Benefit of genetic progress in sheep from Argentina

Transcription

1 Benefit of genetic progress in sheep from Argentina MUELLER, J.P. 1 ; VOZZI, P.A. 2 ; GIOVANNINI, N. 3 ; A LVAREZ, J.M. 4 ABSTRACT We analyzed the economic benefit of genetic improvement observed in stud flocks of the Horned Merino, Corriedale and Polwarth breeds using Provino, the Argentine sheep genetic evaluation service. With population statistics and reproduction parameters estimated from reference production systems of each breed we calculated the number of stud ram progeny. Using gene flow methodologies we calculated the genetic expressions of each lamb throughout his life and that of its descendants. The benefit of 10 years of genetic improvement in the stud flocks, accumulated for over 20 years in the general commercial flocks and discounted at an annual rate of 5% turned out to be 7.70, 0.93 and 0.12 million USD for the Horned Merino, Corriedale and Polwarth populations, respectively. For the first two breeds we considered a three tier structure was modelled and for the Polwarth breed a two tier structure was modelled. Sensitivity analysis indicated an ample scope to increase the economic benefit of genetic improvement in Argentina. Promotion policies for further genetic improvement in the stud flocks and increased dissemination of rams should be encouraged. Keywords: selection, genetic structure, Merino, Corriedale, Polwarth, Provino. INTRODUCTION The genetic improvement of domestic livestock allows producers to increase their incomes and to eventually reduce costs. In the case of structured populations, such as sheep in Argentina, genetic improvement depends on a few breeding sites that produce genetically superior breeding stock that then disseminates the genetic differences to their own flocks or their own customers. The production genetic potential of the entire population depends then on the effectiveness of the selection in those sites. In Argentina several breeding sites use the national genetic evaluation service of sheep Provino to obtain estimates of genetic merit that are then used in the selection process (Giovannini et al., 2015). Provino, in its Avanzado (Advanced) version, uses a prediction methodology of genetic merit based on mixed statistical models that allow to compare animals born in different years and sites, thus also enabling the monitoring of genetic progress in the evaluated population. Provino Avanzado has been used by breeding sites for the most important sheep breeds for several years, so their rates of genetic progress are known (Provino, 2015). Genetic improvement of traits of economic importance translates into additional incomes that reach the population that receives genes from the sites. The economic contribution of this gene flow in the population may sometimes be very important (Amer et al., 2007; Banks, 2000). In this paper we reviewed the genetic progress achieved in traits of economic importance in three sheep breeds in Argentina (Merino, Corriedale and Polwarthl), we assigned them an 1 INTA EEA Bariloche. Modesta Victoria 4450, San Carlos de Bariloche, (8400) Río Negro, Argentina. Correo electrónico: mueller.joaquin@inta.gob.ar 2 INTA EEA Chubut. 25 de Mayo 4870, Trelew, (9100) Chubut, Argentina. Correo electrónico: vozzi.alejandro@inta.gob.ar 3 INTA EEA Bariloche. Modesta Victoria 4450, San Carlos de Bariloche, (8400) Río Negro, Argentina. Correo electrónico: giovannini.nicolas@inta.gob.ar 4 INTA EEA Valle Inferior. Ruta Nacional 3 km 971, Viedma, (8500) Río Negro, Argentina. Correo electrónico: alvarez.juan@inta.gob.ar Received December // Accepted April 27 Aceptado // Published online November

2 December 2016, Argentina economic value and we estimated the economic benefits of genetic improvement for the commercial population of these breeds. MATERIALS AND METHODS Analytical model In order to translate the genetic changes into economic benefits it is necessary to track the improved genes in the population and to quantify their dispersion over time. Hill (1974) and Elsen and Mocquot (1974) developed a recursive methodology to describe the gene flow over time. This methodology facilitates the analysis of genetic progress in populations of overlapping generations and their economic evaluation. FAO (2010) presented an example of an analysis of the benefit of genetic improvement in a structured spreadsheet-based system. Amer et al. (2007) developed a parametrized model that allows describing gene flow in structured populations of sheep and cattle, and they applied it to direct and maternal expression traits. Here the Amer et al. (2007) model is reduced to the case of sheep and is adapted to include characters of repeated expression, such as those related to annual shearing. The parameter symbols used in the model were chosen