Climate Change: Defining Reference Conditions and Restoring Freshwater Ecosystems - Climate Change Impacts on Freshwater Ecosystems - page 224

populations. However, even when abiotic conditions have been successfully
restored, the re-colonization of reference species may be disappointing if dispersal
is a limiting factor (e.g. Donath et al . 2003; Jähnig 2007).
Dispersal of organisms to restored sites is dependent on (i) the connection of
the site to source populations or the 'connectivity' of the site; and (ii) the dispersal
abilities of the organisms themselves. The connectivity of a site is dependent not
only on the distance between the site and potential source populations, but also
on the pathways through which organisms may reach the site. Sites located at
different positions in the landscape may be connected to each other, dependent
on position and type of connections (e.g. Soons et al . 2005; Soons 2006).
Generally, organisms with source populations nearby are the first to colonize
restored sites (e.g. Donath et al . 2003), although the dispersal ability of freshwater
organisms varies widely. A main categorization is often made between organisms
that can actively move substantial distances, such as macroinvertebrates with an
aerial life-stage and fish, and those that are predominantly passively dispersed
over larger distances, such as phytoplankton, macrophytes and most invertebrate
fauna lacking an aerial life-stage. Dispersal of active dispersed species is dependent
on species movement ecology and landscape structure and spatial configuration.
In addition, the dispersal of individual species can be enhanced by the construction
of corridors or the removal of barriers (e.g. fish passages in rivers, Schilt 2007).
Passively dispersed species in freshwater ecosystems can be transported by three
main dispersal vectors (namely, water, wind and water birds - each with their
own dispersal potential) either in their adult life stages or as dispersal propagules.
Dispersal by wind is generally confined to relatively short- to landscape-scale
distances (e.g. Nathan et al . 2002; Soons et al . 2004a, b), whereas dispersal by
water and water birds can occur over much longer distances, from landscape to
regional and even continental scales (e.g. Clausen et al . 2002; Figuerola & Green
2002; Charalambidou et al . 2003; Boedeltje et al . 2004; Soons et al . 2008).
While dispersal distances are highly species-specific, in general it can be expected
that freshwater species that produce many small and drying-tolerant dispersal
propagules have the greatest dispersal abilities.
When restoring sites, connections to potential source populations and the
dispersal abilities of reference organisms need consideration. The often much
delayed re-colonization of restored sites, because of spatial isolation and/or low
dispersal abilities of colonizers, may result in failure to reach the species composition
characteristic of reference conditions. Therefore, it has been argued that restoration
should occur in a landscape context, where connections in the landscape are taken
into account (e.g. Verhoeven et al . 2008 and see below). A useful approach for this
is the operation landscape unit (OLU) approach (described below), which considers
how disrupted connections in the landscape may impede restoration success, and
provides a tool to include crucial landscape connections in a restoration plan. This
is especially important in fragmented landscapes where natural connections
between freshwater ecosystems have been disrupted by diversion and canalization
of water flows and extraction of water for human purposes (e.g. Gordon et al .
2008). Reference communities derived from less fragmented or natural landscapes
may exist due to connections between populations throughout the landscape and
have natural spatial species dynamics (such as metapopulation dynamics or other