Graft inoculations.

The original inoculum of CLas was from an infected ‘Lisbon’ lemon (C. limon) from a commercial grove near Ft. Pierce, FL (Albrecht and Bowman, 2008), which has been used extensively in experiments, with infected trees sustained through serial graft transmission. Trees of rough lemon (Citrus jambhiri, nucellar seedlings) and citron (Citus medica, cultivar Mhac Nao from rooted cuttings) were graft inoculated on 20 Oct. 2010 with CLas using infected ‘Valencia’ (Citrus sinensis) budwood at three buds per tree. These rough lemon and citron trees developed distinctive HLB mottling and high CLas titer and provided materials for graft inoculations for the experiment reported here (Stover et al., 2016a). Sixty trees each of ‘Valencia’ (budded onto rough lemon nucellar seedlings), nucellar seedlings of rough lemon, and nucellar seedlings of ‘Kuharske’ (C. sinensis × P. trifoliata) were graft inoculated on 26 Mar. 2012 with CLas using a combination of infected rough lemon and citron budwood at two buds per tree. Buds were cut from the branches clearly showing blotchy mottle symptoms, and buds with attached leaves were randomly assigned to target trees, with associated leaves of most buds having a cycle threshold (Ct) value of 22–25. All trees received at least one bud from rough lemon. Target trees were 4–8 mm in diameter when bud inoculated and 0.5–0.9 m in height. Thirteen trees (all but one were rough lemon) were rebudded in the same manner on 30 May 2012 when both inoculum buds were observed to be dead on 1 May 2012.

Illumination and scoring treatments.

Trees were placed in walk-in growth rooms on 4 June 2012 (Environmental Growth Chambers, GR Series with TC-2 Microcontrollers, Chagrin Falls, OH; one at 12 h light: 12 h dark periods and one with constant light. Both rooms were 16 h at 30 °C and 8 h at 25 °C). Ten trees of each genotype in each growth room were scored (through the bark but not the wood) with a knife on 18 July 2012 and the scoring was repeated on the same trees at the same scoring wounds on 30 Aug. 2012. Trees were removed from growth rooms on 12 Dec. 2012 and subsequently maintained in a greenhouse with ambient lighting and temperatures ranging from 15.6 to 30.6 °C (mean 22.8 °C).

Data collection.

At 2 to 3 month intervals between June 2012 and May 2013, HLB symptoms and stem diameter at 5 cm above the soil were assessed, and three leaves per tree were collected and pooled for qPCR determination of CLas titers. Presence of blotchy mottle was visually estimated for each tree as the percentage of total canopy leaf area affected. During 29–31 Oct. 2012, chlorophyll levels were measured using a SPAD meter (Konica Minolta Inc., Tokyo, Japan) on all nongirdled trees. Three of the most symptomatic leaves were measured on each tree and the three chlorophyll measurements were then averaged.

Determination of CLas titer.

All leaf samples were stored at −20 °C until processed. For each tree, the lower 2 cm from the leaf petiole base into the midrib were macerated in a Fast Prep 24 (MP Biomedicals, Santa Ana, CA) with DNA extracted from 200 mg of macerated tissue using cetyl trimethylammonium bromide (CTAB) extraction followed by column clean up and concentration (Genesee Scientific, San Diego, CA). Each sample was assayed for CLas and citrus dehydrin via qPCR using an ABI 7500 thermocycler (Applied Biosystems, Foster City, CA). Each 25 µL qPCR reaction was conducted using 100 ng of DNA template with primers specific to 16s rDNA for CLas detection (USHRL-CL1-L 5′-CTTACCAGCCCTTGACATGTATAGG-3′ and USHRL-CL1-R 5′-TCCCTATAAAGTACCCAACATCTAGGTAAA-3′) or Citrus dehydrin (CD-L 5′-TGAGTACGAGCCGAGTGTTG-3′and CD-R 5′-AAAACTTCACCGATCCACCAG-3′). Citrus dehydrin was used as reference control to verify DNA quality and PCR amplification. The master mix consisted of Promega (Madison, WI) GoTaq qPCR with BRYT Green and 0.2 µm of each primer using cycle parameters of 2 min 95 °C start, 40 cycles of 15 s at 95 °C, 20 s at 60 °C, and 20 s at 72 °C, with a melt curve analysis (to verify product identity). From Ct values, numbers of CLas genomes were calculated per ng DNA extracted, based on previously published laboratory analyses (Stover et al., 2015).

