J. Anim. Sci. 2003. 81:2915-2922
© 2003 American Society of Animal Science
Bioeconomic evaluation of sow longevity and profitability1
S. L. Rodriguez-Zas*,2,
B. R. Southey*,
R. V. Knox*,
J. F. Connor
,
J. F. Lowe and
B. J. Roskamp
* Department of Animal Sciences, University of Illinois at Urbana-Champaign, Urbana 61801; and
and
Carthage Veterinary Services Ltd., Carthage, IL 62321
 |
Abstract
|
|---|
Sow production indicators, including litter size, litter weight, and the length of time that sows remained in the herd (sow longevity), were used to characterize sow performance and profitability. Sow longevity and production records from 148,568 sows in 32 commercial herds from Central Illinois from January 1995 to May 2001 were analyzed using survival and repeatability models, respectively. The factors studied included sow genetics (32 genetic lines), with eight major lines present in multiple herds, and the combination of herd and year of entry in the herd. The largest difference in longevity between the major genetic lines was approximately one parity. There were differences (P < 0.05) in the instantaneous sow removal rate or hazard from the major lines. These differences constitute evidence that sow longevity could be improved by using replacements from specific genetic lines. The net present value per sow (present value of future cash flows and the present value of the sow) was used to evaluate the effect of sow longevity and production traits on economic returns. Assuming a zero discount rate per parity, genetic lines with longer herd life resulted in greater profit than genetic lines with shorter herd life. This difference was reduced with increasing discount rates and was reversed with high discount rates and low net income per litter. These results suggest that the magnitude of the economic improvement attained through the use of sow genetic lines with longer longevity depends on the economic context under which the evaluation is made.
Key Words: Analysis • Economics • Genetics • Sows • Survival
|
Introduction
|
|---|
The swine industry has a hierarchical structure, with many pork producers purchasing out-of-herd replacement gilts. This investment constitutes a major budget decision and ideally, the replacement gilts should be maintained until this investment is recuperated. In practice, 40 to 50% of the sows are removed before three to four parities (D’Allaire et al., 1987
; Boyle et al., 1998), an age when most replacements recover their initial cost (Stalder et al., 2003). Incentives to increase sow longevity include larger litters with heavier pigs in older parities, fewer unproductive days, acquired immunity to herd diseases, higher sow salvage value, and lower replacements costs (D’Allaire et al., 1987; Lucia et al., 2000).
Sow longevity is an example of survival or time-to-event data (Allison, 1997), with time typically being the number of days that a sow remains in the herd and event being the removal from the herd (e.g., culling or death). These data are always positive, exhibit a non-normal distribution, and can be censored (Allison, 1997). Censored records occur when the event of interest is not observed during the period studied. For example, sows that are present at the end of the period studied are censored since they have not been culled yet. These records should not be removed from the analysis, as is commonly done (e.g., Holder et al., 1995; Kirkwood et al., 2000), since they contain information on longevity. Linear mixed model with censoring (Guo et al., 2001) and survival analysis (Brandt et al., 1999; Yazdi et al., 2000) have been used to investigate swine longevity. The objectives of this study were to examine records from commercial U.S. herds for sow longevity using survival analysis, production traits, and combine the results in economic terms.
|
Materials and Methods
|
|---|
Data
Records of 32 herds from Central Illinois in the U.S. Midwest were obtained from January 1995 to May 2001. A total of 148,568 sows that farrowed 2.2 times per year on average, with an average of 10.4 live piglets per litter during this period were analyzed. The average replacement rate, culling rate, and death rate were 59.8, 41.6, and 9.7%, respectively. The information available for each sow and parity included dates of service (regardless of success), farrowing, and weaning, date of removal, and genetic line.
