Tomato yellow leaf curl virus (TYLCV) is a member of the genus Begomovirus in the family Geminiviridae that is transmitted in a persistent, circulative manner by its vector, Bemisia tabaci (Gennadius) (Hemiptera: Aleyrodidae) (Brown et al. 1995). TYLCV is one of the most broadly distributed and economically damaging species in the virus complex (Glick et al. 2009). Bemisia tabaci, the only known vector, can acquire TYLCV as an adult or nymph, and adults remain infective for the duration of their life (Cohen & Nitzhany 1966; Jones 2003). Bemisia tabaci adults can acquire the virus from feeding on an infected plant for as little as 15 min, and can transmit the virus less than 24 h after acquiring it (Mansour & Al Musa 1992; Mehta et al. 1994; Czosnek et al. 2002). TYLCV produces distinctive symptoms: upwardly curled leaves that are reduced in size, brightly chlorotic leaf margins and interveinal regions on leaves, shortened petioles and a bushy appearance to plants which are reduced in size if infected early. Viral symptoms typically appear in the plant about two weeks after it has been infected; the earlier the plant is infected, the greater the impact on yield (Levy & Lapidot 2008).

TYLCV causes significant crop losses in tomato growing regions worldwide (Moriones & Castillo 2010; Czosnek & Ghanim 2011). It is a primary production constraint for tomato growers in Florida (Mosler et al. 2009; Schuster et al. 2007), which is a major producer of fresh market tomatoes in the United States (USDA 2013). In Florida, tomato is produced almost year-round in the western, central and southern portions of the state with the exception of the hottest summer months (Jul-Aug). Whitefly pressure is typically higher in the fall crop than the spring early in the season because the pest builds up during the summer on alternate weed and crop hosts. Both B and Q biotypes of B. tabaci are present in Florida, and while both biotypes can transmit TYLCV, field-grown horticultural crops are affected almost exclusively by the B biotype (McKenzie et al. 2009). Management of TYLCV in Florida tomato involves destroying inoculum sources, using reflective mulches to repel the vector, planting TYLCV-tolerant varieties, rogueing infected plants, and using insecticidal control (Schuster et al. 2008).

Protecting the tomato crop from viral infection during the first 6 weeks after transplanting is crucial to mitigate yield loss (Lapidot & Levy 2010). At-plant and early season applications of insecticides have been key components to managing TYLCV and B. tabaci in Florida tomato and elsewhere (Gilbertson et al. 2011; Schuster et al. 2008). However populations of B. tabaci biotype B have developed resistance to several modes of action globally (Castle et al. 2010; Ma et al. 2010; Rao et al. 2012), and resistance to neonicotinoids, one of the most useful insecticide groups for managing B. tabaci, has been documented in Florida (Schuster et al. 2010). Tomato transplants in Florida are typically treated with a neonicotinoid insecticide in the nursery, at plant, and in the first weeks after establishment either through drip irrigation or foliar application. Imidacloprid (IRAC group 4A) is the most commonly used material. The development of resistance to neonicotinoid insecticides by whiteflies in Florida has led to guidelines emphasizing the use of neonicotinoids during the first five weeks after transplanting but no later (Webb et al. 2013).

The registration of endosulfan for use on tomato will be withdrawn after 2014. Endosulfan is a cyclodiene insecticide that many tomatoes growers in Florida have relied on over the years to suppress whitefly adults. In order to offset the loss of control options for B. tabaci due to the development of resistance and registration withdrawals, materials with alternative modes of action are needed. Insecticides with novel modes of action that are registered or nearing registration for use on B. tabaci include cyantraniliprole, flupyradifurone, and pyrifluquinazon. Cyantraniliprole, also called Cyazypyr™, is in the anthranilic diamide group of insecticides that disrupt the functioning of ryanodine receptors. It is a Group 28 insecticide in the International Resistance Action Committee (IRAC) mode of action classification system. Flupyradifurone is a butenolide insecticide (IRAC group 4D) that functions as a nicotinic acetylcholine receptor agonist (Velten et al. 2013). Cyantraniliprole and flupyradifurone are both systemic and can be applied to the roots or foliage. Pyrifluquinazon is a translaminar quinazinalone insecticide which interferes with androgen receptor function (Kang et al. 2012; Yasunga et al. 2012). Its mode of action has not been fully characterized and it does not yet have an IRAC mode of action classification. Flupyradifurone is a Bayer product that will be sold as Sivanto. Cyazypyr™, produced by DuPont, is available in a soil-applied formulation (Verimark) and a foliar formulation (Exirel) for use on vegetables.

