Page  57



Katherine Moesker RoyalBotanical Gardens 680 Plains RoadWest Burlington, Ontario L7T 4H4


MagnoliakobusDC., also known as Kobushi Magnolia, Northern Japanese Magnolia, or Kobus Magnolia, is a medium-sized deciduous tree native to the temperate forests of Japan. It is a popular choice in horticulturallandscapes, due in partto its showy white flowers and attractive foliage. Some ofthe characteristics that makethis species desirable for gardeners arethe samefeatures thathaveallowedfor thedevelopment ofits highinvasivepotentialinnative ecosystems. Magnoliakobusisvery tolerant of a variety of climate and soil conditions, experiences little threat from disease and insect pests, and contains allelopathic compounds that can hinder the growth of surrounding native vegetation. As a result, once the plant becomes established in a manicured landscape or in natural areas, it can thrive relatively unhindered unless it is intentionally targeted for removal. This study researched the life history and invasive characteristics of M. kobus and the potential impacts of its invasion throughout the natural lands in the Royal Botanical Gardens in Hamilton, Ontario, Canada. Hypotheses for the recent establishment and spread of M.kobuswere also developed and are discussed in this study. The population of M.kobusat Royal Botanical Gardens is the first documented occurrence of a naturalized population in Canada.

KEYWORDS: Magnoliakobus, invasive species, invasion lag time, species adaptability, climate change


The Royal Botanical Gardens (RBG) in Hamilton and Burlington, Ontario, Canada contains over 900 hectares of protected natural lands that are at a high risk of invasive plant introductions due in part to their close proximity to horticultural collections. One example of an invasion by a non-native plant species is the spreading of MagnoliakobusDC. beyond the area in which individuals were planted. Several trees of M.kobusare located within the Magnolia Collection in the RBGArboretum, the majority of which were plantedin the 1960s and1970s. In 2013, several thousand M.kobusseedlings were discovered throughout Hickory Valley, with the invasion radiating out from the edge of the Magnolia Collection into the adjacent forested ravines. As a result of intermittent removal efforts and the continual sprouting of new seedlings, a more specific quantification of seedlings was not feasible.

Magnolia kobus is recognized as an invasive species in the State of Delaware (McAvoy 2016), and naturalized populations have been observed in the southeastern Pennsylvania counties of Chester, Montgomery, and Bucks. In


2016, the first known record of a naturalized Magnolia kobus population in western Pennsylvania was reported in Butler County (Krayesky and Chmielewski 2016).The naturalization of this species is not yet recordedin authoritative sources, including the Invasive PlantAtlas of the United States (IPA 2018), Flora of North America North of Mexico (FNANM) (Flora of North America Editorial Committee 1993+), and Flora Ontario Integrated Botanical Information System (FOIBIS) (Newmaster and Ragupathy 2012), nor in the FieldManualofMichiganFlora(Voss and Reznicek 2012). The population of

M. kobus at RBG is the first documented occurrence of a naturalized population in Canada. This report examines the native range and habitat type of Magnoliakobus, its pollination ecology, and its invasive characteristics. The research provided insight for the development of hypotheses with respect to invasion and potential impacts of invasion, which are also discussed. The information is intended to be used as a reference for the development of a management strategy for M.kobus and to increase awareness of the high invasive potential of this species.

Study Species

Magnoliakobusis a medium-sized deciduous tree with 4to 8inch leaves that are simple, obovate, and abruptly acuminate. One of the key identifying features of this species is the presence of long hairs scattered along the sides of the veins and/or in the axils of the major lateral veins on the underside of the leaf (Figure

1) (Cullen et al. 2011). The buds on the tips of the twigs are covered with fine, silky hair. The twigs themselves have a strong sweet smell when broken. The white flowers are ca. 10 cm across and appear in the spring prior to leafing out (Gilman and Watson 1994).Young M. kobus trees do not flower, as this species takes about 15 years before it blooms for the first time (Randy Stewart Landscape Designs 2010). The pink fruit is an aggregate of follicles called a follicetum (see Figure 2). The individual follicles usually contain one or two bright red seeds, and the total number of seeds contained within the follicetum ranges from 2 to 60. When the seeds are ripe, the follicles split open and the fleshy seeds are suspended on a thin string-like membrane called a funiculus (Barbour 2008).

Birds are attracted to the brightly coloured seeds of Magnolia kobus, the appeal of which is accentuated by the funiculus, which allows the seeds to dangle from the follicetum (Barbour 2008). Predation by mammals, such as mice and chipmunks, causes a reduction in seed survival. Since mammals are able to collectseedsthatfalltotheground, itisadvantageousfortheplanttohavetheseeds eaten by birds while they remain on the tree (Stiles 1980). In this way, M.kobus has a means of ensuring the dispersal of its seeds.

Magnolias are dichogamous (Lloyd and Webb 1986), that is, the pistils and stamens mature at different times, and most Magnolia species, including Magnoliakobus, areprotogynous(Sedgley andGriffin,1989).When theflowersfirst open, the female organs are mature and receptive.After a brief period, the flowers close, and the female phase ends. The flowers reopen when the male organs are mature and able to shed pollen (Rose and Dosmann 2014).


FIGURE 1. Underside of a leaf of Magnolia kobus showing long hairs along the edge of the veins and/or in the axils of the major lateral veins.

FIGURE 2. The fruit of Magnoliakobus, an aggregate of follicles called a follicetum. Each fruit contains 2–60 bright red seeds.


Magnoliaspecies rely on beetles for pollination, and beetles in the family Nitidulidae are the main pollinators (Fulcher and White 2012). The flowers do not containnectar, butinsteadsecreteafragrantsugarysubstancethat attracts theinsects. When beetles feed on newly-openedflowers in the spring, they land on the stigmas that are located in a central column above the stamens. As the insects crawldownintotheflowerinsearchoffreshpollen, anypollen grainsfromother flowers in the male phase that are attached to them are removed by the sticky stigmas (Essig 2015).

