Evaluation of Silica-Coated Insect Proof Nets for the Control of Aphis fabae, Sitophilus oryzae, and Tribolium confusum

Insect proof nets are widely used in agriculture as mechanical and physical barriers to regulate pest populations in a greenhouse. However, their integration in the greenhouse ventilation openings is highly associated with the decrease of air flow and the adequate ventilation. Thus, there is need for alternative pest management tools that do not impair adequate ventilation. In the present study, we tested four net formulations of relatively large mesh size coated with SiO2 nanoparticles, namely, ED3, ED3-P, ED5, and ED5-P to evaluate their insecticidal properties against adults of Aphis fabae and Sitophilus oryzae and larvae of Tribolium confusum. ED3 and ED5 nets were coated with SiO2 nanoparticles of different diameter, while in the case of ED3-P and ED5-P, paraffin was added to increase the mass of the deposited particles on the net’s surface. In the first series of bioassays, the knockdown and mortality rates of these species were evaluated after exposure to the aforementioned net formulations for 5, 10, 15, 20, 25, 30, 60, 90, and 180 min. In the second series of bioassays, knockdown and mortality of these species were recorded after 1, 7, and 10 days of post-exposure to the nets for different time intervals (15, 30, and 60 min). Based on our results, all nets significantly affected A. fabae, since all insects were dead at the 1-day post-exposure period to the silica-treated nets. Conversely, at the same interval, no effect on either S. oryzae adults or T. confusum larvae was observed. However, in the case of S. oryzae, the efficacy of all nets reached 100% 7 days after the exposure, even for adults that had been initially exposed for 15 min to the treated nets. Among the species tested, T. confusum larvae exhibited the lowest mortality rate, which did not exceed 34% at the 10 days of post-exposure interval. Our work underlines the efficacy of treated nets in pest management programs, under different application scenarios, at the pre- and post-harvest stages of agricultural commodities.


Introduction
Despite the association of many insecticidal compounds with mammalian toxicity and increased environmental footprint, conventional pesticides are widely used as a means of pest management. Over the years, many insects species have developed resistance mechanisms to the vast majority of control different major stored-product insects such as the rice weevil, Sitophilus oryzae (L.) (Coleoptera: Curculionidae), since 98% mortality was recorded two weeks after the exposure to the treated net.
The use of nanoparticles as a novel biological insect control agent has been well established in the recent years. Silicon is known as the most frequent metalloid and the most copious element on earth after oxygen [21][22][23]. The mode of action of silicon dioxide (SiO 2 ) begins with the adherence of the nanoparticles on the cuticle of the insect, inactivating epicuticular lipids, which leads to death through desiccation [24,25]. Furthermore, Rastogi et al. [22], suggested that the mortality of pests could be related to congestion of spiracles and tracheas caused by the nanosilica particles.
Diatomaceous earth (DE) that consists mainly of amorphous hydrated silica has been extensively used in stored product insect management. Vayias and Athanassiou [26] stated that the effectiveness of DE is associated with the developmental stage of the insects. Indeed, the management of the confused flour beetle adults, Tribolium confusum Jacquelin DuVal (Coleoptera: Tenebrionidae) is considerably difficult compared to soft-bodied greenhouse pests such as aphids. On the other hand, aphids are among the most common pests in horticultural crops and are related to the transmission of many economically hazardous viruses even after injecting their sucking mouthparts on the leaves for a very short time interval [18,27]. Unlike hard-bodied beetles, aphids are considered to be much more vulnerable to desiccation from sorptive materials and micro-wounds. For instance, Singh and Singh [28] found that DEs could cause rapid mortality of the aphid Rhopalosiphum padi (L.) (Homoptera: Aphididae), when applied on wheat plants. However, in that study, the authors stood eventually against these DE applications, because of negative photosynthetic rates [28].
Recently, Shoaib et al. [29], indicated that the exposure of larvae of Plutella xylostella (L.) (Lepidoptera: Plutellidae) to 1 mg × cm −2 of a siliceous dust formulation resulted in 85% mortality after 72 h. The authors suggested that the mortality rate is improved as dose and exposure interval are increased. Moreover, silica nanoparticles, concerning their physical and chemical properties, differ significantly from their bulk formation [19]. According to Debnath et al. [30], nanoparticles proved to be much more effective when compared with bulk silica against S. oryzae. At the post-harvest stages of agricultural commodities, stored product protection could be enhanced by the use of silicon-based nanoparticles as non-conventional pesticide agents [2,31].
To date, little attention has been given to the incorporation of physical inert dusts to the yarns of insect proof nets. The application of non-toxic and environmental-friendly nets could replace the insecticidal treated nets leading to insecticide-free products. Thus, the objective of our study was to assess the effectiveness of silicon dioxide-treated nets under laboratory conditions for the control of the black bean aphid, Aphis fabae Scopoli (Hemipetra: Aphididae), and the stored product pests T. confusum and S. oryzae. The aforementioned species were exposed to insect proof nets coated with two different silicon dioxide dust formulations, Syloid ® ED3 and Syloid ® ED5 with and without the addition of paraffin in order to achieve better particle adherence to the net, respectively.