Research has implicated treatments of cyantraniliprole, flupyradifurone and pyrifluquinazon with reduction in the incidence of whitefly-transmitted viruses and densities of immature whiteflies (Schuster et al. 2008b; Palumbo 2012a,b; Smith & Giurcanu 2013; Smith & Giurcanu in press). Cyantraniliprole, flupyradifurone and pyrifluquinazon were compared in greenhouse studies at University of Florida, Gulf Coast Research and Education Center (GCREC) in 2011 and 2012 as single applications on tomato seedlings (Smith & Giurcanu 2013; Smith & Giurcanu, in press). These studies confirmed a role for each material in suppressing B. tabaci and reducing incidence of TYLCV under controlled conditions.

Field trials were carried out at GCREC to evaluate the effect of applying these new materials in combination during the first five weeks after transplanting on the number of whiteflies and TYLCV incidence. Venom 70 SG (dinotefuran, Valent Corporation; IRAC group 4A) was included as an at-plant material for comparison with at-plant applications of Sivanto and Verimark. Venom is a neonicotinoid insecticide that has been used by Florida tomato growers for whitefly management since 2007. In order to determine the effect of early, complete physical exclusion of vectors from the crop, each treated plot was split into an exposed or completely covered sub plot, by applying an insect-proof row cover for the first two weeks after transplanting. The purpose of these trials was to determine the most efficacious early season combinations of these new materials with the aim of integrating novel modes of action into established insecticide rotations for management of B. tabaci and TYLCV.

Materials and Methods

Field trials were carried out during the fall of 2012, and spring and fall of 2013, at the University of Florida Gulf Coast Research and Education Center (GCREC), Wimauma, Florida (N 27° 45.599′, W 82° 13.446′) to evaluate novel insecticides for suppression of Bemisia tabaci and TYLCV on tomato (variety ‘Florida 47’).

Treatment Applications

Soil drench treatments were hand ladled on the day of transplanting at the rate of 118 mL (4 fl. oz.) of preparation per plant (1698 L per ha/ 181.5 gal. per acre). Foliar treatments were applied with a hand-held sprayer with a spray wand outfitted with a single nozzle containing a 45° core and a D-5 disk. The sprayer was pressurized by CO₂ to 60 psi and calibrated to deliver 561 L per ha (60 gal. per ac) (Table 1). DuPont Pointbond All Purpose Row - Seed Bed - Insect Covers were applied to 1 sub-plot within each main plot immediately after transplanting and soil treatments were accomplished. A full day was required to establish 2 replications; therefore the trials were set up over a 2-day period. Row covers for an entire trial were removed on the same day, either 14 or 15 days after covering (Table 2).

Table 1.

Insecticide treatment rates and schedules for three seasons of trials: Fall 2012, spring 2013 and fall 2013.

Insect Samples

Bemisia tabaci sampling began soon after the row covers were removed and before any foliar applications of insecticides were made. Sampling continued weekly until 1 week after the last foliar treatments were made. All B. tabaci sampling occurred on the middle 10 plants of each sub-plot. Adult B. tabaci densities were sampled in the field by examining the third leaf from the top on 1 stem per plant and recorded as the number per 10 leaves. Immature B. tabaci densities were sampled by removing the terminal leaflet of a 7th or 8th leaf from the top of each plant. The leaflets were examined using a stereo microscope, and data were recorded as B. tabaci eggs, 1st, mid (2nd & 3rd), and 4th instars per 10 leaflets. Analysis based on combined nymphal counts is reported.

