Introduction

Wintercress, Barbarea vulgaris Aiton (Brassicaceae), and other Barbarea spp. have been tested as traps crops and insectary plants for the management of diamondback moth, Plutella xylostella L. (Lepidoptera: Plutellidae) (Idris and Grafius 1994, 1995; Shelton and Nault 2004; Badenes-Pérez et al. 2005, 2014, 2017; Shelton and Badenes-Pérez 2006; Badenes-Pérez 2019). Barbarea vulgaris has two morphologically distinct forms, which have glabrous (G type) and pubescent (P type) leaves, respectively, and which also differ in their glucosinolate and saponin content, number of leaves, leaf size, and habitat adaptation (Agerbirk et al. 2003a, 2015; Kuzina et al. 2009; Nielsen et al. 2010b; Hauser et al. 2012, 2021; Badenes-Pérez et al. 2014; Christensen et al. 2014, 2016; Heimes et al. 2016; Badenes-Pérez and López-Pérez 2018). G-type B. vulgaris (henceforth the G type) contains the saponins 3-O-β-cellobiosylhederagenin and 3–0-β-cellobiosyloleanolic acid, which act as feeding deterrents for diamondback moth larvae, while P-type B. vulgaris (henceforth the P type) does not contain these saponins and is susceptible to the diamondback moth (Agerbirk et al. 2003a; Badenes-Pérez et al. 2014). Barbarea verna (Mill.) Asch. also contains sufficient amounts of these triterpenoid saponins to make this plant resistant to the diamondback moth (Badenes-Pérez et al. 2014). Because of this resistance to the diamondback moth, the G type and B. verna have been proposed as dead-end trap crops for this insect pest (Badenes-Pérez et al. 2004, 2005, 2014; Lu et al. 2004; Shelton and Nault 2004). Barbarea rupicola Moris contains low amounts of these saponins and some diamondback moth larvae can survive on it (Badenes-Pérez et al. 2014). On the other hand, these Barbarea spp., including the P type, are highly attractive to ovipositing diamondback moth and have, thus, potential to be used as trap crops, though only the G type and B. verna can be used as dead-end trap crops for this insect (Badenes-Pérez et al. 2014).

Besides the differences in resistance to the diamondback moth, the G and P types show different levels of resistance to other herbivores. The G type has shown resistance to the flea beetle Phyllotreta nemorum L. (Coleoptera: Chrysomelidae) (Agerbirk et al. 2001; Nielsen et al. 2010a; Kuzina et al. 2011; Hauser et al. 2021), to the western flower thrips Frankliniella occidentalis Pergande (Thysanoptera: Thripidae), and to the ascomycete fungus that causes powdery mildew on Brassicaceae Erysiphe cruciferarum Opiz ex L. Junell (Erysiphales: Erysiphaceae) (henceforth powdery mildew) (Badenes-Pérez and López-Pérez 2018). Compared to the P type, the G type is a worse host to the root-knot nematode Meloidogyne incognita (Kofoid & White) Chitwoodi (Tylenchida: Heteroderidae) (Badenes-Pérez and López-Pérez 2018). The G type is also less susceptible to the specialist turnip sawfly Athalia rosae L (Hymenoptera: Tenthredinidae) and to some species of generalist molluscs than the P type, but both types are equally attractive to the generalist cabbage moth Mamestra brassicae L. (Lepidoptera: Noctuidae) (Christensen et al. 2018). In contrast, the P type shows resistance to the oomycete pathogen that causes white rust Albugo candida (Pers.) Kuntze (Peronosporales: Albuginaceae), while the G type is susceptible to it (van Mölken et al. 2014; Christensen et al. 2014; Heimes et al. 2014; Hauser et al. 2021).

