Geoscience Reference
In-Depth Information
was facilitated by a newly developed method
that, rather than treating the river environments
as a set of unrelated entities, took full account
of their inter-group similarities (Arponen
et al
.,
2008; Leathwick
et al
., 2010); this allowed the
calculation of more robust, complementarity-based
rankings than is possible when using a scoring-
based algorithm (Ferrier and Wintle, 2009). In
calculating these rankings, an approach was used
that is designed specifically to account for the
functional implications of longitudinal connectivity
within catchments (Moilanen
et al
., 2008), enabling
consideration of both the downstream impacts
of stressors in headwater catchments, and the
requirements for upstream and downstream access
by New Zealand's extensive migratory fish fauna.
However, while this ranking process was essentially
objective in its application, subjective decisions
were still required not only about how best to
balance the influence of these various factors, but
also in identifying subtly different sets of rankings
to identify the best conservation opportunities
across landscapes, despite widely varying patterns
of current protection and condition.
Current work is exploring further improvements
to the approach described here. For example,
use of these rankings to identify a set of
high value sites for targeted management made
us aware of a need to balance better the
trade-off between selecting good condition sites
and achieving even representation across all
environment types. Recent rankings recalculated
with more muted condition scores have enabled
better identification of lowland sites that, while
degraded, are the best-condition examples of
those river environments suffering the most
widespread impacts from human activity, and
which were otherwise under-represented in our
high value sites. Similarly, in other recent analyses,
rankings have been calculated successfully that
integrate across rivers and streams, lakes and
wetlands, identifying significant opportunities to
target river management into planning units where
collateral benefits for lakes and wetlands can
also be expected. Finally, trial analyses including
distributional data for threatened species also show
considerable promise, having particular value for
their consideration both of broad-scale ecosystem
patterns and of rarer biodiversity elements that are
often not well captured by generic surrogates. In
the longer term, greater gains are likely through
the development of approaches that shift the focus
to the prioritization of 'management actions for
sites' using information about the likely benefits,
cost and feasibility of different candidate actions
to maximize the retention as opposed to the
protection of biodiversity (Moilanen
et al
., 2009;
Ferrier and Drielsma, 2010). In a recently modified
version of Zonation, such an approach has been
implemented successfully, identifying the most
cost-effective set of management actions across a
landscape, while also balancing the distribution
of these actions with the broader objective of
maintaining representation across a full range of
environments (Moilanen
et al
., in press).
Practicalities of implementation
At a practical level, these results highlight in
particular the difficulty in achieving representative
protection of a full range of river ecosystems in
landscapes subject to intensive land management.
This difficulty is compounded further by both
the marked disparities in patterns of protection
and condition across different environments, a
common feature in many countries (Pressey
et al
.,
1996), and the frequent failure of terrestrially
focused protected areas to adequately represent
freshwater values (Herbert
et al
., 2010). In
New Zealand, conservation lands predominantly
contain those montane and/or wild landscapes
that were left undeveloped after the imposition
of European patterns of settlement; conversely,
lowland environments suitable for agriculture
are severely under-represented (Leathwick
et al
.,
2003). This imbalance is particularly problematic
for riverine ecosystems, given their longitudinal
connectivity (Abell
et al
., 2007; Nel
et al
., 2009a). As
a consequence, the functional integrity of pristine
headwaters can be effectively undermined, despite
their protected tenure, if key functions such as
migration are prevented by impeded access or
habitat degradation in lower reaches. Conversely,
intensive land-uses in the upper parts of larger