as to be consistent with those used by Amer et al. (2007). - Progeny of breeding rams: we assume a sheep population that corresponds to the set of breeding stock of a breed that produces breeding rams that are used in commercial or general flocks. The number of lambs with genes from those rams can be modeled assuming that there is a total (c) of sheep on the sites and that from each sheep are produced and sold a number (b) of rams per year. Then, each of these rams is used in general flocks for a number of years (u) with a constant year-to-year survival probability of (λ). In each year, each ram serves a number (f) of sheep that produce, on average, (p) lambs per year each. Then the number of lambs born in the following year (k) of the (u) years of use of the breeding rams is Ok = c b λk-1 f p para 0 < k u. - Expression of genetic superiority: a lamb that is the offspring a breeder ram will express the genetic superiority of its father and transmit it to its own offsprings, in case it eventually becomes part of the reproductive population. Genetic expression (whether from the lamb itself or its offspring) can be divided into a direct and an adult component. A direct expression is, for example, weaning weight, and an adult expression is, for example, fleece weight or calving. The term d i,j refers to the direct genetic expressions and m i,j to the the adult genetic expressions of a lamb and the expressions of its offsprings in successive generations j. The lamb, son of a breeding ram, is born in year 0 and sold at weaning in year 1 or is maintained until its useful life. Then, the lamb of a ram breeder expresses half of its genes directly at weaning only in year 1 (i = 1), such that d 1,1 = 1/2, and the rest of the column of direct expressions over time is zero, that is to say d i,1 = 0 for i> 1. If the lamb is not slaughtered, it expresses genes as an adult over the years such that m i,1 = di, 1 wk where wk is the probability of that lamb to express in the year k after birth. Expressions of the offspring of that lamb are presented as generation 2 (j= 2) and so on. In general, m i,j = Σ di + 1-k,j wk where the sum is for k = 1ai. From the expression of adult genes it is possible to recursively calculate the expression of direct genes in the next brood, such that di + x, j = mi, j-1 1/2 p for j> 1, where x is the number of years between birth and first service (or first calving, x= 2 in typical sheep systems). The 1/2 factor expresses the ratio of genes passed to the progeny. The sum of expressions (through j generations) of the genes of a breeder ram over the years (i) can be calculated as Σ di= Σ di,j for direct traits and as Σ mi= Σ mi,j for adult traits, with summations ranging from j= 1 to jmax, where jmax is the maximum number of generations over which expressions are counted. As in the fourth or fifth generation the expressions of the initial lamb become insignificant. For the purposes of this work jmax was set in 5. - Genetic expressions in additional strata: so far a twostrata structure has been modeled: stud flocks and general flocks. Medium-sized commercial establishments often have multiplier studs to produce their sheep and eventually sell surpluses. Large commercial establishments usually have three strata: stud flocks, multiplier flocks and general flocks. To model the gene flow with an additional layer is simple. Lambs born in the additional stratum express half of the genes of the rams produced in the multiplier stratum (d1,1= 0.25). If the number of rams produced by the multiplier sheep (b), its survival (λ) and the number of sheep served by each (f) is similar to the reference parameters used in the upper stratum of the stud flocks, then the number of multiplier rams produced in u years of each breeder ram is c b2 Σλk-1 f with summation ranging from k= 1 to u. - Incorporation of genetic tendency: it is assumed that s is the genetic superiority in economic terms by traits expressed in the slaughter (directly) of a lamb born in year i with respect to economic value in year i-1. It is also assumed that a i is the genetic superiority in economic terms for expression traits in adults (or in maternal form). We define T as the number of years the genetic improvement program lasts and H as the number of years to be considered in the evaluation of the impact of the breeding program, such that T H. Then, the accumulated direct genetic progress in year i of the analysis period (i = 1 to H) is yi = Σ sj and the cumulative adult zi = Σ aj with sums ranging from j= 1 to i for i= 1 to T. If T<H, yi= yt and zi= zt for i of T + 1 to H. - Discounted total economic benefit: the economic benefit of genetic progress e i,j where i is the year of the first mating of breeding rams and j is the year in which the benefits of such mating materialize in the general MUELLER, J.P. 1 ; VOZZI, P.A. 2 ; GIOVANNINI, N. 3 ; ÁLVAREZ, J.M. 4

