Introduction
The Mediterranean fruit fly, Ceratitis capitata (Wiedemann, 1824) (Diptera: Tephritidae), is a major quarantine pest across the globe (Silva et al., Reference Silva, Lemos, Zucchi, Silva, Lemos and Zucchi2011). In Brazil alone, management costs added to the production and commercialisation losses due to damage by fruit flies are estimated to be around US$120 million per year (Mendes, Reference Mendes2015). Chemical control is the most frequently used management strategy, mainly using insecticides as cover sprays or toxic bait formulations (Baronio et al., Reference Baronio, Bernardi, Schutze, Baldin, Machota, Garcia and Botton2019). However, the frequent use of insecticides can result in the selection of resistant populations (Kakani et al., Reference Kakani, Zygouridis, Tsoumani, Seraphides, Zalom and Mathiopoulos2010; Arouri et al., Reference Arouri, Le Goff, Hemden, Navarro-Llopis, M'saad, Castañera, Feyereisen, Hernández-Crespo and Ortego2015) and unacceptable chemical residues on fruits. Thus, evaluating alternative fruit fly management strategies is greatly needed (Dias et al., Reference Dias, Zotti, Montoya, Carvalho and Nava2018).
Particle film technology is an alternative to conventional insecticides for controlling infestation by C. capitata (Palma et al., Reference Palma, Cicilloni, Satta, De Pau and D'Aquino2020). It is especially promising for organic fruit production (Sharma et al., Reference Sharma, Datta and Varghese2020) because it neither contaminates the environment nor leaves toxic residues in treated products (D'aquino et al., Reference D'aquino, Cocco, Ortu and Schirra2011; Lo Verde et al., Reference Lo Verde, Caleca and Lo Verde2011). Particle film technology relies on kaolin's properties (Glenn and Puterka, Reference Glenn and Puterka2005), a mineral mainly composed of aluminium silicate that, when suspended in water, rapidly forms a chemically inert and non-expanding solution with white colour and porous texture (Puterka et al., Reference Puterka, Glenn and Sekutowski2000). Abrasive mineral particles, such as kaolin, change the colour of host plants, thereby repelling pests and disrupting their feeding and oviposition (Showler, Reference Showler2002). For example, applying mineral particle films with different kaolins to guava and grape significantly reduced the oviposition of fruit fly pests Anastrepha obliqua (Macquart, 1835) and C. capitata under laboratory conditions (Costa et al., Reference Costa, Leite, Santos, Coelho, Moreira, Silva, Joachim-Bravo and Castellani2021; Da Costa et al., Reference Da Costa, Leite, Moreira, Perez-Maluf, Novaes, Santos, Santos, Joachim-Bravo and Castellani2021).
Most studies that assess the efficacy of particle films with kaolin focus on bitrophic interactions (Mazor and Erez, Reference Mazor and Erez2004; Lemoyne et al., Reference Lemoyne, Vincent, Gaul and Mackenzie2008; Leskey et al., Reference Leskey, Wright, Glenn and Puterka2010; D'aquino et al., Reference D'aquino, Cocco, Ortu and Schirra2011, Reference D'aquino, Palma, Chessa, Satta, De Pau, Inglese, Ochoa and Glenn2021; Yee, Reference Yee2012), and the extent to which the films affect natural enemies, such as predators and parasitoids, remains poorly understood. However, detailed knowledge of kaolin's lethal and sublethal effects on non-pest arthropods is needed before the mineral can be used in integrated pest management programmes. For example, Bengochea et al. (Reference Bengochea, Budia, Viñuela and Medina2014) assessed the lethal and non-lethal effects of kaolin on olive trees, the fruit fly Bactrocera oleae (Rossi, 1970), and the parasitoid Psyttalia concolor (Szèpligeti, 1910).
