Introduction
Conservative agriculture seeks efficient agricultural production through the optimisation of resource use, implementation of precision technologies, and integrated crop management, and has proven to be an effective strategy for the sustainability of agricultural production (Pretty et al., Reference Pretty, Benton, Bharucha, Dicks, Flora, Godfray, Goulson, Hartley, Lampkin, Morris, Pierzynski, Prasad, Reganold, Rockström, Smith, Thorne and Wratten2018). This approach promotes crop diversification and the adoption of agroecological practices, such as agroforestry and green manuring, significantly contributing to food security and the resilience of production systems (Altieri, Reference Altieri2018). In this mode, no-till farming presents itself as an essential tool that aligns with the conservationist practices of agricultural production (Vezzani and Mielniczuk, Reference Vezzani and Mielniczuk2011).
In no-till systems, the remaining plant cover on the soil surface provides shelter and food for natural enemies of pests, such as predators and parasitoids, aiding in the control of pest populations (Adhikari et al., Reference Adhikari, Adhikari, Weaver, Bekkerman and Menalled2019). However, some pests, such as stink bugs, may benefit from this environment (Chocorosqui and Panizzi, Reference Chocorosqui and Panizzi2004). Crops like cotton can benefit from the presence of plant residues, which in turn, makes it difficult for some species of pests to infest (Varela et al., Reference Varela, Savinelli, Mccarthy, Norton and Leonard2013).
The applicability of the NTS in cotton farming assumes a strategic mission in the long-term sustainability of agricultural production (Delaune et al., Reference Delaune, Mubvumba, Alea and Kimura2020). No-till farming associated with integrated pest management (IPM) presents itself as an effective strategy to control the spread of harmful organisms to cotton crops (Gossypium hirsutum L.). This approach incorporates varied tactics, such as the diversity of cover crops, crop rotation (Busch et al., Reference Busch, Douglas, Malcolm, Karsten and Tooker2020), the use of natural enemies, and the judicious use of pesticides, aiming to minimise environmental impacts and pest resistance (Jaworski et al., Reference Jaworski, Xiao, Q, Romero, Guo, Wang and Desneux2019).
The variability in insect population responses in different crops is evident, as evidenced in studies on corn (Zea mays L.) and brachiaria (Urochloa ruziziensis Germain & Evrard), where plant cover can affect the larval activity of lepidopteran pests, such as Spodoptera frugiperda (Smith, 1797) (Lepidoptera: Noctuidae) (DIAS et al., Reference Dias, Marucci, Mendes, Moreira, Araújo, Santos and Barbosa2016), for example. According to that context, the present study selected plants such as millet (P. glaucum L.), corn (Z. mays L.), brachiaria (U. ruziziensis), black velvet bean (S. aterrimum), forage sorghum (S. bicolor L.), and white oat (A. sativa I.) to compose the plant covers, not only for their compatibility with the NTS in the cotton-producing regions but also for the potential capacity of these plants to interfere in the population dynamics of pests, either disadvantaging or benefiting the productive system as a whole.
Considering the scarcity of studies dedicated to the influence of cover crops on the abundance of pest insects, especially in contexts where cotton is the main crop, the NTS stands out as an essential tool for sustainable management, aligned with the principles of IPM. Given the circumstances, the present research had as its general objective to evaluate the interference of different cover crops, both living and decomposing, in the population dynamics of pest insects in cover crops and in the cotton crop (G. hirsutum L.), as well as, to analyse the diversity indices to provide insights about the behaviour of phytophagous insect species in different crops, that kind of metric provides valuable information about the rarity and commonness of species in an agroecosystem, and reflect the many different species in the collected dataset.
Materials and methods
Location
The experiments were conducted at the Experimental Farm of Agricultural Sciences (EFAC) at the Federal University of Grande Dourados (UFGD), located in the municipal district of Dourados (Mato Grosso do Sul State, Brazil), under the geographical coordinates 22o12′ S and 54o56′ W, at an altitude of 452 m. The climate, classified as Am (Alvares et al., Reference Alvares, Stapes, Sentelhas, Gonçalves and Sparovek2013) according to the Köppen classification, is characterised as humid mesothermal with a rainy summer. The soil of the experimental area is classified as very clayey Dystroferric Red Latosol. The present study was conducted in two identical experimental fields, implemented in the same plots from May to March, in the 2020/2021 and 2021/2022 cropping seasons.
