HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pollott, G. E. Articles by Greeff, J. C. Search for Related Content PubMed PubMed Citation Articles by Pollott, G. E. Articles by Greeff, J. C.

Genetic parameters for indicators of host resistance to parasites from weaning to hogget age in Merino sheep¹

G. E. Pollott^*,,2, L. J. E. Karlsson^,, S. Eady^, and J. C. Greeff^,

^* Imperial College London, Department of Agricultural Sciences, Wye Campus, Ashford, Kent TN25 5AH, U.K.; and Great Southern Agricultural Research Institute, Katanning, WA 6317, Australia; and CSIRO McMaster Laboratory, Armidale NSW 2350, Australia; and and Australian Sheep Industry CRC, Armidale NSW 2350, Australia

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

 
Fecal egg count (FEC) has been widely used as an indicator of host resistance to gastrointestinal parasites in sheep and has been shown to be a heritable trait. Two other possible indicators of parasites, dag score (DS; accumulated fecal material) and fecal consistency score (FCS), were investigated in this study, along with BW. All four traits were studied to see how heritability and genetic correlations varied with age from weaning (4 mo) to hogget age (approximately 400 d). More than 1,100 lambs, the offspring of 37 rams, were recorded eight times between weaning (3 to 5 mo of age) and hogget age (13 to 18 mo of age) on two farms. Sire models were fitted to the data from each trait at each recording and in a repeatability model involving the whole data set. Overall, the heritabilities were 0.28 ± 0.072 (FEC), 0.11 ± 0.036 (DS), 0.12 ± 0.036 (FCS), and 0.23 ± 0.070 (BW). By fitting random regression models to the time-series data, it was possible to see how these heritability values varied as the lambs aged, from weaning to hogget age. The heritability of FEC rose from 0.2 at weaning to 0.65 at 400 d. Dag score had a higher heritability (0.25) in the middle of the age range and a low value at weaning (<0.1) and hogget age (0.16). The heritability of FCS was low, with a value of 0.2 at weaning reducing to 0.05 as the animals aged. Body weight had zero heritability at weaning, which rose to greater than 0.6 at hogget age. Most traits had low genetic correlations between them, the only exception being that between FCS and DS (0.63). Most genetic correlations varied little over the age range with the exception of FEC and BW, which fell from 0 at weaning to –0.63 at hogget age. Whereas FCS and DS may be good indicators of scouring, they are very different from FEC as an indicator of host resistance to gastrointestinal parasites.

Key Words: Body Weight • Fecal Egg Count • Gastrointestinal Parasites • Genetic Parameters • Random Regression • Sheep

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

 
The most commonly used indicator of host resistance to parasites in sheep is fecal egg count (FEC). The genetics of this trait have been reported from a number of resource flocks in New Zealand (Morris, 1998) and Australia (Greeff et al., 1995; Woolaston and Piper, 1996; Greeff and Karlsson, 1999) and more recently from industry data in Australia (Pollott and Greeff, 2004a,b).

In long-wool breeds of sheep, fecal contamination of the wool in the breach area is a direct problem by devaluing the wool. It is also an indirect problem as a major predisposing factor to blowfly strike. The accumulated fecal material is referred to as dags, which can be subjectively scored. Dag scores (DS) and fecal consistency scores (FCS) have been investigated as possible indirect indicator traits to breed for decreased worm-related diarrhea or scouring (Greeff and Karlsson, 1997; McEwan et al., 1997). These studies have shown that DS and FCS were heritable at weaning and at hogget age and genetically negatively correlated with FEC (–0.14 between DS and FEC, and –0.06 and –0.25 between FCS and FEC at weaning and at hogget age, respectively; McEwan et al., 1997). This suggests that selection for decreased FEC will result in a moderate increase in DS and/or scouring. In a larger study using a flock unselected for FEC, Greeff and Karlsson (1999) found a small positive genetic relationship of 0.1 between FCS and FEC.

The genetic parameters of FEC in Merino sheep have commonly been estimated at one of three age classes: weaning, postweaning, and yearling (Clarke, 2002; Safari and Fogarty, 2003). The objective of this study was to investigate the genetic parameters of FEC, DS, and FCS and to see how they, along with those of BW, varied with age over the first year of the lamb’s life.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

 
The trial used to provide the data for these analyses, part of the Australian Sustainable Control of Internal Parasites in Sheep program, was designed to determine whether selection for low FEC would result in unfavorable or favorable correlated responses in DS and/or FCS in both a summer and winter rainfall environment. This was achieved using the same sires at both sites by AI. The sires used were randomly selected from two lines of sheep, bred for low FEC, and their equivalent controls, namely the Rylington flock from Western Australia (Greeff et al., 1999), and the Haemonchus (Hc) flock from Chiswick (Woolaston and Piper, 1996).

This trial was conducted in 2000 and 2001 at two sites: one owned by the Department of Agriculture Western Australia at the Mount Barker Research Station (MBRS), and the other by CSIRO at the Chiswick Research Station near Armidale on the northern tablelands of New South Wales.

Each trial site used 400 randomly selected Merino ewes per year typical of their respective regions. The Rylington Merino ewes, used at MBRS, were born in 1997, and the Chiswick ewes were of a mixed age. A total of 37 sires (17 Rylington, 20 Chiswick) were used in the trial; 29 were used at both sites, one at MBRS only, and seven at Chiswick only. Thus, the lambs used as the basis for this study were located at two contrasting sites but were the offspring of the same group of sires.

