SEM Scanning electron micrographs of Tasmanian tiger hair

Introduction

In 1950, Lyne & McMahon published a description of Tasmanian tiger hair surface morphology. They identified three types of hair in the thylacine coat, in contrast to all other species in the family Dasyuridae occurring in Tasmania, which have two. The three hair types were described further by Taylor in 1985 who published a photograph showing gross morphological differences. Taylor referred to these, in decreasing length, as the over hair (O), guard hair (G) and under hair (U). Lyne & McMahon provided illustrations of the hair surface scale patterns but, despite noting the thylacine has three types of hair in its coat, referred to only two types of hair in labelling these illustrations: straight and wavy.

Taylor's photograph of the three hair types gives some indication of the shape of each type of hair. The table below gives a basic description of the gross morphology of each hair type identified by Taylor.

In 2017 Rehberg (author), in collaboration with CSIRO, produced a series of scanning electron micrographs of thylacine hair, this being the first time such high-resolution imagery of thylacine hair is being published.

The hair sample was acquired through purchase from a dealer in antique microscope equipment and specimens who advised it was found in a labelled paper envelope in a cabinet of specimen slides and materials purchased through auction in the north of England. In the dealer's opinion, the handwriting on the envelope dated the specimen to the 1850s, while other material in the cabinet had written dates in the 1880s. As such, this sample is likely to date between 1850 and 1890. Given the full provenance of the sample is unavailable, there remained the possibility that the hair was not thylacine, through mislabelling or other circumstance. The examination by Rehberg and Colin Veitch of CSIRO Manufacturing, Waurn Ponds, using scanning electron microscopy revealed that all three hair types described by Taylor were present and that surface scale patterns matched those illustrated by Lyne & McMahon. Together with a cursory examination of the medullary structure under a light microscope (not presented here), these results confirm that the hairs are from a thylacine, as described on the paper envelope in which the sample was found.

The full set of scanning electron micrographs produced during this examination is presented here.

Procedure

On opening the envelope containing the hair sample, visual examination revealed that some fibres were longer and straighter than others, which were shorter, wavy and thinner. It was soon apparent that the three hair types described by Taylor were present. One of each hair type was removed from the sample via tweezers and mounted onto conductive carbon tape which itself was first adhered to an aluminium specimen base (called a stub). Due to the limited quantity of sample material available, a decision was made not to wash or clean the fibres before examination. As such, dirt particles can be readily seen adhered to the fibres in all of the imagery. For this initial examination, no sample coating was undertaken. (See Australian Museum for an overview of scanning electron microscopy.) Imaging was done in an Hitachi S4300 SE/N Scanning Electron Microscope (SEM). The samples were imaged at 50 Pa pressure. Beam energy was 20 kV and the working distance was nominally 6 mm.

The specimen base was inserted into the SEM and a vacuum created. The hair types were examined in the order: guard; under; then over hair. During examination of the guard hair, some electron charge began to accumulate within the SEM, interfering with the imaging and a decision was made to continue the examination without vacuum. As a result, the images captured  from this point onward have a different appearance due to variance in contrast and brightness, but show no less detail.

Scanning electron microscopy uses a stream of electrons to illuminate the sample - and not visible light. As a result, the images produced are greyscale; not colour. For each of the hair types, images were captured depicting:

• general scale pattern at a mid-point of the fibre
• terminations of the fibre - often showing cut, broken or damaged ends
• close up profile (at the visible edge of the fibre) of scale joins (showing one scale raised above the next)
• extremely close up view of the join between two scales

Samples

The three samples used during scanning electron micrography are shown above, on green backgrounds. Taylor's samples are reproduced below each of the present samples.

Taylor's photograph used a 10mm scale, while the photographs in this study use a 5mm scale. Images have been scaled such that all fibres shown are to the same scale. The difference in over hair length is attributable to the present sample being broken (see image below, file ab). Difference in length for the under hair sample may be attributable to broken ends (not verified), minor variation during scaling of images or original placement of scales, natural variance within the species or between individual thylacines of different size, or natural growth rate of individual hairs.

All features described in the introduction above, for each hair type, are seen in the samples used for scanning electron micrography.

To produce the above images, the fibres were placed on a neutral background and imaged using a Wild Makroskop M420 stereo microscope with the images captured using a Luminoptic 16 Mpixel CMOS CCD camera. Two images are composites due to length of the fibres and all images have been compressed to reduce file size.

Images

Below are the first published scanning electron micrographs of Tasmanian tiger hair. For each image, two scans were performed and the second of each is presented here.

Note: images are labelled with a magnification (such as "Scanned at 1,000x"). Actual magnification will vary depending on your device display. Images also present a scale (such as "50um", meaning 50 microns, or micrometres) - above this measure is a series of 11 squares. The measure is the distance from the first to the last square. The space between two adjacent squares is one tenth of the measure (ie. if the label shows "50um" then the distance between adjacent squares is 5 microns).

Images can be uniquely referenced by the file name (eg. 'file h').

All images are copyright.