Design
The experiment consisted of eight treatments, composed of different cover species for the Direct Planting of the cotton crop, notably those with the highest potential for use in the Midwest region, namely: (1) millet (Pennisetum glaucum L.), (2) corn (Zea mays L.), (3) brachiaria (Urochloa ruziziensis), (4) black velvet bean (Stizolobium aterrimum), (5) forage sorghum (Sorghum bicolor L.), (6) white oat (Avena sativa L.), (7) mix in the proportion of (92:8) between white oat + brachiaria (A. sativa + U. ruziziensis), and (8) fallow.
The experimental design adopted was randomised blocks (RBD) with four blocks. The experimental units were organised into homogeneous blocks and then the treatments are assigned at random to these units within these blocks. Each experimental unit (plot) had dimensions of 10 m in length by 10 m in width, totaling 100 m2. The total experimental area was 3200 m2, containing 32 plots with a spacing of 2 m between the plots, resulting in a total area of 4608 m2. The preparation of the experimental area included conventional plowing and levelling harrowing for the installation of the cover plants that preceded the cultivation of cotton.
Inputs and equipment
Cover crops were sown in the first half of September, with the exception of white oats, brachiaria, and the mix of white oats + brachiaria, which were sown in the second half of May. A mechanical seeder (SEMEATO® – SHM 15/17ROTO) was used for the installation of the covers. Plant densities were 13, 3, 10, 15, and 50 per linear metre for millet, corn, black velvet bean, forage sorghum, and white oats, respectively, with a spacing of 0.50 m for all covers, except for white oats, which had a row spacing of 0.17 m (Aguiar et al., Reference Aguiar, Gonçalves, Ayres, Paterniani, Tucci and Castro2014).
Brachiaria density was approximately four plants per metre (De Maria et al., Reference De Maria, Di Trochio, Piedade and Duarte2012). In the case of the mix of oats + brachiaria, the quantities of plants per linear metre and the spacing were the same as for oats and brachiaria, as previously mentioned, only proportionally mixed to achieve the desired plant population. The sowing of cotton (FM 944GL®) was carried out with a vacuum seeder (JUMIL® - JM EXACTA 2680 PD), in the second half of November, at least 20 days after the desiccation of the cover plants, carried out during the flowering of each crop, with the application of 1440 g i.a. ha-1 of glyphosate. The spacing was 0.90 m, and the density was approximately nine plants per metre, according to Lamas et al. (Reference Lamas, Ferreira, La Torre and Staut2016).
All cultural treatments, including seed treatment and/or inoculation and the use of herbicides, were carried out following good agricultural practices, with products registered with the Ministry of Agriculture, Livestock and Supply – AGROFIT. No pest controls were carried out; only soil analyses were carried out before the installation of the cover crops, with subsequent correction, along with base and top dressing fertilisation, only for the cotton crop, according to the needs indicated in Bulletin 200 of the Agronomic Institute of Campinas (Aguiar et al., Reference Aguiar, Gonçalves, Ayres, Paterniani, Tucci and Castro2014).
The rainfall index was recorded through an acrylic cup-type rain gauge, installed at the trial site, while temperatures were obtained from the Agro Meteorological Station of Embrapa Agropecuária Oeste – Dourados-MS, located approximately 18 km from the experiment. When the precipitations were not sufficient to meet the pluviometric needs of the installed crops, the study was uniformly irrigated.
Attributes/variables evaluated
Evaluations were conducted at each seven-day interval in each plot, starting at the stage of total emergence of the cover plants and cotton. These evaluations comprised the installation of a rebar cube with dimensions of 1 m on each side, thus forming a three-dimensional structure of 1 m3. Inside, this structure delimited a cage covered by fabric mesh. In this cage, the insects present were later captured through a device consisting of a domestic vacuum cleaner coupled to a transparent plastic container. This container included a small mesh bag, in which the insects were collected during the suction process. This limited the number of plants sampled in the area, which were inspected and quantified for the population of pest insects present in the useful area of each plot, following the method similar to that used by Silvie et al. (Reference Silvie, Belot, Martin, Seguy, Bouzinac, Silva and Marques2005), which used a wooden frame (1 × 1 m), generating 1 m2. After this process, the insects were captured and stored in containers containing 70% alcohol. Subsequently, they were identified and catalogued according to the methods of Soria and Degrande (Reference Soria and Degrande2011).
