Average air temperature at 2 m (°C) and calculated daylight integral (DLI) (mol·m−2·d−1) for the growing seasons 2022, 2023, and 2024. Climate data were collected using the on-farm weather station from the University of Georgia Automated Environmental Monitoring Network Information in 15 min intervals, which determined total solar radiation. Total solar radiation was converted to photosynthetic active radiation for DLI calculations.
Fig. 2.
Representative plots directly after harvest for 2023 group I flowers. Jars measure 17 cm in height. (A) 0% shadecloth. (B) 30% shadecloth. (C) 50% shadecloth.
Fig. 3.
Representative plots directly after harvest for 2023 group II flowers. Jars measure 17 cm in height. (A) 0% shadecloth. (B) 30% shadecloth. (C) 50% shadecloth.
Fig. 4.
Cumulative stems harvested and growing degree days (base temperature 14 °C) for group I lisianthus (mean ± standard deviation). Values on the same day after transplant with the same letter are not significantly different from one another (P < 0.05) according to the Tukey-Kramer Least-square means test.
Fig. 5.
Cumulative stems harvested and growing degree days (base temperature 14 °C) for group II lisianthus (mean ± standard deviation). Values on the same day after transplant with the same letter are not significantly different from one another (P < 0.05) according to the Tukey-Kramer Least-square means test.
Fig. 6.
Relationship of growing degree day (GDD), cumulative daylight integral (DLI), and cumulative stems harvested for all years and groups. Shade represented by different symbols (● ■ ▲) for 0%, 30%, and 50% shade, respectively, and years and groups represented by different colors. Data fit to multiple regression using SAS Proc Reg.
Effect of Shade on Lisianthus (Eustoma grandiflorum) Stem Length, Flowering Date, and Quality for Organically Field-Grown Cut Flower Production in Georgia, USA
Authors:
,
, and
Click on author name to view affiliation information
The cut flower industry is growing in the United States and farmers are looking to expand production and enter larger markets. However, farmers growing lisianthus (Eustoma grandiflorum) in the field often see reduced stem length and flower quality. Limited work has been conducted to optimize flower quality of field-grown lisianthus, especially in the southeastern United States, which is crucial for growers to maximize profits. In this study, lisianthus (group I and II ‘Arena’ white) was grown under field conditions with three shade treatments (0%, 30%, and 50%) to determine the effect on flower development, flower quality, marketability, and yield. Results showed a delay in the number of stems harvested with the addition of shade. Group I was more impacted by heavy shade treatments in terms of overall flower quality and delayed the timing of flower development by 10 to 15 d. Although group II was less affected by the addition of shade, both groups show trends of maintained quality marketability, and increased stem length with the 30% shade treatments by increasing stem length on average of 6.3 and 9.0 cm for group I and group II, respectively. These results indicate that inexpensive shadecloth may improve stem length and marketability while highlighting the need for further studies to examine lower shade intensities for cut flower production under field conditions for southern growers.
The US domestic cut flower and florist greens production was valued at more than $763 million in 2022 (Scott 2024); however, much of the cut flower industry is still driven by international imports, and the total floral retail market is valued at $6.43 billion in the United States (Kenner et al. 2023; Scott 2024). In recent decades there has been increased interest in local, organic, sustainable, and specialty cut flowers (Armitage and Laushman 2003; Darras 2021; Loyola et al. 2019; Scott 2024). This interest has led many small to midsized growers adding flower production to their diversified vegetable farms and the growth of more farms focusing specifically on cut flower production (Lewis et al. 2021). The number of US domestic farms with cut flower production increased almost 50% from 2017–22 (Scott 2024). Cut flower production offers a high dollar per land area value for small growers with limited field space (Wien 2009) and may be a profitable niche for local growers if they can enter wholesale and florist markets. In a survey of commercial florists located in Utah, florists are willing to pay more for local cut flowers, but florists voiced the need for the same quality as imported and commercially grown cut flowers (Curtis and Stock 2023); therefore, it is crucial for farmers in cut flower production to focus on optimizing quality to maximize profits.
