34. Analysis for Extraneous Matter

34.2.1 Definition of Terms

34.2.2 Diagnostic Characteristics of Filth

There are certain qualities characteristic to extraneous materials that serve as proof of presence of foreign or objectionable matter in food. Examples include specific diagnostic characteristics of molds (i.e., parallel hyphal walls, septation, granular appearance of cell contents, branching of hyphae, blunt ends of hyphal filaments, non-refracted appearance of hyphae); diagnostic characteristics of insect fragments (i.e., recognizable shape, form, or surface sculpture, an articulation or joint, setae or setal pits, sutures), rodent hairs (i.e., pigment patterns and structural features), and feather barbules (i.e., structural features); diagnostic characteristics of insect-damaged kernels (IDK) and packaging materials; and chemical identification of animal urine and excrement. These diagnostic characteristics are outlined by AOAC International (formerly Association of Official Analytical Chemists) for positive identification of extraneous matter or filth [5].

The AACC International (AACCI) (formerly American Association of Cereal Chemists, AACC) publishes a methods book that includes a section on extraneous matter, containing descriptive material helpful in identifying insect and rodent contaminants [6]. Several microscopic and radiographic illustrations are provided by the AACCI as authentic reference materials to help analysts to identify filth. AACCI Method 28-95.01, “Insect, Rodent Hair, and Radiographic Illustrations,” provides a series of colored pictures representative of insect fragments commonly found in cereal products and pictorial examples of rodent hair structure.

Kurtz and Harris [7] provide a virtual parts catalog of insect fragments with a series of micrographs. Gentry et al. [8], in an updated version of the Kurtz and Harris publication, include colored micrographs of common insect fragments. Also included in AACCI Method 28-95.01 are radiographic examples of grain kernels that contain internal insect infestation. AACCI Method 28-21.02, “X-ray Examination for Internal Insect Infestation,” provides an outline of the apparatus and procedure for X-ray examination of internal insect infestation in grain [9].

34.4.1 Sieving Method

Separation is based on difference in particle size of product and contaminant using standard test sieves (Fig. 34.1). For instance, the insects are (larger) separated from spices (smaller) using a 20-mesh sieve, and wheat grains (larger) are separated from insects (smaller) using a 10- or 12-mesh sieve. Then the contaminant is identified using a widefield stereomicroscope.

34.4.2 Sedimentation Method

Separation is based on different densities of product, contaminant, and immersion fluid. Specific gravity of immersion solution (carbon tetrachloride/chloroform) allows heavier shell, sand, glass, metal, or excreta contaminants to settle; less dense product floats. Apparent lower specific gravity of internally infested wheat kernels, for instance, allows them to float, while sound wheat kernels settle in 1.27 specific gravity solution. Contaminants are then identified using a microscope.

Analysis of high-fat-containing samples such as nuts requires defatting using petroleum ether prior to filth analysis (AOAC Method 968.33A). The chloroform and chloroform/carbon tetrachloride solvents allow pieces of shell, sand, and soil to settle at the bottom of the beaker on the basis of specific gravity and cause the defatted nut meats to float and be decanted. Essentially, the same procedure is suggested to isolate pieces of rodent excreta from corn grits, rye and wheat meal, whole wheat flour, farina, and semolina in AOAC Method 941.16A. It should be noted that the use of the more toxic solvents such as carbon tetrachloride, chloroform, and petroleum ether is avoided in most contemporary analytical methods.

34.4.3 Flotation Methods

Flotation methods are designed to isolate microscopic filth by floating the filth upward, typically in an oil/water-phased system. Insect fragments, mites, and hairs are lipophilic and like to be in the oil phase, thus they float to the surface with the oils. Plant tissues and most related tissues are hydrophilic, and they tend to stay in the water phase. Therefore, separation is based on the principle of affinity for oleophilic solvents. Gravity further helps this process, and larger particles sink. To accomplish the separation of filth from food, a number of solution systems are used to ensure that the majority of the product sinks, while the oils with trapped filth float. The oil phase is trapped off with a Wildman trap flask (Fig. 34.2), filtered, collected on a filter paper, and examined microscopically to determine the amount and kinds of filth present [13].

Flotation is a common method used to determine insect fragments, rodent hairs, and other forms of light filth in wheat flour (AOAC Method 972.32). The acid digestion is used to break down the starch in the flour and allows the other flour constituents to more cleanly separate from the dilute acid solution. Although the AOAC method calls for digestion by autoclaving, AACCI Method 28-41.03 provides for an alternative hotplate digestion, which might be more convenient for some laboratories. The oleophilic property of insect fragments, rodent hairs, and feather barbules allows them to be coated by the mineral oil and trapped in the oil layer for separation and collection on ruled filter paper. The heavier sediments of the digestion are washed and drained from the funnel. Fragments and rodent hairs are reported on the basis of 50 g of flour.

