Project title: Assessment of thrips populations and Tomato spotted wilt virus

Transcription

1 Project title: Assessment of thrips populations and Tomato spotted wilt virus (TSWV) incidence and development of a TSWV risk index and IPM strategy for processing tomatoes in the Central Valley of California () Principal investigator: Robert L. Gilbertson Department of Plant Pathology, University of California, Davis Cooperating personnel: Ozgur Batuman, Postdoctoral Researcher, UC Davis Li Fang Chen, Postdoctoral Researcher, UC Davis Michelle LeStrange, Farm Adviser, Kings County Tom Turini, Farm Adviser, Fresno County Scott Stoddard, Farm Adviser, Merced County Gene Miyao, Farm Adviser, Yolo County Neil McRoberts, Department of Plant Pathology, UC Davis Diane E. Ullman, Department of Entomology, UC Davis Summary The goal of this ongoing project is improved understanding of thrips population dynamics and Tomato spotted wilt virus (TSWV) incidence in processing tomatoes in Central California and application of this knowledge to the development of an IPM strategy for TSWV. Population densities of western flower thrips (WFT; Frankliniella occidentalis) and TSWV incidence were monitored in processing tomato fields in the Central Valley of California for a fifth year in, which turned out to be abnormally cool early in the season. Thrips were monitored with yellow sticky cards and in tomato flowers, whereas as TSWV incidence was assessed visually and infection was confirmed by testing selected plants. In, monitoring of representative tomato fields in, Kings, Merced, Yolo and Colusa Counties revealed a delay in the build-up of thrips populations (started in May) compared with years in which temperatures were normal or above normal early in the season (March-April). However, by June, populations reached high levels, which persisted into September. Populations gradually declined in October. In late-planted fields, thrips populations continued to increase through October and then slowly decreased in November. This situation was slightly different in Fresno County, as thrips populations built-up earlier (mid April), and then populations fluctuated during the season. In, the first detection of TSWV in tomato plants was in Kings County in late April, but this was in a relatively early transplanted fresh market tomato field and the virus likely was introduced with the transplants. TSWV was first detected in a fresh market tomato field in Merced about two weeks later (early May). However, in Fresno, Yolo and Colusa counties, TSWV was not detected until mid-may, which also represented a delay relative to other years. Although, TSWV was eventually detected in

2 most monitored fields, overall incidence was relatively very low in (i.e., -%, - %, -7% and -% in monitored fields in Merced, Yolo & Colusa, Fresno and Kings Counties, respectively). In small number of monitored fields in Colusa and Merced Counties, incidences reached up to -%. However, these levels of TSWV did not cause yield losses in the monitored fields. Inoculum sources for thrips/tswv vary depending on the county and agricultural production area. Thus, there is no single main source of the virus throughout the state. As in the last - years, fall crops (lettuce and radicchio) were monitored during fall/winter seasons. In general, these potential TSWV bridge crops had low thrips populations and TSWV incidence, although some fall-planted lettuce fields in Fresno had high incidences of TSWV infection. As in previous seasons, winter and spring weed surveys revealed very low levels of TSWV infection (<.%). In Merced, the implementation of effective thrips and TSWV management practices in radicchio has helped to reduce the importance of this thrips and TSWV inoculum source, resulting in low TSWV incidences in monitored tomato fields through out the season. In Yolo County, thrips and TSWV were not detected in winter-planted fava beans or daikon radish, or in most of the weeds sampled. Furthermore, results of inoculation experiments in our laboratories to assess whether daikon radish, a proposed cover crop, can be a host for TSWV were negative. RT-PCR testing of thrips revealed that most thrips were not carrying the virus in early in the season, however, many of the thrips samples collected from tomato flowers after mid-june/early July were positive for the TSWV. This is consistent with the notion that TSWV builds-up in tomato fields later in the season. Thrips-transmission efficiency experiments revealed that male adult thrips transmit TSWV more efficiently than female adult thrips. Remarkably, the overall transmission efficiencies of colonies from Fresno and Yolo were different, with thrips from the Fresno colony having a higher transmission rate (.6%) than thrips from the Yolo colony (9.%). Results of our greenhouse experiments for the assessment of the role of the soilemerging adult thrips as an inoculum source for early season tomatoes revealed very low levels of thrips emergence from soil samples and no TSWV in these thrips. In insecticide trials in, materials that reduced thrips numbers included Radiant either with or without Pre-Am, Athena with Beleaf and Venom. Results of the trial generally support results of earlier trials in that Radiant is among the top performing materials, and that Beleaf either alone or tank-mixed with a pyrethroid can provide some control. In, we developed a TSWV Risk Index (TRI), for predicting potential losses due to TSWV in Central Valley processing tomatoes. Based on gathered information for each monitored field, the TRI for all monitored fields in was moderate. We also developed a phenology model, based on degree day accumulation, for predicting thrips