Lisianthus (E. grandiflorum) has quickly become a flower of interest since its reintroduction from native plant to cut flower in the 1980s (Halevy and Kofranek 1984; Harbaugh et al. 2000). Lisianthus has gained popularity because of its long vase life, rose-like bloom, and broad range of color variations (Halevy and Kofranek 1984; Harbaugh 2007; Lugassi-Ben-Hamo et al. 2010). Lisianthus are considered one of the top cut flowers, but the overall price and the ability to be sold on a wholesale market will depend on flower quality, flower type (single, double, frilled, etc.), and stem length. The Dutch Flower Auction Association (VBN 2025) grades the quality of lisianthus by stem length (ranging from 65 to 80 cm), the number of flower buds, and maturity stage for wholesale. Direct to florist or farmer markets likely allow for shorter stems (40 to 50 cm) but stem length and quality are still of importance (Curtis and Stock 2023). At the Boston Terminal Market, wholesale prices for lisianthus (Dec 2024) also vary by sourcing location, with Canada long valued at $2.25 to $2.50 per stem and Netherlands long valued at $2.70 to $3.00 per stem (USDA 2024).
Although lisianthus is a native species to the southern United States and Mexico (Halevy and Kofranek 1984), little work has been done in maximizing quality for field production in the region. Hot summers and the potential for high light intensity in the southeastern region can reduce stem length and flower quality of lisianthus, which may reduce profits for field-grown lisianthus. Previous research (across multiple cut flower species) has shown shorter stems and lower-quality flowers in field compared with greenhouse or high tunnel production (Kelly 1991; O’Connell 2018; Ortiz et al. 2012; Owen et al. 2016); however, for space limited or greenhouse/high tunnel–limited growers, maximizing field production quality can increase profits and space. Increased field production allows for more crop rotation within high tunnels (to potentially reduce disease and pest pressure), as well as allowing for higher-dollar crops to be grown in greenhouse/high tunnels when space is limiting. Given the issues of reduced stem length and marketability in the Southeast, maximizing stem length of lisianthus whilemaintaining/increasing quality under field conditions could be useful to growers looking to enter the wholesale market.
Cultivar selection of E. grandiflorum will have an impact on flower quality and marketability. Significant differences in plant height and stem length were observed for 49 cultivars tested under greenhouse conditions (Harbaugh et al. 2000). Beyond morphological variations by cultivar, lisianthus are also grouped by bloom timing (groups I–IV). Groups can help determine flowering time, growing location, stem length, and plant date (where group I has earlier flowering time and group IV has the latest flowering). Bloom timing groups are in response to light and heat and can lead to different stem length, flower size, and quality (Stock et al. 2022). Group I will reach reproductive stage earlier than group II, which could impact stem length at harvest and often flower quality and/or doubling depending on the cultivar. Although groups help growers achieve successive flowering and harvests, plant response is dependent on environmental conditions and regional research is needed (Lewis et al. 2021; Ohkawa et al. 1991).
Shadecloth has commonly been used to affect light, quality, stem length, and other parameters in horticultural crops (Armitage 1991; Li et al. 2017; McElwee 1949; Stamps 2009). Low (1 m tall) shade structures composed of conduit pipes and shadecloth are inexpensive structures (∼$2.5 per foot of shadecloth; $6 per hoop) and fit into current production practices. Previous studies have shown an increase in stem length, but a delay in flowering and often quality and yield for cut flowers. Armitage (1991) determined an increase in stem length with the use of 55% and 67% shade for Echinops, Anemone, Centaurea, Eryngium, and Zantedeschia, but decreased yields and delayed flowering time for Centaurea, Eryngium, and Zantedeschia. Lugassi-Ben-Hamo et al. (2010) used 67% or 88% shade in intervals from 3 to 8 weeks for greenhouse-grown lisianthus. The group found significant delay in floral transition, reduced yield, and reduced stem length for ‘Mariachi White’, ‘Rosita White’, and ‘Echo Champaign’ lisianthus when shadecloth was used with significant interaction with cultivar. Shading decreases photosynthesis and lessens plant growth while increasing stem length, so finding a balance of shade intensity is necessary (Lugassi-Ben-Hamo et al. 2010). For field-produced flowers, shadecloth also may provide protection from environmental elements by reducing light, changing soil temperature, and changing relative humidity (Kittas et al. 2012; Li et al. 2017; Stamp 2009).
