Monitoring the effect of scrapie control policies that use genetics in different countries

Six scrapie flocks (A - F) were monitored intensively after the use of resistant breeding rams, respectively during 6 years (farms B-E) and two others during 4 years (farm A and F). On the basis of an increase in presence of the ARR/ARR genotype and the ARR allele during the monitoring period, it is expected that the susceptibility for scrapie at the flock level is decreased for the five farms A, B, C, D and E. For farm F the genotype and allele distribution did not change much up until 2004, as resistant rams were not used consistently, so no genetic improvement was achieved here with respect to the scrapie susceptibility. For this reason, monitoring of farm F was ceased at the end of 2004 when the Dutch research project ended. At that moment the monitoring of farm A was ceased too, as on that farm the number of scrapie cases was too low. The total number of scrapie cases per farm varied from 1 to 31 in the period of 6 years. The prevalence on farms A - E decreased and no cases were found during the sampling moments of the last 3 years, i.e. in the tested animals of the yearly tonsil biopsy plus in animals sold in the same week (and subsequently bought by the project). For farm F no decrease in prevalence was observed. On farm A no new scrapie cases were detected at all from the start of the project. For the 4 farms B - E also the incidence rate decreased during the monitoring period, but this was not the case for farm F. During the last 3 years of monitoring no new cases were detected on farm B, D and E, and one each year on farm C. No scrapie was ever detected in birth cohorts from 2001 onwards on any of the farms, i.e. in animals born after the start of the control programme in 2000.

Six scrapie flocks (A - F) were monitored intensively after the use of resistant breeding rams, respectively during 6 years (farms B-E) and two others during 4 years (farm A and F). The following conclusions can be drawn from the modelling analysis of the monitoring data: Our results show for the first time that selective breeding can be effective in combating scrapie in a commercial-flock setting. Our analysis supports the conclusion that in these sheep flocks the reproduction number drops significantly below one, such that scrapie transmission is brought under control, within 1 to 7 years of selective breeding. The time delay until control is achieved depends on the initial transmission levels within the flock and on the initial genotype profile. -To reach the situation where R0<1 the number of non-ARR/ARR animals does not have to be reduced to zero in the flock. However, the allowed frequencies of animals of slightly susceptible to highly susceptible genotypes are likely to differ both between breeds and, due to differences in farm management, between farms. The clues pertaining to this issue as obtained from this study will be useful input for future studies aimed at estimating the expected breeding effort required in a national breeding programme to control the spread of scrapie. Such studies will consist of modelling analyses of forthcoming genotyping and test data on flocks found to be scrapie affected in the Dutch active surveillance (i.e. rapid testing) programme. - As a result of the control programme not only the reproduction number, but also the infection pressure (or force of infection) in the field will decrease in time. However, due to the long incubation time of a scrapie infection a delay of a few years is expected between the reduction in R0 and in infection pressure. This is in line with the detection of new scrapie cases during the years after the start of the control programme (animals born before the start of the programme). The implemented acceleration of the control programme where scrapie-susceptible ewes are removed on the basis of their genotype, will lead to a positive effect on reduction of both R0 as well as the infection pressure.

A monitoring programme was implemented during 4 years in an experimental Sardinian flock of P6 naturally scrapie affected (classical strain). From the initial affected flock called "progenitor group", 2 lines called "resistant" and "susceptible" groups (RG and SG) have been procreated during 5 cohorts or generations using respectively only ARR/ARR rams or ARQ/ARQ rams. Moreover, the 2 groups have been bred together, being in contact specially at lambing. The occurrence of both clinical (animals with neurological signs) and sub-clinical scrapie (fallen stock and culled ewes) of the females and all the male off-springs (slaughtered at 24 months old) were monitored with appropriate Western Blotting and immunohistochemistry on nervous and lymphoid tissues. The sheep of the 5 "resistant" cohorts, RG1 to RG5, were composed only of at least heterozygous ARR (ARR/ARQ or ARR/AHQ) or homozygous ARR (ARR/ARR) sheep. The sheep of the 5 "susceptible" cohorts, SG1 to SG5, were composed mainly of homozygous ARQ (ARQ/ARQ) or partly of heterozygous ARR (ARR/ARQ) sheep. All scrapie-affected sheep carried the susceptible ARQ/ARQ or ARQ/AHQ ("progenitor group") genotypes, and belonged either to the "progenitor group" (mainly at the beginning of the experiment in 2001), or to the "susceptible cohorts", SG1 to SG5, (specially in 2005). Thus, in this experiment, it appeared that the ARR allele was strongly associated with resistance to (classical) scrapie, even for heterozygous ARR sheep. These results showed that selective ARR breeding can be an efficient tool to eradicate scrapie in an affected flock, even if the presence of scrapie agent was kept in the flock during the lambing period due to the susceptible groups ("progenitor group" and SG1 to SG5).

