Microstructure and mechanical properties of different keratinous horns

Abstract

Animal horns play an important role during intraspecific combat. This work investigates the microstructure and mechanical properties of horns from four representative ruminant species: the bighorn sheep (Ovis canadensis), domestic sheep (Ovis aries), mountain goat (Oreamnos americanus) and pronghorn (Antilocapra americana), aiming to understand the relation between evolved microstructures and mechanical properties. Microstructural similarity is found where disc-shaped keratin cells attach edge-to-edge along the growth direction of the horn core (longitudinal direction) forming a lamella; multiple lamellae are layered face to face along the impact direction (radial direction, perpendicular to horn core growth direction), forming a wavy pattern surrounding a common feature, the tubules. Differences among species include the number and shape of the tubules, the orientation of aligned lamellae and the shape of keratin cells. Water absorption tests reveal that the pronghorn horn has the largest water-absorbing ability due to the presence of nanopores in the keratin cells. The loading direction (compressive and tensile) and level of hydration vary among the horns from different species. The differences in mechanical properties among species may relate to their different fighting behaviours: high stiffness and strength in mountain goat to support the forces during stabbing; high tensile strength in pronghorn for interlocked pulling; impact energy absorption properties in domestic and bighorn sheep to protect the skull during butting. These design rules based on evolutionary modifications among species can be applied in synthetic materials to meet different mechanical requirements.

1. Introduction

Animal horns are mainly found in the Bovidae family (e.g. cattle, sheep, goats, buffalo, antelope) and Antilocapridae family (e.g. pronghorn), and serve as attack and defence weapons during intraspecific combat [1,2]. These weapons have diverse forms and shapes arising over millions of years of evolution [3], while the fundamental material component remains consistent, which is α-keratin, a structural protein synthesized by keratinocytes [4,5]. Keratin, typically mineralized to less than 1% (e.g. 0.05 wt.% in ox and rhino horns) [6], is one of the most common biopolymers found in nature, showing remarkable mechanical efficiency because of its relatively high fracture toughness and Young's modulus, but is much lighter than other highly mineralized biological materials, such as bone and teeth [7–9]. Keratinous materials are formed by laminated keratin cell layers, with crystalline keratin intermediate filaments (IFs) embedded in an amorphous keratin matrix as the main constituent of the cells [4,5]. The crystalline IFs can be classified into two categories: α-helix and β-sheet. Wool, hair, hooves and horns of mammals are composed of α-keratin, while bird feathers, reptilian claws and beaks are all made of β-keratin. In terms of the process of biosynthesis and amount of disulfide bonds formed from the presence of cystine, keratinous materials can be classified into soft or hard keratin. Soft keratins such as stratum corneum contain a lower amount of sulfur, while hard keratins such as wool and hair show more sulfur cross-links leading to higher hardness and durability [10,11]. These keratinous materials serve particular functions such as protection (e.g. waterproof feathers), defence (e.g. horns) and predation (e.g. claws) [12–15].

Mammal horns contain a bony core and an exterior keratin sheath. The sheath develops slowly and attains a definitive size and shape by growing from the epidermal layer surrounding the bone core from the frontal region of the skull [16,17]. Given the lifespan of mammals and the inability of most horns to regrow (the pronghorn is an exception), all horns must be resistant to fracture [2]. As the main weapons employed in fights, the horns should meet specific mechanical requirements (impact resistance, energy absorption, high compressive and tensile strength) to accommodate the particular fighting behaviour of their species [18,19]. Thus, the study of hierarchical structure and mechanical properties of keratinous horns can provide principles and design rules for making bioinspired materials with superior mechanical properties and durability.

The hierarchical structure and mechanical properties of bighorn sheep horn have been previously investigated by Horstemeyer and colleagues [20,21]. Prior studies [22,23] showed that tubular and lamellar structures exist in the bighorn sheep horn. Flattened and cornified epithelial cells (keratinized cells) grow from dermal papilla that are arranged in concentric layers around a central medulla, forming the tubular structures [24–26]. Tubular structures in other keratinized tissues such as equine hooves and rhinoceros horns were identified and characterized in previous studies [24,27,28]. The tubular structures in hooves were identified to be continuous hollow medulla cavities, from its origin at the coronet to the ground surface [27,29]. No clear evidence has shown that the tubular structures could benefit the hydration and thermal insulation properties of hoof [30]. However, it is well established that tubules can redirect crack propagation, thus increasing the fracture toughness of biological materials [28,31,32]. In bighorn sheep horns, the tubules are hollow and elliptically shaped (in cross section) with a major axis approximately 100 µm and minor axis approximately 40 µm. Surrounding the tubules are concentrically and tightly packed keratin cells (approx. 20 µm in diameter, approx. 2 µm in thickness) [23]. In hooves, the keratin cells firmly cement together preventing the passage of water inwards and the loss of body fluids outwards by forming a tough protective barrier [29].