3 population dynamics in the Central Valley. This model was able to predict the timing of adult thrips generations with overall accuracy of up to 8%. Therefore, integrating the thrips predictive model with TRI may enhance the capacity to predict and manage thrips and TSWV without extensive sampling. In conclusion, research efforts have revealed a good understanding of the biology of thrips and TSWV in Central California and IPM strategy based on these findings appears to be helping reduce TSWV incidences. We now hope to be able to validate the model and risk index to improve the IPM strategy and make it sustainable and grower friendly. Objectives The objectives of this project were ) to determine thrips populations and TSWV incidences associated with processing tomato fields in, ) to gain insight into potential sources of thrips and TSWV for tomatoes in the Central Valley, ) to assess the role of soil emerging thrips in TSWV epidemiology, ) to evaluate transmission efficiency of TSWV for thrips populations from different geographic origins, ) to assess various thrips control methods, 6) to develop a phenology model and a risk assessment system for thrips and tomato spotted wilt disease, respectively, in processing tomato fields and 7) to develop and validate an integrated pest management (IPM) strategy for TSWV in the Central Valley. Materials and Methods Thrips monitoring in representative fields. Table lists the fields () that were monitored for thrips and TSWV in. In, an additional 6 tomato fields (referred as secondary fields) were regularly surveyed for TSWV incidence but without monitoring thrips populations. Furthermore, 6 fields planted with other crops ( wheat, onion, radicchio and lettuce fields) in the same areas were monitored for thrips and/or TSWV. Yellow sticky cards were placed in each of the four corners of each field, just above the canopy. For the tomato fields, cards were changed weekly or biweekly beginning in March and up to harvest (August-October). Population densities were estimated by counting thrips on yellow sticky cards in the laboratory with a dissecting microscope at x magnification. Population densities of thrips were also estimated weekly or biweekly by randomly collecting flowers, placing these into vials with 7% ethanol and returning vials to the laboratory for counting thrips. Flower samples were collected from the same sites where yellow sticky cards were placed (four sites per field and flowers per site). Total numbers of thrips adults and larvae were counted and identified to species.

4 TSWV incidence and detection. Percent TSWV incidence in tomato, lettuce and radicchio fields was determined by visually examining plants at the four locations in each field. At each location, all plants in yards (meters) of each of randomly selected rows (each separated by rows) were examined. An overall incidence of tomato spotted wilt at each site of the field (four per field) was calculated (presented as number of infected plants per row feet and % incidence). Disease incidence was assessed weekly and selected plants tested with ImmunoStrips (AgDia) and RT-PCR by using N gene-specific primers to confirm TSWV infection. Assessment of the potential role of the soil-emerging thrips as sources of early TSWV infection for processing tomatoes. We collected samples early in (January-March) by taking the upper inches of the soil from selected locations where high incidences of TSWV were observed in. Soil was collected ( samples/location) from a total of 6 sites, each from Fresno, Kings, Merced and Yolo/Colusa counties (Table ). In addition, a control soil sample (sterile greenhouse soil) was also included. Fig. depicts the procedure used in these experiments. Briefly, these soil samples were placed in containers ( liter) with insect-proof mesh covers. These containers were maintained in greenhouses at UC Davis for - months to follow thrips activities. Thrips that were emerging from these soils were detected by yellow sticky cards that were placed inside the containers. Then, the emerging thrips that were captured from these soils were tested for TSWV with the RT-PCR test. Additionally, fava bean indicator plants were placed in these cages with these soil samples as another way to detect emerging viruliferous thrips. Isolate collection and genetic diversity of TSWV. In, selected crop and weed plants with TSWV symptoms, or in a few cases without obvious symptoms, were confirmed to be infected with TSWV by immunostrips or RT-PCR and then used to further investigate the nature of the TSWV isolate. To assess the genetic diversity of these TSWV isolates from the Central Valley, the fragment of RNA encoding the N gene was amplified by RT-PCR and the sequence of the N gene determined and compared with sequences of isolates previously collected from tomato and other crops. Comparison of thrips transmission efficiency for the TSWV-California isolates between two thrips colonies from Fresno and Yolo. In 9, we found that thrips populations were relatively high in Yolo County processing tomato fields, but that TSWV incidence was similar or lower than in other monitored fields in other counties

5 where thrips populations were substantially lower. These results suggested that thrips populations in different areas may have different virus transmission efficiency. To test this idea and develop a system to study thrips biology more thoroughly, we established a thrips rearing system in our laboratory and established colonies from Fresno and Yolo Counties. These two populations of western flower thrips were collected from tomato plants in Fresno and Yolo counties, and maintained on fresh green bean pods in the laboratory. Two samples of TSWV-infected tomatoes from Fresno and Yolo counties were also collected and the virus from these samples was maintained on Datura stramonium (Jimsonweed) plants via sap-inoculation or thrips transmission. For these experiments, first instar larvae were allowed to acquire TSWV from infected Datura plants for at least days and then reared to adulthood on green beans. Transmission assays were performed with - male or female adult thrips in ml conical polystyrene tubes with a -cm-square wet filter paper and single Datura cotyledon leaf for a 8-hr inoculation access period. Leaves were incubated on the surface of water in multiwell plates for days and TSWV titer was determined with the enzyme-linked immunosorbent assay (ELISA). Comparison of insecticides for control of thrips on tomato. The evaluation of insecticides for thrips control was conducted at the UC West Side Research and Extension Center. The processing tomato variety H8 was transplanted and irrigated with buried drip (depth of in.). The selection of materials tested was based on results of previous thrips insecticide trials, and communication with Pest Control Advisors and chemical company representatives (Table.). The experimental design was a four replication randomized complete block. Materials were applied with a CO -pressurized backpack sprayer at psi with equivalent of gallons of water per acre with surfactant (Induce.%). flower samples per plot were collected and kept in 7% EtOH, then thrips nymph and adults were counted with a dissecting scope. Development a phenology model and a risk assessment system for thrips and tomato spotted wilt disease in processing tomato fields in Central Valley of California. To obtain a risk index (RI) value for a particular field, point values were assigned to critical components of tomato production practices according to their relative influence on spotted wilt incidence (Table 8). For establishing the TSWV risk index (TRI) we have emphasized factors that have been found play important role on disease development in processing tomatoes, based on our research conducted thus far. Examples of these factors are shown in Table 8.

