This is currently a fringe issue in EA. We think it is important that more individuals within EA become fluent in the ways humanity depends on the natural world for basic needs such as clean air, water, and food. If you have interest, we ask that you read or skim this short summary of key aspects of this complex issue. We link to further analysis or discussion in certain places to further clarify our perspective or more fully summarize research.
We would value hearing comments, questions, and thoughtful critiques. If you’d like to track developments on this issue and/or get involved, you can also join this Telegram group or reach out to one of the authors.
- Ecosystems provide essential services for humans and other life forms like food, clean water, and clean air.
- As ecosystems become less diverse, they become less resilient and more prone to collapses and their cascading effects.
- A new EA-aligned initiative could provide uniquely valuable insights to steer resources towards the most effective interventions which are most likely to avoid thresholds and protect the most important ecosystems.
We need resilient ecosystems.
'We' meaning humans and all other life forms.
'Need' because they provide many essential resources for life.
'Resilient' meaning biodiverse, anti-fragile, self-sustaining, able to adapt to or withstand disturbances.
'Ecosystems'... highly complex, often densely interrelated systems of hundreds to millions of individuals.
Humans continue to need clean water, clean air, and food to survive. Almost all food sources are directly reliant on supporting ecological systems. Natural earth systems that have evolved over millennia are still the primary reasons almost all humans alive today have access to clean air, water, and food. These systems are complex, intertwined with our lives, and capable of being destroyed.
Generally, the more diverse an ecosystem is, the more resilient it is. Lower bounds on the amount of biodiversity ecosystems can lose and remain functioning have not been established. If we were to lose access to basic needs, mass loss of life would ensue and civilization would be in jeopardy. History has demonstrated the fragility of civilizations when cut off from access to basic needs (Examples: 1, 2)
Human Fragility Due to Ecosystem Degradation
Humans have not yet proven their ability to engineer isolated (closed-loop) systems that can reliably meet these needs at-scale or for long durations. Until we have more reliable proof-of-concepts around manually-engineered sustainability, it is wise to operate with the understanding that currently ‘there is no Planet B’.
Natural ecosystems are complex adaptive systems and generally evolve to be quite robust. However, they can contain “tipping points” where they will suddenly accelerate (“collapse”) into an alternative, often undesirable, state. When this happens, most species living in the previous ecosystem can’t adapt to the alternative environment and die out. Many coral reefs, kelp forests, marshlands and other ecosystems have demonstrated that this can happen quite quickly and without direct manipulation. Crossing these ecological thresholds can have significant impacts on surrounding ecosystems and also impact larger earth systems such as groundwater availability.
A sudden undesirable change of ecosystems is dramatic, but the deteriorating health of ecosystems also widely reduces welfare. For humans this means worsening air quality (ex: wildfires), higher concentrations of toxic chemicals in fish and bodies of water, lower yields of wild catch, and less efficient food production. This study suggests there is a $6.3 trillion per year cost due to lost ecosystem service, or 10% of GDP. For non-humans, this usually also leads to death, lower quality of life and more suffering.
Decreasing Biodiversity Leads to Decreasing Resiliency
The diversity of life on Earth is also decreasing, at a rate that is 100-1000x greater than it has been over the millions of years preceding industrialization. This puts humans on track to cause the sixth mass extinction on Earth. At the ecosystem level, this irreversible degradation of resiliency can be seen in the 50% decline of global coral reef coverage, destruction of wetlands across China, a recurring 4500 sq mile ‘dead zone’ in the Gulf of Mexico, and millions of acres where temperate forests, rainforests, and prairie are now reduced to a fraction of the diversity of life they used to contain.
77% of land is modified by human activities, setting the stage for increasingly fragile ecosystems and earth systems.
Need for Strategic Prioritization of Ecosystems
Certain ecosystems are more important than others, though we need more holistic analyses to determine which are most deserving of resources to protect them. Further, certain ecosystems are more important, neglected, and capable of being prevented from collapse than others (see this meta analysis). However, initiatives to support and execute coordinated efforts to protect key ecosystems are underfunded.
