The plant bugs (Miridae) are the largest family of true bugs (Hemiptera: Heteroptera), comprising nearly 10 000 described species in approximately 1400 genera (Schuh 1995). Within the Miridae family there is enormous variation in diet and feeding behaviour, including species that are major crop pests as well as important predators of crop pests (Gross & Cassis 1991; Wheeler 2001). Several predatory mirids have been used as biological controls for pest control, while some omnivorous species have been considered both a pest and an important predator of other pest species, depending on conditions (Wheeler 2001). Many phytophagous mirids are globally important pests of crops such as cotton, lucerne, soybean, mungbean, strawberry, sorghum, cocoa, apples and tea, and these species show enormous variation in diet, ranging from monophagy to polyphagy (Wheeler 2001).
The green mirid, Creontiades dilutus (Stål) (Hemiptera: Miridae), is a polyphagous and predominantly phytophagous pest species, endemic to Australia (Malipatil & Cassis 1997; Khan 1999). Creontiades dilutus is considered a major pest of cotton, particularly in Bollgard II® cultivars which are genetically engineered to express the Cry1A(c) and Cry2A(b) δ-endotoxin proteins of Bacillus thuringiensis Berliner (c.v. kurstaki). They are also recognised as pests, to varying degrees, of lucerne, pulses and many fruit and vegetable crops. Creontiades dilutus is common and widely distributed within Australia, having been recorded from all states and territories (Malipatil & Cassis 1997). The majority of published work has focused on C. dilutus as a pest of cotton in Queensland and northern New South Wales. Miles (1995) conducted the first extensive study on C. dilutus in Australia. This work focused mainly on species identification, industry perception about the pest (through consultant questionnaire interviews) and sampling methods. Khan (1999) conducted a comprehensive study on some aspects of the biology and management of C. dilutus in Australian cotton. Prior to this, there had been little research attention directed towards C. dilutus, probably because they are difficult insects to study and have not always been considered a major agricultural pest (Khan 1999).
Although integrated pest management (IPM) is becoming a popular approach in the Australian cotton and grain industries, current control options for C. dilutus remain heavily reliant on the use of broad-spectrum insecticide sprays (Khan et al. 2004b; Whitehouse 2007; Brier et al. 2008). These chemicals are known to have negative impacts on beneficial invertebrate species, and, when applied under certain conditions, can result in flare-ups of other pests (Khan et al. 2004b; Bailey 2007; Knight et al. 2007; Whitehouse 2007; Whitehouse & Grimshaw 2007). As a result there is a need for the development of ecologically sustainable, targeted control strategies for C. dilutus.
Here, we review the current knowledge of C. dilutus. We examine various aspects of the biology and ecology of this pest including the feeding behaviour, host plants and known life cycle. We also examine current control methods and laboratory-rearing techniques, which will be important for the development of novel control strategies in cotton and other cropping environments. Relevant information from work on other key mirid pests from Australia and around the world is also included, particularly where literature on C. dilutus is sparse. The information provided is largely centred on C. dilutus as a pest of cotton in northern Australia, particularly in north-western New South Wales and south-eastern and central Queensland, where the majority of research has been conducted.
Systematic studies and morphology
The genus Creontiades contains over 50 described species and has a worldwide distribution (Malipatil & Cassis 1997). The taxonomic status of Australian species has been surrounded by some confusion. Cassis and Gross (1995) assigned six Australian species to the Creontiades genus: Creontiades angulifer, Creontiades dilutus, Creontiades pallidifer, Creontiades virescens, Creontiades vitticollis and Creontiades vittipennis. The closely related Creontiades and Megacoelum genera were then reviewed by Malipatil and Cassis (1997). Using detailed morphological examination the authors concluded that C. dilutus and Creontiades pacificus (the brown mirid) were the only Creontiades species present in Australia. Furthermore, C. virescens and Megacoelum modestum (used by Bishop 1980) were deemed synonymous with C. dilutus, and C. pallidifer was found to be a synonym of C. pacificus. Recently, C. dilutus was thought to have been present and damaging cotton crops in southern Texas, USA. However, molecular comparison to Australian collected specimens of C. dilutus and C. pacificus found the Texan species was neither of these, but rather another closely related Creontiades species (Coleman et al. 2008).
