| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ANIMAL GENETICS |
* Institute of Genetics, Vetsuisse Faculty, University of Berne, 3001 Berne, Switzerland;
and
Institute of Animal Sciences, Federal Institute of Technology and Vetsuisse Faculty, University of Zurich, 8092 Zurich, Switzerland;
and
SUISAG, 6204 Sempach, Switzerland; and
The Seeing Eye Inc., Morristown, NJ 07963-0375
 |
Abstract |
|---|
 Top Abstract INTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
Key Words: cryptorchidism • dog • pig • sex ratio
|
INTRODUCTION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
). The present state of knowledge suggests that cryptorchidism is caused by the interaction of genetic, epigenetic, and environmental factors (Amann and Veeramachaneni, 2006). Although a whole range of genes have been implicated in the regulation of testicular descent (Ivell and Hartung, 2003; Klonisch et al., 2004; Yoshida et al., 2005), in humans, no single gene is mutated in more than 10% of cryptorchids (Amann and Veeramachaneni, 2007). The role of environmental factors affecting testicular descent is not yet clear, because only a few conclusive studies are available (Thonneau et al., 2003).
Prevalences of cryptorchidism in domestic animals have a wide range depending on species and breed (Amann and Veeramachaneni, 2007). Compared with the human situation, the genetic components of cryptorchidism in domestic animals still remain unclear (Amann and Veeramachaneni, 2007). In animal breeding, the occurrence of cryptorchidism leads to economic loss and decreased selection potential of male breeding stock. Breeding organizations are well aware of this problem, but as long as the genetic background is not better known, no great effort is made to eliminate the occurrence of cryptorchidism.
Investigating cryptorchidism in pig and dog, we observed that the sex ratio in litters with cryptorchids seemed to be unbalanced in favor of male progeny. Because this phenomenon has not yet been addressed in the literature, we decided to investigate whether the occurrence of cryptorchids in litters of dogs and pigs significantly affects the sex ratio in these litters.
|
MATERIALS AND METHODS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
Dog Data
All dog breeds entering this study were routinely screened for cryptorchidism. The Seeing Eye Inc. provided the data on Golden Retrievers (GR), Labrador Retrievers (LR), and German Shepherds (GS). These dogs were bred in closed colonies where each colony formed a large family. Puppies were raised in homes by volunteers. Diagnoses were established by the veterinarian of the company when these puppies returned from puppy raiser homes to begin training. In most cases of the diagnoses (72%), dogs were aged between 1 and 2 yr. In 11% of the diagnoses, they were younger, the youngest being 34 d of age, and in 17% of the diagnoses they were older, the oldest being 643 d of age. The data on the Swiss White Shepherds (WS) were collected in the field. Cryptorchidism was diagnosed by 1 veterinarian when puppies were 8 wk of age. Dogs only entered the study as cryptorchids if the diagnosis could be confirmed at the age of 52 wk.
Only litters with at least 1 live male and 1 live female puppy entered the analyses. Because there were very few litters with more than 2 cryptorchids, they were excluded from the analyses. A total of 1,339 litters, of which 12.8% had 1 cryptorchid and 3.1% had 2 cryptorchids, entered the analyses. The data set comprised information on 639 litters in GS, 563 litters in LR, 86 litters in GR, and 51 litters in WS (Table 1).
|
Field data for the analyses in pigs covering the years from 2000 to 2006 were provided by SUISAG, a Swiss company for services in pig production, for populations that were either purebred or crossbred. Data are recorded and reported to SUISAG by the breeders. Cryptorchids are diagnosed by the breeders during the suckling period. The breeds involved were Large White (L), Swiss Landrace (S), Hampshire, Large White Sire line (W), and Duroc (D), also a Sire line. Before the analyses, the data were edited to remove erroneous or incomplete records. Only records from farms that had at least 30 litters were considered, which resulted in 279 farms with an average of 869.1 litters, ranging from 30 to 3,027. Only litters with at least 2 live male and 2 live female piglets entered the analyses. Because there were few litters with more than 2 cryptorchids, they were excluded from the analyses. Litters with more than 17 live and stillborn piglets were excluded from the analyses, as were litters with more than 4 developed stillborn or more than 3 undeveloped stillborn piglets. This data editing aimed at minimizing management effects, effects caused by maternal problems, and effects due to incorrect reporting of the sex of the offspring. A total of 119,920 litters, of which 2.2% had 1 cryptorchid and 0.2% had 2 cryptorchids, entered the analyses. The data set comprised information on 83,034 litters in L, 5,627 litters in S, 3,048 litters in W, 6,106 litters in D x L, 485 litters in D x S, 749 litters in D x (S x L), 12,715 litters in W x L, 511 litters in W x S, 1,655 litters in L x (S x L), 864 litters in L x W, 4,379 litters in L x S, 154 litters in (W x L) x L, 337 litters in Hampshire x L, and 256 litters in S x (S x L) (Table 1
).
