Growth and carcass composition from birth to maturity in relation to feeding level and sex in Dutch landrace pigs

<p/>An experiment was carried out to study growth from birth to maturity in Dutch Landrace pigs based on complete anatomical dissection. The assessment of a detailed description of the compositional changes during growth was the primary objective of this study. In order to examine whether growth patterns would be influenced by feeding level and sex, six treatment groups were formed, being the combinations of the three sexes (entire males, castrated males and females) and two feeding levels <em>(ad libitum vs</em> a restricted level).<p/>The literature was comprehensively reviewed with regard to relative growth of the main components and their constituent parts. Since the principles by which the changes in the body occur and the laws to which the processes of growth obey originate from a general and basic concept, the studies carried out in other farm animals were discussed as well.<p/>The experiment was set up as a cross-sectional design in which 7 dissection stages were involved. From dissection stage 1 (at an age of 10-12 weeks and about 25 kg live weight) onwards animals wore slaughtered at 6 week intervals up to dissection stage V, followed by an interval of 14 weeks up to dissection stage VI, while the last dissection (stage VII) was carried out at or near maturity. Littermates were allotted to the treatment groups and they were assigned at random beforehand to the different dissection stages to avoid if possible that they should be slaughtered at the same stage. Later on dissection stage 0, involving new-born animals of different weights and of both sexes, was added for completion of a total view over the whole trajectory. The experiment was carried out in 4 replicates that partly ran alongside each other. In principle each dissection stage comprised 8 animals per treatment group. A number of animals had to be withdrawn from the experiment because of leg weakness or various other reasons. Ultimately 283 animals were dissected.<p/>The difference in feeding level was introduced at the actual start of the experiment at dissection stage I. The restricted feeding level imposed was directly related to the <em>ad libitum</em> level, in that restrictedly fed groups in addition to the maintenance requirement received half of the feed for production of the <em>ad libitum</em> fed groups of the same weight and sex. The <em>ad libitum</em> fed animals had free access to feed during the whole 24 hours period. Beyond 225 kg live weight the restricted level was no longer associated to the <em>ad libitum</em> level and was fixed at 3.1 kg per day up till mature weight.<p/>The carcass halves were dissected in different ways. The right half was dissected according to the institute's (I.V.O.) standard method into commercial joints. The left half was used for complete anatomical dissection, except for the head and feet, into individual bones and muscles and into three fat depots (subcutaneous, intermuscular and flare fat). In addition some linear measurements were taken, such as carcass length, leg length, thoracic depth, length and thickness of bones, backfat depth at the midline, while also some meat quality characteristics were measured.<p/>The relative growth of parts, <em>y</em> , to given entities, <em>x</em> , was described by means of the allometric equation <em>y</em> = <em>ax <sup><font size="-1">b</font></SUP></em> in its logarithmic form, where <em>b is</em> the growth coefficient describing growth of <em>y</em> proportionate to <em>x</em> . Since primary attempts in ordering the material had revealed that high power terms would be needed, such terms (up to and including a fourth degree term) were added to the equation. This high power polynomial was ultimately used for regressions through dissection stages I to VII. An indicator variable was incorporated in the model for dissection stage 0. The power terms were orthogonalized on all preceding terms. The <em>b</em> -values presented in various tables then correspond to the <em>b</em> of the linear model, even when a high power term was needed.<p/>A first classification of the growth patterns was achieved according to the high power term needed in each of the six treatment groups. It determined whether sex groups (the combination of the treatment groups of the same sex) or feeding level groups (the combination of the treatment groups with the three different sexes of the same feeding level) could be formed, but with the restriction that the proportion of the residual variance (after taking into account the linear term) for each of the treatment groups at issue had to be 10% at least. When all 6 or 5 treatment groups exhibited a quadratic term, the pattern was considered quadratic irrespective of the proportion of the residual variance. A similar procedure was followed within the linear pattern based on significant differences in slope between treatment groups. The <em>b</em> -values were further classified into classes of 0.05 unit proceeding from an average class (A) between 0.975 and 1.025 to both sides, an upper end with the classes A <sup><font size="-1">+</font></SUP>, H, H <sup><font size="-1">+</font></SUP>and H* and a lower end having A <sup><font size="-1">-</font></SUP>, L, L <sup><font size="-1">-</font></SUP>and L <sup><font size="-1">=</font></SUP>growth patterns. Bone, muscle and fat weight distribution were studied within each of the respective tissues.