Previous studies with shade have shown an increase in stem length but a delay in flowering and often quality and yield for cut flowers. These studies focus high shade intensity shadecloth (greater than 50%) and greenhouse production; however, data are lacking for lighter shadecloth, field studies (where light intensity is less limiting), and regionally based studies for southern lisianthus growers. Therefore, the objective of this trial was to determine the effects of three shade treatments (0%, 30%, and 50% shade) on organically, field-grown lisianthus (group I and II ‘Arena’) flower development, yield, and quality.
Materials and methods
Site preparation and planting
This study was conducted at the University of Georgia Durham Horticulture Farm in Watkinsville, GA, USA (lat. 33°50 N, long. 83°30 W) in Spring/Summer of 2022, 2023, and 2024. The study was conducted on uncertified land (certified organic plugs were unavailable to be used on organic land) but grown using organic practices. Cultural practices followed typical field production techniques for both large and small growers. Soil was classified as a Cecil Sandy Loam with a 0% to 2% slope (USDA 2024). The soil had 2.8% organic matter and pH of 6.15 (water equivalent). Cover crops (mustard mix in 2022 and clover in 2023/2024) were incorporated 4 weeks before planting and Nature Safe 10N–2P2O5–8K2O (All Season Organic Fertilizer; Nature Safe, Irving, TX, USA) fertilizer was applied and incorporated just before planting to supply 56 kg N, 11 kg P, and 45 kg of K per hectare. E. grandiflorum ‘Arena Pure White’ F1 (America Takii Inc., Salinas, CA, USA) group I (2022 and 2023) and group II (2023 and 2024) 10-week-old plugs were sourced from Gro n’ Sell (Chalfont, PA, USA) in 128-cell trays and hardened off for 3 days before planting. Plots were organized in split-plot randomized complete block design with flower group as main effect and three treatments of black shadecloth (0%, 30%, and 50%; Johnny’s Seeds, Fairfield, ME, USA) and randomized within blocks. There were a total of four reps per group × shade treatment. Plots were set on a raised bed with 1.5 m length × 0.9 m width allowing for 60 plants per plot (43 plants/m2) at 15 × 15 cm spacing (Stock et al. 2022). A mulched, 1.5 m buffer was placed between each plot. On 20 Apr 2022, 18 Apr 2023, and 9 Apr 2024, group I (2022 and 2023) and group II (2023 and 2024) plugs were planted with three lines of drip (Toro, Aqua-Traxx, 0.75 lph, 30 cm emitter spacing) in 15 × 15 cm floral netting (Tenax Hortonova, Baltimore, MD, USA). At planting, plugs were watered by hand. Irrigation was then run daily for the first 4 weeks, followed by every other day to supply ∼2.5 cm of irrigation per week. Shade tunnels were constructed using conduit (1.2 m tall), placed 3 weeks after planting (to allow for plant establishment and reduce the effects on initial growth), and left until last harvest. Plots were fertilized with Neptune’s Harvest fish emulsion (1:50 ratio with water; 2N–4P2O5–1K2O; Gloucester, MA, USA) 1, 2, 3, and 4 weeks after planting and weeded by hand as needed. Climate data were collected using the on-farm weather station from the University of Georgia Automated Environmental Monitoring Network Information in 15 min intervals, which included soil temperature/moisture, air temperature, relative humidity, total solar radiation, and growing degree day (GDD) calculations (UGAAEMN 2024). GDDs were calculated with a base temperature of 14 °C according to work from Höhn et al. (2023). Solar radiation for the shade treatments was estimated by multiplying the measured total solar radiation by the reduction of light from the addition of shadecloth (1% shade for each treatment). Daylight integrals (DLI) were calculated (Spall and Lopez 2024) with the amount of photosynthetically active radiation (PAR) estimated as 0.45 of the total cumulative radiation (Meek et al. 1984). Cumulative DLIs (mol·m−2·d−1) were calculated as the sum of DLIs each day and correlated with stems harvested for each shade treatment. Although microclimate and PAR under shadecloth were not measured directly because of equipment restraints, the aforementioned parameters allow for best estimate on shadecloth impacts on light and temperature.