Statistical Analyses
All statistical analyses were carried out using the SAS 9.1.3 software package (SAS Institute Inc., Cary, NC). Before comparing the sex ratio of litters with cryptorchids with litters with no cryptorchids, we wanted to see whether the presence of cryptorchids affected litter size or the stillborn rate, or both, because in multiparous species, stillborn progeny are not unusual, and a detailed diagnosis of these animals is not common practice.
The main target trait in the analyses was defined in 2 ways, as the sex difference in a litter (SEXD) and as the sex ratio in a litter (SEXR). The trait SEXD was the number of live males minus the number of live females in a litter, and SEXR was the number of live males divided by the number of live offspring in a litter. A significance level of 5% was chosen. When multiple tests were performed, Bonferroni corrections with the STEPDOWN option were applied to adjust P-values for multiple testing and to remove some of the conservativeness of the Bonferroni procedure (Holm, 1979). Other target traits were the number of live offspring in a litter at birth (LIVE) and the stillborn rate (STILL), which is the number of stillborn offspring divided by the total number of offspring in a litter at birth. The target traits LIVE, STILL, SEXD, and SEXR were analyzed using the GLIMMIX procedure to be able to accommodate nonnormal response distributions where appropriate. Because the information available for dog and pig was not the same, the models differed in some of the independent variables. For LIVE and SEXD, the response distribution was chosen to be normal, because the distribution of both LIVE and SEXD, was roughly normal in our data. For STILL and SEXR, we used the events-trials syntax, in which case the GLIMMIX procedure automatically selects the binomial distribution as the response distribution.
Dog. The first question addressed was whether there were significant differences in LIVE between litters with and litters without cryptorchids. The independent variables in the model were CRYP, BREED, YEAR, PARITY, and SEASON. The variable CRYP was 0, 1, or 2 if there were no, 1, or 2 cryptorchids in the litter. The variable BREED was the dog breed, and YEAR was the year of birth, ranging from 1978 to 2005. The variable PARITY was the parity of the dam, ranging from 1 to 8. Finally, SEASON was set to 1 if puppies were born November through March and set to 2 if puppies were born April through October.
The second question, whether CRYP increases STILL, was addressed using the same model. The third question, whether CRYP affects SEXD and SEXR, was investigated using the same procedure with the only difference that besides the fixed effects CRYP, BREED, YEAR, PARITY, and SEASON, STILL and STILL squared entered as covariates.
Pig. For the analyses of LIVE, STILL, SEXR, and SEXD, the fixed effects CRYP, BREED, YEAR, and SEASON, defined as above, entered the model. For the analyses of SEXR and SEXD, STILL and STILL squared were added as covariates in the model. Farm was added as a random effect in the analyses of LIVE and SEXD. In the analyses of STILL and SEXR, the RANDOM INTERCEPT statement was used to add a random intercept for each farm.
|
RESULTS AND DISCUSSION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
Dog.
In dog, litter size was not affected by the presence of cryptorchids in a litter (P = 0.45; Table 2). The effects of BREED and PARITY were highly significant (P < 0.01). The average litter size in GS was 6.2, in LR 6.5, in WS 6.6, and in GR 8.0. At first parity, the average litter size was 7.1. From the second parity, in which the average litter size was greatest with 7.4, the average number of live puppies gradually decreased to parity 8 and greater, with an average of 5.9 live puppies. The variable YEAR had a significant (P = 0.01), but not systematic, effect on litter size. The variable SEASON had no effect on litter size (P = 0.83).
|
). The effects of BREED and YEAR were highly significant (P < 0.01). In the 14 pig populations (BREED), the average litter size ranged from 10.9 to 11.6. The average litter size gradually increased from the year 2000 with 11.0 to the year 2006 with 11.6, which is the result of selection for number of live born piglets. In the populations investigated, the number of live born piglets makes part of the aggregate breeding value for reproduction, which is weighted more than 50% in the overall breeding values for non-Sire lines (http://www.suisag.ch). The difference between the 2 seasons was 0.03 live piglets in favor of winter, but the difference was not significant (P = 0.06). The extent of between-farm effects corresponded to 5.2% of the residual variance. Stillborn Rate
Dog.
