The Trouble with Ecosystem Services
When Pricing Everything Means Valuing Nothing
In the early 1990s, “bioprospecting” captured the imaginations of both the public and policy makers. Merck, the large US pharmaceutical company, entered into an agreement with Costa Rica’s National Biodiversity Institute (INBio), through which Merck would receive exclusive access to INBio’s natural product collections in exchange for a million-dollar up-front payment, in-kind contributions to research, and a promise of royalties in the event that commercial products were identified.1 INBio’s director speculated that Costa Rica’s bioprospecting earnings would soon exceed the $300 million per year it earned from coffee exports.2
Other companies followed suit, expanding their natural products research operations. US government agencies combined to sponsor an International Cooperative Biodiversity Groups program under whose auspices American pharmaceutical researchers were paired with source-country partners.3 Bioprospecting even inspired a movie — in the 1992 film Medicine Man, Sean Connery, then only three years removed from having been named People magazine’s “sexiest man alive,”4 played an ethnobotanist racing to find a cancer cure before the rainforest was felled around him. Bioprospecting was touted as a “win-win-win proposition”: new drugs would be found, developing countries would prosper, and the rainforests would be saved.
As the enthusiasm for bioprospecting grew, academics and advocates volunteered their recommendations for how the profits that would soon begin to flow could be divided equitably between the research scientists who developed new products from natural sources and the owners of the ecosystems that provided them. Google Scholar lists more than 150 works published before 2000 on “access and benefit sharing” for genetic resources.5 How were the nations of the global South, which had historically been exploited for their labor and raw materials, to avoid a new form of expropriation, “biopiracy,” the theft of their valuable biological patrimony?
Such concerns might have been well founded had the case for bioprospecting stood on solid ground. Regrettably, however, bioprospecting didn’t make economic sense: it rested on the belief that what was decidedly not scarce — countless genetic resources that were to be found in tropical rainforests the world over — was valuable. An asset is only valuable because it is scarce, though. The failure to appreciate this basic economic principle turned what was touted as a “win-win-win” into a “lose-lose-lose” for conservationists, developing countries, and pharmaceutical participants.
Bioprospecting was an early example of an appeal to an “ecosystem service” in an effort to motivate conservation. In its wake, the conservation community has turned its energies to other ecosystem services that place more emphasis on the benefits that preserving relatively undeveloped habitats would bring to the communities living in or adjacent to them. These include services such as water purification, pollination, pest control, and flood and storm protection. But the economic case for this tack is often weak also. When development pressures are high, it tends to be more cost-effective to rely on artificial substitutes for ecosystem services than to forgo converting land to agricultural or residential uses. Even when the argument can be made to retain some remnant areas of natural habitat to provide ecosystem services, it’s not clear that much meaningful conservation results. Trying to make nature valuable, it turns out, has had a disappointing track record.
Like many other young researchers, I was motivated by the desire to help save nature when I set out to work in environmental and resource economics in the early 1990s. The term “biodiversity” had been coined a few years earlier,6 and concern was growing over what was feared to be a “sixth extinction” crisis.7 At five points in the geological record, mass extinctions had been tied to astronomical or volcanic cataclysm. In the sixth extinction event, it was feared that millions of species might perish due to the increasing dominance of one: Homo sapiens.
What was to be done? At the time I began working at Resources for the Future, a nonprofit organization that does research on environmental and resource economics, in the fall of 1991, a new approach to conservation had been picking up steam for about a decade. In 1980, the International Union for Conservation of Nature released an influential report announcing that conservation for conservation’s sake was out; “explain[ing] the contribution of living resource conservation to human survival and to sustainable development”8 was in. The report argued that natural habitats and the organisms they support are worth more alive, as the sources of sustainable goods and services, than they are dead, stripped of timber and converted to other purposes. In the years that followed, conservation advocates and scholars gravitated toward the idea that, since appeals to altruism hadn’t worked, perhaps appeals to tangible self-interest would.
