PAHs) by 80%, and the carcinogenic nitrated PAHs by 90% on
an average. The use of biodiesel complements the working of the catalysator and can
help a current EURO-1 motor attain the EURO-111 standards.
7. Plantations for Biodiesel provide the following benefits in addition:
8. Fixation of up to 10 t/ha/yr CO2 that could be internationally traded
9. Production of 1 t/ ha/yr of high protein seed cake (60% crude protein) that can be
potentially used as animal and fish feeds and organic matter that could be used as
organic fertilizer particularly in remote areas
10. Various other products from the plant (leaf, bark and seed extracts) have various other
industrial and pharmaceutical uses (Heller, 1996)
11. Localised production and availability of quality fuel
12. Restoration of degraded land over a period of time
13. Rural employment generation
14. Comparatively equitable benefit distribution
Major negative aspects attributed to use of biodiesel include:
1. High cost of production: will eventually solve itself when large-scale production and
use starts. Also, the price of petro-diesel does not take into account its actual cost; the
US General Accounting office estimated that the true cost for a barrel of crude oil to the
US citizen is more than USD 45 (current market price USD 21) when environmental
and military costs are included.
2. Modifications are required to the automobiles for use of biofuel: many automobile
brands are currently marketed ready for use of bio diesel.
3. High CFPP (cold filter plugging point) values and hence solidification and clogging of
the system at low temperatures: this problem occurs only in places where the
temperature goes down to around 0°C, even here the problem is currently solved by
adding additives.
A further method of extracting energy from biomass is the production of vegetable oils as fuel.
The process of oil extraction is carried out the same way as for extraction of edible oil from plants.
There are many crops grown in rural areas of the developing world which are suitable for oil
production - coconut, cotton seed, groundnut, palm, rapeseed, soy bean, and more.
There are two well-established technologies for oil extraction:
1. The simple screw press, which is a device for physically extracting the oil from the
plant - this technology is well suited to small-scale production of oil as fuel or as
foodstuff in rural areas. The press can be motorised or hand-operated.
2. Solvent extraction is a chemical process, which requires large, sophisticated equipment.
This method is more efficient - that is, it extracts a greater percentage of the oil from the
plant - but is less suited to rural applications.
The oil, as well as being used for lighting and heating, can be used as a fuel in internal
combustion engines. Sunflower oil, for example, has an energy content about 85% that of diesel fuel.
There appears to be little uptake of this fuel as a source of energy so far.
Performance of Diesel Engine

Engine performance is an indication of the degree of success with which it does its assigned job i.e.
conversion of chemical energy contained in the fuel into the useful into the useful into the useful
mechanical work.
The basic performance parameters are numerated and discussed below:
1. Brake Power(B.P.): The power developed by an engine at the output shaft is called the
brake power.
B.P.=((√3)VIcosφ)/ ηa watts, where
ηa =2πNT/60, 0.8(assume), is efficiency of alternator,
Vin volts from voltmeter, I in amperes from ammeter
2. Total fuel consumed: TFC= 10cc×0.8×3600/(t×1000) Kg/hr
Where ‘t’ is time to consume 10cc of fuel from burette
3. Spefic brake fuel consumed:
It is the mass of fuel consumed per kw developed per hour, and is a criterion of
economical power production.
i.e. s.b.f.c.=(TFC/BP) Kg/Kw/hr
4. Indicated power: The total power developed by combustion of fuel in the combustion
chamber is called indicated power.
I.P.= BP+FP
5. Brake thermal efficiency: It is the ratio of brake power done to energy supplied by the fuel.
i.e. ηBt= BP×3600×100/(TFC×C.V.)%
6. Mechanical effiency:
η m=BP/IP ×100 %
7. Indicated effiency:
ηi= IP×3600×100/(TFC×C.V.) %
8. Volumetric effiency: It is defined as the ratio of actual volume of the charge drawn in
during the suction stroke to the swept volume of the piston.
ηv=Va(Actual volume of air admitted) / Vs(swept volume)
Vs =πd²/4 ×L×N/60 m³/s
Heat Balance Sheet

