|
Testing Barley Samples for Potential Insect Infestations with a Conductance Mill
Daniel L. Brabec1,*, Sophia Grothe1, Christos Athanassiou2
Published in Applied Engineering in Agriculture 39(5): 535-541 (doi: 10.13031/aea.15663). 2023 American Society of Agricultural and Biological Engineers.
1Center for Grain and Animal Health Research, USDA ARS, Manhattan, Kansas, USA.
2University of Thessaly, Dept of Agr Crop Production.Volos, Greece.
*Correspondence: Daniel.Brabec@usda.gov
The authors have paid for open access for this article. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License https://creative commons.org/licenses/by-nc-nd/4.0/
Submitted for review on 5 May 2023 as manuscript number ITSC 15663; approved for publication as a Research Article by Associate Editor Dr. Carol Jones and Community Editor Dr. Seung-Chul Yoon of the Information Technology, Sensors, & Control Systems Community of ASABE on 18 September 2023.
The mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the USDA or the by University of Thessaly. The USDA and the University of Thessaly are equal opportunity employers and providers.
Highlights
Abstract. A laboratory mill was developed by Pearson and Brabec (2007) which typically detects 50% to 80% of infested kernels of wheat, brown rice, or popcorn. Barley is another cereal grain of similar size as wheat. Barley is normally sold with its hull attached to the seed, which makes detection of insect infestations more difficult. The objective of this study was to determine the potential of using the conductance mill to detect barley kernels infested by the lesser grain borer, Rhyzopertha dominica (F.), in comparison to detection in wheat. As in previous studies, these experiments found that the conductance mill could detect infested kernels of wheat containing large and medium larvae at a rate of ~90% and ~65%, respectively. For barley, the detection of infested kernels was over ~80% for large larvae and ~40% for medium larvae. Also, we showed that adults that were freely moving throughout the grain mass could also be detected. Approximately ~65% of the adults were detected in wheat while that percentage was reduced to ~35% in barley. The hull on the barley seems to function as an insulator during the conductance measurement and thus reduces detections. Also, the hull seems to affect the feed rate of material through the mill. The feed-rate for wheat was 500 g in 50 s, while the feed-rate for barley was 500 g in ~80 s. Despite these pitfalls, the conductance mill could still be considered as a useful tool with inspecting barley sample, because there was significant detection of insects in the samples and 1000 g sample of barley could be processed in ~4 min.
Some stored product insects, such as the lesser grain borer, Rhyzopertha dominica (F.) (Coleoptera: Bostrychidae) and the rice weevil, Sitophilus oryzae (L.) (Coleoptera: Curculionidae), are internal feeders whose larvae develop while living inside grain kernels. This characteristics impede detections of insects in bulk grain and often require adequate/large sample size to reliably detect infestations. An ability to reliably detect infestations is important for producers because the discovery of two live weevils by a federal grain inspector within a shipment will cause the entire shipment to be graded as “infested” (USDA, 2020) resulting in a lower sale price. The insects per kilogram of grain for barley to be designated as infested has to contain more than 1 weevil, or 1 weevil and any 5 or more other live stored grain insects, or no live weevils but 10 or more other live stored grain insects Infestation levels such as these are relatively common during transport. Grain is regularly hauled in large semi-trucks and railcars capable of carrying 900 and 3000 bushels of grain respectively. Perez-Mendoza et al. (2004) found that 20 of 24 railcar compartments surveyed for insect averaged less than one insect per sample of wheat (Perez-Mendoza et al., 2004). However, four compartments averaged 2, 6, 17, and 19 internally feeding insect species per 3 kg sample.
Several techniques that are currently available for insect detection in bulk grains can be utilized for determining infestation and implementing control measures. Entomologist fundamentally store bioassay samples in environmental chambers for 4-6 weeks and then sift off and count the emerged adults. Regarding sampling for insects within grain bins, Loschiavo (1974) and Athanassiou and Buchelos (2001, 2020) have proposed the use of probe traps, which consist of perforated tubes that are inserted into the top of the stored grain and can provide reliable data about the current infestation levels. When looking at just grain samples received, more technologically advanced methods for determining insect infestation are available and include near infrared technologies, acoustic monitoring, detecting carbon dioxide level changes, X-ray scanning, molecular indicators, and ELISA (Haff and Slaughter, 2004; Perez-Mendoza et al., 2005; Fornal et al., 2007; Chotikasatian et al., 2017; Jian et al., 2016; Thanushree et al., 2018). Nevertheless, some of these methods are only applicable for the detection of certain stored product species or useful for certain types of grains or bulk commodities. Other techniques can detect the presence of insect fragments, which could include both live and dead individuals. Some of these methods are rather laborious, time-consuming and require specialized personnel.