Miles (1995), Khan (1999) and Khan and Quade (2008) provided morphological descriptions of C. dilutus. Adults are approximately 7–9 mm in length with an elongated body and long antennae. They possess long legs (especially the hind legs) and have sparse setae on their scutellum and pronotum. Newly hatched nymphs are approximately 1.5–2 mm in length, with antennae much longer than their body (Table 1). As nymphs grow, they typically change colour from very pale green to yellowish green. Adults typically have a light yellow-green coloured body, although this is somewhat variable (Malipatil & Cassis 1997; Khan & Quade 2008). Malipatil and Cassis (1997) found colour variation to be most apparent on the hemelytra and hind femora, and this occurred even within populations from the same location. Females can be differentiated from males in the final instar and as adults by the presence of a median cleft which runs along the mid ventral line of the last abdominal segment (Khan 1999). Antennae are four segmented with a red coloured tip. Wing pads start to develop at the third instar. By the fifth instar, insects develop a brown tinge on the hind legs (Khan 1999; Khan & Quade 2008).
|Stage||Body length (mm) (mean ± SE)||Antennal length (mm) (mean ± SE)||Head capsule width (mm) (mean ± SE)|
|First instar||2.07 ± 0.09||2.58 ± 0.07||0.59 ± 0.01|
|Second instar||3.10 ± 0.08||3.77 ± 0.03||0.82 ± 0.01|
|Third instar||4.40 ± 0.11||5.29 ± 0.07||1.02 ± 0.01|
|Fourth instar||5.84 ± 0.06||6.75 ± 0.07||1.22 ± 0.02|
|Fifth instar||7.46 ± 0.15||9.33 ± 0.08||1.50 ± 0.02|
|Adult male||8.90 ± 0.18||10.90 ± 0.11||1.80 ± 0.03|
|Adult female||8.50 ± 0.23||10.90 ± 0.15||1.70 ± 0.02|
Life cycle and population genetics
The majority of mirid species (including C. dilutus) reproduce sexually, although asexual reproduction through parthenogenesis is known in some species in which males are rare (Wheeler 2001). The life cycle of C. dilutus consists of an egg, five nymphal stages and an adult (Khan 1999). Foley and Pyke (1985) found that on rare occasions, individuals reached the adult stage in either four or six nymphal stages. In summer, a generation (egg to adult) can be completed in approximately 3 weeks, with adults able to live for 3–4 weeks. Females lay their eggs singly within the plant tissue, leaving just an oval shaped egg cap (operculum) protruding above the plant surface through which respiration occurs and nymphs emerge (Khan 1999). Egg cap morphology was described by Khan (1999) as being light blue to hyaline in colour after oviposition, changing to a pale yellow colour prior to hatching. Eggs are approximately 1.5 mm long, banana shaped, slightly curved, tapering towards the posterior end, narrowing to a neck below the operculum (Miles 1995). Females can lay up to 80 eggs in a lifetime (Khan 1999).
The oviposition pattern of C. dilutus on cotton plants at various growth stages has been studied in detail. Females show a strong preference for oviposition at the distal end of the plant petioles and prefer to lay eggs on cotton plants at the squaring to boll formation stages, rather than younger or older plants. Insects also show a strong tendency to oviposit on the fourth to eighth main stem nodal petioles (counting downwards from the first unfolded leaf) on plants at any developmental stage. Khan (1999) suggested this pattern could be explained in terms of petiole hardness and hairiness. The top petioles (1–3) are covered in dense hairs, which may provide a mechanical barrier to oviposition, and the lower petioles (9+) are possibly too hard to allow penetration of the insect's ovipositor. Benedict et al. (1981) found similar patterns of oviposition for the western tarnished plant bug (Lygus hesperus) on cotton plants. This species showed a tendency to lay eggs at the distal end of the petioles on the upper one third of the plant.
