Full sun trial gardens (full sun) were established at Leon County (LC) and Santa Rosa County (SRC), Florida, to evaluate the effect of site on horticultural performance traits pertinent to landscape use (long-term growth, flowering, vigor, overall quality, and survival) of native and nonnative warm season grasses when grown under low-input landscape conditions over a 3-year period. The gardens contained landscape fabric- (LC) or plastic (SRC)-covered rows, with the synthetic mulch at each site covered by 4 inches of hardwood chip mulch. Fifteen native and eight nonnative grass species, and cultivars were evaluated as were the two grass-like species black flowering sedge (Carex nigra) and narrowleaf silkgrass (Pityopsis graminifolia), a native dicot with grass-like foliage. Many of the native species were derived from seeds or plants collected from naturally occurring populations in Florida, and other species or cultivars were obtained from commercial sources. Based on quality ratings and survival, a majority of the species and cultivars were rated as at least being marginally acceptable for 2 years or more, but only six species and cultivars were rated as excellent or good over all 3 years. Four of these six species were native, with 100% survival at both sites occurring only for purpletop tridens (Tridens flavus). The other three native species, gulf hairawn muhly, (Muhlenbergia capillaris var. filipes), ‘Alamo’ switchgrass (Panicum virgatum), and indian woodoats (Chasmanthium latifolium) had high rates of survival. Porcupine grass (Miscanthus sinensis ‘Strictus’) and ‘Cabaret’ silver grass (M. sinensis ssp. condensatus) were the only two nonnative species demonstrating potential for long-term performance in a low-input landscape at both sites. Three of the four cultivars of miscanthus (Miscanthus spp.) survived the entire evaluation period in SRC, while two cultivars suffered losses only in LC, demonstrating the importance of site effects on the long-term performance of individual species or cultivars of grasses. Chalky bluestem (Andropogon capillipes) (Orange County, FL), sand lovegrass (Eragrostis trichoides), giant silver grass (Miscanthus ×giganteus), and bamboo muhly (Muhlenbergia dumosa) each survived 3 years at one site but only 1 year at the other site. Pineland threeawn (wiregrass) (Aristida stricta), ‘Red Baron’ cogon grass (japanese blood grass) (Imperata cylindrical), ‘Hameln’ fountain grass (Pennisetum alopecuroides), black flowering fountain grass (P. alopecuroides ‘Moudry’), and ‘Feesey Form’ ribbon grass (Phalaris arundinaceae) were categorized as marginal for low-input landscapes and could only be considered short-term perennials under the conditions of this test. ‘Morning Light’ silver grass (M. sinensis), coastal bluestem (Schizachyrium maritimum), and ‘Lometa’ indiangrass (Sorghastrum nutans) had good 2-year landscape performance and survived at least 2 years at both sites. Bigtop lovegrass (Eragrostis hirsuta), silver plumegrass (Saccharum alopecuroides), and lopsided indiangrass (Sorghastrum secundum) were categorized as having excellent 1-year landscape performance and have potential for short-term performance under low-input conditions. Chalky bluestem (Andropogon capillipes) (SRC), black flowering sedge, ‘Heavy Metal’ switchgrass (Panicum virgatum), and narrowleaf silkgrass were categorized as having good 1-year landscape performance.
Ornamental and native grasses are increasing in popularity in southern United States gardens, as exemplified by evaluations in Florida (Wilson and Knox, 2006), Georgia (Corley and Reynolds, 1994; Ruter and Carter, 2000), and South Carolina (Aitken, 1995). The use of native grasses and sedges will most likely increase as preservation of natural areas and their restoration expands. The introduction of more native grasses into the ornamentals industry is a natural fit given the increasing demands for native plants in landscapes. Grasses derived from natural populations of Florida may exhibit some degree of tolerance to insects, disease, drought, and other forms of stress common in Floridian landscapes. Evaluation of material selected for horticultural use from within the natural diversity of the native flora is needed, and identification of better performers will result in even wider use in the ornamentals industry (Dana, 2002). However, these species must be evaluated outside of their natural habitat to determine their suitability for use in landscapes. Similarly, ornamental grasses have been generally regarded as problem free; the ultimate in low maintenance (Dana, 2002). Many ornamental grasses presently in the market may possess a similar degree of tolerance to insects, disease, or drought and may be suitable for low-input or large-scale landscapes where supplemental irrigation is not possible or practical. These species must also be evaluated outside of the traditional irrigated landscape to determine their suitability for use in a low-input landscape.
Aside from identifying grasses that may perform well in low-input conditions, these trials may prove useful in identifying species and cultivars with excessive growth potential under low-input conditions. The invasive potential of ornamental grasses into natural areas is an issue gaining importance as examples of pest potential are recognized, such as with pampas grass (Cortaderia selloana) in California (Bell et al., 2003), chinese silvergrass in North Carolina and neighboring states, and japanese blood grass, which has been banned in many states due to the invasive quality of the species (Dana, 2002). Results of recent studies strongly suggest that dispersal through local landscape plantings has contributed to the range expansion of invasive pampas grass in California (Okada et al., 2007), and the absence or presence of visible signs of recent disturbances have been shown to be the main habitat differentiation criterion associated with invasive success (Domènech and Vilà, 2007).
The objective of this 3-year study was to evaluate the ornamental potential of Florida native and nonnative grasses when grown under low-input landscape conditions. Specifically, we evaluated the effect of site on long-term growth, flowering, vigor, overall quality, and survival—horticultural performance traits that are pertinent to landscape use. Low-input conditions were used because typical garden conditions, in which moisture and nutrients are nonlimiting, might override potential performance differences among the species and cultivars due to variability in stress tolerance among them; to be consistent with environmentally friendly landscape practices, which include the use of mulch to minimize weed growth; and to identify grasses with potential for use in high stress environments such as roadside plantings, rough areas of golf courses, or landscapes such as schools, parks, or large-scale municipal plantings.
Materials and methods
Trial gardens (full sun) were established at two sites in northern Florida: Leon County Extension in Tallahassee, FL (Leon County) [lat. 30.4°N, long. −84.3°W; American Horticulture Society (AHS) Heat Zone 9; U.S. Department of Agriculture (USDA) Hardiness Zone 8b; Orangeburg fine sandy loam (2% organic matter, 2%–5% slope, typic paleudults)], and West Florida Research and Education Center in Jay, FL (Santa Rosa County) [lat. 31.0°N, long. −87.2°W; AHS Heat Zone 9; USDA Hardiness Zone 8a; Orangeburg sandy loam (2.1% organic matter, 1% slope, typic paleudults)]. The garden rows contained 3-ft-wide synthetic mulch [plastic (SRC) or landscape fabric (LC)] with a 2-ft-diameter circle removed at the site of each planting hole. All synthetic mulch and the bare soil around each plant were covered by 4 inches of hardwood chip mulch.
Twenty-five species and cultivars were evaluated in this study—15 native and 8 nonnative warm season grass species and two grass-like species; black flowering sedge and narrowleaf silkgrass (Table 1). Many of the native species were derived from seeds or plants collected from wild, naturally occurring populations in Florida (Table 1). Origins of these seeds or plants, as well as the USDA Cold Hardiness and AHS Heat Zones for these sites, are noted in Table 1. All other entries were obtained from commercial sources.
Landscape performance of native and nonnative grass species based on quality ratings and survival over three growing seasons under full sun, low-input conditions in Leon County and Santa Rosa County, FL.
