Introduction
Storage pests can potentially cause extensive economic damage to stored grain (Mahroof and Hagstrum, Reference Mahroof, Hagstrum, Hagstrum, Phillips and Cuperus2012) and contamination from fragments and allergenic cereal dusts (Farant and Moore, Reference Farant and Moore1978; Trematerra et al., Reference Trematerra, Stejskal and Hubert2011; Hubert et al., Reference Hubert, Stejskal, Athanassiou and Throne2018). The actually realized pest potential is determined by the interplay between the biological abilities of each species and currently existing biotic and abiotic environmental factors. In a regulated warehouse environment, the relevant abiotic factors, such as temperature and humidity, are relatively limited and stable (Maier et al., Reference Maier, Adams, Throne and Mason1996; Stejskal et al., Reference Stejskal, Vendl, Li and Aulicky2019). Biotic interactions are much more complex and dynamic (Arbogast and Mullen, Reference Arbogast and Mullen1988) since they usually involve multiple relationships among arthropods, microorganisms, and different types of stored commodities (Crombie, Reference Crombie1941; Sinha, Reference Sinha1969; LeCato, Reference LeCato1975; Sinha and Sinha, Reference Sinha and Sinha1990; Dukic et al., Reference Dukic, Radonjic, Levic, Spasic, Kljajic and Andric2016; Fleurat-Lessard, Reference Fleurat-Lessard2017; Rumbos et al., Reference Rumbos, Pantazis and Athanassiou2019). Complexity is also increased by diverse arthropod reactions to chemical changes in grain caused by previous grain infestations (e.g., Stewart-Jones et al., Reference Stewart-Jones, Hodges, Birkinshaw and Hall2004; Trematerra, et al., Reference Trematerra, Valente, Athanassiou and Kavallieratos2007; Stewart-Jones et al., Reference Stewart-Jones, Stirrup, Hodges, Farman and Hall2009) and the effects associated with the gut enzymes and microbiome composition of arthropods (e.g., Osipitan et al., Reference Osipitan, Akintokun, Odeyemi and Bankole2011; Hubert et al., Reference Hubert, Stejskal, Nesvorna, Aulicky, Kopecky and Erban2016; Naseri et al., Reference Naseri, Borzoui, Majd and Mansouri2017).
With the advent of integrated and holistic approaches for commodity protection (Jayas et al., Reference Jayas, White and Muir1995; White, Reference White1995; Arbogast and Throne, Reference Arbogast and Throne1997), there has been significant renewed interest in interspecific interactions. Negative interactions, such as predation, parasitism, and competition, are a particular focus of research (Lukas et al., Reference Lukas, Stejskal, Jarosik, Hubert and Zdarkova2007; Athanassiou et al., Reference Athanassiou, Kavallieratos and Campbell2017; Quellhorst et al., Reference Quellhorst, Athanassiou, Bruce, Scully and Morrison2020). However, interspecific interactions are not necessarily negative. For example, Nansen et al. (Reference Nansen, Phillips and Palmer2004) reported a positive commensal relationship between Rhyzopertha dominica (F.) and Tribolium castaneum (Herbst). They did not illuminate the causes of this association, but one of the common explanations for positive interactions among stored-product insects is that substrate colonization by internally feeding species can disintegrate hard seeds, making cereal food available to another species (e.g., Arbogast and Mullen, Reference Arbogast and Mullen1988). It has been shown that decreasing proportions of sound grain to dockage, cereal dust and flour can significantly affect the survival, multiplication, and interspecific interactions of a number of species (McGregor, Reference McGregor1964; LeCato, Reference LeCato1975; Sinha, Reference Sinha1975; Locatelli et al., Reference Locatelli, Limonta and Stampini2008). Internally feeding pests may contribute profoundly to the increase in dockage and dust contents in grain mass since they produce fine small dusty particles known as frass. Specifically, the feeding of bostrichid beetles such as R. dominica and Prostephanus truncatus (Horn) may result in large amounts of frass in an infested commodity (Stewart-Jones et al., Reference Stewart-Jones, Hodges, Birkinshaw and Hall2004; Stewart-Jones et al., Reference Stewart-Jones, Stirrup, Hodges, Farman and Hall2009; Edde, Reference Edde2012; Kavallieratos et al., Reference Kavallieratos, Athanassiou, Guedes, Drempela and Boukouvala2017; Nyabako et al., Reference Nyabako, Mvumi, Stathers, Mlambo and Mubayiwa2020). It was recently demonstrated that externally feeding pests such as T. castaneum (Bekon and Fleurat-Lessard, Reference Bekon and