I have read many submissions by environmental groups (eNGOs) protesting against seismic surveys which, given the content, certainly suggests manipulation is happening.

As this is a long article, the following is a summary of the key points that will be made:

i). eNGOs generally “cherry-pick” or manipulate their document search to only refer to sections or studies that support their claims. They rarely, if ever, produce a balanced review of the research and appear unable to differentiate between good science and poor science lacking in rigour (ie pseudo-science). In fact, they appear to thrive on pseudo-science.

ii). There are, unfortunately, many examples of pseudo-science in the literature, in which researchers have: set out to prove a particular hypothesis; utilised poor methodology with limited controls; exposed subjects to unrealistic stimulae and made claims that are not supported by the data.

While many eNGO submissions would probably look quite impressive to a reader not familiar with the facts and science, most of the claims in them appear to deliberately misrepresent the research and/or rely on what can only be described as poor science lacking in rigour with no application to the real world.

The following is just one example, of many that could be cited, as to why most of the ‘scientific claims’ of eNGOs are dubious and should be viewed with suspicion. It also demonstrates why some of the studies eNGOs cite should be critically evaluated (by them!) before the studies are incorporated (or not) into their claims.

This recent example used the work of Aguilar de Soto et al (2013) entitled “Anthropogenic noise causes body malformations and delays development in marine  larvae”  to make spurious claims of impacts from seismic surveys. Aguilar de Soto et al claimed that their study provided the first evidence that noise exposure during larval development [of scallops] produces body malformations in marine invertebrates. The essence of their results, which are now cited by eNGOs, are that  “scallop larvae exposed to playbacks of seismic pulses…showed significant developmental delays in the animals and 46 percent developed body abnormalities”.  HOWEVER, this study was NOT representative of the sound exposure in the vicinity of a seismic survey IN A REAL LIFE SITUATION. Furthermore, it could not possibly, as the authors claim, provide “understanding (of) the impact of noise on marine fauna at the population level“.  Strangely, even the authors stated that “the inaccuracies arising from measuring the sound field in a small tank as in our experiment, makes it difficult to predict if larvae were subjected to higher exposures than might arise from anthropogenic noise sources in the oceans”.

Given the parameters used for this study, it is rather obvious that the sound levels to which the scallop larvae were exposed are totally unrepresentative of sound levels from seismic surveys to which larvae in the open ocean would be exposed.  The following are a few key differences:

1. Pulses were at 3 seconds intervals, not 10 seconds as in most seismic surveys.

2. As per Fig 2 of the paper, the “seismic” pulses, which had been amplified from seismic pulses recorded at “tens of kilometres” from a seismic vessel, were 1.5 seconds long. Due to limitations used in the amplifier/speaker used in the study, the pulse is very different from a normal seismic pulse as shown in an article on this website comparing a seismic pulse with the sound level of a breaching humpback whale. This means that primary sound exposure in this study was for 50% of the time (1.5 seconds duration divided by 3 seconds interval), whereas primary sound exposure in seismic surveys is for about 0.5% of the time (0.05 seconds duration divided by 10 seconds interval). This figure of 0.5% is also supported in an article by John Hildebrand of the Scripps Institution of Oceanography, which shows in Table 1 the duty cycle (or exposure) time for seismic surveys is 0.3%.

3. Exposure was in a small tank (2m diameter by 1.3m deep), not in the open ocean. Therefore the larvae in the test were exposed to very strong particle motion (near-field effects), which may have had more damaging effects than the measured sound pressure. The authors themselves stated that “the sound field experienced by an organism is a complex function of its location with respect to the sound source and acoustic boundaries in the ocean necessitating in situ measurements to establish the precise exposure level. This, and other inaccuracies arising from measuring the sound field in a small tank as in our experiment, makes is difficult to predict if larvae were subjected to higher exposures than might arise from anthropogenic noise sources in the oceans.

4. The source was stationary, unlike in seismic surveys, which move at about 2m/sec. Hence, exposure times for relatively stationary larvae in the ocean would be further reduced below 0.5%. This is an extremely important difference.

