Technical Bulletin
The use of regression analysis offers a means of evaluating response to an ionophore as compared to a non-medicated control across a wide range of experimental conditions. Huntington (1996) used linear regression analysis to evaluate weight gain response from cattle fed ionophores (Y; lasalocid, monensin, tetronasin, and lysocellin) as a function of weight gain for control (X; non-medicated) cattle. Data for this analysis was derived from published research studies for cattle grazing a diverse variety of forages and gaining between 0.77 and 3.48 lb/hd/day. The results of this analysis indicated the following relationship:
Y (kg/day) = 1.059(± 0.034)X + 0.047(± 0.026): R2 = 0.986
This relationship indicated a potential positive response (P < 0.09) of approximately 0.047 kg/day (0.10 lb/day) from ionophores under maintenance conditions or when cattle were not gaining any weight. This regression relationship likewise predicts an overall improvement in weight gain response (P < 0.01) of approximately 6% across the range of animals and conditions used in this database.
The objective of this Technical Bulletin was to use regression analysis to further evaluate the weight gain response of pasture cattle provided Bovatec or Rumensin and develop equations which can be used to predict the growth response of ionophores.
Linear regressions were conducted to compare response in ADG for control vs Bovatec; control vs Rumensin; and control vs Bovatec and Rumensin combined. Means used in regression analysis were based on individual studies within each database with each study given equal weight in developing the regression models.
The range in average daily gain (ADG) for control (nonmedicated) cattle ranged from 0.6 to 2.82 lb/hd/day with mean and median values of 1.47 and 1.36 lb/hd/day, respectively. The range in ADG for cattle fed Bovatec was 0.66 to 3.17 lb/hd/day, with mean and median values of 1.59 and 1.52 lb/hd/day, respectively. When ADG for Bovatec was regressed with ADG for control cattle, the resulting linear equation was:
Y = 1.0528(X) + 0.0433 (Sxy = 0.0253; Sy = 0.0391; and R2 = 0.968)
where Y is the predicted ADG (lb/hd/day) of cattle fed Bovatec and X is the observed or assumed ADG for cattle fed a non-medicated supplement The regression coefficient (P = 0.0001) was different from zero, however, the y-intercept was not (P = 0.27) different from zero. Subsequent plots of the residuals did not indicate a bias within this data set and confirmed that the data was normally distributed. Figure 1 graphically depicts this linear relationship. Because the true mean response in growth from an ionophore is never known with absolute certainty, the upper and lower 95% confidence intervals are likewise indicated in Figure 1. The confidence band for predicted ADG defined by the upper and lower limits indicate the most probable (95% confidence level) range of values containing the true growth response from Bovatec.
Rumensin database
Average daily gain for control cattle ranged from -0.02 to 3.19 lb/hd/day with mean and median values of 1.29 and 1.46 lb/hd/day, respectively. The range in ADG for cattle fed Rumensin was 0.05 to 3.15 lb/hd/day with mean and median values of 1.46 and 1.52 lb/hd/day, respectively. When ADG for Rumensin was regressed with ADG for control cattle, the resulting linear equation was:
Y= 0.9371(X) + 0.2534 (Sxy = 0.0235; Sy = 0.0337; and R2 = 0.961)
where Y is the predicted ADG (lb/hd/day) of cattle fed Rumensin and X is the observed or assumed ADG for cattle fed a non-medicated supplement. Both the regression coefficient and y-intercept were different (P = 0.001 from zero. As with the Bovatec regression, the residuals did not indicate a bias or a departure from normality within the database Figure 2 depicts the linear relationship of control vs Rumensin with the upper and lower 95% confidence intervals.
Combined lonophore Database
The mean and median values in the combined database were 1.37 and 1.33 for control cattle and 1.52 and 1.52 for ionophore-fed cattle. Using this data, the following linear regression equation was calculated:
Y = 0.9727(X) + 0.1854 (Sxy = 0.0179; Sy= 0.0267; and R2 = 0.960)
where Y is the predicted ADG (lb/hd/day) of cattle fed either Bovatec or Rumensin and X is the observed or assumed ADG for cattle fed a non-medicated supplement. Both the regression and y-intercept were different (P = 0.001) from zero with the residuals indicating no bias and the data being normally distributed. The standard errors for the regression coefficient (Sxy) and standard error for the y-intercept (Sy) were lower for the combined ionophore database regression than either the Bovatec or Rumensin database. This most likely is a consequence of the larger number of treatment means that were used to develop this regression. Figure 3 depicts the linear relationship of control vs the combined Bovatec and Rumensin data.
