‘Enhancement’ of fossil fish – recognition through EDX mapping

Article image: 
Jim Buckman
Institute of Petroleum Engineering, School of Energy, Geoscience, Infrastructure and Society, Heriot-Watt University, Edinburgh Campus, Edinburgh, EH14 4AS, Scotland
Currently managing and operating the ESEM facility within IPE at Heriot-Watt University. Research concentrating on biomineralization (ie brachiopods, serpulids, testate amoeba), shales and carbonate rocks, with special interests in techniques such as charge contrast imaging (CCI), digital rock characterisation and automation in SEM. Porosity and permeability modelling using digital rock physics (DRP), based on SEM images. Longtime interest in the use of ESEM for wettability studies.
Automated scanning electron microscopy (SEM) combined with energy dispersive x-ray (EDX) analysis, can be used to collect large scale elemental maps over several square centimetres. The technique was utilized to investigate the alteration to fossil fish, which were suspected of having been enhanced through the painting on of detail to fins and tails. Maps for phosphor (P) were found to represent an excellent proxy for bones and scales, while elemental maps for iron (Fe) clearly demonstrated areas of later artistic additions. The technique is non-destructive, and therefore has the potential to examine a range of museum specimens, without causing damage.
Institute of Petroleum Engineering, Heriot-Watt University, for access to SEM facilities.
Corresponding Author
The Eocene Green River Formation of Wyoming USA is well known for its diverse exceptionally preserved fauna of vertebrates (fish, sharks, rays), insects and associated flora1, forming death assemblages within a series of associated lake deposits1, commonly occurring within finely laminated sediments. In particular many fish species have been identified and form the most prolific aspect of the fauna1. Such fossil fish are highly coveted collectables, and can be found in their hundreds to thousands on numerous well known online auction sites, as well as readily available from museums and gift shops.
The value of such fishes varies considerably, but more complete, detailed, or unusual forms typically attract higher prices. This may inevitably lead to the temptation to “enhance” poorer specimens by the addition of detail. The question of fossil fish “enhancement” or “faking” has been raised within a number of online blogs2,3.
With this in mind, energy dispersive x-ray (EDX) mapping, carried out within a scanning electron microscope (SEM) was used to investigate a number of such fish, with the aim of testing the hypothesis that many aesthetically pleasing samples may have undergone enhancement to increase their monetary value, but at the cost of any scientific value.
For the purpose of the investigation, two samples of fossil fish (Fig. 1) were purchased from a retail outlet. These were believed to be from the Eocene, Green River Formation, Wyoming, USA, and appear to closely conform to other known Green River materials. 
Figure 1. (a) and (b) optical images of the two fossil fish samples examined.
Samples are 50 and 60 mm in length, and have been professionally prepared for the resale market. Discussion with staff indicated a suspicion that the professional preparation may have included a degree of enhancement. Therefore, both samples were investigated using Oxford Instruments AZtec large area maps (AZtec LAM) software, to provide elemental maps across the whole of each fossil. The latter software can produce large area elemental maps over areas of many square centimetres, by producing an image montage comprising a mozaic of multiple tiled images.
The limiting constraints on montage size are the available SEM chamber size and time (=cost). The analysis was carried out in an FEI Quanta FEG 650 SEM under low-vacuum conditions, which did not require the samples to be coated with a conductive medium (carbon or gold), and utilized an Oxford Instruments X-MaxN 150 mm EDX detector.
EDX mapping of both samples was successfully carried out, with maps collected for a range of elements, of which the most informative were for phosphor (P) and iron (Fe), both of which are shown in Fig. 2a-b, 3a-b. Although both P and Fe are clearly concentrated within fish-shaped outlines, the two elemental maps are distinctly different in detail. In both cases, maps for Fe display clearer detail on fin morphology (Fig. 2b, 3b), and in the case of Fig. 2b the presence of a tail that is not particularly apparent from the P map (Fig. 2a).]
The overlaying of Fe and P maps (Fig. 2c-f, 3c-f) are particularly informative. Within fin structures, there are no clear correlations in terms of shape or orientation of structural elements, as expressed by the distribution of the two elements, and in virtually all cases the outline of the fins extends considerably further in the Fe maps than for the P maps (Fig. 2c, d, 3c, d, f). 