34.4.3.2 Light Filth Flotation Method

Oleophilic filth is defined as light filth. Examples of light filth include insects, insect fragments, hairs, and feather barbules which can be detected in a food product by separating them from the food in the oil phase of an oil-aqueous mixture. The analysis of light filth is accomplished through a series of steps, starting with a pretreatment that removes fats, oils, soluble solids, and fine particulate matter to enhance the wettability of the food. The second step requires mixing the food with a water and oil mixture. The food will remain in the aqueous phase and the light filth will rise to the top with the oil phase. In the third step, the extract with filth elements is poured onto a ruled filter paper using a filter flask and funnel (Fig. 34.3) and examined line by line under a stereomicroscope (Fig. 34.4). After identification and enumeration, the results are reported to provide the following information: (a) whole or equivalent insects (adults, pupae, maggots, larvae, cast skins); (b) insect fragments, identified; (c) insect fragments, unidentified; (d) aphids, scale insects, mites, spiders, etc. and their fragments; and (e) rodent hairs (state the length of the hairs).

Insect fragments, rodent hairs, and other light filth can be isolated from flour samples by an acid hydrolysis method. Sediment products of the digestion are allowed to settle out in a separatory funnel and are drained away. The remaining oil layer is filtered through a ruled filter paper and the contaminants identified microscopically. However, certain grain products such as whole wheat flour contain amounts of bran particles that may result in excessive amounts of material being trapped in the oil layer, making it difficult to identify particles of filth. The “Tween-Versene Method for Insect Fragments and Rodent Hairs in Rye Flour” (AACCI Method 28-60.02) utilizes two chemical agents that tend to suppress bran accumulation in the heptane recovery layer. Tween 80 (polyoxyethylene sorbitan monooleate) is a nonionic agent that appears to have certain surface-active properties that make it a useful adjunct to Na₄EDTA (tetrasodium salt of ethylenediaminetetraacetic acid). In the presence of Tween 80, Na₄EDTA appears to be a depressor for food materials (such as bran and other light plant matter), which otherwise tend to float. It has been suggested that the chelating properties of Na₄EDTA may result in its adsorption onto the surfaces of food particles along with the surfactant Tween 80, thereby preventing an attraction of food particles to oils used to isolate light filth. By preventing plant material from being collected in the heptane layer that is trapped off, contaminants such as oleophilic insect parts (exoskeleton) that are contained in the separating oil are much easier to distinguish and identify. AACCI Method 28-95.01 provides a description of insect fragments and rodent hair characteristics with illustrations [6].

34.4.4 Objectivity/Subjectivity of Methods

Insect parts, rodent hairs, and feather barbules in food products are generally reported as the total number of filth elements counted of each kind encountered per sample unit. They are identified on the basis of objective criteria. However, identifying insect fragments is not a simple task. Training and supervised practice are required to achieve competence and consistency. Some fragments are easily identified on the basis of structural shape and form. Mandibles, for example, are quite distinctive in their shape and configuration; certain species of insects can be determined on the basis of this one structure. In other instances, fragments may be mere chips of insect cuticle that have neither distinctive shape nor form but can be identified as being of insect origin if they have one or more of the characteristics given in Sect. 34.2.2. Experienced analysts should rarely misinterpret fragments.

Isolation of extraneous material from a food product so that it can be identified and enumerated can be a very simple procedure or one that requires a series of several rather involved steps. In the process of isolating fragments from flour by the acid hydrolysis method, for instance, the sample is transferred from the digestion container to the separatory container and then to the filter paper for identification and enumeration. At each of those transfers, there is an opportunity for loss of fragments. Although the analyst may have made every effort to maintain the isolation “quantitative,” there are opportunities for error. Both fragment loss and analyst variation are minimized by common use of standard methods and procedures and by proper training and supervised practice.

Another concern involves the significance of insect fragment counts (as well as particles of sand, pieces of rodent excreta, rodent hairs, etc.) in relation to fragment or particle size. Fragment counts are reported on a numerical basis; they do not reflect the total contaminant biomass that is present. A small fragment is counted the same as a large fragment. The size of the fragment may be a reflection of the process to which a common raw material (e.g., wheat) has been subjected; a more vigorous process produces more and smaller fragments than a less vigorous process. The state of insects may also be a factor. Dead (dried) forms produce greater numbers of fragments than live forms. These factors have been of concern to food processors for some time and have prompted the search for more objective means of determining insect contamination.