In all regressions, the linear relationships between growth rate of control cattle and those of cattle fed either Bovatec or Rumensin were highly significant and of a positive nature. The R2 values for Bovatec, Rumensin and the combined ionophore database are very similar indicating the strong relation between ADG of control and ionophore-fed cattle across a wide variety of conditions. A high R2 value (<90%) indicates that the associated regression relations have a high degree of accuracy for predicting the growth response from an ionophore when given the ADG for non-medicated cattle. The linear the regressions defined by the parameters indicated above may be used to estimate expected growth resulting from feeding an ionophore-medicated supplement across a variety of forage and supplement types.
The Bovatec relationship agrees with that Huntington (1996) developed using several different ionophores in addition to Bovatec and Rumensin. However, the Rumensin relationship and that resulting from the combined Bovatec and Rumensin data sets differ from that noted by Huntington (1996). Rumensin, in particular, seemed to have a greater positive effect on the maintenance component of the equation (y-intercept) and a negative effect on the gain component of the equation (regression coefficient). Specifically, this is indicated where a non-medicated supplemented animal at maintenance or zero growth, the addition of Rumensin to the supplement would equate to a gain of approximately 0.25 lb/day. Alternatively, because the regression coefficient for Rumensin is less than 1.0, as growth rate of control cattle increases, the predicted response from Rumensin progressively decreases. In comparison, Bovatec appears to have its greatest effect on the gain component of the regression equation and a marginal effect on the maintenance component. Because the regression coefficient for Bovatec is greater than 1.0, as growth rate of control cattle increases over maintenance, the predicted growth response from Bovatec progressively increases. These relationships are indicated in Figure 3, where at a control ADG value of approximately 1.50 lb/hd/day, the predicted linear response line from Bovatec crosses and exceeds the predicted linear response line for Rumensin.
1Bovatec, which is approved to increase rate of weight gain at the levels of 60 to 200 mg/head/day, was fed at these levels.
----
Bovatec® is a registered trademark of Alpharma Inc.
Rumensin® is a registered trademark of Elanco Animal Health.
Copyright © 2002 Alpharma Inc.
Additional technical information is available at www.alpharma.com
Alpharma Inc.
One Executive Drive
Fort Lee, NJ 07024 USA
1-888-897-8657
Technical Bulletin No. CD 0333
Regression Analysis Evaluating Effects of Bovatec® and Rumensin® on Rate of Growth of Cattle Grazing Pasture1
Technical Bulletin No. CD 0333Summary
A comparative regression analysis was conducted to evaluate the effect of Bovatec (lasalocid) or Rumensin (monensin) on growth rate of cattle grazing pasture as compared to cattle receiving a non-medicated supplement. Results indicated a strong and positive relationship between ADG for cattle provided a non-medicated supplement and cattle fed a similar supplement with Bovatec or Rumensin. Regression equations were developed which could be used to determine the expected growth rate of cattle fed Bovatec or Rumensin medicated supplements. Regression equations indicated that Rumensin had a greater response for cattle at maintenance or a low level of growth while response from Bovatec was greatest as basal growth rate increased. In addition to the positive growth response, criteria such as palatability in a wide variety of supplements and safety continue to make Bovatec the ionophore of choice for use in pasture supplementation programs.Introduction
Supplemental feeding of cattle maintained on rangeland or pasture has commonly been used to correct nutrient deficiencies occurring in forage and sustain profitable levels of production. lonophores, such as Bovatec, incorporated into various supplements alter ruminal fermentation, promoting more efficient utilization of forages to subsequently improve growth rate and productivity from pasture or range. Research has also indicated that ionophores can have beneficial effects in promoting health of grazing cattle by reducing incidence and severity of coccidiosis and other health-related problems. The fermentative processes that are affected by ionophores within the rumen, such as ammonia and volatile fatty acid production, are of direct importance to cattle growth since they greatly influence overall energy and protein metabolism. Direct comparisons to evaluate relative growth response either within the same ionophore or among different ionophores are difficult because of the diversity of cattle type, forage species, supplement types, and environmental effects.The use of regression analysis offers a means of evaluating response to an ionophore as compared to a non-medicated control across a wide range of experimental conditions. Huntington (1996) used linear regression analysis to evaluate weight gain response from cattle fed ionophores (Y; lasalocid, monensin, tetronasin, and lysocellin) as a function of weight gain for control (X; non-medicated) cattle. Data for this analysis was derived from published research studies for cattle grazing a diverse variety of forages and gaining between 0.77 and 3.48 lb/hd/day. The results of this analysis indicated the following relationship:
Y (kg/day) = 1.059(± 0.034)X + 0.047(± 0.026): R2 = 0.986
This relationship indicated a potential positive response (P < 0.09) of approximately 0.047 kg/day (0.10 lb/day) from ionophores under maintenance conditions or when cattle were not gaining any weight. This regression relationship likewise predicts an overall improvement in weight gain response (P < 0.01) of approximately 6% across the range of animals and conditions used in this database.
