Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Insect Attraction versus Plant Defense: Young Leaves High in Glucosinolates Stimulate Oviposition by a Specialist Herbivore despite Poor Larval Survival due to High Saponin Content

Insect Attraction versus Plant Defense: Young Leaves High in Glucosinolates Stimulate Oviposition by a Specialist Herbivore despite Poor Larval Survival due to High Saponin Content

  • Francisco R. Badenes-Perez, 
  • Jonathan Gershenzon, 
  • David G. Heckel
PLOS
x

Abstract

Glucosinolates are plant secondary metabolites used in plant defense. For insects specialized on Brassicaceae, such as the diamondback moth, Plutella xylostella L. (Lepidoptera: Plutellidae), glucosinolates act as “fingerprints” that are essential in host plant recognition. Some plants in the genus Barbarea (Brassicaceae) contain, besides glucosinolates, saponins that act as feeding deterrents for P. xylostella larvae, preventing their survival on the plant. Two-choice oviposition tests were conducted to study the preference of P. xylostella among Barbarea leaves of different size within the same plant. P. xylostella laid more eggs per leaf area on younger leaves compared to older ones. Higher concentrations of glucosinolates and saponins were found in younger leaves than in older ones. In 4-week-old plants, saponins were present in true leaves, while cotyledons contained little or no saponins. When analyzing the whole foliage of the plant, the content of glucosinolates and saponins also varied significantly in comparisons among plants that were 4, 8, and 12 weeks old. In Barbarea plants and leaves of different ages, there was a positive correlation between glucosinolate and saponin levels. This research shows that, in Barbarea plants, ontogenetical changes in glucosinolate and saponin content affect both attraction and resistance to P. xylostella. Co-occurrence of a high content of glucosinolates and saponins in the Barbarea leaves that are most valuable for the plant, but are also the most attractive to P. xylostella, provides protection against this specialist herbivore, which oviposition behavior on Barbarea seems to be an evolutionary mistake.

Introduction

According to the optimal defense theory, the most valuable parts of a plant should also be the most protected against herbivores [1], [2]. Young leaves are supposed to be more valuable than older ones because they can make a higher contribution to the fitness of the plant as a result of having relatively higher photosynthetic potential [3]. In agreement with this theory, it has been found that, within a plant, different organs and leaves can contain different concentrations of defense compounds [4], [5]. This the case for glucosinolates, plant secondary metabolites used for defense and found mainly in plants of the order Brassicales [6], [7], which have been found in higher concentrations in younger compared to older leaves within the same plant [4], [8][11]. At the whole plant level, glucosinolate content also changes over time, but not in a linear manner [4], [8], [12]. Like glucosinolates, saponins are plant secondary metabolites used in plant defense [13][15]. With the exception of insects specialized on saponin-rich plants [16], saponins act as feeding-deterrents and are toxic [17][19]. We have not found any studies addressing changes in saponin content with leaf age in Brassicaceae, but saponin content in leaves has been shown to decrease with leaf age in American holly, Ilex opaca Aiton (Aquifoliaceae) [20]. Saponin content also changes over time at the whole plant level, often increasing with plant age, although decreases with plant age and seasonal fluctuations have also been recorded [21][23]. In the genus Barbarea R. Br. (Brassicaceae), the only one in Brassicaceae where saponins have been found so far [24][28], seasonal fluctuations in saponin content seem to occur as inferred by changes in resistance to the flea beetle Phyllotreta nemorum L. (Coleoptera: Chrysomelidae) [29], [30].

The diamondback moth, Plutella xylostella L. (Lepidoptera: Plutellidae), is an insect herbivore specialized on glucosinolate-containing crucifers [31][33]. Specialist insects like P. xylostella have evolved mechanisms to avoid the toxicity of glucosinolates, which are used in host plant recognition and act as feeding and oviposition stimulants [7], [33][38]. Larvae of P. xylostella cannot survive on some varieties and types of B. vulgaris despite these plants being highly preferred for oviposition by P. xylostella adults [24], [25], [28], [39][42]. This oviposition mistake of P. xylostella has been investigated in pest management to use Barbarea plants as a dead-end trap crop for P. xylostella, which is considered one of the most damaging insect pests of cruciferous crops throughout the world [28], [32], [40], [43]. Among the varieties and types of B. vulgaris on which P. xylostella cannot survive are B. vulgaris var. variegata and G-type (glabrous) B. vulgaris var. arcuata, while P-type (pubescent) B. vulgaris var. arcuata allows survival of P. xylostella larvae [24], [25]. The resistance of B. vulgaris to P. xylostella is caused by the triterpenoid saponins 3-0-β-cellobiosylhederagenin (saponin 1) and 3-0-β-cellobiosyloleanolic acid (saponin 2), which act as feeding-deterrents or are correlated with deterrency in P. xylostella larvae [24], [25]. Other B. vulgaris types and other Barbarea spp. containing saponins 1 and 2 are also resistant to P. xylostella [28].

Adult oviposition and larval feeding preference for younger over older leaves of a particular host plant is a common trend among many herbivorous insects, especially in specialists, including P. xyllostella [9], [44], [45]. Within the same plant, two-choice oviposition preference tests have shown that P. xylostella prefers to oviposit on younger leaves of <3.0 maximum leaf diameter compared to older leaves of >6.0 maximum leaf diameter of B. vulgaris [44]. Given the type of rosette growth of Barbarea plants, at the age of the plants used in the study by Badenes-Perez et al (2005) and here, leaf size was correlated with leaf age. Although containing lower content of toxic metabolites than younger leaves, older leaves may be less nutritious for insects than younger leaves [46], [47]. Feeding on older leaves can also increase insect mandibular wear more than feeding on younger leaves because of the increased toughness of older leaves [48]. At the whole plant level, the plant phenological age hypothesis also predicts that herbivores prefer and perform better on younger compared to older plants [49]. However, there are many cases in which insects prefer older plants over younger ones [50], [51]. Among Brassica oleracea L. and B. vulgaris plants aged between 6 and 14 weeks old, P. xylostella also preferred to oviposit on older versus younger plants, even though survival and development of P. xylostella larvae can be negatively affected by plant age [52].