Statistical analyses.

All data including qPCR results (analyzed as Ct) were assessed using the nonparametric Kruskal–Wallis test using Statistical Analysis Software version 9.4 (SAS Institute Inc., Cary, NC). Where CLas was found to be nondetectable, a Ct value of 45 was assigned for use in the analyses (which is less than one copy per sample analyzed). Regressions of SPAD readings vs. CLas Ct in Nov. 2012 were conducted using Excel (Microsoft, Redmond, WA).

Development of CLas titer.

The main effect of genotype, averaged across all treatments, was significant at three of four dates where titers were compared, with ‘Kuharske’ showing the lowest titer and ‘Valencia showing the highest titer (Table 1), as expected based on previous studies. The main effect of scoring was never significant and illumination averaged across genotypes and scoring was significant only in Nov. 2012, when CLas titer was significantly enhanced by continuous illumination with ≈150-fold more CLas than in 12 h light: 12 h dark.

Table 1.Cycle threshold (Ct) values (translation into bacterial titer in footnote) for three leaf samples from each tree at each date, collected for quantitative polymerase chain reaction (qPCR) determination of titers of huanglongbing (HLB) pathogen Candidatus Liberibacter asiaticus (CLas). Sixty trees of each of three citrus genotypes (‘Kuharske’, previously shown to be HLB resistant; rough lemon, previously shown to be HLB tolerant; and ‘Valencia’, HLB susceptible) were bud grafted using two CLas-infected buds per tree on 26 Mar. 2012 and were placed in controlled growth rooms (one 12 h light: 12 h dark and one constant light, both rooms were 16 h at 30 °C and 8 h at 25 °C) on 4 June 2012. Ten trees of each genotype in each growth room were scored with a knife on 18 July 2012 and the scoring was repeated on the same trees at the same scoring wounds on 30 Aug. 2012. Trees were removed from growth rooms on 12 Dec. 2012 and subsequently grown in a greenhouse.

View Fullscreen

Three-way interactions of genotype/illumination/scoring revealed several interesting points (Table 1). On Sept. 2012, at 6 months after inoculation, both ‘Kuharske’ and ‘Valencia’ displayed markedly greater CLas titer in continuous illumination, except that ‘Kuharske’ displayed this increase only when scored, and ‘Valencia’ only when not scored. On Nov. 2012, at 8 months after inoculation, all genotypes displayed markedly greater CLas titer in continuous illumination compared with 12 h light: 12 h dark, whether scored or not, though ‘Valencia’/continuous illumination/scored had higher titer than ‘Valencia’/continuous illumination/not scored. Most markedly, all three genotypes had titers which were not statistically different from each other when exposed to continuous illumination and sampled at 8 months after inoculation. Rather than facilitating differentiation between HLB-susceptible and HLB-resistant genotypes, exposure to continuous illumination accelerated pathogen development but compressed differences in susceptibility. With a few exceptions, differences between treatments within genotypes were few and inconsistent on Mar. 2013 and May 2013 after trees had been removed from the growth rooms and returned to ambient lighting in the greenhouse.

Development of distinctive HLB mottle.

