Efficacy of visual evaluation of insect-damaged kernels of malting barley by Sitophilus granarius from various observation perspectives

Highlights

• Six basic locations of Sitophilus granarius exit holes on kernel surfaces were identified.

• Upper view-inspection of a kernel-sample reveals only up to 50% of exit holes.

• The ventral and dorsal exit holes are significantly bigger than the apical exit holes.

• Shaking effect increased the efficacy of visual (upper.-view) detection; but not profoundly.

• The automated methods independent of observation perspectives should be implemented.

Abstract

Although there is growing evidence that arthropod detection methods are not sensitive enough to meet current quality demands, few adequate validations are available for the detection of insect-damaged kernels (IDK- quality parameter). Therefore the aims of this work were to (i) describe categories of exit holes of Sitophilus granarius (L.) in terms of their location on the grain surface; (ii) analyse spatial distribution of exit holes on kernels of three varieties of malting barley; (iii) analyse the sensitivity of visual laboratory methods for IDK detection, and (iv) test if the turning effect of automatic shaking can increase damage visibility from the upper-view perspective of an inspector. Six topological categorises of exit holes were described and photo-documented. It was found that the relative distribution of exit holes on the kernel surface was in decreasing order from lateral, brush, ventral or dorsal (depending on variety), subapical, to germ area. The ventral and dorsal exit holes were significantly bigger than the apical exit holes. It was found that distribution patterns of exit holes on kernels affected their proportional visibility from various observation perspectives. Only 50% of lateral or dorsal exit holes was visible from top or crease views, respectively. The most effective was the side (brush end) view perspective, with visibility of up to 90% of exit holes. Although the shaking effect was statistically significant, it only increased the efficacy of upper view detection perspective from 54% to 58%. This is the first study showing that the inspector upper view of inspected grain samples (without grain rotation) poses a risk of underestimating actual insect-damaged kernels by 50%. Thus, advanced methods that are independent of observation perspectives (X-ray, NIRS, etc.) should be promoted for routine use in commercial laboratories, despite higher cost.

Introduction

According to FAO statistics (http://www.fao.org/faostat), more than 140 million tonnes of barley are harvested at a global scale annually. Consequently, barley grain is traded and processed as feed, food, and malting products. Each year industrial laboratories around the world analyse hundreds of thousands of barley kernel samples for their quality before grain storage or procession. These inspection facilities are either state-supervised laboratories or laboratories operated by national and international commodity trade-companies and food industry facilities (e.g., mills, distilleries, malting and brewing companies). Stored malting barley grain is prone to infestation and damage caused by multiple pest insect (Tebb, 1968; Buchelos and Athanassiou, 1999; Hagstrum and Subramanayam, 2009; Castane and Riudavets, 2015) and mite (Hubert et al., 2009, 2013) species. Specific damage appearing on stored grain (e.g., dark pin-sized puncture marks) can originate from the field where it is inflicted by sucking field insects (e.g., Armstrong et al., 2019a). Therefore, apart from other quality parameters (e.g., dockage, mycotoxins, moisture and protein content, germination, sprouting, etc.), the laboratories also evaluate arthropod infestation and number of insect-damaged kernels (IDK) (Armitage et al., 2004; Fleurat-Lessard et al., 2005; USDA, 2009, 2016a). In USA, if the number of IDK exceeds the legal threshold (i.e., 32 IDK per 100 g sample; USDA, 2009, 2016a), the inspected batch of cereal commodity must be rejected for food or malting purposes. In addition to kernel damage, pest infestation causes grain weight losses, decreases the germination of grain, and affects malting properties of the stored grain (Lazzari and Lazzari, 2012; MacLeod, 2018). Insects and mites affect food safety as they can introduce physical and chemical contaminants or protein allergens (Trematerra et al., 2011; Hubert et al., 2018). Kernels damaged by internal stored product beetles are susceptible to colonization and infestation by other - notably secondary - pest species (Shah et al., 2020). The stored grain infestation by the granary weevil Sitophilus granarius (L.) may be associated with the specific kernels’ volatiles (Piesik and Wenda-Piesik, 2015; Wenda-Piesik et al., 2018).

