501 Crossbreeding experiment with mixed semen of boars of different breeds Elsi Ettala Agricultural Research Centre, Department of Animal Husbandry Received September 17, 1973 Abstract. Sows of the Finnish Landrace and Yorkshire breeds (L and Y) were artificially inseminated with mixed semen of Finnish Landrace and Yorkshire boars. Purebred and crossbred progeny were identified by blood group testing. The pairs of boars were selected to be as identical as possible in regard to semen quality. Mixing of the two diluted semen fractions did not affect unfavourably the viability and motility of the sperm. Normal conception rates and litter sizes were obtained. The progeny fell into groups as follows: LL 34, YL 28, YY 24, and LY 20 piglets. The distri- bution, however, varied greatly within the various litters. One litter consisted merely of purebreds. The sires were successfully determined by blood group testing in 95 % of the Cases. The crossbred piglets grew very significantly (P < 0.001) faster than the purebreds. They also were more vital than the purebreds. In carcass characteristics the purebreds and crossbreds were either similar to each other or the crossbreds represented intermediate forms between the two parent races. Purebred and crossbred piglets born in the same litters were compared with each other in crossbreeding experiments in the 1920’s and 1930’ s (Lush et al. 1939, Roberts and Carroll 1939). The young were produced by mating sows with two boars of different breeds in the same heat period. Such races were used which made it possible to distinguish between purebred and crossbred progeny on the basis of colour. Such races were Poland China, Duroc Jersey and Yorkshire. Now that instead of natural service sows can be inseminated artificially and the progeny identified by blood group testing, the method has become even more applicable. It can be extended to all races available, and the fact that AI allows semen to be mixed eliminates the effects of order of service and intervals between the services. In this kind of crossbreeding experiment where the purebreds and crossbreds have the same dams and shared foetal and litter periods, the differences due to dams are reduced to a minimum. It can be expected therefore that by this method crossbreeding studies could be carried out with smaller numbers of animals than before without impairing the reliability of the results. Since https://www.c-info.fi/en/info/?token=WQ8BOOntVGG5XhcY.J2juK9fS3dkrcdBWOmOc_A.fx0pnQ92BBCtSeIZl3N8JoOFovXFYKpgfHjy5Jk42z-V5L9pilzouU2YBbyijMFwcY318ub1JkPhGtxaziFWpu4FfLeV8LbDmdquH8fvXNw4VLHSZQ3jyAfb5zsrrrB-XrBlSoB1lX_mUe0ZQV_kT6oRFjDsRocLKw 502 interest in hybrid breeding is steadily increasing, there are good reasons for studying the usability of this method. Material and methods Pure Landrace and Yorkshire sows were inseminated with mixed semen of Landrace and Yorkshire boars. The aim was to obtain four different groups of progeny: 1) purebred Landrace (LL) 2) crossbred Yorkshire boar x Landrace sow (YL) 3) purebred Yorkshire (YY) 4) crossbred Landrace boar x Yorkshire sow (LY) The pairs of boars were selected from animals in whose semen, according to previous experience, sperm density, motility and viability were approximately equal. The collections of semen were arranged one after the other to ensure that the age of both fractions was as similar as possible. Dilution was carried out with hen’s egg dilute. The ratio of dilution was determined by the semen quality and was mostly 1:3 or 1:4. The two semen fractions were mixed together immediately after dilution. The vitality