In relation to the preference of P. xylostella for younger leaves and older plants of B. vulgaris [44], [52], it is not known whether there could be an association between this preference and plant concentrations of glucosinolates and saponins, the former being oviposition and feeding stimulants, and the latter preventing the survival of the insect on the plant. We hypothesize that, given the known attraction of P. xylostella to glucosinolates, if glucosinolate content in Barbarea leaves is higher in younger compared to older leaves as it happens in other Brassicaceae, P. xylostella would preferentially oviposit on young leaves. It is not known how saponins vary with leaf and plant age in Barbarea spp. and whether they are correlated with changes in glucosinolate content. A correlation between plant content of glucosinolates and saponins could protect Barbarea plants from specialist insects adapted to glucosinolates. Furthermore, as both glucosinolates and saponins are plant defense compounds, their co-occurrence would have implications for the protection of Barbarea plants, not only against P. xylostella, but against other herbivores as well. To test our hypotheses, we conducted two-choice oviposition preference tests and measured glucosinolate and saponin concentrations in Barbarea leaves of different size to test the association between leaf size, oviposition preference, and glucosinolate and saponin concentrations. We also measured glucosinolate and saponin content in plants of different age. Besides analyzing true leaves, we analyzed the glucosinolate and saponin content of cotyledons. Larval survival and oviposition preference tests were also conducted with cotyledons and true leaves within the same plant.

Results

Analysis of Glucosinolates and Saponins in Barbarea spp

A significant negative relationship was found between leaf size and content of glucosinolates for both G-type (y = 13.01–0.95x; n = 100; r = 0.39; F1,98 = 17.01; P≤0.001) and P-type B. vulgaris (y = 14.81–0.86x; n = 20; r = 0.44; F1,19 = 4.20; P = 0.050) (Figs. 1A and 1B). In G-type B. vulgaris, the glucosinolate that decreased the most with increasing leaf size was the dominant glucosinolate (S)-2-hydroxy-2-phenylethylglucosinolate (S2OH2PE) (y = 11.95–0.89x; n = 100; r = 0.38; F1,98 = 16.13; P≤0.001), but concentrations of (R)-2-hydroxy-2-phenylethylglucosinolate (R2OH2PE) (y = 0.26–0.17x; n = 100; r = 0.31; F1,98 = 10.34; P = 0.002), indol-3-ylmethylglucosinolate (I3M) (y = 0.64–0.04x; n = 100; r = 0.42; F1,98 = 21.01; P≤0.001), and 4-methoxyindol-3-ylmethylglucosinolate (4MOI3M) (y = 0.09–0.01x; n = 100; r = 0.29; F1,98 = 8.93; P = 0.003) also decreased with leaf size. In P-type B. vulgaris, only the dominant glucosinolate R2OH2PE (y = 4.25–1.07x; n = 20; r = 0.46; F1,19 = 4.69; P = 0.044) decreased significantly with leaf size; concentrations of the other glucosinolates found did not vary significantly with leaf size.

thumbnail
Figure 1. Total glucosinolates (µmol/g of plant fresh weight) in leaves of different size in G-type Barbarea vulgaris var. arcuata (A) and P-type B. vulgaris var. arcuata (B).

For G- and P-type B. vulgaris leaves, maximum leaf width ranged from 3.0 to 11.4 cm (n = 100) and from 4.4 to 12.2 cm (n = 20), respectively.

https://doi.org/10.1371/journal.pone.0095766.g001

In G-type B. vulgaris, a significant negative relationship was also found between leaf size and content of saponin 1 (y = 7.50–0.61x; n = 100; r = 0.51; F1,98 = 34.89; P≤0.001) and saponin 2 (y = 2.50–0.23x; n = 100; r = 0.47; F1,98 = 27.99; P≤0.001) (Fig. 2). In these same leaves of different size, there was a significant positive relationship between saponin and glucosinolate content for both saponin 1 (ln (y+1) = 1.36+0.26ln (x+1); n = 100; r = 0.21; F1,98 = 4.56; P = 0.035) and saponin 2 (ln (y+1) = 1.47+0.43ln (x+1); n = 100; r = 0.25; F1,98 = 6.44; P = 0.013).

thumbnail
Figure 2. Relative content of 3-0-β-cellobiosylhederagenin (saponin 1) and 3-0-β-cellobiosyloleanolic acid (saponin 2) in leaves of different size in G-type Barbarea vulgaris var. arcuata.

Maximum leaf width ranged from 3.0 to 11.4(n = 100). Units of peak areas for the signal of the molecular ion in the negative-ion mass spectrum [M-H] divided by 100,000/mg of leaf fresh weight.

https://doi.org/10.1371/journal.pone.0095766.g002

When leaves of different sizes were grouped into three groups according to maximum leaf width (large, >50 mm; medium, 20–50 mm; and small, <20 mm), there were significant differences in the content of glucosinolates (F2,108 = 224.31; P≤0.001) and saponins 1 (F2,81 = 19.36; P≤0.001) and 2 (F2,81 = 9.06; P≤0.001) among the three different leaf sizes (Tables 1 and 2). Small and large leaves had, respectively, the highest and the lowest concentrations of both glucosinolates and saponins. For these leaves of different size, there was a significant positive relationship between saponin and glucosinolate content for both saponin 1 (y = 2.10+1.63*10−5x; n = 30; r = 0.40; F1,28 = 5.41; P = 0.027) and saponin 2 (y = 1.97+1.02*10−5x; n = 30; r = 0.45; F1,28 = 7.22; P = 0.012) in G-type B. vulgaris; for saponin 1 (y = 1.81+9.91*10−5x; n = 30; r = 0.79; F1,28 = 54.55; P≤0.001) in B. verna (the relationship was not statistically significant for saponin 2); and for saponin 1 (y = 6.61+1.80*10−4x; n = 30; r = 0.48; F1,28 = 8.29; P = 0.008) in B. rupicola, which did not contain any saponin 2.

thumbnail
Table 1. Mean ± SE glucosinolates (µmol/g of leaf fresh weight) concentrations in large, medium and small leaves of B. rupicola, B. verna, and G- and P- type B. vulgaris var. arcuata.

https://doi.org/10.1371/journal.pone.0095766.t001

thumbnail
Table 2. Mean ± SE 3-0-β-cellobiosylhederagenin (saponin 1) and 3-0-β-cellobiosyloleanolic acid (saponin 2) in large, medium, and small leaves within the same plant.

