Abstract
Perennial ornamental grasses are often recommended for rain gardens, but few data support their use. We conducted two experiments to evaluate the ability of ornamental grass cultivars to grow while subjected to cyclical flooding, submergence, and drought typical of rain gardens. Our objectives were to determine the effects of cyclical flood and drought (Expt. 1) and submergence depth and duration (Expt. 2) on grass growth and survival. Seven cultivars were evaluated during greenhouse trials, including Pixie Fountain tufted hairgrass [Deschampsia cespitosa (L.) P. Beauv.], Northwind switchgrass (Panicum virgatum L.), Red October big bluestem (Andropogon gerardii Vitman), Purpurascens Chinese silvergrass (Miscanthus sinensis Andersson), Blue Heaven® little bluestem [Schizachyrium scoparium (Michx.) Nash], Blonde Ambition blue grama grass [Bouteloua gracilis (Kunth) Lag. ex Griffiths], and Karl Foerster feather reed grass [Calamagrostis ×acutiflora (Schrad.) DC]. During Expt. 1, grasses underwent four cycles of flooding duration (2 days or 7 days) followed by drought (drying to volumetric soil water contents of 0.14 or 0.07 cm3·cm−3). During Expt. 2, grasses were cyclically submerged at 15 or 30 cm above the soil surface for 2, 4, or 7 days, followed by floodwater removal and drainage for 2 days before being resubmerged. Cyclical submergence continued until the 7-day submergence treatments completed four cycles. Both experiments were replicated in a full factorial randomized complete block design. Controls were included in both experiments. Plants were measured to determine plant height, shoot count, visual damage rating, shoot dry weight, and root dry weight. Floodwater chemistry and soil reducing conditions were measured during Expt. 2. Chinese silvergrass and switchgrass survived cyclical soil flooding/drought and submergence for 7 days at a depth of 30 cm while maintaining acceptable foliar damage. All grasses survived cyclical flood and drought when the soil volumetric water content was maintained at 14%, suggesting they can withstand periodic soil flooding as long as the water is not too deep. As water depth and duration increased from 4 days to 7 days, little bluestem, blue grama grass, and feather reed grass experienced significant foliar damage. Tufted hair grass and big bluestem experienced significant foliar damage when submerged for 2 days. Our results showed that perennial ornamental grasses can tolerate cyclical flood and drought and periodic submergence, but that plant conditions and survival vary, which can inform strategic plant placement within rain gardens, bioretention basins, and other stormwater management systems.
Before 2008, the majority of the world population lived in rural areas. By 2018, 55% of the world population lived in urban areas, and this number is projected to increase to more than 67% by 2050 (United Nations Department of Economic and Social Affairs Population Division 2019). In the United States, the urban population grew 12% between 2000 and 2010, making the United States one of the most urbanized countries in the world, with more than 80% living in urban areas (US Census Bureau 2010). As urban areas expand, so does the area covered by impervious surfaces (i.e., buildings, driveways, roads, and parking lots), making stormwater management a priority. Stormwater runoff entering waterways (e.g., lakes, rivers, and streams) can increase water temperature, bank erosion, flooding, and pollutant levels, thus compromising water quality (US Environmental Protection Agency 2008). However, holding stormwater in the urban landscape by using bioretention structures, such as rain gardens, reduces runoff and pollutant loads entering waterways. Currently, the use of bioretention structures in many urban areas is increasing, but their use and capacity are still often insufficient to convey the excess stormwater (Li et al. 2019; Vijayaraghavan et al. 2021).
Rain gardens are used for commercial and residential development to reduce flooding, increase water infiltration, and improve stormwater quality by removing pollutants (Asleson et al. 2009; Hunt et al. 2008). A rain garden is a shallow basin that is planted with herbaceous and sometimes woody perennial plants, and it is often mulched with shredded wood. Rain gardens collect stormwater from roofs, roads, or parking lots, and they are commonly 20% to 30% of the impervious surface area (Jennings et al. 2015). The ponding depth varies from 15 to 46 cm, depending on the area of impervious surface that drains into the rain garden and soil hydraulic conductivity (Davis et al. 2009; Minnesota Pollution Control Agency 2021). To prevent mosquito breeding, rain gardens are designed so that ponded water will drain within 24 h, and so that soil pore space drainage will occur within 48 to 96 h (Davis et al. 2009; Minnesota Pollution Control Agency 2021).
Rain garden plants rely on precipitation. Depending on the rainfall frequency, water may be in excess or scarce. As the basin fills with stormwater, plants are exposed to partial or complete submergence. Subsequently, additional water is limited until the next stormwater event because rain gardens typically receive minimum or no irrigation. Therefore, desirable plants must tolerate periodic flooding, submergence, and drought.
Perennial ornamental grasses are often recommended for rain gardens; however, few scientific studies support their recommendation, and, often, only the species is recommended (Hausken and Thompson 2018; Meyer 2004, 2012; Steiner and Domm 2012). Several ornamental grass cultivars that offer improved form and foliage coloration when compared with the species are currently available. These ornamental grasses include ‘Pixie Fountain’ tufted hairgrass [Deschampsia cespitosa (L.) P. Beauv.], ‘Northwind’ switchgrass (Panicum virgatum L.), ‘Red October’ big bluestem (Andropogon gerardii Vitman), ‘Purpurascens’ Chinese silvergrass (Miscanthus sinensis Andersson), Blue Heaven® little bluestem [Schizachyrium scoparium (Michx.) Nash], ‘Blonde Ambition’ blue grama grass [Bouteloua gracilis (Kunth) Lag. ex Griffiths], and ‘Karl Foerster’ feather reed grass [Calamagrostis ×acutiflora (Schrad.) DC] (Thetford et al. 2009). Chinese silvergrass is also called Miscanthus sinensis var. purpurascens and is frequently sold in the horticultural trade under this name (Integrated Taxonomic Information System 2022) (see Supplemental 1 for photographs of each ornamental grass).
The available literature that evaluated the flooding, submergence, and drought tolerance of the seven listed species is limited and virtually nonexistent for the cultivars. Yuan and Dunnett (2018) repeatedly flooded Chinese silvergrass to the substrate level for 1 or 4 d, and then they allowed drainage for 4 d. They observed that 1 d of flooding significantly increased the plant height compared with that of the control, but no significant effects on shoot and root dry weight were observed. Barney et al. (2009) flooded four switchgrass cultivars, two from lowland environments Alamo and Kanlow and two from upland environments Cave-In-Rock and Blackwell, for 11 weeks (2–5 cm above the soil surface). All grasses survived continuous flooding. Plant height, shoot count, and shoot and root weights were similar for grasses in the flooded treatment when compared with their respective controls. However, the lowland cultivars had higher values of the measured parameters when compared with the upland cultivars.
Using the wetland indicator status (Lichvar et al. 2016; US Army Corps of Engineers 2022) for the species and cultivar as a guide (Table 1), tufted hairgrass and switchgrass are facultative wetland and facultative, respectively, and they are more likely to tolerate flooded conditions. Big bluestem and little bluestem are facultative upland plants that are likely able to tolerate limited flooding, whereas blue grama grass and feather reed grass are upland plants and less adapted to flooding. Nelson et al. (2018) evaluated seven sedge (Carex L.) species with a wetland indicator status of obligate, facultative wetland, facultative, facultative upland, or upland to cyclical flood and drought stress. Sedges with an intermediate wetland indicator status of facultative upland, facultative, or facultative wetland were best able to manage repeated stress from flooding and drought that simulated rain garden environments.
Ornamental grass species and cultivar, common name, photosynthetic pathway, wetland indicator status, origin of plant material, and origin of cultivar. Wetland indicator status, description, and designation are from Lichvar et al. (2016) and the US Army Corps of Engineers (2022).
Few studies have evaluated the effects of flooding depth and duration on the growth and survival of grasses. Fraser and Karnezis (2005) evaluated 14 wetland species in seven water depths (i.e., ranging from 6 cm below to 6 cm above the soil surface) and found that the total biomass and survival rate decreased with the increasing flooding depth (Fraser and Karnezis 2005). Three species were perennial grasses: blue-joint grass [Calamagrostis canadensis (Michx.) Beauv.], Virginia wildrye (Elymus virginicus L.), and rattlesnake mannagrass [Glyceria canadensis (Michx.) Trin.] (Fraser and Karnezis 2005).
