SUMMARY

Barley yellow dwarf virus (BYDV) is a phloem-limited luteovirus transmitted by several species of grass-feeding aphids to cultivated and wild grasses, and barley yellow dwarf is one of the most economically important diseases of grasses worldwide. Recently, maize, barley, and oats have been transformed with BYDV coat protein genes, and some of these transgenic cereals are at the point of being field tested. Once stable, effective virus resistance has been achieved, cultivars with these resistance genes are certain to be adopted rapidly by farmers. Yet our ability to predict the ecological consequences of the release of these transgenic crops is severely limited, because we know little about the spread and impact of plant viruses in natural plant populations.

A survey of BYDV infection in 16 common species of annual and perennial grasses at 10 sites in Tompkins County, New York, showed that BYDV infection was common at all sites sampled. At each site, 30-77% of randomly sampled plants were infected with BYDV. Overall infection levels varied among host species, and the proportion of plants that were infected by different virus strains (PAV, RPV, RMV) differed significantly among host species. The grass subfamilies Pooideae and Panicoideae also differed significantly in the amount of infection by different virus strains, with RPV occurring more often in pooid grasses and RMV occurring more often in panicoid grasses. This apparent host specificity of virus strains could be a result of either differential aphid preference for host species, differential probabilities of transmission, or differential host susceptibility. These possible explanations are being investigated in ongoing research. Sites also differed in the amount and type of BYDV infection. PAV was found more often at agricultural sites than nonagricultural sites, whereas RPV was most common at nonagricultural sites. Future work will address the impact of infection on these wild grass hosts.

Key Words: Barley yellow dwarf virus, small grains, Poaceae, grasses

INTRODUCTION

Barley yellow dwarf is one of the most economically important diseases of grasses worldwide (Irwin and Thresh, 1990; D'Arcy and Burnett, 1995) and is among the most prevalent of all viral diseases (Thresh, 1980). The barley yellow dwarf virus (BYDV) infects a wide range of cultivated grasses, such as wheat, oats, barley and rice, as well as many wild grasses (D'Arcy, 1995). Several strains of BYDV have been characterized according to their transmission by different aphid vectors (Rochow, 1969), their cytopathology in host plants (Gill and Chong, 1979), and their molecular genetics (Martin and D'Arcy, 1995). In spite of the economic importance of barley yellow dwarf, few natural resistance genes have been identified in any of the major crops. A single well characterized resistance gene, Yd₂, exists in barley (Schaller et al., 1964), but it does not protect against all strains of BYDV (Skaria et al., 1985). No natural resistance genes have been identified in oat or wheat. Some lines show tolerance to infection with BYDV (Burnett et al., 1995), but tolerance is multigenic and difficult to manipulate using traditional breeding techniques. For some time, there has been interest in transferring resistance genes from other grass hosts, through traditional plant breeding and genetic engineering (Burnett et al., 1995). In addition, researchers have begun using biotechnology to insert viral RNA into the genome of crops plants in order to induce resistance against BYDV (Miller and Young, 1995).

These artificial resistance genes have been successful in disrupting replication of a range of viruses in a number of crops (Gadani et al., 1990). The most common strategy is to introduce viral coat protein into the plant genome (Beachy et al, 1990). Although the mechanisms that result in protection are incompletely understood, plants expressing coat protein genes require greater inoculum for infection, show infection delays, and exhibit milder disease symptoms if inoculation is successful (Miller and Young, 1995). This strategy has been successfully used for many viruses, including potato leaf roll virus (PLRV), a luteovirus closely related to some strains of BYDV (e.g., Kawshuk et al., 1990). Recently, maize (Wyatt et al., 1993), barley (Wyatt et al., 1993; Lister et al., 1994b) and oats (Lister et al., 1994a) have been transformed with BYDV coat protein genes. Lab tests suggest that BYDV titer is suppressed in at least some of the transgenic lines, and workers are in the process of field testing promising lines.

Once stable, effective virus resistance has been achieved, crop cultivars with these resistance genes are certain to be adopted rapidly by farmers. These cultivars would be expected to spread rapidly throughout the cereal growing regions of the U.S. Yet our ability to predict the ecological consequences of the release of these transgenic crops is severely limited, because we know little about the spread and impact of any plant viruses in natural plant populations (Power, 1992).

