Livestock Research for Rural Development 12 (2) 2000

Citation of this paper

The effect of tropical ambient temperature on productive performance and grazing behaviour of sows kept in a outdoor system

R H Santos Ricalde and I  J Lean*

Facultad de Medicina Veterinaria y Zootecnia - UADY, Apartado Postal 4-116 Itzimna, C.P. 9700, Merida, Yucatan, Mexico.

* Wye College, University of London, Ashford, Kent, TN25 5AH, UK.


The effects of two ambient temperature regimes (Low, 19.0 to 33.0 C° vs High, 27.0 to 40.0 C°) on productive performance and grazing behaviour of sows kept in an outdoor system under tropical conditions was investigated in a completely randomised experiment with twelve replicates in each treatment.

Mean weight at weaning was higher (P<0.05) in sows kept in treatment Low (L) than in sows kept in treatment High (H). Similarly, sows in treatment L had a positive weight gain (P<0.05) and backfat gain (P<0.05) from farrowing to weaning and from mating to weaning in comparison to sows in treatment H. There was a  reduction (P<0.05) in milk fat and an increase (P<0.05) in milk ash and milk lactose in sows kept in treatment H in comparison to sows in treatment L. Sows in treatment H had a lower (P<0.05) feed intake during lactation than sows in treatment L. Sows in treatment H had lower values (P<0.05) for time spent grazing, grazing activity and distances walked during the day in comparison to sows in treatment L

The results obtained in this experiment suggest that there was a negative effect of extremely high ambient temperatures on feed intake of the lactating sows and consequently on their weight gain and backfat gain from farrowing to weaning. Also, the extremely high ambient temperatures reduced grazing activities of the pregnant sows and modified the maternal milk composition.

Key words: Sows, Outdoor system, grazing behaviour, productive performance, milk composition, ambient temperature.


Los efectos de dos regímenes de temperatura ambiental (Bajo, 19.0 to 33.0 C° vs Alto, 27.0 to 40.0 C°) sobre el comportamiento productivo y la conducta pastoril de marranas mantenidas en un sistema de producción en exterior bajo condiciones tropicales fueron estudiados en un experimento completamente al azar con doce replicas en cada tratamiento. El peso promedio al destete fue mas alto (P<0.05) en marranas mantenidas en el tratamiento Bajo (B) que en marranas mantenidas en el tratamiento Alto (A). Similarmente, las marranas del tratamiento B tuvieron una ganancia positiva de peso (P<0.05) y grasa dorsal (P<0.05) del parto al destete y de la monta al destete en comparación a las marranas del tratamiento A.

Se observó una reducción significativa (P<0.05) en el contenido de grasa de la leche y un incremento significativo (P<0.05) en ceniza y lactosa en la leche de las marranas del tratamiento A comparado con marranas del tratamiento B. Marranas en el tratamiento A tuvieron un consumo de alimento menor (P<0.05) que las marranas en el tratamiento B durante la lactancia.Marranas en el tratamiento A redujeron (P<0.05) el tiempo pastoreando, actividades pastoriles y distancias caminadas durante el día en comparación a marranas en el tratamiento B.

Los resultados obtenidos en este experimento sugieren que hubo un efecto negativo de las extremadamente altas temperaturas ambientales sobre el consumo de alimento de las marranas

durante la lactación y consecuentemente sobre su ganancia de peso y grasa dorsal del parto al destete. También las extremadamente altas temperaturas ambientales redujeron la actividad pastoril de las marranas gestantes y modificaron la composición de la leche materna.

Palabras claves: Marranas, sistema de producción en exterior, conducta pastoril, comportamiento productivo, composición de la leche, temperatura ambiental.


In Yucatán the peasant families usually keep pigs outdoors in the backyard. The peasant pig production has a target which is the saving of money (Richards and Leyva 1985). The pig diets are based mainly on maize, by-products from agriculture, domestic by-products and forages that they get from grazing activities.

Yucatán is situated in a tropical area in the Southeast of Mexico. In tropical climates the pigs are exposed to ambient temperatures above their thermo-neutral zone during the day (Holmes and Close 1977; McGlone et al 1988; Christon 1988; Serres 1992; Knap 1999). However, during the night they are close to the thermo-neutral zone. This diurnal variation in ambient temperature means that the pigs can accommodate to tropical climates (Giles et al 1988).

