Seasonal Abundance of Aphids (Homoptera: Aphididae) in Wheat and Their Role As Barley Yellow Dwarf Virus Vectors in the South Carolina Coastal Plain

FIELD AND FORAGE CROPS Seasonal Abundance of Aphids (Homoptera: Aphididae) in Wheat and Their Role as Barley Yellow Dwarf Virus Vectors in the South Carolina Coastal Plain

1 1 2 2 JAY W. CHAPIN, JAMES S. THOMAS, STEWART M. GRAY, DAWN M. SMITH, AND SUSAN E. HALBERT3

J. Econ. Entomol. 94(2): 410Ð421 (2001) ABSTRACT Aphid (Homoptera: Aphididae) seasonal ßight activity and abundance in wheat, Triticum aestivum L., and the signiÞcance of aphid species as vectors of barley yellow dwarf virus were studied over a nine-year period in the South Carolina coastal plain. Four aphid species colonized wheat in a consistent seasonal pattern. Greenbug, Schizaphis graminum (Rondani), and rice root aphid, Rhopalosiphum ruﬁabdominalis (Sasaki), colonized seedling wheat immediately after crop emergence, with apterous colonies usually peaking in December or January and then declining for the remainder of the season. These two aphid species are unlikely to cause economic loss on wheat in South Carolina, thus crop managers should not have to sample for the subterranean R. ruﬁab- dominalis colonies. Bird cherry-oat aphid, Rhopalosiphum padi (L.), was the second most abundant species and the most economically important. Rhopalosiphum padi colonies usually remained below 10/row-meter until peaking in February or March. Barley yellow dwarf incidence and wheat yield loss were signiÞcantly correlated with R. padi peak abundance and aphid-day accumulation on the crop. Based on transmission assays, R. padi was primarily responsible for vectoring the predominant virus serotype (PAV) we found in wheat. Pest management efforts should focus on sampling for and suppressing this aphid species. December planting reduced aphid-day accumulation and barley yellow dwarf incidence, but delayed planting is not a practical management option. English grain aphid, Sitobion avenae (F.), was the last species to colonize wheat each season, and the most abundant. Sitobion avenae was responsible for late-season virus transmission and caused direct yield loss by feeding on heads and ßag leaves during an outbreak year.

KEY WORDS barley yellow dwarf, wheat, Rhopalosiphum padi, Sitobion avenae, Rhopalosiphum ruﬁabdominalis, Schizaphis graminum

APHIDS CAUSE SIGNIFICANT economic loss on wheat, biology and epidemiology which cannot be extrapo- Triticum aestivum L., and other cereal grains by direct lated from region to region (Burnett 1991). feeding injury (Kieckhefer and Kantack 1980, 1988; There have been few studies of the cereal aphid Pike and Schaffner 1985; Johnston and Bishop 1987; vectors of barley yellow dwarf virus in the southeast- Voss et al. 1997) and by vectoring the viruses that ern United States. McPherson and Brann (1983) found cause barley yellow dwarf (Gill 1980, Jensen and four species of aphids in Virginia wheat and barley: DÕArcy 1995). Barley yellow dwarf is a worldwide greenbug, Schizaphis graminum (Rondani); English economic problem on cereal grains (Plumb 1983) that grain aphid, Sitobion avenae (F.); R. padi; and corn leaf can reduce grain yields by stunting root and shoot aphid, Rhopalosiphum maidis (Fitch). These same growth; as well as increasing host susceptibility to species were reported in a survey of Kentucky wheat fungal pathogens, drought, and other environmental (Johnson and Hershman 1996). This article presents stress (Burnett 1984). Barley yellow dwarf epidemics information from the South Carolina coastal plain on in winter wheat in North America have generally been the annual ßight activity of cereal aphids, aphid sea- associated with fall virus transmission by the bird cher- sonal abundance in wheat, the effect of planting date ry-oat aphid, Rhopalosiphum padi (L.) (Halbert and on species abundance, barley yellow dwarf virus trans- Pike 1985, Clement et al. 1986, Araya et al. 1987, Hal- mission assays, and the correlation of aphid species bert et al. 1992). However, barley yellow dwarf man- abundance with barley yellow dwarf incidence and agement requires an understanding of local vector yield loss.

1 Department of Entomology, Edisto Research and Education Cen- Materials and Methods ter, Clemson University, 64 Research Road, Blackville, SC 29817. 2 USDA-ARS, Department of Plant Pathology, Cornell University, Suction Traps. All studies, except where noted, Ithaca, NY 14853. 3 Florida State Collection of Arthropods, Division of Plant Industry, were conducted at the Edisto Research and Education Florida Department of Agriculture and Consumer Services, 19111 SW Center near Blackville (Barnwell County), SC. Two 34th Street, Gainesville, FL 32614-7100. suction traps (Allison and Pike 1988) located within

0022-0493/01/0410Ð0421$02.00/0 ᭧ 2001 Entomological Society of America April 2001 CHAPIN ET AL.: APHID ABUNDANCE AND VECTOR ROLE 411

Table 1. Comparative aphid-day accumulation for species found on wheat; barley yellow dwarf incidence and percentage yield loss associated with aphid incidence; and the effects of planting date at Blackville, SC

Aphid-days (ϮSEM)/row-mb a Barley yellow Planting date R. c % yield loss S. avenae R. padi S. graminum dwarf /row-m ruﬁabdominalis 1990Ð91 Early-Nov. 3,404 Ϯ 496aA 629 Ϯ 98aB 740 Ϯ 304aB 33 Ϯ 22aC 2.0 Ϯ 0.16a 2.9 Ϯ 5.4a Mid-Nov. 2,478 Ϯ 476aA 962 Ϯ 168aB 113 Ϯ 25abC 0 Ϯ 0bD 2.9 Ϯ 0.44a 4.2 Ϯ 4.1a Early Dec. 2,699 Ϯ 445aA 333 Ϯ 47bB 42 Ϯ 18bC 0 Ϯ 0bD 2.1 Ϯ 0.14a 1.9 Ϯ 2.9a 1991Ð92 Early-Nov. 24,202 Ϯ 3155aA 1,174 Ϯ 184aB 1,582 Ϯ 229aB 474 Ϯ 138aC 0.8 Ϯ 0.11a 6.5 Ϯ 5.0a Mid-Nov. 23,930 Ϯ 3313aA 565 Ϯ 47bB 463 Ϯ 32abB 150 Ϯ 56aC 1.3 Ϯ 0.21a 6.6 Ϯ 2.8a Early Dec. 2,230 Ϯ 1054bA 276 Ϯ 26cAB 241 Ϯ 115bB 24 Ϯ 24bC 0.2 Ϯ 0.04b 1.9 Ϯ 1.2a 1992Ð93 Early-Nov. 8,232 Ϯ 498aA 5,868 Ϯ 275aA 508 Ϯ 32aB 41 Ϯ 20aC 8.0 Ϯ 0.31a 11.4 Ϯ 4.1ab Mid-Nov. 8,235 Ϯ 1210aA 8,203 Ϯ 1139aA 290 Ϯ 128aB 0 Ϯ 0bC 9.0 Ϯ 0.77a 20.1 Ϯ 3.8a Early Dec. 5,249 Ϯ 327bA 1,805 Ϯ 434bA 187 Ϯ 34aB 19 Ϯ 19bC 3.3 Ϯ 0.35b 6.5 Ϯ 6.1b 1993Ð94 Mid-Nov. 11,070 Ϯ 4027A 343 Ϯ 74B 447 Ϯ 103B 38 Ϯ 22C 0.1 Ϯ 0.01 4.0 Ϯ 1.3 1994Ð95 Mid-Nov. 2,794 Ϯ 360A 297 Ϯ 44B 180 Ϯ 60B 0 Ϯ 0C 0.2 Ϯ 0.03 2.7 Ϯ 1.6 1995Ð96 Mid-Nov. 65,290 Ϯ 21743A 1,897 Ϯ 418B 823 Ϯ 88B 2,282 Ϯ 1023B 1.5 Ϯ 0.15 16.7 Ϯ 1.1 1996Ð97 Mid-Nov. 6,500 Ϯ 835A 1,858 Ϯ 350B 659 Ϯ 85C 153 Ϯ 47D 4.1 Ϯ 0.20 11.9 Ϯ 1.6 1997Ð98 Mid-Nov. 1,376 Ϯ 266B 3,655 Ϯ 375A 228 Ϯ 61C 0 Ϯ 0D 1.9 Ϯ 0.22 10.6 Ϯ 1.5 1998Ð99 Mid-Nov. 9,390 Ϯ 1198A 2,738 Ϯ 228B 288 Ϯ 78D 794 Ϯ 269C 2.9 Ϯ 0.38 8.7 Ϯ 2.2 Pooled Mid-Nov. 13,782 Ϯ 3622A 2,293 Ϯ 391B 377 Ϯ 42C 371 Ϯ 149C Ñ Ñ

