Reprod Dom Anim 50, 632–636 (2015); doi: 10.1111/rda.

12540
ISSN 0936–6768

Management Strategies Aiming to Improve Horse Welfare Reduce Embryonic Death

Rates in Mares
E Malschitzky1, AM Pimentel2, P Garbade2, MIM Jobim2, RM Gregory2 and RC Mattos2
1 2
ULBRA, Curso de Medicina Veterin
aria, Canoas, Brazil; Reprolab, Faculdade de Veterin
aria, UFRGS, Porto Alegre, Brazil

Contents Disrupted social orders such as dominance rank and

The objective of this retrospective study was to evaluate the variations in population density predispose to reduced
effect of management strategies aiming to improve animal well-being (Fraser 1992). Horses confined with new and
well-being on pregnancy and embryonic death (ED) rates. hostile equine companions respond either by withdraw-
Breeding records of a cohort of 1206 Thoroughbred mares ing from confrontation or by fighting for a place in the
brought to a stallion station facility, to be bred with the hierarchy (Mills and Nankervis 1999). The effects of
stallions housed there, were evaluated during ten breeding
seasons. Mares were blocked according to management
isolation, of crowding into yards, rough handling at
strategies in two groups: Stress and Relax. Strategies used to teasing or other yarding operation increases the level of
improve animal well-being (Relax group) were as follows: stress (Fraser 1992; Irvine and Alexander 1994). Also,
stopping the teasing routine, reducing or eliminating stall management of mares for artificial insemination may act
confinement, reducing the number of mares per group and as a stressor, resulting in higher concentrations of
maintaining herd stability during the breeding season. In cortisol secretion (Berghold et al. 2007).
barren mares, the pregnancy rate was higher in the Relax The stress response depends on the intensity and
group (91.8%) when compared to the observed in Stress group duration of the stressor (Dobson and Smith 2000).
(84.7%). However, no difference in pregnancy rates were Prolonged or chronic stress results in suppressed
observed (Stress = 85.2% vs. Relax = 86.2) in foaling mares. gonadotrophin secretion and inhibition of reproduction
ED rate was higher in barren and foaling mares of the Stress
group mares (25.5% and 26.8%, respectively) compared with
(Tilbrook et al. 2000; Breen and Mellon 2014) due to an
the Relax group (16.1% and 14.7%, respectively). No signif- increase in plasma concentrations of the glucocorticoids,
icant differences were observed on foal heat pregnancy rate cortisol and corticosterone. The effect of endogenously
between groups; yet, the embryo loss on foal heat was or exogenously raised cortisol levels on the minute-to-
significant reduced in Relax mares (Relax = 8.7% vs minute gonadotrophin secretion patterns has not been
Stress = 24.5%). In conclusion, management strategies aimed studied in horses (Alexander and Irvine 2011). In ewes,
to reduce social stress can reduce early pregnancy losses and cortisol reduces the amplitude of LH and probably acts
the average cycles per pregnancy, improving reproductive at the pituitary to lower the responsiveness to GnRH
performance in mares. pulses (Breen and Karsch 2006).
Detecting behavioural stress in domestic animals is
extremely difficult, and the conclusions drawn from such
Introduction determinations are frequently controversial. One
Embryonic loss is an important factor in infertility in approach is to monitor reproductive success (Moberg
mares with a reported incidence ranging from 5% to 1991). Management practices severely limit the expres-
24% (Ball et al. 1986). Early embryonic failure repre- sion of full range of reproductive behaviours by
sents a considerable economic loss to the equine focusing attention on what is necessary for breeding
industry in the form of increased costs and decreased rather than what optimizes reproductive success (Curry
foal production (Vanderwall 2008). Factors that may et al. 2007). Besides, the inclusion of potentially positive
contribute to the occurrence of embryonic loss in the environmental signals and practices into the mare
mare have been classified as intrinsic, extrinsic and management programme is also recommended to
embryonic (Ball 1988). Extrinsic factors include stress exclude potentially negative signals (Irvine and Alexan-
that decreases circulating progesterone apparently med- der 1994). It is hypothesized that mare management
iated through adrenal corticosteroids (Van Niekerk and changes aimed to reduce social stress improve the
Morgenthal 1982). reproductive performance.
Horses are a social species that readily form a social The objective of this retrospective study was to verify
order. When they live in bands, a clear social hierarchy the effect of management changes aimed to reduce stress
becomes established (Mills and Nankervis 1999). Mares situations on the reproductive performance of the mares.
form strong positive social attachments with a small
number of similar ranking peers, which are reinforced
with increasing residency time within the group (Curry Materials and Methods
et al. 2007). Mare–mare bonds are very stable and Breeding records of a cohort of 1206 Thoroughbred
persist for years (Van Dierendonck et al. 2004). mares, aged between 4 and 24 years were evaluated

