Herd life, stayability to fourth calving and reasons for disposal of Bos taurus x Bos indicus cows on dual purpose farms in Venezuela 

Lucía Vaccaro and Jazmín Florio* 

Universidad Central de Venezuela, Facultad de Agronomía, Maracay, Venezuela
*Present address: INIA Barinas, Carretera Barinas – Torunos, Km 10;
Sistema de Riego Santo Domingo, Barinas, Venezuela. 

Abstract 

Herd life (HL, lactations), stayability (ST, %) to 4th calving and causes for disposal (CD) were analysed using data obtained in 1989-98 from 11 Bos taurus x Bos indicus dual purpose herds in the Venezuelan lowlands. Unadjusted mean HL (± S.E.) was between 2.8 ± 0.3 (n = 43) and 2.7 ± 0.1 (n = 457), with adjusted means between 2.5 ± 0.1 and 2.6  ±  0.2, depending on restrictions set on the data base. Mean ST  (n = 1785) to 2nd, 3rd and 4th calving was 72%, 56% and 43%. Of all CD (n = 2450), 51% were involuntary (Dinv), 46% voluntary (Dvol) and 3% unknown. The main single CD were slaughter for reproductive problems (25%), culling for low milk yield (21%) and death (14%), with other reasons explaining individually  ≤ 10% total losses. Highly significant differences in HL, ST and CD were observed due to farm (F), year of first calving, breed group (BG), expected breeding value for milk yield/lactation  (EBVm ) and reproductive efficiency (EBVre), but the effects of season of first calving and EBV for calf weight (EBVcw ) were generally not significant. For HL, the interactions of BG with F, EBVm , EBVre and  EBVcw  were all significant. Year of disposal and calving number also influenced CD. Adjusted mean HL (± S.E.) for cows with ½ B. taurus inheritance (BG2: 2.7 ± 0.1) significantly exceeded  > ½ B. taurus (BG1: 2.2  ±  0.1), but not < ½  B. taurus (BG3: 2.5 ± 0.2).  BG2  had higher ST to 2nd, 3rd and 4th calvings than BG1  or BG3  by 8-11 percentage points (P < 0.01), and significantly lower Dinv  than BG1. Cows with positive EBVm  and EBVre  had 0.8 and 1.2 longer HL (P <0.01),  and 21-30 percentage points higher ST to all calvings (P <0.01) than negative cows, with significant differences in the proportions of most CD. It was concluded that HL was highly variable but, on average, short in these herds, partly as a consequence of high levels of Dvol.  BG2  in general and cows with positive EBVm  and EBVre  had the highest survival rates and most favourable levels of Dvol .

Keywords: herd life, stayability, disposal, Bos indicus x Bos taurus, dual purpose, Venezuela
 

Introduction 

Higher survival rates are well documented for Bos taurus x Bos indicus crossbred cattle in the tropics, compared with pure  European dairy breeds, but little is known about the factors which affect the survival of adult cows  or the reasons for culling  in  dual purpose herds which predominate in tropical Latin America (Vaccaro 1990; Madalena 1993). The objective of the present paper is to quantify herd life, stayability and reasons for disposal of crossbred cows in Venezuelan dual purpose herds, and determine some of the  causes of variation. 

Materials and Methods 

Records were collected over the period 1989-1998 from 11 herds in the states of Apure (1), Guárico (6), Falcon (3)  and Táchira (1) which participated in a cooperative research project on genetic improvement, coordinated from the Universidad Central de Venezuela. The location and main characteristics of the farms are summarized in Table 1. The cows were predominantly Bos taurus x Bos indicus crossbreds. All herds were kept on grazing but management varied widely between the locations. The main differences related to the predominance of cultivated pastures, supplementary feeding, twice daily milking and preventive programs of disease control in Falcon and Táchira, whereas the herds in Guárico and Apure used mainly native pastures with little or no supplementation, milked once daily and had more sporadic veterinary care. Criteria for culling varied between herds and were at the discretion of the farmer. However, lists of  cows ranked for estimated breeding value for three traits (ALPA 1988) were supplied to each farm by the project every six months as an aid to selection.  Records on dates of calving, dates and causes of disposal were collected by the project technician during monthly visits. Additional information on herd and data management is given by Florio (2000).

