Characterizing biodiversity in urban areas using remote sensing - Urban Remote Sensing: Monitoring, Synthesis and Modeling in the Urban Environment - page 278

present situation and to future scenarios, such as urbanization
scenarios with different intensities and patterns. In addition, the
use of agent-based models has been explored both for modeling
biodiversity components such as individuals (e.g., Topping et al .,
2003) and for modeling urbanization and land use change (e.g.,
Fontaine and Rounsevell, 2009).

countryside (McDonnell and Hahs, 2008). These types of studies
are known as urban gradient or urban-rural gradient studies
(McDonnell et al ., 1997). The basic idea is to compare species
composition in different parts of the city along transects that
could be of any size stretching from city centers to the more
natural or rural areas surrounding them. In this way differences
in, for example, bird species richness in a city could be compared
to that inmore pristine areas. However, the ecological definitions
of urban-to-rural areas differ between studies and vary depending
on perspective. Some studies refer to urban, suburban and
rural areas, but in regions of urbanization the land outside the
urban and suburban areas tends to be heavily influenced by the
proximity of the city, which also affects biodiversity; these areas
have, therefore, sometimes been referred to as periurban areas
(e.g., Mortberg, 2009).
Previous studies has revealed a general pattern of increased
urbanization leading to the presence of fewer species (Jokimaki
and Kaisanlahti-Jokimaki, 2003; Luck, 2007; Garaffa, Filloy and
Belloq, 2009) and higher densities of birds (Chace and Walsh,
2006). McKinney (2006) argued that cities are great homog-
enizing forces, where some species become so called urban
adaptors (Blair, 1996) and are more common in cities world-
wide; it was also suggested that subsets of native species become
locally and regionally abundant at the expense of indigenous
species. However, the picture is more complex than a sim-
ple negative relationship between human density and native
bird diversity (Turner, 2003; McDonnell and Hahs, 2008).
Thus, the general patterns differ depending on definitions of
scale, species and habitats that are compared. McKinney (2006)
concluded that species richness, biotic interactions and ecosys-
tem complexity decline with increasing urbanization whereas
biomass, total organism abundance and ecosystem reliance
on external subsidies all

20.3 Hierarchical levels
and definitions of urban
ecosystems

Urbanization affects ecosystems and biodiversity in ways that
have not been fully studied. According to Breuste, Niemela
and Snep (2008) urban ecosystems are highly complex, even
more complex than many natural systems; this highlights the
need to define clearly the components of urban ecosystems
when comparing different cities and their biodiversity across
the globe. There is, however, a lack of consistent definitions
and terminology relating to the urban landscape and urban
green habitats, making it difficult to compare related studies.
For instance, Florgard (2007) found more than 10 different
definitions of urban forest fragments. Nevertheless, from a broad
perspective, certain terminology describing urban ecosystems is
used frequently in the literature although not always with exactly
the same meaning.
Moreover, different mechanisms affect species at different
scales. It can be difficult to state categorically whether the quality
of the local habitat or features of the adjacent landscape have
the greatest effect on species richness of a population breeding in
a fragmented urban habitat (Hedblom and Soderstrom, 2010).
By subdividing urban areas into hierarchical levels (multiple
geographical scales) with different ecological functions, a more
accurate explanation for recorded species distributions could be
found. Thus, as suggested by Clergeau, Jokimaki and Snep (2006),
local habitats between houses, larger conglomerations of parks or
houses, the whole town and differences between the northern and
southern parts of continents are examples of hierarchical levels.
Furthermore, natural areas that become incorporated into
urban or suburban settlements will change, even if they are
protected, as a result of recreation pressure, urban predators, air
pollution, traffic noise or other disturbances (Mortberg, 2009).
The joint impacts of such urban disturbances have only recently
been studied in detail, but are receiving increasing research
attention (McDonnell, Breuste and Hahs, 2009). Two major
general patterns of urbanization affect ecological processes in
cities. The first is expansion of the city into surrounding areas;
when the area of the cities increases faster than the urban
population growth this is known as urban sprawl. The second
is fragmentation of existing green areas within the city, which is
known as compaction, infill or urban condensation. The impacts
of these urbanisation patterns are yet to be explored.

increase with increasing urbaniza-

tion.
Blair (1996) suggested that the highest species richness of birds
was at the border of the city where the rural landscape meets
the urban landscape, i.e. suburbia. Blair referred to this as the
intermediate-disturbance theory (Hansen et al ., 2005). It means
that there are more heterogeneous landscapes in the suburban
area (compared to periurban and rural areas, open grassland,
golf courses, parks and town center) that allow a diversity of
possible habitats for species such as birds. Plant species have
been shown to exhibit the opposite trend from vertebrates and
invertebrates in that species richness and evenness increases in
cities compared to periurban areas. This is probably because of
the highly heterogeneous patchwork of habitats coupled with
human introductions of native and exotic species and the human
preference for few individuals of many species in close proximity
(Grimm et al ., 2008; McDonnell and Hahs, 2008).
One further example of the complexity associated with the
effects of urbanization on biodiversity is the different responses
of different vegetation types, as illustrated by an example from
Stockholm, the capital of Sweden (Mortberg, 2009). The total
volume of forest per hectare has been found to decrease along
the urbanization axis from periurban to urban, as can be seen in
Fig. 20.2. Along the same axis, the volume of deciduous forest only
decreased slightly, and the volume of oak ( Querqus robur ,anative
species) per hectare even exhibited a small increase. Similar pat-
terns have been reported, for example, from Finland (Jokimaki
and Suhonen, 1998). Possible explanations are: that conifers are
particularly sensitive to air pollution; that deciduous trees and
shrubs are favoured in urban planning and park management;

20.3.1 Flora and fauna along
urban gradients

A large number of studies of urban areas have compared how
biodiversity changes from the city center to the surrounding