Control of Stored Agro-Commodity Pests Sitophilus granarius and Callosobruchus chinensis by Nitrogen Hypoxic Atmospheres: Laboratory and Field Validations

: Given the complexity of the practical usage of controlled atmospheres for the protection of agro-commodities, several researchers have pointed out that there is not enough robust scientiﬁc documentation regarding the usage of inert gases for their widespread practical application. Therefore, this work evaluated various regimes of hypoxic and anoxic nitrogen atmospheres for the control of two key stored-product pests, in laboratory and under ﬁeld conditions in silos. Sitophilus granarius and Callosobruchus chinensis were selected as the tested species since they are important pests of grain/rice or legumes in Europe and Asia. Under laboratory conditions, we tested nitrogen (N 2 ) concentrations (from 95 to 100%) and exposure times (1–20 days) on the developmental stages of both pest species. In most developmental stages of S. granarius and C. chinensis, the shortest effective exposure was found for nitrogen concentration of 99%. Based on our laboratory tests, validation studies were subsequently carried out in semi-hermetic steel silos (25t) using continuous nitrogen saturation by on-site built swing pressure generators. It was found that a full control of all stages of S. granarius and C. chinensis was achieved in 11 days of nitrogen exposure, using concentrations ranging above 99% and below 100%. Our work shows that hypoxic nitrogen treatment can be effectively achieved in small steel silos under proper technological and environmental conditions.


Introduction
Storage arthropods can cause multiple types of damage to agro-commodities during their storage [1], import and export [2].Storage pests are also able to invade and spoil processed and packaged food [3][4][5].In addition to economic quality damage, the infestation of commodities and processed foods by storage insects and mites is a problem in terms of their contamination of food with allergens [6,7].Currently, storage pest risks are also increasing due to climatic temperature changes [8].Another risk factor for effective pest control is the increasing resistance to phosphine, that is currently the most frequently used fumigant at a global scale [9][10][11].The use of low temperatures [12], liquid nitrogen as freezing agents [13], alternative fumigants, ozone, or inert gases, are proposed as solutions [14,15].In terms of societal requirements for residue-free exposure to products, yet effective pest control effects, inert gases and anoxic/hypoxic atmospheres are among the most promising [16,17].Controlled atmospheres aim at creating a low-oxygen (hypoxia) or zero-oxygen (anoxia) environment that is lethal to pest insects and mites.Anoxic and hypoxic atmospheres, as eco-friendly pest control solution, can be applied using different procedures or technologies in various types of stores and chambers [15,17,18], or commodity and food packaging [19][20][21].Some types of anoxic and hypoxic atmospheres not only control pests, but they also may maintain the quality of the products [22,23].The physiological effects of inert gases are in many aspects different from most other gaseous insecticides [24][25][26].Thus, in addition to the ecological aspects, an advantage of inert gases is their effective action on pest populations that are already resistant to traditional insecticides.For example, Sakka et al. [27] and Agrafioti et al. [9], have recently demonstrated the high potential of hypoxic atmospheres to control populations of a number of phosphine-resistant storage pest species.In addition, controlled atmospheres are starting to be component of phyto-sanitary and phyto-quarantine treatment strategies [28,29].
Despite the above advantages of inert gases and controlled anoxic atmospheres, their use in practice may not be entirely straightforward, both from a technical and biological point of view.Regarding inert gases, and nitrogen in particular, a limited number of laboratory studies and practical applications are available.They describe their use in gas-tight chambers, laboratory experimental facilities and field stores and silos [30][31][32][33][34][35][36][37].These works show the significant influence of different types of storage and the exposure regimes of inert gases, with respect to their effectiveness on different storage arthropod pest-species [38].Commenting on the biological efficacy of anoxic atmospheres, Bailey and Banks [39] noted that the contemporary literature is apparently conflicting on the relative susceptibilities of various species, and the relative speed of action of N 2 -oxygen or CO 2 -oxygen mixtures.Even the more recent study by Liu [26] confirmed that there were considerable differences between stored product insect species and stages in susceptibility to low oxygen treatment.Not only pest species, but also various life stages of storage pests, may respond differently in terms of their susceptibility to various concentration of CO 2 and N 2 [26].Contrary to what conventional intuition would suggest, Navarro [40] found that for certain storage pest species, a shorter controlled atmosphere (CA) exposure time is required if CA is composed of somewhat less than 100% nitrogen concentration.Therefore, depending on the particular pest species, effective nitrogen dosage and exposure must be maintained within a narrow range of effective concentrations [17].The biological efficacy of CA is linked to the physical characteristics of the storage conditions and the technical characteristics of the stores.Differential gas tightness of various types of storage structures is among the key factors influencing procedures for the effective use of controlled atmospheres.Different construction types of commodity silos may also have different gas sorption characteristics, that influence the requirements for CA operational procedures.Technical structures and the method of filling inert gases affect the uniformity of their distribution within the treated objects [17,37].Operational procedures of CA and exposure regimes can also be affected by the geographical location of the stores, as different areas have different thermal conditions.Areas with low ambient temperatures may require extended exposure times and increased gas consumption.Given the complexity of the practical usage of controlled atmospheres, Athanassiou [41] pointed out that there is still relatively little scientific documentation regarding usage inert gases for silo applications under field conditions.This implies that there is still a need to investigate the use of gases in different pest species and developmental stages, under different technical storage conditions and exposure regimes.
The present work was aimed at investigating the lethality of controlled hypoxic atmospheres on two species of primary pests of stored products, under laboratory conditions and in small silos.Sitophilus granarius L. (Curculionidae) and Callosobruchus chinensis L. (Chrysomelidae: Bruchinae) were selected as the tested species since they are important pests of grain/rice or legumes in Europe and Asia.Moreover, beetles of the genus Sitophilus and Callosobruchus are generally suitable model species for CA testing and validation, as they are among the less susceptible storage arthropods to anoxic atmospheres [42][43][44][45].Based on the findings of Navarro [40], the specific objective of the laboratory part of the work was to determine whether the tested pest species belong to the group of storage pests for whose control it is optimal to reach concentrations around 99%, or to the second group of species, for which the optimal control is 100% anoxia.Under laboratory conditions, we thus tested N 2 concentrations of gaseous nitrogen, on various developmental stages of S. granarius and C. chinensis.Jay [38] commented on the fact that there might not always be complete relevance between laboratory studies and the field application of CA to control stored-product insects.Therefore, based on our laboratory tests, validation studies were subsequently carried out in smaller semi-hermetic silos (25t) with continuous nitrogen saturation using on-site swing pressure nitrogen generators.Experiments were conducted in terms of the pest protection of imported agricultural commodities using standard shipping containers; therefore, tests were conducted in a set of small steel silos, each corresponding to the contents of one standard shipping container.The aim was to determine whether it was possible to achieve complete control of the tested pest species in 11 days under realistic operating conditions, or whether it was necessary to expose the pests for 21 days.

