1. Introduction
Onion (common onion or bulb onion) (
Allium cepa L.) is one of the most important vegetable crops worldwide with a global total production of 93,226,400 tones and it accounts for about 24% of the world’s total vegetable production [
1]. Onion production and breeding encounter many challenges. Like other crops, onions are susceptible to insect, fungal, bacterial, viral, and nematode pests [
2,
3]. One of the main insect pests is the onion thrips (
Thrips tabaci Lind.; Thysanoptera: Thripidae). It is a phytophagous and polyphagous, invasive, cosmopolitan and highly fecundating insect pest with a rapid development rate, and vector of several onion pathogens and tospoviruses [4-8]. Attack by
T. tabaci not only leads to a complete loss of onion seedlings but also may cause damage to older crops by feeding on leaves as well as onion bulbs [
9]. The poor effectiveness of insecticides at controlling thrips – and thrips’ increased resistance to them – leads to high losses in the cultivation of onions [
10,
11]. A reduction in yield of about 40-65% has been reported due to attacks by these pests [
12,
13]. Full resistance (or tolerance) of onion to
T.
tabaci would be very beneficial, however, in the available literature, we have not found information about the existence of such cultivars and breeding lines of onion. However, in recent years, some plant characteristics that are responsible for the partial resistance of onions to
T. tabaci have been identified. The differences in resistant and susceptible cultivars have been associated with leaf color [
14], amounts and types of epicuticular leaf wax [
15,
16], plant architecture and anatomy [
17,
18] and total phenol content [
19,
20]. Recently, it was also possible to select some onions genotypes that showed a lower number of thrips compared to susceptible ones [
19,
21], and those produced large bulb yields under thrips pressure [
17,
22]. Onion breeding programs currently focus mainly on the improvement of existing cultivars; however, wild
Allium species possess many desirable traits such as disease and pest resistance [
23].
Allium roylei Stearn and
Allium fistulosum L. have been recognized as the most important gene pools of onion [
24,
25].
A. roylei is a wild species originating from the Indian subcontinent which possesses genes imparting resistance to various harmful fungal diseases and pests [26-29].
A. fistulosum (Welsh onion, Japanese bunching onion, spring onion) is widely cultivated in Japan, China and Korea. Its origin is unknown; it is believed to be of Chinese origin [
30]. This
Allium sp. also possesses many traits agronomically useful for onions, due to its resistance to onion leaf blight, pink root, anthracnose, and onion fly [
23,
31,
32].
Our previous field studies, in which
A. roylei and
A. fistulosum were less colonized and damaged by
T. tabaci compared to
A. cepa cultivars and breeding lines [
33], encouraged us to continue research on them. Therefore, this research under laboratory conditions was conducted to prove
T. tabaci settlement and oviposition activity on leaves of
A. roylei and
A. fistulosum compared to three cultivars of
A. cepa that differed in their susceptibility to colonization and feeding of thrips under field conditions [
20].
2. Materials and Methods
2.1. Test plants
Three onion species were used in the laboratory experiments: Allium cepa L. (cvs. Alibaba, Bila and, Tęcza.), Allium fistulosum (cv. Kroll), and Allium roylei (ecotypes 333). All the A. cepa and A. fistulosum cultivars used in the trials are recommended for cultivation in central Europe and are commercially available. The seeds were obtained from Polish breeding companies, namely PlantiCo Zielonki in Stare Babice (cvs. Alibaba, Bila, and Kroll) and Spójnia in Nochowo (cv. Tęcza). A. roylei seeds we obtained from the bank Plant Genetic Resources Laboratory, Research Institute of Vegetable Crops in Skierniewice in Poland. The onion plants used in the experiments were grown in a standard substrate in trays within a plant growing room at 24 ± 1°C, 35 ± 5% relative humidity and a photoperiod of 16:8 h light:dark. All plants were watered regularly, only with tap water. Leaves from onion plants that were approximately four weeks old were used in all bioassays.
2.2. Test insects
Using a rearing method adapted from Loomans & Murai (1997) [
34], a stock culture of a thelytokous
T. tabaci strain was maintained on white cabbage leaves in 0·75-liter glass jars covered with a fine mesh to ensure ventilation. The rearing was kept in a climate chamber at 24 ± 1°C, 35 ± 5% relative humidity and a photoperiod of 16:8 h light:dark. White cabbage was purchased regularly, and fresh pieces of leaf were added two to three times a week.
