Population Dynamics in Grapevine Genotypes Differing in Susceptibility to Pierce's Disease

Felix B. Fritschi,l Hong Lin,l* and M. Andrew Walker2

Abstract: The xylem-limited bacterium Xylellajastidiosa is the causal agent ofPierce's disease (PD) in grapevines, for which breeding resistant cultivars will be a long-term management strategy that involves the identification and characterization of resistant germplasm. A genetically diverse group of species and selections was mechani­ cally inoculated with X. jastidiosa, grown in a greenhouse for 113 days after inoculation, and evaluated for the levels ofbacterial concentrations in stem and leaf tissues by quantitative enzyme-linked immunosorbent assay (ELISA). Concentrations of X. jastidiosa were affected by genotype, tissue, position on the plant relative to the point of inoculation, and interactions among these factors. Based on estimated concentrations of X. jastidiosa in stem samples at 113 d postinoculation, 9621-67, Muscadinia rotundifolia, arizonica/candicans, V. arizonica/girdiana, V. candicans, V. girdiana, V. nesbittiana, and V. shuttleworthii were resistant to PD. In contrast, V. vinifera, V. aestivalis, 9621-94, and V. champinii had very high X. jastidiosa concentrations in stem tissues. Sequential sampling of leaf blades at 34, 77, and 113 d postinoculation revealed different temporal patterns in X. jastidiosa concentrations among the grape genotypes. Estimates ofX. jastidiosa concentrations decreased after the first sampling in M. rotundifolia, 9621-67, V. girdiana, and V. arizonica/candicans but increased in all other genotypes. The characterization of X. jastidiosa concentrations in a broad range of grape genotypes allows for the selection of promising genetic back­ grounds capable of greatly limiting the population size and development of X. jastidiosa in stems, a critical trait in the breeding of PD-resistant grapevines. Key words: disease resistance, Pierce's disease, Xylellajastidiosa

Pierce's disease (PD) of grapevines is caused by the http://www.cdfa.ca.gov/gwssl). the disease and its vectors xylem-inhabiting bactl'

