This is an excerpt from a zine, “The Troubles of ‘Invasive’ Plants.” Check out the whole project here.
By Nicole Patrice Hill & Kollibri terre Sonnenblume
Calls-to-action often rely on claims of damage, past or potential. As a particular call is repeated, these claims can be inflated, conflated, jumbled, mistaken or otherwise misconstrued. The further one gets from the source, the more likely it is that the facts have been lost. It’s like that game, telephone, where a message passed around a circle of people in a whisper is all mixed up by the time it completes a circuit.
Unfortunately, scientific literature is not immune from this syndrome. A study will be cited by subsequent studies that use the preliminary—and often particular—conclusions of the original to back up new conclusions, often increasingly generalized. Statements are quoted out of context, paraphrased further out of context, and reiterated even further out of context until “many authors feel comfortable in reporting it without citing any source at all.” Eventually, such statements are simply accepted as fact, especially by non-scientists like journalists, policy makers, and non-profit grant seekers.
Therefore, investigating the original source of a frequently made claim—of any kind, be it scientific, political or historical—is important and can be an illuminating experience. After following the thread back to its source, and understanding the original findings or data for what they are, one can decide whether the current call-to-action’s prescribed course is appropriate or practical or pencils out financially.
Wittingly or not, the field of invasion biology has played the telephone game with some of its major claims. We will look at just two here: 1) The estimated financial costs of invasives and 2) the level of threat to endangered species by invasive species as compared to other threats.
First, it is frequently claimed that invasives cost the global economy $1.4 trillion annually with $137 billion of that footed by the U.S. The original source of this figure is a paper by David Pimentel, et al., in 2001, “Economic and environmental threats of alien plant, animal, and microbe invasions.” Many, if not most claims about the monetary costs of invasives, ultimately refer back to this source. Let’s take a look at some of his claims and methodologies.
First off, how is Pimentel defining invasive species? He uses several terms interchangeably: “alien species,” “invading alien species,” “introduced,” “non-native” and “non-indigenous.” Is he considering all non-native (or whatever) species “invasive”? Because that’s certainly not the consensus within invasion biology or elsewhere.
Regardless, according to a table of “alien species,” the US hosts 25,000 plants, 308 animals, 4500 arthropods and 20,000 microbes that are alien. Note that microbes are over 1/3 of the total. For people who associate “invasives” with Zebra Mussels, Kudzu and the like—as we would venture to guess most people do—that might seem a little strange.
Table 2 shows “Economic losses to introduced pests in crops, pastures, and forests in the United States, United Kingdom, Australia, South Africa, India, and Brazil” [our emphasis] and racks up a total $228.22 billion annually. Are any of these designations made up of native plants? With “crops,” decidedly “no,” as Pimentel makes clear (see below). Though the term “pasture” is “primarily used for the production of adapted, domesticated forage plants for livestock” [our emphasis] as opposed to “rangeland” which hosts “native vegetation,” Pimentel seems to be including $5 billion in the cost of controlling weeds on “rangeland” in the figure, so we will say, “yes,” this category includes native species. The final category, “forests,” are highly likely to be populated with native tree species, though “managed” ones might be replanted as monocrops of such. Even so, the total losses to pastures and forests put together add up to merely 5%.
So let’s be clear what that 95% ($216.50 billion) is measuring: the cost of damage by one set of introduced species (“weeds”) to another set of introduced species (agricultural crops), which are themselves a major—if not the leading—cause in the decline of native species. You want to talk about habitat loss? How about the one fifth of the lower 48 US states devoted to field crops? Or the 40% (overlapping with the former) used for animal agriculture? But to Pimentel (and every author who cites these numbers), such grievous truths are irrelevant.
Pimentel is not ignorant of the “alien” nature of agricultural species, either. He states: “Nearly all crop and livestock species are non-indigenous. These alien crops (e.g. corn and wheat) and livestock (e.g. cattle and poultry) are vital to maintaining world agriculture and the food system. However, these benefits do not diminish the enormous negative impacts of other non-indigenous species on agricultural and other managed and natural ecosystems.”