Three orthogonal directions were defined in previous works to show the orientation-dependent mechanical properties: the longitudinal direction is parallel to the tubule direction (growth direction), the radial direction is the impact direction and the transverse direction is perpendicular to both [22]. The compression and tension properties of bighorn sheep horn in different orientations and hydration states were also studied. Anisotropic compression behaviour in the different orientations was observed: the Young's modulus and strength were smaller in radial direction, but the energy absorption was the largest [22]. Hydration significantly decreased both the Young's modulus and compressive strength in different loading states and strain rates [20,21]. It was reported that water molecules could affect keratin matrix in three ways: water molecules act as a swelling agent and extend the space between molecular chains, thus reducing the interaction between chains; water molecules may replace the secondary hydrogen bonding, thereby increasing the chain mobility; water molecules and matrix proteins bind to form a network that acts as a plasticizer, thus the stiffness is reduced and mobility increased [5].

The mechanical properties of other horns have also been examined in previous studies. Experiments revealed that along the length of fresh waterbuck horns, the work of fracture (calculated as the energy differences in loading and unloading curves divided by the area of material fracture in extending the length of the crack [33]) ranged from 10 to 80 kJ m⁻² [18]. The specific work of fracture (normalized by density) of the horn is 32 kJ m⁻², which is greater than most other biological and synthetic materials (antler 6.6 kJ m⁻², bovine femur 1.6 kJ m⁻², glass 5 kJ m⁻², mild steel approx. 26 kJ m⁻²) [34,35]. Three-point bending tests of cattle horn show that the mechanical properties are position dependent. From the distal to the proximal regions, the stiffness and strength decreased. This is likely due to the age of the horn. The distal region is the oldest, while the proximal region is the youngest. This difference may help during fights, because the distal part needs to be hard enough to stab the opponent, whereas the proximal part should be tough to absorb energy. The gradient of mechanical properties could also be related to water content and the proportions of crystalline and amorphous phases in keratin [36].

The horn shapes of different species have been hypothesized to fit their fighting behaviour [2,3]. It is worth noting that the microstructures and mechanical properties of horns can be related to their fighting behaviours. Fighting between bighorn sheep (Ovis canadensis) entails extremely high energy impact on the horns [37,38], while these collisions in domestic sheep (Ovis aries) are much more moderate with a lower impact energy [39–41]. The mountain goat (Oreamnos americanus) uses a different fighting style, stabbing. The opponents initially whirl in anti-parallel position relative to each other, and then stab using the horn tips. The fights between goats can, in some cases, cause severe injuries [42,43]. The pronghorn (Antilocapra americana) uses two fighting modes. One is the sudden clashing together of two bucks; in the other, two bucks gradually approach each other, lower their heads and then slowly engage horns. Fights involve thrusts and counterthrusts with the horns. When the horns interlock, pronghorns use their body strength to twist the opponent's neck and push to make their rival lose balance. Unlike the other species of horned ruminants (family Bovidae), the keratinous horn sheath of the pronghorn (family Antilocapridae) sheds and then regrows every year. Unlike the unbranched horny sheath of bovids, the pronghorn has two branches [44,45].

For the present study, the above four representative species were selected (bighorn sheep, domestic sheep, mountain goat and pronghorn). Both bighorn and domestic sheep are in the same genus (Ovis), and the mountain goat is a close relative, but belongs to a different genus (Oreamnos) in the family Bovidae. The pronghorn is in the family Antilocapridae that evolved keratinous horns independently of bovids. The phylogenetic tree and divergence times of these four species are shown in figure 1 [46,47]. The species selected in this work sample the diversity of microstructures and mechanical behaviours of various horns. This study aims to understand the following points: the similarities and differences of the microstructures and mechanical properties between horns from different ruminant artiodactyl families; the correlations between microstructures and mechanical properties in different species; and whether the microstructures that evolved in different species are optimized to accommodate specific fighting behaviours.

Figure 1.

Phylogenetic tree and estimated divergence times of four ruminant species with keratinized horns [46,47]: bighorn sheep (Ovis canadensis), domestic sheep (Ovis aries), mountain goat (Oreamnos americanus) and pronghorn (Antilocapra americana). The evolution of important traits is mapped onto branches of the tree. ‘Cranial appendages’ refers to the horns, ossicones and antlers of ruminant mammals; keratinous horn sheaths likely evolved independently in the families Bovidae and Antilocapridae [48]. Paintings are by C. Buell (Copyright © J. Gatesy). (Online version in colour.)

