Image for Fox Scientists: Both Sides Stumped?

PREAMBLE:

Officially, the Invasive Animals Cooperative Research Centre (IA-CRC), has seemed unwilling to consider that fox scats, imported from the mainland for use in ongoing training of detector dogs, may have subsequently been ‘discovered’ in the field and identified as scat coming from a fox living wild in Tasmania. Since 2003, the CRC, through the work of Professor Stephen Sarre, has been involved in work associated with the incursions of foxes onto our island.

The former CEO of the IA-CRC, Professor Tony Peacock, told the Tasmanian Parliament’s Public Accounts Committee (PAC) [Here] that he was

dismissive of claims that somebody is planting scats; in fact I’m extremely dismissive of it because evidence was never presented to me that this is the case……I can’t see how it could happen without many people becoming aware of it

Professor Stephen Sarre, whose ‘Foxes are now widespread…’ 2012 paper seemed predicated on the local provenance of the collected scats [Here] , seems to have moved to dispel these claims by recently mentioning that his team are now looking for contents of prey in collected scats. In particular they are looking for DNA of endemic species or of endemic species groups. Said Sarre recently (on Catalyst Nov 2014 [Here] )

One of the main objections they’ve had to it [the reliability of the DNA evidence] is that there’s a risk that scats [have] actually ... been planted by people

[...and (concerning the Great Poo Hunt 2014)...]:

If we find in these scats that they contain prey that are unique or an assemblage of prey that is unique to Tasmania then I’d say that the evidence that they’ve been free-living in Tasmania is indisputable.
[…]
So far the evidence is very strong that foxes are, or have been, present in Tasmania. I think, if you took that evidence and then didn’t act, that would be a far bigger risk for people of Tasmania and indeed the world because of the unique bio-diversity in Tasmania.

INTRODUCTION:

Forensic Science International (FSI) has recently published a third letter from the authors of conflicting scientific papers pertaining to the Tasmanian identification of animal scats as testing positive to fox DNA. These letters follow publication of a number of scientific papers, chiefly Berry (2007) “Faecal DNA detection of invasive species”, and Sarre (2012) “Foxes are now widespread in Tasmania” These articles had sought to establish a scientific basis which would allow for the identification of any foxes in Tasmania and to provide for a means of undertaking their removal. There have been a series of other papers which can be characterised as responses to these two: they come from the so-called ‘independent scientific review’ group of scientists and fox experts who host a website called “Tasmanian Fox? : A scientific review of the Tasmanian fox eradication program”.

Veterinary pathologist David Obendorf is central to this ‘sceptics’ group while on the other hand, Stephen Sarre, from Canberra University, who works with the other group of scientists and experts and has received funding from the Invasive Animals Cooperative Research Centre, holds the standard for the ‘fox believers’ brigade. These competing groups, both trying to win the day via science rather than rhetoric, have joined battle, while many of the onlookers to the jousting, view it either from the perspective of nonplussed audience or from partisan boxes on each side of the tourney.

The nonplussed audience accepts that individual live foxes by accident and/or by design, have actually reached our island, It accepts that there have been sporadic incursions from across Bass Strait yet it does not seem to accept that there might have been deliberate introductions from the mainland. These ‘nonplussees’ have become over years, predisposed to question the competency of the former FEP task force, and to question the authenticity of various sightings and of various physical evidences. What follows is a summation of three journal letters from these two scientific groups, and my own commentary on the contents of those letters…

THE 2014 LETTERS (See endnote i):

1. 7 Mar. 2014: Gonçalves et al. “The risks of using ‘species-specific’ PCR assays in wildlife research…” [Here]

2. 26 Jun. 2014: Sarre et al. “Defining specificity in DNA detection of wildlife…”. [Here]

3. 24 Oct. 2014: Gonçalves, Marks and Obendorf “Reply to Sarre et al. ...” [Here]

DETAIL:

1. THE ORIGINAL GONCALVES LETTER:

Gonçalves et al.“sought to better evaluate the specificity of the putatively fox-specific pair of PCR primers (VV-cytb F and VV-cytb R) designed by Berry et al. [16] using DNA from a wider range of species”.