Let’s not get romantic about the “world agriculture and the food system” either. Pimentel is referring to an industrial sector first and foremost, not to Old MacDonald growing juicy tomatoes and raising happy pigs on a family farm or even to Jose and Maria toiling in the citrus orchards, lettuce fields and vineyards, which is more likely. No, he means Cargill, Conagra, Archer Daniels Midland, Tyson and other corporate giants, who don’t give a damn about our native flowers, birds or butterflies.
Moving on to table 3, “Environmental losses to introduced pests,” Pimentel ascribes damages by the following agents: Mammals (Rats & “Other”), Birds, Reptiles & Amphibians, Fishes, Arthropods, Mollusks, Livestock Diseases and Human Diseases. The total damages for the US are $58.299 billion. Since this table is titled, “environmental losses,” you might assume we are now looking exclusively at harms to native species rather than crops (to which the previous table was dedicated, after all). But you would be wrong.
The single largest item is “Rats” at $19 billion. Is this Rats gobbling up native bird eggs, devouring native flower seeds or decimating native insect populations? Nope. It’s rats “on farms, in industries, and in homes.”
How does arrive at $19 billion? First, he calculates the total rat population of the US by adding two numbers: the ratio of rats to chickens—1 to 5—and of rats to people (“in homes and related areas [whatever those are]”)–1 to 1. Then:
“If we assume, conservatively, that each adult rat consumes and/or destroys stored grains (Chopra 1992; Ahmed et al. 1995) and other materials valued at $15/yr, then the total cost of destruction by introduced rats in the United States is more than $19 billion per year. In addition, rats cause fires by gnawing electric wires, pollute foodstuffs, and act as vectors of several diseases, including salmonellosis and leptospirosis, and, to a lesser degree, plague and murine typhus (Richards 1989). They also prey on some native invertebrate and vertebrate species like birds and bird eggs (Amarasekare 1993) [our emphasis].
The damage to native species is apparently too small to bother estimating, as he offers no number for it. Is this paper about ecological damage or not?
The next highest number in table 2 is $17 billion, which Pimentel attributes to the domestic cat, both homed and feral. He estimates that cats kill an astonishing 465 million birds in the United States annually. He cites two papers in extrapolating this number. The first estimates the number of birds killed by feral cats in the states of Wisconsin and Virginia.[1] The second, “The Population Origins and Expansion of Feral Cats in Australia,” which we reviewed, is focused on the genetics of the domestic cat on that continent and mentions diet only peripherally. The figure $17 billion is based on a per bird cost of $30. Explains Pimentel: “This cost per bird is based on the literature that reports that a bird watcher spends $0.40 per bird observed, a hunter spends $216 per bird shot, and specialists spend $800 per bird reared for release.”
The third highest cause of environmental loss in Pimentel’s report is disease, of both livestock and humans, with $15.50 billion in damages.
Introduced plant species come in a distant ninth in his ranking, with $148 million in damages. This is made up of “US$ 45 million per year in purple loosestrife control plus US$ 100 million per year in aquatic weed control.” Which leaves a whopping $3 million per year in “environmental losses” due to all the other dreaded invasive plant villains like Kudzu, Tamarisk, Garlic Mustard, etc., which are regularly presented as the banes of civilization. Has anyone making the case agains invasive plants has ever cited this particular statistic?
If this paper was a boat, you wouldn’t dare get into it, it’s so full of leaks. Yet, it remains a popular “go to” for invasion biology papers.
The level of threat to endangered species by “invasive species” as compared to other threats
The second major claim we will address is that invasives are the “second greatest threat to endangered species worldwide.” This claim was originally made by Wilcove, et al. in their 1998 paper, “Quantifying Threats to Imperiled Species in the United States.” Over the following decade this claim was cited over 700 times, and quickly became accepted as fact, though “article of faith” is a more accurate characterization, as we will show.
In their defense, the authors of the original paper were upfront about limitations to their data set. To wit:
“We emphasize at the outset that there are some important limitations to the data we used. The attribution of a specific threat to a species is usually based on the judgment of an expert source, such as a USFWS employee who prepares a listing notice or a state Fish and Game employee who monitors endangered species in a given region. Their evaluation of the threats facing that species may not be based on experimental evidence or even on quantitative data. Indeed, such data often do not exist. With respect to species listed under the ESA, Easter-Pilcher (1996) has shown that many listing notices lack important biological information, including data on past and possible future impacts of habitat destruction, pesticides, and alien species. Depending on the species in question, the absence of information may reflect a lack of data, an oversight, or a determination by USFWS that a particular threat is not harming the species. The extent to which such limitations on the data influence our results is unknown” [our emphasis].