6 The phenology model is based on a degree day model that utilizes the effects of weather (i.e., temperature) on thrips biology. Currently, this model generates the degreeday accumulations for each year from January to October in a map format (heat availability). We are comparing results of this model with actual population data collected from monitored fields in Fresno, Kings, Merced, Yolo and Colusa counties in 9-. To do this, weather stations were selected from these Counties, and weather parameters were used to map climate conditions for each county. We then used this data and the phenology model to predict appearance of thrips generations in a particular area and year. The predicted data were compared with actual thrips population dynamics that were documented during our surveys in 9-.

7 Viruliferous Thrips Virus-free Thrips Are they carrying the virus? Fig.. Flow chart depicts our experimental procedures for assessment of the potential role of the soil-emerging thrips as sources of early TSWV infection for processing tomatoes.

8 RESULTS Field Monitoring In, all of the monitored fields were established with transplants as there were few direct-seeded fields. Because of delays in tomato planting due to cool weather conditions early in the season, field monitoring for thrips and TSWV in tomato fields was initiated in mid-april in Fresno and early May in the other Counties. Table lists the fields that were monitored in and includes their locations, sampling dates, harvest dates and TSWV incidence. Fresno County Early in the growing season, the average thrips populations in Fresno County were considerably higher than those detected in 7-. However, subsequent thrips populations fluctuated considerably, with mid-season (June/July) populations relatively high, but lower than those detected in 8. Towards the end of the season (August/September) additional population peaks (presumably new generations) were detected, and then populations decreased dramatically in early September (Fig. ). These late season peaks also were observed in another year where early season temperatures were lower than normal (Fig. ). Thrips populations continued to be detected into October in these locations, but then began to decline (Fig. ). However, note that this allows for overlap with fall-planted lettuce fields. Kings County In, there was a striking delay in the build-up of thrips populations in Kings County, with early season populations considerably lower than those detected in 7- (Fig. ). However, populations quickly increased starting around May (a month later than in 7-), and then remained high in August and early September. Furthermore, these populations were much higher than those detected in 7-. Populations remained high (,-,/card) through September and into October before eventually dropping in November. This was similar to the situation in, and was probably due to the delay of growing season in these years due to cool spring weather (Fig. ). Merced County In, the build-up of thrips populations was also delayed due cool weather, and was not detected until mid-may in Merced County (Fig. ). A similar delay was

9 observed in, another year in which early season temperatures were below normal. This contrasted with 8 and 9, years in which the early season temperatures were typical, and populations increased in late March/early April (Fig. ). Thus, depending on season weather conditions there was a month difference in the timing of thrips population build-up. Interestingly, in, thrips populations increased steadily from mid-may through June, eventually reaching populations that were considerably higher than those detected in 8-. The peak populations in were detected around July, similar to, but were two-fold higher (, vs., thrips/card). Furthermore, peak thrips populations in and were detected two months later than those in 8 and 9, again revealing the influence of the cool temperatures early in the season (Fig. ). In most locations, average thrips populations remained at high levels (,-,/card) through September (Fig. ). In all years, thrips populations remained high until early October and then quickly dropped through November. Yolo and Colusa Counties In, the thrips dynamics in Yolo and Colusa Counties were similar to those detected in 9 and (Fig. ). However, in 9, the average thrips populations were higher than those detected in and, especially in August-September (Fig. ). In, similar to, populations in early-planted fields started to increase in late May, which was three weeks later than those in 9 (Fig. ). Again, this can be attributed to the cool early season temperatures in and. Peak populations were detected in mid- to late August. In and, peak populations were similar (,/card), whereas these were.-fold lower than the peak detected in 9 (Fig. ). In, average thrips populations in most fields remained at high levels (,-,/card) through September (Fig. ). Thrips populations began decreasing in late September and by mid-october had decreased considerably. Flower sampling, another method to assess thrips populations, was initiated at the bloom stage of development. Overall, thrips populations in flowers were detected soon after flower emergence, and thrips continued to be detected in flowers for the rest of the season. In, average thrips populations in flowers in July was consistently lower than those detected in May, June or August in all Counties (Fig. 6a). In, the average thrips population in flowers was higher in early season and then dropped in Merced; however, in the other Counties, average thrips populations were highest in August. In, similar to other years, most of the monitored fields had populations of - thrips per flower throughout the blooming stage (Fig. 6a). Regardless of field location, average

10 thrips populations in flowers were generally high in early season (consistent with extensive blooming in developmental stage of the tomato plants) and low during fruiting stage in July as flower numbers decreased (Fig. 6a, b, c and d). Overall, it appears that the populations determined by yellow sticky cards may be more informative than populations from flowers. In, thrips larvae were commonly found in flowers, similar to 7-. This indicates that thrips are reproducing on tomatoes, and that there is the potential for secondary spread of TSWV within fields by viruliferous adults coming from juveniles that fed on TSWV-infected plants within a field. Evidence for this came from the observation of a significant increase of late season TSWV infections in which only one or a few shoots on a plant developed spotted wilt symptoms in and. This was also observed in 8 and 9 but to a lesser extent. All thrips captured in the monitored fields in were identified as western flower thrips and, consistent with previous results, female thrips populations were about three-fold higher than male populations.

