untitled

malfunction is probably the second most important but it usually requires that the

population is driven to low numbers before demographic stochasticity can operate.

Many examples have been documented, particularly for introductions. Extinction by

genetic malfunction appears to come a distant third. Genetic problems have a low

priority in saving a natural population from extinction. They are more relevant to

managing a population in captivity or one whose size is so small and its future so

bleak that it should be in captivity.

We will now look at a few examples of species that have become extinct, or came

close to extinction, to give us a feel for the range of possibilities.

Many extinctions appear to have been caused by habitat changes (Griffith et al. 1989;

Brooks et al. 2002), but the precise mechanism of population decline is usually difficult

to determine retrospectively. One form of habitat loss is through fragmentation of

continuous habitat into patches. Over time these patches become smaller and the

gaps between them become larger. The ratio of edge to interior habitat of the patches

becomes larger (Temple 1986). We have seen this clearly in the fragmentation of

the eastern hardwood forests of North America since settlement in the 1600s, and

in the eucalypt woodlands of Australia in the last century (Saunders et al. 1993).

This process occurred in the 1300s in New Zealand with the arrival of Maori

(Worthy and Holdaway 2002) and much earlier in Europe during medieval times.

Fragmentation is seen most commonly in the transformation of forest or woodland

into farmland, but also in the change from native grassland into agriculture. The

hostility of the matrix is important too. Thus, a matrix of young regenerating forest

or even exotic plantation is less hostile for animals in old growth forest patches than

a matrix of agriculture. Human residential development is yet more hostile (Friesen

et al. 1995). A further aspect we need to consider is the type of forest involved. In

northern boreal forests of Europe and Canada, containing widespread, migratory bird

species, there is much less effect of fragmentation on species richness and ability

to colonize than in tropical forests with their highly restricted distributions of birds

(Haila 1986).

Fragmentation of habitat has a number of consequences:

1 Species that require interior forest habitats (many bird species), away from the edge,

experience reduced habitat and hence population reductions (Saunders et al. 1993).

In a long-term experiment where forest fragments of different sizes were constructed

in the Amazon forest, the ecosystem showed aspects of degradation within the

patches (Laurance et al. 2002). Many bird species avoided even small clearings less

than 100 m across. Edges were avoided and the type of matrix affected movement.

In both England and the eastern USA, extinctions of bird species occurred once a

critical percentage of the original habitat was destroyed (McLellan et al. 1986).

2 Species that need to disperse through intact habitat (many reptiles, amphibians,

ground-dwelling insects) are prevented from doing so and their populations are reduced

to isolated pockets with potential demographic and genetic consequences. In frag-

mented parts of the northern boreal forests of North America, the foraging move-

ments of the three-toed woodpecker (Picoides tridactylus) are highly constrained because

this species strongly avoids open areas (Imbeau and Desroches 2002). Dispersal of

crested tits (Parus cristatus) in Belgium was restricted in pine forest fragments rela-

tive to continuous forest, and this probably reduced their ability to settle in preferred

habitats (Lens and Dhondt 1994). Northern spotted owls (Strix occidentalis) also

CONSERVATION IN PRACTICE 313

18.2.1Effect of
habitat change and
fragmentation