Susceptibility of kale cultivars to the wheat bug, Nysius huttoni (Hemiptera: Lygaeidae) in New Zealand

ABSTRACT

Seedlings of kale cultivars in New Zealand are highly susceptible to direct feeding by the wheat bug Nysius huttoni (Hemiptera: Lygaeidae), an endemic insect pest.

Two assays (choice and no-choice) were conducted to compare the relative susceptibility of seedlings of the six most popular kale cultivars in New Zealand (Kestrel, Gruner, Sovereign, Regal, Corka and Coleor). The earliest occurrence of feeding damage in the choice assay was on cv. Kestrel, significantly earlier than on Corka and Gruner. In the no-choice assay, significantly more wheat bugs were found on Kestrel than on Corka. Damage to Kestrel occurred significantly earlier than on all the other cultivars except Corka. Reduction in plant dry weight was significantly higher on Coleor and Kestrel. These results are important for developing integrated pest management protocols for kale pests.

KEYWORDS:

• Integrated pest management
• wheat bug
• Nysius huttoni
• susceptibility cultivar
• Kestrel
• Coleor

Introduction

The wheat bug, Nysius huttoni White 1878 (Hemiptera: Lygaeidae), is an endemic New Zealand pest (Eyles 1960b) and is widely distributed in both the North and South Islands (Myers 1926; Eyles 1960a; Eyles and Ashlock 1969). Three other Nysius species belonging to the family Lygaeidae (Hemiptera) are recorded in New Zealand. Nysius convexus Usinger 1942 and N. liliputanus Eyles and Ashlock 1969 are endemic, whereas N. caledoniae Distant 1920 is an adventive species from Australia (Gurr 1957; Eyles and Ashlock 1969). Nysius huttoni has a wide host range comprising almost all brassicas, other cultivated crops and a wide variety of weeds, but N. convexus and N. liliputanus are recorded only in moss habitats.

Kale (Brassica oleracea L.), rape (Brassica napus L. var. napus), turnip (Brassica campestris L.) and swede (Brassica napus L. var. napo-brassica) are widely grown brassica crops for animal production systems in New Zealand (PGG 2009; Speciality Seeds 2016). Forage brassicas are normally drilled in summer (November to December) in New Zealand (PGG 2009) at which time the bug’s populations are at peak levels (Wei 2001). In New Zealand, about 400,000 ha of brassicas are grown annually (Horrocks et al. 2018). Forage brassicas have a high feeding value for ruminants, grow rapidly and have a high dry matter content (Speciality Seeds 2016). The wheat bug is a threat to 4- to 6-week-old brassica seedlings and plant populations can be reduced as a result of feeding damage at the base of the plants (Eyles 1965), causing a cankerous growth of tissue that can kill them or make them susceptible to breakage from wind and stock movement (AgPest 2016). There has been up to 90% damage in brassica crops in severe situations (AgPest 2016; Speciality Seeds 2016).

Forage brassicas may require several insecticide sprays/season to prevent damage (PGG 2009). Management of this pest in New Zealand relies on seed treatment with neonicotinoids and spraying with chlorpyrifos and permethrin insecticides (Goldson et al. 2015; Young 2018). All these pesticides are broad spectrum in nature (AgPest 2016; Horrocks et al. 2018). Neonicotinoids are systemic and traces of them have been found in pollen and nectar, which impact pollinators including wild bees (Goulson et al. 2008; Pook and Gritcan 2017), bumble bees (Whitehorn et al. 2012) and many other insect pollinators (Godfray et al. 2014). Hence, these pesticides are under increasing environmental pressure in Europe and many other countries (Cressey 2017; Woodcock et al. 2018). Furthermore, concern exists about non-target effects on insect natural enemies (Goulson et al. 2008; Goulson 2013), as well as birds and many fish species (Gibbons et al. 2015). Hence, these pesticides need to be replaed by environmentally friendly and an agro-ecological pest management strategies.