Magnolia kobus is native to temperate forests of Japan and is commonly found on all four of the main Japanese islands of Hokkaido, Honshu, Shikoku, and Kyushu. It grows in hills and piedmont regions (Ohwi 1965), which are defined as the areas at the base of mountains (National Geographic 2016). The species is tolerant of a variety of soil conditions, including clay, loam, and sand. It prefers well-drained soils that are slightly alkaline to acidic (Gilman andWatson 1994). Rich and moist forest soils with leaf litter are ideal for the germination of Magnoliaspecies (United States Department ofAgriculture 1948).


Site Description

The “MagnoliaWest” collection of Magnoliakobusin the RBG arboretum, above HickoryValley where this study was conducted, was established in 1970, although the species had been present elsewhere at RBG since at least 1958. Although RBG records for that period do not indicate when plants were moved from propagation or holding sites to their permanent locations, it is probable that about five flowering-size plants of M. kobus were planted at the site above Hickory Valley in 1970. A later addition, planted in 2014, was the cultivar M. kobus ‘Edward A. Kehr.’ Magnolia × loebneri Kache, a hybrid between M. kobus and M. stellata (Siebold and Zucc.) Maxim., is also represented in the collection, having been planted between 1970 and 1995. Cultivars include ‘Merrill,’ ‘Leonard Messel,’ and ‘Ballerina’ (J. Peter, personal communication; RBG plant records).

Several thousand small seedlings are located throughout Hickory Valley (Figures 3 and 4), but specific quantification was not feasible, as seedlings were removed by RBG staff throughout the duration of this study. As of 2015, the majority of the individuals of Magnoliakobus were 1 or 2-year old seedlings, but there were a few trees that were approximately ten years old. A larger individual that was cut down in Hickory Valley was in its 19th growing season. (Voucher specimen: Aug. 17, 2015, Moeskers.n; HAM, accession no. 59122).

In 2014, several Magnoliakobustrees were girdled by RBG staff, since manual removal was not possible due to their large size.These efforts were conducted prior to the beginning of this study, and I observed the results of the girdling the following year. The results and consequential implications are discussed in this study.

Ecological Land Classification

Ecological Land Classification (ELC) is a system that delineates natural regions based on ecological factors including bedrock, climate, physiography, and vegetation (Government of Ontario 2019). RBG utilizes this system to classify habitat types within its natural area to identify rare habitats and potential locations of flora and fauna of interest. ELC has also been used to build upon RBG’s inventory of plant species within the nature sanctuaries, and to create distribution maps of invasive non-nativeplantsforthepurposeofinvasive species management.Theresultisamapofsmall polygons delineating ecologically distinct areas throughoutthe property.As of2019, ELCis ongoing within RBG’s nature sanctuaries.

The soil analysis portion of ELC is conducted by RBG staff and follows the protocol outlined in the FieldManualforDescribingSoilsinOntario(Denholm and Schut 2009).This analysis includes


FIGURE 3. Seedlings of Magnoliakobus.Several thousand have been discovered throughout RBG’s natural lands.

testing soil moisture, drainage, depth of mottles or gley, presence and depth of carbonates, and effective soil texture.

For this study, I analyzed the data associated with the ELC polygons in which Magnolia kobus was found, focusing on the vegetation list and the results of the soil analysis for each. I used this informationtocomparesimilarities between the species’nativerange andtheenvironmentinwhichthe M.kobusseedlings were discovered. I also used the data to develop hypotheses for the invasive ability of this species.

Weed Risk Assessment System

Objective and reliable systems are required in order to better predict the invasive potential of a species prior to its introduction into an area (Pheloung et al. 1999). One such tool is theWeed Risk Assessment (WRA) system, developed by Dr. Paul Pheloung of the Western Australian Department ofAgriculture (Australian Government Department ofAgriculture 2015).This system evaluates species’ properties such as climate requirements, naturalization in other regions, domestication history, and other ecological factors. Each property has an associated value, and the sum of the values associated with all the properties assessed provides a score that is used to assess the in


FIGURE 4. Fast-growing seedlings of Magnoliakobuscreating a dense wall near the edge of the natural area.

FIGURE 5. Sample Weed Risk Assessment scoring sheet.


vasive potential of a plant species, the impacts of an invasion, and the consequential acceptance or rejection of introduction (Koike and Kato 2006). A score of less than one indicates a low risk of invasion, and the plant may be accepted for introduction.A score of one to six poses a possible risk of invasion, and further evaluation is necessary. Finally, a score of six or greater suggests a high risk of invasion, and rejection of the species is recommended. The Weed RiskAssessment scoring sheet provides instructions for determining the score for each property, and is displayed in Figure

5. The first column of the scoring sheet consists of the letters A, E, and C, representing “agricultural”, “environmental”, and “combined”, respectively. This refers to the type of ecosystem that is likely to be affected by the species assessed (Australian Government Department of Agriculture 2015). This latter aspect of the assessment was not related to the context of this study and is therefore not discussed further. The WRA can be adapted for use in regions outside ofAustralia, as its accuracy in the identification ofinvasive plant species remainshighbeyondits countryoforigin(Gordon etal. 2008).When completing the risk assessment of Magnolia kobus for this study, the majority of the questions remained unaltered, but I modified three that pertain to the Australian climate in order to reflect the conditions that exist in Southern Ontario. Question 2.01 had originally read “Species suited to Australian climates” and was altered to “Species suited to Southern Ontario climates”. Question 2.04 read “Native or naturalized in regions with extended dry periods” and was edited to read “Native or naturalizedin regions with mean annualprecipitation of 40–50inches”, referringto the mean annual precipitation in Hamilton, Ontario of 46.5 inches (The Weather Network, 2016). Finally, question

8.05 had read “Effective natural enemies present inAustralia” and was changed to “Effective natural enemies present in Ontario”. INVASIVECHARACTERISTICS

Lack of Predators

Magnoliaspecies face few threats from pests anddiseases.The plants contain compounds that have antimicrobial, nematicidal, and insecticidal properties (Fulcher and White 2012). Magnolia kobus contains several lignans, including kobusin and sesamin, whose biological activities protect the plant from insect pests (Kamikado et al. 1975;Trifunovi.. et al. 2003).The ability of this species to deter insect pests by chemical means increases its ability to overcome ecological barriers as it invades natural areas.