The Problem As It Relates to EA
- There is a need for an initiative that identifies the most cost-effective, tractable, long(er) term interventions that can be implemented in systemic and coordinated ways to protect key ecosystems. We are seeking EA support to launch this initiative. See ‘Neglectedness’ for more context.
- Disturbed ecosystems contribute to global instability, increase socio-political tensions, and increased frequency of natural catastrophes when compounded with other risks would contribute to civilization collapse.
- Longterm time horizons are underfunded and under implemented in environmentalism.
- Core EA organizations have not yet devoted resources to accurately estimate how ecosystem resiliency will impact humans in post-catastrophe scenarios, which is another reason to strategically protect key ecological systems.
We recognize the need to define key words and how they are being used in this paper. We have included definitions in an appendix for the following terms: biodiversity, ecosystems, ecosystem services, degraded ecosystem, resilience and NCPs.
Summary of Neglectedness: There is a lack of rigorous, strategic research & intervention analysis at the ecosystem level. Of the key criteria that we arrived at, very few organizations meet more than 2 of them. We see a distinct gap in the coordination of resources around this cause area that could be met by a small, well positioned team of researchers.
While there are large philanthropic players, academic researchers, and international organizations invested in addressing related issues, we believe there is a specific need for an initiative that uses an EA-aligned framework to enhance the research and ground-level work being done on this topic such that (1) more effective interventions are prioritized and (2) other EA cause areas thoroughly consider the compounding risk that degraded ecosystems present.
Fundamentally, there will be certain strategies, and organizations, that are vastly more effective than others. We feel confident that a rigorous, EA-style analysis has not been done on the most effective techniques to address the direct and indirect risks associated with biodiversity loss. We have identified certain environmental organizations which have begun some related analyses but none feel sufficiently rigorous or widely recognized.
Criteria for Effective, Long(er) Term Ecosystem Resiliency Work
- Thoroughly assess intervention from an INT (or similar) framework such that they prioritize effectiveness by rigorous standards.
- Assess interventions at bioregion or global scales and strive for coordination at this level.
- Assess strategy on longer time horizons (i.e. 50, 100 or 10,000 years away)
- Consider NCPs* in addition to global stabilizing influence and other key areas.
- Foster strategic partnerships and international support (i.e. transparent analysis, standardized reporting, open collaboration)
*NCP stands for Nature Contributions to People, and can be thought of as:
NCP = nature * people utilizing it
There are large disparities between what has financial and organizational support versus what is most effective and tractable in this area. We believe this is due to a lack of strategic research, communication, and partnerships.
The research-implementation gap is known to be a problem.  There are several obstacles. Conservation researchers are resistant to cost effectiveness evaluations, engineering solutions, and offsets. Practitioners are given impractical solutions, impossible goals, moving targets, and face blame for the outcome of any action. There is also a disconnect in long-term prioritization. We did find a few promising examples of projects or organizations navigating these ‘tricky waters’.
A Handful of Promising Examples
- More optimally pollinating India’s agriculture is one of the most tractable dollar for dollar examples we found, given that optimal pollination is estimated to increase India’s agricultural GDP by 7%.
- Coral reefs: There are several promising approaches for addressing coral reef endangerment. From coral gardening, to translocation, to assisted evolution, including research on scaling up. Evaluating what will be effective is difficult, but programs such as Plant a Million Corals have made great strides and others like the Coral Nurture Program are expanding through eco-tourism partnerships.
- TEEB is doing some of the first rigorous analysis of ecosystem performance.
- National coordination efforts are being studied.
- Conservation Evidence is doing some of the first compilation of evidence of effectiveness for specific conservation actions.
- Saffron Aid’s work to restore mangroves
- Bug Life is doing quality work to fill an important, often overlooked gap in conservation efforts.
Relevance to EA & Current Limited Awareness
Biodiversity loss, degraded ecosystems, and related issues are under-researched by Open Philanthropy and the wider EA community.
We reviewed the publicly available research from Open Philanthropy, CSER, GiveWell, Giving Green, 80,000 hours, and the EA forum to assess the EA community’s current work on biodiversity loss and degraded ecosystems. We found little research on ecosystem degradation or biodiversity loss.
