Coffee Filter

It’s hard obviously to imagine a house which doesn’t have a door. I saw one one day, several years ago, in Lansing, Michigan. It had been built by Frank Lloyd Wright…[T]here appeared something like an open-work roof that was practically indissociable from the vegetation that had invaded it. In actual fact, it was already too late to know whether you were indoors or out…A dozen more or less similar houses were scattered through the surrounds of a private golf club. The course was entirely closed off. Guards…were on duty at the one entrance gate. –Georges Perec (1974)

Dosage and tolerance mark the thin line between palliative and poison.

The caffeine that perks up one patron in the coffeehouses of snowbound Minnesota can rocket another into rare tachycardia and cardiovascular collapse. It’s a chance many are willing to take. Even the jitters tell us we’re still alive in -40 wind chill. And look on the bright side, as one must here under the penalty of death, should a slurper keel over, a table in a popular joint is suddenly free for the rest of the day.

The devotional’s exploding global appeal—and increasing consolidation—obfuscates its modest origins. The genus Coffea grew naturally in the Horn of Africa, its purine alkaloids caffeine and theobromine herbivore irritants and insecticides. Lure and lore parlayed the bean into a regional then an imperial prerogative and today a record 150 million-60-kg-bag and US$100-billion-a-year global industry, second only to Big Oil, employing, across production, trade and retail, as many as 500 million people.

Blue Mountain, Colombian, Ethiopian Harar, Hawaiian Kona, Java, SL28, and on and on, manifold varieties and hybrids are grown across seventy countries. At $1000 per kg, the priciest is an Indonesian bean swallowed and shat out by a caged luwak or Asian palm civet. The kind of arbitrary appreciation slash commoditization that’ll eventually slash slash slash the civet population to oblivion.

The differences in species, soils, sunlight and cultivation help produce beans of a variety of balance, bouquet, bright and body. According to an extraordinary line of research by University of Michigan’s Ivette Perfecto, John Vandermeer and their colleagues, coffee ecosystems also differ in their capacity to naturally control pest insects and plant diseases that can devastate a Coffea crop.

Control in Coffea canephora, the major Latin American variety, emerges from more than the plant’s biochemistry and bred-in disease resistance. The thatch of ecological relationships—predation, mutualism, competition, etc.—up and down the food web in which the plant finds itself can box out pest damage. Resistance and resilience are found in the field rather than the object, emerging out of these interconnections and their redundancy. Should one control cascade fail, another steps up or steps in.

For ten years plus, on a 300-hectacre organic coffee farm in operation for nearly 100 years in the Soconusco region of Chiapas, Mexico, Perfecto and Vandermeer’s team have worked to tease out the multiple spatioecological layers that buffers shade coffee from the worst of pest outbreaks.

Coffee rust disease fungus Hemileia vastatrix, the coffee berry borer Hypothenemus hampei, the green coffee scale Coccus viridis, and the leaf-mining moth Leucoptera coffeella are four of potentially 200 now-endemic pests, each alone capable of destroying a coffee crop, and yet, here, have not. The PVC team identified a web of dynamic and contingent relationships across, if you’re keeping score at your local, thirteen kinds of organisms and six ecological processes, keeping the four pests largely in check.

*

The swarming Azteca instabilis ant serves as the keystone species for the control network.

Queens of the polygynous ant ‘bud’ off and with some of their brood colonize new nests on coffee plants nearby. Colony diffusion is constrained in part by a Pseudacteon phorid fly, which lays brain-eating offspring inside the worker ants. Phorid attacks are nest-density-dependent. The more Azteca nests in the vicinity, the more attacks in the area, producing a power law distribution of nests across the farm.

That distribution, in turn, helps a lady beetle species Azya orbigera control the green coffee scale, our first pest. How? It gets all Robert Altman, so pay attention.

The ants and green scales are—perhaps a surprise—mutualists. Azteca offer the scale protection, including against the adult beetle, in return for honeydew the scales secrete. Protection Azteca cannot provide, however, against the beetle larvae. The larvae’s waxy protuberances gum up Azteca mandibles and the young’uns chaw on the scale to their heart’s content. The larvae score a daily double as Azteca also scares away parasitic wasps that feed on—and would control—the larvae in the ant’s favor.

Without Azteca’s indirect protection, the beetle larvae wouldn’t be able to survive its own parasitic tormenters in order to control the green scale. In short, as Perfecto and colleagues describe, the beetle helps produce the very spatial distribution it needs to survive. Dialectical biology in action.