Fleurat-Lessard1992; Arthur et al., Reference Arthur, Starkus, Gerken and Campbell2019) and Trogoderma granarium Everts (Kavallieratos et al., Reference Kavallieratos, Athanassiou, Guedes, Drempela and Boukouvala2017) may also produce some amount of frass. Not only insect species and its developmental stage (i.e. larva/adult) but also the type of food can influence the amount, form, and composition of the resulting faecal/frass materials (Weiss, Reference Weiss2006; Kavallieratos et al., Reference Kavallieratos, Athanassiou, Guedes, Drempela and Boukouvala2017). Earlier resources claimed (e.g., Potter, Reference Potter1935) that frass of R. dominica mainly consists of food material that is chewed off but not eaten. It is currently known that frass contains not only the original non-ingested food material (grain) itself, but also faeces and additional components such as microorganisms and various chemical compounds (Breese, Reference Breese1960; Osipitan et al., Reference Osipitan, Akintokun, Odeyemi and Bankole2011; Edde, Reference Edde2012; Boiocchi et al., Reference Boiocchi, Porcellato, Limonta, Picozzi, Vigentini, Locatelli and Foschino2017). Several previously published works (e.g., Bekon and Fleurat-Lessard, Reference Bekon and Fleurat-Lessard1992; Kavallieratos et al., Reference Kavallieratos, Athanassiou, Guedes, Drempela and Boukouvala2017) listed frass of primary colonizers among the factors positively influencing colonization and population performance of secondary colonizers. However, these works were not designed to distinguish the positive effects of the frass from the positive effect of physical disintegration of the sound grain. Thus, the specific effects of frass of R. dominica and other primary colonizers on the development of other primary or secondary stored product species have remained incompletely understood. Therefore, the goal of this work was to separately evaluate the impacts of the frass produced R. dominica on three species of externally feeding pests: Oryzaephilus surinamensis (L.), Tribolium confusum (Jacquelin du Val), and T. castaneum. The selected species are important pests in grain stores in the Czech Republic (Stejskal et al., Reference Stejskal, Aulicky and Kucerova2014, Reference Stejskal, Hubert, Aulicky and Kucerova2015) as well in many other countries (e.g., Hagstrum and Subramanyam, Reference Hagstrum and Subramanyam2006; Trematerra and Throne, Reference Trematerra, Throne, Sissions, Abecassis, Marchylo and Carcea2012). The hypothesis was that frass is a more supportive food for each of the tested species of externally feeding pests than simple cereal substrates (irrespective of their physical state, such as whole grain, broken kernels, or flour). The reason for our hypothesis was that previous studies have indicated (Breese, Reference Breese1960; Osipitan et al., Reference Osipitan, Akintokun, Odeyemi and Bankole2011; Edde, Reference Edde2012; Boiocchi et al., Reference Boiocchi, Porcellato, Limonta, Picozzi, Vigentini, Locatelli and Foschino2017) that in addition to undigested cereal starch particles, frass contains other chemical and microbial components. To test our hypotheses, we specifically compared the effects of frass on secondary pest (i.e. T. castaneum, T. confusum, and O. surinamensis) performance with the effects of various physical qualities of grain (intact grain, grain fragments, flour, grain + frass) and artificially enriched control diets (milled wheat kernels, oat flakes, and yeast).
Materials and methods
Beetle rearing
Tribolium castaneum, T. confusum (Coleoptera: Tenebrionidae), O. surinamensis (Coleoptera: Silvanidae) and R. dominica (Coleoptera: Bostrichidae) used in the study came from the laboratory cultures reared for more than 20 generations under close-to-optimal conditions (i.e., sensitive to insecticides and natural environmental stresses) at the Crop Research Institute in Prague, Czech Republic. The strain of O. surinamensis was kept in thermochambers (Aviko, Czech Republic) at 27 (±0.5) °C with 75 (±5) % relative humidity (r. h.). The food was a mixture of powdered wheat (80%), oat flakes (16%), germ (3%), and yeast (1%). Both Tribolium species were also kept in thermochambers (Aviko, Czech Republic) at 26 (±0.5) °C and 75 (±5) % r. h. The rearing boxes contained a mixture of milled wheat kernels, oat flakes, and yeast (10:10:2), and the substrate was covered with moistened filter paper. The rearing conditions of R. dominica were identical as described in Kucerova and Stejskal (Reference Kucerova and Stejskal2008).