Two groups of 25 Manech red- faced flocks have been selected on the basis of the following criteria: - willingness of the breeders to participate to a longitudinal monitoring during several years - to be in milk recording to get relevant data regarding pedigrees (dams and sires), dates of birth and mode of rearing, dates of lambing, dates and causes of culling, lactations results, dates of purchasing (if bought from another flock). - group I: composed of 15 flocks in official milk recording with an average flock size of 350 ewes (in 2003), the year of the index case between 1988 and 1999 according to the flock, a full (and often long) scrapie history and a high scrapie incidence, a first year of using ARR/ARR rams between 1997 and 2000. - group II: composed of 10 flocks, 5 in official milk recording and 5 in simplified milk recording, with an average flock size of 320 ewes (in 2003), the year of the index case between 1996 and 2001 according to the flock, a short scrapie history and a medium to high scrapie incidence, a first year of using ARR/ARR rams between 2000 and 2002. Accounting for the scrapie status given by the new French TSE legislation in 2003 and the type of milk recording, the groups I and II are splited in 2 sub-groups, respectively group I a (7 flocks) and I b (8 flocks), and group II a (5 flocks) and group II c (5 flocks). The monitoring of these 25 flocks of Manech red faced breed in milk recording has been started before the "scrapiefreesheep" contract, which represented the last 4 years of monitoring (2003-2006) of a longitudinal survey beginning about 10 years ago and already funded by previous contracts as European "sheeprion" one between 1999 and 2003. Thus, before the use of ARR/ARR rams, the scrapie monitoring concerned several birth cohorts (5 to 9) for the flocks of the groups I a and I b, and only a few ones (0 to 2) for the flocks of the groups II a and II c. On the other hand, 2 to 4 birth cohorts may be analysed for the 2 groups of flocks, after the use of ARR/ARR rams, given the end of monitoring in July 2007 and the constraint that the monitoring length of a given birth cohort until July 2007 is over 5 years: A total of 76 cohorts born after using ARR/ARR rams were available in these 25 flocks representing 5570 sheep. Blood samples of all the present sheep were collected. Then DNA was purified by Labogena from blood leukocytes using the alkaline lysis method. PrP genotyping at the 3 codons 136, 154 and 171 was performed using a PCR method which did not distinguish alleles H and Q at the codon 171, so that only 4 alleles were described at the 3 codons: ARR, AHQ, ARQ and VRQ. They are 2 key years for each scrapie affected flock involved in the monitoring, first the year of the index case (or first outbreak), second the first year of using ARR/ARR rams: according to the flock, the year of the index case occurred between 1988 and 2001, and the first year of using ARR/ARR rams between 1997 and 2002. Since the scrapie history was very different from one flock to another, we decided to express the results in relative years of birth of the sheep (or relative birth cohorts) within a flock, the year 0 corresponding to the year of the event of interest, i.e. the year of the index case or the first year of using RR rams. As expected, compared to the initial PrP genotype frequencies of the Manech red faced breed (3% homozygous ARR/ARR, 28% heterozygous ARR/ARQ or ARR/AHQ and 64% susceptible sheep as ARQ/ARQ, ARQ/AHQ, AHQ/AHQ, ARQ/VRQ and so on), the PrP genotype structures of the ewes born after using ARR/ARR rams were clearly more resistant : respectively about 93%, 89%, 82%, 85% of ewes at least heterozygous ARR in flocks of groups I a, I b, II a, II c, compared to 31% for the initial structure. Such an efficient result was the consequence of increasing artificial insemination with ARR/ARR rams, when the breeders decided to improve the scrapie resistance of their flock. Since the breeders of the groups I have been using AI resistant rams for a longer time that the breeders of the groups II, as a consequence the cohorts born using ARR/ARR rams appeared as more resistant in the groups I than groups II. The favourable evolution of the PrP genotype structures, for the ewes born after using ARR/ARR rams, resulted in practice from a continuous improvement: after 2 or 3 years of using ARR/ARR rams (relative birth cohorts 1 or 2), most of all the ewes born in these 25 flocks were at least heterozygous ARR. But it was only a trend: when looking carefully at all the results, cohort-by-cohort, it appeared that 83 % of the cohorts presented at least 80% ewes at least heterozygous ARR. In other words, 13 cohorts (17%) had less than 80% ewes at least heterozygous ARR, respectively 8, 4 and 1 cohorts between 76, for relative birth cohorts 0, 1 and 2, i.e. mainly the first year of using ARR/ARR rams.

