Expected Genetic Response for Oleic Acid Content in Pork 1

Intramuscular fat (IMF) and oleic acid (C18:1) content in pork are important issues for the pig industry and consumers. Data from a purebred Duroc line were used to i) estimate the genetic parameters of IMF and C18:1 and their genetic correlations with lean growth components, and ii) evaluate the opportunities for genetically improving C18:1 in IMF. The data set used for estimating genetic parameters consisted of 93,920 pigs, from which 85,194 had at least 1 record for BW or backfat thickness (BT) at 180 d and 943 for IMF and C18:1 at 205 d. Intramuscular fat content and C18:1, expressed as percentage of total fatty acids, were determined in the gluteus medius muscle by gas chromatography. Genetic parameters for C18:1 were estimated under a Bayesian 4-trait multivariate animal mixed model. Heritability of C18:1 was 0.50, with a probability of 95% of being greater than 0.37. Genetic correlations of C18:1 with BW, BT, and IMF were 0.11, 0.22, and 0.47, respectively (with a probability of 95% of being greater than –0.07, 0.04, and 0.27, respectively). Genetic responses were evaluated by deterministic simulation using a half-sib recording scheme for C18:1 and the previously estimated parameters. The C18:1 content is expected to exhibit only minor changes in selection programs directed at growth rate but to decrease in those focusing on lean content. Maximum expected response in C18:1 at no lean growth loss (i.e., at no change in BW and BT) was 0.44%, with a resulting correlated response in IMF of 0.15%. However, because lean growth is emphasized in the breeding goal, the resulting response scenarios are more constrained. We concluded that there is evidence to support the idea that C18:1 in IMF is genetically determined and defined selection strategies can lead to response scenarios in which C18:1, IMF, BT, and BW can be simultaneously improved. However, if adopted, the potential for lean growth would be reduced. The extent to which it is affordable relies on how much consumers are prepared to pay for high-C18:1 pork products.


INTRODUCTION
Fat content and composition are important issues for the pig industry and consumers.Intramuscular fat (IMF) and oleic acid (C18:1) contents are two of the traits that have attracted greatest interest in the last few years.The IMF content has been favorably related to tenderness and juiciness of cooked meat (Wood et al., 2008), as well as to technological and sensorial proper-ties of dry-cured products (Ruiz-Carrascal et al., 2000).The C18:1 content has been traditionally considered a key quality criterion in dry-cured products because of its positive role in the manufacturing process and in flavor (Toldrá, 2002, Pages 27-62).More recently, because of its associated benefits for human health (Christophersen and Haug, 2011;Jiménez-Colmenero et al., 2010), C18:1 has become an appreciated trait in some niche markets of fresh meat.
Both IMF and C18:1 are affected by dietary and genetic factors (De Smet et al., 2004).It is known that IMF, despite being unfavorably correlated with carcass lean content, can be efficiently selected (Suzuki et al., 2005).However, there is little evidence on the opportunities for genetic change in fatty acid (FA) composition.Recent studies in this regard, although promising, were either based on small and heterogeneous data sets (Ntawubizi et al., 2010;Sellier et al., 2010) or, regarding C18:1, not conclusive (Casellas et al., 2010).Moreover, because the challenge for the industry is to develop selection criteria not only aimed at increasing C18:1 but at the whole profit of a line, the genetic correlation structure of C18:1 with other economic traits, particularly with lean growth, is needed.Therefore, the aims of this study were to i) estimate the heritability of C18:1 and its genetic correlations with IMF and lean growth in pigs from a Duroc line primarily used for producing high quality dry-cured hams, and ii) discuss the opportunities for genetically improving C18:1 under different selection scenarios.

MATERIALS AND METHODS
All experimental procedures were approved by the Ethics Committee for Animal Experimentation of the University of Lleida.