Substrate preparation
Five types of substrates were tested in the experiment. They were whole undamaged wheat grains, cracked wheat grains, milled wheat grains (flour), frass (originating from the chewing activity of R. dominica on wheat grains), and a mixture of the frass and whole grains. In addition to these substrates, a control diet was used. In both Tribolium species, the control diet was composed of a mixture of milled wheat kernels, oat flakes, and yeast (10:10:2). The diet of O. surinamensis contained whole wheat kernels, oat flakes, and a mixture of milled wheat kernels, oat flakes, and yeast (10:10:2). All substrates were frozen for at least 10 days prior to the experiment to exclude all unwanted organisms. Cracked grains were milled using a centrifugal mill (ZM 200, Retsch) with a 10 mm sieve size, and the cracked kernels were sieved with a 2 mm-pore size sieve. The flour was bought from a shop. The frass originated from a laboratory culture of R. dominica reared on wheat grains. It was obtained by sieving the rearing medium using a 0.5 mm-pore size sieve using automatic elaborate shaking and vibratory sieving equipment (vibratory sieve shaker AS 200, Retsch, Germany). The ratio of whole wheat and frass in the mixed substrate was 1:1. Seven days before the experiment, all substrates were placed in a Lock & Lock plastic box (180 × 110 × 110 mm3) with a saturated solution of NaCl (75%) to ensure uniform relative humidity. The substrate moisture content was measured using a moisture metre (Agromatic, Farmer Tronic, Denmark); the measured moisture content was 15.7%.
Study design
To study the population growth of the tested species (i.e., T. castaneum, T. confusum, and O. surinamensis), 10 grams (weighed on an Adventurer Pro analytical scale, Ohaus, USA) of a particular substrate was placed in a 100 ml plastic vial with a perforated cap covered with nylon mesh. The top of the inner walls of the cup were treated with synthetic fluoropolymer lubricant powder (polytetrafluoroethylene - Fluon) (Sigma-Aldrich Co., St. Louis, USA) to prevent beetles from escaping. Ten unsexed adult beetles were introduced into each cup and kept in a climatic chamber at 27 (± 0.5)°C and 75 (± 5)% r.h. in randomized positions. After 10 days, the cup contents were spilled out on Petri dishes, and all beetles were carefully removed by hand. For the determination of development time, the vials were checked for the presence of adults every other day (starting on the 33rd day). To prevent the disruption of beetle development, the vials were checked visually, and the substrate was not sieved (thus, especially at low adult densities, the method may not precisely reflect the real development time). The development time for each vial was recorded as the time at which at least one individual adult was present. After 8 weeks, the cups were open, and the numbers of larvae, pupae, and adults were counted under an Olympus SZX10 binocular microscope. Five replicates per species/substrate were performed.
Statistics
The effects of the substrate on the population size and development time were tested by one-way analysis of variance followed by Tukey's HSD post hoc test, where the substrate type was used as the factor. Because the data regarding the numbers of individuals (tested by the Kolmogorov–Smirnov test) were non-normally distributed, the data were log(x + 1) transformed to ensure normality. For simplicity, untransformed means and standard errors are reported. The analyses were carried out using Statistica 12.0 (StatSoft Inc., 2010).
Results
Oryzaephilus surinamensis
The population growth of O. surinamensis on various substrates is summarized in table 1. Apparently, this species was able to thrive on frass produced by R. dominica; there were no significant differences in the numbers of larvae or adults or the total number of individuals developing in frass alone or in a mixture of frass and whole grains. However, the total number of individuals in frass alone and in a mixture of frass and whole grains were 3x and 4.6x less than in a control substrate (table 1). Moreover, population growth on substrates containing frass was higher than that on flour and cracked grains. In contrast, the species was not able to develop on undamaged grains. There was a trend of decreasing development time with increasing suitability of the substrate (as identified by the population growth on the respective substrate, see table 1), although the differences were not statistically significant.
