Emergent and vanishing biodiversity, and evolutionary suicide

Simon A. Levin
James S. McDonnell Distinguished University Professor in Ecology and Evolutionary Biology, Princeton University
Director of the Center for BioComplexity in the Princeton Environmental Institute

1. Introduction:  The importance of biodiversity to humanity

Why should we care about biodiversity? The easy answer is that we should respect all species, the diversity within them, and the intricate relationships among them, as we should respect even the inanimate features of our planet. We are just one of millions of species that evolution has created, and we must share Earth with those other inhabitants.

But there is a more selfish reason why we should care about biodiversity: We could not survive without it. Biodiversity provides us with food and fiber, with fuel and pharmaceuticals. It helps sequester toxic materials, mediates climate and cycles nutrients crucial to life. It is the source of a vast array of goods and services that Nature provides us (so-called “ecosystem services”), and that are essential to a sustainable future for humanity. The nature of these services has many dimensions, including aesthetic and ethical aspects.

Much of the focus of biodiversity protection is on individual species, and especially charismatic ones like tigers and elephants. This is all well and good, but it is important to recognize that much of our dependence on natural systems relies not only on what species and the genetic diversity within them represent, but on how they function together to deliver those critical ecosystem services on which all of humanity depends. There are reasons to argue that all species should be preserved if possible; in the words of Paul Ehrlich, intelligent tinkering requires maintaining all the parts. On the other hand, ecological systems go through natural development cycles; and, over evolutionary time, species (including potentially our own specie) disappear and are replaced by others. We too might go this route; but we have an interest in avoiding that fate, and hence in maintaining the features that sustain us.

Are all species equally important to preserve1? Without doubt, special attention must be paid to keystones (see below) and to microbes. Why? Keystones are by definition species that have effects on ecosystems structure and functioning far beyond their abundances, and their loss could cause fundamental shifts in the nature of the ecosystems; microbes govern the key processes of the ecosystem, like decomposition, production and fixation, and thereby regulate nutrient cycling. But how do we go beyond that observation? Does the gradual erosion of biodiversity reduce the robustness and resilience of ecosystems, even when the systems appear to continue functioning? Are there early-warning indicators of shifts in which the fundamental nature of the ecosystem changes, for example transitioning from an oligotrophic lake to a eutrophic one2? I will explore these questions in the sections that follow.

2. Vanishing biodiversity

Over evolutionary time scales, species come and go; some degree of replacement is just a manifestation of natural selection at work.  But net biodiversity loss on ecological time scales is something else altogether. Indeed we are losing species at unprecedented rates due to climate change, habitat degradation, pollution, overexploitation, invasions by alien species, and other factors3. Even when species are not in immediate threat of global extinction, their decline in biomass globally is generally associated with disappearance from particular areas4 and 5.  Observations of losses of vertebrate species have led to this era being associated with a “Sixth Mass Extinction6 and 7”. Indeed, no taxonomic groups are safe from the ravages of our activities.  The loss of biodiversity at such a massive scale already is having and will in the future have drastic effects on the nature and functioning of ecosystems, and on the services we derive from them8.

3. Robustness and resilience

One of the most evocative of images in On the Origin of Species is Darwin’s view of the biosphere and its communities as tangled banks, emphasizing both the importance of processes acting at the level of individual agents, and the emergent properties of ecosystems that cycle nutrients and provide the context in which species compete for resources and exploit one another, but also cooperate through direct and indirect mutualisms that sustain the biota. In some sense then, the ecosystem is self-regulating, but without purpose and indeed without guarantee of long-term persistence. Indeed, all ecosystems are dynamic in that they go through development processes termed succession9 and 10; on multiple scales regular disturbance restarting local succession is essential for the robustness of these systems11, 12 and 13

A rural hedge on a sunny September day. The hedge contains all sorts of young trees, bushes, brambles, ferns and various creepers, including ivy and wild hops. ©Linda Steward
Image description: A rural hedge on a sunny September day. The hedge contains all sorts of young trees, bushes, brambles, ferns and various creepers, including ivy and wild hops. ©Linda Steward

In ecological communities, emergent regularities allow for classification of kinds of ecosystems, matching ecosystem types to environmental conditions, and the relationships among ecological diversity, complexity, stability and robustness14, 15, 16, 17 and 18. A particularly important concept is that of the keystone species, introduced by R.T. Paine in 196619 and 20, species whose disappearance is likely to have outsized effects on the persistence and character of the ecosystems they call home. Paine focused on the top predator, a starfish, in his intertidal communities; but there are numerous other examples21, including notably the sea otter on the west coast of the United States22. The loss of these keystone species, often at the tops of food chains, may be some of the first casualties as biodiversity is lost, leading to cascading collapses of those communities23 and 24.

More generally, much recent research has examined the relationship between the structure and functioning of ecosystems, and looked for early warning indicators of system sudden transitions25. This is a promising area of research not yet mature. By any indications, however, we have eroded the robustness of the systems on which we depend, and the potential for catastrophic shifts cannot be ignored.