Animals and Sample Collection
Data from a purebred Duroc line were used for the analyses.The line was completely closed in 1991 and since then it has been selected for an index including BW, backfat thickness (BT), and IMF (Solanes et al., 2009).The data set used for the estimation of genetic parameters consisted of 93,920 pigs, from which 85,253 had at least one recorded trait.Pigs with records were born from 1996 to 2009.At approximately 75 d of age, pigs were moved to the fattening units, where they were penned by sex (8 to 12 pigs/pen) until slaughter.All pigs were performancetested at an average age of 180 d for BW and BT.Backfat thickness was ultrasonically measured at 5 cm off the midline at the position of the last rib (Piglog 105,Herlev,Denmark).During the test period, pigs had ad libitum access to commercial diets.Since 2002, a sample of the purebred barrows used for producing dry-cured ham was taken for recording IMF and C18:1.Two barrows per litter were taken from fixed litters chosen either at random or selected according to the mid-parent BLUP breeding values for BW and BT at 180 d.These barrows were raised in 12 batches until slaughter at around 205 d.From 160 d onward, barrows were fed a commercial pelleted finishing diet (Esporc, Riudarenes, Girona, Spain) with an average composition of 16.9% crude protein, 6.59% fiber, and 6.66% fat (C16:0, 20.8%; C18:0, 7.1%; C18:1, 35.4%; and C18:2, 27.4%).Feed in each batch was analyzed in triplicate as described in Cánovas et al. (2009).At the end of the finishing period, the barrows were slaughtered in a commercial slaughterhouse.After chilling for approximately 24 h at 2°C, a sample of at least 50 g of the gluteus medius muscle was taken from the left side ham, immediately vacuum packaged, and stored in deep freeze until required for IMF and C18:1 determination.A summary of the population characteristics and number of records, sires, dams, and litters used for each analyzed trait is given in Table 1.

Fat Analysis
After gluteus medius samples were completely defrosted and vacuum drip losses were eliminated, the dissected muscle, trimmed of subcutaneous and intermuscular fat, was minced.A representative aliquot from the pulverized freeze-dried muscle was used for fat analysis.Intramuscular fat content and composition was determined in duplicate by quantitative determination of the individual FA by gas chromatography (Bosch et al., 2009).Fatty acid methyl esters were directly obtained by transesterification using a solution of 20% boron trifluoride in methanol (Rule, 1997).Methyl esters were determined by gas chromatography using a capillary column SP2330 (30 m × 0.25 mm, Supelco, Bellefonte, PA) and a flame ionization detector with helium as carrier gas.Runs were made with a constant columnhead pressure of 172 kPa.The oven temperature program increased from 150 to 225°C at 7°C/min, and injector and detector temperatures were both 250°C.Quantification was carried out through area normalization after adding into each sample 1,2,3-tripentadecanoylglycerol as internal standard.Intramuscular fat content was calculated as the sum of each individual FA expressed as triglyceride equivalents (AOAC, 1997).Oleic acid content was calculated as the percentage of C18:1 relative to total FA in IMF.Fatty acids were identified by comparing their relative retention times with those of the external standard and confirmed by comparing their mass spectra to the computer library of the GC/MS database Wiley 275.L and NBS 75 K.L. Fatty acids were analyzed on a simple quadrupole instrument (GC/ MSD 6890N-5973N, Agilent Technologies, Wilmington, DE) equipped with an electron ionization source using the same temperature program as described above.Scanned mass range of FA was m/z 35 to 450 and the scanning rate 3.46 scans/s.

