Susceptibility of flexible plastic foodstuffs packaging against Monomorium indicum (Hymenoptera: Formicidae) household ants

Study location and ant specimens

This study was conducted at the faculty building of the agriculture and environment department on the Baghdad campus of The Islamia University of Bahawalpur, Pakistan. The natural movement of ants was used as an infestation source in experiments. Experiments were performed during summer vacations, when academic activities were limited and the academic building was usually quiet. Ants were abundant at the experimental locations within the agriculture and environment faculty building. Ant specimens were collected with moist camel hair brush, preserved in 70% ethanol and subsequently identified as Monomorium indicum Forel at the Insect Biodiversity Laboratory, Department of Entomology at The Islamia University of Bahawalpur.

Experiment cages

Cages were built for this experiment with wood and two mm iron gauze, with each cage 8 × 8 square inches in size. In total there were nine such cages to retain three replicates for testing each packaging thickness. Cages were built so that ants could enter the box, but no other damaging pests (rodents, lizard, cats and squirrels) could enter.

Packaging materials

The most common flexible plastic foodstuff packaging types being used in Pakistan include opaque polyethylene (high density), transparent polyethylene (low density) and polypropylene. For this study, plastic materials were purchased from a wholesale plastic market in Lahore at rate of 400 rupees per kg. Mean thickness of the different packaging materials was identified using a digital micrometer (Mitutoyo Corporation, Kawasaki, Japan). All three types of materials were available in 0.02 mm thicknesses. Transparent polyethylene and polypropylene were also available in 0.04 mm and 0.06 mm thickness levels, but not high density or opaque polyethylene. These plastic packaging films were purchased and used in the experiments to evaluate their susceptibility against M. indicum household ants. Small bags of these plastic films (8 × 10 cm) were prepared in the laboratory using a pair of scissors and an impulse (heat) sealer.

Packaged food

Fresh fruit cake, purchased from the local market, was used inside the packaging to check the susceptibility of the packaging types against household ants. Prepared bags of each type and thickness of packaging were filled with one 18 g fresh fruit cake slice, which was weighed on an electric scale. After adding the fruit cake slices, the plastic bags were sealed with a heat sealer.

Experimental setup

Three types of 0.02 mm thick plastic bags—opaque polyethylene (high density), transparent polyethylene (low density) and polypropylene—were filled with fruit cake and sealed with a heat sealer. There were no prior vents or holes in the bags that would allow insects to enter. These packages were then placed inside a cage. Three similar packaging types but without fruit cake (control treatments) were also placed in the experimental cage. The cage was then closed and locked to restrict entry of any foreign objects. Two other cages were prepared in the same manner to maintain three replicates for 0.02 mm thick packaging. Similar methods were used to evaluate 0.04 mm and 0.06 mm thickness packaging, with three replicate cages prepared with 0.04 mm thick transparent polyethylene and polypropylene plastic bags, both with and without food, and three replicate cages prepared with 0.06 thick transparent polyethylene and polypropylene plastic bags, both with and without food.

One cage for each thickness of packaging was placed at each of three different locations near a water source where ants had previously been observed in the faculty building of the agriculture and environment department. Each of the three locations had three cages, one for each thickness level of packaging, for a total of nine cages. This experiment was under observation for the whole study period to reduce any possible outside disturbances. The cages were visited daily and the number of holes in the packaging data was collected after every five days for fifteen days for this experimental setup. This first experiment lasted from 20th June to 5th July 2022.

Every five days, the cages were opened, the bags were removed, and damage to the packaging, measured by number of holes, was observed. Packages displaying any sealing defects were immediately replaced with the same type of packaging to avoid ant invasions through holes not created by the ants (Mullen, Vardeman & Bagwell, 2012). The number of holes in the packaging was measured and then the bags were opened to measure weight loss of the fruit cake caused by ants, using the following formula:

${\displaystyle \text{\% age weight loss}=\frac{\text{Initial weight}-\text{Final weight}}{\text{Initial weight}}\times 100.}$ This experimental setup was maintained for fifteen days and data was gathered for each of the three thicknesses. Following the first experiment, the same experimental setup was repeated from 10th July to 25th July 2022 (second experiment) and finally from 1st August to 15th August 2022 (third experiment) (Table 1). For the second and third experiments, new packages were used along with newly-packaged fruit cake (18 g) for each packaging type and thickness.

Study of insect mandibles

Mandibles of M. indicum and three common pests of stored products, Rhyzopertha dominica, Tribolium castaneum, and Trogoderma granarium were all studied under microscope. Three specimens of each insect type were selected. The head region of each specimen was separated using fine forceps and surgical blade no. 14 and then mounted on a clean glass slide in glycerin (50%). Mandibles were oriented under a camera (Model HD 1500 T, Meiji, TECHNO, Saitama, Japan) fitted with a trinocular stereoscope microscope (Labomed, CXR3, Labo America, Inc., Fremont, CA, USA) with installed software (T Capture Version 3.9 digital software; T Capture, 2017) on a laptop computer (Dell Core i3, 10th Gen). The mandibles of the insect specimens were orientated for proper measurements and visual comparisons, photo captured and saved with proper labelling for future reference. The images were opened with T Capture software, which was calibrated using the micrometer scale (1 mm). The mandibles, as well as mandibular frontal tooth of three specimens for each insect type, were measured. The images, along with the measurements, were saved and the respective values were tabulated in Microsoft Excel 2021 (Version 2019; Microsoft, Bellevue, WA, USA) for further data analyses.

