1 and 3 can also be found. Although most of the species are similar
in both plots, the dominant species are different between plots
(Table 3).
On the other hand, the number of species changes along the
year, having a peak in spring and declining through the year until
reaching a minimum in winter (Fig. 6). Plot 1 and 3 (green areas)
reach a maximum of 16e18 species during spring, but these values
progressively decrease through the year, being 8e10 species in
winter. Plot 2 presents a lower number of species and it never
contains more than 10 species (gravel area).
Regarding the Shannon index (Fig. 7), maximum values are in
the beginning of the spring and decline along the year. The values of
Shannon in plot 2 are more heterogeneous due to the low number
of species that it contains.
Concerning the cover values of species, cover of Sedum is less
variable along the year than the cover of the rest of the species.
These latter species reach the maximum cover at the beginning of
the spring and summer in plots 1 and 3 and the minimum in winter.
In the plot 2, cover of species is practically zero. Thus, temporal
changes are not detected.
Results show that there are two main factors behind the differences
in floristic composition, richness and diversity found
among the plots. On the one hand, the temporal factor and on the
other hand the spatial factor, represented by the location of the
plots.
Temporally, phenological changes over the year in all the plots
can be observed. These temporal changes are reflected in the cover
values of the species (Fig. 8). In summer, high temperatures and
lack of precipitation provoke the removal of practically all the cover
of annual species in all plots, whereas in the beginning of spring
there is a quick increase of cover of plants due to the seedling of
seeds of the seed bank.
Plot 2 shows the highest differences in all the analysed variables.
The location of this plot, in an area without organic substrate,
seems to be the factor behind the scarce number of species and the
low value of diversity. Plot 1 and 3, although less than plot 2, also
shows differences in composition. These results agree with those of
Madre et al. [33], supporting the major role of the substrate in the
diversity and composition of wild species on the green roofs.
The cover of colonizing species and the total richness of species
seem to be inversely related with the cover of Sedum spp. This
result seems to contradict previous studies that show that Sedum
species favour the growth of nearby plants on green roofs under
drought conditions [34], but is the annual life form of the colonizing