Estimation of Genetic Parameters
Genetic parameters for BW, BT, IMF, and C18:1 were estimated fitting a 4-trait multivariate animal model.In matrix notation, the model was: where y i is the vector of observations for Trait i (BW, BT, IMF, and C18:1); b i , a i , c i , and e i are the vectors of systematic, additive genetic, litter, and residual effects, respectively; and X i , Z i , and W i , the known incidence matrices that relate b i , a i , and c i with y i , respectively.Systematic effects for BW and BT were the batch (1,039 levels), gender (3 levels; males, females, and castrates), and age at measurement as a covariate.Pigs tested at the same time and in the same unit were considered as one batch.The same model was used for IMF and C18:1 but with systematic effects only including batch (12 levels) and age at measurement (or carcass weight).Because there were only 1.7 piglets/ litter with records on IMF and C18:1, litter was dropped from the model for these 2 traits.Intramuscular fat content and C18:1 were analyzed using either the raw data or the following u 1 and u 2 isometric log-ratio (ILR) transformed variables (Egozcue et al., 2003): where (100 -IMF) + [C18:1 + (100 -C18:1)] × IMF/100 = 100.
Genetic parameters were estimated in a Bayesian framework using Gibbs sampling with the TM software (Legarra et al., 2008).The traits were assumed to be conditionally normally distributed as follows: , where R was the (co)variance matrix.Sorting records by pig, and trait within pig, R could be written as R 0 Ä I, with R 0 being the 4 × 4 residual (co)variance matrix between the 4 traits analyzed and I an identity matrix of appropriate order.Flat priors were used for b i and residual (co)variance components.Additive genetic and litter val-ues, conditionally on the associated (co)variance components, were both assumed multivariate normally distributed with mean 0 and with (co)variance G Ä A and C ÄI, respectively, where A was the numerator relationship matrix, G was the 4 × 4 genetic relationship matrix between the 4 traits, and C was the 2 × 2 (co)variance matrix between litter effects of BW and BT.The matrix A was calculated using all the pedigree information summarized in Table 1.Flat priors were used for additive and litter (co) variance components.Statistical inferences were derived from the samples of the marginal posterior distribution using a unique chain of 2,000,000 iterations, where the first 250,000 were discarded and 1 sample out of 100 iterations was retained.Statistics of marginal posterior distributions and the convergence diagnostics were obtained using the BOA package (Smith, 2005).Convergence was tested using the Geweke Z-criterion and visual inspection of convergence plots.

Prediction of Expected Responses
The expected genetic response for C18:1 from a simulated breeding program was compared in 2 recording scenarios.In the first, it was assumed that records on C18:1 were not available and selection was only directed at either increasing BW (or IMF) or decreasing BT, while in the second, records on C18:1 were available and C18:1 was proactively selected.The selection objective in each case was derived as the linear combination of the appropriate breeding values weighted by their economic values.Economic weights were determined iteratively using a desired-gains approach until the desired combination of genetic gains was achieved.For simplicity, only some illustrative cases in each scenario are presented.A population with discrete generations was simulated in which 40 boars were randomly mated to 400 sows with a mating ratio of 1 boar to 10 sows.The breeding scheme consisted of 2 selection stages resulting in the top 25% males and 50% females, with the same selection pressure in each stage.Two males and 2 females from the offspring of each sow were performance-tested at 180 d for BW and BT.In the second stage, 3 of the culled individuals per sire family were slaughtered to determine IMF, in the first scenario, and also C18:1, in the second.Pigs in the first stage were selected on the individual, full-sib and half-sib phenotypic performance of BW and BT, and the pedigree information (BLUP) of all recorded traits.Selection on the second stage was additionally based on the new half-sib records on IMF and, if available, C18:1.Only the first stage, but with the whole selection pressure, was applied in cases where neither IMF nor C18:1 were recorded.Selection response was predicted by deterministic simulation of a 2-stage selection scheme with discrete generations using the program SelAction (Rutten et al., 2002).The program accounts for reduction in variance due to selection (Bulmer, 1971) and corrects selection intensities for finite population size and for the correlation between index values of family members (Meuwissen, 1991).

Phenotypic Values and Environmental Effects
The average phenotypic value of C18:1 in IMF was 44.8%, with an IMF content of 4.9% (Table 1).The effects of batch and age at slaughter on C18:1 are given in Table 2. On average, the variation among batches accounted for 2.4% of C18:1, with a maximum difference between batches of 7.5%.The effect of age at slaughter on C18:1 was small but negative (-0.02%/d).There was not much evidence for the environmental effect of the IMF content on C18:1, with a mean value of 0 but showing a large highest density interval at 95%, ranging from -0.31 to 0.26%/percentage unit of IMF.The environmental effect of carcass weight was positive, with a mean value of 0.02%/kg, with a probability of 95% of being greater than 0.

Genetic Parameters
Estimates of the variance components and heritabilities for BW, BT, IMF, and C18:1, together with the respective genetic and residual correlations among each other, can be seen in Table 3. Specific features concerning the posterior distribution of the heritability of C18:1 and the genetic and phenotypic correlations of C18:1 with BW, BT, and IMF are given in Table 4.The correlation between litter effects in BW and BT was 0.58 (SD 0.02).The heritability for C18:1 was 0.50 (SD 0.08) and similar to that for IMF (0.56, SD 0.09), with a probability of 95% of being greater than 0.37.The genetic and phenotypic correlations of C18:1 with IMF were moderate and positive, with a 95% probability of being greater than 0.27 and 0.29, respectively.The genetic and phenotypic correlations with BW and BT were also all positive, although less, with values in the range of 0.11 to 0.22.Results did not provide conclusive evidence concerning the sign of the genetic correlation between C18:1 and BW, where the associated highest posterior density interval at 95% ranged from -0.10 to 0.31.No substantial deviations in the estimates were observed after adjusting C18:1 for carcass weight or IMF content, or when the ILR-transformed variables u 1 and u 2 were used in the analyses instead of IMF and C18:1 (Table 5).Compared with the reference case, where C18:1 was only adjusted for age at slaughter, the estimates of the heritability of C18:1 after alternatively adjusting C18:1 for carcass weight, age plus IMF content, or carcass weight plus IMF content were only slightly greater,   with a maximum value of 0.55.Similar values were obtained for the differently adjusted genetic correlations of C18:1 with BW, BT, and IMF, except for the correlation between C18:1 adjusted for carcass weight and BW, where, as expected, values decreased to almost 0. When the ILR-transformed variables were used, the maximum change occurred for the genetic correlation between C18:1 and BT, which decreased from 0.22 to 0.18.Because only minor changes were seen across models and data transformation, responses below were calculated using the estimates in Table 3.