In dog, STILL was not affected by the presence of cryptorchids in a litter (P = 0.69; Table 2). Also, BREED had no effect on STILL (P = 0.25). The effect of PARITY (P = 0.04) was significant, but not systematic, although STILL seems to increase in parity 7 and greater. The effect of YEAR was highly significant (P < 0.01). Up to 1992, STILL was around 10% and often greater, then gradually decreased, and since 1997 has fluctuated around 5%. Although the effect of SEASON was not significant (P = 0.17), STILL in winter (8.9%) seems to be somewhat greater than STILL in summer (7.9%).
Pig.
In pig, STILL was 0.5% greater in litters with 1 cryptorchid than in litters with no cryptorchids (P < 0.01; Table 2). The comparisons of 0 vs. 2 (P = 0.69) and 1 vs. 2 (P = 0.91) cryptorchids in a litter were not significant. The effects of BREED, YEAR, and SEASON all were highly significant (P < 0.01). In the 14 pig populations, STILL ranged from 4.7 to 6.4%. The STILL gradually increased from 5.1% in 2000 to 6.0% in 2006, which could be due to an increase in litter size during this period. This relationship has also been observed in other studies (Leenhouwers et al., 2003; Arango et al., 2005). The STILL in winter (5.7%) was greater than in summer (5.4%). The extent of the between-farm effects was close to zero.
Sex Difference
Dog.
In dog, the effects of PARITY (P = 0.30), YEAR (P = 0.34), SEASON (P = 0.94), STILL (P = 0.59), and STILL squared (P = 0.63) on SEXD were not significant (Table 2). The effect of BREED was significant (P = 0.03), but only the difference in SEXD between LR and GS was significant (P = 0.03), with LR having 0.39 more male puppies than GS. The effect of CRYP on SEXD was highly significant (P < 0.01). The SEXD in litters with 1 cryptorchid increased by 0.93 in favor of male puppies compared with litters without cryptorchids (P < 0.01; Table 3). Litters with 2 cryptorchids had 1.50 male puppies more than litters without cryptorchids (P < 0.01). The increase of 0.56 male puppies in the SEXD from litters with 1 cryptorchid to litters with 2 cryptorchids was not significant (P = 0.17).
|
). The partial regression coefficients of the covariates STILL (0.0068) and STILL squared (–0.0003) on SEXD were both significant (P < 0.01 and P < 0.01, respectively). In the 14 pig populations, the excess of males ranged from 0.33 to 1.24 (P = 0.02). Significant differences could be observed for D x S, in which the excess of males (0.33) was smaller than in D x L (0.90; P = 0.02), L (0.87; P = 0.03), W x L (0.87; P = 0.04), S (0.88; P = 0.04), and S x (S x L) (1.24; P = 0.01). The differences in the comparison of D x S to the remaining 8 populations were not significant, probably due either to the small differences or to the small sample sizes. Although significant (P < 0.01), the seasonal effect in favor of male piglets was very small with 0.08 more piglets in winter than in summer. The effect of CRYP on SEXD was highly significant (P < 0.01). The SEXD in litters with 1 cryptorchid increased by 0.69 in favor of male piglets compared with litters without cryptorchids (P < 0.01; Table 3). Litters with 2 cryptorchids had 1.06 more male piglets than litters without cryptorchids (P < 0.01). The increase of 0.37 male piglets in the SEXD from litters with 1 cryptorchid to litters with 2 cryptorchids was not significant (P = 0.08). The extent of the between-farm effects corresponded to 1.3% of the residual variance. Sex Ratio
Dog.
In dog, the effects of PARITY (P = 0.28), YEAR (P = 0.38), SEASON (P = 0.99), STILL (P = 0.61), and STILL squared (P = 0.63) on SEXR were not significant (Table 2). The effect of BREED was significant (P = 0.03), but only the difference in SEXR between LR and GS was significant (P = 0.03), with LR having 2.9% more male puppies in excess than GS. The effect of CRYP on SEXR was highly significant (P < 0.01). The SEXR in litters with 1 cryptorchid increased by 7.1% in favor of male puppies compared with litters without cryptorchids (P < 0.01, Table 3). Litters with 2 cryptorchids had 10.9% more male puppies than litters without cryptorchids (P < 0.01). The increase of 3.8% male puppies in the SEXD from litters with 1 cryptorchid to litters with 2 cryptorchids was not significant (P = 0.20).
Pig.