and motility of the sperm was closely observed even after a portion of the semen had been dispatched to the inseminators. The artificial inseminations took place on contract farms with three Landrace and four Yorkshire herds. A total of 21 inseminations were performed. Semen of eight Landrace and eight Yorkshire boars was used Fig. 1. Purebred (YY) and crossbred (LY) progeny of a Yorkshire sow. 503 The piglets farrowed were immediately earmarked and weighed when three weeks old. A veterinary surgeon took blood samples of the progeny when they weighed 15—20 kg and of their parents. Blood group testing was carried out in the Blood Group Laboratory of the Federation of the Artificial Insemination Societies (Keinosiemennysyhdistysten liitto). An attempt was also made to visually distinguish between the purebreds and crossbreds by external iden- tification marks such as the shape, position and size of ears (Fig. 1). When the pigs weighed ca. 20 kg, they were all taken to the same pig-rearing unit where each particular litter was reared in one pen. They were given commer- cial multiple feeds. Feed I with 16.0 % digestible crude protein was used until the liveweight of 30 kg was reached and feed II with 14.0 % digestible crude protein thereafter. Rationing was performed according to the average weekly litterweights using the feeding norms shown in Table 1. Table 1. Feeding standards for the experimental pigs. Pig weight, kg FU/pig/day 20.01.0 25.01.2 30.01.3 35.01.5 40.01.7 45.01.9 50.02.1 55.02.3 60.02.5 65.02.7 70.02.8 75.03.0 80.03.2 85.03.3 The pigs were sent for slaughter when they reached the liveweight of 88 kg. Dispatches were made once weekly. The pigs were weighed when sent for slaughter (liveweight) and at the slaughterhouse as cold carcasses without lard and kidneys (slaughterweight). Half-carcasses were cut according to the normal Finnish system (Uusisalmi 1969, Fig. 2). The statistical analysis was carried out with an IBM 1130 computer. The effects of variation in initial weight, initial age and final weight upon growth during the experimental period, and the effect of carcass weight upon carcass characteristics were eliminated by least squares variance analysis (Harvey 1966) using linear regressions. The general statistical model was as follows: y = a o + at + • • • + a n + bjXj + . . . + b mxm + e in which y = dependent variable, a 0 = constant, a x ... an = class variables, bj ... b m = regression coefficients, Xj ... xm = regression variables, and e = normally distributed random error with the expected average of O. The effect of sex was eliminated by pooling the data for gilts and barrows. Dif- ferences between means were tested by the Tukey’s test (Steel and Torrie 1960). Results Insemination results After combination of the two semen fractions nothing extraordinary was noted in terms of sperm vitality and motility. The percentage of conception was 67. The number of piglets per litter varied from 2 to 17. The two smallest litters were omitted from the experiment which thus came to comprise 12 litters, six by Landrace sows and six by Yorkshire sows (Table 2). In these the average numbers at birth were 11.8 live piglets and 1.3 stillbirths, at the age of three weeks 9.5 piglets. The average litterweight at three weeks was 53.6 kg. Table 2. Distribution of purebreds and crossbreds within the litters according to blood group testing. Number of piglets , ... Herd book nos.Litter of sires LL YL YY LY Unsolved Total 1 8246 + 2701 7 3 - - 1 11 2 8244 + 2250 7 1 - - - 8 3 8277 + 2431 9 2 - - - 11 4 8337 + 2774 4 6 - - - 10 5 8342 + 2001 3 10 - - - 13 