And later:
Again, we note some caveats with respect to the data in this phase of the analysis. Species added to the endangered list prior to 1980 (238 species) tended to have fewer threats delineated in the listing notices than species listed in later years. Although there may be a biological basis for this difference, we strongly suspect that it reflects the less controversial nature of endangered species protection at that time. Before 1980, USFWS probably was under less pressure to produce detailed justifications for its listing decisions. We do not know how this pattern may have influenced our results. Also, as noted in our coarse-scale analysis, assessments of the threats to individual species are often based on the subjective opinions of knowledgeable individuals, rather than experimental evidence or quantitative data [our emphasis].
In short, don’t bet the farm on these numbers. But those who repeated the dire “second greatest threat” ranking either ignored these qualifications or (more likely) didn’t read the paper past the abstract or (even more likely in the case of journalists and other non-scientists) at all.
But that’s not the biggest issue with this paper’s most publicized claim. The main factor that should have limited the use of this citation as a global reference was how the numbers were skewed upwards by the inclusion of Hawai’i with the continental US. The ecology of islands is “vastly different” from the ecology of continents, especially with respect to introduced species, and, as would be expected, the number of species under threat from invasives was much higher in Hawai’i than on the mainland.
Again, the authors were forthcoming about these discrepancies. They devoted one table and two figures to laying out and illustrating them. We summarize below:
(Any given plant species can face multiple threats, which is why the numbers in a column add up to more than 100.)
Note that the percentage of native plant species in the continental US threatened by alien species falls from 57% to 30% when data from Hawai’i is removed. That’s a difference of nearly 50%, which is surely “statistically significant,” as they say.
But this isn’t the end of the story. The authors broke down “habitat degradation and loss” into eleven categories and reanalyzed a subset of their aggregate data (species added to the endangered species list since January 1, 1996, which comprises 723 rather than 1055 plants). Unfortunately, they did not provide the new “alien species” percentage for this subset, or break out Hawai’i, but four of the categories exceeded 30% for plants: development 36%, agriculture 33%, grazing 33%, and outdoor recreation 33%.
In other words, it is possible that for native plant species in the continental US, alien species are the fifth greatest threat, not the second, according to their (admittedly limited) data.
In a paper criticizing invasion biology that reviewed this paper, noted by ecologist Mark A. Davis wrote: “when the paper was written there was no evidence that a single native North American plant species had been driven to extinction, or even extirpated within a single US state, by competition from an introduced plant species.”
As a side note, when the Wilcove paper was published in 1998, the issue of climate change as a threat to native species was still emerging. Showing some respectable prescience, the authors added an entire section covering the issue as discussed by other researchers. They stated: “Although climate change was not listed as a current threat to any species in our databases, it is almost certain to become one in the foreseeable future due to increasing concentrations of greenhouse gases from fossil-fuel use, land-use changes, and agriculture.”
Given the increasing threats of climate change, and the shortcomings in data and aggregation detailed above, we respectfully submit that the Wilcove paper should no longer be cited without qualification in reference to the topic of invasive species.
In his book, “The New Wild: Why Invasive Species Will be Nature’s Salvation,” ecologist Fred Pearce followed the trail of other bold claims that demonize non-native species. On fact-checking frequently repeated claims, he stated that “there is a threadbare laxness in the use of statistics by many invasion biologists. Almost wherever I pursued a key claim, the trail fizzled out in obfuscation, false citations, unverified judgment calls, and absurd leaps from the specific to the general and the local to the global.”[2]
We suspect that most proponents of invasion biology would prefer a higher burden of proof for themselves before, say, being imprisoned or put on death row. (Unfortunately, many people in the US do suffer this fate, more often than not due to discrimination based on their perceived otherness in the society. As a side note, journalist Andrew Cockburn noted that Pimentel’s “dislike of aliens apparently extends to the human variety, as evidenced by his public opposition to both legal and illegal immigration.”)