By the early 1990s, “integrated conservation and development projects,” or ICDPs, had come to embody these hopes. ICDPs often focused on schemes to market the products or services of tropical rainforests and other biodiversity hotspots to would-be consumers around the world. The most exciting ICDPs had to do with the use of genetic resources in pharmaceutical research. Nature is a treasure trove of genetic resources, a library of biochemical blueprints for the production of miracle drugs. Cancer drugs had been developed from the rosy periwinkle and the Pacific yew; diabetes medication from Gila monster spit; heart drugs from foxglove; quinine from the bark of the cinchona tree. Even good old aspirin traces its origins to willow bark.9 If pharmaceutical researchers could be compelled to pay for their access to genetic resources, the argument went, the rainforests could be saved.
The problem of biopiracy remained, however. That was the question to which I turned at the start of my career. I tried to approach the question logically: if the difficulty lay in compensating providers for the value of the genetic resources they offered, the first thing to think about was the value of genetic resources.
Economists argue that value is determined by scarcity. If there isn’t much of something relative to the demand for it, people will pay a lot for more of it. If it’s relatively abundant, they won’t. An argument I often heard as I began my work is that we should save the rainforests because they are likely home to millions of as-yet undiscovered species,10 and each one of those species might be the source of the next miracle drug. As I thought about this idea, though, it struck me that something wasn’t right. If there were literally millions of undiscovered species, pharmaceutical researchers would have lots and lots of different places to look for their next blockbuster. Some experts have written that the stock of genetic resources that could be employed in drug research “is so vast as to seem unlimited.”11 An “unlimited” resource is necessarily a cheap one. No one would be willing to pay a lot for access to samples from Costa Rica if she could explore equally promising leads from Nicaragua or Bolivia, Cameroon or Malaysia. Nor would she be willing to pay much to preserve one hectare of rainforest in Costa Rica if identical or similar samples could be found in thousands of other hectares of rainforest. It is evidence of the redundancy of many natural sources that one of the crucial steps in natural product research is “dereplication”: determining if a promising new “discovery” is, in fact, just another occurrence of a product whose attributes are already well known.12
In 1996 two colleagues and I translated these intuitions into a mathematical model with numerical estimates of value and published our results in a leading economics journal.13 From there, a typical pattern unfolded: People argued over our claim. A couple of authors revisited our calculations and said they were flawed.14 Another pair of researchers revisited our critics’ calculations, argued that they’d made a serious error, and found that our original results were actually pretty close to the truth.15 Researchers continued to cite and dispute this growing body of work, and the academic-industrial complex went on spinning out research on bioprospecting and genetic resources.
But an interesting thing was happening in the real world while we academic scribblers were squabbling amongst ourselves: nothing.
Bioprospecting didn’t take off. The Merck–INBio agreement was terminated many years ago.16 Merck gave away the collections it had acquired from INBio in 2011, 20 years after having contracted to acquire them, “as if,” one commentator remarked, “to mark the anniversary of its Costa Rican folly.”17 By 2013 it was reported that INBio was in deep financial trouble and was soliciting public donations to meet its payroll.18 Costa Rica never made more than a tiny fraction of what it earns exporting coffee from marketing its genetic resources. Eli Lilly, Bristol-Myers Squibb, and other large multinational companies also abandoned their international bioprospecting ventures.19 The titles of papers that appeared over the years track the progression of this disillusionment: in 1998, an article asked what happens “When rhetoric hits reality in [the] debate on bioprospecting”20; by 2003, another piece inquired, “Bioprospecting — Why is it so unrewarding?”21 In 2012, it was “A Bitter Pill . . . why haven’t drug companies — or conservationists — been able to cash in on nature’s pharmacy?”22
Maybe a more accurate summary of recent bioprospecting history would be to say that nothing good happened. Merck may have been lucky not to have found anything of value in Costa Rica. Advocacy groups were already launching a publicity war against them. Had Merck discovered a blockbuster drug, it likely would have been impossible to arrive at a compromise that would have satisfied their shareholders, their Costa Rican partners, and maybe most importantly, international NGOs who remained convinced that bioprospecting was going to provide hundreds of millions of dollars to poor tropical countries. Merck’s “win-win” had turned into a “lose-lose”: they came out empty-handed but still found themselves lambasted for alleged biopiracy.23
A similar furor arose over the issue of “prior informed consent” with one of the US government–sponsored International Cooperative Biodiversity Groups. Indigenous groups in the Mexican state of Chiapas protested that they had not given their permission for the collection and testing of samples. The venture was dissolved three years into its planned five-year course.24
In these and other cases, many people were upset by the idea that drug companies stood to make billions from the things they found in less wealthy countries while the suppliers of their samples were promised only a pittance. These arguments often revolved around the idea that multiple groups should be compensated for providing samples, based on the further claim that useful compounds might be identified in more than one place. But this reasoning just underscores the notion that if multiple potential sources exist for something, that thing is not scarce and, hence, not valuable.
