PHYSIOLOGICAL AND PRODUCTIVE RESPONSES OF ENVIRONMENTAL CONTROL ON HOUSED SOWS

Swine housing must promote an adjusted environment for thermal comfort and high animal productivity without negatively affecting the sow performance and reproductive response. This study evaluated the use of distinct environmental cooling equipments on sow performance, both on the gestation and on nursing in open sided housing. Two treatments were tested in the gestation building: natural ventilation and mechanical ventilation associated to fogging; while in the nursing rooms three treatments were tested: natural ventilation; mechanical ventilation; and evaporative cooling with forced ventilation. Sows were randomly chosen from the same genetic lot form six combined treatments. The evaporative cooling system in the farrowing room differed for piglet performances, at birth (4% higher) and on daily weight gain (15% higher), and also for sow physiological response improving the respiratory rate (8%) and back fat thickness (3%), without influencing skin temperature. The use of evaporative cooling directed to the sow head during nursing improved the physiological and productive results.


INTRODUCTION
Since the 1960's there have been significant changes in swine housing and management to improve production related to restrictions on animal movement, social interactions, thermal comfort, and welfare (Esmay, 1969;Pinheiro et al., 2002;Tolon, 2002).Current knowledge states that appropriate housing should meet the physiological needs of sows (Rapp et al., 1988;Turner et al., 1998;Bridges et al., 1998;Nääs, 2000).
Environmental control alleviates heat stress improving productive responses (Müller, 1982;Bortolozzo et al., 1997;Hannas, 1999).Reduction of the heat load can be achieved by properly designed ventilation and cooling systems, in both gestating and nursing housing (Bull et al., 1997;Gates et al., 1991a).Room evaporative cooling and head ventilation systems help reducing the negative effects of high temperatures in farrowing rooms (Dong et al., 2001;Barbari & Guerri, 2005).Misting systems are lower in efficiency when compared to conventional pad systems (Roller & Goldman, 1969;Bottcher et al., 1991) but the equipment noise may cause animal agitation (Barbari & Guerri, 2005).Thermal index evaluation in swine confinement has been used to identify regions suitable for a particular housing design (Turner et al., 1998).
This research aimed to identify and measure physiological and productive responses of sows exposed to distinct environmental cooling systems both in gestating and farrowing houses.The open sided gestating building was 10 m wide, 50 m long, 3 m high, and the long axis was East-West oriented, having natural ventilated opened sidewalls, with a fiber-cement roof painted white inside and outside, and the floor and walls were made of masonry.Sows were kept in metallic crates from pregnancy to seven days prior to delivery.In this building the sows were exposed to two treatments: T1 = natural ventilation and T2 = cooling system (mechani-cal ventilation with two axial 372.8 J s -1 fans, associated to fogging with ten fogging nozzles operating with 6.9 10 6 Pa pressure and 7 L h -1 distributed in two pipelines).The temperature for the fans to initiate was 25ºC and for the fogging systems was 27ºC.

MATERIAL AND METHODS
The nursing room in the building was 12.5 m long, 10 m wide and 3 m high, covered with metallic tile roof painted white with walls made of concrete bricks covered with mortar and painted white, and had fourteen individual pens (each 1.7 m wide and 2.8 m long).In this housing three environment control treatments were evaluated: T3 = natural ventilation (the walls were 1.7 m high with lateral openings of 1.3 m width and 12.5 length); T4 = mechanical ventilation (one axial 372.8 J s -1 fan); T5 = evaporative pad cooling with forced ventilation equipment (745.7 J s -1 ).The equipment pumped the cooled air through a 0.15 m PVC pipe directly 0.40 m above sow heads.
The sows used for the experiment had the same genetics.At gestation, the animals were treated two times per day with a balanced feed (16% of brute protein and 3.2 kCal of metabolized energy).At farrowing the sows were under a standard lactation diet.The maximum feed intake was 6 kg day -1 .One hundred and twenty six sows that were previously under treatments T1 or T2 were randomly chosen.These sows were equally distributed to three farrowing rooms (T3, T4 or T5).Table 1 shows the treatments Ti-j combined where i represents the treatment in the gestation phase and j the treatment in the nursing stage.
Climatic data inside both gestation and nursing rooms were collected using data loggers.Skin temperature (ST) on the back of the sows was daily measured at 10h00 using an infra red thermometer, and the back fat thickness (BFT) was measured using a digital equipment, approximately at the 10 th rib, 6 cm from the middle back line, in all sows at entering and leaving each treatment in the nursing house.The respiration rate (RR) was recorded for the nursing sows through the observation of the number of flank move- *i represents the treatment in the gestation phase and j in the nursing stage.T1-3 = Natural ventilation both in the gestation stage (T1) and in the nursing stage (T3); T1-4 = Natural ventilation in the gestation stage (T1) and mechanical ventilation in the nursing stage (T4); T1-5 = Natural ventilation in the gestation stage (T1) and evaporative cooling in the nursing stage (T5); T2-3 = Cooling system in the gestation stage (T2) and natural ventilation in the nursing stage (T3); T2-4 = Cooling system in the gestation stage (T2) and mechanical ventilation in the nursing stage (T4); T2-5 = Cooling system in the gestation stage (T2) and evaporative cooling in the nursing stage (T5).ments per minute, with three replicates between 10h00 e 11h00.The piglets were weighed at birth and weaning (21 days).
Descriptive and statistical analysis of mean comparison using the Tukey test was performed using the Minitab ® software.