Potential risks associated with the widespread adoption of crops with BYDV coat protein genes fall into three major categories: the movement of resistance genes from transgenic crops to wild relatives, such that wild relatives become resistant to BYDV (Ellstrand and Hoffman, 1990; Kareiva et al., 1994); RNA recombination between wildtype virus and viral RNA in transgenic plants (de Zoeten, 1991; Mayo and Jolly, 1991; Greene and Allison, 1994; Miller and Young, 1995); and transcapsidation or heterologous encapsidation of wildtype virus RNA by the coat protein expressed by transgenic plants (Rochow, 1977; de Zoeten, 1991; Miller and Young, 1995). Although all of these risks are quite probable, the rates at which they will occur are unknown. Even at quite low rates of occurrence, movement of resistance genes to wild host plants and RNA recombination may have significant evolutionary consequences if these changes provide a selective advantage to either plant hosts or to viral recombinants. In addition, our ability to predict the ecological and evolutionary consequences of these events is severely limited, because we know little about the spread and impact of any plant viruses in natural plant populations.

Although there is increasing evidence that pathogens can have dramatic impacts on host populations (Anderson and May, 1982; May and Anderson, 1983) and community structure and dynamics (Harper, 1977; 1990; Silander, 1985), very little is known about the ecology and evolution of plant viruses in wild plant populations (Harper, 1990). Most information on virus-plant interactions in natural systems derives from studies of wild plants that serve as reservoirs of viruses that induce crop diseases (Bos, 1981; Thresh, 1981). These studies typically focus on the effects of wild hosts on virus epidemiology in crops, rather than the effects of the virus on the wild hosts themselves.

Understanding the impact of viruses on wild host fitness and on plant community structure is essential if we are to make informed assessments of the potential ecological consequences of genetically engineered virus resistance in crops. The first step is to understand the extent of infection of wild host plants by agriculturally important viruses. This study was designed to assess the distribution of BYDV in a range of annual and perennial grasses in two subfamilies of the Poaceae in upstate New York. In addition, we investigated whether the proximity of cultivated hosts of BYDV, such as oats, wheat, barley and corn, influenced virus infection levels in these wild grasses.

MATERIALS AND METHODS

Study system. BYDV is a phloem-limited luteovirus obligately transmitted in a persistent manner by several species of grass-feeding aphids to cultivated and wild grasses (D'Arcy and Burnett, 1995). In the vicinity of Ithaca, New York, where this study was carried out, prevalent strains of the virus include PAV, RPV, and RMV, and the most important vectors are the bird cherry-oat aphid, Rhopalosiphum padi, the corn leaf aphid, R. maidis, and the English grain aphid, Sitobion avenae. BYDV has been documented to infect more than 150 species of grasses (D'Arcy, 1995), but studies on virus host range can be difficult to interpret, because host range is typically determined by experimental inoculations of potential hosts (e.g., Oswald and Houston, 1953; El Yamani and Hill, 1990). This procedure can identify species capable of becoming infected, but it does not establish that these species are hosts of the virus in nature (D'Arcy, 1995). In contrast, field collections can be used to determine virus incidence in existing plant communities. Field surveys of the incidence of BYDV in pastures grasses and wild grasses in agricultural fields have shown considerable variance in the incidence of BYDV in the species tested (Fargette et al., 1982; Grafton et al., 1982; Guy et al., 1986; Griesbach et al., 1990). However, no previous studies have examined BYDV in wild grasses in natural, nonagricultural habitats.

Methods. During the summer of 1994, we conducted a survey of infection by BYDV in 16 of the most common species of wild grasses in Tompkins County, New York. Sampled species included both annual and perennial grasses in two subfamilies of the Poaceae, Pooideae and Panicoideae. Grasses were sampled at 10 sites that varied in their proximity to fields sown with wheat, oats, barley, and corn. Sites that included grain fields were considered to be agricultural sites. Sites that were located more than two kilometers from grain fields were considered to be nonagricultural sites. Where possible, twenty or more individuals of each species was collected at each site, though this number differed substantially among sites and no species was found at every site (Table 1). Within each population sampled, individual plants were collected at intervals of 1 m or more, to maximize the probability of sampling genetically distinct individuals. Collected plants were identified and tested for infection with four strains of BYDV (PAV, MAV, RPV, and RMV) using ELISA (enzyme-linked immunosorbent assay; Rochow, 1986).