croscopy, enzyme-linked immunosorbent assay (ELISA), peat (1:1:1). After four weeks, plants in these small pots and combinations of these techniques. Most of these were transplanted into 1-L plastic pots. To ensure uniform studies examined X. fastidiosa population dynamics and shoot growth and age, plants were cut back to two buds plant responses in Vitis vinifera and/or Muscadinia rotun­ and regrown for about 6 weeks before inoculation. Within difolia genotypes (Hill and Purcell 1995, Fry and Mil­ one week of inoculation, plants were transferred from holland 1990, Raju and Goheen 1981, Hopkins et al. 1974, greenhouses at UC Davis to a greenhouse at the USDA­ Hopkins and Thompson 1981, Milholland et al. 1981). ARS in Parlier, CA. Greenhouse temperatures were main­ Recent studies expanded on earlier studies of X. tained between 20 and 32°C at both locations. An average fastidiosa accumulation in grapes by including two sus­ day/night cycle of 18/6 hr was obtained by a combination ceptible and six field-resistant Vitis selections from di­ of ambient and supplemental lighting. Plants were irrigated verse genetic backgrounds (Krivanek et al. 2005b, Kri­ twice per day using 1.9 L/h emitters for 2 min with occa­ vanek and Walker 2005). The levels of X. fastidiosa in the sional adjustments in irrigation duration to the increasing susceptible genotypes did not differ among nodes, inter­ requirements of the developing plants. The plants were nodes, petioles, and leaf blades. On the contrary, in the fertilized every 3 weeks with 50 mL of diluted Miracle-Gro resistant genotypes, X. fastidiosa populations were great­ (N:P:K = 15:30:15; Scotts, Marysville, OH). In order to in­ est in leaf blades, followed by petioles, and lowest in crease air circulation and light penetration, lateral and api­ stem nodes and internodes. The differences between X. cal shoot tips were removed every 4 weeks once the main fastidiosa populations in the resistant genotypes versus shoot reached 0.9 m. susceptible genotypes were greatest when the stem inter­ Bacterial culture and plant inoculation. A strain of X. node tissues were compared and were not significant when fastidiosa, originally isolated from the Stag's Leap area of leaf blade tissues were compared. The inheritance of resis­ Napa Valley, CA, was used for this study. To ensure viru­ tance to X. fastidiosa in a V. rupestris x V. arizonica popu­ lence, the strain was maintained in greenhouse-grown lation was examined by measuring disease development Chardonnay plants and reisolated from these plants for using ELISA and symptomology (Krivanek et al. 2006), in­ each round of inoculations. Bacteria for inoculations were cluding leaf scorch and a cane maturation index (Krivanek grown on solid periwinkle medium at 29°C according to et al. 2005b). One gene with a single dominant allele ap­ Davis et al. (1980). Actively growing bacteria (-lO-d old) peared to control PD resistance in that population and X. were washed from culture plates with ddHzO, and bacterial 8 fastidiosa levels gave the highest broad-sense heritability suspensions were adjusted to 6 x 10 cfu/mL (A6oo nm = estimate for resistance (Krivanek et al. 2006). In addition, X. fastidiosa levels in stem tissue were highly correlated Table 1 Grape genotypes examined in this study with their with field resistance to PD (Krivanek and Walker 2005). genetic background and place of origin. (Plant material Thus, quantification of X. fastidiosa populations in grape from the vineyards of the Department of Viticulture genotypes appears to be a useful tool to differentiate sus­ and Enology, UC Davis.) ceptibi1itylresistance of grape genotypes to PD. Investiga­ Genotype Species/parentage Origin tions into PD resistance of M. rotundifolia, V. labrusca, V. californica, V. girdiana, V. arizonica, and V. vinifera used Cowart Muscadinia rotundifolia Cultivar from Georgia ELISA quantification of X. fastidiosa and disease sympto­ DVIT 1394" Vitis aestivalis Southcentral U.S. mology (Ruel and Walker 2006). Results revealed a corre­ b43-17 V. arizonica/candicans Monterrey, Mexico lation between PD resistance and relationship to geo­ b42-26 V. arizonica/girdiana Loreto, Mexico graphic areas of known high PD pressure among and T56 V. candicans Bell County, within species, suggesting that resistance to PD evolved DVIT 1588" V. champinii CH3-48 Texas in response to disease pressure. 'DVIT 1379" V. girdiana Santa Ana Canyon, Calif. V. monticola Central Texas The objectives of the study presented here were to DVIT 1371" compare a genetically diverse range of grape genotypes in DVIT 2235.4" V. nesbitt/ana Veracruz, Mexico DVIT 1416" V. rufotomentosa Florida respect to X. fastidiosa levels in leaves over time and to X. 08909-15" V. rupestris A. de Serres x Walker breeding program, fastidiosa levels in stems and leaves. V. arizonica/girdiana UC Davis b42-26 9621-67 and 08909-15 x F8909-17 (V. Walker breeding program, Materials and Methods 9621-94 rupestris A. de Serres x UC Davis V. arizonica/candicans Plant material. Eighteen grape genotypes (17 within b43-17) Vi tis and one M. rotundifolia cultivar) were evaluated in Haines City V. shuttleworthii Florida this study (Table 1). Herbaceous cuttings, obtained from C52-94 V. simpsonii Lady Lake, Florida field-grown grapevines from the vineyards of the Univer­ B02SG V. smalliana Geneva, Florida sity of California (UC) Davis, were rooted in cellulose L28:8 V. tiliifolia Southern Florida sponges on intermittent mist beds with 27°C bottom heat. Chardonnay V. vinifera France 3 Once rooted, plants were transplanted into small 600 cm "DVIT: selections from the USDA National Clonal Germplasm Re­ plastic pots with a mix of Yolo sandy loam, perlite, and pository, Davis,CA.

Am. J. Enol. Vitic. 58:3 (2007) 328 - Fritschi et al.