The cabbage whitefly Aleyrodes proletella L. (Hemiptera: Aleyrodidae) (henceforth cabbage whitefly) and the small white, also known as cabbage white butterfly and imported cabbageworm, Pieris rapae L. (Lepidoptera: Pieridae) (henceforth small white), can be economic pests in cruciferous crops (Shelton et al. 1982; Cartea et al. 2009; Badenes-Pérez et al. 2017). The cabbage whitefly is a polyphagous insect (Martin et al. 2000), while the small white is a specialist that feeds on crucifers (Verschaffelt 1910; Renwick 2001; Badenes-Pérez 2023). For the cabbage whitefly, glucosinolate content in cruciferous host plants does not seem to be a deterrent factor, rather the opposite, as higher content of the aliphatic glucosinolate sinigrin and total glucosinolate content have been shown to be associated to higher densities of cabbage whitefly in Brassica oleracea L. (Brassicaceae) (Newton et al. 2010; Hondelmann et al. 2020). Larvae of the small white possess a nitrile-specifier protein that directs glucosinolate hydrolysis to the formation of the less toxic nitriles, which allows them to feed on glucosinolate-containing crucifers (Wittstock et al. 2004; Agerbirk et al. 2010; Jeschke et al. 2016; Mesimeri et al. 2024). Glucosinolates and their glucosinolates products can affect plant resistance to small white larvae (Giamoustaris and Mithen 1995; Agrawal and Kurashige 2003; De Vos et al. 2008; Santolamazza-Carbone et al. 2016; Badenes-Pérez 2023) and can also act as host recognition cues for small white butterflies prior to ovipositing on plants (Renwick et al. 1992; Müller et al. 2010; Badenes-Pérez 2023). When making intraspecific comparison among plant accessions of Arabidosis thaliana L. and Brassica oleracea L. (Brassicaceae) that had different glucosinolate content, small white butterflies preferred to oviposit on accessions with higher total glucosinolate content and lower concentrations of certain aliphatic glucosinolates (Poelman et al. 2009; Müller et al. 2010; Coolen et al. 2022). Ovipositing small white females prefer indol-3-ylmethylglucosinolate and 2-phenylethylglucosinolate, an indolic and a benzenic glucosinolate, respectively, over allylglucosinolate, an aliphatic glucosinolate (Traynier and Truscott 1991; Renwick et al. 1992; Städler et al. 1995). Indolic glucosinolates are more important than aliphatic glucosinolates in determining small white oviposition preference (Huang and Renwick 1994; Badenes-Pérez 2023). Aliphatic glucosinolates can be detrimental in the ovipositional preference of the small white (Coolen et al. 2022).

Barbarea spp. have different glucosinolate concentrations (Agerbirk et al. 2003b; Badenes-Pérez et al. 2014). In the G and P types, the dominant glucosinolates are glucobarbarin and epiglucobarbarin, respectively, which are both benzenic glucosinolates, but the P type contains higher total glucosinolate content than the G type (Agerbirk et al. 2003b; Badenes-Pérez et al. 2014), which could affect oviposition preference in the small white. Unlike the diamondback moth, the small white can use the G type as a host plant (van Leur et al. 2008; Badenes-Pérez 2023).

The use of trap crops and insectary plants to manage a particular insect pest species can be affected by how susceptible they are to other pests, especially if the purpose is to eliminate or reduce pesticide use. For this reason, Barbarea spp. that show resistance to other insect pests besides the main target pest should be preferred over the ones that are more susceptible to other pests. In this study, Barbarea rupicola, B. verna, and the G and P types of B. vulgaris were tested for their susceptibility to the cabbage whitefly, while the oviposition preference of the small white was tested with the G and P types.

Materials and methods

Seeds of B. rupicola and B. verna were purchased at B & T World Seeds, Aigues-Vives, France, and Johnny´s Selected Seeds, Albion, ME, USA, respectively. Seeds of the G and P types were originally from Denmark; they were donated by Drs. Niels Agerbirk and Jens K. Nielsen and were later multiplied in Spain (see accessions B44 and B4 for G- and P-type seeds, respectively, in Agerbirk et al. (2003a, b)). Plants used in the experiments were grown in a peat moss substrate that included 1 kg of tref base fertilizer 17–10-14 per 70 l of substrate (Jiffy Tref GO V4, Jiffy Group, the Netherlands). Plants were grown in 15-cm pots. All plants grew as rosettes at a vegetative (pre-flowering stage) and B. rupicola plants appeared to be smaller than the plants of the other Barbarea spp. (personal observation).