When searching for hosts, female parasitoids respond to chemical, visual, and mechanical stimuli (Vinson, Reference Vinson1976; Segura et al., Reference Segura, Viscarret, Paladino, Ovruski and Cladera2007; Sharma et al., Reference Sharma, Sandhi and Reddy2019). When applied to crops, kaolin particle films form a protective barrier that creates a hostile environment for insects, makes plants visually or tactually unrecognisable, and prevents the oviposition of pest insects (Glenn et al., Reference Glenn, Puterka, Vanderzwet, Stern and Feldmhake1999; Bürgel et al., Reference Bürgel, Daniel and Wyss2005), and may affect the behaviour of predators and parasitoids (Vincent et al., Reference Vincent, Hallman, Panneton and Fleurat-Lessard2003). We hypothesise that films of mineral particles and biomaterials can change the physical characteristics of grapes, and such changes will reduce oviposition by C. capitata and the parasitism of C. capitata larvae by D. longicaudata. This work aimed to assess the effects of mineral particles and biomaterial films on the tritrophic interactions of grape, the fruit fly C. capitata and the natural enemy D. longicaudata.
Material and methods
Insect rearing
The Mediterranean fruit fly used in our study was obtained from a colony maintained at the Fruit Fly Laboratory of the Universidade Estadual do Sudoeste da Bahia on the campus of Vitória da Conquista, Bahia, Brazil. Routine colony procedures included maintaining adults in wooden cages (50 × 45 × 40 cm) with two sides lined with voile fabric, one inclined for oviposition and the other for insect manipulation. Eggs laid on the side of the cage were collected daily, cleaned, transferred and maintained in plastic containers (500 ml) containing an artificial diet adapted from Zucoloto (Reference Zucoloto1987) for larval development and pupation (~10 days). Pupae were collected, placed in plastic containers (500 ml) with vermiculite, and maintained until the emergence of adults. Emerged adults were then transferred to cages for mating and oviposition with free access to water and an artificial diet (3 sugar:1 yeast extract; Silva Neto et al., Reference Silva Neto, Santos, Dias, Joachim-Bravo, Benevides, Benevides, Silva, Santos, Virgínio, Oliveira, Walder, Paranhos and Nascimento2012). The cages were maintained in a climatised room at 25 ± 2 °C, 70% relative humidity and 12L:12D photoperiod.
The D. longicaudata colony used in our study was established from parasitised C. capitata pupae obtained from the Laboratório de Entomolgia da Embrapa Mandiocultura Fruticultura Tropical (Embrapa/CNPMF). The parasitoid colony was maintained following the methodology of Carvalho et al. (Reference Carvalho, Nascimento and Matrangolo1998). Adult parasitoids were kept in acrylic cages (30 × 30 × 30 cm) with free access to water and an artificial diet made of distilled water, honey, agar-agar, ascorbic acid and nipagin (Carvalho and Nascimento, Reference Carvalho, Nascimento, Parra, Botelho, Corrêa-Ferreira and Bento2002). Third instar C. capitata were offered to adult wasp oviposition in ‘parasitism units’, which included 100 C. capitata larvae packed in organza fabric and placed on top of the acrylic cages containing parasitoids. Parasitism units were periodically exposed for 1 h to 5-day-old parasitoids. After this period, the larvae were removed from the parasitism units and maintained on a layer of fine vermiculite in plastic containers (500 ml) for pupation and, subsequently, adult emergence.
The grapes characterisation
The grape (Vitis vinifera L. ‘Italia’) used in the bioassays were obtained from fresh fruit markets and, later, selected for uniformity of maturation, size and lack of punctures by fruit flies. The biometrical and physical characteristics of the grapes (i.e. mass, length, diameter and colour) were measured before conducting the bioassays. Both colour and firmness, the latter of which requires destructive sampling, were also measured at 24 h after the initiation of the bioassays. Grape mass was determined using an analytical balance (AUY 220, Shimadzu) with a precision of 0.1 mg. The diameter and length of the grapes were measured using a digital pachymeter (Model MPD-200; Metrotools, São Paulo, Brazil) with a precision of ± 0.02 mm. Fruit firmness for each grape (n = 20) was determined after the treatment application using a penetrometer (model WA68; TR, Italy) equipped with an 8 mm-diameter tip, and results were expressed in Newton (N). Luminosity (L*), a* and b* (from white to black, green to red and blue to yellow, respectively) colour coordinates were determined using a Minolta colorimeter CR 400 (Minolta, Osaka, Japan). The colour of the grapes was measured twice, before and after the treatment, always in the same position (opposite sides), using four fruits per treatment. The chroma, C* = (a2 + b2)1/2, and the hue angle, arc tangent of (b/a), were also calculated (Lemoyne et al., Reference Lemoyne, Vincent, Gaul and Mackenzie2008).