A total of 32 genetic lines was originally identified and grouped based on breed composition to ensure adequate representation across multiple herds. Specific genetic lines that were rarely used (less than 50 sows) and of ambiguous coding were grouped into herd-specific categories. Eight genetic lines were found in multiple herds and were designed as major genetic lines. Although not all major genetic lines were present in all herds, all herds were connected directly by at least one major genetic line or indirectly with an immediate herd that shared a major genetic line with the two herds. The major genetic lines were Camborough 15, Camborough 22, Camborough Blue, Camborough unspecified, Large White x Hampshire cross, Large White x Landrace cross, Landrace, and an unspecified multiple breed cross.
Two indicators of sow longevity, herd life and productive days, were calculated. Herd life was defined as the total number of days from first service (regardless of success) until removal from the herd. Productive days were defined as the total number of days the sow gestated and lactated until removal from the herd. Approximately 1% of all lactation records were edited based on observed ranges of gestation and lactation lengths to provide a minimal gestation length of 105 d, a maximal gestation length of 124 d, or a maximal lactation length of 30 d. The number of observations and measures of longevity by herd size and major genetic line are summarized in Tables 1 and 2. The measurements of sow productivity studied were the total number of piglets born, the number of piglets born alive, litter size at weaning, total weight of litter at birth, and total weight of litter at weaning.
View this table:
[in this window]
[in a new window]
|
Table 1. Number of herds, sows, genetic lines, mean and standard deviation of herd life, parities and productive days by herd size
|
|
View this table:
[in this window]
[in a new window]
|
Table 2. Number of herds, sows, genetic lines, mean and standard deviation of herd life, parities, and productive days (PD) by the major sow genetic lines
|
|
Models and Methods
Survival Analysis.
Longevity is an example of time-to-event data, with time being the number of days (or other time unit) that a sow remains in the herd and the event being the removal (e.g., culling or death) from the herd. Longevity data is not normally distributed with values ranging from 0 to infinity (exclude negative values) and usually exhibits censoring. Censoring occurs when data from sows not exhibiting the event before the end of the period is available. Survival analysis is the most appropriate method to analyze this type of data (Allison, 1997). Longevity was characterized with a hazard function that represents the instantaneous removal rate for a sow remaining in the herd to a particular time point (Allison, 1997). The hazard function for individual i at time t, hi(t), can be described using a linear fixed effects proportional hazards model with k explanatory variables (xi1, ..., xik):
 | [1] |
where ln represents the natural logarithm, h0(t) denotes an unspecified baseline hazard function, and ß1, ..., ßk are the regression coefficients associated with the k explanatory variables. The baseline hazard function is an arbitrary function common to all observations.
Complementary survival analysis models were considered. The semiparametric Cox proportional hazards (Cox, 1952) and the Weibull models were used to describe the hazard function using the PHREG and LIFEREG procedures of SAS (SAS Inst., Inc., Cary, NC), respectively. In the Weibull model, time of removal is assumed to follow a Weibull distribution with scale parameter 
. The natural logarithm of the baseline hazard function was modeled as ln(t)/(-1) (Allison, 1997). The hazard is accelerated by a constant factor ( < 1), degraded by a constant factor ( > 1) or constant ( = 1) in the LIFEREG parameterization. The Cox model is more flexible because it does not require the specification of the baseline hazard function (Allison, 1997).
The likelihood-based Schwarz’s criterion (Schwarz, 1978) was used to remove nonsignificant explanatory variables and interactions from the model. This criterion was selected over criteria such as likelihood ratio tests since Schwarz’s criterion accounts for the number of parameters and observations simultaneously. The explanatory variables included in the final model were herd-year of entry (32 herd levels and seven year levels) and genetic line (32 levels). Estimates are provided in terms of hazard ratios that denote the relative risk that a sow will be culled due to changes in the explanatory variable once all other factors in the model have been adjusted for.
Repeated Measures Analysis.