Data analysis
To compare the population dynamics of insects among treatments over time and within each sampling interval, we chose the samples randomly to promote their independence. Thus, a generalised linear model was programmed to analyse the effect of each treatment on the population dynamics of insects over time. For this, the goodness of fit of the models for count data without overdispersion was tested, using the Poisson model and with overdispersion (negative binomial). The over-dispersion test was conducted with the help of the overdose package (Souza et al., Reference Souza, Favero, Belfiore, Correa, Cameron and Trivedi2020), available in the R program (R Core Team, 2020).
The effect of the treatment factor was analysed through deviance analysis (difference of deviations). Depending on the result of the deviance analysis, the abundance of insects was compared among treatments within each condition (before or after cotton cultivation) through contrasts generated by the glht function of the multcomp package (Hothorn et al., Reference Hothorn, Bretz and Westfall2008). We used functions from the vegan package for diversity analysis. We estimated the following indices: Shannon's H’; species richness and Pielou's evenness.
Results
Pest population dynamics in the first cropping season: winter/summer 2020/2021 cover crops
In the first cropping season evaluated (BC/Winter/Summer 2020/2021), before the sowing of cotton, the presence of the caterpillar S. frugiperda was mainly detected in corn, with significant intensity, and low infestation in sorghum and brachiaria. There was no significant difference between sorghum and brachiaria (F = 6.23 df = 1, P > 0.05) (table 1).
Table 1. Average number (±s.e.) of insect pests on cover crops before and after cotton sowing, in the first cropping season 2020/2021 (winter/summer). Dourados-MS, Brazil. 2022
BC, Before sown cotton; AC, After sown cotton; s.e., Standard error.
Capital letters compare columns. *There was no variability in these treatments, therefore, it was not considered in the contrast analysis.
The species Dalbulus maidis (DeLong & Wolcott, 1923) (Hemiptera: Cicadellidae) was more abundant in fallow and brachiaria areas, with no significant difference between these two treatments (F = 6.9, df = 7, P > 0.05). There was no significant difference in D. maidis infestation between fallow, corn, and mix, while millet and oats differentiated themselves by presenting lower levels of infestation (table 1).
When comparing all treatments, fallow stood out significantly (F = 6.9, df = 7, P < 0.05) for the highest abundance of Diabrotica speciosa (Germar, 1824) (Coleoptera: Chrysomelidae). However, most crops did not differ significantly from each other, although black velvet bean and sorghum stood out for presenting lower levels of D. speciosa infestation (table 1).
Considering stink bugs individually, Piezodorus guildinii (Westwood, 1837) (Hemiptera, Pentatomidae) and Euschistus heros (Fabricius, 1798) (Hemiptera: Pentatomidae), no significant differences (F = 3.5, df = 7, P > 0.05) were observed in relation to their population abundances for the plant species studied (table 1). The oat crop expressed a larger number of Lagria villosa (Fabricius, 1783) (Coleoptera: Lagriidae) compared to black velvet bean. However, there was no significant difference between these treatments and in relation to the others (F = 1.63, df = 7, P > 0.05) (table 1).
Pest population dynamics in the first cropping season: winter/summer 2020/2021 cotton crop
Still in the first cropping season, after the sowing of cotton (AC/Winter/Summer 2020/2021), polyphagous species specific to the cotton crop emerged. The brown stink bug E. heroes, Spodoptera eridania (Cramer, 1782) (Lepidoptera: Noctuidae), and Alabama argillacea (Hübner, 1818) (Lepidoptera: Noctuidae) occurred, but they did not show significant differences between treatments (F = 0.2, df = 7, P > 0.05) (table 1).
