Pruning of Meadowsweet and Steeplebush Landscapes, It Is Important to Know How to Manage Them for Optimal Kelly M
If they are to be used in more formal Pruning of Meadowsweet and Steeplebush landscapes, it is important to know how to manage them for optimal Kelly M. Stanton1, Sally S. Weeks2, Michael N. Dana1, and growth, form, and flowering. 1,3 Meadowsweet and steeplebush Michael V. Mickelbart produce summer flowers from new growth. Catchpole (1963) gives rec- ADDITIONAL INDEX WORDS. Spiraea alba, Spiraea tomentosa, hardhack, native plants, ommendations for pruning japanese growth, flowering spirea (Spiraea japonica), margitae spirea (Spiraea ·margaritae), and SUMMARY. Two native shrubs, meadowsweet (Spiraea alba) and steeplebush menzie’s spirea (Spiraea douglasii Spiraea tomentosa ( ), are currently grown for use in habitat restoration. Pruning var. ), all of which have could improve their form and encourage use in more formal gardens. Two seedling menziesii populations of each species were grown in the field or in containers outdoors. In similar growth and flowering habits Mar. 2009, plants were pruned to 3 or 15 cm or were left unpruned. By to meadowsweet and steeplebush. midsummer, there was no effect of pruning on plant height in field-grown plants. Catchpole (1963) recommended that However, pruning did eliminate lateral growth in field plants and, therefore, plants be pruned to ground level in improved overall form of the plants. Meadowsweet pruned to 3 cm had about half as April to promote the growth of new, many inflorescences as did meadowsweet pruned to 15 cm, with unpruned plants flower-bearing shoots. If they are not producing an intermediate number of inflorescences. Pruning container plants to 3 pruned, they accumulate dead wood, cm resulted in plants that were about half the height of unpruned plants, and total and flowering is reduced. Pruning biomass was greatly reduced. In general, plants that were pruned had fewer them to half-height still encourages inflorescences, although the total number of inflorescences in all plants was small. new growth but allows the develop- Pruning resulted in lower quality plant form in container plants. Species and seed sources within species responded similarly to pruning. Based on the data collected in ment of a larger plant. To date, this study, newly planted meadowsweet and steeplebush should be pruned to however, there are no similar pub- achieve good form in the first year. Despite the sometimes leggy growth habit in lished reports or recommendations containers, these species should not be pruned before growing out in the spring, or on pruning strategies for meadow- maximum growth and good form will not be attained. sweet or steeplebush. Pruning could eliminate apical dominance, resulting in shrubs with ative plants are a potential wide. They produce long-lasting ter- more branches and thus a fuller, more source of new plants for man- minal inflorescences on their vertical attractive appearance. Because the Naged landscapes. There is in- stems from mid-June (meadowsweet) inflorescences of meadowsweet and terest in gardening with natives not and mid-July (steeplebush) through steeplebush form on the stem tips, only for their novelty, but also to take August in the upper midwestern U.S. increasing the number of branches advantage of their adaptation to local (Swink and Wilhelm, 1994). Mead- could also increase the number of climates, to reflect regional differ- owsweet produces white flowers and flowers. ences, and to foster a connection with steeplebush has pale pink flowers. Meadowsweet and steeplebush nature (West, 2001). Native plants Both species send up numerous basal can tolerate mowing (White, 1955). can also be grown as an alternative stems, with little branching. Mead- In Quebec, when established plants to potentially invasive exotic land- owsweet is thicket forming (Meilleur of steeplebush were mown before scape plants. Consumers are willing et al., 1997). Meadowsweet has also spring growth was initiated, they to pay premium prices for an attrac- been found to be highly attractive to grew 30 cm by late July (Hay, tive garden featuring native plants insects that prey upon or parasitize 1958), indicating that these species (Helfand et al., 2006). insect herbivores (Fiedler and Landis, might respond well to severe pruning. The shrubs meadowsweet and 2007). These species