Finding patterns in forest loss

Alto Mayo forests, Peru. Bruno Locatelli/CIFOR

If there’s one lesson to be learned from the many conservation efforts implemented around the world, it’s that what works in one place may have no impact in another — or worse, it might have a negative effect.

Why are the results so different, and what can researchers and policymakers learn from those differences?

One reason for the disparity is that “context matters,” says Arild Angelsen, a senior associate with the Center for International Forestry Research and World Agroforestry (CIFOR-ICRAF) and an economics professor at the Norwegian University of Life Sciences in Aas, Norway.

But it’s also important to understand which aspects of the context influence the results, he says. And lamenting that every case is different, so there can be no general lessons learned, is as unhelpful as trying to find a one-size-fits-all solution for the planet’s forest loss, which amounts to some 100,000 square kilometres (38,610 square miles) a year, mainly in tropical countries.

Angelsen and Julia Naime, a CIFOR-ICRAF postdoctoral researcher at the same university, set out to look for patterns in deforestation as part of the Phase 4 of CIFOR-ICRAF’s Global Comparative Study on REDD+ (GCS REDD+). Working with the Wageningen University and Research in the Netherlands, where researcher Niki De Sy spearheaded the analysis, their goal was to see if they could help policymakers by determining what kinds of strategies might work best in certain situations, and why.

The first step was to analyse the amount of forest loss and the reasons, or drivers, behind the deforestation. This resulted in a series of seven deforestation ‘archetypes’, or situations with common social and ecological characteristics.

“We know there are many different drivers of deforestation – this has been studied for many years,” Naime says. “There are different contexts, different drivers and different actors. The goal of archetype analysis is to find general patterns across different cases of deforestation.”

The research is pioneering in applying the archetype approach to identify patterns of tropical deforestation globally (with a more detailed focus on Peru, Brazil, DRC, and Indonesia), said Naime at a recent meeting of the Science and Public Policy Advisory Group of the GCS REDD+ in Peru, where preliminary findings were presented.

To create the archetypes, the researchers calculated how much forest had been lost over the two decades. They also looked at the annual rate of forest loss, and whether it was higher than 1% (in a 5-kilometre-by-5-kilometre area) before and after 2015, to determine whether it was relatively new activity, and whether it was accelerating or slowing.

Understanding local trends in forest loss is important for the design of effective policies, they say.

The patterns ranged from intact (or conserved) forest at one end of the scale to once-forested land now used for agriculture or other human activity at the other, with a series of land-use-change scenarios in between.

Based on that information, the researchers created a series of pattern types, which they call “deforestation archetypes,” each with its own characteristics and risks, making it possible to compare situations in different areas of the world’s tropical countries.

In some places, forest cover is still intact or only slightly degraded, but there is a risk of deforestation in areas that are accessible by road or attractive for farming or ranching, Naime says.

Areas currently undergoing deforestation are considered deforestation fronts, with varying degrees of severity. “Emerging” fronts had high forest cover before 2015, but have lost more than 1% of that cover annually since then. “Active” fronts have lost forest at that rate before and after 2015, indicating a longer trend. “Fading” fronts lost more than 1% of forest cover annually before 2015, but not since then; and “gradual” or “inactive” fronts lost forest at a rate of less than 1% annually before and after 2015.

“Once we describe the patterns of deforestation, the next thing we try to do is identify the drivers, such as the main crops being grown in those places,” Naime says. “That is limited by the availability of data, but we have identified some basic categories.”

Land use may include plantations, crops, crop regrowth, and grassland or pasture, for example, or forest loss may be due to urban sprawl or flooding, she says.

When the researchers assigned colours to the archetypes and applied them to a map of Peru, one of the four countries in the GCS REDD+, land-use change patterns jumped out.

In Peru, nearly 70% of the area they mapped still contains well-conserved tree cover, which they termed ‘core forest’. Nevertheless, there are large swaths of active fronts with moderate to high forest loss, including a crescent of unregulated gold mining in the south and expanding agriculture in the centre. Moderate forest loss along the winding Ucayali River shows where the meandering, seasonally flooded waterway has changed course over the years.

Emerging fronts — areas of forest loss that have appeared since 2015 — are fewer, but show where the rapid implementation of effective policies could stem further deforestation. In the central Ucayali region, for example, an active area of high to moderate forest loss is bordered on the east by a fringe of emerging fronts pushing into intact forest — a warning sign for policymakers seeking to keep deforestation in check.

When the researchers applied the same criteria to other tropical countries, common patterns appeared in certain regions — grasslands in South America, plantations in Southeast Asia, and a mix of pasture, agriculture and shrubs in Africa, for example. In some places, deforestation stops at a national boundary, implying that conservation measures are working, or core forest crosses national borders, pointing to a place where binational collaboration might be possible.

The researchers plan to refine the archetype analysis by looking at why changes have occurred. Is a once-forested area now being used for pasture or grassland, tree plantations or crops? Has it been eaten up by urban sprawl? Have the trees been swept away by flooding?

The next step will be to break down the elements of public policies, classify them and examine them in the context of the archetypes, to see which policies have an impact, and why.

So far, the researchers have been surprised to find that South America has few ‘fading’ frontiers — places where deforestation was high before 2015 and low in recent years.

“It seems that once you start the process, it gets momentum, and there are self-reinforcing feedback loops that may be related to development of infrastructure that comes with people,” Angelsen says. “So it becomes more attractive, and more people move in, and the deforestation continues.”

That’s a warning sign.

“Once it’s started, it’s hard to stop,” he adds, “so it’s important to be careful not to start it.”

This work was carried out as part of the Center for International Forestry Research’s Global Comparative Study on REDD+ (www.cifor.org/gcs). The funding partners that have supported this research include the Norwegian Agency for Development Cooperation (Norad, Grant No. QZA-21/0124), International Climate Initiative (IKI) of the German Federal Ministry for the Environment, Nature Conservation and Nuclear Safety (BMU, Grant No. 20_III_108), and CGIAR Research Program on Forests, Trees and Agroforestry (CRP-FTA) with financial support from CGIAR Fund Donors.