Production of two composite beef cattle herds of Zebu x Criollo Yacumeño and Charolais x Zebu x Criollo Yacumeño on floodable savanna in Beni, Bolivia
II. Cow performance

D Plasse, E Galdo*, B Bauer* and O Verde

Universidad Central de Venezuela, Facultad de Ciencias Veterinarias, Maracay, Venezuela
*Estancias Espíritu, La Paz, Bolivia
dplasse@intercable.net.ve

Abstract

The objective of the present study was to avaluate pregnancy (P) and weaning (W) status as well as weight weaned per cow in the herd (WWCH) over four generations of two composite herds founded in 1975 at a private ranch on floodable savanna in Beni, Bolivia: Herd A with F₁ Zebu x Criollo cows and bulls and Herd B with 1/2Charolais1/4Zebu1/4Criollo, F₁ Zebu x Criollo and 3/4Zebu1/4Criollo cows bred to 1/2Charolais1/4Zebu1/4Criollo bulls. They were maintained in multi-sire herds on native grass in a 4-month breeding season and their calves weaned and weighed at 8 months. Pregnancy (P) and weaning (W) status was classified 0 (failure) and 1 (success) and analyzed by least squares procedures. WWCH was calculated from the respective adjusted means. The analysis included 19483 and 13051 cow-years for herds A and B, respectively.

The effect of generation (G; n = 4) was significant for P and W in Herd A (P<0.01) and Herd B (P<0.05). Year of breeding season (n = 14 ) and age class of dam (n = 11) were important (P<0.01) in both herds for both traits. Non-adjusted and adjusted means in Herd A were: P, 80.6±0.28, and 79.6±0.89%; W, 72.0±0.36 and 72.2±1.04%; WWCH: 126.4 kg at 248±0.3 days; and in Herd B: P, 77.6±0.37 and 79.6±1.13%; W, 66.4±0.47 and 68.8±1.32%; WWCH, 119.4 kg at 250±0.3 days. In Herd A, F₁ (G₁, foundation cows) had 88.1% P which decreased 11% in G₂ and a further 5% in G₃. In Herd B, P in G₁ was 83.0% and decreased 5% in G₂ and a further 3% in G₃. W behaved similarly. In Herd A WWCH declined 19% from G_1/2 (foundation cows with progeny) to G_2/3 and 6% in Herd B.

The adjusted mean WWCH over all generation was 5.9 % greater in Herd A than in Herd B. Between G_1/2 and the mean of G_2/3 and G_3/4 there was a great decrease of 18 and 8 % in WWCH in Herds A and B respectively. When comparing the mean WWCH of composites in G_2/3 and G_3/4 to results published from a selected Zebu herd at the ranch in the same years, Herd A and Herd B had a 3 % and 6 % lower output, allowing the conclusion that neither of the composite herds proved to be an alternative for weaner production under the prevailing conditions, compared to Zebu with a modern selection program.

Keywords: Beef cattle, composite herds, pregnancy percentage, total production, weaning percentage.

Resumen

El objetivo del presente estudio fue evaluar el estatus de preñez (P) y destete (D) como también el peso destetado por vaca en rebaño (PDVR) durante cuatro generaciones en dos rebaños compuestos fundados en una estancia particular en sabana inundable del Beni, Bolivia en 1975: Rebaño A con vacas y toros F₁ Cebú x Criollo y Rebaño B con vacas 1/2Charolais1/4Cebú1/4Criollo, F₁ Cebú x Criollo y 3/4Cebú1/4Criollo apareadas con toros 1/2Charolais1/4Cebú1/4Criollo. Las vacas fueron mantenidas en rebaños multitoro en pasto natural con una temporada de servicio de 4 meses y sus becerros fueron destetados y pesados a los 8 meses. P y D fueron clasificadas 0 (falla) y 1 (éxito) y analizadas por la técnica de cuadrados mínimos. PDVR fue calculado a partir de los promedios ajustados respectivos. Para rebaños A y B, los análisis incluyeron 19483 y 13051 vacas-años, respectivamente.

