Examination and Analysis: Feathers and Fur
© AMNH
Examining a feather object or bird specimen involves investigation not only of its form and construction, but also its history of use and condition. Conducting background research and talking to source communities, curators, and collection managers can help us better understand the object or identify species. Those conversations, coupled with examination for evidence in and on the object, can also help us determine how the object was used. Looking for signs of deterioration on the object leads to an assessment of condition. Together this understanding informs decisions about treatment, rehousing and storage/display conditions.
Conservators do not rely solely on eyesight and normal illumination under the fluorescent lights of a conservation lab. Additional techniques can be used to understand feathers and fur in a way that is not possible using the human eye alone, to answer research questions, and to inform conservation judgements and decisions.
Learn more about how these techniques are used in the examination of fur and feather objects:
Fourier-Transform Infrared Spectroscopy (FTIR)
Multiband Imaging for Fur and Feathers
Digital imaging is an essential technique for graphic representation in conservation, and is readily used to create detailed, accurate, and durable records describing objects and specimens composed of fur and feathers.
Multi-band imaging (MBI) extends the capabilities of digital imaging, enabling the visualization of surface information beyond the visible region of the electromagnetic spectrum, including reflected ultraviolet (UVR), ultraviolet-induced visible fluorescence (UVF), visible light (VIS), and often infrared (IR). This offers the ability to observe material differences and monitor changes over time that may otherwise be difficult to detect under normal lighting conditions.
Multi-band imaging can be used to visualize and capture degradation in keratin, as well as changes in certain associated materials like preening oils and biopigments. Unpigmented keratin in good condition has a low blue-white fluorescence. Previous study has shown that fluorescence in feather keratin increases in brightness with exposure to light and oxygen (photo-oxidation). Change can be observed using ultraviolet before it can be detected in the visible range. Conservators can use MBI images to document and track this chemical change.
Some natural materials such as the preening oils and biopigments found in or on bird feathers fluoresce under ultraviolet (UV) radiation. Examination using UV aids in their detection and identification, which may in turn facilitate better decision-making about treatment. For example, UVF can be used to confirm or exclude the presence of certain fluorescent pigments, like porphyrins, a group of fugitive biopigments present in some owls, grouse, and pigeon species.
Ultraviolet reflectance (UVR) images, which capture wavelengths below the visible range (<400nm), are rendered in black and white to produce a visible image with high contrast that helps to show small structural features and fine texture, such as the topography on a feather’s surface. Physical details which are otherwise difficult to distinguish are sometimes more readily apparent in the UVR image.
These applications of multi-band imaging have been useful to us in research conducted to better understand the impacts of treatment on feather preservation, and we have relied in MBI to build qualitative visual data sets from much of our experimental work.
Our investigation of the impacts of different cleaning techniques on feathers is one example. The study was underpinned by a series of carefully controlled cleaning tests conducted on feathers of different types and condition states. Using MBI, we captured images before and after each cleaning test using UVR, UVF, and VISlighting conditions. Our workflows were designed to minimize camera and filter handling, so that each image set is perfectly aligned for easy correlation of features detected using different wavelengths. Standardizing our capture and color processing made it easy to compare image sets taken before and after cleaning with one another, or with image sets from another cleaning technique. This standardization meant that even subtle changes in feather structure or surface, such as individual barb losses or partial removal of a localized preen oil deposit, could be observed and documented.
Residual dirt left on some feathers after cleaning was readily visible in UVR images as a dark deposit on the white feather. This made its detection much easier, particularly when feather and dirt were similar in color under normal viewing conditions. We used the same approach (in conjunction with UVF) to look for particulate residues left behind by dry cleaning materials like sponges and dusting cloths, as well as wet cleaning solutions that rely on detergents.
Digital Microscopy for Fur and Feathers
Microscopy is a powerful tool that conservators often rely on to identify the animal origins of feathers and fur, and to inform a nuanced assessment of a specimen’s construction and condition.