Native grasses produced from seeds were germinated and grown under greenhouse conditions, and the remaining grasses were produced from divisions. Single seedlings or divisions were transplanted into a soilless substrate in 1-gal containers. The substrate was composed of 3/4-inch shaker-screened pine bark (Georgia-Florida Bark and Mulch, Capps, FL), Canadian sphagnum peat (Berger Peat Moss, St. Modeste, Quebec, Canada), and rescreened 6B gravel (Martin Aggregates, Chattahoochee, FL) at a ratio of 3:1:1 (by volume). Incorporated into this soilless substrate were Osmocote® 18–6–12 (18N–2.6P–10K, 8- to 9-month formulation at 70 °F; Scotts Co., Marysville, OH) at 6.0 lb/yard3 and Micromax® (Scotts Co.) at 1.6 lb/yard3. Pots were placed outdoors on a full-sun bed that was covered in black plastic. Plants were overhead irrigated (pH 7.8) daily to ensure that substrate moisture was nonlimiting. Containers were hand weeded as necessary.
In the latter half of Jan. 2000, six 1-gal plants of each of the 25 species and cultivars were transplanted in a randomized complete block design in each evaluation garden. Plants were spaced 5 or 6 ft apart (medium to small grasses, and large grasses, respectively) within each row, with 5 to 6 ft between rows. In late Feb. 2001 and 2002, plants were pruned back to 6 to 12 inches, with the final height directly related to plant size. Plants were irrigated with drip irrigation providing ≈1 gal of water at each watering as needed only during the first month after transplanting to aid establishment. No supplemental fertilizer was applied at any time during the study. No pesticides were applied except for fire ant (Solenopsis spp.) control and to control weeds around the fence and in the aisles. Plots were hand weeded as needed.
Once per month (from April to October) in 2000, 2001, and 2002, Master Gardeners (LC and SRC) and Master Wildlife Conservationists (LC) determined overall visual rating (1 = very poor, 2 = marginally acceptable, 3 = acceptable, 4 = good, 5 = very good, 6 = excellent), vegetative height, vegetative width, inflorescence height (from soil level), and basal width. For vegetative and basal widths, two measurements were made: widest point and perpendicular to widest point. During August and September, inflorescence height was recorded weekly. At the end of the study, species and cultivars were categorized into landscape performance groups based on overall quality ratings and survival. Species and cultivars were deemed as having survived at a site for “n” number of years if three or more plants were living at that site for that “n” number of years. Selections were classified in the highest landscape performance category for which the following criteria were attained for 1, 2, or 3 years: Excellent = mean rating of 3 or greater for at least four consecutive months at both sites and survived at each site for “n” years; Good = mean rating of 3 or greater for at least four consecutive months for “n” years at one site and “n minus 1” years at the other site and survived at each site for “n” years; Marginal = mean rating of 3 or greater for at least four consecutive months for 2 or 3 years at one site and survived at 2 or 3 years at that site. For example, a selection classified as having excellent performance for 3 years would have attained a mean rating of 3 or greater for four consecutive months for all 3 years at both sites and 3 years survival at both sites. A selection classified as having good 2-year performance would have achieved a rating of 3 or greater for 4 consecutive months for 2 years at one site and 1 year at the other site and survival at both sites was at least 2 years. The threshold of four consecutive months was chosen because 4 months was greater than 50% of the time when plants were evaluated each season.
Daily rainfall and average temperatures were based on data recorded by the National Oceanic Atmospheric Association weather station at the Tallahassee Regional Airport (6.6 km from the LC site) and the Pensacola Regional Airport (34.6 km from the SRC site).
Data were analyzed using mixed model methods. The random effects were blocks-within-sites. The fixed effects were sites. Means were separated using least squares means (with PDIFF option) as part of the mixed model analyses in SAS (version 8.01; SAS Institute, Cary, NC). Probability values were adjusted using the Tukey method when sample sizes were equal, and by the Tukey-Kramer method when sample sizes were unequal (by default in SAS, version 8.01). Differences in survival between sites and within selections precluded detection of significant sites differences at any point in time. Therefore, differences at specific points in time were deemed significant when the mean differences were at least two times greater than the standard errors. Differences in survival were determined using Fisher's exact test because of the small number of plants at each site. Results are graphically presented only for vegetative height and width and quality ratings; all other results are mentioned in the text.
Results and discussion
Florida was subjected to frequent rain from June through Aug. 2000 and 2001, although for SRC, rainfall declined Aug. 2001 and remained low through December (Fig. 1). Rainfall was highest in July and Sept. 2002 (Fig. 1). Overall, LC received more rain than SRC, with higher rainfall in late summer/early fall 2000, June and Aug. 2001, and Mar. 2002. Average temperatures were similar at SRC and LC (Fig. 1), although higher average temperatures were evident in SRC during the summer months compared with LC.
Invasiveness was not formally evaluated in our study. Determining invasiveness requires more than a 3-year study because the process of naturalizing in new sites (a requirement to be deemed invasive) might only begin under certain environmental conditions, such as above-normal temperatures and precipitation (Alpert et al., 2000). Moreover, we did not examine adjacent habitats.
However, we have noted observational data within the discussion for each species in this test that is pertinent to the potential invasiveness and weediness of these species under low-input landscape conditions of the coastal plain in the lower south. Our conclusions are specific to this region because invasiveness and weediness frequently are regional concerns (Bell et al., 2003). Weediness potential under landscape conditions probably is of more immediate concern to homeowners, and is possibly an indicator of invasiveness. However, the potential for seedling recruitment within planting sites was low given that the only microsites available for seedling recruitment were the small amounts of disturbed soil around each plant; these microsites and the reminder of the planting sites were also covered with 4 inches of hardwood chip mulch.
Based on quality ratings and survival, the majority of species and cultivars were rated as marginally acceptable for 2 years or more (Table 1). Descriptions of each species and cultivar and their landscape performance are detailed below.
Formerly known as A. virginicus var. glaucus, this species is native to the United States, and occurs from New Jersey to Texas (USDA, 2006a) and throughout Florida (Wunderlin and Hansen, 2004a) in moist areas. The typical height of this tufted, perennial bunch grass is from 60 to 180 cm (Tobe, 1998). The light blue foliage is covered with a white bloom that easily rubs off, hence the name chalky bluestem. In Florida, the long-lasting blue foliage, which eventually becomes reddish brown when it dies in late fall after flowering, is one of the reasons chalky bluestem is recommended for landscape use (Nelson, 2003). The plants produce pedicillate spikelets of inconspicuous purplish brown flowers along the terminal portions of the tillers, therefore maximum inflorescence height is equal to the height of the vegetative growth. The inflorescences and pale brown seed heads add ornamental interest but are not showy, mainly because they extend little if any above the foliage. Seed were collected from Orange County, FL (OC) and SRC.
At LC, both selections only survived the first season with similar height (125 and 126 cm), width (95 and 78 cm) (Figs. 2 and 3), and basal width (15 and 12 cm). For both selections, quality ratings were low (2.0–2.8) from April to July, after which ratings were high (3.0–4.5), which coincided with inflorescence development and flowering (Figs. 2 and 3).
However, at SRC both chalky bluestem selections survived all 3 years with no mortality of the OC selection and the loss of a single plant each season for the SRC selection. Maximum heights and widths ranged from 89 to 126 cm and 52 to 75 cm, respectively (Figs. 2 and 3). Quality ratings were generally lower for the SRC selection early in the year (1.0–2.4) but increased with the onset of inflorescence development and flowering in July/August through October. Quality ratings for the OC selection remained high in year 1 (3.0–5.2), while in years 2 and 3, ratings were below 2.0 in April and May followed by high ratings (3.0–4.8) the remainder of the season.
This native species, commonly known as wiregrass, occurs in southeastern coastal plains from North Carolina to eastern Mississippi and throughout Florida (Clewell, 1989; Hitchcock, 1924; Wunderlin and Hansen, 2004b). Wiregrass taxonomy has been the subject of recent debate with respect to wiregrass that occurs in the extreme northeastern part of its range versus wiregrass in the remainder of its range (Peet, 1993; Kessler et al., 2003; USDA, 2006b, 2006c; Ward, 2001). Based on empirical evidence, Kessler et al. (2003) concluded only one species exists, and that evidence was lacking to justify varietal status for wiregrass that occurs in all but the extreme northeastern part of its range (Ward, 2001).
