Recommendations and Critical Factors in Determining Moisture in Animal Feeds Midwest AOAC June 6, 2018

Surveys were distributed Spring 2017 Aglabs AAFCO LMSC list NFTA Individual contacts

13 regulatory labs 7 research labs 16 private labs 36 total responses

please indicate how you determine moisture on the following matrices Matrix Temp Drying Time Oven Type Method Reference dry feeds forages, dry grains forages, wet oilseeds, unground liquid molasses feeds oilseeds, ground molasses block pet food, dry milk replacer pet food, canned suet, fats, oils pet food, semi-moist other distillers grains, dry distillers grains, wet

Red font – Official method or suggested method Blue font – regulatory lab responses Green font – research lab responses Purple font – private lab responses *labs who are following official or suggested method

DRY FEEDS AOAC 930.15 135C 2 hr or AOAC 934.01 95-100C vac 5 hr Should be AOAC 935.29 (NFTA 2.2.2.5) 104C 3 hr *4 labs 135C MC 2 hr *6 labs 103-105C MC 3 hr *2 labs 95-100C vac 5 hr *3 labs 100-105C MC 3-4 hr *1 lab 100C vac time not stated 1 lab 105C 16 hr *7 labs 135C MC 2 hr *3 labs 105C MC 3 hr 1 lab 105C MC 6 hr 1 lab 105C MC 16 hr 1 lab 80C vac 20 hr 1 lab Infrared analyzer

OILSEEDS - GROUND AOCS Ba 2a-3B 130C MC 2 hr *1 lab 130C MC 2 hr 1 lab 135C MC 2 hr 4 labs 104-105C MC 3-4 hr 1 lab 95-100C vac 5 hr 2 labs 105C MC 3 hr 1 lab 105C MC 16 hr 1 lab 100C vac time not stated *2 labs 130C MC 2 hr 4 labs 135C MC 2 hr 2 labs 100 - 105C MC 3 hr 1 lab 105C MC 6 hr

PET FOOD – CANNED  none specified 1 lab 125C MC 4 hr 1 lab 101C MC 16-18 hr 4 labs 103-105C MC 3-4 hr 2 labs 135C MC 2 hr 1 lab 100C MC 24-48 hr 1 lab 85C MC 16 hr 1 lab 95-100C vac 5 hr 1 lab Karl Fischer 1 lab 100C vac time not stated 2 labs 100-104C MC 16-18 hr 1 lab 125C MC 4 hr 2 labs 105C MC 3 or 6 hr 1 lab 135C MC 2 hr 2 labs 65 or 60C MC 5 or 16 hr 1 lab 145C halogen moisture analyzer

DISTILLERS GRAINS - DRY  AOAC 935.29 104C MC 3 hr *6 labs 103-104C MC 3 hr 1 lab 105C MC 4 hr 1 lab 105C MC 2 hr 1 lab 95-100C vac 5 hr 1 lab Karl Fischer AOAC 2001.12 *2 lab 105C GC 3 hr 1 lab 105C MC 16 hr 1 lab 2-step method 55C & 105C *8 labs 105C MC 3 hr 1 lab 105C MC 6 hr 1 lab 105C MC 16 hr 2 labs 135C MC 2 hr

LIQUID MOLASSES FEEDS  AOAC 969.35 60C vac 18 hr or AOAC 966.20 Karl Fischer *1 lab 60C vac 18 hr 1 lab 70C vac 5 hr 1 lab 100C vac 16 hr 1 lab 100C MC 24-48 hr 1 lab Karl Fischer AOAC 2001.12 1 lab 105C FD >12 hr *2 labs 60-65 C vac 16-18 hr 1 lab 80C vac 16 hr 3 labs 65-70C vac 4-5 hr 2 labs 105C FD 3 hr 1 lab 105C MC 6 hr 1 lab 105C FD >16 hr

MILK REPLACERS  AOAC 927.05 100C vac 5 hr *1 lab 100C vac 5 hr 1 lab 95C vac 5 hr 3 labs 103-105C MC 3 hr 1 lab 105±5C FD 4 hr±5 min 1 lab 100C MC 24-48 hr 1 lab 135C MC 2 hr *2 labs 100C vac 5 hr 1 lab 70C vac 4 hr 3 labs 105C MC 3 hr 3 labs 100-105C FD 5-6 hr

16 labs use the same method for all or most of their matrices 6 labs 103-105C 3 hr 1 lab 105C 4 hr 1 lab 105C 6 hr 1 lab 105C 16 hr 2 labs 135C 2 hr 2 labs 95-100C vac 5 hr 1 lab 100C vac time not specified 1 lab 55C >12 hr