https://doi.org/10.1371/journal.pone.0095766.t002

There were significant differences in the content of glucosinolates in cotyledons and true leaves of plants of G-type B. vulgaris, P-type B. vulgaris, NAS-type B. vulgaris, B. vulgaris variegata, B. rupicola, and B. verna (F1,134 = 261.79; P≤0.001) (Table 3). True leaves contained approximately 2.3, 7.4, 2.9, 3.0, 3.5, and 4.2 times more glucosinolates than cotyledons in G-type B. vulgaris, P-type B. vulgaris, NAS-type B. vulgaris, B. vulgaris variegata, B. rupicola, and B. verna, respectively. Present in higher concentrations in true leaves than in cotyledons were the individual glucosinolates S2OH2PE (F1,134 = 51.45; P≤0.001); R2OH2PE (F1,134 = 44.22; P≤0.001); I3M (F1,134 = 18.63; P≤0.001); and 2-phenylethylglucosinolate (2PE) (F1,134 = 399.78; P≤0.001). Concentrations of S2OH2PE were 2.6 and 3.3 times higher in true leaves than in cotyledons in G-type B. vulgaris and B. vulgaris variegata, respectively. Concentrations of R2OH2PE were up to 18.8 times higher in true leaves than in cotyledons in P-type B. vulgaris. In B. verna, concentrations of I3M were up to 4.8 times higher in true leaves than in cotyledons. In G-type B. vulgaris, concentrations of 2PE were up to 36.0 times higher in true leaves than in cotyledons. Concentrations of 4MOI3M, however, were lower in true leaves than in cotyledons (F1,134 = 113.77; P≤0.001). In B. verna, concentrations of 4MOI3M were up to 24.0 times lower in true leaves than in cotyledons.

thumbnail
Table 3. Mean ± SE glucosinolates (µmol/g of leaf fresh weight) concentrations in cotyledons and true leaves in plants of B. rupicola, B. verna, G- and P-type B. vulgaris var. arcuata, NAS-type B. vulgaris, and B. vulgaris var. variegata.

https://doi.org/10.1371/journal.pone.0095766.t003

There were significant differences in the content of saponins 1 (F1,35 = 32.48; P≤0.001) and 2 (F1,35 = 5.49; P = 0.025) in true leaves and cotyledons of plants of G-type B. vulgaris and B. verna (Table 4). No saponins were found in cotyledons. Similarly, when comparing true leaves and cotyledons of plants of NAS-type B. vulgaris, B. vulgaris variegata, and B. rupicola, we found significant differences in the content of saponins 1 (F1,40 = 48.54; P≤0.001) and 2 (F1,40 = 1400.14; P≤0.001) (Table 4). No saponins were found in true leaves and cotyledons of B. rupicola. In NAS-type B. vulgaris, saponin 1 was found in all true leaves and in 3 out of 13 cotyledons analyzed (the peak areas of the signal of [M-H] were on average 9.7 times smaller for cotyledons than for true leaves in the plants which cotyledons contained saponins). In B. vulgaris variegata, saponin 1 was found in all true leaves and in 1 out of 5 cotyledons analyzed (the peak areas of the signal of [M-H] were 7.1 times smaller for cotyledons than for true leaves in the plant of B. vulgaris variegata which cotyledon contained saponins). Saponin 2 was not detected in cotyledons. In plants of G- and P-type B. vulgaris, B. rupicola, and B. verna, 5 h after cutting all leaves except either one cotyledon or one true leaf, there were no differences in the content of glucosinolates (F1,64 = 0.11; P = 0.743) (Table S1) and saponins 1 (F1,48 = 0.42; P = 0.436) and 2 (F1,48 = 0.52; P = 0.473) (Table S2) compared to the same type of leaves in intact plants with all their leaves remaining (Table S1). The glucosinolate content of Barbarea seeds is shown on Table 5. Barbarea seeds did not contain saponins 1 and 2.

thumbnail
Table 4. Mean ± SE 3–0-β-cellobiosylhederagenin (saponin 1) and 3–0-β-cellobiosyloleanolic acid (saponin 2) in cotyledons and true leaves in plants of B. rupicola, B. verna, G-type B. vulgaris var. arcuata, NAS-type B. vulgaris, and B. vulgaris var. arcuata.

https://doi.org/10.1371/journal.pone.0095766.t004

thumbnail
Table 5. Mean ± SE glucosinolates (µmol/g of seed) in seeds of B.verna (n = 4), G-type B. vulgaris var. arcuata (n = 2), P-type B. vulgaris var. arcuata (n = 4), and B. vulgaris var. variegata (n = 2).

https://doi.org/10.1371/journal.pone.0095766.t005

When analyzing the whole plant foliage, there were significant differences in glucosinolate content among plants of different age in both G-type (F2,27 = 10.70; P≤0.001) and P-type plants (F2,27 = 56.29; P≤0.001) (Fig. 3). In both G- and P-type B. vulgaris plants, total glucosinolate content was highest in 8-week-old plants and lowest in 4-week-old plants. Among the individual glucosinolates in G-type B. vulgaris, those that varied the most with plant age were I3M (F2,27 = 43.87; P≤0.001), S2OH2PE (F2,27 = 7.58; P = 0.002) and 4MOI3M (F2,27 = 4.49; P = 0.021), while 2PE (F2,27 = 0.15; P = 0.865) and R2OH2PE (F2,27 = 0.30; P = 0.741) did not show significant variation (Table 6). Among the individual glucosinolates that varied the most with plant age in P-type B. vulgaris were R2OH2PE (F2,27 = 66.32; P≤0.001), I3M (F2,27 = 30.78; P≤0.001), and 4MOI3M (F2,27 = 4.63; P = 0.018), while 2PE (F2,27 = 3.33; P = 0.051) and S2OH2PE (F2,27 = 2.14; P = 0.138) did not vary significantly (Table 6). There were also significant differences in the content of saponins 1 (F2,27 = 8.51; P = 0.001) and 2 (F2,27 = 3.86; P = 0.034) among the G-type B. vulgaris plants of different age (Fig. 4). As in the case of total glucosinolate content, the content of saponins 1 and 2 was highest in 8-week-old plants and lowest in 4-week-old plants. For G-type B. vulgaris plants of different ages, there was also a significant positive relationship between saponin and glucosinolate content for both saponin 1 (y = 1.14+30.35x; n = 28; r = 0.65; F1,26 = 19.05; P≤0.001) and saponin 2 (y = 1.90+62.90x; n = 28; r = 0.68; F1,26 = 22.12; P≤0.001).

thumbnail
Figure 3. Mean ± SE glucosinolates (µmol/g of plant fresh weight) in the whole foliage of G- and P-type Barbarea vulgaris var. arcuata plants of 4-, 8-, and 12-weeks old plants (n = 10).

https://doi.org/10.1371/journal.pone.0095766.g003

thumbnail
Figure 4. Mean ± SE 3-0-β-cellobiosylhederagenin (saponin 1) and 3-0-β-cellobiosyloleanolic acid (saponin 2) in the whole foliage of G-type Barbarea vulgaris var. arcuata plants of 4-, 8-, and 12-weeks old plants (n = 10).