There is increasing pressure from consumers, media and governments to reduce pesticide use, but no practical alternatives are currently being offered to manage the wheat bug in New Zealand. It is therefore necessary to develop a cost-effective and sustainable method of pest management. Encouraging farmers to use integrated pest management (IPM) strategies that combine biological, cultural and chemical approaches in a compatible way could be one strategy to reduce pest damage and pesticide use in forage brassicas (Horrocks et al. 2018). Pest resistant/tolerant cultivars are vital component of many IPM programmes. The present study was undertaken to screen kale cultivars on the basis of susceptibility to wheat bugs and the results could guide the development of a future IPM programme for kale.

Materials and methods

Wheat bug collection and identification

Adult wheat bugs were collected from the weed shepherd’s purse Capsella bursa-pastoris (L.) Medik. (Brassicaceae) from the Iversen Field Plant Science Research Unit (43° 38′ 50.4″ S, 172° 27′ 29.9″ E) at Lincoln University in the spring of 2016. The collected bugs were preserved in 1.7 ml microtubes (MCT-175-C) that contained 70% ethanol and delivered to a Hemiptera taxonomist (Dr Marie-Claude Larivière, Landcare Research, Auckland, New Zealand) for confirmation and preserved in the Lincoln University Entomology Research Museum. After the confirmation of the species, N. huttoni were collected from the same field (see above) and used in laboratory rearing.

Wheat bug cultures

A wheat bug breeding colony was established following methods based on those of Burgess and Weegar (1986) and He and Wang (2000). Field-collected wheat bugs were released inside a transparent rectangular plastic container (29 × 19 × 10 cm) that contained a mesh-covered lid for air circulation. Fifty mating pairs were transferred to individual 50 ml polypropylene centrifuge tubes (11.0 cm length × 2.0 cm diameter) using a fine-hair brush. Food for Nysius consisted of fruiting twin cress (Lepidium didymum L.) (Brassicaceae) and hulled organic sunflower seeds (Helianthus annuus L.) (Asteraceae) (BioGro organic certified) which were replaced daily. Males were removed from the pair when the female began to lay eggs on the moistened cotton dental roll (10 mm × 38 mm) that was also included (Yang and Wang 2004). The tubes were checked daily and freshly laid eggs were removed using the brush and transferred to Petri dishes (5 cm diameter). The newly emerged nymphs along with the cotton dental rolls were transferred to another Petri dish (14 cm diameter) in which partially moistened filter papers had been placed. The colony was maintained in a controlled temperature (CT) room at the Bio-Protection Research Centre (http://bioprotection.org.nz), Lincoln University, New Zealand. The ambient temperature, humidity and photoperiod were 23°C with a 4°C range, 65% relative humidity and 16L: 8D.

Plant selection and cultivation

The uncoated seeds of the six most commonly used kale cultivars (Table 1) were obtained from PGG Wrightson (http://www.pggwrightsonseeds.com) and Speciality Seeds (http://www.specseed.co.nz/). Their growth habits are described in Table 1.

Table 1. Phenological characteristics of the six most popular kale cultivars in New Zealand (data from www.pggwrightsonseeds.com and www.specseed.co.nz).

SN	Cultivar	Height	Phenological characteristics
1	Kestrel KE35 TC	Medium	High leaf-to-stem ratio, palatable and digestible thick stems
2	Coleor	Small-medium	High leaf-to-stem ratio, winter hardy and high yield potential
3	Sovereign SOV 27 AC	Intermediate	High leaf-to-stem ratio and high yield potential
4	Regal KBG 01 AC	Intermediate	High leaf-to-stem ratio, winter hardy but low yield potential
5	Gruner	Giant	Tall and high yield potential
6	Corka	Intermediate	High leaf-to-stem ratio, palatable, winter hardy and high yield potential

The seedlings were grown in a glasshouse with a mean temperature of 22°C and relative humidity of 40%. Seeds were direct-seeded into pots containing a potting mix made by mixing 400 L composted bark, 100 L pumice (1.0–7.0 mm), 1500 g Osmocote (slow, 3- to 4-month release plant food), 500 g horticultural lime and 500 g HydraFLO (wetting agent, www.solutions4earth.com). The seedlings were 9 days old and approximately 3.5 cm high when used in the bioassays. Those grown in the glasshouse were transferred to a CT room with a temperature of 21°C with a 4°C range and a day length of 16 h.

