The distribution of diversity: challenges and applications
149
Using simulations, they showed that minimum
extent of vegetation types ranged from 10 km
2 of tall
shrub to 1,948 km
2 of open spruce, while the mean
extent of the fi ve communities available to burn in the
study area varied from 118 km
2 of mixed - wood to
3,407 km
2 of open spruce. Using these thresholds, they
showed that their minimum dynamic reserve maintained its recolonization sources through time, suggesting that minimum dynamic reserves may provide
an operational framework for determining reserve size
in dynamic landscapes under the infl uence of large
natural disturbances such as fi re. Of course, it is very
diffi cult to validate such an estimation – hence the use
of simulations.
Another way in which systematic conservation
planners have attempted to achieve persistence is to
develop a form of redundancy within the plan, i.e. to
set multiple representation targets. Here, the idea is
that reserve planning algorithms are set with the goal
of selecting a network of areas that ensures, for
example, that each species occurs in a minimum of fi ve
separate sites. Building in this degree of redundancy
may be desirable to provide the protected area networks with a degree of resilience to ensure that a
species (or other desired biodiversity attribute) survives
in the face of natural catastrophes, disease epidemics,
the chronic ecological and genetic effects of small population size, or the loss of a reserve to legal or illegal
human intervention.
It should be noted that this use of the term ‘ redundancy ’ has somewhat negative connotations in conservation planning, as it was used as a key theme in
criticisms of formerly widely used scoring procedures
that disregarded complementarity and yielded systems
of protected areas that had high redundancy and were
ineffi cient (i.e. they were expensive and achieved few
targets – see Pressey & Nichols, 1989 ; Pressey, 1994 ).
Thus, the term redundancy is rarely used and ‘ multiple
representation ’ is the favoured expression. This is considered more appropriate because multiple representations are not a by - product of the selection process but,
rather, they are actively pursued.
Rodrigues et al . (2000) provide a useful demonstration of the potential advantages of multiple representations. They used presence/absence data from the
Common Birds Census (CBC) in the UK to test the effectiveness of three families of selection models:
i Single and multiple representations. Single representations calculated the minimum area such that
each species was represented in at least one site. The
environmental surrogates were used. The research
team formed a scientifi c advisory team that gave them
advice on what would be a good target for reservation
for each habitat and water depth class. They were
advised that the key habitats and different depth classes
had to be captured and replicated at least three to fi ve
times to achieve an adequate outcome. See Box 6.3
for the results of this exercise.
Despite their continued use, there has been a large
amount of criticism over the use of simple percentage
targets in systematic conservation plans (Soul é &
Terborgh, 1999 ; Recher, 2004 ; Watson et al ., 2008 ).
The main criticism is that fi xed percentages do not
account for landscape context. The habitat fragmentation literature (Chapter 8 ; and see Lindenmayer &
Fischer, 2006 ) reveals that the size and isolation of the
protected area, its ‘ shape ’ in terms of edge to core ratio,
and also the similarity (or ‘ hostility ’ ) of the matrix
habitat surrounding the protected area, can each
affect the chances of persistence for many species.
Fixed percentage targets do not take these patch - and
landscape - scale effects into account.
There have been a number of recent analyses in the
systematic conservation literature to address this
problem. Specifi c design criteria based on the characteristics of environmental surrogates (e.g. a specifi c
habitat type) have been incorporated into the persistence objective in some systematic conservation plans.
For example, Leroux et al . (2007) introduced a framework for determining a minimum reserve size required
to incorporate natural disturbance and maintain ecological processes by identifying criteria for estimating
the size, location, and effi cacy of a minimum dynamic
reserve . The size and location of such a reserve is determined by the estimated maximum extent of the largest
disturbance event, and by the extent and distribution
of communities of species that are differentially
affected by disturbance.
They illustrated their approach using a study of the
Mackenzie Valley region of Canada, where forest fi re is
the major natural disturbance that infl uences vegetation community dynamics and dependent fauna. In
this research, Leroux et al . (2007) designed and evaluated a candidate minimum dynamic reserve using a
spatially explicit dynamic simulation model that incorporates locally calibrated fi re and the vegetation
dynamics (i.e. the minimum area they need for persistence and recolonization following a fi re event) of fi ve
broad vegetation types (closed spruce, open spruce,
mixed - wood, tall shrub, small shrub).