Not long ago, I gave a talk that outlined these points. After I spoke, a man in the audience sought me out to set me straight. Just that week, he informed me, there had been a report of the development of a promising new compound that might meet the urgent need to develop effective treatments against antibiotic-resistant infections.25 What did I have to say about the fact that an immensely valuable new compound had just been isolated from a soil sample?
I thought for a moment before responding. “Are you familiar with the phrase ‘cheap as dirt’?”
In the decades since I first wrote on bioprospecting, I have continued to study the broader, and evolving, field of ecosystem services. I’ve served as an author in the Millennium Ecosystem Assessment, a review editor for the United Nations–sponsored project on The Economics of Ecosystems and Biodiversity, and an advisor to the World Bank’s program on Wealth Accounting and the Valuation of Ecosystem Services. I’ve continued to research topics such as the pollination, storm protection, and pollution treatment services of ecosystems, and served for several years as director of ecosystem economic studies in the United States Environmental Protection Agency’s National Center for Environmental Economics.
Over the years I’ve seen a lot of ideas come and go. Take ecotourism. As with bioprospecting, because conservation advocates saw that people spent a lot of money to visit natural wonders that housed unique forms of life, they inferred that the habitats supporting those natural wonders must also be tremendously valuable. But where are all those tourist dollars really going? Generally, not to compensate people for maintaining another hectare of rhinoceros or resplendent quetzal habitat. Rather, that money is usually directed toward the investments that airlines and hoteliers make to transport tourists to a remote locale and provide them with soft beds and hot showers when they arrive. Habitat, per se, is often sufficiently plentiful that marginal changes wouldn’t affect tourists’ experience perceptibly. I have known well-intentioned people interested in setting up ecotourism ventures in Venezuela, Sri Lanka, Thailand, and a host of other places around the globe. The list of fascinating places to visit remains almost limitless. As a result, like anything else that is available in almost limitless quantities, exotic destinations in and of themselves cannot command high values.
A similar critique could be made of nontimber forest product marketing, another ICDP case. People’s willingness to pay for products like tagua nut buttons has been sharply limited by the fact that synthetic substitutes are so abundant and cheap. Both ecotourism ventures and forest product sales also prompted concerns that they might lead people to “love nature to death.” A sort of Catch-22 arose: operations at a scale required for commercial viability would result in degradation of the very systems that the project was intended to protect.
As conservation advocates began to appreciate both the economic limitations and the potential unintended consequences of ICDPs, their focus began to shift. With ICDPs, the idea was to identify some product or service that large areas of extant habitat could provide to international markets. In more recent years, ecosystem services have been touted more for the benefits they could provide people who live in, or adjacent to, patches of remnant ecosystems. In the words of a commentary written by several leading ecologists:
Conserving nature out of moral obligation is a luxury most simply cannot afford. . . . In the face of a sea of poverty, demonstrating the ignored links between nature and elements of well-being — safe drinking water, food, fuel, flood control, and aesthetic and cultural benefits that contribute to dignity and satisfaction — is the key to making conservation relevant.26
Much of this new emphasis rested on tangible benefits that accrue to those in the vicinity of the ecosystems providing them. In fact, those ecosystem services might become all the more important and valuable the more people there are near the ecosystems providing the services, or the more intensive their activity. When is safe drinking water important, in other words? When there are large numbers of people who need it, and large numbers of people (and/or herds of their domestic animals) despoiling it. When is flood control most valuable? When there are large concentrations of people and/or valuable structures at risk.