Tested Species and Cultures
The insects used for the study included two internally feeding stored-product Coleoptera species: Sitophilus granarius and Callosobruchus chinensis.Both populations (strains) tested were sensitive strains to commonly used insecticides, that were kept under laboratory conditions for at least 5 years.They were maintained in a rearing room at 25 • C and 75% RH.As feeding substrate was used wheat grain for S. granarius and mung beans for C. chinensis.

Testing Apparatus
All pest species and their developmental stages were tested in a thermostatically controlled N 2 apparatus and the scheme is depicted in Figure 1.It is the modified version of the testing device proposed by Navarro et al. [45].It consisted of 6 small plastic hermetic boxes (Lock and Lock HPL834, 2600 mL) in temperature 25 • C.These boxes were serially connected by transparent plastic hoses (PVC; diameter of 6/9 mm, Deutsch and Neumann GmbH; Hennigsdorf, Germany).The target nitrogen concentration (95%, 99% or 100%) was delivered from a pressurized metal cylinder (Linde Gas a.s.; Prague; Czech Republic) using a C200/2B-3SS outlet valve (Linde Gas a.s.; Czech Republic).A gas washing bottle containing an aqueous NaCl saturated solution was positioned between the metal cylinder and the initial small plastic box to humidity the nitrogen to 75% RH.The concentration of oxygen was determined using a GMH 3691 oxygen sensor (Greisinger Electronic GmbH; Regenstauf, Germany) connected behind the last experimental plastic box.The flow rate of nitrogen was measured using a Flow Sensor SFAB (Festo AG and Co. KG; Esslingen Germany).

Bioassay Experimental Protocol
In the experiments, the efficacy was validated on two species, S. granarius and C. chinensis.Validation of efficacy was performed for all developmental stages (adult, pupa, larva and egg).Each development stage was tested separately.The immature development stages were prepared according to the modified methodology of Aulicky et al. [37] (certified methodology for controlling the effectiveness of controlled atmospheres and fumigation in silos using bio-tests), as follows: (i) Eggs-The adults of both species were placed in a breeding container with substrate for 72 h.Then, the adults were removed and the substrate with the deposited eggs was used for further testing.(ii) Larvae and pupae-Adults were placed in the breeding container with substrate for 1 week.Then, the adults were removed and the substrate was stored at 25 • C in a thermo-chamber.For S. granarius, the substrate was used for larvae between 21-35 days, and for pupae between 35-48 days.For C. chinensis, the substrate was used for larvae between 14-21 days and for pupae between 28-35 days.