Semen was collected from all the rams in early 2000. Mating at MBRS was scheduled for the second half of February to fit in with what was considered the optimum lambing time for this region. Mating in Armidale was scheduled for May. All matings at MBRS were carried out using AI, whereas at Chiswick AI was only used for the Rylington rams.

Measurements

Before lambing, the ewes at both sites were sorted into four groups. At lambing, lambs were tagged daily with a unique identification number, and pedigree, single or multiple birth type, BW, and any other notable features were recorded.

At MBRS lamb tagging took place approximately 2 wk after the finish of lambing. The four groups were then amalgamated into one group. At Armidale, the Hc lines were removed from the lambing plots into one management group and the Rylington lines were treated in the same way. The Hc lines and Rylington lines were amalgamated at weaning in the first year and at lamb marking in the second year.

Weaning took place at 3 to 4 mo of age. At weaning, the measurement cycle started with BW, DS, fecal sampling, and FCS. Any other notable features were recorded. The individual fecal samples were then processed for fecal worm egg counts using the modified McMaster method (Whitlock, 1948).

Fecal consistency was scored on a 1 to 5 scale, where 1 indicated hard fecal pellets and 5 indicated watery fluid faeces. Dags were scored visually from 0 for no dags, to 5 for a large amount of dags, similar to Larsen et al. (1994).

In addition to the weaning sampling and data collection, seven more monthly sampling times were made until the termination of the trial. The last sampling for each yearly progeny group was at approximately 400 d of age. Thus, there were eight sets of measurements used in these analyses, repeated on the same group of lambs at each of the eight recording times.

In addition to the four traits (FEC, DS, FCS, and BW), lamb data were collected on the year of birth, site, sire, sex, date of birth, birth type, and date of sampling/weighing.

In this article, the terms weaning, postweaning, yearling, and hogget refer to lambs aged 3 to 5, 5 to 8, 9 to 12, and 13 to 18 mo of age, respectively.

Methods of Analysis

Heritabilities at Each Recording. Because ewes were drawn randomly each year from commercial flocks kept on the two farms, all analyses reported were carried out using a sire mixed model to calculate (co)variance components, using ASREML (Gilmour et al., 2002). Initial analyses were undertaken to investigate the heritability of each trait at each of the eight recordings. Fixed effects of year, site, birth type, age (fitted as a cubic polynomial), and sex were fitted initially to all traits at each of the eight recordings, and the final model included only the significant fixed effects (Model 1) arrived at using a step-down. Additive genetic variances were calculated as four times the sire component of variance from further Model 1 analyses containing the significant fixed effects and the random effect of sire. Phenotypic correlations between the four traits were calculated at each recording. The standard errors of the heritabilities were calculated as described by Gilmour et al. (2002).

Repeat Records. On the assumption that any trait measured at the eight recordings was in fact a repeated measurement of the same trait, further estimates of heritability, genetic and phenotypic correlations were undertaken using all repeated measurements (i.e., from the eight recordings) in the same analysis. Once again, the same fixed effect model was used as described above but a random term for each lamb was included in the sire model (Model 2) and a fixed effect term for recording number, one to eight for weaning and the seven other recording times. In addition, age was fitted as a cubic polynomial. Genetic and phenotypic correlations were calculated from bivariate analyses between each pair of traits in turn. In addition, the repeatability of each trait was estimated as the ratio of the between-lamb plus sire and phenotypic variances. In these analyses, the phenotypic (co)variances were calculated as the sum of the sire, residual, and lamb components of (co)variance. The standard errors of the heritabilities and genetic and phenotypic correlations and repeatabilities were calculated as described by Gilmour et al. (2002).

Random Regression. The repeated measurements of the four traits up to hogget age meant that it was possible to investigate how the sires’ values varied over the recording period and to use the results to estimate how the heritabilities and genetic correlations changed with age of lamb. Further analyses were undertaken using the random regression of sire value on lamb age for each trait. Inspection of the data indicated that both the means and variances of each trait varied considerably between the eight recording times. Previous experience with random regression models of this type (Pollott and Greeff, 2004b) indicated that convergence of the analyses might be difficult under these circumstances. Standardization of the data and the use of a different residual term for each of the eight recording times are methods that may help overcome the problems of using data from heterogeneous variances. In addition, standardization also eliminates differences in mean value between recordings. Model 3 (below) shows the complete model used for all traits in the initial analyses. The actual fixed effects used for any given trait were derived using a step-down approach, eliminating any fixed effect that was not significant (P < 0.05).

The initial terms in Model 3 were as follows: Y_ijklmn = a lamb’s record for one of FEC, DS, FCS, BW; SYR_i = the effect of the ith site/year/recording time in which the lamb was born (2 yr and two sites) and recorded (eight recording times); SX_j = the effect of the jth sex of lamb (1 to 2; males and females); B_k = the effect of the kth birth type (one to three lambs per birth); ID_l = the effect of the lth individual lamb fitted as a random effect; A = age at recording, in days, fitted as a covariate; Sa_m = the effect of the mth sire (1 to 37) with variance ; Sb_m(A) = the regression of the mth sire effect on age of its offspring at recording, with variance Sc_m(A²) = the regression of the mth sire effect on age² of its offspring at recording, with variance ; residuals_ijklmn = eight normally distributed error terms, one for each recording time.

This model attempted to separate the effects of lamb age (A + A² + A³) and the eight different recording times from the data, independently, using a polynomial regression term. This would then leave the sire effect to vary over age and give a good estimate of the way it changed with time, modeled using quadratic polynomial random regression terms. The three covariances between the level and slope of each sire’s regression on age and age² (cov_Sab, cov_Sac, and cov_Sbc) were also fitted in the model.