Guard hair - mid region - scanned at 1,000x (file h)

This is the first scanning electron micrograph of Tasmanian tiger hair. The scale pattern is visible as lighter lines traversing the fibre. Dirt particles are seen adhering to the fibre. The background texture derives from the specimen mount.

Guard hair - mid region - scanned at 1,000x (file i)

The scale pattern is visible as lighter lines traversing the fibre. Dirt particles are seen adhering to the fibre. The background texture derives from the specimen mount.

Guard hair - distal end - scanned at 3,500x (file j)

The fibre end appears to be broken. Scales are seen "peeling away" from the shaft of hair, in a process called fibrillation. Some dirt particles are seen adhering to the fibre. The background texture derives from the specimen mount.

Guard hair - proximal end - scanned at 600x (file k)

The fibre end appears to be cut or broken. The cuticle (ie. outer layer) has chipped away at this point, revealing the structure of the hair's cortex. The following file examines this region in detail. The scale pattern is visible as lighter lines traversing the fibre. Dirt particles are seen adhering to the fibre. The background texture derives from the specimen mount.

Guard hair - proximal end - scanned at 4,000x (file l)

This image examines the chipped cuticle observed in the prior file. The structure of the cortex is visible, both in longitudinal perspective and cross section. Some of the scale pattern is observed as the lighter line traversing the cuticle.

Guard hair - mid region scale detail - scanned at 3,500x (file n)

This image was captured after file m, but is presented before file m, as it provides context for the detail presented in file m. In this image, the edge of the hair fibre is visible. Scale pattern depth is visible as one scale overlaps the next. In mammalian hair the overlapping scale is closer to the root end of the fibre; thus the root (proximal) end of this hair is identified as being to the left while the distal end is to the right. The bright white streak across the lower portion of the image is due to charge buildup within the SEM. Dirt particles are clearly visible, adhered to the fibre. The patchy background texture in the upper half of the image derives from the specimen mount. The scale overlap and dirt particles seen at centre of this image are enlarged further in the following image (file m).

Guard hair - mid region scale detail - scanned at 20,000x (file m)

An extremely close up view of a scale overlap seen in the prior image (file n). Note the scale masure of this image - 2 microns. The width of this image is approximately 6 microns, which is 6 millionths of a metre, or 6 thousandths of a millimetre. The upper scale is at left. Dirt particles are clearly visible on the lower scale in the upper portion of the image and also on the upper scale. A repetitive curved texture is visible in the join between the two scales, in the lower half of the image (where it is not obscured by dirt build up). This is believed to be an oil or wax residue that is a natural component of hair.

Recalling that file n (presented above but taken after this image) began to exhibit a flare due to electron charge, the subsequent images were obtained utilising a slightly different technique which is explained further, below (under file o).

Artistically, micrographs at this scale begin to approach the appearance of extraterrestrial landscapes! This is the highest magnification micrograph of thylacine hair published.

Guard hair - mid region scale detail - scanned at 3,500x (file o)

During the prior capture (file n, two images before this one on this page), a flare developed that created a white streak across the lower half of the image. This was due to an electron charge build up within the vacuum containing the specimen. A decision was made to fill the chamber with air in order to flush the electron charge. All subsequent images were taken with air in the chamber which leads to images of a different appearance, as seen here and below.

This image shows the edge of the hair fibre, similarly to file n and was captured at the same magnification. File n provides a good comparison for the differing effect due to the two techniques (in vacuum - file n; and without vacuum - file o). The overlapping scale pattern is again visible, as are dirt particles adhered to the hair fibre.

Guard hair - mid region - scanned at 1,000x (file p)

The scale pattern is visible as lighter lines traversing the hair fibre. The bands of scales are themselves considered narrow; that is, the distance from one scale overlap to the next is small as you travel the length of the hair fibre. Each scale generally spans the full width of the hair. This is in contrast to the over hair, which will be seen below (although the under hair is presented next, commencing with file q). Dirt particles are visible, adhering to the hair fibre.

Under hair - mid region - scanned at 1,000x (file q)

The under hair is the thinnest of the three hair types. The scale pattern is visible as light lines traversing the hair fibre. The distance between scale lines, with respect to the width of the hair, is greater than that seen in the guard hair (shown above). However, in common with the guard hair, each scale generally spans the full width of the hair. Dirt particles are seen adhering to the hair fibre.

Under hair - mid region - scanned at 2,500x (file r)

This image presents a closer view of the mid section of the under hair. The overlapping of the scales is visible along the upper edge of the hair fibre. In this case the upper scale is on the right, indicating the root of the hair is to the right. Dirt particles are seen adhering to the hair.

Under hair - mid region scale detail - scanned at 7,000x (file s)

This image shows scale overlap in detail. The hair is oriented from top-right corner down and left to the centre of the bottom edge. The change in orientation relative to the last image (file r) is due to the waviness of the fibre. The upper scale occupies the top-right corner of the image indicating the root of the hair is to the right (as per previous image). Dirt particles are visible adhered to the fibre, predominantly below the scale overlap.

