Materials and methods

This study was conducted at the University of Florida Plant City Campus (U.S. Department of Agriculture cold hardiness zone 9a, American Horticultural Society heat hardiness zone 8) from Feb. 2008 to Jan. 2010. Six bare-root plants of the following cultivars were installed on 23 Jan. 2008: Belinda’s Dream, BUCbi (Carefree Beauty™), Duchesse de Brabant, WEKcisbako (Home Run^®), RADrazz (Knock Out^®), Mrs. B.R. Cant, Mutabilis, Old Blush, Perle d’Or, Spice, and “Bailey Red” (it should be noted that some rose authorities now believe “Bailey Red” is actually ‘Karen Poulsen’, but we have assigned the first name for the purpose of this study). These cultivars were selected because they have been deemed low-maintenance roses by Florida gardeners, the Texas Earth-Kind® program (Harp et al., 2009) or both. The experimental design was a randomized complete block with two plants per plot in three replications. The Seffner fine sand soil was previously amended with composted cow manure (Black Gold Compost Co., Oxford, FL) at a rate of 50 lb every 1.5 ft along planting beds. Soil pH was 7.7 after amendments. The compost was applied 3 inches thick in a 2-ft-wide central band and then incorporated into the upper 6 inches of soil. Plants were placed 8 ft apart on 8-ft-wide beds covered with ground cloth topped with a 4-inch layer of fine ground pine bark mulch. The irrigation regime included an establishment period providing water by drip emitters (two per plant) at a rate of 0.5 gal per plant every day for the first 2 weeks and gradually decreasing to every other day for the next two months and once per week for the following 3.5 months. During the maintenance period, 0.5 gal was provided twice per week from October to May and once per week from June to September. Top-dress applications of 15N–3.9P–10K controlled-release fertilizer (Osmocote Plus 15–9–12; Everris International, Geldermalsen, The Netherlands) were performed as follows: three applications in 2008 on 27 Mar., 9 June, and 8 Sept. at ≈4.5, 7.0, and 11.7 g/plant, respectively. During the second year, fertilizer amounts were increased in proportion to plant growth and ≈23.0 g/plant were applied on 28 Feb. and 25 July. No pesticide applications or regular grooming was done except for one major pruning on 25 Feb. 2009 when the plant canopy was reduced by one-third.

Incidence of black spot, cercospora leaf spot, and foliar damage caused by chilli thrips (Fig. 1A–C) was assessed as these affected plant vigor and caused a decline in flower production in many cultivars. Assessments to quantify damage by these pests were conducted on all cultivars twice a month from Sept. 2008 to Apr. 2009 and weekly from June 2009 to Dec. 2009. A scale from 0 to 5 was used to determine the percentage of foliage with black spot and cercospora leaf spot symptoms where 0 = no symptoms, 1 = 1% to 20%, 2 = 21% to 40%, 3 = 41% to 60%, 4 = 61% to 80%, and 5 = 81% to 100% of symptomatic leaves (Moyer et al., 2008). A similar scale was used to evaluate the percentage of foliar curling, bronzing, or both, caused by chilli thrips feeding damage. Flower coverage was assessed as the percent of canopy covered by flowers using a scale from 1 to 5 where 1 = 0% (no flowers present), 2 = 1% to 25%, 3 = 26% to 50%, 4 = 51% to 75%, and 5 = 76% to 100% flower canopy coverage (Wilson and Knox, 2006). A similar scale (1–5) was used to measure overall visual plant quality as follows: 1 = very poor quality, plants not acceptable, severe leaf necrosis or yellowing; 2 = poor quality, not acceptable, large areas of necrosis or yellowing, poor form; 3 = fair quality, marginally acceptable, somewhat desirable form and color; 4 = good quality, very acceptable, nice color, good form, desirable; and 5 = excellent quality, very desirable landscape performance (Wilson and Knox, 2006). Weekly evaluations of plant quality and flower coverage were conducted for all cultivars from Mar. 2008 to Feb. 2009 and from Mar. 2009 to Jan. 2010 for 84 weeks total.