11 Table. List of monitored fields and their: locations, survey and harvest dates and TSWV incidence in. Yolo/Colusa Counties TSWV % RO (Winters) Harvesting Harvested AO (County Line) Harvested BF (County Line) Harvested DB (Rd / Hwy ) Harvested EG (Sutter Co.) Harvested TP (Wilsonbend Rd) Harvested. MS (Lone Star Rd & Meyer Rd) Harvested Merced County GC (Gun Club Rd- Gustine) Harvested 7.7 HT (Hunt Rd-Gustine) Harvested 6 BC (Bert Crane Rd) Harvested FR (Franklin Rd) Harvested BU (Burchel Rd) -Fresh Market To Harvested PT (Rogers Rd-Paterson) Harvested 9. Fresno County /9/ /6/9/ 6//8/ 6/9/ 7/7// 7//8/ 8//8/ 8// 9// // North -Firebaugh area Harvested 6. Oakland -Five Points area Harvested. Shields -Firebaugh area (Fresh Market To) Harvested Mt. Whitney -Five Points area Harvested Tranquility -Tranquility area Harvested. Nees -Firebaugh area Harvested Kings County // // 6// 6/6/ 7// 7// 7/9/ 7//- 8/6/9/ 9// 9/6/ Tomato # Gale & Jameson Harvested. Tomato # Lassen & Phelps Harvested. Tomato # Dorris & El Dorado Harvested Tomato # Lassen & Jayne Harvested Tomato# Hwy 98 & Avenal Cutoff Harvested.8

12 Thrips counts per yellow sticky card 6 Average Thrips Populations per Card Fresno Co. 7- -Jan -Feb -Apr -Jun -Jul 9-Sep 9-Oct 8-Dec Sampling Dates Build up Peak Drop Highest in the peak 7 6-Apr -May 8 9-Apr 9-May -Oct Mar -Jul -Aug -May -Jul -Oct 9 -Apr 8-Aug -Sep 7 Fig.. Average thrips counts per yellow sticky card in monitored fields in Fresno County in 7-. Note the thrips population build up, peak and drop dates are indicated below graphs. Highest thrips populations during peaks are ranked (high to low) from top to bottom.

13 Thrips counts per yellow sticky card 6 Average Thrips Populations per Card Kings Co. 9- -Jan -Mar -May -Jun -Aug 9-Sep 8-Nov 7-Jan Sampling Dates Build up Peak Drop Highest in the peak 7 -Mar 9-Jun August 8 -Apr -May October 8 9 -Apr -Aug October -Apr 6-Aug November 9 -May 6-Jun November 7 Fig.. Average thrips counts per yellow sticky card in monitored fields in Kings County in 7-. Note the thrips population build up, peak and drop dates are indicated below graphs. Highest thrips populations during peaks are ranked (high to low) from top to bottom.

14 Thrips counts per yellow sticky card Average Thrips Populations per Card Merced Co. 8-6-Feb 8-Mar 7-May 6-Jul -Aug -Oct -Dec -Jan Sampling Dates 8 9 Build up Peak Drop Highest in the peak 8 -Apr 6-May -Oct 9 6-Apr -May -Oct 9 -May -Jul -Oct 8 -May -Jul -Oct Fig.. Average thrips counts per yellow sticky card in monitored fields in Merced County in 8-. Note the thrips population build up, peak and drop dates are indicated below graphs. Highest thrips populations during peaks are ranked (high to low) from top to bottom.

15 Thrips counts per yellow sticky card 8 6 Average Thrips Populations per Card Yolo/Colusa Co. 9- -Jan -Mar -May -Jun -Aug 9-Sep 8-Nov Sampling Dates 9 Build up Peak Drop Highest in the peak 9 8-May -Aug -Sep 9 -May -Aug -Oct 6-May -Jul -Sep Fig.. Average thrips counts per yellow sticky card in monitored fields in Merced County in 8-. Note the thrips population build up, peak and drop dates are indicated below graphs. Highest thrips populations during peaks are ranked (high to low) from top to bottom.

16 A Average Thrips Populations per Flower in B Average Thrips Populations per Flower Fresno & Kings Counties 7- Thrips counts per flower May June July August Fresno&Kings Co. Yolo&Colusa Co. Merced Co. Thrips counts per flower 8 6 May June July August Sampling Dates Sampling Dates C Average Thrips Populations per Flower Merced County 8- D Average Thrips Populations per Flower Yolo & Colusa Counties 9- Thrips counts per flower May June July August 8 9 Thrips counts per flower 8 6 May June July August 9 Sampling Dates Sampling Dates Fig. 6. Average thrips populations in flowers in monitored fields (A) in, (B) in Fresno and Kings Counties in 7-, (C) in Merced County in 8-, and (D) in Yolo and Colusa Counties in 9-.

