Movement and distribution of Sitophilus zeamais adults and relationship between their density and trapping frequency in wheat bulks under different grain temperatures and moisture contents

Movement and distribution of Sitophilus zeamais adults and relationship between their trapping frequency and insect density were determined in one tonne wheat with 11.5 ± 0.2, 12.5 ± 0.1 and 13.5 ± 0.2% moisture contents (wet basis) inside cylindrical bins at 20, 25 and 30 °C. The introduced adult densities were 0.5, 1.0, and 2.0 adults/kg wheat (A/kg). Inside each bin, 15 probe traps were vertically installed in three layers. In each layer, five traps were installed at the center and half radii of the bin individually. The trapped adults were counted daily up to 7 d, and then 15 kg wheat was sampled at each trap location. The insect number in each sample was counted.The recovered adults in the top layer were less than that in the middle layer, while no significant difference between at middle and bottom layers. In each layer, adults preferred the center location at 0.5 A/kg, while this trend was unidentified at other densities. There was an autocorrelation of insect densities at adjacent locations at ≥ 1.0 A/kg, while no autocorrelation in the vertical direction. The adults had an aggregation distribution at any temperature, moisture content and adult density. The normal and Poisson model could not be used to describe the distribution of the count frequency of the adults in the samples, while the negative binomial model could describe this distribution in 18 out of 27 sampling sets at ≤ 1.0 A/kg. The trapping frequency was strongly influenced by the trapping period and increased with the increase of insect densities. The relationship between adult densities and trapping frequencies at the sampling locations had a significant difference at different grain temperatures, while no difference at different moisture contents, grain depths, and locations.     

Geostatistical analysis of adult Rhyzopertha dominica (F.) (Coleoptera: Bostrichidae) in wheat stored at constant temperatures

Insect monitoring and sampling programmes are used in the stored grains industry for the detection and estimation of insect pests. At the low pest densities dictated by economic and commercial requirements, the accuracy of both detection and abundance estimates can be influenced by variations in the spatial structure of pest populations over short distances. Geostatistical analysis of Rhyzopertha dominica populations in 2-dimensions showed that, in both the horizontal and vertical directions and at all temperatures examined, insect numbers were positively correlated over short (0–5 cm) distances, and negatively correlated over longer (≥10 cm) distances. Analysis in 3 dimensions showed a similar pattern, with positive correlations over short distances and negative correlations at longer distances. At 35 °C, insects were located significantly further from the grain surface than at 25 and 30 °C. Dispersion metrics showed statistically significant aggregation in all cases. This is the first research using small sample units, high sampling intensities, and a range of temperatures, to show spatial structuring of R. dominica populations over short distances. This research will have significant implications for sampling in the stored grains industry.     

Delineating the effect of gaseous ozone on disinfestation efficacy,protein quality,dehulling efficiency,cooking time and surface morphology of chickpea grains during storage

Ozone is one of the acceptable and economically viable techniques for treating grains during storage for its residue-free and environment-friendly nature. A laboratory-scale and pilot-scale ozone disinfestation system for chickpea grains was developed. Chickpeas conditioned to different moisture levels (10, 12, 14% wb), infested with Callosobruchus maculatus were stored at different temperatures (15, 23, 31 °C), with varying bed thickness (50, 75, 100 mm), and samples were disinfested by ozone gas at different concentrations (500, 750, 1000 ppm). Ozone concentration, moisture content of the grain, grain bed thickness, and storage temperature had significant (p < 0.01) effects on the disinfestation of C. maculatus insects. 100% insect mortality (adult and egg) and 79% germination were achieved when treated at 1000 ppm ozone for 5 consecutive days. The optimum treatment parameters were tested in pilot-scale bins (250 kg) and the germination ability, protein, milling, and cooking qualities of chickpea grain were evaluated after 6 months storage period. The protein content, germination ability, surface morphological analysis, milling, and cooking time were shown to profoundly deteriorate in the untreated samples as a contrast to that of the ozone-treated samples. Ozonation at higher doses resulted in the crack generation on the surface of the insect eggshell (5–38 μm) and chickpea pericarp (326–711 nm), with added advantages of improved dehulling efficiency of chickpea by 3.1% and reduction in the cooking time by 15 min. The SDS PAGE profiling exhibited major deviations in the protein bands in the stored chickpea. The essential amino acids (EAA) and total amino acids (TAA) decreased in the ozone-treated sample and control sample with storage; however, major deterioration was observed in the control sample stored for 6 months.     