The problem with bioprospecting and some other types of ICDPs was that the marginal value of the habitats it was hoped they would preserve was negligible. They remained so extensive globally that they were not scarce relative to demand. The opposite problem can beset ecosystems preserved to provide local services. Ecosystem services may be valuable, yes, but they may be most valuable when it would also be most expensive to set aside land to provide them. If ecosystem services are valuable enough to justify setting aside some land to provide them, then it may be because it’s not necessary to set aside much land to provide them.
Pollination, an ecosystem service that has been studied extensively, illustrates this point.27 Wild insects still pollinate crops in some areas, but in much of the world, farmers whose crops require pollination (many important food crops don’t) rely on rented European honeybees (Apis mellifera) that are moved from farm to farm as crops flower and require their services. A large fraction of commercially available honeybee colonies in the United States, for instance, are trucked to California every February to serve the state’s almond crop.28 In hopes of motivating conservation by appeals to ecosystem services, however, some biologists and advocates have argued that such farmers should instead set aside land to support native insects that could pollinate almonds.29
While honeybees are trucked away to serve the next crop that comes into bloom, native pollinators would need to have habitat maintained for them year-round. Land in California’s almond-growing areas can sell for $25,000 an acre or more, and almond farmers pay about $450 per acre to rent bees to pollinate their crop.30 In deciding whether to maintain remnant areas of natural habitat on their land or to instead rent honeybees, therefore, the almond farmer must ask herself, “How much $25,000-per-acre land am I willing to take out of almond production so I can save $450 per acre every year?” The economically rational answer is probably “not much.”31 Intuitively, farmers wouldn’t be willing to take much land out of almond production if their reason for doing so was to enhance almond production. In many instances, producers may find that employing substitutes such as truck-borne honeybees rather than native insects to pollinate crops, levees rather than riparian vegetation to control floodwaters, and wastewater treatment plants rather than wetlands to treat pollution is more cost-effective than setting aside extensive areas of expensive land to provide such services. When natural measures are cost-effective, on the other hand, it’s likely because not much land is required to provide them.
I have just argued that artificial substitutes may be more cost-effective than natural ecosystems when the opportunity cost of retaining land in natural ecosystems is high. What about where land remains cheap? While one might ask how badly we need to devise sophisticated arguments to “save” places that may not be under much threat, natural ecosystems are threatened in some areas where land remains cheap. In parts of the developing world, for instance, people are razing forests to establish subsistence farms. But will encouraging them to rely on the contributions of natural ecosystems stem the biodiversity loss they are causing? The result could be a patchwork of fields, interspersed with remnants of natural habitat, that spans more total terrain for each kilogram or calorie of food grown. It would likely be wiser to instead hasten subsistence farmers’ integration into the modern economy, in which high-tech agriculture constitutes a lower overall threat to nature.
It is true that there is one instance where the provision of ecosystem services has justified setting aside very large, and often fairly expensive, portions of the landscape. These services are provided by a relative handful of species in ecosystems that have been specialized for their production. Such ecosystems are known as “farms.”
I offer that observation with tongue in cheek. The expansion of farms is not generally regarded as a boon for conservation — quite the opposite. If a continuum exists between “natural” and “human” landscapes, those who favor the preservation of more of the former should consider what outcome an emphasis on ecosystem services might motivate.