17 In, the first detection of TSWV in tomato was in Kings County in late April. However, this was in an early season fresh market variety and, based on the severe symptoms in the young plants, the virus was probably brought in with transplants. In Merced, TSWV was also first detected in a fresh market tomato field, about two weeks later (early May) than in Kings County. In Fresno, Yolo and Colusa counties, TSWV was detected in processing tomatoes in mid-may, two weeks later than in Merced. The first detection of TSWV in Fresno in was similar to, but was much later than in previous years (7-9). For example, TSWV was first detected on April in 7, and in the first week of May in 8 and 9. The first detection of TSWV in in Yolo and Colusa Counties was two months later than in 9 and. In, the overall incidence of TSWV in processing tomato fields remained, with a few exceptions, very low (e.g., -%, -%, -7% and -% in monitored fields in Merced, Yolo & Colusa, Fresno and Kings Counties, respectively [Fig. 7]). In Merced, TSWV incidence was very low in monitored fields throughout the season, with the exception of the HT processing tomato field (6%). Similarly, the TSWV incidences in monitored fields in Yolo and Colusa Counties were very low except for a single field in Sutter County that had higher incidences (% incidence in parts of the field by the end of the season [Fig. 7]). The high TSWV incidences in these two fields in Sutter and Merced Counties were observed at certain areas, typically near the edge of the field, suggesting the presence of a nearby inoculum source. Higher rates of TSWV infection were reported or observed in some fields that were not monitored, including some of our secondary fields. For example, a acre late-planted processing tomato field in Yolo/Colusa had a TSWV incidence as high as %. However, in Fresno and Kings Counties, TSWV incidences in few of the monitored and secondary fields reached -7%, whereas in the rest of the fields, incidences remained very low (-% and Fig. 7). Although TSWV eventually appeared in all monitored fields in, the overall incidence was less than % in most monitored fields, similar to 8 and 9. In, higher levels of infection developed later in the season, similar to in, but even tehse levels were not sufficient to cause economic losses. Thus, the overall incidence of TSWV in monitored fields in (-%), similar to, was slightly higher than in 7-9. However, the overall pattern of disease development was similar in all of these years: low TSWV incidence in early-planted fields and higher incidences in late-planted fields.

18 TSWV incidences in monitored fields in TSWV % 6 7 Monitored Fields Kings County Merced County Fresno County Yolo/Colusa Counties Kings County Yolo/Colusa Counties Fresno County Merced County Fig. 7. Final percent TSWV incidences in monitored fields in. Detection of other tomato-infecting viruses In, the incidence of other viruses in processing tomato fields was also determined. These viruses included the new tomato-infecting ilarvirus, Tomato necrotic spot virus (ToNSV), curly top disease (caused by Beet mild curly top virus [BMCTV] and Beet severe curly top virus [BSCTV]), Alfalfa mosaic virus (AMV) and Pelargonium zonate spot virus (PZSV). This activity is very important because it prevents misdiagnosis of TSWV (some of these viruses cause TSWV-like symptoms), and gives an idea of the overall importance these other viruses in processing tomatoes in California. Early in the growing season in, we detected more ToNSV than in. In some tomato fields in Merced, ToNSV was widespread. Furthermore, we have now proven that ToNSV is readily transmitted by infected pollen via thrips feeding activities, and that is likely enter to processing tomato fields via pollen from near-by asymptomatic reservoir plants. However, overall incidence of ToNSV was <% and it did not appear to spread within the field. In addition, tomato plants seem to recover from infection by this virus. This supports our previous conclusion that the virus does not pose a serious threat to tomato production. The presence of these viruses complicates disease diagnosis, especially early in the season when symptoms induced by TSWV and some of these other viruses (especially curly top viruses) can be very similar. In addition, some fungal diseases (e.g., Verticillium and Fusarium wilt and Fusarium foot-rot) were common in some fields, and

19 symptoms of these diseases sometimes made it difficult to accurately identify TSWV infection in tomatoes. We have now developed specific primers for all of these other tomato-infecting viruses, which allows us readily identify plants infected with these viruses with PCR (curly top viruses) or RT-PCR (AMV, PZSV and ToNSV) assays. This is important because in some fields, these other viruses can be more prevalent that TSWV. For example, in some parts of the PT field in Paterson, a transplanted processing tomato field in Merced County, the incidences of AMV and ToNSV were %, whereas no TSWV infections were detected. Consistent with the results, the PT field was planted next to an old alfalfa field (a source of AMV). Interestingly, this field also had high populations of thrips (based on yellow sticky cards monitoring and feeding damage on leaves), which persisted throughout the season. We believe these thrips came from alfalfa and were not carrying TSWV. Finally, mixed infections of more than one virus can also occur in the field and this can complicate diagnosis. In 7 and 8, during the course of testing many plants with virus symptom, we found infections by either TSWV, BMCTV/BSCTV or AMV; however, we never observed mixed infections of these viruses. However, as we have tested additional samples in subsequent years we have found a small number of plants with mixed infections of TSWV and BMCTV/BSCTV, TSWV and AMV, BMCTV/BSCTV and AMV, or TSWV and ToNSV. Together, these results indicate the need for testing for multiple viruses to confirm the identity of the virus(es) in a particular field and to allow for the selection of appropriate management strategies. Survey of potential hosts for TSWV and thrips To search for reservoir hosts of TSWV and thrips before, during and after the processing tomato season in, we again surveyed spring- and fall-planted lettuce in Fresno, spring- and fall-planted radicchio in Merced, and numerous weeds collected in the winter and spring. In -, TSWV levels were relatively high in a number of surveyed fallplanted lettuce fields (). Importantly, no TSWV was detected in the spring-planted lettuce (). Thus, thrips populations in monitored spring-planted lettuce fields in Fresno County were low (- thrips/card), and TSWV was not detected in these fields. This is very important as it appears that spring-planted lettuce was not an inoculum