Susceptibility of Barbarea plants to cabbage whitefly and powdery mildew

The experiment was conducted to test differences in resistance and susceptibility to cabbage whitefly among B. rupicola, B. verna, and G- and P-type B. vulgaris plants. The experiment was conducted in the greenhouse at 25 ± 3 ºC at the Institute of Agricultural Sciences in Madrid, Spain. The experiment relied on outbreaks and background infection pressure of whitefly and mildew populations that were naturally occurring in the greenhouse when the experiment was conducted. Plants were grown in the same greenhouse all the time and they were randomly located on a bench at the greenhouse and the minimum distance between adjacent plants was 15 cm. A total of 24 plants of each species and type were used. Plants were 12 weeks old when the observations were made. The whole plant was inspected and observations recorded the presence (any amount)/absence of any life stages of the cabbage whitefly in each plant. During the experiment, the percentage of plants affected by powdery mildew was also recorded.

Oviposition preference of the cabbage white butterfly for G- and P-type B. vulgaris

The experiment was conducted to test if female small white butterflies showed an ovipositional preference for either the G or the P type. Small white butterflies were collected in Jena, Germany, and were successively reared on cabbage plants. Insects were reared in environmental growth chambers (16:8 h light:dark, 21 ± 2º C and 55 ± 5 RH). Plants were 7 weeks old when the experiments were conducted. Two-choice oviposition experiments (i.e., one plant of the G type versus one plant of the P type) were conducted in 32.5 × 32.5 × 32.5 cm polyester cages that had a 96 × 26 mesh (MegaView Science Education Services Co., Ltd., Taichung, Taiwan). Five cages were used, each of which was considered a replicate. Two pairs of small white butterflies (two females and two males) were released in each cage and a small plastic cup with a 10% sugar solution on cotton was placed in the middle of the cage as a food source for the butterflies. Two days after releasing the butterflies, the number of eggs on the plants was counted. The location of eggs, either on the abaxial or the adaxial leaf sides of each plant, was also recorded in order to determine if the small white had a particular oviposition preference for either abaxial or adaxial leaf surfaces in the G and P types. Number of eggs on abaxial versus adaxial leaf surfaces was compared within each plant type. Additionally, the rate of abaxial/adaxial oviposition was calculated for each plant as the number of eggs laid on the abaxial leaf side divided by the number of eggs laid on the adaxial leaf side. This abaxial/adaxial oviposition rate was used to compare the two plant types in terms of oviposition preference for either abaxial or adaxial leaf surfaces. On abaxial and adaxial leaf surfaces in P-type plants, the number of trichomes was counted on leaves of different sizes, with maximum leaf widths ranging from 10 to 61 mm, by randomly selecting three leaves in each of four plants. A measure of trichome density was obtained by dividing the number of trichomes on each leaf sampled by the maximum leaf width of the same leaf. The maximum leaf width (mm) was measured in the large end-lobe of the compounded leaf and also coincided with the maximum leaf width of the whole compound leaf.

Statistical analysis

Data comparing percentage of plants with presence of cabbage whitefly and powdery mildew were analyzed using a one-tailed, two-sample test of proportions using STATA® version 15.1 with significance at P ≤ 0.05. To compare the oviposition preference of the small white between G- and P-type plants and between abaxial and adaxial leaf surfaces within the same plant type, data were analyzed using ANOVA with SPSS®. Prior to analysis, the suitability of the data for ANOVA was checked with the Kolmogorov–Smirnov normality test and the Levene test using SPSS®. The rates of abaxial/adaxial oviposition of the small white on G- and P-type plants and the number of trichomes/leaf width in P-type plants, data were analyzed using the Mann–Whitney U-test with SPSS®.