Bioassay Experimental Protocol
In the experiments, the efficacy was validated on two species, S. granarius and C. chinensis.Validation of efficacy was performed for all developmental stages (adult, pupa, larva and egg).Each development stage was tested separately.The immature development stages were prepared according to the modified methodology of Aulicky et al. [37] (certified methodology for controlling the effectiveness of controlled atmospheres and fumigation in silos using bio-tests), as follows: (i) Eggs-The adults of both species were placed in a breeding container with substrate for 72 h.Then, the adults were removed and the substrate with the deposited eggs was used for further testing.(ii) Larvae and pupae-Adults were placed in the breeding container with substrate for 1 week.Then, the adults were removed and the substrate was stored at 25 °C in a thermo-chamber.For S. granarius, the substrate was used for larvae between 21-35 days, and for pupae between 35-48 days.For C. chinensis, the substrate was used for larvae between 14-21 days and for pupae between 28-35 days.
The tested stages were placed together with a substrate weighing 15 g (S. granariuswheat grain and C. chinensis-mung beans) in Petri dishes (diameter 60 mm) that had a mesh lid (diameter of the hole in the lid 50 mm).Fifty adult individuals were placed into each dish.The age of adults of S. granarius was 7-14 days and C. chinensis 1-5 days.
All pest species and stages were enclosed and exposed in small plastic hermetic boxes; each box contained 8 petri dishes; one dish from each species and stage.In each treatment, 6 small plastic boxes were connected in series and the application of nitrogen from a pressurized metal cylinder was started.After reaching the target oxygen concentration, the pests were separately exposed for 1, 4, 8, 12, 16 or 20 days.After each exposure The tested stages were placed together with a substrate weighing 15 g (S. granarius-wheat grain and C. chinensis-mung beans) in Petri dishes (diameter 60 mm) that had a mesh lid (diameter of the hole in the lid 50 mm).Fifty adult individuals were placed into each dish.The age of adults of S. granarius was 7-14 days and C. chinensis 1-5 days.
All pest species and stages were enclosed and exposed in small plastic hermetic boxes; each box contained 8 petri dishes; one dish from each species and stage.In each treatment, 6 small plastic boxes were connected in series and the application of nitrogen from a pressurized metal cylinder was started.After reaching the target oxygen concentration, the pests were separately exposed for 1, 4, 8, 12, 16 or 20 days.After each exposure trial, the last small plastic box with the samples was always disconnected and the controlled atmosphere was vented.Subsequently, the treated box was placed in a thermostat with a temperature of 25 • C. Adult mortality was checked 4 days after the end of exposure.It was because it was previously described so called delayed mortality effect in some types of controlled atmospheres [44].Check of efficacy for other developmental stages was performed using adult hatching.The inspections were carried out regularly once a week until the adults stopped hatching.The control samples were prepared in the same way as the experimental ones, but they were not exposed in boxes by hypoxic nitrogen atmosphere.

Silos and Source of N 2
Trials using N 2 were conducted in four identical steel silos with a capacity of 25 tons, fully loaded with a polished rice grain.Three silos were concurrently used as test silos filled with controlled nitrogen atmosphere (CA-N), whereas one silo was used as an untreated control without CA-N.The silos were shaded and thermally insulated by the roof and wall construction.The silos were not completely hermetic.However, they were sealed by glue (metal silo seals) and rubber (silo top and bottom covers) and they were equipped with a gastight press plenum through which the nitrogen was introduced.Due to these additional adaptations, we called the silos "semi-hermetic" further in the text.The silos were located inside the building of a food processing facility located in the Czech Republic (Podravka-Lagris a.s., south Moravia).The treated semi-hermetic silo bins included the discoid panel with a hole and plug for nitrogen filling at the top of the silo (Figure 2D).The hermetic discoid panel-located at the bottom of the silo-has a hole with a plug for nitrogen/oxygen concentration measurement (Figure 2B).For further technical description of semi-hermetic silos, see Aulicky et al. [37].The nitrogen (N 2 ) controlled atmosphere was produced from two on-site installed sets based on Swing Pressure Nitrogen Generators (N 2 -NITRO Source PSA, Parker) (Figure 2A).Each set consisted of the following units: compressor, air receiver, adsorption dryer, nitrogen buffer vessel, low-pressure N 2 storage unit and piping.

Bioassay Exposure
Validation of the effectiveness of the controlled atmosphere with nitrogen was car-

Bioassay Exposure
Validation of the effectiveness of the controlled atmosphere with nitrogen was carried out in four metal silos (each capacity of 25 tonnes) containing polished rice-silos are described above.All developmental stages (adults, pupae, larvae and eggs) of the two primary pest species (S. granarius and C. chinensis) were used in the tests.Biological samples were prepared following the same procedure as for the laboratory tests (Section 2.2.2).In the case of adults, 20 individuals were used per dish.Samples were located in two positions in each silo.One group of samples was placed at the top and the other at the bottom of the silo unit.A total of 32 dishes (4 dishes for each species and developmental stage) were placed at each position in the case of CA-N treated cells.A total of 48 dishes (6 dishes for each species and developmental stage) were placed in the control silo unit without CA-N treatment.Treatments were based on N 2 purging and N 2 maintenance phases.After the start of nitrogen application, there was a so-called purging or filling phase, which lasts 24 h.The target oxygen concentration was below 1% through the entire silo profile, including the bottom of the silo unit.Reaching this concentration, an exposure period of 10 or 20 days was calculated.Thus, the total exposure periods were either 11 days (1 day N 2 purging; 10 days of N 2 maintenance) or 21 days (1 day N 2 purging; 20 days of N 2 maintenance).During the treatment period, the oxygen concentration at the bottom of all silos was checked at regular intervals.At the end of each of both exposure periods, the differentially exposed samples were removed and transferred to the laboratory.The procedure for evaluating the biological efficacy of CA-N on the tested pest species and developmental stages was the same as described above for the laboratory trials.