Expected Responses
Indirect expected responses in C18:1 to selection for BW, BT, or IMF are given in Table 6.In the first scenario, where records on IMF are not available, at best no change in C18:1 is expected.In most sire lines, the breeding goal is directed at increasing lean growth.According to the emphasis put on each of the 2 components of the trait, the selection objective in these lines can be placed in-between maximizing BW at restrained BT, in one extreme, and minimizing BT at restrained BW, in the other.Thus, within this scenario, the best situation occurs when selection is for BW at restrained BT, in which case only little changes in C18:1 are expected.However, as selection against BT emphasized, C18:1 decreases by up to 0.2% per generation when BW is constrained to remain unchanged.This decrease in C18:1 can be minimized if records on IMF are available.Thus, in this new scenario, if IMF is also restrained, the decrease in C18:1 is reduced 3-fold.Moreover, if IMF is proactively selected, there is room for favorable responses in C18:1.Increasing IMF at restrained BW and BT led to similar but opposite response in C18:1 than decreasing BT at restrained BW.Response in C18:1 can be further improved if it is directly selected (Figure 1).There are selection scenarios leading to favorable responses in all traits; for instance, 1 kg in BW, -0.25 mm in BT, 0.06% in IMF, and 0.25% in C18:1.Maximum expected response in C18:1 at no lean growth loss (i.e., at no change in BW and BT) is 0.40%, with a resulting correlated response in IMF of 0.15%.Increasing the emphasis on BW and against BT constricts the response curves.
The high value for the heritability of C18:1 is maintained even when adjusted for IMF, showing a negligible probability of being less than 0.28.This finding removes the concerns raised by Casellas et al. (2010) about the genetic determinism of C18:1 at fixed IMF.However, this result contrasts with the dramatic reduction, from 0.58 to 0.18, observed by Ntawubizi et al. (2010) for the heritability of C18:1 after adjusting for IMF.These later authors suggested that this might be due to the low IMF content showed by their experimental crossbred pigs (1.2%), a situation where small variations in IMF, mostly dominated by changes in PUFA, may have a great impact.Our results, which were obtained in a population displaying 4-fold greater IMF than theirs, would support this hypothesis.However, note that here, because estimates are based on a 4-trait analysis, with IMF being one of the traits, and not on a series of univariate analyses, the genetic effect is subtracted from IMF when acting as a covariate, then giving as a result a lesser effect of IMF on C18:1.In fact, the effect of IMF on C18:1 was much greater in a 3-trait analysis excluding IMF (0.33, SD 0.04) than in the full 4-trait analysis (0.01, SD 0.14).The heritability of C18:1 in the 3-trait analysis was less (0.45, SD 0.08) but still conclusive with respect to the genetic determination of C18:1.Taken as a whole, the results indicate that C18:1 displays a moderate-tohigh heritability and suggest that there is potential for improving C18:1 in IMF by selection.
Selection responses in C18:1 should be put into context with the correlated genetic change in other economic traits.In this study, C18:1 showed a favorable and moderately high genetic correlation with IMF, in accordance with the observed trend of FA composition with IMF in this line (Bosch et al., 2012), but much greater than that reported by Suzuki et al. (2006), the only other study that examined the genetic relationship between C18:1 and IMF, which was 0.10.Although positive and low, there is less evidence on the magnitude of the genetic correlations of C18:1 with BW and BT, particularly for BW, where negative values cannot be discarded completely.Reported estimates for the correlation between C18:1 and BW are more consistent with the values encountered here than those for the correlation between C18:1 and BT (Suzuki et al., 2006;Ntawubizi et al., 2010).Suzuki et al. (2006) observed that C18:1 and BT are almost uncorrelated, but Ntawubizi et al. (2010) found that they are positively correlated (0.40).Because genetic correlation among C18:1 at different fat depots is approximately 0.7 (Suzuki et al., 2006), complementary information can be retrieved from results on C18:1 in fat tissues other than IMF.Results for backfat C18:1 give a similar contradictory picture; some authors (Cameron, 1990;Fernández et al., 2003) found that C18:1 and BT are hardly correlated (around 0.10), and others reported that they are unfavorably related (Gjerlaug-Enger et al., 2011).Intramuscular fat content showed a similar genetic correlation structure with BW and BT as C18:1, in agreement with previous results in the same Duroc population (Solanes et al., 2009).
Discrepancies in the above estimates may arise because of differences in the age or weight at test and in the muscle where IMF and C18:1 were measured.In the present study, pigs were tested for BW and BT at 180 d, and IMF and C18:1 were determined analytically in the gluteus medius muscle at 205 d. Results in Solanes et al. (2009) showed that the genetic correlation of IMF with BW and BT, both traits measured at 180 d, were greater than those found here for IMF at 205 d.This could indicate that the genetic relationship between performance traits, particularly for BW and IMF-related traits, including C18:1, decreases as age increases.In fact, in heavy Iberian pigs, Fernández et al. (2007) found that the correlation between BW and IMF was negative.This might be interpreted in light of the fact that C18:1 evolved linearly with age throughout the period studied, whereas BW and BT did not (Bosch et al., 2012).The muscle and the determination method of IMF may also influence the relationship among fat depots.Here C18:1 was measured in the gluteus medius instead of LM, as in most reported estimates, because sampling from gluteus medius is easier and cheaper, compared with LM.Muscles behave differently in terms of both IMF content and composition and, because gluteus medius is fatter than LM at a given age (Casellas et al., 2010), IMF in gluteus medius may be more correlated to overall fatness (Solanes et al., 2009).Variations in age, slaughter weight, and IMF content are commonly adjusted including a covariate in the model describing the data.The magnitude of these covariates for C18:1 in the 4-trait analysis was very small, and therefore inferences concerning C18:1 did not relevantly change across models.Major differences occurred when adjusting for carcass weight, likely because, in this case, the covariate is capturing part of the deviations between BW at 180 d and carcass weight at 205 d.Similarly, no relevant changes in the estimates of the genetic parameters were observed after the isometric log-ratio transformation of IMF and C18:1.Note that both IMF and C18:1 are compositional data in nature (Aitchison, 2003), so conceptually they cannot be used in real space unless they are previously transformed (Egozcue et al., 2003).However, Estany et al. (2011), using real and simulated data, have already shown that, in regard to IMF and C18:1, transformed values only performed a little better when predicting future records of IMF.
Data on FA composition have often been obtained from experiments designed for other purposes or from culled pigs, and, therefore, they are not necessarily randomly chosen.In such cases, data may be subjected to selective recording and inferences on genetic parameters may be biased.However, if the history of the selection process is contained in the data used in the analysis, the posterior distribution has the same mathematical form with or without selection (Gianola and Fernando, 1986).In this study, pigs in which IMF and C18:1 were determined were chosen exclusively on the BLUP of the breeding values of BW and BT from the pedigree and records used in the present analysis.All estimates shown here were derived under this principle, and they were implicitly adjusted for selective recording.Inferences obtained using only data from pigs with records on C18:1, although they did not affect the estimate of the heritability of C18:1, underestimated the genetic correlations of C18:1 and IMF with BW and BT, even suggesting a negative genetic relationship of BW with IMF and C18:1 (results not shown).Including all data in the analysis removed the effect of selection and revealed the risks of estimating genetic parameters, particularly correlations, using data recorded for other purposes.
Expected responses suggest that breeding programs directed at increasing C18:1 are feasible but also that this genetic progress is achieved at the expense of decreasing lean content.In many instances, the correlated change in C18:1 to selection for production traits is likely more important than the execution of direct selection.In this scenario, our results show that selection for lean growth will not lead to favorable changes in C18:1, which will only be indirectly improved in breeding regimens selecting proactively for IMF.Some experiments have already demonstrated that it is possible to increase IMF through selection (Suzuki et al., 2005;Schwab et al., 2009).The low expected responses in C18:1 and IMF to selection for BW at restrained BT indicate that, if selection gives a great emphasis on growth rate, little changes in both IMF and C18:1 should be expected.This result is consistent with experimental evidence indicating that continuous selection for lean growth did not necessarily lead to decreased IMF (Oksbjerg et al., 2000;Tribout et al., 2004).
Direct selection for C18:1 allows for convenient scenarios in which C18:1, IMF, BW, and BT can be simultaneously improved.A desired-gain approach was used to determine the weights for traits in the breeding objective.This is a useful approach for traits not yet included in the payment system or subjected to restrictions, as established in some labeled products.In fact, restricted values on FA are a common feature in regulations for foods bearing nutritional or health claims concerning fat properties and, for example, when minimum C18:1 and maximum C16:0, C18:0, and C18:2 contents are required in grading Iberian cured products.However, proper economic weights are needed to achieve the optimum response profile in each situation.It has been proposed to use interviews with experts or market surveys as input for developing a pricing system based on a quantitative differentiation of willingness-to-pay values for carcasses of different qualities (von Rohr et al., 1999).The method has been used in the Swiss breeding program for calculating the economic value of fat quality, indirectly measured as the amount of double bonds in FA in the outer layer of backfat (Hofer et al., 2006).To our knowledge, this is so far the only published attempt to select for fat composition in pigs, although no realized responses have been reported yet.A similar approach can be used to elucidate the economic value of traits, such as C18:1, reflecting possible future trends in the pork market.