The laboratory mill, developed by Pearson and Brabec (2007) overcomes many of these pitfalls. The instrument detects stored product insects in bulk commodities by monitoring for changes in electrical conductance of crushed wheat and can process a kilogram (~30,000 kernels) of wheat in about 3 min. In brief, this device mills kernels and then uses an electronic circuit and software to detect moisture differences between uninfested and infested kernels. This method has been evaluated with success for the detection of various stored product insect species on different grains including infested kernels containing larvae of R. dominica (Pearson and Brabec, 2007; Brabec et al., 2010, 2012, 2017). Previous research with wheat, rice, and popcorn found ~80% of the large larvae and medium larvae in internally infested kernels and ~40% of the smaller larvae. (Brabec et al., 2010, 2012, 2017). In addition, the instrument can successfully detect grain infested by other internal feeders, including the maize weevil, Sitophilus zeamais Motschulsky (Coleoptera: Curculionidae) and Sitophilus oryzae (L.) (Coleoptera: Curculionidae). Moreover, given that the method is based on discriminating moisture content between infested kernels versus the average moisture content of the grain sample, the original intended operation was based on moisture levels for safe storage of wheat, which is estimated to be at ~12.0%. This method has not been evaluated on barley. In this context, our main objective was to test the conductance mill method for detecting infested barley samples in comparison to wheat. Another objective of this study was to investigate if adult insects, which feed outside the kernels, would be detected compared to internally infested grain at normal and higher levels of grain moisture content.
Materials and Methods
Conductance Mills
The earlier design of the USDA-ASR conductance mill included 9.8cm wide rolls with the two rolls turning as the same rotational speed of 30 rpm (Pearson and Brabec, 2007) which basically crushed the seeds. The current conductance mill design uses narrower rolls which turn at different speeds which shears the seeds and increases the detection of smaller larvae. The rolls of these mills were 3.5 cm wide and 7.6 cm in diameter. The gap between the rolls was ~0.53 mm (~0.021 in.). The rolls operated in a differential speed ratio of 1:1.5 with the main roll operated at 47 rpm while the second roll turned at 70 rpm. One roll of each mill was electrically insulated and connected to a supple voltage of 5 VDC. Electronic circuitry was mounted on each mill (AEW Consulting, Lincoln, Neb.) to monitor the voltage between the two rolls (fig. 1) and output this signal to software on a personal computer. Two electrically conductive grain mills were assembled for this study in the USDA-ARS, Center for Grain and Animal Health Research metal shop in Manhattan, Kansas (Mill 1-A and Mill 2-B).
 |
| Figure 1. Laboratory conductance mill (left) and the electronic circuitry (right) on the side of the frame for monitoring voltage and conductance between the rolls. |
Experimental Overview
Soft red winter wheat and barley grain were used for the experiments. The soft red winter wheat (Coker 9436) was obtained from Cherry Farms, North Carolina in 2013. The barley (ND Genesis) was obtained from North Dakota State Foundation at Carrington, North Dakota, in 2021. The first phase of the experiment was to test grains having a wide range of moisture content, because the conductance of milled grain is fundamentally related to the moisture content of the grain. Testing at higher moisture levels were done to show the upper limits of detection. Grain samples were prepared with varied moisture contents, between 11% and 17%. The samples were passed through the laboratory mill and the data points for conductance signal were collected at a rate of 450 points/s for all samples. For example, a 60 s milling time would collect 27,000 data points for the conductance signal. The conductance voltage signal has a full range of 5 VDC for 100% of scale and the voltage is converted to %full scale. The %voltage data points were averaged and the standard deviation of the data points were determined and reported for each wheat sample at each moisture level.
The second phase tested the ability of each mill to detect R. dominica adults and kernels infested with medium and large larvae. The adult insects and the infested kernels were added to barley having either 12% or 13.5% of moisture content. Grain moisture content of 12% is considered the basic long term moisture content for cereal grains. The higher moisture content, 13.5%, was used to test the potential robustness of the method with barley. Experimental factors included three grain samples (12% wheat, 12% barley, 13.5% barley), four levels of infestation (0, 10 adults, 10 large larvae, or 10 medium larvae), two conductance mills (Mill 1-A and Mill 2-B), and five replicates for each factor combination which resulted in 120 observations in total.