0.25). Two-month-old plants grown from cuttings as de­ water-inoculated control plants were used to set tissue scribed above were inoculated. Mechanical inoculation as specific positive thresholds for each genotype by adding described elsewhere (Krivanek et al. 2005b) was per­ three standard deviations to the mean. formed by placing a lO-jlL droplet of bacterial suspension Statistical analyses. All bacterial concentration data on the stem 10 cm above the base, which was pierced by were natural log-transformed for statistical analyses. a needle, allowing the droplet to be taken into the stem Analysis of variance was conducted using the SAS soft­ vascular system by transpirational flow. Each plant was ware package (version 9.1; SAS Institute, Cary, NC) with inoculated twice to ensure successful inoculation. Control the MIXED procedure in a repeated measures analysis. All plants were treated in the same manner, but the bacterial pairwise comparisons were conducted by Tukey-Kramer's suspension was substituted with water. HSD. Experimental design. Plants were arranged in a com­ pletely randomized design on greenhouse benches, each Results individual plant representing an experimental unit. For each genotype, three to four X. jastidiosa-inoculated Water-inoculated control plants and determination of plants and up to three water-inoculated plants were used. positive threshold. Control plants were inoculated with The statistical design was treated as unbalanced, since water to determine representative background levels for the limited availability of certain genotypes resulted in an the X. jastidiosa-inoculated plants. Initial statistical analy­ unequal number of replications among genotypes. sis across both X. jastidiosa-inoculated and water-inocu­ Collection and grinding of tissue samples. For sam­ lated plants revealed significant differences for genotype, pling purposes grapevine stems were divided into bottom, infection, tissue, genotype x tissue (all p < 0.0001), and middle, and top portion. The bottom portion included the genotype x infection interaction (p = 0.03). There were sig­ 25-cm segment of the plant immediately above the point of nificant differences among ELISA absorbance values for inoculation, the middle portion consisted of the next 25 the water-inoculated control plants, and they were signifi­ cm, and the top section consisted of the remaining por­ cantly affected by genotype, tissue, time, genotype x tis­ tion of the plant. Leaf blade samples were collected three sue, genotype x time (all p < 0.000l), and the three-way times: 34 d, 77 d, and 113 d postinoculation from the three interactions of genotype x tissue x position and genotype stem portions. At each sampling time, one leaf was x position x time (p < 0.05). The significant main and inter­ sampled from every plant. Stem samples were only col­ action effects of time reflect the dynamics in the leaf tis­ lected at 113 d postinoculation. A minimum of two inter­ sue only, since stem tissue was only sampled at 113 d nodes and nodes were sampled from each of the three postinoculation. Separate analyses of background levels portions. Plant samples were placed into air-tight plastic in leaf tissues for each sampling revealed that while the ef­ bags and stored at -20°C. Leaf samples were finely ground fect of position was not significant at the 34-d sampling, in liquid N using a mortar and pestle. Stem samples were it was significant for leaves sampled 77 and 113 d postin­ pulverized using a KLECO 4200 ball mill (Kinetic Labora­ oculation (p < 0.0l). As revealed by a significant geno­ tory Equipment, Visalia, CA) by cooling metal canisters type x position interaction (p < 0.001), background levels and tissue in liquid N before processing for 25 to 30 sec. at 77 d did not follow the same trends in all genotypes. Ground leaf and stem samples were weighed (0.1 g) and However, at 113 d, background levels were significantly extracted in l-mL grape extraction buffer (0.5 M tris­ lower for the bottom than the top positions (p = 0.01), hydroxymethyl-aminomethane, 137 mM NaCl, 2% polyvi­ and intermediate for the middle positions. Leaf back­ nylpyrrolidone (PVP-40), 1% polyethylene glycol, 3 mM ground levels averaged across all sections and samplings 5 5 , NaN3 0.05% Tween-20, pH 8.2; AGDIA, Elkhart, IN). were lowest for V. nesbittiana (4.52 x 10 ± 0.31 x 10 cells Enzyme-linked immunosorbent assay. Quantitative go!) and greatest for V. aestivalis (10.40 x 105 ± 0.92 x 105 ELISA using custom antibodies was conducted as de­ cells go!). All significant pairwise comparisons of leaf scribed elsewhere (Krivanek and Walker 2005). Briefly, a background levels across the three sampling times and standard curve was generated by diluting X. jastidiosa in sections revealed that all differences were related to either 6 plant tissue extracts to reach concentrations of 6.5 x 10 , V. nesbittiana or V. aestivalis. Separate statistical analyses 6 5 5 4 4 3.25 X 10 ,6.5 X 10 ,3.25 X 10 ,6.5 X 10 , and 3.25 x 10 cfu/ for stem tissues of water-inoculated plants also revealed mL. Separate standard curves were generated for leaf and significant genotype (p < O.OOQl), position (p = 0.01), and stem tissues of every genotype using healthy plant mate­ genotype x position (p = 0.0005) effects. The average of rial from water-inoculated plants. As with the samples, the calculated theoretical X. jastidiosa concentration lev­ standard curves and negative controls consisting of X. els in stems of control plants was about 1.78 x 106 ± 0.15 x jastidiosa-free tissue extracts and extraction buffer alone 106 cells g-! and was lowest for V. shuttleworthii and were run in duplicate wells. Predicted concentrations per greatest for V. champinii. milliliter of extract were converted to cells per gram of The threshold values for positive samples for each tis­ plant tissue. Absorbance readings from water-inoculated sue of each genotype were calculated separately by the control plants were used to calculate theoretical concen­ addition of three standard deviations to the mean. The trations of bacteria per gram of plant tissue. Samples from computed positive thresholds were applied individually for