Standardization of the data was achieved by adjusting each record to produce a mean of zero and a variance of unity for a contemporary group. Site, year, and recording number defined a contemporary group. Hence, each record was standardized by subtracting the appropriate contemporary group mean and dividing by the contemporary group standard deviation (Pollott and Greeff, 2004b). Model 4 (below) was used for analyses involving standardized data. In these analyses, the only fixed effects fitted were those found to be significant, not including site, year and record number, the factors used to standardize the data.

The sire component of variance at a given age (2_S|A) was calculated from the (co)variances of sire and the slope of the regression of each sire on age of its offspring as shown in Model 5 (below; Kolmodin et al., 2002). Age (A) in this case was the age at which the variance component was calculated and for illustrative purposes the mean age of lambs at the eight recordings was used in the results section.

In addition, covariance random regression analyses were undertaken to investigate how the genetic correlations between the four traits varied with age utilizing the methodology outlined by Kolmodin et al. (2002), as used by Pollott and Greeff (2004b). For these analyses, linear random regression terms for the effect of sire on age were fitted using standardized data with Model 4 and the appropriate fixed effects derived above.

The question of whether the repeated measurements, taken over 8 mo, were in fact the same trait, was further addressed by calculating phenotypic and genetic correlations between the same trait measured at the eight recordings. The same fixed-effect models were used as described for Model 1, and all variances and covariances were calculated in a series of bivariate analyses.

	Results

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

 
Trends in the Four Traits with Age

The trends shown in Table 1 clearly reflect the way the immune competence to internal parasites changes with age and anthelmintic treatment. The decrease at the second recording period was the result of the weaning anthelmintic treatment. Dag score was approximately 0.5 at weaning, but it increased with age from approximately 250 d onwards to 1.56 at the eighth recording. Fecal consistency score followed a similar pattern to DS but varied between 2.02 at weaning and 3.24 at 400 d. Body weight reflected a slow growth rate from weaning, which accelerated from approximately 1 yr of age (sixth recording).

Heritability of each Trait at each Recording

The heritabilities of the four traits at each of the eight recordings are shown in Table 2 calculated from the sire model (Model 1). The heritability of FEC varied over the eight recordings, starting low (0.19) at weaning and rising thereafter. The value at the fifth recording stood out as being somewhat different from the general trend, due to a sharp increase in the sire variance. Also, the heritability of the eighth record was somewhat lower than the general trend, once again due to a change in the sire variance and a larger residual variance. Both the sire and residual variances increased with recording number, with the exceptions noted previously. Also the postweaning anthelmintic dose decreased both the residual and sire variances.

Fecal consistency score had a low heritability, with its highest value at weaning (0.22), lower value at intermediate ages (0.03), and an increase at the older ages (0.15). These values did not differ by more than the sum of their standard errors and so may be considered as similar.

The heritability of BW increased with age from effectively zero at weaning to 0.57 at the oldest age. The sire variance increased as the lambs became older, whereas the residual variance for BW was similar for the first six records but increased thereafter.

Overall Heritabilities and Correlations

Table 3 shows the number of records used, mean, standard deviation, and an ANOVA summary for the four traits used in the overall analyses. Heritability and repeatability of each trait were calculated from the complete data set, using the data from each recording as a repeat record. Results from using the original data (Table 4) show that FEC and BW had moderate heritabilities, whereas those of DS and FCS were low. Body weight was a highly repeatable trait, FEC and DS had moderate repeatabilities, and FCS was effectively not repeatable. Results from using standardized data (not shown) were calculated in order to compare them with the random regression results, presented later. They were similar to those using the original data. Both FEC and FCS repeatabilities were similar to their corresponding heritability estimates. This indicates a very low level of permanent environmental or nonadditive effects for these traits.

At the phenotypic level, very few of the traits were correlated with each other (Table 4), the exception being FCS, which was moderately correlated with DS. Dag score and FCS showed a highly positive genetic correlation. Indications from the other traits were that both FEC and FCS were moderately genetically correlated with BW. There seemed to be no relationship between FEC and either FCS or DS. The results of the phenotypic and genetic correlations using the standardized data (not shown) were similar to those calculated from the original data.

Random Regression Estimates of Heritability

Fecal Egg Count. The way the sire and residual variances varied at the eight recordings using the random regression models is shown in Table 5. The heritability derived from fitting the two random regression models to the original and standardized FEC data is shown in Figure 1, along with the individual recording time results from Table 2, for comparison.

View larger version (12K): [in this window] [in a new window]   Figure 1. The heritability of fecal egg count using two different random regression models and heritability calculated at each individual recording time. The SE of the random regression heritabilities were 0.06 at 110 d, 0.07 at 250 d, and 0.13 at 405 d.

Dag Score. The heritability of DS calculated from the original data followed the individual record results rising from about 0.07 at weaning to 0.3 at the fifth record and then back down to <0.1 by 400 d of age (Figure 2). Heritability from the standardized random regression was 0 at weaning, approximately 0.25 in the middle age range, and then fell again at 400 d to 0.18. The quadratic polynomial models with variable residuals clearly had the necessary flexibility to reflect the changes in the variances over time. The additive variance from the standardized data had a similar pattern to the heritability changes with time but the residual variance was lowest in the middle of the range and rose at both the youngest and oldest ages (Table 6).

View larger version (12K): [in this window] [in a new window]   Figure 2. The heritability of dag score using two different random regression models and heritability calculated at each individual recording time. The SE of the random regression heritabilities were 0.04 at 110 d, 0.04 at 250 d, and 0.07 at 405 d.

