Performance of Selected Herbaceous Annual Ornamentals Grown at Decreasing Levels of Irrigation

This study was conducted to evaluate the growth, visual quality, and stress response of 17 species of bedding plants andKentucky bluegrass (Poa pratensisL.) grown outdoors for 10 weeks during the summer of 2003 at three locations in Colorado. Plants were irrigated at 100% of the reference evapotranspiration (ET0) (amount required to maintain Kentucky bluegrass in an optimum condition) for 2 weeks followed by 8 weeks at five irrigation levels: 0%, 25%, 50%, 75%, and 100% ET0. Begonia carrieri Hort. Vodka , Lobelia erinus L. Cobalt Blue , and Viola ·wittrockiana Gams. Crown Gold grew well with a minimum of 50% or more ET0 based on Kentucky bluegrass. Impatiens walleriana Hook. fil. Tempo White grew well only with 100% ET0. Antirrhinum majus L. Sonnet Yellow , Dianthus L. First Love , Lobularia maritima (L.) Desv. Carpet White , and Pelargonium ·hortorum L.H. Bailey performed well with 25% to 50% ET0. The species Catharanthus roseus (L.) G. Don Peppermint Cooler , Rudbeckia hirta L. Indian Summer , Senecio cineraria D.C. Silver Dust , Tagetes erecta L. Inca Yellow and T. patula L. Bonanza Gold , Zinnia angustifolia Kunth., and Salvia farinacea Benth. Rhea Blue , which are adapted to midsummer heat and low water, performed well with 0% to 25% ET0. Species considered to be heat or drought tolerant—Petunia ·hybrida hort. ex. E. Vilm. Merlin White and Glandularia J.F. Gmel. Imagination —required little or no irrigation. The bedding plant species evaluated in this study that required 25% or less ET0 are well adapted for low-water landscape installations. Water conservation is becoming an increasingly pertinent issue in Colorado and the intermountain Rocky Mountain west. Colorado has a semiarid to arid climate with recurrent drought periods (Doesken, 2004; Tweit, 2004) and limited water resources (Oad and DiSpigno, 1997). Add to this population growth (Ellefson and Winger, 2004; Oad et al., 1997; Tweit, 2004), which in turn increases the strain on water resources. There is a great deal of information reported in the literature quantifying the water requirements of food crops and turfgrass; however, data quantifying the irrigation requirements of ornamental plants required to maintain healthy growth and acceptable quality is minimal at present. Past reports have classified ornamental plants as low, medium, and high water users based on anecdotal evidence. A California study with groundcovers reported that 50% of evapotranspiration (ET0) (level of irrigation required to maintain Kentucky bluegrass in an optimum condition) was the minimum amount of irrigation to maintain acceptable plant quality (Pittenger et al., 1990). Yet, there was considerable variability among species. A later study indicated that for Vinca L., Gazania Gaertn., and Potentilla L., there was no improvement in aesthetic appearance when irrigated at more than 30% ET0. Baccharis L., Drosanthemum Schwant., and Hedera L. showed no improvement in appearance when grown above 20% ET0 (Pittenger et al., 1992). Other California investigations concluded that irrigation less than or equal to 14% of ET0 can be applied to established groundcovers and shrubs with no adverse affects on health and appearance (Sachs, 1991). Previous work reported from Colorado, where 45 herbaceous perennial plants native to the western United States were subjected to irrigation and no irrigation, demonstrated no differences among irrigation treatments (Cox and Klett, 1984). Staats and Klett (1995) reported that for Kentucky bluegrass (Poa pratensis L. Challenger ), 50% bluegrass ET0 was the optimum irrigation level, whereas Cerastium tomentosum L. required irrigation between 50% and 75% bluegrass ET0 in one season and 25% ET0 during the following season. Potentilla tabernaemontani Asch. required irrigation at 75% ET0 and Sedum acre L. did well at 25% ET0 (Staats and Klett, 1995). Taken together, these findings demonstrate substantial variation in water requirements for ornamental plants grown in Colorado. This study was conducted to determine water requirements and quantify the growth, visual quality, and health of 17 popular species of bedding plants grown outdoors under Colorado conditions compared with Kentucky bluegrass (Poa pratensis L). The objectives of this research were to provide information on minimum irrigation requirements for annual herbaceous ornamental plants to be used by landscape professionals. Materials and Methods The bedding plant species for this study were chosen based on their status as topselling bedding plants in Colorado according to local greenhouse professionals (D. Gerace, Welby Gardens, Inc., Denver, Colo., pers. comm.). Poa pratensis L. (Kentucky bluegrass) was included in the study for