Measurements of sow productivity including total number born, born alive and weaned across parities, total birth and weaning weight are examples of repeated measures since the characteristic is measured repeatedly on the same sow across time. Repeated models assuming unstructured variance-covariance structure were considered to account for the correlation among measurements from the same sow. The repeated measurement model used was:
 | [2] |
where yij is the trait (e.g., total born alive) recorded on sow i at parity j and eij is the residual or remainder not explained by the model and the rest of the terms are as defined before. Nonsignificant explanatory variables and interactions were removed from the model using Schwarz’s criterion (Schwarz, 1978). The final explanatory variables included in the model were herd-year of entry, genetic line, month of farrowing (12 levels), and parity (1 to 10 parities). The analysis was conducted using the MIXED procedure of SAS (SAS Inst., Inc.).
Economic Analysis.
The comparison of genetic lines requires the simultaneous consideration of longevity and performance since lines differ in longevity and performance ranking. Economic indicators that reflected the economic impact of longevity were expressed in terms of net present value dollars. The net present value combines the length and amount of investment, the time necessary for an investment to be profitable, and the cost adjusted by a discount (a combination of inflation and interest rates and risk) rate per parity. The net present value per sow was computed using the median longevity of the genetic line of interest, different discount rates, and net income per litter values. The median longevity was converted into number of parities based on the values from Table 2
. The net income value per litter, defined as the difference between revenue and costs (feeding, health, and others), was computed using the formula and prices of Lacy et al. (2000) and Stalder et al. (2003). Each parity was assumed to have the same net income (either $10 or $50 per parity sow), and this income was prorated for partial parities. All genetic lines had the same net gilt replacement cost ($250) and salvage cost ($150).
|
Results and Discussion
|
|---|
Survival Analysis
Herd and year of entry effects were combined into a contemporary group or herd-year effect based on the unique herd performances across years. This specification is not only consistent with the behavior of the data but also with other studies of sow performance (Tummaruk et al., 1998). This factor had a significant (P < 0.0001) influence on sow longevity but did not have a clear trend across the period studied. The average predicted herd life adjusted for genetic line differences ranged between 500 to 600 d from 1995 to 1999 but dropped to 400 d by 2001 (Figure 1). The lower herd life observed in the last 2 years of entry may be due to the high percentage of censored observations since actual herd life has not been recorded for these sows. The maximal predicted herd life after adjustment for genetic line ranged between 700 and 900 d, whereas minimal predicted herd life after adjustment for genetic line ranged between 150 and 300 d (Figure 1). Considerable variability was observed between herds and year of entry, although the difference between most herd and year of entry combinations was approximately one parity. Most of the variation was due to extreme values from few herd-year combinations since the average longevity was similar within herd size and year of entry group. The variation in longevity is likely to be primarily related to management and economic factors. Change in genetic line composition of a herd is another likely factor since early-parity sows would be removed earlier than expected. However, the interaction between year of entry or herd and genetic line did not significantly differ (P < 0.10).
View larger version (14K):
[in this window]
[in a new window]
|
Figure 1. Mean, maximum, and minimum predicted herd life (number of days from first service until removal from the herd) by year of entry across all herd.
|
|
The estimated hazard ratios for herd life corresponding to the major genetic lines, relative to the Camborough 15, are provided in Table 3. The Camborough 15 was selected as the common reference because it was the most frequent genetic line and was well represented across herd-year groups. Sows from the Camborough Blue genetic line had a 22% lower probability of being removed than did sows from the Camborough 15 genetic line. In contrast, sows from the Large White x Hampshire genetic line had a 20% greater chance of being removed than sows from the Camborough 15 genetic line. The Camborough Blue, Landrace, Camborough 22, and Camborough unspecified genetic lines formed a group that was significantly less likely (P < 0.05) to be culled than the group formed from by the Multiple cross (unspecified), Large White x Landrace, and Large White x Hampshire genetic lines (P < 0.05), whereas the Camborough 15 was intermediate between these groups. Sows from the genetic lines in the first group are more likely to be kept in the herd than the other lines. Xue et al. (1997) also reported a significant effect of genetic line on longevity.