Am. J. Enol. Vitic. 58:3 (2007) Grapevine Genotypes Differing in Susceptibility to Pierce's Disease - 329

each genotype and tissue. Averaged across all genotypes, highly significant genotype x tissue interaction, X. fasti­ the positive threshold was 2.76 x 106 ± 0.43 x 106 cells g-l diosa populations in stems and leaves were differentially for stem tissue and 1.56 x 106 ± 0.15 x 106 cells g-l for leaf influenced by genotype. The ratios of leaf:stem X. fastidi­ tissue. osa concentrations varied between 0.32 for M. rotundi­ Xylella fastidiosa populations in leaves and stems. folia and 8.25 for V. simpsonii. At 113 d postinoculation, Genotype, tissue, time, genotype x tissue, genotype x time three genotypes (M. rotundifolia, V. girdiana, and 9621­ (all p < 0.0001), and position (p = 0.05) significantly af­ 67) had the lowest concentrations of X. fastidiosa in both fected X. fastidiosa population size. Large differences in X. leaves and stems, while V. vinifera, V. aestivalis, and 9621­ fastidiosa populations among genotypes for both leaf and 94 had the highest concentrations in both tissue types. stem tissues were evident (Figures 1 and 2). Ten of the Development of leaf X. fastidiosa populations over investigated genotypes exhibited X. fastidiosa concentra­ time. Xylella fastidiosa levels in leaves of infected plants tions below the positive threshold value for at least one of were significantly affected by genotype, time, genotype x the sampling times and tissues (Table 2). The X. fastidiosa time (all p < 0.0001), and position (p == 0.008). Although concentrations in stems and leaves never exceeded the foliar symptoms of PD were not recorded, obvious symp­ positive threshold in M. rotundifolia. As indicated by the toms including advanced marginal leaf necrosis on lower leaves were observed on Chardonnay plants at the second

100x106 c::::J 34 d 100x1 0" (j) 1!!!1!R1177d c::::J Bottom ::J (fj _113d (j) I!!!I!RII Middle (fj ::J _Top (fj (fj '7"" OJ '7"" .!!2 10x106 OJ a; .!!2 ,£. a; 10x106 ,£. '" Bl i(fj ~ ~ >< 1x106 ~ >< ij 6 ~~~~~ 1x10 100x1 0' I~ 10ox10'

Figure 1 Concentration of Xylella fastidiosa in leaf blades of 18 grape genotypes sampled at 34 d, 77 d, and 113 d postinoculation. Samples Figure 2 Concentration of Xylella fastidiosa in stems of 18 grape were collected from bottom, middle, and top sections of the plants, genotypes sampled at 113 d postinoculation. Samples were collected analyzed separately by ELISA to determine the average bacterial from bottom, middle, and top sections of the plants, analyzed sepa­ concentration. Error bars represent standard error of the mean. rately by ELISA. Error bars represent standard error of the mean.

Table 2 Grape genotypes categorized based on X. fastidiosa concentrations in respect to the genotype and tissue specific positive threshold level. The positive threshold was established by adding three standard deviations to the mean X. fastidiosa concentrations.