RESULTS AND DISCUSSION
A direct correlation was found between piglet live weight and the heat loss to the environment by the sows in the nursing building.The more comfortable the sows were the better the piglet performance was, as already pointed in the literature (Mount, 1974;Hannas, 1999;Nääs, 2000;Pinheiro et al., 2002).
The best result of total weight gain at weaning (6.27 kg) as well as the average daily gain (0.32 kg day -1 ) were found for sows housed under natural ventilation at the gestation building (T1) and had evaporative cooling in the nursing house (T5) (Table 2).Similar results were found by Nunes et al. (2003), who registered a better environmental condition at the gestation housing with a forced ventilation system.Nevertheless this system was not sufficient to change sow reproductive efficiency.The evaporative cooling system in the farrowing room was determinant for piglet performance.These results are in accordance with Wagenberg et al. (2006), who showed that cooling the floor under the sow shoulder using a farrowing crate increased piglet performance (22 g day -1 per piglet, 9% higher than in the ordinary system).Similar results were found by Tavares et al. (2000) who recorded lower weight gain under heat stress, even though the carcasses were not affected, as well as Nääs (2000) got analogous results when observing the increase of productivity of sows housed in the same region during the cooler months of the year.
Even though expected from literature data, the best performance of the combined treatment T2-5, indicates that the forced ventilation system (T2) used during the gestation housing did not influence piglet performance (Table 2).Probably the position of the fans and consequently the air flow distribution had negative effect on the forced ventilation system efficiency in the gestation room.The treatment using natural ventilation (T1-3) produced the lowest piglet weight gain (Table 1), in agreement with Wagenberg et al. (2006).
Mean values of ST, RR and BFT are shown in Table 3, which were measured in order to estimate thermal comfort.The sensible thermal losses decrease with heat stress exposure, and a change in skin temperature was expected as mentioned by Brown-Brandl & Eigenberg (2000), Yan & Yamamoto (2000), and Brown-Brandl et al. (2004).This fact was however not observed, probably due to the fact that all authors    worked in controlled environments, while this present study was carried out in a swine farm; exposed to the local climatic parameters (Table 4).The difference was not in favor of the evaporative cooling system (37°C, higher average of skin temperature).These results were representative of the animal back, while the evaporative cooling system was directed to the sow head.
The lowest value in respiratory rate was found for the nursing treatment T5 (50.2 breaths min -1 ) and the highest for T4 (57.8 breaths min -1 ), showing a distinct effect on the respiratory rate of the animals as a function of the environmental conditions, as suggested by Yan & Yamamoto (2000).The results also differ in part from Hannas (1999), who found that the respiratory rate is the first response of animals when exposed to thermal stress.The respiratory rate for the forced ventilation system (T4) in the farrowing room were probably affected by the high restlessness of the animals associated to a high noise level of the fan in this treatment, when compared to the natural ventilation system treatment (T3).The results were similar to those obtained by Gates et al. (1991b) and Bull et al. (1997) who found that growing-finishing swine as well as sows present highest performance when housed in an environment with evaporative cooling, when compared to both misting and mechanical ventilation systems.
When exchanging heat by conduction swine reach the thermal neutral balance and their respiratory rate tends to become normal, even after exposure to acute heat stress (Bridges et al., 2000;Brown-Brandl & Eigenberg, 2000).Under treatments T3 (natural ventilation) and T4 (mechanical ventilation) the sows presented a certain degree of heat stress exposure, in agreement to Esmay (1969).Tavares et al. (2000), evaluating the effect of heat stress on swine performance concluded that the respiratory rate increased considerably when the environmental temperature was above 26°C, which is above the thermal comfort zone.
Results on back fat thickness (Table 3) point to difference between treatments, and both treatments T4 and T5 presented the best result (14.9 mm), in which the evaporative cooling system (T5) had the lowest variation.These results follow the same trend found by Derno et al. (1995), they however conflict with the results found by Quiniou et al. (2000), who did not find variations in back fat thickness due to change in environmental conditions.This indicates the possibility that the most effective acclimatization systems used during nursing were those that used forced ventilation.In this study there was no change in the room temperature, however the wind speed had a positive effect in alleviating heat stress by increasing the upper limits of the critical temperature during the nursing stage, and influencing the results of back fat thickness, as pointed out by Aherne & Foxcroft (2000).

CONCLUSIONS
Considering piglet performance at weaning as well as the measured physiological response of the sows, the treatment that presented the most effective combination for alleviating heat stress was the use of natural ventilation in gestating buildings, associated to evaporative cooling directed to the sow head during nursing.The use of this type of cooling device represents a positive investment since it results in both the highest piglet mean daily weight gain and total weight at weaning.
Results of piglet performance during lactation.a, b, c distinct letters mean statistical difference (α = 0.05).T1-3 = Natural ventilation both in the gestation stage (T1) and in the nursing stage (T3); T1-4 = Natural ventilation in the gestation stage (T1) and mechanical ventilation in the nursing stage (T4); T1-5 = Natural ventilation in the gestation stage (T1) and evaporative cooling in the nursing stage (T5); T2-3 = Cooling system in the gestation stage (T2) and natural ventilation in the nursing stage (T3); T2-4 = Cooling system in the gestation stage (T2) and mechanical ventilation in the nursing stage (T4); T2-5 = Cooling system in the gestation stage (T2) and evaporative cooling in the nursing stage (T5). t Sow skin temperature (ST), respiratory rate (RR), and back fat thickness (BFT) for all nursing treatments (Mean ± standard deviations).a,b, c distinct letters mean statistical difference (α = 0.05) for each column.

Table 1 -
Treatments distribution in the experiment.

Table 4 -
Mean environmental variables during the experiment inside the gestation and farrowing housing, and outside.