View Figure

• Download as Powerpoint
• Download Figure

Plant quality and flower coverage were considered categorical variables and thus, were analyzed using Fisher’s exact test each week separately. This test allows detecting statistical differences on the rating value’s frequencies for each variable among cultivars. Statistical comparisons between cultivars and plant quality and flower coverage were obtained with the FREQ procedure in SAS (version 9.2; SAS Institute, Cary, NC). Disease severity ratings for both diseases were analyzed by obtaining the area under the disease progress curve (AUDPC) according to Madden and Campbell (1990) for an interval of 67 weeks. AUDPC values were analyzed with the SAS MIXED procedure and differences among cultivars were compared using orthogonal contrasts and determined significant at P ≤ 0.05. The Pearson correlation coefficient between AUDPC values for black spot and cercospora leaf spot was calculated using the SAS CORR procedure. A damage severity index (DSI) was calculated to analyze chilli thrips foliar damage to the rose plants. The index was calculated for each plant and was obtained by multiplying the number of times each plant was evaluated as symptomatic by its corresponding number on the damage scale (within 1 to 5, rating of zero is not included since that indicated healthy plants) for 57 weeks total. The result was then divided by the total number of times a plant was rated as damaged. The formula can be summarized as DSI = {[∑ (frequency₁ × rating value₁) + (frequency_n × rating value_n)] ÷ (total number of symptomatic readings)}. The damage severity index was subjected to an analysis of variance using the ANOVA procedure in SAS. Tukey’s test was used to compare differences among DSI means (at P ≤ 0.05).

Results and discussion

This study evaluated the performance of 11 own-root OGR and MR shrub rose cultivars grown in central Florida with minimal inputs. Cultivars were grouped according to their overall performance as recommended, cautiously recommended, and not recommended. Rating frequencies for flower coverage and plant quality for the cultivars Mrs. B.R. Cant, Spice, and Belinda’s Dream are shown in Fig. 2A–F and are characteristic for the three groups (a complete list of the flower coverage and plant quality ratings for the 11 cultivars can be found in Supplemental Fig. 1). Flower coverage and plant quality data were analyzed each week to explore the statistical differences among cultivars in different weeks. Significant differences among cultivars occurred in 60 weeks for flower coverage and in 80 weeks for plant quality (data not shown). During 2008–09, differences in flower coverage among cultivars were significant in 26 weeks, whereas differences in flower coverage were significant in 34 weeks during 2009–10. Fewer differences were observed during 2008–09 because of lack of full establishment and little expression of flowering traits until 2009–10.

View Figure

• Download as Powerpoint
• Download Figure

All cultivars were repeat-flowering roses, but individual flowering characteristics of the class or cultivar also affected flower coverage ratings (Fig. 2A, C, E). For instance, MR were specifically bred and selected for increased flower display in the garden (Marriott, 2003); thus, ‘Belinda’s Dream’, ‘RADrazz’ (Knock Out^®), ‘BUCbi’ (Carefree Beauty™), ‘WEKcisbako’ (Home Run^®), and ‘Perle d’Or’ are well known for producing large clusters of flowers. However, with the exception of ‘RADrazz’ (Knock Out^®), these cultivars frequently received low flower coverage ratings (two or less) in our evaluation. Similarly, all OGR in this study characteristically produce flowers in clusters and coverage ratings were, in general, similar to those of the MR with the exception of ‘Mrs. B.R. Cant’. Both ‘Mrs. B.R. Cant’ (Fig. 2A) and ‘RADrazz’ (Knock Out^®) received flower coverage ratings of 75% or higher. The cultivars Spice (Fig. 2C) and Duchesse de Brabant had continuous flower coverage between 1% and 50%, but occasionally higher than 50%. Poor flowering cultivars such as Belinda’s Dream (Fig. 2E), “Bailey Red”, Old Blush, Perle d’Or, WEKcisbako (Home Run^®), Mutabilis, and BUCbi (Carefree Beauty™) had 25% or less flower coverage over 75% of the time. All cultivars showed decreased flowering during winter in both periods. Flower coverage could be influenced by plant vigor, particularly on own-root plants. None of our cultivars were grafted, thus in addition to pest pressure, this possibly contributed to a reduction in plant vigor and flower production. As previously mentioned, roses grafted onto ‘Double Cherokee’ perform better on Florida’s soils. However, grafting increases labor and costs in the nursery; therefore, selecting good performing own-root, low-maintenance roses could promote commercial production of such cultivars.

Cultivars showed significant differences in plant quality from the beginning of the evaluation as the plants came from different sources and were of different sizes. A general improvement of quality was observed in all cultivars in the first months of evaluation; however, various factors soon began to affect plant appearance. This trend can be observed on the plant quality ratings for our cultivar examples as these changed through the seasons (Fig. 2B, D, F). Cultivars of OGR occasionally produced sprawling canes, which downgraded their appearance, whereas cultivars of MR generally had neat, compact forms. Black spot, cercospora leaf spot, and chilli thrips were responsible, to a lesser or greater degree depending on cultivar, for defoliation and reduced plant growth and vigor resulting in poor plant quality. Only ‘Mrs. B.R. Cant’ (Fig. 2B) showed an outstanding improvement in plant quality after the first year. This cultivar received good to excellent plant quality ratings consistently during 2009–10. ‘Spice’ (Fig. 2D), ‘Duchesse de Brabant’, and ‘RADrazz’ (Knock Out^®) were rated as fair to good quality at least 50% of the time. ‘Belinda’s Dream’ (Fig. 2F), “Bailey Red”, ‘Old Blush’, ‘Perle d’Or’, ‘WEKcisbako’ (Home Run^®), ‘BUCbi’ (Carefree Beauty™), and ‘Mutabilis’ were rated as very poor to poor for most of both periods.