Herd life was quantified in terms of the number of lactations completed before leaving the herd, for all cows which calved  at least once. When cows left the herd during lactation, the corresponding fraction was estimated by expressing the actual days in milk as a proportion of the herd’s overall mean lactation length. Three data bases (B) were constructed. The first one (B100) included only records from those herd-years in which 100% of the cows calving for the first time had left the herd by December, 1998, when the data base was closed. Since the number of records resulting from this restriction was unexpectedly low (n = 43), second (B95) and third (B85) data bases  were made, which covered herd-years of first calving where 95% and 85%, respectively, of cows had left the herd by the end of 1998.  The numbers of herds, cows and years of first calving included in the three data sets are shown in Table 2. 

Stayability was estimated as the proportion of cows which, having calved for the first time, calved again for the second, third or fourth time.  The data set in each case included only those herd-years of first calving which ensured that all cows had either left the herd or calved for the second (or third, or fourth) time by December, 1998. Table 2 shows the years of first calving and numbers of records included.    

Reasons for disposal  were available for 2450 cows leaving the herds from 1991 onwards as they were not recorded systematically in the early years of the project. Disposal causes  were transcribed by the project technician as described by farm staff. They were defined as voluntary (Dvol), involuntary (Dinv) or unknown.  All Dinv cases either died or were sent for slaughter, as irrecoverable. Of a total  of 18 Dinv and 8 Dvol described by the producers, those which accounted for < 5% of the total losses were grouped together as ‘others’. In the Dvol  category, ‘sale’ refers to  stock defined by the farmer as surplus to requirements which were sold for breeding to other farms,  but the distinction between this and culling for performance could not always be clarified.

Data were classified according to farm (F), first calving year (Y), first calving season (S), breed group (BG) and estimated breeding value (EBV) for milk yield/lactation (EBVm), reproductive efficiency (EBVre) and calf weight at 4 months of age (EBVcw).  In addition to the above factors, reasons for disposal were classified according to year of disposal (YD), season of disposal (SD) and calving number (CN).  

The classification for Y varied with each trait and data set, as shown in Table 2. S and SD were defined  as  ‘wet’ or ‘dry’ for each farm, grouping together as ‘dry’ months for which local climatological records showed  evapo-transpiration to have exceeded precipitation.  CN grouped separately cows of 1st, 2nd - 4th  and ≥ 5th calvings. Three  BG were formed according to the degree of  Bos taurus (European breed)  inheritance in the crosses, using the genealogical information available on the farms: > ½ Bos taurus (BG1),  ½ Bos taurus (BG2) and  < ½ Bos taurus  plus pure zebus (BG3). The predominating Bos taurus breeds, Holstein and Brown Swiss, were evenly represented in each BG.  BG1  cows included 78% high grade crosses and the rest pure European breeds. The  BG2 group consisted of 62% F1, and 38% inter se, three or four breed crosses, while BG3  contained crossbreds with high levels of Bos indicus inheritance and a few (13%) pure zebu cows. A small number of cows were of unknown BG. Their records comprised 0.3% - 1.1% of the total, according to the trait considered, and were excluded (herd life) or included (stayability, causes of disposal), depending on the type of statistical analysis applied to each trait. This explains the slight discrepancy between the total number of observations and the sum of the observations for each BG, in the analyses of stayability and causes of disposal. The EBV values were taken from the project archives and were available for cows which left the herds after the end of 1991, provided they had at least one complete record for the trait in question. The values were independent of BG and classified as positive or negative for each trait, according to whether they were above or below the herd average. Since EBVre was based on records of calving interval, the sign was reversed to give greatest merit to the lowest values, and so that positive EBV’s would signify superiority consistently for each of the three traits considered.   

Linear models were used to analyse  herd life data, including the effects of  F (1,.....,11), Y (1989,.....,1992), S (wet, dry),  BG  (1, 2, 3),  EBVm (+,-), EBVre (+,-), EBVcw (+,-) and the interactions BG x F, BG x EBVm, BG x EBVre and BG x EBVcw. The complete model was modified according to the restrictions of  each of the data sets. When the analysis of variance detected significant effects, differences between means were determined by Duncan’s test.  The significance of differences in stayability to 2nd , 3rd and 4th  calvings, and in the proportions of disposals from different causes, was determined using contingency tables and Chi-squared tests (Steel and Torrie 1960). The factors taken into account were F (1,…..,11), Y (1989,….,1994), S (wet, dry) , BG (1,2,3) , EBVm, (+,-) EBVre (+,-) and EBVcw, (+,-) and the interactions BG x F, BG x EBVm, BG x EBVre and BG x EBVcw. The significance of the interactions was determined using tables of independence (G test), according to Sokal and Rohlf (1969). In addition, for cause of disposal, the effects of  YD (1989,…..,1998) , SD (wet,dry) and  CN (1, 2…4, ≥ 5th) were determined.    