5. Minimum exposure period was 24 hours of effectively continuous exposure if reverberations in the tank are considered. This is very different from seismic surveys in which exposure periods would be far less.  An article on this site demonstrates that, during a typical 2D seismic survey, the exposure levels at one location within the survey approaches approx. 140 dB re 1 µPa2/Hz for about 0.5 hours only 5 times during a period of 120 hours or approximately 2% of the overall time.  In this case the one location is that of a sound recorder, which could be equated to the location of a marine organism if it had stayed in the same location for 5 days. Another article on the same site shows in Fig 3 that 140 dB re 1 µPa2/Hz is equivalent to approximately 178 dB re 1 µPa(rms).

6. The authors state that the intended exposure sound pressure level is 160 dBrms re 1 µPa but that the overall exposure level due to near field effects is closer to 195–200 dBrms re 1 µPa. Thus, not only is the exposure level in this study significantly higher than scallop larvae would be exposed to during seismic surveys, it would be for significantly longer!

7. Finally, as stated by the authors “Malformations were observed in all flasks of the noise-exposed group starting in the sample corresponding to 66 hours post-fertilization.”  It would be impossible to expose scallop larvae to such intense sounds for this period during a seismic survey. In fact, it would be impossible to expose scallop larvae to such intense sounds for even the 24hr minimum exposure period used in this study. It is noted that no malformations were recorded for the 24hr exposure sample.

This laboratory study, at best, represents a rudimentary effort to expose larvae to an intense noise regime that is not representative of the open ocean or a typical seismic survey. Significantly more refined methods would be needed to support the conclusions made by these researchers and hence the claims of eNGOs who refer to this study.

Thus, what did this study provide as “first evidence” of the impacts of seismic surveys on scallop larvae? Very little!  However, even though the study is totally unrepresentative of seismic surveys, the authors  claimed that the “results call for applying the precautionary principle when planning activities involving high-intensity sound sources, such as explosions, construction or seismic exploration, in spawning areas of marine invertebrates with high natural and economic value”.

What is more disturbing is that the authors, having presented a study of dubious scientific value to the understanding of the impacts of seismic surveys on larvae in the open ocean, also referred to a 1992 study (published in 1993 and 1996) that is of equally dubious scientific value.  This earlier study to which the authors refer, has not been validated during further investigations even by the same researchers.  In their article, Aguilar de Soto et al refer to the work of Engas et al (1996)  and state “studies on the impact of seismic surveys on fishing captures have reported…..reductions of 70% in the catch rates of mobile and valuable fin-fish species such as cod (Gadus morhua) and haddock (Melanogrammus aeglefinus)”. When one looks at the methodology used by Engas et al, it is very surprising that researchers and eNGOs still cite this work in 2013.

The field work carried out by Engas et al consisted of fishing using trawls and longlines in a 40 x 40 nautical mile (nm) investigation area of the Barents Sea for a period of 7 days before, 5 days during and 5 days after a 3 x 10 nm area of seismic activity occurred in the centre of the investigation area.

The questions that immediately come to mind are:

A.    What impact on catch rates would there be if fishing was conducted in such a restricted area every day for 17 days? One would have thought there would have been a significant reduction in catch without seismic activity being present.

B.    Where is the “control area”, that provides a comparison of the impact on catch rates of fishing the same area every day for 17 days without seismic activity being present?  No control area is mentioned in the paper

Figure 7 of the Engas et al paper, reproduced below, shows the obvious effect on catch rates claimed as a result of the seismic activity.  It can be seen that the impact within the survey area (near the survey vessel) is actually significant at about 70%.  However, is that not expected?  After all, standard mitigation measures for seismic surveys involve a ramp up (or soft start) in which the source is increased from lowest power to full power over a period of 30 minutes to ensure that marine life (eg. marine mammals and fish) have the opportunity to move away to a “comfortable” distance.  It is noted that this “comfortable” distance varies from species to species, with some species, such as dolphins, often riding the bow wave of the vessel and the trailing equipment while the source is active.