Statistical analyses. We first asked about the infection status of individual plants. Each plant was classified as uninfected or infected based on the presence of any strain of BYDV. For infected plants, the infection was characterized as PAV only, RPV only, RMV only, or mixed infection (including any combination of two or more strains present in the plant). Because no definite infections with MAV were identified, this strain was omitted from the classification scheme. The log likelihood ratio, or G-test, was used to compare the overall BYDV infection rate (infected versus uninfected) and infection by different BYDV strains (including the class of mixed infections) in different host plant species and among subfamilies of grasses. In addition, BYDV infection at different sites and between agricultural and nonagricultural sites were compared. Because not every grass species was sampled at every site, we were unable to examine multiple factors simultaneously.

We next asked where particular virus strains were most likely to occur. In this case, we used the data on each individual plant where a virus strain was detected, regardless of whether it was found in a single or multiple infection. Again, G-tests were used to compare the occurrence of an individual BYDV strain among different host plants, among subfamilies of grasses, among sites and among habitats (agricultural versus nonagricultural sites).

RESULTS

We found an overall incidence of BYDV of 41%, ranging from 11% to 100% infection depending on the species (Figure 1). Overall infection levels varied significantly among species (G = 125.23, P < 0.0001), and the proportion of plants that were infected by different strains of BYDV differed significantly among species (G = 324.60, P < 0.0001). For example, Agrostis gigantea was primarily infected with RPV, while only RMV infections were found in Andropogon scoparius.

The subfamilies Pooideae and Panicoideae differed significantly in the amount of overall infection by BYDV (Figure 2a; G = 6.16, P = 0.0131) and in the proportion of plants infected with different BYDV strains (Figure 2b; G = 42.82, P < 0.0001), with the Pooideae showing greater infection with RPV and the Panicoideae showing greater infection with RMV. Analysis of the occurrence of individual strains indicated RPV (G = 28.64, P < 0.0001) was more likely to be found in pooids, whereas RMV was more likely to be found in panicoids (G = 17.49, P < 0.0001). PAV was found in both subfamilies at low rates (G = 2.171, P > 0.10).

Infection with BYDV was common at all the sites sampled; 30-77% of plants sampled at each site were infected with the virus (Figure 3). The overall proportion of infected plants (Figure 4; G = 64.49, P < 0.0001), and the proportion of plants infected with different BYDV strains (G = 202.83, P < 0.0001), differed significantly among the sites. Interestingly, a greater proportion of plants were infected at nonagricultural sites than agricultural sites (Figure 5a; G = 13.98, P = 0.0002). In addition, the distribution of BYDV strains differed among agricultural and nonagricultural sites (Figure 5b; G = 38.63, P < 0.0001). In terms of the occurrence of individual virus strains, PAV was significantly more common at agricultural sites (G = 15.83, P < 0.0001), whereas RPV was significantly more common at nonagricultural sites (G = 11.85, P = 0.0006). RMV was equally common at both types of sites (G = 0.13, P > 0.10).

DISCUSSION

Infection with the barley yellow dwarf virus was extremely common in all grass species tested (Figure 1) and at all sites (Figure 4); on average 41% of all 706 individuals sampled were infected with one or more strains of BYDV. Despite this overall prevalence of BYDV, there were significant differences between host species and between sites in the proportion of plants infected and in the strains of BYDV that they carried.

We found higher overall infection rates, as well as higher occurrence of RPV, in Pooid grasses compared to Panicoids. RMV was more likely to be found in Panicoid grasses than in Pooid grasses. In contrast, Guy et al. (1987) reported greater incidence of PAV in the Pooideae and greater incidence of RPV in the Panicoideae. Griesbach et al. (1990) found no trend in the specificity of particular BYDV strains for grass subfamilies. Overall, the evidence is mixed for host specificity at the subfamily level among the BYDV strains. Furthermore, any apparent specificity could be a result of either aphid preference or host susceptibility.

In this study, PAV was more common in wild grasses growing in agricultural sites than in nonagricultural sites (Figure 5b), suggesting that the epidemiology of this strain in wild grasses may be linked to their buildup in crops. It is interesting that infection by PAV was relatively uncommon in all grasses (Figure 3, 3.8% of infected individuals), despite its prevalence in cultivated grains in this area (Miller et al., 1991; Power et al., 1991). On the other hand, RPV was significantly more common in grasses growing in nonagricultural sites, where it was the most prevalent strain of BYDV. In the case of RPV, this result may be influenced by the fact that most of the grasses sampled at nonagricultural sites were pooids, and RPV was more likely to be found in pooid grasses. In contrast, PAV was more prevalent in agricultural sites despite a tendency to occur more often in pooid grasses. Previous studies have shown that the prevalent viruses in wild grasses may differ from those in nearby crops (Rochow and Muller, 1976; Fargette et al., 1982), but no previous study has compared infection rates in agricultural and nonagricultural habitats.