Choice tests

Two seedlings of each kale cultivar were arranged in a circular fashion around the perimeter of a 23.0 cm diameter × 5.0 cm depth pot with cultivars arranged approximately 5.0 cm apart and 5.0 cm from the pot wall. Each kale cultivar with wheat bugs (treatment) being compared with a control with no bugs, so that two adjacent cylindrical sleeves (/pot) served as a block, with a total of ten blocks. Each cultivar was ‘marked’ on the outer wall of the choice pots. The pots were enclosed in cylindrical sleeves made of flexible transparent PVC sheets (1 mm thickness). The dimension (diameter x height) of the cylinders were 23.5 cm × 14 cm. The tops of the sleeves were covered with fine white mesh and Fluon® (BioQuip, fluoropolymer resin, PTFE-30) was used on the inner surface of the sleeves to prevent Nysius from climbing.

The study comprised a randomised complete block design (RCBD), with ten replicates. Twenty adult wheat bugs of the same age were starved for 12 h, then introduced into the centre of each cylinder. The times to first settlement (mins) and first obvious feeding damage on a seedling for each pot were recorded. Feeding damage was assessed by recording the presence or absence of girdling of the stem and/or discoloration of the leaf. These were the most common damage symptoms along with leaf distortion, twisted leaf veins and petiole, and finally collapse of the seedlings. Then the number of bugs on seedlings within each cylinder were counted at different time intervals following introduction (0.5, 1, 2, 4, 8 12, 24, 48, 72, 96, 120, 144, 168, 192 and 216 h). At the conclusion of the assay the survival rate of the bugs was assessed.

No-choice tests

Two seedlings of each cultivar were grown per 6.5 cm diameter × 5.0 cm depth pot, giving a total of six treatments. There were twenty replicates of each treatment, ten with wheat bugs and ten as controls. The pots were covered with 7 cm × 12 cm cylindrical sleeves constructed as above and seven bugs were introduced into the treatment cylinders. Treatments were randomised and assessed as above. At the completion of the assay, the dry weights of the seedlings including roots were measured and the percentage weight change calculated.

Statistical analysis

The mean numbers recorded in each cultivar at different time intervals were integrated over the 216 h period by the area under the curve (AUC) method (Hanley and McNeil 1983). Time data (mins) obtained from the experiments were first normalised by using the log₁₀-transformation, and count data were normalised by using a square root transformation. The percentage reduction in plant dry weight (compared with the control) was not transformed. After normality checking, data were subjected to two-way (treatments and blocks) analysis of variance (ANOVA) and means were separated by unprotected least significance difference (LSD) at P < 0.05 (Saville 2015).

Results

Choice tests

For choice tests, settling time of the wheat bug on seedlings did not differ significantly between cultivars (Table 2). The time to first-feeding damage by the bugs across the kale cultivars varied significantly for the choice tests (Table 2). First-feeding damage occurred on Kestrel followed by Coleor, Sovereign and Regal, respectively, all of which were not significantly different from one another. However, feeding damage on Kestrel was significantly earlier than on Gruner and Corka.

Table 2. For the choice tests, mean time (Log₁₀ transformed) required for settling and first-feeding damage on different kale cultivars (n = 10). Back-transformed means are given in brackets.

Kale cultivars	Settling time (Log10 minutes ± SE)	First feeding damage (Log10 hours ± SE)
Kestrel	0.93 ± 0.106 (8.5) a	1.98 ± 0.038 (96.2) a
Coleor	1.13 ± 0.116 (13.2) a	1.99 ± 0.043 (99.3) ab
Sovereign	1.16 ± 0.170 (14.5) a	2.10 ± 0.063 (125.0) abc
Regal	1.30 ± 0.200 (19.9) a	2.12 ± 0.038 (131.2) abc
Gruner	1.16 ± 0.167 (14.5) a	2.12 ± 0.043 (131.5) bc
Corka	1.22 ± 0.159 (16.6) a	2.15 ± 0.045 (141.3) c
LSD (5%)	0.406	0.134

Means within a column with no letters in common are significantly different (Unprotected LSD; P < 0.05).