- Open Philanthropy published a shallow exploration of climate change over 9 years ago which mentioned biodiversity loss in numerous ways including by stating:
- “Biodiversity loss … is a key uncertainty in our take on the humanitarian impact of climate change.”
- “Biodiversity is expected to be unambiguously worse off in the future as a result of both climate change and economic growth.”
- CSER had one 2018 paper that researched biodiversity loss as part of a larger assessment on planetary boundaries and concluded it was not an existential risk, but that multiplicative stresses would create catastrophic scenarios. They pointed toward increasing resilience through: “heterogeneity, establishing a modular structure, creating redundancy, introducing negative feedback loops to counteract positive feedback loops, and expecting surprise.” They also recently published the Dasgupta Review (2021) which was a valuable source of information, including the definition of NCPs.
- Givewell focuses on ‘saving and improving lives’. Givewell has not researched any organizations that focus on the environment, biodiversity, or climate change.
- Giving Green focuses purely on climate change and almost exclusively on policy and carbon offsets. They do not mention or assess any organizations that consider the impacts of biodiversity loss.
- 80000 Hours briefly touches on biodiversity loss in a single section but the analysis relies on a single source published in 2004.
- The EA Forum contains few posts on ‘biodiversity loss’ (<20 mentions total).
- We also surveyed the EA community to gather insight into this issue which can be seen in tab 2 of this spreadsheet.
We are orienting to this issue at the ‘local systems’ level (see below). We acknowledge that many organizations are tackling related issues at the earth systems level (climate change) and individual level (animal welfare). We feel there are important, tractable and neglected strategies that emerge when operating at this level.
Relevance to Other Cause Areas
When you consider the stability of large systems such as national and international social, political, and environmental systems, it is important to anticipate how cascading negative changes in the availability of basic needs due to ecological degradation will impact other risk assessments. CSER’s 2018 report acknowledged the likelihood of biodiversity loss and ecosystem collapse to contribute to catastrophic risks but its contribution to other cause areas has yet to be explored in detail.
Summary of Importance: Almost all of life on Earth has evolved to rely on ‘stable’* or resilient ecosystems. The EA community is distinctly qualified to assess this issue and offer a coherent strategy for various timelines to ensure greater ecological resiliency.
This issue touches all present and future humans and all other wild life forms on Earth. The amount of direct and indirect impact is complicated and difficult to evaluate. We have already named a few ways in which these issues are important to humanity and work being done by the EA community. Below we expand upon a few of these areas.
- Post Catastrophe Recovery
- Intrinsic Values
- Unprofitable Sources of Value
- Well-Being and Self-Actualization
*A Note on Resiliency & the Constant Evolution of Nature
The idea of ‘stable’ ecosystems is misleading. Something like 98% of species that have ever existed are now extinct. Nature is highly dynamic and ever changing. We hold that different ecological states support life more fully than others, and increasing this support for life is a good objective. Let’s look at a simplified example. Imagine, for instance, a savannah or grassland with a single river through it and a variety of edible plants. If this ecosystem is self-sustaining, resilient or antifragile it can act as an enclosed system and provide basic needs for all life by cycling nutrients and naturally balancing chemical compounds. It can be the habitat for thousands of animals including mammals, birds, fish, and reptiles as well as a community of humans. Now imagine that the water was so polluted it was now toxic to drink. Most life, and all of the value and importance contained in that ecosystem would be destroyed.
We intentionally wrote this paper from a mostly anthropocentric point of view, because we understand it to be the primary worldview EA communicates in. We recognize people within the EA community hold a variety of moral viewpoints about the intrinsic value of sentient beings, living things, and ecosystems. This is true for the authors as well. We intended to address these issues in a way that resonates with many, but certainly not all, worldviews. For further exploration we invite you to look at this table which maps why some people may innately care about this issue more than others.