Moisture Method by Matrix Table https://www.aafco.org/Laboratory Recommendations and Critical Factors for Determining Moisture in Animal Feed Moisture Method by Matrix Table

Karl Fischer titration DISCUSSION of METHODS Direct Water Methods Karl Fischer titration Toluene Distillation

Loss on Drying AOAC 934.01 VACUUM OVEN DISCUSSION of METHODS Loss on Drying AOAC 934.01 VACUUM OVEN Used by most labs in 1920s & 1930s Became official in 1934 by establishing temperature to be between 95-100C and time to be 5 hours

DISCUSSION of METHODS Loss on Drying AOAC 930.15 135C / 2 hr Adopted tentatively as an alternative method to be used when a vacuum oven was not available. Matrices evaluated were: Oilcake meal Fish meal Meat meal Bran cereal products (wheat) shorts

DISCUSSION of METHODS Loss on Drying AOAC 930.15 Limitations 1931 – cannot be used on feeds with sugar content due to caramelization 1935 – not to be used when a fat determination is to be made on the same test portion (fat oxidation will occur) Not applicable to feeds containing urea (decomposes at 132C) AAFCO PT data – this is the most widely used LOD method

DISCUSSION of METHODS Loss on Drying Thiex and Van Erem, SDSU, 1999 10 each partially dried hays, haylages, corn silage Karl Fischer 135C 2 hr -- over estimated moisture 104C 3 hr -- most closely matched KF 104C 6 hr

DISCUSSION of METHODS Loss on Drying Thiex and Van Erem, SDSU, 1999 10 different feeds containing no urea Karl Fischer 135C 2 hr -- over estimated moisture 104C 3 hr 104C 6 hr 110C 3hr -- most closely matched KF Vacuum 95C 5 hr

DISCUSSION of METHODS Loss on Drying Thiex and Van Erem, SDSU, 1999 6 different feeds containing urea Karl Fischer 135C 2 hr -- over estimated moisture 104C 3 hr -- most closely matched KF 104C 6 hr 110C 3hr Vacuum 95C 5 hr

DISCUSSION of METHODS Loss on Drying Thiex and Richardson, J Animal Sci, 2003, evaluated various moisture studies and concluded: Drying at 135C should be eliminated or restricted to very few materials for which it has been validated. Drying at 104 or 105C for 3 hr most closely matched KF moisture values.

DISCUSSION of METHODS Loss on Drying National Forage Testing Assn Task Force in 2000 Concluded that drying at 105C for 3 hr provided the most accurate value. NFTA Official Method 2.1.4

DISCUSSION of METHODS Loss on Drying Thiex SDSU 2009 study Analyzed 30 DDGs using 4 LOD & KF methods Drying at 105C for 3 hr most closely matched KF Poorest performing method 135C 2 hr

DISCUSSION of METHODS Loss on Drying PET FOOD No official methods Published data includes only two pet food samples. Dry cat food - 110°C/3 hr LOD most closely matched the KF moisture value. Dry dog food --135°C/2 hr LOD closely matched the KF value. Obviously more data is needed on pet foods.

DISCUSSION of METHODS Loss on Drying Most LOD methods are designed for low moisture (<15%) materials that can be easily comminuted to 1 mm.

DISCUSSION of METHODS Other Methods NIR Infrared moisture analyzer Halogen moisture analyzer  

RECOMMENDED METHODS Bidwell and Sterling stated in the JAOAC in 1925 “The ideal moisture method, which will probably never be found, would be applicable to all substances, would be rapid, would require little skill, would separate and determine uncombined water and nothing else, and preferably, would determine water directly and not by difference.”  

Karl Fischer: the gold standard RECOMMENDED METHODS Karl Fischer: the gold standard Use to validate all other moisture methods AOAC 2001.12 – dry animal feeds, pet foods, grains and partially dried forages  AOAC 991.02 -- soft-moist & semi-moist pet foods AOAC 966.20 -- liquid molasses AOAC 984.20 -- oils and fats

RECOMMENDED METHODS Loss on Drying The working group is recommending the discontinuation of the commonly used AOAC 930.15 (135°C / 2 hr) method for animal feeds, forages and grains due to gross overestimation of water in most feed materials. It is impossible to recommend one universal LOD method due to the diverse nature of animal feeds. The working group is recommending four different methods for the following matrix types.

