Mapping the Economic Costs and Benefits of Conservation

Picture yourself standing at the edge of Paraguay's Atlantic forest, satellite map in hand, looking at three possible routes for a wildlife corridor — three ribbons of land that could reconnect two large forest patches. On the map, they look nearly identical: same general length, same rough terrain, and the same green-brown patchwork of trees and farms. Then you run the numbers. One of those corridors is worth three times more than the other two, not in some vague ecological intuition but in dollars calculated hectare by hectare. That gap is the argument Robin Naidoo and Taylor Ricketts make in their spatial cost-benefit analysis of conservation in eastern Paraguay. And once you understand how they got there, it's hard to imagine making a corridor decision any other way. The core tension Naidoo and Ricketts identify is this: conservation resources are severely limited, yet conservation planners rarely use the cost-benefit frameworks that are standard in health, transport, and development policy. The reason isn't laziness — it's that valuing ecosystems is genuinely difficult. Services are spatially complex. Beneficiaries may be local, regional, or global. Land price data are often missing and must be modeled. All of this requires simplifying assumptions, and the authors are upfront about that. But their argument is that imperfect spatial cost-benefit analysis beats none at all. Their test case is the Mbaracayu Forest Biosphere Reserve in eastern Paraguay, within the Upper Paraná Atlantic Forest — one of the most threatened forest ecosystems on earth. The reserve spans roughly two thousand nine hundred and twenty square kilometers. In nineteen seventy-three, about ninety percent of the watershed was forested. By two thousand four, that had fallen to fifty-six percent, and outside the core protected area, the landscape was heavily fragmented. Three land uses dominated the unprotected areas: smallholder agriculture, large-scale cattle ranching, and soybean production. Large landowners — many of them Brazilian nationals — held thousands of hectares for ranching and soy. Indigenous Ache and Guaraní communities also lived in the region, with the Ache continuing to hunt and gather within the core reserve where permitted. The forest was under pressure from multiple directions, and the opportunity costs of keeping it standing varied enormously across the landscape. Those opportunity costs are the foundation of the analysis. Naidoo and Ricketts modeled the agricultural income foregone by keeping each hectare forested. Essentially, for each parcel, they summed across alternative land uses the probability of conversion multiplied by the expected net benefit of that use. They validated the resulting surface against actual property sale prices for twenty properties inside the reserve and found the model was well-calibrated using a discount rate between fifteen and twenty-five percent. They settled on twenty percent for all valuations. The resulting cost surface ranged from essentially zero in the core protected area, where no detectable deforestation occurred over the prior fifteen years, up to nine hundred and twenty-seven dollars per hectare in the eastern reserve, where soybean farming made conversion highly profitable. The average across all forests was about sixty dollars per hectare. That's a nearly three-order-of-magnitude range across a landscape only three thousand square kilometers in size. The spatial heterogeneity isn't noise — it's the whole point. Against those costs, the authors mapped five ecosystem services. The first was sustainable bushmeat harvest — the potential supply of wild meat to local hunters. They modeled habitat associations for game species using logistic regression and constrained predicted occurrence by minimum forest area thresholds so that small fragments couldn't support species that need large patches. They valued the resulting biomass at the avoided cost of replacing wild meat with domestic beef: one dollar and forty-four cents per kilogram. Mapped bushmeat values ranged from zero to eighteen dollars and fifty cents per hectare, averaging about fifteen dollars and sixty cents per hectare across the landscape. Sustainable timber harvest used published estimates of standing merchantable value for sixteen tree species, arriving at an average per tree value of six dollars and eighty-seven cents, then applied a conservative guideline of four trees per hectare on a thirty-year rotation. That produced a spatially uniform average of about twenty-seven dollars and sixty cents per hectare for forests where logging was permitted. Bioprospecting for pharmaceuticals — the potential value of plant species as drug precursors — drew on a benefits-transfer approach based on pharmaceutical companies' willingness to pay for marginal species. This came out small and spatially uniform: just two dollars and