Oviposition bioassay
The oviposition experiments were performed in a completely randomised design with eight treatments and four repetitions. A total of 12 grape per treatment were used, which were distributed across four replicates over 3 days (N = 96). The treatments included Surround WP (NovaSource, Phoenix, AZ, USA), kaolin 605 white (BrasilMinas, Guarulhos, SP, Brazil), kaolin 607 cream (BrasilMinas), kaolin 608 white (BrasilMinas), kaolin 611 grey (BrasilMinas), cassava starch, potato starch and control (distilled water). The particles were dispersed in distilled water (200 g l−1) with guar gum (~5 g l−1) to improve formulation viscosity and stability (Campos et al., Reference Campos, Oliveira, Fraceto and Singh2015; Gao et al., Reference Gao, Guo, Li, Li, Wu, Hu, Hu, Mo and Wu2020; Costa et al., Reference Costa, Leite, Santos, Coelho, Moreira, Silva, Joachim-Bravo and Castellani2021; Da Costa et al., Reference Da Costa, Leite, Moreira, Perez-Maluf, Novaes, Santos, Santos, Joachim-Bravo and Castellani2021). The kaolin and biomaterial concentrations used in our treatments were based on previous studies (Costa et al., Reference Costa, Leite, Santos, Coelho, Moreira, Silva, Joachim-Bravo and Castellani2021; Da Costa et al., Reference Da Costa, Leite, Moreira, Perez-Maluf, Novaes, Santos, Santos, Joachim-Bravo and Castellani2021). The biomaterial particles, cassava and potato starches, were obtained from a natural products market in Indianapolis (SP, Brazil). Before starting the bioassays, the grapes were sanitised for 30 min in sodium hypochlorite (0.5%) and then individually immersed for 10 s in a beaker that contained 60 ml of the corresponding treatment solution. After immersion, the grapes were dried at 25 ± 2 °C for 1 h.
The plot comprised a field cage (2 × 2 × 2 m), supported by a metal structure covered by a nylon fabric, containing a potted Spondias tuberosa L. plant with ~1.20 m height and radius canopy around 30 cm. Eight treated grapes, corresponding to one grape per treatment, were hung on top of the field cage, spaced 33 cm apart, and then exposed to 50 pairs of 7-day-old C. capitata for 48 h. After exposure, each grape was dissected to count the total number of punctures with eggs, punctures without eggs and eggs. During the bioassay, the cage conditions were maintained at a temperature of 27.08 ± 1.5 °C (min and max of 13.6 and 37.4 °C, respectively), relative humidity of 51.6 ± 5.85 (min and max of 29.8 and 78.8%, respectively) and luminosity of 19.894 lux.
Parasitism bioassay
The parasitism of D. longicaudata on C. capitata larvae was evaluated using choice tests in a completely randomised design with eight treatments and four repetitions. A total of 12 grapes per treatment were used, which were distributed across four replicates over 3 days (N = 96). The treatments were the same as those used in the oviposition bioassay. Before starting the bioassays, the grapes were sanitised for 30 min in sodium hypochlorite (0.5%) and then individually immersed for 10 s in the suspensions. After drying at room temperature, the treated grapes were artificially infested with C. capitata larvae using a methodology adapted from Pires et al. (Reference Pires, Sant'Ana and Redaelli2021). Briefly, the grapes were perforated to a depth of 1.5 cm using a 1.5 mm-diameter needle, and any pulp residue formed during the penetration was removed to prevent orifice obstruction. Two third-instar C. capitata larvae were then inserted into the orifice of each grape using a fine-tipped brush tool, and the orifice was closed using a small cotton ball. After 1 h, the grapes were finally exposed to the parasitoids.
Similar to the oviposition bioassay, the parasitism bioassays were performed in field cages (2 × 2 × 2 m), each containing a potted plant. For each bioassay, eight artificially infested grapes were treated with particle suspensions or water and arranged as previously described for the oviposition bioassay. Then, 50 pairs of 5-day-old D. longicaudata were released into the field cage. The grapes were removed after 2 h of parasitoid exposure. In the lab, the larvae exposed to the parasitoids were removed from the grapes and kept in plastic containers containing a vermiculite layer until adult emergence. The numbers of emerging parasitoids and flies, larval viability (VL% = no. parasitoid pupae × 100/total fly larvae), pupal viability (VP% = no. emerged parasitoids + no. emerged flies × 100/total fly pupae) and parasitism index (IP% = no. emerged parasitoids × 100/no. emerged flies thinsp;+ no. emerged parasitoids) were determined (Matrangolo et al., Reference Matrangolo, Nascimento, Carvalho, Melo and Jesus1998).