Statistical Procedures
Data from the laboratory test were analyzed as a multifactorial experiment design for each species separately.In adults, concentration and exposure times were assessed for survival and mortality.For the other developmental stages (eggs, larvae and pupae), concentration, exposure time and developmental stage for adult pupation were evaluated.For statistical purposes, each species was evaluated separately.Exposure time and concentration were entered as factors in the ANOVA.Concentrations also included controls without nitrogen.Each exposure time included separate evaluation of two concentrations.Data were transformed and evaluated using parametric multi-factorial ANOVA test (Statistica statistical software, version 12; StatSoft CR s.r.o., Prague, Czech Republic).The homogeneity of the groups was further evaluated using the Tukey-Kramer HSD post-hoc test at the 0.05 level of significance.

Laboratory Trial
Laboratory results showed that the effectiveness of three different concentrations of nitrogen atmospheres varied for the tested pest species and their developmental stages, depending on the duration of exposure.Parameters of ANOVA for the main effects of controlled atmospheres in Sitophilus granarius and Callosobruchus chinensis are summarized in Table 1.Table 2 shows the effect of various exposure times on adult mortality of S. granarius adults at three concentrations of nitrogen atmosphere.In the controlled atmosphere of 100% N 2 , the complete adult mortality was achieved at an exposure time of 8 days, whereas for 99% N 2 atmosphere it required 12 days' CA-N exposure for the same biological efficacy.The lowest used concentration of nitrogen atmosphere (i.e., 95% N 2 ) failed to deliver complete adult mortality in even 20 days.Table 3 presents data on the emergence of adults from previously treated grain containing eggs, larvae, or pupae of S. granarius by two concentration of nitrogen atmosphere.Controlled atmosphere of 95% CA-N did not lead to suppression of all treated stages even within 20 days, except one case regarding pupae treatment.Both N 2 concentrations (i.e., 99 and 100%) completely suppressed the development of eggs and larvae equally rapidly.The complete inhibition of egg and larval emergence was achieved at 12 day exposure time at both concentrations.However, higher concentration of N 2 (100%) requires longer exposure time (16 days) to suppress development of pupae than the slightly lower N 2 concentration (99%) (12 days).Table 4 shows adult mortality and Table 5 shows the emergence of adults from previously treated eggs/larvae/pupae of C. chinensis following their exposure to three concentration of controlled nitrogen atmosphere (CA-N).For 99% and 100% CA-N the complete C. chinensis adult mortality was reached in 8 days, whereas 95% CA-N required 12 days of exposure.However, the latter was associated with 100% mortality in the untreated control since C. chinensis adults are short-lived.Controlled nitrogen atmosphere (CA-N) achieved more rapid action (12 days) at a slightly lower CA-N concentration (99% N 2 ) than at the highest concentration (100% N 2 ) (16 days) in all immature stages except pupae.At a CA-N concentration of 99% N 2 , 100% larval mortality was attained at the 12 day exposure, while at a higher concentration up to the 16 day exposure.Similar to S. granarius, the lowest used concertation of CA-N (95%) did not lead to the suppression of all treated stages of C. chinensis within 20 days of exposure, except in one case regarding pupae treatment.

Field Trials in Silos
The initial N 2 purging phase of nitrogen into the commodity-filled silos required 24 h for both types of exposures, i.e., 1 + 10 or 1+ 20 days.The average nitrogen consumption was 148.5 L*min −1 in the shorter exposure time and 147.3 L*min −1 in the longer exposure time.From Figures 3 and 4, it is apparent that in all silos oxygen concentrations below 2% was achieved (at the bottom of each silo unit) within 12 h, and it took a further 12 h to achieve oxygen concentrations below 1%.In the subsequent saturation and maintenance phase, despite slight diurnal fluctuations in temperature and atmospheric pressure, oxygen concentrations between 0% and 1% were maintained for both the 10 day exposure (Figure 3) and the 20 day exposures (Figure 4).The data summary on the mortality of the controlled nitrogen's atmospheres (below 1% of oxygen) on adults of S. granarius and C. chinensis in metal silos, under two exposure regimes of 10 or 20 days, can be found in Table 6.Table 7   The data summary on the mortality of the controlled nitrogen's atmospheres (below 1% of oxygen) on adults of S. granarius and C. chinensis in metal silos, under two exposure regimes of 10 or 20 days, can be found in Table 6.Table 7 shows the efficacy of the tested nitrogen atmospheres on eggs, larvae and pupae of both tested species.All field validation tests, carried out in small semi-hermetic silos, showed 100% mortality (Table 6) or 100% development suppression (Table 7) in both exposure regimes and all tested species and their developmental stages under given conditions.