twenty-one cents per hectare. Existence value, the non-use willingness to pay of households for forests to simply exist, was estimated at five dollars per hectare per year based on debt-for-nature swap values, discounted to a present value of twenty-five dollars per hectare. And then there's carbon. This is where the analysis pivots. Naidoo and Ricketts measured tree plots across forest types to estimate aboveground biomass, converted to carbon using a factor of zero point five, and applied a conservative market price of two dollars and fifty cents per tonne of carbon dioxide, noting that European carbon markets were trading as high as twenty-five dollars and fifty cents per tonne at the time. Even at the conservative price, carbon storage dominated everything else. Vine undergrowth forests held enough carbon to be valued at over one thousand three hundred dollars per hectare. High forest came in at seven hundred and eighty-eight dollars. The average across all forest types was three hundred and seventy-eight dollars per hectare — more than ten times the next largest service. When you put the costs and benefits together spatially, the pattern is stark. Using only the locally accruing services — bushmeat, timber, and bioprospecting — benefits exceeded opportunity costs only in the core protected area and indigenous reserves. Adding existence value helped modestly, raising to nineteen percent the share of forests outside the core that passed the cost-benefit test. But once carbon entered the picture, benefits exceeded costs across ninety-eight percent of all forested areas. Carbon wasn't just the largest service — it was the one that made conservation economically defensible almost everywhere. Sensitivity tests confirmed the conclusion was solid: a twenty percent reduction in key parameters for local services shifted the result by only about five percent, and only a dramatic drop in carbon prices would substantially alter the spatial outcome. Now back to those three corridors. Naidoo and Ricketts applied their cost-benefit maps to the real decision of which corridor route to pursue between two large forest patches. Corridor one covered one thousand three hundred and ninety-eight hectares with conservation costs of about one hundred and fifteen thousand dollars. Total ecosystem service benefits came to roughly one point eight million dollars, leaving a net benefit of nearly one point seven million. Corridor two was slightly smaller in both area and cost, with net benefits of about one point five million. Corridor three, despite being the largest in total area, had costs of only thirty-seven thousand dollars, driven by low conversion pressure, and net benefits of nearly one point four five million. When carbon was included, all three corridors showed positive net benefits. The more revealing comparison came from looking at only the locally accruing services. Here, none of the corridors covered their costs from local benefits alone. But the shortfalls differed dramatically. Corridor one had a local cost gap of nearly ninety thousand dollars. Corridor two's gap was about sixty-two thousand. Corridor three had the highest local benefits at twenty-eight thousand dollars and the smallest shortfall — just nine thousand dollars. That's the corridor where local actors could almost cover conservation costs through their own use of the forest, and where international payments for carbon or existence value would close the remaining gap most efficiently. Naidoo and Ricketts are clear about what their analysis omits: watershed services, recreation, option value, and the spatial interdependencies between parcels were all left out. They couldn't model how human behavior would change over time, and their benefits-transfer approach blended marginal and average values in ways that introduce uncertainty. But they argue convincingly that the directional signal survives these limitations. The spatial map doesn't just answer which corridor to pick. It shows where development and conservation genuinely align, and where real trade-offs require targeted finance or policy action. They point to mechanisms like payments for ecosystem services, carbon markets, and international transfer schemes — already operating in Costa Rica, New York, and Australia — as tools to translate the value of distant beneficiaries' preferences into dollars that reach local landowners. The broader lesson is methodological. Conservation decisions have always involved implicit trade-offs. Naidoo and Ricketts made those trade-offs explicit, mapped them, and showed that the map changes the answer. Even with incomplete data, across a three thousand square kilometer landscape, spatial cost-benefit analysis picked out the right corridor — three times over. This lecture was created by ennepō. Go to https://ennepo.ai to Discover, Create and Follow the latest research in your field. Read when you can. Listen when you want to.