The number of wheat bugs on seedlings across kale cultivars over 216 h were not significantly different in choice tests (Figure 1). However, the largest number of bugs was recorded on Coleor followed by Gruner with the lowest on Sovereign. The mean survival rate in the choice test was about 53%, averaging 10 wheat bugs/cylinder.

Figure 1. Choice tests. Mean numbers (√ transformed) of adult wheat bugs recorded in each of six kale cultivars over 216 h. The vertical bar is the least significant difference, LSD (5%) (n = 10).

No-choice tests

Settling time of the wheat bug did not differ significantly between cultivars (Table 3). In no-choice tests, feeding damage was detected earliest on Kestrel followed by Corka, both of which were significantly more susceptible than Gruner, Sovereign and Regal. Coleor was the third earliest for feeding damage and differed significantly only from Kestrel (Table 3).

Table 3. For the no-choice tests, mean time (Log₁₀ transformed) required for settling and first feeding damage on different kale cultivars (n = 10). Back-transformed means are given in brackets.

Kale cultivars	Settling time (Log10 minutes ± SE)	First feeding damage (Log10 hours ± SE)
Kestrel	1.19 ± 0.036 (15.5) a	1.99 ± 0.028 (98.2) a
Coleor	1.29 ± 0.113 (19.3) a	2.08 ± 0.020 (120.5) bc
Sovereign	1.27 ± 0.137 (18.6) a	2.12 ± 0.037 (132.4) c
Regal	1.38 ± 0.126 (24.2) a	2.13 ± 0.022 (136.1) c
Gruner	1.21 ± 0.072 (16.1) a	2.11 ± 0.031 (130.0) c
Corka	1.33 ± 0.131 (21.2) a	2.01 ± 0.018 (102.3) ab
LSD (5%)	0.300	0.074

Means within a column with no letters in common are significantly different (Unprotected LSD; P < 0.05).

In the no-choice tests, the number of wheat bugs observed on Kestrel seedlings was significantly higher than on Corka but not significantly different from Gruner, Sovereign, Coleor or Regal (Figure 2). Kestrel, Gruner, Sovereign, Regal and Coleor were not significantly different from each other.

Figure 2. No-choice tests. Mean numbers (√ transformed) of adult wheat bugs recorded in each of six kale cultivars over 216 h. Means with no letters in common are significantly different (Unprotected LSD; P < 0.05). The vertical bar is the least significant difference, LSD (5%) (n = 10).

Survival rate was low, averaging between one and two bugs/cylinder. The highest survival rate occurred on Coleor, followed by Kestrel and Gruner, all of which were not significantly different from one another. Also, the survival on Kestrel, Gruner, Sovereign, Regal and Corka did not differ significantly. Furthermore, the survival rate on the latter three cultivars was significantly lower than on Coleor (Figure 3).

Figure 3. No-choice tests. Mean survival (%) of N. huttoni adults on six kale cultivars at 216 h. Means with no letters in common are significantly different (Unprotected LSD; P < 0.05). The vertical bar is the least significant difference, LSD (5%) (n = 10).

Seedling dry weight reduction by the bug, compared with controls was significantly higher in Kestrel and Coleor than on the other four cultivars. The lowest reduction was recorded on Corka which was not significantly different from that on Sovereign, Gruner and Regal, respectively (Figure 4).

Figure 4. Mean percentage dry weight reduction over control from adult wheat bugs on six kale cultivars in no-choice tests. Means with no letters in common are significantly different (Unprotected LSD; P < 0.05). The vertical bar is the least significant difference, LSD (5%) (n = 10).

