Approaches to landscape conservation have generally focused on
either community representation or habitat occupancy. Representation
approaches use computerized optimization techniques to select a subset of
available habitats that maximize the number of species whose distributions
fall within the reserve network [see Cabeza and Moilanen (2001) and
references therein]. Additional criteria, such as the total cost of land
acquisition, may be included in the optimization (Ando et al. 1998). Habitat
occupancy approaches attempt to select a set of habitat reserves that
will maximize habitat occupancy and hence minimize extinction risk.
Occupancy approaches typically focus on the maintenance of habitat
connectivity to preserve potential recolonization routes after local population extinctions.
Landscape conservation issues illustrate the need for new approaches to
deal with the complexity of resource management issues. Network theory
is a developing technique that uses recent advances in computer technologies. The discussion in this chapter of an applied problem that uses network
theory thus serves as an introduction to the evolving modeling approaches
for resource management.
In the remainder of this chapter, I discuss effects of habitat fragmentation on metapopulation survival and illustrate the importance of habitat
connectivity using a simple metapopulation model. I then develop a
mathematical framework for studying dispersal networks in a landscape
and apply this framework to Mexican spotted owl (Strix occidentalis lucidia)
habitat in the southwestern United States.
7.2 Habitat Connectivity and Occupancy in a Simple
Metapopulation Model
The degree to which habitat may be interconnected via dispersal among
populations is a key determinant of species survival in fragmented landscapes. Therefore, understanding how connectivity changes as habitat is
lost is important to conservation-planning efforts. As habitats are lost
from landscapes, typically one observes that the remaining habitats become increasingly isolated, a process referred to as fragmentation. From
percolation theory (Stauffer and Aharony 1985), we know that habitat
connectivity in randomly fragmented habitats exhibits abrupt nonlinear
changes when the amount of habitat lost reaches a critical value (Gardner
et al. 1987). The situation is illustrated in the left-hand column of Figure 7.1.
If our goal is to walk from one side of the grid of habitat patches to the
other by crossing from one white (habitat) cell to another vertically or
horizontally to one of the four nearest neighbor cells, then theory tells us
that the probability of doing so becomes vanishingly small when the amount
of habitat lost is about 60%. Thus, for species that exhibit metapopulation
126
Timothy H. Keitt
either community representation or habitat occupancy. Representation
approaches use computerized optimization techniques to select a subset of
available habitats that maximize the number of species whose distributions
fall within the reserve network [see Cabeza and Moilanen (2001) and
references therein]. Additional criteria, such as the total cost of land
acquisition, may be included in the optimization (Ando et al. 1998). Habitat
occupancy approaches attempt to select a set of habitat reserves that
will maximize habitat occupancy and hence minimize extinction risk.
Occupancy approaches typically focus on the maintenance of habitat
connectivity to preserve potential recolonization routes after local population extinctions.
Landscape conservation issues illustrate the need for new approaches to
deal with the complexity of resource management issues. Network theory
is a developing technique that uses recent advances in computer technologies. The discussion in this chapter of an applied problem that uses network
theory thus serves as an introduction to the evolving modeling approaches
for resource management.
In the remainder of this chapter, I discuss effects of habitat fragmentation on metapopulation survival and illustrate the importance of habitat
connectivity using a simple metapopulation model. I then develop a
mathematical framework for studying dispersal networks in a landscape
and apply this framework to Mexican spotted owl (Strix occidentalis lucidia)
habitat in the southwestern United States.
7.2 Habitat Connectivity and Occupancy in a Simple
Metapopulation Model
The degree to which habitat may be interconnected via dispersal among
populations is a key determinant of species survival in fragmented landscapes. Therefore, understanding how connectivity changes as habitat is
lost is important to conservation-planning efforts. As habitats are lost
from landscapes, typically one observes that the remaining habitats become increasingly isolated, a process referred to as fragmentation. From
percolation theory (Stauffer and Aharony 1985), we know that habitat
connectivity in randomly fragmented habitats exhibits abrupt nonlinear
changes when the amount of habitat lost reaches a critical value (Gardner
et al. 1987). The situation is illustrated in the left-hand column of Figure 7.1.
If our goal is to walk from one side of the grid of habitat patches to the
other by crossing from one white (habitat) cell to another vertically or
horizontally to one of the four nearest neighbor cells, then theory tells us
that the probability of doing so becomes vanishingly small when the amount
of habitat lost is about 60%. Thus, for species that exhibit metapopulation
126
Timothy H. Keitt