Cristina Herrero-Jáuregui (Complutense University of Madrid), Merryn Hunt and Simon Smart (UK Centre for Ecology & Hydrology), and Lenore Fahrig (Carleton University) discuss their article: Richness of forest specialist plant species increases with forest fragmentation per se but decreases with proximity to forest edge and reduced forest patch size
Walk into a large, ancient British woodland and you might come across bluebells carpeting the ground, delicate wood anemones nodding in the spring breeze, or clumps of hart’s tongue ferns tucked along a shady bank. These plants — known as ancient woodland indicator species — are the understory jewels of Britain’s forests. They are slow-colonisers that have persisted in woodland habitats for a long time, hence their presence is sometimes a sign that a forest has remained intact and been a forest for centuries. Unsurprisingly, we tend to assume these understorey plants are among the losers when forests become fragmented into smaller, isolated patches. Our new research suggests that the picture is far more complicated.
The fragmentation puzzle
When a large forest is broken into smaller pieces — a process called habitat fragmentation — ecologists generally worry. Smaller patches mean smaller populations, more exposure to harsh edge conditions (think increased wind, light, and disturbance where forest meets field), and greater isolation from other woodlands. Thus, according to logic, species that thrive in the shaded interior of forests should struggle in a fragmented landscape.
But here is the catch: most previous studies have looked at individual patches in isolation, asking how species respond to patch size or how close they are to the forest edge. Very few have stepped back and asked how the whole mosaic of patches across the landscape affects biodiversity. These two questions can give very different answers.
What we did
We surveyed plants in 97 broadleaved woodland patches distributed across Britain, each visited between 2020 and 2022. Within every woodland, we placed 16 small sample plots and recorded which ancient woodland indicator species — from a list of 95 in total, excluding trees and shrubs — were present. We then mapped the surrounding landscape for up to 5 km around each plot, capturing how much forest there was, how many patches existed, how large they were, and how fragmented the landscape was.
To try and tease apart cause and effect in this complex web of interacting variables, we used a statistical approach called Structural Equation Modelling (SEM) to test a series of hypotheses. This allowed us to track both the direct and indirect pathways through which landscape structure affects the plants on the ground.
A surprising result
We confirmed what many previous studies have found at the patch scale: woodland specialist plants are more abundant towards the interior of a forest, away from the edge. Being close to the forest edge — where conditions are brighter, windier, and drier — clearly reduces the diversity of these shade-loving, humidity-dependent specialists.
Yet when we looked at the landscape scale, we found that landscapes with more forest edge — that is, more fragmented landscapes with many smaller patches — actually supported greater richness of woodland specialist plants. This is what ecologists call a positive effect of fragmentation per se (fragmentation independent of how much total forest area there is).
How can this be? We think the most likely explanation is that a fragmented landscape, where the same area of forest is spread across many small patches, encounters a greater variety of local soil types, moisture levels, and topographic features. This exposes the forest to a wider range of micro-habitats, which in turn supports a broader pool of specialist plant species across the landscape as a whole. Even if diversity is reduced nearer to edges, and hence larger patches have a greater area of favourable interior, the landscape-scale gain in habitat diversity more than compensates when the same area of forest is distributed among more patches and an overall larger spatial range.
What does this mean for conservation?
Our findings carry a clear message: you cannot always predict what happens at the landscape scale by looking at individual patches. Local edge effects — the kind studied in most woodland surveys — do not simply add up to produce the same effect across the landscape. The processes at play are different and sometimes work in opposite directions.
This matters for woodland management. Policies that focus only on increasing total forest cover, without considering its spatial arrangement, may not be enough to protect specialist plant diversity. At the same time, creating several well-connected small woodland patches — for example through agri-environment schemes that restore hedgerows and copses and allow them to expand in area — could be just as beneficial for these plants as establishing a single large block of equivalent wooded area, provided the patches are close enough together for species to move between them. Many of Britain’s woodland wildflowers have survived centuries of change. Understanding the full complexity of how landscape structure shapes their fate is essential if we want them to survive the next century too.