Khan et al. (2009) examined the effects of temperature on egg and nymphal development, fecundity and survival of C. dilutus. These parameters were investigated using a series of temperatures, and a modified day-degree model was then fitted to the data. The optimum range for female egg production was found to be 26–32°C, and no eggs were produced at 11°C or 38 °C. Egg development rate increased significantly with temperature up to 30°C, where eggs took an average of 4.9 days to hatch; however, development was only slightly slower at 32°C (5.1 days). Percentage of egg survival was the highest at 26°C but did not differ significantly between 23°C and 32°C. No eggs survived below 15°C, or at the highest temperature tested (38°C). Total nymphal development time was the shortest at 30°C and 32°C, averaging 10.7 days at each of these temperatures. Using the day-degree model, optimum temperatures for egg and nymphal development were found to be 30.5°C and 31.5°C, respectively. Development times do not appear to differ significantly between males and females. The first four instars last approximately 2 days each, and the fifth instar lasts approximately 3 days. In both field and laboratory experiments, the observed sex ratio of C. dilutus is very close to 1:1 (Khan 1999).
Little is known about the population genetic structure of C. dilutus. Researchers have recently utilised mitochondrial DNA sequences to investigate genetic diversity in the closely related Lygus genus in the USA. Similar levels of intraspecific genetic variation were found within regional populations and between widely dispersed populations of Lygus lineolaris (the tarnished plant bug) suggesting a lack of geographically based population structure (Burange et al. 2007). Neighbour-joining trees showed that most individuals belong to two closely related, sympatric clades and suggested the possible existence of cryptic species. Researchers have recently identified polymorphic microsatellite loci in L. lineolaris (Perera et al. 2007), Lygus lucorum (the green leaf bug) (Liu et al. 2007) and L. hesperus (the western tarnished plant bug) (Shrestha et al. 2007). These molecular markers are likely to be employed to investigate the temporal and geographic population structure of these pests, and could prove useful in mapping insecticide resistance (Liu et al. 2007; Perera et al. 2007; Shrestha et al. 2007).
Distribution, host plants and pest status
Creontiades spp. are distributed throughout the world including parts of South America (Lukefahr 1981), the USA (Armstrong et al. 2009), Africa (Ratnadass et al. 1994), Asia (Patil et al. 2006), Europe (Efil & Bayram 2009) and Australia (Malipatil & Cassis 1997) (Fig. 1). Many of these species are pests, to varying degrees, of a wide range of agricultural crops around the world. To date, Creontiades spp. have not been recorded in the cold temperate zones in the North Hemisphere including Scandinavia, Canada, Russia and Great Britain.
Creontiades dilutus is endemic to Australia and has been recorded from all states and territories, including Tasmania (Malipatil & Cassis 1997). They have been found on a variety of habitat types including grasslands, dry sclerophyll, spinifex plains and beech forest (Malipatil & Cassis 1997). Creontiades dilutus was first recognised as a pest of Australian cotton around 1980 (Bishop 1980; Adams & Pyke 1982), with several authors in the 1960s and 1970s failing to attribute any crop damage to this species (e.g. Room & Wardhaugh 1977). Following this, C. dilutus were still not considered a major pest of cotton as they were coincidentally controlled by applications of broad-spectrum insecticides targeted at Helicoverpa spp. and other insect pests (Fitt et al. 1994; Khan 1999; Khan et al. 1999).
More recently however, Bt-cotton varieties encoding transgenes of B. thuringiensis have been introduced to the Australian cotton industry as a means of suppressing Helicoverpa spp. and other lepidopteran pests. Single-gene Bt-cotton (Ingard®) was rapidly superseded by two-gene Bt-cotton (Bollgard II®), which provides very effective season long control of Helicoverpa spp. (Pyke 2007). This has led to a drastic reduction in insecticide usage for controlling Helicoverpa spp. (Knox et al. 2006; Pyke 2007); however, Bt-cotton does not control sucking pests (Fitt 2000), and as a result C. dilutus has now become a key focus of pest control. This is also true for Lygus spp. in the USA (Snodgrass et al. 2009) and China (Lu et al. 2010). In northern Australia, cotton production during winter has been investigated as one approach to minimise lepidopteran pest issues (Yeates 2001); however, sucking insects including C. dilutus are still problematic at this time of the year (Ward 2005).