20 source for early season tomato fields. The high incidences in fall-planted lettuce were attributed to high incidences of TSWV in late-planted tomatoes, which were in the field after than usual due to the delayed planting of the tomato crop because of cool early season temperatures. This resulted in greater overlap between late-planted tomatoes and the fall lettuce crop, allowing viruliferous thrips to move from tomato to lettuce and spread TSWV. Furthermore, this overlap occurred over a long period of time: August to early-october. In, where there was again a delay planting due to cool spring temperatures, thrips populations in some fields were still at very high levels in August and September (>, thrips per card, and Fig. ), and many late-planted tomato fields had high incidences of TSWV (>%). While these late infections did not appear to cause yield losses in tomatoes, they appeared to be sufficient to provide inoculum for fall-planted lettuce. Fortunately, in / the TSWV from fall-planted lettuce did not jump into the spring-planted crop, possibly due to very low activity of thrips in winter, thereby not making spring lettuce a potential inoculum source in. We are currently surveying three fall-planted lettuce fields in Five Points and Huron locations, to see if high TSWV incidences are associated with the delay planting of tomatoes in. In these fields, thrips populations have been moderately high (-, thrips/card) in October, and the TSWV incidence was already -%. Some isolated August-planted lettuce blocks have even higher percentages of plants with TSWV symptoms. It will be very important to monitor spring-planted lettuce to see if TSWV develops in these fields and whether these fields may serve as TSWV bridge crops. In the winter of, radicchio fields in Fresno and Merced, which were initially under plastic, did not have plants infected with TSWV, and thrips populations were very low (- thrips/card, and data not shown). In spring, TSWV was detected in low incidences in some monitored radicchio fields in Merced, but the incidence was sporadic and did not cause economic losses in this crop. More importantly these fields did not appear to serve as sources of inoculum for tomatoes in Merced County. Furthermore, thrips management efforts combined with effective sanitation of harvested radicchio crops has reduced the role of radicchio as a TSWV bridge crop in Merced County, and this has been reflected by the low incidence of TSWV in tomatoes in 8- (Fig. 7, and data not shown). Thus, through these efforts the importance of radicchio as a TSWV inoculum source for tomato in Merced has been substantially reduced. In Fresno, a single

21 radicchio field was surveyed for TSWV in spring. A low incidence (<%) of TSWV was detected in this field. In, in areas with recent outbreaks of TSWV, weeds (and plants other than tomato) were collected and tested for the virus (Table ). Most samples were negative for TSWV; however, some radicchio, cardone, lettuce, pepper, prickly lettuce, sowthistle, groundsel (pineapple weed), bindweed, Malva, Datura (jimsonweed) and black nightshade plants tested positive for the virus. Furthermore, most of the symptomless weeds, collected before and during tomato growing season, were negative for TSWV. Only a few symptomatic weeds were found and these were found to be infected with TSWV. In addition, most of the weeds that tested positive were collected during the tomato growing season and came from fields of tomato or other TSWV-susceptible crops in which symptomatic plants were present. For example, the symptomatic Datura plants were collected from fields with TSWV-infected tomatoes. Therefore, it appears that the virus infecting these weeds may have actually come from the infected tomato plants rather than vice versa. Thus, the overall incidence of TSWV infection in weeds was very low (8/6 or <.%) and this is similar to results from previous years. To date, we have not found evidence of any weed that is extensively infected by TSWV in the Central Valley of California. Table. Weed survey results for TSWV incidence during. Weed a Tested (+) Weed a Tested (+) Black nightshade () Lambs quarters () Bindweed () Malva 67 () Bur clover () Datura -Jimsonweed () Pineapple weed () Goose foot 9 () Sowthistle 8 () Shepherd purse () Prickly lettuce 9 () Fiddler neck 9 () Russian thistle () Wild radish and Mustard () (+) number of plants tested positive for TSWV by immunostrips and RT-PCR. a, Total weed samples from all counties Early in, surveys were conducted in Yolo and Colusa Counties (January- March) in order to try to identify inoculum sources of thrips and TSWV for processing tomato fields. We found a couple of fava bean fields in one location, but surveys revealed

22 very low thrips populations and no TSWV infections. Daikon radish, a recently proposed a winter cover crop, was assessed as a potential TSWV in. Here, we sap-inoculated plants with TSWV. Despite numerous attempts, daikon radish did not become infected with TSWV, whereas positive control N. benthamiana (a type of tobacco) and Datura plants did become infected (indicating that these method was successful in delivering the virus to the inoculated plants). This indicated daikon radish is not a host. Consistent with these results, daikon radish crops inspected in the field did not show symptoms of TSWV infection and representative plants also tested negative for TSWV. Thrips populations associated with daikon radish in these fields were also very low. Before the growing season in, wheat and onion fields were monitored for thrips and collected thrips from these cards were counted and tested for TSWV with the RT-PCR assay. Until late May, thrips population densities on wheat were low (average - thrips/card) on yellow sticky cards placed in these fields. In onions, thrips population densities were also low (average - thrips/card) until in early April, and then populations rapidly increased in late April (,-, thrips/card). To date, TSWV has not been detected in thrips collected from wheat and onions, which are non-hosts of the virus. Assessment of the potential role of the soil-emerging adult thrips as inoculum sources of TSWV for early planted tomatoes In February, we collected the upper inches of the soil, with minimal disruption, from locations known to have had late-season TSWV outbreaks. Table indicates when and where these soil samples were taken, and provides summaries of the results of these experiments. In Yolo and Colusa Counties, we collected soil samples from a late-planted processing tomato field that had very high TSWV incidences (- % in affected parts of the field) in. Also, this location (sample # ) is a known TSWV hot-spot. Sample # was collected from a relatively weedy area in this hot-spot (olive orchard, near sample #). Samples # and were collected from a fall-planted radicchio crop having TSWV outbreak and a large field of fava bean, respectively. In Merced County, a soil sample was collected from the corner of a late-planted fresh market tomato field with a very high TSWV incidence (~%). This location (sample #7) is also a TSWV hot-spot. Soil from a neighboring radicchio field also was collected (Sample #6). In Fresno and Kings Counties, soils were collected from late-planted tomato fields with relatively high late-season TSWV incidences. These locations also have had history of TSWV (samples #9, and ). In these Counties, we also collected soil