Table 1. Population growth and development time (means ± SEs) of Oryzaephilus surinamensis on various substrates in comparison with the control diet.
Within each column (except for the control: experimental substrate ratio column), means followed by the same letter are not significantly different; in cases with no letters, no significant differences exist; Tukey's HSD post hoc test at 0.05, in all cases df = 5, 24.
a The ratio was computed as (total number of individuals in control substrate + 1)/(total number of individuals in experimental substrate + 1).
Tribolium castaneum
The population growth of T. castaneum on various substrates is summarized in table 2. In general, the results were similar to those for the above species. Substrates containing frass (i.e., frass alone and frass + grains) proved to be suitable for the development of T. castaneum (there were no significant differences from the control diet – the population growth was only 1.3x and 1.8x less than in the control diet for both experimental substrates). All developmental stages (i.e., larvae, pupae, and adults) were present in the two substrates in comparable numbers. Population growth in frass alone was also greater than in flour and cracked grains. In undamaged grains, the species was not able to develop. The development times showed a similar pattern to those of O. surinamensis (although, again, the differences were not statistically significant). The development time was probably greatest in the presence of cracked grains, so only larvae and no adults were present in this substrate.
Table 2. Population growth and development time (means ± SEs) of Tribolium castaneum on various substrates in comparison with the control diet.
Within each column (except for the control:experimental substrate ratio column), means followed by the same letter are not significantly different; in cases with no letters, no significant differences exist; Tukey's HSD post hoc test at 0.05, in all cases df = 5, 24.
a The ratio was computed as (total number of individuals in control substrate + 1)/(total number of individuals in experimental substrate + 1).
Tribolium confusum
The population growth of T. confusum on various substrates is summarized in table 3. Similar to both of the other species, T. confusum was able to thrive on frass, but in this case, population growth was not higher in frass than in flour and cracked grains. In contrast to the other two species, the performance of T. confusum on whole grains + frass was slightly (albeit nonsignificantly) better than that on frass alone. Generally, there were similar ratios of number of individuals in four experimental substrates (whole grains + frass, frass alone, flour, cracked grains) compared with control diet (ranging from 2.6 to 5.6). In undamaged grains, the species was not able to develop. The development time in both frass substrates and flour was similar but was slightly (although not significantly) shorter than that in cracked grains.
Table 3. Population growth and development time (means ± SEs) of Tribolium confusum on various substrates in comparison with the control diet.
Within each column (except for the control: experimental substrate ratio column), means followed by the same letter are not significantly different; Tukey's HSD post hoc test at 0.05, in all cases df = 5, 24.
a The ratio was computed as (total number of individuals in control substrate + 1)/(total number of individuals in experimental substrate + 1).
Discussion
The presence of broken kernels and frass is reported to affect the behaviour, interaction, and population growth of secondary pests (LeCato, Reference LeCato1975; Sinha, Reference Sinha1975; Arbogast and Mullen, Reference Arbogast and Mullen1988; Beckett and Evans, Reference Beckett and Evans1994; Kavallieratos et al., Reference Kavallieratos, Athanassiou, Guedes, Drempela and Boukouvala2017). Although there are available data on the effects of frass and dust as behavioural modifiers since they may contain kairomones (Stewart-Jones et al., Reference Stewart-Jones, Stirrup, Hodges, Farman and Hall2009), there have been no data showing how frass can affect the populations and speed of development of any secondary pest species. This work, therefore, constitutes the first comparison of the separately analysed effects of frass with the effects of cereal substrates (whole kernels, cracked kernels, and wheat flour) on the speed of development and productivity (performance) of three secondary pests, O. surinamensis, T. castaneum, and T. confusum. There was no (T. castaneum and T. confusum) or minimal (O. surinamensis) development of the tested pest species on whole undamaged seeds. Conversely, all tested species (with varying productivity) multiplied on seed fragments, flour, frass, and the mixed control diet. These results confirmed the results of