Sunday, 13 October 2019

Looking back on the eDNA results

It has now been just over a month since Professor Neil Gemmell announced the results of his environmental DNA studies of Loch Ness to the world. I thought I would wait for the media frenzy to abate before I digested what data was available to the general public as well as asking Professor Gemmell to clarify some points for me.

So, as a believer in the Loch Ness Monster, what can I conclude from these results? In some sense, the survey follows on from two large scale searches of the loch over the last five decades and they feed into one another. First, we had the extensive surface watches of the Loch Ness Investigation Bureau in the 1960s. This was followed by the sonar surveys typified by the large Operation Deepscan in 1987. Thirty years later, science has progressed to the point where actual animal material in microscopic quantities can be extracted, sequenced and matched to known animals.

Unlike the previous two searches, the eDNA survey was focused on a study of the flora and fauna of the loch, the other two were concerned with finding evidence of the Loch Ness Monster. Nevertheless, there was the hope to many that the creature would intersect with the study in some way.


But before I compare and contrast these searches, what can be said about the results of the DNA survey? At the time of writing, a series of results have been published by Otago University at their Loch Ness Hunters website. These results can be searched by species or location and it is the species-oriented results that have got the most attention. If you did not know by now, here is the pie chart of species found.

I base my comments on the assumption this is a complete record of vertebrate matches. As I understand it, Arctic Char, Sea Trout and Brown Trout are not explicitly mentioned as they are likely included under "Salmon" as members of the salmonidae family alongside Atlantic Salmon. Perch are not explicitly mentioned either, but will come under the family name of percidae (which suggests an exact species match was not achieved). Some species appear to have been missed, namely otters, newt, carp and roach. The last two are actually unconfirmed by "official" sources, but people claim they are there.  So does their absence from the DNA record dismiss their existence? That depends if you believe people's anecdotes about them!

No reptile DNA was found though reptiles do live by the loch such as adders. lizards and slow worms. They can enter the water, but it is assumed that such ventures are too rare to leave any traceable DNA. The same could go for the amphibian newts. However, we know otters live by the loch and enter the waters, but they have been missed by the sampling regime. This will simply be down to the fact that sampling did not occur near places of otter activity.

Some results were surprising such as sample "Ness 10" in the middle of the loch being identified as 100% pheasant DNA! It was proposed by Professor Gemmell that this was likely due to some birds defecating into the loch at that point.

You will note there is no mention of plant life or microscopic life, one assumes they were not included in the public results. The main point here is that most but not all species were detected and that was down to the element of uncertainty in choosing the best sampling sites. In total, about 250 samples were taken at the loch at depths from 0.5 to 200 metres. However, if one looks at the public data, only 56 samples are listed.

That perhaps should give me pause for thought if not all the data is available. Professor Gemmell told me all the data would come in the scientific paper he intends to publish around the end of the year. But with that limitation in mind, I will proceed.

I also note that some of the sample results are not complete. For example, if one looks at sample location "Ness 2", there were three samples taken at 0.5m, 100m and 200m. The results are stated as 321 for humans and 1 for toads, but it does not state which applies to what depths. Clearly, the human result would be for the surface (perhaps human waste from a cruise boat), but at what depth was the toad DNA retrieved?

This brings to mind a headline from 2007 when a deep water survey team found a toad in Loch Ness at a depth of 98 metres (above). The depth of the loch at the point of sample "Ness 2", north of Inverfarigaig, was about 220 metres, so the team must have sampled the water at best 20 metres above the surface. One wonders how common this amphibian is at these depths? But until I see the depth data, the toad DNA could have been retrieved at the surface.

Looking at the public database, 11 of the 56 samples were taken at depths of 100-200 metres. I am not sure if this is a proportionate representation, in which case there were about 50 deep water samples taken. If not, we have 11 to work with and it is my guess, having eliminated terrestrial and pelagic creatures that next to no vertebrate matches were made at these depths which we call the Profundal Zone.


However, it may be that any data for these 11 samples are simply missing or too scarce because it has been established that fish do inhabit this zone as per the capture of a potentially benthic form of Arctic Char at 200m by the Loch Ness Project. Now the reason I am most interested in this depth data is because I believe the creature spends most of its time at these depths, generally not moving around, with occasional forays up the sides of the loch walls.

This can be established from the witness database. If we filter out the spike years of 1933-1934, we are averaging about 10 sightings per year since then - and that is before one decides what proportion of these are misidentifications and hoaxes. Clearly, this is a creature which does not surface very often which leads to the conclusion that it is not a creature that spends much time in the shallower pelagic zone.

The surface watches of the 1960s confirmed this and the more penetrating sonar surveys, though they do produce occasional sonar hits of the creatures, it is clear that it is in the deepest depths that these creatures must be sought. This is not so easy for sonar due to beam attenuation and widening, but this has been my belief since 2012 that any searches in the pelagic zone will be generally fruitless. 