comparison because the calculated reference evapotranspiration (ET0) value was based on this species. The 17 herbaceous annual ornamental species tested included Antirrhinummajus L. Sonnet Yellow (snapdragons); Begonia carrieriHort. Vodka [syn. Begonia semperflorens Vodka (red-leaved wax begonias)]; Catharanthus roseus (L.) G. Don Peppermint Cooler (Madagascar periwinkle or vinca);Dianthus L. First Love ; Impatiens walleriana Hook. fil. Tempo White (impatiens); Lobelia erinus L. Cobalt Blue ; Lobularia maritima (L.) Desv. Carpet White (sweet alyssum); zonal geraniums Pelargonium ·hortorum L.H. Bailey Americana Red (Fort Collins), Pelargonium ·hortorum L.H. Bailey Kim Red (Denver), and Pelargonium ·hortorum L.H. Bailey Americana Pink (Grand Junction); Petunia ·hybrida hort. ex. E. Vilm. Merlin White ; Rudbeckia hirta L. Indian Summer (black-eyed Susan); Salvia farinacea Benth. Rhea Blue (mealycup sage or salvia); Senecio cineraria D.C. Silver Dust (dusty miller); Tagetes erecta L. Inca Yellow (African marigold); Tagetes patula L. Bonanza Gold (French marigold); Glandularia Imagination [Glandularia J.F. Gmel. Imagination (syn. Verbena Imagination ) is either a species of Glandularia tenuisecta (Briq.) Small (syn. V. tenuisecta Briq.), moss verbena or a hybrid between Glandularia tenuisecta and Glandularia tenera (Spreng.) Cabera (syn. V. tenera Spreng.), Latin American mock vervain. For clarification, this plant will be referred to as Glandularia in this manuscript.]; Viola ·wittrockianaGams. CrownGold (pansies); and narrow-leaf zinnias Zinnia angustifolia Kunth. Crystal White (Fort Collins and Denver) and Zinnia angustifolia Crystal Orange (Grand Junction). Cultivar differences among locations for Pelargonium and Zinnia were the result of issues with availability. Similar trends were observed for each of the cultivars among locations and the data were consequently pooled. All herbaceous ornamental plants used in the field trials were commercially grown from F1 hybrid seed or cuttings (Welby Gardens, Inc., Denver, Colo.) using a traditional sphagnum peat moss and vermiculite Received for publication 14 Feb. 2006. Accepted for publication 24 Mar. 2006. This work was funded byWelby Gardens, Inc., the Colorado State University Specialty Crops Program, the Colorado Floriculture Foundation, and the Colorado Agr. Expt. Sta. Former Graduate Research Assistant, Extension Specialist, and Associate Professor. To whom reprints requests should be addressed; e-mail [email protected]. HORTSCIENCE VOL. 41(6) OCTOBER 2006 1481 substrate, and were transplanted to plastic 804 trays, 606 trays, or 10-cm pots (zonal geraniums). Initially plants were greenhouse grown and subsequently hardened by limiting water before transplanting into the trial beds. The three locations used included a campus site, an urban site, and a site in western Colorado. The urban location was in Denver, Colo., at elevation 1553 m; latitude, 39.83 ; longitude, –104.95 . The campus location was in Fort Collins, Colo., at elevation 1522 m; latitude, 40.57 ; longitude, –105.09 . The western Colorado location was in Grand Junction, Colo. at elevation 1416 m; latitude, 39.04 ; longitude, –108.54 . Each location required some soil amendment. The Denver location was a clay soil and was amended with 0.018m of sphagnum peat moss m. The Fort Collins location was a loam soil, which had been amended during previous years with various organic amendments. Grand Junction soils are typically very saline; therefore, the Grand Junction location was a raised bed and was filled with a sandy loam native topsoil. The raised bed was 1 m high, with 10-cm-thick cement walls and a bottom layer of gravel covered by landscape fabric. One application of a full-season, complete nutrient (7% N–0.874% P–4.98% K), and combined soluble and slow-release fertilizer (HARDY START; Welby Gardens) was applied at a rate of 122 g m and lightly tilled into each plot before planting at each location. Each bedding plant species was planted in a 1-m square with nine plants spaced 25.4 cm on center. Kentucky bluegrass sod was planted in 61 · 76-cm blocks (Denver and Fort Collins) and 76 · 76-cm blocks (Grand Junction) by placing it on top of the drip tapes and building soil up around the edges. Planting was completed 28 May 2003 in Denver, 29 May 2003 in Fort Collins, and 17 July 2003 in Grand Junction. The Grand Junction location was planted late to accommodate for the travel distance and to schedule the peak bloom, because late summer and early fall temperatures would be more suitable for bedding plant performance. The beds were mulched with 1.25 cm cedar bark in Denver and Fort Collins and 1.25 cm shredded wood from pallets in Grand Junction. All weeding was done by hand at all locations. Insecticides were used only if severe