The objective of this Technical Bulletin was to use regression analysis to further evaluate the weight gain response of pasture cattle provided Bovatec or Rumensin and develop equations which can be used to predict the growth response of ionophores.
Summary of the Experiment
Data used in this analysis were obtained from published research studies (data not shown) and developed into two databases which either compared Bovatec with non-medicated controls or Rumensin with non-medicated controls. The Bovatec database was developed from 47 studies that incorporated 59 comparisons between control and Bovatec. The Rumensin database was developed from 66 studies comparing non-medicated controls with Rumensinfed cattle. Within each database, average Bovatec dose fed was 168.5 mg/hd/day (range 100 to 325 mg/hd/day) and average dose fed for Rumensin cattle was 178.4 mg/hd/day (range 44 to 200 mg/hd/day). lonophores were provided in medicated supplements which were either hand-fed or self-fed. Control cattle were fed the same type and level of supplement without ionophore. Hand-fed supplements typically consisted of a grain-based or protein- based supplement, while self-fed supplements were generally a free-choice mineral. Forage types across all studies consisted of a variety of cool-season, warm-season, or mixed species. Cattle across all studies were either stockers (steers/heifers) or replacement heifers. A summary of parameter means in both databases appears in Table 1.Linear regressions were conducted to compare response in ADG for control vs Bovatec; control vs Rumensin; and control vs Bovatec and Rumensin combined. Means used in regression analysis were based on individual studies within each database with each study given equal weight in developing the regression models.
Results and Discussion
Bovatec databaseThe range in average daily gain (ADG) for control (nonmedicated) cattle ranged from 0.6 to 2.82 lb/hd/day with mean and median values of 1.47 and 1.36 lb/hd/day, respectively. The range in ADG for cattle fed Bovatec was 0.66 to 3.17 lb/hd/day, with mean and median values of 1.59 and 1.52 lb/hd/day, respectively. When ADG for Bovatec was regressed with ADG for control cattle, the resulting linear equation was:
Y = 1.0528(X) + 0.0433 (Sxy = 0.0253; Sy = 0.0391; and R2 = 0.968)
where Y is the predicted ADG (lb/hd/day) of cattle fed Bovatec and X is the observed or assumed ADG for cattle fed a non-medicated supplement The regression coefficient (P = 0.0001) was different from zero, however, the y-intercept was not (P = 0.27) different from zero. Subsequent plots of the residuals did not indicate a bias within this data set and confirmed that the data was normally distributed. Figure 1 graphically depicts this linear relationship. Because the true mean response in growth from an ionophore is never known with absolute certainty, the upper and lower 95% confidence intervals are likewise indicated in Figure 1. The confidence band for predicted ADG defined by the upper and lower limits indicate the most probable (95% confidence level) range of values containing the true growth response from Bovatec.
Rumensin database
Average daily gain for control cattle ranged from -0.02 to 3.19 lb/hd/day with mean and median values of 1.29 and 1.46 lb/hd/day, respectively. The range in ADG for cattle fed Rumensin was 0.05 to 3.15 lb/hd/day with mean and median values of 1.46 and 1.52 lb/hd/day, respectively. When ADG for Rumensin was regressed with ADG for control cattle, the resulting linear equation was:
Y= 0.9371(X) + 0.2534 (Sxy = 0.0235; Sy = 0.0337; and R2 = 0.961)
where Y is the predicted ADG (lb/hd/day) of cattle fed Rumensin and X is the observed or assumed ADG for cattle fed a non-medicated supplement. Both the regression coefficient and y-intercept were different (P = 0.001 from zero. As with the Bovatec regression, the residuals did not indicate a bias or a departure from normality within the database Figure 2 depicts the linear relationship of control vs Rumensin with the upper and lower 95% confidence intervals.