The performance of an engine is generally given by heat balance sheet.
The quantity of fuel used in a given time and its calorific value, the amount, inlet and outlet
temperatures of cooling water and the weight of exhaust gases are recorded.After calculating IP and
BP, the heat in different items is found as follows:
1. Energy (HEAT) input in kg/hr
Input energy in the heat energy in the form of stored chemical energy of fuel.
The Heat input= C.V.*T.F.C. Kj/hr
2. Energy (heat) dissipation in 1hour
2.1) Output Energy (heat) brake power in kg/hr= BP*3600 kj/hr
2.2) Heat carried by exhaust gases in kj/hr =m*Cpg*(Tout-Tambient)gases
=heat taken out by water + heat carried by exhaust gases from calorimeter
=(mCp(Tout-Tin))cal.water +((mCp(Tout-Tamb))cal.gas
Cpg, is the specific heat of gases,
M in Kg =mass of air + mass of fuel consumed
To determine mass of air consumed:
Mass of air =volume× density
Volume of air sucked in by the engine V=Cd A(√2gha) where
A(area of the orifice)= π×d²/4,
Ha(head in column of air)=(h1-h2)×ρ water/ρ air
2.3) Heat carried by the jacket cooling water,
Qjw = mjw Cpw(Tout-Tin).
3. Uncounted Losses.
Unaccounted /miscellaneous losses=Heat input- Σheat dissipated KW/Kg-hr
Result and Analysis

For checking the power output of an engine for different oils used, initially it is required to check the
properties of the oils and to compare them with each other. Based on this objective we have tested the
properties of different oils used as bio diesel like Mohua, Castor, Karanj, Jatropa and Karanj. The
different properties that are experimentally found out are Density, Calorific value, Viscosity, Flash
point and Fire point. The comparison of the different oil densities is shown in Table.1 & Fig.1.
Table 1:
Name of the oil used for test Mohua oil Karanj oIL Castor oil Jatropa oil Diesel oil
Specific gravity of oil 0.83 0.9 0.79 0.92 0.84
Figure 1: Comparison of specific gravity of different oils
0.95
Specific gravity of oils

