PLANT-INSECT INTERACTIONS

Winter Weeds as Inoculum Sources of Tomato Spotted Wilt Virus and as Reservoirs for Its Vector, Frankliniella fusca (Thysanoptera: Thripidae) in Farmscapes of Georgia RAJAGOPALBABU SRINIVASAN,1,2 DAVID RILEY,1 STAN DIFFIE,1 ANITA SHRESTHA,1 3 AND ALBERT CULBREATH

Environ. Entomol. 43(2): 410Ð420 (2014); DOI: http://dx.doi.org/10.1603/EN13288

ABSTRACT Thrips-transmitted Tomato spotted wilt virus (TSWV) has a broad host range including crops and weeds. In Georgia, TSWV is known to consistently affect peanut, tomato, pepper, and tobacco production. These crops are grown from March through November. In the crop-free period, weeds are presumed to serve as a green bridge for thrips and TSWV. Previous studies have identiÞed several winter weeds as TSWV and thrips hosts. However, their ability to inßuence TSWV transmission in crops is still not completely understood. To further understand these interactions, population dynamics of two prevalent vectors, viz., Frankliniella fusca (Hinds) and Frankliniella occidentalis (Pergande), on selected winter weeds were monitored from October through April in four counties from 2004 to 2008. Peak populations were typically recorded in March. F. fusca and F. occidentalis adults were found on winter weeds and their percentages ranged from 0 to 68% in comparison with other adults. Immatures outnumbered all adults. Microcosm experiments indicated that the selected winter weeds differentially supported F. fusca reproduction and development. The time required to complete one generation (adult to adult) ranged from 11 to 16 d. Adult recovery ranged from 0.97 to 2.2 per female released. In addition, transmission assays revealed that thrips efÞciently transmitted TSWV from peanut to weeds, the incidence of infection ranged from 10 to 55%. Back transmission assays with thrips from TSWV-infected weeds resulted in up to 75% TSWV infection in peanut. These whole-plant transmission and back transmission assays provide the basis for TSWV persistence in farmscapes year round. KEY WORDS tospovirus, alternate host, vector reservoir, inoculum source

Tomato spotted wilt virus (Family Bunyaviridae; Genus Tospovirus) is transmitted by thrips in a persistent and propagative manner (German et al. 1992; Ullman et al. 1993, 2002; WhitÞeld et al. 2005; Pappu et al. 2009). Approximately 10 species of thrips are known to transmit Tomato spotted wilt virus (TSWV; Mound 1996, Riley et al. 2011). Of those, the tobacco thrips, Frankliniella fusca (Hinds), and the western ßower thrips, Frankliniella occidentalis (Pergande), are considered as important vectors of TSWV in Georgia (Todd et al. 1995, 1996; Riley et al. 2011). TSWV is predominantly a spring season threat in Georgia, and F. fusca and F. occidentalis populations in crops are often associated with early season TSWV infections and late season infections, respectively (Riley et al. 2012). TSWV has been increasing in importance since its introduction to the southeastern United States in the 1980s (Black et al. 1986). Crop losses in Georgia alone, due in large part to TSWV, have exceeded ⬎US$326 million from 1996 to 2006 (Riley et al. 2011). 1 Department of Entomology, College of Agriculture and Environmental Sciences, 2360 Rainwater Rd., Tifton, GA 31793. 2 Corresponding author, e-mail: [email protected]. 3 Department of Plant Pathology, University of Georgia, 2360 Rainwater Rd., Tifton, GA 31793-5766.

TSWV has a broad host range extending ⬎900 plant species in multiple families (Tsompana and Moyer 2008, Pappu et al. 2009). It severely affects crops such as peanut (Arachis hypogaea L.), pepper (Capsicum annuum L.), tobacco (Nicotiana tabacum L.), and tomato (Lycopersicon esculentum Miller) (Gitaitis et al. 1998, McPherson et al. 1999, Garcia et al. 2000, Riley and Pappu 2000, Culbreath et al. 2003, Culbreath and Srinivasan 2011). Besides the crops mentioned above, it is also known to infect numerous weed species in the southeastern United States (Groves et al. 2001, 2002; McPherson et al. 2003; Mullis and Martinez 2009). The role of weed ßora as hosts of the vector or the virus could be critical to virus incidence in crop hosts (Duffus 1971). TSWV is transmitted by thrips only if acquisition occurs at the Þrst-instar larval stage (van de Wetering et al. 1996). Therefore, besides being infected with the virus, it is necessary for any plant host to support thrips vector populations for at least a generation to function as a TSWV inoculum source. TSWV occurs annually in Georgia. Nonetheless, the crops are only grown from March through November and TSWV is not seed-transmitted. This implies that the virus and the vector should survive on alternate hosts or winter weeds during the crop-free period. A