After an account of the various annealing temperatures used by Gonçalves, there followed a discussion of the significance of mismatches in the primer-template duplexes, and Gonçalves concluded: “Therefore, the putatively fox-specific primers VV-cytb F and VV- cytb R did not specifically hybridize to fox DNA, despite being designed with the intention to specifically exclude species with morphologically similar scats to the fox” [20].

Following discussion of the possibility of contamination of faecal samples due to the presence of prey DNA within the scats, Gonçalves concluded: “Consequently, the species-specific assay proposed by Berry et al. [16] must contend with the likelihood of strong PCR amplification arising from extremely abundant forms of non-fox DNA. Irrespective of what post-PCR laboratory methods are used, VV-cytb F and VV-cytb R do not permit sufficient specificity for the unequivocal determination of fox DNA.”

2. THE SARRE RESPONSES TO THE GONCALVES LETTER:

Sarre et al. made the following responses to the Gonçalves et al. letter:
“Unfortunately, Gonçalves et al. [1] have undertaken an analysis of an incomplete imitation of the test we applied.”
1. Sarre said his approach involves a “sequential two phase approach”, using both a multiplex of specific and universal primers and a

“direct sequencing of the PCR product obtained from a second amplification using the VV-cytb F and VV-cytb R primers only. A scat is considered to contain fox DNA only when a direct match is found between the sequence obtained from the fragment amplified and published fox sequences [2,3,6].
This sequential two phase approach, which is described explicitly in Sarre et al. [6], was adopted to make cost effective the screening of predator scats in the large numbers typically required for the detection of rare fauna. The specificity of this approach is a product of the combination of the two phases of the test and cannot be meaningfully estimated without considering both phases.


2. That the specific and universal primers were not considered as a multiplex, but were treated separately. Thus, according to Sarre, Gonçalves had not evaluated the test in its entirety, but only one part of it.
3. Sarre said of recent blind trials which included 527 scats of known origin, [of “foxes, cats, dogs, Tasmanian devils, and two species of quoll (Ramsey et al. in review).”] of trials which incorporated Berry 2007 [and implicitly the sequential two-phase approach] that they didn’t identify as fox any scats which were not fox: he wrote that his “data therefore run counter to the findings of Gonçalves et al.” [1]
4. Sarre wondered if the difference in results might be due to different lab procedures and conditions.

3. THE OBENDORF (see endnote ii) RESPONSES TO THE SARRE LETTER:
Gonçalves, Marks and Obendorf made a number of responses to the Sarre et al. letter:
1. Gonçalves, Marks and Obendorf rejected the Sarre et al. claim that they were in error due to not having applied a ”sequential two phase approach”, and they explained that the Berry 2007 method (referred to by Sarre et al.) did not describe such an approach.
2. Gonçalves, Marks and Obendorf, referring to being denied access to data derived from “approximately 10,000 putative fox scats” (see endnote iii), rejected a supposed implication from Sarre et al. that they had claimed that these data were generated by the “exclusive use of [Berry’s] putatively fox-specific PCR”.
3. Gonçalves, Marks and Obendorf rejected the Sarre et al. suggestion that variation in the PCR conditions in their lab and possible contamination within the lab might account for Gonçalves, Marks and Obendorfs’ results.

DISCUSSION:

Berry (2007) had created a laboratory test which he claimed could accurately identify fox DNA present in predator scats presented for examination. In my opinion, his development of ‘universal’ and ‘fox-specific’ primers represented an advance in the fight against the fox in Tasmania and are to be applauded.