Statistical models to predict densities of Sitophilus zeamais adults in wheat warehouse using probe traps

Statistical models were developed to estimate the density of Sitophilus zeamais adults in a flat storage warehouse holding 230 tonne of wheat with 12.4 ± 0.2% (wet basis) moisture content at 18–33 °C. Published data were used to develop the models. To valid the developed models, trapping and manual sampling with 15 kg of sampling unit at five locations were conducted in the warehouse for more than 10 wk. Insect densities were calculated by using the developed models and the trapping frequency and temperatures at the sampling locations. The developed models were validated by calculating the average magnitude of relative error (MRE), regression slope, and residues between the calculated and measured insect densities. Three models were developed, and the best model was a linear model. The linear model could explain 96% of the measured insect densities. The temperature on the density prediction has a natural exponential effect.     

储粮扦样环节中扦样点数分析研究

为了监测粮仓中储藏小麦在分区后选取不同扦样点数时的质量指标状况，开展平房仓储粮分区分层扦样试验，确定扦样点数的选择及其布点方案，探索所扦取样品的代表性且能客观有效反映分区内储粮质量的可行性。方法:以平房仓中储藏小麦为研究对象，选取两个仓房中的3处区域作为平行实验单元，按标准进行分层分区后，依次选取3~9个扦样点数分别扦样，并对其容重和水分进行测定，采用SPSS进行单因素方差分析，依据不同扦样点数之间的质量指标差异，探讨储粮扦样时的扦样点数设置。 结果 小麦质量指标变化波动状况与扦样点数之间无明显联系；当选取的扦样点数不少于5个时，检测结果之间均无显著性差异；当选取扦样点数为5个时，能较好反映该分区内的小麦容重状况；选取扦样点数为5~8个时，能较好反映该分区内的小麦水分状况。结论 试验数据统计分析结果证明了每个分区选取 5 个扦样点（中心1个、四角各1个），这样既可减少了样本量，又能扦取有代表性的样品，能合理有效反映该区域内小麦的质量状况。     

Insect population dynamics in commercial grain elevators

Data were collected in 1998-2002 from wheat stored in commercial grain elevators in south-central Kansas. Bins at these elevators had concrete walls and were typically 6-9 m in diameter and 30-35 m tall. A vacuum-probe sampler was used to collect grain samples in the top 12 m of the wheat in each bin. The primary insect species found in the wheat samples were: Cryptolestes ferrugineus, Rhyzopertha dominica, and Tribolium castaneum. In the top 3.7 m of grain, R. dominica, C. ferrugineus, T. castaneum and Sitophilus oryzae made up 44, 36, 19 and 1% of the insects found in the samples, respectively. From 3.8 to 12.2 m, R. dominica, C. ferrugineus, T. castaneum and S. oryzae were present at 84, 8, 8, and 1%, respectively. The most prevalent species also changed over time. In June, the start of wheat harvesting and storage in Kansas, insect density was low in the bins. At this time, C. ferrugineus was the most common insect, and it was found mostly in the top grain sample (0-1.2 m). In September through November, C. ferrugineus and R. dominica were at similar densities; however, from February to March, R. dominica was more common.Generally, insect density was greatest at the top and decreased with grain depth. Very few insects were found in samples collected from greater than 12 m (most of the bins contained grain to depths of 24-36 m). Insect density for all species increased rapidly from June through October. During this period less than 20% of the bins had economically significant insect densities (>2 insects/kg). From October until February, the average insect density remained fairly constant but it was greatly reduced in April, May, and June. Bins that had insect densities >2 insects/kg tended to be located adjacent to other heavily infested bins.     

Development and evaluation of sequential sampling plans for Cryptolestes ferrugineus (Stephens) (Coleoptera: Cucujidae) infesting farm-stored wheat

The development and evaluation of appropriate sampling plans are needed for cost-effective management of stored-product insects. Sequential sampling plans, which are based on a variable sample size, are generally more cost effective than plans based on a fixed sample size. For adults of the rusty grain beetle, Cryptolestes ferrugineus (Stephens), we developed sequential sampling plans based on complete counts and the presence/absence of insects in 0.5 kg grain samples removed with a trier from the top 1 m of wheat stored in farm bins. Insect count data were used to develop a sampling plan for estimating the density of C. ferrugineus with a fixed level of precision. The presence/absence data were used to develop a sampling plan for classifying the C. ferrugineus infestation level relative to an action threshold. The performance of these sampling plans was evaluated (validated) using independent data sets and an IBM-PC software program specifically designed to test the plans. This is the first paper illustrating the development and evaluation of sequential sampling plans for a stored-product insect.     