Surveys of natural infection rates such as this one still do not provide information about impacts on fitness of wild hosts. However, BYDV can be quite virulent in cultivated grasses like oats and barley and often induces disease symptoms (yellowing, reddening, and stunting) in the wild grasses that they are known to infect. In a sample of 36 annual and perennial grasses that were hosts to the viruses, 55% of the species showed distinct yellow-dwarf symptoms (Oswald and Houston, 1953). A separate study of 37 species of grass hosts found that 38% had symptomatic infections (Griesbach et al., 1990). Some species thought to be asymptomatic based on greenhouse studies have been found to exhibit symptoms in the field (Catherall, 1966). Although other field surveys have found symptoms of barley yellow dwarf to be rare in wild grasses (Fargette et al., 1982; Grafton, 1982), these samples have consisted largely of a few species known to be asymptomatic. There is some evidence that the probability of showing symptoms of virus infection is correlated with plant family. Pooid (= festucoid) grass hosts appear to be more likely to show symptoms than hosts from other grass subfamilies such as the panicoids (Watson and Gibbs, 1974; Guy et al., 1987). Since symptoms such as stunting and yellowing indicate physiological damage to plant function, this may suggest that infection with BYDV would be expected to cause more damage to pooid hosts than panicoid hosts. However, the lack of severe symptoms is not evidence that plant fitness is unaffected.

Although plant pathologists have suggested that most wild plants should have evolved resistance to virus infection (van der Plank, 1975,; Smith, 1977; Matthews, 1981; Cooper and MacCallum, 1984), studies of host symptomology suggest that this is not necessarily true for grass hosts of BYDV (e.g., Beuve and Lapierre, 1992; Makkouk et al., 1993). While disease symptoms imply that hosts are physiologically altered by virus infections, and BYDV is known to affect nitrogen relations in oats (Markkula and Laurema, 1964; Ajayi, 1986), there is little or no information on the fitness effects of virus infection in wild grasses. From data on cultivated hosts, we would expect that seed output, and thus reproduction, would be reduced as a result of any virulent infection (Plumb, 1983). Vegetative growth is also likely to be affected. For example, Kolb et al. (1991) reported significant reductions in oat root growth as a result of infection with the PAV-IL strain of BYDV. Catherall (1966) showed that BYDV-infected Lolium perenne produced more, but shorter, tillers than healthy plants. Additional work on BYDV infection of L. perenne indicated that infection may either reduce or increase above-ground biomass, depending on the particular genotype (Catherall and Parry, 1987). However, consistent with the data on oats, all L. perenne exhibited reduced root growth when infected.

The results of these studies suggest that we are likely to find significant effects of BYDV on at least some of the wild grasses that it infects. Where such effects do exist, the potential impacts of transgenic plants containing viral RNA may be significant. Where movement of the resistance gene into wild grasses results in increased fitness, these species may be released from ecological constraints normally imposed by infection with BYDV (Schmitt and Linder, 1994). In addition, transgenic plants expressing the coat protein of one strain of BYDV could allow the transmission of a second BYDV strain by an aphid that would not normally be a vector. Since each species of aphid vector has a unique plant host range, this increase in vector capability could allow movement of the virus into a plant that would not normally be exposed. The probability of this occurring would depend on the level of coat protein expression by transgenic plants and on the particular BYDV coat protein, since transcapsidation does not occur among all BYDVs (Creamer and Falk, 1990; Wen and Lister, 1991). The recombination of viral RNA with transgene RNA would have similar risks. Increased host range, modifications in virulence, and any changes in aphid transmission could provide a selective advantage that would allow the recombinant to spread. Studies of recombination between DNA viruses and transgenic hosts have demonstrated expansions of viral host range and changes in symptom severity, an indicator of viral virulence, as a result of the recombination (Schoelz and Wintermantel, 1993). Only by understanding the distribution and impact of viruses in wild plant populations can we begin to estimate the risks of releasing transgenic plants with viral RNA.