As for the cotton aphid Aphis gossypii (Glover, 1877) (Hemiptera: Aphididae), millet presented the highest infestation, while black velvet bean and corn showed lower infestation. Oats, sorghum, and brachiaria presented high infestation, however, below the highest concentration of aphid, with no significant difference between them (F = 17.93, df = 6, P > 0.05) (table 1).
Considering that D. speciosa is a polyphagous pest, a high variability was observed between treatments before and after the sowing of cotton. D. speciosa expressed the highest abundance in all treatments, except in sorghum, differentiating significantly from the others (F = 1.38, df = 7, P < 0.05) during the period when cotton was present (table 1).
When comparing pests in all types of cultivation with the infestation of Frankliniella schultzei (Trybom, 1910) (Thysanoptera: Thripidae), the difference in the high population abundance of thrips in the treatments (table 1) became evident. Sorghum and mix stood out among the highest average values, followed by oats. While millet, black velvet bean, and fallow presented significantly equal levels of infestation; likewise, they did not differ from the corn, brachiaria, and fallow treatments (F = 29.61, df = 7, P > 0.05) (table 1).
Pest population dynamics in the second cropping season: winter/summer 2021/2022 cover crops
In the evaluation of the second cropping season that preceded the sowing of cotton (BC/Winter/Summer 2021/2022), the presence of D. maidis occurred equally in the treatments, except in sorghum, which only differentiated from the mix treatment by presenting a higher number of the species in question (table 2). In the fallow and brachiaria treatments, higher levels of D. speciosa infestation were observed, however, brachiaria was also significantly equal to sorghum, millet, and corn, which did not differ from the mix treatment (F = 2.04, df = 5, P > 0.05) (table 2).
Table 2. Average number (±s.e.) of insect pests on cover crops before and after cotton sowing, in the second cropping season 2021/2022 (winter/summer). Dourados-MS, Brazil. 2022
BC, Before sown cotton; AC, After sown cotton; s.e., Standard error.
Capital letters compare columns. *There was no variability in these treatments, therefore, it was not considered in the contrast analysis.
# Data not analysed due to non-response to all treatments, except fallow in the Acrosternum heegeri column and black velvet bean in the Lagria villosa column. SE: Standard error.
When comparing the treatments where the appearance of the stink bug E. heros occurred, it was found that in the plots of fallow and brachiaria, there was the highest population density of this insect, followed by the mix which, in turn, was significantly equal to the other treatments (F = 5.63, df = 5, P > 0.05) (table 2). In relation to the infestation of P. guildini, there was a low variation among the treatments analysed, with only the fallow treatment having a higher incidence, while the other plots remained without significant difference (F = 1.83, df = 5, P > 0.05) and with lower infestation of this stink bug (table 2).
The caterpillar S. frugiperda expressed its highest infestation in the corn crop, standing out for the highest number of individuals compared to sorghum and millet, respectively. There was a statistical difference between the treatments (F = 56.58, df = 5, P < 0.05) (table 2). In the case of L. villosa, samples of this beetle indicated a low concentration and variability between the treatments, mainly because it occurred with significantly equal average values (F = 1.02, df = 7, P > 0.05) and in only two (oats and corn) of all the treatments evaluated (table 2).
Pest population dynamics in the second cropping season: winter/summer 2021/22 cotton crop
Considering the samplings carried out in the second cropping season after the establishment of the cotton crop (AC/Winter/Summer 2021/2022), the pests A. argillacea and F. schultzei demonstrated higher population abundance compared to most pests, with no significant differences between all treatments for these two species (F = 0.67, df = 7, P > 0.05) (table 2).
Aphis gossypii presented considerable average values, standing out in treatments with lower infestation, in combined crops, equal to the mix, with the exception of oats, which expressed a higher incidence of individuals. However, it remained similar (F = 1.67, df = 7, P > 0.05) to the treatments with brachiaria, fallow, sorghum, and corn (table 2).
As for the corn and oat crops, the infestation by D. speciosa differed significantly (F = 6.49, df = 7, P < 0.05), while brachiaria and fallow did not differentiate from any of the treatments (table 2).