Preference of Tribolium confusum (Coleoptera: Tenebrionidae) adults to grain patches under constant and rising temperatures

Heat treatment, an environmental friendly insect control method, has been applied to disinfest grain storage and food processing facilities. Movement of Tribolium confusum adults into grain piles (0.1, 0.5, 1.0, or 5.0 kg oat groats) was investigated at constant temperatures (30, 35, 37.5, 40, 45, or 50 °C) or rising temperatures (at a rate of 0.05 °C/min from 24 to 54 °C in laboratory, or from approximately 30 to 60 °C in a mill during a heat treatment). Adults moved among locations, paper surface, surface of grain pile and inside grain pile. There was preference for the grain at ≤37.5 °C. No single adult always stayed inside grain piles at 30 °C. At rising temperatures, adults preferred grain piles (on the surface of or inside grain piles) and moved into cold grain regardless of adult introduction methods (introduced on surface of groats piles, on paper, or premixed with groats). Rising temperatures did not drive adults out of grain piles during heat treatment.     

Mortality and movement of Cryptolestes ferrugineus and Rhyzopertha dominica in response to cooling in 300-kg grain bulks

We examined the effect of low temperatures on insect population density and mortality of adult and immature stages of Cryptolestes ferrugineus and Rhyzopertha dominica in three barrels holding 300 kg of wheat each. Barrels were infested with 2 insects/species/kg and left to develop at 30 °C for 5 months. Temperature of the grain mass before cooling was 38 °C in the centre to a low of 28 °C on the periphery. On 11 February the building door was opened to allow for cooling of the grain. During the 3 weeks of cooling, temperatures in the centre reached as low as −5 °C. Segregated (11 segments) grain trier samples were collected from three locations in each barrel, from February until March 2013 (6 dates), sieved, and the number of live and dead adults counted. Sieved grain was held for 5 weeks 30 °C and emerged adults counted to give an estimation of immature stages within the grain at the time of sampling. Rhyzopertha dominica adults were found mainly in the top 40 cm of the barrel, and there was little movement to the centre as the grain mass cooled. The initial distribution of C. ferrugineus adults was more complex; in two barrels adults were mostly in the centre of the grain mass, while in the other barrel, more adults were found on the periphery. When the grain mass cooled, C. ferrugineus adults in all barrels were found mostly in the centre. On day 0, the average number of live adults of C. ferrugineus and R. dominica were 232 ± 56 and 414 ± 48 insects/100 g, respectively. Rhyzopertha dominica adults were more susceptible to low temperatures than C. ferrugineus. The density of immatures of C. ferrugineus and R. dominica were similar at 0 d, there was a decline over the 21 d, but there was some survival at day 21.     

Rice weevil management through application of silica nano particle and physico-chemical and cooking characterization of the treated rice

Nano technology is an alternative method for pest management and can be applied to minimize the post harvest storage losses due to insect infestation. A study was undertaken with an aim to control the stored grain pests, particularly rice weevil using silica nano particle (SNP). The nano-silica of 30 nm size was taken for the experiment. Entomotoxic effect of five different doses of SNP was tested against rice weevil Sitophilus oryzae. The effects of SNP mixing with grain at the rate of 0.25, 0.5, 0.75 and 1.0 g/kg of rice, stored in glass jars and gunny bags were examined. The cumulative mortality rate was studied at 1, 2, 4, 7 and14 days of exposure. Mortality rate of 80 and 97.4% were achieved after 7 and 14 days of storage by application of SNP at the rate of 0.5 g per kg of rice. Residual persistency of samples treated with optimum dose of SNP was studied by washing with distilled water at the ratio of 1:2 (sample to water ratio) for three times and the samples were studied under Scanning Electron Microscope (SEM). Single washing prior to cooking was found to be sufficient enough for removing the SNP from treated rice and can be recommended safe for cooking and consumption. This process can also prove to be an eco-friendly and efficient pest management technology. The technology is suitable for industrial application in large scale warehouses and storage godowns to control stored grain insect pests.     