View this table:
[in this window]
[in a new window]
|
Table 3. Estimated hazard ratios for herd life and productive days for the main genetic lines relative to the Camborough 15 line from the Cox and Weibull survival models
|
|
Figure 2 depicts the survival curves representing the probability that a sow will remain in the herd until a certain parity for selected genetic lines. Each point of the survival curves represents the probability that an average sow will remain in the herd for each parity. For example, a Camborough 15 sow has an 89% probability of remaining in the herd for one parity, and this probability drops to 32% and 3% by the fourth and eighth parities, respectively. The Large White x Hampshire genetic lines provided the worst herd life since over 50% of sows would be removed by end of the third parity. In contrast, four parities would be necessary to remove 50% of sows from the other genetic lines.
View larger version (20K):
[in this window]
[in a new window]
|
Figure 2. Estimated survival functions of herd life (number of days from first service until removal from the herd) for Camborough 15, Camborough 22, Landrace, and Large White x Hampshire cross genetic lines.
|
|
The estimated hazard ratios for productive days corresponding to each genetic line, relative to Camborough 15, are summarized in Table 3. Sows from the Camborough 22 and Camborough Blue genetic lines were less likely to be removed from the herd (16% and 14%, respectively) than were sows from the Camborough 15 genetic line. The Camborough unspecified genetic line was very similar to Camborough 15 in productive days, and the Landrace line was significantly different from Camborough 15. The Large White x Landrace, Large White x Hampshire, and Multiple cross (unspecified) lines had significantly greater hazard than Camborough 15 indicating sows from these lines were more likely to be culled with fewer productive days than sows from the Camborough 15 genetic line.
The probability of a sow reaching 600 d of herd life (Table 4) ranged from 29 to 48%, indicating that more than half the sows are expected to be removed by four parities. More sows were expected to reach 520 productive days than 600 d of herd life (Table 4). The magnitude of the differences among genetic lines is evident from the differences in average median herd life and productive days (Table 4). The average median time was adjusted for influences of herd, year of entry, and the differences in the number of sows within each line. The most extreme genetic lines differed in herd life by 158 d, or approximately one parity. This result suggests that sow longevity could be improved by replacing a low-longevity genetic line with another genetic line with higher expected longevity. The maximal length of the 95% confidence intervals was 57 and 93 d for herd life and productive days, respectively, indicating little variation in the differences between genetic lines.
View this table:
[in this window]
[in a new window]
|
Table 4. Probability that a sow reaches a herd life of 600 d or 520 productive days (four parities approximately) and median and 95% confidence interval for longevity and productive days for the major genetic lines
|
|
Assuming an average of 2.35 parities per year, and that four parities are required to recuperate the investment cost, a sow must remain in the herd for approximately 600 d. The probabilities of reaching this age ranged between 0.31 and 0.48 (Table 4). This range suggests that most sows are likely to be culled before recuperating the investment cost. Considering two 1,000-sow herds using the highest and lowest lines for herd life in Table 4, it is expected that 170 more sows from the highest line for herd life than the lowest line for herd life are expected to recuperate their initial purchase costs. The longer a sow must remain in a herd to reach a set return on investment, the larger the difference among lines. The shorter the time a sow is required to remain in a herd, the lesser the impact of genetic line on profitability. These results exemplify the potential impact of genetic line decisions on sow profitability.
Repeated Measure Analysis
The total number of piglets born and the number of piglets born alive were similar among genetic lines (Table 5). Although genetic line was significant, the difference in total born between the most extreme genetic lines was 0.46 of a piglet after adjusting for herd-year, parity, and month. The difference between total born and total born alive indicated that approximately 1.5 piglets were born dead. The difference between the best and worst genetic lines for litter size at weaning was 0.64, and on average, one to two piglets were lost from birth to weaning. The differences between genetic lines for litter performance traits indicated that the lines differed in maternal ability. The lines with the highest litter size at birth had the lowest litter size at weaning. This association may be biased by the different cross-fostering practices that may occur in the different herds.