Data collection and analysis
Preharvest plant height was determined from soil to plant tip. Stems were harvested when two flowers were completely open and one bud was showing color. At harvest, total number of stems per plot was determined, as well as number of buds and flowers per stem, stem length, and overall quality (visually rated on a scale from 1 to 5 by same rater each year). Overall quality was visually assessed based on flower color, petal density, flower fullness, and pest/environmental damage. A quality rating of 1 would indicate a highly damaged flower (rain/pests, excessive browning, and or reduced flower fullness). A quality of 3 would indicate some flowers with mild damage or minimal fullness, but overall quality would be adequate for sale. A quality of 5 would indicate white blossoms, flower fullness, straight stem, and upright flowers. Quality and quantitative attributes (stem length, flower count, bud count) were assessed at each harvest, for each stem harvested by the authors. For this study, “marketability” was defined as greater than 40 cm stem length for group I flowers, greater than 50 cm for group II flowers, and a quality rating of 3 or higher for both groups. These stem-length values are lower than wholesale market standards (VBN 2025) but reflect higher-quality flowers meant for direct sale from farm to florist. Plants were harvested and the first 40 flowers from each plot were used for statistical analysis of quality parameters. Although this decision could lead to bias of earlier-maturing stems, it was done to standardize harvests across treatments and account for plots with higher stem loss due to biotic and abiotic factors. Data were analyzed using a linear mixed model using PROC GLMMIX in SAS 9.4 (SAS Institute, Cary, NC, USA). Stem length, bud number, flower number, quality, preharvest height, and total stems per plot and group being fixed effects and year and block considered a random effect. Least-square means comparisons were performed using the Tukey-Kramer (P < 0.05) test when appropriate. During statistical analysis, year and group were significant, therefore data have been shown separately by both year and flower group (I or II). SAS PROC REG was used to determine linear regression fit and parameters for cumulative stem harvest, cumulative DLI, and GDDs.
Results
Weather data
Historical rainfall from 1 Apr to 1 Aug from 1949 to 2016 for the region is considered 516 mm with average maximum temperatures of 29 °C and minimum temperatures of 17 °C [University of Georgia (UGAAEMN) 2024]. Both 2022 and 2024 had less than 200 mm of rainfall during the trial compared with more than 300 mm in 2023 (Table 1). Average air temperatures were similar between studies; however, differences in cumulative temperature during the growing season can be observed through GDDs (Table 2). The studies in 2022 and 2024 had a larger number of GDDs to first harvest for both group I and group II lisianthus. Relative humidity and solar radiation were similar throughout study years with higher average relative humidity in 2023. First flowers for unshaded group I received a cumulative DLI of 2906, 3042, and 2901 mol·m−2·d−1 before first harvest for 2022, 2023, and 2024, respectively (Table 2). Group II received a cumulative total radiation of 3317 and 3417 mol·m−2·d−1 before first harvest for 2023 and 2024, respectively. Average DLI ranged from 43.5 to 46.9 mol·m−2·d−1 until first harvest with greater than 50 mol·m−2·d−1 immediately after planting for each year for the 0% shade treatment (Fig. 1). The addition of 30% shade led to an estimated average maximum DLI of 30.7 and minimum of 23.5 mol·m−2·d−1 across years and groups and the addition of 50% shade led to an estimated average maximum DLI of 23.5 and minimum of 21.8 mol·m−2·d−1 across years and groups.
Table 1.Air temperature, relative humidity, solar radiation, soil temperature, and rain collected at 2 m height at 15 min intervals from Watkinsville, GA, USA,i during the growing season for both groups. Values have been averaged over plant date to last harvest.
Table 2.First harvest, last harvest, harvest window duration, days after transplant (DAT), and growing degree days (GDDs) calculated for each group and year across all shade treatments (group I and II ‘Arena’ white) grown in Watkinsville, GA, USA.
Fig. 1.Average air temperature at 2 m (°C) and calculated daylight integral (DLI) (mol·m−2·d−1) for the growing seasons 2022, 2023, and 2024. Climate data were collected using the on-farm weather station from the University of Georgia Automated Environmental Monitoring Network Information in 15 min intervals, which determined total solar radiation. Total solar radiation was converted to photosynthetic active radiation for DLI calculations.
