The route to collect data at a remote jungle outpost passes through a landscape devastated by illegal mining and a string of mining boom towns far from the reach of national authorities.
(This story describes field work in the Amazon rainforest carried out by SkyTruth’s Bjorn Bergman along with the Peruvian Service for Natural Protected Areas by the State, thanks to support from Conservation X Labs’ Amazon CoLab. It is also available in Spanish.)
On the wall of the headquarters for Peru’s Amarakaeri Communal Reserve, park staff had posted a map peppered with colored squares, which identified areas under threat. Most areas were on the edges of the reserve where miners and coca growers pressed against its boundaries.
From only four control posts around the reserve’s 300-mile perimeter, a small team of dedicated park staff is tasked with protecting this vast area.
The Amarakaeri Reserve not only boasts some of highest biodiversity in the Amazon, but also has a number of enigmatic archeological sites such as a stone face looking solemnly out from a cliff deep in the jungle. The reserve also has the important role of supporting ten surrounding native communities of the Harakbut, Yine, and Matsiguenka ethnic groups.
I had been invited to the reserve by Peru’s Service for Natural Protected Areas (SERNANP)—a Peruvian government agency that works on conservation issues—along with SERNANP’s remote sensing specialist for a field testing plan supported by Conservation X Labs as part of their Amazon CoLab program. In 2020 our Project Inambari won support at the Artisanal Mining Challenge run by Conservation X Labs to create a satellite image-based monitoring platform to help communities, governments, and conservation groups detect mining operations in critical forest environments in the Amazon Basin.
Since then, we have been developing an automated detection model for mined areas while also building user functions that make monitoring easier. We hoped to create a model that would distinguish areas cleared for mining from those cleared for logging and agriculture, allowing regulators to focus specifically on new mining threats where timely intervention is critical. So it was exciting to have the opportunity to observe mining activity firsthand and compare our platform with data collected in the field. Our goal for the trip was to acquire very high-resolution drone imagery and then compare that with our satellite data sources. These measures would help validate our analysis of satellite imagery and support training a model to automatically detect such sites.
On the afternoon before our trip to the reserve we were sitting at a conference table at SERNANP’s regional headquarters. SERNANP staff had disassembled a large, fixed-wing drone and were busy checking each component and charging up the drone’s batteries. The SERNANP remote sensing specialist had long experience flying drones over areas of southeast Peru affected by informal mining. This particular drone, called a VTOL (vertical take-off and landing), had the advantage of being able to launch vertically from a confined area and then fly horizontally at high speed on fixed wings. This was one of several drones donated by the nonprofit Conservación Amazónica (ACCA) to help monitor reserves in the area.
Reserve staff debated some possible flight plans. They had to keep in mind both the total range of the drone (about 60 kilometers, or 37 miles) and carefully consider any areas where changes in elevation could threaten the drone. They settled on some areas of concern within range of one of SERNANP’s control posts on the eastern edge of the reserve.
It was the end of October 2021, and we were racing to complete flights before the rainy season; once the rains started, it could take until May for clear weather to return. Logistics for the trip were complicated. In addition to hauling the drone in its bulky case out to the control post, we would need to load up on supplies. These included cooking gas and fuel for a generator—the only functioning source of electricity at the control post. All of this would have to be repeatedly packed and unpacked as we crossed the Inambari River by boat and switched vehicles.
After spending the afternoon and evening driving around Puerto Maldonado—the sweltering capital of Peru’s southeast Madre de Dios department, Peru’s term for state—to make purchases we got all of the supplies stowed at SERNANP headquarters and turned in to get some rest before the next day’s early start.
The following morning we headed out from SERNANP headquarters, five of us packed into a double cab pickup with the drone and supplies tied down in back. Heading southwest, we began our trip along a modern two-lane road dubbed the Interoceanic Highway. Since its completion in 2010 the highway has transformed this once remote region by opening it up for thousands of migrants from the Peruvian Andes. Seeking economic opportunity, these migrants mostly take on work that exploits newly cleared forest lands. The highway continues to the Brazilian border where it connects with a Brazilian road network going all the way to the Atlantic coast. Costing almost $2 billion dollars to build, the highway has failed to fulfill its proponents’ expansive visions of stimulating trade between the Atlantic and Pacific coasts. Moreover, it has had many unintended consequences, including pushing the agricultural frontier into a once pristine area of the Amazon where new colonists are coming into contact with some of the Amazon’s few remaining isolated tribes.