Using this, Sarre (2012) identified as ‘fox’, a number of collected scats and analysed their collection locations to develop a model of suitable fox habitat and thence of likely distribution within Tasmania. He argued for a greater effort to eliminate the foxes which he said were ‘widespread in Tasmania’. In my opinion, Sarre’s study failed – not because of the science and argument – but because he took without question the provenance of the scats that came to his team for analysis. Any fair person must hold up to the light the record: As Ian Rist (farmer-hunter and critic of the former Fox Eradication Program) has pointed out, since the furore of 2001, we yet have no photographs of live foxes here in Tasmania, none. And neither in that ensuing period have any foxes been caught by being shot, trapped or baited (see endnote iv).

On the other hand, there have been some highly credible sightings, there have been discoveries of physical remains (the authenticy of which is disputed by some) and there has been the identification of some 60 or less scats from around the state as containing fox DNA. This is over a period of some 13 years. We know that a large number of fox scats (1,200?) have been imported from the mainland for various purposes, principally I think, for use in training fox-scat sniffer dogs. Records obtained under FOI showed that a number of those scats held by the DPIPWE were ‘written off’ as destroyed due to poor condition. Furthermore, at the time of the FOI disclosure, there were some records of a number of scats which were signed out to members of staff and which had not been signed back in. In the period 2008- 2010, 44 scats were signed out by a number of staff, including one of the Sarre 2012 co-authors, - but not signed back in. It might be the case that the last of those sign-outs were subsequently returned after the FOI (RTI?) disclosure. Yet, they were still in breach of the commitment that scats would be signed out for a period of 1-2 weeks and then returned. Given these shortcommings it would be helpful if DPIPWE released all records pertaining to the use of imported scats in Tasmania. (For example: imported-scat ‘running sheets’ which were not released under that original FOI request)

Gonçalves 2014 letter, speaking for the ‘TasmanianFox?’ group of scientists, concluded that the 2012 Sarre paper was flawed, not for the bureaucratic and in-the-field discrepancies, but for the science. In her letter she canvassed the chance of contamination from (ingested?) prey and thus, the specificity of one of the primers:

“Consequently, the species-specific assay proposed by Berry et al. [16] must contend with the likelihood of strong PCR amplification arising from extremely abundant forms of non-fox DNA. Irrespective of what post-PCR laboratory methods are used, VV-cytb F and VV-cytb R do not permit sufficient specificity for the unequivocal determination of fox DNA.”

.

Be that as it may, I’d like to point out that Gonçalves 2014, with her “Irrespective of what post-PCR laboratory methods are used” has shown knowledge of a Sarre 2007 paper (which had followed Berry’s 2007): [Here: DNA Detection of foxes to prevent establishment in Tasmania (2007)]. In this paper, Sarre had explained the second phase of Berry’s test. This is some five years or so before his “Foxes are now widespread in Tasmania” 2012. and seven years before Gonçalves’ 2014 letter. In my opinion, Sarre, in his 2014 letter, is entitled to reply, as he does, that Gonçalves has examined one part of his (and Berry’s) test, but not the whole. To some extent I therefore think that it is irrelevant for her to assert that “Irrespective of what post-PCR laboratory methods are used, VV-cytb F and VV-cytb R do not permit sufficient specificity for the unequivocal determination of fox DNA.” without dealing with the second-phase sequencing part of the test. Her letter shows no evidence of the TasmanianFox? group having conducted any post-PCR laboratory work.

Sarre’s letter response to Gonçalves, reminding the Tasmanian Fox? group of the sequential two phase aspect of his and Berry’s DNA PCR approach, of the multiplex nature of their test primers and of their follow-up testing with a range of native and introduced animal had some force. His questioning of the possibility of differing lab conditions (and contamination?) contributing to the differing results, was in my opinion, fair enough, though admittedly speculative: for example, in regard to the variety of annealing temperatures used by the TasmaniaFox? group, while yet claiming to be conforming to Berry’s process. To me, it does not seem that they did conform. I am willing to discuss the disparity in the annealing temperatures with anyone who is interested. Sarre had made his point. In my opinion, it would need some clear argument to counter it.

