Production efficiency of fattening pigs in terms of average daily gain, feed efficiency and carcass composition is a heritable trait complex (Aymerich et al., 2019; Cámara et al., 2016; Latorre et al., 2003; Serrano et al., 2008), for which considerable heterosis effects are harnessed in crossbreeding schemes (García-Casco et al., 2012; Katoele et al., 2002; Luo et al., 2018). In commercial pig production, the dam line is usually fixed, and as a consequence of artificial insemination (AI) sire line selection is flexible. According to the BMEL (2019), currently nine pig breeding companies are operating in the European Union. The fact that each company improves and offers at least one or two basic sire lines, and selects these in different ways, leads to a variety of sire lines. This results in a complex choice for the pig producers.

Individual piglet birth weight is important for subsequent growth performance (Fix et al., 2010; Vázquez-Gómez et al., 2020). Sire line impacted piglet birth weight (Vermeulen et al., 2016); hence, a selection of purebred sires on birth weight is useful (Dufrasne et al., 2013), and the importance of sire lines in breeding programs is increasing (Dodenhoff et al., 2019). Regarding gender, controversial results with respect to growth performance can be found (Franco et al., 2014; Lee et al., 2019; Morales et al., 2011). Hence, a closer examination of sire line effects on offspring birth weight and of the implications of gender for further performance is needed.

Several studies have analyzed offspring performance in the period of fattening with various sire lines, in terms of variables such as feed composition, different breeding values and selection strategies, or various possibilities of male castration (Cámara et al., 2016; Cilla et al., 2006; De Cuyper et al., 2019; Gilleland et al., 2019; Lowell et al., 2019; Morales et al., 2011). Against this background, an integrated view on productivity from birth to finishing seems necessary. Hence, in the current study, different sire lines where mated to the same sow genotype in an integrated production system. The aim was to investigate the effect of sire line, birth weight, and gender on growth performance, carcass and meat quality traits.

The descriptive statistics of all investigated traits according the sire line are in Table 3. A total of 73 pens were used to calculate growth performance. ADGnet and carcass traits were based on 3437 carcasses. Overall, 672 pH45 min and 1895 pH24 h measurements were collected. The LSQ means according to the sire line are reported in Table 4. Offspring of both sire lines exhibited the same BWinitial (p=0.743). During the fattening phase, pigs sired by line A grew 48.8 g d-1 faster (p<0.01) and reached 16.5 g d-1 higher ADGnet (p<0.001) as compared to pigs sired by line B. A 60 g d-1 higher ADFI (p<0.01), but the same feed efficiency as sire line B (p=0.179) was detected at sire line A. These offspring were 3.3 kg heavier in BWfinal (p<0.01). Based on a 90.0 kg weight gain at finishing, 6 fewer fatting days were needed. Regarding the carcass traits, HCW was 1.3 kg higher for offspring of sire line A based on the same age at slaughtering. Sire line A had 340 g more BE (p<0.001) but no difference in FT (p=0.423) compared to pigs from sire line B. Pigs sired by line B had a leaner carcass with heavier primal cuts (p<0.001). No difference between sire lines (p=0.061) was detected for pH45 min, but offspring sired by line B had a lower pH24 h (p<0.001).

Males grew faster (ADG: 883.37 vs. 847.83 g d-1, p≤0.05; ADGnet: 532.64 vs. 513.11 g d-1, p<0.001) and had a higher feed intake (ADFI: 2.31 vs. 2.16 kg d-1, p<0.001) but were less efficient (FCR: 2.62 vs. 2.51 kg kg-1, p<0.01) compared to females. Consequently, males reached a higher BWfinal (115.56 vs. 111.80, p<0.001) and a 2.05 kg heavier HCW (p<0.001). Females had a better carcass value and were 3.04 % leaner with 2.39 mm thicker LD and 396 g less BE (p<0.001) but had heavier primal cuts (p<0.001). Neither pH45 min nor pH24 h were affected by sex (p≥0.05). The LSQ means of the significant interactions between sire line and sex are presented in Table 5. Males and females sired by line A did not differ in ADGnet and ADG (p≥0.05). In sire line B a faster growth (p<0.01) and higher feed intake (p<0.01) were detected for males compared to females. HCW of males sired by this line did not differ from males and females of sire line A. Females of sire line B were lighter in HCW compared to all other sex and genotype variations. Regardless of sire line, similar ADGnet, ADG, ADFI, HCW, LM, FT, BLM and BSH between males was observed (p≥0.05). Females sired from line B reached heavier BH, LO, BLO and BSH compared to the other sex and genotype variations.

Growth rate and carcass traits according to the birth weight classes are presented in Table 6. Low birth weight pigs (BWQ1) had the lowest ADGnet compared to pigs of the middle (BWQ2) and the heaviest (BWQ3) birth weight classes. Pigs in the BWQ3 reached the fastest growth rate (p<0.001). Moreover, pigs HCW differed correspondingly among the birth weight groups (p<0.001). Pigs in BWQ1 were 4.66 kg lighter compared to pigs of BWQ2 and 5.82 kg lighter than pigs belonging to BWQ3. Fatter carcasses based on a thicker FT and a heavier BE, as well as a lower percentage in LM and BLM, were found for pigs of BWQ1 but not for pigs of BWQ2 and BWQ3 (p<0.001). Furthermore, birth weight was positively related to the weight of BH, LO, BLO and BSH, and a difference was detected among the BWQ (p<0.001). The interaction between BWQ × sire line was significant for LD (p<0.01), in sire line A, similar LD was detected for pigs of BWQ2 and BWQ3 (62.79 vs. 62.98 mm, p≥0.05), whereby all other BWQ and genotype variations differed (p<0.001).

Higher ADG was observed for pigs sired by sire line A, which is a reason for the heavier BWfinal at a fixed slaughter age and may enable more pigs per fattening place and year. This is in line with several other studies applying the same approach (Aymerich et al., 2019; Cámara et al., 2016; Lee et al., 2019; Serrano et al., 2008), in which differences between 2 and 6 kg for BWfinal (p≤0.05) and 41 and 152 g d-1 for ADG (p≤0.05) among crossbred pigs were observed. It is to be expected that the higher ADFI for pigs sired by line A is linked to the faster growth rate of these offspring. This assumption is generally in line with Kavlak and Uimari (2019) and Labroue et al. (1997), who reported a positive correlation between feed consumption per visit and growth rate. A greater emphasis on growth in line A in competition with line B is presumed. As a consequence of faster growth, a heavier belly weight and a fatter carcass of crossbreeds sired by line A were detected. According to Kavlak and Uimari (2019), a positive genetic correlation between ADFI and back fat thickness (r=0.60–0.72) does exist. Consequently, pigs consuming more feed tended to gain more fat. A limited protein disposition capacity of pigs sired by line A with respect to the high feed intake could provide an explanation for the fat accumulation in the carcass and is generally in agreement with finding of Hermesch et al. (2000). Additionally, a higher ADFI is linked with a poorer feed efficiency (Godinho et al., 2018a). Feed efficiency was equal for both lines (p=0.179), which indicates comparable energy demands for maintenance. But in consideration of the difference in carcass composition it is to expect that the protein requirements differ between the sire lines. In particular, it is well known that growth-emphasized genotypes had a higher protein requirement and need a closer energy–protein ratio to exploit their growth potential (Emmans and Kyriazakis, 1997; Quiniou et al., 1996; Whittemore et al., 2001). Metabolic differences in protein disposition and energy metabolisms between the sire lines could be an explanation for the findings of the current study. Saintilan et al. (2015) reported 13 % higher average digestible lysine requirements for the 25 % most efficient pigs compared to the 25 % least efficient pigs. Furthermore, the effect of sire line might be attributable to differences in additive genetic variance and thus heritability between lines, leading to different breeding progress. It can be assumed that the genetic progress realized under the diet used on the test farm is not completely transferable when using other diets, like in the current trial. In this context, Godinho et al. (2018b, c) reported impairment of lipid disposition and residual energy intake in pigs on different diets due to genotype by feed interactions, which could be a reason for the observed differences in growth performance and carcass value. Moreover, findings of Škorput and Luković (2018) suggest a range of additive genetic variance for ADG from 586.09±55.73 g d-1 in the Large White population to 1190.03±139 g d-1 in Piétrain populations.

As expected, heavier HCW was detected for sire line A due to faster growth rate. This is in line with Miar et al. (2014), who reported a strong correlation between growth rate and HCW (r=0.75±0.28; p≤0.05). According to Dufrasne et al. (2013) and Zumbach et al. (2007), with 2.03 % and 6.7 %, respectively, sire line explained only low proportions of genetic variance for final live body weight. It is even known that heritability ranges between 0.15 and 0.39 depending on the estimation method and the population (Dufrasne et al., 2014; Fragomeni et al., 2016; Holl et al., 2008; Khanal et al., 2019; Miar et al., 2014). Against expectations, the FT was equal in both sire lines. However, the ratio of LD and FT in consideration of sire line suggested a lower FT, which is resulting from a higher lean growth potential in pigs sired by line B. Additionally, the current results are in line with van Wijk et al. (2005), who reported that selection for higher growth led to effects decreasing most primal and sub-primal cut weights. Selection for leanness in Piétrain sire has resulted in reduced feed intake capacity (Labroue et al., 1999; Lean et al., 1972). Hence, the leaner carcass with lighter BE is linked to a higher muscle growth potential of pigs sired by sire line B. Findings of Correa et al. (2008) and Lowell et al. (2019) support these results. However, as discussed above, a higher protein requirement in combination with an inappropriate energy level could also be an explanation for the poorer carcass value of the offspring sired by line A. Furthermore, based on a nonlinear association between daily growth and muscle growth, pigs with more than 60 % lean meat tend to grow slower (Argemí-Armengol et al., 2019), which is consistent with the results of the current study.

Males reached a higher BWfinal at the same slaughtering age as compared to females, which is linked to 35.5 g d-1 faster growth rate. Additionally, FCR was better by 0.11 in females, which agrees with the findings of Serrano et al. (2008) and Renaudeau et al. (2006), but is in contrast to Lee et al. (2019) and Peinado et al. (2008). The poorer FCR of males is related to the moderately faster growth rate and the 160 g d-1 higher ADFI. The PDmax is the point of maximum protein accretion of a pig and up to this plateau, protein growth responds linearly to feed intake (Campbell et al., 1985; Whittemore and Fawcett, 1976). Consequently, pigs with a high feed intake reach this point earlier and tend to accumulate fat tissue as the PDmax is exceeded (Lewis and Lee Southern, 2000; Tullis, 1982). This might explain the fatter carcasses of male pigs in the current study. Generally, it is expected that females have a higher ratio of protein accretion relative to lipid accretion, and this is in agreement with Noblet et al. (1999), Schinckel et al. (2008) and Serrano et al. (2008). Further insight into biological differences between the sire lines can be drawn from the significant interactions between sire line and sex, which were also found by Lowell et al. (2019) and Morales et al. (2011). Generally, these interactions clearly indicate that pigs sired by line A grew faster than pigs sired by line B and that pigs from sire line B showed a leaner carcass with heavier primal cuts compared to the other sire line. Similar growth rate (ADGnet, ADG) of males and females sired by line A suggested a more homogenous growth performance in offspring of this sire line. However, the differences in carcass traits between males and females were smaller in sire line A and supported the homogenous growth of this line. Moreover, only females of sire line B differed in growth rate (ADGnet, ADG) compared to all other variations, which is explained by the significant lower ADFI in comparison with males regardless of the sire line. Neither a difference in growth rate (ADGnet, ADG) nor in ADFI between both male groups was detected, but females differed significantly. Regarding a different protein requirement between the sire lines, an explanation could be that the females sired by line A better exploited their growth potential compared to the males of the same sire line, despite the fact that the energy–protein ratio does not seem to be optimal for this genotype. Findings by Chen et al. (1999) and Hansen and Lewis (1993) indicated different responses with deficient or excessive protein intake between castrated males and females, and Martins et al. (2012) reported differences in protein requirements between different genotypes. Hence, a better adaptability to the available level of protein and energy content of females sired by line A in contrast to the males of the same genotype is an explanation for the interactions between sire line and sex regarding the growth rate.