View this table:
[in this window]
[in a new window]
|
Table 5. Litter size at birth for total born and born alive, litter size at weaning, and total weight of litter at birth and weaning for the major genetic lines
|
|
The genetic lines had very similar litter weights at birth and at weaning when unadjusted for litter size (Table 5). The genetic lines with larger litters at birth tended to have higher weights, but this difference was not present at weaning. This discrepancy may be due to cross-fostering or other management practices. The Camborough Blue genetic line had the greatest change in litter weight from birth to weaning due to its low ranking for the number of piglets weaned. Apart from this genetic line, there appears to be little difference in litter size or litter weight between genetic lines when analyzed as traits of the sow. The differences between genetic lines, though relatively small compared with the overall litter sizes, were significant. This may be due in part to the large number of observations analyzed. Consequently, the choice of genetic line based on litter performance traits is likely to have less impact on economic returns than choices based on longevity traits.
Litter size at birth and litter size at weaning generally increased from the first to the fourth parity and decreased in subsequent parities (Table 6). Hughes (1997) and Tummaruk et al. (1998) also reported a significant effect of parity on litter size at birth. A similar trend was also observed in the litter weight at birth and weaning (Table 6). Comparison of litter size and weight shows that after the seventh parity, performance is significantly worse than with first-parity sows. Parity three and four sows had the best performance, and parity seven and older sows tend to have the worst performance. The difference between the first parity and parities three and four was 0.7 piglets born alive and 3 kg of litter weight at birth. The difference at weaning was 0.2 piglets and 4 kg of litter weight. These results are consistent with Koketsu and Dial (1997), who reported significant parity effects on litter weight at weaning.
View this table:
[in this window]
[in a new window]
|
Table 6. Litter size at birth for total born and born alive, litter size at weaning, and total weight of litter at birth and weaning for each parity
|
|
Economic Analysis
For a net income of $50 per parity, 2.0 and 3.68 parities are required to recuperate the initial investment (replacement) cost of the gilt under zero and 10% discount rates, respectively, whereas 10 parities are required to recuperate the investment cost if the net income is $10 per litter and the discount rate is zero. However, it is virtually impossible to recuperate the investment cost for any discount rate greater than 4% with a net income of $10 per litter only by keeping the sow in the herd. Consequently, maximizing the net income per litter is important, not only to the overall profitability but also to maximize the benefit of longevity.
The net present value estimates per sow for different discount rates per parity and net income per litter equal to $50 and $10 are given in Tables 7 and 8, respectively. Assuming a discount rate of zero, the difference between genetic lines was reduced to the difference in longevity multiplied by the net income per litter. Considering a $50 net income per litter, the difference between the best and worst longevity lines was $52.39, and the difference between the best two longevity lines was $13.94. Considering a $10 net income per litter, the difference between the best and worst longevity lines was $10.48, and the difference between the best two longevity lines was $2.79. This reflects the economic advantage of considering sow longevity in culling practices, in that more income will be generated from sows remaining in the herd for a longer period. Under the scenarios considered, higher profits can be obtained by increasing sow longevity through careful selection of the genetic line of the replacements.
View this table:
[in this window]
[in a new window]
|
Table 7. Net present value of the income from one sow assuming an income of $50 per parity (high return market) for each genetic line and different discount rates per parity
|
|
View this table:
[in this window]
[in a new window]
|
Table 8. Net present value of the income from one sow assuming an income of $10 per parity (low return market) for each genetic line and different discount rates per parity
|
|
The relative economic value of the genetic lines varies with the discount rate. The relative value of the high-longevity line decreased with increasing interest rates since the value of income decreases with increasing time. The difference between the best and worst lines in terms of longevity was $16.37, and the difference between the best two lines was $0.20, considering a $50 net income per litter level and a discount rate of 10% per parity. In contrast, the difference between the best and worst lines in terms of longevity was -$12.37 and the difference between the best two lines was -$7.41, considering a $10 net income per litter level and a discount rate of 10% per parity.