<p/>Although it will discriminate against a number of significant differences, it is impractical to refer to all discernible results. Therefore of the results obtained only the more general trends and conclusions will be mentioned. This is done point by point and generally following the various headings of the preceding chapter depicted by an extra interruption in the text. In cases where a linear growth pattern is mentioned it means linear according to the logarithmic form of the allometric equation through stages I -VII.<p/>1. Pigs grew well up to dissection stage V which is at about 125 to 165 kg live weight depending on feeding level. The maximum live weight growth for boars and sows was about 825 en 750 g/day respectively between stages II and III, and for castrates about 845 g/day between stages I and II. At the restricted feeding level the maximum growth rate shifted to a further stage and was about 80, 110 and 210 g/day lower than at the <em>ad</em><em>libitum</em> feeding level for the sexes respectively. The feed conversion ratios in general increased with the increase in live weight and were the most favourable at the restricted level as well as in boars, while castrates exhibited the highest ratios.<p/>The weighings every fortnight allowed the drawing of weight-age relationships for a diminishing number of animals. From the Text-figures the inflexion point was estimated at about 30% of mature weight. The growth curves seemed to flatten at about 320 kg in boars and castrates, and also <em>ad libitum</em> fed sows approached this weight, while the restrictedly fed females reached about 260 kg live weight. Remaining single animals of all three sexes reached higher final weights.<p/>2. The carcass grew linearly and at a faster rate than live weight. After birth the dressing percentage decreased at first, but it increased with 10% through stages I-VII. The dressing percentage in general was higher in the fatter animals.<p/>3. The linear A <sup><font size="-1">-</font></SUP>-pattern and the negative cubic term in the males indicate that carcass length relatively diminished with increasing carcass weight. Boars were longer than castrates when related to carcass weight, but at a given muscle + bone weight the reverse was found. <em>Ad libitum</em> fed sows based on carcass weight had a smaller carcass length than restrictedly fed ones, but when based on muscle + bone weight they were longer. This is another example in addition to others mentioned in the literature that the choice of the independent variate may determine the direction of the effect, especially when such a variable factor as fat is involved. This influence of fat was illustrated also in the thorax depth. The higher amount of fat caused that for a given carcass weight restrictedly fed groups had a greater depth than <em>ad libitum</em> fed ones, but when related to carcass length the reverse was found.<p/>4. Thorax depth showed an A <sup><font size="-1">+</font></SUP>or a concave pattern with respect to carcass length. Since furthermore leg length related to carcass length at birth was relatively greater than in later stages, it is clear that during growth animals change in form, they relatively deepen. This was confirmed in that the increase in length of the long bones of the limbs lagged behind to that of carcass length.<p/>5. The composition of the carcass markedly changed during growth. Without reference to quadratic patterns found, common <em>b</em> -values for each of the sexes were established. Castrates and sows resembled each other. In the order of faster growth the values were about 0.70, 0.75, 0.85, 0.90 and 1.35 for the components offal, bone, muscle, skin and fat respectively. Muscle thus grew faster than bone, but both tissues grew slower than the carcass to which they were related. In boars the same order was found, but at a clearly different level: about 0.75, 0.80, 0.90, 1.10 and 1.25 respectively. Only fat increased at a faster rate than the carcass, except the striking higher growth of skin in boars in which this characteristic deviates from growth in males of other species.<p/>Carcass composition also differed with regard to feeding level. The unfavourable influence on the amount of fat after <em>ad libitum</em> feeding disappeared when muscle or bone weight was regressed on muscle + bone weight. For a given muscle + bone weight castrates had more bone than sows. Boars had less muscle than sows due to a higher bone weight, but more muscle when related to carcass weight.<p/>6. The muscle to bone ratio increased during growth, especially in the early stages because of the relatively well-developed skeleton at birth. The ratio is not influenced by feeding level. Sows had the most favourable ratio, followed by boars.<p/>7. The muscle to fat ratio decreased during growth, for castrates and sows to below unity. Castrates had the most unfavourable ratio, followed by sows. <em>Ad libitum</em> feeding unfavourably influenced the ratio.