tend to have asym- steeplebush are native to eastern North Steeplebush was used in the land- metrical, lateral growth in the nurs- America, and can be found growing scape as early as 1736 (Symes, 1983), ery. If these plants are to be sold in from the Gulf States into Canada (U.S. and meadowsweet and steeplebush retail nurseries, this appearance would Department of Agriculture, 2007). have potential as landscape plants. likely deter consumers from purchase. Meadowsweet and steeplebush grow Today, both species are used pri- If pruning after 1 year could im- to be 1 to 1.5 m tall and equally as marily in habitat restoration and prove plant appearance, these spe- less intensively managed areas where ciesmightbemoresuitableforretail We thank JFNew and Possibility Place Nursery for growth form is relatively unimportant. sales. Likewise, recommendations for supplying plants. Rob Eddy and Daniel Hahn assisted with container plant growth and maintenance. Jay Young and Nathan Linder assisted with setup and plot maintenance in the field. Mike Gosney provided Units technical advice and help. Meghan Honerlaw explained To convert U.S. to SI, To convert SI to U.S., statistics and SAS. multiply by U.S. unit SI unit multiply by 1 Department of Horticulture and Landscape Archi- 0.3048 ft m 3.2808 tecture, Purdue University, West Lafayette, IN 47907 3.7854 gal L 0.2642 2Department of Forestry and Natural Resources, 2.54 inch(es) cm 0.3937 Purdue University, West Lafayette, IN 47907 16.3871 inch3 cm3 0.0610 3Corresponding author. E-mail: mickelbart@purdue. 28.3495 oz g 0.0353 edu. ( F – 32) O 1.8 F C(1.8· C) + 32
700 • August 2010 20(4) early development of plant form in were transplanted into 1.5-L con- July 2009 and was dried to a constant the landscape might make these tainers on 11 Oct. 2008. The con- weight at 60 C for determination of plants more attractive to landscapers tainers were overwintered outside dry weight. and homeowners. To determine suit- under mulch to a depth of at least DATA ANALYSIS. Data were ana- able pruning-management strategies 5 cm above soil level. The plants were lyzed using SAS (version 9.1; SAS for young meadowsweet and steeple- pruned 3 cm above media level or left Institute, Cary, NC). The significance bush, plants in containers and the field unpruned as a control. Pruning was of the differences among the treat- were pruned to various heights, and done on 24 Mar. 2009 immediately ments or seed sources was determined the resulting growth, form, and flow- after removing mulch from the con- from analysis of variance and Tukey ering of the plants were measured. tainers. Two weeks after pruning, 7 g tests were conducted with PROC of the 22N–2.2P–5K fertilizer (1.5 g GLM. The significance of the effects Materials and methods of N) was applied to each container. of the treatments on the frequencies of FIELD EXPERIMENT. All plants The plants were irrigated by hand as form rating was determined by using used in the experiments were supplied needed. Fisher’s exact test in PROC FREQ. as 15-inch3 plugs grown by commer- The experiment was a two-way cial nurseries from seeds germinated factorial arrangement on a random- Results in Spring 2008. Seeds of both species ized complete block design with 12 FIELD EXPERIMENT. Meadow- were collected by the nurseries in Will blocks. Each block contained two sweet and steeplebush produced ir- County, IL, and Jackson County, MI pruning treatments applied to plants regular and lateral growth in the (hereafter IL and MI, respectively). of both seed sources of both species. crowded conditions of the plug trays. Both seed sources of each species were MEASUREMENTS. Growth, flow- However, when plugs were planted used in the container experiments, but ering, and form ratings were collected in the field, most of the new growth only the IL source of steeplebush was for both experiments from 2 to 4 June produced in Summer 2009 was verti- used in the field experiment. and 20 to 24 July 2009. By early June, cal. Pruning to 15 or 3 cm did not The experiment was conducted most of the plants had produced affect the midseason height of plants at the Meigs Farm of the Throckmor- vigorous spring growth, but few had from different seed sources of either ton Purdue Agricultural Center in begun flowering. July measurements species (Table 1). However, because Tippecanoe County, IN (lat. 40.29 N, were conducted