The efficacy of sieving, filth flotation and Tullgren heat extraction for detecting various developmental stages of Tribolium castaneum and Ephestia kuehniella in samples of wheat grain, flour and semolina

A prerequisite for effective pest risk management in food is the unbiased interpretation of results obtained by various detection methods. In this study we compared the sensitivity of filth flotation tests, sieving and heat extraction in Tullgren–Berlese funnels for detecting insect contaminants. Samples of wheat grain, flour and semolina were contaminated with eggs, juveniles and adults of Tribolium castaneum, and eggs or larvae of Ephestia kuehniella. Calibration methods were applied for every detection method, and total and sample recoveries and detection limits were calculated for each method, food substrate and contaminant type. The tested contaminants were not detected on a qualitative level by any single technique, instead a combination of techniques was necessary for detection. Sieving was the method with the highest total recoveries, ranging from 90 to 100%. Filth flotation was a uniquely effective for egg detection, with total recoveries ranging from 65 to 95%. The extraction of adults and larvae of both species in Tullgren–Berlese funnels failed in semolina and flour, and was of very limited success in grain. The detection limits for sieving were from 1 to 16 contaminants/kg commodity. The detection limits for filth flotation were from 224 to 508 eggs, and 58 to 507 adults or larvae/kg commodity. The sample recoveries were usually influenced by sample size, species, stadium and their interactions, and indicated how to optimize method protocols. The calibration of methods provided estimates of contaminant densities different from those obtained without calibration. Our work revealed that some currently used methods are not sensitive enough to detect all stages of insect pests, or in some cases, low levels of pest infestation. This lack of sensitivity potentially enables the infested cereal food product to continue down the food processing chain even after laboratory inspection.     

Improving vineyard sampling efficiency via dynamic spatially explicit optimisation

Background and Aims: Environmental variables within vineyards are spatially correlated, impacting the economic efficiency of cultural practices and accuracy of viticultural studies that utilise random sampling. This study aimed to test the performance of non‐random sampling protocols that account for known spatial structures (‘spatially explicit protocols’) in reducing sampling requirements versus random sampling. Methods and Results: Canopy microclimate data were collected across multiple sites/seasons/training systems. Autocorrelation was found in all systems, with a periodicity generally corresponding to vine spacing. Three spatially explicit sampling models were developed to optimise the balance between minimum sample sizes and maximum fit to a known probability density function. A globally optimised explicit sampling (GOES) model, which performed multivariate optimisation to determine best‐case sampling locations for measuring fruit exposure, reduced fruit cluster sample size requirements versus random sampling by up to 60%. Two stratified sampling protocols were derived from GOES solutions. Spatially weighted template sampling (STS) reduced sampling requirements up to 24% when based on probabilistic panel weighting (PW), and up to 21% when preferentially selecting specific locations within canopy architecture (AW). Conclusions: GOES, PW STS and AW STS each reduced required sample size versus random sampling. Comparative analyses suggested that optimal sampling strategies should simultaneously account for spatial variability at multiple scales. Significance of the Study: This study demonstrates that dynamically optimised sampling can decrease sample sizes required by researchers and/or wineries.     

A novel strategy for improving radio frequency heating uniformity of dry food products using computational modeling

This study attempted to quantify effects of dielectric properties (DPs) and densities of a surrounding container and treated food products on heating uniformity in a 6 kW, 27.12 MHz parallel plate radio frequency (RF) system. A computer simulation model was established with finite element-based commercial software, COMSOL Multiphysics®, and experiments with 1.5 kg soybean flour packed in a rectangular polystyrene container were performed to validate the developed model. Surface temperature distributions of soybean flour in three different horizontal layers were obtained with an infrared camera, and temperature–time histories at two representative locations inside the container were monitored with two optical fiber sensors. The uniformity index (UI) was used as a criterion to evaluate the RF heating uniformity within food products. Results showed that the RF heating uniformity in food samples was clearly influenced by DPs and density of the surrounding container. UI was the lowest when the surrounding container dielectric constant was in a comparable range of the sample's, with the loss factor values of surrounding container lying between 0.01–0.1% of the sample's. The optimum RF heating uniformity in food products could be achieved with a smaller density value of the surrounding container. The correlations of DPs and density between surrounding container and food products derived from the validated simulation model could provide valuable information and strategy to improve the RF heating uniformity in low moisture foods for insect or microbial control. Thus, the established strategy can further be used for developing effective industrial-scale RF treatment protocols after optimization of this process by the food industry.Industrial relevanceAlthough the most important characteristic of radio frequency (RF) treatments is fast and volumetric heating generated by dipole rotation and ionic conduction, edge over-heating is still a major problem for foods heated in rectangular containers. The validated model was used to study the effects of dielectric properties and density of sample and surrounding container on sample uniformity index. Simulated results illustrated that the RF heating uniformity could be improved when the dielectric constant and density of surrounding container and sample were in accordance with the established relationships. The established strategy may provide valuable optimized methods to ensure RF heating uniformity in industrial applications.     