Table 2.

Key dates for each trial to evaluate novel insecticides for suppression of Bemisia Tabaci and tylcv on tomato (variety ‘Florida 47’).

Results

Bemisia tabaci Response

Average densities of whitefly adults remained less than 4.0 per 10 leaves per week in fall 2012 and less than 1.0 in spring 2013. There were no statistical differences in adult densities between covered and non-covered treatments on any week in the fall 2012 or spring 2013 trials. Trial-total adult densities were lowest in fall 2012 with treatments of Sivanto, alone or followed by pyrifluquinazon, or Verimark, alone or followed by pyrifluquinazon and there were no significant sub plot (cover) or interaction effects (Table 3). No significant differences between chemical treatments were found in spring 2013; however, adult densities were lower in the covered plots than in the non-covered plots that season and there was significant interaction between chemical and row cover effects (Table 3). Because of overall lack of significance and low counts in the fall of 2012 and spring of 2013, adult counts were not collected in fall 2013.

With the exception of the Venom drench, trial-total egg densities of B. tabaci were lower in both fall 2012 and fall 2013 trials with any chemical treatment than in the untreated check (UTC) (Table 4). No difference due to row cover was observed in the fall 2012 or spring 2013 trials for eggs; high plant mortality due to excessive heat under the covered plots in the fall 2013 trial caused the covered portion of the experiment to be abandoned (Table 4). When examined on a week by week basis, egg densities were significantly higher 6 weeks after transplanting in the UTC (4.5 per 10 leaflets) (SE = 1.3) than all other treatments (∼0.5 per 10 leaflets) (SE = 0.1) (F_9,27 = 9.27; P < 0.01), and higher in the UTC (1.3 per 10 leaflets) (SE = 0.6) than all other treatments except the Venom drench (0.6 per 10 leaflets) (SE = 0.4) (F_9,27 = 3.01; P < 0.05) 7 weeks after transplanting in the fall 2012 season. In fall 2013, egg densities were statistically higher in the UTC (3.5 per 10 leaflets) (SE = 1.2) than all treatments except Venom followed by pyrifluquinazon (1.5 per 10 leaflets) (SE = 0.5) 4 weeks after transplanting (F_9,27 = 2.84; P < 0.05). In that trial, all treatments had statistically fewer eggs than the UTC (12.0 per 10 leaflets) (SE = 4.2) 6 weeks after transplanting with the exception of Venom treatments (Venom drench: 7.0 per 10 leaflets) (SE = 3.0); Venom fb Exirel: (6.0 per 10 leaflets) (SE = 3.0); Venom fb pyrifluquinazon: (4.5 per 10 leaflets) (SE = 2.7) and Verimark followed by pyrifluquinazon (7.3 per 10 leaflets) (SE = 5.4) (F_9,27 = 2.40; P < 0.05). There were no statistical differences among treatments with regard to egg densities on any week in spring 2013.

Mean B. tabaci combined instar densities, combined over all sample weeks, were lower with any chemical treatment than UTC in all three trials, except Venom alone in fall 2012 (Table 4). In fall 2012, nymphs were detected on the first sample week, 21 days after transplant (DAT), on non-covered plants only. Trial-total combined instar densities were lower in covered plots than in non-covered in the fall 2012 trial only (Table 4). In the fall of 2013 combined instar densities were significantly higher in the UTC than in any other treatment for each sample week, except 6 weeks after transplant (F_9,27 = 1.98; P = 0.08), (Data not presented).

Table 3.

Mean number (±sem) of Bemisia tabaci adults per ten leaves totalled over all sample dates (fall 2012 consists of 5 weekly samples; spring 2013 consists of 6 weekly samples).