Digital photomicroscopy couples a microscope with a digital camera, providing the ability to capture what you see through the optics of the microscope in still and/or video formats. Many higher end digital microscopes integrate Z-stacking capabilities that are beneficial in examining and imaging the textured and topographical surfaces of fur and feathers under magnification. Microscopy is best used in combination with other techniques that help to contextualize information gathered at different scales.
With sufficient magnification (100x – 400x), characteristic anatomical features in fur and feathers may be observed and used to make a general or exact species identification. In fur, such features include the shape and size of scales on the fiber cuticle; the size and character of the medulla (a central open network present in some animal fibers); and the average diameter of the fiber overall. Diagnostic features can be also observed in feathers. Villi, fine protrusions at the barbule base, may be present in some species and absent in others; and nodes, minute features at the junction between adjacent cells on the plumulaceous barbules of certain feathers, are characteristic in their shape, pigment location, and pigment distribution. These clues can be used in conjunction with reference libraries to narrow down the animal from which the fiber or feather originated.
Microscopy also aids in identifying coloration mechanisms. Both non-iridescent and iridescent structural colors found in feathers are resilient to color fading because they derive from optical phenomena rather than light-sensitive biopigments. To determine whether structural color is present, a feather can be viewed in transmitted light, neutralizing the structural colors and leaving only the biopigmentation visible. In reflected light, one can also clearly see whether coloration is located in the feather barb (non-iridescent structural color), the barbule (iridescent structural color), or is dispersed in both (biopigmentation).
Minute anatomical features like feather barbules and hooklets are too small to see clearly without magnification, but they impact how the feather behaves on the macroscopic level. Is the vane uniform in its opacity, or does it have thin, translucent areas? Does the vane “zip” together into a cohesive sheet, or do the barbs splay apart? Translucency and unzipping occur when barbules in an area are broken off or clumped in a way that prevents the barbs from properly engaging with one another. Understanding structure on the microscopic level supports a more informed assessment of condition issues visible to the naked eye, and makes for better informed judgements.
We have relied heavily on digital microscopy in our research to better understand the impacts of treatment on feather preservation, providing a record of the minute changes that occur in feathers when they are cleaned in different ways. This technique has allowed us to observe and easily capture damage to small structural components, even when no change was visible to the naked eye.
Monitoring change on this level may require one to reliably locate the same small area of the feather before and after each cleaning test. To do this, we developed a simple jig to hold the feather on the microscope stage. The jig consisted of two flexible magnetic sheets, stuck to one another and hinged at the top. A small window cut through both layers defined the viewing area, and markings were added above and below the window to register with the distal tip of the feather’s central rachis. The jig was placed on the microscope stage and the objective aligned with the window cut in the magnetic holder.
The window in our jig was large enough to reveal a narrow zone extending halfway across the feather, from the rachis to the outside edge of the vane. That area was much larger than our microscope’s viewing area, so our capture method relied on automated image stitching to produce a single photomicrograph of the whole region. This approach allowed us to observe differences in the distribution of minute damages across the feather. For example, in our tests, some cleaning techniques caused barbules to clump only near the feather rachis where they are more robust and well-engaged. Other techniques caused barbule damage only near the edge of the vane, where structures are more fragile and loosely engaged. These localized differences in behavior were not something that we had anticipated.
FTIR Spectroscopy for Fur and Feathers
Fourier-Transform Infrared Spectroscopy (FTIR) measures the intensity of absorbed or transmitted infrared (IR) radiation as it passes through a sample. Peaks in the resulting FTIR spectrum represent chemical groups present within the sample, and collectively are characteristic for any given material. In many materials, aging and degradation are accompanied by characteristic changes in the relative intensity or position of peaks. By measuring these changes, FTIR may be used to monitor deterioration.