Regarding perennial grass drought tolerance, Stavridou et al. (2019) found no difference in the gravimetrically measured shoot or root biomass of two Chinese silvergrass genotypes grown under 80% soil field capacity and under 15% of soil field capacity. Clifton-Brown and Lewandowski (2000) found similar results when Chinese silvergrass was grown at 16% to 18% soil gravimetric water content (GWC) (0.0 MPa; no drought), 9% soil GWC (moderate drought; approximately −0.5 MPa), and 6% soil GWC (severe drought; approximately −0.8 MPa). The authors noted that no leaf senescence was observed on Chinese silvergrass regardless of soil GWC. Dougherty et al. (2015) found similar gains in plant height and shoot count within each of seven cultivars of Chinese silvergrass Adagio, Autumn Light, Dixieland, Gracillimus, Graziella, Variegatus, and Zebrinus grown for 16 weeks under four soil matric potential treatments ranging from −0.02 MPa to −4.05 MPa. Barney et al. (2009) also evaluated drought stress of four switchgrass cultivars. Grasses maintained at 25% to 30% soil volumetric water content (VWC) (0.0 MPa; saturated) and 5% soil VWC (−4.2 MPa) survived the 11-week experiment. Grasses maintained at 5% soil VWC had reduced plant height, shoot count, and shoot and root weights when compared with plants kept at 25% to 30% soil VWC. Interestingly, there were no differences between the lowland and upland cultivars. Similar results were found by Mann et al. (2013) when shoot and root mass of ‘Alamo’ switchgrass were reduced under drought conditions (≤−1.0 MPa at a depth of 30 cm) when compared with a well-watered control (≥−0.01 MPa). Because C4 plants are more drought-tolerant than C3 plants, the C4 grasses (switchgrass, big bluestem, Chinese silvergrass, little bluestem, and blue grama) should be more tolerant than the C3 grasses (tufted hair grass and feather reed grass) (Taylor et al. 2014). Additionally, tufted hairgrass is a facultative wetland plant and should be less adapted to drought than the other seven grass species that have a wetland indicator status of facultative, facultative upland, or upland.
Regarding the aforementioned seven grass cultivars, no scientific literature has reported the real-world combination of flooding and drought or submergence depth and duration. Therefore, we conducted two experiments to answer our research objectives. Expt. 1 was performed to determine the effects of cyclical flood and drought on the growth and survival of the grasses. Expt. 2 was performed to determine the effects of the submergence depth and duration on the growth and survival of the grasses.
Materials and Methods
Plant material.
For both experiments, grass plugs were purchased from commercial greenhouses (Table 1) and represent four of the five wetland indicator categories in the National Wetland Plant List for the Great Plains region (Lichvar et al. 2016; US Army Corps of Engineers 2022). Grasses were transplanted into 1.07-L [10.7 cm (width) × 8.7 cm (height)] square pots (T.O. Plastic, Clearwater, MN, USA) filled with Pro Mix BRK (Premier Tech Horticulture, Quakertown, PA, USA) containing 45% to 55% sphagnum peatmoss, processed pine bark, perlite, and limestone (to adjust the pH). The potting medium was amended with 5 g of Multicote 14–14–16 (Haifa North America, Savannah, GA, USA) per 1.07-L pot. Grasses were kept in a greenhouse located on the North Dakota State University Campus, Fargo, ND, USA (lat. 46°52′38″N, long. 96°48′18″W), and maintained at a minimum of 21 °C with a 14-h photoperiod until needed for experiments.
Grasses were cut to 25 cm and roots were washed free of potting media and planted into 2.9-L [16.5 cm (width) × 17.8 cm (height)] nursery containers (Meyers Industries, Akron, OH, USA) filled with a mixture of all-purpose play sand, topsoil from a Barnes soil series (fine-loamy, mixed, superactive, frigid Calcic Hapludolls) (Soil Survey Staff 2011), and peatmoss (5:4:1 by volume). Shoot counts at planting for tufted hair grass, switchgrass, big bluestem, Chinese silvergrass, little bluestem, blue grama grass, and feather reed grass were 40, 15, 10, 25, 15, 25, and 20, respectively (Expt. 1), and 40, 10, 7, 12, 15, 25, and 20, respectively (Expt. 2). Grasses were placed in a greenhouse, fertilized every 2 weeks with a water-soluble fertilizer (20N–8.7P–16.6K; JR Peters Inc., Allentown, PA, USA); each pot received 200 mg·L−1 (N), 87 mg·L−1 (P), and 166 mg·L−1 (K) and was allowed to establish for at least 3 months before starting the experiment. During establishment, grasses were treated with a one-time application of 1.0 g FeDTPA (Sprint 330; BASF Corporation, Triangle Park, NC, USA) applied to each container to correct iron deficiency chlorosis. Supplemental heat was provided when temperatures dropped below 18 °C, and the air cooled when temperatures reached 25 °C. Expt. 1 was started on 6 Oct 2019 (run 1; 124 d), 6 Jan 2020 (run 2; 111 d), and 10 Apr 2020 (run 3; 106 d). The average temperature during the study was 24.7 °C. Expt. 2 was started on 20 Dec 2020 (run 1) and 6 Feb 2021 (run 2), with each run lasting 34 d. The average temperature during the study was 24.2 °C.
Preliminary study of drought set points.
A preliminary study using tufted hair grass, switchgrass, Chinese silvergrass, little bluestem, and feather reed grass was conducted to determine drought set points for the cyclical flood and drought study using three soil mixes. Grasses were cut to 15 cm, potting medium was washed from the roots, and plants were potted into 2.8-L [16.5 cm (width) × 17.8 cm (height)] nursery containers (Meyers Industries, Akron, OH, USA). Nursery containers were filled with a mixture of topsoil from a Delamere soil series (coarse-loamy, mixed, superactive, frigid Typic Endoaquolls) (Soil Survey Staff 2005), all-purpose play sand (TCC Materials, Mendota Heights, MN, USA), and peatmoss (Premier Tech Horticulture, Quakertown, PA, USA) mixed 4:5:1 or 1:8.5:0.5 (by volume) or a soil mixture containing topsoil from a Barnes soil series (fine-loamy, mixed, superactive, frigid Calcic Hapludolls) (Soil Survey Staff 2011), all-purpose play sand, and peatmoss mixed 4:5:1 (by volume). Grasses were allowed to establish for 3 months and fertilized every 2 weeks as previously described. After establishment, plants were dried down over 10 d, with soil volumetric water content (VWC) readings taken daily using a handheld GS3 VWC sensor connected to a ProCheck sensor readout storage system (Decagon Devices, Pullman, WA, USA). After 10 d, grasses were watered and dried down for another 10 d, with daily soil VWC readings. Daily visual observations of the foliage were performed to note leaf roll, leaf wilt, and leaf dieback. No visible growth differences among grasses growing in the three soil mixes were noted during the study (data not shown). The soil mixture containing all-purpose play sand, Barnes soil, and peatmoss mixed 5:4:1 (by volume) was selected for both studies because it is similar to a well-draining rain garden mix recommended by the Minnesota Pollution Control Agency (2021) with 50% to 65% coarse sand, 25% to 35% topsoil, and 10% to 15% compost. Based on the preliminary study, visual plant damage did not occur until the soil VWC was less than 0.14 m3·m−3 (no visible plant damage), whereas severe visual plant damage occurred at 0.07 m3·m−3 (leaf wilt and leaf dieback) (Fig. 1). Based on preliminary results, the drought set points of 0.14 m3·m−3 (drought onset) and 0.07 m3·m−3 (severe drought) were selected.
Typical plant damage rating (1 = no plant damage; 2 = beginning leaf roll; 3 = leaf roll or leaf roll and leaf wilt; 4 = leaf dieback) for a given substrate volumetric water content (VWC). Visual damage ratings and VWC were obtained daily from 9 to 29 Sep 2019 from perennial ornamental grasses growing in a greenhouse at the North Dakota State University campus, Fargo, ND, USA.
Citation: HortScience 59, 8; 10.21273/HORTSCI17800-24
Sensor calibration.
A water retention curve was developed for the soil mix containing all-purpose play sand, Barnes soil, and peatmoss mixed 5:4:1 (by volume) as described by Nelson et al. (2018) (Fig. 2). Briefly, soil was placed into pressure chambers, pressure plates, and a dew point potentiometer to determine soil moisture between matric potentials of −10 to −300, −500 to −1500, and <−1500 kPa, respectively. Once removed from pressure chambers and plates, soil was weighed and oven-dried for 48 h at 105 °C. After drying, the soil was weighed and the GWC was determined by subtracting the soil wet weight from the soil dry weight and dividing by the dry weight. The GWC was converted to VWC by multiplying the weight by the bulk density of the soil mix (i.e., 1.25 g·cm−3). To obtain the bulk density, three containers of known volume were filled with soil mix, placed in the greenhouse, and watered for 60 d. After 60 d, the soil was weighed and dried at 105 °C for 48 h and reweighed. The mean soil bulk density was determined by dividing the dry weight of the soil mix by the volume of the soil mix for the three samples. Dewpoint potentiometer (WP4C; Decagon Devices Inc., Pullman, WA, USA) readings were performed in a room with a 20 °C constant temperature, and the GWC and VWC of soil samples were determined as previously described. The measured VWC data were fitted to the van Genuchten (1980) model. The particle density values of the soil mix comprising sand, Barnes soil, and peatmoss were 2.64, 2.68, 2.62, and 1.58 g·cm−3, respectively. The particle density was determined by the pycnometer method, as described by Blake and Hartge (1986).
Water retention curve of a soil mixture containing all-purpose play sand, Barnes soil, and peatmoss mixed 5:4:1 (by volume).
Citation: HortScience 59, 8; 10.21273/HORTSCI17800-24
Soil and media mix.