2. Experiments and methods

2.1. Materials preparation

Four pairs of horns (bighorn sheep, domestic sheep, mountain goat and pronghorn) were purchased from the Wilderness Trading Company (Pinedale, WY, USA). All horns were stored in air at room temperature. We did not have access to fresh horns, but it has been reported that there is a 20 wt.% water content in fresh oryx horns. It is impossible to obtain this value by partial hydration because the water would preferentially be absorbed in the tubules, as well as in the keratin matrix. Thus, the controllable conditions of ambient dried and fully hydrated were tested. Figure 2 shows the combat styles and horn morphologies of different horns: (i) bighorn sheep; (ii) domestic sheep; (iii) mountain goat; and (iv) pronghorn. The keratin sheaths and bony cores are indicated in the cross-section images in figure 2. Compression and tension samples were acquired from the midsection of the horns, indicated with yellow and red colours. We choose samples from representative areas that sustain the highest impact force and will most likely be damaged in these extreme conditions. For bighorn and domestic sheep horns, the main forces are impacts and compressions on the middle and outer area as shown in figure 2; thus we took samples for the impact and compression areas. For consistency, the tension samples were also obtained from the same area. For mountain goat, because it is very symmetric and the forces come from the tip, we took samples in the central part. Owing to the complicated structure of pronghorn, we took samples under the branch, because samples in this area supported most forces.

Figure 2.

Combat styles and horn morphologies for the different species: (a) bighorn sheep, (b) domestic sheep, (c) mountain goat and (d) pronghorn. Growth direction is indicated by a black arrow for each horn. Compression and tension samples taken from the keratin sheath are indicated with yellow blocks and red strips, respectively. Combat photographs of different species are taken from: theholepicture.photoshelter.com; gudmann.photoshelter.com; gettyimages.com and nationalgeographic.com. (Online version in colour.)

2.2. Structural characterization

The microstructures were characterized by optical microscopy and scanning electron microscopy (SEM). A Leica Ultracut UCT Ultramicrotome (Leica Microsystems, Wetzlar, Germany) was used to get fine-polished ultra-thin horn slices. Samples were obtained from the central region of the horn and approximately 5–10 mm away from superficial surfaces. Two samples of each horn were prepared with one cross section parallel and one cross section perpendicular to the longitudinal direction. The thickness of horn slices was 1–1.5 µm. Slices were stained by toluidine blue (a basic thiazine metachromatic dye with high affinity for acidic tissue components). Optical micrographs of stained horn slices were taken from a Zeiss AxioImager M1 optical microscope (Zeiss, Oberkochen, Germany). SEM images of freeze fracture surfaces in liquid nitrogen were acquired using an ultra high-resolution scanning electron microscope (FEI, Hillsboro, OR, USA).

2.3. Water absorption and Fourier-transform infrared spectroscopy

Water absorption tests were performed on two samples (6 × 6 × 6 mm) from each species. Samples were air-dried for 24 h before weighing. The samples were then soaked in purified water and weighed every 4 h. The water was changed every 24 h. After 96 h, the samples were taken out and the surface water was removed with a cloth. The samples were then placed into a pre-heated oven (110°C) for 5 days to fully dehydrate. During dehydration, the samples were weighed every 12 h. Fourier-transform infrared spectroscopy (FTIR, Perkin-Elmer, Waltham, MA, USA) was used to characterize the chemical compositions of horns in different species. Three samples with dimension 5 × 3 × 0.3 mm (length × width × thickness) were acquired from each species and characterized by FTIR.

2.4. Compression and tensile tests

Cubic samples with dimensions of 6 × 6 × 6 mm were prepared for compression tests. A section saw with circular diamond blade was used to cut the samples and obtain parallel surfaces. All samples were cut from the central region of the horn (shown in figure 2). For each species, three sets of samples were prepared with the loading direction along longitudinal, radial and transverse directions. For each species, a total of 30 samples were prepared; 15 were tested in ambient dried condition and 15 in fully rehydrated condition. For each condition, five samples each were tested in longitudinal, radial and transverse directions [22]. The rehydrated samples were pre-soaked in purified water 3 days before testing with the water changed daily and kept wet until testing. Compression test experiments were conducted on a universal testing machine equipped with a 50 kN load cell (EM Model 5869, Instron, MA, USA). The testing crosshead speed was controlled at 6 × 10⁻³ mm s⁻¹, which corresponds to a 1 × 10⁻³ s⁻¹ strain rate. The machine automatically stopped when strain reached 0.7.