Let me illustrate what I mean with another example drawn from pollination. A California almond grower recently experimented with indigenous pollinators as an alternative to continuing to rent transient colonies of European honeybees.32 The Wonderful Company produces almonds on nearly 20,000 hectares of land in California, and they have historically rented about 80,000 colonies of honeybees every year, at an annual cost of $10 million.33 Given those costs, the company looked for an alternative in the blue orchard bee (Osmia lignaria). The blue orchard bee is a prolific pollinator, but it has some drawbacks. Unlike honeybees, which return to their hives, blue orchard bees tend to disperse. And unlike honeybees, which, after they have returned to their hives, are trucked around the country from one crop to another, blue orchard bees must be provided with alternative fodder sources when almonds are not in season. Blue orchard bees are also prey for mice, toads, and other animals.
So in order to rear blue orchard bees to provide the ecosystem service of pollination, the Wonderful Company: carefully weeded an eight-hectare area and planted a select set of the types of flowers that blue orchard bees prefer to feed on; enclosed the area with netting to keep the blue orchard bees in, and other insects out; established nesting areas for the bees; controlled mice and toads that might otherwise have eaten them; and commenced a selective breeding program in hopes of creating a line of domesticated blue orchard bees that will not stray from the area where farmers would like them to make their home.
What would you say? Was the company setting aside natural habitat to provide an ecosystem service? Or might this outcome be better described as a farm-within-a-farm for the propagation of a new breed of livestock?34
I have a quarter-century of experience now in studying why and how appeals to ecosystem services might motivate conservation. Sadly, such appeals don’t seem to have accomplished much so far. Programs to capitalize on ecosystem services have often proved ineffectual, and sometimes even counterproductive.
One might entertain two hypotheses based on this history. The first is that conservation advocates simply haven’t yet hit on the right idea. In this view, ecosystem services are systematically underappreciated, and if we just keep looking, we’ll identify values that will convince decision makers to save far more of the natural world.
My experience leads me to an alternative hypothesis. Ecosystem services are not generally underappreciated. What is underappreciated is, rather, basic economics. This is the common thread running through all the disappointing experiences to date with finding ways to make nature pay for itself.
Conservation advocates haven’t necessarily gotten it all wrong. If by “nature” we mean the ineffable grandeur and exquisite wonder of creation, nature is becoming more scarce in a world increasingly dominated by humans. Because nature is becoming more scarce, it’s becoming more valuable. But nature’s services, the demonstrable and tangible benefits we get from harnessing natural assets for human purposes, are, by and large, not growing more scarce. In fact, it’s generally the opposite.
Bioprospecting couldn’t live up to its billing because the biological diversity it sought to exploit remains abundant. That’s only half the story, though. Synthetic chemistry also provides ever more substitutes for natural sources for drug discovery, making those sources less and less valuable. It’s the same for a variety of other ecosystem services. Can natural ecosystems filter water, protect against hurricanes, and pollinate crops? Sure they can, but there’s a good reason we see so many sewage treatment plants, seawalls, and peripatetic honeybees. These constructed or managed alternatives often provide the same services at a lower cost, particularly when that cost includes the value of land that would otherwise have to be devoted to providing them.
The response to such facts has too often been to elide them. A common trope in commentaries on ecosystem services is that they are “externalities.” In economics, an externality arises when one person acts and others experience costs or benefits of her action: if a landowner cuts down the trees on her land, others suffer from the loss of ecosystem services they provided. Just because there might be externalities doesn’t mean that there are, though, and just because there are externalities doesn’t necessarily mean that they are large enough to justify major changes in land use policy. Despite the reams of papers written on ecosystem services, hard evidence concerning their value is difficult to come by. By my reading there are very, very few instances in which appeals to ecosystem service values would have the significant impact on conservation practice some advocates have imagined.
I do not want to overstate my point. Ecologists, economists, and others should, of course, study natural assets closely when we risk losing some forever. Nor is there anything necessarily wrong with adopting “natural” approaches if they can be demonstrated to make agriculture and industry more productive with fewer harmful side effects.