Initially, we passed through farm fields and cattle pastures; by midmorning the scenery took a dramatic turn. Vast mining pits opened up on either side of the highway with groups of standing dead trees. In some of the pits, miners with high pressure hoses blasted away at the thin tropical soil, dislodging a flow of sediment from which tiny specks of gold would be extracted. This area, called La Pampa, is perhaps Peru’s most infamous jungle mining site.
In 2019 when Peru hosted the Third Latin American and Caribbean Congress of Protected Areas, I watched former Peruvian President Martin Vizcarra take the stage to pledge an end to mining in La Pampa. A swift and effective major military operation followed his declaration. However, now four years and four presidents later, it’s clear that maintaining the political will and financial resources to keep a military presence in La Pampa will be a challenge.
The Interoceanic Highway we were traveling along also marked a regulatory divide: Mining is strictly prohibited to the south, whereas a legalized mining corridor to the north contains concessions in the process of formalization. Inaction by some government agencies has resulted in repeated extensions of the permitting process, effectively allowing mining operations to remain in a legal limbo as they operate without regulations. The fourth extension, for three additional years, was approved in November 2021.
As we drove past the gaping pits of semi-legalized mining operations on the north side of the highway, I asked about the prolonged formalization process. One unfortunate consequence is that mining operators have an incentive to deforest as large an area as possible before an environmental plan is approved. After approval, they would be free to work on already disturbed land but would need to conserve undisturbed forest.
Particularly shocking, given the destruction of the landscape and forest cover, was a complete lack of any mandate for post-mining restoration. Peru’s forest service (SERFOR) estimated that it would take 12 years and $29 million to restore 11 hectares (27 acres) of mined lands and toxic pools in La Pampa. If miners were actually forced to set money aside to pay these restoration costs, it’s doubtful any of their operations would be economically viable. Without considering the real costs of land recovery, the current formalization process is largely farcical and represents a sad capitulation in the face of pressure from miners, as well as limited and inconsistent government regulation.
As we continued down the highway into the heart of La Pampa, a vast informal settlement hugged both sides of the road. A shantytown grown into a small city, the settlement operates with almost no government presence or services. Inhabitants had drilled shallow wells for drinking water, pulling up water polluted by the toxic mining pits surrounding the city.
We sped through La Pampa—the park rangers preferred to stop for food and further supplies in a remote town where there was less risk of antagonism directed at those traveling in a government vehicle. The risks were very real: Just a few months before this trip a park ranger confronting miners over an incursion into the Amarakaeri Communal Reserve had been beaten. Our route took us on to Mazuko, an older mining town on the Inambari River near the confluence of two highways descending from the Andes—one route climbed steeply into the high mountains of Peru’s southern Puno department, while the other continued west toward Cuzco. At the time, a strike by coca growers had blocked the highway up to Puno. Coca has been grown in the high Andean slopes for millennia as a traditional stimulant. However, at lower elevations it’s processed into paste for the cocaine trade. Clandestine plots in the jungle, and the violence associated with the drug trade, now pose another major threat to remote, protected areas in Peru.
At the banks of the Inambari River we had to leave SERNANP’s truck behind since there was no bridge. We loaded the drone and our supplies into a boat that would ferry us across. On the stony bank on the far side we quickly found a local truck driver to take us on to Huepetuhe, the last mining town before our destination in the Amarakaeri reserve. We then were off the highway on a gravel track, which rose steeply from the river. Despite being off the highway, traffic was thick and seemed to consist almost exclusively of fuel tanker trucks.
Continuing eastward, the park rangers explained that we had entered another mining area where, over several decades, heavy machinery had torn through the landscape of the Huepetuhe River Valley. As we climbed up into the hills on the north side of the valley, the vegetation parted, giving us a stunning view down into the valley below. A wide area of barren mud flats surrounded the tiny trickle of water from the river. On the southern bank, mountains of dirt and gravel stretched out as far as the eye could see. Most were abandoned, the remnants of decades of mining activity. But a few showed active mining with massive excavators tearing through the soil; a few blue tarps nearby offered miners shelter from the beating sun.
A few hours later we descended into the town of Huepetuhe. Starting at the edge of Huepetuhe, I counted more than a dozen gas stations, an astounding number for a small town at the end of a dirt road. Near the town square we changed vehicles and picked up food for lunch, in addition to other important supplies. For the last leg of the trip we left the road entirely, driving along the wide mud flats on the river bank.