Risk of negative effects of ARR/ARR sheep on production or functional traits (including disease resistance) and called (PF) traits may come from 2 genetic causes: - the PrP gene itself could have a pleiotropic effect on other production or functional (PF) traits, with a negative direct effect of the ARR allele. Such a pleiotropic effect, if it exists, should be stable across sheep breeds or populations. The evaluation of this risk is based on association analysis, i.e. on direct comparison, at the breed or population level, of different PrP genotypes for any (PF) trait of interest. When evaluation those associations, a crucial point is to make sure there is no confusion between the putative PrP effect and a family effect, particularly when evaluations are operated on groups of animals selected on PrP: indeed, at the beginning of ARR selection (when the ARR allele is not fixed), a transitory unfavourable association between the PrP gene and the genetic merit for other selected (PF) traits may occur, since as a trend many of the ARR/ARR candidates are kept while only the non ARR sheep showing a high genetic value for the PF traits are selected. Thus such a bias has to be removed when estimating the possible pleiotropic effects of PrP. - the PrP gene may be genetically linked to another polymorphic gene controlling the genetic variability of a PF trait (PF gene). Most often, the both loci (PrP and PF) will not be in linkage disequilibrium, meaning that, from one family to another one, the ARR allele may be on the same chromosome that a favourable or unfavourable allele of the PF gene. Thus the evaluation of this risk has to be based on within family evaluation of the effect on PF of the PrP allele received by its progeny from a parent heterozygous for the PrP gene, i.e. it corresponds to a linkage analysis. Alternatively, any evidence effect on a PF trait of a genetic marker located near the PrP locus on the chromosome 13 is an indication of possible co-selection of the ARR allele with a favourable or unfavourable PF allele. Such a situation introduces a need for the monitoring of this co-selection in the breeding schemes. - The pleiotropic or direct effect of the PrP gene on the production and functional traits has been analysed, both by the partners P1, P5 and P6 of this project and also by other colleagues as summarised in the chapter 4 of the final report, for a large number of meat, hardy and dairy sheep breeds, and also for synthetic lines. The traits analysed were numerous: reproduction (male and female, as fertility, litter size or ovulation rate), growth and carcass traits (ultrasonic fat depth, ultrasonic muscle depth&), dairy traits (milk yield, fat and protein contents, machine milking ability), wool and disease resistance (mastitis via somatic cell count, salmonella resistance, nematodes resistance&). The global conclusion was the absence of direct or pleiotropic effect of the PrP gene on the other production or functional traits. The contrast between non PrP genotyped sheep and some genotyped PrP class of sheep were significant in a few cases, the most relevant explanation being a selective PrP genotyping of the individuals according to a high breeding value for the considered production traits. The only trait showing a significant effect of the PrP genotype was the litter size, but the results were inconsistent according to the studies, ARR allele being the more or the less prolific one. Thus the conclusion was the absence of a pleiotropic effect of PrP on litter size, the significant results being probably caused by sampling bias not removed in the analysis. - Regarding linkage analysis, a few quantitative trait loci (QTL) were found on the chromosome 13 in sheep in the 11 scans available in the literature, including the 3 analysis performed by the partners of this project. The same trend was described in cattle. In sheep, evidence for QTL affecting wool quality, somatic cell count (mastitis resistance) or nematodes resistance were found, respectively at 25cM, 37cM and 34cM from the PrP gene. Thus it is rather far from the PrP gene, so that the within family association between the PrP locus and the QTL were probably weak. To conclude, none of the most important traits as milk production, growth and meat traits, wool or litter size, was pointed out as genetically determined by a locus located near the PrP gene. Therefore no linkage effect was expected between the PrP gene and other genes controlling the important traits remembered above.