Most importantly, nature is worth saving. But this means it is far too important a challenge not to address it seriously. To the extent that preserving nature means more than incorporating its vestiges in managed landscapes, we’ll likely find it more effective to continue to decouple human activity from nature than trying to find and exploit synergies between the two. Misunderstanding problems can lead to proposing solutions that don’t work. Too many approaches to conservation come off as ecological get-rich-quick schemes, wishful long shots with little realistic hope of motivating the types of large-scale conservation efforts that we need.
Read more from Breakthrough Journal, No. 9
Featuring pieces by Rachel Laudan, Alan Levinovitz,
Mark Sagoff, Fred Block, Julie Guthman,
Brandon Keim, and more.
 Michael D. Coughlin, Jr., “Using the Merck–INBio Agreement to Clarify the Convention on Biological Diversity,” Columbia Journal of Transnational Law 31, no. 2 (1993): 337–75.
 INBio director Rodrigo Gámez Lobo, interview by Bob Carty, “Pharmaceutical Prospecting,” Living on Earth, September 4, 1992, http://loe.org/shows/segments.html?programID=92-P13-00036&segmentID=1.
 See “International Cooperative Biodiversity Groups (ICBG),” NIH Fogarty International Center, https://www.fic.nih.gov/Programs/Pages/biodiversity.aspx.
 People, December 18, 1989.
 Google Scholar search for “access and benefit sharing” and “genetic resources,” restricting the time period to 1999 and before.
 The history of the term and idea are explored by David Takacs, The Idea of Biodiversity: Philosophies of Paradise (Baltimore: Johns Hopkins University Press, 1996).
 See Elizabeth Kolbert, The Sixth Extinction: An Unnatural History (New York: Henry Holt, 2014). While Kolbert’s work may have popularized the notion of a “sixth extinction,” she certainly did not originate the idea. References to a “sixth extinction” event were already common when I began working on biodiversity.
 International Union for Conservation of Nature (IUCN), World Conservation Strategy: Living Resource Conservation for Sustainable Development (1980), 4, https://portals.iucn.org/library/efiles/documents/wcs-004.pdf.
 Richard Conniff provides a list of examples of products derived from natural sources in “A Bitter Pill,” Conservation, March 9, 2012, http://www.conservationmagazine.org/2012/03/a-bitter-pill/.
 It seems that estimates of the numbers of species fluctuate between sources and over time. At the time I was working, taxonomists were excited by the discovery of numerous new species in rainforest canopies, and I would see some estimates in the tens of millions. It appears numbers may have come down somewhat recently. A 2011 paper estimates about 6.5 million on land, with another 2.2 million in the oceans. See Camilo Mora, et al., “How Many Species Are There on Earth and in the Ocean?” PLoS Biology 9, no. 8 (2011), https://doi:10.1371/journal.pbio.1001127.
One thing that doesn’t seem to have changed, though, is the extrapolation from incomplete data. Mora et al. estimate that 86 percent of terrestrial species and 91 percent of marine species have yet to be described.
 Quoted in David G. I. Kingston, “Modern Natural Products Drug Discovery and Its Relevance to Biodiversity Conservation,” Journal of Natural Products 74, no. 3 (2011): 496–511.
 Ibid. “One of the biggest concerns in natural products research is that after so much study many of the compounds in a given extract may well be known compounds, leading to much wasted effort in the search for new bioactive compounds.”
 R. David Simpson, et al., “Valuing Biodiversity for Use in Pharmaceutical Research,” Journal of Political Economy 104, no. 1 (1996): 163–185.
 Gordon C. Rausser, et al., “Valuing Research Leads: Bioprospecting and the Conservation of Genetic Resources,” Journal of Political Economy 108, no. 1 (2000): 173–206.
 Christopher Costello, et al., “Search, Bioprospecting and Biodiversity Conservation,” Journal of Environmental Economics and Management 52, no. 3 (2006): 615–626.