0.9
used for test

0.85
0.8
0.75
0.7
Mohua oil Karanj oIL Castor oil Jatropa oil Diesel oil
Name of the oils used for test
The specific gravity of Karanj oil (0.9) and Jatropa oil (0.92) was found to be more than
Mohua(0.83 and Castor at room temperature. The density of Mohua was found to be more than that of
Castor oil. The specific gravity of Castor and Mohua are found to be in close conventional diesel fuel.
From the comparison of different specific gravities it can be observed that the blending of these oils
with conventional diesel oil will be different for different oils. There can not be a particular proportion
which can be said the optimum blend to give optimum flow ability, maximum calorific value and
maximum power output.
The comparison of the calorific value of different oils is done & shown in the Table:2 &
Figure:2 below indicates that the castor oil has the highest calorific value as compared to the other oils.
The calorific value of castor oil is close to the diesel oil commonly used for automobiles. The calorific
value of Karanj oil was found to be more than Mohua oil. Similarly the other properties of the fuels or
oils are tabulated and can be compared.
Table 2:
Name of the oil used for test Mohua oil Karanj oil Castor oil Jatropa oil Diesel oil
Calorific value of oil in Kj /kg 37,328 38,683 38,800 37,328 42,714
Viscosity of Oil in mm2/sec 5.11 4.27 5.65 5.51 3.60
Flash point in ºC 181 163 172 110 60
Figure 2: Comparison of calorific value of different oils
44,000
Calorific value in kJ/kg
43,000
42,000
41,000
40,000
39,000
38,000
37,000
36,000
35,000
34,000
Mohua oil Karanj oil Castor oil Jatropa oil Diesel oil
Name of the oils used in test
Figure 3: Comaprison of Viscosity of different oils
Viscosity in mm2/s
5
4
3
2
1
0
Mohua oil Karanj oIL Castor oil Jatropa oil Diesel oil
Name of the oils used for Testing
For checking the performance of engine the brake power of the engine is calculated for
different oil composition and is tabulated. The brake power developed for different load conditions is
shown in Fig:4. The Fig:5 shows the comparison of the maximum brake power developed for different
oils under full load condition. It indicates that the power developed under full load condition with
castor oil as fuel is far more greater than the oils like Mohua and Karanj. The generation of higher
brake power can be attributed to higher calorific value of the Castor oil. The brake power developed by
Mohua and Karanj is almost equal to each other. Although the brake power developed in Karanj oil
always remains a little bit higher than Mohua.
Figure 4: Load vs Brake Power
8
Brake power of diesel
7
Brake power in kW
6
Brake power of
5
Castor
4
3 Brake power of
2 Karanj
1 Brake power of
0 Mohua
1 2 3 4 5 Brake power of
Jatropa
Load in kW
Figure 5: Break Power Developed in Full Load Condition
8 7.4kW
Brake Power in kW
6
4.5kW 4.27kW 4.11kW
4kW
4
0
Diesel Castor Karanj Mohua Jatropa
Different type of Oils Used
The variation of the fuel consumed with the increasing load is shown in Fig: 6. as under. It can
be seen that with the increase in load the fuel consumed per hour also increases. It can be also observed
that the variation is not too high for karanj oil. But for high viscous oils the variation is appreciable.
Figure 6: Load in kW vs Total fuel consumed in kg/hr
0.57
Total fuel consumed in kg/hr
0.56
0.55
0.54
0.53
0.52
0.51
0.5
0.49
0.48
1 2 3 4 5 6
Load in kW
The variation of fuel consumed for different type of fuel used is shown in the graph drawn in
figure- 7. It can be observed that the fuel consumption for castor and Mohua is not high but for karanj
it is appreciably large. The high consumption can be attributed to high flowability of oil under suction
pressure of diesel engine. Due to high viscosity of other oils flowability is not high which in turn
effects the fuel consumption.
Figure 7:
Total fuel consumed