El efecto de generación (G; n = 4) para P y D fue significativo en Rebaño A (P<0.01) y Rebaño B (P<0.05). Año de temporada de servicio (n = 14) y edad de la vaca en temporada de servicio (n = 11) fueron importantes (P<0.01) en ambos rebaños para las dos características. Los promedios no ajustados y ajustados en el Rebaño A fueron: P, 80.6±0.28 y 77.6±0.89% y D, 72.0±0.36 y 72.2±1.04%; PDVR, 126.4 kg a los 248±0.3 días. Rebaño B: P, 77.6±0.37 y 79.6±1.13%; D, 66.4±0.47 y 68.8±1.32%; PDVR, 119.74 kg a los 250±0.3 días. En el Rebaño A, las F₁ (G₁, vacas de fundación) tenían 88.1% P que disminuyó 11% en G₂ y 5% más en G₃. En rebaño B, la P en G₁ fue 83.0% y disminuyó 5% en G₂ y 3% más en G₃. D se comportó similar. Entre G_1/2 (vacas de fundación con progenie) y G_2/3 el PDVR disminuyó 19% en el Rebaño A y 6% en el Rebaño B.

El promedio ajustado de PDVR sobre todas las generaciones fue 5.9% mayor en Rebaño A que en Rebaño B. Entre G_1/2 y el promedio de G_2/3 y G_3/4 había una gran disminución de 18 y 8% en PDVR en Rebaños A y B, respectivamente. Comparando el PDVR de los compuestos, en años correspondientes, con resultados publicados de un rebaño Cebú seleccionado de la misma estancia, el Rebaño A y el Rebaño B tuvieron 3% y 6% menos de producción permitiendo la conclusión que ninguno de los rebaños compuestos investigados se perfiló, en las condiciones prevalecientes, como alternativa al Cebú con un moderno programa de selección para la producción de becerros destetados.

Palabras claves: Bovinos de carne, porcentaje destete, porcentaje preñez, producción total, rebaños compuestos.

Introduction

The floodable savannas of Beni, Bolivia, have demonstrated their great potential for an increase of beef production during the second half of the last century. Therefore, beginning 1961, the authors have studied several purebreeding (Bauer et al 1992; Galdo et al 1992; Plasse et al 1993; Verde et al 1993) and crossbreeding (Bauer et al 1997; Plasse et al 1997; Galdo et al 2002; Plasse et al 2002) programmes using as genetic resources mainly Bos indicus and Criollo Yacumeño genes; however, a small amount of Charolais semen was also used. Two composite herds using the available crossbred resources with Criollo, Zebu and Charolais genes were also founded.

The analysis and discussion of the performance of four generations of cows and three of calves of these composite populations as an alternative to other breeding programmes carried out at this location is the objective of the present series. The paper of Plasse et al (2005) refers to weaning weight of calves and the objective of the present paper is the evaluation of cow production over three generations.

Materials and methods

The present study was carried out on a private ranch located in the floodable savanna of Beni, Bolivia, at 15° S and 64° W. The area is influenced by the Yacuma river and its tributaries and characterized by a severe tropical climate with mean temperature of 26.6° C (14 years average) and absolute maximum and minimum temperatures of 7° C and 39° C respectively. In normal years 80 to 90% of the ranch is flooded during the rainy season (October to March) and periodically severe floods make it necessary to evacuate cattle from large areas and interrupt routine work. During the dry season (April to September) severe droughts also occur periodically. Annual precipitation during the years of this experiment was on average 1865 mm.

Three physiographic zones can be distinguished: (1) High areas well drained and not affected by floods, (2) semi - high or transition areas which might be flooded 5 to 40 cm and low areas with floods above 40 cm. Predominating native grass species of interest for beef cattle are were given by Plasse et al (2005). A botanical description of the savanna of the area in which this ranch is locate has been given by Beck and Sanjinés 2005). At the time the present research was conducted, no cultivated pasture existed on the ranch. The adequate carrying capacity was about 0.25 AU ha^-1.

Herd A was founded starting 1975 with 1063 F₁ Zebu x Criollo (1Ze1Cr) cows and the same breed of bulls. It was closed in 1979 and reached 1745 cows when the study was discontinued in 1988. The F₁ animals had been produced out of Criollo cows and Zebu bulls generated in a selected Zebu elite herd established at the ranch in 1962 from Nelore-type Bos indicus cattle and in which later on Brahman bulls were use. After 1964, a selected Criollo herd of 600 cows was also established from a 6000 commercial Criollo herd. This bull producing herd was later named Criollo Yacumeño (Bauer et al 1992).