23 samples from fall lettuce fields with very high TSWV incidences in fall (samples # and ). Other soil samples from Fresno and Kings Counties were collected from fallow fields and weedy orchards (samples #, and 6), which were previously confirmed to harbor TSWV-infected weeds and were nearby tomato fields had TSWV outbreaks in. Soils were placed in large plastic tubs, covered with a mesh top and kept in a greenhouse. Thrips emergence was monitored with yellow sticky cards and indicator plants used to detect TSWV. We captured only a few thrips from some of the soil samples (Table ). Most of the emerging thrips were captured in first or second week of the experiment. Thrips populations on these yellow sticky cards were low (-7 thrips/card), and these came mainly from soil samples collected from fall crops rather than soils from tomato fields that had been plowed. None of the thrips from these soils (or control thrips captured on yellow sticky cards that were placed outside of the cages in the greenhouse) tested positive for TSWV by the RT-PCR technique. Interestingly, in most cages many weeds and volunteer crops germinated and started to grow during the experiment. These plants, as well as the fava bean indicators, did not develop symptom of TSWV infection, consistent with the results indicating that the thrips were not carrying TSWV. For example, soil samples from tomato fields from Fresno and Kings Counties had high populations of volunteer tomato seedlings as well as variety of common weeds. The fact that these plants did not develop TSWV symptoms supports the idea that viruliferous thrips had not emerged from this soil. Furthermore, these results indicated that our experimental conditions mimicked the natural conditions that promote emergence of adult thrips. To further test whether these volunteer tomatoes or weeds were infected with the TSWV, leaf samples from these plants as well as the fava beans indicator plants that were placed in each cage were analyzed with RT-PCR. All of these plants were tested negative for TSWV. However, even though no thrips were captured from soil sample #7 from Merced, two months later, one of the indicator fava bean indicator plants in this cage developed some leaf necrosis. This fava bean plant tested negative for TSWV with immunostrip, but was weakly positive with RT-PCR. Together, these results indicated that thrips can stay dormant in the soil for some period of time, and that some may well retain TSWV based on the single confirmed case of an infected indicator plant, which was likely infected by viruliferous thrips emerging from soil. However, this must be reviewed as a preliminary result, and these experiments should be repeated to conclude whether soil emerging thrips can be an inoculum source

24 for tomatoes. Because, thus far, our results indicated that most of the soil samples had no or very low levels of overwintering thrips, it appears that viruliferous adult thrips emerging from soil may not be an important inoculum source for early-planted tomatoes. However, it is important not to exclude the possibility that such viruliferous thrips may be sources of TSWV for susceptible weeds or bridge crops, in which the virus can be amplified. Table. Summary of the assessment of the potential role of the soil-emerging thrips Sample # Source of the soil samples Collection Date Previous/Current Crop Type Number of captured RT-PCR tests of thrips RT-PCR tests of plants Soils Discarded Yolo & Colusa Counties Yolo Rd -Feb Fall Radicchio Negative Negative -May Yolo/Colusa County Line -Feb Olive Orchard Negative Negative -May Yolo/Colusa County Line -Feb Proc. Tomato Negative Negative -May Yolo Rd 9 -Feb Fava Beans 7 Negative Negative -May Merced County CD Chileds Ave. - Merced -Feb Fall Radicchio 7 Negative Negative 9-May 6 LG La Grand Rd. -Merced -Feb Fall Radicchio 7 Negative Negative 9-May 7 BU Burchell Rd. -Merced -Feb Late Fresh Mark. To N/A Positive 9-May 8 PL Plainsburg Rd.-Merced -Feb Weedy Almond N/A Negative 9-May Fresno County 9 Russel -Fairbaugh -Feb Proc. Tomato N/A Negative 9-May Five Point -Feb Fall Lettuce N/A Negative 9-May North TP -Fairbaugh -Feb Proc. Tomato N/A Negative 9-May Five Star -Five Point -Feb Fallow field/vineyard N/A Negative 9-May Kings County Jones -Feb Late Fresh Mark. To N/A Negative 9-May th & Oniele -Feb Fallow field/vineyard N/A Negative 9-May Westside -Feb Fall Lettuce N/A Negative 9-May 6 Plymouth -Feb Weedy Almond N/A Negative 9-May 7 UC Davis Greenhouse -Feb Sterile soil; (-) control N/A Negative 9-May Genetic diversity of TSWV Isolates from the Central Valley Tomatoes as well as other crops such as pepper, radicchio and lettuce, and weeds with or without obvious virus-like symptoms were collected and tested for TSWV in by RT-PCR. From some isolates, the N gene DNA fragment was amplified, cloned and sequenced to determine the genetic diversity of the TSWV in the Central Valley of California. Sequence analysis of TSWV N genes revealed a very closely related group of isolates. No major differences were found, irrespective of the host, location and year of isolation (data not shown). Similar results were obtained with TSWV sequences amplified from thrips. These results indicate that in Central California, the TSWV