3 flock, then we obtain as e i,j = Ok (yi-k+1 dj-i+1 + zi-k+1 mj-i+1) with summation for k= 1 to i and j i. If k>u, O k = 0. By adding the economic benefits of each year (i= 1 to H) as vj= Σ e i, j and then updating those values by multiplying with the updating factor qj= (1/(1 + r))j+1, where r is the discount rate, we obtain the total discounted benefit as ε= Σ qj vj with sum for j= 1 to H. In the analysis of structures of two strata the benefit is occurs from year 1, considering that in the year 0 commercial sheep mate with rams that the commercial flocks buy at a discount cost. For the three-strata structure analysis it is considered that the first multiplier rams will have their first progeny in the second year of birth, that is to say in year 2 and the first incomes will appear after selling the lambs in year 3. Analyzed systems We analyzed the economic impact of breeding programs in the Merino, Corriedale and Polwarth breeds. The first two are the main breeds of sheep in Argentina while the Polwarth breed is important in the coastal region (litoral) (Mueller, 2005). The Merino breed is bred in its Horned and Mocho type. Here we consider the Horned type, which represents 90% of the breed. In order to estimate the necessary parameters to calculate the number of offspring carrying rams genes produced in the genetically assessed stud flocks, we analyzed the databases of the Provino records were analyzed and typical population statistics were summarized to calculate the number of lambs born from breeding rams in each breed (Table 1). The lamb born from a breeder ram will express the genetic superiority (or inferiority) acquired as lamb (to slaughter) or as an adult (over the years). To determine adult expressions it is necessary to know the probability of that lamb of being present through successive years (wk). In commercial flocks, normally only females are retained and therefore the values of wk are obtained from the age distribution of sheep and the replacement rates in typical commercial flocks. These age distributions and replacement rates depend on the reproductive rate, mortality, and the life span of the sheep. If one wished to consider the case of retention of males as wool capons a vector, por example c i,j, containing the proportion of capons retained by lamb born over the years of life should be added. In general, sheep producers do not retain capons or retain very few and therefore, this case is not considered in this analysis. For the purpose of modeling wk it is necessary to determine the age distribution. If m1 is the mortality from birth to the first year and m2 is the annual mortality from the first year, then the vector w can be modeled by determining the proportions of animals in each age category as qt =1/ (1- m2) t with summation of t= 1 to uf where uf is the number of sheep (=births+ 1) of sheep and q0= q1 (1 + m1). The percentage of replacements is then re= 2q0 /p and wt= qt x re for t of 1 to uf and w0= 0. Note that if deifned as the number of lambs produced by sheep at weaning (year= 1) then m1 can be ignored. In fact, for the purposes of this study, p is defined as percentage of lambs at weaning. Table 1 shows the typical values for the required parameters in the wk model. For Merino and Corriedale breeds the genetic structures of three strata were analyzed and for the Polwarth breed a structure of two strata was analyzed. For the three races we decided to present the economic benefit only of the stratum of general flocks. Sensitivity analyzes were conducted on lambs produced by sheep (b) and sheep served by ram (f), considering as reference the typical values and calculating the effect of smaller and larger quantities. The rate of survival of rams (λ) and the reproductive rate (p) were also analyzed for lower and higher values of the references for each breed. Analyzed traits and their genetical progress We studied the genetic progress in weaning weight (WW) as a direct expression trait, and fleece clean weight (CFL), average fiber diameters (AFD), adult weight (AW) and number of weaned lambs (NWL) as adult expression characters. In the Merino and Polwarth breeds no progress was considered in NWL since this is a characteristic of little Parameter Symbol Merino Corriedale Polwarth Number of sheep in establishments Provino Advanced c Number of rams produced by each sheep of the flock b 0,3 0,25 0,3 Years of use of the rams in commercial flocks u Survival of rams from one service to another λ 0,9 0,9 0,9 Number of ewes served per ram f Number of lambs (born) per ewe p 0,75 0,80 0,75 Annual mortality from the first year m Number of shearing of females (=services +1) uf Table 1. Productive parameters in stud and commercial flocks.