Results

Susceptibility of Barbarea plants to cabbage whitefly and powdery mildew

The percentage of plants with presence of cabbage whitefly was 50% in the G type, 8.3% in B. verna, and no infestation at all in B. rupicola and the P type (Fig. 1). Differences in the presence of cabbage whitefly were statistically significant when comparing the G type with B. verna (z = 3.18; P ≤ 0.001), with B. rupicola (z = 4.00; P ≤ 0.001), and with the P type (z = 4.00; P ≤ 0.001). In contrast, 95.8% of the P-type plants showed symptoms of powdery mildew, while the G type and the other Barbarea spp. were unaffected by this fungus and this difference in infection between the P type and the other Barbarea spp. and types was statistically significant (z = 3.18; P ≤ 0.001). Plants affected by cabbage whitefly and powdery mildew did not appear stunted, indicating that infection was not severe.

Fig. 1
figure 1

Percentage of plants with the presence of cabbage whitefly and powdery mildew among B. rupicola, B. verna, and G- and P-type B. vulgaris. Significant differences at P ≤ 0.05 across different host plants are shown with different letters

Full size image

Oviposition preference of the small white for G- and P-type B. vulgaris

Small white female butterflies showed no oviposition preference between G- and P-type plants (F1, 9 = 0.52, P = 0.491) (P ≤ 0.05) (Fig. 2). In G-type plants, small white showed no significant differences in oviposition on abaxial and adaxial leaf surfaces (F1, 9 = 1.19, P = 0.306), while these differences were significant in the case of P-type abaxial versus adaxial surfaces (F1, 9 = 8.58, P = 0.019) (Fig. 2). When comparing the rate of abaxial/adaxial leaf surface oviposition between the G and P types, small white oviposition on P-type plants tended to occur on the adaxial leaf side and the rate of abaxial/adaxial oviposition was lower on the P type than on the G type (TS = 2.00, P = 0.032). The value of trichome number/leaf width was slightly smaller on the adaxial (3.27 ± 0.82) than on the abaxial (4.38 ± 1.18) leaf surface of P-type plants, but this difference was not statistically significant (TS = 56.00, P = 0.378).

Fig. 2
figure 2

Abaxial and adaxial oviposition preference (mean ± SE eggs per plant leaf surface) of small white on G- and P-type B. vulgaris. Significant differences at P ≤ 0.05 are shown with an asterisk (*)

Full size image

Discussion

This study shows that B. rupicola, B. verna, and the P type are significantly less attractive to cabbage whitefly than the G type. This indicates that in locations were cabbage whitefly is a prevalent economic pests, B. verna, which can also be used as a dead-end trap crop for diamondback moth (Badenes-Pérez et al. 2014), could be chosen over the G type as trap crop. As the presence of trichomes in some Brassica spp. has been associated with resistance to cabbage whitefly, the presence of trichomes in the P type could be responsible for the antixenotic resistance observed in this plant type (Ramsey and Ellis 1996; Pelgrom et al. 2015). Further research is needed to understand the mechanisms of antixenotic resistance to cabbage whitefly observed in B. rupicola, B. verna, and the P type. The apparent glossiness of the G type, B. rupicola, and B. verna seems to indicate that these plant types contain lower amounts of leaf epicuticular wax on the leaf surface than the P type (Badenes-Pérez, personal observation). Leaf cuticular wax has been shown to affect plant susceptibility to whiteflies in cotton (Ali et al. 2021). However, wax removal in two cabbage varieties did not affect cabbage whitefly infestation (Broekgaarden et al. 2012).