Samples for tensile tests were cut from the central region of the horns. Rectangular samples with dimensions of 15 × 3 × 0.3 mm (length × width × thickness) were prepared using a water-jet cutting machine. Care was taken to cut the samples such that tubules were aligned parallel to the longitudinal direction. Then the two ends of each sample were attached to 800 grit sandpapers to ensure that slippage did not occur. These samples had a gage length of 10 mm. For each species, a total of seven samples were prepared. The samples were kept under ambient conditions for 1 day until testing. Tests were conducted on a universal testing machine equipped with a 500 N load cell (Instron 3342 Universal Testing Systems, Instron, MA, USA). The crosshead speed was controlled at 0.001 mm s⁻¹, which corresponds to a 1 × 10⁻³ s⁻¹ strain rate, same as for the compression tests. SEM images of the fracture surfaces after tension tests were acquired from a ultra-high-resolution scanning electron microscope (FEI, Hillsboro, OR, USA).

2.5. Impact tests

The ASTM standard D7136/D7136 M-07 [49] is applied for drop-weight impact tests of layered multidirectional polymer composite matrix, and the standard requires the test machine to have a rectangular free-standing (unclamped) area with the dimension of 125 × 75 mm (length × width). As it was difficult to get such large testing samples from the horns, a modification of the standard apparatus was used. A laboratory-built drop-weight test machine [50] was used with a 1 : 5 scale of the ASTM standard test machine. With a preset round free-standing area, for this apparatus the optimal test sample dimensions are 20 × 20 × 5 mm (length × width × thickness). The drop weight is fixed at 1.2 kg and the drop height varied from 0 to 0.74 m. The energy of impactor can be expressed by: E = mgh; where E is the impact energy imparted to the sample, m is the mass of the impactor, g is the gravitational constant and h is the total drop height. Only bighorn and domestic sheep horn samples were tested because of the limitation of horn size and shape in the other two species.

2.6. Statistical analysis

Statistically significant differences of the Young's modulus and yield strength among the different horns were identified by using one-way ANOVA analysis. The criterion for statistical analyses was p < 0.05.

3. Results and discussion

3.1. Microstructures of the different horns

Low magnification optical micrographs of ultra-thin horn slices cut perpendicular to longitudinal direction are shown in figure 3. Figure 3a,e,i and m are schematics of the distributions, shapes and sizes of tubules in bighorn sheep horn, domestic sheep horn, mountain goat horn and pronghorn horn, respectively. The transverse section images of each horn are shown in figure 3b,f,j and n. As shown, the shape and size of the pores (cross section of the tubules) vary among the different species. Elliptically shaped pores are found in the bighorn sheep and pronghorn. For the bighorn sheep, the pores are more evenly distributed and uniform in size (approx. 70 µm in major axis, approx. 35 µm in minor axis). For the pronghorn, the pores tend to disperse stochastically between the lamellae and have a wider range of size (from 28 × 10 to 97 × 37 µm). The pronghorn horn has several large pores (only one shown in the figure 3n, with a major and minor axis dimension of 105 µm and 84 µm). These large pores were not observed in the other three horns. Pores from the domestic sheep show a deformed oval shape with a large length to width ratio of 5.7 (a major axis dimension of 80 µm and a minor axis dimension of 14 µm). For the mountain goat, tubules were not observed (figure 3i,j).

Figure 3.

Schematic of the tubule distribution, optical microscopy and scanning electron microscopy (SEM) images of horn cross sections for different species. (a–d) Bighorn sheep, (e–h) domestic sheep, (i–l) mountain goat and (m–p) pronghorn. The radial, longitudinal and transverse directions are indicated in the schematic diagrams. Different tubular cross section and pore shapes found in optical microscopy images (b,f,j,n). Lamellae surrounding tubules are identified in the SEM images of the cross sections (c,g,k,o). Higher magnification SEM images reveal the cell lamellae in each species with a thickness approximately 1–2 µm (d,h,l,p). (Online version in colour.)

SEM micrographs of the fracture surfaces of the four species after fracture in liquid nitrogen are shown in figure 3c,g,k and o. Elliptically shaped tubules were found in the bighorn sheep horn cross section (figure 3c). Oval-shaped pores are identified as in the optical microscopy image (figure 3g). No tubules but a wavy lamellar structure are found in the mountain goat horn (figure 3k). Figure 3d, h,l and p are SEM images with a higher magnification, showing the cell lamellae in each horn. The thicknesses of the keratinized cells are approximately 1–2 µm. Tubular density (porosity) in each species was calculated and averaged from five optical microscopy images using ImageJ. The bighorn sheep horn has the highest porosity (11.3%), followed by domestic sheep (7.7%) and pronghorn (5.8%), while the mountain goat has 0% porosity. There were some voids identified in the pronghorn cells, while no voids were found in the other three species. Embedded hair fibres in the keratin sheath also were found in pronghorn (figure 3o), which may anchor the sheath in place during the rapid growth of new keratin layers after shedding the old horn sheath [51]. These hair fibres account for the large pores observed in the previous optical microscopy image (figure 3n).