The exposure to high ambient temperatures has negative effects on pig performance, especially in lactating sows (Lynch 1977; McGlone et al 1988; Black et al 1993; Prunier et al 1994; Azain et al 1996). Reduction in feed intake, losses in liveweight, decreases in backfat depth and changes in behaviour have been reported in pigs exposed to high ambient temperatures in comparison to pigs kept at lower temperatures (Schoenherr et al 1989; Christon 1988; Prunier et al 1997; Knap 1999).

Studies on productive performance of pigs in a hot environment are scarce. In particular, no data are available on the effect of high temperatures on grazing behaviour in pigs. With the importance of the peasant pig production in Yucatan it is necessary to consider the effect of climate on productive performance and behaviour related to grazing activity in order to propose practical schemes of management for particular conditions.

The aim of the experiment described below was to determine the influence of two temperature regimes under tropical conditions on productive performance and grazing behaviour of sows kept in an outdoor system.

Materials and methods


The research was carried out at the FMVZ-UADY nutrition unit, in Mérida, Yucatán, México (29 Lat. N., 90 Long. W), from June 1997 to July 1998. Yucatán has a tropical sub-humid climate. The average rainfall is 928 mm/year and the rainy season is between May and October.

Treatments and diets

Twenty four primiparous pregnant sows were used in this experiment. They had a live weight of 115±11.1 kg and were of similar age at mating. The treatments were two ambient temperature regimes:

The Low (L) temperature regime was recorded between September and February. The High (H) temperature regime was recorded between March and August. The sows were allocated to one of the two ambient temperature regimes with twelve experimental units per treatment. The experimental units were individual sows. The sows in treatment L were mated in September and sows in treatment H were mated in March.

The sows were fed daily at 10:00 hours during pregnancy with a commercial lactation diet (140 g crude protein/kg and 13 MJ DE/kg). In order to be fed individually the sows were kept during feeding time in crates until they finished their respective diets. Fresh water was available ad libitum. In addition the sows had opportunity to graze freely in a paddock of Star grass (Cynodon nlemfuensys). Star grass (SG) is characterised by its high adaptative  capacity to different environmental conditions, growth aggressiveness and its high biomass yield per hectare per year, which has made it into one of the grasses most widely cultivated in Yucatan (Chamorro et al 1995). The SG field was divided into four paddocks by electric fences. A rotational system was used. Each paddock was grazed for seven days followed by 21 days rest. The paddocks of SG were irrigated every third day during the dry season (December to May) to keep the grass green and growing. During lactation the sows were fed ad libitum with a commercial lactation diet (150 g crude protein/kg and 13 MJ DE/kg).

General management

The sows were naturally mated three times. After mating the sows were moved immediately to the SG paddocks. Five days before the estimated date of farrowing the sows were each moved to an individual paddock divided by electric fences to farrow in a shelter. This was to avoid cross suckling between litters. After farrowing the sows were checked for metritis and agalactia. Intramuscular injection of antibiotic was used when necessary. It was not a common practice. Piglet management included umbilical cord disinfection, ear notching at farrowing and iron injection at three days of age. Piglets were weaned at 21 days.


Live weight was recorded by weighing each sow (500 kg capacity scale) at mating, at farrowing and at weaning. Before weighing the sows were keep without food and access to the field for around 12 hours to reduce effects due to gastrointestinal tract fill. Backfat thickness was monitored ultrasonically by a "Renco" lean meter, at 65 mm on each side of the midline at the level of the last rib (P2). These measures were done in the morning, before feeding.

Distance walked throughout the day was recorded by measuring distances walked between marked points in the paddock. Grazing time was measured by direct observation. Activities that related to grazing behaviour were recorded every 10 minutes according to the sample point methodology (Martin and Bateson 1993). These measurements were made during one day at 35, 70 and 105 days of gestation, between 7:00 and 19:00 hours. Ambient temperature was measured with a thermohydrograph during those days. After farrowing,  sow feed intake was monitored. Total feed intake during lactation was recorded from farrowing to 21 days of lactation.

The number of pigs farrowed and weaned was recorded for each litter. Litter weight was measured at farrowing and at 21 days of age, using a scale of 10 kg capacity.

Samples of milk were taken on day 14 of lactation, following the intramuscular injection of two ml of oxyctocin, and isolation from the litter for 60 minutes.

Analytical procedures

Milk samples were analysed for fat by the Gerber method, protein by Kjeldahl, ash by ignition at 450 ° C  and lactose by the "Folin" method.