Data analysis

Data was statistically analyzed using SPSS software (version 2016; SPSS Inc., 2007). Data was analyzed separately for each thickness level using one-way ANOVA in which different packaging types, both with and without food, served as independent variables to see the effect of packaging types of each thickness level on number of holes and percent weight loss of packaged fruit cake, with number of holes and percent weight loss serving as dependent variables. One-way ANOVA was also performed to see the effect of the three experiment dates on number of holes and weight loss of packaged fruit cake, with experiment dates serving as the independent variable and number of holes and percent weight loss of the fruit cake serving as dependent variables. Mean values were separated post hoc at a 5% level of probability using a Tukey HSD test. For each thickness level, correlation (Pearson) was also tested between number of holes created by M. indicum (damage) and weather data, including temperature and relative humidity, of the three experiment dates. To see the overall effect of packaging thickness on damage, a Pearson correlation analysis was performed on all the data between number holes in all thickness levels (omitting high density polyethylene in 0.02 thickness to standardize data along three thickness levels) and packaging thickness. Finally, the measured mandible of M. indicum was statistically compared with the measured mandibles ofthe three frequent pests of packaged foodstuff: Rhyzopertha dominica, Tribolium castaneum and Trogoderma granarium. An analysis of variance one-way ANOVA was also performed in which lengths of the mandible and frontal tooth of M. indicum and three storage pests served as dependent variables, and insect types served as independent variables. Means values were separated post hoc by a Tukey HSD test at a 5% level of probability.

Effect of packaging type on number of holes in packaging and weight loss of packaged fruit cake caused by M. indicum household ants

Figure 1 shows the effect of packaging types with 0.02 mm thickness on damage (holes in packaging and percent weight loss of fruit cake) caused by M. indicum household ants. The results showed ants created the highest number of holes (2.00) in polyethylene high density packaging, followed by low density polyethylene bags (1.56), with the least number of holes found in polypropylene (1.22) packaging. Damage was not observed in any of the packaging types without food material (F_{5, 53}:1.832; P: .124).

Weight loss of packaged fruit cake was recorded in packaging where holes were created by ants. Percent weight loss due to ants feeding was highest in the polyethylene high density (39.64%) packaging, followed by polypropylene (21.03%) packaging, with the lowest weight loss percentage observed in polyethylene low density (18.56%) packaging. Zero weight loss was recorded in packaging without holes (F_{5, 53}: 2.762; P: .028).

In 0.04 mm thick packaging, the highest average number of holes was recorded (0.11) in polypropylene packaging with fruit cake, while no holes occurred in polyethylene packaging and packaging without fruit cake (F_{3, 35}: 1.000; P: .405). Similarly, percent weight loss was only recorded in polypropylene packaging (8.09%) with fruit cake, but no weight loss was recorded in packaging without holes (F_{3, 35}: 1.000; P: .405; Fig. 2).

In 0.06 mm thick packaging, the highest average number of holes was recorded (0.44) in polyethylene packaging with fruit cake, while no holes occurred in polypropylene packaging and packaging without fruit cake. Similarly, percent weight loss was only recorded in polyethylene packaging (5.36%) with fruit cake, but no weight loss was recorded in packaging without holes (F_{3, 35}: 1.000; P: .405; Fig. 3).

Effect of experiment dates on number of holes in packaging and weight loss of packaged fruit cake caused by M. indicum household ants

At a packaging thickness level of 0.02 mm, ants were able to cause damage on all three dates of experiments ranging from 25th June to 5th July, 15th July to 25th July and from 5th August to 15th August during 2022 (Fig. 4). In these date ranges, the highest number of holes was recorded (1.28) during the first experiment, followed by the second experiment (1.06), with the lowest number of holes recorded (.06) in the third experiment (F_{2, 53}: 1.806; P: .175). The highest weight loss percentage of packaged fruit cake was recorded in the first experiment (23.72%), followed by the second experiment (11.78%), with the lowest weight loss percentage (4.11%) recorded in the third experiment (F_{2, 53}: 1.818; P: .173).

At a packaging thickness level of 0.04 mm, 0.08 holes created by M. indicum were recorded in the first experiment, however no holes were created in 0.04 mm thick packaging in the second or third experiments. Accordingly, 6.06% weight loss of packaged fruit cake was recorded during the first experiment, but no weight loss was recorded in the second or third experiments (F_{2, 35}: 1.000; P: .379; Fig. 5).

At a packaging thickness level of 0.06, 0.33 holes created by M. indicum were recorded in the first experiment, however no holes were created in 0.06 mm thick packaging in the second or third experiments. Similarly, 5.36% weight loss of packaged fruit cake was recorded in the first experiment, but no weight loss was recorded in the second or third experiments (F_{2, 35}: 1.000; P: .379; Fig. 6).

Correlation of damage caused by M. indicum with weather factors and packaging thickness

The correlation analysis results showed that in all three packaging thickness levels, temperature had a strong positive relationship with damage to packaging caused by M. indicum, while relative humidity had a strong negative effect on packaging damage. A correlation analysis of overall data for all thickness levels showed packaging thickness had a negative correlation with damage caused by M. indicum (Table 2).

Study of M. indicum mandibles in relation to mandibles of major stored grain insect pests

Figure 7 shows a comparison of mandibular length and frontal tooth length between M. indicum household ants and three major stored product pests. The results showed M. indicum had significantly longer mandibles compared with the mandibles of three common stored product pests (F_{3, 11}: 94.551; P: < 0.001). The highest recorded mean length of mandible was 400.67 µm for M. indicum, followed by 241.67 µm for adult R. dominica, and 201.33 µm for T. castaneum, and the lowest recorded mean length of mandible was 174.33 µm for T. granarium larva. Mandibular frontal tooth length was highest in M. indicum (124.00 µm), followed by both R. dominica and T. castaneum at 81 µm, and the lowest length was 32.00 µm in T. granarium larva (F_{3, 11}: 68.601; P: < 0.001).