In its native range, Magnoliakobusis a host plant for Caloptiliamagnoliae and Gibbovalva kobusi, both of which are moths in the family Gracillariidae (De Prins and De Prins 2019). It is also host to a variety of scale insects in the genera Eulecanium, Pulvinaria, and Pseudaulacaspis (García Morales et al. 2019). Magnolia scale (Neolecanium cornuparvum) is a pest that is managed within the Magnolia Collection at RBG, and some leaf miner damage has also been observed on the leaves of some of the Magnolia species, including M. kobus. However, these pests do not appear to pose a significant threat to the species.

The Magnolia kobus seedlings observed in the natural lands did not exhibit evidence of deer browse, despite the presence of a large local population of white-tailed deer. This suggests that these plants are not a preferred source of food for deer. The M.kobus seedlings are able to grow uninhibited, while many native tree and shrub species are continually stunted. This can result in the species having a competitive advantage over other plants that face pressures


from deer browsing, including witch hazel (Hamamelis virginiana L.) and dogwood species (Cornusspp.).

Hardy and Vigorous Grower

In 1894, about 30 years after Magnolia kobus was introduced to North America, it was described by American botanist Charles Sprague Sargent as being “the hardiest, most vigorous, and most rapid growing of all Magnolias” in New England (Sargent 1894). The average current year’s growth of five M. kobus trees in Hickory Valley was 74.6 cm (29.37 in). These trees were approximately 10 years old, based on the number of tree rings counted in the cross-section of Magnoliatrees that were of similar size.A plant that grows 12 inches or less per year is considered to be slow-growing, 12–24 inches per year is a medium growth rate, and above 25 inches per year is considered fast- growing (Arbor Day Foundation 2015). Based on these measurements, the rate of growth designates this species as fast-growing. The vigorous nature of M. kobus combined with the lack of growth control from deer browse and insect pests is likely to result in it overwhelming and outcompeting native plant species.

Magnolia kobus is more tolerant of cold weather than most species of Magnolia (Utah State University 2016). Its native range includes Hokkaido, the northernmost island of Japan where temperatures can reach –34.4°C (Davis Landscape Architecture 2016). The individuals in the Magnolia Collection at RBG are well within the hardiness range of the species. The property is located in North American Plant Hardiness Zone 6a, where the extreme minimum temperatureis –23.3 °C.MagnoliakobuscantolerateconditionsuptoZone4,where –34.4°C is the extreme minimum temperature (Natural Resources Canada 2016; Brand 2015). Many of the other Magnoliaspecies in the Collection are native to warmer southern latitudes and have lower cold tolerances.


The pollen from Magnolia species may be transported by beetles to a different plant, but insects often move between flowers on the same plant. In Magnolia kobus, this results in self-fertilization, as the species is self- compatible (Azuma et al. 2001). This provides an advantage when the species is established in new habitats. Self-fertilization permits the quick development of a population of individuals that have characteristics adapted to the new environment. This is also advantageous for an invading species as only one or a few individuals are required for the quick colonization of an area (Heiser 1962).

In addition to spreading by seed, Magnolia kobus is also able to reproduce by layering. This is a process by which a lower branch that is still connected to the main stem forms roots where it comes into contact with the soil (Evans and Blazich 1999) (Figure 6). Although the initial spread of a population may be limited by the presence of only a few mature seed-bearing trees, young seedlings are able to reproduce vegetatively. The ability of Magnolia kobus to


FIGURE 6.A branch of Magnoliakobusthat is still connected to the main stem forming roots at the point where it came into contact with the soil.

reproduce both sexually and asexually increases its ability to quickly invade natural areas.


Magnolia kobus plants contain sesquiterpene lactones, chemical compounds that can produce allelopathic effects on surrounding vegetation (Abdelgaleil and Hashinaga 2007). Two of these compounds that are found in M. kobus are costunolide and parthenolide (Park et al. 2010). In a study by Abdelgaleil and Hashinaga (2007), costunolide and parthenolide inhibited seed germination and root length of four test species, viz., wheat, lettuce, radish, and onion. The extent of germination inhibition was dependent upon the concentration of the chemicals added, but shoot growth of the test species was sig


nificantly reduced at all levels of chemical concentration. A similar study by Abdelgaleil et al. (2009) using wild oats showed reductions in growth with increasing concentrations of costunolide and parthenolide. The presence of costunolide and parthenolide in M. kobus can result in allelopathic effects on nearby plant species by the release of the chemicals into the soil via roots and fallen debris (Hampton 2010).


The invasion of non-native plants into natural areas following a lag phase often cannot be explained by a single factor. These events are frequently the result of a combination of stochastic and deterministic factors. Stochastic changes may provide increases in the availability of ideal site conditions that a species can take advantage of, while deterministic factors regarding the life history of a species can contribute to increases in invasive potential (Kowarik 1995). Three hypotheses for the recent establishment and spread of Magnoliakobusin RBG’s natural lands were developed over the course of this study and are discussed below.

Similarities to Native Habitat

Magnoliakobusoccurs throughout Japan, where it can be found in the mixed broadleaved–Abies sachalinensis forest, also called the Pan-Mixed forest. The typical vegetative composition of the mixed broadleaved–Abies sachalinensis forest consists of well-developed tree, shrub, and ground-cover layers. The canopy is dominated by AbiessachalinensisF.Schmidt, AcermonoMaxim., Betula ermanii Cham., Quercus crispula Blume., Picea jezoensis (Siebold & Zucc.)Carr., and Tiliajaponica(Miq.)Simonk.Magnoliakobusexistsbelowthe canopy along with other smaller trees including Acerpalmatumvar. matsumurae (Koidz.) Makino, Cornus controversa Hemsl., Kalopanax septemlobus (Thunb. ex A.Murr.) Koidz, Ostrya japonica Sarg., and Prunus sargentii Rehder (Nakamura and Krestov 2005).

The majority of genera listed by Nakamura and Krestov (2005) as associates of Magnoliakobusin this forest type in its native range are located within RBG’s natural lands. The associated genera found in the forest in which the seedlings are spreading include Acer, Betula, Quercus, Tilia, Cornus, Ostrya, and Prunus.This vegetative community type may therefore be indicative of an environment that is suitable for the growth and establishment of M. kobus.