The bioassays were performed at a temperature of 22 ± 1.5 °C (min and max of 17.1 and 33.9 °C, respectively), relative humidity of 51 ± 8.5% (min and max of 51 and 80%, respectively) and luminosity of 13.586 lux measured during set up of the bioassay (8:00 am).
Statistical analyses
The homoscedasticity and normality of the data for the biometrical and physical characteristics of the grapes and oviposition of C. capitata and D. longicaudata were evaluated using Bartlett and Shapiro–Wilk tests, respectively. Datasets that violated these assumptions (e.g. the number of punctures with and without eggs, number of eggs and number of parasitoids and flies) were square root-transformed and, subsequently, analysed using a generalised linear model (GLM). The GLMs were established using the nlme (Pinheiro et al., Reference Pinheiro, Bates, DebRoy and Sarkar2020) and lsmeans (Lenth, Reference Lenth2016) packages in R. Paired t-tests were used to compare the mean values of pre- and post-treatment L*, C* and hue angle. All analyses were performed using R software (version 3.6.1; R Development Core Team, 2019).
Results
Effect on grapes characteristics
The grapes used for the treatment groups exhibited no significant differences regarding mass (F = 0.22303; df = 7.31; P = 0.97605), length (F = 0.6665; df = 7.31; P = 0.70095), diameter (F = 0.20034; df = 7.31; P = 0.9823), L* (F = 1.0555; df = 7.31; P = 0.42077), C* (F = 1.1042; df = 7.31; P = 0.39) and hue angle (F = 0.5303; df = 7.31; P = 0.80286; Table 1).
The biometrical and physical characteristics of the grapes before application of the suspensions
Means in the same column followed by the same letter do not differ significantly (Tukey test, P ≤ 0.05).
The immersion of grapes in mineral and biomaterial suspensions affected both L* (t = −11,795; df = 31; P < 0.0001) and C* (t = 7.9406; df = 31; P < 0.0001), and the different immersion treatments resulted in significantly different L* (F = 1258.1; df = 7.31; P < 0.0001), C* (F = 183.69; df = 7.31; P < 0.0001) and hue angle (F = 188.71; df = 7.31; P < 0.0001; Table 2). Kaolin and starch treatments increased grape luminosity, with the highest values observed in grapes treated with Surround WP and kaolin 605 and the lowest for control grapes. In contrast, C* values were consistently decreased by immersion in the suspensions, and the hue angle was lower in the Surround WP- and kaolin 605-treated grapes than in the control group. All the mineral films and starches increased fruit firmness (F = 28.554; df = 7.31; P < 0.0001).
The physical characteristics of the after application of the suspensions
Means in the same column followed by different letters differ significantly (Tukey test, P ≤ 0.05).
Effect on C. capitata oviposition
Treatment had no effect on the number of punctures without eggs (AIC = 20.63; df = 31) but did significantly affect the number of punctures with eggs (AIC = 29.58; df = 31) and the number of eggs (AIC = 94.31; df = 31; Table 3). Briefly, both kaolin and cassava starch reduced the number of egg punctures, with fewer eggs in fruits treated with Surround WP, kaolin 605 and kaolin 608. In contrast, the treatment with potato starch yielded the highest mean egg number (3.18 ± 0.46).
Parameter estimates for the number of punctures with and without eggs, and the number of eggs (mean ± SE) laid by C. capitata in grapes exposed to D. longicaudata in field cage conditions
Means in the same column followed by different letters differ significantly (Tukey test, P ≤ 0.05).
Effect on D. longicaudata parasitism
One hundred and fifty-seven (157) of the 172 puparia yielded adult insects (69 fruit flies, 88 parasitoids), with larval and pupal viabilities of 89.6 and 91.3%, respectively. The parasitism index was 56%, ranging from 30% in the potato starch treatment to 69.6% in the control. Treatments did not affect the numbers of parasitoids (AIC = 42.35, df = 31) or flies (AIC = 35.78; df = 31; Table 4).