How grain quality, arthropod infestation, and insect-damaged kernels are most commonly evaluated under practical conditions in laboratories? During the purchase process, the supplier delivers grain using high-volume transport means (e.g., wagon, train, ship, or truck) to the customer’s premises. Before loading and storage of the delivered commodities into silos or floor warehouses, laboratory workers take a limited number of samples from each lot using a manual hollow spear sampler or automatic suction probes. The samples are then subjected to laboratory analysis. If the sample does not have parameters and characteristics according to the negotiated conditions of a particular contract (commodity quality) (Armitage et al., 2004; Freese et al., 2015) or general legislation (food safety), there will be either a price reduction or rejection and return of the whole consignment. Samples are taken from commodity according to ISO standards (ISO 13690, 1999; ISO 6644, 2002) or according to the contracted conditions (Armitage et al., 2004; Freese et al., 2015). Wilkin (1991) and Wilkin and Fleurat-Lessard (1991) were among the first who experimentally demonstrated low sensitivity of ISO-based sampling standards in relation to pest detection. To date, extensive knowledge has accumulated confirming their results (e.g., Elmouttie and Hamilton, 2010; Jian et al., 2014). The additional detection uncertainty regarding sampling is introduced by the specific methods used to extract pests in the laboratory that differ in their sensitivity and efficacy for different pest species (Krizkova-Kudlikova et al., 2007; Stejskal et al., 2008; Hubert et al., 2009; Lukas et al., 2009; Reboux et al., 2019). Clearly, there is growing evidence that methods of infestation detection are not always sensitive enough to meet current quality demands. However, few if any adequate sensitivity validations are available for methods of insect-damaged kernel detection. Traditionally, for insect-damaged kernel detection, simple visual inspection by plain eye has been used. In recent decades, for the detection of internal infestation and presence of cavities inside grain, several technically advanced methods were developed, including X-ray, near infrared spectroscopy (NIRS), conductance mills, magnetic resonance, imaging analysis, etc. (Mahajan et al., 2015; Patrício and Rieder, 2018; Morrison III et al., 2020; Tian et al., 2020). Single kernel assessment of visual cereal grain quality based on automatic image analysis is newly suggested (Armstrong et al., 2019b; McClung et al., 2020). Although the previous methods have been proposed or are under development, the traditional simple visual methods are still prevailing due to their low price and high speed. The procedures follow either various internal company standards or national legislative standards as suggested for example by the USDA in the USA (USDA, 2009, 2016a). Some maximum level of insect damage of cereal grain may also be required (as an additional quality parameter) by customers even when it is not part of any standards or legislation (Flinn et al., 2010).

According to USDA standards, the IDK parameter does not apply to externally feeding grain pests but only to those whose larvae develop inside the kernel, and after eclosion (emergence) the adult beetles leave the kernel through typical round exit (=emergence) holes. For the purpose of grain quality inspection, there are manuals or guidelines with pictures for practitioners for grain sampling and inspecting (e.g., USDA, 2016b). Kernels are inspected by plain eye, simple lens, or under a stereomicroscope (optical camera systems connected with binocular lenses). According to the USDA inspection procedure (2009, 2016a), to detect IDK, a whole surface of the sample kernels must be observed. However, there is little knowledge of impacts on the sensitivity of IDK inspection if the whole surface of each kernel in all samples (or their parts) is not inspected individually and carefully. This situation may occur in the case of misconduct of the official procedure or in a case where IDK is just supportive/side criterion and not the legally required standards, for example, if the kernels can only be observed on a Petri dish or non-transparent tray from the upper side (Fig. 1A). Thorough inspection requires individual spatial rotation of each kernel along two axes (horizontal and vertical) in the sample by tweezers (Fig. 1B and C), which is time-consuming. Occasionally a short time-window may be provided by requirements for the quick processing of multiple samples, containing many hundreds of kernels during a harvest campaign. In addition, a quick upper view inspection approach may be psychologically enforced by “pictorial stereotype” and occasionally the depiction of exit holes of a primary pest, such as Sitophilus spp. or Rhyzopertha sp., mainly on the visible crease/top site of kernels, but rarely on the brush end or germ end.

The objectives of this study included the following goals: (i) To describe and photodocument categories of tunnelling damage (i.e., exit hole; EH) caused by S. granarius; (ii) To estimate and statistically evaluate the distribution of adult exit holes on the kernel surface in three malting barley varieties; (iii) To estimate the size of exit holes at different parts of kernels; and (iv) To compare visibility of the various locations of exit holes from different perspectives; we specifically addressed problems resulting from the impossibility of inspecting the quality of the whole surface of a kernel from one perspective, without grain rotation. (v) The last goal was to test whether automatic shaking could turn kernels with their damaged site up-side to increase visibility.