Means Ϯ SEM within a column, in the same year, followed by the same lower case letter are not signiÞcantly different. Means within a row followed by the same upper case letter are not signiÞcantly different; FisherÕs protected LSD (P Ͻ 0.05). Aphid-day data were analyzed as log(x ϩ 1), but reported in original scale. a Early-Nov. ϭ 1 Nov; mid-Nov. ϭ 15 Nov. (1990Ð1995), 13 Nov. (1996), 20 Nov. (1997 and 1999), and 19 Nov (1998); and early-Dec. ϭ 1 Dec. (1991), 2 Dec (1992), and 4 Dec (1993). b Cumulative seasonal aphid-days per row-meter; conversion factor for aphid-days per m2 ϭ 4.92; conversion factor for aphid-days per plant ϭ 0.017 (59.6 Ϯ 3.4 plants per row-meter across all tests); conversion factor for R. padi aphid-days per stem during jointing ϭ 0.0059 (169.7 Ϯ 4.8 stems per row-meter); and conversion factor for S. avenae aphid-days per headed stem ϭ 0.0087 (114.7 Ϯ 4.3 heads per row-meter). c Barley yellow dwarf incidence measured as symptomatic stems (mean Ϯ SEM) per row-m at kernel formation (Zadoks growth stage 70).