As we drove through the mud flats, we frequently passed heavy machinery. Once we had turned away from the river, our track cut through areas of active mining with backhoes shifting mounds of dirt and gravel just a short distance away.
Eventually, the track narrowed, and we ducked into the cover of the trees. Vines and branches pressed in from all sides, and the truck drove in and out of a stream bed. After a short ascent up through the forest, we saw a clearing where a simple roof covered a storage area stacked with gas cans and boards. There the track ended, and we would have to continue on foot.
We unloaded the supplies and ate a late lunch with the food we had picked up in Huepetuhe. We sat on logs and plastic barrels, passing around cups of coca cola and trying to ignore the biting insects. A placard warned that we were at the edge of the Amarakaeri reserve, but we still had a steep climb ahead of us before we reached the control post. We ate quickly then began carrying the supplies up in shifts, leaving them where the trail flattened out.
As we walked back to pick up another load, I saw three men coming up the trail; they had picked up some of the cases of water we had left below. One was dressed formally in a dress shirt and trousers, which made him stand out oddly in the heat and dirt of the jungle path. I asked—hopefully—if some of the technicians from Administration Contract Executor or ECA (the community management organization that partners with SERNANP to administer the reserve) had been able to join us for the drone flight. No, they replied, those are miners.
We had passed many miners working near the reserve, but I hadn’t expected to meet them in person at the park control post. I wondered if our trip had been getting some unwanted attention as we had made our way through the mining towns. The drone itself, in a huge oblong case, was now being carefully carried up to the small grassy area in front of the control post. Now that we were on site, we could begin the drone mapping flights that were the core mission of our trip. Clear flying weather was crucial, so SERNANP’s remote sensing specialist decided to immediately unpack the drone for launch. This would be an anxious moment for us all: Any mishap on this first flight could result in damage or loss of a valuable piece of equipment—our entire trip would be wasted. In the lengthening afternoon shadows, an unexpected audience of three visiting miners watched as we prepared to send the drone skyward.
The coming days at the outpost would not only test our ability to put our drone mapping plan into practice, but also would show the realities of being on the front lines in one of the world’s most critical environmental conflicts. A conflict far from the public eye, where park rangers at a remote outpost remain determined to hold the line despite limited resources and being vastly outnumbered.
Over the next few days we would wait with bated breath as the drone went on six long distance mapping flights beyond communication range with the outpost relying on preprogrammed instructions for completing its survey grid and returning. Then on dark tropical evenings we would pour over the photos collected by the drone on our laptops, the rangers checking for recent incursions and expanding mining within the park buffer area.
I heard from the park rangers about their experiences living at this beautiful and isolated location and the vast remote park they guarded. They told me about the long trek to visit the mysterious stone face in the forest and accounts of other faces deep in the jungle from the local Harakbut. They mentioned ideas for opening an adventure tourism circuit to bring revenue to this rarely visited park. They talked about the difficulty of access and resupply of the outpost, particularly in the rainy season, and occasionally having to ask for assistance with transport from the nearby miners. We discussed the ten native communities surrounding the reserve and the internal divisions they now faced as some community members looked to profit by cooperating with miners or even taking up mining themselves.
One afternoon we climbed up to a lookout point and gazing southwest above the reserve’s expanse of pristine forest we saw the distant snow-clad peaks around Ausangate, the sacred mountain of the Incas. Looking in this direction it was easy to imagine all of the hidden wonders that might lie below those miles of shimmering green canopy and not think of the sterile mud and gravel pits devouring the park’s buffer zone from the other side.
The next morning we were back on the track leaving the reserve. As our truck bounced across a streambed the park rangers remarked at the now reddish brown water which had been clear only a few days before. While we’d been doing the drone flights the new miners we’d seen had also been busy. The new sediment in the water was another stark reminder that the threat from mining was drawing closer day by day, and here at the edge of the wilderness only a handful of determined park staff in a remote outpost stood in its way.
The Amazon CoLab was made possible through a Global Development Alliance with the support of the Gordon and Betty Moore Foundation, Microsoft, Esri, and the United States Agency for International Development (USAID) and centers its efforts on solutions for the Amazon.
I am grateful for the opportunity to accompany SERNANP staff on this trip. For their security, we are not identifying them.
The opinions expressed in this article are the sole responsibility of the author and do not necessarily reflect the views of SERNANP, or the Amazon CoLab Program of Conservation X Labs.