There is a second, if indirect, means by which the distribution of Azteca is circumscribed to 3-5% of the farm. The white halo fungus Lecanicillium lecanii attacks the scale on which Azteca depends when the scale is locally abundant (which occurs largely under Azteca protection).

White halo also attacks the coffee rust, our second pest, but, as we see, does so only because Azteca protects scales to densities white halo attacks. In other words, the scale and rust are by indirect means mutually constraining.

*

Something puzzled the PVC team. How do adult beetles—viciously attacked by Azteca—oviposit their similarly vulnerable eggs on a plant on which only their larvae survive?

Remember the phorid fly whose larvae feed on Azteca? What do we want? Brains! When do we want it? Brains! The fly offspring locate the ant colony by detecting its alarm pheromone and any one individual host by its movement. The ants respond by retreating to the nest or standing stock still in such a way as to avoid the fly’s motion detection and, when able, to attack its phorid tormenter.

The ants produce a second ‘phorid’ pheromone alerting other ants to enter their defensive catatonia. The female beetles looking to oviposit their eggs unmolested can detect this second pheromone, finding areas of the plant in which Azteca have entered their collective freeze frame.

It appears, then, that Azteca distribution, natural pest control, and likely other such distributions in the forest and farm, arise from no single cause but a nonlinear complex of interactions distributed across the ecological network, an important lesson for those of us in the fields of livestock disease and public health.

The complications pile on, however.

If the coffee scale needs Azteca’s protection, how does a new ant colony find scale elsewhere? PVC discovered that although not nearly as effective as Azteca, at least five other ant species that forage in the area tend the scale. In essence, the various species, occupying different parts of the farm canopy act as indirect mutualists maintaining scale densities across the farm, including the local outcrops a new Azteca colony needs.

One ground ant Pheidole cfp, which while feeding on scales (and like Azteca on leaf miners and berry borers, our final two pests), offers Azteca additional help by outcompeting a third ant, an Azteca competitor, Pseudomyrmex simplex.

Other ant species meanwhile act as Azteca antagonists, if only because they do not co-tend scales. Pheidole protensa, for one, which also feeds on berry borers in old fallen seeds that offer borers off-season refuge, outcompete Azteca ally Pheidole ctp on the ground.

We are speaking here of an ecological guild of more than eighty ant species that engage in complex interactions of various—and at times simultaneous—mutualisms and competition across canopy niches.

*

We—or rather nature—can add yet another layer.

To test the effects birds have on pest numbers, the Perfecto-Vandermeer team conducted an exclosure experiment across shade and intensive farms in Soconusco. With 5-cm mesh fishing nets, the researchers excluded birds from 10 x 5 x 3m plots of at least ten coffee plants. They selected control plants open to the elements from parallel rows nearby.

The team placed third and fourth instar larvae of the salt marsh moth and fall armyworm ten per plant in the experimental and control plots, modeling a sudden pest surge from a 2.1 average larvae density.

For each of four days the team placed larvae on the plants before sunrise and counted every three hours until 2pm. The researchers also identified the birds feeding at the coffee layer.

The censuses showed significant differences in the number and density of birds feeding on coffee plants between the shade and intensive plants, with a significant synergistic effect for treatment and site. Many more birds and bird species fed on the shade site, with a significant difference between exclosure and control plots not found on the intensive farm.

In other words, larvae were being removed from the shade controls in a way they were not from those on the intensive farm.

Behavioral observation qualified the results. Contrary to expectations, bird diversity did not appear the direct mechanism by which shade coffee was better protected. Instead, it appeared particularly effective insectivores—including the rufous-capped warbler—foraged repeatedly in shade coffee.

That is, there may be a third effect. Despite the traditional troubles in segregating the effects of bird diversity and density out in the field, it appears the more birds feeding here, the more likely one or a few will be particularly effective, if by chance alone. A sampling effect.

*

Coffee plants may keep the North Country—and the birds south—awake during the day. But there’s no rest for pests at night. While the Rufous-capped warbler and other birds coop, bats—many seed dispersers and pollinators—take to the air, some, here, to eat pest insects.

To weigh the predatory effects of bats and birds, Perfecto and Kim Williams-Guillén’s team set up a series of exclosure treatments in Soconusco: birds-only during the day, bats-only during the night, both sets day and night, and a control of no netting. The group censused noncolonial arthropods—insects, spiders, harvestmen and mites—every two weeks over a seven-week period during the dry season and over eight weeks during the wet season.