200 m of the experimental plots were operated con- applied in the same form in mid-February at ZadoksÕ tinually from September 1992 until June 1996 at growth stage 30 (Zadoks et al. 1974). Plots were heights of 2 m and 8 m. The high trap caught fewer of treated with propiconazole 0.12 kg(AI)/ha (Tilt, 42% each cereal aphid species than the low trap, but caught EC, Novartis Crop Protection, Greensboro, NC) at full them at the same time. Therefore, a combined catch ßag leaf emergence (GS 47) to prevent leaf rust, Puc- from both traps is presented. Trap catches, preserved cinia recondita f.sp. tritici. Bromoxynil 0.56 kg(AI)/ha in ethylene glycol, were removed weekly and brought (Buctril 33%, Rhone-Poulenc, Research Triangle, NC) into the laboratory for sorting. All aphids collected or tribenuron methyl 0.0115 kg(AI)/ha (Express 75% from suction traps were identiÞed to species by S.E.H.; DF, DuPont Agricultural Products, Wilmington, DE) however, only data for the Þve cereal aphid species were applied in January for weed control. that were also found on the wheat crop are reported Aphids were counted on at least ten 15-cm row here. lengths of wheat (two samples per experimental unit) Aphids in Wheat and Planting Date Effect. Aphids at approximately biweekly intervals. After examining were sampled in wheat, ÔCoker 9835Õ, over nine grow- the aboveground portion of plants and the soil surface ing seasons from 1990 to 1999 in a series of insecticide for aphids, we uprooted separate 15-cm row lengths efÞcacy tests. The aphids were collected from un- and examined the underground wheat stem and roots. treated control plots arranged in a randomized com- When the soil was too hard to readily uproot the plete block design with Þve replicates. The plot width plants, we used a small trowel to lift the sample. Sam- ranged from 1.6 m (eight rows on 20.3-cm spacing) to pling for subterranean aphids was conducted on ex- 4.8 m (24 rows). Plot length was 15.2 m. The seeding terior plot rows to avoid destructive sampling on the rate was 72/row-meter for the Þrst 4 yr and 79/row- interior rows, which were harvested for yield. Aphids meter thereafter. The soil type was either a Marlboro that could not be identiÞed to species in the Þeld with or Dothan loamy sand, depending on test location the unaided eye or a 10ϫ hand lens were returned to from year to year. The crop rotation for all tests was the laboratory for identiÞcation (J.W.C.). Voucher corn, peanut, corn, or wheat. Plots were disked, and specimens were deposited in the Clemson University then a Terramax II (Worksaver, LitchÞeld, IL) was Arthropod Collection and the Florida State Collection used to provide broadcast deep tillage. Potassium was of Arthropods. Wheat growth stage was recorded on applied preplant as 0-0-60 according to a soil test. each sample date. During the 1990Ð1991, 1991Ð1992, Residual phosphorus was always at the high level in and 1992Ð1993 seasons, three planting times were sam- the test Þelds, and none was applied to the wheat crop. pled: early-November, mid-November, and early- Before emergence, 33 kg/ha nitrogen was applied as December. In subsequent years only a mid-November S-25 (32% urea, 20% ammonium nitrate, 14.5% ammo- planting was sampled. Actual planting dates are foot- nium sulfate). An additional 77Ð99 kg/ha nitrogen was noted in Table 1. Plant populations were estimated 412 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 94, no. 2 before tillering by counting plants in two 1-m locations Symptomatic plants were tested by enzyme-linked per control plot. Stem counts were taken during joint- immunosorbent assays (ELISA) (Gray et al. 1998) ing by counting stems within the 15-cm row length in to determine the barley yellow dwarf virus sero- which aphids were sampled. Head counts were taken type. Serotypes are identiÞed by acronyms derived from two 1-m row samples per plot. Aphid-days were from their most efÞcient vectors. Thus, PAV isolates calculated (Ruppel 1983) for each species and plant- are primarily vectored by R. padi and S. avenae, and ing date. Analysis of variance (ANOVA) and a pro- RPV is vectored by R. padi (Lister and Ranieri tected least signiÞcant difference (LSD) test (P Յ 1995). 0.05) (PROC GLM, SAS Institute 1985) were used to Barley Yellow Dwarf Incidence and Yield Loss. compare cumulative aphid-days among species within Barley yellow dwarf incidence was estimated during planting dates and among planting dates for each spe- kernel Þll (ZadoksÕ GS 70) by two observers, each cies. Aphid-day data were transformed to log(x ϩ 1) scanning two adjacent interior plot rows (61 row- before analysis to homogenize variances, but results meter per plot) and recording the number of symp- are reported in the original scale. Pearson correlation tomatic stems. Symptomatic and asymptomatic virus coefÞcients (PROC CORR, SAS Institute 1985) were incidence in the untreated control was conÞrmed by calculated by species for total seasonal aphid-days and ELISA (Gray et al. 1998) at GS 70 in 1997, 1998, and monthly rainfall, and total aphid-days and average 1999. In 1997, 25 symptomatic stems (Þve per plot) daily temperature for each of the preceding 12 mo (1 were arbitrarily collected, and an additional 100 stems May to 30 April). We also grouped rainfall and tem- (20 per plot) were blindly sampled without regard to perature data into 3-mo seasonal intervals (spring ϭ symptoms. Thirty symptomatic stems (six per plot) March through May, summer ϭ June through August, were collected in 1998, and 200 random stems were fall ϭ September through November, winter ϭ De- sampled in 1998 and 1999. cember through February) for correlation with total Yield loss was measured as the difference between aphid-day accumulation by species. A sequential Bon- the mean yield of the untreated control and the mean ferroni test (Rice 1989) was used to evaluate the yield of plots that received an insecticide seed treat- signiÞcance of correlation coefÞcients based on the ment and mid-February foliar application of insecti- more restrictive test-wide values of P appropriate for cide. The interior six rows were harvested with an simultaneous inference. Almaco plot combine (Almaco, Nevada, IA). Grain In 1991Ð1992, six grower Þelds in the northern weight and moisture were measured, and yields were coastal plain (Darlington County) and four Þelds in adjusted to 13.5% moisture content. In 1990Ð1991, the southern coastal plain (Bamberg and Barnwell 1991Ð1992, and 1992Ð93, the insecticide treatment Counties) were also surveyed approximately biweekly used for comparison with the control was imidacloprid for aphids. In each Þeld, aphids were identiÞed and seed treatment (0.125 gm [AI]/kg seed, NTN33893 counted from a minimum of ten 15-cm row samples 240 FS, Miles Chemical, Kansas City, MO) plus foliar taken arbitrarily along a diagonal transect. disulfoton (0.82 kg [AI]/ha, DiSyston 8, emulsiÞable Virus Transmission Assays. During the 1997Ð1998 concentrate, Miles Chemical). In 1993Ð1994 and and 1998Ð1999 seasons, live aphids were collected and 1994Ð1995, the insecticide treatment was imidacloprid assayed for virus transmission approximately weekly seed treatment (0.31 g [AI]/kg seed, Gaucho 480 FS, from seedling emergence in early December until Bayer, Kansas City, MO) plus foliar disulfoton (0.82 kg mid-April. We collected the Þrst alates encountered [AI]/ha, DiSyston 8, emulsiÞable concentrate, Miles regardless of species and attempted to collect at least Chemical). In the Þnal four seasons, the insecticide 100 alates per week. If 100 alates were not found, treatment was imidacloprid seed treatment (0.31 g apterae were collected. No aphids were collected in [AI]/kg seed, Gaucho 480 FS, Bayer) plus foliar January 1999 due to very low alate availability. Aphids lambda-cyhalothrin 0.03 kg [(AI)]/ha, Karate 1, emul- were removed from wheat with an artistÕs brush or by siÞable concentrate, Zeneca, Wilmington, DE). Foliar grasping