Regardless of taxonomy issues, wiregrass is a perennial bunch grass whose name is derived from its tightly involuted blades, which appear to be narrow and wiry. It grows in moist and dry sites. Unlike many native grasses, dead foliage is retained over several seasons, which is the main reason it carries fire so well in the fire-adapted longleaf pine (Pinus palustris)/wiregrass ecosystem. Growth habit of this fine-textured species can vary from upright divergent to tufted; the typical height is up to 90 cm. Narrow panicles of pale brown, one-flowered spikelets with long awns appear in late summer/early fall and extend well above the foliage. While the inflorescences/seed heads add interest, they are not ornamental except in large populations, as would be seen in the longleaf pine forests.
In general, plants performed similarly at both sites during the first two seasons (Fig. 4), but wiregrass survived only 2 years in LC, while all but two plants survived up until the final month of the experiment in SRC. Maximum basal width ranged from 10 to 16 cm over the 3-year evaluation, with maximum plant height of 54 to 73 cm and maximum width of 65 to 92 cm. For all years and both locations, plants flowered from August to October and flower heights ranged from 55 to 98 cm. In SRC, quality ratings were high (3.0–4.0) later in the growing season, while in LC quality ratings remained low (below 3.0) for most of the study with the exception of September of year 1 (rating = 3.3), which corresponded to the peak flower period for that year (Fig. 4).
Black flowering sedge.
This member of the Cyperaceae occurs from Maine southward to Connecticut, and then westward to Wisconsin (USDA, 2006d). It is often found in moist, fertile sites in light shade. Black flowering sedge is slow-growing, creeping, and about 15.2 to 22.9 cm tall. Flowers appear in spring and are hidden within the soft, blue-gray foliage that is narrow and curled. In winter, the foliage can be killed by freezing temperatures, becomes tan in color, and is retained until covered by new foliage in the spring.
This species survived only two seasons at both sites. In SRC, plant mortality was evident in July and by October year, 2 only two plants remained. Shoot height increased to 16 cm in year 1 and only 9 cm in year 2 (Fig. 5). In year 1, basal width increased to14 cm, but by the end of year 2, plants had declined and basal width was only 7 cm in October. Initially, quality ratings were very high but plants quickly became discolored and visual ratings dropped below 2 by the end of year 1 and were never above 2.5 in year 2 (Fig. 5).
In LC, black flowering sedge performed very well in year 1, achieving a height of 34 cm, a width of 65 cm (Fig. 5), and a basal width of 22 cm. In addition, mean quality rating was three or greater the entire season. In year 2, end of season height (39 cm), width (63 cm), and basal width (30 cm) were similar to year 1, but plant quality was poor, as in SRC (Fig. 5).
Also known as mountain oats and northern sea oats, and formerly Uniola latifolia, this native species occurs throughout much of the United States, the main exceptions being New York, New England, and states from Minnesota to California (USDA, 2006e). In Florida, indian woodoats is only found in 11 north Florida counties from Pensacola to Jacksonville (Wunderlin and Hansen, 2004c). This tufted, perennial bunch grass with short rhizomes and determinant culms occurs in moist, shaded sites in Florida. Often only a single flush of growth occurs each season. Normally, it is about as tall (60 cm) as it is wide. Indian woodoats has good ornamental value over a relatively long period for a native grass as conspicuous, open panicles or racemes of green, oat-like spikelets appear in the summer, with the pale brown seed heads retained for many weeks thereafter.
Indian woodoats was one of only four species classified as having excellent performance for 3 years (Table 1). Mean quality ratings were three or greater most of the time at both sites, with the highest plant ratings coinciding with flowering in late spring and early summer (Fig. 6). In general, quality ratings were lower in April when culms were still expanding and in the fall when foliage began to senesce. Flowers extended above the foliage, with inflorescence height (from soil level) ranging from 61 cm to 94 cm over the 3-year period. The early flowering ended significant increases in plant height for the remainder of the growing season (Fig. 6). Only one plant at each site died: in July of year 1 at SRC, and in September of year 3 at LC. Plants in LC were wider than plants in SRC for much of the first 2 years, and they also appeared taller, but differences were not significant due to variation among the limited number of replications (Fig. 6). The greater vegetative growth at LC might have been due to better moisture-holding capacity of the soil at LC. The soil at LC had a high percentage of clay. Maximum foliage heights of 44 cm (SRC) and 55 cm (LC) in year 1 increased to 68 cm by year 2. Plants appeared shorter as the season progressed as the weight of the seed-filled florets increased and the culms began to bend downward from July to September. Maximum foliage widths of 41 cm (SRC) and 66 cm (LC) in year 1 increased from 83 cm to 131 cm by year 3. Basal diameter for plants in SRC increased from 12 cm to 24 cm by year 3, while basal width in LC more than doubled (17 vs. 34 cm).
This native species occurs mainly from Illinois and Ohio southward to Florida and Texas (USDA, 2006f). In Florida, it occurs throughout much of the northern and central part of the state (Wunderlin and Hansen, 2004d). Bigtop lovegrass is a tufted bunch grass classified as a perennial (Hall, 1978; USDA, 2006f), but in a preliminary evaluation of its performance under low-input landscape conditions, bigtop lovegrass was an annual to short-lived perennial (J.G. Norcini, unpublished results). However, what this species lacks in longevity is more than offset by its showy, airy panicles of tiny, purplish pink spikelets that appear in late summer/early fall. The inflorescences, which nearly hide the foliage, are especially ornamental when backlit or when covered with morning dew. The pale brown seed heads are retained for a few more weeks and provide additional ornamental value. In a preliminary study, we observed some reseeding.
Bigtop lovegrass survived two seasons in both locations, which concurs with the previous observation of this species being a short-lived perennial under northern Florida conditions. In SRC, plant mortality increased throughout the 2-year period, with only two plants surviving by July of year 2. However, in LC, only a single plant died over the winter between the first and second seasons. Plants in SRC were smaller (height and width) than those in LC during year 1, with maximum plant height of SRC plants peaking out at only 37 cm, while plants in LC were nearly twice as tall (Fig. 7). In year 2, plants at LC were smaller than in year 1; however, variability increased, and there were fewer plants between the two sites, therefore differences in height and width among the two sites were not significant. Basal width of plants in SRC was 21 cm in year 1 and further increased to 30 cm in year 2, while in LC, basal width was 31 cm to 32 cm in years 1 and 2, respectively. Quality ratings were very high in year 1 at both sites but in year 2, the only ratings of 3 or more were in SRC, but not until July when flowering began (Fig. 7). Peak quality ratings coincided with flowering from July to October. Inflorescence heights from soil level ranged from 42 to 86 cm at both sites.
This native bunch grass occurs throughout a large portion of the United States from Wyoming to New York, and southward to New Mexico and Georgia (USDA, 2006g). The species is mostly deciduous in colder climates and prefers sand barrens and open sandy woods. While it tolerates hot dry sites, sand lovegrass needs some moisture during the summer in hot, arid climates (Greenlee, 1992). This short-lived perennial has an upright growth habit and normally grows about 80 to 150 cm tall (USDA, 2005, 2006g). Sand lovegrass produces large, open panicles of four to six purplish flowered spikelets in late spring (USDA, 2005, 2006g).
Sand lovegrass survived 3 years in SRC, with all but one plant living until very late in the third season. In contrast, survival in LC was only 1 year. Landscape performance was very similar among the sites in year 1, although height and width were notably greater in LC early in the season (Fig. 8). The foliage reached a maximum height of 100 to 118 cm and the width of bunches were from 130 to 175 cm (Fig. 8), while basal width was from 36 to 48 cm. This species flowered from April to October at both sites. Initial flower heights were from 58 to 65 cm the first month of each season, while the inflorescences were from 100 to 168 cm in height the remaining months. The season-long flowering had a great visual impact on the plants and this was reflected in high visual quality ratings (3–6) for all but one rating period (Fig. 8).