To obtain groups of females of known age, thrips pupae were randomly collected from the rearing jars and transferred to Petri dishes (diameter 90 mm) with sections of leek (Allium ampeloprasum L.) leaves. The dishes were closed with lids with central holes covered with a fine mesh to allow air circulation and were sealed with sealing film to prevent the thrips from escaping. After 48 h, adult females were checked, and any remaining pupae were removed. Following an additional pre-oviposition period of 48 hours, single females were used in the bioassays.
2.3. Oviposition rate
The oviposition rate of female onion thrips on Allium species/cultivars was evaluated using a no-choice test. For this bioassay, four-centimeter sections of onion leaves were cut from the middle part of the leaves of the respective test plant. To protect the leaf sections from desiccation and to prevent the thrips from getting inside the leaves, both ends of the leaf sections were briefly dipped in warm liquid paraffin wax. After the wax had solidified, the leaf sections were placed singly on a thin film of 1 % water agar (Agar – Agar, Kobe I, Carl Roth, Karlsruhe, Germany) in glass Petri dishes (60 mm diameter). Subsequently, single females of known age were transferred to each glass Petri dish and placed on the leaf section. To prevent the escape of the T. tabaci females, the dishes were covered with thin (14 µm) clear plastic film (Carl Roth, Karlsruhe, Germany), which was perforated (one hole per cm2 on average) using insect pins (0·4 mm diameter). The bioassay units were kept in a climate chamber at 24 ± 1°C, 35 ± 5% relative humidity and a photoperiod of 16:8 h light:dark. After 24 hours, the females were removed. The plant sections with eggs were kept in the climatic chamber for another five days. After that, the hatched larvae were counted under a stereoscopic microscope.
2.4. Settling preference
The settling preference of T. tabaci females for leaf sections of the different Allium species/cultivars was determined by means of a choice test. Eight pairs of species/cultivars were compared against each other: Tęcza × Alibaba; Tęcza × Bila; Tęcza × Kroll; Tęcza × A. roylei, Alibaba × Bila; Alibaba × Kroll; Bila × Kroll; and Kroll × A. roylei. Each pair consisted of two four-centimeter sections of onion leaves, sealed with wax at both ends (as described above). These sections were placed parallel and equidistant to each other in the center of the bottom of a 90 mm diameter glass Petri dish. Subsequently, ten females of unknown age were placed at the starting point in the center, between the parallel-lying onion leaf sections of the tested pair of cultivars. Each bioassay unit was covered with perforated plastic film and completely randomized under an artificial light source in a climate chamber at 24 ± 1°C and 35 ± 5% relative humidity. Ten, 30, 60, 120, and 180 minutes after the female thrips had dispersed from the starting point, the number of thrips were counted on each of the two sections of onion leaves of the tested cultivars, as well as in the surrounding space. The experiment was replicated 10 to 12 times resulting in preferences recorded for 100 to 120 female thrips for each pair of onion cultivars/species.
2.5. Statistical analysis
Analyses of the bioassay data were performed using Statistica 13 software (TIBCO Software Inc., 2017). The results of the settling preference tests were analysed with Student’s t-test with a significance level of P < 0.05. The data obtained from the oviposition tests with hatched thrips larvae were subjected to a one-way analysis of variance (ANOVA), with the factor onion cultivar/species. Residual plots were checked prior to data analysis. In cases where the data did not show a normal distribution, they were normalized by log(x+1) transformation. The Tukey test was used to compare means at a significance level of P < 0.05.
3. Results
In the no-choice experiments on the oviposition rate of female onion thrips, significant variability was observed among cultivars/species regarding the mean number of thrips larvae that had hatched from eggs laid inside the tissue of the onion leaves (F = 27.158; df = 4; p < 0.000).
T. tabaci females laid significantly more eggs on leaves of
A. cepa cultivars compared to
A roylei (almost three times more) (
Figure 1). Additionally, a significantly higher number of
T. tabaci larvae hatched from eggs laid on cv. Tęcza and Bila compared to
A. fistulosum (cv Kroll). Moreover, a significantly higher number of
T. tabaci larvae hatched from eggs on cv Kroll than on
A. roylei (
Figure 1).