engas Fig7
Fig1 – Figure 7 extracted from Engas et al paper

It is very concerning that the catch results for 7 days before, 5 days during and 5 days after the seismic activity have be averaged into just 3 data points at each location relative to the seismic activity.  Has this been done deliberately?  I don’t believe so because individual catches, which would contribute to daily catches, are plotted in figures 8 and 10 of the Engas et al paper.  However, 50% of these plots appear to indicate there is a downturn in the catch rate BEFORE seismic activity begins (ie Figs 8A – trawl catches of cod & 10B – long-line catches of haddock).

As a result of all the questions this paper raised, requests were made to the Norwegian Institute of Marine Research (IMR) for access to the data from the study.  Finally, 18 years after the fieldwork was conducted, on June 30, 2010, the IMR, Bergen, released the data and the following is a brief summary courtesy of Ingebret Gausland.

The files covered both trawl and long line catches, but only the trawl data gave sufficient information to be analysed.  The long line data was in a format that could not be read by available programs.

On analysis, these data showed a very different conclusion from that portrayed in the original paper.  Of course, there was a significant drop in catch within the survey area.  What a surprise!  This is to be expected and, in fact, is what is wanted as part of the mitigation measures. 

However, when the daily catch statistics for each sample distance in the 7 days before seismic activity are displayed, it can be seen that the before sample for each location, including the location within the area of seismic activity, is highly variable, heavily weighted by a couple of good days’ catch early in the 7-day period and show a downturn BEFORE commencement of seismic activity.

There is insufficient space in this article to show all the analyses but the most relevant one would be the plot for 1-3 nautical miles (1.85-5.55 km) from the seismic activity shown as Fig 2 below:

IG analysis
Fig 2 (Courtesy of Ingebret Gausland). Average catch per haul for each day at 1-3nm from the seismic activity. Red bars indicate days in which seismic activity occurred. (Trawling vessel was in port on 4 and 5 May).

Perhaps all that the original study shows is that fishing in the same area for many days leads to low catch rates.  I would suggest this is already known by the fishing industry.  Similarly, low catch rates in the same area as a seismic survey is something that would be expected due to the avoidance response by the fish and the mitigation effect of the seismic source.  The key issue that both the fishing industry and petroleum industry needed to understand is whether catch rates are affected outside the seismic survey area but this study did nothing to add to our scientific knowledge. This is because the lower catch rates could very well be due to the fishing frequency and not the seismic survey.  In addition, I understand that this was not a fish feeding area (all the fish stomachs were empty) or fish spawning area, implying that fish were in transit through the area. Also, fishing activity in the area had been heavy even before the study commenced on 1 May.  Thus, there could very well be other reasons for the downturn in catch before seismic activity commenced and for the daily variability of the catch.

The study is, of course, now over 20 years old and has been superceded by more recent studies (Slotte et al 2003, Ona et al 2007, Lokkeberg et al 2009 and Pena et al 2013), which, in most cases, have contradicted this work.  Unfortunately, many organisations (including researchers!) still cite this work as evidence of the impact of seismic surveys on catch rates in areas surrounding seismic surveys when clearly it is not.

To conclude, I would like to use the following quote by Geoffrey Harold Sherrington, a scientist who has been concerned about some of the pseudo-science used in the climate change debate: “Good science is simply work that can be replicated, work that has not been falsified, work that considers and quantifies all possible variables, work that advances the state of knowledge, work that is complete before release – not incomplete and calling for the Precautionary Principle.” 

I would suggest that an additional requirement of good science is that it explains and confirms the facts and observations we see everyday as part of the laws of nature.  For example, fish seen in close to proximity to operating seismic vessels (including predatory fish and sharks that bite the depth controllers on seismic streamers, as they look like fins), dolphins riding the bow wave of the seismic vessel and seismic trailing equipment and even humpback whales getting very close to seismic arrays.

The work of Engas et al and Aguilar de Soto et al do not meet the above criteria.