0046-225X/14/0410Ð0420$04.00/0 䉷 2014 Entomological Society of America

April 2014

SRINIVASAN ET AL.: WINTER WEED, THRIP, AND TOMATO SPOTTED WILT VIRUS

number of winter annual weed species have been identiÞed as hosts of TSWV in the southeastern United States (Groves et al. 2001, 2002; Mullis and Martinez 2009). Groves et al. (2002), through sampling adults and immatures from winter weeds and rearing immatures to adulthood on green bean (Phaseolus vulgaris L.) pods, demonstrated that winter weeds could serve as hosts of F. fusca. Other Þeld studies indicated that the winter weeds differed in their abilities to support thrips populations (Groves et al. 2001, Morsello and Kennedy 2009). However, biological Þtness parameters of F. fusca on various winter weeds over an entire generation (adult to adult) were not exclusively studied in the laboratory. The potential of these weed hosts to serve as TSWV inoculum sources and affect TSWV transmission in crop hosts is not completely understood either. To our knowledge, only one study had conducted controlled transmission assays involving weeds as inoculum sources, tobacco as recipients, and onion thrips (Thrips tabaci Lindeman) as the vector (Chatzivassiliou et al. 2007). Even in that study leaf discs were used instead of whole plants to conduct transmission assays. Previous surveys have revealed high levels of TSWV infection (up to 32%) in several weeds such as narrow leaf cudweed (Gamochaeta falcata (Lamarck) Cabrera), chickweed (Stellaria media L.), VA pepperweed (Lepidium virginicum L.), and Carolina geranium (Geranium carolinianum L.) (Groves et al. 2001, 2002; Mullis and Martinez 2009). Similarly, one of our preliminary winter weed sampling surveys, followed by serological testing of foliar samples, also indicated that a high percentage (up to 20%) of certain species of winter weeds were infected with TSWV. However, when an attempt was made to phenotype the TSWV isolates from the weed samples through mechanical inoculation of an indicator host (tobacco), numerous inoculations did not produce TSWV infection. Careful examination with other detection techniques revealed that several of those samples were not infected with TSWV and were indeed false positives. TSWV infections in weed hosts are typically asymptomatic, and double antibody sandwichÐ enzyme-linked immunosorbent assay (DASÐELISA) is routinely used for TSWV detection in weeds. It is not uncommon to observe high background absorbance in microtiter plates while attempting to test samples from multiple plant species or families (Timmerman et al. 1985, Smith et al. 2006). Even though there are suggestions to ameliorate the background issue (Towbin and Gordon 1984, Smith et al. 2006), it is recommended that ELISA results be conÞrmed by other concrete detection methods or through transmission assays (Timmerman et al. 1985). Susceptible crops grown in spring, especially in their early stages, are very vulnerable to TSWV and seldom recover from infection (Culbreath et al. 2003, Joost and Riley 2004, Riley et al. 2012). Their yields are typically unmarketable. Thus, it is essential to understand how the green bridge of winter and spring weeds functions as a thrips reservoir and as a TSWV inoculum source. Very few studies in the laboratory have ex-

411

amined the biological Þtness parameters of F. occidentalis and onion thrips (T. tabaci) on weeds (Bautista and Mau 1994, Chatzivassiliou et al. 2007), even fewer with F. fusca. However, a number of Þeld studies have documented differential abilities of winter annual weeds to support F. fusca populations (Groves et al. 2001, Morsello and Kennedy 2009). To better understand the ability of winter weeds to serve as reservoirs of TSWV vectors in GeorgiaÕs farmscapes, we monitored the population dynamics of thrips across four counties from 2004 to 2008. Further, we conducted microcosm studies to compare thrips Þtness parameters on selected winter weeds in the laboratory. In addition, we conducted transmission assays using TSWV-infected peanut as an inoculum source and selected winter weeds as recipients. In these assays, to circumvent the background effects associated with DASÐELISA, we used reverse transcriptase polymerase chain reaction (RT-PCR) and employed other techniques such as mechanical inoculation of an indicator host to conÞrm TSWV-infection in winter weeds. To examine the potential of TSWV spread from weeds to crops, we conducted back transmission assays with TSWV-infected weeds as inoculum sources and peanut plants as recipients. Materials and Methods Thrips Sampling in Winter Weeds. Five species of weeds prevalent in GeorgiaÕs farmscapes and known to serve as TSWV hosts were selected for sampling in six ßagged locations (⬇6 m apart) adjacent to production Þelds in four counties (Brooks, Colquitt, Decatur, and Tift) @ two Þelds per county. The weed species sampled from 2004 to 2008 were G. carolinianum, S. media, G. falcata, sowthistle (Sonchus asper (L.) Hill), and L. virginicum. In the Þrst season (2004 Ð 2005), the samples were collected in monthly intervals from October to April. In the second season (2005Ð 2006), the samples were collected in monthly intervals from December to April. In the third (2006 Ð2007) and the fourth (2007Ð2008) seasons the samples were collected in January and March alone, as they were considered as key indicator months for winter and spring thrips. A maximum of six samples (one from each ßagged location) of fresh foliage of each weed species, enough to Þll to capacity a 3.785 liters (1-gallon)-clear ziplock plastic bag, were collected from each farm site in monthly intervals and returned to the laboratory. If a weed species was not present or not sufÞciently abundant within a 6 m diameter from a sample location ßag to Þll a sample bag, then whatever amount was present was collected. In the laboratory, the fresh samples were weighed. The weed samples were subsequently placed in BerleseÕs funnels for 7 d and thrips were collected in 70% ethanol. The collected adult thrips were identiÞed to species and immature thrips were counted as a group. Maintenance of Nonviruliferous F. fusca. A colony of F. fusca was established in 2009 on noninfected peanut leaßets of the cultivar Georgia Green with thrips collected from peanut blooms from the Bel-

412

ENVIRONMENTAL ENTOMOLOGY

ßower Farm, UGA Tifton Campus. Since then, thrips were maintained in Munger cages (Munger 1942) on noninfected Georgia Green leaßets. Munger cages were maintained in a growth chamber (Thermo scientiÞc, Dubuque, IA) at 25Ð30⬚C with a photoperiod of 14:10 (L:D) h. The leaßets in each cage were dusted with ⬇0.05 g of pine (Pinus taeda L.) pollen (Angelella and Riley 2010). All laboratory and greenhouse experiments were performed only with F. fusca. F. fusca Reproduction and Development on Winter Weeds. F. fusca reproduction and development were evaluated on four winter weed species, viz., G. carolinianum, G. falcata, S. asper, and S. media. Seedlings of weeds were collected from the borders of crop Þelds during the months of January and February. The seedlings were held in the greenhouse in 47.5 cm3 thrips-proof cages (Megaview Science Co., Taichung, Taiwan) for approximately 2 wk. If thrips larvae, adults, or feeding injuries were noticed before commencement of the experiment, then the plants were discarded. Ten Munger cages were set up with foliage (⬇2 g per cage) from each weed species. Ten nonviruliferous female adults (up to 2 d old) were released in each Munger cage. F. fusca are parthenogenic and females will begin to lay viable eggs without mating (Eddy and Livingstone 1931). Five days postrelease, adult thrips from all the cages were removed. The cages were monitored daily under a compound microscope (Meiji Techno, Santa Clara, CA) and newly hatched larvae were recorded. The number of adults emerging from each cage was recorded at 24-h intervals and removed. The cages were monitored until there were no more larvae or adults in each cage. The experiment was repeated once (N ⫽ 20 cages for each genotype). Data were pooled from both repeats of the experiment using experiment as the blocking variable. Treatments (weeds) were considered as Þxed effects and replications were considered as random effects. Data were analyzed by generalized mixed linear models using PROC GLIMMIX in SAS (SAS Enterprise 4.2, SAS Institute, Cary, NC). A completely randomized design was used. Treatment means were separated using the least squares means option in SAS. The median developmental time required for the adults to develop was estimated on each weed species. Statistical signiÞcance of differences in developmental time on weed species was estimated by KruskalÐWallis test using PROC NONPAR1WAY in SAS at a 95% CL. TSWV Transmission to Winter Weeds. The four winter weed species described in previous experiment served as recipients in F. fusca-mediated transmission assay. Seedlings of weeds were collected from crop Þelds and maintained in the greenhouse in thripsproof cages as previously described. Ten plants of each weed species were selected for the transmission assay and the assay was repeated once (N ⫽ 20 plants for each weed species). A subsample from each weed species (⬇50%) was tested for the presence of TSWV by RT-PCR before the commencement of the transmission assay. Ten potentially viruliferous F. fusca that developed on TSWV-infected peanut leaßets were transferred to a 1.5-ml microcentrifuge tube (Fisher,