34.5.1 Overview

Methods described in Sect. 34.3 are directed primarily at routine quality control efforts to determine if the level of natural or unavoidable defects is below the defect action level. To a certain extent, those routine methods can be used to identify the source of contaminants in processed foods. For example, the identification of certain insect fragments can indicate infestation in the raw commodity rather than in the processing system. However, other more sophisticated techniques offer opportunities to pinpoint the nature and source of other contaminants that may exist unavoidably or due to mistakes, accidents, material or equipment failures, or intentional adulteration.

The detection of insects in stored grain and the quantitation of insect parts present in grain products represent a serious and continuing problem for the grain industry. Approved methods of detection primarily involve visual and microscopic inspection and X-ray analysis which require trained personnel and are time consuming, difficult to standardize, and expensive. The assays for insect contamination are preferred to be highly specific, sensitive, rapid, and inexpensive. Moreover, it ideally should be employable by persons having minimal training, particularly in nonlaboratory settings such as at grain elevator and processing sites.

There are several attempts to develop rapid and efficient methods including the use of nuclear magnetic resonance, sound amplification, and infrared spectrometry as alternatives to presently used chemical techniques mentioned in preceding sections. Most of these techniques are expensive and challenging due to difficulty in quantification and identification of specific infestations. Immunological assays, which have found widespread use in clinical diagnostic settings and also in home use, have been explored to detect insect contamination. These methods are described below as they relate to detecting an infestation.

34.5.2 X-Ray Radiography

X-ray radiography is widely used as a test reference method [14]. Grain processors use it as a means of inspecting wheat for internal insect infestation, which is the main the source of insect fragments in processed cereal products (Fig. 34.5).The existing X-ray techniques enable the classification of at least four stages of insect development by measuring the area occupied by the insect, and an accurate classification is also possible based on visible insect morphology [15]. The use of real-time digital imaging instead of X-ray radiographs to discriminate the infested kernels significantly shortened the X-ray procedures. Conventional film observations give, however, better accuracy (3 % error rate) than the digital images (11.7 %) for infestation by third-larval instars, while the error is less than 1 % for both methods with a more advanced stage of larval development [16].

34.5.3 X-Ray Microtomography

X-ray microtomography (XMT) is an emerging 3-D imaging technique that operates on the same basic principles as medical computed tomography (CT) scanners but has much higher resolution. It is very effective in characterizing various internal structural features, which are not possible with conventional 2-D imaging methods (Chap. 32, Sect. 32.​5.​2). Conventional imaging techniques such as light microscopy, scanning electron microscopy (SEM), and digital video imaging have some limitations: they are destructive in nature, as sample preparation involves cutting to expose the cross section to be viewed. High-resolution XMT has a wide range of applications in science and engineering for which accurate 3-D imaging of internal structure of objects is crucial. XMT is able to capture several features of the internal structures of grain kernels, which are not possible with the conventional imaging methods [17]. Figure 34.6 is demonstration of shadow images taken at several step angles during scanning. A typical scan creates 200–400 of those images that are then used to create axial, sagittal, and coronal views shown in Fig. 34.7.

34.5.4 Electrical Conductance Method

The electrical conductance method is based on monitoring the conductance signals for each single kernel during milling in a single-kernel characterization system (SKCS) which is commonly used for wheat hardness determination [15]. This method is highly accurate for detecting older developmental stages of insects: the percentage of properly classified cases for small, medium, and large larvae and pupae are 24.5, 62.2, 87.5, and 88.6, respectively. The accuracy of this method depends also on insect species (rice weevil and lesser grain borer) and wheat type (soft or hard red winter wheat).

34.5.5 Impact-Acoustic Emission

Acoustic detection is a very promising method for early detection of insects in bulk grain. Acoustic identification of insects is based on their ability to generate sound by eating, flying, egg laying, or locomotion [18]. Despite the fact that it has been a known pest detection method in storage facilities for a long time, reliability and efficacy of acoustic pest detection have been improved only recently as a result of the development of improved acoustic devices and signal processing methods [18].