Fecal Consistency Score. Results from analyzing FCS from both the standardized and original data (Figure 3) indicated a heritability of about 0.18 at weaning decreasing to 0.1 by 300 d of age and increasing again at the oldest age. The pattern of sire variance with age reflected that of the heritability, higher at both extremes and lower in the middle (Table 6). The pattern of residual variances was consistent throughout with a slight drop at the oldest age.

View larger version (12K): [in this window] [in a new window]   Figure 3. The heritability of fecal consistency score using two different random regression models and heritability calculated at each individual recording time. The SE of the random regression heritabilities were 0.06 at 110 d, 0.04 at 250 d, and 0.05 at 405 d.

View larger version (12K): [in this window] [in a new window]   Figure 4. The heritability of body weight using two different random regression models and heritability calculated at each individual recording time. The SE of the random regression heritabilities were 0.04 at 110 d, 0.08 at 250 d, and 0.15 at 405 d.

Correlations between FEC and FCS were low and mostly negative and only significant for Recordings 1, 6, and 8 (Table 7). This is not surprising given the number of interactions involved including different worm species (Haemonchus vs. Trichostrongylus), two very different environments and the potential for both low and high worm egg count scouring.

The FCS and DS correlation was positive, highly significant and generally moderately high. The decline for Recording 8 may be a reflection of DS accumulating over time and FCS staying relatively constant. The relationship between FCS and BW exhibited low correlations, which were only significant for Recordings 4 and 6. The correlation between DS and BW was low and negative.

The genetic and phenotypic correlations between any given traits measured at each of the eight recordings were calculated (results not shown). As with all tables of this type, the closer the two traits are recorded in time, the higher the correlation. Fecal egg count at the different recordings showed high genetic correlations, suggesting that it was the same trait at all times. However, phenotypic correlations were lower and the earlier recordings were not well correlated with later ones at the phenotypic level. Body weights were highly correlated at both the genetic and phenotypic level. This was not surprising because they were not independent measures. Dag scores were poorly correlated at the phenotypic level and moderately correlated at the genotypic level. Fecal consistency scores showed little phenotypic correlation between subsequent records but were highly correlated genetically.

Change in Genetic Correlations with Age

Figure 5 shows the way the genetic correlations involving standardized FCS, FEC, and DS changed as the animals aged, as calculated using the random regression models. The standard errors of all genetic correlations were high, but FEC and DS had a consistent genetic correlation (approximately 0.11) at all ages. The genetic correlation between FEC and FCS fell from 0 at weaning to –0.3 at 400 d, whereas that between FCS and DS was consistently highly positive over the age range shown. The overall genetic correlations shown in Table 4 were in the middle of these ranges.

View larger version (13K): [in this window] [in a new window]   Figure 5. The genetic correlations of fecal egg count (FEC), dag score (DS), and fecal consistency score (FCS) as they change with age, using standardized data. The SE of the correlations between FEC and DS were 0.26 at 110 d, 0.24 at 250 d, and 0.26 at 405 d. The SE of the correlations between FCS and DS were 0.16 at 110 d, 0.12 at 250 d, and 0.26 at 405 d. The SE of the correlations between FEC and FCS were 0.29 at 110 d, 0.25 at 250 d, and 0.24 at 405 d.

View larger version (14K): [in this window] [in a new window]   Figure 6. The genetic correlations of fecal egg count (FEC), dag score (DS), and fecal consistency score (FCS) with BW as they change with age, using standardized data. The SE of the correlations between FEC and BW were 0.30 at 110 d, 0.18 at 250 d, and 0.15 at 405 d. The SE of the correlations between FCS and BW were 0.40 at 110 d, 0.23 at 250 d, and 0.22 at 405 d. The SE of the correlations between BW and DS were 0.29 at 110 d, 0.19 at 250 d, and 0.16 at 405 d.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

Genetics of Fecal Egg Count

The heritability of FEC (0.28 ± 0.07) estimated here was similar to several other reports in the literature (see Clarke, 2002, and Safari and Fogarty, 2003, for a summary) for this trait in resource flocks and also the value of 0.26 ± 0.02 from the industry data reported by Pollott and Greeff (2004a). The estimate from the standardized data was similar to that of the original data but the heritability of FEC from standardized data reported by Pollott and Greeff (2004b) was 0.23 ± 0.02. Using an animal model, Pollott and Greeff (2004a) found values of 0.34 ± 0.03 and 0.32 ± 0.03 for original and standardized data respectively.

The repeatability of a trait gives an indication of the extent to which successive measurements of the trait on the same animals can be considered to be related to each other. The difference between the repeatability and the heritability measures the permanent environmental effect of the animal on the trait and any nonadditive effects. The repeatability of the FEC data was 0.22. This indicates that there is little relationship between successive FEC measurements other than those of genetic origin. Morris et al. (1997a) reported repeatability estimates in Romney sheep of 0.46 and 0.43, giving permanent environmental effects of 0.12 and 0.10, respectively. Similar permanent environmental effect values were reported by Bishop et al. (1996) from Scottish Blackface sheep (0.06, 0.15, and 0.16) from records taken postweaning at 2-mo intervals.

Evidence of low phenotypic relationships between successive FEC measurements but a high level of similar genetic control was found from the phenotypic and genetic correlations between the eight recordings calculated from this data set (not shown). Genetic correlations ranged from 0.55 to 1.11, whereas phenotypic correlations ranged from 0.11 to 0.54. The closer the age at which the correlated records were taken, the higher the correlation; the major exception being correlations involving FEC measurements at weaning, which were less well correlated with adjacent FEC measurements, possibly due to a postweaning stress suppression of the acquired immune response. A similar pattern of results was reported by Bishop et al. (1996).