Vol. 43, no. 2

Pittsburgh, PA) by a paintbrush. Thrips were subsequently released on weeds (10 potentially viruliferous thrips per plant) that were dusted with ⬇0.05 g of pine pollen per plant. The plants at the time of inoculation were ⬇5Ð10 cm in height with up to 10 fully expanded true leaves. All the weed species that were inoculated remained in the vegetative stage. Each plant was enclosed in a Mylar Þlm (GraÞx, Cleveland, PA) cage with a copper mesh top (mesh pore size-170 ␮m; TWP, Berkeley, CA). Plants were maintained in the greenhouse for 3 wk, after which TSWV infection status was assessed by RT-PCR (Jain et al. 1998, Shrestha et al. 2013). Total RNA was extracted from foliar weed samples using RNeasy kit (Qiagen, Valencia, CA) according to the manufacturerÕs recommendation. RT-PCR was performed by OneStep RT-PCR kit (Qiagen, Valencia, CA). The reaction volume was 50 ␮l, which included 0.6 ␮M of each forward and reverse primer (Jain et al. 1998), 2 ␮l of one-step RT-PCR enzyme mix, 10 ␮l of RT-PCR buffer, 10 ␮l of Q solution, and 400 ␮M of each deoxyribonucleotide triphosphates. Reverse transcription was performed in an automated thermal cycler (Eppendorf, Hamburg, Germany) programmed at 50⬚C for 30 min. Initial PCR activation was conducted at 95⬚C for 15 min followed by 35 cycles at 94⬚C for 1 min, 56⬚C for 45 s, and 72⬚C for 1 min, and a Þnal extension step at 72⬚C for 10 min. The amplicons were assessed by electrophoresis on a 1% agarose gel. Statistical analysis was performed to compare the incidence of TSWV infection among different weed species. A completely randomized design was used. Treatments (weeds) were considered as Þxed effects and replications were considered as random effects. Data were pooled from both repeats of the experiment using experiment as the blocking variable. TSWV infection was treated as a binomial response (positive or negative), and differences among treatments were estimated by logistic regression analysis using PROC GENMOD with logit link function in SAS. The statistical signiÞcance of differences between treatment pairs was estimated using pairwise contrasts. Mechanical Inoculation of Tobacco Plants. Foliar samples from Þeld-collected, TSWV-infected weeds (infection conÞrmed by RT-PCR) in the Þrst repeat of the transmission assay were used as inoculum sources for mechanical inoculation (Mandal et al. 2002). Foliage from each weed species was ground in 1:10 ratio (wt:vol) in 0.1 M phosphate buffer pH 7.0, containing 0.2% sodium sulÞte, 0.01 M mercaptoethanol, Celite 545 (Acros Organics, Geel, Belgium), and 0.01 g/ml of carborundum (320 grit, Fisher, Fair Lawn, NJ). The extract was then rubbed onto leaves of 3-wk-old tobacco seedlings of the cultivar NC-71 that had been previously dusted with 1 ml carborundum per plant . Tobacco seedlings are very susceptible to TSWV and are often used as indicator hosts. Inoculated tobacco seedlings were placed in thrips-proof cages at 25Ð30⬚C, 80 Ð90% relative humidity, and a photoperiod of 14:10 (L:D) h in the greenhouse. The presence or absence of TSWV symptoms was visually observed at 1 and 2 wks postinoculation.

April 2014

SRINIVASAN ET AL.: WINTER WEED, THRIP, AND TOMATO SPOTTED WILT VIRUS

413

Fig. 1. Total thrips counts from foliar samples of weeds. Total counts represent thrips collected from all weed species and all counties and locations in a sampling month. Samples were collected from October to April over 4 yr. Sampling months were not uniform every year. Foliar samples were collected from Þve weed species known to be hosts of TSWV. Upon collection they were placed in BerleseÕs funnels for 7 d. Adults and immatures were counted, but only adult thrips were identiÞed.

Back Transmission of TSWV from Weeds to Peanut. Four TSWV-infected winter weed species described in the previous experiment served as inoculum sources and 1-wk-old peanut seedlings served as recipients. Ten potentially viruliferous second-instar F. fusca larvae that developed on TSWV-infected weeds were transferred to a 1.5-ml microcentrifuge tube (Fisher, Pittsburgh, PA) by a paintbrush. Thrips were subsequently released on noninfected 1-wk-old peanut seedlings (10 potentially viruliferous thrips per plant) that have been dusted with ⬇0.05 g pine pollen per plant. Each plant was enclosed in a Mylar Þlm cage as described above. Plants were maintained in the greenhouse for 3 wk, after which TSWV infection status was assessed by RT-PCR. Ten peanut plants were inoculated per each inoculum source (weed species) and the experiment was repeated once (N ⫽ 20 peanut plants for each weed species). Total RNA extraction, RT-PCR, and statistical analyses were conducted as described previously. Results Thrips Sampling in Winter Weeds. Adult TSWV vectors, viz., F. fusca and F. occidentalis, were collected

