«Community ecology and invasion of natural vegetation by Cynanchum rossicum (Asclepiadaceae) in the Toronto region, Canada 1 VLADIMIR V. KRICSFALUSY & ...»
Thaiszia - J. Bot., Košice, 20: 53-70, 2010 THAISZIA
Community ecology and invasion of natural
vegetation by Cynanchum rossicum
(Asclepiadaceae) in the Toronto region, Canada
VLADIMIR V. KRICSFALUSY & GAVIN C. MILLEREcology Division, Toronto and Region Conservation Authority 5 Shoreham Dr., Downsview, ON M3N1S4, Canada; firstname.lastname@example.org 1 Current address: School of Environment and Sustainability, University of Saskatchewan 117 Science Place, Saskatoon, SK S7N5C8, Canada; email@example.com Kricsfalusy V. V. & Miller G. C. (2010): Community ecology and invasion of natural vegetation by Cynanchum rossicum (Asclepiadaceae) in the Toronto region, Canada. – Thaiszia – J.
Bot. 20: 53-70. – ISSN 1210-0420.
Abstract: Habitat preferences of the invasive alien species Cynanchum rossicum (KLEOPOW) BORHIDI in northeastern United States and southeastern Canada are characterized on the basis of data from both field studies and literature. C. rossicum behaves more as a habitat generalist in North America, compared to its native range in Europe, particularly with respect to shade tolerance and soil type. It is prevalent in forest habitats as well as open meadow and savannah or woodland; and occurs mainly on loam and sandy loam soils. The ecology and structure of vegetation communities affected by C.rossicum is analyzed in the Toronto region, Canada where the phytocoenological optimum of C.
rossicum occurs in semi-open communities. C. rossicum tends to be the primary or secondary dominant species in infested vegetation community polygons (43% and 26%, respectively). The area infested by C. rossicum includes 1813.21 ha or 7.25% of surveyed natural cover in the Toronto region. Young-to-mid-aged forests and plantations as well as semi-open successional communities tend to have the most severe infestations. C. rossicum is a serious threat to rare plant communities such as alvars, tallgrass oak savannahs and woodlands and their associated species.
Keywords: conservation biology, habitat preferences, infestation, invasive species, vegetation cover.
53 Introduction Invasive alien species are increasingly being recognized as one of the most important threats to biodiversity worldwide (VITOUSEK et al. 1996; ADAIR & GROVES 1998; PARKER et al. 1999; MACK et al. 2000). In the USA, invasive alien species are ranked as one of the top threats to endangered species (WILCOVE et al. 1998; SALA et al. 2000). The proportion of endangered species in Canada that is threatened by alien invasions has reached 22% (VENTER et al. 2006).
There has been much research on invasive alien plants' behaviour and control methods, but often relatively little detailed investigation of these species’ community ecology. Only recently have researchers started to analyse possible relationships between invasive alien plant distribution, species traits (RICHARDSON, PYŠEK 2006; PYŠEK et al. 2009) and habitats (HEJDA et al. 2009) in native ranges and new invaded regions. Comparison of plant ecology in a species’ home and introduced ranges may reveal both the plant’s strategies of invasion and help to illuminate possible methods of control. This paper examines the ecology of one invasive species from eastern Europe that has affected parts of eastern North America: Cynanchum rossicum (KLEOPOW) BORHIDI (syn.
Vincetoxicum rossicum (KLEOPOW) BARBARICH).
As concern about the invasiveness of C. rossicum in North America rises, an increasing amount of information is available on its ecology and biology. Recent comprehensive reviews of this species have been published (MILLER & KRICSFALUSY 2008; DOUGLASS et al. 2009). Nonetheless, significant gaps in knowledge remain, particularly regarding this species’ habitats, community ecology and impact on natural vegetation.
Our previous study of C. rossicum was undertaken to assess its distribution patterns in southeastern Canada as well as estimate the rate of spread and colonization success of this alien species in the Toronto region given that C.
rossicum is one of the most severe threats to natural heritage and biodiversity there (KRICSFALUSY & MILLER 2008).
The main goal of this paper is to provide additional data on C. rossicum that could help with development of effective methods for its control. This study attempts to analyze the species’ geographic range parameters, habitat preferences, ecology and the structure of communities at a regional level. An additional goal was to assess the current intensity of infestation of vegetation cover in the Toronto region by C. rossicum.