https://doi.org/10.1371/journal.pone.0095766.g004

thumbnail
Table 6. Mean ± SE glucosinolates (µmol/g of plant fresh weight) in foliage of G- and P- type B. vulgaris var. arcuata plants 4-weeks old, 8-weeks old, and 12-weeks old (n = 10).

https://doi.org/10.1371/journal.pone.0095766.t006

Oviposition Preference Tests between Leaves of Different Size within the Same Plant

There was a significant negative relationship between leaf size and number of eggs laid by P. xylostella per leaf area (y = 1.84–0.17x; n = 84; r = 0.58; F1,83 = 42.15; P≤0.001) (Fig. 5A). The number of eggs laid per leaf area was also positively correlated with leaf glucosinolate content (y = 0.32+0.04x; n = 84; r = 0.34; F1,83 = 10.40; P = 0.002) (Fig. 5B). In the case of P. xylostella oviposition preference between true leaves and cotyledons within the same plant, significantly more eggs were laid on true leaves than on cotyledons (F1,18 = 12.62; P = 0.002). When considering the numbers of eggs laid by P. xylostella per leaf area, however, these differences were not statistically significant (F1,18 = 0.47; P = 0.502) (Table 7).

thumbnail
Figure 5. Correlation between leaf size and number of eggs laid by P. xylostella per leaf area (A) and between leaf glucosinolate content and number of eggs laid by P. xylostella per leaf area (B). Leaves of G-type Barbarea vulgaris var. arcuata plants were used (n = 84).

https://doi.org/10.1371/journal.pone.0095766.g005

thumbnail
Table 7. Leaf areas of cotyledons and true leaves of B. rupicola, B. verna, and G-type B. vulgaris, and numbers of eggs laid by P. xylostella on cotyledons and true leaves in two-choice preference tests (n = 4 for each plant and leaf type).

https://doi.org/10.1371/journal.pone.0095766.t007

Survival of Larvae on True Leaves and Cotyledons within the Same Plant

On true leaves of plants, 100% and 80% of P. xylostella larvae survived after 5 days on P-type B. vulgaris and B. rupicola, respectively (for each Barbarea tested, n = 5 plants, each with 5 larvae). No larvae survived the 5-day period on true leaves of G-type B. vulgaris, B. vulgaris variegata, NAS-type B. vulgaris, and B. verna. On cotyledons, however, survival of P. xylostella larvae after 5 days was high for all Barbarea plants tested: 100%, 100%, 80%, 100%, 100% and 100% on G-type B. vulgaris, P-type B. vulgaris, B. vulgaris variegata, NAS-type B. vulgaris, B. rupicola, and B. verna, respectively.

Discussion

Our research shows that in Barbarea, the only genus of the Brassicaceae family known to simultaneously contain glucosinolates and saponins, content of these two plant defense compounds are negatively correlated with leaf size. Oviposition preference by P. xylostella was also negatively correlated with leaf size because P. xylostella laid more eggs per leaf area on smaller leaves than on larger ones. P. xylostella and many other herbivores use plant secondary metabolites as “fingerprints” to recognize hosts and oviposit on them [34], [53]. In our study, attraction to glucosinolates seemed to be more important for ovipositing P. xylostella than avoidance of saponins, which is consistent with the presence and absence of glucosinolates and saponins, respectively, on the leaf surface of Barbarea in concentrations perceivable by P. xylostella [34]. In Barbarea plants with saponins, these were found in true leaves of all sizes, while no saponins (or very small amounts of them) were found in cotyledons. Larvae of P. xylostella could survive on cotyledons, even in those Barbarea plants whose true leaves contained enough saponin concentrations to prevent their survival.

In Lepidoptera, survival is greatly determined by the oviposition behavior of adult females, as immature stages have limited mobility [54]. Consequently, most ovipositing Lepidoptera prefer to oviposit on hosts where their larvae are able to survive, but there are cases in which the correlation between oviposition preference and larval performance is poor, and several hypotheses have been put forward to interpret this apparently non-adaptative behavior [55][57]. With few exceptions, such as ovipositing on Barbarea [28], [39], [40], P. xylostella oviposition preference and larval performance are positively correlated [58]. However, the oviposition preference for smaller Barbarea leaves over larger ones demonstrated here for P. xylostella seems to be a non-adaptive mechanical response to cues given by plant secondary metabolites (glucosinolates) specific from their cruciferous host plants. Given that no P. xylostella larvae survive on resistant Barbarea, and that survival of larvae is less likely on small leaves that contain high concentrations of saponins, there cannot be any selective advantage in the oviposition behavior of P. xylostella on Barbarea. The relatively low content of saponins in larger leaves of Barbarea would make P. xylostella more likely to survive on the plant, yet larger leaves also have relatively low concentrations of glucosinolates, which make them less stimulatory for P. xylostella larvae [35]. The preference of P. xylostella moths for younger leaves within a Barbarea plant represents a second “oviposition mistake”, on top of the known “oviposition mistake” of P. xylostella preferring resistant Barbarea plants over other host plants that allow survival of its larvae [28], [40][42].

Cotyledons serve as a storage of nutrients for the growing plant and they are the first photosynthetic tissue appearing above-ground when germination occurs [59]. Cotyledons of brassicaceous plants usually contain variable amounts of glucosinolates [8], [60]. In Barbarea plants, glucosinolates, which can defend the plants against generalist herbivores, were present in cotyledons, but saponins, which could also protect the plant against specialist herbivores like P. xylostella, were not (or were present in very low concentrations). No saponin 2 was detected in cotyledons, indicating that synthesis of saponin 2 could be subsequent to that of saponin 1 (assuming that saponins are not translocated from cotyledons to other parts of the plant). Unlike glucosinolates, saponins were not found in the seeds of Barbarea, indicating that saponins may start being produced in cotyledons and true leaves after some time once true leaves appear. Lack of saponins in seeds and cotyledons indicates that, for some time, at the seedling stage, the plant may not be protected against P. xylostella and other herbivores. However, given the small size of cotyledons, they do not provide sufficient food for a P. xylostella larva to develop from first instar to pupa (Badenes-Perez, personal observation). Even though P. xylostella is known to oviposit on cotyledons of crucifer seedlings in the field, upon egg hatch, larvae move to true leaves, where they prefer to feed [61]. With the exception of 4MOI3M, we found that cotyledons contained lower concentrations of glucosinolates than true leaves in the same plant. This, together with the low frequency of cotyledons with saponins, the low concentrations of saponins found in those cotyledons with saponins, and the ensuing survival of P. xylostella larvae on cotyledons, indicates that cotyledons are not as protected from herbivory as true leaves. However, as Barbarea spp. are early successional biennial plants that appear early in the season [62], cotyledons might be important for the plant only for a relatively short time, when the presence of herbivores and the visibility of the plant as a seedling may be relatively low. This would also be in agreement with the plant apparency hypothesis [63].