As keratin degrades, the protein undergoes chemical changes in its primary and secondary structure. One of those changes is the oxidation of the amino acid cystine to form cysteic acid, which correlates with loss of strength and breakage of the disulphide bonds that stabilize its helical form. Because both cystine and cysteic acid are represented by peaks in keratin’s IR spectrum, infrared spectroscopy can be used to track degradation over time by looking at relative change in these markers. Their ratio to one another can be compared before and after an event (such as a chemical treatment, light exposure, etc.) to reveal how that event has impacted the chemistry of the keratin.
It is always important when using a new material in conservation, or an old material in a new context, to have as complete an understanding as possible of that material’s potential long-term impact (negative or positive) on the object. Our research into both fur and feathers has focused on how particular treatment methodologies (i.e. cleaning methods, pesticides, and colorants) impact deterioration processes. To do this, we have coupled FTIR analysis with real-time and accelerated aging to research how markers of photo-oxidation change with aging in treated and untreated samples.
Studies probing the long-term consequences of treatment on highly variable materials like keratin can present interesting challenges in the course of experimental design. Lightweight samples need to be secured within the aging chamber where there is high air flow, and stabilized so that consistent measurement locations can be defined. Even simply collecting a high-quality spectrum from fur and feather samples can present issues. Our Bruker Alpha FTIR-ATR unit requires that samples have good contact when pressed against the small diamond crystal plate used in measurement. For fur, that means layering fibers to completely cover the crystal. For feathers, it means taking care that barbs are fully engaged and carefully positioned to stay within the measurement area when pressed. Poor contact will result in a noisy spectrum, and lower quality data.
Another complicating factor may be the management of competing spectral contributions from different materials present in a single sample. In our research considering the use of metal-complex solvent dyes for recoloring faded taxidermy, we investigated whether dyes applied to fur accelerate or slow normal degradation behaviors in keratin. When the dyed fur sample was analyzed, the dye spectrum overlapped with and obscured the cystine and cysteic acid markers of interest in the keratin. Capturing useful data required us to find a means to resolve this.
We developed a novel sample preparation method for fur to resolve both problems:
- Fur cut from a hide was carded and needle felted, making a uniformly heterogeneous arrangement of fibers, and immobilizing them in a sheet of consistent thickness.
- The felt sheet was cut into smaller samples that were sewn onto a wool flannel substrate to provide further stability, easier handling, a place to write the sample number, and a means to stitch lines of thread into each sample to define dyed and undyed (control) areas.
- For each sample, dye was airbrushed onto one area while another was left undyed, and the sample was given six weeks of accelerated light aging.
- After aging, the dye was washed from the felt so that it would not interfere with analysis of the keratin.
- Samples were successfully analyzed with FTIR by laying each felt sample across the sampling area for measurement.
Spectrophotometry for Fur and Feathers
Spectrophotometry measures the amount of light reflected or transmitted by a material at individual wavelengths of the spectrum. Using algorithmic conversions, the measured reflectance spectrum can be translated into a scientific description of color, or colorimetry. This non-invasive analytical technique allows the conservator to describe color in a sample quantitatively, which in turn enables us to monitor color change, such as fading from light exposure, over time.
The complex structure and topography of fur and feathers makes them difficult candidates for measurement with typical fiber optic reflectance spectrophotometry probes, which were designed to measure the color of flat surfaces. Structural color in feathers, which changes with orientation and viewing angle, is particularly difficult to measure. While some ornithologists have successfully adapted this equipment to measure feather color in the field and lab, it requires one to be meticulously consistent in the orientation of the sample relative to the probe, and even then, error can be high enough to overshadow change.
In place of a fiber optic probe, an integrating sphere streamlines capture and improves the repeatability of measurements from bulk fur and feather samples. The integrating sphere provides a perfectly diffuse interior chamber where light reflected from the sample is equally distributed by multiple scattering reflections. Through diffusion, the light’s intensity becomes uniform before it enters the detector, minimizing the impact of directionality in the sample.