Overall, flower growth was good in all 3 years. Both year (P < 0.0001), group (P < 0.0001), and year × group (P < 0.0001) were significant, therefore data were separated for further analysis of shade treatments (Tables 3 and 4). Significant differences were determined in preharvest plant height by shade treatment for group I in 2022 and 2024 and group II in 2023 and 2024 (Table 3), with shade increasing plant height (Figs. 2 and 3). Tallest preharvest plants were noted in the 30% shade treatments across all years and groups that had significant shade effects (group I in 2023 had no shade effect on preharvest heights). No significant differences were determined in the total number of stems harvested by treatment over each season (Table 3). For group I, 35 stems/m2 were harvested on average (across all treatments) in a total of six harvests beginning 21 Jun 2022 and ending 12 Jul 2022; 32 stems/m2 in seven harvests from 27 Jun 2023 and ending 18 Jul 2023; and 33 stems/m2 from 14 Jun and ending in 1 Jul. For group II, 36 stems/m2 were harvested in seven harvests from 3 Jul 2023 to 25 Jul 2023 and 39 stems/m2 were harvested in seven harvests from 24 Jun 2024 to 8 Jul 2024.
Table 3.Preharvest plant height, total number of stems harvested per plot each growing season (at 2 open flower, 1 colored bud stage), and the percent marketability of the total stems harvested. Marketable stems were determined based on a quality rating of 3 or higher AND a stem length of at least 40 cm for group I and 50 cm for group II.
Table 4.Quality parameters for the first 40 harvested flowers. A quality rating of 1 would indicate a highly damaged flower (rain/pests, excessive browning, and/or reduced flower fullness). A quality rating of 3 would indicate some flowers with mild damage or minimal fullness, but overall quality would be adequate for sale. A quality rating of 5 would indicate white blossoms, flower fullness, straight stem, and upright flowers.
Fig. 2.Representative plots directly after harvest for 2023 group I flowers. Jars measure 17 cm in height. (A) 0% shadecloth. (B) 30% shadecloth. (C) 50% shadecloth.
Fig. 3.Representative plots directly after harvest for 2023 group II flowers. Jars measure 17 cm in height. (A) 0% shadecloth. (B) 30% shadecloth. (C) 50% shadecloth.
For group I, days after transplant (DAT) to first harvest were 62, 70, and 66 for 2022, 2023, and 2024, respectively (Table 2, Fig. 4). Harvest durations ranged from 21 to 17 d for group I flowers. Although there were differences in the timing of the number of stems harvested due to shade (discussed later in this article), there were no differences in first harvest date due to shade treatments for group I flowers. For group II flowers, the first harvest occurred 76 DAT in both years for 0% and 30% shade, despite 918 GDDs measured in 2023 compared with 1112 in 2024 (Table 2, Fig. 5). A much shorter harvest duration occurred in 2024 (14 d) for group II compared with 2023 (22 d). In contrast to group 1, 50% shade did affect first harvest date and the number of stems harvested for group II (Fig. 5, Table 3) The 50% shade treatment delayed first harvest by 11 d (238 GDD) in 2023 and 4 d (96 GDD) in 2024.
Fig. 4.Cumulative stems harvested and growing degree days (base temperature 14 °C) for group I lisianthus (mean ± standard deviation). Values on the same day after transplant with the same letter are not significantly different from one another (P < 0.05) according to the Tukey-Kramer Least-square means test.
Fig. 5.Cumulative stems harvested and growing degree days (base temperature 14 °C) for group II lisianthus (mean ± standard deviation). Values on the same day after transplant with the same letter are not significantly different from one another (P < 0.05) according to the Tukey-Kramer Least-square means test.