However the ‘Obendorf’ reply was in my opinion, not much more than a restatement of what had gone before. To assert that Berry 2007 didn’t use a sequential two-phase approach so that they, the TasmanianFox? group needn’t do so even as they rejected Sarre’s 2012 paper which did use a two-phase approach … is not convincing. Especially since Sarre 2007 had published material to accompany and expand on Berry’s work, and especially since Gonçalves’ Letter had shown awareness of it, while yet dismissing it without discussion. Here is some of Sarre 2007 (the emphasis-formatting is mine, please read endnote 5 regarding the sometimes confusing use of the word ‘test’):

With appropriate verification of the methodology ((Taberlet and Luikart 1999), non-invasive DNA-based methods could provide the high-quality distribution data that is required for effective control. To this end, we have developed a PCR-based test specific to foxes that excludes the amplification of other carnivore DNA and provides a rapid initial screen of all scats collected. Further verification is required following the initial identification of a scat positive for fox DNA. The full details of this test and its development can be found in Berry et al. (2007).
[…]
3. The test was implemented by the Department of Primary Industry and Water in 2004 for scat identification activities on contract to the Wildlife Genetics Laboratory at the University of Canberra. Following the initial
implementation, additional analytical procedures were added to the test to minimise the risk of false positive identifications of fox traces (see below).
[…]
The implementation of the test for samples derived from Tasmania necessitated the development of a series of protocols to maximise the value of each sample collected and to minimise the risk of Type 1 errors (false identification of a sample as fox when it is not)
[...]
To minimise the risk of contamination by prey items, the primers used to amplify the fox specific cytochrome b fragment in our study were designed to specifically exclude the amplification of homologous sequence from other large carnivores in Tasmania (Berry et al. 2007). However, occasionally, non-target fragments of a size similar to the target fox band are amplified and require investigation. Such fragments are sequenced and compared to sequence contained in the world-wide database GenBank (http://www.ncbi.nlm.nih.gov/
blast/Blast.cgi) and from our own sequence records to determine the species of most likely origin. Only when a sequence from a scat matches specifically to fox do we consider a scat to be positive for fox (Figure 2).

CONCLUSION:

1 Those scientists claiming that ‘foxes are now widespread in Tasmania’ have failed to authenticate the provenance of the scats on which they base their claims. Thus their conclusion is invalid.
2 Those fox-sceptic scientists who have challenged Berry and Sarre’s science, have likewise failed in the laboratory, to make their case.
3. Live foxes have been here in Tasmania. It is possible that some still persist, but all things considered, that seems unlikely.

ENDNOTES:

i Please note that these are the dates on which the letters were received by the Editor, rather than publication dates.

ii While Gonçalves is the first-cited author here, it seems reasonable to associate David with the construction of the letter - likewise Sarre with the second letter, (earlier). The alternative – referring to (the) two Gonçalves’ letters – would perhaps be too confusing.

iii These c. 10,000 scats were not putative fox scats, e.g., the Great Poo Hunt had invited collection of any predator scats (from both introduced and native species) in order to gain knowledge on the distribution of these various species and on other matters, such as dietary information.

iv The so-called ‘Bosworth fox’ is a possible exception: Eric Bosworth in a signed statement 13Sept2001 said that he found a dead fox in the same place as he had spotlighted and shot at an animal 10 days earlier. The fox carcass then became the subject of a police investigation.

v Sometimes it is not clear that Sarre, when referring to the Berry 2007 test, is distinguishing that PCR test from the second-phase sequencing comparison step. Sometimes, it seems as if it is included as part of the ‘Berry 2007’ test, at other times it is clear that the PCR test is the test being referred to. However, it seems clear to me, that from 2004 (three years prior to Berry 2007) post PCR lab. work was to be undertaken: the PCR phase was to be followed by “additional analytical procedures” … which I take to mean DNA sequencing comparisons.