Analysis of the diversity of pests and natural enemies found in different cover crops across two cropping seasons
The analysis of the diversity indices obtained in the survey of phytophagous insects revealed a heterogeneous pattern in relation to the treatments, considering the average of the two cropping seasons evaluated (BC/Winter/Summer 2020/2021 and 2021/2022). Although the variable referring to species richness presented a clear separation into two distinct groups: one group with a high index, containing the crops of brachiaria, sorghum, corn, and millet, and another group with a low index, composed of the crops of oats, black velvet bean, mix, and fallow (fig. 1).
Figure 1. Population fluctuation of phytophagous arthropods, in cover plants evaluated under diversity indices before the sowing of cotton in two consecutive cropping seasons (BC/Winter/Summer 2020/2021 and 2021/2022).
When considering the abundance of individuals, the treatment with sorghum had the highest Shannon–Weaver index, while the black velvet bean culture had the lowest concentration of pests, followed by oats. The sorghum crop stood out for presenting the highest indices in all evaluated items, including the variable uniformity of species averages, where it was found close to the fallow condition. On the other hand, millet stood out from the others for the highest non-uniformity of species averages (fig. 1).
In the diversity indices of the collected data on phytophagous insects, referring to the average of two seasons after the sowing of cotton culture (AC/Winter/Summer of 2020/2021 and 2021/2022), the treatment corresponding to the cultivation of sorghum during this period presented a diversity significantly different from the previous phase of the main culture establishment (BC). The levels of pest concentration were the lowest recorded, except for species richness, where all treatments expressed equally high values, as shown (figs 1 and 2).
Figure 2. Population fluctuation of phytophagous arthropods, in cover plants evaluated under diversity indices after the sowing of cotton in two consecutive cropping season (AC/Winter/Summer 2020/2021 and 2021/2022).
Brachiaria and millet presented similar averages, with high values in relation to the abundance and intrinsic uniformity of the species. All treatments expressed similar dispositions in the figures referring to the Shannon–Weaver indices and Pielou's Evenness, with different averages in absolute values (fig. 2).
Generally, an intense activity of pests was verified, mainly during the period in which the crops remained alive. This fact occurred in the two crops, both in the covers and in the main crop plants. The movement of herbivores also affected their natural enemies, resulting in variations in the quantity and number of arthropod species observed in the trial (fig. 3).
Figure 3. Population fluctuation of natural enemies, in cover plants evaluated under diversity indices before the sowing of cotton in two consecutive cropping seasons (BC/Winter/Summer 2020/2021 and 2021/2022).
When interpreting the data corresponding to the Shannon–Weaver index in the phase (BC/Winter/Summer 2020/2021 and 2021/2022), it was verified that the mix demonstrated the highest concentration of predators, even being composed of brachiaria with 50% of oats, which in contrast exhibited the lowest rate in the population abundance of natural enemies. Millet, fallow, and sorghum exhibited proximities in their averages just below the first place, while corn and black velvet bean were distributed separately in the figure with the index around 1.5 and 1.4, respectively (fig. 3).
Regarding the balance between the averages, most treatments presented indices between 0.80 and 0.90, with the exception of oats, which remained at a lower rate. However, when observing the richness of species, there was a categorical inversion of oats, revealing the highest indices together with the treatments of millet, mix, fallow, sorghum, and brachiaria. Such as, they differ from corn and, especially, from black velvet bean, which stood out with the lowest level among all treatments (fig. 3).
The treatments subjected to cover with the corn crop in the period when the main crop was already installed (AC/Winter/Summer 2020/2021 and 2021/2022) expressed the highest indices in the three highlighted parameters (fig. 4), even though they were accompanied by the covers of oats, millet, black velvet bean, fallow, and sorghum, in species richness and with brachiaria in the category of average uniformity.
Figure 4. Population fluctuation of natural enemies, in cover plants evaluated under diversity indices after the sowing of cotton in two consecutive cropping seasons (AC/Winter/Summer 2020/2021 and 2021/2022).
After analysing the number of predator species, it was found that the cover composed of black velvet bean, although associated with other crops in high values, showed the lowest diversity index in relation to abundance and uniformity of averages. Consequently, the treatment that obtained the lowest diversity rates was the one that belonged to the mix cover (fig. 4).
