| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ANIMAL GENETICS |
* IPG, Institute for Pig Genetics B.V., PO Box 43, 6440 AA Beuningen, the Netherlands; and
Animal Breeding and Genomics Centre, Wageningen University, PO Box 338, 6700 AH Wageningen, the Netherlands
 |
Abstract |
|---|
 Top Abstract INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Key Words: heat stress • reproductive performance • sow • sow line • upper critical temperature
|
INTRODUCTION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). It is affected by several factors such as breed, parity, lactation length, nutrition, season, and temperature. Several authors described an adverse effect of high temperature on reproduction (Peltoniemi et al., 1999; Tummaruk et al., 2004; Suriyasomboon et al., 2006). Sows are exposed to heat stress when temperature exceeds the upper critical temperature (UCT) of the thermo-neutral zone of the sow. This zone lies between 12 and 22°C (Black et al., 1993) for lactating sows.
Pig production occurs on commercial farms all over the world including in hot climates. Genetic selection of pigs for increased (re)production occurs in nucleus herds in mainly temperate climates. As a consequence, pig breeders face a wide variety of environmental conditions in which their pigs are required to perform (Knap, 2005). When evaluating breeding goals for dam lines, we were confronted with differences in appreciation of 2 sow lines by farmers in the Netherlands (a temperate environment) and in Spain (a warmer environment). The line preferred by Dutch farmers was not preferred by the Spanish and vice versa. Reproduction differences between Dutch and Spanish farms substantiated this opinion. In dairy cattle, sufficient genetic variation has been found in heat stress tolerance for milk, fat, and protein production and nonreturn rate to allow genetic selection (Ravagnolo and Misztal, 2000, 2002). To our knowledge no study has focused on the genetic variation in heat stress tolerance in sows.
The objectives of our study were 1) to investigate if there were differences in the relationship between temperature and reproductive performance traits in 2 different sow lines, a Yorkshire line producing mainly in temperate climates and a Large White line producing mainly in warm climates, and 2) to determine the UCT for the reproductive performance of these 2 lines.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
Data
Data were composed of 39,038 first insemination records from 12,442 sows inseminated from January 2003 until December 2005. Sows were kept on 20 sow farms in Spain; 2,850 sows originated from a Dutch (D) purebred Yorkshire sow line and 9,592 sows originated from an International (I) purebred Large White sow line. Both sow lines belonged to genetic lines of the TOPIGS breeding company (Vught, the Netherlands). For line I a nucleus farm with related multiplication farms was operated for more than 15 yr in Spain. The nucleus farm provided its own female replacements; male replacements came from the Netherlands. Similar structures existed for Italy, the Philippines, and Brazil. Data for EBV estimation came from all of these farms. For line D a similar structure existed over these years with solely Dutch nucleus and multiplication farms. Selection in both lines was based on BW gain, backfat, litter size, and litter mortality. In Spain the D-line was located on 8 farms and the I-line on 12 farms.
In this study only first insemination results per parity were considered. Sows ranged in parity from 1 to 14. For each sow, sow identification number, birth date, parity, farm, sow line, first insemination date, service sire, and farrowing rate were available. For the sows that farrowed from first insemination, gestation length, farrowing date, litter size (LS), number of live born piglets, number of stillborn piglets, number of mummified piglets, and weaning date were available.
Farrowing rate (FR) was recorded as a binomial trait, with an underlying normal variation. It was defined as 1 if the first insemination resulted in a pregnancy and gestation length was longer than 108 d or if litter size was at least 1; otherwise, FR was defined as 0. Litter size was recorded immediately after farrowing and was defined as the sum of number of live born piglets, number of stillborn piglets, and number of mummified piglets. For sows that did not farrow from first insemination LS was missing. Litter size ranged from 1 to 25. Total number of piglets born per first insemination (TNBF) was defined by multiplying FR with LS. Therefore, TNBF was equal to LS when the sow farrowed from first insemination, and TNBF was defined as 0 when first insemination did not result in parturition. Thus, TNBF ranged from 0 to 25.