Results and  Discussion 

Herd life 

The unadjusted and adjusted mean values (± S. E.) for herd life were: 2.8 ±   0.3, 2.6 ±  0.2 (B100);  2.3  ±   0.1, 2.5  ± 0.1 (B95) and 2.7 ±  0.1, 2.5 ± 0.1 (B85 ) lactations, respectively, with  coefficients of variation of 58% (B100), 68% (B95 ) and 67% (B85).   

The unbiased estimate, derived from B100, was based on very few observations, while B95 and B85 contained more records but underestimated  herd life by excluding records of a fraction of cows which survived longest. However, the mean values obtained in the three cases were very similar, suggesting that B85 may not have underestimated mean herd life too seriously. Assuming that the true mean value on these farms was probably 2.5 – 2.8 lactations, the results are comparable with the estimates of 2.0 to 3.1 calvings obtained for different groups of Bos taurus x  Bos indicus crossbreds in single herds in Venezuela (Cardozo and Vaccaro 1983, 1984; Chirinos 1998). However, they are much lower than the 4.7 –5.8 calvings reported from other field studies of crossbred herds in Venezuela  (Gonzalez Stagnaro 1990; Contreras and Blake 1998; Rizzi et al 2002).  Other higher values have been reported from Latin American countries, in the range 3.2 to 8.5 lactations and 3.6 to 8.4 years  (Evora 1988; Noguera 1985; Rodriguez 1993; Lemos et al 1996; Planas 1996), and even the values reported in 15 studies of pure European breeds (Vaccaro 1990; Rizzi et al 2002) tend to equal or exceed the values obtained in the present case.   

The results of the analysis of variance using B85  are shown in Table 3. Preliminary analyses had shown that only F and BG x F were significant ( P  < 0.01) using all three data sets. With the increased numbers of observations in B95, the effects of  Y, BG,  EBVre  and the three BG x EBV interactions also reached significance. Exactly the same  significance levels were found for these factors and, in addition,  EBVm   (P < 0.01) using B85. For this reason and to avoid repetition, only the  results of B85 given in Table 3 are discussed exclusively in the rest of the paper.  

Table 4 shows the adjusted means, according to the significant sources of variation. A very large difference (2.9 lactations) was found between F with extreme values, and to a lesser extent (1.3 lactations) between the  maximum and minimum Y.  

The major differences between farms show that herd life is an economically important factor which may be critical in some cases. Four farms had averages below three lactations, with a minimum of 1.5 lactations/cow. The worst problems occurred where pure European breed cows and high grade crosses were used on the extensive farms, or on the more intensive farms during a period in which the subsidy to concentrate feeds was suddenly eliminated, resulting in nutritional levels which were completely inadequate for high yielding cows. However, herd life was not clearly related to level of management, thus fitting with the conflicting evidence in this regard (Gonzalez Stagnaro 1990; Lemos et al 1996; Rizzi et al  2002).  The differences due to Y reflect the volatility of the social and economic condition in the decade of the 1990’s,  which sometimes forced farmers to drastically change management and even reduce their herds. Cows calving for the first time in 1989 had a significantly longer herd life than those calving later, which follows the consistently decreasing trend over time recorded in another  region of the country (Rizzi et al 2002).  

Table 4 shows that BG2 exceeded BG1 by 0.5 lactations (P < 0.01), with no significant difference  compared to  BG3. Although BG x F was significant (Table 3), the means of BG2 equaled or exceeded those of BG1 and BG3 on 10 of the 11 farms, with differences of up to 1.5 lactations.  The one exception was the herd in Táchira where the BG1  cows survived 0.8 lactations longer (P < 0.01). This may have been due to the more intensive conditions on this farm or the fact that some BG1  cows had been imported and were probably treated more favourably.  

The generally longer herd life of  BG2 cows fits with evidence from Brazil (Lemos et al 1996) and India (Katpatal 1977), referring to F1 cows. In the present case,  about one third of the BG2 group were  inter se and multi-breed crosses,  so that heterosis may  have been lower. This suggests that breed additive effects are also important,  and the results complement information on other traits showing that mixed 50% Bos taurus x Bos indicus crosses, as well as F 1’s,  provide a valuable option for use under lowland tropical conditions  (Vaccaro et al 1997).  

