The effect of constant dietary energy supply during late gestation and early lactation on performances of prolific D’man ewes
The effect of constant dietary energy supply during late gestation and early lactation on performances of prolific D'man ewes
Mokhtar MAHOUACHI 1
Mourad REKIK 1
Narjess LASSOUED 0
Naziha ATTI 0
0 Laboratoire de Productions Animales et Fourragères, Institut National de la Recherche Agronomique de Tunisie , rue Hédi Karray, 2049 Ariana , Tunisia
1 Département de Production Animale, École Supérieure d'Agriculture du Kef , 7100 Le Kef , Tunisia
- A total of 42 prolific D'man ewes were used to study the effect of the dietary energy supply on their performances during the last 45 days of pregnancy and the first 60 days of the suckling phase. The ewes were fed 1 kg of hay and supplemented with either 200 (treatment L), 570 (treatment M) or 840 g (treatment H) of a barley and soyabean-based concentrate. The proportion of soyabean was adjusted so that the crude protein (CP) content of the three diets was similar (120 g CP·kg-1). The energy contents of the diets were 1.2, 1.8 and 2.2 of the daily net energy requirements for maintenance in respectively the L, M and H treatments. Daily hay intake averaged 829 ± 18 g dry matter (DM) per ewe and was not affected by the level of the concentrate distributed. The body condition score of the ewes in all treatment groups was low at the beginning of the experiment (1.9 ± 0.3) and continued to decline throughout the period of treatment application. Nevertheless, the rate of decline was higher for the L and M ewes (P < 0.05). The live weight of the ewes after lambing was higher (P < 0.05) for ewes in the H treatment group (43.2 ± 6.9 kg) in comparison to the other ones (37.7 ± 7.8 and 40.0 ± 7.3 kg for the L and M groups respectively). Daily milk production over the first 40 days of lactation (959 ± 399 g·day-1) was not affected by the energy level of the diet. Similarly, fat (8.9 ± 2.4%) and protein (4.5 ± 0.9%) contents of the milk were not affected by the diet energy content. The weights of the lambs at birth, at 10, 30 and 50 days of age were not influenced by the diet energy content. In contrast, average daily growth (ADG) between 10 and 30 (ADG10-30) and also between 30 and 50 days of age (ADG30-50) were significantly affected by energy allowances in the diet (95 ± 23, 133 ± 25, and 152 ± 27 g·day-1 for ADG10-30 and 76 ± 22, 135 ± 24 and 136 ± 24 g·day-1 for ADG30-50 in L, M and H groups respectively). It is concluded that the energy requirements of prolific D'man ewes during late gestation and early lactation are likely to be higher than the levels attempted in this study.
ont été utilisées pour étudier les effets du niveau énergétique de la ration sur leurs performances de
production pendant 45 j en fin de gestation et 60 j après la mise-bas. Les brebis ont reçu 1 kg de foin
complémenté avec 240 g (lot bas : B), 570 g (lot moyen : M) ou 840 g (lot haut : H) d’un concentré
à base d’orge et de tourteau de soja. La proportion de ce dernier a été ajustée pour obtenir 3 rations
iso-azotées (120 g MAT·kg–1). Les apports énergétiques sont estimés à 1,2, 1,8 et 2,2 fois le besoin
d’entretien pour les lots L, M et H respectivement. L’ingestion du foin (829 ± 18 g MS·j–1) n’est
pas affectée par la quantité de concentré distribuée. La note d’état corporel des brebis est très faible
au début de l’essai (1,9 ± 0,3) et continue à baisser tout au long de l’expérience. Cependant, la baisse
est plus prononcée (P < 0,05) chez les brebis qui reçoivent les niveaux L et M. Le poids vif des brebis
après la mise-bas est aussi affectée (P < 0,05) par le niveau énergétique de la ration (37,7 ± 7,8, 40,0 ±
7,3 et 43,2 ± 6,9 kg pour les lots L, M et H respectivement). La production laitière (959 ± 399 g·j–1),
les teneurs en matière grasse (8,9 ± 2,4 %) et protéique (4,5 ± 0,9 %) ne sont pas influencées par le
niveau énergétique de la ration. Les poids à la naissance, à 10, 30 et 50 j sont similaires pour les trois
traitements. En revanche, les gains moyens quotidiens entre 10 et 30 j (GMQ10–30) et entre 30 et
50 j (GMQ30–50) sont significativement (P < 0,05) améliorés par l’accroissement du niveau
énergétique de la ration (95 ± 23, 133 ± 25 et 152 ± 27 g·j–1 pour GMQ10–30 et 76 ± 22, 135 ± 24
et 136 ± 24 g·j–1 pour GMQ30–50 respectivement dans les lots L, M et H). Il est conclu que les besoins
énergétiques des brebis prolifiques D’man sont vraisemblablement plus élevés que les niveaux testés
dans cet essai.