ACKNOWLEDGMENTS

We are grateful to S. Tom for valuable technical assistance throughout this project. S. Gray (USDA-ARS, Cornell University) kindly provided advice and antibodies for the various BYDV strains.

REFERENCES

Ajayi O. 1986. The effect of barley yellow dwarf virus on the amino acid composition of spring wheat. Ann. Appl. Biol. 108:145-149.

Anderson RM, May RM. 1982. Coevolution of hosts and parasites. Parasitology 85:411-426.

Beachy RN, Loesch-Fries S, Tumer NE. 1990. Coat protein-mediated resistance against virus infection. Annu. Rev. Phytopathol. 28:451-474.

Beuve M, Lapierre H. 1992. Resistance to RPV barley yellow dwarf virus in the genus Bromus. Canadian J. Botany 70:32-37.

Bos L. 1981. Wild plants in the ecology of virus diseases. In: Maramorosch K, Harris KF (eds) Plant diseases and vectors: ecology and epidemiology. Academic Press, New York, pp. 1-33

Burnett PA, Comeau A, Qualset CO. 1995. Host plant tolerance or resistance for control of barley yellow dwarf. In: D'Arcy CJ and Burnett PA (eds) Barley yellow dwarf: 40 years of progress. Am. Phytopathol. Soc. Press, St. Paul, MN, pp. 321-343.

Catherall PL. 1966. Effects of barley yellow dwarf virus on the growth and yield of single plants and simulated swards of perennial rye-grass. Ann. Appl. Biol. 57:155-162.

Catherall PL, Parry AL. 1987. Effects of barley yellow dwarf virus on some varieties of Italian, hybrid and perennial rye grasses and their implications for grass breeders. Plant Pathol. 36:148-153.

Cooper J I, MacCallum FO. 1984. Viruses and the environment. Chapman and Hall, London.

Creamer R, Falk BW. 1990. Direct detection of transcapsidated barley yellow dwarf luteoviruses in doubly infected plants. J. Gen. Virol. 71:211-217.

D'Arcy CJ. 1995. Symptomatology and host range of barley yellow dwarf. In: D'Arcy CJ and Burnett PA (eds) Barley yellow dwarf: 40 years of progress. Am. Phytopathol. Soc. Press, St. Paul, MN, pp. 9-28.

D'Arcy CJ, Burnett PA (eds). 1995. Barley yellow dwarf: 40 years of progress. Am. Phytopathol. Soc. Press, St. Paul, MN.

de Zoeten GA. 1991. Risk assessment: Do we let history repeat itself? Phytopathology 81:585.

Ellstrand N, Hoffman C. 1990. Hybridization as an avenue of escape for engineered genes. Bioscience 40:438-441.

El Yamani M, Hill JH. 1990. Identification and importance of barley yellow dwarf virus in Morocco. Plant Dis. 74:291-294.

Fargette D, Lister RM, Hood EL. 1982. Grasses as a reservoir of barley yellow dwarf virus in Indiana. Plant Dis. 66:1041-1045.

Gadani F, Mansky LM, Medici R, Miller WA, Hill JH. 1990. Genetic engineering of plants for virus resistance. Arch. Virol. 115:1-21.

Gill CC, Chong H. 1979. Cytopathological evidence for the division of barley yellow dwarf virus isolates into two subgroups. Virology 95:59-69.

Grafton KF, Poehlman JM, Sehgal OP, Sechler DT. 1982. Tall fescue as a natural host and aphid vectors of barley yellow dwarf virus in Missouri (Rhopalosiphum padi). Plant Disease 66:318-320.

Greene AE, Allison RF. 1994. Recombination between viral RNA and transgenic plant transcripts. Science 263:1423-1425.

Griesbach JA, Steffenson BJ, Brown MP, Falk BW, Webster RK. 1990. Infection of grasses by barley yellow dwarf viruses in California. Crop Sci. 30:1173-1177.

Guy PL, Johnstone GR, Duffus JE. 1986. Occurrence and identity of barley yellow dwarf viruses in Tasmanian pasture grasses. Aust. J. Agric. Res. 37:43-53.

Guy PL, Johnstone GR, Morris DI. 1987. Barley yellow dwarf viruses in, and aphids on, grasses (including cereals) in Tasmania. Aust. J. Agric. Res. 38:139-152.