TYLCV Response

Fall 2012: TYLCV pressure was moderate, reaching 65% in the chemically untreated plots 9 days before harvest. In the covered plots, symptoms were first observed 35 DAT, 18 days after the covers were removed; at that time there were no differences in % virus incidence between any chemical treatment and the UTC (Table 5). Chemical treatments of Sivanto, alone or followed by pyrifluquinazon, resulted in the lowest % virus incidence observed on each of the 9 weeks of assessment, although these were not always significantly different from other treatments (F_9,27; P > 0.05). The % virus incidence was greater in the non-covered than in the covered plots on each of the 9 weeks that assessments were made (F_1,30; P ≤ 0.016) and there were no interactions between chemical treatment and cover effects (F_9,30; P ≥ 0.11).

Table 4.

Mean number (±sem) of Bemisia tabaci eggs and nymphs per ten leaflets totalled over all sample dates (fall 2012 consists of 5 weekly samples; spring and fall 2013 consists of 6 weekly samples).

Continued

Table 5.

Mean cumulative % (±sem) of plants (per 14-plant plot) with tylcv symptoms in the fall 2012 trial to evaluate novel insecticides for suppression of Bemisia tabaci and tylcv on tomato (variety ‘florida 47’).

Continued

Spring 2013: Virus pressure was negligible, probably due to unusually cold weather, and TYLCV symptoms were not observed in any plot before 54 DAT (Data not presented). Nine days before harvest, mean virus incidence was not greater than 4% in any chemical treatment and there were no significant differences between covered and non-covered treatments or interactions between chemical treatment and cover effects.

Fall 2013: Virus pressure (Table 6) and B. tabaci levels (Table 4) were at their highest of the three trials; also there were no data from covered plots in this trial, which tended to make virus levels appear larger when compared with the fall 2012 trial. Nevertheless, by 27 DAT, virus incidence was lower in all chemical treatments, except Venom, alone or followed by pyrifluquinazon, than in the UTC (Table 6). Sivanto followed by pyrifluquinazon was the only chemical treatment which resulted in significantly lower virus incidence than the UTC at 47 DAT (Table 6). This combination tended to have the lowest percent virus in 2012 also (Table 5).

Discussion

Whitefly numbers overall tended to be low during these 3 trials, yet during the 2 fall seasons, virus incidence was substantial. These results are not atypical, as it is not unusual for TYLCV incidence to be high in commercial tomato fields when whitefly counts are low. This phenomenon may be explained by the influence of the virus on the behavior of B. tabaci. There is evidence that TYLCV affects the settling, probing and feeding behavior of B. tabaci in ways that enhance transmission of the virus (Liu et al. 2013; Moreno-Delafuente et al. 2013). In addition, Liu et al. (2013) determined that viruliferous tomatoes may be more attractive to B. tabaci than uninfected tomato.

The exclusion tunnel treatment failed in fall 2013 because of high temperatures under the tunnel. Data from previous seasons indicate that no at-plant systemic treatment was comparable to mechanical exclusion of the pest. The ability of cyantraniliprole, flupyradifurone and pyrifluquinazon to reduce transmission of TYLCV in greenhouse studies has been demonstrated (Smith & Giurcanu in press), and the antifeedant properties of some of these compounds are discussed below. Under the actual field conditions of our trials significant levels of TYLCV infection were observed on treated plants. Bemisia tabaci can infect a tomato plant with TYLCV with as little as 15 min of feeding (Mehta et al. 1994; Czosnek et al. 2002). Our data indicate that no mode of action can consistently prevent transmission of virus under field conditions and that growers must employ strategies in addition to chemical control in order to protect tomato crops from moderate and severe virus pressure, such as planting TYLCV-tolerant varieties of tomato.