In addition to damaging cotton, C. dilutus also feed on lucerne, several pulse crops, sunflowers, and many fruits and vegetables including apples, tomatoes, grapes, citrus, potatoes and stone fruits (Hely et al. 1982; Gross & Cassis 1991; Hori & Miles 1993; Wheeler 2001; Bailey 2007). There are several non-crop plant species, which act as alternative hosts for C. dilutus over the winter period and during the warmer months of spring and summer. Table 2 details the non-crop host plants identified in northern Australia, although the complete number of species across Australia is likely to be greater than those listed. Primary hosts are those capable of sustaining C. dilutus throughout their full life cycle, whereas incidental hosts are those on which C. dilutus has been found but not as breeding populations. Creontiades dilutus may also be predators of some mite species, aphids and eggs of Helicoverpa spp.; however, this has not been studied in detail (Khan et al. 2004a). Hori and Miles (1993) observed mirids occasionally feeding on dead or incapacitated individuals of their own species but concluded that any regular predatory activity appeared unlikely, and this would probably not offset the damage caused to crops.
|Family||Common name||Scientific name||Host status|
|Aamranthaceae||Common joyweed||Alternanthera nodiflora||Primary host†|
|Apiaceae||Wild parsnip||Trachymene glaucifolia||Primary host‡|
|Asteraceae||Variegated thistle||Silybum marianum||Primary host†|
|Noogoora burr||Xanthium occidentale||Incidental host†|
|Wild sunflower||Verbesina encelioides||Incidental host†|
|Common white sunray||Rhodanthe floribunda||Primary host‡|
|Burr daisy||Calotis multicoulis||Incidental host‡|
|Speedy weed||Flaveria australasica||Incidental host‡|
|Ixiolaena||Ixiolaena chloroleuca||Incidental host‡|
|Slender groundsel||Senecio glossanthus||Incidental host‡|
|Aizoaceae||Hairy carpet weed||Glinus lotoides||Primary host†|
|New Zealand spinach||Tetragonia tetragonoides||Incidental host†|
|Brassicaceae||Turnip weed||Rapistrum rugosum||Primary host†‡|
|Chenopodiaceae||Saltbush||Atriplex sp.||Incidental host‡|
|Fabaceae||Burr medic||Medicago polymorpha||Incidental host†|
|Sesbania||Sesbania cannabina||Incidental host†‡|
|Rattlepod||Crotalaria sp.||Primary host‡|
|Hairy indigo||Indigophera hirsuta||Primary host‡|
|Hexham scent||Melilotus indicus||Primary host‡|
|Rhynchosia||Rhynchosia minima||Primary host‡|
|Annual verbine||Psoralea cinerea||Primary host‡|
|Siratro||Macroptilium atropurpureum||Incidental host‡|
|Goodeniacea||Serrated goodenia||Goodenia heterochila||Incidental host‡|
|Haloragaceae||Glauca||Haloragis glauca||Primary host†|
|Malvaceae||Marshmallow||Malva parviflora||Incidental host†|
|Solanaceae||Black-berry nightshade||Solanum nigrum||Primary host†|
|Thornapple||Datura inoxia||Incidental host†|
|Umbellifereae||Coriander||Coriandrum sativum||Incidental host†|
|Verbenaceae||Trailing verbena||Verbena supine||Primary host†|
|Mayne's pest||Verbena tenuisecta||Primary host‡|
|Common verbena||Verbena littoralis||Incidental host‡|
|Zygophyllaceae||Caltrop||Tribulus terrestris||Incidental host‡|
The closely related C. pacificus is a common pest of pulse crops, cotton, sunflowers and lucerne in Australia (Bailey 2007). In Australian cotton, C. pacificus is typically more common in mixed cropping systems (e.g. cotton, soybean, mungbean, pigeon pea) than in cotton monoculture systems; however, this species is still far less prevalent than C. dilutus in both instances (Khan & Quade 2008). Unlike C. dilutus, which is endemic to Australia, C. pacificus has also been recorded in China, several countries in the Oriental region and various islands in the south-west Pacific (Malipatil & Cassis 1997). Other Creontiades species, C. debilis, C. rubrinervis and C. signatus, are pests of cotton in the USA, while C. pallidus is widely distributed in Asia, Africa and Europe (Wheeler 2001). Other Australian pest mirids include the broken-backed bug (Taylorilygus pallidulus), the apple dimpling bug (Campylomma spp.) and the Australian crop mirid (Sidnia kinbergi); however, these are far less prevalent than C. dilutus (Bailey 2007).