Tempering Grains and Milling at Varied Moisture Content
About 40 kg of wheat and 80 kg of barley were acquired for the experiments and oven moisture content was determined on their bulk using whole grain procedure (ASABE Standards. 2017). From the initial moisture content, the quantity of tempering water was determined for reaching seven levels of moisture: 11.1%, 12.1%, 13.1%, 14.1%, 15.1%, 16.1%, and 17.1%. For each of the seven moisture levels, ~1.0 kg wheat or barley were placed into a mixer, water was added, and then the grain was tumbled for 20 min. All tempered samples were placed into plastic bags and held for over 24 h in a room at approximately 72°F. Oven moisture content was measured after 24 h on the tempered set of samples. If a sample was not within 0.2% of the specified level, additional cycle of conditioning followed. Including the days needed to determine moisture contents, the tempered grain was held over 2 days before testing with the conductance mill which allowed the grain to fully equilibrate.
The wheat and barley samples of varying moistures were subdivided into 250 g samples. For each mill, duplicate samples or two × 250 g samples were run and the voltage signals from the conductance circuit was collected with “Bugsmart” software (AEW Consulting, Lincoln, Neb.). The conductance signal data were processed to determine the average%voltages and the standard deviations of the voltage signals for each grain at all seven levels of moisture content.
Preparation of Insect Colonies
The R. dominica colonies used for the experiments were obtained from USDA-ARS-CGAHR (Manhattan, KS). Colonies were prepared by placing ~250 adults of R. dominica on 400 g portions of either wheat or barely that was conditioned to ~13% moisture. These colonies were stored in an environmental chamber maintained at 27°C and 60% relative humidity. After 6 weeks, the colonies were sieved in a number 10 U.S. Standard Testing Sieve to remove the adults and fines. The grain and infested kernels were returned to the jars and allowed to incubate until needed.
X-rays of Lesser Grain Borer Colonies
Grains from these colonies were arranged in a single layer on small grids and x-rayed to identify internally infested kernels. (MX20-dc44, Faxitron X-ray Corp., Wheeling, Ill.). Due to the size of the instrument, only ~7.0 g portions of grains or ~150 seeds could be x-rayed at the same time and each portion generally only contained 3 to 10 infested kernels. To find the kernels infested with medium or large larvae, over 100 x-rays were required to inspect and pick the infested seeds used in the experiments. Once the images were obtained, larvae inside infested seeds were classified as either large (fourth instar) or medium (third instar) (Kirkpatrick and Wilbur, 1965) (fig. 2). The infested seeds were packaged with 10 infested seeds per bag. These experiments included five replications on two conductance mills. Thus, 10 bags of large larvae and 10 bags of medium larvae infesting wheat were prepared, along with 20 bags of large larvae and 20 bags of medium larvae infesting barley.
 |
| Figure 2. X-rays of infested wheat and barley. (a) infested wheat with large larvae, (b) infested wheat with medium larvae, (c) infested barley with large larvae, and (d) infested barley with medium larvae. |
Trials for Infested Grain Detection
The main experiment included four levels of infestation: control, 10 adults, 10 kernels with large larvae, or 10 kernels with medium larvae. The adults or infested kernels were carefully mixed with 500 g of grain immediately prior to analysis on the conductance mill. The “Bugsmart” software (AEW Consulting, Lincoln, Neb.) was initialized, then the grain samples were ground through a conductance mill. The 12% and 14% moisture grain samples passed through the mill freely. The mill rolls are fitted with scrapers to help dislodge occasional material and no additional cleaning was required between samples. A total of 120 × 500 g samples were processed in these trials. One 500 g sample contained either 10 adults or 10 large larvae or 10 medium larvae or 0 for control. And the 500 g infested samples were replicated five times per grain type and moisture content and insect life stage.
One of the basic signal processing in the software included computing the derivative of the voltage signal. The derivative of the voltage removed variability caused by varied moisture contents. The infested seeds produced a distinctly different voltage spike above the baseline signal from the grain. A derivative threshold of 2.0 was used and any voltage spikes which exceeded the threshold were counted as infested kernels. This threshold level was set to minimize false detects from non-infested barley samples at 14% moisture. The detection threshold can be adjusted depending on the typical grain and moisture conditions of the samples being evaluated and a good example of testing varied threshold levels is given in the earlier test with the conductance mill and popcorn (Brabec et al., 2017).