from the foliage of selected winter weeds by using BerleseÕs funnels during the months of October through March. Other adults and immatures were also collected from the weeds during that period in all 4 yr. Other adults included Frankliniella tritici (Fitch), Frankliniella bispinosa (Morgan), and members of Phlaeothripidae (Thyansoptera: Tubulifera). Although the densities of thrips varied from year to year, their temporal distribution pattern was similar (Fig. 1). The number of thrips (adults and immatures) was typically greater in March than in any other sampling month in all 4 yr (Fig. 1). More thrips (adults and immatures) were found on G. carolinianum than on other weeds in all 4 yr (Fig. 2). In general, the number of immatures was typically much greater than the number of F. fusca or F. occidentalis adults (Fig. 2). The immatures were not identiÞed to species in any of the weed species sampled. Among adults, irrespective of the host, on several occasions, the percentages of F. fusca or F. occidentalis adults were less than the other adults (Fig. 3). The percentages of F. fusca and F. occidentalis varied with weed species and sampling year (Fig. 3). The weights of plant materials collected in clear plastic bags indicated that the relative abundance of winter weeds varied with time (Fig. 4). At the

414

ENVIRONMENTAL ENTOMOLOGY

Vol. 43, no. 2

Fig. 2. Total thrips counts from foliar samples of Þve weed species collected from October to April from all counties and locations. Samples were collected over a 4-yr period. Upon collection they were placed in BerleseÕs funnels for 7 d. Adults and immatures from each weed species were counted, but only adult thrips were identiÞed.

time of peak thrips incidence (in March and April), the foliar weight of G. carolinianum was more than any other weed species collected in all 4 yr of sampling (Fig. 4). F. fusca Reproduction and Development on Winter Weeds. The number of adults produced per 10 adults released varied with winter weed species (F ⫽ 4.54; df ⫽ 3, 76; Pr ⬎ F ⫽ 0.0090). The number of adults ranged from 6.06 ⫾ 1.25 (mean ⫾ SE) to 22.18 ⫾ 4.02 (Fig. 5). The number of adults produced on S. media foliage was signiÞcantly greater than the number of adults produced on G. carolinianum and G. falcata foliage (Fig. 5). Few adults emerged in Munger cages with G. falcata foliage; however, it was not different from the number of adults that emerged from Munger cages with G. carolinianum and S. asper foliage (Fig. 5). Median developmental time also varied with weed species (␹2 ⫽ 9.01; df ⫽ 3, 76; Pr ⬎ ␹2 ⫽ 0.0292). The developmental time on G. carolinianum, G. falcata, S. asper, and S. media foliage were 12, 14, 16, and 11 d, respectively. The time required to complete one generation (adult to adult) was the shortest on S. media and the longest on S. asper. TSWV Transmission to Winter Weeds. Potentially, viruliferous F. fusca that developed on TSWV-infected peanut foliage transmitted TSWV to winter weeds. TSWV infection in winter weeds was identiÞed by ⬇770 bp RT-PCR amplicons (Fig. 6a). The incidence of TSWV infection varied with weed species (␹2 ⫽

12.40; df ⫽ 3, 76; Pr ⬎ ␹2 ⫽ 0.0061). The percentages of TSWV infection in G. carolinianum, S. media, G. falcata, and S. asper were 10.00 ⫾ 10.00 (mean ⫾ SE), 55.00 ⫾ 5.00, 40.00 ⫾ 14.14, and 20.00 ⫾ 10.00, respectively. The incidence of TSWV infection in S. media was greater than in G. carolinianum and S. asper (Table 1), but was not different from the incidence of TSWV infection in G. falcata (Table 1). The incidence of TSWV infection in G. falcata was greater than in G. carolinianum (Table 1). The incidences of TSWV infection in G. carolinianum and S. asper were not different from each other (Table 1). TSWV infection in individual weeds was detected by RT-PCR and conÞrmed by mechanical inoculation of tobacco seedlings. Tobacco seedlings inoculated with foliage from all TSWV-infected weeds species, except for S. media, produced symptoms suggestive of TSWV infection. Detection of TSWV by RT-PCR and conÞrmation by mechanical inoculation were in agreement for G. carolinianum, G. falcata, and S. asper. The symptoms included necrotic spots, which later coalesced and spread systemically (Fig. 7). Infected plants were stunted or dead. Back Transmission of TSWV from Weeds to Peanut. Potentially viruliferous thrips that developed on TSWV-infected winter weeds transmitted TSWV to peanut. TSWV infection in peanut was identiÞed by ⬇770 bp RT-PCR amplicons (Fig. 6b). Typical TSWV symptoms, such as necrotic spots, yellowing, and ter-

April 2014

SRINIVASAN ET AL.: WINTER WEED, THRIP, AND TOMATO SPOTTED WILT VIRUS

415

Fig. 3. Percentages of adult thrips categories from foliar samples of Þve weed species collected from October to April over 4 yr. Adult thrips were identiÞed and separated into three categories.

minal wilting, were also observed in infected plants. The incidence of infection varied based on the inoculum source (␹2 ⫽ 14.18; df ⫽ 3, 70; Pr ⬎ ␹2 ⫽ 0.0027). The percentages of TSWV infection in peanut after inoculation with viruliferous thrips from G. carolinianum, S. media, G. falcata, and S. asper were 62.50 ⫾ 12.50 (mean ⫾ SE), 14.28 ⫾ 0, 60.00 ⫾ 10.00, and 75.00 ⫾ 15.00, respectively. The infection percentage was the lowest when potentially viruliferous thrips from S. media were used for inoculation (Table 2). The incidences of TSWV infection in peanut plants did not vary from each other when potentially viruliferous thrips that developed on G. caroliananum, G. falcata, and S. asper were used for inoculation (Table 2). Discussion Adult TSWV vectors, viz., F. fusca and F. occidentalis, were found on all the sampled winter weeds over a 4-yr period. Groves et al. (2002) sampled numerous species of winter weeds and collected adult thrips and immatures from them using BerleseÕs funnels in North Carolina. They also reared immatures to adulthood using green bean pods. Their results indicated that