 Edward Hammond writes that “by 2008 Merck walked away from the agreement” (“Amid Controversy and Irony, Costa Rica’s INBio Surrenders Biodiversity Collections and Lands to the State,” Third World Network, April 2, 2015, https://www.twn.my/title2/biotk/2015/btk150401.htm), and Conniff (see note 9) writes, “In 2008, Merck quietly abandoned its search for new drugs from the natural world.” Others report that Merck was no longer contracting for new samples after 1999: “A second two-year contract, with similar terms, was signed in 1994 and a third in 1997. However, in 1999, Merck ended its contract with INBio to focus on the analysis of samples.” In Life as Commerce: Bioprospecting in Costa Rica, Global Forest Coalition (2008), http://globalforestcoalition.org/wp-content/uploads/2010/12/Costa-Rica-summary1.pdf.
 Conniff, “A Bitter Pill.”
 Hammond, “Amid Controversy and Irony, Costa Rica’s INBio Surrenders Biodiversity Collections and Lands to the State.”
 Conniff, “A Bitter Pill.”
 Colin Macilwain, “When Rhetoric Hits Reality in Debate on Bioprospecting,” Nature 392 (1998): 535–540.
 Richard D. Firn, “Bioprospecting — Why Is It So Unrewarding?” Biodiversity and Conservation 12, no. 2 (2003): 207–216.
 Conniff, “A Bitter Pill.”
 See, e.g., Vandana Shiva, Biopiracy: The Plunder of Nature and Knowledge (Boston: South End Press, 1997).
 “Mexico Biodiversity Project Cancelled Following NGO Criticism,” International Centre for Trade and Sustainable Development, November 22, 2001, https://www.ictsd.org/bridges-news/biores/news/mexico-biodiversity-project-cancelled-following-ngo-criticism.
 Jenna Gallegos, “In the Hunt for New Antibiotics, Scientists Hit Pay Dirt, Washington Post, June 15, 2017, https://www.washingtonpost.com/news/to-your-health/wp/2017/06/15/in-the-hunt-for-new-antibiotics-scientists-hit-pay-dirt/?utm_term=.cca52a9ae5a8.
 Paul R. Armsworth, et al., “Ecosystem-Service Science and the Way Forward for Conservation,” Conservation Biology 21, no. 6 (2007): 1383–1384.
 Pollination is mentioned prominently in Armsworth et al.; there are around 42,000 Google Scholar citations of “ecosystem services” and “pollination.”
 Tom Philpott, “Holy Shit! Almonds Require a Ton of Bees,” Mother Jones, May 25, 2015, https://www.motherjones.com/food/2015/05/almonds-now-require-85-percent-us-beehives/.
 Claire Brittain, et al., “Synergistic Effects of Non-Apis Bees and Honey Bees for Pollination Services,” Proceedings of the Royal Society B: Biological Sciences 280 (2013).
 Roger A. Duncan, et al., Sample Costs to Establish an Orchard and Produce Almonds, University of California Cooperative Extension (2016), https://coststudyfiles.ucdavis.edu/uploads/cs_public/27/59/27599258-3e53-4e5f-aefb-08d12f5b0cb9/16almondssjvnorthfinaldraft81116.pdf.
 In a recent paper, I find that under what I consider to be very generous assumptions an almond farmer might devote, at most, about one-eighth of her land to native pollinators. R. David Simpson, “Conservation Incentives from an Ecosystem Service: How Much Cropland Should Be Set Aside for Wild Pollinators,” forthcoming in Environmental and Resource Economics.
 Todd Fitchette, “Large Almond Grower Looks to Wild Bees for Pollination,” Western Farm Press, March 31, 2015, http://westernfarmpress.com/tree-nuts/large-almond-grower-looks-wild-bees-pollination.
 Paige Embry, “Building a Backup Bee,” Scientific American, March 2018, 67–71.
 As it turns out, the question is now moot: in February of 2018 — just as a new article was going to press in Scientific American on its efforts — the Wonderful Company announced it was ending its eight-year-long effort to domesticate blue orchard bees.