Our analyses show that total glucosinolate content in 4-, 8-, and 12-week-old plants varies, but not in the same linear manner as oviposition preference by P. xylostella varies among plants of similar ages as described by Badenes-Perez et al (2005). Non-linear ontogenetic changes in the content of defense compounds have been interpreted as part of a dynamic pattern, also affected by the development of herbivore tolerance and resource allocation constraints in the plant [59]. Besides plant glucosinolate content, the increase in leaf area and leaf number that occurs with plant age may affect oviposition preference by P. xylostella [52]. In B. vulgaris, the increase in number of leaves and total leaf area when comparing 6- and 12-week-old plants was 11.6 and 42.2 fold, respectively [52].

The simultaneous presence of high content of glucosinolates and saponins in small/young leaves of Barbarea, which are the most valuable for the plant, but also the most attractive to ovipositing P. xylostella, provides protection against this specialist herbivore. The association between glucosinolates and saponins could indicate that, from an evolutionary point of view, in Barbarea, saponins might have appeared after glucosinolates, enabling plants to be defended against insects that had adapted to glucosinolate-defended plants. Saponins would then be what has been called a “second line of defense”, appearing as a response to herbivores that have overcome the “first line of defense” provided by glucosinolates [25], [64].

Materials and Methods

Ethics Statement

Insects collected in Kenya were collected at Athi River, 40 km southeast of Nairobi, Kenya, in 2005 by Dr. Bernhard Löhr, and sent by him in July 2005 to MPICE in Jena under EU permit number EG-D-TH1-390390 AG39/2005. Insects collected in Spain were collected in 2013 in Arganda del Rey, Madrid, at the experimental farm “La Poveda”, which belongs to the Institute of Agricultural Sciences (CSIC). A permit was not required for the collection of insects at the collecting site in Spain.

Plant Growth and Insect Culture

Experiments were conducted in the laboratory at the Max Planck Institute for Chemical Ecology in Jena, Germany. Barbarea rupicola Moris, B. verna (Mill.) Asch., and B. vulgaris var. variegata seeds were purchased from B & T World Seeds (Aigues-Vives, France). Barbarea vulgaris var. arcuata G-type seeds were purchased from Rieger-Hofmann GmbH (Blaufelden-Raboldshausen, Germany) and P-type seeds were collected in Tissø (Denmark) and donated to us by Dr. Jens K. Nielsen. Seeds of NAS-type B. vulgaris were collected in The Netherlands and donated to us by Dr. Hanneke van Leur. The NAS-type of B. vulgaris was not classified varietally, although morphologically they would belong to what botanists consider var. arcuata or var. vulgaris [65]. Additional G-type Barbarea vulgaris var. arcuata seeds from Jena (Germany) were provided by Dr. Tamara Krügel. All plants used in the experiments were grown in pots in the greenhouse using a substrate of peat moss with clay. In the experiments testing the effect of different leaf size on glucosinolate and saponin content, plants were approximately 10 weeks old at the time when the experiments were conducted and they were grown in 20-cm-diameter pots. To compare differences in glucosinolate and saponin content between true leaves and cotyledons, 4-week-old plants were used and they were grown in 8-cm-diameter pots. Glucosinolate and saponin content was also compared among plants that were 4, 8, and 12 weeks old and grown in 15-cm-diameter pots. All plants used in the experiments were grown in the glasshouse at 22–28°C under 16 h supplemental light from Master Sun-T PIA Agro 400 or Master Sun-T PIA Plus 600 W Na lights (Philips, Turnhout, Belgium). P. xylostella used in the experiments were either collected in Kenya (provided by Dr. Bernhard Löhr) or collected in Spain. Insects were later reared on cabbage plants in a climate-controlled chamber (16∶8 h light:dark, 21±2°C and 55±5 RH).

Analysis of Glucosinolates and Saponins in Barbarea spp

Glucosinolate and saponin content in individual cotyledons, in individual true leaves, in foliage of whole plants, and in seeds was determined as in Badenes-Perez et al. (2010). Cotyledons and leaves were cut approximately from the middle of their petiole. In the experiment comparing cotyledons and true leaves, the first (largest) true leaf in 4-week-old plants was used in the analyses. Foliage of whole plants was harvested by cutting approximately half of the plant from the crown in the case of the comparison of foliage among plants 4, 8, and 12 weeks old. In the experiment testing whether changes in glucosinolate and saponin content occurred when only one cotyledon or one true leaf was left per plant, all the leaves of the plant, except either one cotyledon or the largest true leaf, were cut with scissors. The content of glucosinolates and saponins in the remaining cotyledon or true leaf was compared to a control of similar leaves in intact plants (having no leaves cut) 5 hours after removal of leaves. This experiment was arranged to assess whether an associated experiment conducted to test P. xylostella oviposition preference between true leaves and cotyledons could be influenced by glucosinolate and saponin induction at the time as a result of the mechanical removal of leaves. The time period of 5 hours to study changes in glucosinolate and saponin content in the plants was set to coincide with the time during which P. xylostella had laid most of its eggs in the course of the experiment [31], [66]. For glucosinolate and saponin analysis in Barbarea seeds, 20 mg of seeds were analyzed for each plant type and replicate. Glucosinolates and saponins were extracted with 80% aqueous methanol (methanol:water 80∶20, v:v). For glucosinolate determination, 4-hydroxybenzylglucosinolate was added as an internal standard. The methanolic extract was loaded onto DEAE Sephadex columns, followed by washing steps and by sulfatase treatment and elution of desulfoglucosinolates. Desulfoglucosinolates were separated on reversed-phase chromatography and quantified with a diode array detector at 229 nm (Agilent 1100 HPLC system, Agilent Technologies, Waldbronn, Germany), using a relative response factor of 2.0 for aliphatic and 0.5 for indole glucosinolates. The response factors we used were based on Brown et al. (2002). For saponin determination, the HPLC system indicated above was coupled to an ESI ion-trap mass spectrometer (Esquire 6000, Bruker Daltonics, Bremen, Germany) operated in negative mode in the range m/z 250–1700, with skimmer voltage, −40 eV; capillary exit voltage, −150.6 eV; capillary voltage, 4,000 V; nebulizer pressure, 35 psi; drying gas, 10 l min-1; and gas temperature, 330°C. Saponins were quantified by the peak areas for the signal of the molecular ion in the negative-ion mass spectrum [M-H] and in some of the analyses we used a standard curve created with an isolated standard of saponin 2.