Materials and methods To compare conditions in native and introduced ranges of C. rossicum we gathered and analyzed available geographic and climatic data. Geographical coordinates of C. rossicum ranges in Europe and North America were estimated by the authors based on published literature data and herbarium records (KRICSFALUSY & MILLER 2008). Climate data (temperature, precipitation) were obtained from different sources on the Internet: WORLD CLIMATE (2009), WORLD CLIMATE INDEX (2009) and CANADIAN CLIMATE DATA (2009).
54 The current prevalence of C. rossicum infestation in the Toronto region was assessed using queries of Geographical Information System (GIS) data that had been collected by Toronto and Region Conservation Authority (TRCA) biologists over the period 2000-2005 with a few additional records dating back to 1996.
This data covers about 40% of the total natural cover in the TRCA jurisdiction.
In these surveys, vegetation communities were delineated as polygons in ArcView GIS software and categorized according to the Ecological Land Classification (ELC) for southern Ontario (LEE et al. 1998). The ELC data collection protocols were adapted by TRCA and are regularly updated (TRCA 2010). Each vegetation community was divided into different layers (up to four layers depending on the complexity of the community: canopy, middle/subcanopy, lower/understorey, and ground). Dominant species (up to four) present in each layer were recorded. Those polygons that included C.
rossicum on the list for any of the vegetation layer dominants were identified as being infested. This method identifies large, established populations, but does not capture small initial invasions where the species does not dominate in any polygon layer. The majority of sites that do have C. rossicum usually have abundant populations; thus, most of them would include C. rossicum as a dominant species within at least one polygon.
To estimate the intensity of infestation of natural cover in the study area we analysed frequency of C. rossicum occurrence in the UTM (Universal Transverse Mercator) mapping grids (2 x 2 km). This method has recently been applied to assess distribution patterns of C. rossicum in the Toronto region (KRICSFALUSY & MILLER 2008).
The ecological requirements of C. rossicum in the Toronto region were inferred from the vegetation community data of infested polygons together with North American literature sources. Soils information for the Toronto region was provided by an overlay of Ontario Ministry of Agriculture, Food and Rural Affairs (OMAFRA 1990) digital soil mapping onto TRCA information for the ELC polygons.
Results and discussion
Fig. 1. The geographic extent of the native and introduced ranges of C. rossicum.
Habitat preferences The native habitats of C. rossicum in southeastern Ukraine and southwestern Russia are located in forest-steppe and steppe zones on the slopes of ravines, sandy hills and scrub habitats (VISIULINA 1957; POBEDIMOVA 1978). KLEOPOW (1990) characterized the original habitat in which the plant probably evolved as subxerophyllic (i.e. fairly dry) oak woodland.
Comparative analysis of C. rossicum habitat preferences in its native range in Europe and introduced range in North America shows a remarkable similarity between them: semi-open scrub or woodland on calcareous, often light-textured soils. “Forest-steppe” and “steppe” in Europe are approximately equivalent to “woodland/savannah” and “prairie” (or grassland) in North America. These habitat preferences are consistent with the plant’s threat to limestone-based alvar ecosystems and oak savannah or woodland in the Great Lakes region.
One possible difference is that C. rossicum is more generalist in its habitat associations in North America, particularly with respect to shade tolerance and soil type. It is prevalent in more-or-less shaded forest habitats in the Toronto region as well as open meadow and savannah or woodland. It can occur on clay soils and occasionally even in moist-to-wet communities.
57 While there is similarity in the habitat of C. rossicum in Europe and North America, the plant’s behaviour is radically different. It is not aggressive or invasive in its native range. In fact it is considered to be rare, or even endemic to southeastern Ukraine and southwestern Russia (KLEOPOW 1929). Because of its limited native range C. rossicum actually needs local protection in southeastern Ukraine (OSTAPKO 1995).
In North America C. rossicum is associated with disturbed and waste areas, such as transportation corridors, limestone quarries, abandoned pastures, hedgerows, pastures and old fields (DITOMMASO et al. 2005). According to ERNST & CAPPUCCINO (2005) it is abundant in sunny undisturbed old fields and along railways in the Ottawa region. It often grows along open rocky or gravelly shores.
C. rossicum has been reported in Ontario by MOORE (1959) and KIRK (1985) from streambanks, edges of alluvial woods, woods (maple, beech, oak, ash), grassy slopes, as well as gardens, fencerows and railroad embankments.