Negligible mottle was observed before Sept. 2012. ‘Kuharske’ trees displayed slight mottle symptoms when exposed to continuous illumination (at 6–8 months following inoculation, Sept. 2012 and Nov. 2012), which were only statistically significant when combined with scoring (Table 2). Rough lemon also displayed only minor levels of mottling with inconsistent treatment differences, but with higher mottling at 12 h light: 12 h dark vs. continuous lighting with no scoring in Nov. 2012 and Mar. 2013. ‘Valencia’ consistently showed the greatest numerical percentage of leaf area with mottle symptoms compared with the other genotypes. At almost every date of comparison, within ‘Valencia’, the greatest level of mottling was observed in trees with continuous lighting and no scoring.

Table 2.Percentage of leaf area showing typical huanglongbing (HLB) symptom of blotchy mottle. Sixty trees of each of three citrus genotypes (‘Kuharske’, previously shown to be HLB resistant; rough lemon, previously shown to be HLB tolerant; and ‘Valencia’, HLB susceptible) were bud grafted using two Candidatus Liberibacter asiaticus (CLas)-infected buds per tree on 26 Mar. 2012 and were placed in controlled growth rooms (one 12 h light: 12 h dark and one constant light, both rooms were 16 h at 30 °C and 8 h at 25 °C) on 4 June 2012. Ten trees of each genotype in each growth room were scored with a knife on 18 July 2012 and the scoring was repeated on the same trees at the same scoring wounds on 30 Aug. 2012. Trees were removed from growth rooms on 12 Dec. 2012 and subsequently grown in a greenhouse.

View Fullscreen

SPAD assessment of chlorophyll.

SPAD readings correlated with CLas titer on the one date when tested at 8 months after inoculation (Table 1). An examination of the means in Table 1 shows significant differences between SPAD readings corresponded to significant differences between CLas titer from the same date. When individual tree data for SPAD and Ct were compared, the correlations were significant and positive across all three genotypes combined (P = 0.0008), for ‘Valencia’ alone (P = 0.03) and for rough lemon alone (P = 0.002) but not Kuharske alone (P = 0.12). However, the r² for regression across all genotypes is only 0.09 (r² is 0.06 in Kuharske, 0.23 in rough lemon, and 0.11 in ‘Valencia’) making SPAD assessments of questionable value in comparing CLas titers between genotypes.

Tree growth.

Tree response to illumination treatments, as measured by change in trunk diameter, varied between genotypes. Both ‘Kuharske’ citrange and rough lemon displayed greater growth in the constant illumination condition (Fig. 1), but following removal to the greenhouse in Dec.2012, growth rates were very similar between treatments, essentially maintaining differences that developed in the growth room. ‘Valencia’ sweet orange displayed differences between treatments in mean tree diameter from the first measurements, indicating our visual observation of uniformity was inaccurate when trees were assigned to treatments. The ‘Valencia’ trees exposed to 12 h light: 12 h dark, combined with scoring, were the largest throughout the study, even preceding scoring. The rate of growth, however, was different between the treatments with the trees exposed to continuous light growing more slowly than the trees with 12 h light: 12 h dark. As in the other two genotypes, following removal to the greenhouse, growth rates were very similar between treatments. Scoring treatments had no observed effect on tree growth.

View Figure

• Download as Powerpoint
• Download Figure

HLB tolerance.

Recent reports from Florida have documented tolerance to HLB in a number of genotypes including rough lemon that was shown to display less phloem disruption than the highly susceptible sweet orange (Fan et al., 2012). Several other genotypes with citron in the pedigree have also been documented to display markedly greater growth and canopy health than most HLB-affected accessions in the genus Citrus (G.P. Miles et al., unpublished data). It was noted that HLB tolerance in mandarin hybrids was not associated with delayed symptoms or lower CLas titers than highly HLB-susceptible sweet orange and grapefruit (Stover et al., 2016b). Response of HLB-tolerant rough lemon in this study was generally similar to ‘Valencia’ in CLas titer (Table 1), but rough lemon displayed markedly greater growth than ‘Valencia’ (Fig. 1), which is consistent with HLB tolerance.