The distribution of P has a clear affinity with the occurrence of bones (Fig. 2a, 3f) and scales (Fig. 3d), while that for Fe is present more as a broad wash and highlights areas such as fin rays (Fig. 2b, 3b) and the tails of the fish (Fig. 2g).
Figure 2. montage maps for fish in figure 1(a). (a) = phosphor, (b) = iron. Arrow in (a) indicates position of ribs. In (b) C-G represent areas of subsequent detailed images. (c) – (f) Combined P and Fe maps of fin structure, and in (g) of tail area.
Figure 3. montage maps for fish in figure 1b(). (a) = phosphor, (b) = iron. In (b) C-G represent areas of subsequent detailed images. (c) – (f) Combined P and Fe maps of fin structure, and in (g) of tail area. Arrow in (d) indicates position of scales, and in (f) fin ray bones.
As indicated, phosphor maps clearly correlate with bones and scales. The fact that the outlines of fins generally extend further in Fe maps than in P maps, does not necessarily imply that the Fe maps represent artificial enhancement, as fin spines (boney parts) do not have to extend the full length of the fin; as the outer fin can comprise solely of soft rays1. Nevertheless, disparity in the orientation of ‘fin-components’ between P and Fe (Fig. 2c, f, 3d, f) suggest that features observed in the Fe maps are enhanced painted on fins. 
This is also supported in Fig. 3c, where the pattern of boney spine material (P map) suggests post-mortem disturbance and reorientation, whereas potentially painted on fin structures (Fe map) are simple and straight. A similar painted on argument is also supported in the tail region. In one case, there is a clear mismatch between the distribution of Fe and P within the tail region (Fig. 3g), while in the other a tail is clearly marked in the distribution of Fe, while P is patchy and faint, with no obvious bone structure (Fig. 2g).
The possibility of enhancement of fossil materials destined for the collector’s market may seem a trivial matter, at worst parting a few extra “dollars” from the buyer for what would otherwise be a more mediocre less desirable specimen. Hardly a major crime, and given that the average fossil purchaser is unlikely to recognise enhancement, and is primarily interested in the aesthetics of their prized specimen (which has been improved), surely no problem exists? 
However, the number and position of fins, fin morphology and importantly the number of associated fin spines and soft fin rays, are important aspects used in species identification, as also are details of tail structure. If such details have undergone enhancement the scientific value of any such specimen will be significantly reduced, and may potentially lead to erroneous species classification. 
This technique is comparable to X-ray fluorescence (XRF) mapping4, although benefits from the addition of being able to simultaneously collect images using secondary electron (SE) and backscattered electrons (BSE), as illustrated from a close-up BSE montage of the head area (Fig. 4) of the sample in Fig. 1a. 

Figure 4. backscattered SEM image montage of the head of fish illustrated in figure 1(a).
The enhancement of fins and tails, and the application of an overall brown wash, has most likely been achieved by the painting on of a liquid medium loaded with crushed fossil material (bones and scales) obtained from more poorly preserved specimens. This would explain the occurrence of a ghostly phosphor tail, contrasting with the bright iron-rich fish tail in sample 1 (Fig. 2a, b, g).
Individual fine details, such as extended fin support spines, would likely have been applied with a fine paintbrush. Such application would be susceptible to removal, either accidental or through purposeful testing for fakery, which may explain the occurrence of the addition of a surface lacquer / varnish across the whole specimen surface (Fig. 5).
Figure 5. backscattered SEM image of surface lacquer / varnish.
1. Grande, L. 1984. Paleontology of the Green River Formation, with a review of the fish fauna.  The Geological Survey of Wyoming, Bulletin 63.
4. Frese, M., Gloy, G., Oberprieler, R.G., and Gore, D.B.  2017.  Imaging of Jurassic fossils from the Talbragar Fish Bed using fluorescence, photoluminescence, and elemental and mineralogical mapping. PLoS ONE 12 (6): e0179029. https://doi.org/10.1371/journal.pone.0179029
Website developed by S8080 Digital Media