Combined lonophore Database
The mean and median values in the combined database were 1.37 and 1.33 for control cattle and 1.52 and 1.52 for ionophore-fed cattle. Using this data, the following linear regression equation was calculated:
Y = 0.9727(X) + 0.1854 (Sxy = 0.0179; Sy= 0.0267; and R2 = 0.960)
where Y is the predicted ADG (lb/hd/day) of cattle fed either Bovatec or Rumensin and X is the observed or assumed ADG for cattle fed a non-medicated supplement. Both the regression and y-intercept were different (P = 0.001) from zero with the residuals indicating no bias and the data being normally distributed. The standard errors for the regression coefficient (Sxy) and standard error for the y-intercept (Sy) were lower for the combined ionophore database regression than either the Bovatec or Rumensin database. This most likely is a consequence of the larger number of treatment means that were used to develop this regression. Figure 3 depicts the linear relationship of control vs the combined Bovatec and Rumensin data.
In all regressions, the linear relationships between growth rate of control cattle and those of cattle fed either Bovatec or Rumensin were highly significant and of a positive nature. The R2 values for Bovatec, Rumensin and the combined ionophore database are very similar indicating the strong relation between ADG of control and ionophore-fed cattle across a wide variety of conditions. A high R2 value (<90%) indicates that the associated regression relations have a high degree of accuracy for predicting the growth response from an ionophore when given the ADG for non-medicated cattle. The linear the regressions defined by the parameters indicated above may be used to estimate expected growth resulting from feeding an ionophore-medicated supplement across a variety of forage and supplement types.
The Bovatec relationship agrees with that Huntington (1996) developed using several different ionophores in addition to Bovatec and Rumensin. However, the Rumensin relationship and that resulting from the combined Bovatec and Rumensin data sets differ from that noted by Huntington (1996). Rumensin, in particular, seemed to have a greater positive effect on the maintenance component of the equation (y-intercept) and a negative effect on the gain component of the equation (regression coefficient). Specifically, this is indicated where a non-medicated supplemented animal at maintenance or zero growth, the addition of Rumensin to the supplement would equate to a gain of approximately 0.25 lb/day. Alternatively, because the regression coefficient for Rumensin is less than 1.0, as growth rate of control cattle increases, the predicted response from Rumensin progressively decreases. In comparison, Bovatec appears to have its greatest effect on the gain component of the regression equation and a marginal effect on the maintenance component. Because the regression coefficient for Bovatec is greater than 1.0, as growth rate of control cattle increases over maintenance, the predicted growth response from Bovatec progressively increases. These relationships are indicated in Figure 3, where at a control ADG value of approximately 1.50 lb/hd/day, the predicted linear response line from Bovatec crosses and exceeds the predicted linear response line for Rumensin.
Conclusions
These results confirm the ability of ionophores to positively affect ADG for cattle that are being supplemented on pasture. However, based on the linear regression relationships developed in this Technical Bulletin, differences may exist between where Bovatec and Rumensin have their greatest effect. For Rumensin, this point appears to be where cattle are at maintenance or marginal level of growth. Conversely, Bovatec appears to have a greater response on the growing animal with an increased response from Bovatec resulting from each incremental increase in ADG of non-medicated cattle. Because obtaining an optimal level of growth during the time cattle are on pasture is a major management objective for stocker cattle or replacement heifers, the use of Bovatec in a pasture supplementation program will help insure that these goals may be achieved in the most economical and efficient manner. As a result of these attributes combined with a wide safety margin in both cattle and horses and compatibility with a variety of supplementation programs, Bovatec is the ionophore of choice for use with pasture cattle.1Bovatec, which is approved to increase rate of weight gain at the levels of 60 to 200 mg/head/day, was fed at these levels.
Literature Cited
Huntington, Gerald B. 1996. Grazing ruminant response to ionophores affected by management and environment. Feedstuffs, October 21, 1996.----
Bovatec® is a registered trademark of Alpharma Inc.
Rumensin® is a registered trademark of Elanco Animal Health.
Copyright © 2002 Alpharma Inc.
Additional technical information is available at www.alpharma.com
Alpharma Inc.
One Executive Drive
Fort Lee, NJ 07024 USA
1-888-897-8657
Technical Bulletin No. CD 0333
© Copyright 2006 - 2007 Alpharma Inc. all rights reserved
Home | About Us | Cattle News Room | Industry Links | Need Help? | Health Tools | Health Topics | Contact Us | Site Map | Privacy Policy | Trademarks