RECOMMENDED METHODS Loss on Drying Dry animal feeds, grains, partially dried forages Drying in forced draft oven at 104°C for 3 hours (NFTA Method 2.1.4, AOAC Official Method 935.29 & 945.15)

RECOMMENDED METHODS Loss on Drying Forages, ensiled materials, wet feeds and grains Two-step moisture method as follows: Determine partial dry matter by drying the entire laboratory sample in a forced draft oven at 55°C overnight or until dry (NFTA 2.2.1.1) Comminute the entire laboratory sample to 1 mm particle size Determine laboratory dry matter by drying a test portion in forced draft oven at 104°C for 3 hours (NFTA Method 2.1.4) Calculate total dry matter: % total dry matter = (% partial dry matter x % laboratory dry matter) / 100 Calculate total moisture: % total moisture = 100 - % total dry matter

RECOMMENDED METHODS Loss on Drying Liquid feeds Drying in vacuum oven at 60°C for 18 hours (AOAC Official Method 925.45D)  

RECOMMENDED METHODS Loss on Drying Ingredients Most ingredients used in feeds and pet foods can be analyzed for moisture using the 104C for 3 hours LOD method without introducing sources of error such as caramelization or loss of other volatiles. In cases where a specific method has been validated or published for an ingredient by a body, and the customer requires, then the matrix specific method should be used and cited. Oilseed meal – forced draft 130C for 2 hr (AOCS Ba2a-38) Dried milk products – vacuum 100C 5 hr (AOAC 927.05) Fish, seafood – forced draft 100C 1 hr (AOAC 952.08) Meat & poultry – forced draft 125C 4 hr (AOAC 950.46B(b) 

RECOMMENDED METHODS Loss on Drying Pet Food The moisture working group is in the process of evaluating data on LOD methods for pet foods and will be making a recommendation in the near future.

RECOMMENDED METHODS Reporting of Results Moisture, loss on drying at 104°C / 3 hr Moisture, loss on drying, vacuum 60°C / 18 hr Moisture, Karl Fischer AOAC 2001.12 Moisture, NIRS legume hay calibration

CRITICAL FACTORS Loss on Drying Temperature Must be consistent Controlled to within ±1C Check and record oven temp each time using a thermometer or thermocouple, not the digital display unless it has been certified annually Temperature monitoring device needs to be calibrated against NIST thermometer at least annually.

CRITICAL FACTORS Loss on Drying Time Do not shorten specified time. Can extend by 10-15 minutes Time starts when oven returns to designated temp

CRITICAL FACTORS Loss on Drying Correct Method Many laboratories inappropriately use the same drying temperature and time for all of their test materials. This is acceptable only if all materials tested are similar, such as in a manufacturing facility. The analyst needs to keep in mind when encountering a different matrix that it may require the use of a different method.

CRITICAL FACTORS Loss on Drying Oven Forced draft or mechanical convection Control air velocity Shelves and containers cannot inhibit air circulation Short recovery time Avoid overloading

CRITICAL FACTORS Loss on Drying Drying Containers Sufficiently large to provide a surface area that allows the test portion to be spread out to provide maximum exposure for drying Should be shallow which will bring the heated oven air more closely over the surface of the test material Aluminum is ideal since it has low specific heat resulting in rapid cooling Covers not needed with electronic balances

CRITICAL FACTORS Loss on Drying Desiccation Desiccant needs to be changed or regenerated periodically Use desiccant with color indicator Desiccate for short time only – until cool and not overnight Keep desiccator closed between weighings

CRITICAL FACTORS Loss on Drying Balance Verify calibration each day Use weights that bracket the mass of the drying container & test portion Use same balance Calibrate annually

CRITICAL FACTORS Karl Fischer Maintain dry conditions for titration vessel and reagents Standardize against known quantity of H2O Ensure all H2O is released from test material special solvents – formamide or octanol homogenizer for solids thermal extraction Replicate test portions

CRITICAL FACTORS NIRS Equations standardized against KF or KF validated LOD method Update equations with recently analyzed test materials Matrix match calibration equations to test material

CRITICAL FACTORS All Methods Particle and Test Portion Sizes and Comminution Comminute to small particle size to provide maximum surface area Small particle size ensures more representative test portion of adequate mass Avoid heat generation during comminution Store in air-tight container

CRITICAL FACTORS All Methods Quality Assurance Validate moisture method Proficiency testing program QC material in each run Control charting to monitor QC material

Questions? https://www.aafco.org/Laboratory