The heritability of FEC varied with age starting from approximately 0.2 at weaning and rising to over 0.6 at hogget age. The few comparable reports from Merino sheep, summarized by Clarke (2002), show a more confusing pattern than this with no clear trend with age. Greeff et al. (1995) found that the heritability of FEC fell from 0.18 at weaning to 0.08 and then rose again to 0.25 at 12 mo of age. Greeff and Karlsson (1997) reported a similar trend with a value of 0.40 at weaning and 0.22 at hogget age. On the other hand, Greeff and Karlsson (1998) found similar FEC heritabilities at 3 and 15 mo of age. Recent industry data analyses of Merino sheep found the heritability of FEC to be 0.24 at yearling age and 0.38 at hogget age (M. Khurso, Armidale, Australia, unpublished data). Reports from other breeds suggest an increase in heritability of FEC with age (Morris et al., 1997a, 2000), and these reports from Romney sheep in New Zealand demonstrated a somewhat higher value for the heritability of FEC than reported here.

No other reports discuss the use of random regression methods to model the change of FEC heritability with age. The heritability of FEC increased from weaning to hogget age, and in this data set, it rose from 0.2 to about 0.7. There seems to be little permanent environmental effect of the lamb on FEC, but the same genes largely control the trait at different ages.

These analyses have not considered the effect that different species of parasitic worms may have had on the results. There was variation both between and within sites, at different times of the year, in the types of worm present but there was no quantitative data on this aspect of FEC. If different species of worms produce different numbers of eggs per individual parasite, this will lead to variation in the results not considered here. There is evidence from New Zealand that there is a moderate genetic correlation of 0.43 between strongyle (mainly Trichostrongylus and Ostertagia) and Nematodirus FEC (Morris et al., 2004), which indicates that different genes and thus different defense mechanisms may operate against different parasite species. More detailed recording of worm species needs to be undertaken in order to elaborate on this point.

The Genetics of Dag Score

Dag score had a low overall heritability at 0.11. Few comparable estimates appear in the literature for Merino sheep. Larsen et al. (1995) reported repeatabilities of DS in seven groups of Merino ewes from different farms. The repeatability of DS over successive seasons varied from 0.38 to 0.61, but Karlsson and Greeff (1996) reported a value of 0.08 at yearling age. Results from Romneys in New Zealand estimated the heritability of DS at 0.31 (Meyer et al., 1983), 0.50 (Watson et al., 1986), 0.06 (McEwan et al., 1992), 0.24 (Bisset et al., 1992), 0.12 and 0.28 (Bisset et al., 1994), 0.36 and 0.22 (Bisset et al., 1996), and 0.40 (Shaw et al., 1999).

Dag score was a moderately repeatable trait (0.25), and the permanent environmental effect plus nonadditive gene effect was 0.14. The phenotypic correlations between the DS recorded at the eight recording times ranged from 0.09 to 0.65, with the corresponding range of genetic correlations being 0.28 to 1.12.

The heritability of DS varied between weaning and hogget age, with low values at the extremes (0.07 and 0.16) and rising to 0.25 (standardized data) at the fourth recording period. The results from the random regression models using standardized data gave results similar to the individual recording time estimates. Meyer et al. (1983) also made repeated measurements of DS and found a similar pattern and also reported a phenotypic correlation of 0.30 and a genetic correlation of 0.75 between scores within a year. Dag score seems to be a moderately heritable trait, with higher values of heritability at the yearling age than either weaning or hogget age.

The Genetics of Fecal Consistency Score

The overall heritability of FCS in this study was low (0.12). The heritability of FCS in Merinos has been previously reported by Greeff and Karlsson (1997, 1998, 1999) at 0.38, 0.04, and 0.13, respectively, and by Karlsson and Greeff (1996) as 0.23 and 0.17 in October and June, respectively. Other literature results for Romneys in New Zealand quote a heritability of 0.27 (McEwan et al., 1992), 0.11 (Bisset et al., 1994), and 0.05 (Bisset et al., 1996).

The repeatability of FCS was also low (0.12). Further evidence for the low level of the relationship between successive measurements of FCS was found in the phenotypic correlations between all eight recording times. These ranged from 0.08 to 0.23, and there was no clear pattern of closer measurements being more highly correlated. The genetic correlations between repeated measurements of FCS were consistently high, ranging from 0.53 to 1.30 (SE 0.30 and 0.51), indicating that the same genes controlled FCS throughout this recording period.

The pattern of FCS heritability estimates by age of lamb was quite consistent. The individual record estimates of heritability (Table 2) indicated a relatively high level of genetic variability at weaning, which fell at yearling age and then rose again at hogget age. This pattern was reflected in the random regression estimates of heritability with age. Using standardized data, there was a decrease in heritability from weaning (0.2) to yearling age (0.1), and a rise was apparent at hogget age (0.2). However, the values at the extremes rarely differed by more than the sum of their respective standard errors, indicating that there may be little difference in the heritability of FCS across the age range studied. The other reports of repeated FCS measurements were reported by Greeff and Karlsson (1997), who found the same value of heritability at both weaning and hogget age (0.37 and 0.38 respectively), and Karlsson and Greeff (1996), who found values of 0.23 at weaning and 0.17 at yearling ages.

