Piloting a Proficiency Testing Program for Laboratory Sampling of Animal Feed Materials.

BACKGROUND: Laboratory sampling is a significant source of error in feed testing. Proficiency testing programs such as the Association of American Feed Control Officials Proficiency Testing Program are an effective means of assessing error in and among analytical methods. However, all proficiency test items are comminuted and blended to control variability among items, effectively minimizing sampling error. Currently there is no mechanism for monitoring sampling error among laboratories. OBJECTIVE: The objective of this work was to investigate the feasibility of a proficiency testing program for laboratory sampling methods and provide insight into a program to advance the performance of sampling in laboratories. METHODS: The study involved the fabrication of identical feed test items from feed ingredients and shipping the uncomminuted materials to volunteer laboratories. The volunteer laboratories followed in-house procedures for selecting test portions for routine feed tests. Tests on all the test portions for a single analyte were performed by a single laboratory, so that the variability in test results could be attributed to laboratory sampling processes to select test portions. RESULTS: The average RSD, %, for Item A and Item B, respectively, were as follows: protein, 5.08 and 5.23; non-protein nitrogen, 8.90 and 16.6; crude fat, 3.45 and 5.67; vitamin A, 33.9 and 26.9; calcium, 21.9 and 23.6; zinc, 17.9 and 27.9; and copper, 17.4 and 27.9. CONCLUSION: This study suggests that a proficiency testing program for laboratory sampling is feasible with manual manufacture of the test items, and data can be used to monitor laboratory sampling proficiency and also to compare the performance of different laboratory sampling methods. HIGHLIGHTS: The data illustrates that each analyte has unique distributional and compositional heterogeneity, thus unique sampling error, even when multiple analytes are determined from a single test portion.

Effect of Test Portion Mass on Vitamin A Testing in Animal Feed Materials.

BACKGROUND: Vitamin A test results have historically been notorious for poor repeatability and reproducibility. This problem has been discussed at length in Association of American Feed Control Officials Laboratory Methods and Services Committee meetings. OBJECTIVE: The objective of this work was to assess the effect of test portion mass on the repeatability of vitamin A test results. METHODS: The study was conducted in two parts. In Part I, fundamental sampling error (FSE) was determined experimentally through replicated (n = 16) vitamin A testing of three animal feed materials. The testing followed rigorous test portion selection for 10 g and 100 g test portions. In Part II, FSE calculations were made (1) using theoretical equations based on vitamin A as a liberated analyte and (2) on representing the particles in feed materials. Particle size characterization of vitamin A ingredients was estimated by microscopy and further evaluated by particle size analysis. RESULTS: RSDs, % for vitamin A determinations ranged from 10.5-24.7, and 2.26-10.7 for 10 g and 100 g test portions, respectively. FSE calculated for Ingredient A ranged from 18.3-101% and 5.79-32.0% for 10 g and 100 g test portions, respectively, and for Ingredient B, ranged from 10.2-56.2% and 3.21-17.8% for 10 g and 100 g test portions, respectively. CONCLUSION: Test portion mass has a substantial impact on FSE and is an important factor in controlling the random error in vitamin A testing. FSE equations are useful to approximate minimum test portion mass. HIGHLIGHTS: Vitamin A method development should use theoretical predictions and experimental verification to guide test portion mass. Strategies to deal with the larger test portion masses will be key to validating new methods.

Considerations for sampling contaminants in agricultural soils.

Sampling agricultural soils for contaminants is relatively new. Existing standard sampling protocols used for the evaluation of soil nutrients are likely insufficient for contaminants. The main reasons are the very low analyte levels and differences in heterogeneity between nutrients and contaminants. To evaluate the adequacy of existing sampling protocols or to develop new protocols, a systematic scientific approach is needed. This approach begins with the development of the Sample Quality Criteria followed by a realistic understanding of the properties of the material to be sampled, most notably its heterogeneity. The Sample Quality Criteria and material properties are inputs into the Theory of Sampling. With these inputs, the Theory of Sampling can be used to determine the specifics of the sampling protocol (e.g., mass, number of increments, tool selection) that must be implemented to control error to reliably estimate the concentration of the analyte(s) of interest. Development of sampling protocols in this manner will ensure sample representativeness and therefore improve data equivalency among various parties involved. This is the only way to provide a sound technical basis for defensible decision making to ensure increased safety of food and feed, specifically with respect to contaminants in agricultural soils.

QC of sampling processes- a first overview: from field to test portion.

