The research area lies mostly within the municipality of Rio Preto da Eva, with a smaller slice in Manaus itself, beginning about 80 kilometers north of the city. The BR-174 highway, which leads from Manaus toward Boa Vista and eventually Venezuela, divides the site in two. Most of the area sits east of the highway. Small segments lie west, inside the Rio Negro Left Bank Environmental Protection Area. The Rio Urubu State Forest borders the research zone to the north. Inside these boundaries sit 23 nature reserves distributed across 11 forest fragments, each fragment isolated from its neighbors by pasture or secondary regrowth. Seven field camps, each with sleeping quarters, a small laboratory, a kitchen and basic toilets, serve the researchers who rotate through. The camps are rough by any standard. The scientific yield is among the highest per hectare on Earth.

The experiment asks what happens to biodiversity when continuous forest is fragmented. The answer, it turns out, is complicated and unhappy. Small fragments lose species faster than anyone expected. A one-hectare island loses its canopy birds, its primates, its large predators within years of isolation. The forest edge dries out, sunlight penetrates the understory, wind-fallen trees multiply, and a cascade of local extinctions follows. Even 100-hectare fragments, which seem generous at first glance, shed species decade after decade. The work, led by Thomas Lovejoy with the Smithsonian Institution and the Brazilian National Institute of Amazonian Research, demonstrated a pattern now recognized worldwide: fragmentation is not just habitat loss by another name. It is a second, compounding disaster. The insight has shaped conservation planning from the Appalachian forests to Borneo.

The Biological Dynamics of Forest Fragments Project Area of Relevant Ecological Interest, known by its Portuguese acronym ARIE-PDBFF, was formalized by federal decree 91.884 on November 5, 1985, six years after the research began. Under Brazilian conservation law, an Area of Relevant Ecological Interest is a small protected area where the only permitted activity is scientific research. The Chico Mendes Institute for Biodiversity Conservation, ICMBio, manages the ARIE in cooperation with the National Institute of Amazonian Research and the Smithsonian. The area became part of the Central Amazon Ecological Corridor in 2002, a larger effort to link protected zones across the state. A consultative council was established in 2015. The arrangement has given the project something most tropical research sites lack: continuous access over many decades, without land-use changes undermining the baseline.

The average elevation is 80 meters, on plateaus cut by small streams and creeks that flood seasonally. The climate is hot and wet, the Koppen classification Afi, with average temperatures around 26 degrees Celsius and annual rainfall near 1,900 millimeters. Rainy season runs December through April. The vegetation is dense rainforest with a fairly uniform canopy at about 30 meters and occasional emergent trees reaching 55 meters. Researchers have identified 56 families of trees with over 1,000 species in the study area. Animal counts are staggering for a patch of this size: 51 frog species, 24 lizards, 63 snakes, 370 birds and 52 mammals. Each of these figures represents decades of patient field work, often in conditions that defeat less committed biologists. What Lovejoy and his successors have built is not only a protected area but a detailed, comparative record of what ecology actually looks like when you leave some forest standing and cut the rest.