Discussion
Good technological practice for commodity storage on farms, and their onward export and import, requires the availability of robust, rapid and environmentally safe pest control procedures.Compared to toxic fumigants, the use of modified atmospheres (= controlled atmospheres or hermetic storage) is considered challenging in terms of the requirements for relatively long exposure times [16,17,46,47].For a long-term storage of agricultural commodities at large farms or national strategic reserves facilities, the length of exposure to controlled atmospheres used in silos is usually not a major technological constraint [26].However, the length of exposure to controlled atmospheres can play a significant role in the short-term storage of commodities on small farms with limited storage capacity, in ports [39], or in commercial organizations that process agro-commodities (cereals, rice, legumes, dry/dried fruits) into small food packages for retail sales [44,48].For such companies, minimizing concentrations and exposure duration while achieving high efficacy is an essential constraint for the implementation of controlled atmospheres for routine practical use.This work therefore addressed the problem of optimization of the length of exposure by hypoxic nitrogen atmospheres for the control of two important storage pests, under laboratory and small-silos conditions.

Laboratory Trials
In the laboratory part of our work, new biological data were obtained regarding the biological efficacy of controlled nitrogen atmospheres on C. chinensis and S. granarius species under different exposure lengths (1-20 days) and different concentrations (95-100%).From a practical point of view, an important finding of our work was that for most of the tested exposures up to 20 days, the concentrations 99% and 100% nitrogen were highly effective for the control of all developmental stages (eggs, larvae, adult pupae) of both species.In contrast, at 95% nitrogen concentrations, 100% mortality was not achieved for all life stages of both species tested for most of the tested exposure times.The published work on laboratory exposures of C. chinensis and S. granarius with controlled atmospheres is relatively sparse [49][50][51][52][53][54][55].As a result, a direct comparison of our results with data from other authors is also relatively limited.Cui et al. [49] investigated survival, development time delay and metabolomics changes in larvae (4 th instar) of C. chinensis under two hypoxia regimes (2% O 2 = hypoxia or 2% O 2 + 18% CO 2 = hypoxia/hypercapnia).They found that the development of C. chinensis was significantly suppressed by both hypoxia conditions and showed profound differences in metabolites between the treatment and control groups.Cui et al. [49] concluded that C. chinensis has high tolerance to hypoxia since some proportion of 4th-instar larvae survived (i.e., emerged as adults after 24 days) more than 20 days under hypoxia/hypercapnia.In our experiments, we achieved 100% mortality of C. chinensis larvae when exposed in nitrogen atmosphere 99% for 12 days.However, the larval survival and adult emergence of C. chinensis was observed for all tested exposure times at 95% nitrogen atmosphere.Regarding S. granarius, Lindgren and Vincent [50] reported 95% mortality of S. granarius adults after 5.3 day exposure by 100% nitrogen exposure (70 • F and 60-70% RH).Adler [54] observed 99% mortality from 10 strains of S. granarius in 6.6 days (20 • C and 75-99% nitrogen and 1% oxygen).The results by Adler [54,55] are comparable with the results obtained later by Conyers and Bell [56] on the sensitive and resistant strains of S. granarius (99.3-100% mortality, in 8 days at 99% nitrogen atmosphere-50% RH and 20 • C), and with the results obtained in our study (100 % mortality in 8 days at 100% nitrogen atmosphere, in 12 days at 99% nitrogen atmosphere).
Although direct comparison with available literature on the tested species is limited, it is still interesting to compare our results with abundant publications concerning the effectiveness of controlled atmospheres on the related pest species.They mainly include Callosbruchus maculatus [44,[57][58][59][60][61][62], Acanthoscelides obtectus [57,63], Sitophilus zeamais [34], and S. oryzae [33,35].Comparisons of data on C. chinensis and S. granarius with data on related species allow further generalization of the biological effects of inert gases on pests [24,64], and thus the construction of generally applicable atmosphere-controlled technologies and procedures for stored product pests [17].Our laboratory data show that C. chinensis and S. granarius adults were the most susceptible developmental stages, which is in agreement with many findings obtained by other authors for several species of storage pests.For example, Hashem et al. [63] found that hypoxia/hypercapnia effects were more severe in adult bruchids A. obtectus, which led to 100% mortality after 3 days of exposure, but the full hatchability suppression required 7 days.However, for S. zeamais, Williams et al. [35] found that not only adult stages but also eggs were more sensitive to anoxia than larval and pupal stages.The comparison of our data with the previously published results also suggested that the tested species of internal-feeding pests might be less sensitive to controlled atmospheres compared to some species of externally-feeding pests, such as Tribolium sp. and Oryzaephilus sp.[37,44,56,65,66].For example, Sakka et al. [44] found that C. maculatus is more tolerant to nitrogen atmosphere than T. castaneum.They hypothesized that the difference in their sensitivity could be because the immature development of C. maculatus occurs in the inner part of the seeds, where nitrogen penetration may be more gradual, and the T. castaneum individuals are more directly exposed to nitrogen.However, several authors have shown that some externally developing pests, such as Cryptolestes spp.and Trogoderma genera [29,44,46,56,66], can exhibit tolerance to controlled hypoxic/anoxic atmospheres.For example, Conyers and Bell [56] found that adults of C. ferrugineus appeared to be most tolerant to an atmosphere with 1% oxygen, followed by S. granarius, while adults of O. surinamensis were the most sensitive.Moncini et al. [33] reported that an atmosphere containing 98.5% N 2 caused the complete mortality of S. oryzae adults after 3 days of exposure on wheat kernels, whereas the complete mortality of T. confusum adults required 7 days on flour.This is surprising, because it was previously reported [65] that externally developing T. confusum was very sensitive to nitrogen atmosphere.In this case, the question is whether sensitivity may be influenced not only by pest species, but also by the type of food commodity, such as grain or flour [33].All the discrepancies found in the above-mentioned published works suggest that the affiliation of a particular species of a storage pest to an ecotype, with development inside or outside the seed commodity, is not a completely reliable predictor of its sensitivity or tolerance to anoxic/hypoxic atmosphere.In addition, the complex biotic and abiotic relationships discussed above indicate that practical procedures and exposure regimes for the use of a controlled atmosphere in practice should be more or less specific to each pest species.
Another practical objective of our laboratory work was to explore the mortalityconcertation relationship in the tested species for the 95%-100% range of atmospheric nitrogen concentrations.Navarro and Navarro [17] pointed out that effective nitrogen dosage and exposure must be relatively precise and within a certain range of concentrations, because mortality may not have a positive linear relationship with increasing nitrogen concentration over the entire range of values.The basis for this caveat was previous laboratory experiments conducted by Navarro [40].He found that, contrary to intuitive expectation, the necessary exposure time for some species increases when the atmospheric oxygen concentration approaches zero.In other species, however, 100% anoxia may be more effective than various levels of hypoxia.For example, Navarrro [40] described that Ephestia cautella pupae and Tribolium castaneum adults reacted to increasing concentrations of N 2 in a similar pattern, whereas S. oryzae adults reacted in a different way.The lower the oxygen concentration was, the shorter the exposure time needed to produce 95% mortality of E. cautella pupae and T. castaneum.However, S. oryzae adults showed greater sensitivity at 1% oxygen than at the zero level of oxygen concentration.In our study, we found that responses to different oxygen concentrations differed not only between the two species tested, but also among developmental stages within a species.In S. granarius, lethal exposure was shorter at 100% nitrogen concentration than at the lower concentration (99%), but the opposite was true for the pupal stage of this species.For the species C. chinensis, the exposure time leading to 100% mortality was the same for the 99% and 100% concentrations, whereas for the immature stage, the exposure time leading to 100% mortality was higher for the lower of the two concentrations tested, 99%.