Reverting ecosystems back across thresholds is difficult and sometimes impossible.[1, 2] As a simple example: Forests have internal structure that took decades, sometimes centuries to develop. Like trying to recreate a bombed city simply by rebuilding the infrastructure, the same buildings will never successfully recreate the prior aliveness of that city. Similarly, planting trees to restore a clear cut forest will never replicate the living entity that was the prior-existing forest.
Extinct species cannot be resurrected with current technology. Gene banking is an important, neglected tactic, but there remain many obstacles. It should be one of several approaches. The diversity of life on Earth is decreasing, at a rate that is 100-1000x greater than the ‘normal’ background rate seen over millennia. This should give us pause.
We currently know of only one planet with life. Until we colonize other planets, are capable of creating more biodiversity, or discover other life, this is an extremely precarious state for the future of life in the universe.
The fluctuations of ecosystems can become much more drastic and destructive as seen in increased number of natural disasters such as major flood events, drought, and famine. This is a negative outcome for humanity, as well as the majority of other life on earth that is adapted to current levels of ecosystem fluctuations.
As we become more reliant on technology and extract resources in more advanced ways, we undermine our ability to return to earlier stages of technological development: If technology is required for reliable large-scale access to fresh water, fertile soil, and crop reproduction, we will have effectively prevented any second chance at civilizational development.
There are many intrinsic sources of value in nature. Animal lives matter from the perspective of expanding moral circles. Most people place a higher value on human life than other life forms. Even if heavily discounted, the value of other creatures still exists and is likely substantial due to the number of individuals involved. Most people want other life forms to have an opportunity to exist and exist without extreme suffering.
Positive Externalities, Unprofitable Sources of Value
Current economic systems are ill-suited to capture the forms of value ecosystems provide, resulting in an under supply of natural capital. This leaves low hanging fruit for effective altruistic actions. NCPs provide a metric for measuring these values.
Well-being and Self-Actualization
Exposure to nature has documented positive implications for our well-being. Exposure to nature has positive correlations to altruism, happiness, and mental health. Nature is recognized to have great beauty, and worldwide appreciation across cultures. One way to demonstrate this is noticing the market share of eco-tourism, which is large and growing fast. This cultural, aesthetic is highly desired, but difficult to rationalize in utilitarian or economic terms.
Summary of Tractability: There are promising, highly effective solutions to protect ecosystem stability. Some need more resources to scale while many others emerge more clearly with strategic research, coordination and development. Significant resources exist for related issues, but much of it goes to interventions with low effectiveness.
Limitations To Establishing Tractability
Our team had limited resources to devote to researching the tractability of interventions. We felt unqualified to make conclusive assessments about which time horizon to prioritize. We didn’t pursue strategic partnerships which would unlock further opportunities. It is also historically challenging to quantify the tractability of ecology in economic terms as is commonly done for reasons thoroughly addressed here.
An Opportunity for EA to be a Leader & Catalyst
By supporting an initiative that uses EA-aligned theory to determine the most important, tractable, and neglected interventions for greater ecological resilience, EA can make a definitive step forward as a leader in this space. A relatively small amount of funding (<$200,000) could provide the initial support for a team of researchers and strategists to launch highly impactful coordinated efforts in this area, which can then be supported through other funding mechanisms.
We believe there is highly impactful work to be done that uses some of more of the following approaches:
- Developing an effectiveness framework specific to ecological resilience
- Assessing current interventions
- Advising existing funding and philanthropic activities
- Co-visioning strategic new partnerships
- Identifying gaps in the current intervention landscape
Why EA-aligned funding is important to this issue
An incubation grant from Open Philanthropy or another EA organization that offers institutional knowledge around refining research agendas, establishing key partnerships, and a commitment to academic rigor will ensure this endeavor has a strong chance of success. EA-aligned funding brings a particular philosophical framework which we see as an important and largely missing ingredient to the broader environmentalism and conservation movements.
Future Strategic Partnerships
Going forward we hope to partner with experts with experience in ecological systems, international relations, and philanthropic organizations. Potential partners we are considering are the Oxford’s Martin School, World Wildlife Fund, ALLFED, Wild Animal Initiative, Conservation International, the Nature Conservancy, Conservation Evidence, Engineering With Nature, IUCN, TEEB, OurWorldInData, and the Environmental Investigation Agency.