6 8342 4- 2351 4 6 - - - 10 7 8246 4- 2701 - - 4 5 1 10 8 8340 4- 2351 - - 1 8 - 9 9 8641 4- 2353 - - 4 - 3 7 10 8342 4- 2001 - 6 2 8 11 8244 4- 2250 - - 7 - - 7 12 8698 4- 2773 - - 2 5 1 8 Total 34 28 24 20 6 112 Fig. 2. Cuts of porcine carcass (Uusisai.mx 1969). The most valuable parts: 1 ham 2 = loin 3 = back 4 = foreback 5 = shoulder 504 505 The distribution of purebred and crossbred piglets within the litters varied greatly (Table 2). Higher conception rate was in some cases achie/ed with the boar of the same race as the sow, in other cases with the different one. When on three occasions the same pair of boars were used on both Yorkshire and Landrace sows (1 and 7, 2 and 11, 5 and 10), a better result was in two cases obtained with the boar of the same race. One litter consisted of purebred progeny only, and in another, purebreds only could be determined. Total distribution was as follows: LL 34, YL 28, YY 24, and LY 20 piglets. Blood group testing was successful in identifying the sires of 94.7 % of the piglets (Table 2). Visual identification on the basis of external appearance was successful in 74.3 % of the cases. Rearing results Viability of the piglets came on trial during the rearing period due to an attack of pneumonia caused by the bacterium Pasteurella multocida. The losses in the experimental period were as follows: Paternity Purebreds Crossbreds , ,unsolved Deaths during experimental period 6 1 1 Sent for slaughter before reaching normal weight ... 5 Excluding all those animals which succumbed to illness or were un- identified, as well as those litters (Nos. 9 and 11) which did not contain cross- breds, growth results (Table 3) were obtained for 41 purebreds and 47 cross- breds. The crossbreds grew significantly faster than the purebreds (difference in the experimental period 76.8 g/day***). There were no weight differences at the age of three weeks nor at the start of the experiment. In the Landrace sow group the crossbred progeny increased their growth rate remarkably from the fourth week of the experiment (Fig. 3). Both of the two groups of purebreds grew at an almost equal rate and both of the crossbred groups faster than the purebreds (Fig. 3). In the litters of the Landrace sows the difference became significant from the sixth week of the experiment; in the litters of the Yorkshire sows it was significant in the seventh and eighth week only. Expressed in weight units the difference was larger in the latter group than in the former, but due to the small number of piglets in the YY group and the relatively wide dispersion of the results within the group (Table 3), the statistical significance of the results remained low (Fig. 3). It was not possible to determine the feed conversion efficiency separately for the purebreds and the crossbreds since all pigs belonging to the same litter were reared together in the same pen. An indication of this can be obtained by comparing the feed consumption of the various litters in view of the relative numbers of purebreds and crossbreds in them. The comparison reveals that on the average the crossbreds were slightly more efficient feed consumers than 506 Table 3. Growth data for purebred and crossbred pigs. Purebreds Crossbreds IX YL YY I.Y ~ Differences „„ . „ n = 41 n = 47 n = 29 n = 28 n = 12 n = 19 Characteristics purebreds- , , crossbreds ,, , . mean s.d. mean s.d. mean s.d, mean s.d. mean s.d, mean s.d. Weight at 3 weeks, kg 5.8 1.1 5.6 1.7 + 0.2 5.5 0.8 5.1 1.4 6.3 1.6 6.2 2.1 Initial weight, kg 21.8 3.5 21.1 4.8 + 0.7 22.2 3.5 20.8 3.7 20.7 3.6 21.6 6.2 Final weight, kg 89.3 3.5 89.7 3.4-0.4 89.2 2.7 89.8 3.6 89.4 5.2 89.6 2.9 Initial age, days 62.9 5.6 65.2 5.7-2.2 63.5 4.4 64.6 4.3 61.5 8.1 66.0 7.4 Final age. days 172.1 10.6 163.9 11.1 + B.2** 173. 