Threshold Stem 113 d Leaf 34 d Leaf 77 d Leaf 113 d All samples M. rotundifolia M. rotundifolia M. rotundifolia M. rotundifolia V. rufotomentosa V. girdiana V. girdiana V. girdiana 8909-15 V. ariz/cand a V. ariz/cand V. ariz/cand 9621-94 V. candicans V. candicans V. tiliifolia V. shuttleworthii V. shuttleworthii V. monticola V. nesbittiana V. champinii V. ariz/gird b V. aestivalis V. simpsonii V. vinifera V. smalliana 9621-67 aForm of V. arizonica with introgression from V. candicans (V. mustangensis). bForm of V. arizonica with introgression from V. girdiana.

Am. J. Enol. Vitic. 58:3 (2007) 330 - Fritschi et al.

sampling (77 d postinoculation). Marginal leaf necrosis A was also observed on 9621-94 at that time. Typical PD c:::J Bottom 10x10' _ Middle symptoms such as leaf necrosis, matchsticks, and green ill _Top ::l islands were observed on all infected Chardonnay and UJ ~ 9621-94 plants at 113 d postinoculation. However, no vis­ , Q) ible symptoms were observed at the first sampling (34 d !!! OJ postinoculation) on any genotype. The genotypes can be !:3- < the second and third sampling compared to the initial sampling (M. rotundifolia, 9621-67, V. girdiana, V. ~ ~ ~~ i~'~ ;~~ ;~ arizonica/candicans) and a second group consisting of II! ! ! ! all other genotypes in which X. fastidiosa levels in­ creased after the 34-d sampling (Figure 1). Additional dis­ tinctions could be made within the second group; in par­ ticular, genotypes that exhibited a pronounced peak at 77 d postinoculation (V. shuttleworthii, V. nesbittiana, V. candicans, V. champinii) may be of interest and could be B categorized into a third group. These four genotypes all 100x1O. fell within the lower half of the second group when ar­ c:::J Bottom _ Middle ill ranged according to X. fastidiosa concentration at 113 d ::l _Top UJ UJ postinoculation. Average leaf X. fastidiosa levels at 113 d ,:;:; of the four genotypesin the first group (no increase in X. Q) !!! 1Ox1 O· fastidiosa over time) were less than 1.0 x 106 cells g-l tis­ OJ !:3- < and V. shuttleworthii (third group) were below their posi­ 100x1 03 ~~I i~ i~~ tive thresholds. In all other genotypes, X. fastidiosa lev­ ~! els in leaf tissues collected 113 d postinoculation ex­ ceeded the positive thresholds. Influence of position on X. fastidiosa populations in leaves and stems. The distribution of X. fastidiosa within a plant was investigated by determining the concentra­ tions of X. fastidiosa in stems and leaves of bottom, middle, and top portions of the plants. For the stem c 100x10· samples collected from X. fastidiosa inoculated plants, the c:::J Bottom interaction of genotype x position was significant (p = _ Middle _Top 0.0001), as was the simple effect of genotype (p < 0.0001) but not that of position (Figure 2). Pairwise comparisons between the three positions conducted for each genotype only revealed significant (p < 0.05) effects for V. can­ dicans and V. nesbittiana, both of which had greater X. fastidiosa levels in the bottom position than in the middle and top positions. Significant effects of genotype, time, genotype x time (all p < 0.0001), and position (p =0.008) ~IIII on X. fastidiosa populations in leaves were observed (Fig­ 3 100x1 0 iii. III 1161 ure 3). Although the interaction of position x time was not significant, separate analysis by sampling revealed that differences between positions were significant at 34 d, but not at 77 d and 113 d. Averaged across all genotypes, the X. fastidiosa concentration in the leaves from the bottom third of the plants was 3.29 x 106 ± 0.72 x 106 cells g-l Xylella fastidiosa 6 6 Figure 3 Concentration of in leaves from the bottom, compared with 1.01 x 10 ± 0.09 x 10 cells g-l in the middle, and top section of the plants sampled (A) 34 d, (8) 77 d, and (C) middle portion, and 0.93 x 106 ± 0.09 x 106 cells g-l in the 113 d postinoculation. Error bars represent standard error of the mean.