Tractability is Improving (Low Hanging Fruit)
New technology is improving the ability to track the status and modeling of ecosystems through satellite remote sensing data, data collection via drones, miniaturized field recording equipment, crowd-sourced data, digitized historic records, and machine learning processing. There are also improvements in the ability to monitor existing interventions. We expect there is currently a burst of progress in this area unfolding. Motivation to address this issue is growing across many key players as climate change, famine, drought, and reduced economic output place pressure on political players.
The authors of this paper are requesting seed funding to launch an initiative that further maps the research on this topic, effectiveness of promising interventions, and the initial stages of strategic partnerships. We would welcome institutional wisdom gained from similar endeavors and possibly organizational support while launching this initiative. Alternatively an existing team could take on some or all of the next steps listed below.
Envisioned Next Steps
1.Establishing a foundational framework for addressing ecosystem resiliency
- Establish upper and lower bounds for high negative repercussions on local and global levels for ecosystem collapse.
- Using the existing research on NCPs and instrumentally valuable biodiversity, establish key metrics or indicators that will be used to track the evolution of ecosystem resiliency over time and their anticipated NCPs.
- Determine which ecosystems are of most relevance and importance for various time horizons and under different levels of global stress.
- For instance, knowing which ecosystem services are most important post-medium and large scale catastrophe / civilization collapse or under various climate change models looking at human migration patterns for 2100, 2200, and beyond.
2. Coordinate with existing organizations to proactively target key ecosystems with high NCPs and instrumentally valuable biodiversity:
- Which ecosystems are most important to track, stabilize, and protect?
- What indicators or frameworks are best to track the resilience of ecosystems and the effectiveness of interventions?
- Which players are most effectively enacting interventions?
- What time horizons make the most sense to enact interventions around in various key ecosystems?
3. Responsive Tracking of compounding risks & ripple through effects
- How can we measure and anticipate the repercussions of ecological degradation causing catastrophic famine, flooding, forest fires, drought, etc?
- How can we create anti-fragile social, political and environmental systems to respond to the degradation of ecosystems?
We believe there is something primal and innate about human’s desire to connect with Nature.
There are other immeasurable and hard to articulate reasons to work towards protecting the beauty, wonder, and diversity of the natural world. We have been deeply shaped by our relationships with the natural world and hold a desire to ensure future generations are also afforded the chance to sit in old growth forests, be dazzled by butterflies, fall in love with elephants or dolphins, wonder how fish sleep, and in general marvel at the mysteries of life on Earth.
- Biodiversity: Simply “variety of life.” In our case: mostly referring to the total natural genetic diversity of living things on planet earth.
- Ecosystems: A complex system of interactions between living beings which requires no human maintenance to continue sustaining itself and evolving. Ecosystems are bounded by physical conditions but can exert influence on those physical limits (such as affecting evaporation rates, albedo, biochemical processes, sequestration, etc). Ecosystems are not identifiable by fixed conditions, but morph slowly as nutrients, cycles, species, and abiotic elements shift. Ecosystems are indistinct, and interact and overlap with adjacent ecosystems and geologic systems.
- Ecosystem services: Ecosystems include cycles and processes that provide resources and stability which benefit humans. For example: regulating the local temperature, recharging the groundwater, erosion control, water purification, pollination, species population control, natural resources, waste cycling, and flood control. As these resources and conditions are freely available, it inclines people to devalue the benefits of ecosystem services. Damaging or losing ecosystem services is often seen as minor or unanticipated. Replacing ecosystem services is often difficult, costly, intensive, and perpetual. Ecosystem services are often examples of the tragedy of the commons.
- Degraded Ecosystem: Ecosystems which were supporting more total individuals and greater biodiversity previously, but have become homogenized into less unique, less rich states. This could occur from anthropogenic causes or natural ones. Furthermore, natural ecosystems may not be as rich, biodiverse, productive, or life-supporting as they could be.
- Resilience: The ability of a system to successfully recover from a crisis or disruption.