5 a 11.3 163. 8 b 11.7 168. sab5 ab 8.4 163.9 b 10.1 Experimental period, days 109.1 11.9 98.7 10.9 +lo.4*** 110.0° 12.2 99.2 d 9.7 107.0cd 11.2 98. 0 d 12.5 Growth in experimental period, g/day ...624.6 69.6 701.4 68.0 -76.8*** 614.9° 62.3 702.5 d 65.4 648.2° d 86.1 699.8 d 71.9 Growth in lifetime, g/day 521.2 42.8 550.5 41.9 -29.3*** 516.6 ac 38.0 552.0d 44.5 532.3 abcd 53.7 548.4b 37.7 Statistical treatment by least squares analysis with initial weight, initial age and final weight as regression variables for the analyses of experimental period and qrowth rate in the experimental period, and final weight as regression variable for the analyses of final age and growth rate in the lifetime. * P < 0.05, •• P < 0.01, *** P < 0.001. The differences between group meanstreated by Tukey’s test, a—b: P < 0.05, c —d: P < 0.01. Data for barrows and gilts combined by pooling. 507 Weight differences kg YL-LL LY-YY LY-YX, initial weight —1,4 +0.9 +O.B Ist-week weight —1.6 +0.7 +l.B* 2nd- » • -1.6 +1.6 +2.6* 3rd- » » -1.6 +2.0 +3.5»* 4th- » • -0.7 +2.8 +3.5 sth- * » +0.4 +3.7 +2.8 6th- » * +l.s** +4.3 +1.7 7th- » » +l.B** +5.4* +2.9 Bth- » * +3.7»* +6.2» +1.7 9th- » * +4,3** +5.9 +0.7 10th- * * +4.4** +6.5 +1.9 11th- » » +s.4** + 6.2 +l,B :2th- * » +s.9** +5.3 +l.O Fig. 3. Average liveweights of purebred and crossbred piglets and the significance of weight differences up to the 12th week when the first experimental animals were submitted for slaughter. 508 the purebreds (Table 4). The average feed efficiency of all the test animals during the experimental period was 3.21 FU/kg of weight gain. Table 4. Feed conversion efficiency of pigs of various litters, a) litters with more than 50 % purebreds, b) litters with more than 50 % crossbreds. a b No. of pigs No. of pigs FU/kg of FU/kg of pure- cross- growth pure- cross- growth breds breds breds breds 9 2 3.90 1 8 2.94 7 1 3.39 4 6 2.98 6 2 3.36 4 5 3.05 7 3 3.13 3 10 3.12 4 6 3.14 - - . 2 5 3.16 29 8 3.45 18 40 3.07 Carcass characteristics Carcass section measurements and lean cuts analyses were obtained for 72 animals. Of these 34 were purebreds and 38 crossbreds (Table 5). The table shows that in all groups the carcass characteristics were quite similar. The only statistically significant differences between purebreds and crossbreds were found in weight distribution between the fore- and rear part of the car- cass and in the fat -f- skin fraction of the ham-loin section. The amount of meat -f- bone produced per day was significantly larger for the crossbreds (94.1 g/animal/day) than for the purebreds (88.5 g) but the difference was a result of faster growth rather than an indication of a larger amount of meat + bone in the carcass. Differences were found between the purebred groups in the amount of meat + bone in the back section as well as in side length. The crossbreds inherited these characteristics in intermediate form. Discussion The use of mixed semen of boars of different races in crossbreeding experi- ments proved practicable in the sense that it was possible to mix the semen without complications and to achieve normal conception rates and litter sizes. On the other hand, the great variability of the distribution of purebreds and crossbreds within the litters was a factor reducing the practicability of this method as compared with normal crossbreeding experiments with separate litters. The same difficulty emerged, even more clearly, in the earlier