Table 4 also shows that cows with positive EBVm  and EBVre had a significantly longer herd life than negative cows (P < 0.01). The interactions BG x EBV were due to differences in the magnitude of the advantage of the positive cows, since they were superior in all BG by 0.3 – 1.3 (EBVm )  and  0.8 – 1.5 (EBVre) lactations. In general, EBVcw  had no effect (Table 3) , but BG x EBVcw, was significant. Positive animals  completed fewer lactations in all groups, but the difference only reached significance in BG1  (0.7 lactations, P < 0.05).  The results show that both milk yield and fertility were taken into consideration by farmers in the their culling decisions for all breed groups, while EBVcw  was less important.   

Stayability 

The proportions of cows which calved for the 2nd, 3rd and 4th time were 72%, 56% and 43% (Table 3).  As  would be predicted from the herd life results, these proportions are all low compared with other estimates from crossbred cattle in Latin America (Cardozo and Vaccaro 1984;  Lemos et al 1996; Cerutti et  al 1995). 

Table 3 also shows that survival to each stage was affected by F, Y, BG, EBVm and EBVre, but not by S. EBVcw only affected survival to 2nd calving. The interactions were not consistently significant. 

Table 4 gives the mean values obtained according to the significant sources of variation. The most marked effect was due to F, with extreme values varying by 32, 46 and 51 percentage points between farms in survival to 2nd, 3rd and 4th calving, again reflecting the differences observed in herd life.  

BG2 had consistently higher survival rates to all calvings than BG1 or BG3  (P  < 0.01). The BG x F interaction was only significant for survival to 2nd calving (Table 3), and BG2 cows achieved the highest survival rates to this stage on 8 of the 11 farms. Two exceptions occurred in favour of BG1, under intensive conditions in Falcon and Táchira, while  BG3  cows had highest survival rates on one farm in Guárico with more extensive management, where zebu-type cows may be more resistant. The results are consistent with those from Brazil where high grade European–zebu crossbreds only had comparatively high survival rates under good management conditions (Lemos et al 1996).   

Cows with positive  EBVm and EBVre survived very much better than negative cows, with differences in the range 21-30 (EBVm) and  27-28 (EBVre) percentage points (P < 0.01), according to calving number (Table 4). There was no BG x EBVm  interaction, and the  interaction BG x EBVre for survival to 3rd and 4th calvings, shown in Table 3,  was due to a difference in magnitude of the advantage of the positive cows. The difference  was most marked in BG3  reaching 21 (3rd) and 27 (4th) percentage points in favour of the positive cows (P < 0.01), but was not significant in BG1. Table 4 also shows a small difference in favour of the  positive EBVcw  cows at 2nd calving (6%, P < 0.05), but this disappeared in later calvings. 

Causes of disposal   

The main causes of disposal were 51% Dinv, 46% Dvol, with 3% unknown (Table 5). Slaughter due to reproductive problems was the most important single reason of loss (25% of all disposals), followed by culling for poor yield (21%), death  (14%) , culling for combined causes (10%) culling for low fertility (9%), sale for breeding (7%) and  slaughter for udder problems (5%). The ‘other’ category of Dinv included accidents; diseases such as brucellosis, leptospirosis and tuberculosis; extreme debilitation; foot problems and combined causes. In the Dvol group, the ‘other’ category included behavioural problems of the cow or her calf, usually related to breaking fences or refusal to be milked; old age and combined causes.   

The level of Dvol  reported here is unusually high and of  Dinv correspondingly low. Other studies of crossbred cattle in Latin America have shown voluntary losses in the range 5-27%  (Cardozo and Vaccaro 1983 1984;  Lemos et al 1996), although Gonzalez  Stagnaro (1990) reported  22.6 – 33.9% due to low milk yield,  and all of these are clearly  favourable compared with results from purebred. European breeds, where 50-75% of all losses have commonly been due to reproductive causes alone (Vaccaro 1990). However, many of these studies refer to experimental herds or herds kept as sources of breeding stock, in which voluntary culling would be expected to be low compared with commercial farms. The high level of Dvol  obtained in this study partly explains the low average herd life, and may have been associated with the fact that the herds received  EBV lists regularly and were encouraged to cull unproductive cows.  

The effects of different sources of  variation on the proportion of Dvol  compared with Dinv,  are given in Table 3. Significant differences were observed  due to F, Y, YD, BG, EBVm  and EBVre (P < 0.01). S and CN were significant at the level P < 0.05, and no effect was observed due to SD or EBVcw. The two significant interactions were BG x F and BG x EBVcw (P < 0.01). 