Magnolia seeds germinate well in rich, moist soils with alayer ofleaf litter in forested areas (United States Department ofAgriculture 1948).According to the results of soil tests completed under Ontario’s Ecological Land Classification (ELC) system at RBG, the average depth of organics was 3.09 cm in the polygons where Magnolia kobus was found, and the moisture regime of the soils ranged from Freshto Moist.This supports my conclusion that the soil in the area


where the M.kobusseedlings were first discovered provides ideal conditions for seed germination.

Some invasive species retain the same habitat requirements over time, thereby limiting the range that they are able to occupy. The range of abiotic conditions that can be tolerated by a species depends on the amount of genetic diversity within the species and on the number and movement of propagules (Olyarnik et al. 2009). As noted above, Magnolia kobusgrows in hills and piedmont regions in its nativerange (Ohwi1965).AccordingtoWilliamMcAvoy, abotanistfor the Delaware Department of Natural Resources and Environmental Control (personal communication), the invasive M.kobusis currently limited in Delaware to the piedmont regions in the northern part of the state. Based on the locations in which M.kobusnaturalization has occurred, it appears that this species is notyet capableofadaptingtoawide varietyofhabitattypes. InRBG’snaturallands,the species is spreading throughout ravines, which is consistent with its preferred habitat type. These ravines, in combination with the presence of associated vegetative community types and soil conditions, provide an ideal environment for the further establishment of M.kobus.

Ideal Climate Conditions

Abiotic factors, such as climatic conditions, are an ecosystem’s first line of defense against invasions by limiting the establishment of non-native species to regions that provide conditions similar to those in their native range (Olyarnik et al. 2009). The establishment of Magnolia kobus at RBG may be attributed in part to changes in Hamilton’s average spring temperatures, as described below.

The reproductive capacity of a plant depends on its ability to produce quality viable seed.This, in turn, depends on genetic, physiological, and ecological factors. The seed productivity of a plant generally depends on the viability of pollen. The main cause of reduced pollen viability may be low temperatures that occur during the time of flowering, which results in the inability of the pollen to fully mature (Kameneva and Koksheeva 2013). Magnolia kobus blooms in early spring when there is still potential for cold temperatures in southern Ontario. However, as a result of changes in climate conditions, the warmer weather has begun to arrive earlier in the year. Over the past several decades, the temperatures observed in Canada during the spring season have increased (Redmond andAbatzoglou 2014).The average spring temperature in Hamilton has increased 0.7°C since 1970, and the average temperature in the winter has increased 1.7°C within the same time period (Ontario Centre for Climate Impacts and Adaptation Resources 2011). This warming trend may allow for the maturity of M.kobuspollen, thereby increasing the probability of the development of viable seeds.

The average spring temperature and precipitation of a representative city on each of the main islands of Japan and, for comparison, of Hamilton, Ontario, are shown inTable 1. Hamilton’s average temperature for the months of March andApril is 2.95°C, well below that of three of the other Japanese cities. However, there is only a 0.2°C difference between the average spring temperatures


TABLE 1. Average spring temperature and precipitation of four representative Japanese cities and of Hamilton, Ontario. Data from Norwegian Meteorological Institute and Norwegian Broadcasting Corporation (2016) and Weatherbase (2016).

Average spring temperature Average spring precipitation City (March&April)(°C) (March&April)(mm)

Sapporo, Hokkaido, Japan 3.15 71.5 Tokyo, Honshu, Japan 11.30 110.0 Matsuyama, Shikoku, Japan 11.35 100.0 Nagasaki, Kyushu, Japan 12.65 155.0 Hamilton, Ontario 2.95 73.8

of Hamilton and Sapporo. The latter city is located in the northernmost range of Magnolia kobus in the northern Japanese island of Hokkaido. Thus, this species is evidently capable of producing mature pollen grains at temperatures of approximately 3°C. The warming trend in the city of Hamilton has brought its spring temperature to a level that is just within the tolerance range of M. kobus.

In order for Magnolia seeds to maintain their viability, they must remain moist (Barbour 2008). A lack of sufficient snowfall can result in the seeds drying out and losing their viability. Data collected from an Environment Canada weather station in Hamilton shows a decline in winter precipitation of 9 mm since the 1970s (Ontario Centre for Climate Impacts and Adaptation Resources 2011). However, the winters of 2013/14 and 2014/15 saw more total hours of snowfall than would be expected for the climatic average, and the temperatures remained below 0°C for most of the season (WeatherSpark 2016). The cold weather causedthemajority ofthe snow toremainon thegroundfor the duration of each of those two winter seasons. The blanket of snow would have protected the seeds against winter desiccation until the spring thaw (Shimano and Masuzawa 1998).

Disturbance and Adaptability

Shelford’s Law of Tolerance states that the distribution of a species is controlled by the environmental factor that the species can tolerate the least. Environmental factors that can limit plant species development include temperature, soil pH, and moisture. Three steps must be taken in order to determine whether or not a factor will limit the range of a species. First, the stage of the plant life cycle that is most sensitive to the factor must be determined. Second, the level of tolerance of that stage must be determined. Finally, it must be shown that the variations of the factor are within the species’ tolerance limits in its native range and that they are beyond the limits outside of its range (Krebs 2008).

Magnolia kobus requires gaps in the forest canopy for regeneration (Nakamura and Krestov 2005), which indicates that light availability is a growth- limiting factor for this species. Disturbance may also directly influence the success of an invasive species by altering the suitability of the environment, or


indirectly by increasing the availability of resources through the reduction of competing species. The mortality and canopy dieback of ash trees (Fraxinus spp.) due to the invasive emerald ash borer (EAB) has created gaps in the canopy of RBG’s forests. These canopy gaps provide an opportunity for M. kobus to become established as the availability of light increases. All of the ELC polygons that contained M. kobus included ash species. Dieback of ash species due to EAB was first acted upon at RBG in 2012 by pruning and by the removal of hazard trees. Biennial insecticidal injections of healthy ash trees (where 90% or more of the canopy remains intact) also began in 2012 (d’Entremont and Burtenshaw 2013). As of 2016 these management actions were ongoing, and the peak loss of ash trees occurred between 2014 and 2015 (Theysmeyer 2016).