alates by the wings with Þne forceps. Alates treatments were applied with a tractor-mounted, CO2- were difÞcult to Þnd on jointed wheat and therefore, pressurized boom using Tee Jet 11003 nozzles (Spray- beginning on 11 March 1999, alates were collected by ing Systems, Wheaton, IL) delivering 187 liters/ha at beating a row of wheat against a 32 by 22 by 5-cm 2.1 kg/cm2. Seed treatments were applied with a Hege aluminum baking pan which was painted yellow for 11 seed treater (Hege, Colwich, KS) using a 25-ml total aphid visibility. Individual aphids were placed directly volume aqueous dilution on 3.2 kg seed lots. in 15 by 100-mm plastic tubes containing a ÔCoast ANOVA and a protected LSD test (P Յ 0.05) BlackÕ oat (Avena byzantina Koch) seedling which had (PROC GLM, SAS Institute 1985) were used to com- been germinated on moist sterile cotton, and shipped pare barley yellow dwarf incidence and yield loss overnight to S.M.G. Plants were then transplanted to among planting dates. Pearson correlation coefÞcients soil, and the aphid was caged on the plant for a 96- to (PROC CORR, SAS Institute 1985) were calculated 120-h inoculation access period. Subsequently all for peak seasonal aphid species abundance and barley aphids were removed and identiÞed to species yellow dwarf incidence, peak seasonal species abun- (D.M.S.). The plants were fumigated, moved to a dance and yield loss, species aphid-day accumulation greenhouse, and evaluated over a 4- to 6-wk interval and barley yellow dwarf incidence, species aphid-day for the stunting and marginal leaf yellowing or accumulation and yield loss, and barley yellow dwarf reddening characteristic of barley yellow dwarf. incidence and yield loss. Correlation coefÞcients were April 2001 CHAPIN ET AL.: APHID ABUNDANCE AND VECTOR ROLE 413 also calculated for barley yellow dwarf incidence and shown). Four apterous specimens of R. maidis were the monthly and seasonal rainfall and temperature collected in 1991, but colonies never became estab- variables previously described. lished on wheat. A total of four apterous specimens of Sipha sp. were also collected on wheat. Results Sitobion avenae was the most prevalent aphid species on wheat based on peak abundance (Fig. 2D) and Suction Traps. The three ßight activity periods de- aphid-day accumulation (Table 1). The only excep- tected for S. graminum are illustrated in the 1993Ð1994 tion occurred in 1997Ð1998 when R. padi accumulated season (Fig. 1A) when there was a fall ßight in Oc- more aphid-days. When mid-November planting dates tober through December; a spring ßight in March and were pooled over years, S. avenae accumulated about April; and a late summer ßight in July through Sep- six times as many aphid-days as R. padi (Table 1). tember. Rice root aphid, Rhopalosiphum rufiabdomi- Rhopalosiphum padi was typically the second most nalis (Sasaki), alates were trapped throughout the abundant aphid species (Fig. 2; Table 1). Comparing year except for a consistently low level of ßight activity mid-November planting dates, R. padi accumulated from January through March (Fig. 1B). The most signiÞcantly more aphid-days than S. graminum or R. consistent seasonal ßight activity period for R. padi rufiabdominalis in Þve of nine seasons, and when mid- was during March and April (Fig. 1C). A late-fall to November planting dates were pooled, R. padi accu- early-winter peak was also detected during December mulated about six times as many aphid-days as S. and January of 1992Ð1993, and in November and De- graminum. Schizaphis graminum had greater aphid- cember of 1994. A deÞnitive summer peak was re- day accumulations than R. rufiabdominalis in eight of corded only in late July of 1994. Rhopalosiphum maidis nine seasons, however this comparison was based, by had two consistent annual ßight activity periods: a necessity, on different sampling procedures. summer peak in July to August and a lesser fall peak We found no signiÞcant correlation, for any species, in October to November (Fig. 1D). The few R. maidis between total seasonal aphid abundance and monthly alates collected in April to May suggests the presence rainfall or temperature (␣ ϭ 0.05, k ϭ 12, P Ͼ 0.004). of a spring ßight as well. The only consistent S. avenae Nor were there any correlations between aphid abun- ßight activity detected was a spring ßight in March dance and seasonal (3-mo interval) rainfall or tem- through May (Fig. 1E). perature (␣ ϭ 0.05, k ϭ 4, P Ͼ 0.0125). Aphid Colonization of Wheat. Schizaphis graminum In the 1991Ð1992 survey of grower Þelds, seasonal was typically the Þrst aphid species detected in wheat, species abundance followed the same general pattern and they were generally found within 7Ð10 d of crop from Þeld-to-Þeld and therefore the data were com- emergence (Fig. 2A). S. graminum populations usually bined to show a composite of seasonal abundance peaked in December or January and then declined to (Fig. 3) which was similar to that of our experimental Ͻ10/row-meter for the remainder of the growing season. R. rufiabdominalis was also an early colonizer, plots. We collected no R. maidis from grower Þelds. with alates found immediately after plant emergence Planting Date Effect. Aphid-day accumulation for (Fig. 2B). Colonies of apterous R. rufiabdominalis S. graminum was signiÞcantly greater for early- were nearly always found on plant stems just below November planted wheat than December-planted the soil line, and populations typically peaked in De- wheat in two of three seasons evaluated (Table 1). cember and then declined for the rest of the season Early-November planting increased R. rufiabdomina- (Fig. 2B). An exception occurred in 1995Ð1996 when lis abundance relative to mid-November planting in apterous R. rufiabdominalis populations of Ͼ80/row- two of three seasons, and in all three seasons relative meter were found on the above ground stems and to December planting. Planting in December reduced foliage in March. Rhopalosiphum padi alates usually R. padi abundance compared with early-November arrived on wheat seedlings after S. graminum and R. plantings in all three test seasons and reduced abun- rufiabdominalis alates. There were two periods of alate dance compared with mid-November plantings in two R. padi activity on wheat, a fall to early winter alate of three seasons. Delaying planting from early to mid- population usually peaking in December or early Jan- November reduced R. padi abundance only in 1991Ð uary, and a spring population typically peaking in late 1992. Similarly, S. avenae abundance was reduced in February or March (Fig. 2C). The fall alate activity December-planted wheat relative to early or mid- corresponded with development of apterous R. padi November plantings in two of three seasons. Delaying colonies in December (GS 21). The numbers of these planting from early-November to mid-November did apterous R. padi colonies generally remained below not reduce S. avenae abundance. Delaying planting 10/row-meter until early February (Fig. 2C), then until December reduced barley yellow dwarf inci- increased during February (GS 23Ð30), and usually dence in both 1991Ð1992 and 1992Ð1993, but had no peaked in March (GS 32). S. avenae was the last aphid effect in 1990Ð1991 (Table 1). The percentage yield species to colonize wheat. In most years, apterous S. loss from virus was signiÞcantly affected by planting avenae were Þrst found in January, increased rapidly date in 1992Ð1993, when the early-December planting in March and April, and peaked in April or May (Fig. had less yield loss than mid-November planted wheat 2D). A few alate specimens of R. maidis were collected (Table 1). The1992Ð1993 early-November planting in December and February, but these collections were suffered Ϸ40% yield reduction from March freeze sporadic and never exceeded 0.7/row-meter (data not injury. The severity of this cold injury minimized the 414 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 94, no. 2