Harper JL. 1977. Population biology of plants. Academic Press.

Harper JL. 1990. Pests, pathogens and plant communities: an introduction. In: Burdon JJ, Leather SR, (eds) Pests, pathogens and plant communities. Blackwell, London, pp. 3-14.

Irwin ME, Thresh JM. 1990. Epidemiology of barley yellow dwarf virus: a study in ecological complexity. Annu. Rev. Phytopathol. 28:393-424.

Jedlinski H, Rochow WF, Brown CM. 1977. Tolerance to barley yellow dwarf virus in oats. Phytopathology 67:1408-1411.

Kareiva P, Morris W, Jacobi CM. 1994. Studying and managing the risk of cross-fertilization between transgenic crops and wild relatives. Molec. Ecology 3:15-21.

Kawshuk LM, Martin RR, McPherson M. 1990. Resistance in transgenic potato expressing the potato leafroll virus coat protein gene. Mol. Plant-Microbe Interact. 3:301-307.

Kolb FL, Cooper NK, Hewings AD, Bauske EM, Teyker RH. 1991. Effects of barley yellow dwarf virus on root growth in spring oat. Plant Disease 75(2):143-145.

Lister RM, Vincent JR, Lei C-H, McGrath PF. 1994a. Coat protein mediated resistance to barley yellow dwarf virus in oats. Phytopathology 84(12).

Lister RM, Vincent JR, McGrath PF. 1994b. Coat protein mediated resistance to barley yellow dwarf virus in barley. Phytopathology 84(12).

Makkouk KM, Comeau A, Ghulam W. 1994. Resistance to barley yellow dwarf luteovirus in Aegilops species. Canadian J. Plant Science 74:631-634.

Markkula M, Laurema S. 1964. Changes in the concentration of free amino acids in plants induced by virus diseases and the reproduction of aphids. Ann. Agric. Fenn. 3:265-271.

Martin RR, D'Arcy CJ. 1995. Taxonomy of barley yellow dwarf viruses. In: D'Arcy CJ and Burnett PA (eds) Barley yellow dwarf: 40 years of progress. Am. Phytopathol. Soc. Press, St. Paul, MN, pp. 203-214.

Matthews REF. 1981. Plant virology, 2nd ed. Academic Press, New York.

May RM, Anderson RM. 1983. Epidemiology and genetics in the coevolution of parasites and hosts. Proc. R. Soc. London Ser. B 219:281-313.

Mayo MA, Jolly CA. 1991. The 5'-terminal sequence of potato leafroll virus RNA: evidence of recombination between virus and host RNA. J. Gen Virology 72:2591-2595.

Miller NR, Bergstrom GC, Gray SM. 1991. Identity, prevalence, and distribution of viral diseases in winter wheat in New York in 1988 and 1989. Plant Disease 75:1105-1108.

Miller WA, Young MJ. 1995. Prospects for Genetically Engineered Resistance to Barley Yellow Dwarf Viruses. In: D'Arcy CJ and Burnett PA (eds) Barley yellow dwarf: 40 years of progress. Am. Phytopathol. Soc. Press, St. Paul, MN, pp. 345-370.

Oswald JW, Houston BR. 1953. Host range and epiphytology of the cereal yellow dwarf disease. Phytopathology 43:309-313.

Plumb RT. 1983. Barley yellow dwarf virus--a global problem. In: Plumb RT, Thresh JM (eds) Plant virus epidemiology. Blackwell Scientific, Oxford, pp. 186-198.

Power AG. 1992. Patterns of virulence and benevolence in insect-borne pathogens of plants. Critical Reviews in Plant Sciences 11:351-372.

Power AG, Seaman AJ, Gray SM. 1991. Aphid transmission of barley yellow dwarf virus: inoculation access periods and epidemiological implications. Phytopathology 81:545-548.

Rochow WF. 1969. Biological properties of four isolates of barley yellow dwarf virus. Phytopathology 59:1580-1589.

Rochow WF. 1977. Dependent virus transmission from mixed infections. In: Harris F, Maramorosch K (eds) Aphids as virus vectors. Academic Press, New York, pp. 253-277

Rochow WF. 1986. Barley yellow dwarf virus. Methods for Enzyme Analysis 11:420-430.

Rochow WF, Muller I. 1976. Gradual change in predominating isolates of barley yellow dwarf virus in New York. Plant Dis. Rep. 60:387-390.