Under hair - mid region - scanned at 2,000x (file t)

The scale pattern is visible in this image. The root of the hair is toward the upper-right. A considerable amount of dirt is visible adhered to the hair fibre.

Under hair - mid region - scanned at 1,800x (file u)

The scale overlaps are visible in this image. Through much of this region the overlaps are perpendicular to the fibre edge but there are also notable sections that are curved and at a diagonal to the fibre edge. Dirt particles are visible adhered to the fibre.

Over and under hair - size comparison - scanned at 100x (file v)

Few morphological details are visible in this image but the image serves to give a size comparison between over and under hair diameters. The over hair is above; the under hair is below. The scale pattern is faintly observed on the over hair and this is explored further in remaining images. The curviature of the under hair is clearly visible at low magnification. No further images of the under hair ends were obtained due to the fact that on mounting the specimen, the hair ends were left "free floating", away from the carbon tape adhesive. This results in microscopic movement of the hair end. During the scanning process, this movement blurs the resulting image, especially at higher magnifications.

Over hair - distal end - capture of scan at 700x (file w)

This image is not a high resolution scan. Rather, the default mode of operation of the scanning electron microscope is to continually update the monitor in real time, but at low resolution. Each high resolution scan takes seconds to complete, while low resolution scans are used to obtain near-real time updates of the monitor whilst examining the specimen. This image is a capture of the real-time monitor. The reason for this approach is that the end of the over hair fibre was "free floating" away from the carbon tape adhesive (similarly as described for the under hair in the prior image, file v).

The end of this hair fibre is frayed - essentially this Tasmanian tiger had split ends. Note also the diameter of the fibre is considerably thinner than the mid section shown in the previous image (file v). The background texture derives from the specimen mount.

Over hair - mid region - scanned at 1,200x (file x)

The scale pattern is observed as lighter lines traversing the hair fibre. The pattern revealed in this image of the over hair is markedly different than that observed on the guard and under hairs. Scale lines are more jagged, less parallel and not as perpendicular to the fibre edge. Some dirt particles are visible adhered to the fibre.

Over hair - mid region - scanned at 1,800x (file y)

Scale pattern is visible as pale lines traversing the hair fibre. Some dirt particles are visible adhered to the fibre.

Over hair - mid region - scanned at 1,200x (file z)

The scale pattern is visible as pale lines traversing the hair fibre. The horizontal white lines are artefacts of the scanner during this capture. The apparent shadow of the hair is likely due to the hair compressing against the carbon tape adhesive during placement of the fibre on the mount. Dirt particles are visible adhered to the fibre.

Over hair - proximal end - scanned at 800x (file aa)

The proximal end of the hair fibre is visible at left edge of the image and appears broken irregularly. Along the top edge of the hair can be seen the scale pattern, with the upper scale lying to the left of the lower scale leading to the identification of the left end of the fibre being at the root of the hair. As this hair is broken, this is not the hair root. Dirt particles are visible adhering to the fibre.

Over hair - scale detail - scanned at 7,000x (file ab)

This image shows detail of one scale overlapping the next. The upper scale is at left while the lower scale is at right, indicating the root end of the hair is to the left. Dirt particles are visible adhered to the fibre.

Further research and acknowledgements

This scanning work was carried out by CSIRO under a generous offer of support in response to ABC News coverage of Where Light Meets Dark's crowd funding campaign to produce high quality light microscopy images of Tasmanian tiger hair. The author would like to acknowledge the generosity of Colin Veitch, CSIRO Manufacturing - Waurn Ponds Laboratory and the coverage of ABC.

The stated goal of the crowd funding project is to produce light microscopy images of Tasmanian tiger hair that may become a reference base against which researchers may compare samples collected in the field for the purpose of a potential thylacine match. Although the scanning electron microscopy has produced unprecedented detail and magnification for thylacine hair, it is limited to surface morphology only. Light microscopy can reveal internal features of hair fibres, including the medulla (core) and this is also diagnostic in identifying unknown hair fibres. Further, SEM resources are expensive and not necessarily accessible to most researchers (particularly field researchers in search of the thylacine). Light microscopy is affordable at a much more local scale and so there is still a need to complete the light microscopy work before a useful catalogue of images is available for comparative analysis.

The author would also like to acknowledge the many backers of the light microscopy project, and the project's sponsor, Dr Stephen Sleightholme, without whom there would not have been news coverage and the generous offer from CSIRO for SEM.

Citing this article

Rehberg, C. (2017). SEM Scanning electron micrographs of Tasmanian tiger hair, Accessed: date, http://www.wherelightmeetsdark.com.au.

References

• Lyne, AG and McMahon, TS (1950) Observations on the surface structure of the hairs of Tasmanian monotremes and marsupials. Papers and Proceedings of the Royal Society of Tasmania. pp. 71-84. ISSN 0080-4703
• Taylor, Robert J (1985) Identification of the hair of Tasmanian mammals. Papers and Proceedings of the Royal Society of Tasmania 119, 1985. pp.69-82. 0080-4703