Statistical analysis included the Generalized Linear Mixed (GLM) model and was fit using the GLIMMIX Procedure in SAS® 9.4, SAS/STAT 15.1 (© 2016 by SAS Institute Inc., Cary, N.C.). Insect detects were considered a binomial response variable. The variable “Detects / 10” was analyzed for differences between grain types and insect life stages. The parameters of the GLM model included the fixed treatment effects of grain type with moisture contents and insect treatment groups, and their respective interaction. The two mills in the study represent a random source of variation and are included in the GLM model as a random effect along with the interaction with Grain and Insect. In the GLIMMIX Procedure, the binomial distribution is fit with the logit link, and least squares means estimates and standard errors were back-transformed to proportions (values between 0 and 1) with the ILINK option in the LSMEANS statement.
  | ||||
| (a) | (b) | |||
| Figure 3. Average voltage (% full scale) resulting from different moisture content in wheat and barley (a) for Mill 1-A and (b) for Mill 2-B. The vertical error bars are ±standard deviation of the voltage signal. | ||||
Mills Feed-Rates and Ground Material Particle Distributions
The feed rate through each mill was determined by processing 500 g of material and timing the process to mill the grain. The grain used was non-infested grain. Secondly, two 50 g samples of wheat and barley were processed through each mill, Mill 1-A and Mill 2-B, and the ground material was collected and sifted for 60 s with a vibratory shaker to determine the distribution of particles. The sieve stack included a Tyler #20, Tyler #40, Tyler #100, and pan. The fractions of materials over each sieve were weighed, and the percent computed for each grain and mill. The particle distribution data obtained from barley and wheat after they had been passed through the two mills were compared using a one-way ANOVA (P>0.05).
Results and Discussion
Conductance Mill Response to Varied Moisture
There was a nonlinear relationship between the average voltage level and moisture content for both wheat and barley (fig. 3). The average conductance voltage was around 1.8% (full scale) with wheat near 12% moisture content, while the average voltage for barley at 12% was around 8.4% (full scale). The standard deviation of the voltage for wheat at 12% and 14% was 0.4% and 1.8%, respectively. The standard deviation of the voltage for barley at 12% and 14% was larger and was of 2.8% and 8.2%, respectively. The figure shows that barley has marginally higher average voltages than wheat for similar moisture grain. Grain elevators in the U.S accept harvested wheat with moisture levels lower than 13.5% into the grain marketing channels. Grains with moisture levels above 13.5% require additional handling and blending as serious quality deteriorations could occur during long term storage (Cahagnier et al., 1993; Athanassiou and Arthur, 2018).
Voltage Chart vs. Derivative Chart
Typically, the conductance mill software displays the conductance voltage levels in real time as the grain is processed (fig. 4). When an infested seed is detected between the rolls, the fluid from the insect causes a temporary short circuit and the voltage circuit produces a distinct spike. The conductance mill can discriminate these voltage spikes when the grain moisture content is at normal storage moisture content of 14% or less. As seen in figure 4, there are many shorter peaks in the voltage signal; however, taking the derivative of the voltage attenuates most of the short peaks and the larger peaks are more distinct (fig. 5).
Conductance Mill Basic Operation vs. Detection of Infested Grain
There were 30 control samples out of 120 samples (each 500 g) in these trials and only 1 false-detect was recorded. If we estimated that 1000 seeds weigh ~33 g, then 500 g contains ~15,000 seeds. And 30 samples × 15,000 seeds/sample equal 450,000 seeds which were crushed in the mills for the control samples, but only one seed was detected falsely. Hence, the instrument produced minimal false positives in these series of tests. However, earlier studies by Brabec et al. (2010, 2017) showed that small pieces of dirt could produce false-positive results. False-positive from bits of soil could be minimized by careful cleaning methods prior to testing with the conductance mill.