Fig. 2. Occurrence of Cynanchum rossicum in different habitat types in the Toronto region (ELC community classes as adapted by TRCA).
In the Toronto region according to our analysis C. rossicum thrives in forest (deciduous, mixed, coniferous), successional (thicket, hedgerow, savannah, cultural woodland), plantation (deciduous, mixed, coniferous), cultural meadow, and dynamic (beach, sand dune, bluff, sand barren, tallgrass prairie, savannah, woodland) habitats (Figure 2). It is even occasionally present in wetlands (swamp, meadow marsh, shallow marsh) usually along edges where conditions are less saturated.
Almost half (45%) of all recorded ELC polygons containing C. rossicum were forest communities that, together with plantations (12%), account for 57% of the total (another 27% are successional). From this we can assume that the essential ecological requirements of C. rossicum are being met in different forest stands. However plant density is usually lower in shaded forest habitats than in 58 sunny locations (CHRISTENSEN & STROBL 1999), or even may be considerably less according to DITOMMASO et al. (2005). Overall, C. rossicum shows a very high degree of plasticity and a ruderal strategy in all types of habitat in the Toronto region.
C. rossicum is typically associated with calcareous soils. In its native range in southeastern Ukraine, the species grows on stony soils in steppes, on open calcareous screes usually rich in calcium and carbonates (VISIULINA 1957, 1965;
OSTAPKO, pers. comm.). In Canada C. rossicum occurs primarily on shallow soils over limestone bedrock, silty and sandy loams, glacial till, deep loams of upland woods, rocky or clay loam based ravines (DITOMMASO et al. 2005).
Fig. 3. Soil types of vegetation polygons with Cynanchum rossicum in the Toronto region: S – sand; SL – sandy loam; L – loam; CL – clay loam; C – clay; VARI – polygons spanning 1 soil type; UNCLAS – no soil data.
Analysis of TRCA vegetation polygon and OMAFRA soil data show that vegetation communities with C. rossicum in the Toronto region tend to occupy loam and sandy loam soils (discounting the polygons that either have no soil layer data or which span more than one soil type) (Figure 3). Almost half of all infested ELC polygons fall in these soil categories. Nonetheless, there are numerous populations on clay and clay loam as well. According to CHRISTENSEN (1998) populations of C. rossicum in the Toronto region were found growing on sand loams and loamy sands overlying glacial till with carbonate deposits in the upper layers indicating a fluvial origin.
Community structure A wide range of vegetation communities with numerous plant species are found with C. rossicum in North America. Our observations, in conjunction with the literature (CHRISTENSEN 1998; DITOMMASO 2005) show that forests in which 59 this species grows include such trees as red cedar (Juniperus virginiana), white cedar (Thuja occidentalis), white ash (Fraxinus americana), ironwood (Ostrya virginiana), Manitoba maple (Acer negundo), sugar maple (Acer saccharum), basswood (Tilia americana), as well as plantations of white spruce (Picea glauca) and Scots pine (Pinus sylvestris). In the Toronto region poplar (Populus spp.) stands also often appeared to be infested with C. rossicum.
The most common shrub species associated with C. rossicum are thicket creeper (Parthenocissus quinquefolia), poison ivy (Rhus rydbergii), European buckthorn (Rhamnus cathartica), itself highly invasive, grey dogwood (Cornus foemina), staghorn sumach (Rhus typhina), wild raspberry (Rubus idæus ssp.
melanolasius) and riverbank grape (Vitis riparia).
Associated herbaceous layer species include European cool-season grasses (Agrostis, Bromus, Phleum spp.), tufted vetch (Vicia cracca), herb Robert (Geranium robertianum), the invasive garlic mustard (Alliaria petiolata), Jack-inthe-pulpit (Arisæma triphyllum), May-apple (Podophyllum peltatum) and goldenrod (Solidago altissima, S. canadensis and S. gigantea).
In examining the structure of ELC vegetation units recorded for C. rossicum in the Toronto region the spectrum includes 26 community classes, 43 community series, 165 ecosites and 1936 vegetation types (ELC polygons). From the Toronto area data we might assume that meadow communities, followed by successional, have a relatively low structural diversity as attested by the fact that a large number of its vegetation types fit in a low number of higher-level community classes (Figure 4). The opposite is true of the dynamic and wetland communities whose wide range of structure is reflected in their greater division at the community class and series levels.