77% of total thrips trapped comprised TSWV vectors, of which 96% was F. fusca. Our data over the sampling periods indicated that only up to 68% of the trapped adults were TSWV vectors. Of which, 0 Ð 67% were F. fusca and 0 Ð 47% were F. occidentalis. In our case, the immatures were not reared to adulthood. This precluded us from evaluating what percentage of trapped immatures were F. fusca or F. occidentalis. Several preliminary attempts to rear immatures from weed foliage to adults using green bean pods resulted in the death of a very high percentage of immatures before reaching adulthood. To further examine the ability of these weeds to support thrips reproduction and development, Munger cage studies were conducted in the laboratory with weed foliage using F. fusca. F. fusca reproduction and development were monitored for an entire generation (adult to adult) in this study. Although both F. fusca and F. occidentalis can efÞciently transmit TSWV (WhitÞeld et al. 2005, Pappu et al. 2009, Riley et al. 2011), F. fusca is documented to colonize crops, such as peanut and tobacco, in their early stages in Georgia (Chamberlin et al. 1992, McPherson et al. 1999). However, F. occidentalis populations on crop hosts seem to

416

ENVIRONMENTAL ENTOMOLOGY

Vol. 43, no. 2

Fig. 4. Percentages of weights of foliage from Þve weed species. Foliage from six sampling points in two Þelds from each county was brought to the laboratory and weighed.

Fig. 5. Percentage survival to adult F. fusca on weed foliage in Munger cages. Adults were removed daily beginning 5 d after release of thrips. Observations were taken until there were no alive larvae.

April 2014

SRINIVASAN ET AL.: WINTER WEED, THRIP, AND TOMATO SPOTTED WILT VIRUS

417

Fig. 6. (A) A representative photograph of an electrophoresis gel displaying ⬇770 bp RT-PCR amplicons indicative of TSWV infection in foliar samples from weeds. The lanes in the gel photograph represent: 1) 1 kb ladder, 3) TSWV-infected peanut, 4) noninfected peanut, and 5Ð12) inoculated winter weeds. (B) A representative photograph of an elecrphoresis gel displaying ⬇770 bp RT-PCR amplicons indicative of TSWV infection in peanut foliar samples inoculated with potentially viruliferous thrips from weeds. The lanes in the gel photograph represent: 1) 1 kb ladder, 3) TSWV-infected peanut, 4) noninfected peanut, and 5⫺10) peanut inoculated with potentially viruliferous thrips from winter weeds.

be closely associated with ßowers later in the season in Georgia (Riley and Pappu 2000, Joost and Riley 2004, Riley et al. 2010). Early season TSWV infections are known to cause more yield losses in crops than late season infections (Chaisuekul et al. 2003, Culbreath et al. 2003, Riley et al. 2012). Because of these reasons and the maintenance of a F. fusca colony in our laboratory, Þtness studies were only conducted with F. fusca. Munger cage studies indicated that winter weeds differentially supported F. fusca reproduction and development. The number of adults developed per female adult released was greater on S. media than on G. carolinianum, G. falcata, and S. asper. These results should be cautiously interpreted, as the Þtness experiment was conducted using detached foliage. Foliage detachment could lead to induction of defense responses otherwise absent in intact leaves and affect insect feeding (Reymond et al. 2000). Despite that, these results reiterate that winter weeds could support F. fusca populations for at least a whole generation. Our Þeld sampling counts indicated that numerically more F. fusca, and thrips in general, were recovered from G. carolinianum than from other weed Table 1. Differences in the incidence of TSWV infection in winter weeds with TSWV-infected peanut as an inoculum source Treatment pairs

df

␹2

P ⬎ ␹2

S. media versus G. falcata S. media versus G. carolinianum S. media versus S. asper G. falcata versus G. carolinianum G. falcata versus S. asper G. carolinianum versus S. asper

1, 38 1, 38 1, 38 1, 38 1, 38 1, 38

0.96 10.34 5.65 5.26 2.02 0.82

0.3273 0.0013 0.0174 0.0218 0.1548 0.3653

Ten potentially viruliferous F. fusca adults that developed on TSWV-infected peanut plants were used for inoculating weeds. TSWV infection status of inoculated plants was tested by reverse transcriptase polymerase chain reaction 3 wk post inoculation. Differences in the incidence of TSWV infection between treatment pairs were assessed by pairwise contrasts following logistic regression in SAS using PROC GENMOD.

species. This might be due, in part, to the amount of samples available and collected from each location. The sample weights clearly indicated that the weight of G. carolinianum foliage was greater than the foliage from any other weed species sampled in all 4 yr. Thrips median developmental time (adult to adult) was the shortest on S. media and the longest on S. asper. This revealed that not only was the reproduction rate greater on S. media, but also the median developmental time was the shortest on S. media, suggesting that S. media might be a more suitable host for F. fusca than the others tested in this study. Transmission assays with F. fusca indicated that they transmitted TSWV from peanut to all weed species tested. The TSWV infection percentages ranged from 10 to 55%. The infection percentages in S. media and G. falcata were greater than the other two weed species tested. S. media, in addition to being a very suitable host for thrips, was also very susceptible to TSWV. Because the seedlings of weeds used for transmission assays were Þeld-collected, to assess any initial TSWV infection, we tested at least 50% of the number of plants in each weed species by RT-PCR before inoculation. None of the plants tested were positive for TSWV infection before inoculation. Considering the infection rates obtained after inoculation and that the infection rates before inoculation was zero, it is highly likely that the observed TSWV infection in weeds was a direct result of thrips-mediated inoculations in the greenhouse. In general, serological detection techniques have been known to overestimate virus incidence in some crop and noncrop hosts (Timmerman et al. 1985, Smith et al. 2006). High levels of background absorbance induced by nonspeciÞc binding or because of other reasons have been associated with DASÐELISA testing. To avoid overestimation of TSWV infection in inoculated weed hosts all the weeds were tested only by RT-PCR. Further, to conÞrm TSWV infection in inoculated weeds, mechanical

418

ENVIRONMENTAL ENTOMOLOGY

Vol. 43, no. 2

Fig. 7. ConÞrmation of TSWV infection in winter weeds by mechanical inoculation of tobacco seedlings. (A) noninoculated tobacco plant and (B) tobacco plant inoculated with TSWV-infected winter weed foliage. (Online Þgure in color.)

