TECHNICAL SERVICES BULLETIN, October 2001
(3 of 4)

PERFORMANCE OF COMMERCIAL LAYING HENS WHEN FED DIETS WITH VARIOUS SOURCES OF ENERGY
R. H. Harms & G.B. Russell
University of Florida

Summary: Two experiments were conducted to determine the influence of adding various fats sources to the diet of the commercial laying hen. In one experiment a corn soybean meal control diet was fed with diets containing 0, 2 and 4% corn oil. Another diet containing 13.85% wheat bran was fed with and without 4% corn oil. There was no difference in egg production among the six treatments. The addition of corn oil produced a non-significant increase in egg weight. The same corn-soybean meal diet was used in the second experiment. Seven diets with 4% fat using soybean oil, corn oil, tallow, brown grease, yellow grease, choice white grease and poultry fat were prepared and fed. There was no difference in egg production among the eight treatments. A increase in egg weight was obtained from the addition of all fats sources except tallow.

FPRF Comments: Though little differences were observed in this experiment all sources of animal fats except tallow increased egg weights and performed equally to that observed with plant oil sources at a lower feed cost. This project was initiated and completed in response to Dr. Harms interest and FPRF members supplying the respective animal fats.

ROUTINE TECHNIQUES FOR MONITORING THE NUTRITIONAL VALUE OF ANIMAL MEALS
Dr. Theo van Kempen
North Carolina state University

INDUSTRY SUMMARY

Introduction: To prevent the large variation in nutritional quality of animal meals from becoming a bottleneck for the incorporation in feeds, techniques are needed that allow for the rapid quantification of their actual nutritional value. A method that was previously pioneered relied on using in vivo digestibility data to train Near Infrared Spectroscopic (NIRS) instruments such that the NIRS could predict the in vivo response. Although this method was proven to work, it is practically too costly and cumbersome as it relies on in vivo digestibility assays and a bare minimum 40 to 50 samples are needed to train the NIRS, which would mean that a method for a single feedstuff would cost over k$100. A more practical alternative would be to develop a validated in vitro method on which techniques such as NIRS could be based. Using an in vitro method 50 to 100 samples can be assayed in a short period of time for a fraction of the cost of the in vivo assays, and if the data are biologically relevant, indirect methods can be based on the in vitro data.

Objectives: In brief, an in vivo validated in vitro technique is to be developed, and the in vitro digestible amino acid content is quantified for approximately 70 animal meal samples. These samples are subsequently used to develop NIRS, Fourier Transform Infra Red spectroscopy (FTIR), and Raman spectroscopy calibrations relating spectral data to total amino acids, and in vitro and in vivo digestible amino acids. These calibrations can then be used to predict the nutritional value of an animal meal sample in 2-5 min with a variable cost that is not much more than the labor required for the handling of the sample.

Industry Summary: An in vitro digestibility method has been developed that yields data that approximate in vivo digestibility better than existing in vitro techniques. Nevertheless, on average, the in vitro digestibility was a couple percent (2-10%) lower for lysine than the in vivo measurements, making extrapolation from in vitro to in vivo digestibility unreliable. The reasons are possibly 1) solubility problems of the in vitro digested sample; 2) lack of intestinal peptidases in the in vitro system; 3) particle size effects of the meat and bone samples on digestibility, and 4) lack of removal of end products of digestion. The in vitro method thus requires additional work before it can be used as a reference method for the development of spectroscopic prediction equations.

Using spectroscopic methods such as NIRS and FTIR, calibrations could be developed for total, in vitro, and in vivo digestible amino acids. The prediction error for lysine was approximately 0.25 (%) and the r2 was 0.85 for both methods. Using additional samples, further improvements in this prediction error are very feasible.

Upon careful analysis of in vivo digestibility data it was found that digestibility of the samples is actually a minor component of the variation in digestible amino acids (less than 10% of the total variation). This suggests that a method for predicting total amino acids and using a fixed digestibility coefficient for calculating digestible amino acids is nearly as valuable as a method for measuring digestible amino acids. The estimated prediction error for this method is 0.21 (%) for digestible lysine.

MANUSCRIPT

Scientific Abstract: The animal industry is in need of methods that can assess the nutritional value of feed ingredients such as animal meals, which are suitable for routine use. Infrared spectroscopy can be trained to predict digestible amino acids. However, the use of in vivo reference data makes this an expensive and time-consuming proposition. As an alternative the possibility to develop an in vitro digestibility assay that could be used for training infrared spectroscopy was investigated.

The in vitro digestibility assay was developed on the premise that in vitro digestion should be maximized using pepsin and pancreatin as the digestive enzymes but using minimal enzyme levels such that contamination from enzyme catalysis can be minimized. The method developed uses 500 mg of meat and bone meal incubated with 0.7 mg of pepsin for 24 h at pH 2, followed by incubation with 13 mg trypsin-enriched pancreatin for 96 h at pH 8. Although this method maximally digested animal meal samples, the actual digestibility coefficients were still somewhat lower than the in vivo digestibility coefficients (up to 10%), making this method not yet suitable as a reference. The reason for this is not known, but one possibility is inadequate solubilization of the digested material prior to analysis.

Using in vivo poultry digestibility data obtained on 25 animal meal samples and total amino acid data on over 70 animal meal samples, calibrations using both NIRS and infrared calibrations could be developed that explained approximately 85% of the variation in total and digestible lysine. These data also showed that digestibility was responsible for less than 10% of the variation in digestible amino acids; thus, total amino acid data together with fixed digestibility coefficients can be used to predict digestible amino acids with good accuracy.

In conclusion, this research confirmed that infrared spectroscopy could be trained to predict digestible amino acids based on in vivo digestibility data. Digestible amino acids, though, can also be predicted with good accuracy from total amino acids for which infrared calibrations can be developed with great ease.

FPRF Comments: Though the hoped for success of the nuclear magnetic resonance (NMR) assay did not materialize as a current viable assay procedure for in vitro nutrient contents, the progress made in refining those associated with the infrared spectroscopy (NIRS) procedure was evident in this project. This indirect method of analysis has been used extensively for plant proteins but has been inconsistent for analyzing animal proteins. The samples used in this project including those in which in vivo (cecectomized rooster assay) are being retained. Subsamples can be obtained for those wishing to calibrate their company NIRS equipment.

< previous   next >

BACK TO TECHNICAL SERVICES BULLETINS INDEX