4 December 2016, Argentina importance in the breeding plans in those breeds and for which there is no enough information on the genetic tendency. The annual average genetic merit of the traits of interest in the three breeds in their catalogs (Provino, 2015) are shown in Figure 1. With the aim of showing the average annual genetic progress we used the linear regressions of the breeding values for the last 10 years as indicators of annual genetic progress (Table 2). Genetic progress in economic terms In order to determine the economic value of the annual genetic progress observed in each trait, we used the prices in force in November 2015 and the production costs published by Álvarez et al. (2014). These costs were applied as a percentage of the value of the product in 2014 and that percentage was then applied to the year 2015 to bring the monetary values to the same reference level. The result obtained from the prices minus the costs per unit of change in each trait is presented in Table 3. The exchange rate used for all calculations was 10 pesos per USD. The genetic progress in economic terms achieved in WW by lamb and in NWL, CFL, AFD and AW by adult sheep is presented in table 4. In commercial flocks we evaluated the economic benefit for the three breeds over a 20 year horizon (H= 20) considering the effect of 10 years of reeding (T= 10) and a discount rate of 5% (r= 5) that is suitable for programs of breeding or national interest (Bird and Mitchell, 1980). The Trait Merino Corriedale Polwarth Weaning weight (kg/year) 0, ,138 Clean fleece weight (kg/year) Fiber diameter (μm/an o) ,101 Number of weaned lambs (lambs/year) Adult weight (kg/year) 0, ,162 Table 2. Annual genetic progress in units of traits (slopes in the regressions of Figure 1). Trait Merino Corriedale Polwarth Weaning weight (kg/year) 2,37 0,53 2,10 Clean fleece weight (kg/year) 0,52 8 0,12 Fiber diameter (μm/an o) 0,13 0,10 0,48 Number of weaned lambs (lambs/year) 0,51 Adult weight (kg/year) 1,52 0,30 0,73 Direct expression traits (USD/year) Adult or maternal expressions traits (USD/year) 2,37 0,53 2,10 2,18 0,98 1,33 Table 4. Annual genetic progress in economic units. Results Merino Corriedale Polwarth Number of sheep in Provino stud flocks E1 rams produced per year in E1 stud flocks E1 x Rams lifespan E2 sheep mated by E1 rams E2 lambs produced by E1 rams E2 rams produced by E2 sheep E2 rams x lifespan E3 sheep paired by E2 rams E3 lambs produced by E2 rams Accumulated economic benefit (in thousands of USD) Table 5. Number of animals in genetic improvement and economic benefit accumulated in 20 years for genetic improvement in reference systems. Note: E1, E2 and E3 are the strata of stud flocks, multipliers and general flocks, respectively. Trait Merino Corriedale Polwarth Weaning weight (kg/year) 19,46 19,46 15,20 Clean fleece weight (kg/year) Tabla 3. Benefit (price-cost) of unit changes. 64,82 39,76 59,71 Fiber diameter (μm/an o) -3,51-2,01-4,72 Number of weaned lambs (lambs/year) 486,53 505,99 364,87 Adult weight (kg/year) 5,40 5,40 4,51 Source: IPCG-based meat prices (2015) of November 20, 2015 and wool prices based on SiPyM (2015) of November 26, Cost as a percentage of prices based on Álvarez et al. (2014). sensitivity of the results was also analyzed for one-year (T = 1) or twenty years (T = 20) breeding programs and for programs without discount (r= 0.0) and with an appropriate discount for programs of individual interest with a higher opportunity cost (r = 0.1). RESULTS AND DISCUSSION Benefits according to the breed The discounted benefits accumulated in 20 years (specifically from 2004 to 2023) due to 10 years (specifically from MUELLER, J.P. 1 ; VOZZI, P.A. 2 ; GIOVANNINI, N. 3 ; ÁLVAREZ, J.M. 4

5 Horned Merino: weight at weaning (WW) and adult weight (AW) Horned Merino: clean fleece weight (CFL) and average fiber diameter (AFD) 4,0 3,5 3,0 2,5 2,0 1,5 1,0 0,5 y = 0,282x + 0,228 y = 0,122x ,2 0,1-0,1-0,2-0,3-0,4-0,5-0,6 y = 08x + 69 y = -38x - 73 WW, kg AW, kg PVL, kg PDF, µm 1,6 Corriedale: weight at weaning (WW) and adult weight (AW) Corriedale: clean fleece weight (CFL), number of weaned lambs (NWL) and average fiber diameter (AFD) 1,4 1,2 1,0 0,8 0,6 0,4 0,2 y = 55x + 0,615 y = 27x + 0,437 0,3 0,2 0,1-0,1-0,2-0,3-0,4 y = 02x + 36 y = 01x + 01 y = -49x + 0,211 WW, kg AW, kg NCD, cord PDF, µm PVL, kg Polwarth: weight at weaning (WW) and adult weight (AW) Clean fleece weight (CFL) and average fiber diameter (AFD) 1,6 1,2 0,8 0,4 y = 0,162x - 0,277 y = 0,138x - 0,339-0,2-0,4-0,6-0,8-1,0 y = 02x - 31 y = -0,101x ,4-1,2 WW, kg AW, kg PVL, kg PDF, µm Figure 1. Genetic progress in traits of economic interest of three breeds in the last ten years (adapted from Provino, 2015) to 2013) of genetic improvement in stud flocks resulted in USD 7.70 million for the reference system of Horned Merino, USD 0.93 million for Corriedale and USD 0.12 million for Polwarth (table 5). The difference in the benefit between Merino and Corriedale is mainly due to the size of the population and the progress rates achieved. The greater number