The Barnes soil used for both experiments was collected from a commercial farm field located in Fergus Falls, MN, USA, on 22 May 2019, that was planted with soybean [Glycine max (L.) Merr.] and managed without residual herbicides during the previous year. Soil was collected from the 0- to 15-cm profile, screened through a 9.5-mm sieve to remove large clods, and placed into 114-L plastic containers. The soil was dried in a greenhouse by spreading soil over a tarp laid over the greenhouse floor. Soil was raked every 2 d to facilitate drying. Soil was air-dried for 18 d, and three soil samples were randomly collected and sent to Agvise Laboratories, Northwood, ND, USA, to analyze the pH, organic matter, electrical conductivity (EC), NO3, Olsen soil test P, K, Ca, Mg, Zn, S, Cl, Cu, Fe, Mn, B, Na, CaCO3, and cation exchange capacity (Table 2). Dry soil was screened through 6.4-mm sieve to remove any remaining large clods and placed into 114-L plastic containers until needed for the experiment. Sand, Barnes soil, and peatmoss (5:4:1 by volume) were added to a cement mixer (38-L volume) and allowed to mix for 5 m. After mixing, soil was placed into 11-L plastic containers until needed for the experiment. Three soil samples were collected from mixed soil and sent to Agvise Laboratories for the same analysis as that previously described (Table 2).
Soil chemical properties of Barnes soil and mix containing 50% sand, 40% Barnes soil, and 10% peatmoss (by volume).
Expt. 1.
Grasses were provided with a 12-h photoperiod in the greenhouse with supplemental lighting using 400-W high-pressure sodium lights (P.L. Light Systems, Beamsville, ON, Canada) that produced ≈139 μmol·m−2·s−1 irradiance. Grasses were flooded for 2 d or 7 d and allowed to dry down to one of two soil VWC set points for a total of four treatments plus a well-watered control. Control plants were watered as needed to maintain a soil VWC higher than 0.14 m3·m−3. The soil VWC set points of 0.14 m3·m−3 (drought onset) and 0.07 m3·m−3 (severe drought) had soil matric potentials of −40 kPa and −2500 kPa, respectively, and were selected based on data from the preliminary study. Flooding for 2 d represented a functioning rain garden, whereas flooding for 7 d represented a poorly drained rain garden. Flooding treatments were performed by placing the 2.9-L container with a plant into another 2.9-L container lined with a 26.8- × 27.3-cm plastic bag (SC Johnson, Racine, WI, USA).
Tap water was added to maintain a 1.25-cm layer of water on the soil surface. Water was added daily as needed to maintain the depth. After the flood duration was met, grasses were removed from flooding and allowed to freely drain for 24 h. After 24 h, soil VWC readings were performed daily until the respective drought set point was reached. When the drought set point was reached, the flood cycle was repeated. A grass species was removed from the trial when the 7-d flood and 0.07 m3·m−3 severe drought set point went through four cycles of flood and drought. Individual plants were removed from the trial if no living tissue was visible. After removal from the study, the following data of each plant were collected: plant height, shoot count, visual damage rating, shoot dry weight, and root dry weight. Plant height was measured from the soil surface to the height of the highest living leaf (grasses were pulled straight and then measured). Shoot counts were performed by counting all living shoots that were at least 1.25 cm above the soil surface. A visual damage rating was assigned to each plant using a scale of 1 to 10 (1 = 0% to 10% dieback; 4 = 31% to 40% dieback; 7 = 61% to 70% dieback; and 10 = 91% to 100% dieback). Shoots were cut off at the soil line, and roots were washed free of soil, placed into separate paper bags, and set in a 65 °C dryer for 96 h. After drying, the dry weights of shoots and roots were determined using an electronic balance (LP6200S; Sartorius AG, Gottingen, Germany).
Expt. 2.
Grasses were provided with a 16-h photoperiod in the greenhouse with supplemental lighting using 40-W fluorescent lights (Signify North America Corporation, Somerset, NJ, USA) that produced ≈13 μmol·m−2·s−1 irradiance. The slight differences in the photoperiod and irradiance as compared with those of Expt. 1 were attributable to greenhouse space limitations that required Expt. 2 to be performed in a different greenhouse. Because the light intensity in the greenhouse used during Expt. 2 was lower than that during Expt. 1, the photoperiod was extended to provide the plants with a longer duration of light. Grasses were submerged in tap water at 15 cm or 30 cm above the soil surface for 2, 4, or 7 d. After the submergence duration was met, grasses were removed and allowed to drain for 2 d; then, submergence was repeated. The experiment was ended after the 7-d submergence duration treatments went through four cycles. Individual plants were removed from the trial if no living tissue was visible. Control plants were watered as needed to maintain a soil VWC higher than 0.14 m3·m−3. The same datapoints of Expt. 1 were collected during Expt. 2.
The soil VWC of the control plants was periodically monitored using a GS3 sensor as described previously. Submergence treatments were conducted by placing the 2.9-L container with grass into a 26.5-L pail (ULINE, Pleasant Prairie, WI, USA). Tap water was added to bring the water depth to 15 cm or 30 cm above the soil surface. Water was added as needed every 2 d to maintain the respective depth. Starting 24 h after submergence, water temperature, EC, and dissolved oxygen were measured daily, and water pH, nitrate, ammonium, and chloride levels of all submergence treatments were measured periodically. All water measurements were taken ≈7.6 cm above the soil using a HI9829 multiparameter probe (Hanna Instruments, Smithfield, RI, USA) fit with sensors to determine the previously described parameters.
Reducing soil conditions (i.e., substrate oxygen status) were measured in one replicate of each run by placing a section of IRIS (indicator of reducing conditions in soil) tube (InMass Technologies, West Lafayette, IN, USA) measuring ≈15.25 cm × 2.2 cm in each container. Briefly, IRIS tubes were made by lightly sanding a PVC pipe and applying iron oxide paint. Under reducing conditions, Fe(III) was converted to Fe(II) and removed from the IRIS tube (Castenson and Rabenhorst 2006; Rabenhorst 2008). Before placement, a soil core measuring ≈15 cm × 2.54 cm was removed from each pot, allowing for placement of the IRIS tube. The IRIS tube was located ≈3.75 cm from the edge of the container. After placement, containers were watered to insure good soil contact with the tubes. At the end of the study, IRIS tubes were gently pulled from containers, washed clean of soil using tap water and a soft bristle brush, and allowed to air-dry. After drying, IRIS tubes were photographed (Nikon D5000 digital camera with a 18- to 55-mm lens; Melville, NY, USA) with a black background, rotated 180 degrees, and photographed again. The amount of iron removal from each tube was calculated using ImageJ Software (Schneider et al. 2012). Each photograph was analyzed to determine the total pixel area of the IRIS tube and the total pixel area of the removed iron. For each tube, the total pixel areas of iron removed from both photographs were added together and divided by the sum of the total pixel area of the IRIS tube to determine the percentage of iron removal. Each IRIS tube was given a rating using a scale of 1 to 10, where each integer represented 10% additional iron removal (i.e., 1 = 0% to 10% removal, 2 = 11% to 20% removal, …, and 10 = 91% to 100% removal).
Experimental design and statistical analysis.
The experiments were arranged as a randomized complete block design with a 7 × 2 × 2 factorial arrangement consisting of seven species, two flood durations, and two drought set points with three single plant replicates (Expt. 1), and with a 7 × 3 × 2 factorial arrangement consisting of seven species, three submergence durations, and two submergence depths with three single plant replicates (Expt. 2). Expt. 1 was conducted three times (i.e., runs) and Expt. 2 was conducted two times (i.e., nine and six total replicates, respectively). Error mean square values were within a factor of 10 among runs. Therefore, the variance of each run was considered similar, and those data were pooled for both experiments. All data were expressed as a percent of the control, except for the visual damage rating and iron removal from IRIS tubes, and subjected to a mixed linear model analysis of variance (Proc MIXED; SAS 9.4; SAS Institute, Cary, NC, USA). Root mass data were square-root-transformed before the analysis to standardize the variance and back-transformed for data presentation (Expt. 1). The experimental run and repetition were considered random effects. Species, flood, and drought (Expt. 1) and species, submergence duration, and submergence depth (Expt. 2) were used as fixed effects. The Tukey-Kramer honestly significant difference test was used to separate treatment means. Means were considered significant when P < 0.05.
Results
Expt. 1. Relative shoot and root characteristics.