The analysis of diarrhea, or scouring, as measured by DS or FCS is complicated by fact that this trait is made up of two subtypes, namely "high-FEC scouring" (low immune competence) and "low-FEC scouring" (enhanced immune competence). Postweaning high-FEC scouring is the predominant type. At yearling age, low-FEC scouring is the predominant subtype, especially in the winter rainfall environment (Larsen et al., 1994, 1995).

The genetics of FCS, as analyzed in this report, seem to show a low level of genetic control with little real change over the age range studied. Successive measurements of the trait are not well related, although the genes controlling the repeated measurements seem to be similar. However, the small amount of evidence from other reports does little to confirm this picture. There are other reports of FCS having a low heritability but a similar number have estimated higher values, two of which were taken close to weaning.

The Genetics of Body Weight

Body weight is one of the most highly researched traits (see Safari and Fogarty, 2003, for an extensive list of genetic parameters). The overall heritability of BW was 0.23, a result commonly found in the literature.

The repeatability of BW was 0.57 indicating a high level of combined permanent environment effect and nonadditive genetic effects. Phenotypic correlations between the BW records taken at the eight recording times ranged from 0.56 to 0.88 and the corresponding range of genetic correlations was very high at 0.89 to 1.03.

The heritability of BW was effectively zero at weaning but rose to greater than 0.6 at hogget age. This pattern was reflected in the random regression estimates of BW heritability with age. The trend found in other repeat-record studies is also toward a higher heritability as animals become older (e.g., Maniatis and Pollott, 2002), which is reflected here.

Relationships Between the Four Traits

If FEC, DS, and FCS were all similar indictors of the host/parasite interaction, then they should be well correlated. The correlations between FEC, FCS, and DS were all effectively zero, with the exception of FCS and DS, which were genetically highly correlated (0.63) and moderately correlated at the phenotypic level (0.33). These three genetic correlations varied little with age; FEC and DS changed from 0.1 at weaning to –0.05 at hogget age, FEC and FCS ranged from 0.01 at weaning to –0.26 at hogget age, and DS and FCS varied from 0.72 to 0.59 between weaning and hogget age.

There are very few other reports involving these three traits. Baker et al. (1991) found that a low phenotypic (–0.09 and 0.11) and moderate genetic correlation (–0.26 and 0.34) existed between FEC and DS in two Romney sheep flocks. Greeff and Karlsson (1997) found a correlation of –0.06 and –0.25 between FCS and FEC at weaning and hogget age, respectively. Also Greeff et al. (1999) found a genetic correlation of 0.10 at hogget age. Karlsson and Greeff (1996) found FCS and FEC traits to have a negative correlation as yearlings (–0.12) and a positive correlation at weaning (0.25). Watson et al. (1986) stated that FCS was negatively correlated with FEC but gave no figures in their paper. A value of 0.23 was reported by McEwan et al. (1992). The trend of a positive correlation postweaning and a negative correlation at yearling to hogget age fits with the theory of an age-related change from high to low FEC scouring. Overall it seems FCS and FEC show a low negative genetic correlation and little change with age.

Morris et al. (1997b, 2000) found a correlated response in DS to selection for low FEC. Genetic correlations between the two traits reported in the literature were 0.21 (Bisset et al., 1996), 0.1 (Bisset et al., 1994), 0.44 (Bisset et al., 1992), –0.42 with FEC at weaning, 0.24 with FEC at hogget age (Karlsson and Greeff, 1996), and –0.55 (McEwan et al., 1992). These results show little consistency and mostly differ from the results presented in this article.

Dag score and FCS have been found to be highly correlated in all reports, although no other authors have repeatedly dag-scored and calculated its correlation with FCS (Bisset et al. [1996], 0.77; Bisset et al. [1994], 1.17; Karlsson and Greeff [1996], 0.93; and McEwan et al. [1992], 1.37). These two traits seem to be controlled by the same genes, although their phenotypic correlation (0.33) indicates a somewhat different physical expression. These results suggest that FCS could be an indicator trait for DS because wool is normally removed from around the breech area, where the dags accumulate to make the animals less susceptible to blowflies. This practice removes the variation in dags between animals and thus FCS could make a contribution as an alternative indicator trait for DS.

Body weight and FEC were found to be negatively genetically correlated (–0.32), with a range from 0 at weaning to –0.63 at hogget age. This contrasts with the results reported by Pollott and Greeff (2004a) from industry Merino data, which showed a low negative correlation (–0.05 to –0.09) between the two traits. The summary produced by Safari and Fogarty (2003) suggests that estimates lie in the low to moderately negative range with very little change over the age range. The exceptions to this were genetic correlations at weaning (0.09) and yearling age (–0.26) from Bisset et al. (1992) and positive results from McEwan et al. (1992) at 0.37 postweaning, and from Greeff and Karlsson (1998), who reported low positive correlations at a range of ages.

There are few reports of genetic correlations between BW and FCS. In this study, an overall value of 0.29 and a range from 0.29 at weaning to 0.5 at hogget age was found. Values of –0.04 (McEwan et al., 1992) and –0.17 (Bisset et al., 1996) were the only other ones reported in the literature.

Body weight and DS had an overall genetic correlation of 0.11 and ranged from –0.18 at weaning to 0.25 at hogget age. The opposite effect was found by Bisset et al. (1992), with 0.28 at weaning and 0.03 at yearling ages, and with –0.46 (hogget), 0.14 (weaning), –0.18 (yearling), and 0.06 (hogget), as reported by Bisset et al. (1994). A further study by the New Zealand group (Bisset et al., 1996) found genetic correlations of –0.41 (yearling), –0.17 (yearling), and –0.22 (postweaning). One other estimate of –0.12 was found by McEwan et al. (1992).