The related species L. lineolaris, L. hesperus and Lygus rugulipennis (the European tarnished plant bug) are major pests in the USA, and the latter is also an important pest across much of Europe and Asia (Wheeler 2001). These species are all highly polyphagous; L. hesperus and L. rugulipennis are known from over 100 host plants, and L. lineolaris has over 300 recorded plant hosts (Young 1986; Wheeler 2001). Some of the main crops damaged by Lygus spp. in different parts of the world include cotton, lucerne, canola, some pulses and many fruits and vegetables (Wheeler 2001). Levels of damage caused by these pests vary depending on the species, the host crop and the geographic region. Lygus spp. have a long history of damaging lucerne seed crops in the USA (Wheeler 2001), and are considered the most important pest of cotton in some regions of the USA (Fournier et al. 2008; Williams 2008; Musser et al. 2009).
Other economically important mirid pests include the rapid plant bug (Adelphocoris rapidus), the superb plant bug (Adelphocoris superbus) and the alfalfa plant bug (Adelphocoris lineolatus). Several mirid species are major pests of sorghum grown in Africa and Asia, where they feed upon the flowers and grain of the plant. The head bug (Eurystylus oldi) is a major pest of sorghum in Africa, causing significant losses in grain quality and quantity (Wheeler 2001). In India, the sorghum earhead bug (Calocoris angustatus) was first recognised as a pest in the late 1800s, and remains a constant threat to production in several states today (Wheeler 2001).
Feeding mechanisms and plant damage
All mirids have piercing and sucking mouthparts. They feed by piercing the plant tissue with their sharp stylet and removing the contents of adjacent cells using a ‘lacerate and flush’ mechanism (Miles 1972). Phytophagous mirids typically feed preferentially on the growing points of plants, and are not generally well adapted to feeding on hard or dry plant tissues (Miles 1972). Digestive enzymes present in the watery saliva of mirids are thought to be involved in both penetrating the surface of plant tissue and the pre-oral digestion of cells (Taylor 1995; Colebatch et al. 2001). Damage to cotton plants by C. dilutus is thought to be caused by the salivary enzyme pectinase (Khan 1999). In young plants this chemical destroys cells surrounding the feeding point, which hinders the movement of nutrients up the plant and leads to wilting (Hori & Miles 1993; Khan 1999). In squaring plants, pectinase may cause an imbalance of the plant hormones auxin and ethylene, which leads to the shedding of squares (Khan 1999). Damage to lucerne plants is also thought to be caused by pectinase, which is discharged into the feeding sites (Hori & Miles 1993).
The feeding behaviour of C. dilutus on Australian cotton has been studied in detail. Creontiades dilutus cause both direct damage (destruction of terminals, leaves and branch primordia), and indirect damage (deformed plants) to cotton plants (Khan 1999). Adults and nymphs feed mainly on the growing points of cotton, particularly the terminals, squares and young bolls (Bishop 1980; Foley & Pyke 1985; Khan 1999; Khan et al. 2004a). Feeding damage is cumulative and, depending on the severity, can cause terminals to wilt, the abscission of squares and young bolls, and damage to lint in developing bolls (Adams & Pyke 1982; Chinajariyawong et al. 1988; Khan 1999; Khan & Bauer 2001). Early season damage can delay plant growth and maturity, which may increase management costs through additional insecticide sprays or water for irrigation (Khan 1999). Later in the season, losses in yield quantity and quality can occur through insect feeding on cotton bolls (Khan 1999).
In lucerne, C. dilutus target the flowering heads and seed pods. This can cause pods and buds to wither and under high pest pressure cause flowers to drop from the heads (Hori & Miles 1993). Mirid feeding can result in severe losses to lucerne seed crops; however, C. dilutus is unlikely to be a major pest in lucerne crops grown for hay, as they are unable to complete their development when provided with foliage alone (Hori & Miles 1993). Feeding on pulse crops by C. dilutus (and C. pacificus) can cause buds, flowers and small pods to abort, and result in significant production losses through a reduction in seed size and quality when large pods are damaged (Bailey 2007).