inoculation was adopted. Mechanical inoculation using TSWV-infected foliage from weeds produced infection in tobacco, except when TSWV-infected S. media foliage was used as an inoculum source. It is not clear as to why that was the case even though the incidence of TSWV infection in S. media was greater than all the other weed species. TSWV-infected S. media plants had extensive thrips feeding injuries. They were also substantially senescing when compared with other weed species that were inoculated at the same time. These factors might have contributed to a reduction in virus load in S. media and needs to be investigated in detail. In the transmission assay, thrips introduced to inoculate weeds were not removed from those plants. Second-instar larvae that developed on TSWV-infected weeds were used for back transmission assay using peanut plants as recipients. Back transmission assay indicated that F. fusca were able to transmit TSWV from all the selected weed species to peanut. The incidence of TSWV infection varied from 14 to 75%. The incidence of infection was the lowest when S. media was used as the inoculum source. The incidence of TSWV infection in peanut was 3Ð5 times Table 2. Differences in the incidence of TSWV infection in peanut plants with TSWV-infected winter weeds as inoculum sources Treatment pairs

df

␹2

P ⬎ ␹2

S. media versus G. falcata S. media versus G. carolinianum S. media versus S. asper G. falcata versus G. carolinianum G. falcata versus S. asper G. carolinianum versus S. asper

1, 32 1, 32 1, 32 1, 38 1, 38 1, 38

7.67 7.67 13.17 0.04 1.03 1.03

0.0056 0.0056 0.0003 0.8410 0.3903 0.3903

Ten potentially viruliferous F. fusca second-instar larvae that developed on TSWV-infected weeds were used to inoculate non-infected peanut plants. TSWV infection status of inoculated plants was tested by reverse transcriptase polymerase chain reaction 3 wk post inoculation. Differences in the incidence of TSWV infection between treatment pairs were assessed by pairwise contrasts following logistic regression in SAS using PROC GENMOD.

more when potentially viruliferous larvae from other weed species were used for inoculation. Despite the ability of S. media to serve as a good thrips reservoir and support high levels of TSWV infection, thrips were able to transmit TSWV from S. media to peanut only at a low efÞciency when compared with other weed species. Extensive thrips feeding injuries observed on S. media foliage and increased senescence than the other weed species could have suppressed virus multiplication and accumulation, which in turn, could have negatively affected TSWV acquisition from S. media and inoculation of peanut plants by F. fusca. A previous survey on winter weed infection percentages revealed that TSWV infection in S. media was lower than several other winter weeds sampled (Groves et al. 2002). On the contrary, Groves et al. (2001) was able to recover more viruliferous F. fusca from caged S. media plants in the Þeld than from Scleranthus annuus L. and S. asper. Planting of S. media as an inoculum source in the Þeld also resulted in increased TSWV spread to Ranunculus sardous Crantz plants. More than 40% of the R. sardous samples were infected with TSWV (Groves et al. 2001), suggesting that S. media could be an efÞcient TSWV inoculum source. From our experiments, it was evident that thrips were able to transmit TSWV to weed hosts and from weeds to peanut efÞciently. Again, the caveat here is that the experiments were conducted in the greenhouse and in the laboratory where environmental conditions were relatively optimal for plant growth, thrips reproduction, and TSWV symptom expression. Experiments conducted by Morsello and Kennedy (2009) indicated that weather factors, such as rainfall timing and amount and temperature, could play a major role in the extent of TSWV spread in a weed species and the subsequently inßuence of TSWV infection in crop hosts. For instance, fewest number of TSWV-infected S. media plants were found in patches that received the highest rainfall. Further, rainfall also negatively affected the immature F. fusca populations (Morsello

April 2014

SRINIVASAN ET AL.: WINTER WEED, THRIP, AND TOMATO SPOTTED WILT VIRUS

and Kennedy 2009). Not much information is available on how weather factors could inßuence thrips development and TSWV spread in other weed species. However, rainfall in general, is known to be a key factor in regulating the population dynamics of numerous thrips species including F. fusca (Kirk et al. 1997; Morsello et al. 2008, 2010). A recently developed epidemiological model for TSWV incidence in tobacco suggests a strong link between previous years thrips abundance, winter annual weeds, and weather factors such as winter temperature and precipitation (Chappell et al. 2013). The results from this study clearly indicate that TSWV infection rates could be high in winter weeds and that F. fusca could efÞciently transmit TSWV from peanut to weed hosts and vice versa. Besides F. fusca other vectors, such as F. occidentalis, could also inßuence TSWV epidemics in the southeastern United States by using weeds as hosts and as inoculum sources (Kahn et al. 2005). Given that TSWV is transmitted by more than one vector species and that several winter weeds can serve as a green bridge for TSWV vectors and the virus, this information could be used to reduce TSWV incidence in agricultural crops. Information on abundance of winter weeds in conjunction with weather parameters could be used in epidemiological models to forecast TSWV incidence in crops (Chappell et al. 2013). Management of winter annual weeds to mitigate TSWV incidence would be another pragmatic option to consider. For instance, a multidisciplinary effort including weed management was helpful in the management of TSWV in Hawaii (Cho et al. 1989, Bautista et al. 1996). However, studies addressing timing of removal, removal distance from crops, and economic feasibility should be undertaken to assess the usefulness of this tactic in the southeastern United States. The Þndings in this manuscript substantiate the role of winter weeds as thrips reservoirs and TSWV inoculum sources. The experiments conducted in this manuscript clearly demonstrate that winter weeds in Georgia farmscapes could serve as hosts of TSWV vectors; however, their ability to support thrips populations could vary with species. The transmission and back transmission assays provide direct experimental evidence for movement of TSWV from crops to weeds and vice versa. The Þndings also emphasize the importance of suitable pathogen detection techniques and conÞrmatory tests while assessing the role of nonconventional hosts in virus epidemics.

Acknowledgments We wish to express our sincere thanks to the County extension agents who assisted with sample collection. We also extend our gratitude to Simmy Mckeown and Sheran Thompson for their assistance with thrips rearing and in conducting transmission assays. This research was partially supported by research grants from U.S. Department of Aagriculture Southern Region Integrated Pest Management Center and the National Peanut Board.