The dual exclusion left the greatest density of arthropods on individual coffee plants, 46% greater than on the controls. Bats had a significant effect in the wet season, their exclusion leaving 89% greater arthropod density than controls, but less so in the dry. Finally, there appeared no significant interaction between birds and bats, indicating their predation is additive and each predates on different types of pests.

The seasonal difference may arise in part from the influx of overwintering songbirds during the dry season and an increase in bat abundance during the wet season when mothers, doubling their typical food intake, must nurse their offspring.

Which bats are gleaning what? By netting bats over 44 nights and acoustically monitoring echolocation calls sensitive enough to detect a caterpillar chewing a leaf, the team identified 24 insectivores across a continuum of shade and intensive coffee plantations.

Few species were captured on a single type of farm, but they did differ in their preferences. Indeed, while species richness differed little across the farm gradient, open-space bats, such as the greater sac-winged bat, appeared most frequent in the more intensive farms while forest bats, including the Argentinean brown bat, appeared more shade-prone.

A follow-up PCR study identified DNA of the berry borer and cicada Idiarthron subquadratum in bat fecal samples. So, yes, the bats are indeed consuming the insects.

While forest bats fed less, as measured by their feeding buzzes, the more intensive the farming, open-space bats did not feed more along the gradient, indicating intensive coffee, some plantations with higher abundances of pests, scored little protection across the bat ensemble.

What’s the take-home? The team concluded that even areas dominated by intensive agriculture would benefit from forest fragments, which offer roosts for all bat insectivores, including open-air species. However, they continue, fine-grain, spatially contiguous agricultural matrices, including shade cultivation, would offer forest bats and other insectivores the kinds of wildlife-friendly refugia in which they could better survive.

Indeed, Guillén-Williams and Perfecto write, with the pressures of poverty and food insecurity also in the mix, blocking off agroecological landscapes into patches of forest and intensive farming, at the heart of much conservation modeling, can cause declines in local biodiversity. When boxed out of all available land, the poorest farmers clear cut the forest, while the largest operations, Perec’s guard at the entrance, surf their own destructive production along an ever-expanding forest edge.

*

A self-organized pest control emerges here out of ecological interactions.

Such systems are neither preplanned nor static projects. They’re historically contingent. As PVC describe, coffee plants, and their rusts and berry borers, were imported from Africa. White halo fungus are common to the tropics, the leaf-mining moth to Western tropics, and the ants native to southern Mexico. By dint of conscious cultivation and chance biogeography, this particular combination of organisms happened to converge upon this specific and—at the geological time scale—likely passing control program.

While nature bears a connotation of ancient origins, over geologically short, if anthropologically long, intervals, functional ecologies are time and again disassembled and reconstituted.

To scale us back into humanity’s present and pressing needs, in the short term, say, the next few hundred years, if ecosystems were to be conserved and agriculture integrated into local forest matrices, farmers can enjoy such autonomous ecosystem services largely free of charge.

How exactly such services—soil enrichment, water conservation, pest control, etc.—emerge from place to place and how farmers might harness them, for lack of a better term, will require much more of the kind of research the Perfecto and Vandermeer team has pursued. Forget genetic engineering. That’s all cave man brushing his teeth with a smartphone. This, on the other hand, is the cutting-edge research of the 21^st century.

Think on what modern agriculture does in contrast. It strips out the forest and destroys the kind of self-integrated services nature often offers, or, better said, embodies. Agribusiness acts as one big exclosure keeping larger fauna out while soil degrades and bugs continue to munch on. Farmers are left to reproduce these services by firebombing their crops and soils with destructive petrochemicals.

Along with the model’s unsustainability, which serves as a tautological rationale for cutting deeper into the forest that remains, intensive agriculture assigns farmers the absurdist task of nailing—like little Maxwell’s demons—every miniscule pest that comes along. What a waste of time and effort, especially as pests evolve resistance to pesticides as a matter of course. Instead farmers could have a whole ecosystem moving that bit of bother off their margins.

Farmer convenience, however, was never really the point of corporate agriculture. On the contrary. “Cargill is engaged in the commercialization of photosynthesis,” CEO Gregory Page said in 2008, “That is at the root of what we do.” By dispossession, monetizing the sun and soil and air out from underneath the farmer and the forest.