In addition to feeding damage, several mirid species can vector important plant viruses and bacterial pathogens. Examples include the transmission of cotton boll rot by L. lineolaris and C. pallidus in the USA and Africa, respectively. Engytatus nicotianae (the tomato mirid) is a vector of the velvet tobacco mottle virus in tomato plants (Wheeler 2001). Creontiades dilutus is not known to vector any important plant viruses.
Monitoring methods and action thresholds
Determining the level of plant damage caused by a given number of a particular insect pest is critical prior to implementing a control strategy. For C. dilutus, obtaining accurate population estimates is difficult as both adults and nymphs have a tendency to run or fly off quickly when disturbed (Bodnaruk 1992). Khan (1999) also noted that adults generally occur on the outsides of cotton terminals and squares, whereas nymphs are generally found inside these structures, which may have important implications for sampling and population estimates.
Creontiades dilutus populations vary both spatially and temporally in their distribution within cotton crops. Bodnaruk (1992) found significant differences in the numbers of mirids captured by sweep netting at different times throughout the day, and concluded that this species exhibits daily activity peaks during the mid–late morning and late afternoon. Similar behavioural patterns have been shown for Lygus spp. (Mueller & Stern 1973). Green mirid populations can also increase rapidly within a short period of time when weather conditions remain cloudy and temperatures are around 32°C for several days (Khan & Quade 2008). Using long-term average temperature data, Khan et al. (2009) predicted the months of November–January to be most favourable for the development of C. dilutus populations in the major cotton growing areas of Australia. December–January is when many Australian cotton crops are typically at the early boll stage and this is when C. dilutus can cause significant damage to crops (Khan et al. 2006a).
Appropriate methods for sampling mirids have been investigated in detail in cotton and some pulse crops. Sweep netting and ‘beat sheets’ are two quick and easy methods that provide consistent results between samples and operators (Khan et al. 2004b; Threlfall et al. 2005). However, sweep nets are unsuitable for use when plants are small and tender because they can damage the plants (Khan 1999). Additionally, sweep nets cannot be accurately relied upon for counting nymphs (Wilson & Gutierrez 1980). Vacuum sampling is another relatively quick and easy method; however, noise from the machine, plant phenology, position of insects on the plant and diurnal patterns of behaviour all affect the efficiency of this technique (Stanley 1997). Visual counts of adult and nymphs are accurate; however, they are also very time consuming and become less efficient as plants reach later developmental stages (Khan 1999; Khan et al. 2004b).
In addition to the difficulties associated with estimating mirid numbers, determining the amount of damage being caused to cotton plants is further complicated by the ability of cotton to physiologically compensate for some early season loss of squares (Brook et al. 1992). It is recommended that control decisions for C. dilutus in Australian cotton are based on a combination of plant damage and insect population data (Khan et al. 2004b). Economic thresholds have been refined within the cotton industry (Khan et al. 2006a), and the most recent information is provided in the Cotton Pest Management Guide 2010/11 (Maas 2010). Thresholds for C. dilutus range from 0.5 to 4 mirids per metre of row, depending on the crop stage, sampling technique (visual or beatsheet) and climatic conditions. Plant damage indices that are used in conjunction with insect count data are percentage fruit retention, percentage boll damage and percentage tip damage.
In mungbeans, Creontiades spp. can reduce yields at a rate of 60 kg/ha per mirid per square metre of crop, with C. dilutus and C. pacificus being equally damaging (Bailey 2007; Brier et al. 2008). Thresholds are generally based on this value and are typically in the order of 0.5 mirids per square metre using the beat sheet technique (Brier et al. 2008). Beat sheet sampling has been adopted as industry best practice because it is more effective than other methods for a wide range of pests (Brier et al. 2008, 2010). Similar to cotton, mungbeans have been shown to compensate well for early mirid damage (i.e. at early budding) and application of insecticides can be delayed until mid-flowering without impacting yield (Brier et al. 2010). Soybeans, however, are more tolerant of mirid feeding. Populations of C. dilutus as high as 5 per square metre have only little impact on crop maturity and do not appear to reduce yield (Brier et al. 2008, 2010).
Laboratory rearing methods
Foley and Pyke (1985) were the first to successfully rear C. dilutus in the laboratory. Adults and nymphs were collected from lucerne and housed in 4 L plastic containers with nylon gauze lids, and kept at 28°C with 14 h light : 10 h dark. Washed green beans were placed on top of the gauze as a food source and an oviposition site. Lucerne clippings were placed inside the containers for shelter and this was found to reduce adult mortality. Approximately 60% of nymphs survived to adulthood and just over half of these were female.