6 December 2016, Argentina A Two-strata reference system 60 Benefit in millions of USD Years from the first mating with breeding rams B Three-strata reference system 700 Benefit in millions of USD Years from the first mating with breeding rams Figure 2. Discounted annual benefit in the strata of commercial flocks of Horned Merino that uses rams from the Provino Avanzado ranches stratum or from an MPR multiplier stratum (B). Parameters of the reference system: years of breeding= 10, discount rate= of sheep in the Merino ranches results in a greater spread to more sheep in multiplier and general flocks (Table 5). For Merino the annual rates of genetic progress are much higher than the rates of Corriedale - in all the characteristics except for AFD, where the difference was lower (Table 2). When compared with the other breeds, the difference in the benefit of Polwarth - asides from the difference in the size of the stud flocks (Table 1) - is mainly due to the 2 stratum structure considered for this breed. The rams produced by Polwarth producers - grouped in Caban as Integradas (integrated ranches) are mostly used in general flocks. If those rams were to be used in multiplier flocks the progress would multiply to more flocks and the benefit in that stratum would reach approximately USD 1.55 million, proportionally greater than Merino and Corriedale, which have more sheep in their flocks but lower rates of progress than the population of Polwarth. In any case, the economic benefits of genetic improvement are remarkable. For example, the benefit per Horned Merino sheep was USD 5.36 (USD 7.69 billion / million sheep, Table 3), 32% more than the value of 1 kg of wool (USD 4.06/Kg for pre-lambing Merino wool (Fineness: 20 μm, 60% yield after combing, SiPyM, 2015). The calculated benefit is also consistent with similar analyzes performed in sheep populations in Australia and Great Britain. For example, Greeff (1997) estimated in one million Australian MUELLER, J.P. 1 ; VOZZI, P.A. 2 ; GIOVANNINI, N. 3 ; ÁLVAREZ, J.M. 4

7 A Years of breeding (T) B Discount rate (r) y = -2733x x y = 36540e -0,767x T=1 T=10 T=20 r=0 r=5 r=0,10 C Rams produced per sheep (b) D Ram s survival (λ) y = 5131x y = 2233x b=0,2 b=0,3 b=0,4 λ=0,8 λ=0,9 λ=1,0 E Sheep served per ram (f) F Reproduction (p) y = 5131x y = 1266x f=20 f=30 f=40 p=0,65 p=0,75 p=0,85 Figure 3. Discounted and accumulated benefit in 20 years resulting from 10 years of genetic progress in Horned Merino (in thousands of US Dollars). Sensitivity to changes in the assumed parameters of the reference system: years of breeding= 10, discount rate= 0.05, rams produced per mother= 0.3, sheep served per ram= 30, reproduction= 0.75, annual rams survival= 0.9. dollars the discounted economic benefit accumulated in 14 years for an improvement program of sheep in western Australia, whose average annual genetic progress over 6 years was -0.1 μm in AFD, 0.04 kg in CFL and 0.26 kg in body weight per year. Atkins (1993) estimated that the 30-year benefit of sheep breeding in Australia results in 3.5 billion accumulated USD. For the calculations Atkins (1993) assumed an annual progress rate of 1 Australian dollar, equivalent to 2.7% of the annual gross margin. Amer et al. (2007) estimated the 10-year benefit of sheep breeding in

