The assessments carried out in this study revealed significant differences in the attractiveness and oviposition of B. tabaci MEAM1 in different weed and cultivated species, as well as variable colonization rates under no-choice conditions.
In the free-choice test (Table 2), tomato showed to be a highly attractive plant species to B. tabaci MEAM1, differing from all other species in the 2nd, 3rd and mean of the evaluations. Emilia sonchifolia, S. obtusifolia, soybean and E. heterophylla were also highlighted as attractive species. With respect to E. heterophylla, Sottoriva et al. (2014) found significantly higher infestation in this species compared to six other weed species and one cultivated plant (soybean). Their study found that over three evaluation periods, similar to those of the present study, E. heterophylla surpassed the soybean infestation in addition to presenting the highest number of eggs/cm². In the present study, E. heterophylla also presented a higher number of adults of B. tabaci, being among the most infested in the average of the evaluation periods in a free-choice test, similar to that verified with tomato. This species stood out even more in the no-choice test (Table 6), where it presented the highest averages of eggs and nymphs/cm² among all the evaluated species.
Other authors observed a higher incidence of eggs and nymphs of B. tabaci MEAM1 in E. heterophylla among the weed species found in cotton fields in the Midwest region of Brazil (Rodrigues & Silva, 2018). Among the evaluated species, C. benghalensis, Conyza spp., E. heterophylla, Ipomoea spp., R. brasiliensis and S. latifolia were considered alternative hosts for the insect. Of these species, E. heterophylla was the most prominent in the free-choice and no-choice assays of the present study, followed by S. latifolia, which also showed high densities of eggs and nymphs in the no-choice assay.
Evaluating the density of eggs and nymphs of B. tabaci MEAM1 on weeds present in Florida, Smith et al. (2014) observed higher rates in S. obtusifolia and in plants of the genus Emilia. In the free-choice test performed in the present study, S. obtusifolia and E. sonchifolia stood out as the most infested and with the highest number of eggs after tomato, corroborating the results of those authors.
In another study evaluating weeds as alternative hosts for B. tabaci, Barman et al. (2022) included Ipomoea quamoclit, Ipomoea cordatotriloba, Sida spinosa and E. heterophylla as suitable hosts for the insect. Among the whitefly’s unfavorable hosts, the species R. raphanistrum, Amaranthus palmeri and Richardia scabra stood out in Barman’s study. This corroborated the present study, where R. raphanistrum, and species of the Amaranthus and Richardia genus also emerged as unattractive hosts to B. tabaci MEAM1.
Given the variations in insect preference for different plant species, it should be noted that the process of identification and selection of a host plant by the insect involves a series of external stimuli, including visual, morphological, olfactory and gustatory aspects (Smith, 2005). In this sense, the color of the plant tissue, as well as the presence of secondary compounds and the density of trichomes can play an important role in the host selection behavior of the insect.
In the case of the whitefly, there is a strong dependence on visual stimuli for the orientation and movement of the insect (Mound, 1962; Isaacs et al., 1999). Color is considered one of the most important factors involved in host selection, according to Van Lenteren and Noldus (1990). Considering the sessile habit during most of the nymphal period of B. tabaci, the choice of a suitable host for oviposition is fundamental for successful insect colonization (Novaes et al., 2020). In the present study, significant correlations were found between the number of adults in the free-choice test (24 h) and the three colorimetric parameters evaluated in the materials (L*, a* and b*).
The correlation between the number of insects and luminosity was negative in this study, indicating a greater preference of adults of B. tabaci MEAM1 for plants with darker leaves (lower L* values). The species E. heterophylla, which presented the lowest light index, was among the most infested plants in the free-choice evaluations, and stood out with the highest averages of eggs and nymphs/cm² in the no-choice test. However, tomato and the species S. obtusifolia and E. sonchifolia, which were the most infested in the free-choice trial, showed intermediate values of L*. In a study evaluating cucumber genotypes, Novaes et al. (2020) found no significant correlation between this parameter and the presence of adults of B. tabaci MEAM1. Similarly, in a preference study with different genotypes of cabbage and the same whitefly species, no significant influence of luminosity on the insect selection behavior was found (Domingos et al., 2018). It is likely, therefore, that in studies in which genotypes of the same plant species are evaluated, smaller variations in the luminosity of the plant tissue should be observed in relation to studies involving different plant species, as is the present case.
Regarding the chromatic coordinates evaluated in this study, there was a negative correlation between the number of adults and the a* index, indicating a preference of B. tabaci MEAM1 for plant species with greater intensity of green. Among the evaluated species, tomato and E. sonchifolia showed the highest green intensities, standing above the most infested species. This same trend (intensity of green) was observed by Prado et al. (2015) who, in an evaluation of the active cotton germplasm bank of the Instituto Agronômico (IAC) for resistance to B. tabaci MEAM1, found that the IAC PV 010-175 strain, the most attractive host plant, presented the highest intensity of green.