infestations of insects occurred. No dead heading of blooms was done except on zonal geraniums. Differences between mulch colors used at the Denver and Fort Collins locations compared with the Grand Junction location were negligible. The irrigation treatments included five levels, determined as 0%, 25%, 50%, 75% and 100% of the reference evapotranspiration (ET0) required for optimum Kentucky bluegrass growth. Daily weather data (temperature, solar radiation, wind speed, and relative humidity) were collected from local weather stations, which were used to calculate locationspecific ET0 using the Penman–Montieth equation (Qian et al., 1996). Each irrigation zone (irrigation level based on percent ET0) was operated by a separate solenoid valve. Flow meters were installed to measure the actual water volume of irrigation water applied to each zone. Water flow was then calibrated to ensure accuracy among treatments. Two lines of drip tape were installed 25.4 cm apart along the center of all experimental units from end to end of each whole plot. The drip tape was 1.6 cm, 8 mm polyethylene (Chapin Watermatics, Inc., Watertown, N.Y.) with emitters every 20.3 cm and had an output of 203 mL min m. Irrigation was automatically applied to whole plots using a WeatherTRACK ET Controller (HydroPoint Data Systems, Petaluma, Calif.). Daily weather data for the Denver location was collected from an onsite weather station from which ET0 was determined. Daily weather data for the Fort Collins location was collected from a nearby Northern Colorado Water Conservancy weather station located in central Fort Collins from which ET0 was determined. Daily weather data for the Grand Junction location used a network of local weather stations to obtain an ‘‘ET everywhere’’ value. HydroPoint Data Systems distributed these daily ET0 values to each ET Controller via cellular technology. The ET controller automatically adjusted the irrigation output for each irrigation zone (%ET0) accordingly. Each ET controller was programmed for each location’s soil type, type of landscape plant material (flowers), depth of irrigation, slope and exposure, and time of day to begin irrigation. Temperature data from each location are included in Figs. 1 through 3. Each zone was entered into the controller by subtracting out 75% for the 25% zone, 50% for the 50% zone, and 25% for the 75% zone. The 0% ET0 zone received only natural precipitation (no irrigation). The 100% zone was 100% of the reference ET (ET0) for Kentucky bluegrass (i.e., the ET required to grow Kentucky bluegrass to optimum—up to 3.28 cmwater m week). Irrigation applied to each zone as well as natural rainfall is included in Table 1. A rain sensor shutoff valve (Hunter Mini-Click, Cary, N.C.) was connected to the ET controller to initiate an automatic rain pause when 0.32 cm rain was detected. The daily data sent to the ET controller would initiate a rain pause for a calculated number of days after a rain event as well. All plots were irrigated to 100% ET0 for 2 weeks after planting and were then subjected to the irrigation treatment levels for the following 8 weeks. Percent cover, plant biomass, leaf temperature, and visual evaluations were collected to contrast the growth and health of each species at the different levels of irrigation at all three locations. Percent cover using the point intercept method (Daubenmire, 1959) was measured every other week for 8 weeks at each location using a grid constructed from a 1-m frame constructed from polyvinyl chlorine pipe. An acceptable final percent cover was calculated for each species by taking its highest cover, dividing it by 93 (the highest average percent cover achieved by a plant species in the study), and then multiplying that by 75. This formula was used based on 75% being Fig. 3. Average daily temperature with maximum and minimum temperatures for Grand Junction, Colo., collected during the first 2-week establishment period followed by the 8-week study period. The temperature data were collected from the National Oceanic & Atmospheric Administration weather station located at Grand Junction Walker Field, Grand Junction, Colo. Fig. 1. Average daily temperature with maximum and minimum temperatures for Denver, Colo., collected during the first 2-week establishment period followed by the 8-week study period. The temperature data were collected from the National Oceanic & Atmospheric Administration weather station located at Denver International Airport, Commerce City, Colo. Fig. 2. Average daily temperature with maximum and minimum temperatures for Fort Collins, Colo., collected during the first 2-week establishment period followed by the 8-week study period. The temperature data were collected from the Northern Colorado Water Conservancy District weather station located in central