As so many of us now know, fragmented forests, grasslands and prairies lose an average of half of their plant and animal species within twenty years........... Some continue to lose species for thirty years or more.............. In all of the cases examined, the worst losses occurred in the smallest habitat patches and closest to a habitat edge................. More than seventy per cent of the world’s forest now lies within one kilometre of such an edge................Large intact core reserves, interconnected via uninterrupted "wild paths" are the only way to put a halt to species loss from "swiss chesse" punched natural regions

http://www.newyorker.com/tech/elements/roads-habitat-fragmentation

What Roads

 Have Wrought

BY MICHELLE NIJHUIS

The first paved highway across 
the Brazilian Amazon began, 
in the nineteen-seventies, as a
 narrow, hard-won cut through 
dense rainforest. The road, 
which connects the northern
 port city of Belém with the
 country’s capital, Brasília, 
twelve hundred miles away, 
was hailed as a huge step in 
the region’s development,
 and so it was: it quickly 
spawned a network of smaller
 roads and new towns, drawing
 industry to the Brazilian
 interior. But the ecological
 price was high. Today, much
 of the Belém-Brasília highway 
is flanked by cattle pastures—
a swath of deforestation some
 two hundred and fifty miles
 wide, stretching from horizon 
to horizon. 

“Roads scare the hell out of
 ecologists,” William Laurance,
 a professor at James Cook 
University, in Australia, said.
 “You can’t be in my line of 
business and not be struck by 
their transformative power.” 
Laurance has spent most of
 his career studying that power.
 Beginning in 1979, not long 
after the Belém-Brasília highway 
took shape, one of Laurance’s 
colleagues, Thomas Lovejoy,
 helped direct the selective 
clearing and burning of nearly 
four hundred square miles of 
intact forest in northwestern 
Brazil, near the city of Manaus
—a deliberate act of habitat 
fragmentation that would 
become the world’s largest 
and longest-running experiment 
in tropical ecology. (Laurance
 joined the project in 1996.) 
Today, a study in Science 
Advances synthesizes results
 from Manaus with those from 
similar experiments worldwide, 
confirming what scientists 
have long suspected: no matter
 the ecosystem—forest, prairie,
 patch of moss—the effects of
 habitat fragmentation are ruinous.