4. Untangling the bank; order out of chaos

The most crucial insight from the tangled bank metaphor is that Nature is not a random collection of species, but rather associations that engage in competition, exploitation and cooperation. Most strikingly, however, a view of ecological communities as tightly coevolving sets of species that replace each other as units along environmental gradients26 has largely been replaced by an individualistic perspective27, in which individual species come and go somewhat independently along those gradients28 and 29.

Still, despite this element of apparent randomness, communities have emergent characteristics, including the cycling of nutrients that sustain functioning. Given that these features change more slowly along gradients than do the names of the species involved, this implies that multiple species can perform similar functions, with slight modification under different temperature and humidity regimes, for example along geographical or altitudinal gradients. Along such gradients, over broad regimes, change may be almost imperceptible because of the continuous replacement of species by those more fine-tuned to new environments. One sees a similar phenomenon at particular locations undergoing temporal fluctuations, and indeed it is this smooth substitutability that conveys robustness to systems30. Tilman and Downing, in a long-term experiment in Minnesota grasslands31, showed that, in response to a major drought, the most diverse communities were the most robust in terms of maintaining biomass and productivity, for the same reasons that evolution has developed mechanisms like mutation and sexual recombination to generate variation: That variation is important in providing adaptive capacity from which better candidates can be selected to exploit changing environments most efficiently. The system has built in not only variation, but a kind of functional redundancy; both are essential for maintaining robustness and resilience.

Space and time are inseparable in understanding ecological communities. Communities and ecosystems are not spatially uniform, but exhibit heterogeneity and patchiness on multiple scales32. In forests, local disturbances like treefalls, and even burns, create opportunities for appropriately named opportunistic species to colonize and thrive for limited times, only to be replaced by better competitors in a broadly predictable successional sequence, often called an r-K spectrum, where species trade off colonization ability (r) with competitive advantage (K) until the so-called climax species take over. Under high disturbance regimes, one finds only the early successional species; on the other hand, if disturbance is suppressed, only the climax species persist. This has led to the conjecture that diversity is maximized at intermediate levels of disturbance33, 34 and 35, a generalization that has broad support across a range of systems.  Humans, unfortunately, are increasing levels of disturbance in a wide variety of systems, driving those systems back towards the low-diversity early-successional phases.

Healthy, diverse ecological communities are patchworks in continual flux, mosaics of assemblages in various stages of succession36, 37 and 38. The modular nature of these systems, and of ecological networks more generally, provides the third leg in conferring robustness on systems, joining redundancy and diversity39 and 40. One often hears of a “balance of Nature” as stabilizing systems, but that is to some degree an illusion. Equilibrium and stasis in any system implies reduced variation, and reduced capacity for adaptation, hence a loss of robustness.  For healthy ecological communities, the relative robustness at broad scales is due to the absence of it at local scales, the spatial analogue of Tilman’s temporal results41.

5. Critical transitions in space and time

Interest in sudden transitions in biological systems is not new, and has led to the development of a suite of mathematical approaches42, 43 and 44. Such transformations, like the eutrophication of lakes, or the desertification of fertile land, or the defoliation of forests, may change ecosystems important to us in ways that render them less able to support the services we have become used to.

Perhaps the most familiar biological example is that of the outbreak of a disease epidemic. If a population is well-protected, say through vaccination, there are not enough susceptibles to feed an epidemic, and an introduced disease will die out after a limited number of cases. But, for example, if vaccination levels decline, as they have in some communities for measles, the number of susceptibles will increase until it possibly reaches a threshold level sufficient to trigger an epidemic. If the decline in vaccinations is local, the outbreak will not spread beyond the initial community; again, modularity has given the system robustness by walling off the cases that have arisen. This of course is what we see in public health practice repeatedly, in terms of quarantines and travel restrictions that maintain the modularity of the system; we are seeing all of these features play out in the current pandemic. Mathematical and computer models are playing a fundamental role in management, because they help us determine the time course of the epidemic without interventions, and what effects management measures like quarantining, mask wearing, social distancing and vaccination might have on reducing disease spread. Disease spread is of course just one example of the consequences of introduction of a novel agent into a new environment, and similar considerations apply whenever non-native species enter new environments. 

But what if susceptibility in neighboring areas is also high? Then the initial outbreak will trigger contagious spread, analogous to what we saw in the spread of bank failures during the financial recession of a decade ago. To address such issues, we need to expand our models to include both spatial and temporal dimensions, for example as Grenfell and others have done in considering the dynamics of measles in the UK and other outbreaks45. Similar approaches are needed as well in trying to anticipate how species will respond to climate change in terms of their range shifts, and interactions with other species.