Although light is becoming increasingly available in the forest, much of the forest floor, where the M.kobusseedlings are located, is in full shade.The successful establishment of the large number of M.kobusseedlings may be attributed to the ability of the species to adjust its phenotype, or that of its offspring, in response to environmental conditions. This ability is known as phenotypic plasticity (Sultan 2000). Invasive species generally exhibit significantly higher phenotypic plasticity than non-invasive species (Davidson et al. 2011). Phenotypic plasticity results in the adaptability of a species to different environments if the genotype maintains fitness in poor environmental conditions or increases fitness in ideal conditions, thereby permitting the species to be successful in a variety of environments. By altering its phenotype, M.kobuswould be capable of overcoming abiotic barriers (such as light availability) that protect an ecosystem against invasion by non-native species. The reallocation of resources to growth and reproduction in Magnoliakobusis also likely as a result of a lack of predators, as the plants are no longer required to use resources for defense purposes. For example, a plant can increase the allocation of nitrogen to photosynthesis and reduce allocation to the development of defense structures (Gioria and Osborne 2014).This would allow M.kobusplants to allot energy towards adapting to the new environment, resulting in successful establishment.


Results of the Weed Risk Assessment System

According to the Weed Risk Assessment scoring system, a score that is greater than six suggests that the species being assessed poses a high invasive potential, and rejection of its introduction is recommended. Following the completion of the WRA for Magnoliakobus, the species obtained a score of 11.According to this assessment, M.kobusposes a high risk of invasion, and rejection of future introductions of the species into the gardens at RBG is recommended. The complete assessment results can be viewed in detail inTable 2.


TABLE2.ResultsoftheWeedRiskAssessmentforMagnoliakobus.Thecolumns,fromlefttoright, are:(i)thecategoryofthecriterion,whereA=agricultural,E=environmental,andC=combined; (ii)thesubjectareaofthecriteria;(iii)thenumberofthecriterion;(iv)statementofthecriterion;(v) response;and(vi)score.Thetotalscoreis11,whichindicatesthespeciesexhibitsahighriskofinvasionandistoberejected.

History/Biogeography A Domestication/ 1.01Isthespecieshighlydomesticated?Ifansweris‘no’ Y–3 cultivation gotoquestion2.01 C 1.02Hasthespeciesbecomenaturalizedwheregrown? Y1 C 1.03Doesthespecieshaveweedyraces? ? Climateand 2.01SpeciessuitedtoSouthernOntarioclimates(0-low; 22 distribution 1-intermediate;2-high) 2.02Qualityofclimatematchdata(0-low;1-intermediate;2-high)22 C 2.03Broadclimatesuitability(environmentalversatility) Y1 C 2.04Nativeornaturalizedinregionswithmeanannual Y1 precipitationof40–50inches 2.05Doesthespecieshaveahistoryofrepeatedintroductions outsideitsnaturalrange? Y C Weed 3.01Naturalizedbeyondnativerange Y2 E elsewhere 3.02Garden/amenity/disturbanceweed N0 A 3.03Weedofagriculture/horticulture/forestry N0 E 3.04Environmentalweed Y2 3.05Congenericweed ? Biology/Ecology A Undesirable 4.01Producesspines,thornsorburrs N0 C traits 4.02Allelopathic Y1 C 4.03Parasitic N0 A 4.04Unpalatabletograzinganimals Y1 C 4.05Toxictoanimals N0 C 4.06Hostforrecognizedpestsandpathogens N0 C 4.07Causesallergiesorisotherwisetoxictohumans N0 E 4.08Createsafirehazardinnaturalecosystems N0 E 4.09Isashadetolerantplantatsomestageofitslifecycle Y1 E 4.10Growsoninfertilesoils N0 E 4.11Climbingorsmotheringgrowthhabit N0 E 4.12Formsdensethickets N0 E Planttype 5.01Aquatic N0 C 5.02Grass N0 E 5.03Nitrogenfixingwoodyplant N0 C 5.04Geophyte N0 C Reproduction 6.01Evidenceofsubstantialreproductivefailureinnativehabitat N0 C 6.02Producesviableseed Y1 C 6.03Hybridizesnaturally ? C 6.04Self-fertilization Y1 C 6.05Requiresspecialistpollinators Y–1 C 6.06Reproductionbyvegetativepropagation Y1 C 6.07Minimumgenerativetime(years) >4–1 ADispersal 7.01Propaguleslikelytobedispersedunintentionally N–1 C mechanisms 7.02Propagulesdispersedintentionallybypeople Y1 A 7.03Propaguleslikelytodisperseasaproducecontaminant N–1 C 7.04Propagulesadaptedtowinddispersal N–1 E 7.05Propagulesbuoyant N–1 E 7.06Propagulesbirddispersed Y1 C 7.07Propagulesdispersedbyotheranimals(externally) N–1 C 7.08Propagulesdispersedbyotheranimals(internally) N–1 C Persistence 8.01Prolificseedproduction Y1 Aattributes 8.02Evidencethatapersistentpropagulebankisformed(>1yr) ? A 8.03Wellcontrolledbyherbicides ? C 8.04Toleratesorbenefitsfrommutilation,cultivationorfire Y1 E 8.05EffectivenaturalenemiespresentinOntario N1


Threats to Native Vegetation

The leaves of Magnoliakobusremain green until late autumn, well after the leaves of native plants have changed colour or have dropped. This gives M. kobus the ability to produce and store more energy than native plant species (Wenning 2014). This factor, in addition to the species’lack of predators, vigorous growth habit, allelopathic properties, ability to reproduce vegetatively, and tolerance of a variety of light conditions, contributes to the competitive advantage held by M.kobus.Such an advantage threatens the growth of other less competitive species, thereby threatening to alter the structure of a vegetative community.

American robins in the Magnolia Collection at RBG were observed in October 2015 consuming several seeds at a time before flying into the adjacent forest. The birds returned within a few minutes to continue feasting on the bright red fruit before again retreating into the forest. This pattern continued until the Magnolia trees were stripped of their fruit. Native fruiting species such as gray dogwood (Cornus racemosa Lam.) are used as a food source by birds in the fall months when Magnoliaseeds are also available (United States Department of Agriculture 2002). Since birds are drawn to the large and attractive Magnolia seeds, they consume smaller amounts of the fruit of native species, thereby contributing to the dispersal of M. kobus while limiting the spread of native species.