In pig, only YEAR did not affect the SEXR (P = 0.35); all other effects were significant (Table 2). The partial regression coefficients of the covariates STILL (0.0015) and STILL squared (–0.0001) on SEXR were both significant (P < 0.01 and P < 0.01, respectively). In the 14 pig populations, the SEXR ranged from 52.0 to 55.6% (P < 0.01). A significant difference (P = 0.03) could only be observed for D x S, in which the percentage of males in the litter was 3.6% smaller than in S x (S x L). Although significant (P < 0.01), the difference in the seasonal effect in favor of male piglets was small, with 1.4% more piglets in winter than in summer. The effect of CRYP on SEXR was highly significant (P < 0.01). The SEXR in litters with 1 cryptorchid increased by 3.2% in favor of male piglets compared with litters without cryptorchids (P < 0.01; Table 3). Litters with 2 cryptorchids had 4.6% more male piglets than litters without cryptorchids (P < 0.01). The increase of 1.4% male piglets in the SEXR from litters with 1 cryptorchid to litters with 2 cryptorchids was not significant (P = 0.11). The extent of the between-farm effects was close to zero.
The analyses of the SEXD and the SEXR in dog and pig led to very similar results. The observations in dog and pig clearly show that the occurrence of cryptorchids in a litter goes together with a shift in the sex ratio-difference, in favor of males. Although not significant, the results suggest that an increasing number of cryptorchids in a litter goes together with an increasing shift of the sex ratio-difference in favor of males, comparable to a dose effect. The obvious explanation would be that the chance of observing a cryptorchid in a litter increases with an increasing number of males in a litter. This hypothesis was tested in 2 ways. First, litters in dog, and pig as well, were pooled according to the number of live males in a litter. This led in dog to 8 classes with the number of live males per litter ranging from 1 to 8 and in pig to 14 classes with the number of live males ranging from 2 to 15. Regression analyses of the fraction of cryptorchids in a litter over the classes (data not shown) showed in dog a regression coefficient of –0.00696 (P = 0.02) and in pig of –0.00029 (P < 0.01). There is no evidence that litters with more live males have a greater fraction of cryptorchids than litters with fewer live males. Second, for pig and dog each, 20 replicate data sets were generated that were identical to the original data sets with the exception that CRYP was simulated. The simulation was done by setting CRYP to zero for all litters and then randomly assigning cryptorchids to litters so that each replicate had the same number of litters with 1 or 2 cryptorchids as the original data sets. The simulated data sets were analyzed for the target trait SEXR using the same models for pig and dog that were used for the original data sets. The results (Table 4) clearly show that the association between SEXR and CRYP is not a statistical artifact. In each, pig and dog, in only 1 out of the 20 replicates, the effect of CRYP was significant (pig P = 0.03, dog P = 0.02), although much smaller than in the original data sets. The estimates of the effects other than CRYP turned out to be very similar in the original and the simulated data sets with the exception of BREED in dog, which probably is due to the considerable differences in the prevalences for CRYP in the original data set.
|
investigated cryptorchidism and intersexuality with respect to litter size and SEXR in various domestic species. Although his data editing and statistical analyses do not allow for a direct comparison, some of his results are of interest to our study. Comparing 1,529 litters of GS sires that were known not to be carriers for cryptorchidism with 3,684 litters of GS sires that were known carriers for cryptorchidism, he found no significant difference in the sex ratio. Looking at the litters of the sires known to be carriers only, he found in the 404 litters with cryptorchids an excess of 5.1% males (P < 0.01) compared to the 3,280 litters without cryptorchids. Comparing the litter sizes of the carrier sires and the noncarrier sires, he found that on average, carrier sires had smaller litters than noncarrier sires. He then speculated that there must be a gene leading to cryptorchidism in males and embryonic death in females. In our data, we found no evidence that litter size was negatively affected by the presence of cryptorchids in a litter. Therefore, we are not able to comment on his speculation on female embryonic death, because our data did not allow us to assign litters to carriers or noncarriers. He also reanalyzed published data on 362 piglets and found that litters with cryptorchids had on average 4.3% males in excess compared to litters without cryptorchidism. Our results confirm the observations of Brandsch (1964) that litters with cryptorchids have on average more males than litters without cryptorchids in both pig and dog. We do not claim that the occurrence of cryptorchids in litter is leading to the observed shift in the sex ratio, but it seems reasonable to assume that a common cause influences both, the occurrence of cryptorchidism in a litter and the SEXR in such litters. An overlap in the genetic background of the 2 traits seems to be more likely than a causative environmental factor, because there is a considerable difference in the prevalence of cryptorchidism in GS compared to LR and GR (Table 1), although they all shared a common environment.
|
Footnotes |
|---|
2 Corresponding author: dolf.gaudenz{at}itz.unibe.ch
Received for publication September 25, 2007. Accepted for publication May 23, 2008.
|
LITERATURE CITED |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTS AND DISCUSSIONLITERATURE CITED |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