But if only 11 deep water samples were taken, would much be picked up? The public data suggests nothing, but I await further information as surely some known lifeforms were detected. I would point out a further potential problem with eDNA surveys at these depths. The relatively inert nature of the abyssal plain does not lend to distribution and scattering of DNA due to the higher water pressure at those depths and the lack of disruptive water movements such as the higher thermocline. As Adrian Shine says:

In contrast to the turbulence and variety of physical conditions among the stones of the shoreline, the fine and relatively rich silts of the abyssal regions offer remarkable stability. In an environment of great hydrostatic pressure, constant darkness, and a scarcely changing low temperature of 5.6C, high oxygen levels (over 80% saturation), permit surprising variety in the profundal community of the 200m deep basin floors.

I am not sure how one gets around this more difficult sampling region. I had suggested prior to the team's arrival that they actually sample the silt itself which may preserve more DNA, but that did not happen. Note higher oxygen levels are a boon to the survival of bottom dwelling creatures. So the jury is out for me on what was detected and what was detectable at these extreme depths and I await Professor Gemmell's scientific paper.


Which brings us to the data that was certainly not in the data. It was stated that 25% of the DNA was not amenable to identification. When I asked Professor Gemmell about this his answer was that unexplained DNA sequences were generally short DNA sequences that can not be accurately attributed to any specific species or taxonomic group with statistical certainty. Most metabarcoding and eDNA studies have portions of these sequences so he did not see any significance in that data.

The more relevant piece of data in regard to this was the study conducted at the nearby lochs of Cluanie, Oich, Duntelchaig and Ashie. When I asked how much unidentified DNA was present in those loch samples, he said it was largely the same. In other words, we should not look for monsters in that unprocessed set of data.

It also has to be said that there was another indeterminate region in which low samples of DNA, though processable, were too small to produce enough confidence and hence were discarded. In other words, anything that came up with fewer sequences than was detected for any species in the negative control were discarded. 

As an aside, it had been mentioned in pre-trip publicity that Loch Morar would form part of the "control" lochs. Professor Gemmell informed me that the loch in the end did not get sampled, which was a pity given its monster tradition.


As said before, no reptilian DNA was detected and that would eliminate extinct animals such as plesiosaurs right away. But if reptile DNA was detected, how could one zero in on a plesiosaur identification? Professor Gemmell's approach was to use a rough DNA composite somewhere between crocodiles and birds. Some have suggested plesiosaurs lie closer to turtles by relation and this was conceded by Neil, though it is a moot point given the absence of reptilian DNA.

The matter of giant eels was the main takeaway from the conference given by Professor Gemmell. However, the truth of the matter is that the eDNA survey had only failed to eliminate giant eels as a monster candidate as giant eels could have the same DNA sequence as smaller ones. Indeed, it could never eliminate giant eels in the same way it could not eliminate giant salmon, giant dogs or giant toads! I suspect that conclusion was more a sop to the worried Highland tourism industry. 

My own take is that giant eels are not the main explanation for Loch Ness Monster sightings. They could not possibly account for long necks, land sightings, semi spherical humps or sustained surfacings. This would only be possible if these features are explained by other non-eel causes. I do concede the historical possibility of large eels in the loch, perhaps of the order of two or even three metres long. Whether these have played a part in surface sightings is indeterminate, but certainly not the corpus of accounts.

My own question to Professor Gemmell was how the eels had been identified in the loch as some identifications from DNA had only been accurate to the family level and not the species level. His confirmation was that the species anguilla anguilla had indeed been matched and that is where I think I will leave that theory.


In the broader scheme of things, Professor Gemmell's eDNA survey did not produce anything unusual from a cryptozoological point of view. He did mention some "surprising" results, but one must assume those surprises were confined to the microscopic level. I was not expecting anything from the majority half meter samples unless he struck lucky and a benthic monster had passed by that way in the last few weeks. Indeed, one may have passed that way close to June 2018 according to Gary Campbell's sightings register:

28 May - Morag Connor and her friend were driving north out of Drumnadrochit between 11 and 11.30 am. They saw a creature with a long neck with some humps behind it sticking about 7-8 feet out of the water and about 50 m from the shore. The creature had an all dark body but with no discernible head. They were unable to stop as they were driving and there was no place to pull over.

One suspects that if this was genuine, the creature's DNA had degraded by the time Professor Gemmell's team had started, even if he had intersected with its widening DNA dispersion trail back into the abyss. Be that as it may, we know that not all creatures were identified and that was purely down to the coverage of the loch not being complete and no one is blaming anyone else over that herculean task.

The Loch Ness Monster is a non-abundant species, just as I suspect the missing otters, newt, carp and roach are. It likewise did not intersect with the survey and that's just the way it statistically falls out,  especially if this species spends most of its time in the inert silt 200 meters or more below. But I will defer final judgement until I see the complete data when Professor Gemmell's paper comes out.

The author can be contacted at