Stored Grain Advisor Pro: Decision support system for insect management in commercial grain elevators

A decision support system, Stored Grain Advisor Pro (SGA Pro) was developed to provide insect pest management information for wheat stored at commercial elevators. The program uses a model to predict future risk based on current insect density, grain temperature and moisture. A rule-based system was used to provide advice and recommendations to grain managers. The software was tested in a research program conducted at commercial grain elevators in Kansas and Oklahoma, USA. A vacuum-probe sampler was used to take ten 3-kg grain samples in the top 12 m of each bin that contained wheat. After the insect species and numbers were determined for each sample, the data were entered into SGA Pro. A risk analysis and treatment recommendation report for all bins was presented to the grain managers every 6 weeks. SGA Pro correctly predicted for 71–80% of bins whether the grain was safe or at high risk of dense infestation and grain damage. SGA Pro failed to predict “unsafe” insect densities in only 2 out of 399 Kansas bins (0.5%) and in none of 114 bins in Oklahoma. Grain managers who followed SGA Pro's recommendations tended to fumigate only the bins with high insect densities instead of fumigating all bins at their facility. This resulted in more efficient insect pest management because fumigating bins only when insect densities exceeded economic thresholds and treating only the bins that required fumigation minimized the risk of economic losses from insects, reduced the cost of pest management, and reduced the use of grain fumigant.     

Simulation model of Rhyzopertha dominica population dynamics in concrete grain bins

Rhyzopertha dominica is one of the most damaging insect pests in grain elevators and causes millions of dollars worth of stored grain losses annually in the USA. A simulation model was developed for predicting R. dominica population dynamics in concrete grain bins. The model used a two-dimensional representation of a cylindrical concrete bin (33 m tall×6.4 m wide), and used hourly weather data to predict changes in grain temperature. Output from the grain bin temperature and moisture module was used by the insect module to predict changes in insect density in 32 different bin regions. When compared to validation data from nine grain bins, the model accurately predicted insect vertical distribution and insect density. In December, the highest insect density was in the top center of the grain mass, and decreased steadily with increasing depth and towards the periphery of the grain mass. R. dominica attains this spatial distribution because immigration is primarily through the top of the bin, and higher populations occur in the interior of the grain mass because of warmer temperatures there. Initially, the model underestimated actual insect density in the grain bins. We increased the immigration rate by 50% and this resulted in a much better prediction of R. dominica density by the model. From 20 September to 14 December, populations of R. dominica increased from 0.1 to 3.5 insects per kg of wheat.     

Detection and quantification of the insect pest Rhyzopertha dominica (F.) (Coleoptera: Bostrichidae) in rice by qPCR

The early detection of insects during grain storage and processing remains a major issue for the cereal industry, especially when immature stages are hidden inside the grain kernels. For this reason, we developed a qPCR method to detect and quantify one of the main pests of stored products in rice: the coleopteran internal feeder Rhyzopertha dominica. For that purpose, a specific primer set was designed to amplify artificial infestations of this pest in rice. Then, using a regression model, a standard curve was generated that correlated individuals to adult equivalent DNA quantity (inverse of the C_t value). Results revealed that the designed primer set was specific for R. dominica when tested against the other 4 common internal feeders in grain. The technique showed to be accurated (DNA was detected in more than 73% of the samples) and sensitive to insect presence (i.e. from 0.02 adults, 0.1 3rd instar to pupae or 13 egg to 2nd instar detectable per kg of rice). Moreover, the detection of R. dominica was strongly associated with a given infestation size: DNA quantity increased along with the size of the population. The use of the described qPCR protocol in grain and milling factories may enhace the critical detection and quantification of R. dominica populations in raw materials and processed food.     

