P.PSH.1202 Final Report

Summary

Forage nutritive value (NV) is the principal determinant of voluntary intake and livestock productivity. Accurate NV estimation is required to predict animal performance, manage diets to achieve production goals, as a selection tool for higher quality forages, and to develop forage quality curves which underpin systems models. Animal house feeding experiments and wet chemistry laboratory analysis are expensive and time consuming. Near Infrared Spectroscopy (NIRS) allows the rapid and inexpensive prediction of nutritional traits. The use of NIRS requires the development of statistical relationships between measured traits and light absorbance in the near-infrared region of the electromagnetic spectrum (wavelength range 700–2,500 nanometres). Some nutritional and production traits are easier to predict by NIRS than others and existing calibrations have generally been developed for a narrow set of species (plants) or for only northern cattle (faecal). There are very few methods to predict intake, diet quality and performance of ruminants in extensive grazing systems. Intake markers are difficult to administer and achieve uniform release rates. Methods based on animal sensors can be very inaccurate when applied to different species, breeds and feedbase systems. Faecal NIRS offers a tool to predict aspects of selection, intake and diet quality from a faecal sample. This methodology has been developed for tropical cattle systems. It is unclear if it can be applied to southern cattle and sheep systems.

Digestibility is a key nutritional measure used to evaluate the quality of feeds and is used to estimate the metabolizable energy (ME) available for sheep and cattle (CSIRO 2007). Feed laboratories regularly use NIR calibrations to predict digestibility at maintenance feeding level, based on in vitro digestibility using either rumen fluid (Tilley et al. 1963), or pepsin cellulase (PC, Clarke et al. 1982) methods, with digestible organic matter in the dry matter (DOMD) used to estimate ME of the feed (CSIRO 2007). NIR calibrations based on primary in vivo standards (either faecal or forage) potentially bypasses the error associated with in vitro predictions (Kitessa et al. 1999). One of the objectives of this project was to determine whether NIR calibrations could be created directly for in vivo DOMD in sheep and cattle that covered a range of species at maintenance and ad lib feeding levels.

The main target audience was red meat producers, the feed testing industry, consultants, forage breeders and researchers. This work has potential national and international impact. The outputs from this project will be used to deliver more accurate predictions of nutritive value for all species within the Australian feedbase. With refinement, the faecal NIRS work will deliver an ability to predict diet quality, intake, aspects of diet selection and potentially methane emissions for sheep and cattle across Australia.

Objectives

The aim of this project was to bring together two large research and commercial delivery laboratories to combine, test and refine existing plant NIRS calibrations with samples that are taxonomically, spatially and temporally diverse.

Including:

- Expand the calibration database to accurately predict novel forages in the Australian feedbase (including forage brassicas, subtropical species and native shrubs).

- Develop calibrations for nutritional parameters that have not been predicted within the Australian feedbase were also developed (e.g.in vivo DOMD, oxalate or fibre digestion fractions).

Investigate veracity of claims that mineral content (e.g. phosphorus) can be predicted accurately were also investigated.

Using some samples that were generated in the project, as well as historical samples, expand faecal NIRS calibrations to predict intake, diet selection and performance traits of southern sheep and cattle was explored.

Test in field hand held NIRS units. For two commercially available portable NIRS units, loss of accuracy associated with in-field scanning of swards was quantified.

Finally, tropical cattle faecal calibrations were future-proofed by transferring them to modern NIRS machines.

Key findings

• Developed a broader range of forage (forage and silage) NIRS calibrations to predict nutritional values
• In the majority of cases, taxonomically broad calibrations predicted the nutritional traits of samples with greater accuracy than calibrations developed specifically for a plant or taxonomically similar group of similar plants.
• These broad NIRS calibrations represent a valuable tool for Australian researchers, feed testing agents and livestock producers, as they encompass nearly all of the species that appear in monocultures or mixed swards.
• Commercial NIRS labs offer predictions of calcium, magnesium, phosphorus, sodium, potassium, sulphur, chloride and dietary cation-anion difference (DCAD) which is calculated from these minerals.
• The capacity of high quality, laboratory grade, hand-held NIRS sensors was tested to provide accurate in-field predictions of plant nutritional factors. There is significant loss of accuracy. Predictions do not offer useful tools for animal management or plant improvement practices.
• Diet CP of sheep could be predicted through fecal samples with fair screening potential. There is a high prospect for NIRS calibrations to predict dietary CP intake for sheep and cattle. This offers a management tool for extensive production systems where diet selection is variable and nitrogen supplementation provides production benefits.
• Diet ADF, NDF and Ash content were predicted with excellent results, for sheep alone or sheep and cattle combined.
• Methane could be predicted with faecal NIRS, using cattle grazing tropical forages.
• Prediction of the dietary concentrations of some minerals from faecal samples was possible.
• ‘Fair’ screening potential for organic matter intake by sheep was demonstrated in this study. Digestible organic matter intake was predicted to a quality control standard.
• Tropical cattle faecal calibrations were successfully transferred to a modern SpectraStar XTR and FOSS XDS instrument.

Benefits to industry

Accurate, rapid and inexpensive predictions of a wider range of nutritive characteristics for the vast majority of plant species in the Australian feedbase. This will provide producers, consultants, researchers and plant breeders information that will allow them to make management decisions that will optimise red meat production and animal welfare.

NIRS-based tools to predict diet selection, diet quality (energy, fibre, crude protein, some antinutritional compounds and minerals), and digestible organic matter intake. This offers opportunities for producers and researchers to assess performance of grazing animals at the individual or herd/flock level in extensive systems. This will enable more timely supplementation decisions and offers an accurate and inexpensive tool for phenotyping individual animals grazing pasture.

Improved profitability and welfare of livestock industries through better management of the feedbase and livestock. This will translate to improved welfare outcomes during times of nutrient shortfalls.

Greater awareness of the opportunity to improve profitability by matching the nutritional needs of different classes of livestock and with current and emerging feedbase species.

MLA action

To support delivery on commercialisation plan for forages and fecal predictions of nutritive value. To maintain support of licensing for commercial fecal NIRS providers.

Future research

The broad multi-species NIRS calibrations for many of the forage quality traits for the Australian feedbase (160 plant species) are near-commercial and could be considered for release. This includes calibrations for novel traits. Some may require a little more work to optimise the product across the Bruker and Foss/SpectraStar methods. At this stage, there will be separate calibrations for the platforms with different predictive potential. This may cause some confusion to industry and requires further consideration during commercialisation.

The faecal NIRS calibrations for diet selection, diet quality, intake and possibly methane are showing much more promise than anticipated at the start of this project. It is exciting that evidence has been generated to indicate that sheep and cattle faeces can be used in the same calibration (as sheep are a much easier and less expensive model for generation of new samples for validation and expansion). These faecal NIRS calibrations could be a game-changer for the red meat industry and should be considered a priority for future investment. In the latter stages of the project, the team co-designed and gained animal ethics approval for a metabolism crate/methane feeding study to generate faecal samples to fill the critical gaps in the data.

The delivery of information regarding diet selection, diet quality, intake and possibly methane from faecal NIRS calibrations could be linked to animal nutrition models to provide producers with information regarding growth rates, efficiency and the need for strategic supplementation. If delivered through a web-based platform, this information could be provided as quickly as samples can be dried, ground and scanned (24-48h).

The researchers have considered – in consultation with MLA and industry – the potential commercialisation pathways of plant and faecal calibrations developed during this project.