Although no differences in total stems harvested was determined, significant differences were determined by day during the first few harvests across years and groups. For group I, the 50% shade treatment had significantly fewer flowers for the first four harvests, whereas no difference was observed between the 0% and 30% shade treatments (Fig. 4). In 2023, group II had no significant differences between treatments for the first two harvests, averaging 0.8 and 3.0 stems/m2 76 and 80 DAT, respectively, but had significant differences from 3 to 93 DAT. In 2024, group II had significant differences in cumulative number of stems for the first harvest (76 DAT) with 11.3, 3.2, and 0 stems/m2 from the 0%, 30%, and 50% shade, respectively (Fig. 5). Significant differences continued through DAT 80 but 30% was not significantly different from 0% shade by DAT 83 with an average of 30.8 stems/m2 harvested. The 50% shade continued to have significantly fewer stems harvested through 87 DAT. The number of stems harvested was correlated as GDDs and DLI with an R2 = 0.7615 with all variables significant at P < 0.001 (Fig. 6). GDDs described 70% of the model and DLI described 6% of the model. Fit was good and followed linear trends for 2022 and 2023 group I, regardless of shade treatment, but differences in group II and the impact of shade could be seen as a source of variation in fit (Fig. 6). Like stems vs. DAT (Figs. 4 and 5), a lag in flowering with the addition of shade can be observed with group II even when expressed as a function of GDDs and DLI.
Fig. 6.Relationship of growing degree day (GDD), cumulative daylight integral (DLI), and cumulative stems harvested for all years and groups. Shade represented by different symbols (● ■ ▲) for 0%, 30%, and 50% shade, respectively, and years and groups represented by different colors. Data fit to multiple regression using SAS Proc Reg.
The addition of shade significantly affected stem length, bud number, overall quality, and the percent of marketable flowers for both groups, with greater effects observed in group I flowers (Table 4). Across all treatments, the 50% shade had the longest stems measured. As expected, group I had shorter stems than group II, with an average across all treatments and years of 46.8 cm compared with 68.1 cm. In group I, stem length increased 11.3, 11.8, and 9.4 cm on average between the 0% and 50% shade treatments in 2022, 2023, and 2024, respectively. The 0% shade treatment consistently led to significantly shorter stems. Trends were similar in group II, with significantly shorter stems in the 0% shade treatment and a difference of 15.3 and 7.7 cm between the 0% and 50% shade treatments in 2023 and 2024, respectively. Although the desired trait of stem length increased with increased shade intensity, the number of buds and overall quality and marketability decreased from 30% to 50% shade (Table 4). Across all treatments, group II had more buds counted than group I flowers. For both groups, there was a significant reduction in number of buds with the 50% shade treatment compared with 0% shade intensity. The significance of the 30% shadecloth on buds was variable (with some years showing a significant decrease and others not), but in all years, 0% shade maintained higher bud counts than the 50% treatment. Overall quality was significantly reduced with the 0% and 50% shadecloth compared with 30% shadecloth for all years and groups with the exception of group II in 2023, in which no significant difference was determined. With that exception, there was an increase in quality with the addition of 30% shadecloth. The percentage of marketable stems (quality greater than 3; group I stem length >40 cm and group II >50 cm) was not significantly different for group II flowers either year (Table 4) and averaged 76% of stems in both years. Differences were determined for group I flowers by year and shadecloth treatment. Overall, marketability was higher in 2022 for group I flowers, with increasing marketability with increasing shade. In both 2023 and 2024, there was a significant increase with the 30% shadecloth compared with the 0% and 50% shade treatments. Marketability was lowest in 2023, with very few harvested flowers deemed marketable for the 0% shadecloth (13.7%) and 50% shadecloth (15.2%).