AUDPC values for disease severity were significantly different among the 11 cultivars for both diseases (Table 1). Cultivars with low AUDPC values for each disease were considered resistant when compared with the other cultivars, whereas those with the highest values were considered susceptible. Lowest black spot AUDPC values were observed for the MR ‘WEKcisbako’ (Home Run^®) and ‘RADrazz’ (Knock Out^®) (Table 1), which were regarded as more resistant than the other cultivars. ‘RADrazz’ (Knock Out^®) has been previously reported showing low levels of defoliation because of black spot in north-central Texas under minimal inputs (Mackay et al., 2008). In addition, resistance to black spot and cercospora leaf spot on ‘RADrazz’ (Knock Out^®) has been reported also in Tennessee, MS (Mynes et al., 2007, 2009), and Alabama (Hagan and Arkidge, 2010). Even though yellowing and defoliation in the Knock Out^® series has been attributed to an unknown disorder (Hagan and Arkidge 2009), cercospora leaf spot was continuously observed in ‘RADrazz’ (Knock Out^®) in our study. However, all plants of this cultivar retained most of their foliage year-round with no yellowing present. It is possible that the foliage density of ‘RADrazz’ (Knock Out^®) could have contributed to the persistence of cercospora leaf spot on this cultivar. ‘WEKcisbako’ (Home Run^®) was selected from a cross involving ‘RADrazz’ (Knock Out^®) as the pollen donor (Weeks Roses, 2010), which suggests that the high level of resistance to black spot could have come from this parent. The OGR ‘Mutabilis’ was the most susceptible to black spot. For cercospora leaf spot, the lowest AUDPC values were found for the OGR ‘Old Blush’ (Table 1), which was significantly more resistant than the other cultivars. The most susceptible cultivar was the MR ‘WEKcisbako’ (Home Run^®).

Table 1.Mean values of the area under the disease progress curve (AUDPC) for the severity of black spot and cercospora leaf spot in 11 cultivars of old garden (OGR) and modern (MR) roses in central Florida from Sept. 2008 to Dec. 2009.

View Fullscreen

Manners (1999) noted that horticultural classes of MR are generally more susceptible to black spot than those of OGR. In contrast, we found that most of the MR cultivars in our study were significantly more resistant to black spot than the OGR (Table 1). The exception was ‘Belinda’s Dream’. Although previously reported as resistant (Manners, 1999), it was susceptible to black spot in our evaluation. Manners (1999) reported that OGR ‘Mrs. B.R. Cant’, ‘Duchesse de Brabant’, and ‘Mutabilis’ were resistant to black spot, but these cultivars were susceptible in our evaluation. There are more than 20,000 rose cultivars registered worldwide, thus neither Manners’ cultivar list nor ours are totally representative of any particular group. Hence, it is important to remark that evaluating cultivars for suitability in specific areas is significant. Differences to black spot susceptibility among the cultivars screened in this study compared with previous reports might be explained by the occurrence of races of the fungus as previously documented (Wenefrida and Spencer, 1993; Whitaker et al., 2007; Yokoya et al., 2000). In addition, Whitaker et al. (2007) indicated that genetic variation of black spot pathogen can be found even within isolates of the same race, although races and specific geographical locations might not be significantly associated. The rose cultivars used in our trial were obtained from nurseries located in three different states (California, Alabama, and Texas) presenting the possibility that the plants were previously infected by pathogenic races at those sites other than those already present in central Florida. Individual cultivar characteristics might have also contributed to susceptibility levels. ‘Mrs. B.R. Cant’ has a large, open canopy and did not produce regular flushes of new foliage. These characteristics might have reduced the probability of continuous infection compared with cultivars with denser canopies (such as ‘Mutabilis’) that may retain humidity and inoculum for longer periods. Cercospora leaf spot on roses has been regarded nationwide as less important than black spot and thus, information is lacking on the epidemiology of this disease. Although some cultivars have been screened for tolerance to cercospora leaf spot, possible resistance mechanisms in rose cultivars have not been established.

