Brazil uses a large quantity of herbicides based on nicosulfuron and atrazine in the production of maize sold locally as well as for export, although restrictions are reducing or even eliminating the use of these products in certain areas. The expiration of patents has increased the number of commercial herbicide products in Brazil [13, 14, 26, 27], where crop management aims to reduce the amount of a.i. applied per hectare through the use of modern cultivation techniques, including the no-tillage system and transgenic plants [28]. The sugarcane borer, Diatraea saccharalis (F., 1794) (Crambidae), Dichomeris famulata Meyrick, 1914 (Gelechiidae), the corn earworm, Helicoverpa zea (Boddie, 1850) (Noctuidae) and S. frugiperda, are the main lepidopteran defoliators in maize crops in Brazil, and several species of Trichogrammatidae are used in the biological control of their eggs [29]. Herbicide solutions applied to the host eggs pre-parasitism can kill females of these parasitoids before or during oviposition or inhibit oviposition by repelling them. The selectivity of the mixture of the herbicides Sanson 40 SC® (a.i. nicosulfuron; dose: 30 g.ha− 1) + Gesaprim 500 Ciba-Geigy® (a.i. atrazine; dose: 1500 g.ha− 1) to the emergence rate of adults and the sex ratio of most Trichogrammatidae species evaluated is due to the low penetration rate of these products into the chorion of the host eggs (i.e., E. kuehniella), thus protecting the immature stages of these parasitoids, as reported in a no-choice test in the laboratory in Montes Claros, Minas Gerais state, Brazil [21]. The thick chorion of E. kuehniella eggs (2.63 ± 3.23 µm), with four layers, protects the parasitoids inside, whose resistance increases with their development [30]. In addition, the mixture of Sanson® (a.i. nicosulfuron; concentration: 15.0 g.L− 1) + Primóleo® (a.i. atrazine; concentration: 0.3 g.L− 1) applied to the dorsum of the generalist predator Podisus nigrispinus (Dallas, 1851) (Hemiptera: Pentatomidae) reduced its survival by about 50% in a laboratory in Diamantina, Minas Gerais state, Brazil [11]. The pesticides can penetrate into the body of this insect, causing deformities, mainly during ecdysis [31]. Pesticides decreased the testicular area and the reproductive potential of surviving Pentatomidae individuals when applied to the dorsum of these insects. The elimination rate of these chemicals, via secretion, was high, but insufficient for reducing their mortality [32]. The mixture of Sanson 40 SC® (a.i. nicosulfuron; dose: 1.5 L.ha− 1) and the insecticide Engeo Pleno® (a.i. λ-cyhalothrin + thiamethoxam; dose: 0.25 L.ha− 1) reduced the parasitism and emergence rates of T. pretiosum in treated eggs of the host E. kuehniella in a no-choice test in a laboratory in Januária, Minas Gerais state, Brazil [25]. This impact may be lower in the field, as this parasitoid hatches in periods during and outside of the planting season, but it needs alternative hosts to survive [33].
The similar parasitism rate by T. pretiosum between the two treatments shows the resistance of this species, the most distributed and important of its genus in biological control programs in the Americas, where chemicals are widely used [34]. The greater resistance of T. pretiosum is also due to its high parasitism capacity in hosts with eggs with different physical characteristics, adaptation to alternative hosts in the absence of preferred ones [35] and survival and reproduction in environments with high temperature variation [36]. In addition, resistance and parasitism by T. pretiosum are higher for well-fed females with larger body sizes [37]. The lower parasitism rate by T. acacioi, T. annulata, T. atopovirilia, T. bennetti, T. bruni, T. brasiliensis, T. demoraesi, T. galloi, and T. soaresi with the solution of nicosulfuron + atrazine shows the susceptibility of these species to this herbicide mixture. This low rate suggests female mortality pre-parasitism due to the knockdown effect or inhibition of oviposition due to herbicide repellency. This is suggested by the fact that the parasitism rate by T. atopovirilia (70.14%) was higher than that by T. pretiosum (29.23%), with the same number of adults of both species per container competing for S. frugiperda eggs, without application of chemicals in the laboratory [38]. The higher parasitism rate by T. soaresi than that of T. acacioi, in contact with the dry residues of the herbicide solutions Callisto® (a.i. mesotrione), Equip Plus® (a.i. foramsulfuron + iodosulfuron-methyl), Extrazin SC® (a.i. atrazine + simazine), Primoleo® (a.i. atrazine), Provence 750 WG® (a.i. isoxaflutole) and Siptran 500 SC® (a.i. atrazine) in a laboratory in Pelotas, Rio Grande do Sul state, Brazil, shows the resistance of the first species to these herbicides [39]. This is due to the fact that T. acacioi needs a highly favorable environment and maximum vigor of its individuals to exert parasitism because it is more susceptible to pesticides [21, 40, 41, 42] and inadapted to high temperatures [43]. In addition, parasitism by non-fed females is low, producing only males when not mated in the laboratory in Alegre, Espírito Santo state, Brazil [44]. The mixture of two a.i. herbicides can increase the efficiency of the solution for weed control and be less toxic than those with only one a.i. This was shown for the higher adult mortality rate of Trichogramma chilonis Ishii, 1941, exposed to the a.i. insecticides acetamiprid, fipronil or spinetoram (> 95%) followed by abamectin (71.3–83.5%), than with mixtures of pyraclostrobin + metiram and trifloxystrobin + tebuconazole (≤ 36.4%) in a laboratory in Peshawar, Khyber Pakhtunkhwa, Pakistan [20]. Trichogramma chilonis has good tolerance to pesticides and normal reproductive capacity, mainly because it has two types of parthenogenesis, i.e. arrhenotokous and thelytokous, producing females, males or both sexes with or without reciprocal or backcross mating [45, 46]. Pesticides applied on Spodoptera eggs can control them or inactivate their natural chemical protection against parasitism by T. chilonis [47]. The toxicity of different commercial products with the same a.i. may vary, as reported for the mortality and longevity of T. pretiosum adults exposed to maize leaves treated with different commercial products based on glyphosate (i.e. Glyfosato Atanor 48®, Gli-Up 480 SL®, Roundup Original®, Roundup Transorb®, Roundup WG®, Shadow 480 SL®, Stinger®, and Trop®) in a laboratory in Pelotas, Rio Grande do Sul state, Brazil [48]. This is due to additives, such as the first and third products having 679 and 751 g.L− 1, respectively, of “other ingredients”. The type and quantity of toxic salts and adjuvants added to herbicides can increase their toxicity to parasitoids [49]. Herbicide solutions of Alteza 240/30 SL® (a.i. glyphosate + imazetapyr), Gamit 500 EC® (a.i. clomazone), Gliz 480 SL® (a.i. glyphosate), and Roundup Ready 480 SL® (a.i. glyphosate) applied on caterpillars and pupae of S. frugiperda were slightly harmful to pupae of the parasitoid Telenomus remus Nixon, 1937 (Hymenoptera: Platygastridae) in host eggs, which was attributed to the physical protection to this parasitoid inside them in the laboratory in Rio Verde, Goiás state, Brazil [50]. The chorion of S. frugiperda eggs is thick (2.50–4.40 µm) and deposited in multiple layers (i.e., generally two or three). Natural enemies can also parasitize those in lower layers, even those covered with thicker hairs deposited by the moth as physical protection [30].
The similar emergence rate of T. pretiosum females between the two treatments agrees with that of the oviposition of this natural enemy. Trichogramma pretiosum is tolerant to herbicides and their additives and received a low impact from entomopathogens like Bacillus thuringiensis (Berliner, 1915) var. aizawai, Bacillus thuringiensis var. kurstaki (Bacillales: Bacillaceae), Baculovirus anticarsia (Baculoviridae), Beauveria bassiana (Bals.-Criv.) Vuill. (1912) (Hypocreales: Cordycipitaceae), Metarhizium anisopliae (Metchnikoff) Sorokin (1883) (Hypocreales: Clavicipitaceae), and Trichoderma harzianum Rifai (1969) (Hypocreales: Hypocreaceae) [51]. The lower emergence rate of T. soaresi, T. acacioi, T. annulata, T. atopovirilia, T. bruni, T. brasiliensis, T. demoraesi, and T. galloi females also followed a pattern similar to that of the oviposition for these species, indicating high mortality of immature stages (i.e., larva and/or pupa) of these parasitoids inside the host eggs. The emergence of Trichogrammatidae was also lower in hosts treated with herbicides, such as T. bruni, 7.3% with a Sanson 40 SC® solution (a.i. nicosulfuron; dose: 1.5 L.ha− 1) [41] and T. annulata and T. brasiliensis, 16.2% and 4.0%, respectively, with a Gesaprim 500 Ciba Geigy® solution (a.i. atrazine; dose: 150 L.ha− 1) [40], in a no-choice test in the laboratory. The lower emergence rate of Trichogrammatidae females reduces the efficiency of biological control [18, 24]. The