A significant species × drought set point interaction occurred for relative shoot height, count, mass, and relative root mass (Figs. 3–5). For these plant characteristics, the drought set point treatments caused between >30% in relative gains and >90% in relative losses to the control. Across all grass species, the 7% soil VWC drought set point resulted in a lower relative shoot height compared with the 14% soil VWC drought set point, but the difference was only significant for switchgrass, Chinese silvergrass, blue grama grass, and feather reed grass (see Supplemental 2 for example photographs). No significant difference between the 7% and 14% soil VWC drought set points for tufted hairgrass, big bluestem, and little bluestem were observed. Within a grass species, except for big bluestem (shoot count) and little bluestem (shoot mass), the 7% soil VWC drought set point resulted in a significantly lower relative shoot count and mass when compared with the 14% soil VWC drought set point (Figs. 3 and 4). Although the 7% soil VWC drought set point had a lower relative shoot count compared with that of the 14% soil VWC drought set point for big bluestem, the difference was not significant. The 14% soil VWC drought set point treatment resulted in a positive relative shoot count for switchgrass, Chinese silvergrass, and little bluestem, whereas all other grasses had a negative relative shoot count. Although the 7% soil VWC drought set point had a lower relative shoot mass than that of the 14% soil VWC set point for little bluestem, the difference was not significant. For tufted hair grass, Chinese silvergrass, blue grama grass, and feather reed grass, the 14% soil VWC drought set point had a positive relative shoot mass, but the other grasses had a negative relative shoot mass. Within a grass species, except for big bluestem, little bluestem, and blue grama grass, the 7% soil VWC drought set point had significantly less relative root mass when compared with that of the 14% soil VWC drought set point (Fig. 5). For big bluestem, little bluestem, and blue grama grass, the 7% soil VWC drought set point had lower relative root mass compared with that of the 14% soil VWC drought set point, but the difference was not significant.
Relative shoot count as a percentage of the control for seven perennial ornamental grass species subjected to cyclical flood and drought periods. Mean values are averaged across flood duration treatments. Grasses were flooded for 2 d or 7 d and allowed to dry down to 0.07 or 0.14 m3·m−3 substrate volumetric water content (VWC). Mean (n = 18) values labelled with different lowercase letters were significantly different according to the Tukey-Kramer honestly significant difference test at P < 0.05. The wetland indicator categories are as follows: FACW = facultative wetland; FAC = facultative; FACU = facultative upland; and UPL = upland.
Citation: HortScience 59, 8; 10.21273/HORTSCI17800-24
Relative shoot mass as a percentage of the control for seven perennial ornamental grass species subjected to cyclical flood and drought periods. Mean values are averaged across flood duration treatments. Grasses were flooded for 2 d or 7 d and allowed to dry down to 0.07 or 0.14 m3·m−3 substrate volumetric water content (VWC). Mean (n = 18) values labelled with different lowercase letters were significantly different according to the Tukey-Kramer honestly significant difference test at P < 0.05. The wetland indicator categories are as follows: FACW = facultative wetland; FAC = facultative; FACU = facultative upland; and UPL = upland.
Citation: HortScience 59, 8; 10.21273/HORTSCI17800-24
Relative root mass as a percentage of the control for seven perennial ornamental grass species subjected to cyclical flood and drought periods. Mean values are averaged across flood duration treatments. Grasses were flooded for 2 d or 7 d and allowed to dry down to 0.07 or 0.14 m3·m−3 substrate volumetric water content (VWC). Mean (n = 18) values labelled with different lowercase letters were significantly different according to the Tukey-Kramer honestly significant difference test at P < 0.05. The wetland indicator categories are as follows: FACW = facultative wetland; FAC = facultative; FACU = facultative upland; and UPL = upland.
Citation: HortScience 59, 8; 10.21273/HORTSCI17800-24
In addition to the species × drought interactions, a flood main effect occurred for relative shoot count and root mass. The 7-d flood duration resulted in a significantly lower relative shoot count and root mass compared with the 2-d flood duration. The relative shoot count was reduced more than 20% and almost 30% with the 2-d and 7-d flood durations, respectively. Relative root mass was reduced by more than 50% and almost 60% with the 2-d and 7-d flood durations, respectively.
Expt. 1. Relative total biomass, visual damage rating, and survival.
A three-way interaction occurred for relative total biomass among species, flood duration, and drought set point. For tufted hair grass, Chinese silvergrass, and blue grama grass, the 7% soil VWC drought set point had significantly lower relative total biomass when compared with that of the 14% soil VWC drought set point, regardless of flood duration (see Supplemental 2 for example photographs). For switchgrass, there was no significant difference between the 7% and 14% soil VWC drought set points for the 2-d flood duration. However, these did differ with the 7-d flood duration, with the 7% soil VWC drought set point having significantly less relative total biomass compared with that of the 14% soil VWC drought set point. For big bluestem and feather reed grass, the 7% soil VWC drought set point had significantly less relative total biomass when compared with that of the 14% soil VWC drought set point for the 2-d flood duration. However, there was no significant difference between the 7% and 14% soil VWC drought set points for big bluestem and feather reed grass for the 7-d flood duration. For little bluestem, there was no significant differences among treatments.
The main effects of species, flood duration, and drought set point were significant to the visual damage rating. The visual damage rating was significantly higher for tufted hair grass when compared with that of all other grass species except for blue grama grass (Fig. 6). Switchgrass, big bluestem, Chinese silvergrass, and little bluestem had significantly less visual damage when compared with that of all other grasses except for feather reed grass. The 7-d flood duration resulted in a significantly higher visual damage rating (≈4.75) when compared with that of the 2-d flood duration (4.0). The 7% VWC drought set point had a significantly higher visual damage rating (≈ 7.0) when compared with that of the 14% VWC drought set point (∼1.8).
Visual damage rating (scale of 1 to 10: 1 = 0% to 10% dieback; 2 = 11% to 20% dieback; 10 = 91% to 100% dieback) of perennial ornamental grasses subjected to cyclical flood and drought periods. Mean values are averaged over flood duration and drought set point treatments. Grasses were flooded for 2 d or 7 d and allowed to dry down to 0.07 or 0.14 m3·m−3 substrate volumetric water content (VWC). Mean (n = 36) values labelled with different lowercase letters were significantly different according to the Tukey-Kramer honestly significant difference test at P < 0.05.
Citation: HortScience 59, 8; 10.21273/HORTSCI17800-24
All plants survived the 14% soil VWC drought set point regardless of the flood duration. The 7% soil VWC drought set point resulted in a few blue grama grass, tufted hair grass, feather reed grass, and little bluestem plants dying (Table 3).
Cyclical flood and drought experiment treatments with at least one plant dying during the experiment conducted in Fargo, ND, USA, from Oct 2019 to Jul 2020 (top). The experiment was repeated three times during the stated period. Grass species listed had nine plants per treatment. A total of 11 of 252 plants died during the experiment. Submergence experiment treatments with at least one plant dying during the experiment conducted at Fargo, ND, USA, from Dec 2020 to Feb 2021 (bottom). Grass species listed had six plants per treatment. The experiment was repeated two times during the stated period. A total of 26 of 252 plants died during the experiment (bottom).
Expt. 2. Floodwater chemistry and substrate oxygen status.
The average temperature of floodwater during the experiment was 21.8 °C. As the submergence duration increased from 2-d to 7-d, the dissolved oxygen and pH of floodwater decreased for all species regardless of the water depth. The decreases in dissolved oxygen and pH were most pronounced for grasses submerged for 7 d and ranged from ∼11.0 mg·L−1 and 9.1 pH at the beginning of the submergence duration to 0.5 mg·L−1 and 6.7 pH at the end of the submergence duration. The greatest decrease in dissolved oxygen occurred for blue grama grass submerged for 7 d at a depth of 30 cm (7.0 mg·L−1 to 1.8 mg·L−1), whereas tufted hair grass submerged for 7 d at a depth of 15 cm lost the least amount of dissolved oxygen during submergence (5.7 mg·L−1 to 2.9 mg·L−1). Little bluestem submerged for 7 d at a depth of 30 cm had the greatest reduction in pH from day 1 to day 7 (8.7 to 7.5), whereas tufted hairgrass submerged for 7 d at a depth of 15 cm had the least change in pH from day 1 to day 7 (7.9 to 7.2). The EC increased as the submergence duration increased from 2 d to 7 d for all species. The increase in EC was most pronounced for grasses submerged for 7 d and ranged from 0.3 dS·m−1 at the beginning of the submergence duration to 0.7 dS·m−1 at the end of the submergence duration. Nitrate, ammonium, and chloride in the floodwater remained low for all grasses, with average values of 1.2 ± 1.1, 0.8 ± 0.5, and 25.8 ± 5.6 mg·L−1, respectively.
The main effects of species and submergence duration were significant to the substrate oxygen status, as measured by iron removal from the IRIS tubes. Feather reed grass had significantly higher iron removal (≈60%) from IRIS tubes compared with that of all other grasses except for little bluestem (≈50%). Chinese silvergrass had significantly lower iron removal from IRIS tubes (30%) compared with little bluestem and feather reed grass, but not with big bluestem (≈40%), switchgrass (≈40%), tufted hair grass (≈35%) and blue grama grass (35%). The 2-d submergence duration had significantly less iron removal from IRIS tubes (≈30%) compared with that of the 4-d submergence duration (≈40%) and 7-d (≈50%) submergence duration. No significant difference occurred between the 4-d and 7-d submergence duration treatments (see Supplemental 3 for example photographs).
Expt. 2. Relative shoot and root characteristics.