Previous studies reported that cultivars highly susceptible to black spot show more resistance to cercospora leaf spot and vice versa (Mynes et al., 2007). We observed a similar tendency as highly resistant cultivars to black spot such as ‘RADrazz’ (Knock Out^®) and ‘WEKcisbako’ (Home Run^®) were also among the most susceptible to cercospora leaf spot as shown by the AUDPC. The opposite was also noted as ‘Old Blush’ and ‘Mutabilis’, two OGR cercospora leaf spot-resistant cultivars, were among the most susceptible to black spot. Evaluating the AUDPC values, we obtained a Pearson correlation coefficient (−0.77, P < 0.001) suggesting a significant negative relationship between susceptibility to black spot and cercospora leaf spot within the cultivars in our study.

Chilli thrips were reported first in Florida in 2005 on Knock Out^® roses but were considered established as a pest in 2007 (Hodges et al., 2005; Ludwig and Bográn, 2007). Rose cultivars have not yet been evaluated for tolerance to this new pest. Although typical damage from chilli thrips was observed on all cultivars during our evaluation, mean values of damage severity (measured by the DSI) were significantly different among rose cultivars (Table 2). ‘Mrs. B.R. Cant’ had the lowest DSI, but it was not different from ‘Old Blush’ and “Bailey Red”. The remaining three OGR were the most susceptible to the feeding injury: ‘Spice’, ‘Duchesse de Brabant’, and ‘Mutabilis’ had the highest DSI values. Chilli thrips cause damage principally on the young leaves and flower buds of their hosts (Seal and Klassen, 2005). It is conceivable that cultivars with abundant and recurrent new foliage might be preferred and, in fact, those cultivars with the highest DSI values also produced new growth continuously. ‘Mrs. B.R. Cant’, ‘Old Blush”, and “Bailey Red”, which produced less profuse foliage and were more cyclical in their foliage production, had lower DSI values.

Table 2.Mean values of the damage severity index (DSI) caused by chilli thrips in 11 cultivars of old garden (OGR) and modern (MR) roses in central Florida from Sept. 2008 to Dec. 2009.

View Fullscreen

Cultivars were grouped according to their overall performance as recommended, cautiously recommended, and not recommended. Two years after establishment, ‘Mrs. B.R. Cant’ appeared to be the most suited for central Florida. This cultivar had low susceptibility to chilli thrips (Table 2) and was moderately susceptible to black spot and cercospora leaf spot (Table 1). Plant quality of this cultivar improved greatly during the second year of evaluation and flower production was fairly constant (Fig. 1A and B).

‘Duchesse de Brabant’ and ‘RADrazz’ (Knock Out^®) were susceptible to cercospora leaf spot but less prone to defoliation than the susceptible cultivars. ‘RADrazz’ (Knock Out^®) also was moderately susceptible to chilli thrips, whereas ‘Duchesse de Brabant’ was highly susceptible. Despite these problems, these plants produced well-structured shrubs. Similarly, ‘Spice’ had an overall fair quality rating, maintained a well-rounded plant structure, and defoliated less even though it was highly susceptible to chilli thrips and moderately susceptible to black spot and cercospora leaf spot. Flower production on these three cultivars did not appear to be greatly reduced. Frequencies of plant quality and flower coverage ratings for ‘Spice’ are shown in Fig. 1C and D, which are representative for this group. These cultivars are cautiously recommended as low-maintenance, own-root cultivars in central Florida, particularly where flower production is valued over foliage appearance.

The OGR “Bailey Red” and ‘Old Blush’ and the MR ‘Belinda’s Dream’ were highly susceptible to black spot (Table 1). ‘Perle d’Or’, ‘BUCbi’ (Carefree Beauty™), and ‘WEKcisbako’ (Home Run^®), all MR, were highly susceptible to cercospora leaf spot (Table 1). All these cultivars suffered severe defoliation, poor growth, and reduced vigor, consequently failing to form well-structured shrubs. Frequencies of plant quality and flower coverage ratings for ‘Belinda’s Dream’ are shown in Fig. 1E and F representing the low performance of cultivars in this group. In addition, ‘Perle d’Or’, ‘WEKcisbako’ (Home Run^®), and ‘Belinda’s Dream’ were also prone to dieback. Although plants of these cultivars received high flowering ratings, these were due to the small size of the plant in comparison with the large clusters of flowers produced. ‘Mutabilis’ was the cultivar most susceptible to chilli thrips damage showing damaged foliage throughout both years. Chilli thrips damage and high susceptibility to black spot resulted in severe defoliation on this cultivar. However, all ‘Mutabilis’ plants survived and continued to grow, whereas those of ‘BUCbi’ (Carefree Beauty™), “Bailey Red”, ‘Perle d’Or’, ‘WEKcisbako’ (Home Run^®), and ‘Belinda’s Dream’ were greatly weakened after the second year and were removed from the trial. Consequently, these cultivars are not recommended as low-maintenance, own-root shrub roses for central Florida.

Units

View Fullscreen