Field Validations
The required effective pest control exposures of commodities to hypoxic atmospheres can exceed one month, depending on the technological and environmental conditions in commodity stores and silos [35].However, such long exposure times may not be acceptable for many companies for operational and economic reasons [44].Users of controlled atmosphere usually balance between minimizing the length of the hypoxic exposure of infested commodities and maximizing its biological efficacy on pests.Unfortunately, practitioners frequently do not have detailed information to optimize these control processes for many pest species when controlling them in small silos.Most practical validations are either from large silos [67,68], laboratory small silos [34], mini-silos [35], glass desiccators [56] or hermetic chambers [44,48].Published data regarding the use of controlled atmospheres for the treatment of commodities in smaller silos (25t) is lacking, or is very limited [35,37,69].At the same time, detailed technological information is very important because the treatment of small silos is challenging; even small gas leaks cause a significant reduction in the concentration of inert gases [17,70].For potential users, compensating for nitrogen leakage from silos by continuously saturating the silo-atmosphere with inert gases may seem economically and operationally demanding.However, our fieldwork has shown that good results can be achieved even in small metal silos in the normal operating conditions of a commercial commodity store (Podravka-Lagris, CZ).Our validations in these silos demonstrated that these types of silos could be saturated by the 98% nitrogen within 12 h, and a target hypoxic concentration above 99% can be well achieved within 24 h.It was further demonstrated that under a period of 10-20 days, a relatively stable level of over 99% nitrogen can be maintained; oxygen concentration oscillating between 0.3-0.7%.However, it should be noted that the treated silos were located indoors.Thus, the large fluctuations in temperature and the effects of direct sun-glare were reduced, that likely buffered large fluctuations in controlled atmosphere concentrations.Indeed, other authors have documented that the method of application, the type of facility and/or temperature/sunshine fluctuations can cause concentration fluctuations and uneven distribution of inert and toxic gases in treated stores, chambers and transport vehicles [11,44,71,72].Regarding the importance of the necessity of an even concentration distribution of nitrogen in the treated chamber, Sakka et al. [44] stressed that: "pest location is a crucial parameter that may determine the insecticidal effect of nitrogen.In our study, we saw that despite variations in insect survival in the different locations, some of the insects that survived we found in the vials that had been placed in the "heart" of the pallet.This is somehow expected, as nitrogen penetrates into the pallet from the outside to the inside, and thus, "oxygen nests" can be created in the internal part of the product mass."From this point of view, it would be interesting to compare our results (concerning the stability and concentration variations of a controlled hypoxic atmosphere) with experiments carried out in small silos located outside shielded and tempered spaces, in the future experiments.
Concurrently with the validation of nitrogen filling and concentrations maintenance in the silos, bio tests were placed in the upper and lower parts of the silos.This is because our previous work [37] revealed that there might be differences in the biological efficacy achieved between the top and bottom of the silo when filling silos by nitrogen from top to bottom, and using very short exposures.However, in the current validation experiments, it was shown that both exposure regimes (11 and 21 days) resulted-under given environmental conditions-in 100% control of all developmental stages of C. chinensis and S. granarius at both the top and bottom of the treated silos.Thus, these tests yielded optimistic results, acceptable for the use of relatively short exposure times of controlled atmospheres in small silos.On-site nitrogen production using a swing-pressure-generator is economically challenging; especially in the current global energy crisis.Thus, the energy savings made possible by short exposure times, and thus reduced nitrogen production needs, is a very important and timely result of this work for practice.However, we believe that the results obtained for C. chinensis and S. granarius still need to be validated for other species.Indeed, there is an experimentally supported suspicion that some species, such as T. granarium, may be highly tolerant to hypoxic [44,46] and hyper-carbic [46] atmospheres.To control resistant species such as T. granarium, it will likely be necessary to increase the length of exposure, manipulate temperature/pressure [48,73], or use a combination of stressors such as radiation and controlled atmospheres [29].Other aspects that merit future research attention may be effects caused by differences in the sensitivity of different geographic strains of pests to anoxic and hypoxic atmospheres.The differences in the tolerance of different strains of pests to controlled atmospheres has been noted by Adler [54] and Conveyers and Bell [56], who reported differences in the responses to hypoxia of S. granarius populations from different countries and geographic regions.