<p/>8. Boars had a heavier head than the other sexes. It may be a secondary sex characteristic as the difference remained after regression on muscle + bone weight. The head at birth had a relatively high percentage, about 18 % of carcass weight, which swiftly decreased and stabilized at about 4 to 5%. Within total offal the head grew faster than the feet, while boars had a relatively lighter head than the other sexes. This may be due to a somewhat less deposition of fat in the head region of boars.<p/>9. The kidneys followed the decreasing pattern of the total of organs and viscera when related to live weight. Within total offal boars differed from the other sexes. Growth of testes was distinctly different from that of kidneys. Initially they grew fast in relation to live weight, because they were relatively undeveloped at birth, and reached mature proportions already in the intermediate stages after the prepubertal growth.<p/>10. The ratios of the left and right testis, kidney and feet revealed that after regression on live weight, the <em>b</em> -value as well as the <em>ln a</em> -value did not deviate from zero, meaning that the ratio not only remained constant but also that the weights were equal, indicating symmetric growth.<p/>11. Bone weight distribution alters during growth. Bone groups hardly changed, but within the groups some bones had an A <sup><font size="-1">+</font></SUP>-pattern: scapula, <em>os coxae</em> , ribs and sternum, while the long bones tended to an A <sup><font size="-1">-</font></SUP>-pattern. Although the feet and the proximal bones of the limbs had distinct <em>b</em> -values growth gradients down the limb were disturbed by the ranking of the long bones, though the differences were small. A somewhat higher growth pattern in the thoracic vertebrae disturbed a growth gradient along the axial skeleton. The growth coefficients of the bones of the thoracic limb tended to be higher than those of the corresponding bones of the pelvic limb.<p/>12. Feeding level and sex influenced bone weight distribution. After <em>ad libitum</em> feeding a higher proportion was found in humerus, radius + ulna, total of pelvic limb bones and ribs, but a lower proportion in scapula and <em>os femoris</em> . The higher proportion in ribs is given a functional meaning. The lower percentage of total bone in the carcass in sows was reflected in the long bones as well as in the total of the bones of both limbs. Sows had a greater proportion of ribs and scapula. The proportion of long bones in the thoracic limb was higher in boars than in castrates, while the opposite was found in those of the pelvic limb.<p/>13. A variable picture was found in the vertebrae. As this might influence the distribution and patterns of other bones a new total bone weight excluding the vertebrae and sternum was tried as the independent variate. Almost nothing changed in the situation and no further insight was gained. The higher variation was ascribed to length of the vertebrae rather than to errors in splitting of the carcass, because in general no systematic differences between weights of the carcass halves were established based on the result of regression of the ratio of both halves on the sum of them.<p/>14. The relative diminishing increase in bone length in later stages led to the conclusion that bone length is earlier maturing than bone weight. Because of a proceeding mineralization also, the weight to length ratio increased. Since the scapula alongside the ribs had lengthened more than other bones, the ratio was lowest in this bone. <em>Ad libitum</em> feeding resulted in higher ratios at a given total bone weight. The radius was relatively thin, but its length kept pace with the others, hence resulting in a length to thickness ratio standing out. Increase in length and thickness of other long bones occured in a regular way. It means that an alleged decreasing length to thickness ratio could not be confirmed.<p/>15. Bone thickness was hardly influenced by feeding level or sex. At a given total bone weight as well as at a given carcass length, however, the bones were longer at the restricted feeding level. Boars had the longest bones when compared at the same carcass length, while at the same total bone weight castrates had the longest bones.<p/>16. In the weight to length ratios the third root of weight was incorporated in the variable. Only in a few cases the 1/3 power of weight was exactly justified. In most cases a weight proportionate to a power of somewhat below 0.4 would have led to independence of total bone weight.<p/>17. Muscle weight distribution also changed during growth. About the same course in muscle groups was exhibited for feeding levels and sexes. The main changes occur during the early stages, already before dissection stage 1. In many cases, therefore, the neonates appeared not to be adapted to the regression lines through the later stages, but they were in line with the fully quadratic positive and negative pattern for the group of muscles of the neck region and that of back and loin respectively. The group of muscles of the proximal thoracic limb showed a positive quadratic term for the restrictedly fed groups and castrates, while castrates also had a negative quadratic term in the group of muscles of the proximal pelvic limb.