Dr. Rob Dillon, Coordinator





Monday, September 9, 2013

Plagiarism, Paul Johnson, and The American Fisheries Society

If you’re just joining us.  About halfway through this essay I am going to assume that you have previously read last month’s post, regarding Delaware and the nonprofit organization NatureServe.  If you have not already done so, please take a minute to back up to [19Aug13], and then read forward.

My first impression of Dr. Paul D. Johnson was that of a “hard-charger.”  It was November of 1998, and I had been invited to Chattanooga to join a committee primarily composed of natural resource managers, fresh from organizing two successful meetings on unionid mussel conservation in the Midwest, interested in expanding their portfolio to include freshwater gastropods and going national.  That weekend we drafted a constitution for The Freshwater Mollusk Conservation Society and made plans for a first general meeting, to be hosted by Dr. Johnson in Chattanooga four months later.  I was pleased to accept the chairmanship of the FMCS Gastropod Committee that Saturday afternoon in November.  And I was honored to nominate Dr. Johnson to the office of President-elect of the entire society in March, from which he ascended to the presidency in 2000.

Even as early as 1998, Dr. Johnson was advocating a “national strategy” for the conservation of freshwater gastropods, to be modeled after a mussel strategy then nearing completion by the group.  This project would involve the development of a list of North American freshwater gastropods prioritized for conservation purposes.  And of course, Dr. Johnson envisioned that such a list would arise from a collaborative effort, presumably coordinated by the FMCS Gastropod Committee. 

Although I was not opposed to the idea (15 years ago), it was my strong opinion that our committee’s first order of business ought to be a comprehensive survey of the continental freshwater gastropod fauna, only after which conservation priorities might be assigned.  I have also developed moral scruples regarding the admixture of science, politics, and public policy, which have deepened in recent years, but no point in going down that road here [1].

In any case, I declined to become involved with Dr. Johnson’s “national strategy,” passing the chairmanship of the FMCS Gastropod Committee to him in 2002.   The effort seems to have subsequently shifted home, from the FMCS to the American Fisheries Society Endangered Species Committee, which has in recent years become a center for such work on the aquatic biota in general.  I have remained on the sidelines, hoping for the best while fearing the worst.  And in June, alas, my worst fears were realized.

In June Dr. Johnson and 13 of our friends and colleagues published a feature article in Fisheries, the peer-reviewed journal of the American Fisheries Society.  It is entitled, “Conservation status of freshwater gastropods of Canada and the United States” [2].

Although the paper extends to 36 journal pages, details regarding the development of the data upon which the Johnson/AFS recommendations of “conservation status” were based are extremely vague.  Here is the single relevant sentence from the methods section, quoted in its entirety:  “Species occurrences within provincial and state boundaries were generated using primary literature, including provincial and state checklists where available, as well as personal communications with professional who are knowledgeable about certain groups or regions.”

Now I have some very, very bad news to report.  If you open a new window in your browser today (9Sept13), go to the USGS website hosting the Johnson/AFS database, and execute a map query for Delaware, you will find almost exactly the same list of 8 species you received from your identical query of the NatureServe Explorer database last month.

This is a peculiar list.  Missing from it are the four most common gastropod species actually inhabiting the freshwaters of Delaware: Physa acuta (aka P. heterostropha), Menetus dilatatus, Ferrissia fragilis, and Lymnaea (Pseudosuccinea)columella.  All four of these species are very nearly cosmopolitan in their distribution throughout eastern North America, and simple reference to the collections of either the DMNH or the ANSP would have returned numerous Delaware records for most of them.

The Johnson/AFS report also includes one species that our extensive field surveys of Delaware and attendant reviews of systematic collections have failed to uncover, Physa gyrina.  Populations of Physa gyrina do inhabit Ridley Creek in Delaware County, southeastern Pennsylvania, so on first reading it certainly seems possible that the Johnson/AFS record might be bona fide.  Or might this record represent a misidentification of Physa acuta?  The DMNH collection does hold a single undated lot of P. acuta (locality just “Wilmington”) misidentified as P. gyrina.

On 3Aug13 I sent an email inquiry to Dr. Johnson, asking if he could provide a reference to the primary literature or any other source available to him supporting his report of P. gyrina in Delaware, with 12 of his coauthors on the CC line.  Dr. Johnson has not favored me with the courtesy of a reply [3].

The match between the Johnson/AFS database and the NatureServe database is simply too close to be coincidental.  But at no point in his paper does Dr. Johnson acknowledge NatureServe as the origin of his primary data – not in methods, results, or acknowledgments.  The NatureServe organization is mentioned only on pages 250, 252 and 263 with regard to its system of conservation ranking, and cited only with respect to conservation ranking in the reference section.  The Johnson paper does not include a citation to the NatureServe Explorer as explicitly required by NatureServe for the fair use of its data.

The match is not perfect.  The exotic Bellamya (“Cipangopaludina”) chinensis was deleted from the Johnson/AFS Delaware list, and Helisoma (Planorbella) trivolvis added, indeed #11 on the confirmed list soon to appear on the FWGNA website.  The generic nomina of Physa gyrina and Fossaria obrussa have been emended to Physella and Galba.

