Environmental Factors and Habitat Preferences During the Invasion

In the course of the invasion of the Czech Republic, this species shifted to lower altitudes; at the beginning of the 1970s, 28.5% of the localities were above 600 m a.s.l., but only 14.7% in 1990 (Pysek, 1994). This suggests that the inherent preferences for a cooler climate that H. mantegazzianum acquired at higher altitudes in the Caucasus Mountains affected its ability to invade warmer areas at the beginning, but this constraint was overcome and is no longer present. Moreover, the above frequency distribution of altitudes in the Czech Republic is similar to a recent distribution of altitudes at which the localities of H. mantegazzianum are recorded, indicating that its occurrence is no longer affected by altitude (Pysek, 1994).

Changes in the representation of habitats occupied in the course of the invasion of the Czech Republic are rather profound, indicating that the habitat preferences of H. mantegazzianum changed as the invasion progressed (Fig. 3.3). Because of its popularity as a garden ornamental, this species was more or less confined initially to parks and gardens and their immediate surroundings. Later on the importance of these habitats decreased. That 'semi-natural' habitats (Fig. 3.3) were occupied to the same degree from the very start of the invasion indicates that H. mantegazzianum was able to invade such less disturbed habitats and some grasslands, wetlands and scrub. The beginning of the exponential spread in the 1930s-1940s (Pysek and Prach, 1993; Williamson et al., 2005) is associated with an increase in the proportion of riparian corridors and other linear habitats (roads, railways), suitable for efficient seed dispersal (see Moravcová et al., Chapter 5, this volume). Major changes in habitat spectra are detectable in the 1970s, when profound landscape changes are thought to have encouraged the spread of neophytes in the country in general (Pysek et al., 1998b; Williamson et al., 2005). In this period, H. mantegazzianum started to spread into urban habitats and linear dispersal along corridors, such as roads and railways, became relatively more important than along rivers and water courses (Fig. 3.3). That is, large rivers in particular acted as efficient dispersal vectors in the early stages of the invasion, but later on the species started to spread to more distant areas not associated with rivers (Pysek, 1994). Although rivers are in general a good dispersal vector for invasive plants (Pysek and Prach, 1994), the major invasive species in the Czech flora differ significantly in their affinity to riparian habitats and

H. mantegazzianum is less confined to riparian habitats than, for example, Impatiens glandulifera and two Fallopia species (Pysek and Prach, 1993). In general, H. mantegazzianum is able to invade a rather large spectrum of habitats, and the rate of invasion and dates of first records do not differ substantially among these habitats (Pysek and Prach, 1993). This suggests that the nature of recipient habitats is less important than may be the case for most neophytes, and indicates that once H. mantegazzianum enters a habitat, it spreads exponentially regardless of the characteristics of the recipient vegetation (Pysek, 1994). This is strongly supported by ecological studies on its seed bank (see Moravcova et al., Chapter 5, this volume; Krinke et al., 2005), reproductive biology (Moravcova et al., 2005 and Chapter 5, this volume) and habitat occupation on both local (Mullerova et al., 2005) and national (Pysek, 1994) scales; these studies indicate that this species is only slightly limited by the character of the invaded vegetation or specific site conditions - once there are suitable habitats in the landscape it spreads at a constant and rather high rate.

In the mid 1990s, urban sites (including dumps and deposits in open landscapes) and linear habitats (such as roads, paths and railways) were frequently reported as habitats for H. mantegazzianum, accounting for 29.2% and 29.3% of the total records (n = 679 year/habitat records from 603 localities, with some localities assigned to more than one habitat type). The species was also fairly frequently recorded in various less disturbed 'semi-natural' habitats (15.6%) and riparian habitats (14.4%). Over the entire invasion history, 7.5% of the records came from parks and gardens. These numbers are cumulative,

0 Riparian

» Linear (roads, railways)

1 Urban x Parks and gardens

1930 1940 1950 I960 1970 1900 1990 2000

Fig. 3.3. Changes in habitat preferences of H. mantegazzianum during the course of its invasion in the Czech Republic. Habitat preference is expressed as the percentage of the total number of localities reported up to the given year, which are for a particular habitat. 'Semi-natural' habitats include less disturbed sites such as scrub, grassland, wetlands, forest and their margins (for details of the classification of habitats based on floristic records, see Pysek et al., 1998b). Based on data from Pysek (1994) and updated.

1930 1940 1950 I960 1970 1900 1990 2000

Fig. 3.3. Changes in habitat preferences of H. mantegazzianum during the course of its invasion in the Czech Republic. Habitat preference is expressed as the percentage of the total number of localities reported up to the given year, which are for a particular habitat. 'Semi-natural' habitats include less disturbed sites such as scrub, grassland, wetlands, forest and their margins (for details of the classification of habitats based on floristic records, see Pysek et al., 1998b). Based on data from Pysek (1994) and updated.

therefore slightly biased if inferring current habitat spectra, but correspond reasonably well to where the species is currently found. Taking only records from the last decade (1986-1995, n = 344) yields very similar percentages of 28.8% (urban sites), 33.7% (linear habitats), 14.8% ('semi-natural' habitats), 12.5% (riparian habitats) and 5.8% (parks and gardens).

These figures correspond reasonably well with those of a detailed investigation of habitat preferences at a local scale in the Slavkovsky les region (Pysek and Pysek, 1995). Taking the percentage of a habitat invaded by H. mantegazzianum as a measure, 40.4% of path margins, 38.2% of road ditches and 38.0% of the area adjacent to water courses are occupied by more or less dense populations of this species. In addition, 30.7% of willow scrub and 7.9% of forest margins are also invaded, but only 2.3% of dry grassland and 2.2% of wetlands (Pysek and Pysek, 1995).

Was this article helpful?

0 0

Post a comment