To determine the most efficient breeding strategies at a regional or country level for reducing the prevalence of scrapie in slaughter sheep and the flock prevalence of scrapie affected, Partner 7 developed a simulation model. The model was designed as a dynamic Monte Carlo simulation model. Year was used as a time step because of the strong seasonality of the simulated sheep production, as it is the case in Norway. Demographic parameters representing the Norvegian sheep population were used as input to the model applied to Norvegian Rogaland county representing 2705 flocks with a total of 174,000 adult ewes, these county being one the most important sheep counties in Norway and where of the Norvegian classical scrapie cases have been detected. The model was running with a median prevalence of 1%. Each breeding strategy was compared with the scenario where no intervention to control scrapie was performed. Three breeding scenario were compared: - scenario 1: genotyping of all rams and use of ARR/ARR rams only. - scenario 2: genotyping of all rams and preferably use of ARR/ARR rams, thereafter ARR/AXX rams (where AXX=ARQ or AHQ) depending on availability in the flock. If the supply of these rams within flocks was not sufficient, ARR/ARR rams were purchased on average every second year. - scenario 3: genotyping of all rams and culling of XXX/VRQ rams (where WWW=AHQ, ARQ, ARR and VRQ). If the supply of rams within the flock after culling of theses VRQ carrier rams was not sufficient, the flock was supplied ARQ/ARQ rams (the most common genotype in most breeds). The results for each scenario (breeding strategy) were compared with those of the reference with no intervention and they were the following: - in the simulated region, the number of animals delivered for slaughter was approximately 240,000 lambs (males and females) and 40,000 ewes. Without any breeding, the median number of infected flocks was approximately 1% of the total number of flocks. The mean percentage of scrapie infected animals delivered for slaughter was 0.013% and 0.07% of the slaughtered lambs and adult ewes respectively. - in scenario 1, in which only use of ARR/ARR-rams were allowed, the median number of infected flocks was reduced to zero after 9 years. There was an immediate reduction in the number of infected lambs delivered for slaughter while the reduction in the number of infected slaughtered ewes followed the reduction in infected flocks. - in scenario 2, in which ARR/AXX-rams were allowed with a preference for ARR/ARR-rams, the median number of infected flocks was reduced to zero after 17 years. The reduction in the number of infected slaughtered animals follows the reduction in the number of infected flocks. - in scenario 3, in which all rams except XXX/VRQ were allowed, there was a decrease in the number of infected flocks as well as the number of slaughtered infected animals during time. However, there were still infected flocks left in the population after 40 years. The conclusions of these regional or country approaches were finally in good agreement with the results of the contract considering the situation at the flock level when it is infected. Indeed the main conclusions at the regional level presently simulated were the following: - scenario 1: the fact that the fastest reduction was achieved by using ARR/ARR-rams only is in agreement with the current knowledge of scrapie. The placenta of infected ewes is considered as one of the most important sources of the resistant prion protein (PrPSc), and its infectivity depends on the genotype of both the ewe and the progeny. If the progeny is heterozygote for ARR, infectivity has not been detected in the placenta. As sheep heterozygote for ARR have a low susceptibility for scrapie, the introduction of ARR/ARR-rams only will give an immediate reduction in the transmission of scrapie within the flock. That it takes approximately 9 years before the population is free of the scrapie-infection indicates that after the introduction of ARR/ARR-rams very few new infected animals are produced and the time reflects the time it takes before all infected animals are slaughtered. This is supported by the immediate reduction of infected lambs delivered to slaughter. - scenario 2: the availability of rams within the flock was considered when selecting rams for breeding. By limiting the availability of ARR/ARR-rams the eradication of scrapie was delayed with approximately ten years compared to scenario 1. In breeds where the supply of ARRARR-rams is even less than simulated here, the eradication might take even longer then estimated in this model. - scenario 3 where only XXX/VRQ-rams are excluded, scrapie is not eradicated within the 40 years that were simulated. This clearly underlines the importance of implementing a breeding programme based on selection of the most resistant allele and not only removing the most susceptible allele.

Six scrapie flocks (A - F) were monitored intensively after the use of resistant breeding rams, respectively during 6 years (farms B-E) and two others during 4 years (farm A and F). On the basis of an increase in presence of the ARR/ARR genotype and the ARR allele during the monitoring period, it is expected that the susceptibility for scrapie at the flock level is decreased for the five farms A, B, C, D and E. For farm F the genotype and allele distribution did not change much up until 2004, as resistant rams were not used consistently, so no genetic improvement was achieved here with respect to the scrapie susceptibility. For this reason, monitoring of farm F was ceased at the end of 2004 when the Dutch research project ended. At that moment the monitoring of farm A was ceased too, as on that farm the number of scrapie cases was too low. In this monitoring project no indications were found for the presence of new scrapie strains. However, the set up of this study was not adequate for that purpose. Recently in Europe and in the Netherlands atypical scrapie cases were diagnosed. The relevance of these scrapie cases for the current control programme is not clear at all. For a better insight in this field, more experimental research in combination with epidemiological work is necessary, in cooperation with other EU countries.