Quality control (QC) is a systematic approach for estimating and minimizing significant error contributions to the measurement uncertainty from the full sampling and analysis process. Many types of QC measures can be implemented; the three dealt with here are primary sampling reproducibility, sample processing reproducibility, and contamination. Sampling processes can be subject to QC by applying a replication experiment, used either from the top by replication of the entire sampling/ preparation/analysis process, or in a hierarchical fashion successively at each subsequent sampling stage. The analytical repeatability is necessarily always included in either alternative. The replication experiment results in a quality index, the Relative Sampling Variability, which is used to assess the total error associated with the full field-to-analysis pathway. Contamination can occur at essentially all locations in the sampling regimen in the food/feed realm, affecting sample containers, sampling tools, sample processing equipment, environmental conditions, and sampling personnel. QC events to determine contamination should always be included where appropriate, but is of most concern for low concentration and/or volatile analytes. It is also of key importance in the development of new sampling protocols or carried-over protocols intended for use on new types of materials/lots than the ones for which they were originally developed. We here establish a first practical framework for QC as applied to the sampling context.

Considerations for inference to Decision Units.

The goal of sampling is to take a small portion of a target material for analysis instead of collecting all the material. If sampling is done following certain principles, then inference can be made from analytical results of the portion taken back to the entire target material (Decision Unit). There are different sampling strategies that are dependent on the properties of the material being sampled as well as different methods for making inferences from analytical results to the Decision Unit (DU). A thorough understanding of material properties and methods for inference is therefore critical for the development of proper sampling protocols. This paper addresses inferences from analytical test results to DUs and the implications to the development of a sampling protocol.

Sample quality criteria.

The concept of Sample Quality Criteria (SQC) is the initial step in the scientific approach to representative sampling. It includes the establishment of sampling objectives, Decision Unit (DU), and confidence. Once fully defined, these criteria serve as input, in addition to material properties, to the Theory of Sampling for developing a representative sampling protocol. The first component of the SQC establishes these questions: What is the analyte(s) of concern? What is the concentration level of interest of the analyte(s)? How will inference(s) be made from the analytical data to the DU? The second component of the SQC establishes the DU, i.e., the scale at which decisions are to be made. On a large scale, a DU could be a ship or rail car; examples for small-scale DUs are individual beans, seeds, or kernels. A well-defined DU is critical because it defines the spatial and temporal boundaries of sample collection. SQC are not limited to a single DU; they can also include multiple DUs. The third SQC component, the confidence, establishes the desired probability that a correct inference (decision) can be made. The confidence level should typically correlate to the potential consequences of an incorrect decision (e.g., health or economic). The magnitude of combined errors in the sampling, sample processing and analytical protocols determines the likelihood of an incorrect decision. Thus, controlling error to a greater extent increases the probability of a correct decision. The required confidence level directly affects the sampling effort and QC measures.

Considerations in sampling of water.

Sampling water is no different than sampling any other media. It starts with the development of Sample Quality Criteria, understanding of material properties, then application of the Theory of Sampling. The main difference with sampling water as opposed to solids is the material properties. This paper addresses some of the material properties and consequences of those properties for the development of the sampling protocols. Two properties that must be addressed for water are the temporal nature and the inclusion of suspended solids. Examples are provided for three specific water sampling scenarios which may have application to other water sampling scenarios.

Perchlorate contamination from the detonation of insensitive high-explosive rounds.

The insensitive high-explosive PAX-21 was the first of its kind fielded in an artillery munition by the United States military. This formulation contains three main components: RDX, dinitroanisole, and ammonium perchlorate (AP). In March 2012, detonation tests were conducted on PAX-21 60mm mortar rounds to determine the energetic residues resulting from high-order and blow-in-place (BIP) detonations. Post-detonation residues were sampled and analyzed for the three main PAX-21 components. Concentrations of RDX and dinitroanisole in the samples were quite low, less than 0.1% of the munitions' original organic explosive filler mass, indicating high order or near high order detonations. However, disproportionately high concentrations of AP occurred in all residues. The residues averaged 15% of the original AP following high-order detonations and 38% of the original AP mass following the BIP operations. There was no correlation between AP residues and the RDX and dinitroanisole. Perchlorate readily leached from the detonation residues, with over 99% contained in the aqueous portion of the samples. Use of these rounds will result in billions of liters of water contaminated above drinking water perchlorate limits. As a result of this research, PAX-21 mortar rounds are currently restricted from use on US training ranges.