énergie / brebis / prolifique / fin de gestation / allaitement
In several countries, prolific breeds of
sheep are commonly used to improve flock
1, 21, 27
]. In Tunisia, the
Moroccan D’man breed is selected because
of its adaptation to local climatic conditions.
Following its introduction into Tunisia in
1994, the breed population has increased
from 100 to approximately 6 000 breeding
ewes in 2001 according to the latest
published statistics [
There is a wealth of information related to
the breed’s reproductive performances [
] as well as the genetic parameters of
its main production traits [
contrast, its nutritional requirements were not
so thoroughly investigated especially
during critical physiological stages such as late
gestation and early lactation. In the southern
areas of the Mediterranean sea, sheep are
often undernourished during these two
physiological stages as a result of quantitatively
and qualitatively limited feeding resources
 in their traditional management.
Undernutrition during late gestation is known to
diminish foetal growth, weight at birth, dam
milk production and post-natal growth [
and could delay restoration of post-partum
reproductive activity [
]. In addition, the
occurrence of metabolic disorders, such as
pregnancy toxaemia, linked to excessive
mobilisation of the body fat reserves, becomes
more frequent. The likelihood of these
disorders seems to be more important in the
case of prolific breeds [
Tunisian breeders do not generally change
feeding regimens of sheep between
pregnancy and early lactation. Their diets are
based on hay and concentrate in these two
Feeding strategies of prolific sheep
during late gestation and early lactation are to
be different from other non-prolific breeds
and must be in accordance with the climatic
and economic conditions of each area [
The same author reported that during late
gestation and early lactation, prolific ewes
like the Finnsheep and the Outaouais could
be fed diets based on stored forages
supplemented with grains at levels varying between
18 and 30%. However, Guerouali [
pointed out that diets based on bulky
feeding resources are unsuitable for pregnant
D’man ewes if they are not supplemented
with a high proportion of concentrates in
their diets. In all these reported data, no
direct measurements of the effect of nutrient
allowances (energy, nitrogen, minerals…)
during late pregnancy or early lactation on
the performances of the ewes were reported.
Hence, this study was undertaken to study
the effect of maintaining constant different
energy levels, while maintaining the same
nitrogen allowances during late gestation and
the early lactation period on the
performances of prolific D’man ewes.
2. MATERIALS AND METHODS
2.1. Animals and general management
A total of 42 adult pregnant D’man ewes
were used in this study. The animals weighed
45 ± 6.8 kg and were allocated to three
treatment groups balanced for age (2.8 ±
0.9 years), body condition score (1.9 ± 0.5
out of a scale ranging from 0 to 5) and litter
size (1.95 on average) determined by
ultrasonography at approximately 3 months of
pregnancy. The ewes were then
individually housed and fresh water was available
at all times. All the ewes included in the
experiment are regularly run in an
accelerated lambing rhythm (3 lambings in two
years). For the particular reproductive period
that corresponded to the present experiment,
mating took place in July and the lambings
were spread between the 7th and the 30th of
December (23 days). After lambing, all lambs
suckled their mothers and no milk was drawn
out. The ewes were vaccinated against
enterotoxaemia and were drenched for internal
worms and parasites.