Twenty five Manech red faced focks were monitored for a long time, a first group of 15 flocks since the mid-1990s until July 2007 before and after using resistant ARR/ARR rams, a second group of 10 flocks since the beginning 2000s until July 2007 mainly after using resistant rams. From the first (1996-2001) to the second TSE legislation implemented in 2002, the scrapie survey of these 25 flocks moved mainly from passive surveillance (clinical cases) to active surveillance (rapid tests in abattoir or fallen stock). Moreover extra tests have been implemented in these flocks for research purposes (tonsil biopsies of alive sheep or culled animals). When looking at the evolution of prevalence during the 4 years of the contract, there was about no evolution between 2003 and 2004, and the estimated prevalence for the 4 groups of flocks ranged between 4.62% in 2003 and 4.30% in 2004. Then a clear decrease was observed in 2005, since the estimated prevalence was divided about by 3 times, with an estimate equal to 1.60 %. Finally, the scrapie prevalence fell down near zero in 2006 (0.14%). From the starting of using ARR/ARR rams, the ewes present in the 25 flocks belonged to two groups, either the ewes born before using ARR/ARR rams (the surviving ewes of old cohorts), or the birth cohorts of ewes born after using ARR/ARR rams: the first group of old ewes represented 8155 sheep in the 25 flocks, while the second group of ewes born after using ARR/ARR rams was composed of 5855 sheep. From the starting of using ARR/ARR rams until July 2007 (end of the monitoring), 493 new scrapie cases were found, of which 460 were from old ewes born before using ARR/ARR rams, and 33 from ewes of the cohorts born after using ARR/ARR rams. Thus 93% of the new scrapie cases detected after using ARR/ARR rams concerned surviving ewes of old cohorts born before using these ARR/ARR rams, mainly composed of susceptibles ewes with high scrapie incidence as demonstrated before by using "cure models" analysis. In other words, only 7% of the new scrapie cases concerned cohorts born after using ARR/ARR rams, i.e. 33 cases. Thus the scrapie incidence of these cohorts born after using ARR/ARR rams was very low (0.56%) compared to the incidence of the old cohorts born before using ARR/ARR rams, estimated on average to 30%. On the other hand, the 460 scrapie cases from the cohorts born before using ARR/ARR rams were composed of 333 ARQ/ARQ sheep (72,4%), 78 ARQ/VRQ (16,9%), 3 VRQ/VRQ sheep (0,7%), 11 ARR/ARQ sheep (2,4%) and 6 ARR/VRQ sheep (1,3%), plus 29 unknown PrP genotype sheep (6,3%). On the contrary, the 33 cases from the cohorts born after using ARR/ARR rams were composed only of susceptible sheep, i.e. 26 ARQ/ARQ sheep (78,8%) and 7 ARQ/VRQ sheep (21,2%). The 5855 ewes born after using ARR/ARR rams belonged to 76 birth cohorts, whose relative birth years ranged between zero (first year of using ARR/ARR rams) and 3 (fourth year of using ARR/ARR rams). The 33 scrapie cases for the cohorts born after using ARR/ARR rams occured in 12 cohorts (16% of the cohorts) in 10 flocks (table 18). Furthermore, 9 of the 12 cohorts had a relative year "zero", 2 cohorts a relative year "1" and 1 cohort a relative year "2". In other words, 75% of the scrapie cases happened the first year of using ARR/ARR rams, 17% the second year of using ARR/ARR rams, and 8% the third year of using ARR/ARR rams. Thus it appeared that the risk was discreasing with the number of years using ARR/ARR rams. Another main point was that the 33 scrapie cases were composed only of PrP susceptible ewes: 26 ARQ/ARQ sheep and 7 ARQ/VRQ sheep. The 12 cohorts with scrapie cases were not composed only of ARR/ARR, or ARR/ARQ or ARR/AHQ sheep, i.e. at least heterozygous ARR animals, showing that all the sires in these 12 cohorts were not ARR/ARR rams. The proportion of susceptible ewes ranged from 57% (flock C and relative birth cohort zero) to only 3% (flock A and relative cohort 2) in these 12 cohorts. Finally, when focusing on the consequence of using ARR/ARR rams to eradicate scrapie in infected flocks, we observed, in a few 2-3 years, a rapid and strong evolution of the PrP genotype structures toward resistance, with most of sheep at least heterozygous ARR. Under these conditions, the scrapie prevalence fell down near zero after 4 years of using ARR/ARR, and the incidence was reduced to almost zero after 2 years of using resistant rams, providing that almost 100 % of replacement ewe lambs were born only from ARR/ARR rams. Indeed, it appeared mandatory to use only ARR/ARR rams, in the aim to fully eradicate scrapie in a few years.