But the evidence of plagiarism is pervasive.  My thirty years of experience grading the genetics lab reports of lazy college sophomores have (alas!) given me way too much practice identifying the phenomenon [4].  I have footnoted analyses of the situations in West Virginia [5] and New Jersey [6] below.  If these examples do not constitute sufficient evidence to convince my readership that the extensive data table reproduced in the appendix of the paper by Johnson and his colleagues did not originate from the NatureServe Explorer, tell me how many more such examples are necessary, and I will supply them.

Dr. Paul D. Johnson and his 13 colleagues stole a crappy, spurious dataset off the internet, tweaked it to the point they thought nobody would catch them, put their names on it, and transferred it into the peer-reviewed literature without attribution.  Shame on everybody involved: Arthur E. Bogan, Kenneth M. Brown, Noel M. Burkhead, James R. Cordeiro, Jeffrey T. Garner, Paul D. Hartfield, Dwayne A. W. Lepitzki, Gerry L. Mackie, Eva Pip, Thomas A. Tarpley, Jeremy S. Tiemann, Nathan V. Whelan and Ellen E. Strong.  Your mothers taught you all better.

And both the American Fisheries Society and the USGS Southeast Ecological Science Center are now accessories to egregious plagiarism.  The Johnson paper must be retracted, with apologies to NatureServe and to the scientific community at large.

Because the damage extends beyond that done to the professional reputations of Paul Johnson and his 13 collaborators.  The greatest damage is that done to science.  For what was merely the conventional ignorance of the worldwide web has now been transformed, by its publication in what appears to be a reputable journal, into ignorance of a high and aggravated nature, disgorged by 14 professionals whose credentials would lead one to expect some minimum level of scientific rigor, wrongly.  And perhaps a bit of integrity, for a change.


Notes

[1]  The language, culture, and values of science are not incompatible with those of law, politics and public policy, but they are not compatible either.  And over the years it has become clearer to me that much damage is done by workers with either worldview when we try to force a fit with the other, directly analogous to the damage done when a false compatibility is forced between public policy and religion, or science and religion, for that matter.  See any of my essays labeled “Worldview Collision” at right for more.

[2] Johnson, P. D., A. E. Bogan, K. M. Brown, N. M. Burkhead, J. R. Cordeiro, J. T. Garner, P. D. Hartfield, D. A. W. Lepitzki, G. L. Mackie, E. Pip, T. A. Tarpley, J. S. Tiemann, N. V. Whelan & E. E. Strong (2013)  Conservation status of freshwater gastropods of Canada and the United States.  Fisheries 38: 247 – 282.

[3] The complete correspondence record is as follows.  On 2July13 I sent an email to Dr. Johnson inquiring if he might be willing to share his database of occurrences with me, and asking for additional detail on the method by which these data were converted to conservation status recommendations.  Twelve (of his 13) coauthors were on the CC line.  (I have been unable to find an email address for Tarpley.)  I received no reply from any of the 13 recipients.  On 3Aug13 I sent a second email to Dr. Johnson, again with 12 coauthors on the CC line, simply requesting information regarding the occurrence of Physa gyrina in Delaware.  I received one fragmentary reply from Mr. Jay Cordeiro, who abruptly broke off our correspondence when I asked for clarification.  And I have heard nothing since.

[4] The sad science of plagiarism detection focuses on the “shared bonehead error,” or SBE.  If student #2 copies the errorless lab report of student #1, he will not be caught.  If student #2 copies a lab report, finds errors and fixes them, he will not be caught.  In fact, if student #2 copies a lab report containing reasonable errors, for example “three squared equals six,” he will not be caught.  The key to detecting plagiarism is the situation where student #2 copies a bonehead error, for example “three squared equals seven.”  Such “shared bonehead errors” are like fingerprints.

[5] NatureServe’s West Virginia list includes 28 species.  To this list Johnson/AFS added nine nomina – five valid species and four junior synonyms.  But there are two SBE omissions on both lists: Lyogyrus granum and Physa (“Physella”) gyrina.  This despite the fact that reference to the ANSP collection online would return 1 West Virginia lot of the former and 6 lots of the latter.

[6] NatureServe’s New Jersey list includes 22 species, three of which are exotic or introduced, reducing the list to just 19.  To these 19 Johnson/AFS added ten nomina – nine of which are specifically valid.  But again there is one SBE omission [7] on both lists: Helisoma (Planorbella) campanulatum, of which the ANSP collection online holds 15 New Jersey lots.

[7] There is also at least one glaring “shared reasonable error” on the two New Jersey lists.  Neither includes Ferrissia fragilis, which is #8 most common of the 30 species (no not 22, no not 28) soon to be documented on our Freshwater Gastropods of the Mid-Atlantic website.  This omission does not qualify as a “shared bonehead error,” however, because no national collection (to my knowledge) actually holds a single New Jersey record of F. fragilis.  The ANSP does hold 8 lots of F. fragilis from New Jersey, but five are curated as “Ferrissia sp.” in their online database, and three are misidentified as Ferrissia rivularis.  Thus the absence of F. fragilis from the New Jersey tabulation of Johnson/AFS does not constitute evidence of plagiarism.  It is a glaring example of that simple, conventional ignorance which we all ought to be working to fix, together.