Poisson generalised linear model parameter estimates for the numbers of parasitoids and flies (mean ± SE), and the parasitism index in grapes exposed to D. longicaudata in field cage conditions
Means in the same column followed by the same letter do not differ significantly (Tukey test, P ≤ 0.05).
Discussion
The evaluation of the grapes before their use in the bioassays indicated excellent fruit uniformity, preventing the potential influence of fruit characteristics on the results, as suggested by Da Costa et al. (Reference Da Costa, Leite, Moreira, Perez-Maluf, Novaes, Santos, Santos, Joachim-Bravo and Castellani2021). Applying mineral and biomaterial films to the grapes did not affect the number of punctures without eggs, confirming the laboratory-based findings of Da Costa et al. (Reference Da Costa, Leite, Moreira, Perez-Maluf, Novaes, Santos, Santos, Joachim-Bravo and Castellani2021). The mechanical resistance provided by the films might have discouraged flies from ovipositing in the fruit after puncturing it. However, films should ideally inhibit both oviposition and fruit puncturing since puncture injuries in some fruits (e.g. apples) can facilitate the entry of fungi and bacteria (Santos et al., Reference Santos, Corrent, Berton, Schwarz and Denardi2008).
The kaolin and cassava starch treatments reduced the number of eggs, with fewer eggs laid in grapes treated with Surround WP, kaolin 605 and kaolin 608, corroborating previous findings (Costa et al., Reference Costa, Leite, Santos, Coelho, Moreira, Silva, Joachim-Bravo and Castellani2021; Da Costa et al., Reference Da Costa, Leite, Moreira, Perez-Maluf, Novaes, Santos, Santos, Joachim-Bravo and Castellani2021). In previously reported laboratory studies, kaolin reduced fruit fly oviposition in bitrophic interactions of grape × C. capitata (Da Costa et al., Reference Da Costa, Leite, Moreira, Perez-Maluf, Novaes, Santos, Santos, Joachim-Bravo and Castellani2021) and guava (Psidium guajava L.) × A. obliqua (Costa et al., Reference Costa, Leite, Santos, Coelho, Moreira, Silva, Joachim-Bravo and Castellani2021), and the number of punctures in apple (Malus domestica L.) × C. capitata (Leskey et al., Reference Leskey, Wright, Glenn and Puterka2010; Ourique et al., Reference Ourique, Redaelli, Efrom and Pedrini2017), mango (Mangifera indica L.) × C. capitata (Ourique et al., Reference Ourique, Redaelli, Efrom and Pedrini2017) and citrus × C. capitata (D'aquino et al., Reference D'aquino, Cocco, Ortu and Schirra2011). Kaolin has also been reported to reduce fruit fly oviposition in field studies of citrus × C. capitata (Lo Verde et al., Reference Lo Verde, Caleca and Lo Verde2011), apple × Rhagoletis pomonella (Walsh) (Villanueva and Walgenbach, Reference Villanueva and Walgenbach2007), cactus (Opuntia ficus-indica ‘Gialla’) × C. capitata (D'aquino et al., Reference D'aquino, Palma, Chessa, Satta, De Pau, Inglese, Ochoa and Glenn2021) and the number of punctures in citrus × C. capitata (Braham et al., Reference Braham, Pasqualini and Ncira2007). In contrast, the biomaterials failed to protect the fruits from oviposition, and the potato starch treatment yielded the highest mean egg number, appearing to stimulate oviposition. These findings agreed with previous laboratory-based studies (Da Costa et al., Reference Da Costa, Leite, Moreira, Perez-Maluf, Novaes, Santos, Santos, Joachim-Bravo and Castellani2021). However, it is important to note that potato starch was reported to preserve guava peel colour and to protect guava fruits from oviposition by A. obliqua (Costa et al., Reference Costa, Leite, Santos, Coelho, Moreira, Silva, Joachim-Bravo and Castellani2021).