2.2. Nutritional treatments
On a daily basis, the ewes were
individually fed 1 kg of hay (avena sativa) and
200 g (treatment L), 570 g (treatment M) or
840 g (treatment H) of concentrate. The
3 concentrates included barley, soyabean
meal, a supplement of vitamins and
minerals whose composition is shown in Table I.
The quantities of concentrate were
calculated for the energy allowances in the diets
to achieve 1.2, 1.8 and 2.2 of the daily net
energy maintenance requirements estimated
at 0.033 UFL·kg live weight (LW)–0.75 [
] in respectively treatments L, M and H.
The three regimens were kept constant from
late pregnancy (–6 weeks) until 8 weeks of
lactation. The net energy value of the feeds
was determined according to the French
system based on equations established by
Vermorel  and in vitro organic matter
]. The chemical
compositions of the feedstuffs used in this
experiment are reported in Table II. The diets were
fed twice daily at 08.30 and 16.30. The
feeding regimes were distributed for a period of
2 weeks of adaptation prior to the
experimental period that lasted 105 days and
corresponded to the last 45 ± 13 days of
pregnancy and the first 60 days of lactation.
2.3.1. Food intake, body score and ewe live weight
Food intake was determined daily by the
weighing of the distributed and the refused
feeds. Assessment of the ewes body
condition was carried out by attributing a score
(scale from 0 to 5) according to the method
of Russel et al. [
]. In total, 6
measurements were taken from 4 weeks before until
8 weeks after lambing. The ewes were also
weighed 24 hours after lambing and
thereafter at weekly intervals during a month.
2.3.2. Milk production
Daily milk production was determined
using oxytocin (3 IU) as in the method
described by Ricordeau et al. [
]. At each
occasion of milk production recording (hand
milking), samples of 20 mL of milk were
also taken and immediately sent to the
laboratory for analysis of fat and protein
contents using an integrated milk testing machine
(Combifoss 5300, Foss Electric, Denmark).
In total, 4 measurements of milk production
and composition were carried out at 10 day
intervals during the first 40 days of
2.3.3. Growth of lambs
Newborn lambs were weighed at birth
and then at intervals of 21 days. Weights at
10, 30 and 50 days as well as average daily
growths (ADG) between 10 and 30 (ADG10–
30) and between 30 and 50 (ADG 30–50)
days of age were thereafter calculated.
2.3.4. Chemical analysis
The dry matter (DM), minerals, organic
matter (OM) and crude fibre (CB) contents
of feedstuffs were determined by reference
to the official methods [
]. The protein
content was also determined using the Kjeldahl
method (Protein content = 6.25 × Nitrogen
2.4. Statistical analysis
The effect of the level of energy in the
diet on all measured parameters was
determined by an analysis of variance using the
GLM procedure of SAS [
]. Means were
then compared using the Newman and
Keuls test [
]. Data on fertility and litter
size were analysed using a χ2 test.
3.1. Feed intake
On average, the ewes consumed 829 g
DM·day–1 which is equivalent to 50.8 g
DM·kg LW–0.75 of hay with little
differences between late pregnancy (810 ± 18 g
DM·day–1) and early lactation (847 ± 17 g
DM·day–1). This level of intake was not
significantly affected by the level of
concentrate in the diet (Tab. III). In contrast, the
intake of hay tended to decrease during the
last week of pregnancy (–50 g DM·day–1 for
H group), then levelling up immediately
after parturition and remaining stable after
the 4th week of lactation. Moreover, no
significant differences in hay ingestion were
detected between the ewes bearing
multiples or singles during this period.