When collecting measurements from naturally variable materials over time, it may be important to precisely locate the same measurement site. For large samples like whole feathers, Mylar templates can help to clearly define those locations. The sample perimeter can be traced onto the Mylar, and a hole cut to fit the instrument’s aperture, so that when the template is laid onto the sample, the measurement location is consistent and clear. We have used this approach in our long term, real-time aging study of how pesticides impact feathers, where a year may pass between successive measurements. Mylar templates reduce our reliance on memory or instructional workflows that may be difficult to interpret.
Samples that tend to change shape or come apart require a different solution. To test the lightfastness of dyes used to restore faded fur, we developed a method for immobilizing groups of fibers so that they could be airbrushed with dye, exposed to high airflow in our accelerated aging chamber, and analyzed multiple times by spectrophotometry. The technique relies on a two-part sample holder typically used for XRF analysis (Chemplex SpectroCertified Quality XRF Sample Cup No 3115), and not unlike an embroidery hoop in that it is designed to secure material across a circular area through the engagement of larger and smaller cylinders. White Teflon was laid across the open top of the sample holder’s smaller inner circle component as an initial support layer. Fibers were laid on top of the Teflon sheet, brushed and aligned to create a uniform layer with no gaps. Then the larger outer circle component was placed over the smaller one to secure fibers and Teflon in place as a flat sheet. After trimming away the excess, fibers were effectively stabilized during handling, aging, and measurement.
Accelerated Aging for Fur and Feathers
Analytical techniques like FTIR and spectrophotometry may be coupled with accelerated aging to develop a general understanding of how keratin materials will change chemically or aesthetically as they age over time. Accelerated light aging rapidly reproduces the damage in keratin that is caused by light in real environments over longer periods. Aging is conducted inside of a chamber containing high-powered xenon arc lamps. Light output, as well as temperature and humidity, can be set, and filters can be used to adjust the incident spectrum to include or exclude UV.
Accelerated light aging is a very effective means of inducing photo-oxidation in keratin. As aging takes place, a number of physical changes may be observed, including yellowing and/or bleaching of the keratin; increase in ultraviolet-induced fluorescence; fading of organic biopigments, particularly carotinoids, which are not as light stable as melanins; increased sensitivity to water; brittleness and loss of strength. In feathers, loss of strength may initially be associated with breakage and loss of distal barbules. However eventually in both feathers and fur, continued light aging weakens keratin to the point that fibers and feather barbs will simply shatter when handled.
The rate at which accelerated aging takes place is determined by the set points (light output, temperature, humidity) and lighting condition (UV-included vs. UV-excluded) selected. Associated damage can be tracked with complimentary techniques that record or measure those changes directly.
Accelerated aging has been used at the AMNH to answer questions about how keratin interacts with other materials over the long term, as well as the longevity of treatment methodologies. Our research evaluating metal-complex solvent dyes for recoloring faded historic taxidermy is a good example. This effort was conducted in several phases, each coupling accelerated aging with an analytical technique that measured change directly.
In initial phases of the project we established the lightfastness of each dye color, first testing it alone on an inert quartz plate, and later applied to white fur. In both cases, testing relied on spectrophotometry to measure change in color during and after accelerated light aging. Results were benchmarked against a well-understood reference, the ISO Blue Wool Standard, as well as samples aging in real time in our own American Bison diorama. A comparison of the two datasets showed no evidence of dye-keratin interaction impacting the dyes’ lightfastness.
The project also investigated whether chemical degradation behaviors in keratin are accelerated or slowed when fur is dyed with metal-complex solvent dyes. In this case, testing relied on FTIR to measure relative change in chemical markers of photo-oxidation during and after accelerated light aging. In comparing dyed samples to undyed controls, it was evident that the dyes have minimal impact on the rate of photo-oxidation in keratin, and if anything, act as a light filter, slowing degradation rather than accelerating it.