Oviposition Preference Tests between Leaves of Different Size within the Same Plant

Oviposition preference between leaves of different type within the same plant was assessed in plexiglass tubes 3.0 cm (inner diameter) by 10.0 cm (length) with plants of G-type B. vulgaris var. arcuata. Each tube had a 0.5-cm-diameter hole in the middle, through which a piece of dental wick soaked with a 10% sugar solution was inserted into the tube as a food source for the moth inside. One mated female moth was placed in each tube, where it was offered two 7.1 cm2 circular disks of the abaxial side of B. vulgaris leaves. For each tube, the ends of a single tube were attached to two different leaves in the same B. vulgaris plant with the help of rubber bands and parafilm. The leaves compared had a difference in maximum leaf diameter ranging from 0 to 58 mm. A total of 42 comparisons involving 84 leaves were conducted (besides these 84 leaves, 16 additional leaves in which P. xylostella had not laid any eggs were taken to have more data points to analyze the relationship between leaf size and glucosinolate and saponin content). After one day, the number of eggs on each plant was counted in the laboratory using a dissecting microscope. The leaves used in the oviposition preference experiments were photographed with a digital camera and leaf areas were determined using WinFOLIA leaf area analysis software (Regent Instruments Inc., Quebec, Canada).

Oviposition preference experiments were also conducted with cotyledons and true leaves of the same plant for which all other leaves had been removed by cutting them with scissors. Immediately after cutting the leaves, oviposition preference tests were conducted in 32.5×32.5×32.5 cm polyester cages with 96×26 mesh (MegaView Science Education Services Co., Ltd., Taichung, Taiwan). Multiple cages were used, each of which was considered a replicate. One mated female moth was released in each experimental arena containing one Barbarea plant with only one true leaf (the largest) and one cotyledon. The experiment was replicated four times for each comparison. A small plastic cup with a 10% sugar solution on cotton was placed in the middle of the cage to provide a food source for the moths. Moths were allowed to oviposit overnight in the darkness from 19∶00 to 7∶00 h. P. xylostella lays most of its eggs during the first 3 h of scotophase and the peak oviposition occurs between 19∶00 and 20∶00 h [31], [66]. The number of eggs on each plant was counted in the laboratory using a dissecting microscope.

Survival of Larvae on True Leaves and Cotyledons within the Same Plant

Survival of first-instar larvae of P. xylostella was monitored over a period of 5 days. Using a brush, one P. xylostella larva was placed individually on a plant containing either one true leaf or one cotyledon (five plants and five larvae were used in total per treatment). The plants tested were B. rupicola, B. verna, G- and P-type B. vulgaris var. arcuata, B. vulgaris var. variegata, and NAS-type B. vulgaris.

4.4. Statistical Analysis

Differences in eggs laid by P. xylostella per leaf area and in glucosinolate and saponin content among Barbarea leaves of different size were analyzed using analysis of variance (ANOVA) and simple regressions with SPSS. When significant treatment differences were indicated by a significant F-test at P≤0.05, means were separated by Fisher’s Protected least significant difference (LSD). Differences in P. xylostella oviposition preference between cotyledons and true leaves were analyzed with a paired t-test with SPSS. In order to normalize the residuals, data were transformed prior to analysis using a natural log (x+1) function. Although all tests of significance were based on the transformed data, only untransformed data are presented.

Supporting Information

Table S1.

Mean ± SE glucosinolates (µmol/g of leaf fresh weight) concentrations in cotyledons and true leaves in Barbarea plants six hours after removing the rest of the leaves in the plant or leaving them intact. As true leaf, the largest true leaf of the plant was taken. For each plant and leaf type and treatment n = 5.

https://doi.org/10.1371/journal.pone.0095766.s001

(PDF)

Table S2.

Mean ± SE 3-0-β-cellobiosylhederagenin (saponin 1) and 3-0-β-cellobiosyloleanolic acid (saponin 2) in cotyledons and true leaves of Barbarea plants six hours after removing the rest of the leaves in the plant or leaving them intact. As true leaf, the largest true leaf of the plant was taken. For each plant and leaf type and treatment n = 5. Saponin concentrations given as µmol/g of leaf fresh weight.

https://doi.org/10.1371/journal.pone.0095766.s002

(PDF)

Acknowledgments

We thank Dr. Michael Reichelt for help with HPLC and LC-MS analysis of glucosinolates and saponins. Thanks to Jutta Steffen and Christin Heinrich for insect rearing and technical assistance during the experiments. Thanks to Drs. Jens K. Nielsen, Niels Agerbirk, Hanneke van Leur, and Tamara Krügel for providing B. vulgaris seeds. Thanks to Drs. Niels Agerbirk and Tetsuro Shinoda for providing saponin standards. Thanks to Dr. Bernhard Löhr for providing insects. Thanks to Birgit Hohmann and Beatriz Parrado Márquez for help growing plants.

Author Contributions

Conceived and designed the experiments: FRBP JG DGH. Performed the experiments: FRBP. Analyzed the data: FRBP. Contributed reagents/materials/analysis tools: FRBP JG DGH. Wrote the paper: FRBP. Provided comments on the manuscript: JG DGH.