Discussion
Cover crops contribute to the modulation of pest population abundance throughout the phases of the experiment, with the greatest impact observed in the phase preceding cotton cultivation. The results revealed variations in the population dynamics of various pest species, highlighting the expressive influence of cover crop management. Fallow emerged as a conducive environment for the proliferation of various pest species, evidencing a high concentration throughout all phases and crops evaluated. Notably, D. speciosa exhibited exceptional adaptation to fallow, standing out not only for its widespread dispersion but also for its polyphagous feeding habit and reproductive efficiency (Hirose and Moscardi, Reference Hirose, Moscardi, Hoffmann-Campo and Corrêa-Ferreira2012). This trend corroborated previous studies that highlighted the ability of the beetle to adapt to different environmental conditions and food sources (Boiça Júnior et al., Reference Boiça Júnior, Costa, Souza, Forim, Perlatti and Cruz2022).
The frequent emergence of stink bugs, especially E. heros, in fallow during the two crops indicates a significant relationship between the presence of these insects and the diversity of plants in this treatment. The desiccation of the cover plants resulted in a drastic reduction in the activity of the stink bugs, indicating a strong dependence on the availability of food (Panizzi, Reference Panizzi1997; Bundy and Mcpherson, Reference Bundy and Mcpherson2000). The subsequent migration of these insects to other plants in the region, notably soybeans, highlights the importance of crop rotation in integrated pest management (Zerbino et al., Reference Zerbino, Miguel, Altier and Panizzi2020).
The behaviour of the corn leaf hopper (D. maidis) revealed a clear preference for grasses, especially corn, while black velvet bean exerted a significant inhibitory effect on its infestation. This polyphagous pattern of the corn leaf hopper (Nault, Reference Nault1980; Pitre et al., Reference Pitre, Combs and Douglas1996), concentrating on a few grasses (Oliveira et al., Reference Oliveira, Molina, Albres and Lopes2002), was evidenced in different phases of the experiment, contributing to the understanding of specific interactions between this pest and cover plants.
The presence of L. villosa in the oat crop, although with low frequency, highlights the importance of this cover plant as a potential host. This defoliating caterpillar's ability to consume both living and decomposing plants (Link et al., Reference Link, Panassolo and Gausmann1981) highlights its ecological role, although its population growth was not significant throughout the experiment.
Defoliating caterpillars, represented by S. frugiperda and S. eridania, showed distinct behaviours. While S. frugiperda reached a peak of infestation in the BC phase of the two crops, evidencing its preference for corn (Montezano et al., Reference Montezano, Specht, Sosa-gómez, Roque-specht, Sousa-Silva, Paula-moraes, Peterson and Hunt2018; Silva-Brandão et al., Reference Silva-Brandão, Peruchi, Seraphim, Murad, Carvalho, Farias, Omoto, Cônsoli, Figueira and Brandão2018), S. eridania had a practically insignificant presence, suggesting a possible negative interference of cover plants (Santos et al., Reference Santos, Meneguim and Neves2005), especially millet (Jesus et al., Reference Jesus, Sousa, Machado, Pereira and Alves2013), in its reproduction.
The cotton aphid (A. gossypii) and the thrips (F. schultzei) demonstrated occurrence patterns influenced by the phenology of the host plant (Costa et al., Reference Costa, Martins, Busoli and Cividanes2010; Lima et al., Reference Lima, Monteiro and Zucchi2013; Moraes et al., Reference Moraes, Bleicher, Silva and Marques2017) and climatic conditions (Barbosa et al., Reference Barbosa, Sarmento, Pereira, Pinto, Lima, Galdino, Santos and Picanço2019). The fallow treatment in the AC phase of the first crop stood out for a more intense infestation of thrips, indicating the influence of specific environmental factors in this period.
The caterpillar A. argillacea exhibited a uniform attack pattern, regardless of the crop or cover crop used. This consistent behaviour suggests the need for specific management strategies to deal with this pest, given its constant presence in all stages of cotton crop development (Bleicher et al., Reference Bleicher, Jesus, Ferraz and Melo1983; Ramalho, Reference Ramalho1994).