Whatever the level of concentrate given,
the ewes totally consumed their
concentrate. As a result, total intake was different
Means on the same line with different superscripts are different at P < 0.05; * Protein content of feedstuffs
from Nefzaoui and Chermiti [
(P < 0.05) between the treatment groups
(1013 ± 18 (L) vs. 1328 ± 20 (M) and 1572 ±
17 g DM·day–1 (H)) which is equivalent to
62.5, 81.6 and 96.3 g DM·LW–0.75
respectively for L, M and H groups. There were
no signs of acidosis in the case of the ewes
subjected to the H treatment.
3.2. Body condition
At the start of the experiment, the body
condition score was low 1.9 ± 0.6 (L), 1.8 ±
0.5 (M) and 1.9 ± 0.4 (H). The body
condition of all ewes further decreased until week
4 post-partum (Fig. 1). Nevertheless, the
intensity of body condition losses is more
important after lambing. During the two
weeks preceding and the three weeks
consecutive to lambing, the ewes in the H
treatment achieved a higher mean body
condition score of 1.6 ± 0.5 when compared to L
(1.2 ± 0.4) and M (1.2 ± 0.3) ewes.
3.3. Live weight of ewes
Ewe live weight at lambing was
significantly (P < 0.05) different between the
regimens (Fig. 2). The figures reached 37.7 ±
7.8 (L), 40.0 ± 7.3 (M) and 43.2 ± 6.9 kg
(H). Thereafter, live weight continued to
decrease and then remained constant
starting at around week 3 after lambing.
Following a month of lactation, the ewes lost an
average 2.6 ± 2.3, 2.7 ± 1.4 and 4.3 ± 2.6 kg
of live weight in respectively the L, M and
H treatments (P = 0.27). Across treatments,
the loss of live weight was however,
variable ranging from 0.5 to over 9 kg during the
first month. Live weight at the end of the
experiment also showed a large variability
ranging from less than 26 kg in the L
treatment to over 50 kg in the H treatment.
3.4. Distribution of born and suckled litters
All ewes lambed except one in the H treat
ment group, which died during pregnancy
as a result of Coenurosis. Prolificacy (i.e.
Number of lambs born per ewe lambing)
and lamb mortality rate were higher in the
M treatment group in comparison to the
other treatment groups (Tab. IV). As a result,
the average number of suckled lambs per
ewe was similar for ewes in all treatment
groups with, however, important differences
in the distribution of the litter sizes. The
number of ewes rearing more than two lambs
was higher in the H (n = 3) treatment group
when compared to the L (n = 1) and M (n =
3.5. Quantitative and qualitative milk production
Overall, milk production averaged 959 ±
399 g·day–1 (Tab. V) with a large
coefficient of variation particularly at the first
control (49%). As a result of this large
variation, milk production was not significantly
affected by the energy level of the diet. It
tended, however, to be higher in the case
of the most energetic diet (1098 ± 400 vs.
870 ± 369 (M) and 913 ± 400 (L) g·day–1).
This difference between energy levels is
more marked during early lactation (1050
(L) vs. 1385 (M) and 1330 (H) g·day–1) than
at the end of the measurement period (675
level of energy in the diet had a significant
effect (P < 0.05) on milk yields in the first
control only for the ewes rearing multiples.
Milk production was the highest during
the first week of lactation and then decreased
over the next 40 days. Milk of the D’man
breed had a higher fat and protein content
(Tab. V) during the first week of lactation
(11% and 5.4% respectively) when
compared to the last week of lactation (8.2% and
4.6% respectively). The fat and protein
content of milk was not affected by the diet
energy level during late pregnancy and early
lactation. Average fat content was 8.7 ± 2.8,
9.1 ± 2.0 and 9.1 ± 2.3% and average protein
content was 4.6 ± 1.0, 4.6 ± 0.9 and 4.4 ±
0.9% in the L, M and H treatments
respectively (Tab. V).