Efficacy of filter cake and Triplex powders from Ethiopia applied to wheat against Sitophilus zeamais and Sitophilus oryzae

The efficacy of filter cake and Triplex powders applied to wheat was evaluated in the laboratory against the maize weevil, Sitophilus zeamais Motschulsky and rice weevil, Sitophilus oryzae (Linnaeus)—two most common insect pests associated with stored grain in Ethiopia. Efficacy of these powders was determined by exposing 20 adults of each species to 100 g of wheat treated with 0, 100, 500, 700 and 1000 mg/kg of filter cake and Triplex. Adult mortality was determined 7 and 14 d after exposure. In addition, adult progeny production, percentage of insect damaged kernels, and percentage of grain weight loss at each species-powder-concentration-time combinations were determined after 42 d. The 7 and 14 d mortality was 100% for adults of both species exposed to 1000 mg/kg of filter cake; only the 14 d mortality of Sitophilus species was 100% for adults exposed to 700 mg/kg. Mortality of S. oryzae adults was 100% when exposed for 14 d to1000 mg/kg of Triplex. Mortality of S. zeamais never reached 100% in any Triplex treatments. Adult progeny production of S. zeamais was completely suppressed at filter cake concentrations of 700 and 1000 mg/kg, whereas 1000 mg/kg was necessary for complete suppression of S. oryzae adult progeny production. Complete suppression of adult progeny production was not observed in any Triplex treatments. Complete reduction in percentage of insect damaged kernels and percentage of grain weight loss were obtained when S. zeamais and S. oryzae adults were exposed to 1000 mg/kg of filter cake; similar reductions with S. zeamais occurred only at 1000 mg/kg of Triplex. In the case of S. oryzae, complete reduction of insect damaged kernels and grain weight loss were not achieved at any concentration of Triplex. These powders can be used as alternatives to chemical insecticides for management of Sitophilus species.     

Detection of internal insects in wheat using a conductive roller mill and estimation of insect fragments in the resulting flour

A laboratory roller mill that monitors the conductance of kernels that pass through it was tested for its ability to estimate the number of insect fragments in flour after milling. This system can test a kilogram of whole wheat in approximately 1 min and requires little sample preparation. Hard red winter wheat samples were infested with lesser grain borers and stored at 24 °C. Infestations ranged from 12 to over 2000 infested kernels per 1 kg or per 30,000 kernels. After crushing of samples in the conductance instrument, the samples were milled into flour and sub-samples were sent to two laboratories for insect fragment analysis. The insect fragments were proportional to the number of detection incidences obtained using the conductance instrument and X-ray images. Insect fragment counts per 50 g of flour ranged from 0 to over 5000. For insect fragment counts from 0 to 250, correlations between fragment counts and conductance mill detection were 0.75 and 0.80 from two separate cereal chemistry laboratories. Therefore, the conductance mill is potentially a good method for testing incoming grain for live internally infesting insects; it is able to test 1 kg of grain in about 1 min and can detect low levels (as low as three insects) of live internal infestations in a 1- or 2-kg sample.     

Predicting Prostephanus truncatus (Horn) (Coleoptera: Bostrichidae) populations and associated grain damage in smallholder farmers’ maize stores: A machine learning approach

Prostephanus truncatus is a notorious pest of stored-maize grain and its spread throughout sub-Saharan Africa has led to increased levels of grain storage losses. The current study developed models to predict the level of P. truncatus infestation and associated damage of maize grain in smallholder farmer stores. Data were gathered from grain storage trials conducted in Hwedza and Mbire districts of Zimbabwe and correlated with weather data for each site. Insect counts of P. truncatus and other common stored grain insect pests had a strong correlation with time of year with highest recorded numbers from January to May. Correlation analysis showed insect-generated grain dust from boring and feeding activity to be the best indicator of P. truncatus presence in stores (r = 0.70), while a moderate correlation (r = 0.48) was found between P. truncatus numbers and storage insect parasitic wasps, and grain damage levels significantly correlated with the presence of Tribolium castaneum (r = 0.60). Models were developed for predicting P. truncatus infestation and grain damage using parameter selection algorithms and decision-tree machine learning algorithms with 10-fold cross-validation. The P. truncatus population size prediction model performance was weak (r = 0.43) due to the complicated sampling and detection of the pest and eight-week long period between sampling events. The grain damage prediction model had a stronger correlation coefficient (r = 0.93) and is a good estimator for in situ stored grain insect damage. The models were developed for use under southern African climatic conditions and can be improved with more input data to create more precise models for building decision-support tools for smallholder maize-based production systems.