 |
| Figure 4. Chart of conductance voltage for 12.1% barley. In this trial, 10 infested kernels with medium larvae were added to 500 g of barley, and 5 infested kernels were detected. |
 |
| Figure 5. Chart of signal derivative with 12.1% barley. The derivative attenuates the normal voltage signal. In this trial, 10 infested kernels with medium larvae were added to 500 g of barley, At threshold of 2.0, 5 infested kernels were detected. If threshold were lowered to 1.1, potentially more peaks detected as indicated by the ‘z’. |
For this experiment, there were two main factors; grain type and insect infestation life stage (treatment). The statistical evaluation show that the main effect on insect detection with the conductance mill was the insect infestation type; either adults or large larvae or medium larvae. The F ratio for the “insects” or insect life stages was 47.9. Insect life stage was by far the most significant source of variability for these experiment. Grain type was also a significant source of variability as seen with the lower F ratio of 10.5 (table 1). And the wheat and barley are similar in size, however the barley had hull surrounding each kernel. This instrument was initially made to test wheat and performs better with wheat. The two mills were fabricated to be very similar mechanically. The statistic covariance analysis found the mills contributing a minor fraction of the residual estimates. The basic operations did show some differences between the mill’s feed-rates and detections, but goal of manufacturing these units was to minimize gaps and operational differences.
| Table 1. Statistical binomial evaluation on experimental factors grain_MC and insects and their effects. | ||||
| Effect | Num DF | Den DF | F value | Pr > F |
| grain_MC | 2 | 8 | 10.46 | 0.0059 |
| insects | 2 | 8 | 47.89 | 0.0001 |
| grain_MC*insects | 4 | 8 | 1.35 | 0.3325 |
Detection of insects varied with life stage. For wheat, ~6.8 out of 10 adults were detected while for barley only ~3 to 4 of 10 adults were detected (table 2). For both wheat and barley, the detection levels were high for kernels infested with large larvae, with Mill 1-A averaged detecting over 8.8 out of 10 and with Mill 2-B averaged detecting over 7.6 out of 10 large larvae infested kernels. The medium larvae were detected to a lesser extent. The mills averaged 7.0 and 5.8 medium larvae in wheat. The two mills detected medium larvae at a lower rate in barely, 5.4 and 3.8 at 12% moisture and 4.4 and 4.2 at 13.5% moisture. The conductance mill’s electronic circuit detects signal peaks after the fluids of the smashed insect mixes with the crushed grain particles and produces a brief short-circuit or electrical connection between the rolls. Larger life stages have more fluid to contribute to these electrical signal peaks than small larvae as seen by the test data.
| Table 2. Detection of adults and internal insects with Mill 1-A and Mill 2-B. | ||||
| Grain and Moisture Content | ||||
| Treatment | Wheat at 12% | Barley at 12% | Barley at 13.5% | |
| Mill 1-A | Control | 0.0 ± 0.0 | 0.0 ± 0.0 | 0.0 ± 0.0 |
| Adults | 6.4 ± 0.5 | 3.1 ± 1.2 | 2.8 ± 1.1 | |
| Large larvae | 10.0 ± 1.2 | 9.8 ± 0.4 | 8.8 ± 0.4 | |
| Med larvae | 7.0 ± 0.7 | 5.4 ± 1.1 | 4.4 ± 0.9 | |
| Mill 2-A | Control | 0.2 ± 0.4 | 0.0 ± 0.0 | 0.0 ± 0.0 |
| Adults | 7.0 ± 1.6 | 3.8 ± 1.3 | 4.8 ± 0.8 | |
| Large larvae | 8.4 ± 0.9 | 8.0 ± 1.0 | 7.6 ± 1.3 | |
| Med larvae | 5.8 ± 0.8 | 3.8 ± 1.6 | 4.2 ± 1.6 | |
[a] Each trial processed 500 g of grain which contained either 0 or 10 adults or 10 infested kernels. Each factor was replicated 5 times per treatment and the average and standard deviation is shown. | ||||
For the tests with barley, our data show that medium larvae of R. dominica are less likely to be detected than large larvae and adults, which stands in accordance with the data that have been reported for other grains and stored insect species (Brabec et al., 2010, 2012, 2017). When the larvae are the smaller, we expect even lower detection. However, when bulk grain is suspect of infestations, increasing the sample size or number could increase the potential of detections all the life stages which would exist. The conductance mill is easy to use and provides quick indications of potential risk of live infestations.
The threshold detection line for these experiments was set to detect derivative peaks above 2.0. This line was applied to all samples used in these experiments, to both wheat and barley and for grain moisture contents ~13.5% and lower. The threshold line could be changed and lowered to potentially detect smaller peaks. In figure 5, an alternate line was added at a level of 1.1. When using this alternate threshold level, there would have been 8 detects of the medium larvae. Lowering the threshold line is possible if all the grain samples were known to be below 13.5% moisture or if the grain is known to produce a lower baseline signal.