Higher magnification optical micrographs of the cross- and transverse sections from different horns are shown in figure 4. The lamellar structure can be seen more clearly from these images, showing a high similarity between the different horns. The lamellae are layered in the radial direction as illustrated in figure 4a–d. This result indicates there is a common layering pattern of the lamellae in the different horns. Micrographs of the longitudinal sections reveal the keratin cell morphology and relationship between keratin cell alignment and tubule growth direction (figure 4e–h). Keratin cells show a disc-shaped morphology, roughly spherical with the radius (approx. 20 µm) much larger than the thickness (approx. 2 µm) [5]. In the longitudinal direction, the cells are attached edge-to-edge, forming the lamella. The tubules extend along the longitudinal direction with no termination observed. However, differences are found between the species. In the bighorn sheep (figure 4e), the parallel lamellae form an approximately 30° angle to the tubule direction. In the domestic sheep (figure 4f), the lamellae are layered in a more disorderedly manner than in other horns. The lamellae that are close to the tubules tend to be layered parallel to the tubules, whereas away from the tubules, the lamellae form an acute angle to the tubule growth direction. As shown in figure 4g,h, in the mountain goat and pronghorn, the lamellae are layered parallel to each other and to the growth direction. Schematic diagrams in figure 4 show the tubular and lamellar structures in the horns of different species. Laminated cells are indicated with red lines, showing differences among the species. The inclined lamellae in bighorn sheep and domestic sheep could be a consequence of the strong curvature of the horns in these two species.

Figure 4.

High magnification optical microscopy images of cross sections and transverse sections stained with toluidine blue: (a,e) bighorn sheep horn, (b,f) domestic sheep horn, (c,g) mountain goat horn and (d,h) pronghorn horn. Pores and cell lamellae are identified in the cross sections (a–d). Cell lamellae directions show differences in each species: (e) approximately 30° in the direction between cell lamellae and tubule direction (longitudinal). (f) No specific arrangement is found in domestic sheep. (g) Cell lamellae are parallel to the longitudinal direction in mountain goat horn; (h) similar to the mountain goat, the cell lamellae in the pronghorn horn are parallel to the orientation of tubules. (Online version in colour.)

3.2. Compression tests

Compression specimens of the different horns in both ambient dry and fully hydrated conditions were tested. The water content in each species was determined before the compression tests because of the significant effects of hydration on the mechanical properties, as reported in the previous literature [20,22,52]. Figure 5a shows the degree of water absorption in each horn. In the ambient dry condition, the water content of different species varies, ranging from mountain goat (approx. 7 wt.%) to pronghorn (approx. 14 wt.%). After submersion in purified water for 126 h, the water content in all four species increased, and pronghorn showed the highest water content (approx. 52 wt.%). Figure 3p shows that there are numerous nanopores inside the keratin cells, which are not found in the other three species. This indicates that the keratinized cells in the pronghorn are not fully cornified and densified, perhaps due to the rapid annual growth and shedding of the horn sheaths, a unique feature of pronghorns. The mountain goat horn has the lowest water content (approx. 22% after hydration) and almost zero porosity. Therefore, both tubules and the nanopores should be considered as reservoirs for water, because the water content of rehydrated samples from mountain goat, domestic and bighorn sheep is positively related to their porosity. In a previous study, hollow tubules in hoof were hypothesized to serve mechanical functions and not transfer water from the surface of the hoof due to the dense layering of keratinized cells [30]. However, based on the present study on the hydration states of horns with different tubule densities, we suggest that tubules can contribute to water absorption by keratin. Thus, it appears that the overall water contents were correlated with increased density of tubules in the different horns.

Figure 5.

(a) Water content (wt.%) of horns in the ambient dry and fully rehydrated conditions. (b) Fourier-transform infrared spectroscopy (FTIR) spectra of ambient dry horns in the different species. (Online version in colour.)

The chemical compositions of different horns were characterized by FTIR (figure 5b). Bighorn sheep and domestic sheep show very similar spectra, indicating almost the same protein compositions in these two species. This can be explained by the fact they are from the same genus. The mountain goat horn has a higher –S–C-peaks than the other three species, showing a cystine-rich composition [53]. The presence of cystine monoxide (S=O) peak indicates the oxidation of cystine, which could be due to the influence of the mountain goat horn preservation [53–55]. The sharp C=O peak at approximately 1750 cm⁻¹ in the mountain goat horn indicates a higher fraction of aspartic and glutamic acids [56,57]. However, the chemical composition analysis by FTIR is a qualitative way to show the differences between various horn keratins. More quantitative amino acid, peptide and proteomic analyses of different horn keratins are needed to better determine compositional effects on horn properties.