4
in Kg/ hr of karanj oil
Total fuel consumed
3
in kg/hr of castrol oil
Total fuel consumed
2
in Mohua oil
0
0 0.5 1 1.5 2 2.5
It can be also shown that with the increase in load brake thermal efficiency increases and
reaches to maximum value then it starts decreasing. This can be graphically shown as shown in figure-
8.
Figure 8: Variation of Break Thermal Efficiency With Load
0.8
Brake Thermal Efficiency in %
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.5 1 1.5 2
Load in kW
The variation of mechanical efficiency with load also shows the same trend as the brake
thermal efficiency. The mechanical efficiency increases with increase in load and it reaches a
maximum value, then it decreases. So from this observation it can be concluded that for getting
maximum mechanical efficiency and brake thermal efficiency we have to run the diesel engine at the
optimum load condition.
Figure 7: Load v/s Mechanical Efficiency
60
Mechanical efficiency in %
50
40
30
20
10
0
0 0.5 1 1.5 2
Load in kW
A steady drop in exhaust gas temperature was observed for both the rated and peak torque
condition for all blends, indicating a shorter ignition delay thus shifting the peak pressure point toward
top dead center. The lower exhaust gas temperature was a result of increased thermal efficiency.
Conclusions
From the above result tables and the graphs we can conclude the following points:
1) From the properties of fuel analysis it is found that the flash point and fire point of the
commonly used biofuels are almost in the same range as it is found in diesel.
2) The viscosity of Mohua oil was found to be more than karanj and castor oil.
3) The viscosity of diesel was found to be less than Karanj and Castor oil.
4) Similar variations were observed in density of fuels also.
5) The calorific value of castor oil was found to be more than diesel oil as well as karanj and
mohua oil.
6) The calorific value of Karanj and Mohua oil was found to be less than diesel engine.
7) The power developed by different oils were compared. It was found that power developed
by Castor oil is more than all other oils.
8) The fuel consumed increases with the increase in load. But the increase was more in case of
castor oil where as it follows a linear trend for all other oils.
9) The increase in brake power was also linear up to a limit for increasing load.
After attaining a maximum value the brake power decreases.
10) The brake thermal efficiency also shows a similar trend for increasing load conditions. It
increases with increase in load upto a certain limit but after attaining the maximum value it
decreases.
11) The mechanical efficiency of the engine also shows a similar trend for increasing load
conditions.
12) The brake thermal efficiency is found to be more for Mohua and Karanj oil than
conventional diesel oil, where as for Castor oil was found to be less than all oils.
13) The mechanical efficiency of Castor oil was found be more than all other oils.
The mechanical efficiency of karanj and mohua oil was found to be less than conventional
diesel oil.
14) All the properties of the fuels and the performance of engine varies linearly from one trend
to other as the blend changes proportionately.
15) The optimum blend for getting maximum power output and best overall efficiency is yet to
be experimentally decided for future use of biodiesel fuels in the conventional diesel
engines.
References
[1] Feldman, M. E. and Peterson, C. L. (1992). Fuel injector timing and pressure optimization on a
DI diesel engine for operation on biodiesel. Liquid Fuels From Renewable Resources-
Proceedings of an Alternative Energy Conference. Nashville, TN.
[2] Marshall, W. F. (1993). Effects of methyl esters of tallow and grease on exhaust emissions and
performance of a Cummins L10 engine. Itt Research Institute, National Institute for Petroleum
and Energy Research, Bartlesville, OK. (Report prepared for Fats and Proteins Research
Foundation, Inc, Ft. Myers, Beach, FL)
[3] Mittelbach, M. P., Tritthart, P, and Junek, H. (1985). Diesel fuel derived from vegetable oils, II:
emission tests using rape oil methyl ester. Energy Agriculture. 4: 208-215.
[4] Mittelbach, M. P., Tritthart, P. (1988). Diesel fuel derived from vegetable oils, III. Emissions
tests using methyl esters of use frying oil. Journal of the American Oil Chemists' Society. 65(7):
1185-1187.
[5] Niehaus, R. A., Goering, C. E. Savage, L. D., and Sorenson, S. C. (1985). Cracked soybean oil
as a fuel for a diesel engine. ASAE Paper No. 85-1560. ASAE, St. Joseph, MI.
[6] Reece, D. L. and Peterson, C. L. (1993). A report on the Idaho on-road vehicle test with RME
and neat rapeseed oil as an alternative to diesel fuel. ASE Paper No. 93-5018. ASAE, St.
Joseph, MI.
[7] Scholl, K. w. and Sorenson, S. C. (1993). Combustion of a soybean oil methyl ester in a direct
injection diesel engine. SAE paper NO. 930934. SAE, Warrendale, PA.
[8] Schumacher, L. G, Hires, W. G, and Borgelt, S. C. (1992). Fueling a diesel engine with methyl-
ester soybean oil. Liquid Fuels From Renewable Resources- Proceedings of an Alternative
Energy Conference. Nashville, TN.
[9] Schumacher, L. G., Borgelt, S. C., Hires, W. G. and Humphrey, J. K. (1993). Biodiesel on the
road- A report from Missouri. ASAE paper No. 93-5017. ASAE, St. Joseph, MI.
[10] Schumacher, L. G., Borgelt, S. C., and Hires, W. G. (1993). Soydiesel/Biodiesel Blend
Research. ASAE paper No. 93-6523. ASAE, St. Joseph, MI.
[11] Srinivasa, R. P. and Gopalakrishnan, K. V. (1991). Vegetable oils and their methylesters as
fuels for diesel engines. Indian Journal of Technology. 29: 292-297.
[12] Ziejewski, M., Kaufman, K. R., Schwab, A. W, and Pryde, E. H. (1984). Diesel engine
evaluation of an nonionic sunflower oil-aqueous ethanol microemulsion. Journal of the
American Oil Chemists Society. 61 (10): 1620-1626.