The mean percentage losses due to different causes, according to the non-genetic factors which had highly significant effects on disposal cause, are presented in Table 5. Very large differences in Dinv occurred due to F, Y and YD, varying from 26% to 73%, according to F, 24%  to 56% according to Y, and from 27% to 73% according to YD. These reflected major differences in losses from reproductive problems and death, which reached as high as 51% and 27%, respectively, on individual farms with high grade European crosses.  Correspondingly variable values were observed for Dvol, with culling for milk yield, fertility and sale of surplus stock varying from 9 – 36 %, 1- 15% and 0 – 26%, respectively, according to farm.  The highest levels of Dvol were observed on extensive farms with extremely high fertility and survival rates in their young stock. Y and YD effects were probably caused by the unstable socioeconomic conditions referred to above.  In general, however, widely different proportions of losses due to different causes are to be expected due to level of management and farm (Gonzalez Stagnaro 1990; Lemos et al 1996).  

Table 6 shows the proportions of disposals, according to BG, EBVm   and EBVre  , which also had highly significant effects on the  causes (Table 3).  BG2  had the lowest level of Dinv  (47%) although the difference compared with  BG3 was not significant. There were exceptionally high levels of Dinv for BG1 (60%), due to higher levels of losses due to reproductive problems (29%) and death (19%). Although the interaction BG x F was significant,  the proportion of Dinv  in BG2  was similar to, or lower than, other groups in all cases.  Thus, BG2  cows not only had better survival rates but  also left the herd less frequently for involuntary reasons than BG1, supporting earlier evidence from Brazil  (Lemos et al 1996). 

Table 6 also shows that cows with positive EBVm had higher Dinv than negative cows, due to greater losses from reproductive problems and death. They had correspondingly lower Dvol and,  as expected,  comparatively low culling for milk yield.  The  positive EBVre cows had slightly lower Dinv, with lower involuntary losses from reproductive problems than negative cows, although there was no difference in voluntary culling for fertility. They did, however, leave the herd more frequently due to culling for milk yield.

The relatively high rate of Dinv  in the positive EBVm   cows, compared with negative ones, seems more an automatic consequence of their lower culling levels for milk yield than indicative of  undesirable weaknesses, since their rate of survival to 4th calving was 60%, compared with 30% for the negative group.  The  particularly high survival of positive EBVre  cows in the BG3   group suggests that high reproductive efficiency was a major consideration in retaining them in the herd, since their milk yield would generally have been lower than that of the other BG.     

Conclusions 

It is concluded that  10 years’ records  from 11 herds provided a limited data base for estimating herd life, which  emphasizes the difficulty of obtaining reliable estimates of this trait from commercial herds in the tropics where continuous record keeping over a sufficient period of time is uncommon. The use of  B85   provided an option which does not appear to have seriously underestimated the mean,  and provided sufficient numbers of records for detecting important sources of variation in this trait. The mean values, 2.5 - 2.8 lactations, were short compared with evidence from the literature, but were accompanied by relatively high levels of voluntary culling, with  21% of all losses due to culling for low milk yield. However, slaughter due to reproductive problems was the main single source of loss (25%). Wide variation was observed between farms in herd life and causes of disposal, with poor survival due to high involuntary culling being a major problem in some cases. The results point consistently to higher survival to each calving for the BG2 cows, exceeding the other two BG by 8-11 percentage points, and with a significantly longer herd life than  BG1. They also had lower Dinv losses which permitted correspondingly greater opportunities for voluntary culling or sale for breeding. The major contrast was with the BG1 group, which showed a shorter herd life, associated with considerably higher losses from reproductive problems and death. Positive EBVm  and EBVre cows had a significantly longer herd life and higher proportions of survival to 4th calving than negative cows, suggesting that farmers took EBV information into account in culling decisions and that cows with high EBV for these traits were not prone to excessive losses from involuntary causes.  

Acknowledgements 

The authors express their gratitude to the farmers who generously collaborated with this research, to Humberto Mejías and Armando Pérez who were responsible for data collection, and to Jenny López for transcription. They also acknowledge the financial support of the International Development Research Centre (Canada) and the CDCH of the Universidad Central de Venezuela (Proyecto  01.36.4252.99). 