other authors showing that the physical structure of cereal substrates plays an important role in the biology of secondary storage pests. For instance, several studies (Fraenkel and Blewett, Reference Fraenkel and Blewett1943; Turney, Reference Turney1957; Fleming, Reference Fleming1998; Beckel et al., Reference Beckel, Lorini and Lazzari2007) have found that O. surinamensis is mostly unable to develop on whole cereal (e.g., wheat, corn, rice) seeds, while crushed seeds and flour allow population growth in this species. Beckel et al. (Reference Beckel, Lorini and Lazzari2007; Skourti et al., Reference Skourti, Kavallieratos and Papanikolaou2020) compared different flour milling grades (i.e., 1–20) and found that wheat grain milled at grade 20 yielded the largest number of progeny of O. surinamensis. Several works claim that T. castaneum can survive and multiply on whole seeds (Fraenkel and Blewett, Reference Fraenkel and Blewett1943; Daniels, Reference Daniels1956), but other studies do not support these conclusions (White, Reference White1982). Therefore, Li and Arbogast (Reference Li and Arbogast1991) ranked cereal substrates in terms of their suitability for T. castaneum population development as follows: flour>cracked seeds>whole intact seeds. It appears from published works that a finer physical structure provides a more suitable medium for the development of Tribolium spp. and Oryzaephilus spp. Astuti et al. (Reference Astuti, Rizali, Firnanda and Widjayanti2020) found that the particle size variation and protein content of flour products affected the survivorship and development time of T. castaneum. Fardisi et al. (Reference Fardisi, Mason, Ileleji and Richmond2019) claimed that flour particle size affected the suitability of the microclimate for T. castaneum development. Frass, with its fine particle size, is closest to flour, semolina, and environmental dust, which are substrates upon which secondary pests multiply well (e.g., Locatelli et al., Reference Locatelli, Savoldelli, Girgenti, Lucchini and Limonta2017). Therefore, one partial explanation for the beneficial effect of frass on pest development may be its fine physical structure.
On the other hand, even the favourable physical decomposition of grain and other cereal substrates does not inherently result in the most appropriate dietary composition for Tribolium spp. or for Oryzaephilus spp. (LeCato and McCray, Reference LeCato and McCray1973; Shafique et al., Reference Shafique, Ahmad and Chaudry2006; Astuti et al., Reference Astuti, Rizali, Firnanda and Widjayanti2020). Notably, the presence of germ in the cereal diet is important. Armstrong and Howe (Reference Armstrong and Howe1963) and White (Reference White1982) observed low survival of O. surinamensis when the insects fed only the endosperm of grain without the germ. Locatelli et al. (Reference Locatelli, Savoldelli, Girgenti, Lucchini and Limonta2017) found that T. confusum could also multiply on very fine grain silo environmental dust that had been supplemented with microscopic fragments of dietary insects and other organic and inorganic compounds. In our experiments, all three species of pests (T. confusum, T. castaneum, O. surinamensis) showed the highest productivity and development rates by far on the optimized laboratory control diet rather than on flour or frass. Astuti et al. (Reference Astuti, Rizali, Firnanda and Widjayanti2020) stressed that in addition to protein and particle size variation, other predictors (such as water, ash, phenol, and riboflavin contents) also affect the development of T. castaneum. This observation is in concordance with earlier findings by Sokoloff et al. (Reference Sokoloff, Franklin, Overton and Ho1966) showing that the development of T. castaneum requires not only protein but also other nutrients, such as minerals and vitamins. Our results obtained with the optimized diet and the work of other authors (Fraenkel and Blewett, Reference Fraenkel and Blewett1943; LeCato and McCray, Reference LeCato and McCray1973; Beckel et al., Reference Beckel, Lorini and Lazzari2007; Locatelli et al., Reference Locatelli, Savoldelli, Girgenti, Lucchini and Limonta2017) show that most secondary pests exhibit a higher performance on a richer diet than on carbohydrate-based cereals alone (i.e., a diet containing additional components such as proteins, lipids, vitamins, yeasts, etc.). In our experiments, frass was found to be a significantly better food medium than flour or cracked grain for O. surinamensis and T. castaneum, while T. confusum performed equally well on flour and frass. Our results showed that in terms of food quality and suitability for the tested species, wheat frass occupied an intermediate position between the optimal diet mixture and simple uniform cereal diets, such as cracked wheat grain or flour. Our biotests thus