Broadly, black velvet bean emerged as the most effective cover plant, consistently displaying the lowest infestation indices of most of the pest insects analysed in the experiment. In contrast, the fallow (control) treatment stands out for presenting an inverse behaviour, evidencing itself as more prone to high levels of infestation, contrasting with black velvet bean.
It was observed that Brachiaria, in turn, emerges as the culture most susceptible to infestations of D. maidis and D. speciosa. However, when analysing the cover plants after the establishment of cotton in both crops, it was noted that the highest infestations related to D. speciosa occurred in oats, brachiaria, and fallow, while sorghum proved to be the best-performing culture in this aspect. Under similar circumstances, the mix and black velvet bean stand out as treatments with better performance in relation to the attack of A. gossypii. On the other hand, the mix and sorghum revealed the highest infestation of F. schultzei, while corn, fallow, and brachiaria present the lowest infestation indices for this pest.
Thus, this research points out that each culture investigated in this study presents distinct mechanisms that influence the development of specific species of insects. These findings suggest that these cultures can be strategically used as effective alternatives of cover plants, contributing significantly to the integrated management of common pests in direct cotton planting. This study reinforces the importance of the careful selection of cover plants, taking into account their specific properties in the context of integrated pest management for crops such as cotton.
In parallel, the results also reveal a significant variation in species richness among the different agricultural cultures studied. This variation is influenced by various factors, such as the diversity of niches and habitats provided by the structure of the plants, agricultural management practices, and availability of resources for the species (Espírito Santo et al., Reference Espírito Santo, Santos, Lopes, Silva and Lima2022). Notably, vegetable covers, including Brachiaria, corn, sorghum, and millet, stood out for presenting a larger number of species at specific moments.
Brachiaria, in addition to contributing to the formation of a vegetable cover, played a crucial role as a source of food and shelter for various species of phytophagous insects. This finding corroborates previous studies that highlight the susceptibility of Brachiaria to pasture leafhoppers and other pests such as corn earworms and stink bugs (Alvarenga et al., Reference Alvarenga, Auad, Moraes, Silva, Rodrigues and Silva2017; Montezano et al., Reference Montezano, Specht, Sosa-gómez, Roque-specht, Sousa-Silva, Paula-moraes, Peterson and Hunt2018; Tomacheski and Panizzi, Reference Tomacheski and Panizzi2018; Simon et al., Reference Simon, Silva, Medeiros, Lima, Fidelis, Silva, Bendahan and Schurt2021). The significant presence of insects associated with Brachiaria suggests the importance of this culture as a host and highlights the complexity of insect–plant interactions.
Corn, being the second most produced culture in Brazil, faces significant challenges related to pests, as indicated by the high infestation indices in the experiment. The expansion of these cultivated areas can create conducive environments for the development of harmful insects, contributing to the observed infestation. This finding underscores the need for effective integrated pest management strategies in regions where corn is a dominant crop (Resende et al., Reference Resende, Mendes, Marucci, Silva, Campanha and Waquil2016).
Sorghum and millet, although they are relevant crops, face similar challenges related to pest action. The presence of approximately 150 species that can attack sorghum and about 500 species that affect millet highlights the complexity and diversity of insect–plant interactions in these cultures (Sharma and Davies, Reference Sharma and Davies1988; Sharma, Reference Sharma1993). This diversity can directly influence the population dynamics of phytophagous insects and requires specific management strategies.
Crops such as oats and fallow, which showed low infestation rates, proved to be less susceptible to specific pests. This phenomenon can be explained by the preference of pest insects for polyphagous species when they do not find their preferred plants (Southwood and Norton, Reference Southwood, Norton, Geier, Clark, Anderson and Nix1973). Fallow, spontaneously composed of various plant species, reflects the positive influence of plant diversity on the functionality of the environment, impacting the diversity of arthropods (Wan et al., Reference Wan, Zheng, Fu, Kiaer, Zhang, Chaplin-Kramer, Dainese, Tan, Qiu, Hu, Tian, Nie, Ju, Deng, Jiang, Cai and Li2020).