Seed Nutritional Value

Although the ecological threats posed by Magnolia kobus are significant, its presence also has the potential to benefit migratory bird species. The seed coat (testa) of many Magnoliaspecies consists of five layers, including a pellicle, hypodermis, fleshy tissue, inner hypodermis, and a woody inner testa. The middle fleshy layer contains oil and reducing sugars (Barbour 2008). The reducing sugars include glucose, fructose, and galactose, which are the main forms of carbohydrates used by birds for energy (Black 2015). Magnoliaspecies produce fruit with a high fat content that ranges from 33–62% (Frick-Ruppert 2010). Fat is more useful for migrating birds than sugar or protein, since one gram of fat has nine calories, while the same amount of sugar or protein has only four calories (UnitedStatesDepartment ofAgriculture2016).Thefatin the seeds of M.kobus can provide migratory birds with the energy that they need to reach their destination.


Due to the competitive nature of Magnolia kobus and its ability to quickly spread throughout forested ravines, immediate removal of the plants in RBG’s natural lands was recommended.Areas where the species is abundant should not necessarily be labelled as being top priority. An area may contain many young seedlings, but another may have older individuals that are producing fruit. The


removal of M. kobus plants that are flowering and producing seed should be a high priority. Currently, the majority of the M. kobus plants in the RBG natural lands areyoungseedlings.Althoughit willbea number ofyears before theyproduce flowers and seeds, the seedlings are capable of reproducing vegetatively. It was therefore recommended that the young plants be removed as soon as possible.

The high score for Magnolia kobus in the Weed Risk Assessment indicates that it should not be planted in the gardens at RBG in the future. This result indicates a high risk of invasion based on several of the species’ characteristics, which further supports the recommendation for its immediate removal from the gardens at RBG, as the seeds will continue to disperse, and the plants will continue toadaptto thelocal environmental conditions. Failureto remove the source of the problem will serve to exacerbate the ecological risks associated with this Magnoliaspecies,andtheoffspringmaydevelopcharacteristicsthatenablethem to invade a variety of habitats.

Based on observations of Magnolia kobus trees that had been girdled the previous year by RBG staff, this method is not the most effective means of controlling the spread of the species. Some dieback was observed on the trees that had been girdled, but the wounds ultimately stimulated growth via shoots from the root collars. Since the plants are also able to reproduce vegetatively, girdling of the main stem may result in shoots sprouting from the root system. The most effective control method for M. kobus is physical removal of the entire plant, including the roots. Young seedlings can easily be hand-pulled, and trees with a diameter at breast height of up to approximately 5 cm can be removed with a shovel or tree extraction tool. Larger trees can be cut down and the stump treated with an herbicide containing glyphosate (Reynolds 2016).

As previously discussed, many native fruiting species are utilized as a food source by birds in the autumn months when the Magnoliaseeds are also mature (United States Department of Agriculture 2002). Since birds are drawn to the large and attractive Magnolia seeds, they consume less fruit of native species, therebycontributingto the dispersalof Magnoliakobuswhile limitingthespread of native species. Future restoration projects in or near areas where M. kobus is found should include native species that produce fruit at a time that coincides with the appearance of the fruit of M. kobus. The fruit of the native species should also be visually attractive (e.g., brightly coloured) and have a high fat content and high antioxidant concentrations. Fat provides energy for both migratory and non-migratory birds, and high levels of antioxidants protects the birds against oxidative damage. This damage can result from fasting during long- distance flight and from free radicals that are produced during fat metabolism (Bolser et al. 2013). Native species that produce fruit with the desired characteristics should be planted as part of restoration projects in order to compete with M.kobusas a food source for birds. Species selection for restoration projects is site-specific, and the plants that are selected for planting should reflect the species composition of the surrounding area or of a reference ecosystem to which the area is being restored. A list of native species suitable for restoration plantings at RBG is presented inTable 3.According to RBG’s checklist of spon


TABLE 3. Native species that produce nutritious fruit at a time that coincides with M. kobus fruit production in autumn and that are appropriate for restoration plantings at RBG.

Scientific Name (Common Name) Beneficial Fruit Characteristics Additional Information Viburnumrecognitum (SouthernArrow-wood) Very high in antioxidants Viburnumdentatumis described by many sources as being highly favoured by birds due to its very high fat and antioxidant content. Others simply list “Arrow-wood” as being a provider of these benefits. Viburnum.recognitumis the species located within CPNS. Cornusracemosa (Gray Dogwood) High fat content Cornussericea (Red Osier Dogwood) High fat content Cornusalternifolia (Alternate-leaved Dogwood) High fat content Information could not be found regarding the nutritional content of this particular species. However, the fruit of Cornusspp. are generally high in fat, with the exception of C.amomum (Silky dogwood). Linderabenzoin (Spicebush) High fat and protein content Since this is a dioecious species, both male and female plants should be present in a restoration. SambucusCanadensis (American Elderberry) Rich in carbohydrates and protein, high in antioxidants Prunusvirginiana (Choke Cherry) High fat content

taneous flora (Smith 2010), these species occur naturally in RBG’s Cootes Paradise Nature Sanctuary.


Abiotic factors, such as climatic conditions, are an ecosystem’s first line of defense against invasions (Olyarnik et al. 2009). Since Magnoliakobusis tolerant of a wide range of temperatures, soils, and light availability, the plants are likely to overcome the abiotic barrier to invasion once germination has occurred. Once the seeds germinate, the plants will have overcome the greatest barrier to establishment, as climate conditions mayhave successfullyhindered seedgermination until recent years.

Magnolia kobus is a popular choice for horticultural landscapes because of its year-round attractiveness and its ability to tolerate a variety of environmental conditions. This has allowed the species to become established in regions


well beyond its native range, and it is now beginning to display characteristics that suggest a high invasive potential. Due to its popularity in manicured landscapes, the risk of further invasion continues to increase. Public education is critical for enhancing awareness of the ecological threats posed by this Magnolia species. Preventive actions are especially important in these early stages of its invasion, and avoiding the planting of M. kobus will help to protect the biodiversity that would be threatened by its establishment in native ecosystems.


