Population Structure Mortality and Flowering

What can the structure of populations tell us about the species' life history? Can we determine the reasons why H. mantegazzianum is such a successful invasive species? Can an understanding of the main drivers of population dynamics help us develop an efficient control strategy? To answer these questions, the fate of individual plants was followed in permanent plots in the Czech Republic and Germany. Although H. mantegazzianum has been the object of many studies in the past (for review see Tiley et al., 1996), there is little information on its population biology and life history. Moreover, this information was based on a limited number of plants (Otte and Franke, 1998), on observations under artificial conditions (Stewart and Grace, 1984) or was anecdotal (Tiley et al., 1996). Another source of information is short notes on occasional observations by botanists and land managers (Morton, 1978; Brondegaard, 1990) or studies on control (for review see Pysek et al., Chapter 7, this volume). Finally, the studies on H. mantegazzianum and closely related species as fodder plants do not record the fate of individual plants or information on population biology, but focus on standing crop, biomass productivity and suitability for livestock (Satsyperova, 1984).

Population density of H. mantegazzianum is highly variable, ranging from isolated plants in sparse and small populations to dense and large populations covering several hectares. H. mantegazzianum occurs in many habitats, particularly those affected by former or present human activity: along transportation corridors (rivers, roads and railways), and abandoned meadows or forest edges (Pysek and Pysek, 1995; Thiele, 2006; Thiele et al., Chapter 8, this volume). Thus, populations were studied in a range of environmental conditions in both the Czech Republic and Germany. In Germany the populations were classified as open and dense stands on the basis of the density of mature plants and extent of H. mantegazzianum cover (see later for details). In 2002, permanent plots were established in both regions. In Germany, 2-8 replicates of 1 x 2.5 m permanent plots were established at five sites and monitored every year for 3 years. This produced demographic data for two annual transitions. In the Czech Republic, eight 1 x 10 m plots were established and sampled at the beginning of summer and at the end of the vegetation season. These plots were monitored for 4 years yielding data for three transition periods.

When referring to 'population density', it should be noted that, for practical reasons, only plants above a certain threshold size (plants with leaves at least 8 cm long) were considered. This omits the majority of current year seedlings, which are unlikely to survive until the following spring. The stands in the Czech Republic had a mean density across sites and years of 5.4 plants/m2 (min. 0.4; max. 20.2). In Germany, the dense and open stands harboured on average 7.7 (min. 1.3; max. 31.2) and 2.0 (0.3; 7.0) plants/m2, respectively. Changes in density over time for particular localities in the Czech Republic are shown in Fig. 6.3. The decreasing trend in population densities, particularly in the overcrowded populations, agrees with projections of matrix models (see below). However, because the duration of the study was only 4 years, these results must be interpreted with care (Nehrbass et al., 2006).

The proportion of plants that flowered varied considerably between years in the Czech populations and is difficult to interpret. The mean density, pooled across sites and years, is 0.7 flowering plants/m2. In the dense stands in

2002 2003 2004 2005

Year

Fig. 6.3. Changes in population density of H. mantegazzianum over 4 years at eight sites studied in the Czech Republic. Each line represents one site. Number of plants was counted on permanent plots, 1 x 10 m in size. Only plants with leaves longer than 8 cm were counted.

2002 2003 2004 2005

Year

Fig. 6.3. Changes in population density of H. mantegazzianum over 4 years at eight sites studied in the Czech Republic. Each line represents one site. Number of plants was counted on permanent plots, 1 x 10 m in size. Only plants with leaves longer than 8 cm were counted.

Germany it is 0.8 (min. 0.0; max. 2.1) and in the open stands only 0.3, averaged across sites and years (min. 0.05; max. 0.8). Published reports correspond to our observations: one flowering plant per 0.5-1.0 m2 (Tiley et al., 1996) or 4-7 flowering individuals/m2 in established stands (Gibson et al., 1995, cited by Tiley et al., 1996). Although the density of flowering plants is highly variable, the total number of seeds in the seed bank is related to the number of flowering plants (Krinke et al., 2005).

To determine the effect of the size of a plant on its survival, the number of leaves and the length of the longest leaf were used as proxies of plant size in logistic regressions. For populations in the Czech Republic, both factors significantly and positively affected the probability of surviving to the following year. The survival was not dependent on the distance of the tested plant from the nearest neighbour or to the size of its Thiessen's polygon (J. Pergl et al., unpublished). This indicates that survival is similar in the range of habitats studied and does not depend on local conditions.

The survival of H. mantegazzianum individuals in summer and winter was compared for the same stage classes as used in matrix models (see below). Survival of newly emerged seedlings (with the lamina of the largest leaf longer than 8 cm) was on average 22% during summer and 50% during winter. It was higher for larger plants (with 2-4 leaves) and varied between 60% and 67% for newly recorded plants at a given census and between 81% and 85% for those that were recorded previously. Large vegetative plants (with the longest leaf larger than 140 cm or with more than four leaves) have a slightly higher probability of surviving over summer (93%) than over winter (90%).

Similar to survival, flowering in H. mantegazzianum appears to be size dependent, which in turn is closely linked to the age of a plant and the time required to accumulate the necessary resources (Pergl et al., 2006; Perglova et al., Chapter 4, this volume). The results of studies in Germany (Hüls, 2005) (Fig. 6.4) and the Czech Republic (J. Pergl et al., unpublished) suggest that the trigger for flowering is the size of the plant the 'previous' year and that the majority (90-100%) of plants that reach the minimal size flower.

Timing of flowering is crucial for monocarpic plants, so Pergl et al. (2006) used annual rings in the roots (Fig. 6.5) to determine the age structure of H. mantegazzianum populations in its native (Caucasus) and invaded (Czech Republic) distribution ranges. This study revealed that flowering occurred later in the native distribution range (Caucasus) and managed habitats (pastures) than in unmanaged sites in the Czech Republic. The later time of flowering in the native distribution range seems to be due to the higher altitude there, hence shorter period of growth compared to the sites in the invaded range. Grazing significantly prolonged the time needed for accumulating the resources necessary for flowering. The age structures of the populations in the different habitats in the Caucasus and Czech Republic are shown in Fig. 6.6. The number of 1-year-old plants is underestimated, as only plants with leaves at least 8 cm long are included. Interestingly, although plants from unmanaged sites in the Czech Republic flowered on average significantly earlier than those from other habitats, the oldest flowering plant was found in an extremely dry

Fig. 6.4. Average height (A), number of rosette leaves (B), width of largest rosette leaf (C) and petiole length of largest rosette leaf (D) of plants in 2002 and 2003 that did not flower (V, white bars) and flowered (F, black bars) the following year. Error bars denote 95% confidence intervals. Differences between all the groups were statistically significant (logistic regressions: x2 > 34, df = 1, P < 0.001). Data from Hüls (2005).

Fig. 6.4. Average height (A), number of rosette leaves (B), width of largest rosette leaf (C) and petiole length of largest rosette leaf (D) of plants in 2002 and 2003 that did not flower (V, white bars) and flowered (F, black bars) the following year. Error bars denote 95% confidence intervals. Differences between all the groups were statistically significant (logistic regressions: x2 > 34, df = 1, P < 0.001). Data from Hüls (2005).

locality in the Czech Republic. These results suggest that the species is very tolerant and plastic in its response to environmental conditions and is able to postpone flowering for many years (up to 12 years) (for details see Pergl et al., 2006; Perglova et al., Chapter 4, this volume).

Was this article helpful?

0 0

Post a comment