These calculations illustrate the need to consider the time frame in order to account for changes in monetary value due to factors such as inflation. The net present value analysis allowed for the simultaneous consideration of longevity and productivity indicators to maximize the future economic returns. The estimates presented in Tables 7 and 8 show that the evaluation of the genetic lines must also include income per litter as well as the discount rate. As expected, the higher interest rates reduced the value of genetic lines with high longevity. The reduction depended on the net income per litter since the high-longevity genetic lines tended to maintain their superiority over the low lines in the high income per litter scenarios.
The net present value calculations ignored the influence of parity on performance indicators. This simplification is not a major concern in the present study since there is little difference in litter size at weaning between genetic lines at the parity corresponding to the median longevity. A more detailed net present value analysis would account for differences in the income and costs associated with the changes in litter size across parities. The influence of parity and genetic line on litter size had a smaller economic impact than the choice of interest rate and income per parity used. Stalder et al. (2003) also noted that segregated weaning price was the major determinate in their example operation, followed by the number of piglets born alive. At high discount rates, the differences between genetic lines in this study were reduced such that the combination of returns and production differences tended to overcome any economic difference due to differential sow longevity. However, at high net income per litter, the higher longevity lines remained profitable at the discount rates considered. This result indicates that herds with a prevalence of sows with high longevity would be exposed to less economic risk than herds with low longevity.
The calculations presented in this study apply to the evaluation of systems on a per-sow basis. The corresponding calculations at a herd level require the consideration of total operating costs over a representative time period to be considered. For example, for a time horizon of 12 parities, and assuming no generation overlap, the economic computations for sows from a genetic line with an average longevity of four parities should be considered three times vs. four times for sows from a genetic line with an average longevity of three parities. Under no discount rate and using the previous assumptions, the difference between genetic lines was the extra replacement cost of $250. This difference decreases with increasing discount rates, such that at a 10% interest rate per parity, the difference was $134.17 regardless of the assumed net income per litter.
|
Implications
|
|---|
Early removal of sows from the herd due to mortality, health problems, and low production is a major bottleneck in the swine industry that leads to animal welfare and economic concerns due to replacement and veterinary treatment costs. This study provided an understanding of the factors influencing sow longevity and expressed the comparative results in economic terms. Important differences in genetic lines were observed that could be translated in economic benefits, provided that sows remained in the herd for period sufficient to recover the initial investment costs. The benefit of sow longevity was reduced when considering the discount rate used to compute the net present value. The resulting bio-economical characterization of sow longevity can be used to identify bottlenecks and improve the productivity, economic efficiency, and well-being of sow breeding herds.
|
Footnotes
|
|---|
1 This project was funded by the National Pork Board, Des Moines, IA. 
2 Correspondence: 307 Animal Sciences Laboratory, 1207 W. Gregory Dr. (phone: 217-233-8810; fax: 217-333-8286; E-mail: rodrgzzs{at}uiuc.edu).
Received for publication March 21, 2003.
Accepted for publication July 25, 2003.
|
Literature Cited
|
|---|
Allison, P. D. 1997. Survival Analysis Using the SAS System: A Practical Guide. SAS Inst., Inc., Cary, NC.
Boyle, L., F. C. Leonard, B. Lynch, and P. Brophy. 1998. Sow culling patterns and sow welfare. Ir. Vet. J. 51:354–357.
Brandt, H., N. von Brevern, and P. Glodek. 1999. Factors affecting survival rate of crossbred sows in weaner production. Livest. Prod. Sci. 57:127–135.
Cox, D. R. 1952. Regression models and life tables (with discussion). J. R. Stat. Soc. B 34:187–220.