Using Random Regressions to Estimate Genetic Parameters

This article used a range of methods to investigate the genetics of indicators of host resistance to parasites, and the relationships between them and BW. Individual record analyses suggested that most of the traits had a somewhat erratic pattern of sire and residual variance and heritabilities as the animals aged, which may make fitting random regression models difficult. To overcome this difficulty, two different approaches were used. A polynomial random regression model for the sire component of variance incorporating a separate residual for each recording time was used. In one analysis, the original data were "corrected" for site/year/recording time as a fixed effect. In the other, data were standardized to have a mean of 0 and a variance of 1 within each site/year/recording time contemporary group. The two approaches resulted in largely similar estimates of heritability across the age range for all four traits. The standardized data approach fitted simpler models, which reached solutions more readily when computed. The random regression models produced a similar pattern of heritabilities across the age range to those found at individual recording times, with one notable exception. The heritability of FEC was higher at the older ages using the random regression approach.

	Implications

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

 
This work suggests that fecal consistency and dag score are very similar traits controlled by much the same genes. The pattern of correlations over the age range with other traits, their similar overall genetic correlations, and their heritabilities and repeatabilities all suggest this. Perhaps the only difference is the pattern of heritabilities over the age range, particularly the higher heritability of fecal consistency score at weaning and the higher dag score value at the yearling age, and the slightly higher repeatabilities for dag score. Dag score and fecal consistency score are clearly not the same trait as fecal egg count. Thus, although these traits are generally used as indicators of host resistance to parasites, their poor relationship with fecal egg count implies that they would be ineffective to improve host resistance to parasites. However, selection for these traits in their own right is important to decrease the occurrence of scouring in sheep, where it is a problem.

	Footnotes

 
¹ This research was carried out while the first author was on sabbatical leave from Imperial College London. It was funded by the Sheep Cooperative Research Centre, while the first author was based at the Dept. of Agric. of Western Australia. The data used in these analyses originated from a trial on the Sustainable Control of Internal Parasites, funded by Australian wool growers and the Australian government through Australian Wool Innovation Limited. All institutions are acknowledged for their various contributions to this work.

² Correspondence—phone: +44 (0)20 759 42707; fax: +44 (0)20 759 42919; e-mail: g.pollott{at}imperial.ac.uk

Received for publication April 14, 2004. Accepted for publication July 7, 2004.

	Literature Cited

Top Abstract Introduction Materials and Methods Results Discussion Implications Literature Cited

 


Baker, R. L., T. G. Watson, S. A. Bisset, A. Vlassoff, and P. G. C. Douch. 1991. Breeding sheep in New Zealand for resistance to internal parasites: Research results and commercial application. Pages 19–32 in Breeding for Disease Resistant Sheep. G. D. Gray, and R. R. Woolaston, ed. Australian Wool Corp., Melbourne.

Bishop, S. C., K. Bairden, Q. Q. McKellar, M. Park, and M. J. Stear. 1996. Genetic parameters for fecal egg count following mixed, natural, predominantly Ostertagia cicumcinta infection and relationships with liveweight gain of young lambs. Anim. Sci. 63:423–428.

Bisset, S. A., A. Vlassoff, C. A. Morris, B. R. Southey, R. L. Baker, and A. G. H. Parker. 1992. Heritability of and genetic correlations among fecal egg counts and productivity traits in Romney sheep. N. Z. J. Agric. Res. 35:51–58.

Bisset, S. A., C. A. Morris, D. R. Squire, S. M. Hickey, and M. Wheeler. 1994. Genetics of resilience to nematode parasites in Romney sheep. N. Z. J. Agric. Res. 37:521–534.

Bisset, S. A., C. A. Morris, D. R. Squire, and S. M. Hickey. 1996. Genetics of resilience to nematode parasites in young Romney sheep—Use of weight gain under challenge to assess individual anthelmintic treatment requirements. N. Z. J. Agric. Res. 39:314–323.

Clarke, B. 2002. Review of genetic parameters for Australian Merino sheep. Report to Meat and Livestock Australia. Project No. SHGEN.005. Dept. of Agric. Western Australia, Perth.

Gilmour, A. R., B. J. Gogel, B. R. Cullis, S. J. Welham, and R. Thompson. 2002. ASReml Users Guide Release 1.0 VSN Int., Ltd., Hemel Hempstead, U.K.

Greeff, J. C., and L. J. E. Karlsson. 1997. Genetic relationships between fecal egg count and scouring in Merino sheep. Proc. Assoc. Adv. Anim. Breed. Genet. 12:333–337.

Greeff, J. C., and L. J. E. Karlsson. 1998. The genetic relationship between fecal consistency, fecal worm egg counts and wool traits in Merino sheep. Pages 63–66 in Proc. 6th World Cong. Genet. Appl. Livest. Prod., NSW, Australia.

Greeff, J. C., and L. J. E. Karlsson. 1999. Will selection for decrease fecal worm egg count result in an increase in scouring? Proc. Assoc. Adv. Anim. Breed. Genet. 13:508–511.

Greeff, J. C., L. J. E. Karlsson, and J. F. Harris. 1995. Heritability of fecal worm egg count at different times of the year in a mediterranean environment. Proc. Assoc. Adv. Anim. Breed. Genet. 11:117–121.

Greeff, J. C., L. J. E. Karlsson, and R. B. Besier. 1999. Breeding sheep for resistance for internal parasites. Proc. Assoc. Adv. Anim. Breed. Genet. 13:150–155.