In a preference test for onion thrips settling on different species and cultivars of
Allium sp., a significantly higher number of female onion thrips settled on leaf sections of
A. roylei compared to cv. Tęcza (
A. cepa) and cv. Kroll (
A. fistulosum) during each observation period (
Figure 2c, d). In both comparisons, almost twice as many
T. tabaci females were observed on leaf sections of
A. roylei. When
A. cepa cultivars were paired with cv. Kroll (
A. fistulosum), female
T. tabaci demonstrated a significant preference for
A. fistulosum over the cv. Alibaba throughout the test period and cv. Bila throughout the test period except for the first 10 minutes (
Figure 2 f, h). When common onion cultivars Alibaba and Bila were paired with cv. Tęcza, the settling preference of
T. tabaci females was different. A significantly higher number of thrips settled on cv. Alibaba compared to cv. Tęcza, whereas cv. Tęcza was preferred over Bila (
Figure 2 e, g). For the pairs Alibaba x Bila (
Figure 2 b) and Tęcza x Kroll (
Figure 2 a) there were no significant differences in onion thrips’ preference.
4. Discussion
Our experiments were short-term laboratory experiments that only provide information on thrips’ selection behavior at the first stage of colonization and acceptance of the plant for egg-laying. The observed differences in the numbers of hatched larvae indeed reflect differences in oviposition (numbers of eggs from which the larvae have hatched). Among the three onion species tested
(A. cepa, A. fistulosum, and
A. roylei), the most frequently colonized by
T. tabaci females was
A. roylei. While adult
T. tabaci did not respond to
A. cepa plant odor in a Y-tube olfactometer [
14], plant volatiles from
A. roylei leaf sections may have attracted
T. tabaci during the 3 h period in this study. Olfactometer tests using
T. tabaci and
A. roylei plants could help to clarify this possible attraction. On the contrary, in the oviposition bioassay without choice over a 24-hour period, four times fewer larvae hatched on the leaves of
A. roylei than on the common onion cv. Tęcza and almost two times fewer than on Welsh onion. It has been hypothesized by some authors that thrips and other herbivores females would prefer those leaves for oviposition on which they realize the highest reproductive success [35-37]. Lower fecundity of females on leaves of
A. roylei could therefore result from less favorable conditions for feeding and development of offspring. In our earlier field studies we noticed that despite the presence of
T. tabaci on the leaves of
A. roylei, no damage caused by thrips feeding on them was observed, in contrast to the injured leaves of
A. cepa and
A. fistulosum [
33]. The lower oviposition rate of females on
A. roylei leaves under laboratory conditions and the lack of damage to the leaves under field conditions – when
T. tabaci had a choice between many cultivars and the breeding lines of
A. cepa and
A. fistulosum – may suggest that this wild onion species has some traits of resistance to
T. tabaci which discourage thrips from feeding and egg laying.
A. roylei is a plant with narrow, drooping leaves which are sometimes described as filiform. Mature, full-sized leaves are fistulous, though smaller ones are sometimes solid. In turn,
A. fistulosum (cv. Kroll) leaves are wide, erect and pointed hollow tubes and are more similar to the cylindrical, fleshy, and hollow leaves of
A. cepa [
38] (authors’ observations). Differences between resistance to thrips of the tested onion species may be due to certain morphological features of their leaves, which make it difficult for females to lay eggs. They may concern the thickness and rigidity of the cellular wall, the amount of epicuticular waxes, the structure of the epidermis, and the number of stomates [
39,
40]. Post-alighting host acceptance, and also subsequent feeding and reproduction, are strongly influenced by the plant’s nutritional quality and defenses such as secondary metabolites [
41,
42]. Many bioactive metabolites like cysteine, sulfoxides, flavanols, polyphenols, and saponins are synthesized by different organs of
A. roylei to provide defense against a wide range of plant pathogens and herbivores [
43,
44]. Resistance against downy mildew (
Perenospora destructor (Berk.) Casp. ex Berk.) was identified in
A. roylei and successfully transferred to bulb onion [
28]. This wild onion is partially resistant to leaf blight disease caused by
Botrytis squamosa J. C. Walker and basal rot disease caused by
Fusarium oxysporum f. sp.
cepae [
29,
45]. Also,
A. roylei has been proven to be partially resistant to beet armyworm
Spodoptera exigua Hübner. Larval growth and survival of
S. exigua proved to be significantly slower on
A. roylei compared to the
A. cepa, A. fistulosum, and
A. galanthum Kar. et Kir. [
46]. The prospect of using
A. roylei as a source of resistance to
T. tabaci in onion breeding is promising, but further research is needed on how many larvae can complete development, how quickly they will develop, and what will be the final condition and fertility of the next generation of females.