The stress–strain curves of the ambient dry condition and fully hydrated horns are shown in figure 6. The stress–strain curves in all directions exhibited the same trend, which is an initial linear elastic deformation region followed by a long plastic region and ending with a densification region. This type of stress–strain curve is similar to the compression behaviours in typical cellular solids, which have been used for load-bearing and energy absorption purposes [58]. All horns present high strains with a deformation of approximately 70% without failure. However, anisotropic behaviour is evident when comparing the curves loaded in the longitudinal, radial and transverse directions. When compressed along longitudinal direction, as shown in figure 6a, the horns show a long plastic region with nearly constant stress. Based on a previous work [23], this may be attributed to delamination and gradual buckling of lamellae, facilitating a large deformation without the change of stress level. When compressed along other directions, where tubules extend perpendicular to loading direction, the stress plateau is shorter compared with longitudinal loading conditions. This could result from tubule closure during compression, leading to a fast densification of keratin layers [22,23]. Therefore, the compressive stress increases more rapidly, as shown in figure 6b,c. The mechanical properties of ambient dry horns in compression tests are listed in table 1. The yield strength is taken as the point on the strain–stress curve where the slope deviates from linearity, and the initial slope is the Young's modulus. The toughness is measured as the area under the strain–stress curve up to a strain of 70%. For all horns, the longitudinal direction has the highest Young's modulus and yield strength. As the keratin lamellae are almost parallel with the longitudinal direction, this is suggested to result in a higher stiffness and strength, which has been verified by a previous study [23]. When comparing the mechanical properties of different horns loading in different directions, no significant difference is found for the Young's modulus, which indicates that the stiffness is independent of the tubule density, size and shape (electronic supplementary material, figure S1). The compressive yield strength of mountain goat horn is significantly larger than that of the other horns when loading in radial and transverse directions, which is likely due to the existence of tubules in the other three horns. The yield strength in longitudinal direction is not affected by the tubules because there is no significant difference between these four species (electronic supplementary material, figure S2). This may correspond to the fighting styles of the different species. The mountain goat fights involve stabbing and thrusting, and the high yield strength and stiffness make the horns strong enough to maintain structural integrity during the stabbing and penetrating into the dermis of their opponents.

Figure 6.

Comparative compressive stress–strain curves of horns in different directions (longitudinal, radial and transverse) and hydration conditions (ambient dry and fully hydrated). Stress–strain curves are averages of five samples for each species, and error bars indicate standard deviations. (Online version in colour.)

Table 1.

Compressive Young's modulus, yield strength and toughness of the horns under different loading orientations in the ambient dry condition.

species	orientation	pronghorn	bighorn sheep	domestic sheep	mountain goat
Young's modulus (GPa)	longitudinal	1.8 ± 0.2	1.7 ± 0.5	2.0 ± 0.2	2.1 ± 0.6
	radial	0.9 ± 0.3	1.2 ± 0.2	1.3 ± 0.5	1.5 ± 0.2
	transverse	1.2 ± 0.2	1.3 ± 0.6	0.9 ± 0.3	1.4 ± 0.4
yield strength (MPa)	longitudinal	85.5 ± 6.4	67.8 ± 7.9	64.3 ± 13.5	81.6 ± 22.2
	radial	67.5 ± 4.2	57.3 ± 5.3	63.1 ± 8.9	82.3 ± 9.8
	transverse	59.9 ± 7.9	49.3 ± 9.0	59.5 ± 6.1	76.1 ± 2.5
toughness (MJ m−3)	longitudinal	72.5 ± 2.5	92.1 ± 9.5	83.5 ± 5.1	103.4 ± 5.8
	radial	82.0 ± 5.6	117.0 ± 4.5	98.1 ± 5.6	141.0 ± 16.0
	transverse	78.1 ± 5.7	98.4 ± 5.4	85.2 ± 3.0	120.1 ± 8.8

In contrast to the ambient dry horn stress–strain curves, figure 6d–f shows the compressive stress–strain curves in the rehydrated state. The maximum compressive stress drops severely (order of magnitude) for all orientations and has a negative relationship to increasing water content. The mechanical properties of the rehydrated horns are shown in table 2. The average Young's modulus of three directions decreased 97% in pronghorn, 91% in bighorn sheep, 87% in domestic sheep and 84% in mountain goat after being fully hydrated. Given the significant decrease of all mechanical properties over those of the dry horns, hydration is demonstrated to significantly influence the mechanical response. Comparing the four horns, the strength and stiffness in the pronghorn are significantly lower than that in the other three horns, while the mountain goat has the highest strength and stiffness after hydration (electronic supplementary material, figures S3 and S4). These can be related to differences in water content after hydration, because the pronghorn has the highest water absorption while the mountain goat has the lowest (figure 5).