Fig. 1. Seasonal ßight activity of cereal aphids as measured by the combined weekly 2- and 8-m suction trap catch from October 1992 to May 1996 at Blackville, SC. April 2001 CHAPIN ET AL.: APHID ABUNDANCE AND VECTOR ROLE 415

Fig. 2. Seasonal abundance of alate and apterous aphids on wheat in Blackville, SC. Date of growth stage is the 9-yr average (GS 21 ϭ initial tillering, 31 ϭ jointing, 55 ϭ 50% head emergence, and 70 ϭ kernel formation; Zadoks et al. 1974). estimate of yield loss from virus infection for this alate specimen (0.4%) transmitted barley yellow planting date. dwarf virus (Table 2). Similarly, only one of 347 (0.3%) Virus Transmission Assays. Of 246 alate and eight alate R. ruﬁabdominalis transmitted barley yellow apterous S. graminum assayed over two seasons, one dwarf virus. Only eight R. maidis, all alates, were 416 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 94, no. 2

and March, and S. avenae predominated in April (Ta- ble 2). Barley Yellow Dwarf and Yield Loss. Although vi- sual symptoms often underestimate virus infection (Clement et al. 1986), ELISA indicated that the symptomatic incidence we observed in April was a mean- ingful measure of barley yellow dwarf incidence. From 1997 to 1999, 79/80 (99%) of symptomatic stems tested positive for barley yellow dwarf virus (77 PAV, 1 RPV, 1 PAVϩRPV), whereas only 10/492 (2%) of randomly selected asymptomatic stems were positive for barley yellow dwarf virus (all PAV). Eight of the 500 randomly selected stems over this interval were symptomatic, and all tested positive (7 PAV, 1 RPV). In addition, 213/251 (85%) of symptomatic stems and only 15/247 (6%) of asymptomatic stems taken at GS 70 from the experimental plots and farm surveys during 1991Ð1993 tested positive (Gray et al. 1998). Barley yellow dwarf incidence was highly correlated with percentage yield loss (r ϭ 0.67, n ϭ 15, P Ͻ 0.01) over the nine-year test period. Barley yellow dwarf incidence and yield loss were signiÞcantly correlated with both R. padi peak abundance and R. padi aphid-day accumulation (Table 3). There was no correlation of barley yellow dwarf incidence or yield loss with any of the other three aphid species (Table 3). Fig. 3. Composite of seasonal aphid species abundance in We found no correlation of barley yellow dwarf in- 10 coastal plain growersÕ Þelds (Bamberg County, two Þelds; cidence with rainfall or average daily temperature Barnwell County two Þelds; Darlington County, six Þelds); ␣ ϭ 1991Ð1992 season. when months were considered individually ( 0.05, k ϭ 12, P Ͼ 0.004) or when grouped into 3-mo seasonal intervals (␣ ϭ 0.05, k ϭ 4, P Ͼ 0.0125). assayed; but one of these (12.5%) transmitted the A 40-stem sample was taken from control plots in predominant PAV serotype. Of 762 R. padi alates, 37 each test in April or May to conÞrm that Hessian ßy, (4.8%) transmitted barley yellow dwarf (33 PAV, Mayetiola destructor (Say) (Diptera: Cecidomyiidae), 4 RPV) and 9/571 (1.6%) of apterae transmitted the infestation was non-economic. Coker 9835 wheat is PAV serotype. Sitobion avenae transmission rates were resistant to local races of Hessian ßy, and in all of our 27/389 (6.9%) for alates (25 PAV, 2 PAVϩRPV) and tests the Hessian ßy stem infestation of untreated plots 9/351 (2.6%) for apterae (all PAV). In March and was Ͻ3%. Even a 10% level of spring Hessian ßy in- April of 1998, apterous R. padi and S. avenae collected festation would not cause measurable wheat yield from symptomatic plants had high virus transmission reduction (Buntin 1999). Cereal leaf beetle, Oulema rates (Table 2 footnotes). Due to their relative sea- melanopus (L.) (Coleoptera: Chrysomelidae), is not sonal abundance, S. graminum and R. ruﬁabdominalis an economic problem in the study area, and in all tests alates were assayed primarily in December. R. padi cereal leaf beetle infestation remained below one per- was the most available species in January, February, cent of stems. Given the absence of other known pest