<p/>18. A mono or diphasic pattern is largely dependent on the new-born stage. When the neonates were taken into consideration together with the patterns already mentioned, a fully linear pattern could only be established for the group of muscles of the distal thoracic limb and for the sublumbar and skin muscles.<p/>19. In most muscle groups some muscles deviate from the pattern of the group; they were found in the group of abdominal and thoracic muscles in particular. But for a number of exceptions it appeared that more deeply situated muscles in many cases exhibited a lower growth pattern, whereas the more superficially situated ones revealed a higher growth pattern. Muscles in an intermediate position showed an average pattern. Disto-proximal and ventrodorsal growth gradients may interfere in this respect.<p/>20. It is difficult to relate growth of individual bones to that of the muscles surrounding them, because often muscles act over more than one bone. Only a rough relationship existed for regions of the body according to the growth gradients. In accordance with a somewhat higher pattern of the bones of the thoracic limb with respect to that of the pelvic limb, the muscles in the regions did. But the higher growth pattern of the group of neck muscles was not found for the cervical vertebrae.<p/>21. A disto-proximal growth gradient was established in both limbs, proceeding via the sublumbar muscles to the dorsal region, and via the shoulder girdle region to the neck and to the dorsal region too. Furthermore a gradient over the trunk from ventral to dorsal completed the picture. An anterior-posterior gradient was not found.<p/>22. Upon maturing a masculinization process occurred in that the group of muscles of the neck region relatively increased together with some anteriorly situated muscles of the shoulder girdle region. Pronounced examples were found in <em>m. splenius</em> and <em>m. rhomboideus,</em> both bearing a positive cubic term in boars. Castrates were intermediate in this respect. The higher proportion of long bones in the thoracic limb in boars is in accordance with this masculinization process.<p/>23. At birth the relatively heavy head has to be borne which is made possible by a higher proportion of the group of neck muscles to which <em>m. splenius</em> contributed in particular.<p/>24. The most stable group of muscles is that of the proximal thoracic limb. Apart from the average growth pattern it is not influenced by feeding level nor by sex. Feeding level influenced the distribution in other groups in that ad <em>fibitum</em> feeding resulted in a higher proportion of the groups of muscles of the trunk (abdomen, thorax, back and loin, skin muscles), whereas a lower proportion was found in those of the neck region and the limbs including the sublumbar muscles. A functional meaning was given to the higher proportion of the abdominal and thoracic muscle group, while the relatively lighter head might have contributed to the lower proportion of muscles of the neck region.<p/>25. In most muscle groups a sex influence was discernible, mainly due to a difference between boars and the other sexes. The boars had a higher proportion in the muscles of the neck, proceeding to the group of muscles of the shoulder girdle region and in that of the thorax, including the skin muscles. The boars had a lower proportion in the groups of expensive muscles.<p/>26. But for some exceptions the influence of feeding level and sex on the individual muscles largely followed that of the standard muscle group.<p/>27. Fat weight distribution changed during growth as well. Subcutaneous fat and flare fat increased, whereas intermuscular fat decreased. Flare fat had the highest growth pattern.<p/>28. Feeding level hardly influenced the fat weight distribution. <em>Ad libitum</em> feeding resulted in more subcutaneous fat in castrates, and less intermuscular fat in boars and sows.<p/>29. Clear influences of sex on fat weight distribution were established. Boars contained less subcutaneous fat. This is also valid for castrates at later stages with respect to sows. Boars had more intermuscular fat than castrates, but because of the opposite high power terms they meet again at later stages. Boars also produced more flare fat than the other sexes.<p/>30. Backfat thickness as related to total side fat weight showed a concave pattern and increased most in earlier stages. <em>Ad libitum</em> feeding rendered a thicker fat layer. The backfat depth was smaller in boars than in the other sexes, while also sows had a smaller fat depth than castrates in a number of cases.<p/>31. An anterior to posterior growth gradient in backfat depth could not be ascertained, because it would be disturbed in the intermediate locations. Fat depth over the shoulder followed most closely that of total fat weight.<p/>32. There were hardly any changes discernible in the meat quality characteristics measured, while for feeding level and sex no influence was found.<p/>33. In many cases growth patterns of part to whole relationships had to be described by means of high power terms. Unless one does accept unn