419

References Cited Angelella, G., and D. G. Riley. 2010. Effects of pine pollen supplementation in an onion diet on Frankliniella fusca reproduction. Environ. Entomol. 39: 505Ð512. Bautista, R., and R.F.L. Mau. 1994. Preferences and development of western ßower thrips (Thysanoptera: Thripidae) on plant hosts of tomato spotted wilt tospovirus in Hawaii. Environ. Entomol. 23: 1501Ð1507. Bautista, R., R.F.L. Mau, J. J. Cho, and D. Custer. 1996. Thrips, tospoviruses, and host-plant associations in a Hawaiian farm ecosystem: prospects for reducing disease losses. Acta Hortic. 431: 477Ð 482. Black, M. C., P. F. Lummus, D. H. Smith, and J. W. Demski. 1986. An epidemic of spotted wilt disease in South Texas peanuts in 1985. Proc. Am. Peanut Res. Ed. Soc. 18: 66. Chaisuekul, C., D. G. Riley, and H. R. Pappu. 2003. Transmission of Tomato spotted wilt virus to tomato plants of different ages. J. Entomol. Sci. 38: 126 Ð135. Chamberlin, J. R., J. W. Todd, R. J. Beshear, A. K. Culbreath, and J. W. Demski. 1992. Overwintering hosts and wing form of thrips, Frankliniella spp., in Georgia (Thysanoptera: Thripidae): Implications for management of spotted wilt disease. Environ. Entomol. 21: 121Ð128. Chappell, T. M., A.L.P. Beaudoin, and G. G. Kennedy. 2013. Interacting virus abundance and transmission intensity underlie Tomato spotted wilt virus incidence: an example weather-based model for cultivated tobacco. PLOS ONE 0073321. Chatzivassiliou, E. K., D. Peters, and N. I. Katis. 2007. The role of weeds in the spread of Tomato spotted wilt virus by Thrips tabaci (Thysanoptera: Thripidae) in tobacco crops. J. Phytopathol. 155: 699 Ð705. Cho, J. J., R.F.L. Mau, T. L. German, R. W. Hatmann, L. S. Yudin, D. Gonsalves, and R. Provvidenti. 1989. A multidisciplinary approach for management of tomato spotted wilt in Hawaii. Plant Dis. 73: 375Ð383. Culbreath, A. K., and R. Srinivasan. 2011. Epidemiology of spotted wilt disease of peanut caused by Tomato spotted wilt virus in the southeastern U.S. Virus Res. 159: 101Ð109. Culbreath, A. K., J. W. Todd, and S. Brown. 2003. Epidemiology and management of Tomato spotted wilt in peanut. Annu. Rev. Phytopathol. 41: 53Ð75. Duffus, J. E. 1971. Role of weeds in the incidence of virus diseases. Annu. Rev. Phytopathol. 9: 319 Ð340. Eddy, C. O., and E. M. Livingstone. 1931. Frankliniella fusca Hinds (thrips) on seedling cotton. S. C. Agric. Exp. Stn. Clemson Agric. Coll. Bull. 271: 1Ð23. Garcia, L. E., R. L. Brandenburg, and J. E. Bailey. 2000. Incidence of Tomato spotted wilt virus (Bunyaviridae) and tobacco thrips in Virginia-type peanuts in North Carolina. Plant Dis. 84: 459 Ð 464. German, T. L., D. E. Ullman, and J. W. Moyer. 1992. Tospoviruses: diagnosis, molecular biology, phylogeny, and vector relationships. Annu. Rev. Phytopathol. 30: 315Ð 348. Gitaitis, R. D., C. C. Dowler, and R. B. Chalfant. 1998. Epidemiology of tomato spotted wilt in pepper and tomato in southern Georgia. Plant Dis. 82: 752Ð756. Groves, R. L., J. F. Walgenbach, J. W. Moyer, and G. G. Kennedy. 2001. Overwintering of Frankliniella fusca (Thysanoptera: Thripidae) on winter annual weeds infected with Tomato spotted wilt virus and patterns of virus movement between susceptible weed hosts. Phytopathology 91: 891Ð 899. Groves, R. L., J. F. Walgenbach, J. W. Moyer, and G. G. Kennedy. 2002. The role of weed hosts and tobacco thrips, Frankliniella fusca, in the epidemiology of Tomato spotted wilt virus. Plant Dis. 86: 573Ð582.