Regarding the birth weight, the ADGnet increased with increasing BWQ and indicated that heavier birth weight pigs grow faster with improving subsequent performance, which is in line with Fix et al. (2010) and Vázquez-Gómez et al. (2020). However, Douglas et al. (2013) reported that birth weight had less impact on subsequent growth performance compared to weaning weight or initial weight in the finisher barn. Nevertheless, in the current study fatter carcasses in pigs of the BWQ1 and BWQ2 compared to heavier birth weight pigs of BWQ3 were detected and are in agreement with the findings of Nissen et al. (2004) and Rehfeldt and Kuhn (2006). A decreasing proportion of ham, loin and belly in the lightest birth weight classes suggested disadvantages in growth performance, in terms of less ADG, lower ADFI and a poorer FCR. These findings were generally in line with Gondret et al. (2005) and Rehfeldt et al. (2008). However, lower belly weight in pigs with higher birth weight may be explained by further selection for higher proportion of lean meat. In agreement with Kim et al. (2017), stronger selection for lean meat is linked to less fat disposition in belly muscle and leads to lower belly weight. Moreover, no interactions between sire line and BWQ indicated that offspring of both sire lines show the same biological relation between birth weight and carcass traits, except for LD. Hence, similar LD between the BWQ2 and the BWQ3 class in sire line A (p≥0.05) indicated that an additional gain of ≥1.70 kg birth weight would not lead to a further meat gain within this genetic line. Consequently, biological limit in this trait seems to be reached, and further selection in sire line A on birth weight would not increase the LD. According to Beaulieu et al. (2010) and Gondret et al. (2005), it is not unexpected that carcass leanness is not primarily influenced by birth weight. Other factors, such as initial and final body weight, the level of feed intake, growth rate, feed processing and delivery, and genetic potential were also involved (Patience et al., 2015). This interaction supported a difference in protein and energy metabolism between the sire lines.

Based on the findings of Latorre et al. (2003), Lloveras et al. (2008), Lowell et al.  (2019) and Miller et al. (2000), pH values in this study were generally within normal range. Unlike Miller et al. (2000) and Latorre et al. (2003) but in agreement with Kim et al. (2017) and Fecke (2013), pH45 min was not affected by genotype. However, pigs sired by sire line B showed lower pH24 h, indicating poorer meat quality. Drip losses are moderately to strongly negatively correlated with the texture of loin (-0.23±0.06), belly weight (-0.46±0.14) and pH24 h (-0.99±0.49) (p≤0.05) (Miar et al., 2014). Similar results were reported by Kim et al. (2017) and Otto et al. (2006). Consequently, poorer water retention capacity among offspring of sire line B seems likely, and the leaner carcasses of this offspring back up this result. It suggests that higher leanness in line B and in females means poorer meat quality.

This research confirms differences in growth performance and carcass characteristics between crossbreeds of two different sire lines. Pigs sired by line A grow faster and had higher ADFI and the same FCR compared to offspring of line B. This resulted in a fatter carcass and heavier belly weights, which indicated a better meat quality due to a higher pH24 h in crossbreeds sired by line A. Leaner carcasses with heavier primal cuts were detected at offspring of line B and at heavier birth weight pigs. Males grew faster and reached a heavier but also fatter carcass. Regardless of sire line, piglets with heavier birth weight were leaner and had heavier primal and boneless primal weights. Except for LD, no interactions between sire line and BWQ were detected, indicating subsequent growth was equal in both lines, depending on birth weight. The detected interaction for LD between sire line and BWQ as well as sire line by sex interactions indicated different protein requirements between the genotypes. In conclusion, in a fatting production system, paternal line A is preferable if a producer wishes to increase the number of fatting pigs per fattening place and year. Additionally, in the dam line, emphasis should be put on leanness. Breeding with paternal line B is a good option for a finisher producer if the goal is to increase leanness and carcass primal cuts.