Presentation: Breeding Barley for End-Use Quality. Will it change barley marketing?

Introduction

The most limiting and expensive traits to select for, whether you are breeding new varieties or marketing/buying product, are the quality traits. However, these are often the most economically important traits that determine the end use of our grain. Quality traits are not simply inherited; they have other physiological effects on the seed or final product and are significantly affected by the environment. The variation between samples of barley for any one trait is between 20 and 40%. This variability is due to both genetic and environmental factors. Simply measuring test weight (bushel weight) and kernel weight (seed plumpness) do not define nutritional quality.

Phenotyping Breeding Populations

Over the last 40 years, our research program has looked at many different techniques to rapidly screen breeding populations for different perceived or real economic quality traits. Most techniques are too expensive or unreliable to screen large numbers of samples. In about 1978 we began to look into NIRS technology (Near Infrared Reflectance Spectroscopy). However it took until 1995 to see the technology mature to the point where it could be used for screening breeding populations for a significant number of quality parameters. This was due to advancement in both scanning monochromator technology and desktop computer capability. Modern NIRS is accurate, repeatable and rapid as well as non-destructive to the sample, allowing breeding programs to screen large numbers of lines for multiple characteristics at the same time.

NIRS technology uses electromagnetic radiation to measure energy absorption of hydrogen-containing molecules, which produces a ‘fingerprint’ or spectrum of the sample. This spectral data is then matched with data from traditional wet chemistry analyses and/or animal performance data to produce a calibration equation. The NIRS equation can then be used to predict feed quality characteristics such as protein and energy in unknown samples, all in just a matter of minutes. Therefore, with this technology the breeder is able to rapidly phenotype genetic populations for many quality traits at the same time. At the present time The Field Crop Development Centre (FCDC) in Lacombe Alberta is running over 40,000 samples through the laboratory in a 4 month period and screening for over 25 quality characteristics related to malting, feed and food quality (Table 1a & 1b).

Table 1a. Statistics for NIRS equations developed at the FCDC to predict feed barley quality characteristics used for screening in the breeding program.


Table 1b. Statistics for NIRS equations developed at the FCDC to predict food barley quality characteristics used for screening in the breeding program.

Measuring End Use Quality

Measuring and breeding for end-use quality has long been the norm for malting barley and bread wheat. However feed grains have not been measured for quality and are primarily used in animal diets based on price and perceived average quality characteristics for the species. Reports from the corn industry in the USA indicate that the variability for Digestible Energy (DE) in corn is 7 to 8% (FeedInfo News Service 31/05/2005). We have found the variability in Canadian barley to be between 25% and 40% depending on the quality trait. This has tremendous economic impact on the efficiency of feed use in barns and feedlots. Figure 1 is based on data from the breeding program over years and locations and compares the range and mean for DE and % Protein Digestibility (PD) in a pig. This gives a good indication of the genetic variability available to the breeder.



Information in the following figures (2,3,4, and 5) is based on advanced yield trial material grown in one location in 2006. Cultivars A through F are hulless lines and lines G through L are hulled lines. There are two-row and six-row lines of both. What is important in this set of figures is the variability we see for the four traits measured but also the interactions between the traits. On an economic basis cultivar F would be the best line for feeding swine.







The biggest problem in measuring feed quality is that it is not defined and differs depending on the species of animal being fed. The second biggest problem is that the error in animal research is often too great to define the samples characteristics. If you couple that with the sample size needed for animal trials and the cost of these trials, we are limited to bench top techniques that we hope have a relationship to animal performance. As of yet, there are no standardized, in vitro techniques used worldwide for establishing animal feed quality.

Another large source of error in developing NIRS technology for feed quality is in the selection of the sample sets that are used to develop and test the calibration. The variability in a commercial sample of Metcalfe barley compared to a genetically pure sample from a randomized controlled experimental plot is shown in Figure 6.


This represents how environment can introduce variation within a sample thereby impacting the commercial use of NIRS technology. Therefore, depending upon the trait and the genetic and environmental influence on that trait, a special set of samples that contain the genetic and environmental variability represented in a breeding population may be required for the development of a calibration. In fact, a different set of samples may be required for every trait. Good examples are beta-glucan and amino acids in barley. For this we would recommend using the same population building techniques that would be used for building a Molecular marking population; that is, making several crosses between high and low lines and building a calibration sample set from randomly derived F5 lines grown at several locations.

Our goal is to develop a set of in-vitro techniques that can be used to develop robust calibrations for as many feed quality characteristics as possible. These calibrations can be used to phenotype both research and commercial samples. This should lead to further definition of the genetics controlling these quality characteristics and the development of varieties with greater economic feed value.

One of the questions that must be asked is, “Will this methodology lead to significant improvement in quantitatively inherited traits?” Since 1990, FCDC has made significant progress in screening genotypes for swine energy and protein digestibility as shown in Figure 7.



We have not as yet, tried to determine how many genes are involved in this improvement or if the genes are mark-able. We suspect that it is similar to breeding for yield in that you can make more progress selecting for the yield than you can by selecting for individual components of yield.

Commercializing NIRS Technology

We have begun the process of transferring the NIRS technology used in genetic development to commercial partners, thereby allowing for the definition of the major quality components of feed ingredients into a standard format. This will allow both feed and livestock producers to price ingredients according to their true feed value.

Ultimately the use of NIRS technology should result in considerable dollar savings for livestock producers and feed manufacturers by providing quick reliable analysis of feed quality components. These savings could be achieved through formulation of diets more closely related to the actual nutrient requirement of the animal (decreasing safety factors) and/or through better animal performance. Feed grain producers will also benefit from NIRS technology by using the grain analysis to market their grain according to its specific quality traits and to be paid for actual feed grain quality. This is demonstrated in Table 2 that shows the comparative economic value of barley in swine diets by the addition of canola oil to bring the DE content up to 3600kcal/kg that is equal to maize. It is very evident from this comparison that kernel weight and test weight do not affect economic value.



Last but not least, it will define for the plant breeder quality traits of economic importance and the tools to phenotype these traits, thereby developing greater economic and nutritional value in the crop.

Summary

• NIRS is a rapid and cost effective way to screen for feed quality in both research and commercial samples.
• We can build a quality standard that will market feed grains on end-use quality.
• We can build better barley for feed that will increase feed efficiency and reduce environmental impact.
• Two-Row or Six-Row barley doesn’t ensure better quality
• Simply measuring test weight (bushel weight) and kernel weight (seed plumpness) do not define nutritional quality.

• Alberta Agriculture and Rural Development, Field Crop Development Centre (FCDC)
• Alberta Crop Industry Development Fund (ACIDF)
• Alberta Barley Commission (ABC)
• University of Alberta