Table 2.

Compressive Young's modulus, yield strength and toughness of the horns under different loading orientations in the fully hydrated condition.

species	orientation	pronghorn	bighorn sheep	domestic sheep	mountain goat
water content (wt.%)		52	40	29	22
Young's modulus (GPa)	longitudinal	0.11 ± 0.04	0.14 ± 0.08	0.26 ± 0.13	0.50 ± 0.23
	radial	0.02 ± 0.007	0.09 ± 0.04	0.11 ± 0.03	0.14 ± 0.03
	transverse	0.02 ± 0.003	0.16 ± 0.05	0.15 ± 0.06	0.23 ± 0.14
yield strength (MPa)	longitudinal	1.6 ± 0.9	3.4 ± 0.8	4.9 ± 1.1	13.9 ± 4.1
	radial	1.7 ± 0.5	5.2 ± 0.6	5.4 ± 0.5	8.6 ± 0.8
	transverse	1.0 ± 0.3	3.2 ± 0.4	3.6 ± 1.1	7.2 ± 1.7
toughness (MJ m−3)	longitudinal	3.5 ± 0.7	13.3 ± 1.7	16.7 ± 3.6	18.4 ± 2.7
	radial	7.2 ± 1.5	20.1 ± 1.1	21.5 ± 5.3	27.1 ± 2.6
	transverse	6.8 ± 0.4	15.5 ± 1.1	15.4 ± 3.5	26.5 ± 2.0

As shown in figure 7, for the dry horns (solid lines), the radial direction can absorb more compressive energy than in the longitudinal and transverse directions. This is because the lamellae are layered face to face along this direction, which allows them to withstand more compressive stress without overly deforming. The radial direction is also the orientation in which sheep horns clash during combat. In all directions, mountain goat horn exhibits the largest energy absorption compared to the other horns. The nearly zero porosity in the mountain goat horn and parallel alignment of the lamella in the longitudinal direction may be the factors influencing the energy absorption ability. Bighorn sheep also possess high energy absorption capability, especially along the radial direction. During combat, the bighorn sheep butt horns at high speeds. The impressive energy absorption prevents horns from cracking and crushing when subjected to high impact loads. When the horns are fully rehydrated, the energy absorbed decreases significantly (figure 7 dashed lines), which is likely due to the softening of the keratin. Also, the difference of energy absorbing ability between different loading directions diminishes, which indicates a matrix-dominant property for rehydrated horns.

Figure 7.

Energy absorption as a function of compressive strain of dry and rehydrated horns in the (a) longitudinal, (b) radial and (c) transverse directions. For each species, plots are based on an average of five samples for each condition. (Online version in colour.)

In summary, there are many common characteristics of the different horns. Anisotropic mechanical properties are found in all the horns in ambient dry conditions. The longitudinal direction has the highest Young's modulus and yield strength, which indicates the stiffest and strongest direction of the horn. The radial direction (clashing direction) has the largest toughness, which indicates that more energy can be absorbed, compared to the other directions. For all orientations, the mechanical properties decrease with increasing water content. It can be concluded that the change between hard (dry) and ductile (hydrated) horns can be manipulated by changing the water content of the structure. All species present hydration-dominated mechanical properties, which indicates a common characteristic for the horns.

3.3. Tensile tests

Tensile tests were performed along the longitudinal direction because it is the most likely direction to undergo tensile stress when fights occur. Horns were tested in ambient dry conditions. The stress–strain curves are shown in figure 8. The mountain goat has the largest failure strain (14.7%), while the domestic sheep has the smallest (5.9%). The failure strain in tension is much smaller than in compression, which indicates a different fracture mechanism. The mechanical properties of the horns under tensile stress are listed in table 3. Among the four horns, the mountain goat is the stiffest (1.54 GPa), strongest (yield strength of 88.0 MPa and ultimate strength 107.3 MPa) and toughest (11.9 MJ m⁻³) (electronic supplementary material, figure S5). This may be linked to the mountain goat's stabbing combat style. The horns may be subjected to tensile stresses when the goat tries to pull its horns out from the opponent. A high yield strength and stiffness maintains structural integrity while stabbing and avoids buckling and large plastic deformation. Unlike the other three species, the stabbing behaviour requires the mountain goat horns to be thin, sharp and light to penetrate the skin of their opponent. Thus, mountain goat horns are expected to have the highest tensile properties relative to the other species. The pronghorn has the second strongest (yield strength of 69.6 MPa and ultimate strength of 78.3 MPa) and toughest (6.1 MJ m⁻³) horn. The pronghorn performs two methods of combat, one involving thrusting and twisting where tensile stress will be generated. To avoid fracture of horns, these would be important properties. By contrast, bighorn and domestic sheep do not engage in stabbing or wrestling behaviours and their horn yield strengths are lower (38.9 GPa and 41.9 GPa for bighorn and domestic sheep, respectively) compared to mountain goat and pronghorn. No significant difference is found between these latter two species. Mountain goat and pronghorn horns have lamellae that face parallel to the tensile direction (figure 4g,h), while sheep horns have lamellae arranged at angles to tensile direction (figure 4e,f). Therefore, the superior mechanical properties of mountain goat and pronghorn may be a result of the aligned lamellae that are parallel to the tensile direction. The larger amount of disulfide bonds in the mountain goat horn can also be a factor that improves the mechanical behaviour.