3.6. Growth of lambs
Live weights at birth, 10, 30 and 50 days
as well as ADG10–30 and ADG30–50 of
the lambs are shown in Table V. None of the
live weight parameters was significantly
affected by the diet energy level. There was
nevertheless a tendency (P = 0.08) for
weight at 50 days to be higher in the H
treatment group (Tab. V). At 50 days of age,
average live weight difference between the
Means on the same line with different superscripts are different at P < 0.05.
lambs in the L and H treatment groups
reached 2.8 kg. In contrast, ADG10–30 and
ADG30–50 were affected (P < 0.05) by the
energy level of the diet. ADG10–30 and
ADG30–50 were particularly low in the
case of the L treatment group and averaged
95 ± 23 and 76 ± 22 g·day–1 respectively.
Differences up to +38 and +59 g·day–1 were
recorded for respectively ADG10–30 and
ADG30–50 when the energy level of the
diet increased from L to M. When the litter
weights at different ages were considered,
it appeared that litter weights at birth and
10 days after were more influenced by the
corresponding litter sizes than by the
dietary energy supply (Tab. VI). At 30 and
50 days of age, the corresponding litter
sizes still exerted a significant effect on
litter weight. Nevertheless, the dietary energy
supply also became important and
significantly affected litter growth (Tab. VI).
Litter weights were higher (P < 0.05) in ewes
receiving the high energy level in
comparison to litters in both the low and moderate
Using a stepwise regression procedure of
], we demonstrated that for ewes in
all treatments groups, milk production was
correlated (r = 0.46; n = 40; P < 0.01) to the
average ADG10–30 of the litter. There was
a low relationship between both fat (P =
0.36) and protein (P = 0.21) contents and the
ADG10–30 of the litter.
Despite its poor nutritional quality, the
daily allowance of 1 kg·day–1 of hay was
totally consumed by D’man ewes during
late gestation and early lactation. In contrast
to what is commonly reported on ruminants
receiving good quality forages ad libitum
], hay intake was never depressed by the
addition of high levels of concentrate
(840 g·day–1 i.e. 47% total DM intake).
This may be due to the fact that maximum
voluntary intake of this hay was not reached,
since it was not given ad libitum. Hence we
cannot estimate the potential intake of this
hay. Inversely, due to the low crude protein
content of the hay, it is even possible that
the low level of concentrate increased the
digestibility thus preventing any
substitution of hay to concentrate even at the highest
]. This proves that with poor
quality roughage high concentrate
supplementation is at no risk to depress its intake.
Hence, such a diet may be well adapted to
prolific ewes in late gestation and early
The D’man ewes used in this study are
managed in an accelerated rhythm of three
lambings in two years, thus explaining the
particularly low average body condition
score at the start of the experiment. This did
not prevent the ewes, especially those in the
L and M treatment groups to continue the
depletion of their body reserves during late
gestation (a mean reduction of 0.5 points of
body score that is 26% of initial BCS). The
determination of blood metabolite
concentrations (glucose, free fatty acids, β-OH)
used to measure the intensity of body reserve
depletion, would have confirmed our
observation of the ability of D’man ewes to loose
According to Chilliard , Geenty and
] and Khaldi [
], the energy
resulting from mobilisation of body reserves
is used to compensate for the increased
demand for foetal development. For
nonprolific breeds facing an energy deficit, the
ability to deplete their body reserves is well
documented in dry [
] and lactating ewes [
]. Our results suggest that similarly to
these breeds, the prolific D’man ewe, even
with low body condition, could undergo
underfeeding during late gestation with no
apparent signs of sanitary problems (toxemia,
abortions, drastic reduction of intake, …).
Increasing the energy level of the diet
from 1.2 (L) to 1.8 (M) of the energy
requirements for maintenance was not accompanied
by an improvement of the ewes’ body
condition during late gestation. This result may
be accounted for by differences in
prolificacy between the two treatment groups being
200% and 164% respectively for ewes in the
M and L treatment groups. However, when
prolificacy was similar (case of the L (164%)
and H (177%) treatments), increases in the
dietary energy supply were associated to
improvements in the body condition score
during late pregnancy (i.e. + 0.32 points;
P = 0.046). Most probably, the
supplementary energy allowances (+ 0.58 UFL·day–1)
prevented body reserves from being depleted.
Whatever is the level of energy intake
during late gestation, the average birth live
weight of the lambs was similar around 2.5–
2.7 kg (Tab. V). These figures were similar
to those reported by Boujenane [
D’man lambs or by other authors [
for other prolific breeds under diverse
feeding conditions. In this experiment, the lack
of differences between treatments for this
parameter could be attributed to the ability
of D’man ewes to mobilise their body
reserves thus insuring a normal foetal growth.
Following lambing, the body score con
tinued to decline though more intensely in
the case of ewes in the H treatment group.
Ewes receiving the H energy level had more
body reserves at parturition than the ewes
in the other treatment groups (Fig. 1). The
decline in body score was paralleled by a
decrease in live weight indicating that for
the three treatment groups, the dietary energy
supply was lower than the whole
requirements during early lactation.
Average milk production for D’man ewes
as recorded in this study (959 g·day–1), was
similar to other reported figures for the
same breed [
]. There was a tendency for
the energy level of the diet to favour milk
production as stated by Bocquier and Caja
] who indicated that increasing levels of
feeding are generally associated with higher
milk production and inversely. In our study,
however, the effect of feeding level on milk
production was partially confounded by the
number of suckled lambs [
] as suggested
by the analysis of milk production data
according to litter size (analysis not shown
here). Failure to detect a clear effect of the
level of energy on milk production could
also be the result of the quality of proteins
of the different diets. The diet in treatment
L had more soyabean meal, which has a
higher content in essential amino acids [
than the other feeds (i.e. barley and hay).
Furthermore, the difference of AA supply
from feeds may have been used for
neoglucogenesis in L and M treatments.
There is no effect of the level of energy
on milk composition. This is probably the
result of the natural variation of this
parameter during the course of lactation but
cannot be the result of an indirect effect of the
quantity of milk produced [
] since milk
yields were very similar. In comparison to
other reported values for the same breed
], our values of the milk fat content were
higher. This may be accounted for by the
intense depletion of body reserves leading
to increased concentrations of free fatty
]. In contrast, levels of protein
content were within the range of reported
values for D’man ewes [
]. It must be kept in
mind, however, that the absolute values of
milk composition should be taken
cautiously since the oxytocin method may lead
to the overestimation of milk yield as
suggested by Doney et al. [
Average daily growth rates and litter
weights at 30 and 50 days of age were
affected by the diets energy level of their
dams during late gestation and early
lactation. Nevertheless, these growth
performances remained low for the three levels of
energy used, and were consistent with
reported values in the literature for the same
] and, in all case, are much lower
than growth performances of suckling local
autochthonous Tunisian breeds [
Considering the lamb growth
performances, there was practically no linear response
from low to high energy supply. Thus, it can
be stated that lean D’man ewes can support
the lowest level of energy supply in late
gestation and early lactation. However,
considering the body condition score evolution of
ewes, it is clear that D’man ewes need, at
least, to be fed at the highest level of energy
supply in order to limit the use of body
reserves during lactation, especially for ewes
subjected to accelerated lambing rates.
 Aboul-Naga A.M., The use of prolific sheep into various countries: Middle East and North Africa , in: Fahmy M.H. (Ed.), Prolific sheep , CAB International, Wallingford, UK, 1996 , pp. 350 - 359 .
 Aitken I.D. , Diseases associated with prolificacy , in: Fahmy M.H. (Ed.), Prolific sheep , CAB International, Wallingford, UK, 1996 , pp. 485 - 502 .
 AOAC , Official methods of analysis , Association of Official Analytical Chemists , Washington, DC, 1984 .
 Atti N. , Capacité d'adaptation de la brebis Barbarine aux conditions alimentaires difficiles : importance des réserves corporelles et des adaptations digestives , Thèse de Doctorat d'Etat en Sciences Agronomiques , Spécialité : Nutrition Animale, Institut National Agronomique de Tunisie, 2000 , 200 p.
 Atti N. , Kayouli C. , Mahouachi M. , Guesmi A. , Doreau M. , Effect of a drastic and extended underfeeding on digestion in Barbary ewe , Anim. Feed Sci. Technol . 100 ( 2002 ) 1 - 14 .
 Bedhiaf S. , Rekik M. , Aloulou R. , Ben Hamouda M. , Ben Sassi M. , Arous M. , Genetic and phenotypic evaluation of litter size and lambing traits of D'man ewes , Anim. Res . (accepted for publication) .
 Bocquier F. , Caja G. , Production et composition du lait de brebis : effets de l'alimentation, INRA Prod . Anim. 14 ( 2000 ) 129 - 140 .
 Bocquier F. , Chilliard Y. , Effect of severe under nutrition on body weight and fat tissue changes in dry Lacaune ewes , Ann. Zootech. 43 ( 1994 ) 300 .
 Bocquier F. , Thériez M. , Prache S. , Brelurut A. , Alimentation des ovins , in: INRA (Ed.), Alimentation des bovins, ovins et caprins , INRA, Paris, France, 1988 , pp. 249 - 281 .
 Bouix J. , Kadiri M. , Chari A. , Performances enregistrées dans les troupeaux pépinières de la race ovine D'man , Al Awamia 52 ( 1974 ) 67 - 97 .
 Boujenane I. , Prolific sheep breeds: The D'man , in: Fahmy M.H. (Ed.), Prolific sheep , CAB International, Wallingford, UK, 1996 , pp. 109 - 120 .
 Boujenane I. , Kerfal M. , Estimates of genetic and phenotypic parameters for growth traits of D'man lambs , Anim. Prod . 51 ( 1990 ) 173 - 178 .
 Boujenane I. , Kerfal M. , Estimation de la production laitière des brebis D'man , Al Awamia 78 ( 1992 ) 145 - 155 .
 Boujenane I. , Boudiab A. , El Aich A. , Performances de production des races ovines locales marocaines , Actes Instit. Agron. Vét . 2 ( 1982 ) 24 - 48 .
 Caldeira R.M. , Vas Portugal A. , Interrelationship between body condition and metabolic status in ewes , Small Ruminant Res . 6 ( 1991 ) 15 - 24 .
 Chilliard Y. , Doreau M. , Bocquier F. , Lobley G.E. , Digestive and metabolic adaptations of ruminants to variations in food supply , in: Journet M., Grenet E. , Farce M.H. , Thériez M. , Demarquilly C . (Eds.), Recent developments in the nutrition of herbivores , INRA, Paris, 1995 , pp. 329 - 360 .
 Dagnelie P. , Théories et méthodes statistiques , Vol. 2 , Presses Agronomiques , Gembloux, Belgique, 1980 , 463 p.
 Doney J.M. , Peart J.N. , Smith W.F. , A consideration of the techniques for estimation of milk by suckled sheep and a comparison of estimates obtained by two methods in relation to the effect of breed, level of production and stage of lactation , J. Agric. Sci . (Camb.) 92 ( 1979 ) 123 - 132 .
 Fahmy M.H. , Feeding and management of prolific sheep: Under intensive management: the total confinement experiment in Canada , in: Fahmy M.H. (Ed.), Prolific sheep , CAB International, Wallingford, UK, 1996 , pp. 414 - 428 .
 Geenty K.G. , Sykes A.R. , Effect of herbage allowance during pregnancy and lactation on feed intake, milk production, body composition and energy utilization of ewes at pasture , J. Agric. Sci . (Camb.) 106 ( 1986 ) 351 - 367 .
 Gootwine E. , Bor A. , Braw-Tal R. , Zenou A. , Reproductive performance and milk production of the improved Awassi breed as compared with its crosses with the Booroola Merino, Anim . Sci. 60 ( 1995 ) 109 - 115 .
 INRA , Alimentation des bovins, ovins et caprins , INRA (Ed.), INRA, Paris, France, 1988 .
 Guerouali A. , Jonhson E.D. , Robershaw D. , Nutrient intake and digestibility in dry, pregnant and lactating prolific sheep receiving two levels of intake , J. Anim. Sci . 68 ( Suppl . 1) ( 1991 ) 322 .
 Khaldi G. , Variations saisonnières de l'activité ovarienne et du comportement d'oestrus et de la durée de l'anoestrus post-partum des femelles ovines de race Barbarine : influence du niveau alimentaire et de la présence du mâle , Thèse de Doctorat d'État Sciences Naturelles, U.S.T.L. Montpellier , 1984 , 168 p.
 Khaldi G. , Barbary sheep, in: Small ruminant in the Near East , Vol. III, North Africa , FAO , Anim. Prod. Heal. Paper 74 ( 1989 ) 96 - 135 .
 Lahlou-Kassi A. , Berger Y.M. , Bradford G.E. , Boukhliq R. , Tibary A. , Derquaoui L. , Boujenane I. , Performance of D'man and Sardi breeds of sheep in purebred and crossbred matings on an accelerated lambing schedule. I. Fertility, litter size, post-partum anoestrus and puberty , Small Ruminant Res.
 Lassoued N. , Rekik M. ,. Differences in reproductive efficiency between female sheep of the Queue Fine de l'Ouest purebred and their cross with the D'man , Anim. Res . 50 ( 2001 ) 373 - 381 .
 Le Houerou H.N. , Recherches expérimentales sur la tolérance du mouton Barbarin à l'inanition et la restriction alimentaire , in: Le Folch E., Grouzis M. , Cornet A. , Bille J.C . (Eds.), L'aridité : contrainte au développement , ORSTOM, France, 1992 , pp. 369 - 385 .
 Maijala K. , Prolific sheep breeds: the Finnsheep , in: Fahmy M.H. (Ed.), Prolific sheep , CAB International, Wallingford, UK, 1996 , pp. 10 - 46 .
 Ministry of Agriculture, État et perspectives des secteurs ovin et caprin en Tunisie, Document rédigé en arabe , 2002 , 28 p.
 Nefzaoui A. , Chermiti A. , Composition chimique et valeur nutritive pour les ruminants des fourrages et des concentrés d'origine tunisienne (Chemical composition and nutritive value of forges and concentrates of Tunisian origin ), Ann. INRAT 62 (Fasc. b) ( 1989 ) 36 .
 Perrier R. , Doreau M. , Effect of long-term underfeeding and subsequent refeeding on hay digestibility in sheep , Ann. Zootech. 44 ( Suppl . 1) ( 1995 ) 206 .
 Rekik M. , Ben Hammouda M., A steering frame for the genetic improvement of sheep and goats in Tunisia, Opt . Méditerr., Série A : Séminaires Méditerranéens 43 ( 2000 ) 129 - 136 .
 Ricordeau G. , Boccard R. , Denamur R. , Mesure de la production laitière des femelles ovines et bovines pendant la phase d'allaitement , Ann. Zootech. 9 ( 1960 ) 98 - 120 .
 Russel A.J.F. , Doney J.M. , Gunn R.J.G. , Subjective assessment of body fat in live sheep , J. Agric. Sci . (Camb.) 72 ( 1969 ) 451 - 454 .
 SAS, SAS User's guide: Statistics (Version 5 ed .), SAS Inst .Inc., Cary, NC , 1985 .
 Tilley J.M.A. , Terry R.A. , A two stage technique for the in vitro digestion of forage crops , J. Br. Grassland Soc . 108 ( 1963 ) 104 - 111 .
 Vermorel M. , Nutrition énergétique , in: Jarrige R. (Ed.), Alimentation des bovins, ovins et caprins , INRA (Ed.), INRA, Paris, France, 1988 , pp. 57 - 71 .