Most insecticide treatments in these trials reduced densities of whitefly compared to the untreated control. With the exception of Venom treatments, most insecticide treatments did not separate statistically from one another, indicating a similar level of efficacy in suppressing whitefly. During the 2 fall trials, treatments with Sivanto tended to have the numerically lowest percentage of virus, although in 2013 this was only observed earlier in the crop season, and differences were not always statistically significant from other treatments. It should be kept in mind that sizeable differences in virus incidence may not produce statistical differences among treatments because of variability, but may produce economic yield differences in commercially-grown tomato. We did not observe statistically significant yield differences during the 2 fall trials, when virus pressure was either moderate or high. It is not unusual to observe little or no difference in tomato yield due to treatment in small plot B. tabaci management evaluations, even when treatment effects on the pest are significant (Stansly et al. 2008; Schuster et al. 2009a, b, c). We only collected yield data once, while growers harvest multiple times. Our primary goal was to evaluate early season insecticide and row cover effects on whitefly density and virus incidence, not to compare full season whitefly management programs which may have produced greater treatment effects on yield.

On the whole, Venom was less effective than other materials in these trials. Ongoing monitoring of B. tabaci to dinotefuran and other group 4 insecticides indicates that whitefly populations in south Florida continue to be susceptible to dinotefuran, but that susceptibility varies in different populations (Smith & Nagel 2014). The results of these trials indicate a role for each material evaluated in managing B. tabaci and TYLCV. These results also confirm that chemical control alone may not provide sufficient protection from virus transmission and that growers must employ other tactics, including reflective mulches and TYLCV-tolerant varieties.

Table 6.

Mean cumulative % (±SEM) of plants (per 14-plant plot) with tylcv symptoms in the fall 2013 trial to evaluate novel insecticides for suppression of Bemisia tabaci and tylcv on tomato (variety ‘Florida 47’).

Table 7.

Mean yield, lbs./10 plants, (±SEM) from a single harvest (22 May) of the spring 2013 trial to evaluate novel insecticides for suppression of bemisia tabaci and tylcv on tomato (variety ‘Florida 47’).

Florida's subtropical conditions produce year-round pest pressure, which in turn leads growers to spray intensively to manage pests of high value horticultural crops such as tomato. Programs to manage whitefly-transmitted viruses in Florida focus on the integration of tactics to alleviate constant spray pressure and the resulting development of insecticide resistance (Adkins et al. 2011). Crop hygiene, reflective mulches and virus-resistant varieties contribute to whitefly suppression, however chemical control remains the primary tactic to suppress B. tabaci and the viruses it vectors (Webb et al. 2013).

Reliance on insecticides has led to the development of “treatment windows” for resistance management whereby a given mode of action is applied during specific stages of the crop's development and deliberately avoided during subsequent intervals (Flood & Wyman 2005). The treatment window approach aims to avoid treating sequential generations of a given pest with the same mode of action. The treatment window that encompasses the first 5 or 6 weeks after transplanting is possibly the most important for management of B. tabaci and TYLCV. Data presented here confirm that novel modes of action that are newly available or nearing registration for use on tomato can be effectively combined during this treatment window to reduce transmission of TYLCV compared with untreated plants. From the perspective of both resistance management and virus suppression the availability of several insecticides with distinct modes of action and antifeedant properties is advantageous. We deliberately focused on early-season protection rather than season-long whitefly management in our evaluation of new materials because yield losses diminish the later the crop is infected with TYLCV. Our efforts to compare the efficacy of atplant treatments when plants were exposed to viruliferous whiteflies at planting versus 2 weeks after planting by protecting plots with row covers met with limited success because virus pressure was very low in the spring trial and the row cover treatment failed in the fall 2013. However results from fall 2012 indicate that no at-plant treatment was comparable to complete mechanical exclusion of whitefly during the first 2 weeks post-transplant.