References

  1. 1. Rhoades DF (1979) Evolution of plant chemical defense against herbivores. In: Rosenthal GA, Janzen DH, editors. Herbivores: Their Interactions with Secondary Plant Metabolites. New York, USA: Academic Press. 3–54.
  2. 2. McKey D (1974) Adaptive patterns in alkaloid physiology. American Naturalist 108: 305–320.
  3. 3. Harper JL (1989) The value of a leaf. Oecologia 80: 53–58.
  4. 4. Brown PD, Tokuhisa JG, Reichelt M, Gershenzon J (2003) Variation of glucosinolate accumulation among different organs and developmental stages of Arabidopsis thaliana. Phytochemistry 62: 471–481.
  5. 5. van Dam NM, de Jong TJ, Iwasa Y, Kubo T (1996) Optimal distribution of defences: are plants smart investors? Functional Ecology 10: 128–136.
  6. 6. Halkier BA, Gershenzon J (2006) Biology and biochemistry of glucosinolates. Annual Review of Plant Biology 57: 303–333.
  7. 7. Hopkins RJ, van Dam NM, van Loon JJA (2009) Role of glucosinolates in insect-plant relationships and multitrophic interactions. Annual Review of Entomology 54: 57–83.
  8. 8. Petersen B, Chen S, Hansen C, Olsen C, Halkier B (2002) Composition and content of glucosinolates in developing Arabidopsis thaliana. Planta 214: 562–571.
  9. 9. Gutbrodt B, Dorn S, Unsicker S, Mody K (2012) Species-specific responses of herbivores to within-plant and environmentally mediated between-plant variability in plant chemistry. Chemoecology 22: 101–111.
  10. 10. Lambdon PW, Hassall M, Boar RR, Mithen R (2003) Asynchrony in the nitrogen and glucosinolate leaf-age profiles of Brassica: is this a defensive strategy against generalist herbivores? Agriculture, Ecosystems & Environment 97: 205–214.
  11. 11. Shelton AL (2005) Within-plant variation in glucosinolate concentrations of Raphanus sativus across multiple scales. Journal of Chemical Ecology 31: 1711–1732.
  12. 12. Wentzell AM, Kliebenstein DJ (2008) Genotype, age, tissue, and environment regulate the structural outcome of glucosinolate activation. Plant Physiology 147: 415–428.
  13. 13. Osbourn A (1996) Saponins and plant defence – a soap story. Trends in Plant Science 1: 4–9.
  14. 14. Francis G, Kerem Z, Makkar HPS, Becker K (2002) The biological action of saponins in animal systems: a review. British Journal of Nutrition 88: 587–605.
  15. 15. Sparg SG, Light ME, van Staden J (2004) Biological activities and distribution of plant saponins. Journal of Ethnopharmacology 94: 219–243.
  16. 16. Matsuda K, Kaneko M, Kusaka K, Shishido T, Tamaki Y (1999) Soyasaponins as feeding stimulants to the oriental clouded yellow larva, Colias erate poliographus (Lepidoptera: Pieridae). Applied Entomology and Zoology 33: 255–258.
  17. 17. Adel M, Sehnal F, Jurzysta M (2000) Effects of alfalfa saponins on the moth Spodoptera littoralis. Journal of Chemical Ecology 26: 1065–1078.
  18. 18. Jain DC, Tripathi AK (1991) Insect feeding-deterrent activity of some saponin glycosides. Phytotherapy Research 5: 139–141.
  19. 19. De Geyter E, Swevers L, Caccia S, Geelen D, Smagghe G (2012) Saponins show high entomotoxicity by cell membrane permeation in Lepidoptera. Pest Management Science 68: 1199–1205.
  20. 20. Potter D, Kimmerer T (1989) Inhibition of herbivory on young holly leaves: evidence for the defensive role of saponins. Oecologia 78: 322–329.
  21. 21. Pecetti L, Biazzi E, Tava A (2010) Variation in saponin content during the growing season of spotted medic Medicago arabica (L.) Huds. Journal of the Science of Food and Agriculture 90: 2405–2410.
  22. 22. Szakiel A, Pączkowski C, Henry M (2011) Influence of environmental abiotic factors on the content of saponins in plants. Phytochemistry Reviews 10: 471–491.
  23. 23. Teng H-M, Fang M-F, Cai X, Hu Z-H (2009) Localization and dynamic change of saponin in vegetative organs of Polygala tenuifolia. Journal of Integrative Plant Biology 51: 529–536.
  24. 24. Agerbirk N, Olsen CE, Bibby BM, Frandsen HO, Brown LD, et al. (2003) A saponin correlated with variable resistance of Barbarea vulgaris to the diamondback moth Plutella xylostella. Journal of Chemical Ecology 29: 1417–1433.
  25. 25. Shinoda T, Nagao T, Nakayama M, Serizawa H, Koshioka M, et al. (2002) Identification of a triterpenoid saponin from a crucifer, Barbarea vulgaris, as a feeding deterrent to the diamondback moth, Plutella xylostella. Journal of Chemical Ecology 28: 587–599.
  26. 26. Nielsen JK, Nagao T, Okabe H, Shinoda T (2010) Resistance in the plant, Barbarea vulgaris, and counter-adaptations in flea beetles mediated by saponins. Journal of Chemical Ecology 36: 277–285.
  27. 27. Nielsen NJ, Nielsen J, Staerk D (2010) New resistance-correlated saponins from the insect-resistant crucifer Barbarea vulgaris. Journal of Agricultural and Food Chemistry 58: 5509–5514.
  28. 28. Badenes-Perez FR, Reichelt M, Gershenzon J, Heckel DG (2014) Using plant chemistry and insect preference to study the potential of Barbarea (Brassicaceae) as a dead-end trap crop for diamondback moth (Lepidoptera: Plutellidae). Phytochemistry: 137–144.
  29. 29. Nielsen JK, de Jong PW (2005) Temporal and host-related variation in frequencies of genes that enable Phyllotreta nemorum to utilize a novel host plant, Barbarea vulgaris. Entomologia Experimentalis Et Applicata 115: 265–270.
  30. 30. Agerbirk N, Olsen CE, Nielsen JK (2001) Seasonal variation in leaf glucosinolates and insect resistance in two types of Barbarea vulgaris ssp. arcuata. Phytochemistry 58: 91–100.
  31. 31. Talekar NS, Shelton AM (1993) Biology, ecology, and management of the diamondback moth. Annual Review of Entomology 38: 275–301.
  32. 32. Furlong MJ, Wright DJ, Dosdall LM (2013) Diamondback moth ecology and management: problems, progress, and prospects. Annual Review of Entomology 58: 517–541.
  33. 33. Ratzka A, Vogel H, Kliebenstein DJ, Mitchell-Olds T, Kroymann J (2002) Disarming the mustard oil bomb. Proceedings of the National Academy of Sciences of the United States of America 99: 11223–11228.
  34. 34. Badenes-Perez FR, Reichelt M, Gershenzon J, Heckel DG (2011) Phylloplane location of glucosinolates in Barbarea spp. (Brassicaceae) and misleading assessment of host suitability by a specialist herbivore. New Phytologist 189: 549–556.