A significant species × submergence duration interaction occurred for relative shoot height and count. Within a grass species, there were no significant differences among the 2-, 4-, or 7-d submergence durations except for tufted hair grass and feather reed grass (Fig. 7). For tufted hair grass and feather reed grass, the 2-d submergence duration had significantly higher relative shoot height when compared with that of the 7-d submergence duration, but not that of the 4-d submergence duration. All grasses except for switchgrass had reduced shoot height with some or all submergence durations relative to the respective control. For switchgrass, all submergence durations had increased shoot height relative to the respective control. Increased shoot height relative to the respective control also occurred with the 7-d submergence duration of Chinese silvergrass and the 2-d submergence duration of feather reed grass.
Relative shoot height as a percentage of the control for seven perennial ornamental grass species subjected to repeated submergence. Mean values are averaged across submergence depth treatments. Grasses were submerged in tap water at depths of 15 or 30 cm above the soil surface for 2, 4, or 7 d. Mean (n = 12) values labelled with different lowercase letters were significantly different according to the Tukey-Kramer honestly significant difference test at P < 0.05. The wetland indicator categories are as follows: FACW = facultative wetland; FAC = facultative; FACU = facultative upland; and UPL = upland.
Citation: HortScience 59, 8; 10.21273/HORTSCI17800-24
For tufted hair grass and blue grama grass, the 2-d submergence duration had significantly higher relative shoot counts when compared with those of the 7-d submergence duration, but not that of the 4-d submergence duration (Fig. 8). For little bluestem, the 2-d submergence duration had a significantly higher relative shoot count when compared with that of the 4-d and 7-d submergence durations, whereas no difference was observed between submergence for 4 d or 7 d. For feather reed grass, the 2-d and 4-d submergence durations had significantly higher relative shoot counts when compared with those of the 7-d submergence duration. Although the 2-d submergence duration for feather reed grass had a higher relative shoot count compared with that of the 4-d submergence duration, the difference was not significant. Relative shoot counts for switchgrass, big bluestem, and Chinese silvergrass were not significantly affected by submergence duration. All grasses had a lower shoot count relative to that of the respective control, except for the 2-d and 4-d submergence durations for switchgrass, which were higher than that of the respective control.
Relative shoot count as a percentage of the control for seven perennial ornamental grass species subjected to repeated submergence. Mean values are averaged across submergence depth treatments. Grasses were submerged in tap water at depths of 15 or 30 cm above the soil surface for 2, 4, or 7 d. Mean (n = 12) values labelled with different lowercase letters were significantly different according to the Tukey-Kramer honestly significant difference test at P < 0.05. The wetland indicator categories are as follows: FACW = facultative wetland; FAC = facultative; FACU = facultative upland; and UPL = upland.
Citation: HortScience 59, 8; 10.21273/HORTSCI17800-24
A significant submergence duration × submergence depth interaction and species main effect occurred for relative shoot mass. The interaction occurred with the 4-d submergence duration, with the 15-cm submergence depth having ≈20% more relative shoot mass compared with that of the 30-cm submergence depth. The 15-cm submergence depth with a 4-d submergence duration was the only treatment with a positive relative shoot mass (≈2%). Submergence durations of 2 d and 7 d had similar reductions in relative shoot mass (≈4% and 10%, respectively), regardless of the submergence depth. Switchgrass had significantly higher relative shoot mass compared with that of all other grass species except little bluestem and feather reed grass (Fig. 9). Switchgrass, little bluestem, and feather reed grass were the only grasses with a positive shoot mass relative to that the control. Chinese silvergrass and tufted hair grass had the lowest relative shoot mass compared with that of all other grass species except for blue grama grass and big bluestem.
Relative shoot mass and root mass as a percentage of the control for seven perennial ornamental grass species subjected to repeated submergence. Mean values are averaged across submergence duration and depth treatments. Grasses were submerged in tap water at depths of 15 or 30 cm above the soil surface for 2, 4, or 7 d. Mean (n = 36) values labelled with different lowercase letters within shoot mass or root mass were significantly different according to the Tukey-Kramer honestly significant difference test at P < 0.05. The wetland indicator categories are as follows: FACW = facultative wetland; FAC = facultative; FACU = facultative upland; and UPL = upland.
Citation: HortScience 59, 8; 10.21273/HORTSCI17800-24
The main effects of species and submergence duration were significant for relative root mass. Little bluestem had significantly higher relative root mass compared with that of all other grasses except big bluestem (Fig. 9). Little bluestem was the only grass species with a positive relative root mass. Tufted hair grass had significantly lower relative root mass when compared with that of all other grass species except feather reed grass. Relative root mass was similar among switchgrass, big bluestem, Chinese silvergrass, blue grama grass, and feather reed grass, except feather reed grass had a significantly lower relative root mass when compared with that of big bluestem. Submergence durations of 2, 4, and 7 d resulted in root mass loss >10%, >20%, and >25%, respectively, relative to that of the controls. The 7-d submergence duration had significantly less relative root mass compared with the 2-d, but not the 4-d, submergence duration. No significant difference in relative root mass was observed between the 2-d and 4-d submergence durations.
Expt. 2. Relative total biomass, visual damage rating, and survival.
The main effects of species and submergence duration were significant to the relative total biomass. Little bluestem had an increase in relative total biomass (>5%), whereas all other grasses were reduced from >5% to almost 40% compared with the respective control. The relative total biomass was significantly higher for little bluestem compared with that of all other grass species except for switchgrass and big bluestem. Tufted hair grass had the lowest relative total biomass (almost 40%), and it was significantly lower compared with that of all other grass species except for Chinese silvergrass. No significant differences were observed among feather reed grass, blue grama grass, Chinese silvergrass, big bluestem, and switchgrass. Relative total biomass was reduced by 10%, 18%, and >20% for the 2-, 4-, and 7-d submergence durations compared with that of the respective controls. The 2-d submergence duration was significantly higher compared with that of the 7-d, but not the 4-d, submergence duration for relative total biomass.
The visual damage rating had several significant two-way interactions. These included a species × submergence duration interaction, a species × submergence depth interaction, and a submergence duration × submergence depth interaction. For little bluestem, blue grama grass, and feather reed grass, the 2-d submergence duration had a significantly lower visual damage rating when compared with that of the 7-d submergence duration, but not that of the 4-d submergence duration, within each grass species (Fig. 10). Within each species, no significant differences were observed among submergence durations for tufted hair grass, switchgrass, big bluestem, and Chinese silvergrass. All submergence duration treatments for switchgrass and Chinese silvergrass had significantly lower visual damage ratings when compared with those of tufted hair grass and big bluestem submergence durations (see Supplemental 4 for example photographs).
Visual damage rating (scale of 1 to 10: 1 = 0% to 10% dieback; 2 = 11% to 20% dieback; 10 = 91% to 100% dieback) of seven perennial ornamental grass species subjected to repeated submergence. Mean values are averaged across submergence depth treatments. Grasses were submerged in tap water at depths of 15 or 30 cm above the soil surface for 2, 4, or 7 d. Mean (n = 12) values labelled with different lowercase letters were significantly different according to the Tukey-Kramer honestly significant difference test at P < 0.05. The wetland indicator categories are as follows: FACW = facultative wetland; FAC = facultative; FACU = facultative upland; and UPL = upland.
Citation: HortScience 59, 8; 10.21273/HORTSCI17800-24
The species × submergence depth interaction occurred for tufted hair grass, with the 15-cm submergence depth having a significantly lower visual damage rating when compared with that of the 30-cm submergence depth (Fig. 11). For all other grass species, there was no significant difference between the submergence depths of 15 cm and 30 cm. Switchgrass submerged at depths of 15 cm and 30 cm had significantly lower visual damage ratings compared with those of all other grass species submerged at the same depths, except for Chinese silvergrass, blue grama grass, and feather reed grass (see Supplemental 4 for example photographs).
Visual damage rating (scale of 1 to 10: 1 = 0% to 10% dieback; 2 = 11% to 20% dieback; 10 = 91% to 100% dieback) of seven perennial ornamental grass species subjected to repeated submergence. Mean values are averaged across submergence duration treatments. Grasses were submerged in tap water at depths of 15 or 30 cm above the soil surface for 2, 4, or 7 d. Mean (n = 18) values labelled with different lowercase letters were significantly different according to the Tukey-Kramer honestly significant difference test at P < 0.05. The wetland indicator categories are as follows: FACW = facultative wetland; FAC = facultative; FACU = facultative upland; and UPL = upland.
Citation: HortScience 59, 8; 10.21273/HORTSCI17800-24
Visual damage ratings were affected by the submergence duration and submergence depth. For all submergence durations, visual damage ratings were significantly lower for the 15-cm depth compared with those of the 30 cm depth, except for the 2-d submergence duration, for which there was no difference between depths. The 2-d submergence duration at a 15-cm depth had a significantly lower visual damage rating (≈3.5) compared with that of the 7-d submergence duration at a depth of 15 cm (almost 7), but not when the submergence duration was 4 d at a depth of 15 cm (almost 4). When submerged at a depth of 30 cm, visual damage ratings were significantly lower for the 2-d submergence duration (≈3.5) compared with those of the 4-d (≈5.5) and 7-d submergence durations (almost 7). Additionally, the 4-d submergence duration at a 30-cm depth was significantly lower than that of the 7-d submergence duration at the same depth.