investiga- tions. In the work of Roberts and Carroll (1939) nearly 40 % of the litters Table 5. Carcass characteristics of purebreds and crossbreds. Lean cuts from half-carcasses. +■ Purebreds Crossbreds LL YL YY LY Differences „. n = 34 n = 38 n = 24 n = 23 n = 10 n = 15 Characteristics purebreds- crossbreds mean s.d. mean s.d. mean s.d. mean s.d. mean s.d. mean s.d. Liveweight, kg 89.2 3.5 89.3 3.0-0.1 89.4 2.9 89.3 2.8 88.8 5.1 89.4 3.2 Carcass weight, kg 65.4 2.6 65.5 2.8-0.1 65.5 2.4 65.6 2.6 65.2 3.1 65.4 3.3 Slaughter loss, % 26.6 2.4 26.7 1.7-0.1 26.7 2.0 26.6 1.6 26.4 3.1 26.9 2.0 Half-carcass, kg 32.9 1.4 33.1 1.6-0.2 33.0 1.3 33.2 1.4 32.8 1.6 32.9 1.9 foreparts, kg 19.9 0.9 20.4 1.1-0.5*» 19.8C 0.8 20.6 d 1.0 20.1cd 1.2 20.1 cd 1.1 rear parts, kg 13.0 0.7 12.6 0.7 + o.4** 13.2° 0.7 12.6 d 0.7 12.7 cd 0.6 12.8cd 0.9 Main parts of half-carcass, g 20396 961 20398 1314-220457 978 20507 1208 20250 954 20232 1493 shoulder, meat + bone, g 2804 182 2851 205-47 2781 182 2875 204 2859 179 2813 213 fat + skin, g ... 1000 155 972 176 + 28 980 171 977 196 1048 89 964 152 foreback, meat + bone, g 2346 239 2451 245 -105 2316 245 2463 232 2418 152 2434 271 fat -f skin, g ... 544 118 562 152-18 532 118 580 172 570 125 534 115 back, meat + bone, g .... 3965 380 3951 344 + 14 4080“ 373 3975 ab 367 3689b 229 3913 ab 307 fat -f skin, g 1715 310 1709 388 + 6 1735 315 1736 465 1666 305 1669 236 loin + ham, meat + bone, g 6079 479 6067 489 + 12 6100 502 6149 480 6030 457 5940 495 fat + skin, g... 1943 246 1836 271 +lo7* 1932 a 229 1751 b 255 1970a 305 1965 a 243 total, meat + bone, g ....15195 977 15319 1078 -124 15277 1057 15462 1073 14996 734 15100 1079 fat + skin, g 5201 663 5079 741 +122 5179 690 5044 848 5254 665 5132 584 Meat + bone, % of half- carcass 46.2 2.3 46.3 2.1-0.1 46.3 2.5 46.6 2.5 45.8 2.0 45.9 1.4 Meat + bone, % of main part of carcass 74.5 3.1 75.1 2.9-0.6 74.6 3.3 75.4 3.5 74.1 2.7 74.7 1.8 Meat + bone/fat + skin .... 3.02 0.54 3.09 0.46-0.07 3.05 0.59 3.16 0.55 2.92 0.43 2.99 0.26 Meat + bone, g/day 88.5 6.4 94.1 8.5-5.6*» 88.2 a 6.2 95.0 b 8.5 89.3ab 7.6 92.7 ab 8.8 Lard + kidneys, g 2062 357 1950 443 +ll2 2146 368 1883 331 1860 179 2053 564 Side length, cm 77.1 3,1 76.9 2.3 + 0.2 78.0 a 3.1 76.5 ab 2.5 75.0 b 2.4 77.4 ab 1.9 Meat colour, points 2.69 0.31 2.67 0.28 + 0.02 2.73 0.30 2.72 0.29 2.59 0.33 2.60 0.25 OXo CO Statistical treatment by least squares analysis with slaughterweight as regression variable. * P < 0.05, ** P < 0.01. Differences between means tested by Tukey’s test, a—b: P < 0.05, c —d: P < 0.01. 510 consisted of progeny of one boar only, this being in half of the cases the first in order of service, in the other half the second. Lush et al. (1939) arrived at the conclusion that the order of service was unimportant in regard to the conception result, whereas the main factor was the fertility of individual boars. The present study also indicates possible differences in boar fertility despite the fact that the boars were carefully selected and tested for uniform semen quality. This requirement of uniformity, however, appears to have reduced the effects of this detrimental factor, for in no more than one of the twelve litters studied were all the piglets offspring of one boar only. Blood group testing proved an effective method of indentification of the piglets with the only reservation that the procedure of blood sampling from very young piglets was somewhat difficult. Judgment on the basis of external appearance only did not prove accurate enough. The animal material in this study was rather small. This was particularly true