In a detailed study, Khan (1999) trialled four rearing methods with the aim of developing an easy and reliable mass rearing technique. Insects were collected from lucerne and adults were introduced to cages containing either lettuce, sprouted potatoes, green beans or cotton seedlings as a food source and oviposition site. Trials were conducted in a controlled temperature room at 25°C, 40–60% humidity and 14 h light : 10 h dark, with the exception of the cotton seedling experiment which was conducted in a glasshouse at 30–32°C with a natural light–dark cycle. Both the lettuce and potato trials proved unsuccessful. The cotton seedling method was somewhat successful; insects laid eggs and some 20 nymphs emerged, although none reached adulthood. The green bean method was the most successful. All adults survived 22–39 days, females laid high numbers of eggs and the majority of nymphs that emerged reached adulthood. The use of green beans has several advantages over other rearing methods. Beans are inexpensive, readily available throughout the year, and they do not need to be changed every day. Pesticide contamination of purchased beans is a potential problem; however, Snodgrass (1996a) showed that soaking beans in sodium hypochlorite solution, washing with detergent then rinsing in water is able to remove or oxidise any pesticide residues.
Several studies have shown Lygus bugs can also be successfully reared on green beans. Khattat and Stewart (1977) compared the suitability of potato shoots, green beans, wax beans, pea sprouts and bean sprouts as food and oviposition sites for L. lineolaris. Potato shoots and green beans were the most successful. Snodgrass and McWilliams (1992) used green beans for rearing L. lineolaris and also showed that mirid eggs could be stored at 10°C for up to 15 days without affecting hatch rates or development. Several studies have also shown Lygus bugs can be reared on artificial diets. Debolt (1982) reported the first artificial diet for Lygus rearing. Patana (1982) developed a disposable, heat-sealed packet diet that could be successfully used for rearing both L. hesperus nymphs and adults. Later, Cohen (2000) improved the artificial rearing system by developing an oligidic diet, which was used to successfully rear L. hesperus and L. lineolaris.
Current control methods
Insecticide sprays remain the primary method of control for pest mirids in Australia and overseas. For Bollgard II® cotton crops in Australia, an average of 2–4 insecticide sprays per season are usually applied to control mirids (predominantly C. dilutus), with most crops sprayed at least once (Khan et al. 2006a, 2009; Khan & Quade 2008). Industry surveys conducted in the 2005/2006 and 2006/2007 Australian cotton growing seasons found the majority of spray applications were based on the number of pests present rather than the level of plant damage or fruit retention (Whitehouse 2007). Growers were generally not following the industry's recommended action thresholds for mirids but were instead spraying at infestation levels well below these guidelines. The main insecticides used to target C. dilutus were fipronil and dimethoate, which have rankings of moderate and high, respectively, for their overall impact on beneficial species (Wilson et al. 2005).
The continued use of ‘insurance’ sprays for C. dilutus in cotton and pulses is concerning. Whitehouse (2007) identified that a common practice among growers was to apply an insecticide – when conditions would not normally necessitate a spray – but because of other management constraints (e.g. ‘last opportunity to use a ground rig’). Insurance spraying in pulses is likely to be due at least in part to the low cost of dimethoate, which is an effective insecticide against mirids. Additionally, poor pod set in some pulses is often wrongly attributed to mirids when in fact it is due to other factors such as moisture stress and nutrient deficiency (H Brier, pers. comm. 2010).
Although there is currently no known resistance to any insecticides in C. dilutus (Khan et al. 2004b), the overuse of chemicals and repeated applications of particular classes of insecticides will select for resistance. Insecticide resistance is widespread and well documented for Lygus species in other parts of the world. Field populations of L. lineolaris have shown resistance to several pyrethroid and organophosphorous chemicals (Snodgrass 1996b; Snodgrass & Scott 2003). Resistance has decreased the effectiveness of chemical control for L. lineolaris and increased the amount and cost of insecticides required to control this species (Zhu & Snodgrass 2003).