when plants had much of the lateral growth was re- long. 86.88 W). The soil type was attained their full size for the growing moved with pruning, the width of Drummer (silty clay loam) with a pH season and were blooming. pruned plants was less than unpruned of 6.9. Height was determined as the plants (Table 1). More of the pruned Plugs were planted on 28 Oct. vertical distance from the soil (field) plants were characterized as symmet- 2008, and they were pruned on 24 or media (container) level to the tall- rical compared with plants that were Mar. 2009. On 8 Apr. 2009, they est living part of the plant. For the not pruned (Table 2). Although the were fertilized with 12 g of 22N– field experiment, plant width was plants were smaller, they had a more 2.2P–5K 4- to 5-month controlled- measured in north-south and east- compact and full appearance. release fertilizer with micronutrients west axes, and the average was taken. The young plants used in this (2.6 g of N; Scotts, Marysville, OH). Inflorescences on container- study had few inflorescences over the They were irrigated with drip tape as grown plants were counted in July, growing season, and there was great needed throughout the summer. before biomass collection. All inflo- variability within treatments. Pruning Weeds were controlled mechanically. rescences, including those that were did not significantly affect the num- The experiment was designed as post-anthesis, were counted in Sept. ber of inflorescences produced by a two-way factorial arrangement on 2009 to determine the total number plants from any individual seed source a randomized complete block. Each of inflorescences produced during the (Table 1). However, when the data of eight blocks contained one plant of growing season. from the two meadowsweet seed each species, seed source, and prun- The quality and form of each sources were combined, the plants ing treatment combination. The plant was rated in June 2009. Plants pruned to 15 cm produced almost pruning treatments were: unpruned were rated on a scale of 1 to 4 accord- twice as many inflorescences as the (control), pruned to 15 cm above- ing to size (small or large plants) and plants pruned to 3 cm, with the un- ground level, and pruned 3 cm above- symmetry (symmetrical or asymmetri- pruned plants producing an interme- ground level. The field experiment cal). The scale used to rate the plants diate number of inflorescences (P £ was planted in two rows separated was categorical, but the evaluations 0.0143). Plants of the IL seed source by 12 ft with plants in each row were based on typical nursery stan- of meadowsweet were shorter (P £ spaced 4 ft apart. dards for deciduous shrubs (American 0.0001) and produced more inflores- CONTAINER EXPERIMENT. Plants Nursery and Landscape Association, cences (P £ 0.016) than plants of the of the MI seed source of both species 2004), with 1 = an asymmetrical plant MI seed source. were transplanted into 1.5-L con- with few shoots, 2 = a symmetrical CONTAINER EXPERIMENT. Plants tainers on 28 Aug. 2008 containing plant with few shoots, 3 = a small of all species and seed sources in Premier Pro-Mix Bx/Mycorise soil- asymmetrical plant with vigorous containers that were pruned to 3 cm less media (pH 5.2–6.2; Premier shoots, and 4 = a large symmetrical produced less biomass than unpruned Horticulture, Rivie`re-du-Loup, QC, plant with vigorous shoots. plants after about 4 months of growth Canada). Plants from the meadow- The aboveground biomass of (Table 3). By July, the height of sweet and steeplebush IL seed sources container plants was collected on 24 pruned and unpruned plants of the
• August 2010 20(4) 701 RESEARCH REPORTS
Table 1. Size and flowering of field-grown meadowsweet and steeplebush pruned the more asymmetrical growth of to 3 or 15 cm aboveground or left unpruned (control) in Mar. 2009 (n = 8). container plants compared with field Plants were grown from seed sources from Illinois (IL) and Michigan (MI). plants (Tables 2 and 4), width was not Pruning Ht Width Inflorescences measured in these plants. Seed source treatmentz (cm)y (cm)x (no.)w Discussion Meadowsweet Plants from two seed sources of IL 3 cm 40 26 bv 5.2 meadowsweet and steeplebush were 15 cm 38 43 a 13.5 grown in containers and in the field Control 42 39 a 11.3 and were pruned to various heights P valueu NS ** NS to determine the effects of early MI 3 cm 60 23 b 7.4 spring pruning on growth, form, 15 cm 58 32 ab 11.4 and flowering. The growth and form Control 52 37 a 7.4 of control plants in these experiments was, in large part, due to their growth P value NS * NS patterns in the plug trays in which Steeplebush they were produced in 2008. Growth IL 3 cm 29 20 b 7.0 of meadowsweet and steeplebush in 15 cm 33 27 b 8.1 plug trays is often asymmetric and not Control 32 38 a 6.1 upright (personal observations of the P value NS *** NS authors and several growers). This may be due to poor light penetration, leading to weak stems that lodge particularly when inflorescences are Table 2. The percentage in each form rating category in June 2009 of field-grown formed. Production of fuller, more meadowsweet and steeplebush plants pruned to 3 or 15 cm aboveground or left symmetrical plants in containers and unpruned (control) in Mar. 2009 (n = 8). Plants were grown from seed sources in the landscape may lead to greater from Illinois (IL) and Michigan (MI). Plants were placed in one of four categories according to size and symmetry. retail sales and landscape use. When young meadowsweet and Small Large steeplebush plants were pruned to 15 Seed Pruning Asymmetrical Symmetrical Asymmetrical Symmetrical cm before planting in the field, much source treatmentz Proportion in each form category (%) of the initial asymmetric and lateral Meadowsweet growth was eliminated, resulting in IL 3 cm 33 33 0 33 Ax a more symmetrical plant at the be- 15 cm 0 0 13 88 B ginning of the growing season. The Control 0 13 63 25 C irregular shape of the previous sea- P valuey ** son’s growth was also eliminated when the plants were pruned to 3 MI 3 cm 0 80 0 20 A cm, but many basal, vegetative buds 15 cm 0 14 57 29 AB from which new growth could have Control 25 0 75 0 B emerged were also removed. P value ** Plants pruned to 3 cm in con- tainers were smaller than unpruned Steeplebush plants throughout the growing sea- IL 3 cm 29 43 0 29 A son. This may be because the con- 15 cm 13 0 13 75 A tainer plants were grown in an area Control 50 0 50 0 B with only morning full sun, because P value ** we observed that shading can reduce z1 cm = 0.3937 inch. the growth of meadowsweet and stee- yThe significance of the interaction between pruning treatment and form. The data were analyzed using Fisher’s plebush. In the container experiment, exact test to determine if there were differences among the pruning treatments in the frequencies of each form only plants of the IL seed source of rating. ** indicates significance at P £ 0.01. xFor each seed source, different letters indicate significantly different pruning treatments determined from pairwise meadowsweet grew to the same comparisons of the pruning treatments (Fisher’s exact test P £ 0.05). The letters do not indicate ranking because height as control plants in about 3 the test does not indicate in what way the pruning treatments differ. months (Table 3). Plants in the field experiment were in full sun, and IL seed source of meadowsweet was number of inflorescences even further growth in the pruned meadowsweet similar, but the height of pruned in plants of the MI and IL seed was sufficient to eliminate treatment plants of the other seed sources of sources of meadowsweet and steeple- height differences by July. meadowsweet and steeplebush was bush, respectively (Table 3). The Flowers compete with vegetative about half that of unpruned plants container-grown plants tended to be growth for carbohydrates. Therefore, (Table 3). Plants of all seed sources asymmetrical, but the unpruned reproductive and vegetative growth of- produced very few inflorescences in plants tended to be larger than the ten are negatively correlated (Kozlowski containers, and pruning reduced the pruned plants (Table 4). Because of and Pallardy, 2002). Despite the pruned
702 • August 2010 20(4) Table 3. Size and inflorescence production in July 2009 in container-grown flowering, this could explain why the meadowsweet and steeplebush plants that were pruned to 3 cm above media level meadowsweet pruned to 3 cm had the or left unpruned (control) (n = 12). Plants were grown from seed sources from fewest inflorescences. Illinois (IL) and Michigan (MI). However, the removal of apical Seed source Pruning treatmentz Ht (cm)y Biomass (g)x Inflorescence (no.)w dominance through pruning can cause a single shoot to produce mul- Meadowsweet tiple new flowering branches in pri- IL 3 cm 42 13 0.8 mocane-fruiting blackberries (Rubus Control 49 30 1.6 spp.) (Thompson et al., 2007). Mead- P valuev NS *** NS owsweet produces flowers in a panicle MI 3 cm 45 16 0.8 on the current season’s growth. An Control 88 51 4.0 increase in the number of shoots due P value ** *** *** to the removal of apical dominance could explain why the moderately Steeplebush pruned meadowsweet had the most IL 3 cm 26 4 0.7 inflorescences. Control 56 24 4.7 The recommendation that spirea P value *** *** ** (Spirea spp.) that bear flowers on new growth be pruned to the ground to MI 3 cm 27 4 1.2 promote new growth and eliminate Control 46 11 3.2 old dead wood (Catchpole, 1963) P value *** ** NS may apply more to older, larger plants z1 cm = 0.3937 inch. than those used in these experiments. yVertical distance from the soil surface to the highest part of the plant. Thus, the plugs used in the field xDry weight of the aboveground biomass; 1 g = 0.0353 oz. experiment all produced new growth, wThe total number of inflorescences produced through July 2009. vProbability values indicate differences among the pruning treatments within a seed source. NS, **, *** indicate not regardless of treatment, while having significant or significant at P £ 0.01 or 0.001, respectively. little old wood to be eliminated. Plants that have been growing in the Table 4. The percentage of container-grown meadowsweet and steeplebush landscape for several years may have plants in each pruning treatment that were assigned to each form rating category more shoots, dead wood, and seed in June 2009 (n = 12). Plants were grown from seed sources from Illinois (IL) heads from previous seasons, leading and Michigan (MI). Plants were placed in one of four categories according to size to different pruning requirements and symmetry, and the percentage of plants placed in each category for a given than plants in their first season in the seed source is shown. landscape. Small Large Although there were significant differences in size and flowering be- Asymmetrical Symmetrical Asymmetrical Symmetrical Seed Pruning tween plants of the meadowsweet source treatmentz Proportion in each form category (%) seed sources, it was still difficult to Meadowsweet tell the seed sources apart. The differ- IL 3 cm 67 8 25 0 ences were small enough, and the Control 8 0 92 0 variance was great enough that they **y appeared similar, especially when compared across pruning treatments. MI 3 cm 67 0 33 0 In conclusion, pruning of mead- Control 0 0 100 0 owsweet and steeplebush can im- ** prove form. Most healthy plants recover from being pruned to 3 cm, Steeplebush but we do not recommend this in the IL 3 cm 78 11 0 11 first year after planting because it Control 0 0 91 9 significantly reduces size and flowering. ***
MI 3 cm 83 8 8 0 Control 62 8 31 0 Literature cited NS American Nursery and Landscape Associ- z1 cm = 0.3937 inch. ation. 2004. American standard for nurs- yThe significance of the interaction between pruning treatment and form. The data were analyzed using Fisher’s ery stock ANSI Z60.1-2004. Amer. Natl. exact test to determine if there were differences between the pruning treatments in the frequencies of each form Standard. Inst., Washington, DC. rating. NS, **, and *** indicate not significant or significant at P £ 0.01 or £ 0.001, respectively. Catchpole, N. 1963. Flowering shrubs plants starting the growing season If the most severely pruned meadow- and small trees, 2nd ed. Collingridge, shorter than the unpruned plants, by sweet eliminated this initial height London. midsummer, there were no height difference by directing carbohydrates Fiedler, A.K. and D.A. Landis. 2007. differences in the field-grown plants. toward vegetative growth rather than Attractiveness of Michigan native plants
• August 2010 20(4) 703 RESEARCH REPORTS
to arthropod natural enemies and herbi- Meilleur, A., H. Ve´ronneau, and A. patterns of primocane-fruiting black- vores. Environ. Entomol. 36:751–765. Bouchard. 1997. Shrub propagation berries. HortScience 42:1174–1176. techniques for biological control of in- Hay, J.R. 1958. Grass and legume crops. vading tree species. Environ. Manage. U.S. Department of Agriculture. 2007. Proc. Eastern Section Natl. Weed Com- 21:433–442. Spiraea L. The PLANTS Database. 7 mittee (Canada) 11:43–54. Nov. 2007.
704 • August 2010 20(4)