Insects Tested
The rearing of the A. fabae individuals took place at the Laboratory of Agricultural Entomology, Benaki Phytopathological Institute, Attica, Greece. Aphids were reared on Vicia faba L. plants at 20 ± 1 • C, relative humidity (r.h.) of 65 ± 2% and a photoperiod of 16 h in light: 8 h in dark [32]. Aphid colonies were transferred to the Laboratory of Entomology and Agricultural Zoology (LEAZ), Department of Agriculture, Crop Production and Rural Environment, University of Thessaly to carry out the bioassays.
Established laboratory colonies of S. oryzae and T. confusum were maintained in LEAZ under controlled conditions in a growth chamber as suggested by Faliagka et al. [33].

Nettings and Dust Formulations
Experiments were carried out with four samples of the same insect proof net, the textiles of which were performed by the Institut für Textiltechnik (ITA) of RWTH Aachen University. The weaving and the evaluation of the nets was provided from Thrace Nonwovens & Geosynthetics S.A. (Thrace, Greece). All textiles were coated either with Syloid ® ED3 or Syloid ® ED5 which are amorphous silica dust formulations (P&S Powder and Surface GmbH, Salzkotten, Germany). The former silica dust consists of 99-100% SiO 2 with particle size of 5.8 µm, while the latter consists of 95-100% SiO 2 with particle size of 9.0 µm. Paraffin was added as an organic primer in two of the four netting samples (ED3-P and ED5-P) consisting of different silica formulation in order to enhance the adhesion properties of each dust to the net. ED3 and ED5 samples were used without the addition of paraffin. The purpose of the primer is the integration of up to 40% more SiO 2 particles on the net. The mass of the deposited silica particles on the net's surface was fluctuated from 0.4 to 0.9 g × m −2 depending on the dust formulation as well as the addition of paraffin. However, paraffin may reduce the mesh size and therefore, the air permeability properties of the tested nets. In the present study, the insecticidal properties of SiO 2 nanoparticles were evaluated. In the case where adequate results are provided, it is suggested to incorporate silica on nets of larger pores in order to alleviate insufficient ventilation. Table 1 shows the properties of the nets used in the experimental trials. All samples were produced by the same 50 mesh net that was constructed by using high density polyethylene (HDPE) monofilament yarns. SiO 2 free-nettings with the same mesh size as well as clear Petri dishes without the addition of nets were used as a control treatment. The silica coating procedure was repeated two times for each tested net.

Bioassay Series
Nets were fine cut and adjusted at the bottom of plastic Petri dishes (59.4 cm 2 in surface), using a thin layer of silicon. The lids of plastic dishes are usually loose-fitting making it easy for insects to escape, so, the "neck" of each Petri dish was covered with polytetrafluoroethylene dispersion, (Northern Product, Cumberland, RI, USA). A single fine brush was used to insert the species into the Petri dishes to avoid wounding.

Short-Term Effect
In the short-term effect bioassay series, adults of A. fabae and S. oryzae, and larvae of T. confusum were used. Ten insects of each species were transferred to the Petri dishes. All species were exposed to silica-treated nets, clear Petri dishes (without nets), and untreated nets for the intervals of 5, 10, 15, 20, 25, 30, 60, 90, and 180 min. For each exposure interval, mortality and knockdown effect of all species in each Petri dish were recorded. For each of the four nets, the bioassay was replicated eight times per exposure interval. Thus, in this series of bioassays 144 Petri dishes were used in total.

Long-Term Effect
The purpose of this series of bioassays was to evaluate the effect of the nets on the knockdown and mortality rates of the exposed individuals at certain post-exposure intervals. Thus, as above, ten individuals of each species were transferred to Petri dishes, with different plates per species, and exposed to the nets for 15, 30, and 60 min. Knockdown and mortality were recorded immediately after the completion of the exposure periods and then all insects were carefully transferred to new dishes consisting of untreated nets or clear Petri dishes (without nets). When transferring the insects, a clean single fine brush was used to avoid possible contamination of the dust in the treated nets. Within each exposure interval of 15, 30, and 60 min no food was added in the Petri dishes. However, during post-exposure period for S. oryzae and T. confusum, cracked wheat (0.5 ± 0.1 gr) and wheat flour (1.0 ± 0.1 gr) were added in all petri dishes, respectively. In the case of A. fabae no food was supplied, since the bioassay lasted one day, unlike that of the stored product insects which was completed after 10 days. All bioassays were carried out at 25 • C and 65% r.h. Mortality and knockdown effect of S. oryzae and T. confusum were assessed after 1, 7, and 10 days, whereas in the case of A. fabae after 1 day. For each of the four nets, the bioassays were replicated eight times. A total of 624 Petri dishes were used to perform this series of bioassays.

Statistical Analysis
The results of the first series of bioassay (short-term effect) were analyzed using Probit Analysis to estimate the knockdown time, i.e., KD t50 , KD t95 , and KD t99 for each species and net formulation. The intervals tested yielded a variety of knockdown results only for A. fabae, which allowed the calculation of KD t50 , KD t95 , and KD t99 values. For the second series of bioassays, separately of each species, the data (knockdown and mortality) were converted to percentages and then were submitted to a two-way ANOVA, to determine the differences among the exposure intervals (15,30, and 60 min) and among the net formulations (including controls as well), while means of knockdown and mortality were separated using the Tukey-Kramer HSD test at the 5% level. Moreover, the same approach was conducted for the long-term effect for each species tested. The data were analyzed using SPSS 25.0 (SPSS Inc., Chicago, IL, USA).

Short-Term Effect
The mortality rate of all tested species was relatively low. In the first series of tests, KD t values were not found to fit the data well, since P values were <0.01 (Table 2). In most of the cases, confidence interval could not be estimated for A. fabae and S. oryzae, except for ED5-P of A. fabae. For T. confusum, Probit could not be estimated, since the percentage of knocked down individuals was negligible (data not presented).
Overall, KD t99 values for A. fabae at ED3-P resulted in the highest values among treatments, corresponding to 1590.9 min (approximately 26 h) ( Table 2). Contrariwise, the lowest KD t99 value was recorded at ED5-P, corresponding to 363.3 min (approximately 6 h) ( Table 2). Moreover, for S. oryzae the resulting values were similar with those of A. fabae at ED3, which were 1235.7 min (approximately 20 h) and 1281.4 min (approximately 21 h), respectively. For KD t50 the lowest value (51.3 min) for A. fabae was recorded at ED5, whereas the highest value (279.7 min) was noted at ED3-P (Table 2).

Long-Term Effect
The mortality rate ranged significantly between the tested cases (Tables 3-5). Knockdown of A. fabae was low and did not exceed 28.7% (Table 3), whereas mortality was negligible (data not presented). For knocked down individuals, significant differences were noted in all tested net formulations among the time intervals (15,30, and 60 min). Finally, at the 1-day post-exposure period, mortality was 100% in all tested net formulations (Table 3).
Regarding S. oryzae individuals, knockdown and mortality were negligible, in all tested net formulations for the 15, 30, and 60 min intervals (data not presented), while at the 1-day post-exposure period, no significant differences were noted among the exposure intervals (Table 4). At this interval, mortality was generally low, reaching 6.2%. However, in most of the cases, at the 7-day post-exposure period mortality was approximately 87%, even when the initial exposure was 15 min, while after 10 days mortality reached 100% (Table 4).
Regarding T. confusum larvae, knockdown and mortality were generally low (<5%, data not presented). Larval knockdown of T. confusum was also negligible (data not presented). After 7 d of exposure in all formulations, mortality did not exceed 30%, whereas after 10 d the highest mortality level was 34% at ED5-P (Table 5). Significant differences were recorded among the exposure intervals in most of the cases (Table 5). Table 3. Short-and long-term knockdown (KD) and mortality (% ± SE) of Aphis fabae adults exposed for 15, 30, and 60 min to dishes with four different treated nets (ED3, ED3-P, ED5, ED5-P), dishes with untreated net and dishes without net. The first day after exposure for each exposure interval is considered as the long-term effect.     Within each exposure interval, post-exposure interval, and net formulation, means followed by the same lowercase letters (a, b) do not differ significantly according to the Tukey-Kramer HSD test at P < 0.05. Within each exposure interval, post-exposure interval, and net formulation, mortality means followed by the same uppercase letters (A, B) do not differ significantly across treatments according to the Tukey-Kramer HSD test at P < 0.05. Where no letters exist, no significant differences were noted. ANOVA parameters for 1st day were: at 15 min, F = 1.9, P = 0.10, 30 min, F = 1.6, P = 0.18, 60 min, F = 1.3, P = 0.26, for 7th day were: at 15 min, F = 142.2, P < 0.01, 30 min, F = 91.9, P < 0.01, 60 min, F = 79.1, P < 0.01, whereas for 10th day were: at 15 min, F = 92.2, P < 0.01, 30 min, F = 75.4, P < 0.01, 60 min, F = 55.6, P < 0.01, in all cases df = 5, 47. In the case where "-"is indicated, F and P values were not provided in the Table since no variances among the samples were recorded. Table 5. Delayed mortality (% ± SE) of Tribolium confusum larvae exposed for 15, 30, and 60 min to dishes with four different nets (ED3, ED3-P, ED5, ED5-P), dishes with untreated net and dishes without net. The 1st, 7th, and 10th day after exposure for each exposure interval is considered as the long-term effect. Within each exposure interval, post-exposure interval, and net formulation mortality, means followed by the same lowercase letter do not differ significantly according to the Tukey-Kramer HSD test at P < 0.05. Within each exposure interval, post-exposure interval, and net formulation, mortality means followed by the same uppercase letter do not differ significantly across treatments according to the Tukey-Kramer HSD test at P < 0.05. Where no letters exist, no significant differences were noted. ANOVA parameters for 1st day were: at 15 min, F = 1.