Meteorological data were obtained from the European Climate Assessment Data set (Klein Tank et al., 2002) and included daily summaries for the maximum outside temperature for 7 Spanish weather stations and were available for the years 2003, 2004, and 2005. Each farm was assigned to the nearest weather station. Most of the farms had a weather station within 70 km of the farm, with the least separation being 3 km and the greatest separation being 160 km. Freitas et al. (2006) estimated a correlation of 0.9 between on-farm weather data and weather station data even for weather stations more than 300 km away from the farm. Each insemination record in our data set was assigned the daily maximum temperature on day of insemination from the nearest weather station.
Observations with unknown maximum outside temperature at day of insemination were removed, reducing the data set to 38,466 observations. In a preliminary graphical analysis of the data, 10°C seemed to be the lesser critical temperature of the thermo-neutral zone (data not shown). Therefore, observations with a maximum outside temperature below 10°C were removed from the data set. Due to a low number of observations with a maximum outside temperature at day of insemination exceeding 36°C, these observations were also excluded from the data set.
The final data set consisted of 32,631 observations from 11,935 sows on 20 farms with 2,759 D-line sows and 9,176 I-line sows.
Statistical Analysis
The descriptive analysis was performed for the D-line and the I-line separately using the MEANS procedure (SAS Inst. Inc., Cary, NC). For the descriptive analysis maximum outside temperature was divided into 27 temperature classes (10°C, 11°C, ..., 36°C).
To test the effect of maximum outside temperature at day of insemination on reproductive performance, data were analyzed for the D-line and the I-line separately using a 2-step approach. First, data were corrected for systematic effects using the GLM procedure (SAS Inst. Inc.) with the model:
where yijkl is the value of FR, LS, or TNBF; pi is the effect of parity i (14 classes); sj is the effect of service sire j (383 classes); hyk is the effect of herd-year k (56 classes); and eijkl is a random residual term.
Corrected observations (y*) for FR, LS, or TNBF were calculated for each insemination record as
Second, corrected observations y* were used to investigate the effect of temperature on FR, LS, and TNBF. To test whether there was a thermo-neutral zone and a UCT, 2 models were defined: a linear regression model and a plateau-linear model including a thermo-neutral zone (i.e., a plateau). For each sow line both models were tested for goodness of fit using the NLIN procedure (SAS Inst. Inc.). The linear regression model was
where yi* is the corrected observation for FR, LS, or TNBF; i is the intercept; xi is the maximum outside temperature at day of insemination varying from 10 to 36°C; b is the slope of y* when x increases by 1°C (Figure 1); and a random residual term, ei.
|
The first part of the concept estimates the UCT [3] of the thermo-neutral zone of the sow, where c is the constant value of yi* when reproductive performance of the sow is unaffected by temperature; i is the intercept; and b is the slope of the decrease in yi* when yi* was affected by temperature (Figure 2).
|
).
The third part of the concept estimates the decline in reproductive performance when temperature exceeds the UCT [5], where yi* is the corrected observation for FR, LS, or TNBF; i is the intercept; b is the slope of the decrease in yi*; xi is the value of maximum outside temperature at day of insemination; and ei is the random residual term (see also Figure 2). Solutions for [3], [4], and [5] were generated iteratively.