Herd B was also founded in 1975 with 250 1/2Charolais1/4Zebu1/4Criollo (2Ch1Ze1Cr) cows bred to the same type of bulls. In 1976, 250 1/2Zebu1/2Criollo (1Ze1Cr) cows and 350 3Ze1Cr cows were included. Herd B cows were bred to 2Ch1Ze1Cr bulls. Animals of this breed composition had been produced by artificial insemination of 1Ze1Cr cows with semen of Charolais bulls. Herd B was closed in 1978 with 879 cows and maintained until 1988 when it had 1006 cows. The foundation herd was composed of approximately 25% Charolais, 43% Zebu and 32% Criollo Yacumeño genes. Unfortunately no Zebu control group could be included in the design, however there was a selected Zebu herd kept at the same ranch.

Young bulls were selected from the respective herds according to 24-month weight and allowed to breed during three seasons (33% replacement rate. Heifers were selected to replace low producing cows according to a target weight established yearly. Cows were eliminated mainly according to low reproductive performance based on pregnancy diagnosis and abortions. Low maternal ability was considered the second priority for culling. The breeding season lasted from 15 of August to 31 of December and cows were bred in multi - sire herds. Pregnancy was diagnosed by rectal palpation about 45 to 60 days after the end of the breeding season. Cows calved in calving pastures from June through October and their calves were identified, dehorned and received basic health treatment at birth. Calves were kept with their mothers full time on native grass up to weaning and had access to common salt. Weaning took place at 8 months in monthly groups. Unusually heavy floods in 1977, 1981 and 1983 made systematic weaning impossible but in all other years all calves were weighed at weaning. Every effort was made to keep herds A and B under similar environmental conditions.

Cows included in the separate data files for herds A and B where those with a corresponding breed code, belonging to generation 1 through 4, which calved within the normal breeding season, and were 2 to 12+ years old. These final data files included the breeding seasons 1975 through 1988 and 19483 cow-years for herd A and 13051 cow-years for herd B. Pregnancy was coded 1 and failure 0.

In order to construct the two corresponding data sets for the analysis of weaning percentage, the data corresponding to the breeding seasons 1976, 80 and 82 were excluded because the resulting calf crops born in 1977, 81 and 83 were not weaned systematically. Cows having weaned a calf were coded 1 and those which did not 0. These two final data sets for the analysis of weaning percentage included for herds A and B 15515 and 10000 observations respectively.

The binomial variables pregnancy and weaning were analyzed by least squares procedures (Harvey 1990) according to the following model:

Y_ijkl = µ + g_i + a_j + v_k + e_ijkl

where

Y_ijkl is the pregnancy or weaning code of cow "l" belonging to generation "g_i" participating in the breeding season of year "a_i" with an age class of "v_k".
µ = population mean
g_i = effect of generation "i" (i = 1, 2, 3, 4)
a_j = effect of year of breeding season "j" (j = pregnancy: 1, 2, ……14;
weaning: 1,2,….11)
v_k = effect of age class of cow "k" in breeding season (k = 1, 2, ….11)
e_l = residual (with the usual assumptions).

Dividing the adjusted weaning percentages by 100 and multiplying the result by the mean adjusted weaning weights gave the value for weaning weight per cow in the herd.

Results and discussion

Variance analysis

Results of the least squares analysis of variance for the events pregnancy and weaning for Herd A and Herd B (Table 1) show that of the effects included in the model, generation of cow was statistically significant in Herd A (P<0.01) and in Herd B (P<0.05). Year of breeding season as well as age of cow in the breeding season exerted a highly significant P< (0.01) effect on the variation of the two traits in both herds.

Pregnancy percentage

Unadjusted means for pregnancy of 19483 cow-years in Herd A and 13051 cow-years in Herd B were 80.6 ±0.28 % and 77.6 ±0.37 % respectively, while the corresponding adjusted means were 79.6 ± 0.89% and 79.6 ±1.13% (Table 2).

Adjusted mean pregnancy of the Zebu elite herd (Plasse et al 1993) and the Criollo Yacumeño elite herd (Verde et al 1993) at the same ranch from 1967 through 1982 were 80.6% and 86.6% respectively. From the commercial upgrading program of criollo to Zebu at the same ranch, 45882 cow-years from 1971 through 1982 showed an adjusted mean for pregnancy of 78.5% (Plasse et al 1997). This comparison of the pregnancy percentages of the composite herds with the other herds existing at the ranch, might be somewhat biased since the data sets belong to somewhat different time periods and the size of the ranch makes it difficult to manage all herds under the same pasture conditions. However, it indicates that the two composite herds did not perform any better than the others and that the Criollo Yacumeño cows were of higher fertility. But 80% pregnancy as a mean of 14 years, is a good level of performance for beef cows on tropical lowland. Only a few publications from related ecosystems give evidence of similar reproductive efficiency. Plasse et al (2000) found adjusted means of pregnancy percentage to be 78% and 80% in two experimental herds of F₁ Bos taurus x Brahman combined with the respective Brahman control groups on floodable savanna in Venezuela, and Plasse et al (2002) reported the results of a rotational crossbreeding programme of Zebu x Criollo compared to grade zebu, with an adjusted mean pregnancy percentage of 78% from another ranch in Beni, Bolivia. On the other hand, Plasse et al (1988) evaluated 204830 palpations in eight commercial Bos indicus herds on floodable savanna in Venezuela and obtained a non-adjusted mean pregnancy percentage of 62%, while Arango et al (1999) and Plasse et al (1999) in Bos indicus herds on floodable savanna en Venezuela found adjusted means of 61% and 55% respectively.