Am. J. Enol. Vitic. 58:3 (2007) Grapevine Genotypes Differing in Susceptibility to Pierce's Disease - 331

the top portion. Pairwise comparisons among the X. was less than 2.1 x 106 cells g-! tissue. In addition, 9621­ fastidiosa concentrations of the three positions indicated 67, a highly PD-resistant member of a genetic mapping that these differences in X. fastidiosa concentrations population (Doucleff et al. 2004, Krivanek et al. 2006), had among leaves from the bottom position and those from the the lowest average X. fastidiosa concentration among middle and top positions were highly significant (p < stem tissues. Vitis simpsonii also had very low X. fasti­ 0.0001). diosa concentrations «2.1 x 106 cells g-! tissue). Average stem X. fastidiosa concentrations in the remaining geno­ Discussion types exceeded 5.0 x 106 cells g-! tissue. Selection of appropriate reference material for back­ With the exception of V. simpsonii, genotypes with low ground controls when conducting ELISA has been empha­ average stem X. fastidiosa concentrations also had low sized in the past (Sutula et al. 1986). The significant ef­ average leaf X. fastidiosa concentrations. The leaf X. fects of genotype, tissue, sampling time, and position as fastidiosa population size in M. rotundifolia, 9621-67, V. well as various interactions for the calculated mean theo­ girdiana, V. arizonica/candicans, and V. shuttleworthii retical concentration of X. fastidiosa cells (natural log­ was <1.0 x 106 cells g-! tissue at 113 d postinoculation. transformed) of samples collected from water-inoculated We did not confirm previous reports that X. fastidiosa lev­ control plants highlight the importance of using appropri­ els in the stems of infected plants are generally lower ate reference tissue when conducting ELISA. On average, than in leaves (Krivanek and Walker 2005), perhaps be­ the positive thresholds determined from water-inoculated cause of differential extraction efficiencies, since plant tis­ control plants in this study were comparable to those re­ sues were not ground with the same methods. Patterns of ported for a number of grape genotypes (Krivanek and leaf X. fastidiosa concentrations over time suggest a cat­ Walker 2005), thresholds for stem tissues being slightly egorization of the genotypes into the following three greater and those for leaf tissues slightly lower than in groups: D (decrease), concentrations that were lower at 77 that study. d and 113 d than at 34 d postinoculation; P (peak), con­ Based on the genotype and tissue specific positive centrations that peaked at 77 d postinoculation; and I (in­ thresholds determined in this study, M. rotundifolia was crease), genotypes in which concentrations were higher at the only genotype with estimated average leaf and stem 77 d and 113 d than at 34 d but that did not peak at 77 d X. fastidiosa concentrations that never exceeded the posi­ (Figure 1). Multiple scenarios may account for the differ­ tive threshold value. That was not surprising since M. ro­ ences in the observed pattern. In group D (M. rotundi­ tundifolia is native to areas with intense PD pressure, and folia, 9621-67, V. girdiana, V. arizonica/candicans), it may resistance of M. rotundifolia to X. fastidiosa has been be that the X. fastidiosa population does not grow vigor­ shown in both field and greenhouse studies (Ruel and ously and that a dilution effect occurs as plant biomass Walker 2006, Hopkins et al. 1974, Hopkins and Thompson increases and/or that the bacteria do not multiply at all 1984). The high susceptibility of V. vinifera is well known. and eventually die. In group P (V. shuttleworthii, V. nes­ Estimated X. fastidiosa concentrations in stem and leaf tis­ bittiana, V. candicans, V. champinii), the bacteria appear sues collected from inoculated V. vinifera plants were al­ to multiply rapidly initially, but do not continue with the ways among the highest in this study. In addition to simi­ same vigor over time. It may be, for example, that these larities to previously published studies using quantitative plants respond to the increased bacterial population in ELISA to investigate PD susceptibility in grapes (Krivanek some way that limits bacterial growth. In group I (V. ari­ and Walker 2005, Ruel and Walker 2006), the results ob­ zonica/girdiana, V. smalliana, 8909-15, V. tiliifolia, V. tained here for M. rotundifolia and V. vinifera confirm the simpsonii, V. rufotomentosa, V. monticola, V. vinifera, reliability of the approach used in this study. 9621-94, V. aestivalis), it appears that vascular system Recent reports indicate that the concentration of X. conditions in these genotypes are more or less conducive fastidiosa in stem tissue is well suited as a measure of PD for extended growth of X. fastidiosa. Genotypes of group resistance and that it corresponds well to field resistance D seem to be good candidates to develop PD-resistant (Krivanek and Walker 2005, Krivanek et al. 2005a,b, Ruel grapevines via breeding. The reasons for the X. fastidiosa and Walker 2006). Thus, screening grape genotypes uuder population pattern observed for genotypes in group Pare greenhouse conditions allows for rapid and efficient evalu­ unclear. Whether the observed trend continues over time ation of numerous plants for PD resistance. However, the and eventually results in X. fastidiosa concentrations performance of selected genotypes and breeding lines un­ similar to those in resistant genotypes or whether the X. der field conditions will be the ultimate measure for PD fastidiosa population stabilizes itself at a certain level or resistance and is essential. Estimated X. fastidiosa concen­ increases again is uncertain. trations in the stems sampled 113 d postinoculation varied The significantly greater X. fastidiosa concentrations greatly among the 18 genotypes investigated (Figure 2). in the bottom portion leaves at 34 d postinoculation sug­ For M. rotundifolia, V. girdiana, V. arizonica/candicans, gests that under conditions of mechanical inoculation, a V. candicans, V. shuttleworthii, V. nesbittiana, and V. ari­ majority of the X. fastidiosa remain near the point of in­ zonica/girdiana, the estimated X. fastidiosa concentration oculation. However, it has been recently reported that X. in the stems did not exceed the positive threshold and fastidiosa can move passively through long xylem vessels

