Ecologists call the boundary between two habitat types an ecotone, and the instinct to picture it as a thin line on a map is almost always wrong. A real edge has depth, a distinct microclimate, and its own set of species that depend on the boundary itself rather than on either habitat it separates. What that boundary means for the ecosystems on either side of it, though, depends enormously on how the edge came to exist. Some are ancient and stable, built into the landscape over millennia; the animals and plants nearby evolved to use them. Others were cut abruptly by a chainsaw or a bulldozer within living memory, and the forest never had time to adapt.
An edge that feeds both sides
The transition zone connecting mangrove forests, seagrass beds and coral reefs along tropical coastlines is one of the clearest examples of an edge acting as a functioning habitat rather than a mere boundary. Many species of reef fish spend their juvenile phase sheltering among mangrove roots, where tangled structure and murky water offer protection that the open architecture of a reef does not, before migrating seaward to spend their adult lives on the reef itself[1]. That connectivity is not incidental to reef health — it measurably strengthens it. Reefs with intact, nearby mangrove habitat support significantly higher reef fish biomass than comparable reefs without that connection, and the effect holds at a regional scale across the Caribbean[2]. The exchange runs in more than one direction: nutrients, organic matter and mobile fauna move back and forth across the mangrove-seagrass-reef continuum, and disrupting any one link in that chain measurably weakens the resilience of the others, particularly under the added stress of climate-driven disturbance[1].
An edge that costs both sides
A forest edge created by clearing land tells a very different story, and the Biological Dynamics of Forest Fragments Project in the Brazilian Amazon — one of the longest-running fragmentation experiments in the world — has quantified exactly how different. Trees within 60 metres of a fragment edge die, are damaged, or turn over at roughly three times the rate of trees in the forest interior: mean mortality, damage and turnover rates of 4.01, 4.10 and 3.16 percent near the edge, compared with 1.27, 1.48 and 1.15 percent more than 300 metres in[3]. The mechanism is largely physical. A clearing sits hotter and drier than the forest beside it, and that dry air pulls moisture out of the adjacent trees; wind that would otherwise be broken up by continuous canopy instead reaches the fragment edge at full force, increasing windthrow and structural damage. A 2023 analysis extended these findings to show that edge exposure changes tree architecture itself over time — trees near edges grow differently than interior trees in ways that compound biomass loss beyond what the elevated mortality rate alone would predict[4]. What surrounds the fragment matters too: edges bordered by active cattle pasture showed higher mortality than edges bordered by regenerating secondary growth, suggesting the severity of an edge effect is not fixed but depends on what is on the other side of it[3].
How much of the world's forest is now edge
These are not marginal effects confined to a handful of experimental plots. An estimated 70 percent of the world's remaining forest now lies within one kilometre of an edge, and nearly 20 percent lies within 100 metres of one — close enough that the degrading physical effects documented at Amazonian fragment edges are essentially constant features of that forest's condition, not occasional disturbances[5]. Fragmentation has generally been getting worse, not better: more than half of the world's forests became measurably more fragmented between 2000 and 2020, with tropical forests faring worst — as much as 80 percent of some tropical forest regions fragmented further over that same period[6]. The ecological cost compounds with proximity to an edge: across fragmentation studies globally, habitat fragmentation has been associated with reductions in biodiversity of anywhere from 13 to 75 percent, alongside measurable declines in stored biomass and disruption to nutrient cycling[5].
Two edges, one distinction worth keeping
It would be a mistake to conclude from any of this that edges are inherently good or inherently bad for an ecosystem — the mangrove-reef continuum and the fragmented Amazon edge are both, technically, ecotones, and they behave in opposite ways. What distinguishes them is not chemistry or geography but time and origin. The mangrove-to-reef gradient is a boundary the surrounding species evolved alongside, refined over a timescale long enough for entire life cycles to become built around crossing it. A logging or clearing edge is, on an ecological timescale, brand new — decades old at most, cut through a forest whose species have no evolutionary history of coping with sudden desiccation and wind exposure at that scale. The practical implication for conservation is straightforward, even if the fix is not: protecting a fragment's interior means little if the surrounding matrix keeps generating new edges, and restoring a coastline's productivity means treating the mangrove, the seagrass bed and the reef as one connected system rather than three separate habitats that happen to be adjacent.