D’Allaire, S., T. E. Stein, and A. D. Leman. 1987. Culling patterns in selected Minnesota swine breeding herds. Can. J. Vet. Res. 51:506–512.[Medline]
Guo, S.-F., D. Gianola, R. Rekaya, and T. Short. 2001. Bayesian analysis of lifetime performance and prolificacy in Landrace sows using a linear mixed model with censoring. Livest. Prod. Sci. 72:243–252.
Holder R. B, W. R. Lamberson, R. O. Bates, and T. J. Safranski. 1995. Lifetime productivity in gilts previously selected for decreased age at puberty. Anim. Sci. 61:115–121.
Hughes, P. E. 1997. Effects of parity, season and boar contact on the reproductive performance of weaned sows. Theriogenology 47:1445–1461.
Kirkwood R. N., F. X. Aherne, and P. G. Monaghan. 2000. Breeding gilts at natural or a hormone-induced estrus: effects on performance over four parities. Swine Health Prod. 8:177–180.
Koketsu, Y., and G. D. Dial. 1997. Factors influencing the postweaning reproductive performance of sows on commercial farms. Theriogenology 47:1445–1461.
Lacy, R. C., K. J. Stalder, T. L. Cross, and G. E. Conatser. 2002. Breeding herd replacement female evaluator. Available: http://www.agriculture.utk.edu/ansci/swine/gilt_replacement.htm. Accessed Oct. 15, 2002.
Lucia, T., G. D. Dial, and W. E. Marsh. 2000. Lifetime reproductive performance in female pigs having distinct reasons for removal. Livest. Prod. Sci. 63:213–222.
Schwarz, G. 1978. Estimating the dimension of a model. Ann. Stat. 6:461–464.
Stalder, K. J., R. C. Lacy, T. L. Cross, and G. E. Conatser. 2003. Financial impact of average parity of culled females in a breed-to-wean swine operation using replacement gilt net present value analysis. J. Swine. Health Prod. 11:69–74.
Tummaruk, P., N. Lundeheim, S. Einarsson, and A.-M. Dalin, 1998. Reproductive performance of purebred Hampshire sows in Sweden. Livest. Prod. Sci. 54:151–157.
Xue, J. L., G. D. Dial, W. E. Marsh, and T. Lucia. 1997. Association between lactation length and sow reproductive performance and longevity. J. Am. Vet. Med. Ass. 210:935–938.[Medline]
Yazdi, M. H., L. Rydhmer, E. Ringmar-Cederberg, N. Lundeheim, and K. Johansson. 2000. Genetic study of longevity in Swedish Landrace sows. Livest. Prod. Sci. 63:255–264.
This article has been cited by other articles:
|
|
|
|
 
L. Engblom, N. Lundeheim, E. Strandberg, M. del P. Schneider, A.-M. Dalin, and K. Andersson
Factors affecting length of productive life in Swedish commercial sows
J Anim Sci,
February 1, 2008;
86(2):
432 - 441.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. L. Rodriguez-Zas, C. B. Davis, P. N. Ellinger, G. D. Schnitkey, N. M. Romine, J. F. Connor, R. V. Knox, and B. R. Southey
Impact of biological and economic variables on optimal parity for replacement in swine breed-to-wean herds
J Anim Sci,
September 1, 2006;
84(9):
2555 - 2565.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. Piles, J. P. Sanchez, J. Orengo, O. Rafel, J. Ramon, and M. Baselga
Crossbreeding parameter estimation for functional longevity in rabbits using survival analysis methodology
J Anim Sci,
January 1, 2006;
84(1):
58 - 62.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
J. Arango, I. Misztal, S. Tsuruta, M. Culbertson, and W. Herring
Study of codes of disposal at different parities of Large White sows using a linear censored model
J Anim Sci,
September 1, 2005;
83(9):
2052 - 2057.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