Selection for C18:1 leads to an undesired correlated response in BT (i.e., lean content) and to genetic lag in BW (i.e., ADG).Then, for a given scenario, the opportunity cost of selecting for increased C18:1 can be derived by subtracting the total economic response weight in the adopted scenario from the maximum total economic response.Alternatively, in case of being negative, this difference can also be interpreted as an estimation of the societal benefits of selecting for healthiness (Kanis et al., 2005).Other economic traits not included in the present analysis may also show undesired responses.There have not been reported estimates of the genetic correlation of C18:1 in IMF with feed conversion ratio, proportion of premium cuts, or prolificacy.However, results relating to C18:1 (Fernández et al., 2003) and C18:2 (Hofer et al., 2006) in backfat lead to expected unfavorable correlated responses in both feed conversion ratio and proportion of premium cuts, although not to premium pieces weight.By contrast, in accordance with Solanes et al. (2009), who found that IMF was uncorrelated to prolificacy, no relevant genetic change in prolificacy is expected after selection for C18:1.
Genetic differences between individuals for C18:1 in IMF may come from differential ability of pigs either to incorporate dietary C18:1 to IMF or to synthesize C18:1 from C16:0 and C18:0 via increased enzymatic activity of elongases and delta-9 desaturases, respectively.Cánovas et al. (2009) found that selection for decreased BT at restrained IMF led to decreased expression of both enzymes in backfat but not in IMF, giving support to the hypothesis that the metabolic pathways underlying the synthesis of C18:1 are altered by selection.From a practical view, however, the question whether selection for increased C18:1 content is affordable must be contrasted with the cost:benefit ratio of alternative strategies.Diet and age at slaughter, which partly explain the variation among batches for C18:1, are the 2 most used practices to improve both IMF content and composition.However, experimental results indicate that the impact of dietary FA additions mainly affects subcutaneous fat and PUFA rather than IMF and MUFA (Wood et al., 2008).Even though feeding pigs high-oleic acid diets may increase C18:1 in IMF by up to 3% (Mas et al., 2010), this approach has not always been successful (Mas et al., 2011).In general, major changes in C18:1 are achieved indirectly by raising IMF content.Teye et al. (2006), using a low protein diet, and Bosch et al (2012), delaying the age at slaughter, 2 management practices aimed at improving IMF, increased C18:1 by values in the range of 4 to 7%.However, our data indicate that, on average, batch differences only accounted for around 2% of C18:1, approximately the expected genetic change that would be achieved after 5 generations of selection.
A limitation for implementing direct selection for C18:1 is that phenotypes cannot be observed on the selection candidates themselves and are costly to determine.It is difficult to measure C18:1 in live animals unless biopsies (Bosch et al., 2009) or genetic markers (Estellé et al., 2009) are used.However, the first approach is mostly restricted to experimental designs, and the second has not yet been able to translate advances into effective commercial improvements (Dekkers, 2004).The use of increasingly accurate on-line equipment, such as that based on near-infrared spectroscopy (Gjerlaug-Enger et al., 2011;Shackelford et al., 2011), represents an opportunity for systematic recording of C18:1 on the slaughter chain.Due to greater measurement errors, lower heritability values may be expected using such records in relation to analytical methods (Fernández et al., 2003).However, the estimate of the heritability of IMF obtained here is consistent with a previous estimate obtained in the same Duroc population, but using data taken with a near-infrared transmittance spectrometry device (Solanes et al., 2009).Accordingly, no relevant changes should be expected by using on-line measurement technologies.Other direct alternative methods specifically for determining C18:1 content have also been proposed (Muñoz et al., 2011).
Two questions were addressed in this study.i) Is there genetic variation in C18:1 content in IMF? ii) Which response scenarios are expected for indirect and direct selection?We concluded that selection for C18:1 content in IMF can be effective and that there are selection strategies leading to response scenarios in which C18:1, IMF, BT, and BW can be simultaneously improved.However, if adopted, a reduction in the potential for lean growth is also expected.The extent to which it is affordable relies on how much consumers are pre-pared to pay for high-C18:1 pork products.

Table 1 .
Description of the data set used in the analyses

Table 2 .
Features of the posterior distribution of the effect of batch, age at slaughter, and carcass weight on oleic acid content (C18:1) limit for the interval [k, +∞) having a probability of 95%.3 Maximum difference, minimum difference, and SD among batch effects.

Table 4 .
Features of the posterior distribution of the heritability of oleic acid content (C18:1) and the genetic and phenotypic correlations of C18:1 with BW, backfat thickness (BT), and intramuscular fat content (IMF)

Table 6 .
Indirect response per generation in oleic acid content (C18:1) to restricted selection for BW, backfat thickness (BT), or intramuscular fat content (IMF) by availability of IMF records