8 December 2016, Argentina Great Britain as close to 17.8 million British pounds. For the three breeds only the benefit in the stratum of the general flocks was considered, and to include the benefit from upper strata would increase total benefit by 9%. We decided not to report this additional benefit since the structure of incomes and costs tends to be different in the rams-producing strata than in commercial flocks. Although they are quite simple to include in the analyzes, we did not include the costs of genetic improvement, but it can be assumed that they are amply compensated for by the additional incomes not considered in the upper strata. Sensitivity analysis of assumptions For the three breeds we analyzed the discounted benefits accumulated in 20 years (H= 20) after 10 years of genetic improvement (T= 10). Figure 2 shows that in Horned Merino the benefit reaches an annual maximum after the end of the years of the breeding program (years 1 to 10). This is due to the delay in the manifestation of progress through the strata and by the multiplication of the number of animals carrying the improvement as time passes. Later, progressively, the updated annual benefit decreases by the updating factor (qj) which decreases over time. Additional results indicate that the effect of the update is much greater than the dilution effect of genes enhancers over time. The reduction of the benefit over time is observed first in the multiplier stratum and then in the stratum of commercial flocks (gura 2A and 2B). The accumulated discounted benefit also depends in a quadratic form on the modeled breeding years (Figure 3A), whereas the discount rate has an exponential effect on the benefit (Figure 3B). BEsides the population size involved, the number of strata considered, the annual rate of progress, the number of years of genetic improvement and the discount rate, the benefit depends on the parameters of the reference production systems. Several of the assumed parameters linearly modify the results. For example the number of rams produced per mother in the upper strata (Figure 3C), the survival of these rams (Figure 3D), the number of sheep served per ram (Figure 3E) and the reproductive rate (Figure 3F). For example, the reduction of number of rams produced per mother (b) from 0.3 to 0.2 reduces the benefit similarly to the reduction of the number of sheeps served per ram (f) from 30 to 20. The reference models assume a natural service. In the reference system with Horned Merino, each ram ram is used annually on 30 sheep (f= 30), but at least some rams could be used in artificial insemination (AI) programs with fresh semen. If, for example, 5% of the rams born in the upper and in the multiplier strata are used in AI to serve 300 sheep each (f= 300) and the remaining 95% is used in natural service (f= 30) then the average of services per ram climbs to 43.5 (average weighted f= 43.5), the number of lambs with breeder ram genes increases from 1.1 million to 2.3 million and the accumulated discounted benefit in the general sheep flocks increases more than two fold. It is more likely that AI is used to use less but genetically superior rams than to increase the number of lambs. In that case, the rate of progress used in the applied models must be increased depending on the increases in the achieved selection differentials. Genetic progress and its dissemination The benefit was calculated on the genetic progress registered in the establishments using Provino Avanzado. Of the 27 active flocks of Horned Merino, only 5 participate in the population evaluation Provino Avanzado, although these 5 establishments provide 48% of sheep to multipliers (445 of 924, AACM database, births 2014, Epper, C.E., personal communication). We can not infer the genetic progress of the Merino flocks that do not participate in Provino Avanzado. In the event that these flocks had the same genetic progress, the calculated benefit and the improved sheep would increase proportionally. That is, the total benefit would be almost twice the USD 7.70 million calculated for the population of Provino Avanzado. It should be noted that the number of Horned Merino sheep from general flocks that result from the modeling of the reference system (1.44 million, Table 3) accounts approximately for the 46% of total Horned Merino in Argentina, assuming that all the sheep of the provinces of Rio Negro and Chubut and half of the sheep in Santa Cruz (SENASA, 2015) are Merino and 90% of them are Horned. If the entire Merino population of Argentina is genetically linked to the stud flocks, then the 46% is consistent with the 48% of rams provided by the stud flocks using Provino Avanzado. The genetic progress registered in the stud flocks was calculated using the usual prediction methodology for breeding values per year of birth. The methodology depends on the assumed genetic parameters, so the calculated benefit also depends on those parameters. Provino Avanzado uses parameters obtained from the evaluated population that, for Merino, are usually consistent with averages in the literature (Safari et