Statistical analysis

Data of liveweight, backfat thickness, litter performance, feed intake and milk composition variables were analysed using the ANOVA procedure of SAS (1990) as a completely random experiment. Initial weight and initial backfat thickness of each sow were used as covariables in the analysis of weight changes and backfat changes, respectively, during gestation.  Ambient temperature regime was taken as the independent variable.

Data of grazing time, grazing activity, total distance walked and rectal temperature were analysed using the GLM procedure of SAS (1990) as a factorial experiment 2x3. Factors taken into account were ambient temperature regime (L and H) and day of gestation (35, 70 and 105).

Results and Discussion

The adjusted least square means of weight changes during pregnancy and lactation are given in Table 1. Sows in treatment H had a lower weight (P<0.05) at weaning in comparison to sows in treatment L. Similarly, sows in treatment H had a negative weight change (P<0.05) from farrowing to weaning and from mating to weaning. 


Table 1: Weight changes during pregnancy and lactation of sows kept under low and high ambient temperature regimes in an outdoor system
Low High SEM 
Weight changes (Kg)
At mating 119 111 1.1
At farrowing 159 153 2.9
At weaning 164 143 3.6*
From mating to farrowing 43.7 37.7 2.9
From farrowing to weaning 5.1 -9.4 2.5*
From mating to weaning 48.7 28.3 2.2*

Standard error of the mean. * P<0.05)

The adjusted least square means of backfat changes during pregnancy and lactation are shown in Table 2. The sows in treatment H lost more backfat (P<0.05) from farrowing to weaning. From mating to weaning sows in treatment H had a negative  backfat change (P<0.05) when compared with backfat depth of sows in treatment L.


Table 2:  Backfat changes during pregnancy and lactation of sows kept under low and high temperature regimes in an outdoor system
Low  High SEM
Backfat changes (mm)
At mating 11.8 13.0 0.4
At farrowing 14.4 14.6 0.9
At weaning 14.4 12.4 0.6
From mating to farrowing 2.0 2.2 0.6
From farrowing to weaning 0.04 -2.2 0.5*
From mating to weaning 2.0 -0.1 0.7*
SEM Standard error of the mean; * P<0.05

These results suggest that the environmental temperature in treatment H affected negatively the weight gain and backfat depth during lactation. These results are in agreement with other reports where the effect of high temperature on weight and backfat depth of lactating sows was studied (Lynch 1977; McGlone et al 1988; Azain et al 1996). Prunier et al (1994) pointed out that loss of live weight in sows during lactation was higher when they were kept between 25 and 38 °C during the summer than in winter (18 - 25 °C).

The reduction in backfat and liveweight during lactation was associated with reduction in feed intake during lactation in treatment H (Table 4). According to Prunier et al (1994) the consequence of elevated ambient temperature is a marked decrease in the voluntary feed intake and consequently losses in liveweight and backfat depth during lactation. Similar results have been reported by others (Lynch 1977; O'Grady et al 1985; Stansbury et al 1987; McGlone et al 1988; Schoenherr et al 1989; Black et al 1993). According to Black et al (1993) sows exposed to hot conditions mobilised their body reserves to support milk production. Similarly, sows in treatment H had a lower content of fat (P<0.05) in the milk than sows in treatment L. Also, there was observed a significant increase (P<0.05) in ash and lactose content in milk in treatment H compared to treatment L (Table 3).


Table 3:  Consumption during lactation and milk composition of sows kept in low and high ambient temperature regimes in an outdoor system
Low  High SEM
Feed intake during lactation 102.0 85.8 5.3*
Milk composition (%)
Fat 8.3 6.5 0.3*
Protein 5.3 5.4 0.2
Ash 0.8 1.0 0.04*
Lactose 4.1 5.4 0.2

SEM  Standard error of the mean; * P<0.05.

The reduction in milk fat observed in treatment H is not in accord with information reported by Schoenherr et al (1989) and Prunier et al (1997) who did not find an effect of high ambient temperature on milk fat. However, it has been reported that growing pigs kept in hot environments had a reduced thyroid activity (Christon 1988). The reduction in concentration of thyroid hormone in plasma of lactating sows is associated with a decrease in rate of metabolism and in milk production (Hitchcock et al 1972; Wung et al 1977; Cabell and Ebenshade 1990). Also, it has been reported that lactating sows kept in hot ambient temperature (30 °C) had a reduced concentration of free fatty acids in plasma in comparison to lactating sows kept at 20 °C (Prunier et al., 1997). Based on this information it can be proposed that a reduction in thyroid activity represents an efficient strategy used by sows exposed to hot environments to reduce heat production. The reduction in thyroid hormone could reduce free fatty acids circulating in plasma and decrease the metabolic rate in the mammary gland. It could explain the reduction in milk fat concentration in the lactating sows in treatment H.