Table 2. Barley yellow dwarf virus transmission by alate and apterous aphids collected from wheat; Blackville, SC, 1997–99

No. of aphids assayed (no. transmitting, if Ͼ 0) Sample month S. graminum R. ruﬁabdominalis R. padi R. maidis S. avenae Alate Apt. Alate Alate Apt. Alate Alate Apt. Dec. 208 0 287 168 (2) 0 5 18 0 Jan. 13 0 9 (1) 148 (2) 0 0 13 0 Feb. 10 5 33 193 (17) 125 1 75 (1) 92 Mar.a 11 3 17 209 (11) 383 (5) 2 (1) 82 (2) 166 (2) Apr.b 4 (1) 0 1 39 (5) 63 (4) 0 201 (24) 93 (7) Dec.ÐApr. 246 (1) 8 347 (1) 762 (37) 571 (9) 8 (1) 389 (27) 351 (9) % transmission 0.4 0.0 0.3 4.8 1.6 12.5 6.9 2.6 Serotypec 1 PAV Ñ 1 PAV 33 PAV 9 PAV 1 PAV 25 PAV 9 PAV 4 RPV 2 PAV ϩ RPV

a An additional 15/30 (50%) R. padi and 1/4 (25%) S. avenae collected from symptomatic plants on 24 March 1998 transmitted PAV. b An additional 28/52 (54%) R. padi and 3/26 (11.5%) S. avenae collected from symptomatic plants on 1 April 1998 transmitted PAV. c Serotype acronyms originate from their most efÞcient vectors (PAV ϭ R. padi and S. avenae vectored; RPV ϭ R. padi vectored). April 2001 CHAPIN ET AL.: APHID ABUNDANCE AND VECTOR ROLE 417

Table 3. Correlation of aphid species’ peak seasonal abun- Rhopalosiphum maidis. R. maidis is anholocyclic dance and cumulative aphid-days with barley yellow dwarf inci- and found throughout the world on over 30 genera of dence and percentage grain yield loss in wheat; Blackville, SC, 1991–99 Poaceae (Halbert and Voegtlin 1995). The minor April and May ßight activity we found probably represents r (P) alate colonization of maize, Zea mays L., and the major Aphid species abundancea Barley yellow dwarfb Yield loss (%) ßight activity in July and August is the abandonment S. graminum of maize as it matures. We (J.W.C. and J.S.T.) have peak abundance Ϫ0.07 (0.82) 0.18 (0.53) observed large colonies (up to 456.1 Ϯ 212.2 per plant) aphid-days Ϫ0.12 (0.66) 0.18 (0.51) of this species on maize in July and August in South R. rufiabdominalis Carolina. The fall ßight activity of R. maidis occurs peak abundance Ϫ0.13 (0.65) 0.48 (0.08) aphid-days Ϫ0.15 (0.59) 0.45 (0.09) largely before wheat has emerged. Although R. maidis R. padi alates are capable of vectoring the local PAV serotype, peak abundance 0.72 (0.004)* 0.68 (0.007)* as shown by our capture of an infective alate in March aphid-days 0.90 (0.0001)** 0.81 (0.0003)** (and Gray et al. 1998), the consistently low popula- S. avenae peak abundance Ϫ0.16 (0.59) 0.44 (0.12) tions we found on wheat demonstrate that this species aphid-days Ϫ0.17 (0.55) 0.44 (0.10) is not an economic problem in South Carolina wheat production. However, in Virginia, R. maidis comprised Pearson correlation coefÞcients (n ϭ 15). * indicates signiÞcant 50% of fall species abundance, and substantial popu- sequential Bonferroni test (␣ ϭ 0.05, k ϭ 4, P Ͻ 0.0125). ** indicates signiÞcant sequential Bonferroni test (␣ ϭ 0.01, k ϭ 4, P Ͻ 0.0025). lations persisted through March (McPherson and a Peak abundance measured as aphids per row-meter; aphid-days as Brann 1983). deÞned by Ruppel (1983). Rhopalosiphum rufiabdominalis. R. rufiabdominalis b Barley yellow dwarf incidence measured as symptomatic stems is virtually worldwide in distribution and, other than per row-meter at ZadoksÕ GS 70. in Japan, is reported to be anholocyclic on a very wide range of secondary hosts (Jedlinski 1981, Blackman species, we attributed any yield response from insec- and Eastop 1984). This aphid prefers damp roots (Hal- ticide treatment to aphids and barley yellow dwarf. bert and Voegtlin 1995). The almost continuous ßight activity we recorded for R. rufiabdominalis may reßect Discussion its host diversity and the proximity of swamps. Rho- palosiphum rufiabdominalis is infrequently reported Schizaphis graminum. S. graminum occurs on a wide on wheat in the United States, yet we commonly found variety of grasses and cereal crops and is probably alates and apterous colonies both in our experimental anholocyclic (asexual) in the southern United States plots and grower Þelds throughout the coastal plain. It (Blackman and Eastop 1984, Halbert and Voegtlin seems likely that subterranean apterous R. rufiab- 1995). In Virginia, this species was a substantial and rather uniform proportion of the total aphid popula- dominalis have been overlooked (Halbert and Voegt- tion throughout the growing season (McPherson and lin 1995) or perhaps misidentiÞed as R. padi. The alates Brann 1983), whereas we found a consistent seasonal of R. rufiabdominalis also may be misidentiÞed as R. pattern of greatest S. graminum abundance in the fall padi or go undetected if sampling is not conducted and early winter. Schizaphis graminum feeding injury early enough in the fall. The October through De- is seldom reported as an economic problem on wheat cember ßight activity of R. rufiabdominalis indicates in South Carolina, which may be related to the fact that this species, like S. graminum, is available to col- that the crop emerges after most of the fall S. grami- onize fall-planted wheat as soon as the crop emerges, num ßight activity. Earlier plantings of cereals in the and this corresponds with the consistent early pres- coastal plain would be exposed to greater S. graminum ence of R. rufiabdominalis alates on seedling wheat colonization as evidenced by our data on planting date and apterous colonies on subterranean wheat stems. effects. The spring S. graminum ßight activity probably Rhopalosiphum rufiabdominalis is an efÞcient vector represents movement to summer hosts; and the July, of the predominant PAV serotypes in South Carolina August and September ßight activity may be indica- (Gray et al. 1998). Jedlinski (1981) found that sub- tive of crowding on, and movement from, summer terranean apterae of R. rufiabdominalis are capable of grass hosts. One population of S. graminum was found overwintering on seedling wheat and transmitting bar- to be an ineffective vector of the PAV serotype in ley yellow dwarf virus in Illinois and suggested that South Carolina (Gray et al. 1998), and other South undetected colonies may explain barley yellow dwarf Carolina populations have subsequently been found outbreaks in the absence of conspicuous aphid pop- to transmit PAV inefÞciently (S.M.G., unpublished ulations. Rhopalosiphum rufiabdominalis may play data). In our assay, the single S. graminum transmission some role in vectoring virus into wheat Þelds. How- of barley yellow dwarf occurred in April when there ever, the lack of correlation between R. rufiabdomi- would be little yield effect. The lack of correlation nalis abundance and virus incidence or yield loss in between S. graminum abundance and barley yellow our study leads us to conclude that this species is not dwarf incidence would indicate that the consistent fall responsible for most barley yellow dwarf virus trans- colonization of seedling wheat by this species is rel- mission in South Carolina. This is signiÞcant because atively unimportant in local barley yellow dwarf ep- colonies of this cryptic species are difÞcult for crop idemics. managers to sample for and identify. 418 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 94, no. 2