Conclusions
Hypoxic or anoxic atmospheres belong among promising and environmentally acceptable methods of effective stored product pest control, both for farmers and for international phyto-quarantine services [16,74,75].Moreover, they have the potential for the effective control of pest populations resistant to traditional fumigants, such as phosphine.In this paper, data were presented regarding both the laboratory and field efficacy of hypoxic atmospheres on different developmental stages of two primary key pest species (C.chinensis and S. granarius) with intra-grain immature development.The laboratory data indicated that 10-20 days' exposure resulted in 100% control of all developmental stages of C. chinensis and S. granarius when nitrogen concentrations was maintained between 99% and 100%, under defined temperature and humidity.Hypoxic nitrogen concentrations of 99% had even higher efficacy than completely anoxic atmospheres in the laboratory conditions tested.However, the 95% nitrogen concentration atmosphere did not result in 100% control of both species at the exposures given, even under controlled laboratory conditions for exposures of up to 20 days.Experimental validation in the silos, located indoors without large temperature fluctuations and direct sunshine-warming effects, showed that a targeted hypoxic concentration above 99% could be well achieved within 1 day of nitrogen initial purging, and stably maintained for the consequent 10-20 days.Bio tests indicated that both exposure regimes (10 + 1 and 20 + 1 days) resulted in 100% control of all developmental stages of C. chinensis and S. granarius under the given environmental conditions.Therefore, the use of nitrogen atmospheres in smaller silos for relatively short exposure periods can be considered a promising effective method for controlling C. chinensis and S. granarius species in small silos equivalent to the capacity of a standard shipping container.Thus, the results are of interest both for the protection of stored commodities on small farms, and for commercial organizations storing/treating separately a supply of commodities corresponding to the size of standard shipping containers.However, it is recommended to extend the tests and validations to other pest species and geographical strains/populations as they may have different sensitivities [56,66].

Figure 1 .
Figure 1.Testing apparatus for validation of nitrogen control atmospheres in laboratory conditions (A)-technical scheme; (B)-plastic boxes in a climatised thermostatic-chamber; (C)-source of N2 (gas cylinders); (D)-bio-samples of pests and commodites in Petri dishes covered by mesh lid.

Figure 1 .
Figure 1.Testing apparatus for validation of nitrogen control atmospheres in laboratory conditions (A)-technical scheme; (B)-plastic boxes in a climatised thermostatic-chamber; (C)-source of N 2 (gas cylinders); (D)-bio-samples of pests and commodites in Petri dishes covered by mesh lid.

Figure 2 .
Figure 2. Visualisation of the validation of biological efficacy of nitrogen-controlled atmospheres in silos.(A)-the on-site-build pressure swing nitrogen generator (N2-NITRO Source PSA-Parker) consisting of compressor, air receiver, adsorption dryer, nitrogen buffer vessel, low pressure N2 storage unit and piping; (B)-hermetic removable discoid panel at the bottom of the silo with a hole and plug for nitrogen/oxygen measures; (C)-top board of silo bins with rectangular opening for loading the commodity and discoid panels; (D)-detail of a hermetic discoid panel with plug for nitrogen filling at the top of the silo.

Figure 2 .
Figure 2. Visualisation of the validation of biological efficacy of nitrogen-controlled atmospheres in silos.(A)-the on-site-build pressure swing nitrogen generator (N 2 -NITRO Source PSA-Parker) consisting of compressor, air receiver, adsorption dryer, nitrogen buffer vessel, low pressure N 2 storage unit and piping; (B)-hermetic removable discoid panel at the bottom of the silo with a hole and plug for nitrogen/oxygen measures; (C)-top board of silo bins with rectangular opening for loading the commodity and discoid panels; (D)-detail of a hermetic discoid panel with plug for nitrogen filling at the top of the silo.