‘Red Baron’ cogongrass.
This perennial rhizomatous grass, which is native to east and southeast Asia, India, Micronesia, and Australia, was introduced to the United States from Japan. The parent species is considered a nonnative, invasive exotic in Florida (Florida Exotic Pest Plant Council, 2008). Although not evaluated for seed production, ‘Red Baron’ has been noted to flower in garden situations (Raymer and Thetford, 2002).
‘Red Baron’ cogongrass survived one season in both locations and two seasons in SRC. Plant mortality was similar at both locations in year 1, with four plants surviving through October. All four remaining plants survived the entire second season in SRC with maximum shoot height (65 cm), width (144 cm), and basal width (96 cm) nearly twice those of year 1. Quality ratings were initially low but increased by summer (Fig. 9). In LC, plants were less vigorous (shoot height 41 cm, width 42 cm, and basal width 14 cm) and this was reflected in the low quality ratings, which remained low throughout the growing season and never reached an acceptable level (Fig. 9). Less than half of the plants survived beyond year 2 in SRC and year 1 in LC. While this species is noted to be sterile and not produce any inflorescences, the plants in SRC produced inflorescences in both years. Flowers were present in May in year 1 and from May through June in year 2, which contradicts many publications stating that this cultivar does not flower. Hence, further evaluation of ‘Red Baron’ cogon grass as a potential invasive species is warranted. The likelihood of seed dispersal is high if additional genets are in close proximity of this cultivar.
Giant Silver grass.
This silver grass is a hybrid of M. sinensis and M. sacchariflorus (Greef et al., 1997), which unlike some ornamental miscanthus species and cultivars that have escaped and become invasive, does not set viable seeds (Hodkinson and Renvoize, 2001) and so is not a threat to native plant communities. Miscanthus species are of tropical/subtropical origin found from the Pacific islands to the mountain regions of Japan and the Himalayas (Greef and Deuter, 1993).
Plants performed similarly at both sites in year 1 but in year 2, plants at LC were less vigorous. In LC in year 2, maximum height and width declined from August to October (Fig. 10). Less than half of the plants survived the third growing season in LC, while all plants survived for 3 years in SRC. For all years, flower heights ranged from 196 to 296 cm and plants flowered from July to October in both locations. Quality ratings remained high (4–5) for the entire growing season for all years in SRC; however, in LC, quality ratings increased from 3.5 to 5 between April and October year 1 but never exceeded 3.1 the second season.
‘Cabaret’ silver grass.
This broad-leaved, variegated form of purple Japanese silver grass was introduced into the United States from Japan by the U.S. National Arboretum in 1976. The leaves are cream-white in the center with wide dark green margins, and mature specimens may reach 250 cm. ‘Cabaret’ requires a long warm season for flowering, with copper inflorescences appearing in late September. Under Florida conditions, Wilson and Knox (2006) noted that ‘Cabaret’, as well as two other purple Japanese silver grass cultivars (‘Central Park’ and ‘Cosmopolitan’), grew quite large but produced comparatively fewer inflorescences than grasses of similar size. And out of concern about its invasive potential, they reported that 46% to 64% of seeds harvested from Florida plants were viable, and of those seeds, up to 96% germinated. However, there is anecdotal evidence that seedlings of ‘Cabaret’ silver grass routinely lack chlorophyll and do not survive (Darke, 2004); Wilson and Knox (2006) did not assess seedling survival.
This cultivar survived 3 years at both sites with no mortality in SRC and 33% mortality in LC in year 3. Plant vigor was similar among the two locations in year 1, but thereafter, plants in LC did not achieve foliar heights and widths greater than those achieved the first season (Fig. 11). By year 3, basal width for plants in SRC was double the basal width of the first season (32.6 vs. 61.6 cm), while in LC, basal width increased only 31% from year 1 to year 3 (32.3 vs. 41.5 cm). In year 1, plants flowered from September to October and flower heights were about 170 cm at both locations. In years 2 and 3, flower heights were 189 cm in SRC, while flower height decreased each year in LC (152 and 141 cm in years 2 and 3). Quality ratings remained high (4–5) May to October for all years in SRC; however, in LC, quality ratings increased from 3.8 to 5.8 between April and October year 1 but seldom exceeded 3.0 until August and September of year 3 when flowering was present (Fig. 11).
‘Morning Light’ silver grass.
This narrow-leaved, variegated cultivar of silver grass from Japan was introduced to the United States by the U.S. National Arboretum in 1976. The foliage is extremely narrow, with vertical bands of white and green; the inflorescences are initially crimson-red then fade to tan.
‘Morning Light’ silver grass survived 3 years at both sites with increasing mortality after year 1. By the end of year 3, survival for SRC was 50% and for LC, was 66%. Plant vigor was initially similar between the two locations, but plant size began to decline in LC in year 3 as the three surviving plants in LC only achieved heights and widths similar to that of plants in year 1 (Fig. 12). Maximum foliar height and width continued to increase each year at SRC. By year 3, basal widths increased to 36 cm in SRC and to 32 cm in LC. Plants flowered from September to October, with inflorescence heights of 75 to 142 cm. Plants in LC did not flower in year 3. Quality ratings improved each season at SRC, and overall quality was rated as 3 or better for all but April to June of 2000 (Fig. 12). However, in LC, quality ratings remained below 3 after July of year 2. The low quality ratings in LC were attributed to severe infestations of a foliar disease identified as miscanthus blight, a fungal disease incited by Leptosphaeria and its anamorphic state Stagonospora (O'Neill and Farr, 1996). The poor performance of this cultivar in Florida landscapes concurs with the conclusion of Wilson and Knox (2006) that ‘Morning Light’ silver grass is not well-suited for Florida landscapes based on its poor survival, growth, and flowering.
This variegated form of silver grass from Japan is distinguished by horizontal golden bands of variegation throughout the leaf. The leaves are up to 1.3 cm wide and plants may reach 120 to 180 cm in height, with inflorescences up to 50 cm above the foliage (Greenlee, 1992).
Porcupine grass survived 3 years at both sites with the loss of only a single plant in LC at the end of year 3. Plant vigor was similar among the two locations for all 3 years, with maximum foliar heights from 134 to 164 cm (SRC) and 144 to 151 cm (LC), and maximum widths from 141 to 217 cm (SRC) and 142 to 209 cm (LC) (Fig. 13). By year 3, basal width increased to 106 cm in SRC but only 46 cm in LC. Plants flowered from June/July to October, with inflorescence heights ranging from 108 to 220 cm. In year 3, inflorescence height for plants in LC was 10 to 30 cm shorter than in previous years. Although plants achieved similar growth at both locations, quality ratings did not remain high throughout the 3-year period. Quality ratings, which were similar at both sites in year 1, were initially low until May after which ratings remained above 3.0 (Fig. 13). After year 1, quality ratings in SRC remained above 4.1 until September of year 3 when miscanthus blight became a problem on the variegated portions of the leaves. A similar pattern of lower visual ratings occurred with plants in LC at the end of year 2 when ratings dropped below 4, and visual ratings did not recover until plants began to flower in July.
Gulf hairawn muhly.
This increasingly popular native bunchgrass, which is widely referred to as muhly grass, is a perennial found from Connecticut to Texas (USDA, 2007a) and throughout much of Florida (Wunderlin and Hansen, 2004e). In this study, evidence suggests that we evaluated gulf hairawn muhly because the collection site was Levy County, FL, and only gulf hairawn muhly has been documented in that county. Wunderlin and Hansen (2004e) refer to gulf hairawn muhly as M. capillaris var. filipes, but it is referred to as M. filipes by the USDA (2007b). Gulf hairawn muhly is a southeastern United States species found from North Carolina to Texas (USDA, 2007b), and is slightly less widespread in Florida compared with hairawn muhly (Wunderlin and Hansen, 2004e). While both species occur in moist areas, only gulf hairawn muhly has been found growing on beaches and dunes (Hall, 1978). Phenotypically, hairawn and gulf hairawn muhly differ only in awn and lemma characteristics, neither of which would affect the overall appearance of the inflorescences by the casual observer. Hall (1978) even noted that “…in Florida specimens, it proves impossible to separate these two varieties on any consistent basis.” Both species have a tufted to upright divergent growth habit, are 90 to 150 cm tall, and flower in the fall. The showy inflorescences are composed of airy panicles of purplish red spikelets, which fade to pale brown. Like bigtop lovegrass, hairawn muhly inflorescences are quite attractive when back lit and when covered with morning dew. As a native to the pine flatwoods, coastal upland and beach dunes, and sandhill communities of Florida, muhly grass should be well-suited for moist sites as well as beach front landscapes.
Plants performed similarly at both sites, although plants in LC were often larger than plants in SRC in years 1 and 2 (Fig. 14). All plants survived in SRC, while in LC a single plant died in August of year 2. At both sites, plant size increased during the first 2 years, with no increase in maximum height or width during year 3 (Fig. 14). Basal widths at LC and SRC gradually increased over 3 years to 31 cm and 42 cm, respectively.
Plants flowered from June to October, with maximum inflorescence heights in SRC ranging from 113 to 121 cm and in LC from 119 to 146 cm in years 1 and 2. In year 3, maximum inflorescence heights were 115 cm for SRC, but were only 75 cm for LC. In SRC, quality ratings remained above 3 and by July remained high (4–6) for all years (Fig. 14). For all years in LC, quality ratings were initially low in April (<3) but increased by June and remained high (4–6) for the remaining growing season.
This perennial, warm season grass native to southern Arizona and northwestern Mexico at 7000 to 8000 ft elevation has an upward to slightly arching growth habit (Hitchcock, 1950; USDA, 2007c). Flowers are initiated in the spring on stems from the previous season.
The aboveground portions of the plants were not winter hardy in North Florida, therefore no flower data were available. Plants in SRC were vigorous for all 3 years with the loss of only a single plant in year 2, while plants in LC were less vigorous and no plants survived beyond two growing seasons. A single plant died in June of year 1 and emergence the following spring was slow, with three plants emerging in April and an additional plant emerging in June. Increases in maximum height, width, and basal width (27, 37, and 58 cm, respectively) were evident over the 3-year evaluation period in SRC (Fig. 15). In LC, there was no increase in maximum height (124 vs. 123 cm) or width (136 vs. 100 cm) from year 1 to year 2 (Fig. 15), and only a minimal increase in basal width (15 vs. 23 cm) in year 2. Quality ratings remained high (3–5) from May to October for all years in SRC; however, in LC, visual ratings increased from 2.6 to 4 between April and October year 1 but never exceeded 3 in year 2 (Fig. 15).
This common native, perennial bunchgrass is found throughout the United States except on the west coast (USDA, 2007d), and throughout most of Florida (Wunderlin and Hansen, 2004f). It is adapted to lowland and upland conditions. Lowland ecotypes are tetraploids that are coarse-stemmed, upright, and have glabrous foliage. In contrast, upland ecotypes are hexaploids and octoploids, fine-stemmed, semidecumbent and broad-based, and have varying amounts of pubescence on the leaf blades (Hultquist et al., 1997). Two cultivars were evaluated in this study, ‘Alamo’ and ‘Heavy Metal’.
‘Heavy Metal’ switchgrass, also known as blue switchgrass, has stiff, upright, bluish stems and foliage, and was introduced by Kurt Blumel Inc. of Baldwin, MD. It grows 120 to 150 cm tall and has purplish pink panicles of spikelets similar to those of ‘Alamo’ switchgrass. ‘Alamo’ switchgrass is a lowland ecotype that originated from a native stand in south Texas along the Frio River and was released in 1978 by the USDA Soil Conservation Service in Texas and the Texas Agricultural Experiment Station (USDA, 1979). This selection is 90 to 180 cm tall and has an upright divergent growth habit. The airy, pyramid-shaped panicles of purplish spikelets (fading to beige) appear in late summer/early fall and extend above the foliage. While the spikelets are purplish, they are not nearly as showy as those of gulf hairawn muhly or bigtop lovegrass.
‘Alamo’ switchgrass survived 3 years at both sites with minimal mortality. All but one plant survived in SRC, and three plants died in LC between June and August of year 3. Plant vigor was greater in LC mainly during year 1 as plants were taller (191 vs. 150 cm) and wider (338 vs. 224 cm), with the effect carrying over into the beginning of year 2 (Fig. 16). By year 3, basal width increased to 95 cm in SRC and to 84 cm in LC. Plants flowered from July to October and inflorescence heights were from 195 to 250 cm in SRC and 238 to 252 cm in LC. Quality ratings remained high (3–6) throughout the 3-year period for both sites, with the exception of July and August year 2 in LC (Fig. 16). The low quality ratings in LC were at least partly attributed to the lodging of flowering culms, which resulted in shorter and broader plants.
‘Heavy Metal’ switchgrass survived only two seasons in both locations. Although two plants died in LC in September of year 2, plant mortality increased at both sites in year 3 when less than half of the plants emerged. Maximum height was similar for the 2 years in SRC (72 vs. 74), while maximum height of plants in LC increased from 82 cm to 94 cm. Although maximum width increased in SRC (55 vs. 84 cm) and LC (71 vs. 94 cm), maximum width occurred early in the growing season, and heights were reduced after August as most of the foliage dropped from the plants and new tillers emerged (Fig. 17). Basal width continued to increase in SRC (33 vs. 50 cm) but in LC, basal width appeared to decrease after August the second season as two of the larger plants died (38 vs. 32 cm). Quality ratings were rarely high throughout the season and generally only when flowering was present (Fig. 17). Quality ratings remained below the acceptable level of 3 for the majority of the trial. Flowering for both locations began around April and inflorescences remained intact on the plants long after the stems died. Inflorescence heights for the 2 years peaked at 93 and 98 cm in SRC and 106 and 108 cm in LC. The architecture and foliage of ‘Heavy Metal’ switchgrass suggest it is a lowland ecotype.
‘Hameln’ fountain grass.
Chinese fountain grass is a bunch grass native from eastern Asia to Western Australia that has escaped cultivation and established in many areas, including California and parts of the northeastern United States. ‘Hameln’ is a dwarf cultivar that does not reseed, therefore it is propagated by division (Gilman, 1999). The 15- to 18-cm inflorescence resembles a bottle brush. The foliage is bright green during the summer, and turns golden brown in the fall after the flowers begin to die.
‘Hameln’ fountain grass survived two seasons in both locations. All plants survived for 2 years in SRC, while in LC, plant mortality increased beginning in August of year 1 with only two plants surviving by June of year 2. Plants in SRC increased in size the second season (Fig. 18), while plant size could not be accurately evaluated in LC after May with only two plants surviving. Maximum plant height was 39 and 65 cm in SRC (years 1 and 2) and 41 cm in LC in year 1 (Fig. 18). Plant width followed a similar pattern (80 and 119 cm in SRC years 1 and 2 and 84 cm in LC) (Fig. 18). In SRC, maximum basal width was 28 cm in year 1 and 96 cm by the end of year 2, while in LC, maximum basal width was 24 for year 1. Quality ratings were initially low from April to June/July each year but were high from July/August to October (Fig. 18), usually when inflorescences were present. Flowering for both locations was from August to October and inflorescence heights were from 56 to 96 cm.
Black flowering fountain grass.
This fountain grass cultivar was introduced by the U.S. National Arboretum. It is particularly fertile, thus it has the potential to escape cultivation. Black flowering fountain grass survived two seasons in SRC with no mortality and plants increased in size the second season (Fig. 19). In LC, no mortality occurred the first season, but no plants emerged the second season. Maximum plant height was similar in SRC (61 cm) and LC (66 cm) in year 1 and plant width followed a similar pattern (Fig. 19). In SRC, maximum basal width was 23 cm in year 1 and 27 cm by the end of year 2, while in LC, maximum basal width was 23 cm for year 1. Quality ratings remained at 3 or greater at all times, with the highest ratings from July to September (Fig. 19), and corresponded with flowering for both locations. Maximum inflorescence height (57 to 66 cm) occurred September to October.
‘Feesey Form’ ribbon grass.
This is a vegetative propagated cultivar of ribbon grass, a warm season perennial with a spreading growth habit. It is native to northern temperate regions of Europe, Asia, America, and Africa, and is widely distributed throughout those regions. ‘Feesey Form’ ribbon grass can be weedy or invasive in the United States (USDA, 2007e).
Plants in SRC were vigorous for all 3 years with the loss of only a single plant in year 3. Continued vigor was evident in SRC and basal width increased annually (22, 46, and 72 cm) over the 3-year evaluation period. In LC, plants were vigorous year 1 and although maximum height, width, and basal width (20 vs. 39 cm) occurred by August of year 2 (Fig. 20), 50% mortality also occurred by the end of the second growing season. Quality ratings remained acceptable (≥3) from April to July/August for most years in SRC; however, in LC, visual ratings remained below 3 after June of year 1 (Fig. 20). Performance typically declined as temperatures increased and rainfall decreased.
This native dicot (Asteraceae) has grass-like leaves covered with hairs on the upper leaf surface giving the plant a silvery gray appearance. Narrowleaf silkgrass is a rhizomatous species that occurs from Delaware to Texas (USDA, 2007f) and throughout most of Florida (Wunderlin and Hansen, 2004k). In Florida, it is a low-growing perennial from which emerge multistemmed inflorescences (up to 90 cm tall) covered with many small, yellow flowers during late summer and fall.
Narrowleaf silkgrass only survived a single season at both locations and no plants died in SRC the first season. Maximum height was similar at both sites except in July when the single plant died in LC (Fig. 21). Maximum width continued to increase at both sites throughout the growing season, with a peak width of 129 and 101 cm; basal widths for the two sites (21 and 23 cm) were also similar by October. Quality ratings remained above 3 the entire season for SRC, and in LC, the rating fell below 3 only for the month of July corresponding to the plant loss at that time. Flowering began as early as April for SRC and in June for LC, with inflorescence heights ranging from 60 to 70 cm from July to October.
This native, perennial bunchgrass was formerly known as Erianthus alopecuroides and occurs from New Jersey to Illinois and southward to Florida and Texas (USDA, 2007g). In Florida, it occurs mainly in the panhandle (Wunderlin and Hansen, 2004g). Silver plumegrass most often is found growing in damp woods, open areas, and field margins. It has a narrow upright habit and can grow 160 to 180 cm tall. Wide, flat blades emerge from the reed-like culms. Flowering occurs in early to mid-fall. Inflorescences have a silky appearance due to pubescent glumes on the panicles that emanate from the culms. These panicles extend well above the foliage and spikelets have a silvery cast in the sunlight.
Silver plumegrass survived only two seasons in both locations. In SRC, only a single plant did not survive the transplant period, while in LC, a single plant died soon after emerging in April of year 2. Maximum height was similar for the 2 years in SRC (149 vs. 152 cm), while maximum height of plants in LC decreased from 208 cm to 158 cm (Fig. 22). Maximum width did not increase in year 2, although basal width continued to increase in both SRC (22 vs. 25 cm) and LC (23 vs. 28 cm). Quality ratings were below 3 before July in year 1 and remained below the acceptable level of 3 for the majority of year 2. Flowering at both locations began in August/September, with inflorescence heights for the 2 years peaking at 203 and 260 cm in SRC, and 254 and 242 cm in LC.
This native species is relatively uncommon and only occurs along the Gulf Coast of the Florida panhandle to Louisiana. Nomenclature varies as the USDA (2007h) and Clewell (1985) refer to it as Schizachyrium maritimum, while Wunderlin and Hansen (2004h) and Hall (1978) lump it together with other forms of Schizachyrium scoparium. It normally grows in sandy areas, usually at the ocean waterline, but also along roads in low, dune areas. Unlike other native grasses in this study, gulf bluestem is a creeping perennial that spreads by rhizomes. It is easily distinguished by its bluish, glaucous leaves, prostrate growth habit, and terminal inflorescences of small, inconspicuous stalked spikelets (Clewell, 1985).
Plants performed similarly at both sites in year 1. In year 2, plants at LC were less vigorous, and no plants survived the third growing season in LC. In SRC, no plant mortality was evident until after August year 3 when two plants died. Increases in maximum height (84, 76, and 91 cm), width (153, 207, and 173 cm), and basal width (33, 62, and 116 cm) were evident over the 3-year period for SRC (Fig. 23). In LC, plants were vigorous during year 1 but much less so during year 2 (Fig. 23), although basal width in LC increased from 43 cm in year 1 to 83 cm in year 2. Gulf bluestem produces pedicillate spikelets along the terminal portions of the prostrate tillers, thus maximum flower heights are equal to plant height. Quality ratings were high (3.1–5.5) for both sites in year 1. In SRC, quality ratings remained high through September in year 2, but dropped to 2.8 in October, and in year 3, quality ratings dropped below 3 from July to October (Fig. 23). In LC, visual ratings never exceeded 2.8 in the second season.
This indiangrass cultivar originated from a native stand east of Lometa, TX, in the central Texas hill country northwest of Austin. It was released in 1983 by the USDA Soil Conservation Service in Texas and the Texas Agricultural Experiment Station (USDA, 1983). ‘Lometa’ indiangrass has bluish-green leaves and upright growth habit, and “occurs in large bunches joined by short, stout rhizomes” (USDA, 1983). Average height is about 120 cm, at least under the dry conditions of the central Texas hill country (average rainfall of 55.8 cm). In the fall, the narrow densely branched panicles of flowers extend above the foliage and give way to light brown seed heads that are somewhat ornamental.
‘Lometa’ indiangrass performed similarly at both sites in years 1 and 2. In SRC, no plant mortality was evident, while in LC, a single plant died in October year 2; no plants emerged in year 3. In SRC, maximum height and width initially increased and then declined (Fig. 24). Basal width (39, 56, and 62 cm) increased over the 3-year period. In LC, maximum height decreased in year 2, width remained similar, and basal width increased in year 2 (42 and 52 cm). Inflorescences were present from August to September and were from 131 to 176 cm in height. Quality ratings remained high (3.0–4.8) for both sites in year 1 (Fig. 24). In SRC, quality ratings remained high in year 2 through August, but dropped to 2.6 in September, and in year 3, visual ratings dropped below 3 except in July (5.3) and August (3.6) (Fig. 24). In LC, visual ratings only exceeded 2.5 in June. Unlike most other grasses in this trial, the visual rating of ‘Lometa’ indiangrass decreased with the onset of flowering and was attributed to the dramatic decline in foliage appearance.
This native, perennial bunchgrass is found mainly in the lower south, although it does occur as far north as Nebraska (USDA, 2008). In Florida, it occurs throughout the state in >80% of the counties (Wunderlin and Hansen, 2004i). Lopsided indiangrass prefers well-drained, sunny sites (Nelson, 2003), although in Florida and Georgia, it often grows under high pine shade (J.G. Norcini, unpublished observations). This upright perennial, which grows 90 to 180 cm tall (Nelson, 2003), is valued for it showy inflorescences that occur in the late summer or early fall. The large, multicolored spikelets are hues of browns and yellows and have relatively long, twisted awns. However, those familiar with this species are keenly aware that the showy spikelets do not remain on the plant very long, typically about 2 weeks (J.G. Norcini, unpublished observations). Given the limited time that lopsided indiangrass is at peak ornamental value, it is best used as an accent plant or in naturalistic settings (Nelson, 2003).
Lopsided indiangrass performed similarly at both sites in years 1 and 2. In SRC, plant mortality was evident in year 1 with four plants surviving through October and only three plants surviving in year 2. In LC, no mortality was evident in year 1 but by the end of year 2, only two plants survived. No plants emerged in year 3. In SRC, maximum height and width of surviving plants increased in year 2 (Fig. 25). Basal width (11.8 and 16.3 cm) increased over the 2-year period. In LC, maximum height and width decreased in year 2, while maximum basal widths remained similar (14.5 cm). For both sites, few inflorescences were present and occurred primarily in October, ranging from 90 to 167 cm in height. Quality ratings remained high (3.4–4.5) from July to October for both sites in year 1 (Fig. 25). In SRC, quality ratings remained high in year 2 from August to October but dropped to 2.8 in November (Fig. 25). In LC, visual ratings never exceeded 1.3 in year 2.
Often called purpletop, this native, perennial bunch grass is widespread in the eastern United States in areas receiving at least 76 cm of rain annually and that are not extremely cold (USDA, 2002). It is adaptable to a wide variety of soil conditions, and is often seen along roadsides. In Florida, purpletop tridens occurs throughout the state (Wunderlin and Hansen, 2004j) and can be found growing in pine and oak woods as well along roadsides. It has an erect, open, and tufted growth habit, with relatively wide blades (up to 1.3 cm wide) (Hall, 1978); typical height is 90 to 190 cm (USDA, 2002). In the fall, pyramid-shaped panicles of dark reddish purple spikelets extend well above the foliage. The showiness of these inflorescences in the wild is usually quite diminished against the background of shrubs and roadside turf.
Purpletop tridens was vigorous at both sites, with no mortality for all 3 years. At times, plant heights were greater in LC, with foliage width and basal width (26, 36, and 36 cm SRC; 29, 45, and 55 cm LC) greater in LC than in SRC for a majority of the trial period (Fig. 26). The foliage of purpletop tridens was dense and dark green in our trials, which differs from the sparse open nature described by Hall (1978). Spring flowering was present in SRC in years 1 (April to June) and 2 (May to July), while in LC, this occurred only in April year 2. Spring inflorescence height ranged from 18 to 83 cm. For both locations, flowering began in September/October all years and inflorescence heights ranged from 100 to 171 cm. Quality ratings were occasionally low in April (below 3) but remained high (3–5.8) from May to October for all years (Fig. 26).
Life cycle classifications for these species are based on observational data of native stands, and to be classified as perennial, especially a long-lived one, requires monitoring of the same plants for at least 3 years. Being a native, even from Florida, does not mean that all germplasm will survive a long time or be classified as a perennial when grown outside of a native habitat. For example, all of our observational data for bigtop lovegrass shows that most plants are annuals when grown in low-input landscapes, even though it is classified as a perennial (USDA, 2006f). Plants that live only 2 years (and are not biennials) are often classified as short-term perennials (R. Wunderlin, personal communication). Empirical data about landscape performance for an individual species, not to mention ecotypes, might be limited, thus life cycle classification with respect to landscape performance might be based on a congener or only a single ecotype. This would not be surprising for uncommon native species.
Poor performance of Florida native grasses in a low-input garden trial should not be surprising considering most non-USDA releases represent unimproved germplasm. Furthermore, a native propagated by division might be a genotype that does not have much longevity under landscape conditions. These accessions simply might not perform well under landscape conditions. For those natives that show promise as a potential landscape plant, starting with a genetically diverse seed lot should result in the identification of at least a few genotypes that would perform well over the long term.
Somewhat surprising was the lack of data indicating any weedy potential of ‘Red Baron’ cogongrass. In addition, no reversion to the invasive green form (Greenlee, 1992; MacDonald, 2004) was noted. However, ‘Red Baron’ cogongrass flowered in SRC and therefore could hybridize with the green form (Raymer and Thetford, 2002). Equally interesting was the lack of reseeding of two USDA releases (‘Alamo’ switchgrass and ‘Lometa’ indiangrass), both of which are commercially propagated via seeds (USDA, 1979, 1983, 2007d). However, the inherent seed production potential of the USDA releases indicates that they have the potential to spread beyond the landscape plantings. We suggest that landscape plantings of USDA releases be limited to urban environments to minimize potential genetic contamination of indigenous populations of the parent species.
Only six of the species and cultivars evaluated in this 3-year study showed potential as long-term landscape plants under low-input conditions in the coastal plain climate of the lower south (Table 1). Four of the six species were native, with 100% survival at both sites occurring only for purpletop tridens. The excellent performance of purpletop tridens was a pleasant surprise given that in the wild it has an open, airy growth habit with limited flowering, and is sometimes classified as a weed. However, under our low-input conditions, the daylily-like foliage was dense and disease free. The other three native species, gulf hairawn muhly, ‘Alamo’ switchgrass, and indian woodoats, exhibited limited plant mortality. That gulf hairawn muhly was rated high is not surprising given that it has proven to be a popular, reliable plant under garden conditions in the south over the past 5 to 10 years. As with purpletop tridens, indian woodoats performed much better than expected under our full-sun conditions. This species normally occurs under partial to full shade in slightly moist soils. However, because some yellowing of the foliage was observed during the summer, this species would be best suited in areas that are at least shaded during the afternoon.
Porcupine grass and ‘Cabaret’ silver grass were the only two nonnative species demonstrating potential as long-term landscape ornamentals based on their performance under our low-input landscape at both sites. Survival of porcupine grass was nearly 100% with loss of only a single plant at the end of year 3, while survival of ‘Cabaret’ silver grass was more site dependent, with survival of 100% in SRC and 66% in LC. Three of the four cultivars of miscanthus survived the entire evaluation period in SRC, while two cultivars suffered losses only in LC, demonstrating the importance of site characteristics on the long-term performance of individual species or cultivars of grasses.
‘Morning Light’ silver grass, coastal bluestem, and ‘Lometa’ indiangrass may be used reliably as short-term perennials or annuals as each survived at least 2 years at both sites. Although ‘Lometa’ indiangrass had good 2-year landscape performance, the inflorescences often lodged, resulting in lower visual ratings.
Chalky bluestem (OC), sand lovegrass, giant silver grass, and bamboo muhly may be used reliably as annuals as each survived 3 years at one site but only 1 year at the other site. This differential in longevity was site specific and demonstrated a marginal potential of these species as longer-term ornamentals in low-input landscapes.
Pineland threeawn, ‘Red Baron’ cogon grass, ‘Hameln’ fountain grass, black flowering fountain grass, and ‘Feesey Form’ ribbon grass were categorized as marginal for low-input landscapes and could only be considered short-term perennials under the conditions of this test (Table 1). Although these species and cultivars survived at least 2 years and were rated acceptable at times during the trial period, sustained season-long, acceptable ornamental quality was lacking in the second season. In the case of pineland threeawn, plant survival and growth were evident for all 3 years, but plants were rated as marginal because the ornamental quality of the plants was lacking unless inflorescences were present.
Bigtop lovegrass, silver plumegrass, and lopsided indiangrass were categorized as having excellent 1-year landscape performance and hence have potential as short-term ornamentals under low-input conditions and (Table 1). These species and cultivars sustained high quality ratings year 1 and should be considered for use as annuals in low-input landscapes. These species and cultivars continued to grow and flower in year 2 but quality ratings did not remain acceptable in the second season under our low-input conditions.
Chalky bluestem (SRC), black flowering sedge, ‘Heavy Metal’ switchgrass, and narrowleaf silkgrass were categorized as having good 1-year landscape performance. However, only chalky bluestem and narrowleaf silkgrass are recommended for use as an annual in low-input landscapes. These species and cultivars performed well in SRC for the entire first season and for a majority of the first season in LC. However, black flowering sedge and ‘Heavy Metal’ switch grass declined in quality later in the growing season at both locations, perhaps a reflection of their adaptation to moist soils.
ClewellA.F.1985Guide to the vascular plants of the Florida panhandleUniversity Press of FloridaTallahassee, FL
DarkeR.2004Pocket guide to ornamental grassesTimber PressPortland, OR
Florida Exotic Pest Plant Council2008Florida Exotic Pest Plant Council invasive plant lists7 July 2008<http://www.fleppc.org/list/list.htm>.
GilmanE.F.1999Pennisetum alopecuroides ‘Hameln’ Fact SheetFlorida Coop. Ext. Serv. Publ. FPS-461
GreenleeJ.1992The encyclopedia of ornamental grasses: How to grow and use over 250 beautiful and versatile plantsRodale PressEmmaus, PA
HallD.W.1978The grasses of FloridaUniv. of FloridaGainesvillePhD Diss.
HitchcockA.S.ChaseA.1950Manual of the grasses of the United StatesU.S. Dept. Agr. Misc. Publ. No. 200
NelsonG.2003Florida's best native landscape plants: 200 readily available species for homeowners and professionalUniversity Press of FloridaGainesville, FL
TobeJ.D.1998Florida wetland plants: An identification manualFlorida Dept. Environ. ProtectionTallahassee
U.S. Department of Agriculture1979Alamo switch grassU.S. Dept. Agr. Misc. Publ. No. L1774
U.S. Department of Agriculture1983Lometa indiangrassU.S. Dept. Agr. Misc. Publ. No. L-2065
U.S. Department of Agriculture2002Plant fact sheet: Purpletop Tridens flavus (L.) Hitchc10 Oct. 2007<http://plants.usda.gov/factsheet/doc/fs_trfl2.doc>.
U.S. Department of Agriculture2005Plant fact sheet: Sand lovegrass, Eragrostis trichodes (Nutt.) Wood10 Oct. 2007<http://plants.usda.gov/factsheet/pdf/fs_ertr3.pdf>.
U.S. Department of Agriculture2006aPLANTS profile: Andropogon capillipes Nash4 Sept. 2006<http://plants.usda.gov/java/profile?symbol=ANCA4>.
U.S. Department of Agriculture2006bPLANTS profile: Aristida beyrichiana Trin. & Rupr4 Sept. 2006<http://plants.usda.gov/java/profile?symbol=ARBE7>.
U.S. Department of Agriculture2006cPLANTS profile: Aristida stricta Michx4 Sept. 2006<http://plants.usda.gov/java/profile?symbol=ARST5>.
U.S. Department of Agriculture2006dPLANTS profile: Carex nigra (L.) Reichard4 Sept. 2006<http://plants.usda.gov/java/profile?symbol=CANI5>.
U.S. Department of Agriculture2006ePLANTS profile: Chasmanthium latifolium (Michx.) Yates4 Sept. 2006<http://plants.usda.gov/java/profile?symbol=CHLA5>.
U.S. Department of Agriculture2006fPLANTS profile: Eragrostis hirsuta (Michx.) Nees4 Sept. 2006<http://plants.usda.gov/java/profile?symbol=ERHI>.
U.S. Department of Agriculture2006gPLANTS profile: Eragrostis trichodes (Nutt.) Wood4 Sept. 2006<http://plants.usda.gov/java/profile?symbol=ERTR3>.
U.S. Department of Agriculture2007aPLANTS profile: Muhlenbergia capillaris (Lam.) Trin10 Oct. 2007<http://plants.usda.gov/java/profile?symbol=MUCA2>.
U.S. Department of Agriculture2007bPLANTS profile: Muhlenbergia filipes M.A. Curtis10 Oct. 2007<http://plants.usda.gov/java/profile?symbol=MUFI3>.
U.S. Department of Agriculture2007cPLANTS profile: Muhlenbergia dumosa Scribn. ex Vasey10 Oct. 2007<http://plants.usda.gov/java/profile?symbol=MUDU3>.
U.S. Department of Agriculture2007dPLANTS profile: Panicum virgatum L10 Oct. 2007<http://plants.usda.gov/java/profile?symbol=PAVI2>.
U.S. Department of Agriculture2007ePLANTS profile: Phalaris arundinacea L6 Nov. 2007<http://plants.usda.gov/java/profile?symbol=PHAR3>.
U.S. Department of Agriculture2007fPLANTS profile: Pityopsis graminifolia (Michx.) Nutt9 Nov. 2007<http://plants.usda.gov/java/profile?symbol=PIGR4>.
U.S. Department of Agriculture2007gPLANTS profile: Saccharum alopecuroides (L.) Nutt10 Oct. 2007<http://plants.usda.gov/java/profile?symbol=SAAL21>.
U.S. Department of Agriculture2007hPlant fact sheet: Gulf bluestem Schizachyrium maritimum (Chapman) Nash10 Oct. 2007<http://plants.usda.gov/factsheet/pdf/fs_scma3.pdf>.
U.S. Department of Agriculture2008PLANTS profile: Sorghastrum secundum (Elliot) Nash13 Mar. 2008<http://plants.usda.gov/java/profile?symbol=SOSE5>.
WunderlinR.P.HansenB.F.2004aAtlas of Florida vascular plants: Andropogon virginicus var. glaucus4 Sept. 2007<http://plantatlas.usf.edu/main.asp?plantID=2414>.
WunderlinR.P.HansenB.F.2004bAtlas of Florida vascular plants: Aristida beyrichiana. FL4 Sept. 2007<http://plantatlas.usf.edu/main.asp?plantID=698>.
WunderlinR.P.HansenB.F.2004cAtlas of Florida vascular plants: Chasanthium latifolium4 Sept. 2007<http://plantatlas.usf.edu/main.asp?plantID=2109>.
WunderlinR.P.HansenB.F.2004dAtlas of Florida vascular plants: Eragrostis hirsuta4 Sept. 2007<http://plantatlas.usf.edu/main.asp?plantID=817>.
WunderlinR.P.HansenB.F.2004eAtlas of Florida vascular plants: Muhlenbergia capillaris var. filipes11 Oct. 2007<http://plantatlas.usf.edu/main.asp?plantID=2308>.
WunderlinR.P.HansenB.F.2004fAtlas of Florida vascular plants: Panicum virgatum11 Oct. 2007<http://plantatlas.usf.edu/main.asp?plantID=3532>.
WunderlinR.P.HansenB.F.2004gAtlas of Florida vascular plants: Saccharum alopecuroides11 Oct. 2007<http://plantatlas.usf.edu/main.asp?plantID=3537>.
WunderlinR.P.HansenB.F.2004hAtlas of Florida vascular plants: Schizachyrium scoparium11 Oct. 2007<http://plantatlas.usf.edu/main.asp?plantID=2218>.
WunderlinR.P.HansenB.F.2004iAtlas of Florida vascular plants: Sorghastrum secundum11 Oct. 2007<http://plantatlas.usf.edu/main.asp?plantID=2407>.
WunderlinR.P.HansenB.F.2004jAtlas of Florida vascular plants: Tridens flavus<http://plantatlas.usf.edu/main.asp?plantID=3403>.
WunderlinR.P.HansenB.F.2004kAtlas of Florida vascular plants: Pityopsis graminifolia9 Nov. 2007<http://plantatlas.usf.edu/main.asp?plantID=524>.