Am. J. Enol. Vitic. 58:3 (2007) 332 - Fritschi et al. in minutes (Thorne et al. 2006). Such passive but rapid Hopkins, D.L., and AH. Purcell. 2002. Xylellafastidiosa: Cause of movement in response to mechanical inoculation may have Pierce's disease of grapevine and other emergent diseases. Plant affected the dynamics of X. fastidiosa development in the Dis. 86: 1056-1066. middle and top portions of the plants. Nonetheless, X. Hopkins, D.L., and C.M. Thompson. 1981. Multiplication of viru­ fastidiosa concentrations in the middle and top portion lent and avirulent Pierce's disease bacterial isolates in grapevine leaves increased more slowly, but in many genotypes at tissue. In Proceedings of the 5th International Conference on Plant Pathology and Bacteriology. Cali, Colombia, pp. 225-234. almost equal rates, over the course of the 113 days. Time­ course data from leaf samples indicate that there is no Hopkins, D.L., and C.M. Thompson. 1984. Seasonal concentration of the Pierce's disease bacterium in 'Carlos' and 'Welder' musca­ need to wait for 11 weeks postinoculation for purposes of dine grapes compared with 'Schuyler' bunch grape. HortScience genotype screening. Additionally, based on visible symp­ 19:419-420. tom development of highly susceptible genotypes, it is Hopkins, D.L., H.H. Mollenhauer, and J.A Mortensen. 1974. Tol­ likely that the duration may be shortened considerably to erance to Pierce's disease and the associated rickettsia-like bac­ ~6 weeks postinoculation, whereas 4 weeks does not ap­ terium in muscadine grape. J. Am. Soc. Hortic. Sci. 99:436-439. pear to be long enough. Based on the greater X. fastidiosa Krivanek, A.F., and M.A Walker. 2005. Vitis resistance to Pierce's concentrations in the bottom leaves at 34 d postinocula­ disease is characterized by differential Xylellafastidiosa popula­ tion, it appears that sampling of tissues removed from the tions in stems and leaves. Phytopathology 95:44-52. point of inoculation should more accurately reflect the Krivanek, AF., T.R. Famula, A Tenscher, and M.A Walker. 2005a. susceptibilitylresistance characteristics of a genotype, Inheritance ofresistance to Xylella fastidiosa within a Vitis rupestris minimizing the localized influence of inoculum. x Vitis arizonica hybrid population. Theor. Appl. Genet. 111: 110­ 119. Conclusion Krivanek, A.F., J.F. Stevenson, and M.A. Walker. 2005b. Develop­ ment and comparison of symptom indices for quantifying grape­ This investigation into the responses of 18 grape geno­ vine resistance to Pierce's disease. Phytopathology 95:36-43. types to X. fastidiosa infection identified accessions and Krivanek, A.F., S. Riaz, and M.A Walker. 2006. Identification and breeding selections that dramatically differ in their ability molecular mapping ofPdRl, a primary resistance gene to Pierce's to support X. fastidiosa growth. These results were gener­ disease in Vitis. Theor. Appl. Gen. 112:1125-1131. ated under greenhouse screening conditions and will have Milholland, R.D., P.Y. Huang, C.N. Clayton, and R.K. Jones. 