Karlsson, L. J. E., and J. C. Greeff. 1996. Preliminary genetic parameters of fecal worm egg count and scouring traits in Merino sheep selected for low worm egg count in a Mediterranean environment. Proc. Aust. Soc. Anim. Prod. 21:477.

Kolmodin, R., E. Strandberg, P. Madsen, J. Jensen, and H. Jorjani. 2002. Genotype by environment interaction in Nordic dairy cattle studied using reaction norms. Acta Agric. Scand. Sect. A. 52:11–24.

Larsen, J. W. A., N. Anderson, A. L. Vizard, G. A. Anderson, and H. Hoste. 1994. Diarrhoea in Merino ewes during winter: Association with Trichostrongylid larvae. Aust. Vet. J. 71:365–372.[Medline]

Larsen, J. W. A., A. L. Vizard, J. K. Webb Ware, and N. Anderson. 1995. Diarrhoea due to trichostrongylid larvae in Merino sheep: Repeatability and differences between bloodlines. Aust. Vet. J. 72:196–197.[Medline]

Maniatis, N., and G. E. Pollott. 2002. Direct and maternal (co)variances for weight and ultrasonically measured traits of lambs in a closed Suffolk flock. Small Rumin. Res. 45:235–246.

McEwan, J. C., P. Mason, R. L. Baker, J. N. Clarke, S. M. Hickey, and K. Turner. 1992. Effect of selection for productive traits on internal parasite resistance in sheep. Proc. N. Z. Soc. Anim. Prod. 52:53–56.

McEwan, J. C., S. A. Bisset, and C. A. Morris. 1997. The selection of sheep for natural resistance to internal parasites. Pages 161–182 in Sustainable Control of Internal Parasites in Ruminants. G. K. Barrell, ed. Lincoln Univ., Canterbury, New Zealand.

Meyer, H. H., T. G. Harvey, and J. E. Smeaton. 1983. Genetic variation in incidence of daggy sheep – an indicator of genetic resistance to parasites? Proc. N. Z. Soc. Anim. Prod. 43:87–89.

Morris, C. A. 1998. Responses to selection for disease resistance in sheep and cattle in New Zealand and Australia. Proc. 6th World Cong. Genet. Appl. Livest. Prod. Armidale, NSW, Australia. 27:295–302.

Morris, C. A., S. A. Bisset, A. Vlassoff, R. L. Baker, T. G. Watson, and M. Wheeler. 1997a. Yearling and ewe fleece weights in Romney and Perendale flocks selected for divergence in fecal nematode egg count. Proc. Assoc. Adv. Anim. Breed. Genet. 12:50–53.

Morris, C. A., S. A. Bisset, A. Vlassoff, C. I. West, and M. Wheeler. 2004. Genetic parameters for Nematodirus spp. egg counts in Romney lambs in New Zealand. Anim. Sci. 79:33–39.

Morris, C. A., A. Vlassoff, S. A. Bisset, R. L. Baker, C. J. West, and A. P. Hurford. 1997b. Responses of Romney sheep to selection for resistance or susceptibility to nematode infection. Anim. Sci. 64:319–329.

Morris, C. A., A. Vlassoff, S. A. Bisset, R. L. Baker, T. G. Watson, C. J. West, and M. Wheeler. 2000. Continued selection of Romney sheep for resistance or susceptibility to nematode infection: Estimates of direct and correlated responses. Anim. Sci. 70:17–27.

Pollott, G. E., and J. C. Greeff. 2004a. Genetic relationships between fecal egg count and production traits in commercial Merino sheep flocks. Anim. Sci. 79:21–32.

Pollott, G. E., and J. C. Greeff. 2004b. Genotype x environment interactions and genetic parameters for fecal egg count and production traits of Merino sheep kept in a range of environments. J. Anim. Sci. 82:2840–2851.[Abstract/Free Full Text]

Safari, A., and N. M. Fogarty. 2003. Genetic parameters for sheep production traits: Estimates from the literature. Tech. Bull. 49, NSW Agric., Orange, Australia.

Shaw, R. J., C. A. Morris, R. S. Green, M. Wheeler, S. A. Bisset, A. Vlassoff, and P. G. C. Douch. 1999. Genetic and phenotypic relationships among Trichostrongylus colubriformis-specific immunoglobulin E, Trichostrongylus colubriformis antibody, immunoglobilin G₁, fecal egg count and BW traits in grazing Romney lambs. Livest. Prod. Sci. 58:25–32.

Watson, T. G., R. L. Baker, and T. G. Harvey. 1986. Genetic variation in resistance or tolerance to internal nematode parasites in strains of sheep at Rotomahana. Proc. N. Z. Soc. Anim. Prod. 46:23–26.

Whitlock, H. V. 1948. Some modifications of the McMaster helminth egg-counting technique and apparatus. J. Council Sci. Ind. Res. 21:177–180.

Woolaston, R. R., and L. R. Piper. 1996. Selecting Merino sheep for resistance to Haemonchus contortus: Genetic variation. Anim. Prod. 62:451–460.

This article has been cited by other articles:

				C. A. Kerr, K. L. Bunter, R. Seymour, B. Shen, and A. Reverter The heritability of the expression of two stress-regulated gene fragments in pigs J Anim Sci, August 1, 2005; 83(8): 1753 - 1765. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Pollott, G. E. Articles by Greeff, J. C. Search for Related Content PubMed PubMed Citation Articles by Pollott, G. E. Articles by Greeff, J. C.