Impact-acoustic emissions are used as a nondestructive, real-time method for detection of damaged grains and shelled nuts [19]. Kernels are impacted onto a steel plate and the resulting acoustic signal is analyzed to detect damage using different methods: modeling of the signal in the time domain, computing time-domain signal variances and maximums in short-time windows, analysis of the frequency spectrum magnitudes, and analysis of a derivative spectrum. Features were used as inputs to a stepwise discriminant analysis routine, which selected a small subset of features for accurate classification using a neural network. Pearson et al. [19] reported that impact-acoustic emission is a feasible and promising method for detection of IDK, sprout damage, and scab damage. More study is needed to improve accuracy on kernels infested with insects that have not yet emerged from the kernels. The computational cost of classifying a kernel using this technique is very low, allowing inspection of large numbers of wheat kernels very rapidly, ~40 kernels/s. Grain inspectors usually use a 100 g (3,000 kernel) sample to inspect for IDK. This takes an inspector approximately 20 min to analyze manually but can be accomplished in about 75 s with an acoustic system.

According to official standards for wheat grain, a sample as a whole is considered infested when it contains two or more live weevils, or one live weevil and one or more other live insects injurious to stored grain, or two or more other live insects injurious to stored grain [20]. Eliopoulos et al. [18] evaluated the efficacy of bioacoustics in detecting the presence of adult beetles inside the grain mass using a piezoelectric sensor and an acoustic emission amplifier. The accuracy of this system has been reported to be 72–100 % in detecting one to two insects per kilogram of hard wheat, which is the standard threshold for classifying a grain mass “clean” or “infested.” Also, the detection threshold of this system is very low, 0.1–0.5 insects per kilogram of grain, which is not possible with many other detection methods [18].

34.5.6 Microscopy Techniques

Microscopy techniques including light microscopy, fluorescence microscopy, and scanning electron microscopy (SEM) are used to study the structure/function relationships of food but also can be applied to questions of extraneous matter. For example, SEM with energy dispersive spectroscopy (EDS) can be used to determine the nature of metals in products that may be due to equipment failure or intentional adulteration due to tampering [21]. Light microscopy in a polarized mode can be used to distinguish between plastics, glass, and other fiber or crystalline contaminants [22].

34.5.7 Near-Infrared Spectroscopy

Near-infrared spectroscopy (NIRS) is a relatively fast, accurate, and economical technique available to the grain industry for compositional analysis such as water, oil, fiber, starch, and protein in grains and seeds. It also has relatively recent applications in analysis of extraneous matter. NIRS has been used to identify several coleopteran species [23], and to detect parasitized weevils in wheat kernels [24] and external and internal insect infestation in wheat [25–27]. Berardo et al. [28] reported that NIRS predicts the percentage of Fusarium verticillioides infection in maize kernels and the content of ergosterol and fumonisin B1 in meals. In the same way, promising results were obtained when NIRS methodology was applied to detect scab-damaged kernels [29] and estimate deoxynivalenol, ergosterol, and fumonisin in single kernels of wheat [23] and corn [30].

Near-infrared spectroscopy used with a single-kernel characterization system is able to detect later stages of internal insect infestation in wheat with a 95 % confidence [27]. In contrast to other procedures, this system is capable of being automated and incorporated into the current grain inspection process. NIRS also has been compared with the current standard insect fragment flotation method for its ability to detect insect fragments in flour [31]. Fragment counts with both techniques were correlated; however, the flotation method was more sensitive below the FDA DAL of 75 insect fragments per 50 g of wheat flour. NIR spectroscopy was able to predict accurately whether flour samples contained less than or more than 130 fragment/50 g.

Hyperspectral imaging is a recent development that has proven to be effective in grain quality inspection. It is an integration of conventional imaging and spectrometry which involves acquiring images at more than 20 wavelengths within the electromagnetic spectrum [32, 33]. Since it is a complex indirect method, hyperspectral data should be confirmed for homogeneity before statistical modeling through various preprocessing techniques, as described in detail by Ravikanth et al. [32]. In the area of cereal grain quality, hyperspectral imaging has been successfully used to identify fungus-damaged kernels [33–35], insect-damaged kernels [36, 37], defective wheat [29, 38], and wheat classes [39, 40]; to detect contaminants from different origins (animals, other cereal grains, botanical impurities, other contaminants) in cereal grains; [41], and to quantify ergot bodies in cereals [42].

34.5.8 Enzyme-Linked Immunosorbent Assays

To develop an optimal immunological assay for an insect contamination of foodstuffs, antibodies are required that are directed against an insect-specific antigen, preferably protein, likely to be present in any life stage of the contaminating insect or in insect remains. Antigens and antibodies are two key parts of any immunoassay (see also Chap. 27).

Recent research [45] showed that the myosin in fourth instars of the lesser grain borer developing within kernels of wheat degraded within the first 2 weeks when larvae were killed with phosphine, a fumigant commonly used to manage insect infestations in stored grain. Myosin degradation resulted in underestimating insect fragment estimates by about 58 %.