Efficient Method for Assigning Students to Proper Groups

Uploaded by

Copyright:

Available Formats

You might also like

Efficient Method for Assigning Students to Proper Groups

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Efficient Method for Assigning Students to Proper Groups

Uploaded by

Copyright:

Available Formats

European Journal of Scientific Research

ISSN: 1450-216X Volume 21, No 2 August, 2008

Editorial Advisory Board e

Parag Garhyan, Auburn University Teresa Smith, University of South Carolina

Morteza Shahbazi, Edinburgh University Ranjit Biswas, Philadelphia University

Raj Rajagopalan, National University of Chiaku Chukwuogor-Ndu, Eastern Connecticut

Sang-Eon Park, Inha University John Mylonakis, Hellenic Open University

Ajay K. Ray, National University of Singapore Dimitrios Mavridis, Technological Educational

Jatin Pancholi, Middlesex University Leo V. Ryan, DePaul University

Further information is available at: http://www.eurojournals.com/ejsr.htm

Détermination de L’âge et Modélisation de la Croissance D’un Poisson: L’anchois

Efficient Method for Assigning Students to Proper Groups 249-358

Evolution de la Microflore et Caractéristiques Physico-Chimiques D’un Aliment

Study of Performance of Anthropometric Indices in Screening Children with

On-Line Power Quality Monitoring System Using CompactRIO Device: A Review

Earnings, Market Value, and Option-Based Equity Valuation 296-304

Regime Alimentaire du Poisson-Chat Schilbe Mandibularis (Günther, 1867) Dans Deux

Weights Adjustment Neural Networks 314-318

Modified Atmosphere Packaging of Strawberries (Fragaria X Ananassa Duch.)

A Comparative Study of Engine Performance by Using Different Bio-Fuels 365-375

Détermination de L’âge et Modélisation de la Croissance D’un

Figure 1: Zone d’étude: golfe de Skikda (G S), Côtes Est algériennes.

Tableau 1: Mesures effectuées sur les poissons échantillonnés

Log z = log z(xi+h) – log z(xi),

2) Croissance linéaire absolue

4) Croissance pondérale absolue

nous avons donc

Méthode de Battacharya (figure 3)

Les résultats de l’application de la méthode de Battacharya sont consignés dans le tableau 2

Tableau 2: ésultats de l’application de la méthode de Battacharya

2) Croissance linéaire absolue

Femelles → W  3,288 10

4) Croissance pondérale absolue

Figure 4: Variations mensuelles de k

Figure 5: Variations de k en fonction du sexe et des classes de tailles

Tableau 5: Paramètres du modèle de Von Bertalanffy obtenus dans certaines régions.

L∞ (cm) K t0 Pays Localité Auteur

Tableau 6: Paramètres de la relation taille-poids obtenus dans certaines régions de la méditerranée.

a b Sexe n Pays Auteur

L’équation de la croissance pondérale absolue obtenue, montre que l’évolution du poids en

Efficient Fault Tolerant Mapping of Large Three-Dimensional

In this paper we present an efficient 3D large volume simulation decomposition and

Keywords: Parallel Simulation; Computational Modeling; Plasma Physics; Hamiltonian

2. Task assignment, mapping, and scheduling

3. Base-b reflected gray codes

4. The 3D PIC Model

4.2. Partial Charge Scatter Operation

4.3. Gather Operation

4.3.2. Electric Field Calculation

4.4. Particle Advancement Operation

5. Cray T3D Interconnection Network

5.1. Dimension order routing

5.2. 3-D Torus Partitioning

6. Algorithm mapping onto ring topology

6. 1. Algorithm mapping onto fault free ring topology

Table 1: Mapping 4-Cube onto 3-D Torus of Size (2,4,2).

i Vg G(2)(4) T(t3 t2, t1) Vt

001 101 Wraparound

6. 2. Algorithm mapping onto faulty node(s)