On the other hand, there is no clear explanation for the increase in ash and lactose milk content in treatment H in comparison to treatment L. Little information has been generated to study the effect of high ambient temperature on milk composition of lactating sows. The evidence reported in the literature suggests a reduction in milk synthesis in lactating sows as a result of increase in ambient temperature (Schoenherr et al 1989; Black et al 1993; Prunier et al 1997). Vaccaro(1991) mentioned that milk composition values for cows in tropical climates are above the average values for cows in temperate climates. Vaccaro (1991) explained this as an effect of low yields per day and per lactation which tended to concentrate the components in the milk. Even though there significant differences in milk composition between treatments in this experiment, there was no effect (P>0.10) of treatments on litter performance (Table 4).

Table 4:  Effect of two temperatures regime on litter performance of sows kept in outdoor.

Low High SEM
No. of litters 12 12
Piglets born/litter 9.8 9.9 0.7
Piglets born alive/litter 9.4 9.8 0.6
Piglets birth weight (kg) 1.3 1.4 0.04
Litter weight born alive (kg) 12.4 13.3 0.8
Piglets alive 21 days 5.3 6.8 0.6
Piglets weight 21 days (kg) 5.3 5.0 0.3
Litter weight 21 days (kg) 28.2 33.7 3.2
SEM Standard Error of the mean.

The average results for grazing activity (Table 5) show that sows in treatment H reduced (P<0.05) considerably the time spent in grazing. When grazing was measured using the sample point technique the activity was reduced (P<0.05) in comparison with the sows in treatment L. Distance walked during the day was also reduced (P<0.05) for H compared with L sows.


Table 5:  Effect of temperature regime and days of gestation on grazing time, grazing activity and distance walked
. Temperature

Days of pregnancy

Low High 35 70 105 SEM 
Grazing time (min.)  114 82.6 98.9 97.1 98.4 3.1
Grazing activity (%) 14.6 11.8 13.2 12.6 13.9 0.5
Distance walked (m) 269 172 234 203 224 5.8
SEM Standard error of the mean.  No effect of stage of pregnancy (P>0.05); Significant effect of temperature for all traits (P<0.05)

The results in Table 5 suggest that extremely high ambient temperature reduce the grazing behaviour of sows kept outdoors during pregnancy. According to Halter et al (1979) and Schrama et al (1998) minimising physical activities reduces heat production. Thus, a reduction in grazing activities in treatment H in comparison to the treatment L represents a reduction also of heat produced by grazing activities. This reduction in muscular exercise associated with grazing activities represents an efficient mechanism to reduce heat stress in animals exposed to high temperatures (Williamson and Payne 1978; Knap 1999).

The little information about the effect of ambient temperatures higher than 30 °C on pregnant and lactating sows suggest that further research is necessary to establish the influence of extremely hot ambient temperatures on grazing behaviour, milk composition and milk production.

General conclusions


The authors gratefully acknowledge the help of the laboratory and technical staff of the Animal Nutrition Department in the Facultad de Medicina Veterinaria y Zootecnia-Universidad Autonoma de Yucatán for their technical assistance. Financial support was provided by CONACYT (project 0304 -PB)


Azain M J, Tomkins T, Sowiski J S, Arentson R A and Jewell D E 1996  Effect of supplemental pig milk replacer on litter performance: Seasonal variation in response. Journal of Animal Science. 74:2195-2202.

Black J L, Mullan B P, Lorschy M L and Giles L R 1993 Lactation in the sow during heat stress. Livestock Production Science 35:153-170.

Cabell S B  and Esbenshade K L  1990 Effect of feeding thyrotropin-releasing hormone to lactating sows. Journal of Animal Science. 68:4292-4302.

Chamorro O, Sarmiento L and Santos R 1995  Productive performance of finishing pigs fed increasing levels of African star grass (Cynodon nlemfuensis) in the diet. Cuban Journal of Agricultural Science. 29:239-334.

Christon R 1988 The effect of tropical ambient temperature on growth and metabolism in pigs. Journal of Animal Science. 66:3112-3123.

Giles L R, Belinda D E and Lowe R F  1988 Influence of diurnal fluctuating high temperature on growth and energy retention of growing pigs. Animal Production. 47:467-474.