Brabec et al. (2017) found an underestimation of infested kernels in maize; a detection rate of 80% was observed for kernels infested with pupae and medium larvae of S. zeamais, while <50% detection rate was observed for kernels infested with small larvae. This underestimation in detecting R. dominica individuals needs to be accounted for when considering the level of infestation within the grain sample. If the conductance mill detects 10 insects per kilogram, there would be more infested kernels actually in the sample. If the conductance mill detected 60% to 70% of the varied life staged, then the sample probably contained 13 to 15 infested kernels although only 10 were detected. Still, the instrument detections can be linked with specific grain management action thresholds and decision support systems.
 |
| Figure 6. Particle distributions per sieve (%) from milled soft red winter wheat and barley which were at 12.1% moisture and processed with mill 1-a or mill 2-b (df=1, 26; f=0.78; p=0.38). |
Sifted Fractions from Grain Processed Mill 1-A and Mill 2-B
These laboratory mills operate at narrow roll gaps similar to the first break in wheat milling and with differential roll speeds (1:1.5) to increase the shearing of each kernel and the potential detections of internal insects. Although Mill 1-A and Mill 2-B are very similar in all physical dimensions, they still had some operational differences, particularly with barley. The feed rate for wheat through Mill 1-A was ~53 s for 500 g while for Mill 2-B, the feed rate was ~42 s for 500 g wheat. With the barley samples, Mill 1-A required ~100 s to feed 500 g barley, while Mill 2-B required ~70 s. Figure 6 shows that Mill 1-A produced 62% wheat with particles larger than sieve #20 while Mill 2-B produced on 67%. The milled fractions for barley were 75% and 80% over sieve #20. Barley tended to mill into larger particles as compared to the soft wheat sample. However, there were no significant differences (P = 0.38), regarding in the overall particle size data of ground wheat and barley grains produced using Mill 1-A and Mill 2-B.
Conclusions
The conductance signals varied with moisture content, wherein drier samples resulted in lower voltage signals and deviations. These signals were also distinctly different from the signal peaks produced by internally infested seeds. Also, the barley samples produced conductance voltage level and standard deviations which were larger than those with wheat at varied moisture contents. We hypothesize that the presence of hull around the barley kernels might have contributed to differences in detection sensitivity and reduced feed-rate through the mill. At moisture levels higher than 13.5%, the combination of higher average values and higher deviation reduces the instrument’s ability to discriminate exceedingly wet kernels from the internally infested seeds.
For all insect life stages, the conductance mill underestimated the presence of R. dominica individuals in comparison with the actual number of insects or infested seeds included in each sample. The instrument detected in the barley samples ~80% of the large larvae and ~40% of the medium larvae during testing. For wheat, the instrument detected over 90% of the large larvae and ~65% of the medium larvae. This underestimation in detecting infested kernels needs to be accounted for when considering the actual level of infestation within the grain sample. Hence, if the instrument detects 10 insects per kilogram, there are probably 15 to 20 infested kernels per kilogram.
The conductance mill is a potentially useful method for testing incoming grain for live internally infesting insects and can process a 1000g sample within ~5 min. The potential to detect internal insects in barley shows significant detections even though the conductance mill was more effective on wheat. Early detection of infested grain is key for the timely implementation of management practices that will prevent further development of insect populations.
Acknowledgments
This article reports the results of research only. The authors would like to thank Alan Walker (AEW consulting, Lincoln, Neb.) with his assistance in preparing circuit boards and technical support with the Bugsmart software.
References
ASABE Standards. (2017). ASAE S352.2: Moisture measurement — Unground grain and seeds. St. Joseph, MI: ASABE.