Schaller CW, Qualset CO, Rutger JN. 1964. Inheritance and linkage of the yd₂ gene conditioning resistance to the barley yellow dwarf virus disease in barley. Crop. Sci. 4:544-548.

Schmitt J, Linder CR. 1994. Will escaped transgenes lead to ecological release? Molec. Ecology 3:71-74.

Schoelz JE, Wintermantel WM. 1993. Expansion of viral host range through complementation and recombination in transgenic plants. The Plant Cell 5:1669-1679.

Silander JA. 1985. Microevolution in clonal plants. In: Jackson JBC, Buss LW, Cook RE (eds) Population biology and evolution of clonal organisms. Yale University Press, New Haven, CT, pp. 107-152.

Skaria M, Lister RM, Foster JE, Shaner G. 1985. Virus content as an index of symptomatic resistance to barley yellow dwarf virus in cereals. Phytopathology 75:212-216.

Smith KM. 1977. Plant viruses. Chapman and Hall, London.

Thresh JM. 1980. The origin and epidemiology of some important plant virus diseases. Applied Biology 5:1-65.

Thresh JM. 1981. The role of weeds and wild plants in the epidemiology of plant virus diseases. In: Thresh JM (ed) Pests, pathogens and vegetation. Pitman, Boston, pp. 53-63.

van der Plank JE. 1975. Principles of plant infection. Academic Press, New York.

Watson L, Gibbs AJ. 1974. Taxonomic patterns in the host ranges of viruses among grasses, and suggestions on generic sampling for host-range studies. Ann. Appl. Biol. 77:23-32.

Wen F, Lister RM. 1991. Heterologous encapsidation in mixed infections among four isolates of barley yellow dwarf virus. J. Gen. Virology 72:2217-2223.

Wyatt SD, Berger PH, Lemaux PG. 1993. Coat protein gene-mediated resistance to barley yellow dwarf virus in maize and barley. Phytopathology 83(12):1374.

Table 1. Number of individuals of each grass species collected at ten sites in Tompkins County, New York.

Species

Agricultural Game Poultry SPCA Tailby Turkey Warren Farm Farm Farm Farm

Non-Agricultural College Six-Mile Snyder Whipple town Creek Hill Field

Total plants

Pooideae
Agrostis gigantea	-	-	-	-	32	-	-	-	-	40	72
Anthoxanthum odoratum	-	-	-	-	20	-	-	-	-	33	53
Agropyron repens	-	-	-	6	20	23	-	-	-	-	49
Bromus sp.	-	-	-	9	12	-	-	40	-	-	61
Dactylis glomerata	-	-	-	5	10	-	20	-	-	-	35
Festuca arundinaceae	-	-	-	23	-	10	-	-	-	10	43
Glyceria sp.	-	-	-	25	-	-	-	-	-	-	25
Phalaris arundinaceae	-	-	-	-	5	-	-	-	7	-	12
Phelum pratense	-	24	-	19	26	11	-	-	2	40	122
Poa pratensis	-	-	-	25	18	-	-	-	-	-	43
Panicoideae
Andropogon scoparius	-	-	-	-	-	-	-	-	22	-	22
Digitaria sp.	8	-	-	8	-	-	-	-	-	-	16
Echinochloa crus-galli	5	-	-	5	-	-	-	-	-	-	10
Panicum capillare	8	-	-	8	11	-	-	-	-	-	27
Setaria lutescens	18	-	60	12	10	-	-	-	-	-	100
Setaria viridis	8	-	-	8	-	-	-	-	-	-	16
Total	47	24	60	153	164	44	20	40	31	123	706

Figure 1. The frequency of BYDV infection in 16 species of wild grasses in Tompkins County, NY in 1994.

Figure 2. The frequency of BYDV in two subfamilies of grasses, the Panicoideae and the Pooideae. (A) Overall rates of infection by BYDV and (B) the frequency of occurrence of different strains of BYDV.

Figure 3. The frequency of infection by BYDV strains of 16 species of wild grasses in Tompkins County, NY in 1994.

Figure 4. The frequency of BYDV infection in wild grasses in ten sites in Tompkins County, NY in 1994.

Figure 5. The frequency of BYDV in two habitats, agricultural and non-agricultural. (A) Overall rates of infection by BYDV and (B) the frequency of occurrence of different strains of BYDV.