Grass survival was dependent on species, with all grasses surviving the 2-d submergence duration regardless of depth, with the exception of tufted hair grass (Table 3). Tufted hair grass lost one out of six plants when submerged for 2 d at depths of 15 cm and 30 cm. As submergence duration and depth increased, the number of dying tufted hair grass plants also increased (Table 3; Supplemental 4). A total of 39% of the tufted hair grass plants died during the submergence experiment. A few big bluestem (14 plants or 19%), feather reed grass (3 plants or 8%), and little bluestem (2 plants or 6%) plants also died during the experiment (Table 3).
Discussion
Perennial ornamental grasses are often recommended for rain gardens, but there are few data supporting their recommendation. These two experiments are the first to evaluate ornamental grass cultivars regarding their ability to grow while being subjected to cyclical flooding and drought and periodic submergence, which are conditions typically found in a rain garden. During Expt. 1, severe drought (7% soil VWC) reduced the relative shoot height, shoot count, shoot mass, and root mass relative to their respective well-watered controls for all grasses, with the exception of the shoot height for tufted hair grass. Drought was expected to reduce plant growth parameters because drought stress reduces cell elongation and division, resulting in reduced growth (Farooq et al. 2012).
Plant growth is only one measurement of environmental stress. Visual damage ratings, which are a function of plant dieback percentages, help complete the picture. Tufted hair grass had the highest visual damage rating (≈5.5; i.e., 50% dieback), which was consistent with the lower growth parameters observed during the experiment. Severe drought negatively impacted the shoot count and shoot mass of tufted hair grass. Tufted hair grass had the lowest shoot count, but it was not significantly different from that of blue grama grass and feather reed grass. Similarly, tufted hair grass had the lowest root mass under severe drought compared with that of all other grasses, and three plants died during the experiment. The negative response to severe drought was expected because this species is a facultative wetland plant that usually occurs in wetlands (Lichvar et al. 2016; US Army Corps of Engineers 2022). Blue grama grass had a visual damage rating (≈5.2) similar to that of tufted hair grass and feather reed grass (≈4.3). The relative root mass of blue grama grass was reduced >45% and 70% under mild and severe drought, respectively. It is worth noting that no difference in relative root mass reduction occurred between the mild and severe drought set points. Blue grama grass suffered the highest plant mortality, with five plants (i.e., 7%) dying during the cyclical flooding and drought experiment. The high visual damage rating, lack of significant difference between drought set points for relative root mass across flood duration, and high plant mortality can be partially explained by the wetland indicator status of blue grama grass. Blue grama grass is an upland plant (Lichvar et al. 2016; US Army Corps of Engineers 2022) that is able to withstand drought, but it is not adapted to flooding (Wynia 2007), and the Blonde Ambition cultivar originated in New Mexico (Table 1).
The lowest visual damage ratings (≤4.3; i.e., 33% dieback) were observed for switchgrass, big bluestem, Chinese silvergrass, little bluestem, and feather reed grass; however, that of feather reed grass was not significantly lower when compared with that of blue grama grass. Switchgrass, big bluestem, and little bluestem are native to the northcentral United States (US Department of Agriculture, Natural Resources Conservation Service 2024). All switchgrass and big bluestem plants survived the experiment, whereas only two little bluestem plants died. Under mild drought conditions, switchgrass and little bluestem continued to grow because both grasses had a positive relative shoot count regardless of flood duration. The adaptability of switchgrass to flooding and drought conditions was not surprising because switchgrass is adapted to several soil types and conditions, from sandy to clay loam soils, with soil water contents from dry to fully saturated (US Department of Agriculture, Natural Resources Conservation Service 2006), and because the Northwind cultivar originated in Illinois (Table 1). Switchgrass is also a facultative wetland plant, meaning that it is equally likely to be found in a wetland or nonwetland setting (Lichvar et al. 2016; US Army Corps of Engineers 2022). Although the relative root mass was reduced for all grasses under mild and severe drought set points, it is worth noting that within a species, the difference was not significant for big bluestem and little bluestem. The relative root mass was reduced between 45% and 65% for both grasses when compared with that of the respective control. The lack of significant differences between drought set points for these two grasses can be partially explained based on their native habitat. Big bluestem is a tallgrass prairie species widely distributed in the midwestern United States that is known to tolerate annual rainfall ranging from 58 to 116 cm (Gray et al. 2014). Little bluestem is commonly found on dry upland sites such as ridges, and it is known to tolerate sandy to clay-loam textured soils (Tober and Jensen 2013), and the Blue Heaven cultivar originated in Minnesota (Table 1). Both big bluestem and little bluestem are facultative upland plants, meaning that they are usually found in nonwetland environments (Lichvar et al. 2016; US Army Corps of Engineers 2022). In addition to having reduced relative root mass, both big bluestem and little bluestem had reduced relative plant height when the soil VWC drought set point was 14% or 7%. The reduced plant growth for big bluestem and little bluestem under stressful conditions likely led to both plants having a low visual damage rating (≈3).
Chinese silvergrass and feather reed grass were able to tolerate cyclical flooding, drought conditions, and, in some cases, increase growth. Under mild drought, Chinese silvergrass was able to increase the relative shoot count across all flood durations. During the experiment, all Chinese silvergrass plants survived, whereas only one feather reed grass plant died. Similar survival results were observed for Chinese silvergrass (Yuan and Dunnett 2018); all plants survived cyclic flooding for 1 d and 4 d. The adaptability of Chinese silvergrass and feather reed grass can be partially explained by their origins and wetland indicator status. The cultivar of Chinese silvergrass used in our study was Purpurascens, which is considered a hybrid between Chinese silvergrass and Amur silvergrass [M. sacchariflorus (Maxim.) Franch.] (Jiang et al. 2013). Amur silvergrass is typically found near wetlands and along waterways, which can explain the tolerance of ‘Purpurascens’ to flooded soils (Bonin et al. 2014). Chinese silvergrass is tolerant of soil water potentials as low as −4.05 MPa, which can explain the tolerance of ‘Purpurascens’ to drought stress (Dougherty et al. 2015). Clifton-Brown and Lewandowski (2000) found that Chinese silvergrass was able to minimize leaf senescence under drought conditions by reducing stomatal conductance and leaf area, which can explain the low visual damage rating observed during our study (≈4). The feather reed grass cultivar Karl Foerster is as a hybrid of chee reed grass [C. epigejos (L.) Roth] and C. arundinacea (L.) Roth. Both grasses are native to central Europe, where chee reed grass is found in dry, mesic, and flooded soils, whereas C. arundinacea is known to tolerate submergence (Lei et al. 2014; Rebele and Lehmann 2001; US Department of Agriculture, Natural Resources Conservation Service 2024). This can explain the ability of ‘Karl Foerster’ to tolerate both flood and drought stress.
A previous study by Nelson et al. (2018) evaluated seven sedge species to determine their tolerance to cyclical flooding and drought using a similar setup as the present experiment, except the lowest drought set point was −14,800 kPa. Interestingly, no sedges died during the experiment. Sedges with wetland indicators of facultative wetland, facultative, and facultative upland tolerated cyclical flooding and drought better than obligate and upland sedges. A similar result was observed in the present study, but not for tufted hair grass. Although it is a facultative wetland plant, tufted hair grass did not tolerate cyclical flooding and drought. Sedges also had higher relative shoot mass when flooding increased from 2 d to 7 d. This was not observed with the seven species of perennial ornamental grasses used in the current experiment.
During Expt. 2, as the submergence duration and depth increased, the most susceptible perennial grasses experienced decreased relative shoot height and count and relative root mass and increased visual damage ratings. These results were expected because, as the submergence duration increased, the amount of oxygen in submergence water and soil decreased because of oxygen diffusion from the air being 104-fold slower through water (Armstrong and Drew 2002). Low soil oxygen levels reduce plant root respiration and, ultimately, plant growth (Pedersen et al. 2021). The substrate oxygen status was monitored using IRIS tubes; as the submergence duration increased from 2 d, significantly more iron was removed from the tubes, indicating that roots and soil microbes used all the available soil oxygen.
The increases in the submergence duration and depth had little effects on plant growth parameters of switchgrass and Chinese silvergrass. For switchgrass, the relative shoot height, shoot count, and shoot mass were mostly positive, whereas the relative root mass was reduced ≈20% when compared with that of the respective control. Our results were similar to those of Barney et al. (2009), who found that flooded switchgrass plants had similar plant height, shoot count, shoot mass, and root mass when compared with those of the nonflooded plants. It is important to note that Barney et al. (2009) flooded switchgrass continuously for 11 weeks at a depth of 2 to 5 cm above the soil surface, whereas our experiment submerged plants intermittently at depths of 15 cm or 30 cm above the soil surface. Results of Chinese silvergrass were similar to those of switchgrass, but Chinese silvergrass had a greater reduction in shoot mass.