Tomato and E. sonchifolia also showed higher levels of b*, confirming that B. tabaci MEAM1 also shows attraction to yellowish substrates, as documented by Berlinger (1980). Similar to the results obtained in the present study, Domingos et al. (2018) found a preference of B. tabaci MEAM1 for kale genotypes that had leaves with higher intensities of green and yellow, also corroborating previous work on the subject (Van Lenteren & Noldus, 1990).
In addition to the colorimetric factors, the trichomes present on the leaves can exert a strong influence on the colonization of B. tabaci. In this work, a positive correlation was found between the density of trichomes on the leaves of plant species and the density of eggs of B. tabaci MEAM1. Tomato plants showed the highest rates of preference for adults and oviposition in the free-choice test, and also stood out with the highest density of trichomes followed by soybean. The species S. latifolia, with a high density of trichomes, was among the least selected by adults of B. tabaci MEAM1 in the free-choice test, although it presented high averages of eggs and nymphs per cm² in the no-choice test. Biochemical studies indicate the occurrence of several secondary compounds in the Rubiaceae plant family, such as phenolic compounds, flavonoids, tannins, triterpenes and alkaloids (Cosmoski et al., 2015; Martins & Nunez, 2015), which may be related to changes in colonization behavior and the lower preference of adults of B. tabaci MEAM1 for S. latifolia (a member of the Rubiaceae plant family) in a free-choice test.
There are several hypotheses regarding the role of trichomes in the process of host plant selection and colonization by whiteflies. In one of them, it is suggested that the insect's preference for ovipositing at the base of the trichomes is associated with an adaptive advantage over the attack of natural enemies, as well as a possible favorable microclimate in leaves with high densities of non-glandular trichomes (Butter & Vir, 1989; Chu et al., 1995; Heinz & Zalom, 1995; Torres et al., 2012; Miyazaki et al., 2013). In this sense, the positive correlation between plant susceptibility and the presence of trichomes was found in studies with different crops such as soybean, cotton, tomato and eggplant (Heinz & Zalom, 1995; Silva et al., 2012; Hasanuzzaman et al., 2016; Oliveira et al., 2021). On the other hand, several reports in the literature point to negative correlations between the total density of trichomes and the oviposition of B. tabaci (Oriani & Vendramim, 2010; Taggar & Gill, 2012), indicating that the isolated analysis of this morphological factor may be questionable for the classification of susceptibility and/or resistance of plants to the whitefly (Muigai et al., 2003).
Adding to the quantification of these structures, other factors related to trichomes seem to influence the colonization of B. tabaci, such as angle, length and type (Lambert et al., 1995; Lakshminarayan et al., 2008; Valle et al., 2012). Glandular trichomes are capable of releasing allelochemicals involved in non-preference for oviposition, and may also keep the insect attached to the leaf surface through released exudates (Fancelli et al., 2005; Oriani & Vendramim, 2010). In this study, A. viridis showed greater variability of glandular trichomes (types IV, VI and VII) and also low averages of adults and eggs in free-choice assessments, confirming the negative role of this type of structure in whitefly colonization.
In general, the results obtained in this study reveal the expressive potential of some weed species, frequently present in crops in Brazil, to act as alternative hosts of B. tabaci MEAM1. Invasive species such as E. sonchifolia, S. obtusifolia and E. heterophylla proved to be highly attractive to the insect, showing infestation levels close to those observed with cultivated insect host species, such as tomato and soybean.
Although some species did not attract significant numbers of B. tabaci under free-choice conditions, they showed high rates of insect colonization under confined conditions (no choice), as verified with G. parviflora and S. latifolia. In this mandatory condition, E. heterophylla also stood out as the species with the highest density of eggs and nymphs. In this sense, it is important to emphasize that in agricultural scenarios where preferential hosts are not present, as well as in off-season periods, weeds can assist in maintaining pest populations in the cultivation area, thereby contributing to the continuity of the insect cycle.
In regions with high population densities of whitefly year around, it is important to focus on the presence of the weed species highlighted above. In addition to acting as attractive hosts for the whitefly, some invasive species play an important role in the dissemination of viruses transmitted by the insect, making it appropriate to adopt strategies aimed at eradicating and eliminating such plants (Gilbertson et al., 2011). Monitoring and control of these species can be recommended in order to reduce favorable conditions for the insect and possible sources of inoculum, thus contributing to the management of B. tabaci MEAM1 populations and its associated diseases in the field.