The highly significant effect of generation on pregnancy percentage showed maximum performance for foundation cows in both herds (Table 2). In Herd A 1Ze1Cr cows had a 89.1% pregnancy rate, 11 % above the average of all generations. The high performance of these F₁ cows confirmed earlier reports (revised in: Plasse 1983; Plasse 1994). Plasse et al (1975) analyzed data corresponding to breeding seasons 1966 to 1970 at the same ranch where the present experimet was carried out and found that 2948 1Ze1Cr cows had a mean adjusted pregnancy percentage of 86%, while 15243 commercial criollo cows had 76% and 1474 zebu 72% pregrancy. This would result in an estimated heterosis of 16%. Analyzing pregnancy data of the same ranch corresponding to years 1971 through 1982, Plasse et al (1997) found that 17443 1Ze1Cr cows had an adjusted mean of 86 , while Criollo cows with F₁ calves had a 71 % pregnancy rate.

F₁ cows generate, theoretically, the maximum heterosis percentage for reproductive efficiency and under the present conditions they would also benefit from increased fertility of crossbred bulls as was indicated by Plasse (1983). There was an 11% decrease in pregnancy percentage of Herd A from the first to the second generation (Table 2) and a further 5% decrease from the second to the third (Table 2). The third generation had a 15% lower pregnancy percentage (74.9% vs 88.1%) than the foundation cows (G₁). G₄ had a relative small number of observations in both herds resulting in large standard errors for their constants. Theoretically, in Herd A heterosis for pregnancy percentage in G₂ cows should be half of that of the foundation cows. Taking the estimate of 16% heterosis in 1Z1Cr cows (Plasse et al 1975) and considering that half of this heterosis percentage would be lost in the next generation, an expected 7 percentage points loss would occur in G₂, which is less than the 9.5 points realized. The further 4 percentage points loss in G₃ would not have been expected, and must be attributed to reasons other than intra locus heterosis.

In Herd B the adjusted pregnancy percentage of the foundation herd (G₁, Table 2) was 83%. Consequently, 2Ch1Ze1Cr, 3Ze1Cr and 1Ze1Cr cows bred to 2Ch1Ze1Cr bulls had a 6% lower rate than herd A. Less heterosis should be expected in this herd and the 3/4 Bos taurus (2Ch1Cr) would probably respond with diminished reproductive efficiency, because of a high percentage of genes not favouring adaptation to extreme tropical conditions. There was a 5% decrease of performance in G₂, much less than in Herd A. G₃ had 76.5% pregnancy, 8% below the foundation herd and 1% above the same generation in Herd A.

The ranges of constants between years were large for both herds. Age of cow effects showed lowest values for 3 - year - old cows in both herds. These are all cows which are raising their first calf, since all heifers which did not conceive in the first breeding season (2 year old) were culled. This shows the important influence of first lactations in tropical cattle and has been reported elsewhere (Warnick et al; 1960; Linares et al 1976; Plasse et al 1978). The age effect for first service (2- year-old) cows was also evident in both herds, although much stronger in Herd B. Heifers entered the breeding herd with low weights typical for savanna - raised 2 - year - old females. The high negative value for this age group in Herd B might be due to the slower growth rate of heifers with more than 50% Bos taurus genes, especially 2Ch1Cr1Ze in the foundation generation, which might have lower adaptability. The highest pregnancy percentages were related in Herd A to cows between 6 and 10 years and in Herd B to those between 7 and 11 years.

Weaning percentage

Unadjusted and adjusted means for weaning percentage were 72.0 ± 0.36% and 72.2 ± 1.04% in Herd A and 66.4 ± 0.47% and 68.8 ± 1.32% in Herd B (Table 3).