The addition of common salt (NaCl) to reduced rates of fipronil, indoxacarb and dimethoate has been shown to significantly increase the efficacy of these insecticides against C. dilutus and decrease the harmful effects on beneficial arthropods (Khan 2003; Brier et al. 2004; Khan et al. 2006b). The salt itself is not toxic but it increases the palatability of the insecticide, which encourages insects to ingest more of the chemical (Khan et al. 2004b). Although this practice may save money on insecticide costs, care must be taken, as the effect of reduced chemical rates on other pest species is unknown and may contribute to the development of insecticide resistance. Similar to the use of salt as an additive, the addition of a petroleum spray oil to varying rates of fipronil, imidacloprid and indoxacarb has been found to increase the efficacy of these chemicals (Khan et al. 2004b). Petroleum spray oils have also been shown to increase the efficacy of the biopesticides, nuclear polyhedrosis virus (NPV) and Bt, against Helicoverpa spp. (Mensah et al. 2005). Preliminary trials have shown that kaolin, a mineral particle film, may be useful in preventing cotton plants from damage caused by the feeding of C. dilutus, by making them visually or tactually unrecognisable (Khan & Quade 2006). Future work is required to determine the impact of kaolin on beneficial invertebrates and assess the suitability of incorporating this technology into an integrated control strategy (Khan & Quade 2006).
Trap crops work on the premise that pest species prefer particular host plants than others. This knowledge can be utilised to help protect a crop from pest damage, either by preventing the pest reaching the target crop, or by concentrating them in an area where they can be easily controlled (Mensah & Khan 1997). The use of lucerne as a trap crop in Australian cotton crops was investigated by Mensah and Khan (1997). In field experiments, 15 and 35 times more C. dilutus adults and nymphs, respectively, were found on cotton without lucerne strip inter-plantings compared with cotton containing lucerne as a trap crop. Within the cotton/lucerne inter-planting, 18 and 42 times more adults and nymphs, respectively, were found on the lucerne compared with the cotton. In mesh cage tests, C. dilutus showed a strong preference for oviposition on lucerne over cotton when given a choice. This work has since been trialled on numerous occasions; however, difficulties associated with managing cotton and lucerne together (e.g. water limitations, establishment and removal of lucerne) have resulted in this strategy being largely unused (M Miles, pers. comm. 2010). Several studies have also investigated the use of lucerne as a trap crop to manage pest Lygus bugs in cotton (Stern et al. 1969; Sevacherian & Stern 1974; Godfrey & Leigh 1994) and strawberries (Easterbrook & Tooley 1999). Lucerne trap crops along with tractor-mounted vacuums are being successfully used in managing L. hesperus in organic strawberries in California (Swezey et al. 2007).
Creontiades dilutus have proved difficult to control biologically as their temporal abundance is unpredictable and the use of broad-spectrum insecticides in cotton crops generally has a negative impact on beneficial natural enemies (Grundy 2007; Knight et al. 2007). No specific parasitoids have been identified, and although several generalist predators are known to feed on mirids, the level of control they exert is not well understood (Khan et al. 2004a). Grundy (2007) found that inundative releases of Pristhesancus plagipennis (the assassin bug) combined with compatible insecticides of low toxicity provided good control of both C. dilutus and Helicoverpa spp. on cotton. Despite this, the commercial uptake of such practice is unlikely due to the high costs involved in large-scale inundative releases (Grundy 2007). Nabis capsiformes (the damsel bug) kills C. dilutus and is thought to be important in limiting their numbers on bean crops (Hely et al. 1982). A small mirid, Tytthus chinensis, has been identified as a predator of C. dilutus eggs (Khan & Murray 2005), but little work has been carried out to assess the extent of this predatory activity. Several large spiders that occur in cotton are known to prey on mirids (Whitehouse 2006). These include lynx spiders (Oxyopidae), yellow night stalkers (Cheiracanthium spp.) and possibly jumping spiders (Salticidae). Lynx spiders have been shown to rapidly kill high numbers of C. dilutus in laboratory experiments; however, this has yet to be measured under field conditions (M Whitehouse, pers. comm. 2008). Despite the current lack of information regarding the effectiveness of natural predators controlling C. dilutus, the development of future control strategies should place strong emphasis on preserving populations of these beneficial species (Khan et al. 2004a).