Figure 3 .
Figure 3.Oxygen concentration during field test-10 + 1 days of exposure (n = 3 silo units); 1 day N2 purging; 10 days of full N2 concentration exposure.The average oxygen concentration after initial 12 h nitrogen purging was 1.5 ± 0.1%; the average oxygen concentration for the 10 days of nitrogen maintenance was 0.4 ± 0.0%.

Figure 4 .
Figure 4.Oxygen concentration during field test-20 + 1 days of exposure (n = 3 silo units); 1 day N2 purging; 20 days full N2 concentration exposure.The average oxygen concentration after initial 12 h nitrogen purging was 1.5 ± 0.1%; the average oxygen concentration for the 20 days of nitrogen maintenance was 0.5 ± 0.0%.

Figure 3 .Figure 3 .
Figure 3.Oxygen concentration during field test-10 + 1 days of exposure (n = 3 silo units); 1 day N 2 purging; 10 days of full N 2 concentration exposure.The average oxygen concentration after initial 12 h nitrogen purging was 1.5 ± 0.1%; the average oxygen concentration for the 10 days of nitrogen maintenance was 0.4 ± 0.0%.Figure 3.Oxygen concentration during field test-10 + 1 days of exposure (n = 3 silo units); 1 day N2 purging; 10 days of full N2 concentration exposure.The average oxygen concentration after initial 12 h nitrogen purging was 1.5 ± 0.1%; the average oxygen concentration for the 10 days of nitrogen maintenance was 0.4 ± 0.0%.

Figure 4 .
Figure 4.Oxygen concentration during field test-20 + 1 days of exposure (n = 3 silo units); 1 day N2 purging; 20 days full N2 concentration exposure.The average oxygen concentration after initial 12 h nitrogen purging was 1.5 ± 0.1%; the average oxygen concentration for the 20 days of nitrogen maintenance was 0.5 ± 0.0%.

Figure 4 .
Figure 4.Oxygen concentration during field test-20 + 1 days of exposure (n = 3 silo units); 1 day N 2 purging; 20 days full N 2 concentration exposure.The average oxygen concentration after initial 12 h nitrogen purging was 1.5 ± 0.1%; the average oxygen concentration for the 20 days of nitrogen maintenance was 0.5 ± 0.0%.

Table 1 .
Parameters of ANOVA for main effects related to mortality of the exposed adults of Sitophilus granarius (df = 180) and Callosbruchus chinensis (df = 180) or to S. granarius (df = 315) and C. chinensis-(df = 315).adult emergence following their previous N 2 exposure in sub-adult stages (egg, larva and pupa).

Table 2 .
Laboratory efficacy (% of mean adult mortality; Av. % ± SE) of six exposure times (ranging from 1 to 20 days) and three concentrations (ranging from 95% to 100% N 2 ) nitrogen-based controlled atmospheres on the exposed adults of grain weevil (Sitophilus granarius).(Different letters indicate statistically significant differences between variables).

Table 3 .
Laboratory efficacy (Average No. of adults emerged following their treatment in subadult stages; Av.No. ± SE) of six exposure times (ranging from 1 to 20 days) and three concentrations (ranging from 95% to 100% N 2 ) nitrogen-based controlled atmospheres on eggs, larvae and pupae of grain weevil (Sitophilus granarius).(Different letters indicate statistically significant differences between variables).

Table 4 .
Laboratory efficacy (% of mean adult mortality; Av. % ± SE) of six exposure times (ranging from 1 to 20 days) and three concentrations (ranging from 95% to 100% N2) nitrogen-based controlled atmospheres on the exposed adults of adzuki bean weevil (Callosobruchus chinensis).(Different letters indicate statistically significant differences between variables).

Table 5 .
Laboratory efficacy (Average No. of adults emerged following their treatment in subadult stages; Av.No. ± SE) as of six exposure times (ranging from 1 to 20 days) and three concentrations (ranging from 95% to 100% N 2 ) nitrogen-based controlled atmospheres on eggs, larvae of adzuki bean weevil (Callosobruchus chinensis).(Different letters indicate statistically significant differences between variables).
shows the efficacy of the tested

Table 6 .
Field efficacy (% of adult mortality; Av. % ± SE) of controlled nitrogens atmospheres (below 1% of oxygen) on adults of the tested species (S. granarius, C. chinensis) in metal silos under two exposure regimes 11 (1 day N 2 purging; 10 days full N 2 concentration exposure) or 21 days.For each treated silo units (n = 3 silos) 4 bio-assay-dishes with insects were used.

Table 7 .
Field efficacy (Average No. of adults emerged following their treatment in subadult stages; Av.No. ± SE) of controlled nitrogens atmospheres (below 1% of oxygen) on eggs, larvae and pupae of the tested species in metal silos under two exposure regimes of 11 or 21 days.For each treated silo units (n = 3 silos) 4 bio-assay-dishes with insects were used.