Athanassiou, C. G., & Arthur, F. H. (2018). Recent advances in stored product protection. Berlin, Heidelberg: Springer. https://doi.org/10.1007/978-3-662-56125-6
Athanassiou, C. G., & Buchelos, C. T. (2001). Detection of stored-wheat beetle species and estimation of population density using unbaited probe traps and grain trier samples. Entomol. Exp. Appl., 98(1), 67-78. https://doi.org/10.1046/j.1570-7458.2001.00758.x
Athanassiou, C. G., & Buchelos, C. T. (2020). Grain properties and insect distribution trends in silos of wheat. J. Stored Prod. Res., 88, 101632. https://doi.org/10.1016/j.jspr.2020.101632
Brabec, D., Dowell, F., Campbell, J., & West, M. (2017). Detection of internally infested popcorn using electrically conductive roller mills. J. Stored Prod. Res., 70, 37-43. https://doi.org/10.1016/j.jspr.2016.11.002
Brabec, D., Pearson, T., & Flinn, P. W. (2012). Detection of lesser grain borer larvae in internally infested kernels of brown rice and wheat using an electrically conductive roller mill. Cereal Foods World, 61, 65-70. https://doi.org/10.1094/CPLEX-2012-0316-01R
Brabec, D., Pearson, T., Flinn, P., & Katzke, D. (2010). Detection of internal insects in wheat using a conductive roller mill and estimation of insect fragments in the resulting flour. J. Stored Prod. Res., 46(3), 180-185. https://doi.org/10.1016/j.jspr.2010.04.003
Cahagnier, B., Lesage, L., & Richard-Molard, D. (1993). Mould growth and conidiation in cereal grains as affected by water activity and temperature. Lett. Appl. Microbiol., 17(1), 7-13. https://doi.org/10.1111/j.1472-765X.1993.tb01423.x
Chotikasatian, C., Chayaprasert, W., & Pathaveerat, S. (2017). A study on the feasibility of quantifying the population density of stored product insects in air-tight grain storage using CO2 concentration measurements. J. Stored Prod. Res., 73, 21-29. https://doi.org/10.1016/j.jspr.2017.05.005
USDA AMS FGIS. (2020). Grain inspection handbook – Book II: Grain grading procedures. Washington, DC: USDA, AMS, FGIS. Retrieved from https://www.ams.usda.gov/sites/default/files/media/Book2.pdf
Fornal, J., Jelinski, T., Sadowska, J., Grundas, S., Nawrot, J., Niewiada, A.,... Blaszczak, W. (2007). Detection of granary weevil Sitophilus granarius (L.) eggs and internal stages in wheat grain using soft X-ray and image analysis. J. Stored Prod. Res., 43(2), 142-148. https://doi.org/10.1016/j.jspr.2006.02.003
Haff, R. P., & Slaughter, D. C. (2004). Real-time x-ray inspection of wheat for infestation by the granary weevil, Sitophilus granarius (L.). Trans. ASAE, 47(2), 531-537. https://doi.org/10.13031/2013.16022
Jian, F., Doak, S., Jayas, D. S., Fields, P. G., & White, N. D. (2016). Comparison of insect detection efficiency by different detection methods. J. Stored Prod. Res., 69, 138-142. https://doi.org/10.1016/j.jspr.2016.07.008
Kirkpatrick, R. L., & Wilbur, D. A. (1965). The development and habits of the granary weevil, Sitophilus granarius within the kernel of wheat. J. Econ. Entomol., 58(5), 979-985. https://doi.org/10.1093/jee/58.5.979
Loschiavo, S. R. (1974). Laboratory studies of a device to detect insects in grain, and of the distribution of adults of the rusty grain beetle, Cryptolestes ferrugineus (Coleoptera: Cucujidae), in wheat-filled containers. Can. Entomol., 106(12), 1309-1318. https://doi.org/10.4039/Ent1061309-12
Pearson, T., & Brabec, D. L. (2007). Detection of wheat kernels with hidden insect infestations with an electrically conductive roller mill. Appl. Eng. Agric., 23(5), 639-645. https://doi.org/10.13031/2013.23662
Perez-Mendoza, J., Flinn, P. W., Campbell, J. F., Hagstrum, D. W., & Throne, J. E. (2004). Detection of stored-grain insect infestation in wheat transported in railroad hopper-cars. J. Econ. Entomol., 97(4), 1474-1483. https://doi.org/10.1093/jee/97.4.1474
Perez-Mendoza, J., Throne, J. E., Maghirang, E. B., Dowell, F. E., & Baker, J. E. (2005). Insect fragments in flour: Relationship to lesser grain borer (Coleoptera: Bostrichidae) infestation level in wheat and rapid detection using near-infrared spectroscopy. J. Econ. Entomol., 98(6), 2282-2291. https://doi.org/10.1093/jee/98.6.2282
Thanushree, M. P., Vimala, B. S., Moses, J. A., & Anandharamakrishnan, C. (2018). Detection techniques for insect infestation in stored grains. Agric. Eng. Today, 42(4), 48-56.