Rhopalosiphum padi. In northern North America, R. peak in our suction trap data represents the movement padi is holocyclic (an autumnal sexual generation pro- of R. padi off maize as it senesces. It is unclear how R. duces overwintering eggs), using Prunus as a primary padi bridges the period between maize senescence in host and many genera of Poaceae as secondary hosts August and the emergence of wheat and other small (Halbert et al. 1992, Halbert and Voegtlin 1995). An- grains in late October to December. Although volun- holocyclic clones of R. padi have been observed in the teer maize is abundant several weeks after harvest, this midwestern United States (Halbert and Voegtlin resource is not yet available when R. padi alates mi- 1995), and R. padi is probably permanently anholo- grate from the maturing maize crop. Adults of R. padi cyclic in the southeastern United States (S.E.H.). We migrating from maize have a mean life span of only collected a few males (two to nine per yr) in suction 14 d (Blackmer and Bishop 1991). Halbert et al. (1992) traps from late February to May. We also collected a pointed out that in Idaho, volunteer cereals form a single male in December 1992 and November 1995. bridge between maize and fall-planted grain. High The presence of males in the fall indicates that a sexual populations of R. padi have also been found on vol- cycle may be possible, but this is not likely to be a unteer wheat in Indiana (Araya et al. 1987) and else- signiÞcant consideration in local barley yellow dwarf where, but most wheat Þelds in South Carolina are dynamics. double-cropped with soybean, and herbicides used for Rhopalosiphum padi is an efÞcient vector of the grass suppression in soybean typically control volun- predominant local PAV serotype (Gray et al. 1998), teer wheat. We have not found R. padi on volunteer and we collected infective alate R. padi from Decem- wheat or maize in the fall. Either R. padi exists at low ber through April. The 4.8% alate infectivity rates we population levels on these volunteer crops or there is observed is within the 0Ð17% range summarized for another host that bridges the gap before cereal emer- other studies in North America (Halbert and Pike gence. An alternative explanation is that R. padi alates 1985). The high degree of correlation we found be- migrate in the fall from a distant host (Clement et al. tween R. padi abundance and both virus incidence 1986). and yield loss indicates that this species is primarily McPherson et al. (1988) related increased barley responsible for barley yellow dwarf outbreaks in the yellow dwarf incidence and aphid populations to South Carolina coastal plain. Our aphid assays suggest warm fall and winter temperatures in Virginia. We that relatively low levels of viruliferous R. padi alates were unable to Þnd any temperature relationship un- establish infection foci in the late fall and early winter. der South Carolina conditions. Rhopalosiphum padi Rhopalosiphum padi colonies then spread the infec- populations in cereals have also been positively cor- tion during February and March. related with preceding summer rainfall (Plumb 1995, Rhopalosiphum padi alates ßy to wheat after the Maudsley et al. 1996), however, we found no signiÞ- peak of R. rufiabdominalis migration, but the timing cant correlation. In our studies, barley yellow dwarf and abundance of fall R. padi migrants was less con- incidence was primarily dependent on the develop- sistent than for other aphids. The December and Jan- ment of R. padi colonies during February and March. uary ßight detected by suction traps in 1992Ð1993 The factors that regulate aphid populations over this preceded the highest level of R. padi infestation, bar- relatively short interval are apparently more complex ley yellow dwarf incidence, and yield loss that we than our measures of temperature and rainfall. measured. The source of these alates is unknown, and Pike and Schaffner (1985) reported no yield reduc- we did not Þnd many to be viruliferous until the later tions when nonviruliferous R. padi populations ex- ßights recorded in February and March. The consis- ceeded 150/plant and accumulated over 3,400 aphid- tent alate activity in February and March may repre- days per plant on tillered winter wheat. However, Voss sent both intra- and interÞeld movement of R. padi. et al. (1997) reported a 4.7% yield loss when nonviru- We have observed alatoid nymphs of this species in liferous R. padi accumulated 300 aphid-days per stem wheat as early as mid-February. The early part of this during a 10-d interval at the boot stage (Zadoks GS late winter to spring alate activity may be movement 40). In our study, the highest R. padi populations within the Þeld and from one grain Þeld to another, (210/row-meter, 8,200 aphid-days/row-meter) oc- while the latter part of this ßight activity represents curred in the 1992Ð1993 season. This is equivalent to abandonment of wheat, and a migration to maize and only 3.5 aphids per plant, or 137 R. padi aphid-days per other grass hosts. We have found R. padi on maize (up plant, or 48 aphid-days per stem, given our plant pop- to 60.8 Ϯ 21.9 per stalk) in July and August in South ulations of 60/row-meter and stem populations of 170/ Carolina. Maize is known to be an important barley row-meter during jointing. These relatively low levels yellow dwarf virus reservoir and a host of R. padi in of R. padi occurred primarily on jointing wheat, yet we many parts of the world (Brown et al. 1984, Clement measured a yield loss of 20.1% in 1993. It seems highly et al. 1986, Coceano and Peressini 1989, Blackmer and unlikely that the levels of R. padi encountered in our Bishop 1991, Halbert et al. 1992). A diversity of pasture study caused any direct injury to the crop, and we grasses also serve as R. padi hosts and reservoirs of the attribute the yield reduction associated with this pest PAV serotype, but generally these grasses are not to barley yellow dwarf. considered a signiÞcant source of infection due to Sitobion avenae. S. avenae is holocyclic on cereals in their very low aphid populations (Henry and northern North America, but thought to be primarily Dedryver 1991). In South Carolina, maize is typically anholocyclic in the southeastern United States harvested by 1 September. The July to August alate (S.E.H.). We collected two male S. avenae specimens April 2001 CHAPIN ET AL.: APHID ABUNDANCE AND VECTOR ROLE 419 in suction traps during March, but such spring collec- in reduced yield potential and the risk of being unable tions are probably not indicative of sexual reproduc- to plant due to wet soils. Thus, barley yellow dwarf is tion. The single major annual ßight activity period we an additional risk factor to discourage planting before recorded when alates left maturing wheat in April was the optimal agronomic interval, but delaying planting similar to the pattern reported by Halbert et al. (1992) past 1 December is not a pragmatic management op- and others. tion. Sitobion avenae is a less efÞcient vector of the South The degree of correlation between R. padi popu- Carolina PAV serotype than either R. padi or R. rufi- lation levels and yield loss from barley yellow dwarf abdominalis (Gray et al. 1998). The lack of correlation indicates that it may be feasible to develop economic between S. avenae abundance and virus incidence thresholds for R. padi and barley yellow dwarf man- suggests that S. avenae is not an important vector in agement. Kendall and Chinn (1990) and Gillet et al. South Carolina. However, infective alates and apterae (1990) found that measures of aphid infestation gave of S. avenae that we collected in late March and April the best predictions of subsequent virus infection. demonstrate that this species can spread virus within Scouting resources should be focused on the mid- wheat Þelds before senescence. Late-transmitted vi- January to mid-March interval when R. padi colonies rus may not become symptomatic before plant senes- are most likely to spread barley yellow dwarf. Al- cence and therefore could not be correlated with S. though barley yellow dwarf virus infection of younger avenae abundance in our study. Late-transmitted bar- plants generally causes greater yield reductions (Gill ley yellow dwarf also may not give a positive ELISA 1980), early-season scouting should be less critical test based on leaf tissue (S.M.G. and D.M.S; unpub- under South Carolina conditions because rates of pri- lished data). Gill (1980) documented signiÞcant yield mary infection by R. padi are relatively low and S. and quality loss from inoculation of spring wheat at graminum and R. rufiabdominalis populations do not jointing, but in our experiments barley yellow dwarf seem to play a major role in virus spread. Knowing the virus transmission after mid-March (GS 40) has had a source of late-fall and early winter R. padi migrants minimal effect on wheat yield (unpublished data). and an improved understanding of other risk factors Sitobion avenae may play a signiÞcant role, along with such as tillage, crop rotations, row spacing, and R. padi, in movement of barley yellow dwarf virus weather may also help determine the barley yellow from wheat to summer hosts such as maize (Blackmer dwarf risk to individual grower Þelds (Irwin and and Bishop 1991). We have found S. avenae in maize Thresh 1990). Aphid species identiÞcation is likely to at populations of up to 8.8 Ϯ 2.2 per stalk during July be important because including nonvectors or less and August. Although there was no measurable rela- important vectors in threshold counts could lead to tionship between S. avenae abundance and yield loss unnecessary or poorly timed insecticide treatments. in our study, this is not to say that S. avenae cannot Although R. rufiabdominalis colonies are common on cause signiÞcant direct injury loss. The high S. avenae underground wheat stems, our work indicates that in populations in the 1995Ð1996 season corresponded South Carolina, crop consultants can make appropri- with a year of relatively low barley yellow dwarf in- ate barley yellow dwarf and aphid management de- cidence (Fig. 2D, Table 1). Considering this level of cisions on wheat without sampling underground plant barley yellow dwarf incidence, we measured a dispro- parts. Sitobion avenae is secondary in importance to R. portionately large yield loss of 848 kg/ha (16.7%) in padi as a barley yellow dwarf vector, but S. avenae is 1996. We attribute most of this loss to S. avenae pop- responsible for some late-season barley yellow dwarf ulations of 35.8 Ϯ 7.2 per head at GS 70. This would be transmission and may vector virus to summer hosts. consistent with yield losses of 12Ð13.9% reported from Sitobion avenae can also cause substantial yield loss S. avenae populations of 5Ð10 per head at head emer- from direct feeding injury in outbreak years, but other gence to anthesis (Shaoyou et al. 1986, George and cereal aphids are unlikely to cause signiÞcant direct Gair 1979, Oakley and Walters 1994) and a 28% loss injury to wheat in South Carolina. from 44 S. avenae per stem at the dough stage (GS 85) (Johnston and Bishop 1987). Management Implications. In summary, the pri- Acknowledgments mary risk factor for barley yellow dwarf incidence and We thank Manya Stoetzel for verifying identiÞcations of aphid related yield loss on wheat in South Carolina is reference specimens in the Þrst year of this study. We also the presence of R. padi colonies on the crop. There- thank Monica Wieke (University of Idaho, Aberdeen) and fore barley yellow dwarf management programs Gay Donaldson-Fortier and Paul A. Shelley (Florida State should focus on the feasibility of either avoiding or Collection of Arthropods) for technical assistance. Partial detecting and suppressing this species. Delaying plant- support was provided by Smith-Lever 3(d) IPM funding. This ing until December reduced barley yellow dwarf in- is Contribution No. 4026 of the S.C. Agricultural Experiment cidence and aphid-day accumulation of R. padi and Station, supported by the Hatch Act and state funds. other potential pests of wheat in our study, similar to previous reports (Plumb 1988, McGrath and Bale 1990). However, the optimal planting interval in our References Cited area is 15 November to 1 December. Earlier planting Allison, D., and K. S. Pike. 1988. An inexpensive suction trap increases the risk of cold injury in March and April as and its use in an aphid monitoring network. J. Agric. well as Hessian ßy damage, and later planting results Entomol. 5: 103Ð107. 420 JOURNAL OF ECONOMIC ENTOMOLOGY Vol. 94, no. 2