Given that viruliferous B. tabaci can transmit TYLCV within approximately 15 min of feeding, antifeedant properties in an insecticide are essential for managing the vector and the virus (Mehta et al. 1994). Cyantraniliprole and pyrifluquinazon have each demonstrated antifeedant properties in laboratory studies. Kang et al. (2012) observed inhibition of feeding behavior among Myzus persicae (Sulzer) (Hemiptera: Aphididae) treated with pyrifluquinazon. Jacobson & Kennedy (2013a) documented a reduction in feeding probes by M. persicae on pepper (Capsicum annuum) treated with cyantraniliprole 10 days post-treatment. The same authors (2013b) measured a reduction in the number and duration of probes by Frankliniella fusca (Thysanoptera: Thripidae) feeding on pepper treated with cyantraniliprole. Cameron et al. (2013) used fluorescent dye to show feeding reduction on the part of B. tabaci nymphs on cotton treated with cyantraniliprole. In addition, settling and feeding behavior of Diaphorini citri Kuwayama (Hemiptera: Psyllidae) was reduced on citrus treated with cyantraniliprole (Tiwari & Stelinski 2013). Cyantraniliprole, flupyradifurone and pyrifluquinazon have also reduced egglaying by B. tabaci in greenhouse studies (Tokumaru et al. 2010, Smith and Giurcanu 2013).

Antifeedant properties have also been identified in established insecticides that are available for whitefly management. These include pymetrozine (Fulfill®, Syngenta Crop Protection, Greensboro NC; IRAC group 9B) and bifenthrin (many formulations; IRAC group 3A). Pymetrozine interferes with the functioning of the cibarial pump in certain hemipterans. Hai-Hong et al. (2011) demonstrated that at 300 mg/L pymetrozine inhibits stylet penetration by B. tabaci, and Polston & Sherwood (2003) implicated pymetrozine in the reduction of transmission of TYLCV. He et al. (2013) demonstrated that sublethal doses of bifenthrin reduced phloem feeding by B. tabaci on cotton, a behavior change that may reduce transmission of TYLCV and other viruses in susceptible crops. Smith & Giurcanu (in press) found that an insecticide containing bifenthrin and zeta-cypermethrin (Hero®, FMC Corporation, Philadelphia PA) suppressed transmission of TYLCV in greenhouse studies on a level similar to flupyridifurone in tomato that were exposed to viruliferous whitefly 3 and 7 days after treatment. Current recommendations are that broad spectrum insecticides such as pyrethroids be reserved for use later in the tomato cropping season (Mossier et al. 2009). However there may be justification for the targeted use of materials such as bifenthrin within the first 6 weeks of transplanting as part of a diversified rotation of modes of action with antifeedant properties that can reduce transmission of TYLCV.

Additional field trials are needed to determine optimal rotations of new and established materials for suppressing TYLCV while offsetting the development of resistance. Pre-plant, at-plant and early drip injected treatments will employ systemic materials primarily in the group 4 and group 28 categories. The use of alternative modes of action is advised for the second post-transplant treatment window. While dinotefuran continues to offer effective suppression of TYLCV in many regions, its efficacy was not comparable to newer materials in these trials, underlining the need for additional modes of action. Sivanto has a similar mode of action as the neonicotinoid insecticides and should probably be treated as a neonicotinoid from the perspective of resistance management. Its efficacy in suppressing TYLCV, particularly in combination with pyrifluquinazon, is noteworthy in these trials.

The role of Verimark and Exirel in suppressing TYLCV during the first treatment window will influence decisions growers make regarding the use of diamides to manage caterpillars and leafminers in subsequent treatment windows. Resistance management guidelines for diamides have been developed for Florida tomato (Smith 2013). Unlike the other diamides used for management of caterpillars and leafminers on tomato, cyantraniliprole has major efficacy against sucking insects and has been implicated in the reduction of transmission of TYLCV. Growers who plan to use Verimark for protection of tomato against TYLCV during the first five weeks after transplanting are advised to use alternatives to diamides to manage caterpillars and leafminers for the second five-week window should these pests be detected at economic levels during that time frame (Smith 2013). The yield data from spring 2013, when whitefly and virus pressure was low, indicated that cyantraniliprole-treated plants had higher yield than other plants.

The availability of three new modes of action for managing TYLCV provides new tools to growers who rely heavily on insecticide applications to manage the virus. The continued efficacy of these products will require that growers practice good resistance management tactics. This includes integrating chemical control with the use of reflective mulches and resistant varieties, and destructing virus reservoirs such as harvested tomato fields.