Discussion
Shadecloth applications and light reduction have been shown to increase stem length, reduce photosynthesis, affect quality, and alter flowering and fruiting timings for a number of crops (Armitage 1991; Gude et al. 2022; Kabir et al. 2024; Lewis et al. 2021; Li et al. 2017; Lugassi-Ben-Hamo et al. 2010). For this study, significant differences were determined in flower timing, stem length, cut flower quality, and overall marketability in E. grandiflorum ‘Arena White’ group I and group II flowers with 0%, 30%, and 50% shade treatment under field conditions, with no significant differences in total number of stems harvested. Differences varied between year (likely because of weather variations) and were more pronounced in group I flowers compared with group II flowers. For this study, the addition of shade had less impact on the timing of first harvest and total number of stems harvested over the season, but significant effects on the number of stems harvested at each date (Figs. 4 and 5). Similar to our results, Halevy and Kofranek (1984) determined that photoperiods (long day, short day, natural) had no effect on time to flowering or number of flowers per plant produced. The group also determined that increased night temperature (13 °C vs. 18 °C) led to flowering time that was 11 to 23 d earlier. Spall and Lopez (2024) determined that photoperiod had less impact on flowering and finishing on dianthus if moderate DLIs were maintained. In that study, DLI greater than 14 mol·m−2·d−1 decreased time to flower and stem thickness for dianthus. Higher GDDs in 2022 and 2024 may have led to fewer DAT to first harvest observed in group I (Table 2). Interestingly, for group II, GDD accumulation did not seem to impact DAT to first harvest. The fit for predicting cumulative stems harvested by GDD and DLI (Fig. 6) overpredicted for the shadecloth treatments. This could be because temperature estimates for GDD did not take into account shade treatments. The effect of shade on ambient temperature has been variable. Studies have shown both a decrease and no effect on ambient temperature and GDDs with the use of shade (Gude et al. 2022; Kabir et al. 2024). For this study, shadecloth was hung to cover top and sides (to prevent rising/setting sun light), so it is difficult to predict how air flow and other environmental factors may have affected the ambient temperature and subsequent GDDs for shade treatments. In addition, shade treatments likely reduced photosynthesis directly and have been shown to lead to delayed floral transitions in lisianthus (Lugassi-Ben-Hamo et al. 2010). DLI was significant and did increase model performance but by only 6%. It is likely that the DLI observed under field conditions (even taking into account the use of 30% and 50% shade) was likely high enough during the study to have smaller impact on flower timing than GDDs for summer grown crops in the southeastern United States.
Lisianthus stem length increased with both 30% and 50% shade treatments in both groups. Stem length increased with increasing shadecloth from 30% to 50%, but bud number and overall quality decreased from 30% to 50%. Increasing shade percentages have been shown to have similar effects with decreased quality with increased stem length (Armitage 1991) and 30% shade in this study may have balanced protection from shadecloth, increased stem length, and maintained flower quality while increasing stem length. It is important to note that the group I flowers did not reach the length for wholesale flowers set by the Dutch Flower Auction Association of 65 to 80 cm (VBN 2025); however, on average, group II flowers were in the lower end of that range of desired stem length. “Marketability” in this study reflected more potential markets (organic, local florists, some wholesale, and farmers markets) where 40 cm for group I and 50 cm for group II would be deemed marketable and 30 cm stems have been used by other authors to determine marketability (Ortiz et al. 2012). In this study, the addition of 30% shade increased or maintained the overall quality of the flowers and the percentage of marketable total stems harvested. The 30% shade treatment may have led to longer stems because of reduced light and provided protection from rain, which can increase flower quality similar to high tunnels. Lewis et al. (2021) determined a decrease in flowering time and increase in stem length and marketability in high tunnel vs. field production for Utah-produced snapdragon. Similarly, Ortiz et al. (2012) determined an increase in stem length for Matthiola, Eustoma, Zinnia celosia, and Antirrhinum. In their study, Eustoma ‘Marichi Blue’ was 14.7 cm longer on average in high tunnel vs. field. Similar to how high tunnels reduce light and protect flowers from climatic conditions, our results indicate low-density shade in field-grown cut flower production can increase stem length while also maintaining flower quality. Through the use of less-intense shading, the current study increased stem length while not compromising overall flower quality.
Conclusions
Stem length and flower quality are important to growers of all sizes to enter or expand their market footprint. Improving field production by including shadecloth into field production may provide options to increase flower marketability and stem length using low-cost and flexible shade structures for lisianthus. The intensity of shade applied needs to be balanced to allow for stem elongation without reduction in flower yield and quality. This study determined a delay in the cumulative number of stems harvested with the addition of shade, which could be beneficial or detrimental depending on the market and timings of the growers but could be used in succession plantings. Group I ‘Arena’ white was more sensitive to shade treatments (delayed 10 to 15 d with the addition of 50%) and reaffirmed the needs for cultivar- and group-specific trials. Group I had increased stem length without reduced quality for the 30% shadecloth treatments. Although group II was less affected by shading treatments, it also maintained or increased overall quality with the 30% shading treatments, with increased stem length adding to increased marketability for growers. These results emphasize the potential benefits of balancing shade intensity with overall flower quality and the need for further studies with field-grown cut flowers using lower-density shade that have been previously investigated. In addition, further studies, such as plant date and that impact on GDD and DLI should be investigated to determine impacts on stem length and quality for field-grown lisianthus (E. grandiflorum). The timing of flowering will be dependent on temperature conditions.