Field observations of the persistence of Comp B explosives residues in a salt marsh impact area.

Field observations of weathering Comp B (RDX/TNT 60/40) residue were made on a live-fire training range over four years. The Comp B residue was formed by low-order detonations of 120-mm mortar projectiles. Physical changes were the disaggregation of initially solid chunks into masses of smaller diameter pieces and formation of red phototransformation products that washed off with rain or tidal flooding. Disaggregation increased the surface area of the residue, thereby increasing the potential for dissolution. The bulk of the mass of Comp B was in the craters, but solid chunks were scattered asymmetrically up to 30m away.

Leg and calf press training modes and their impact on jump performance adaptations.

To examine the effects of resistance exercise (REX) mode on jump performance, subjects were assigned to one of three groups over a 6-week period with no cross-over. Subjects were assigned to leg and calf press REX on either a standard (n = 10) or ergometer (n = 9) device while a third group (n = 9) served as controls (CTRL). REX subjects worked out twice per week, which consisted of a three-set, 10-repetition paradigm for leg and calf press exercises. Immediately before and after the 6-week period, subjects performed tests that assessed jump (standing vertical jump, four-jump test protocol, depth jump) ability, while a fourth estimated knee extensor fast-twitch percentage (FT%) from fatigue incurred through a 50-repetition isokinetic protocol. Data analyses utilized 3 x 2 (group x time) repeated-measures ANCOVAs. Several dependent variables showed effects by group (standard REX, ergometer REX > CTRL) and time (post > pre). An interaction occurred for explosive leg power factor, a four-jump test variable, with standard REX post-test values as the interaction source. A trend for an interaction occurred for depth jump hang time, as ergometer REX values improved over time. Results suggest that mode-specific adaptations occur with REX training. Thus, athletes are best served with the selection of a REX device that is most specific to the demands of their jump performance task.

Blood lactate responses to leg presses performed against inertial resistance.

INTRODUCTION: A flywheel ergometer has been devised which employs gravity-independent inertial resistance. Concentric and eccentric actions, which are integral to weight-bearing and ambulation activities, may be done on this ergometer. However, blood lactate responses to exercise on the ergometer by contractile mode and training volume are unknown. METHODS: Workout order was balanced so that subjects performed each type of ergometer leg press workout twice in a nonsequential manner. Per workout, 10 repetitions were performed per set. Workouts were as follows: three sets with concentric and eccentric (CE3) actions, three sets with concentric-only (CO3) actions, and six sets with concentric-only (CO6) actions. Pre- and 5-min post-exercise lactate was measured from a fingertip blood drop. Lactate means were compared with a 2 x 3 (time X workout) ANOVA with repeated measures applied to both independent variables and Scheffe's post hoc test. With body mass and performance measures as predictor variables, multivariate regression attempted to explain post-exercise and delta (post-pre) lactate variance. RESULTS: Post-exercise blood lactate values (mmol x L(-1), mean +/- SEM) were as follows: CE3, 8.08 +/- 0.44; CO3, 7.57 +/- 0.49; and CO6, 6.96 +/- 0.43. CE3 and CO6 workouts produced comparable volumes of work. Though several CE3 performance measures were significantly correlated to post-exercise and delta lactate values, power indices had the strongest relationship. DISCUSSION: Factors related to lactate production and clearance caused CE3 values to be highest. Power indices were most correlated to lactate as they denote a higher work rate and reliance on glycolysis.

Identity and distribution of residues of energetic compounds at army live-fire training ranges.

Environmental investigations have been conducted at 23 military firing ranges in the United States and Canada. The specific training facilities most frequently evaluated were hand grenade, antitank rocket, and artillery ranges. Energetic compounds (explosives and propellants) were determined and linked to the type of munition used and the major mechanisms of deposition.

The effect of particle size reduction by grinding on subsampling variance for explosives residues in soil.

Efforts to characterize the surface soil contamination on military training ranges have been compromised by the inability to obtain representative subsamples of soils submitted to analytical laboratories for determination of explosives residues. Two factors affecting subsampling error for explosives residues were examined using soils collected from hand grenade and anti-tank ranges. These factors were increased subsample size and particle size reduction prior to subsampling of soils. Increasing the subsample size from 2 to 50 g did not reduce the soil subsampling error because of the extreme heterogeneous distribution of the solid contaminants. Alternatively, particle size reduction by machine grinding on a ring mill reduced subsampling error to less than 10% relative standard deviation for replicate analyses using 10-g subsamples.