of the litters of the Yorkshire sows. Nevertheless results were obtained which were very similar to those of e.g. Skärman (1965), who studied crosses between Landrace and Yorkshire breeds using a large material of separate litters, and to those described by Glodek (1970) and Lauprecht (1957) in their reviews of crossbreeding experiments. The greater vitality and higher growth rate of the crossbreds emerged clearly from the present experiment as also did the similarity or the intermediate inheritance of the carcass characteris- tics. This study indicates therefore that if mixed semen is used, crossbreeding experiments are possible with very limited animal material even though very variable proportions of purebred and crossbred progeny are obtained. Acknowledgements. My best thanks are due to the Artificial Insemination Society of Salpausselkä (Salpausselän keinosiemennysyhdistys), to the Research Foundation for Agricultural Machinery (Maatalouskoneiden tutkimussäätiö), and to the Meat Processing Plant of Helsingin Kauppiaat OY for all their help in the course of this study. 511 REFERENCES Glodek, P. 1970. Zuchtverfahren zur Ausnutzung der Heterosis und ihre Anwendung in der Schweinezucht. 11. Ergebnisse aus Schweinezuchtversuchen. Z. Tierz. Zucht.biol. 86: 273-288. Harvey, W. R. 1966. Least-squares analysis of data with unequal subclass numbers. ARS 20 —B. Agric. Res. Service, U.S. Dept, of Agric. 157 p. Lauprecht, E. 1957. ttber das Verhalten der Nachkommen aus Paarungen verschiedener Rassen des Schweines. Ein Beitrag zur Heterosisfrage. Z. Tierz. Zucht.biol. 70: 57 76. Lush, J. L., Shearer, P. S. & Culbertson, C. C. 1939. Crossbreeding hogs for pork production, la. Agric. Exp. Sta. Bull, 380: 82 116. Roberts, E. & Carroll, W. E. 1939. A study of hybrid vigor in a cross between Poland China and Duroc Jersey swine. J. Agric. Res. 59: 847 854. Skärman, S. 1965. Crossbreeding experiments with swine. Lantbrukshogsk, Ann. 31:3 92. Steel, R. G. D. & Torrie, J. H. 1960. Principles and procedures of statistics. 481 p. New York, Uusisalmi, U. 1969. Vorläufige Ergebnisse iiber das Messen der Schlachteigenschaften beim Schwein. J. Scient. Agric. Soc. Eini. 41: 50 59. Selostus Eri rotuisten karjujen siemenseoksella suoritettu risteytyskoe Elsi Ettala Maatalouden tutkimuskeskus, Kotieläinhoidon tutkimuslaitos, Tikkurila Maatiais- ja yorkshirerotuisten (M ja Y) karjujen siemenseoksella siemennettiin maatiais- ja yorkshire-emakoita ja veriryhmämäärityksellä tunnistettiin pahnueiden puhdasrotuiset ja risteytysjälkeläiset. Karjupareiksi valittiin sperman laadun suhteen mahdollisimman saman- kaltaiset yksilöt. Laimennettujen sperma-annosten yhdistäminen ei vaikuttanut haitallisesti siittiöiden elä- vyyteen tai liikkeisiin. Tiinehtyminen ja pahnuekoko olivat myös normaaleja. Jälkeläisten jakautuminen eri ryhmiin oli seuraava: MM 34 kpl, YM 28 kpl, YY 24 kpl ja MY 20 kpl. Ja- kautuminen eri pahnueissa oli kuitenkin varsin vaihteleva. Yhdessä pahnueessa oli pelkästään puhdasrotuisia eläimiä. Porsaiden isät kyettiin määrittämään veriryhmämäärityksellä n. 95 %:sesti. Risteytyseläimet kasvoivat merkitsevästi {P < 0.001) nopeammin kuin puhdasrotuiset. Ne olivat myös puhdasrotuisia elinvoimaisempia. Ruho-ominaisuuksiltaan puhdasrotuiset ja risteytysjälkeläiset olivat hyvin samankaltaisia tai risteytyseläimet edustivat rotujen väli- muotoa. Tutkimuksen perusteella näyttää siltä, että seossiementä käyttäen voidaan varsin pienellä aineistolla selvittää eri rotujen risteytystuloksia siitäkin huolimatta, että puhdasrotuisten ja risteytysjälkeläisten lukumäärät pahnueissa vaihtelevat.