Sorghum presented the highest abundance of phytophagous insects and the highest Shannon–Weaver index. This can be attributed to the greater susceptibility of this crop to certain pests present in the trial throughout its cycle (Sharma, Reference Sharma1993). In contrast, black velvet bean presented a lower population density of pests, possibly due to its allelopathic properties, acting as a natural insect repellent. This finding highlights the importance of the chemical characteristics of plants in regulating pest populations (Silva, Reference Silva1997; Isman, Reference Isman2006).
schultzei, in turn, presented a specific attack on cotton, showing variations in abundance in different cover crops. The high population of phytophagous thrips suggests the urgency of specific preventive and control measures for this pest. These results are consistent with previous studies that highlight the adaptability and specificity of these pests in relation to host plants (Oliveira and Rando, Reference Oliveira and Rando2017; Walsh et al., Reference Walsh, Ávila, Cabrera, Nava, Pinto and Weber2020).
Thrips' peculiar behaviour, characterised by sudden peaks and abrupt falls, resembles previous studies that highlight the specificity of these pests in relation to host plants (Silva et al., Reference Silva, Hereward, Walter, Wilson and Furlong2018). The presence of these insects in cover crops before the establishment of cotton suggests the persistence of crop residues in the population dynamics of polyphagous insects and preferential pests.
The presence of predators and natural enemies in cover crops is directly related to the number of pests in the area. The diversity of cover plants can offer shelter, protection, and complementary food for these predators, contributing to regulating populations of phytophagous insects (Batista et al., Reference Batista, Fonseca, Teodoro, Martins, Pallini and Venzon2017; Venzon et al., Reference Venzon, Togni, Amaral, Rezende, Batista, Chiguachi, Martins and Perez2018). The lower infestation of natural enemies in the plots of black velvet bean and oats can be explained by the low infestation rates of pests in these crops.
The comparative analysis of pest infestation indices revealed that sorghum had the highest levels, while black velvet bean had the lowest. Intriguingly, when relating these two covers to the population of predators, a relative stability in the values was observed, indicating a possible compensation between pests and natural enemies. On the other hand, millet and brachiaria stood out for the significant increase in the accumulation of pests after the establishment of cotton, highlighting the need for specific management strategies for these crops.
Natural enemies, analysed before implementing the main crop, exhibited distinct patterns. The oat crop presented lower population densities of these predators, while higher values characterised the mix. The presence of corn as a cover proved conducive to a higher population density in cotton, suggesting a potential beneficial effect of this cover in promoting natural enemies.
Based on the analysis of the data found in the experiment, we can affirm that the dynamic relationship between cover plants, pest insects, specific pests, and natural enemies reveals the underlying complexity of agricultural systems. These results not only emphasise the need for personalised management approaches but also highlight the importance of considering secondary pests and natural enemies in the design of sustainable strategies for pest control. Future studies should explore the mechanisms underlying these interactions, promoting a holistic approach to integrated pest management in agricultural environments.
In summary, this study's results highlight the complexity of interactions between cover plants, phytophagous insects, pests, and natural enemies. Different cover plants influence the population dynamics of insects in distinct ways, evidencing the need for more personalised management strategies. The presence of natural enemies in crops such as black velvet bean and oats until the end of the experiment suggests that these plants can play an important role in pest regulation. Future studies can deepen these investigations, considering population variations and the biochemical mechanisms underlying insect–plant interactions.
Conclusions
– Black velvet bean is the most effective cover plant, consistently presenting lower infestation indices for most pest insects.
– Conversely, Brachiaria is more susceptible to infestations of D. maidis and D. speciosa.
– The data obtained reinforce the importance of carefully selecting cover plants, considering their specific properties in the context of integrated pest management for cotton.
– Upon analysis, it becomes evident that one of the key conclusions of this study is the necessity for more comprehensive and holistic research in this field. This calls for future studies to delve deeper into the dynamics of pest management in cotton cultivation, particularly in conservative agriculture and no-tillage systems.
Acknowledgements
The authors thank CAPES (National Postdoctoral Program/Coordination for the Improvement of Higher Education Personnel).
Competing interests
The authors declare no conflict of interest with the participants or any other collaborator, direct or indirect, for the development of this research essay.