Araya, J. E., J. E. Foster, and S. E. Cambron. 1987. A study Jensen, S. G., and C. J. D’Arcy. 1995. Effects of barley yellow of the biology of Rhopalosiphum padi (Homoptera: Aphi- dwarf on host plants, pp. 55Ð74. In C. J. DÕArcy and P. A. didae) in winter wheat in northwestern Indiana. Gt. Burnett [eds.], Barley yellow dwarf 40 years of progress. Lakes Entomol. 20: 47Ð50. APS, St. Paul, MN. Blackman, R. L., and V. F. Eastop. 1984. Aphids on the Johnson, D. W., and D. E. Hershman. 1996. Survey of cereal worldÕs crops: an identiÞcation guide. Wiley, Chichester. aphids in Kentucky wheat Þelds: common species and Blackmer, J. L., and G. W. Bishop. 1991. Population dynam- distribution. Trans. Ky. Acad. Sci. 57: 15Ð17. ics of Rhopalosiphum padi (Homoptera: Aphididae) in Johnston, R. L., and G. W. Bishop. 1987. Economic injury corn in relation to barley yellow dwarf epidemiology in levels and economic thresholds for cereal aphids (Ho- southwestern Idaho. Environ. Entomol. 20: 166Ð173. moptera: Aphididae) on spring-planted wheat. J. Econ. Brown, J. K., S. D. Wyatt, and D. Hazelwood. 1984. Irrigated Entomol. 80: 478Ð482. corn as a source of barley yellow dwarf virus and vector Kendall, D. A., and N. E. Chinn. 1990. A comparison of in eastern Washington. Phytopathology 74: 46Ð49. vector populations indices for forecasting barley yellow Buntin, G. D. 1999. Hessian ßy (Diptera: Cecidomyiidae) dwarf virus in autumn sown cereal crops. Ann. Appl. Biol. injury and loss of winter wheat grain yield and quality. 116: 87Ð102. J. Econ. Entomol. 92: 1190Ð1197. Kieckhefer, R. W., and B. H. Kantack. 1980. Yield losses in Burnett, P. A. 1984. Preface, pp. 6Ð13. In P. A. Burnett [ed.], spring wheat in South Dakota caused by cereal aphids. Barley Yellow Dwarf, a Proceedings of the Workshop. J. Econ. Entomol. 73: 582Ð585. CIMMYT, Mexico. Kieckhefer, R. W., and B. H. Kantack. 1988. Yield losses in Burnett, P. A. 1991. Breeding for resistance to barley yellow winter grains caused by cereal aphids (Homoptera: Aphi- dwarf, pp. 179Ð193. In D. C. Peters, J. A. Webster, and C. S. didae) in South Dakota. J. Econ. Entomol. 81: 317Ð321. Chlouber [eds.], Proceedings, Aphid-plant interactions: Lister, R. M., and R. Ranieri. 1995. Distribution and eco- populations to molecules. 12Ð17 August 1990, Stillwater, nomic importance of barley yellow dwarf, pp. 29Ð53. In Oklahomas Miscellaneous Publication 132. Oklahoma C. J. DÕArcy and P. A. Burnett [eds.], Barley yellow dwarf: State University Agricultural Experiment Station, Still- 40 years of progress. APS, St. Paul, MN. water, OK. Maudsley, S. J., A. Mackenzie, J. I. Thacker, and A.F.G. Dixon. Clement, D. L., R. M. Lister, and J. E. Foster. 1986. ELISA- 1996. Density dependence in cereal aphid populations. based studies on the ecology and epidemiology of barley Ann. Appl. Biol. 128: 453Ð463. yellow dwarf virus in Indiana. Phytopathology 76: 86Ð92. McGrath, P. F., and J. S. Bale. 1990. The effects of sowing date and choice of insecticide on cereal aphids and barley Coceano, P. G., and S. Peressini. 1989. Colonisation of maize yellow dwarf virus epidemiology in Northern England. by aphid vectors of barley yellow dwarf virus. Ann. Appl. Ann. Appl. Biol. 117: 31Ð43. Biol. 114: 443Ð447. McPherson, R. M., and D. E. Brann. 1983. Seasonal abun- George, K. S., and R. Gair. 1979. Crop loss assessment on dance and species composition of aphids (Homoptera: winter wheat attacked by the grain aphid, Sitobion avenae Aphididae) in Virginia small grains. J. Econ. Entomol. 76: (F.), 1974Ð77. Plant Pathol. 28: 143Ð149. 272Ð274. Gill, C. C. 1980. Assessment of losses on spring wheat nat- McPherson, R. M., T. M. Starling, and H. M. Camper Jr. 1988. urally infected with barley yellow dwarf virus. Plant Dis. Fall and early spring aphid (Homoptera: Aphididae) pop- 64: 197Ð203. ulations affecting wheat and barley production in Vir- Gillet, H., C. A. Dedryver, Y. Robert, A. Gamon, and J. S. ginia. J. Econ. Entomol. 79: 827Ð832. Pierre. 1990. Assessing risk of primary infection of ce- Oakley, J. N., and K.F.A. Walters. 1994. A Þeld evaluation of reals by barley yellow dwarf virus in autumn in the different criteria for determining the need to treat winter Rennes basin of France. Ann. Appl. Biol. 117: 237Ð251. wheat against the grain aphid Sitobion avenae and the Gray, S. M., J. W. Chapin, D. M. Smith, N. Banerjee, and J. S. rose-grain aphid Metopolophium dirhodum. Ann. Appl. Thomas. 1998. Barley yellow dwarf luteoviruses and Biol. 124: 195Ð211. their predominant vectors in winter wheat grown in Pike, K. S., and R. L. Schaffner. 1985. Development of au- South Carolina. Plant Dis. 82: 1328Ð1333. tumn populations of cereal aphids, Rhopalosiphum padi Halbert, S. E., and K. S. Pike. 1985. Spread of barley yellow (L.) and Schizaphis graminum (Rondani) (Homoptera: dwarf virus and relative importance of local aphid vectors Aphididae) and their effects on winter wheat in Wash- in central Washington. Ann. Appl. Biol. 107: 387Ð395. ington State. J. Econ. Entomol. 78: 676Ð680. Halbert, S. E., and D. Voegtlin. 1995. Biology and taxonomy Plumb, R. T. 1983. Barley yellow dwarf virus - a global of vectors of barley yellow dwarf viruses, pp. 217Ð258. In problem, pp. 185Ð198. In R. T. Plumb and J. M. Thresh C. J. DÕArcy and P. A. Burnett [eds.], Barley yellow dwarf: [eds.], Plant virus epidemiology. Blackwell, London. 40 years of progress. APS, St. Paul, MN. Plumb, R. T. 1988. Opportunities for the integrated control Halbert, S. E., B. J. Connelly, G. W. Bishop, and J. L. Black- of barley yellow dwarf viruses in UK. Aspects Appl. Biol. mer. 1992. Transmission of barley yellow dwarf virus by 17: 153Ð161. Þeld collected aphids (Homoptera: Aphididae) and their Plumb, R. T. 1995. Epidemiology of barley yellow dwarf in relative importance in barley yellow dwarf epidemiology Europe, pp. 107Ð127. In C. J. DÕArcy and P. A. Burnett in southwestern Idaho. Ann. Appl. Biol. 121: 105Ð121. [eds.], Barley yellow dwarf: 40 years of progress. APS, St. Henry, M., and C. A. Dedryver. 1991. Occurrence of barley Paul, MN. yellow dwarf virus in pastures of western France. Plant Rice, W. R. 1989. Analyzing tables of statistical tests. Evo- Pathol. 40: 93Ð99. lution 43: 223Ð225. Irwin, M. E., and J. M. Thresh. 1990. Epidemiology of barley Ruppel, R. F. 1983. Cumulative insect-days as an index of yellow dwarf: a study in ecological complexity. Annu. crop protection. J. Econ. Entomol. 76: 375Ð377. Rev. Phytopathol. 28: 393Ð424. SAS Institute. 1985. SAS/STAT guide for personal comput- Jedlinski, H. 1981. Rice root aphid, Rhopalosiphum ruﬁab- ers, version 6 edition. SAS Institute, Cary, NC. dominalis, a vector of barley yellow dwarf virus in Illinois, Shaoyou, L., R. L. Stoltz, and N. Xinzhi. 1986. Damage to and the disease complex. Plant Dis. 65: 975Ð978. wheat by Macrosiphum avenae (F.) (Homoptera: Aphi- April 2001 CHAPIN ET AL.: APHID ABUNDANCE AND VECTOR ROLE 421

didae) in Northwest China. J. Econ. Entomol. 79: 1688Ð Zadoks, J. C., T. T. Chang, and C. F. Konzak. 1974. A decimal 1691. code for the growth stage of cereals. Weed Res. 14: 415Ð Voss, T. S., R. W. Kieckhefer, B. W. Fuller, M. J. Mcleod, and 421. D. A. Beck. 1997. Yield losses in maturing spring wheat caused by cereal aphids (Homoptera: Aphididae) under Received for publication 11 September 2000; accepted 17 laboratory conditions. J. Econ. Entomol. 90: 1346Ð1350. December 2000.