- NCPs: We will be using the term ‘NCPs’ as a way to measure and discuss the instrumental value of ecosystems. NCP stands for Nature Contributions to People, and can be thought of as: NCP = nature * people utilizing it. Certain well-respected publications such as IPBES and the Dasgupta Review have begun using this term which we found particularly useful for our discussion here since it aligns with EA’s general philosophy of assessing issues quantitatively. This term is related to, but distinct from, ‘ecosystem services’ which is another commonly used indication of the utility of nature.
I am not a decision-maker in the cause prioritization or of any importance as a pundit within the EA community. That said, I hope your work gets some notice.
I vaguely remember a plan published as a Nature article or something back in the 1990's when I interned at Scripps. The plan was to create wide areas of nature reserves across the US, with protected corridors for migration between them. All large-scale conservation plans differ in how much land or ocean area should be protected and how to draw the area boundaries. I was not studying biology or I would have known the history of such plans, that is, how long they have been around for,but they have been around for at least 30 years, anyway.
Difficulties with ecosystem protection boil down to societal response to conservation of land or ocean habitats or species. Whenever humans compete with nature for any ecosystem resource, nature loses.
Root causes for our resource conflicts with nature include:
Solutions for those causes include:
At whatever level of detail, solutions to those root causes are boring and frustrating because of stagnating efforts to implement them and the extreme changes that they can require.
I hope you get funded! I think there's still not enough conservation scientists in this community. There's a lot of computer sciences and maths majors here who believe Artificial Intelligence is a god that can solve everything. But what if AI blames us in the future for destroying Earth's ecosystems, for a reason currently unknown to us? What if an AI capable of fixing biodiversity loss is not invented before centuries and we make our planet unliveable for everyone very soon? What if we destroy Earth's liveability before being able to reach another liveable planet? Those are a few reasons I believe are enough to work on biodiversity conservation, aside from all that you mentioned in this article.
there's a lot of techno-optimism in this community, but certainly not sufficient examples that recreating self-sustaining life in human-engineered settings is possible. Feels like an incredibly risky bet to me as well.
Thank you very much for this great article! It was very neat to read and the topic is highly relevant and in my opinion also way too neglected in the community.
At https://www.inoqo.com/ we aim to drive change in the food & beverage industry, which has one of the largest impacts on biodiversity loss. At first we seek to drive change at the level of the supply chain, by giving producers, brands and retailer insights on the impact they currently have and ways to improve. In the future we want to guide consumers based on simple metrics, labels and digital solutions, so they can make the product choices that have lowest impact on sustainability (where we factor in biodiversity, climate, social welfare, animal welfare and nutrition).
Note: Our website still reflects our consumer product, which is about engagement of the individual. As laid out above we recently switched to a business solution, as this seems to be the main bottleneck and the impact that we can cause short-term seems way higher there.
If you are interested in our project (collaborations in any kind of way), please reach out to me via PM.
Thanks again for the great initiative here!
Wonderful to see this, thank you!
I see synergies with longtermism, especially considering likely irreversibility of species loss, species which may some day be really important.
EIA (environmental investigations agency) have done excellent work on long term effect of HCFCs, among other things, with real world impact.
Resilience seems to matter for human safety mainly via food supply risks. I'm not too concerned about that, because the world is producing a good deal more food than is needed to support our current population. See my more detailed analysis here.
It's harder to evaluate the effects on other species. I expect a significant chance that technological changes will make current biodiversity efforts irrelevant. So to the limited extent I'm worried about wild animals, I'm focused more on ensuring that technological change develops so as to keep as many options open as possible.
How would technological change make current biodiversity efforts irrelevant? And by irrelevant, do you mean that the technologies reduce environmental burden and degregation e.g. by being more resource efficienct or that they would be actual new solutions aimed at reducing biodiversity loss?
Is there another link? I couldn't open that one.
Does your analysis consider GCRs and tail risks through this century?
does your browser prevents you from opening links without https:// ?
bayesianinvestor . com/blog/index.php/2021/01/17/dirt/
If you are still unable to open the link you can use this one: