Two noted ecologists – Daniel Simberloff, of the University of Tennessee, and Donald Strong, of the University of California, Davis – offer this critique of Fred Pearce’s recent Yale Environment 360 article, “On a Remote Island, Lessons in How Ecosystems Function.”
Fred Pearce’s previous rosy nostrums on Yale Environment 360 have risen to even greater heights with his August 26 article on Ascension Island. He has mastered the “skeptical environmentalist” tactic of dismissing science in favor of a mishmash of sciency stories. His imputations that there is a lack of scientific work with revegetation and restoration of remote places are carelessly silly. Google Scholar yields within a few seconds a large literature demonstrating the contrary. This literature finds substantial roles for the natural phenomena that Pearce belittles or ignores – long-term natural processes, seed dispersal by animals, species interactions, coevolution, and biotic-abiotic interactions, inter alia – in the reconstitution of biodiversity.
Consider the revegetation of Krakatoa as just one example from this rich literature. Following the volcanic explosion that cleared the island of life in 1883, both native and non-native species dispersed there, mainly from adjacent mainland areas. More than a century later, between 124 and 173 woody plant species have colonized, many probably introduced as seeds carried by native birds and bats from the mainland. Forests are closing to replace savannas that developed earlier after the explosion. And without the aid of the Royal Navy that Pearce cites on Ascension, no less!
Such studies as these, rather than Pearce’s superficial just-so stories, are what inform the science of restoration ecology. His articles are riddled with undefined terms, unsupported statements, vague attribution, and straw men. A close examination of these problems undercuts his optimism, as well as his claim that Green Mountain on Ascension has major practical implications for conservation in general.
Pearce’s argument about Ascension Island in a nutshell is: (1) A lush, pretty forest consisting of trees introduced within the last 200 years occupies Green Mountain today. (2) Some native species appear at least to coexist happily with the exotic vegetation. (3) According to “mainstream ecological theory,” this forest “really should not exist.” (4) This forest suggests that “even highly biodiverse ecosystems may be accidental, temporary, and versatile.” (5) Green Mountain “suggests that restoration could be much easier than many believe. Ecologists may not have to painstakingly reassemble the complex ecosystems that have been lost. They often may be able to let nature take its course.”
We should unpack these components to see how compelling his argument is.
1) Forest doesn’t “cover” Green Mountain, as Pearce states. Patches of grass and dry-adapted shrubs characterize the area below 330 meters. From 330 to 660 meters, there is thicker coverage, consisting of grasses, prickly pear, acacia, juniper, and Australia pine. Above 660 meters it is dominated by denser forest vegetation with some grassy slopes. Throughout the mountain, the plants are very largely introduced species.
2) Some native species do appear to get along with the non-natives. Pearce cites a fern that may grow only on mosses on the branches of exotic trees and native land crabs that shelter in the shade of the forest. This is one of several claims that sometimes an introduced species serves some conservation function. There is nothing surprising or new about this, and many similar examples are known. Any non-native species that becomes common will likely constitute a resource for some other species, even if the net impact on the ecosystem is highly detrimental.
When examined closely, many such cases, including the two Pearce cites, seem reasonable, but an absence of data makes even the limited claim speculative. Where did the fern exist before the exotic trees were present? Do data show whether the crab population is greater or smaller than before the forest existed? What negative impacts on native species have been wrought by the invaders? Pearce concedes, for instance, that three of the ten endemic plant species are now extinct. He could have added that two others are critically endangered, the highest threat level listed by the International Union for the Conservation of Nature, while a third, as he states, is represented by just one individual on the island while being propagated in the U.K. Meanwhile, various combinations of 13 non-native plants severely threaten the existence of every one of the extant endemic plant species. It is extremely clear that native seabirds do not get along with the rats and feral cats that have been introduced to Ascension Island.
The plants are just the best understood part of the story. Insects, much less well studied than plants, have likely disappeared. Although ferns typically suffer less insect attack than angiosperms, many insects do attack ferns, and when plants disappear, they can carry insects dependent on them to extinction as well, though we often lack evidence because the insects were not studied beforehand. When chestnut populations plummeted to near extinction as chestnut blight spread in North America in the 20th century, at least seven moth species were extinguished, but we know this only because Paul Opler had studied them carefully. No analogously intensive research occurred on Ascension. Data are so scarce that Simon Fowler, an entomologist seeking a biological control for invasive mesquite (Prosopis) on the island, discovered seven previously unrecorded insect species on this plant in just one month.
Ascension, a philatelic powerhouse, issued a series of 12 insect stamps in the late 1980s, entirely featuring widespread non-native species. Ascension currently has 53 native insect species and 80 introduced ones, plus 45 species of undetermined origin. Everything we know about the trajectory of remote, species-poor islands with similar proportions of non-native plants and insects (e.g., Kerguelen, St. Helena) suggests that, even if some native species tolerate or even benefit from the one or more invaders, the native biota as a whole is devastated.
3) We cannot imagine, nor does Pearce reveal, what ecological theory says that this forest should not exist. If there were such a theory, science would put it out of its misery quickly. In many parts of the world mixtures of non-native plant species (often from different continents) dominate the landscape. Brazilian pepper (Schinus terebinthifolius) trees with a ground cover of shoebutton ardisia (Ardisia elliptica) from Asia dominate vast tracts of south Florida. Non-native species that have never seen one another even form mutualisms. For instance, in 1958 C.S. Elton noted that in California citrus groves the Asian red scale insect (Aonidiella aurantii) is tended by the Argentine ant (Linepithema humile), which in turn reaps the “honeydew” secreted by the scale.
There has been no substantial ecological research on the Green Mountain forest – there are no data on biogeochemical cycles, the soil biota and its interaction with trees, or on population dynamics or trends of any of the species. Thus, although we can imagine theories that might be contradicted by certain sorts of empirical data on this forest, no data exist to test those theories.
4) We are unsure what Pearce intends by a “versatile” forest. By “accidental,” he appears to mean constructed of species that arrived at different times by different means, and often with human assistance. It has long been known that complex, species-rich communities are sometimes composed of elements that arrived at different times and from different areas. R.H. MacArthur and E.O.Wilson, for example, describe several rich, “radically mixed biotas” in which invertebrates and plants come from one region and vertebrates from another region. Biological invasions have produced many recently assembled species-rich communities comprising elements from different regions. For instance, the opening of the Suez Canal in 1869 allowed numerous Red Sea species (“Lessepsian migrant”) to pass into the Mediterranean Sea, including 57 shallow water fishes, which have joined 324 native shallow water species with no known extinctions of the latter. The mixed scrub community of Assumption Islandin the Aldabra group contains 28 species, almost all introduced. The size of that community is very similar to that of the native mixed scrub community of undisturbed but otherwise similar Aldabra Atoll, consisting of 31 species, including many endemics.
As for whether the Green Mountain forest is “temporary,” this is another of Pearce’s undefined words. How long would the forest have to persist with approximately the same species composition before it would be defined as not temporary? How similar would the composition of the forest have to be for the forest at two different times to be classified as the same forest? After all, all ecosystems are dynamic, and, given enough time, some species disappear, and others arrive or evolve in situ. Would the composition have to be identical for the forest not to be considered changed? And what theory or even vague notion suggests that there would be anything surprising about a large group of species recently thrown together losing or gaining one or several species?
5) Pearce does not understand the goal of ecological restoration. Restoration ecologists today do not “painstakingly reassemble the complex ecosystems that have been lost.” The static balance-of-nature idea has long been succeeded by the understanding that nature is dynamic. Restoration ecologists do not aim to recreate the past, but rather to reestablish the historical trajectory of an ecosystem before it was deflected by human activity, to allow the restored system to continue to respond to various environmental changes. They in fact do aim to achieve a situation that will let “nature take its course.” The arrival by ship of the current biota of Ascension Island is hardly a story of nature taking its course.
Green Mountain certainly does not suggest, as Pearce argues, that restoration could be much easier than many believe (again, who are the “many”?). The forest on Green Mountain is in no sense “restored.” It has undergone revegetation and a measure of what restoration ecologists term “reclamation,” but the forest that is currently present is unlike any forest that existed there previously and whatever trajectory it is on is certainly not historical. It may provide lessons in greening deserts and other barren locations, as D.R. Catling and C.A. Stroud suggest and Pearce reiterates. But greening deserts has nothing to do with restoration. Greening deserts is about constructing new ecosystems to provide services for humans.
Green Mountain seems to fit the definition of a “novel ecosystem.” This construct, which has attracted much attention since it was proposed in 2006, has been variously defined, but perhaps the latest, most authoritative definition is “a physical system of abiotic and biotic components (and their interactions) that, by virtue of human influence, differs from those that prevailed historically, having a tendency to self-organize and retain its novelty without future human involvement.” Pearce is enamored of this concept, which he discussed in another recent Yale Environment 360 article, “True Nature: Revising Ideas On What is Pristine and Wild.” He sees the existence of such ecosystems as “good news,” because it shows that nature is “resilient and adaptable, able to bounce back from the worst we can throw at it.” In what sense has nature “bounced back”? Following another enthusiast for novel ecosystems, writer Emma Marris, Pearce views novel ecosystems as the “new normal,” and he suggests that, in light of this status and the major contribution of introduced species to novel ecosystems, we might well want to aid species introductions rather than trying to prevent them.
So far, the various organizations (IUCN, European Union, the Convention on Biological Diversity) and many nations (e.g., United States, South Africa, Australia, New Zealand, Canada) that have invested heavily in preventing invasions seem not to be swayed by Pearce and Marris, and many advances have occurred in stopping invaders, eradicating them, or minimizing their impacts.26 Whether the many large-scale traditional restoration activities underway today worldwide will come to be seen as futile and old-fashioned, and the new goal will no longer be restoration but construction of ecosystems that provide some services for humans, remains to be seen. But Pearce should understand what ecological restoration is and note the great progress in the field.
Finally, with respect to the “ecological fitting” that Pearce sees in both his articles as somehow an alternative to coevolution, it seems not to have crossed his mind that the two phenomena are not mutually exclusive. Daniel Janzen, observing geographic ranges of some species at a site in Costa Rica, simply noted that some of them thrived together, despite the fact that they had probably not coevolved. He observed that the same is true for some introduced species in their new communities (e.g., kudzu in North America). This was not a new observation. Elton, for example, pointed to numerous examples of introduced species that survived very well in their new homes. Janzen did not say that coevolution does not occur, or that it is not important. He did not propose the relative importance of a “rival model for how ecosystems work,” as Pearce claims. He simply pointed to the obvious, that some species do well in the midst of communities with which they did not coevolve. The brilliant aspect of the paper was rather a series of 10 questions related to this phenomenon that Janzen asked at the end of his paper. Three are very pertinent to our discussion, and they have never been fully answered:
1) What properties would an ecologically fitted community have that differentiate it from a highly coevolved community? Are there any such properties?
2) Can one distinguish a coevolved mutualism between two species from a mutualism that arose instantaneously when they were thrown together because they happened to be preadapted to one another (e.g., if agoutis were introduced from the New World to Africa, would the mutualisms they would form there differ from those that had evolved in their native range?).
3) “What will be the consequences of arbitrarily chopping out a small block of terrain occupied almost entirely by widespread species and overnight converting them to local species (the biological history of almost all the national parks and reserves around the world)?”
The first and third questions are of most concern to people seriously interested in conserving biodiversity rather than simply providing ecosystem services to humanity. The first is the one most relevant to the question of ecological fitting and the nature of and prospects for novel ecosystems, and is the one furthest from an answer. It cannot be answered yet because the properties of so few ecosystems, novel and otherwise, have been comprehensively measured. Those of the Green Mountain forest certainly have not been. Answers to this question would indeed likely shed light on “how ecosystems work.” The answer to the second question is probably that at least some newly formed mutualisms would not differ from coevolved ones (e.g., the red scale and the Argentine ant noted by Elton), but we do not yet know what fraction differ, in what directions, and by how much. A partial answer to the third question is that, despite Pearce’s cheerful view of the prospects for conservation, at least in many cases, the original native community degrades as various species are lost. If one is happy with a haphazard collection of replacements from elsewhere, as Pearce seems to be, this fact may not be as depressing as it seems to us.