emergence rate of T. annulata, T. bennetti, T. bruni, T. brasiliensis, T. demoraesi, T. galloi, and T. soaresi females, similar between the use of Sanson 40 SC® (a.i. nicosulfurom) and distilled water (i.e., the control) in a no-choice test, was attributed to protection of the parasitoids by the host egg chorion and/or the detoxification capacity of these immature parasitoid stages [41]. However, the highest emergence rate of females for T. acacioi, T. atopovirilia and T. pretiosum, 12.76%, from treated eggs of the alternative host E. kuehniella with Sanson 40 SC® (a.i. nicosulfurom) post-parasitism, was due to the hormesis phenomenon [41], when sub-lethal amounts of a stressor (e.g., herbicide) actually benefit an organism [18, 41, 42]. The hormesis phenomenon has also been reported for T. pretiosum, T. acacioi and T. annulata in host eggs treated with Sanson 40 SC® (a.i. nicosulfurom) and for T. pretiosum, T. demoraesi, T. galloi, and T. soaresi with Gesaprim 500 Ciba Geigy® (a.i. atrazine), with higher emergence of their females in a no-choice test [40, 41]. The higher emergence of females of the pupa parasitoid Palmistichus elaeisis Delvare & LaSalle, 1993 (Hymenoptera: Eulophidae) with the pre-treatment of the pupae of the host [i.e., the mealworm beetle, Tenebrio molitor L., 1758 (Coleoptera: Tenebrionidae)] with a Scout® solution (a.i. glyphosate; dose: 200 L.ha− 1) in a laboratory in Diamantina, Minas Gerais state, Brazil, was also attributed to hormesis [52]. This parasitoid has a Neotropical distribution, and is common in crops with reduced use of pesticides, such as sugarcane, Saccharum officinarum L. (Poales: Poaceae) and African oil palm, Elaeis guineenses Jacq (Arecales: Arecaceae), and in urban areas with high non-agricultural contaminants like domestic and industrial waste [52, 53, 54]. The reduction in oviposition and emergence rates of Trichogrammatidae females in treatment with herbicide solutions is, in general, greater in free-choice tests than in those of no choice, due to the fact that the immature stages have the physical protection of the host eggs in the latter.
The greatest impact of the herbicide mixture on the sex ratio, mainly for T. acacioi, T. atopovirilia, T. bruni, T. demoraesi, T. galloi, and T. soaresi, may have been due to the degeneration (i.e., aminoacid degradation) of both the host eggs and immature stages of these parasitoids [55]. Trichogrammatidae females are larger than males, and lay female eggs in bigger hosts and male eggs in smaller ones [56]. The larger size of the host eggs, usually containing female insects, increases the contact and absorption of a higher volume of herbicide. The reduction in the sex ratio of these parasitoids may also be due to the fact that some Trichogrammatidae species require a larger number of fertile females and that their first eggs are deposited in host eggs originating females [57]. The chemical composition of the eggshell and the liquid inside the host eggs also affects the impact of pesticides on the parasitoids inside them [30]. The sex ratio of T. annulata, T. bennetti, T. brasiliensis, and T. pretiosum, similar between the two treatments, corresponds to the fact that these species have a wide geographical distribution and are common in crops with the application of different pesticides [58], especially the first, which has better biological parameters than the other species of Trichogramma [59]. The mixture of Sanson 40 SC® (a.i. nicosulfuron) + Gesaprim 500 Ciba-Geigy® (a.i. atrazine) reduced the sex ratio of T. bennetti by 3% and that of T. galloi and T. pretiosum by 2% [21], Gesaprim 500 Ciba Geigy® alone reduced that of T. bruni by 66.7%, T. atopovirilia by 17.5% and T. bennettii by 15.0% [40], and Sanson 40 SC® alone reduced that of T. galloi by 6.0%, T. bennettii by 4.0% and T. pretiosum by 2.0% [41] in no-choice tests. The reproduction potential of parasitoids of the families Eulophidae and Trichogrammatidae is high, but it has been reduced by the application of herbicides [60, 61]. The mixture of Sanson 40 SC® (a.i. nicosulfuron) + Gesaprim 500 Ciba-Geigy® (a.i. atrazine) in no-choice tests reduced the sex ratio of T. brasiliensis by 5.49% and that of T. atopovirilia by 5.83% [21], Gesaprim 500 Ciba Geigy® (a.i. atrazine) alone reduced that of T. acacioi by − 4.8%, T. brasiliensis by − 4.4%, T. annulata and T. galloi by 0.0% and T. pretiosum by 1.0% [40], and Sanson 40 SC® (a.i. nicosulfuron) alone reduced that of T. atopovirilia by 3.9% [41].