Discussion
The results of our study indicated that even in the case of the longest exposure interval (180 min) no immediate mortality was achieved for either of the stored product species tested. However, for A. fabae, knockdown effect was vigorous and rapid, especially at longer intervals, as KD t50 was high after approximately one hour. In general, from the most to the least susceptible, the species/life stages tested here can be classified as A. fabae adults > S. oryzae adults > T. confusum larvae.
Particle size is highly associated with the insecticidal efficacy of inert dusts. Many authors have emphasized on the importance of the particle size, a significant physical property of dust formulations, underlining that formulations of smaller particles may enhance the insecticidal efficacy on a variety of stored-product beetles [34][35][36][37][38][39]. Korunić [40] found that the efficacy of DEs is promoted when particle size ranges between 1 and 30 µm. Similarly, Vayias et al. [38] indicated that DEs particles, sized up to 45 µm, resulted in higher mortality of stored product beetles, as compared with larger particles. However, Korunić [40] underlined the significance of the DEs origin, and concluded that even a high range between the size of particles (i.e., 0-192 µm) could lead to similar insecticidal results, indicating the significance of the particle shape. In this context, Rumbos et al. [39] found that the insecticidal effect of zeolites on the control of stored-product pests was not highly associated with particle size.
Peng et al. [41] stated that the dispersity and stability of SiO 2 nanoparticles are enhanced when added to liquid paraffin. In the present study, it was concluded that the susceptibility of A. fabae and S. oryzae was high, regardless of the addition of the primer or the size of silica nanoparticles, since no significant differences were recorded concerning the long-term effect. Notwithstanding, our results suggested that the use of paraffin on the net's surface contributes to the reduction of fabrics' shrinkage whereas it increases the rate of the particle deposition. Thus, the knockdown effect of A. fabae was increased when insects were exposed to ED5-P nets since in this case, the deposition of silica particles was the highest among all tested samples due to the addition of paraffin. In particular, it was shown that 99% of the adults were knocked down after six hours when exposed to the net coated with the ED5-P dust formulation, which consisted of the largest silica particles at the highest deposition rate.
Silica particles should not be considered as the only lethal factor in our study, as mortality can be also attributed to the presence of the net itself. As shown in the results corresponding to the control treatments, insects exposed to the SiO 2 -free nets were also affected by the presence of the net as opposed to the insects exposed to the dishes without net. Concerning the above, silica-treated nets could be applied on the vent openings of a greenhouse limiting the number of infestations due to greenhouse pests' invasions, which occurs mechanically, i.e., without the use of the dust. A similar approach has been investigated in the case of stored product insects, with similar results regarding the detrimental effect of the net to the insect bodies [17,20].
Although no immediate mortality was recorded for all the insect species tested, delayed mortality was high, reaching 100% even in the shortest exposure time (15 min). Similarly, Kavallieratos et al. [42] observed a significant delayed mortality on 14 d for S. oryzae and the lesser grain borer, Rhyzopertha dominica (F.) (Coleoptrea: Bostrychidae) when exposed to a mixture of DE and abamectin. Our data show that, despite low initial effect, delayed mortality for the two beetle species was high, which clearly suggests that the exposed individuals were heavily affected by the presence of the dusts in their cuticles. As inert materials are slow-acting, it has been observed that mortality may take several days or even weeks to occur, even if the insects are in continuous contact with the dust particles [43,44]. Nevertheless, the current results suggest that even the shortest exposure to the dust particles tested here is irreversibly lethal, and recovery is less likely to appear. This is particularly important, as insects will eventually die after their removal from the treated substrate, a characteristic that is even more important in the case of aphids, if it is combined with the knockdown that appears rapidly after exposure.
To conclude, this study indicates that all of the nets tested here were effective for the control of aphids and less for beetles (including larval stage), despite variations in efficacy among treatments. Practically, the nets that are coated with inert materials can serve as a good alternatives to traditional pesticides in greenhouses but also in food and processing facilities, significantly moderating insect immigration. This technology can be adopted in a wide range of facilities, and also in food packaging, where the larval stage can penetrate with ease. At the same time, in a "real world" greenhouse scenario, immigration of aphids from the outside toward the plantation will force the insects to pass through the treated net, which is expected to be lethal, even if some aphids eventually pass successfully.