A special thank you to Corey Burt, Colin Chapman, Ryan Godfrey, and Dr. Jim Pringle of the RBG Science department for their work in identifying the Magnolia seedlings to species. I am also grateful to Lindsay Barr, Felicia Radassao, and Dr. Jim Pringle for their invaluable assistance and guidance during this study.


Abdelgaleil, S. A., N. Abdel-Razeek, and S. A. Soliman. (2009). Herbicidal activity of three sesquiterpene lactones on wild oat (Avenafatua) and their possible mode of action. Weed Science 57: 6–9.

Abdelgaleil, S. A., and F. Hashinaga. (2007). Allelopathic potential of two sesquiterpene lactones from MagnoliagrandifloraL. Biochemical Systematics and Ecology 35: 737–742. Arbor Day Foundation. (2015). Tree guide: about growth rate. Available at https://www. (Accessed January 11, 2016).

Australian Government Department ofAgriculture. (2015). Weed risk assessment system. Available at (Accessed June 17, 2019).

Azuma H., M. Toyota, and Y. Asakawa. (2001). Intraspecific variation of floral scent chemistry in MagnoliakobusDC. (Magnoliaceae). Journal of Plant Research 114: 411–422. Barbour, J. R. (2008). MagnoliaL. Pp. 1–416 inThe woody plant seed manual, F. T. Bonner and R.

P. Karrfalt, editors. United States Department ofAgriculture. Black, R. G. (2015). Carbohydrates: The good and bad points. Available at articles/carbohydrates-the-good-and-bad-points. (Accessed January 13, 2016). Bolser, J. A., R. R. Alan, A. D. Smith, L. Li, N. P. Seeram, and S. R. McWilliams. (2013). Birds select fruits with more anthocyanins and phenolic compounds during autumn migration.TheWilson Journal of Ornithology. 125: 97–108.

Brand, M. H. (2015). University of Connecticut. College of Agriculture, Health and Natural Resources. Plant database: Magnolia kobus, Kobus Magnolia. Available at http://www.hort. (Accessed October 6, 2018).

Cornell University. (2016) Top Five Great Berries for the Great Birds ofYour Region. Available at (Accessed January 12, 2016).

Cullen, J., S. G. Knees, andH. S. Cubey. (2011).TheEuropeangarden florafloweringplants:A manualfor the identification ofplants cultivatedin Europe, both out-of-doors and under glass,Volume

1. Cambridge University Press, Cambridge, United Kingdom. Davidson, A. M., M. Jennions, and A. B. Nicotra. (2011). Do invasive species show higher phenotypic plasticity than native species and, if so, is it adaptive?A meta..analysis. Ecology Letters 14: 419–431.

Davis Landscape Architecture. (2016). Japan Hardiness Map. Available at https://davisla.wordpress. com/plant-zones/japan-plant-hardiness-zones-map. (Accessed January 11, 2016). Denholm, K.A. and L. W. Schut. (2009). Field Manual for Describing Soils in Ontario. Department of Land Resource Science, University of Guelph, Canada.


De Prins, J. and W. De Prins. (2019). Global Taxonomic Database of Gracillariidae (Lepidoptera). Available at (Accessed June 16, 2019).

d’Entremont, D. and L. Burtenshaw. (2013). RBG emerald ash borer management strategy. Royal Botanical Gardens, Ontario, Canada.

Essig, F. B. (2015). Plant life:A brief history. Oxford University Press, NewYork, N.Y.

Evans, E., and F. Blazich. (1999). Plant propagation by layering: Instructions for the home gardener. Available at (Accessed January 12, 2016).

Flora ofNorthAmerica Editorial Committee. (1993+). Flora ofNorthAmerica North ofMexico. (21 volumes of 30 projected published to date). Oxford University Press, NewYork, N.Y.

Frick-Ruppert, J. E. (2010). Mountain nature:A seasonal natural history of the southernAppalachians. University of North Carolina Press, Chapel Hill.

Fulcher, A. F., and S. A. White. (2012). IPM for select deciduous trees in southeastern US nursery production. Southern Nursery IPMWorking Group in Cooperation withthe Southern Region IPM Center, Knoxville, Tennessee.

García Morales, M., B. Denno, D. R. Miller, G. L. Miller, Y. Ben-Dov, and N. B. Hardy. (2016). ScaleNet: A literature-based model of scale insect biology and systematics. Available at (Accessed June 16, 2019).

Gilman, E. F., and D. G. Watson. (1994). Magnolia kobus. Available at documents/pdf/tree_fact_sheets/magkoba.pdf. (Accessed December 10, 2015).

Gioria, M., and B. A. Osborne. (2014). Resource competition in plant invasions: Emerging patterns and research needs. Frontiers in Plant Science 5: 501.

Gordon, D. R., D. A. Onderdonk, A. M. Fox, and R. K. Stocker. (2008). Consistent accuracy of the Australian weed riskassessmentsystemacross variedgeographies. Diversity andDistributions14: 234–242.

Government of Ontario. (2019). Ecological land classification. Available at data/ecological-land-classification. (Accessed June 9, 2019).

Hampton, N. (2010). Plants that will grow under a magnolia tree. Available at https://www.wildflower. org/expert/show.php?id=5368. (Accessed January 12, 2016).

Heiser, C. B. (1962). Some observations on pollination and compatibility in Magnolia. Proceedings of the IndianaAcademy of Sciences 72: 259–266.

Herb Society of America. (2010). Essential Facts for Spicebush. Available at https://www.herbsociety. org/herbs/documents/Linderabenzoin_000.pdf. (Accessed March 11, 2016).

IPA. (2018). Invasive Plant Atlas of the United States. Available at https://www.invasive (Accessed June 16, 2019).

Kameneva, L. A., and I. M. Koksheeva. (2013). Reproductive biology of seven taxa of MagnoliaL. in the south of Russian Far East. Bangladesh Journal of PlantTaxonomy 20: 163–170.

Kamikado,T., C. F. Chang, S. Murakoshi,A. Sakurai, and S.Tamura. (1975). Isolation and structure elucidation of growth inhibitors on silkworm larvae from Magnolia kobus DC. Agricultural and Biological Chemistry 39: 833–836.