An F-test was used to test per trait per line the fit of the plateau-linear model compared with the linear regression model.
|
RESULTS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
. Mean parity in the D-line (3.4) was slightly greater than in the I-line (3.2). Farrowing rate was equal in both lines, but D-line sows had greater mean LS than I-line sows. This resulted in a mean TNBF that was greater in the D-line (9.0) than in the I-line (8.7). Sows of the D-line were inseminated at days with, on average, a slightly greater maximum outside temperature than sows of the I-line: 22.4°C compared with 21.8°C (Table 1).
|
. For temperatures below 23°C the FR, LS, and TNBF of sows of the D-line was greater than the FR, LS, and TNBF of sows of the I-line. Above 24°C LS was approximately the same in the I-line and the D-line, but FR and TNBF were greater for the I-line when temperature exceeded 25°C (Figure 3).
|
shows the results of the goodness of fit tests for FR, LS, and TNBF when testing no effect of temperature on reproduction versus a linear effect of temperature on reproduction, no effect of temperature on reproduction versus a plateau-linear effect of temperature on reproduction, and a linear effect of temperature on reproduction versus a plateau-linear effect of temperature on reproduction.
|
). This implies that there is a UCT for the reproductive performance of the D-line. The UCT of this zone varied from 19.2°C for FR, to 21.7°C for LS, to 19.6°C for TNBF. Above these temperatures the estimated decrease in FR was 0.01 per °C, 0.05 per °C in LS, and 0.13 per °C in TNBF (Figure 4).
|
). This implies that when temperature increases, LS and TNBF of the I-line decreases. However, the slope of this decrease was small. The estimated decrease in LS was 0.01 per °C, and 0.02 per °C in TNBF (Figure 4). Consequently, sows of the I-line have a greater reproductive performance for outside temperatures above 25°C than sows of the D-line. |
DISCUSSION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
The results in this study indicate some clear differences in heat stress tolerance between 2 sow lines as measured by differences seen in their reproductive performance. Negative influences of high temperature on reproductive performance of sows have been described by several authors (Wettemann et al., 1988; Tummaruk et al., 2004; Suriyasomboon et al., 2006), but none of them studied differences between sow lines. The line differences found in our study are an indication of genetic variation in heat stress tolerance between animals, which gives possibilities for selection on heat stress tolerance. In cattle and sheep, genetic variation in heat stress tolerance has been found (Ravagnolo et al., 2000; Finocchiaro et al., 2005).
Animals are considered to suffer from heat stress when temperature exceeds the UCT of the thermo-neutral zone (Black et al., 1993). Above the UCT of this zone the animal will reduce both production and reproduction to control body temperature (Bianca, 1976). In our study UCT was defined as the temperature above which reproductive performance starts to decrease. For the reproductive performance of sows of the D-line a clear UCT could be estimated. The UCT of D-line sows lies around 20°C. This is numerically less but in line with the UCT of 22°C reported by Black et al. (1993). For I-line sows no UCT for reproductive performance could be estimated. A significant adverse effect of temperature on LS and TNBF of I-line sows was found (0.01 per °C for LS, 0.02 per °C for TNBF). However, the decrease in reproductive performance of I-line sows with increasing outside temperature was less than in D-line sows.
Temperature Effects on Reproductive Performance
It is well known that elevated temperature is a major factor responsible for reduced reproductive performance in livestock during hot seasons in tropical areas (e.g., Prunier et al., 1997; Peltoniemi et al., 1999; Ravagnolo and Misztal, 2002). Heat stress decreases expression of estrous behavior, alters ovarian follicular development, compromises oocyte competence, and inhibits embryonic development (Hansen et al., 2001). Wettemann et al. (1988) found that reduced reproductive performance due to heat stress during d 8 to 16 after mating was related to an altered maternal recognition of pregnancy. When sows are exposed to high temperatures, the sow responds with elevated blood plasma concentrations of cortisol (Einarsson et al., 1996). Tsuma et al. (1995) showed that sows that failed to return to estrus after weaning had greater cortisol concentrations during lactation and after weaning and had decreased LH secretion after weaning than sows that returned to estrus directly after weaning. High temperatures during lactation have been found to decrease feed intake (Prunier et al., 1997). Reduced feed intake seems to restrict LH-release during lactation, which results in restricted follicle growth during lactation and affects follicle development after lactation. Impaired follicle development results in lesser ovulation rate and impaired quality of oocytes and follicular fluid, which may explain increased embryonic mortality. This can result in decreased LS at farrowing (Kemp et al., 2006). In an in-vitro study it was shown that exposure of porcine oocytes to high temperatures had a dramatic effect on the quality and the meiotic competence of oocytes (Barati et al., 2008).