Halter H M, Wenk C and Schurch A  1979  Effect of feeding level and feed composition on energy utilisation, physical activity and growth performance of piglets. In: Energy metabolism. Proceedings of the eight symposium on energy metabolism held at the Churchill College, Cambridge. September, 1979. (Editor: L E Mount) Butterworths. London. pp. 395-398.

Hitchcock J P, Ai C, Orr D E and Miller E R 1972  Thyroprotein for lactating sows. Journal of Animal Science. 35:1106 (Abstr.).

Holmes C W and Close W H 1977  The influence of climatic variables on energy metabolism and associated aspects of productivity in the pig. In: Nutrition and climatic environment. (Editors: W Haresign, H Swan  and D Lewis)  Butterworths. London. pp. 51-71.

Knap P W  1999  Simulation of growth in pigs: evaluation of a model to relate thermoregulation to body protein and lipid content and deposition. Animal Science. 68:655-680.

Lynch P B 1977  Effect of environmental temperature on lactating sows and their litters. Irish Journal of Agricultural Research. 16:123-130.

Martin P and Bateson P  1993  Measuring behaviour: An introductory guide. Cambridge University Press, Cambridge, UK. pp. 222.

McGlone J J, Stansbury W F, Tribble L F and Morrow J L  1988  Photoperiod and heat stress influence on lactating sows performance and photoperiod effects on nursery pig performance. Journal of Animal Science. 66:1915-1919.

O'Grady J F, Lynch P B  and Kearney P A  1985  Voluntary feed intake by lactating sows. Livestock Production Science. 12:355.

Prunier A, Dourmand J Y and Etienne M  1994. Effect of light regimen under various ambient temperatures on sows and litter performance. Journal of Animal Science. 72:1461-1466.

Prunier A, Messias de Braganca M and  Le Dividich J  1997 Influence of high ambient temperature on performance of reproductive sows. Livestock Production Science. 52:123-133.

Rinaldo D. and Le Dividich J  1991  Assessment of optimal temperature for performance and chemical body composition of growing pigs. Livestock Production Science. 29:61-75.

Richards E M  and Leyva M C 1985  An example of the use of economic analysis in the definition of research and development priorities: Pig production in the zone of Yucatan, Mexico. Research and Development in Agriculture. 2:7-17.

SAS 1990  SAS users guide: Statistics. Statistical Analysis Systems Institute. SAS Inst. Inc., Cary, NC.

Schrama J W, Bosch M W, Verstegen M W A, Vorselaars A H P M, Haaksma J and Heetkamp M J W  1998  The energetic value of nonstarch polysaccharides in relation to physical activity in group-housed growing pigs. Journal of Animal Science. 76:3016-3023.

Schoenherr W D, Stanhly T S and Cromwell G L 1989 The effects of dietary fat or fibre addition on yield and composition of milk from sows housed in a warm or hot environment. Journal of Animal Science. 67:482-495.

Serres H  1992  Manual of pig production in the tropics. CAB International. Wallingford, Oxon. UK. pp. 26-35.

Stansbury W F, McGlone J J and Tribble L F  1987 Effects of season, floor type, air temperature and snout cooling on sow and litter performance. Journal of Animal Science. 65:1507-1513.

Tauson A H, Chwalibog A, Ludvigsen J, Jakobsen K and Thorbek K G  1998  Effect of short-term exposure to high ambient temperatures on gas exchange and heat production in boars of different breeds. Animal Science. 66:431-440.

Vaccaro R  1991  Comportamiento de bovinos para doble proposito en el tropico. In: Memoria del seminario Internacional sobre lecheria tropical efectuado en Villahermosa, Tabasco del 20 al 24 de Noviembre de 1990. FIRA,  Morelia, Mexico. 1991. Vol. 3. pp. 14-35.

Verstegen M W A, Close W H, Start I B and Mount L E  1973  The effect of environmental temperature and plane of nutrition on heat loss, energy retention and deposition of protein and fat in groups of growing pigs. British Journal of Nutrition. 30:21-35.

Williamson G and Payne W J A 1978  An Introduction to Animal Husbandry in the Tropics. Third edition. Tropical Agriculture Series. Longman. London. pp. 755.

Wung S C, Wu H P, Kou Y H, Shen K H, Koh F K and Wan W C M  1977  Effect of thyrotropin-releasing hormone on serum thyroxine of lactating sows and the growth of their suckling young. Journal of Animal Science. 45:299-304.

Received 17 February 2000

Go to top