Further research is needed to test if P-type plants would be equally affected by powdery mildew in the field compared to greenhouse conditions. Several studies have shown that powdery mildew is an emerging threat to some Brassica crops under climate warming conditions, but it is unknown if the P type would be as severely affected as other Brassicaceae species (Enright and Cipollini 2007; Barbetti et al. 2012; Uloth et al. 2018; Runno-Paurson et al. 2021). The fact that the P type has larger leaves than the other Barbarea spp. and types tested (Badenes-Pérez and López-Pérez 2018) might have also affected powdery mildew infection because larger leaves can catch more water and retain it for longer time (Bradley et al. 2003). Trichomes can also capture fungal spores and be a preferred site for fungal infection (Kim 2019). It is also unknown if the presence of powdery mildew had any effect on the absence of cabbage whitefly in P-type plants. Fungal infection can induce plant secondary metabolites and influence subsequent attraction and resistance to insects (Abdel-Farid et al. 2009; Aghajanzadeh et al. 2023; Jindřichová et al. 2024). Testing if the presence of powdery mildew has any effect on the absence of cabbage whitefly infection in P-type plants would require either using a fungicide, which could also affect insect infestation (Sánchez-Bayo 2021; Margus et al. 2023), or conducting the cabbage whitefly infection tests in environmental conditions less conducive to powdery mildew infection than the ones tested in these experiments. In the P-type plants tested here, not a single cabbage whitefly was seen, even in leaves that showed no powdery mildew infection. P-type plants affected by powdery mildew were not stunted nor had all their leaves with visible infection by this fungus. Thus, mildew infection was not as severe as to completely prevent cabbage whitefly from using P-type plants as a host. Absence of cabbage whitefly in the P type could be due to the marked preference of this insect for other host plants, like the G type and B. verna.

There was no difference in ovipositional preference by the small white between G- and P-type plants with different trichome densities and benzenic glucosinolate content. In a different study with two other chemotypes of B. vulgaris that have different benzenic glucosinolate content as dominant glucosinolates, small white also did not show any oviposition preference for any of the chemotypes (van Leur et al. 2008). In the present study, differences in small white oviposition preference between the G and P types occurred at the within-plant level, where small white preferred ovipositing on the adaxial leaf surfaces on P-type plants and showed no preference for either abaxial or adaxial leaf surfaces on G-type plants. In a different study with 16 different species of plants in the order Brassicales, this oviposition preference of small white for adaxial leaf surfaces was shown for two other species, Arabidopsis thaliana L. and Iberis amara L., while for the other plant species tested, the differences between abaxial and adaxial oviposition were not significant. Glucosinolate content on adaxial leaf surfaces was found to be lower than on abaxial leaf surfaces in B. rupicola and B. verna, showing no significant differences on the G type (Badenes-Pérez et al. 2011). However, the content of glucosinolates on the leaf surfaces of P-type plants is unknown and using gum arabic to mechanically remove surface waxes and determine glucosinolate content was not possible because of the removal of fragments of plant tissue with the gum arabic peelings (Badenes-Pérez, personal observation). Thus, it is unknown if the oviposition preference of the small white for the adaxial leaf surface of the P type could be due to the differences in glucosinolate content. Although trichome density was slightly lower on adaxial than on abaxial leaf surfaces in the leaves sampled, this difference was not statistically significant, so an oviposition preference for lower trichome density was not found in the study. The fact that there is no oviposition preference between the P- and G-type plants also seems to indicate that trichome density is not a major factor in the oviposition preference of the small white. Preference for adaxial versus abaxial oviposition preference may affect the management of the small white because parasitism can be higher for small white larvae located on the adaxial side compared to the abaxial side of leaves (Tagawa et al. 2008). Furthermore, some insecticide sprayers deposit more insecticide on the adaxial than on the abaxial leaf side (Maski and Durairaj 2010). Oviposition on the adaxial leaf side, which is more visible to birds and other predators, could also increase predation (Baker 1970; Schmaedick and Shelton 1999) and susceptibility to dislodgement by rainfall, as it happens in P. xylostella (Rahman et al. 2019).

When trap crops and insectary plants are deployed to manage a particular insect pest species, they can be affected by other pests, which could reduce their effectiveness as trap crops. Barbarea spp. that show resistance to other insect pests besides the main target pest should be preferred. Both the cabbage whitefly and the small white are common pests in cruciferous crops. This study shows that different species and types of Barbarea show different levels of antixenotic resistance to cabbage whitefly and different patterns of oviposition preference in the small white. For this reason, depending on the relative economic importance of pests, different Barbarea spp. and types could be chosen as trap crops for P. xylostella. Further research is needed to understand the mechanisms of antixenotic resistance to cabbage whitefly in Barbarea spp. and types, which could be used as a source of resistance to this pest in plant breeding programs.