If risk of negative effects of selecting ARR allele on production or functional traits appeared as negligible due to the lack of pleiotropic or linkage effects as presented in the previous paragraph, on the other hand the loss of selection pressure on production traits due to the introduction of selective breeding for ARR allele must be considered and optimised. Indeed when selecting for scrapie resistance, the aim of the breeder is double: increasing the frequency of the ARR allele at the PrP locus, while maintaining a genetic gain as important as possible on production traits. The general framework is the introduction, within a selection scheme already existing, of an additional selection criterion concerning a new trait controlled by a single major gene. To solve this problem, Partner 1 elaborated a determinist and dynamic model, in order to find optimum rules for large size populations with overlapping generations. The model allowed a global optimisation of a selection scheme in order to maximize the frequency of the ARR allele while minimizing the loss of genetic gain on the initial production traits. This determinist and dynamic model was applied to the co-selection of the ARR allele at the locus PrP and of a production trait determined by a polygenic inheritance, the case of the milk production being considered. The selection scheme corresponded to the current organisation of the selection in French dairy sheep breeds, including for the males 2 steps of selection, first after birth and second after progeny testing; and for the females, 1 step of selection after the first lactation. The selection scheme was supposed to be applied during 15 years of selection. Three initial frequencies of the ARR/ARR sheep were considered: 0,03, 0,30 and 0,69. The chosen objective function was to maximize the frequency of ARR/ARR females constraining the genetic gain for milk trait not to be lower than a given percentage (p) of the possible genetic progress without any selection for the PrP locus. The main results without any constraint on the genetic progress for milk yield (p=100%) were the following: (1) the homozygous susceptible genotypes disappeared at the end of 15 years of selection, and the final ARR/ARR genotype frequencies for the females varied from 0,77 to 0,97, respectively for an initial ARR/ARR genotype frequencies of 0,03 and 0,69. The optimal strategy increased first the resistant genotypes for the males, then for the females. The objective to get a fully resistant population of sires was reached more rapidly with a high initial ARR/ARR genotype frequency (third year) than with a low initial one (eighth year). (2) the loss of genetic gain on the (milk) production trait was rather important (18 %) in the case of a low initial ARR/ARR genotype frequency (0,03), and very small (2 %) if the initial ARR/ARR genotype frequency was high (0,69). If now there was a constraint of a ten percent maximum loss on the genetic gain for milk yield (p=90%), the results were the following: (1) the optimal selection rates of males with resistant genotypes were lower and fluctuated during the first 10 years with cycles of high and low selection rates. However the selection rates for susceptible genotype males were null. The heterozygous genotype sires were conserved latter, and sires were fully resistant at the eleventh year. In other words, during the first 10 years, the strategy to select both heterozygous and fully resistant rams is promoted. (2) at the end of 15 years of selection, the proportions of heterozygous and homozygous resistant dams were about equivalent, i.e. homozygous susceptible dams were almost entirely eliminated (< 5%). To conclude, these modelling demonstrated that the optimal strategy strongly depended on the choice of the objective including or not constraint on the genetic gain for the (milk) production: higher this constraint to minimize the loss of genetic progress, longer the process to fix the ARR resistant allele.

A monitoring programme was implemented during 4 years in an experimental Sardinian flock of P6 naturally scrapie affected (classical strain). From the initial affected flock called "progenitor group" including 74 sheep, 2 lines called "resistant" and "susceptible" groups (RG and SG) have been procreated during 5 cohorts or generations (one per year) using respectively only ARR/ARR rams or ARQ/ARQ rams. Moreover, the 2 groups have been bred together, being in contact specially at lambing. The sheep of the 5 "resistant" cohorts, RG1 to RG5, represented 290 sheep and were composed only of at least heterozygous ARR (ARR/ARQ or ARR/AHQ) or homozygous ARR (ARR/ARR) sheep. The sheep of the 5 "susceptible" cohorts, SG1 to SG5, included 300 sheep and were composed mainly of homozygous ARQ (ARQ/ARQ) or partly of heterozygous ARR (ARR/ARQ) sheep.

Resultado final

Compartir esta página

Descargar