Received: 18 Aug 2025
Accepted: 01 Oct 2025
Published Online: 05 Nov 2025
Published Print: 01 Dec 2025
Fig. 1.
Average air temperature at 2 m (°C) and calculated daylight integral (DLI) (mol·m−2·d−1) for the growing seasons 2022, 2023, and 2024. Climate data were collected using the on-farm weather station from the University of Georgia Automated Environmental Monitoring Network Information in 15 min intervals, which determined total solar radiation. Total solar radiation was converted to photosynthetic active radiation for DLI calculations.
Fig. 2.
Representative plots directly after harvest for 2023 group I flowers. Jars measure 17 cm in height. (A) 0% shadecloth. (B) 30% shadecloth. (C) 50% shadecloth.
Fig. 3.
Representative plots directly after harvest for 2023 group II flowers. Jars measure 17 cm in height. (A) 0% shadecloth. (B) 30% shadecloth. (C) 50% shadecloth.
Fig. 4.
Cumulative stems harvested and growing degree days (base temperature 14 °C) for group I lisianthus (mean ± standard deviation). Values on the same day after transplant with the same letter are not significantly different from one another (P < 0.05) according to the Tukey-Kramer Least-square means test.
Fig. 5.
Cumulative stems harvested and growing degree days (base temperature 14 °C) for group II lisianthus (mean ± standard deviation). Values on the same day after transplant with the same letter are not significantly different from one another (P < 0.05) according to the Tukey-Kramer Least-square means test.
Fig. 6.
Relationship of growing degree day (GDD), cumulative daylight integral (DLI), and cumulative stems harvested for all years and groups. Shade represented by different symbols (● ■ ▲) for 0%, 30%, and 50% shade, respectively, and years and groups represented by different colors. Data fit to multiple regression using SAS Proc Reg.
ArmitageAM
. 1991. Shade affects yield and stem length of field-grown cut-flower species. HortScience. 26(
. 2007. Lisianthus: Eustoma grandiflorum, p 644–663. In: Anderson NO (ed). Flower breeding and genetics: Issues, challenges and opportunities for the 21st century.
. 2016. Late-season high tunnel planting of specialty cut flowers in the midwestern United States influences yield and stem quality. HortTechnology. 26(
. 2024. Daily light integral, but not the photoperiod, influences the time to flower and finished quality of dianthus specialty cut flowers. HortScience. 59(
Average air temperature at 2 m (°C) and calculated daylight integral (DLI) (mol·m−2·d−1) for the growing seasons 2022, 2023, and 2024. Climate data were collected using the on-farm weather station from the University of Georgia Automated Environmental Monitoring Network Information in 15 min intervals, which determined total solar radiation. Total solar radiation was converted to photosynthetic active radiation for DLI calculations.
Fig. 2.
Representative plots directly after harvest for 2023 group I flowers. Jars measure 17 cm in height. (A) 0% shadecloth. (B) 30% shadecloth. (C) 50% shadecloth.
Fig. 3.
Representative plots directly after harvest for 2023 group II flowers. Jars measure 17 cm in height. (A) 0% shadecloth. (B) 30% shadecloth. (C) 50% shadecloth.
Fig. 4.
Cumulative stems harvested and growing degree days (base temperature 14 °C) for group I lisianthus (mean ± standard deviation). Values on the same day after transplant with the same letter are not significantly different from one another (P < 0.05) according to the Tukey-Kramer Least-square means test.
Fig. 5.
Cumulative stems harvested and growing degree days (base temperature 14 °C) for group II lisianthus (mean ± standard deviation). Values on the same day after transplant with the same letter are not significantly different from one another (P < 0.05) according to the Tukey-Kramer Least-square means test.
Fig. 6.
Relationship of growing degree day (GDD), cumulative daylight integral (DLI), and cumulative stems harvested for all years and groups. Shade represented by different symbols (● ■ ▲) for 0%, 30%, and 50% shade, respectively, and years and groups represented by different colors. Data fit to multiple regression using SAS Proc Reg.