1981. to be verified under field conditions. However, this green­ Pierce's disease on muscadine grapes in North Carolina USA Plant house-based evaluation has been validated in previous Dis. 65:73-74. studies, and these newly identified sources of resistance Mortensen, J.A. 1968. The inheritance of resistance to Pierce's dis­ to X. fastidiosa can be used by breeders in their continu­ ease in Vitis. Proc. Am. Soc. Hortic. Sci. 92:331-337. ing efforts to introgress PD resistance into elite cultivars Mortensen, J.A., L.H. Stover, and C.P. Balerdi. 1977. Sources of of V. vinifera. resistance to Pierce's disease in Vitis. J. Am. Soc. Hortic. Sci. 102:695-697. Literature Cited Raju, B.e., and Ae. Goheen. 1981. Relative sensitivity of selected grapevine cultivars to Pierce's disease bacterial inoculations. Am. Blua, M.J., P.A Phillips, and RA. Redak. 1999. A new sharpshooter J. Enol. Vitic. 32: 155-158. threatens both crops and ornamentals. Calif. Agric. 53:22-25. Ruel, J.J., and M.A Walker. 2006. Resistance to Pierce's disease Davis, M.J., A.H. Purcell, and S.Y. Thomson. 1978. Pierce's dis­ in Muscadinia rotundifolia and other native grape species. Am. J. ease of grapevines: Isolation of the causal bacterium. Science 199: Enol. Vitic. 57:158-165. 75-77. Sorensen, J.T., and R.J. Gill. 1996. A range extension ofHomalodisca Davis, M.J., AH. Purcell, and S.Y. Thomson. 1980. Isolation me­ coagulata (Say) (Hemiptera: Clypeorrhyncha: Cicadellidae) to dia for the Pierce's disease bacterium. Phytopathology 70:425-429. Southern California. Pan-Pac. Entomol. 72:160-161. Doucleff, M., Y. Jin, F. Gao, S. Riaz, AF. Krivanek, and M.A Walker. Stoner, W.N., L.H. Stover, and G.K. Parris. 1951. Field and labo­ 2004. A genetic linkage map of grape, utilizing Vitis rupestris and ratory investigations indicate grape degeneration in Florida due to Vitis arizonica. Theor. Appl. Genet. 109:1178-1187. Pierce's disease virus infection. Plant Dis. Rep. 35:341-344. Fry, S.M., and RD. Milholland. 1990. Multiplication and trans­ Sutula, C.L., J.M. Gillett, S.M. Morrissey, and D.e. Ramsdell. 1986. location of Xylella fastidiosa in petioles and stems of grapevine Interpreting ELISA data and establishing the positive-negative resistant, tolerant, and susceptible to Pierce's disease. Phytopa­ threshold. Plant Dis. 70:722-726. thology 80:61-65. Takiya, D.M., S.H. McKamey, and RR. Cavichioli. Validity of Halbrooks, M.e., and J.A Mortensen. 1989. Origin and significance Homalodisca and of H. vitripennis as the name for glassy-winged of Flordia hybrid bunch grapes and rootstocks. HortScience 24: sharpshooter (Hemiptera: Cicadellidae: Cicadellinae). Ann. 546-550. Entomol. Soc. Am. 99:648-655. Hill, B.L., and AH. Purcell. 1995. Multiplication and movement Thorne, E.T., B.M. Young, G.M. Young, J.P. Stevenson, J.M. Laba­ of Xylellafastidiosa within grapevine and four other plants. Phy­ vitch, M.A Matthews, and T.L. Rost. 2006. The structure ofxylem topathology 85:1368-1372. vessels in grapevine () and a possible passive mechanism Hopkins, D.L. 1989. Xylella fastidiosa: Xylem-limited bacterial for the systemic spread of bacterial disease. Am. J. Bot. 93:497­ pathogen of plants. Ann. Rev. Phytopathol. 27:271-290. 504.

Am. J. Enol. Vitic. 58:3 (2007)