al., 2005, Mueller et al., 2003). For other breeds with smaller populations (Polwarth) or with less published information (Corriedale), Provino uses parameters from the literature that may differ slightly from those appropriate for the breed in Argentina and, consequently, generate less accurate progress estimates. In any case, the genetic progress used to calculate the benefit is decisive and Figure 1 shows some traits with important annual fluctuations, particularly for Corriedale. To consider the variance in genetic progress would allow to assign confidence limits to genetic progress and, consequently, also to the calculated economic benefit (Ponzoni et al., 2007). Breeders participating in Provino Avanzado use genetic evaluations together with their own experience and intuition in their selection decisions. These same breeders also usually make introductions of rams from other flocks and even from other countries. In the case of Merino, imported animals are of great genetic importance (Vozzi et al., 2014) and it is not possible to determine conclusively the proportion MUELLER, J.P. 1 ; VOZZI, P.A. 2 ; GIOVANNINI, N. 3 ; ÁLVAREZ, J.M. 4

9 of genetic progress awarded to the highest accuracy of selection by Provino information and the proportion awarded to introductions and other decisions to replace sheep and rams. However, it is possible to estimate the theoretical genetic progress that could be achieved by in-plant selection using the genetic merits estimated by Provino and the differentials and generational intervals of the reference systems. The selection index theory also allows predicting possible genetic progress if breeders based their selection exclusively on selection indexes proposed by Provino (Alvarez et al., 2014). In the case of the Horned Merino, the definition using the selection index, based on measurements of body weight at shear, CFL and AFD used in many establishments and assuming parameters of the reference system, the theory predicts an annual genetic progress of 0.17 kg, 0.03 kg, μm and 0.36 kg, for WW, CFL, AFD and AW, respectively. Progress rates actually achieved (Table 2) indicate that breeders achieved approximately 70% of the possible progress in body weights (WW and AW) and approximately 30% in fleece characteristics (CFL and AFD). The scope for increasing genetic progress is then broad, even without the incorporation of sheep from other regions. in Argentina is the industry that can produce higher value wool coils from a better raw material (Mueller, unpublished). The adoption of genetic improvement technologies depends on the availability of credit, research products, extension services and genetic evaluation services. Banks (2001) estimated the benefit of Lambplan, the Australian sheep genetic evaluation service equivalent to the Argentine Provino, in 71 million australian dollars over a 5-year period. Atkins (1993) estimates that at least half the benefit of Australian sheep breeding is due to research and extension products in these areas. CONCLUSION The accumulated and discounted economic benefits generated by genetic improvement in flocks and their dissemination in general flocks of the Horned Merino, Corriedale and Polwarth breeds in Argentina is substantial as is the margin for increasing that benefit by increasing the rate of genetic progress in the breeding sites and increasing the number and spread of breeding rams. These results should encourage advocacy policies, based on the proven economic benefits of both breeding in flocks and the use of breeding rams. Implications of the benefit The evaluation of a breeding program should not be restricted to the involved animal population but should consider the consequences along the value chain of genetically enhanced products (FAO, 2010). For example, benefits due to the production of more wool are not restricted to breeders and producers of general flocks, but are multiplied in the wool industry. In fact, preliminary analysis indicates that the link most benefited by the thinning of Merino wools ACKNOWLEDGEMENTS To Peter Amer of AbacusBio Ltd for his suggestion to include annual expression characters, such as shearing, in the vector of maternal characters (here adults). To Carlos Epper of the Argentine Association Breeders of Merino (Asociación Argentina Criadores de Merino) for providing statistics and genealogy of the Merino breed. To Daniel Maizon of INTA La Pampa for his suggestions of improvement in preliminary versions of this work.