149
Using simulations, they showed that minimum
extent of vegetation types ranged from 10 km
2 of tall
shrub to 1,948 km
2 of open spruce, while the mean
extent of the fi ve communities available to burn in the
study area varied from 118 km
2 of mixed - wood to
3,407 km
2 of open spruce. Using these thresholds, they
showed that their minimum dynamic reserve maintained its recolonization sources through time, suggesting that minimum dynamic reserves may provide
an operational framework for determining reserve size
in dynamic landscapes under the infl uence of large
natural disturbances such as fi re. Of course, it is very
diffi cult to validate such an estimation – hence the use
of simulations.
Another way in which systematic conservation
planners have attempted to achieve persistence is to
develop a form of redundancy within the plan, i.e. to
set multiple representation targets. Here, the idea is
that reserve planning algorithms are set with the goal
of selecting a network of areas that ensures, for
example, that each species occurs in a minimum of fi ve
separate sites. Building in this degree of redundancy
may be desirable to provide the protected area networks with a degree of resilience to ensure that a
species (or other desired biodiversity attribute) survives
in the face of natural catastrophes, disease epidemics,
the chronic ecological and genetic effects of small population size, or the loss of a reserve to legal or illegal
human intervention.
It should be noted that this use of the term ‘ redundancy ’ has somewhat negative connotations in conservation planning, as it was used as a key theme in
criticisms of formerly widely used scoring procedures
that disregarded complementarity and yielded systems
of protected areas that had high redundancy and were
ineffi cient (i.e. they were expensive and achieved few
targets – see Pressey & Nichols, 1989 ; Pressey, 1994 ).
Thus, the term redundancy is rarely used and ‘ multiple
representation ’ is the favoured expression. This is considered more appropriate because multiple representations are not a by - product of the selection process but,
rather, they are actively pursued.
Rodrigues et al . (2000) provide a useful demonstration of the potential advantages of multiple representations. They used presence/absence data from the
Common Birds Census (CBC) in the UK to test the effectiveness of three families of selection models:
i Single and multiple representations. Single representations calculated the minimum area such that
each species was represented in at least one site. The
environmental surrogates were used. The research
team formed a scientifi c advisory team that gave them
advice on what would be a good target for reservation
for each habitat and water depth class. They were
advised that the key habitats and different depth classes
had to be captured and replicated at least three to fi ve
times to achieve an adequate outcome. See Box 6.3
for the results of this exercise.
Despite their continued use, there has been a large
amount of criticism over the use of simple percentage
targets in systematic conservation plans (Soul é &
Terborgh, 1999 ; Recher, 2004 ; Watson et al ., 2008 ).
The main criticism is that fi xed percentages do not
account for landscape context. The habitat fragmentation literature (Chapter 8 ; and see Lindenmayer &
Fischer, 2006 ) reveals that the size and isolation of the
protected area, its ‘ shape ’ in terms of edge to core ratio,
and also the similarity (or ‘ hostility ’ ) of the matrix
habitat surrounding the protected area, can each
affect the chances of persistence for many species.
Fixed percentage targets do not take these patch - and
landscape - scale effects into account.
There have been a number of recent analyses in the
systematic conservation literature to address this
problem. Specifi c design criteria based on the characteristics of environmental surrogates (e.g. a specifi c
habitat type) have been incorporated into the persistence objective in some systematic conservation plans.
For example, Leroux et al . (2007) introduced a framework for determining a minimum reserve size required
to incorporate natural disturbance and maintain ecological processes by identifying criteria for estimating
the size, location, and effi cacy of a minimum dynamic
reserve . The size and location of such a reserve is determined by the estimated maximum extent of the largest
disturbance event, and by the extent and distribution
of communities of species that are differentially
affected by disturbance.
They illustrated their approach using a study of the
Mackenzie Valley region of Canada, where forest fi re is
the major natural disturbance that infl uences vegetation community dynamics and dependent fauna. In
this research, Leroux et al . (2007) designed and evaluated a candidate minimum dynamic reserve using a
spatially explicit dynamic simulation model that incorporates locally calibrated fi re and the vegetation
dynamics (i.e. the minimum area they need for persistence and recolonization following a fi re event) of fi ve
broad vegetation types (closed spruce, open spruce,
mixed - wood, tall shrub, small shrub).