The relative shoot count was generally reduced as the submergence duration increased for tufted hair grass, little bluestem, blue grama grass, and feather reed grass. Big bluestem did not show a significant difference between 2-, 4-, and 7-d submergence treatments. However, big bluestem had a large reduction in the relative shoot count for all treatments in comparison with that of the control. The results for big bluestem are similar to those during Expt. 1 and were not surprising because big bluestem is a facultative upland plant. The results were unanticipated for tufted hair grass because it is a facultative wetland plant; therefore, it is expected to perform better than the other grasses that are facultative upland or upland plants. The poor performance of tufted hair grass can be partially explained by shoot height. Tufted hair grass had the shortest shoot height of all the grasses in the experiment, and submergence at 15 cm and 30 cm resulted in the majority of the shoots being underwater, likely resulting in greater plant damage. Because of the short stature of tufted hair grass, it was not surprising that relative shoot and root mass were reduced >20% and >50%, respectively. Little bluestem had a positive relative shoot and root mass, which was surprising because of the grass is a facultative upland plant.
Similar to Expt. 1, visual damage ratings were lowest for switchgrass (<2) and Chinese silvergrass (<3.5), regardless of the submergence duration. All switchgrass and Chinese silvergrass plants survived the experiment, suggesting that both are able to tolerate intermittent submergence. Tufted hair grass and big bluestem had reduced growth of all plant parameters, but this did not reduce visual damage ratings or improve survival. The visual damage rating was highest for tufted hair grass and big bluestem, ranging from >6 to approximately 9. During the experiment, 14 tufted hair grass plants and seven big bluestem plants died. Big bluestem was able to tolerate cyclical flood and drought, but it did not tolerate intermittent submergence. As previously stated, tufted hair grass was the shortest grass in the experiment; because most of the foliage was covered by water, its survival was significantly reduced and visual damage rating was increased. Tufted hair grass was the only grass with a better visual damage rating when submerged at a depth of 15 cm (≈6.5) compared with that when it was submerged at a depth of 30 cm (≈8.5).
Visual damage ratings were similar for little bluestem, blue grama grass, and feather reed grass when submerged up to 4 d. Both blue grama grass and feather reed grass had a lower relative root mass compared with that of little bluestem, which may have allowed the plants to withstand submergence better than little bluestem. All three grasses had a visual damage rating of ≥6 when submerged for 7 d. Surprisingly, all the blue grama grasses survived the experiment, whereas only two little bluestem and three feather reed grass plants died. As stated, blue grama grass is known to be intolerant to soil flooding and submergence; therefore, it is interesting that all plants survived the duration of the experiment. It is important to note that grasses only went through four cycles of submergence with a 2-d draining period between each cycle. If the number of submergence cycles had been more than four, then blue grama grass survival may have been reduced.
Practical implications for rain gardens and bioretention systems.
Perennial ornamental grasses are a popular garden and landscape plant because of year-long interest and their minimal maintenance requirement. These two experiments showed that perennial ornamental grasses can survive under tough environmental conditions such as water-logged and drought-stressed soils and when plants are partially submerged by water. The perennial ornamental grasses ‘Northwind’ switchgrass and ‘Purpurascens’ Chinese silvergrass were able to survive cyclical flooding and drought as well as repeated submergence of up to 30 cm for periods of 7 d. The extreme environmental conditions tolerated by these two grasses make them candidates for bioretention systems or portions of rain gardens where periodic drought and submergence are likely to occur. The ‘Blue Heaven’ little bluestem, ‘Karl Foerster feather reed grass, and ‘Blonde Ambition’ blue grama grass were able to tolerate cyclical flooding and drought, especially when soil VWC was at 14%. The grasses experienced significant plant damage when submerged for 7 d. The grasses are well-suited for situations of fluctuating drought and soil flooding, but they should not be planted in locations where prolonged submergence may occur. The ‘Pixie Fountain’ tufted hairgrass and ‘Red October’ big bluestem experienced significant plant damage when submerged for 2 d or longer. Compared to tufted hairgrass, big bluestem was able to manage drought better. The cultivar Red October big bluestem would be suited for areas that experience cyclical soil flooding and drought if plant submergence does not occur. The cultivar Pixie Fountain tufted hairgrass should be used with caution because of its poor tolerance to submergence and drought. The cultivar Pixie Fountain would be suitable for areas where whole plant submergence will not occur and the soil VWC will not decrease below 14%.
Conclusions
Based on these two experiments, ‘Purpurascens’ Chinese silvergrass and ‘Northwind’ switchgrass were able to survive cyclical soil flooding and drought as well as submergence for 7 d at a depth of 30 cm while exhibiting an acceptable amount of foliar damage. All grasses survived cyclical flood and drought when the soil VWC was maintained at 14%, suggesting that all seven grasses can withstand periodic soil flooding if the water is not too deep. As the water depth and duration increased from 4 d to 7 d, ‘Blue Heaven’ little bluestem, ‘Blonde Ambition’ blue grama grass, and ‘Karl Foerster’ feather reed grass experienced significant foliar damage. ‘Pixie Fountain’ tufted hair grass and ‘Red October’ big bluestem experienced significant foliar damage when submerged for 2 d.
Our results showed that perennial ornamental grasses can tolerate cyclical flood and drought and periodic submergence, but that the plant condition and survival vary by species, but not by the wetland indicator status. Future research should focus on evaluating other commonly used perennial ornamental grasses to determine their suitability for use in stormwater management.
References Cited
Armstrong W, Drew MC. 2002. Root growth and metabolism under oxygen deficiency, p 729–761. In: Waisel Y, Eshel A, Kafkafi U (eds). Plant roots: The hidden half (2nd ed). Marcel Dekker, New York, USA. https://doi.org/10.1201/9780203909423.
Asleson BC, Nestingen RS, Gulliver JS, Hozalski RM, Nieber JL. 2009. Performance assessment of rain gardens. J Amer Water Resour Assoc. 45:1019–1031. https://doi.org/10.1111/j.1752-1688.2009.00344.x.
Barney JN, Mann JJ, Kyser GB, Blumwald E, Deynze AV, DiTomaso JM. 2009. Tolerance of switchgrass to extreme soil moisture stress: Ecological implications. Plant Sci. 177(6):724–732. https://doi.org/10.1016/j.plantsci.2009.09.003.
Blake GR, Hartge KH. 1986. Particle density, p 377–382. In: Klute A (ed). Methods of soil analysis: Part 1-physical and mineralogical methods. SSSA, ASA, Madison, WI, USA. https://doi.org/10.2136/sssabookser5.1.2ed.c14.
Bonin CL, Heaton EA, Barb J. 2014. Miscanthus sacchariflorus – biofule parent or new weed? Glob Change Biol Bioenergy. 6:629–636. https://doi.org/10.1111/gcbb.12098.
Castenson KL, Rabenhorst MC. 2006. Indication of reduction in soil (IRIS): Evaluation of a new approach for assessing reduced conditions in soil. Soil Sci Soc Amer J. 70:1222–1226. https://doi.org/10.2136/sssaj2005.0130.
Clifton-Brown JC, Lewandowski I. 2000. Water use efficiency and biomass partitioning of three different Miscanthus genotypes with limited and unlimited water supply. Ann Bot. 86(1):191–200. https://doi.org/10.1006/anbo.2000.1183.
Davis AP, Hunt WF, Traver RG, Clar M. 2009. Bioretention technology: Overview of current practice and future needs. J Environ Eng. 135(3):109–117. https://doi.org/10.1061/(ASCE)0733-9372(2009)135:3(109).
Dougherty RF, Quinn LD, Voigt TB, Barney JN. 2015. Response of naturalized and ornamental biotypes of Miscanthus sinensis to soil-moisture and shade stress. Northeastern Naturalist. 22(2):372–386. https://doi.org/10.1656/045.022.0210.
Farooq M, Hussain M, Wahid A, Siddique KHM. 2012. Drought stress in plants: An overview, p 1–36. In: Aroca R (ed). Plant responses to drought stress: From morphological to molecular features. Springer, Verlag, Berlin. https://doi.org/10.1007/978-3-642-32653-0_1.
Fraser LH, Karnezis JP. 2005. A comparative assessment of seedling survival and biomass accumulation for fourteen wetland plant species grown under minor water-depth differences. Wetlands. 25(3):520–530. https://doi.org/10.1672/0277-5212(2005)025[0520:ACAOSS]2.0.CO;2.
Gray MM, Amand PS, Bello NM, Galliart MB, Knapp M, Garrett KA, Morgan TJ, Baer SG, Maricle BR, Akhunov ED, Johnson LC. 2014. Ecotypes for an ecologically dominant prairie grass (Andropogon gerardii) exhibit genetic divergence across the US Midwest grasslands’ environmental gradient. Mol Ecol. 23(24):6011–6028. https://doi.org/10.1111/mec.12993.
Hausken S, Thompson G. 2018. Rain garden plants. http://www.extension.umn.edu/garden/yard-garden/landscaping/best-plants-for-tough-sites/docs/08464-rain-garden.pdf. [accessed 11 Feb 2018].
Hunt WF, Smith JT, Jadlocki SJ, Hathaway JM, Eubanks PR. 2008. Pollutant removal and peak flow mitigation by a bioretention cell in urban Charlotte, N.C. J Environ Eng. 134(5):403–408. https://doi.org/10.1061/(ASCE)0733-9372(2008)134:5(403).