References 

Asociación Latinoamericana de Producción Animal (ALPA) 1988 Normas de Evaluación Genética de Bovinos de Leche y Doble Propósito en el Trópico Latinoamericano.  Memoria 23 Suplemento 1: 53-85 

Cardozo R and Vaccaro L 1983 Vida útil en sistemas intensivos de producción de leche en el trópico. 2: Hembras europeas x cebú. IX Reunión ALPA, Santiago de Chile, Chile: GM-35 (Resumen) 

Cardozo R and Vaccaro L 1984  Supervivencia de hembras mestizas europeas x cebú en un sistema intesivo de producción de leche en Venezuela. Revista UNELLEZ de Ciencia y Tecnología, Serie: Producción Agricola Año 2, No 5, Barinas, Venezuela. pp 91-95

Ceruttti  F, Rizzi R, Bagnato A, Pagnacco G and Oropeza M 1995 Primeras consideraciones sobre longevidad y “stayability” en la raza Carora. XIV Reunión Latinoamericana de Producción Animal. Mar de Plata,  Argentina, 26 Nov.-01Dic. 1995 

Chirinos Z 1998 Vida útil y longevidad en vacas mestizas de doble propósito. Trabajo de Ascenso. Facultad de Agronomía, Universidad  del  Zulia. 53 p 

Contreras R and Blake R 1998 Vacas Holstein puras y cruzadas para producción de leche en Venezuela. X Jornadas Técnicas de la Ganadería en el Estado Táchira, San Cristobal, Venezuela. pp 95-112 

Evora J 1988 Algunos indicadores de longevidad del genotipo 5/8 Holstein 3/8 Cebú. XI Reunión Asociación Latinoamericana de Producción Animal. La Habana, Cuba (Resumen G 18). 147 p 

Florio J 2000 Vida útil, permanencia y causas de salida en vacas de doble propósito: su relación con factores genéticos, grupo racial y valor genético estimado. Tesis MSc Postgrado de Producción Animal. Universidad Central de Venezuela. Facultades de Agronomía y Ciencias Veterinarias. Maracay. 147 p 

Gonzalez Stagnaro C 1990 Tasa e causales de eliminación en vacas mestizas. 12° Reunión de la Asociación Latinoamericana de Producción Animal. Brasil: 200 

Katpatal B 1977 Dairy cattle crossbreeding in India. I. Growth and  development of crossbreeding. World Animal Review 22: 15-21 

Lemos A, Teodoro R and Madalena F 1996 Comparative performance of six Holstein-Friesian x Guzera grades in Brazil. 9. Stayability, herd life and reasons for disposal. Brazilian Journal of Genetics 19 (2): 259-264 

Madalena F 1993 La utilización sostenible de hembras F1 en la producción del ganado lechero tropical. Estudio FAO No 111 Producción y Sanidad Animal, FAO, Rome. 87 p 

Noguera E 1985  Evaluación del comportamiento productivo y reproductivo mediante análisis de registros del rebaño de una estación experimental dedicada a la producción de leche. Tesis MSc. Postgrado en Reproducción Animal, Facultad de Ciencias Veterinarias, Universidad Central de Venezuela, Maracay. 93 p 

Planas T 1996 Longevidad del Siboney de Cuba. Revista ACPA. Notas Técnicas 2:13 

Rizzi  R, Bagnato A, Cerutti F and Alvarez J C 2002 Life time performance in Carora and Holstein cows in Venezuela. Journal Animal Breeding and Genetics 119: 83-92 

Rodriguez T 1993 Comportamiento de Holstein, ½ Holstein ½ Criollo-Cebú y ½ Pardo Suizo ½ Criollo-Cebú en ambiente cálido húmedo. Trabajo de Ascenso. Universidad de Oriente, Jusepin, Venezuela. 222 p 

Sokal R and Rohlf F 1969 Biometry, The Principles and Practice of Statistics in Biological Research. W H Freeman and Company. San Francisco. USA. pp 601-607  

Steel R y Torrie J 1960 Principles and Procedures of Statistics with Special Reference to Biological Sciences. Mc-Graw Hill Book Company, INC, New York, USA. pp 366-386 

Vaccaro L 1990 Survival of European dairy breeds and their crosses with zebus in the tropics. Animal Breeding Abstracts 58 (6): 475-493 

Vaccaro L, Pérez A, Mejías H, Khalil R and Vaccaro R 1997 Cuantificación de la interacción genotipo : ambiente en sistemas de producción con bovinos de doble propósito. In: Lascano C, Holmann  F, (eds) Conceptos y Metodologías de  Investigación en Fincas con Sistemas de Producción Animal de Doble Propósito. Centro Internacional de Agricultura Tropical (CIAT). Cali, Colombia. pp 67-79 

 Received 4 June 2002

Go to top