Deep dive: forest fragments, long-term data and the extinction debt
In 1975 a scientist named Jared Diamond published a paper suggesting that, given the choice, governments should protect one large block of habitat rather than several small pieces, even if the total area is the same. It was an extension of the theory of island biogeography, a then-decade-old idea that biodiversity is a function of habitat size and connectivity. A year later Daniel Simberloff and Lawrence Abele countered with a paper showing that protecting several smaller conservation areas might be more effective if those areas feature a broader range of species. In academic circles it came to be known as the SLOSS debate: single large or several small. With SLOSS on his mind, in 1976, a young ecologist with the World Wildlife Fund named Thomas Lovejoy initiated the most ambitious experiment in the history of tropical ecology.
The Biological Dynamics of Forest Fragments experiment is still running in the middle of the Brazilian Amazon, and I was lucky enough to visit the site with Lovejoy and a small crew as part of an annual expedition over New Year’s Eve (detail: we observed jungle time, so when 2013 officially arrived I was asleep in my hammock, though surrounded by thousands of lively frogs who no doubt called in the New Year with style). The fruits of that trip as well as further reporting can be found in a recent feature (Splinters of the Amazon). But this week I found myself pondering the way that story unfolded, and the questions that raised about investments in long-term data collection. So I decided to go flip through my reporter’s notebook one more time.
When all of this got started, landowners in the Brazilian Amazon were legally required to maintain forest on 50 percent of their land (compared to 80 percent today). Lovejoy imagined working with ranchers to create an experimental landscape where plots of primary forest could be isolated and studied; by varying the size of plots and comparing to intact rainforest, scientists would be able to test fundamental ecological theories in the field. He convinced the National Science Foundation to buy him a plane ticket to Manaus in 1976, and in a matter of hours he sold the idea to the Brazilian National Institute for Amazon Research. He has been selling the idea ever since, although it has gotten harder, not easier, with time.
The experiment served as a training ground for a generation of tropical ecologists, and it also produced a wealth of papers as the initial wave of changes – local extinctions, biomass shifts, the development of new microclimates and ecosystems along the edges of the plots – swept through the forest. But as so often happens, the experiment has struggled to make the transition from an exciting new experiment to a stable site for long-term monitoring. People often find the latter, in Lovejoy’s words, “kind of boring.”
I’m as guilty as anyone. The National Science Foundation has invested some funds to re-isolate the plots from encroaching forests (the local ranchers gave up their battle against the persistent rainforest long ago). But during my visit I found myself asking Lovejoy why he doesn’t incorporate these secondary forests into the experimental design. From a practical standpoint, it the answer seemed fairly obvious: put your resources into the questions of the day. Scientific interest in secondary forests is only growing as governments seek to manage our increasingly imperfect paradise. But I was also attracted to the pure science, which might yet yield practical knowledge of a different kind, the kind that is steeped in time.
The SLOSS debate wore itself out in relatively short order. The fact is that both sides were right; depending on the situation, one or the other strategy might be better, but more of both is always better. The best policy is might thus be named as many as possible, as large as possible. For convenience and consistency, I will henceforth refer to this idea as AMAPALAP. And in the end, AMAPALAP won out. Conservationists and like-minded politicians realized this long ago, and they have proceeded accordingly. So what’s the purpose of the forest fragments project today?
The answer is in the name. When Lovejoy first proposed the idea, he called it the Minimum Critical Size of Ecosystems Project. What’s more, he thought he would go in, do the work, get the answer and be done in the space of 20 years. Very bold indeed. So bold, in fact, that the Brazilians ended up changing the name. But the original moniker hints at the deeper questions, and the new name – the Biological Dynamics of Forest Fragments – suggests a method. The effects of fragmentation started out along the edges, and even today they are still creeping through the forest. Scientists are now looking at things like seed dispersal in order to understand longer-term changes. Are some trees more likely to be lost than others because the creatures they depend on to propagate have disappeared? How many of the trees hanging on 33 years after the initial isolation will ultimately succumb? And when? How many of species of insect and bird are living out their final days, unaware of the fact that they have already been doomed to extirpation? It’s known as the “extinction debt” – one of many environmental debts that we are passing on to our children.
As discussed in the story, scientists working at the forest fragments site tried exactly once, in 2003, to tackle Lovejoy’s original question (see: Ferraz, et al.). They did it with bird data, documenting a 50 percent decline in understory bird species in the 100-hectare plots within the first 15 years of isolation. But rather than producing an easy number (42, anybody?), they ended up deriving a half-life for extinction in order to calculate how quickly species are likely to disappear from fragmented landscapes of different sizes. By combining forest recovery and forest decay, the team provided an interesting way to home in on a practical minimum (so as to avoid “either ecological irrelevance or practical impossibility”).
For birds, Ferraz et al. suggested that most tropical forest ecosystems would need to be larger than 1,000 hectares in order to retain hope of “rescue” by forest regrowth. If such a rescue is not expected, the team estimated, fragments as large as 10,000 hectares will still lose many species over the course of a century. For his part, Lovejoy uses 100,000 hectares as a rule of thumb for what might constitute a rule of thumb on the minimum size of a truly viable tropical forest ecosystem. Which, he says, is what he might have guessed going into the experiment so many decades ago.
That said, there are enormous uncertainties surrounding these numbers, and they break down if you push them far enough. But this is ecological theory at work, informed by experimental data, and the best way to improve or test the theory is to gather more data. There’s a new experiment, dubbed the Stability of Altered Forest Ecosystems project, starting up in the Malaysian Borneo that could dive even deeper into these questions (and do so in an environment that is already heavily impacted by humans, which is pretty much what we have left today). But only the original forest fragments project has long-term data.
The site is secure for the time being; scientists will re-isolate the plots later this year. But the usual questions about funding will arise again down the road, and Lovejoy is now working with an external board to put the project on a stable footing into the future. We’ll see. Lovejoy will have to sell his proposal once again, this time knowing that the sexy science documenting rapid and dramatic ecological changes has already passed. Now it’s a matter of observing the long tail of ecological decay, but the simple fact is that this is the only long tail we have to observe.