The reduced oviposition of C. capitata in kaolin-coated grapes was likely due to changes in fruit colour and firmness. More specifically, it is possible that the effects of the white mineral particles on the grape peels’ natural green colour interfered with host identification by C. capitata females. Indeed, some studies have demonstrated that fruits or spheres coated with white substances experience reduced fruit fly oviposition (Cytrynowicz et al., Reference Cytrynowicz, Morgante and Souza1982; Katsoyannos et al., Reference Katsoyannos, Panagiotidou and Kechagia1986; López-Guillén et al., Reference Lopéz-Guillén, Valle-Mora, Cazares and Rojas2009; Costa et al., Reference Costa, Leite, Santos, Coelho, Moreira, Silva, Joachim-Bravo and Castellani2021; Da Costa et al., Reference Da Costa, Leite, Moreira, Perez-Maluf, Novaes, Santos, Santos, Joachim-Bravo and Castellani2021). The high reflectance of white surfaces is visually less attractive to fruit flies, as demonstrated in C. capitata (Nakagawa et al., Reference Nakagawa, Prokopy, Wong, Ziegler, Mitchell, Urago and Harris1978; Katsoyannos et al., Reference Katsoyannos, Panagiotidou and Kechagia1986), Bactrocera dorsalis (Hendel, 1912) (Wu et al., Reference Wu, Chen and Yang2007) and A. obliqua (López-Guillén et al., Reference Lopéz-Guillén, Valle-Mora, Cazares and Rojas2009). In addition, the films formed a physical barrier that affected fruit firmness. The epicarp of some fruits provides natural resistance that prevents some species of flies with short aculeus from puncturing and laying eggs (Aluja and Mangan, Reference Aluja and Mangan2008). For instance, C. capitata preferred to oviposit in fruits with an advanced maturation stage and, consequently, lower firmness (Gómez et al., Reference Gómez, Paranhos, Silva, Lima, Silva, Macedo, Virginio and Walder2019). Mineral particles also roughen the fruit's surface, making it inadequate for oviposition (Saour and Makee, Reference Saour and Makee2004). According to Salerno et al. (Reference Salerno, Rebora, Kovalev, Gorb and Gorb2020), the kaolin particle film when applied to natural and artificial substrates, it reduces the insect's ability to attach.
The effect of mineral particles and biomaterial films on grape colour did not affect the parasitism capacity of D. longicaudata. As such, parasitoid females could locate C. capitata larvae in all treatments (Table 4). Messing and Jang (Reference Messing and Jang1992) reported that D. longicaudata females responded to fewer visual stimuli than males since olfactory stimuli (e.g. larvae kairomones) play a more critical role in host localisation (Carrasco et al., Reference Carrasco, Montoya, Cruz-Lopez and Rojas2005). Benelli and Canale (Reference Benelli and Canale2012) reported that naive P. concolor females show no colour preferences. These findings agree with those of Bengochea et al. (Reference Bengochea, Hernando, Saelices, Adán, Budia, González-Núñez, Viñuela and Medina2010), who investigated the effectiveness of kaolin against B. oleae in olive groves and its effect on the parasitoid P. concolor, indicating that the parasitism capacity of P. concolor was unaffected by kaolin treatment. Additional laboratory and semi-field studies have also reported that kaolin is harmless to the fruit fly parasitoid P. concolor (Adán et al., Reference Adán, González, Bastante, Budia, Medina, Del Estal and Viñuela2007; Bengochea et al., Reference Bengochea, Hernando, Saelices, Adán, Budia, González-Núñez, Viñuela and Medina2010). According to Bengochea et al. (Reference Bengochea, Hernando, Saelices, Adán, Budia, González-Núñez, Viñuela and Medina2010), using kaolin in olive crops is promising because it affects beneficial arthropods to a lesser extent than other commonly used compounds, such as dimethoate. However, these findings also contradict those of Bengochea et al. (Reference Bengochea, Budia, Viñuela and Medina2014), who reported that kaolin treatment reduced the rate of parasitism by P. concolor.
Together, our findings support the conclusion that even though mineral films do not entirely prevent damage by fruit flies, they do not interfere with the parasitism of C. capitata by D. longicaudata. Thus, mineral film application may be considered in integrated pest management schemes with a high potential to be successful. The ability of the parasitoid to locate and parasitise C. capitata larvae in kaolin-coated fruits suggests that kaolin films could be used in conjunction with biological agents to control fruit fly pests in organic agriculture operations. Nevertheless, new studies conducted under commercial conditions are needed to support our recommendations fully.
Acknowledgements
We thank the Foundation for Research Support of the State of Bahia (FAPESB) and the CAPES Foundation (Brazilian Ministry of Education; Finance Code 001) for the financial support. We are grateful to the companies, Brasilminas and NovaSource, that donated the kaolin samples.
Competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this article.