The point here is not to pick on Pimentel or Wilcove or et al. (though other writers have been happy to do so) but rather to point out that numbers in papers like these aren’t as “black and white” as they sound when they’re pulled out and paraded around on their own.
We are not saying that the numbers are meaningless. They’re based on something, after all. Just not actual counts of solid objects in the real world. But that’s how statistics work, and there’s nothing new about that. As has been said for over a century: “There are three kinds of lies: lies, damned lies, and statistics.” But most people don’t know that about statistics, so they hear “$1.4 trillion” and assume that’s a genuine total, as in, the added-up sum of other concrete numbers. But it’s not.
For example, let’s look more closely at that $45 million for Purple Loosestrife “control” (eradication) in Pimentel’s paper. In a table of “losses,” does this count? Wouldn’t many people consider control a “cost?” Put it this way: if you go out for a night on the town and someone steals $45 from you, or it slips through a hole in your pocket, that’s a loss. But if you spend it in a bar, that’s a cost. In the first instance, the money is gone against your wishes, but in the second you willingly gave it up. And if the latter is the case—that it all went to beer and smokes—would you complain to your roommates about the $45 you “lost?” If you did, we bet they would not express much sympathy for you as they would for a mugging. Just saying.
But that’s only common sense and the branch of mathematics known as statistics operates in a different world where it is indeed standard practice for a cost to be counted as a loss, as Pimentel did.[3] It is totally allowable to add a number based on a projection (lower yield from weeds) with a number copied from a receipt (cost of herbicides); that is, to treat a guess and an actual outlay as equivalent.
If that’s how statistics operates, fine. Good to know. But again, how many people know that? The impressions these numbers create are real enough, and more so are the subsequent actions such as the use of herbicides. We reckon that very few scientific papers count as a loss the number of native plants killed as collateral damage when glyphosate is sprayed.
Monsanto, a manufacturer of glyphosate, is—like most giant corporate entities—well-practiced at manipulating statistics for their own benefit and are happy to take advantage of those who are less savvy, such as harried elected officials or over-eager conservation officers, to say nothing of the general public. We would guess that if anyone from the biotech industry has ever checked Pimentel’s work, it’s just to make sure he’s not under-estimating. We have a feeling they’ve never bothered, though, since his attitude is exactly what they want.
And what about the 788 papers that cite Pimentel’s work? Or the 700+ that refer to Wilcove’s? Should scientific researchers know better? Is it enough for them to just skim the abstract and the conclusion? Or should they be digging deeper? Following citations to confirm that they check out? Adding up numbers to see that the math is right? (It’s not, by the way, in the grand total on the far bottom right in Pimentel’s table 2, and is off by nearly a factor of 10.)
Are some authors just copying-and-pasting a familiar citation just because it’s familiar? Because dropping that name shows what club they’re in? Because it’ll help in the next round of funding? Scientists are like everyone else and feel the same social pressures to conform, the same fears of alienation. Most of them don’t want to work too hard or stick their neck out too far when it comes right down to it. So maybe we shouldn’t expect too much of them?
Also, many scientists are on corporate payrolls, directly and indirectly, and their words are not to be taken at face value. For three decades, Monsanto was able to conceal the dangers of glyphosate, with the assistance of compromised scientific researchers being essential to that effort (see here, here and here). Now their false narrative is officially falling apart thanks to a high profile court case. But, vitally, the entire time they were spreading lies, other people toiled to bring the real facts to light. Their work and dedication is now being vindicated. We’d like to think that, in the end, “the truth will out.”
Notes:
[1] Luoma, JR. 1997. “Catfight.” Audubon 99 (4): 85-90.
[2] Pearce, Fred. The New Wild: Why Invasive Species will be Nature’s Salvation. (Beacon Press: Boston, MA, 2015), p. 279.
[3] Stolen, Justin. Private correspondence with this retired statistics professor, 12/30/18.
Nicole Patrice Hill holds a bachelors degree in Environmental Science with a minor in Botany. She is a former farmer who has been exploring the wildtending life in the US American west. Ms. Hill can be reached at wildwiskedjak (AT) riseup (DOT) net