Koike, F., andH. Kato. (2006). Evaluationofspeciespropertiesusedin weedriskassessmentandimprovement of systems for invasion risk assessment. Pp. 73–83 in Assessment and control of biological invasion risks, F. Koike, M. N. Clout, M. Kawamichi, M. De Poorter, and K. Iwatsuki, editors. Shoukadoh Book Sellers, Kyoto, Japan.

Kowarik, I. (1995). Time lags in biological invasions with regard to the success and failure of alien species. Pp. 15–38 in Plant invasions: General aspects and special problems, P. Pyšek, K. Prach,

M. Rejmánek, M. Wade, editors. SPBAcademic Publishing, Amsterdam, The Netherlands. Krayesky, D. M., and J. G. Chmielewski. 2016. New Reports of Nonnative Plants in Pennsylvania. Castanea 81: 150–154.

Krebs, C. (2008). The ecological world view. University of California Press, Berkeley.

Lloyd, D. G., and C. J. Webb. (1986). The avoidance of interference between the presentation of pollen and stigmas in angiosperms I. Dichogamy. New Zealand Journal of Botany 24: 135–162.

McAvoy, W. A. (2016). Non-native and invasive plants in Delaware. Available at http://www. Delaware%202016.pdf. (Accessed January 10, 2016).

Nakamura,Y., and P.V. Krestov. (2005). Coniferous forests of the temperate zone ofAsia. Pp. 163–


220 in Ecosystems of the world: Coniferous forests, F. Andersson, editor. Elsevier, Amsterdam, The Netherlands.

National Geographic. (2016). Piedmont. Available at piedmont. (Accessed December 11, 2015).

Natural Resources Canada. (2016). Canada’splanthardiness site.Available at http://www.planthardiness. (Accessed January 11, 2016).

Newmaster, S.G. and S. Ragupathy. (2012). Flora Ontario Integrated Botanical Information System (FOIBIS). Available at: (Accessed June 17, 2019).

Norwegian MeteorologicalInstitute andNorwegian BroadcastingCorporation. (2016).Weather forecasts for Norway and the world from the Norwegian Meteorological Institute and the NRK.Available at (Accessed January 12, 2016).

Ontario Centre for Climate Change Impacts and Adaption Resources. (2011). Final report: Climate change impacts and adaptation in the Hamilton Conservation Authority Watersheds. Available at Authority%20Workshop%20Final%20Report.pdf. (Accessed January 12, 2016).

Ohwi, J. (1965). Flora of Japan. Smithsonian Institution Press, Washington, D.C.

Olyarnik, S. V., M. E. S. Bracken, J. E. Byrnes, A. R. Hughes, K. M. Hultgren, and J. J. Stachowicz. (2009). Ecological factors affecting community invasibility. Pp. 215–238 in Biological invasions in marine ecosystems: Ecological, management, and geographic perspectives. G. Rilov and J. A. Crooks, editors. Springer, Heidelberg, Germany.

Park, H. W., J. H. Lee, S. U. Choi, N. I. Baek, S. H. Kim, J. H.Yang, and D. K. Kim. (2010). Cytotoxic germacranolide sesquiterpenes from the bark of Magnoliakobus.Archives of PharmacalResearch 33: 71–74.

Pheloung, P. C., P.A. Williams, and S. R. Halloy. (1999).A weed risk assessment model for use as a biosecurity tool evaluating plant introductions. Journal of Environmental Management 57: 239– 251.

Randy Stewart Landscape Designs. (2010). Magnolias. Available at http://rslandscapedesign. (Accessed December 10, 2015).

Redmond, K. T., and J. T. Abatzoglou. (2014). Current climate and recent trends. Pp. 53–94 in Climate change in NorthAmerica. G. Ohring, editor. Springer International Publishing, Switzerland.

Reynolds, L. (2016). How to kill magnolias. Available at 29546.html. (Accessed January 15, 2016).

Rose, N., and M. Dosmann. (2014). Early to evolve, early to flower. Available at http://www. (Accessed December 11, 2015).

Sargent, C. S. (1894). Forest flora of Japan: Notes on the forest flora of Japan. Houghton & Mifflin, Boston, Massachusetts.

Sedgley, M., andA. R. Griffin. (1989). Sexualreproduction oftreecrops.AcademicPress, NewYork, N.Y.

Shimano, K., and T. Masuzawa. (1998). Effects of snow accumulation on survival of beech (Fagus crenata) seed. Plant Ecology 134: 235–241.

Smith,T. W. (2010). Checklist of the spontaneous flora of Royal Botanical Gardens’nature sanctuaries. Royal Botanical Gardens, Ontario, Canada.

Stiles, E. W. (1980). Patterns of fruit presentation and seed dispersal in bird-disseminated woody plants in the eastern deciduous forest. American Naturalist 116: 670–688.

Sultan, S. E. (2000). Phenotypic plasticity for plant development, function and life history.Trends in Plant Science 5: 537–542.

Trifunovi.., S., V. Vajs, V. Teševi.., D. ..okovi.., and S. Milosavljevi... (2003). Lignans from the plant species Achillealingulata. Journal of the Serbian Chemical Society 68: 277–280.

Theysmeyer, T. (2016). Information report: Emerald ash borer—February 2016. Royal Botanical Gardens, Ontario, Canada

United States Department ofAgriculture. (1948). Woody-plant seed manual.The United States Government Printing Office, Washington D.C.

United States Department ofAgriculture. (2002). Plant FactSheet: Graydogwood, Cornusracemosa Lam. Available at (Accessed January 12, 2016).

United States Department ofAgriculture. (2016). How many calories are in one gram of fat, carbohydrate, or protein?. Available at or-protein. (Accessed January 12, 2016).


Utah State University. (2016). Magnolia, Kobus. Available at cfm?controller=final&action=details&key=461&sortby=CommonName. (Accessed December 11, 2015).

Voss, E. G., andA. A. Reznicek. (2012). Field manual of Michigan flora. The University of Michigan Press,AnnArbor.

Weatherbase. (2016).Travel weather averages.Available at (Accessed January 12, 2016).

WeatherSpark. (2016). Beautiful weather graphs and maps. Available at (Accessed January 12, 2016).

Wenning, B. (2014). Tree of heaven: An exotic invasive plant fact sheet. Available at (Accessed January 13, 2016).