420

ENVIRONMENTAL ENTOMOLOGY

Jain, R. K., S. S. Pappu, H. R. Pappu, A. K. Culbreath, and J. W. Todd. 1998. Molecular diagnosis of tomato spotted wilt tospovirus infection of peanut and other Þeld and greenhouse crops. Plant Dis. 82: 900 Ð904. Joost, P. H., and D. G. Riley. 2004. Sampling techniques for thrips (Thysanoptera: Thripidae) in pre-ßowering tomato. J. Econ. Entomol. 97: 1450 Ð1454. Kahn, N. D., T. F. Walgenbach, and G. G. Kennedy. 2005. Summer weeds as hosts for Frankliniella occidentalis and Frankliniella fusca (Thysanoptera: Thripidae) and as reservoirs for tomato spotted wilt tospovirus in North Carolina. J. Econ. Entomol. 98: 1810 Ð1815. Kirk, W.D.J. 1997. Distribution, abundance, and population dynamics, pp. 217Ð257. In T. Lewis (ed.), Thrips as crop pests. CAB International, Wallingford, United Kingdom. Mandal, B., H. R. Pappu, and A. K. Culbreath. 2002. Differential response of selected peanut (Arachis hypogaea) genotypes to mechanical inoculation by Tomato spotted wilt virus. Plant Dis. 86: 939 Ð944. McPherson, R. M., H. R. Pappu, and D. C. Jones. 1999. Occurrence of Þve thrips species on ßue cured tobacco and impact on spotted wilt disease incidence in Georgia. Plant Dis. 83: 765Ð767. McPherson, R. M., R. J. Beshear, W. C. Johnson, N. MartinezOchoa, and M. L. Wells. 2003. Winter and early-Spring survey of thrips vectors and host plants of tomato spotted wilt tospovirus in and near a ßue-cured tobacco farmscape. J. Entomol. Sci. 38: 660 Ð 668. Morsello, S. C., and G. G. Kennedy. 2009. Spring temperature and precipitation affect tobacco thrips, Frankliniella fusca, population growth and Tomato spotted wilt virus spread within patches of the winter annual weed Stellaria media. Entomol. Exp. Appl. 130: 138 Ð148. Morsello, S. C., R. L. Groves, B. A. Nault, and G. G. Kennedy. 2008. Temperature and precipitation affect seasonal patterns of dispersing tobacco thrips, Frankliniella fusca, and onion thrips, Thrips tabaci (Thysanoptera: Thripidae) caught on sticky traps. Environ. Entomol. 37: 79 Ð 86. Morsello, S. C., A.L.P. Beaudoin, R. L. Groves, B. A. Nault, and G. G. Kennedy. 2010. The inßuence of temperature and precipitation on spring dispersal of Frankliniella fusca changes as the season progresses. Exp. Appl. 134: 260 Ð271. Mound, L. A. 1996. The Thysanoptera vector species of tospoviruses. Acta Hortic. 431: 298 Ð309. Mullis, S., and N. Martinez-Ochoa. 2009. Non-crop hosts of TSWV, pp. 20 Ð21. In S. W. Mullis (ed.), Tospoviruses in solanaceae and other crops in the coastal plain of Georgia. University of Georgia, CAES Bulletin 1354, p. 51. Munger, F. 1942. A method for rearing citrus thrips in the laboratory. J. Econ. Entomol. 35: 373Ð375. Pappu, H. R., R.A.C. Jones, and R. K. Jain. 2009. Global status of Tospovirus epidemics in diverse cropping systems: successes achieved and challenges ahead. Virus Res. 141: 219 Ð236. Reymond P., H. Weber, M. Damond, and E. E. Farmer. 2000. Differential gene expression in response to mechanical wounding and insect feeding in Arabidopsis. Plant Cell 12: 707Ð719. Riley, D. G., and H. R. Pappu. 2000. Evaluation of tactics for management of thrips-vectored tomato spotted wilt tospovirus in tomato. Plant Dis. 84: 847Ð 852.

Vol. 43, no. 2

Riley, D. G., G. M. Angelella, and R. M. McPherson. 2010. Pine pollen dehiscence relative to thrips (Thysanoptera: Thripidae) population dynamics. Entomol. Exp. Appl. 138: 223Ð233. Riley, D. G., S. V. Joseph, R. Srinivasan, and S. Diffie. 2011. Thrips vectors of tospoviruses. J. Integr. Pest Manage. I1ÐI10. Riley, D. G., S. V. Joseph, and R. Srinivasan. 2012. Temporal relationship of thrips populations to tomato spotted wilt incidence in tomato in the Þeld. J. Entomol. Sci. 47: 65Ð75. Shrestha, A., R. Srinivasan, S. Sundaraj, A. Culbreath, and D. Riley. 2013. Second generation peanut genotypes resistant to thrips-transmitted Tomato spotted wilt virus exhibit tolerance rather than true resistance and differentially affect thrips Þtness. J. Econ. Entomol. 106: 587Ð596. Smith, T. N., S. J. Wylie, B. A. Coutts, and R.A.C. Jones. 2006. Localized distribution of Iris yellow spot virus within leeks and its reliable large-scale detection. Plant Dis. 90: 729 Ð 733. Timmerman, E. L., C.J.D. Arcy, and W. E. Splittstoesser. 1985. Beet western yellows virus in Illinois vegetable crops and weeds. Plant Dis. 69: 933Ð936. Todd, J. W., A. K. Culbreath, J. R. Chamberlin, R. J. Beshear, and B. G. Mullinix. 1995. Colonization and population dynamics of thrips in peanuts in the Southern United States, pp. 453Ð 460. In B. Parker, M. Skinner, and T. Lewis (eds.), Thrips Biology and Management. Plenum, New York, NY. Todd, J. W., A. K. Culbreath, and M. R. Brown. 1996. Dynamics of vector populations and progress of spotted wilt disease relative to insecticide use in peanuts. Acta Hortic. 431: 483Ð 490. Towbin, H., and J. Gordon. 1984. Immunoblotting and dot immunobindingÑ current status and outlook. J. Immunol. Methods 72: 313Ð340. Tsompana, M., and J. W. Moyer. 2008. Tospoviruses, pp. 157Ð162. In B.W.J. Mahy and M.H.V. Van Regenmortel (eds.), Encyclopedia of virology, vol. 5/3. Elsevier Ltd., Oxford, United Kingdom. Ullman, D. E., T. L. German, J. L. Sherwood, D. M. Westcot, and F. A. Cantone. 1993. Tospovirus replication in insect vector cells: Immunocytological evidence that the nonstructural protein encoded by the S-RNA of tomato spotted wilt tospovirus is present in thrips vector cells. Phytopathol. 86: 900 Ð905. Ullman, D. E., R. Meideros, L. R. Campbell, A. E. Whitfield, J. L. Sherwood, and T. L. German. 2002. Thrips as vectors of Tospoviruses. In R. T. Plumb (ed.), Plant virus vector interactions. Adv. Bot. Res. 36: 113Ð140. van de Wetering, F., R. Goldbach, and D. Peters. 1996. Tomato spotted wilt tospovirus ingestion by Þrst instar larvae of Frankliniella occidentalis is a prerequisite for transmission. Phytopathology 86: 900 Ð905. Whitfield, A. E., D. E. Ullman, and T. L. German. 2005. Tospovirus-thrips interactions. Annu. Rev. Phytopathol. 43: 459 Ð 489. Received 12 October 2013; accepted 4 February 2014.

Winter weeds as inoculum sources of tomato spotted wilt virus and as reservoirs for its vector, Frankliniella fusca (Thysanoptera: Thripidae) in farmscapes of Georgia.

Thrips-transmitted Tomato spotted wilt virus (TSWV) has a broad host range including crops and weeds. In Georgia, TSWV is known to consistently affect...
695KB Sizes 2 Downloads 3 Views

Recommend Documents