Discussion

Kale is an important forage crop for ruminants (cattle and sheep), being drilled during the summer for winter feeding in New Zealand (Speciality Seeds 2016). Damage from wheat bugs is obvious during the seedling stage of that crop (PGG 2009; AgPest 2016). The aim of this work was to examine the susceptibility of this bug on a range of commercial kale cultivars. The results confirmed that, in order of preference, wheat bugs favoured the kale cultivars Kestrel, Coleor and Gruner over Sovereign, Corka and Regal (Tables 2 and 3, Figure 2). Significantly higher survival of the bug was recorded on Coleor and Kestrel than on Regal, Corka and Sovereign, respectively. The bug’s preference for Kestrel and Coleor could be partly caused by various cultivar characteristics such as digestibility, palatability, leaf to stem ratio, growth vigour, and concentrations of S-methyl cysteine sulphoxide (SMCO) compared with the other cultivars (PGG 2009). Damage to the bugs’ favoured cultivars can later lead to the death of the seedlings and further reduce their seedling number per unit area in brassica fields. Hence, this pest sometimes called a crop establishment pest (AgPest 2016). The cultivars growth rate in both choice and no-choice tests were similar. However, the first obvious damage was noticed on Kestrel in both choice and no-choice tests. This could be the result of the more prompt settling and higher numbers of N. huttoni on Kestrel. Damage was slowest to appear on Corka in choice tests, and on Regal in no-choice tests. However, high mortality (70%–80%) of the bugs was recorded on all the cultivars, perhaps due to the limited availability of food in these experiments (Wei 2001). The greatest reduction in plant dry weight occurred on Kestrel and Coleor. Higher numbers of bugs settled over time with a high survival rate on these cultivars (Figures 2–4). Gruner was the medium category of cultivar in terms of preference by the bugs (Figures 2 and 3). Although these cultivar rankings imply that Kestrel and Coleor could be avoided by growers, other more important agronomic factors such as yield and diseases resistance can be the main criteria for cultivar selection. For example, past studies on forage brassicas have mostly focused on varietal screening for resistance to clubroot disease and other aspects of varietal improvements (Asrat et al. 2010; Bradshaw and Wilson 2012) but not resistance to the wheat bug. However, there is evidence that disease resistance in brassicas may be negatively correlated with insect resistance (Rostás and Hilker 2002).

The bugs’ preference for some cultivars may be affected by cues comprising volatile plant chemicals or by visual cues (Finch and Collier 2000). Among plant chemicals, glucosinolates have been widely studied in crucifers and they can have feeding deterrent or stimulatory properties on generalist or specialist insects, respectively (Renwick 2002). Further, the variation in glucosinolate profile between cultivars can also affect the host-plant preference (Poelman et al. 2009). However, the chemical basis of resistance to wheat bugs on forage brassicas is not known.

Global agriculture is beginning to adopt ‘sustainable intensification’ approaches (Pretty et al. 2018). Reasons for this include insecticide resistance, along with a decline in the rate at which new insecticide molecules are developed (Nauen and Denholm 2005; Hawkins et al. 2018). There is also increasing consumer resistance to pesticides in some markets (Wollaeger et al. 2015).

An IPM strategy developed with farmer input has been suggested to reduce reliance on insecticides (Warner 2007; Horrocks et al. 2018). While the results in this study show cultivar differences in wheat bug susceptibility, further research is needed to investigate if the best cultivars in this study are also less susceptible to other potential insect pests of kale crops such as aphids, beetles and caterpillars. In the future, insecticide use in kale crops, and the potential environmental impacts, could be minimised by incorporating less susceptible cultivars, such as Corka or Regal, into an integrated pest management programme with other management tools such as biological control and the use of ‘soft’ chemicals (Dent 2000; Horrocks et al. 2018).

Acknowledgements

We are grateful to the New Zealand Ministry of Foreign Affairs and Trade (MFAT), Lincoln University, the Bio-Protection Research Centre and the Agriculture and Forestry University (AFU), Nepal, for their financial, technical and logistical help with this PhD study. Special thanks to Janine Johnson for editorial assistance and Dr Marie-Claude Larivière for help with taxonomic identification of N. huttoni. We also thank Specialty Seeds, and PGG Wrightson (Kimihia Research Centre) for providing seeds for the experiments.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by the Ministry of Foreign Affairs and Trade New Zealand (NZ Aid); the Bio-Protection Research Centre, Lincoln University, New Zealand; and the Agriculture and Forestry University, Nepal.