There are a variety of approaches to describing the spatiotemporal dynamics of populations46.  Any of these must in some way combine the local dynamics, say of the disease, with a redistribution mechanism that describes how individuals or other crucial features move. One of the simplest assumptions is to assume that individuals move via random walks; when combined with local dynamics, these result in models of reaction and diffusion47, 48 and 49. However, there are multiple different modes of movement possible, and so a wide range of models have been considered in the literature50, 51, 52 and 53 to address correlated walks, gradient following, long-distance transport and other mechanisms.

A more ecological application of these ideas involves the transition from forest to grassland, or back, for example in savanna systems. In arid areas, fire may play a crucial role, maintaining systems in grassland mode until trees escape in size between fires and become large enough to withstand damage. Such dynamics give rise to models that admit multiple stable states - one dominated by grass, and one dominated by trees, with the potential for sudden flips as conditions change affecting levels of precipitation, and hence the frequency of fire54 and 55. Classical views of ecological communities56 and 57 allow the identification of classes of communities (tropical rain forest, desert, etc.) with climatic conditions; but the presence of multiple stable states introduces a level of indeterminacy, whose resolution will depend on past history as well as spatial context. Models that incorporate space and time are then essential to describe dynamics. In the face of climate change, attention must therefore turn to the movement of species distributions58, and of the ecotones between forests and savannas59.

Misty day at Shenandoah National Park, Virginia ©sreenath_k
Image description: Misty day at Shenandoah National Park, Virginia ©sreenath_k

6. Biodiversity, and adaptive capacity as public good

As already discussed, biodiversity is a broad source of services to humanity. In addition, even in the absence of immediate benefits, it provides a kind of insurance against the loss of essential services. Individual species, like the American chestnut in northeastern U.S. forests, that provide benefits may be lost, only to be replaced by others, like Chestnut oak, that fill a similar role. However, one can only go to the bank so often to get replacements. Each time a species or genotype is lost and replaced, some of the natural capital that provides the system with adaptive capacity is lost.

If ecosystems were individually and privately owned, the owners would have incentive to preserve this adaptive capacity. In general, however, the adaptive capacity is a public good (or a common pool resource), beneficial to all, but not sufficiently so to offset the individual incentives to overexploit. Fisheries and forests under open access provide striking examples of what Garrett Hardin called the “Tragedy of the Commons,” building on concepts of William Forster Lloyd60 and 61.  Hardin’s solution to the problem was “mutual coercion, mutually agreed upon,” with a large role for top-down control. It was left to Elinor Ostrom to show how such mutual agreement could arise from the bottom up62. I return to this in the last section.

As we will see in the next Section, public goods situations and conflicts between levels of organization have multiple manifestations in the non-human dimensions of ecosystems63. Bacteria produce antibiotics to provide competitive advantage in competition, creating opportunities for free-riding by cheaters; bacterial biofilms rely on the production of extracellular polymers, a costly enterprise; trees convert ambient nitrogen into useable form through fixation, again allowing for free-riding; and collective motions in a wide variety of species, from bacteria and slime molds to fish and birds, relies on the ultimate public good, information64. The maintenance of public goods is a challenge throughout the natural world.

7. Evolutionary suicide

The primary units of evolution are at levels well below those of whole ecosystems, housed in individual genomes and tight coevolutionary relationships. There is no reason to believe that the mechanisms of evolution will preserve the services that maintain a habitat for humanity, or even the general macroscopic features of the ecosystems we know. We need look no further than what we are doing to our planet, in particular in eroding biodiversity and its adaptive capacitance, to find examples. 

A dramatic illustration of what is possible has been termed “evolutionary suicide,” the potential for short-term evolutionary forces to take populations down paths that sow the seeds of their own destruction65.  Gore and his collaborators have demonstrated that this is more than a theoretical possibility, by showing in laboratory experiments that bacteria can change the pH in their environments in ways that ultimately lead to their extinction; a more familiar example perhaps is the profligate spread of tumor cells within our bodies, eventually leading to the death of the host and the population of tumor cells. Suicidal evolution can happen within populations, or in the erosion of the higher-level system features that are crucial to the larger system66. This indeed is what we have been doing to our planet by destroying biodiversity, and what we need to learn to slow, and even reverse.

8. Suicide prevention

For the survival of our world as we know it, we must find pathways to sustainability, and that means preserving the biodiversity that is at the core of the services we enjoy. UN Sustainable Development Goals 14 (Life Below Water) and 15 (Life on Land) make clear the essential nature of the challenges, and Goal 12 (Responsible Production and Consumption) hints at the solutions67.  But how do we overcome the Tragedy of the Commons, and prevent collective suicide?  Economists, ecologists and others must cooperate to find the pathways to success68. As already noted, preserving biodiversity is a public goods problem, and will require collective action as Ostrom suggested69, 70, 71 and 72. To some extent, this will mean recognizing that we are players in a coordination game, and must find the best of available options73 and 74; it will mean recognizing the importance of prosociality (caring for others), and finding ways to encourage it75 and 76; it will mean changing social norms77. How we get there will represent one of the greatest challenges facing humanity in the next decades.