© 2015 Blackwell Verlag GmbH

Management Reduce Embryonic Death in Mares 633

during ten breeding seasons. Mares were brought each

year to Posto de Fomento, a stallion station facility
located in Porto Alegre, south Brazil (parallel 30° S), to
be bred with the stallions housed there, with proven
normal fertility. The percentage of permanent resident
mares and ‘walk-in-mares’ was 30% and 70%, respec-
tively.
All barren mares were submitted to an artificial
photoperiod either from May 15 on, or later on, from
the moment they entered the management centre.
Foaling mare were put under light 45–60 days before
parturition to reduce post-partum anestrus
(Malschitzky et al. 2001). Artificial photoperiod was
from 15/9 h (light/darkness) with incandescent bulb
(10 W/m2).
From August 1 on, all mares in oestrus were examined
by transrectal palpation and ultrasound every day or
Fig. 1. Mares tied during feed to prevent aggressions
every other day until ovulation to evaluate follicular
growth, grade of uterine oedema and presence of
intrauterine fluid (IUF). The grade of uterine oedema
was classified in a scale from 1 (no oedema – dioestrus) foaling mares. The composition of the groups was
to 3 (intense uterine oedema – oestrus). Foaling mares frequently modified according to the reproductive man-
were examined by transrectal palpation and ultrasound agement (oestrus mares, served mares waiting preg-
on day 5 post-partum and every day or every other day nancy diagnostic, early pregnant mares). The mares
until oestrus began. were daily individually teased before transrectal palpa-
Mares were bred when a pre-ovulatory follicle tion and ultrasonography.
(>40 mm) was observed, in association with a reduced
grade of uterine oedema (Ginther 1986). A new
examination was performed 24–48 h after breeding to Relax
confirm ovulation and to detect IUF. Mares showing Breeding records of 443 foaling and 256 barren mares
IUF were treated by means of, at least, three uterine were evaluated. Barren mares were kept in individual
saline lavages (2 l each) (Mattos et al. 1997), followed boxes during the night until day 20 of pregnancy and
by antibiotic uterine infusion [5 000 000 IU Penicillin then left lose in a paddock 24 h a day. When mares were
K (Aricilina- Ariston Ind
ustria Qu
ımica e in the boxes, they received oat and alfafa. Immediately
Farmac^eutica Ltda, Brazil) or 1.6 g Gentamicine after parturition, foaling mares were maintained exclu-
(Gentaflex- Eurofarma Laborat
orios, Ltda, Brazil)] sively in the paddock. Paddocks had more than 10 ha
during 3 days. with good forage quality, and the mares had free access
Pregnancy was diagnosed at day 12 post-ovulation by to water. Animals were fed twice a day in individual
ultrasonography. Pregnant mares were examined at feeders (3–4 kg oat and 1.5–2 kg concentrate feed with
least once a week until day 45 of pregnancy. Twin 14% protein) and tied to prevent aggression (Fig. 1).
pregnancies were manually reduced until the 16th d after The groups from 5 to 10 mares according to breeding
ovulation. Pregnancy losses occurring between the first (barren mares) or foaling date (foaling mares) were
diagnosis and the 45th day of pregnancy were recorded stable, and no newcomers were introduced until the end
as embryonic death (ED). of the breeding season. Teasing was abolished and
The mares were blocked according to the management oestrus detection was based only in transrectal palpation
strategies in two groups as follow: and ultrasonography.

Stress Statistical analyses

Breeding records of 291 foaling and 216 barren mares Pregnancy rates and embryo loss rates were compared
were included. Animals were kept in semi-intensive using chi-square test. A one-way ANOVA was performed
conditions. Mares were released in paddocks (>20 ha) to evaluate the effect of both experimental groups as
with good forage quality and free access to water every main factor arranged in random blocks and the age,
day from 7:00 am to 6:00 pm. At night, mares were number of cycles per pregnancy, mare residency status
housed in individual boxes with free access to water. and stallion as dependent variables. Tukey test was used
Barren and foaling mares received every morning and to compare means when a significant effect was
evening 3 and 4 kg oat, 1.5 and 2 kg concentrate feed obtained. A probability of p ≤ 0.05 indicated that a
with 14% protein and 3 kg alfafa hay, respectively. difference was significant and those with 0.05 > p < 0.1
Mare groups were composed from circa 20 barren or were considered as tendency.

© 2015 Blackwell Verlag GmbH

634 E Malschitzky, AM Pimentel, P Garbade, MIM Jobim, RM Gregory and RC Mattos

Results Table 1. Seasonal pregnancy and embryo death (ED) rates in

Thoroughbred barren mares maintained in different management
The results of reproductive performance of barren and systems
foaling mares are presented in Tables 1 and 2, respec-
tively. In barren mares, seasonal pregnancy rate was Seasonal
significantly higher (p = 0.02) in Relax mares (91.8%) pregnancy
rate
than that in Stress group (84.7%). In foaling mares, no
difference on pregnancy rates was detected between both Management n n % ED (%) Cycles/pregnancy
groups (p = 0.7). In barren and foaling mares, embryo
death was lower in Relax group than that in Stress Stress 216 183 84.7a 25.5c 2.0e
group. In mares served on foal heat, no difference was Relax 256 235 91.8b 16.1b 1.7f
observed in pregnancy rates p = 0.68); however, a
Different characters (a, b) in columns indicate significant difference (p = 0.02).
significant increase (p = 0.02) of ED was detected in Different characters (c, d) in columns indicate significant difference (p = 0.01).
mares bred on foal heat of Stress group in comparison Different characters (e, f) in columns indicate difference (p = 0.07).
with mares of Relax group (Table 3).
No significant difference was detected on average ages
of barren (Stress: 10.5 vs. Relax: 10.8 years old; Table 2. Season pregnancy and embryo death (ED) rates in
Thoroughbred foaling mares maintained in different management
p = 0.31) and foaling mares (Stress: 9.5 vs. Relax: systems
9.8 years old; p = 0.22). No significant differences were
observed in the pregnancy rates obtained in Stress and Seasonal
Relax groups with the different stallions (p = 0.48) and pregnancy
in both mare residency status (permanent and walk-in rate
mares) (p < 0.01).
Management n n % ED (%) Cycles/pregnancy

Stress 291 248 85.2 26.8a 1.91c

Discussion Relax 443 382 86.2 14.7b 1.64d

A high incidence of early pregnancy failure represents a Different characters (a, b) in columns indicate significant difference (p < 0.01).
major loss to Thoroughbred industry (Morris and Allen Different characters (c, d) in columns indicate significant difference (p = 0.02).
2001). Stress is considered one of the factors that may
contribute to the occurrence of embryonic loss in the
Table 3. Pregnancy and embryo death (ED) rates in Thoroughbred
mare (Ball 1988). Social and environmental disturbance mares served in foal heat and maintained in different management
raises free cortisol plasma level in horses and can affect systems
reproductive hormones (Irvine and Alexander 1994
Alexander and Irvine 1998; Berghold et al. 2007). In Foal heat
the present study, mares submitted to management pregnancy rate
strategies aiming to diminish stress presented a reduc-
Management n n % ED (%)
tion of ED and from the pregnancy rate per cycle in
relation to mares maintained in the Stress group. Stress 273 165 60.4 30.1a
The formation of stable groups from 5 to 10 mares Relax 344 214 62.2 17.3b
and the no introduction of newcomers until the end of
the breeding season may result in a lower social stress in Different characters (a, b) in columns indicate significant difference (p = 0.02).
the Relax group from this study in relation to the Stress
group. Feral horses generally live in long-term stable the increase of ED and the raise of the pregnancy rate
social groups consisting of a single adult male and per cycle.
between 1 and 7 adult females (Curry et al. 2007). Social The instability of the Stress group was another factor
attachments formed by the mares are very strong, with a that probably increased the embryonic loss and the
small number of peers and persisting across consecutive pregnancy rate per cycle. While in the Relax group no
years (Van Dierendonck et al. 2004; Curry et al. 2007). newcomers were introduced during the breeding season,
Horses normally do not form social groups of more the Stress group had its composition frequently modi-
than 20 individuals (Tarouco et al. 2009). To live in a fied according to the reproductive management. The
proper social society, individuals must be able to disruption of the social order, such as dominance rank
recognize others and remember their relationship with and variations in population density, predisposes to
them. Probably twenty represent the mental limit of a reduced well-being (Fraser 1992). The stress response
horse’s social memory, and it is likely that a horse’s depends on the intensity and duration of the stressor
social circle tends to be smaller (less than ten horses). (Dobson and Smith 2000), resulting in suppressed
Beyond this, they may not be able to organize them- gonadotrophin secretion and inhibition of reproduction
selves in a proper horse society, especially when space is (Tilbrook et al. 2000; Breen and Mellon 2014) due to an
limited (Mills and Nankervis 1999). Probably the big increase in plasma concentrations of the glucocorticoids,
number of mares in the Stress group contributed with cortisol and corticosterone. These results may explain

© 2015 Blackwell Verlag GmbH

Management Reduce Embryonic Death in Mares 635

the reduced delivery intervals observed in mares with Exercise improves the overall physical and mental
long term and stable relations when compared of those health of the mare (Pugh and Scumacher 1990). How-
groups with unstable relations (Kaseda et al. 1995). ever, the role of exercise in breeding management of
Cortisol reduces the amplitude of LH and probably acts mares is poorly understood. Most farm managers do
at the pituitary to lower the responsiveness to GnRH not exercises mares other than allowing time in pad-
pulses in ewes (Breen and Karsch 2006). In the mare, docks (Hintz 1993). Yet, this procedure permits that
glucocorticoids administration reduces the ability of the mares eliminate uterine accumulated fluid and enhance
hypothalamus to secrete GnRH and alters the produc- fertility when compared with mares restricted to a stall
tion, storage or secretion of LH by the pituitary (Ferris (Chopin 2011). It is possible that the better pregnancy
and McCue 2010). per cycle observed in the Relax mares could be raised
Mares from the Relax management system were kept due to the free movement of the mares inside the
in paddocks 24 h a day, immediately after delivery paddocks eliminating uterine fluids.
(foaling mares) or after the 20th day of pregnancy Management strategies aimed to reduce social stress
(barren mares). Movement and association are required can reduce early pregnancy losses and the average cycles
by horses to develop properly and naturally. They per pregnancy, improving reproductive performance in
benefit greatly by being raised in a suitable, natural mares.
environment (Fraser 1992). Horses are roaming and
grazing animals (Umphenour and Steiner 2000) and
spend 60% or more of their time eating when grazing or Acknowledgements
when feed is available free choice. Interestingly, their The authors are thankful to Coordenacßa~o de Aperfeicßoamento de
intake is stimulated by the presence of other horses Pessoal de N
ıvel Superior, Conselho Nacional de Desenvolvimento
(Houpt 1990). A continuous foraging pattern with hay Cient
ıfico e Tecnol

ogico and FAPERGS for funding this study; to
Erika Weber for photograph cession.
distributed in a paddock increased conception rates in
Arab mares when compared with the same total amount
of hay offered only at night in individual boxes Conflict of interests
(Benhajali et al. 2013). In the present study, mares from None of the authors have any conflict of interest to declare.
the Relax group spent more time in paddocks than the
mares from the Stress group, allowed to graze the entire
day, a factor that could have helped to the better Author contributions
pregnancy per cycle rates. Stressor factors, such as E Malschitzky designed the study, collected data and prepared the
feeding the mares in the paddocks, were minimized in manuscript. AM Pimentel carried out analysis and revised paper. P.
the Relax group by tying the mares in the fence avoiding Garbade, MIM Jobim, RM Gregory and RC Mattos participated in
conflicts. the design of the study and in the preparation of the manuscript.

on cortisol secretion and fertility of oe- Ginther OJ, 1986: Ultrasonic Imaging and
References strous horse mares. Anim Reprod Sci 102, Reproductive Events in the Mare. Equiser-
Alexander SL, Irvine CHG, 1998: The effect 276–285. vices, Cross Plains, 378 p.
of social stress on adrenal axis activity in Breen KM, Karsch FJ, 2006: New insights Hintz H, 1993: Nutrition of the breeding
horses: the importance of monitoring regarding glucocorticoids stress and gon- mare. In: McKinnon AO, Voss J (eds),
corticosteroid-binding globulin capacity. adotropin suppression. Front Neuroen- Equine Reproduction. Lea & Febiger,
J Endocrinol 157, 425–432. docrin 27, 233–245. Philadelphia, pp. 631–639.
Alexander SL, Irvine CHG, 2011: GnRH. Breen KM, Mellon PL, 2014: Influence of Houpt KA, 1990: Ingestive behavior. Vet
In: McKinnon AO, Squires SL, Vaala stress-induced intermediates on gonado- Clin North Am Equine Pract 6, 319–337.
WE, Varner DD (eds), Equine Reproduc- tropin gene expression in gonadotrope Irvine CHG, Alexander SL, 1994: The role
tion. 2nd edn. Wiley Blackwell, West cells. Mol Cell Endocrinol 385, 71–77. of environmental factors in reproduction
Sussex, pp. 1608–1618. Chopin JB, 2011: Preventive medicine for in the mare. Ars Vet 10, 33–41.
Ball BA, 1988: Embryonic loss in mares: broodmares farms. In: McKinnon AO, Kaseda Y, Khalil AM, Ogawa H, 1995: Harem
Incidence, possible causes and diagnostic Squires SL, Vaala WE, Varner DD (eds), stability and reproductive success of Misaki
considerations. Vet Clin North Am: Equine Reproduction. 2nd edn. Wiley feral mares. Equine Vet J 27, 368–372.
Equine Pract 4, 263–290. Blackwell, West Sussex, pp. 2747–2752. Mills D, Nankervis K, 1999: Equine Behav-
Ball BA, Little TV, Hillman RB, Woods Curry MR, Eady PE, Mills DS, 2007: ior: Principles and Practice. Blackwell
GL, 1986: Pregnancy rates at days 2 and Reflections on mare behavior: social and Science Ltd, Oxford, 232 pp.
14 and estimated embryonic little loss sexual perspectives. J Vet Behav 2, 149– Malschitzky E, Schilela A, Meirelles LS,
rates prior to day 14 in normal and 157. Mattos ALG, Gregory RM, Mattos RM,
subfertile mares. Theriogenology 26, Dobson H, Smith RF, 2000: What is stress, 2001: Artificial photoperiod in pregnant
611–619. and how does it affect reproduction? mares and its effect on pregnancy length
Benhajali H, Ezzaouia M, Lunel C, Charfi Anim Reprod Sci 60, 743–752. and postpartum reproductive perfor-
F, Hausberger M, 2013: Temporal feeding Ferris RA, McCue PM, 2010: The effects of mance. Pferdeheilkunde 17, 565–569.
pattern may influence reproduction effi- dexamethasone and prednisolone on pitu- Mattos R, Castilho LFF, Malschitzky E,
ciency, the example of breeding mares. itary and ovarian function in the mare. Keller A, Gregory RM, Mattos RC, 1997:
PLoS ONE 8, e73858. Equine Vet J 42, 438–443. Uterine lavage with saline in mares as
Berghold P, M€ ostl E, Aurich C, 2007: Effects Fraser AF, 1992: The Behavior of the Horse. treatment for endometritis. Pferdeheilkunde
of reproductive status and management CABI Publishing, Oxon, pp. 184—197. 13, 521–524.

© 2015 Blackwell Verlag GmbH

636 E Malschitzky, AM Pimentel, P Garbade, MIM Jobim, RM Gregory and RC Mattos

Moberg GP, 1991: How behavioral repro- non-rodent mammals: the role of gluco- plasma progestagen levels in pregnant
duction in stress disrupts the endocrine corticoids and sex differences. Rev Re- Thoroughbred mares. J Reprod Fertil
control of domestic animals. J Dairy Sci prod 5, 105–113. Suppl 32, 453–457.
74, 304–311. Umphenour NW, Steiner JV, 2000: Breed- Vanderwall DK, 2008: Early embryonic loss
Morris LHA, Allen WR, 2001: Reproduc- ing management of the Thoroughbred in the mare. J Equine Vet Sci 28, 691–702.
tive efficiency of intensively managed stallion. In: Samper J (ed), Equine
thoroughbred mares in Newmarket. Pfer- Breeding Management and Artificial
deheilkunde 17, 548–556. Insemination. W.B. Saunders, Philadel- Submitted: 13 Feb 2015; Accepted: 21 Apr
Pugh D, Scumacher J, 1990: Management of phia, pp. 63–72. 2015
the broodmare. Proc 36th Ann Conv Am Van Dierendonck MC, Sigurjonsdottir H,
Ass Equine Pract 36, 61–78. Colenbrander L, Thorhallsdottir AG, Author’s address (for correspondence):
Tarouco AK, Freitas CC, Neves AP, Greg- 2004: Differences in social behaviour
Rodrigo Costa Mattos, Reprolab, Faculd-
ory RM, Mattos RC, 2009: Sexual and between late pregnant, post-partum and
ade de Veterin
aria, UFRGS, Av. Bento
social behaviors of pony stallions and barren mares in a herd of Icelandic horses. Goncßalves 9090, 91540-000 Porto Alegre,
mares. Anim Reprod 6, 534–544. Appl Anim Behav Sci 89, 283–297. Brazil. E-mail: [email protected]
Tilbrook AJ, Turner AI, Clarke IJ, 2000: Van Niekerk CH, Morgenthal JC, 1982:
Effects of stress on reproduction in Fetal loss and the effect of stress on

© 2015 Blackwell Verlag GmbH