In our bioassay,
A. fistulosum cv. Kroll was more frequently chosen for settling by female
T. tabaci than the two cultivars of
A. cepa i.e., Alibaba and Bila, but a significantly lower number of larvae hatched from eggs laid by females on leaves of
A. fistulosum compared with leaves of two cultivars of
A. cepa, i.e. Bila and Tęcza. In our field study, despite the very high number of thrips caught from the leaves of
A. fistulosum, minor damage was recorded on them [
33]. The leaves of
A. fistulosum were damaged by feeding thrips only in 6.5% of cases in 2015 and 1.5% of cases in 2016 (unpubl. data) while the corresponding figures for leaves of
A. cepa cvs. Alibaba, Bila, and Tęcza were, respectively, 13.5%, 13.9%, and 17.8% in 2015 and 5.3%, 6.8%, and 7.3% in 2016 [
20]. This was also confirmed by a field study by Hudák and Pénzes [
47], where
A. fistulosum showed less damage than
A. cepa under similar field conditions. It confirms that
T. tabaci has shown higher affinity toward
A. fistulosum phenotype during settlement and supports higher densities of thrips, but
A. cepa creates better conditions for foraging and laying eggs. Ren et al. [
48] proved that volatiles are important factors for thrips in host preference. In their study, one of the most attractive volatiles, along with
Medicago sativa, for
T. tabaci and
Frankliniella occidentalis Pergande appeared to be volatiles of
A. fistulosum in its vegetative and flowering stages. In turn, Jones et al. [
49] showed that the Nebuka type of
A. fistulosum has a similar low degree of thrips colonization as the resistant cv. White Persian of
A. cepa. The authors noted that the leaves of both
Allium sp. were circular, and they had a spreading growth habit and a long sheath region. Some authors report that
A. fistulosum can be used for improvement of the common onion, especially for its resistance to pink root (
Phoma terrestris E. M. Hans.) [
50], Fusarium basal rot,
T. tabaci, smut (
Urocystis cepulae Frost) [
51], and onion fly (
Hylemya antiqua Bouche) [
52]. Varietal resistance against
Liriomyza chinensis (Kato) has been reported in
A. fistulosum in Japan. Antibiosis studies revealed a significant difference in survival up to the pupal stage, in the forewing lengths of adults and in the development time from the egg to pupal stages among the resistant and susceptible varieties of
A. fistulosum [
53,
54].
A. fistulosum, as well as wild Allium species like
A. hookeri, A. altaicum, and
A. angulosum, are a rich source of lectins and these compounds have recently been proven to have insecticidal activity against
T. tabaci [
55]. The authors suggest that the high lectin content of
A. hookeri and
A. fistulosum can be correlated with the low amount of thrips damage. Whole plant and detached leaf damage tests revealed that
A. hookeri was resistant to
T. tabaci. However, worse development of this pest was observed not only on
A. hookeri, but also on
A. fistulosum. There are many local and commercial cultivars of
A. fistulosum with distinctive differences in morphological and other traits which are adapted to a variety of climatic conditions. The wide variety of
A. fistulosum phenotypes that exists around the world [
56] suggests that some of them will likely develop traits that will promote resistance or tolerance to
T. tabaci, so further research in this direction appears to be justified.
In a previous field study, cv. Tęcza was resistant to thrips abundance but susceptible to thrips feeding and was more heavily damaged than varieties susceptible to thrips infestation and foraging cv. Alibaba [
20]. Although the laboratory test did not show significant differences in the number of hatched
T. tabaci larvae between tested cultivars of
A. cepa, the highest number of them was recorded on cv. Tęcza. Perhaps the cv. Tęcza, which stimulated
T. tabaci individuals to feed more under field conditions, could also stimulate females to lay eggs more intensively during the bioassay test. In the settlement test in the Alibaba x Tęcza pairing, the most preference was exhibited by female onion thrips to the cv. Alibaba, but in pair of Bila x Tęcza, cv. Tęcza was preferred to cv. Bila. In a field study, Alibaba was also more populated by
T. tabaci than cv. Tęcza but cv. Tęcza appeared to be less attractive than cv. Bila [
20]. Laboratory tests do not always reflect the behavior of insects in the field, where they are influenced by many abiotic and biotic factors, and the results obtained must be interpreted with this in mind.