Meteorological Data
Temperature information at day of insemination originated from weather stations throughout Spain. These stations routinely collect daily information on temperature (Klein Tank et al., 2002). On-farm measurements could have reflected the weather conditions on the farm more accurately. However, for genetic evaluation large data sets are required, and therefore, temperature data need to be recorded on every farm. This information is not directly available. As a substitute for on-farm temperature measurements, records from nearby weather stations could be used. Freitas et al. (2006) concluded that records from nearby weather stations and records from weather stations further away from the farm provide satisfactory information for genetic evaluation of heat stress.
Heat stress refers to those meteorological elements that interfere with heat loss from the animal to the environment (Bianca, 1976). The meteorological element considered in our study was maximum temperature at day of insemination, and a significant adverse effect of high temperatures on reproductive performance was found. Other meteorological elements, such as relative humidity, average temperature, and minimum temperature, were ignored. However, in a study by Suriyasomboon et al. (2006), it was shown that not only elevated temperature, but also elevated humidity, had a negative impact on reproductive performance. Therefore, a worthwhile future study would be to investigate the combined effects of temperature and humidity on reproductive performance.
In our study heat stress was modeled linearly and plateau-linearly, and the cumulative effect of heat stress before and after day of insemination was ignored. However, high temperatures at day of insemination were found to be strongly correlated with temperatures in the period 4 wk before and 4 wk after mating (data not shown). Future studies should test which meteorological element has the largest effect on reproductive performance and should concern models in which the dynamics of heat stress can also be analyzed (e.g., cumulative effect of high temperatures).
Implications for Pig Breeding
Genetic potentials of pigs have changed considerably in the past 50 yr, and resulted in increased pig productivity (Brown-Brandl et al., 2004). In pig breeding, selection takes place on the nucleus level in mainly temperate climates and under improved environmental conditions (Knap, 2005). In general, selection on production under these improved conditions has been shown to lead to increased environmental sensitivity (Van der Waaij, 2004). Analogously here, sows of the D-line were selected on reproductive performance based on data collected in a temperate climate (the Netherlands), and these sows showed considerable reduction in reproductive performance with increasing temperatures in Spain. Selection on reproductive performance in the I-line was based on international data from mainly tropical countries (Brazil, Spain, Italy, Philippines), and these sows showed fewer problems with high temperatures. Therefore, even though D-line sows have greater reproductive performance under temperate conditions than I-line sows, sows of the I-line are superior to sows of the D-line when outside temperatures exceed 25°C. This is a clear indication of a genotype x environment interaction. Our interpretation of these differences between D-line and I-line sows is that for line I families were selected, which did well for reproductive performance in these tropical countries, adapting therefore to the local environments.
At temperatures above the thermo-neutral zone gene networks within and across cells respond with a so-called cellular heat stress response. The heat shock transcription factor has been found to be the first responder for activation of this heat stress response (Sonna et al., 2002). This transcription factor coordinates the cellular response to thermal stress and affects expression of a wide variety of genes during heat stress. This expression has been associated with the regulation and the synthesis of heat shock proteins at the cellular level, which protect cells from damage (see review Collier et al., 2008). The variation in heat stress tolerance between animals and the role that heat shock transcription factor plays in response to heat stress suggests that there is an opportunity to improve heat stress tolerance via genetic selection.
In conclusion, the results from this study imply that there are important differences in the relationship between temperature and reproductive performance traits in 2 genetically different sow lines, and these differences suggest that genetic selection on sow heat stress tolerance may be possible.
1 Corresponding author: Saskia.Bloemhof{at}ipg.nl
Received for publication January 11, 2008. Accepted for publication July 23, 2008.
|
LITERATURE CITED |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