also indirectly indicated that R. dominica wheat frass is not just cereal dust containing waste and ingestible faecal particles but is instead a rich potential diet for secondary pests. According to the limited relevant published data, R. dominica frass consists not only of digested faecal parts but also of undigested food particles containing high nitrogen levels, digestive proteins, and the excreted microbiome (Breese, Reference Breese1960; Edde, Reference Edde2012). There are no available data on the protein and fat contents of frass produced by R. dominica. However, there are data on the frass composition produced by flies fed dried distiller grains (Yildirim-Aksoy et al., Reference Yildirim-Aksoy, Eljack and Beck2020): the larvae produce frass with high protein (216 g/kg) and fat contents (60 g/kg). Stewart-Jones et al. (Reference Stewart-Jones, Stirrup, Hodges, Farman and Hall2009) studied the chemical composition of frass produced by stored insects and identified triglycerides and five free fatty acids as the most abundant chemicals in frass. They also found that Sitophilus spp. did not change the composition of free acids, while R. dominica feeding increased their contents by four–sixfold in frass. The results obtained by Osipitan et al. (Reference Osipitan, Akintokun, Odeyemi and Bankole2011) revealed that frass produced by the P. truncatus contained ten species of bacteria and six species of fungi. A recent analysis (Boiocchi et al., Reference Boiocchi, Porcellato, Limonta, Picozzi, Vigentini, Locatelli and Foschino2017) showed that frass of R. dominica and another tested storage pests was enriched by an extensive microbiome (including bacteria – Enterobacteriaceae, Pseudomonadacae and Bacillaceae; yeasts – Candida; and other fungi – Saccharomycetales, Pleosporaceae, and Nectriaceae).
In conclusion, this work is an experimental proof that wheat frass produced by R. dominica may support the survival and population growth of secondary pests such as T. confusum, T. castaneum, and O. surinamensis. For these species, it may represent an even better diet than plain wheat grains in any physical form (whole grain, cracked grain, and flour). Our results thus imply several practical consequences for pest risk assessment and management in stored grain. First, the presence of frass in stored grain has the potential to affect interspecific interactions and may provide a hypothetical explanation for the previous finding reported by Nansen et al. (Reference Nansen, Phillips and Palmer2004) that the presence of R. dominica has a positive effect on T. castaneum populations. It should be mentioned that the recent observations revealed certain beneficial interactions not only between primary and secondary colonisers but also between primary colonizers such as P. truncatus, R. dominica, and Sitophilus spp. (Athanassiou et al., Reference Athanassiou, Kavallieratos and Campbell2017; Kavallieratos et al., Reference Kavallieratos, Athanassiou, Guedes, Drempela and Boukouvala2017; Quellhorst et al., Reference Quellhorst, Athanassiou, Bruce, Scully and Morrison2020). Frass was also listed among the possible supportive interspecific ecological mechanisms among populations of primary colonisers (Kavallieratos et al., Reference Kavallieratos, Athanassiou, Guedes, Drempela and Boukouvala2017). We therefore suggest that additional experiments should be done to illuminate the exact effects of frass on various species of primary colonizers; e.g. in a similar way as demonstrated in this study. Second, our results imply the necessity of grain cleaning during storage when infested by R. dominica (Flinn et al., Reference Flinn, McGaughey and Burkholder1992; Morrison et al., Reference Morrison, Bruce, Wilkins, Albin and Arthur2019). The correlation of primary pests with the intensity of negative effects on food (i.e., injury and damage to kernels (IDK), contamination by insect fragments) may be low due to their accumulation over time (Stejskal, Reference Stejskal2000; Hagstrum and Subramanyam, Reference Hagstrum and Subramanyam2006). Thus, the presence of dust and frass should be checked regularly even in the absence of primary pests, since frass may accumulate from previous infestations and may support current populations of secondary pests. Finally, strategies for the improvement of the diet of stored pests during their cultivation are also sought for experimental purposes (e.g., Beckel et al., Reference Beckel, Lorini and Lazzari2007), and frass can be a candidate component of enhanced diets for secondary feeding pests.
Acknowledgements
The work was partially funded by projects RO018 and VH20182021038. Shah J.A. was supported by a scholarship provided by the Ministry of Education of the Czech Republic via CZU. The authors thank the reviewers for their help to improve the previous versions of the MS.