25 population is still fairly homogeneous, and likely represents a geographically isolated population. Detection of TSWV in thrips and studies of thrips biology Thrips from flower samples that were collected from each monitored fields were pooled in a way to represent the biweekly population from each field, and tested by RT- PCR throughout the season in. RT-PCR tests performed with thrips in early season revealed that most of the insects were negative for TSWV, whereas lab-reared viruliferous thrips controls and some of the thrips collected from infected tomatoes were positive (data not shown). These results indicate that most of the adult thrips present in processing tomato fields early in the season are not carrying the virus. In contrast, most of the thrips samples that were collected after mid-june and early July, when incidences of TSWV were increasing in fields, tested positive for the TSWV. These results are consistent with our hypothesis that levels of TSWV inoculum early in the season are low, and that most of the thrips migrating to processing tomatoes were not carrying the virus. However, as the season progressed, TSWV was amplified in susceptible crops (e.g., tomato), and then later generations of thrips readily acquire the virus (recall that, larvae, which must acquire the virus for adults to be viruliferous, are common in tomato flowers) and can transmit the virus within or between tomato fields. This is also consistent with the relatively low levels of TSWV in tomatoes in early growing season in, despite high thrips populations in many of these fields (Fig. ). If more of these thrips were carrying the virus early in the season, it is likely that the incidence of TSWV would have been much higher. Thus, there was no correlation between thrips populations and TSW disease incidence. Comparison of thrips transmission efficiency for the TSWV-California isolates between two thrips colonies from Fresno and Yolo In these experiments, laboratory-reared male and female thrips colonies were used to conduct TSWV transmission assays. To assess transmission efficiency of male and female thrips and to compare their transmission efficiency with each other (i.e., gender and geographic origin), progeny thrips reared from Fresno and Yolo colonies were separately tested for the ability to transmit TSWV over five independent experiments. The results are summarized in Table. Based on our results, regardless of the geographic origin, male adult thrips transmitted TSWV more efficiently than female adult thrips. Interestingly, the overall transmission efficiencies of these two colonies were different, with thrips from the Fresno colony having a higher transmission rate (.6%) than thrips

26 from the Yolo colony (9.%). This is an exciting finding and indicated that these two western flower thrips colonies, at least under our experimental conditions, had significant differences in their transmission efficiency. This is also consistent with the finding that thrips populations are often higher in tomato fields in Yolo County than Fresno County, but the incidences of TSWV are higher in Fresno County than Yolo. However, because other factors that may affect the efficiency of thrips to transmit TSWV, further experimentation must be done to conclude whether it is the different transmission rates of the thrips that are responsible for different levels of TSWV in the field in these locations. Table. Summary of assessment and comparison studies in transmission efficiency of male and female thrips collected from Fresno and Yolo Counties. Fresno Yolo Trial Male Female Male Female I /9 / - - II / / / / III 9/ 7/ / / IV / / /7 / V / /9 /8 /6 Subtotal / (%) / (9.%) 7/ (%) 7/7 (8.9%) Total 6/ (.6%) /7 (9.%) Insecticide Trial The thrips insecticide trial to assess the performance of insecticides against Western flower thrips was conducted at the UC WSREC. In, several materials reduced the levels of thrips when compared with the untreated control (Table ). These materials included Radiant (either with or without Pre-Am), Athena with Beleaf and Venom. Results of the trial generally supported results of earlier trials in which Radiant was consistently among the top-performing materials and Beleaf, either alone or tank mixed with a pyrethroid, also has provided a level of control better than the untreated control. However, unlike 7 and 8 results, but similar to results, Venom also reduced thrips population densities in. In, there were clearly differences among the foliar treatments in terms of yield and percentage of fruit with TSWV symptoms (Table 6), as well as symptoms in plants (Table 7). While there also were differences in terms of symptom incidence on plants in 9, we did not see other differences that season. Furthermore, no differences in treatments vs. untreated controls were observed in. We believe that because the

27 level of TSWV around the field station was lower in, that the incidence and spread of TSWV was from mostly from in-field viruliferous thrips, which could be controlled by application the chemical. In, we believe that high amounts of virus brought in from thrips immigrating from surrounding fields may have obscured the effect of the chemicals, whereas this additional virus pressure was not present in 9 or. Regardless of the year we have never seen a reduction in disease associated with the drip-applied materials. Together, these results indicate that some materials can reduce thrips populations to some extent and reduce TSWV incidences but that grower can not rely on insecticide application for efficient control. Table. Insecticide efficacy results against Western flower thrips at WSREC in. Treatment rate fp/acre z Nymphs/ flowers y Radiant 6. fl oz + Prev-Am qt 6. c x Athena 7 fl oz + Beleaf SG.8 oz 7. c Radiant 6. fl oz 7. c Venom 7SG.89 lb 8. c HGW86. fl oz. bc Athena 7 fl oz.8 abc Requiem qts. a Untreated 7. ab z y x Treated 6 Aug with Co -pressurized back pack sprayer at gal/acre. Collected samples per plot in 7% EtOH and counted nymph and adults under dissecting scope. Means followed by the same letter are not different (LSD.. )