Integrated Taxonomic Information System. 2022. Online database. https://www.itis.gov/servlet/SingleRpt/SingleRpt?search_topic=TSN&search_value=531165#null. [accessed 20 May 2022].
Jelitto Perennial Seeds. 2024. Deschampsia cespitosa ‘Pixie Fountain’ tufted hair grass. https://www.jelitto.com/Seed/Ornamental+Grasses/DESCHAMPSIA+cespitosa+Pixie+Fountain+Portion+s.html. [accessed 15 May 2024].
Jennings AA, Berger MA, Hale JD. 2015. Hydraulic and hydrologic performance of residential rain gardens. J Environ Eng. 141(11). http://doi.org/10.1061/(ASCE)EE.1943-7870.0000967.
Jiang J, Zhu M, Ai X, Xiao L, Deng G, Yi Z. 2013. Molecular evidence for a natural diploid hybrid between Miscanthus sinensis (Poaceae) and M. sacchariflorus. Plant Syst Evol. 299(7):1367–1377. https://doi.org/10.1007/s00606-013-0801-2.
Lei S, Zeng B, Yuan Z, Su X. 2014. Changes in carbohydrate content and membrane stability of two ecotypes of Calamagrostis arundinacea growing at different elevations in the drawdown zone of the Three River Gorges reservoir. PLoS One. 9(3):e91394. https://doi.org/10.1371/journal.pone.0091394.
Li C, Peng C, Chiang P, Cai Y, Wang X, Yang Z. 2019. Mechanisms and applications of green infrastructure practices for stormwater control: A review. J Hydrol. 568:626–637. https://doi.org/10.1016/j.jhydrol.2018.10.074.
Lichvar RW, Banks DL, Kirchner WN, Melvin NC. 2016. The national wetland plant list: 2016 wetland ratings. Phytoneuron. 30:1–17.
Mann JJ, Barney JN, Kyser GB, Ditomaso JM. 2013. Root system dynamics of Miscanthus x giganteus and Panicum virgatum in response to rainfed and irrigated conditions in California. Bioenerg Res. 6(2):678–687. https://doi.org/10.1007/s12155-012-9287-y.
Meyer MH. 2004. Ornamental grasses for cold climates. UMN Extension Service, St. Paul, MN, USA. http://purl.umn.edu/51550.
Meyer MH (inventor). 2006. Schizachyrium plant named ‘MinnblueA’. Regents of the University of Minnesota (assignee). US Plant Patent 17,310. (Filed 13 Apr 2005, granted 26 Dec 2006).
Meyer MH. 2012. Ornamental grasses in the United States, p 121–152. In: Janick J (ed). Horticultural reviews, volume 39. John Wiley & Sons, Hoboken, NJ, USA. https://doi.org/10.1002/9781118100592.ch3.
Minnesota Pollution Control Agency. 2021. Minnesota stormwater manual. Minnesota Pollution Control Agency. http://stormwater.pca.state.mn.us/index.php/Main_Page. [accessed 9 Oct 2021].
Missouri Botanical Garden. 2020a. Andropogon ‘Red October’. St. Louis, MO, USA. http://www.missouribotanicalgarden.org/PlantFinder/PlantFinderDetails.aspx?taxonid=299543&force=1. [accessed 7 May 2020].
Missouri Botanical Garden. 2020b. Panicum virgatum ‘Northwind’. St. Louis, MO, USA. https://www.missouribotanicalgarden.org/PlantFinder/PlantFinderDetails.aspx?kempercode=z240. [accessed 7 May 2020].
Myers M. 2024. 2014 perennial plant of the year - Northwind switchgrass. https://www.melindamyers.com/audio-video/melindas-garden-moment-audio-tips/flowers-ornamental-grasses/2014-perennial-plant-of-the-year-northwind-switchgrass. [accessed 15 May 2024].
Nelson RS, McGinnis EE, Daigh ALM. 2018. Rain garden sedges tolerate cyclical flooding and drought. HortScience. 53(11):1669–1676. https://doi.org/10.21273/HORTSCI13234-18.
Pedersen O, Sauter M, Colmer TD, Nakazono M. 2021. Regulation of root adaptive anatomical and morphological traits during low soil oxygen. New Phyologist. 229(1):42–49. https://doi.org/10.1111/nph.16375.
Rabenhorst MC. 2008. Protocol for using and interpreting IRIS tubes. Soil Surv Horiz. 49(3):74–77. https://doi.org/10.2136/sh2008.3.0074.
Rebele F, Lehmann C. 2001. Biological flora of central Europe: Calamagrostis epigejos (L.) Roth. Flora. 196(5):325–344. https://doi.org/10.1016/S0367-2530(17)30069-5.
Salman D (inventor). 2011. Bouteloua grass plant named ‘Blonde Ambition’. US Plant Patent 22,048. (Filed 12 May 2010, granted 26 Jul 2011).
Schneider CA, Rasband WS, Eliceiri KW. 2012. NIH Image to ImageJ: 25 years of image analysis. Nat Methods. 9(7):671–675. https://doi.org/10.1038%2Fnmeth.2089.
Soil Survey Staff. 2005. Delamere soil series official description. https://soilseries.sc.egov.usda.gov/OSD_Docs/D/DELAMERE.html. [accessed 9 Aug 2020].
Soil Survey Staff. 2011. Barnes soil series official description. https://soilseries.sc.egov.usda.gov/OSD_Docs/B/BARNES.html. [accessed 9 Aug 2020].
Stavridou E, Webster RJ, Robson PRH. 2019. Novel Miscanthus genotypes selected for different drought tolerance phenotypes show enhanced tolerance across combinations of salinity and drought treatments. Ann Bot. 124(4):653–674. https://doi.org/10.1093/aob/mcz009.
Steiner LM, Domm RW. 2012. Rain gardens: Sustainable landscaping for a beautiful yard and a healthy world. MBI Publishing Company, Minneapolis, MN, USA.
Taylor SH, Ripley BS, Martin T, De‐Wet L, Woodward FI, Osborne CP. 2014. Physiological advantages of C4 grasses in the field: A comparative experiment demonstrating the importance of drought. Glob Change Biol. 20(6):1992–2003. https://doi.org/10.1111/gcb.12498.
Tober D, Jensen N. 2013. Plant guide for little bluestem (Schizachyrium scoparium), USDA-NRCS, Plant Materials Center, Bismark, ND, USA. https://plants.usda.gov/DocumentLibrary/plantguide/pdf/pg_scsc.pdf.
Thetford M, Norcini JG, Ballard B, Aldrich JH. 2009. Ornamental landscape performance of native and nonnative grasses under low-input conditions. HortTechnology. 19(2):267–285. https://doi.org/10.21273/HORTTECH.19.2.267.
United Nations Department of Economic and Social Affairs Population Division. 2019. World urbanization prospects: The 2018 revision. https://population.un.org/wup/Publications/Files/WUP2018-Highlights.pdf. [accessed 8 Oct 2022].
US Army Corps of Engineers. 2022. National wetland plant list, v. 3.5. https://wetland-plants.sec.usace.army.mil/nwpl_static/v34/home/home.html. [accessed 9 Feb 2022].
US Census Bureau. 2010. Census Data 2010. US Census Bureau. https://www.census.gov/programs-surveys/geography/guidance/geo-areas/urban-rural/ua-facts.html. [accessed 15 May 2022].
US Department of Agriculture, Natural Resources Conservation Service. 2006. Plant fact sheet: Switchgrass. https://plants.usda.gov/factsheet/pdf/fs_pavi2.pdf. [accessed 23 Apr 2020].
US Department of Agriculture, Natural Resources Conservation Service. 2024. The PLANTS Database. National Plant Data Team, Greensboro, NC. USA. http://plants.usda.gov. [accessed 24 May 2024].
US Environmental Protection Agency. 2008. National Research Council Report: Urban stormwater management in the United States.
Van Genuchten MT. 1980. A closed-form equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci Soc of Amer J. 44(5):892–898. https://doi.org/10.2136/sssaj1980.03615995004400050002x.
Vijayaraghavan K, Biswal BK, Adam MG, Soh SH, Tsen-Tieng DL, Davis AP, Chew SH, Tan PY, Babovic V, Balasubramanian R. 2021. Bioretention systems for stormwater management: Recent advances and future prospects. J Environ Manage. 292:112766. https://doi.org/10.1016/j.jenvman.2021.112766.
Wynia R. 2007. Plant guide for blue gramma [Bouteloua gracilis (Willd. ex Kunth.) Lag. Ex Griffiths]. USDA-NRCS, Manhattan Plant Materials Center, Manhattan, KS, USA. https://plants.usda.gov/DocumentLibrary/factsheet/pdf/fs_bogr2.pdf.
Yuan J, Dunnett N. 2018. Plant selection for rain gardens: Response to simulated cyclical flooding of 15 perennial species. Urban Forestry & Urban Greening. 35:57–65. https://doi.org/10.1016/j.ufug.2018.08.005.