  35. 35. van Loon JJA, Wang CZ, Nielsen JK, Gols R, Qiu YT (2002) Flavonoids from cabbage are feeding stimulants for diamondback moth larvae additional to glucosinolates: chemoreception and behaviour. Entomologia Experimentalis et Applicata 104: 27–34.
  36. 36. Spencer JL, Pillai S, Bernays EA (1999) Synergism in the oviposition behavior of Plutella xylostella: sinigrin and wax compounds. Journal of Insect Behavior 12: 483–500.
  37. 37. Renwick JAA, Haribal M, Gouinguené S, Stadler E (2006) Isothiocyanates stimulating oviposition by the diamondback moth, Plutella xylostella. Journal of Chemical Ecology 32: 755–766.
  38. 38. Badenes-Perez FR, Reichelt M, Heckel DG (2010) Can sulfur fertilisation increase the effectiveness of trap crops for diamondback moth, Plutella xylostella (L.) (Lepidoptera: Plutellidae)? Pest Management Science 66: 832–838.
  39. 39. Idris AB, Grafius E (1996) Effects of wild and cultivated host plants on oviposition, survival, and development of diamondback moth (Lepidoptera: Plutellidae) and its parasitoid Diadegma insulare (Hymenoptera: Ichneumonidae). Environmental Entomology 25: 825–833.
  40. 40. Shelton AM, Nault BA (2004) Dead-end trap cropping: a technique to improve management of the diamondback moth, Plutella xylostella (Lepidoptera: Plutellidae). Crop Protection 23: 497–503.
  41. 41. Lu JH, Liu SS, Shelton AM (2004) Laboratory evaluations of a wild crucifer Barbarea vulgaris as a management tool for the diamondback moth Plutella xylostella (Lepidoptera : Plutellidae). Bulletin of Entomological Research 94: 509–516.
  42. 42. Badenes-Perez FR, Shelton AM, Nault BA (2004) Evaluating trap crops for diamondback moth, Plutella xylostella (Lepidoptera : Plutellidae). Journal of Economic Entomology 97: 1365–1372.
  43. 43. Shelton AM, Badenes-Perez FR (2006) Concepts and applications of trap cropping in pest management. Annual Review of Entomology 51: 285–308.
  44. 44. Badenes-Perez FR, Nault BA, Shelton AM (2006) Dynamics of diamondback moth oviposition in the presence of a highly preferred non-suitable host. Entomologia Experimentalis et Applicata 120: 23–31.
  45. 45. Raupp MJ, Denno RF (1983) Leaf age as a predictor of herbivore distribution and abudance. In: Denno RF, McClure MS, editors. Variable plants and herbivores in natural and managed systems. New York, NY: Academic Press. 91–124.
  46. 46. Raupp MJ, Werren JH, Sadof CS (1988) Effects of short-term phenological changes in leaf suitability on the survivorship, growth, and development of gypsy moth (Lepidoptera: Lymantriidae) larvae. Environmental Entomology 17: 316–319.
  47. 47. Kause A, Ossipov V, Haukioja E, Lempa K, Hanhimäki S, et al. (1999) Multiplicity of biochemical factors determining quality of growing birch leaves. Oecologia 120: 102–112.
  48. 48. King BH, Crowe ML, Blackmore MD (1998) Effects of leaf age on oviposition and on offspring fitness in the imported willow leaf beetle Plagiodera versicolora (Coleoptera: Chrysomelidae). Journal of Insect Behavior 11: 23–36.
  49. 49. Karban R (1990) Herbivore outbreaks on only young trees: testing hypotheses about aging and induced resistance. Oikos 59: 27–32.
  50. 50. Heisswolf A, Obermaier E, Poethke HJ (2005) Selection of large host plants for oviposition by a monophagous leaf beetle: nutritional quality or enemy-free space? Ecological Entomology 30: 299–306.
  51. 51. Spangler SM, Calvin DD (2000) Influence of sweet corn growth stages on European corn borer (Lepidoptera: Crambidae) oviposition. Environmental Entomology 29: 1226–1235.
  52. 52. Badenes-Perez FR, Nault BA, Shelton AM (2005) Manipulating the attractiveness and suitability of hosts for diamondback moth (Lepidoptera : Plutellidae). Journal of Economic Entomology 98: 836–844.
  53. 53. Städler E, Baur R, De Jong R (2002) Sensory basis of host-plant selection: in earch of the “fingerprints” related to oviposition of the cabbage root fly. Acta Zoologica Academiae Scientiarum Hungaricae 48 (Suppl. 1)265–280.
  54. 54. Renwick JAA (1989) Chemical ecology of oviposition in phytophagous insects. Experientia 45: 223–228.
  55. 55. Janz N (2002) Evolutionary ecology and oviposition strategies. In: Hilker M, Meiners T, editors. Chemoecology of Insect Eggs and Egg Deposition. Oxford, UK: Blackwell Publishing Ltd. 349–376.
  56. 56. Larsson S, Ekbom B (1995) Oviposition mistakes in herbivorous insects: confusion or a step towards a new host plant? Oikos 72: 155–160.
  57. 57. Thompson JN (1999) What we know and do not know about coevolution: insect herbivores and plants as a test case. In: Olff H, Brown VK, Drent RH, editors. Herbivores: Between Plants and Predators. Oxford, UK: Blackwell Science Ltd. 7–30.
  58. 58. Zhang P-J, Lu Y-b, Zalucki M, Liu S-S (2012) Relationship between adult oviposition preference and larval performance of the diamondback moth, Plutella xylostella. Journal of Pest Science 85: 247–252.
  59. 59. Boege K, Marquis RJ (2005) Facing herbivory as you grow up: the ontogeny of resistance in plants. Trends in Ecology & Evolution 20: 441–448.
  60. 60. Wallace SK, Eigenbrode SD (2002) Changes in the glucosinolate–myrosinase defense system in Brassica juncea cotyledons during seedling development. Journal of Chemical Ecology 28: 243–256.
  61. 61. Uematsu H (1996) Inter-leaf movement of larvae of diamondback moth, Plutella xylostella L. (Lepidoptera: Yponomeutidae) on rape (Brassica napus) seedlings. Japanese Journal of Applied Entomology and Zoology 40: 35–38.
  62. 62. Root RB, Tahvanainen J (1969) Role of winter cress, Barbarea vulgaris, as a temporary host in seasonal development of crucifer fauna. Annals of the Entomological Society of America 62: 852–855.
  63. 63. Feeny P (1976) Plant apparency and chemical defense. In: Wallace J, Mansell R, editors. Biochemical Interaction Between Plants and Insects: Springer US. 1–40.
  64. 64. Feeny P (1977) Defensive Ecology of the Cruciferae. Annals of the Missouri Botanical Garden 64: 221–234.
  65. 65. van Leur H, Raaijmakers CE, van Dam NM (2006) A heritable glucosinolate polymorphism within natural populations of Barbarea vulgaris. Phytochemistry 67: 1214–1223.
  66. 66. Pivnick KA, Jarvis BJ, Gillott C, Slater GP, Underhill EW (1990) Daily patterns of reproductive activity and the influence of adult density and exposure to host plants on reproduction in the diamondback moth (Lepidoptera: Plutellidae). Environmental Entomology 19: 587–593.