Figure 8.

Tensile stress–strain curves in the longitudinal direction of horns in the ambient dry condition. Averages based on seven samples from each species; error bars indicate standard deviations. (Online version in colour.)

Table 3.

Tensile properties of different horns in the ambient dry condition.

species	pronghorn	bighorn sheep	domestic sheep	mountain goat
Young's modulus (GPa)	1.06 ± 0.2	1.15 ± 0.1	1.33 ± 0.1	1.54 ± 0.2
yield strength (MPa)	69.6 ± 12.3	38.9 ± 7.3	41.9 ± 9.9	88.0 ± 12.5
ultimate strength (MPa)	78.3 ± 14.9	45.1 ± 10.5	47.9 ± 11.5	107.3 ± 9.8
failure strain (%)	11.9 ± 1.5	6.5 ± 1.2	5.9 ± 1.4	14.7 ± 1.8
toughness (MJ m−3)	6.1 ± 1.6	2.3 ± 0.7	2.1 ± 0.9	11.9 ± 1.7

Compared to compressive tests at the same hydration level (ambient dry) and load direction (longitudinal), the tensile failure strain and stress decrease significantly. Under compressive stress, all horns can sustain a deformation of 70% without failure (figure 6). However, under tensile stress, all horns fracture when the strain is greater than 15%. Horns can be subjected to a compressive stress of more than 200 MPa without failure, while the ultimate tensile strength of horns is four times lower in bighorn sheep and domestic sheep horns (figure 8). Under tensile loading, the toughness is significantly less (table 3) than under compressive loading (table 1), indicating a more brittle characteristic.

SEM images of the tensile fracture surfaces are shown in figure 9. For all horns, tubule pull-out is rarely observed, which indicates a high tubule-to-matrix adhesion. However, fibre-to-fibre adhesion may be one factor that influences the tensile properties. For horns having a relatively low yield strength (bighorn and domestic sheep, figure 9a–d), detachment of the fibre bundles from the matrix is observed. The detached fibres are thin with a small diameter of 1–1.5 µm. High stresses applied on those thin fibres cause fibre fracture at a relatively low tensile load. For horns having a high yield strength (mountain goat, figure 9e,f), detached fibres are not observed, indicating a good fibre-matrix adhesion. For the pronghorn, which also has a high yield strength, there is separation between the fibre bundles and the matrix, but the detached fibre bundles have a larger diameter (greater than 10 µm). Compared to the thin fibres observed in bighorn and domestic sheep, these thick fibres are presumably stronger under the same load and can withstand more tensile stress.

Figure 9.

Scanning electron microscopy images of tensile fracture surfaces of horns from four species. Images at top (a,c,e,g) are low magnification, and high magnification images (b,d,f,h) are below. (a,b) Bighorn sheep, (c,d) domestic sheep, (e,f) mountain goat and (g,h) pronghorn. (Online version in colour.)

3.4. Drop tower impact tests

According to ASTM standard D7136/D7136 M-07, there are four externally visible damage types and two internal damage types appearing in drop-weight tests. The external damage types include dimple (depression), splits (cracks), combined splits and delamination, and puncture. Internal damage types are sorted into two categories: delamination and splits. Dimpling is the main damage mode on the impact surface for both sheep horns. With an increase in the impact energy, the diameter and depth of dimple increases. On the bottom surface, from low to high impact energy, the domestic sheep horn shows all external damage types. For bighorn sheep, on the bottom surface only split/crack, delamination and puncture were observed.

The normalized impact strength (E_n) is calculated by: E_n = E/d_st, where E is the impact energy derived in §2.5, d_s is the cover plate aperture diameter and t is the thickness of the specimen. For bighorn and domestic sheep, the typical damage modes under low and high normalized impact energy (E_n) are shown in figure 10. At low E_n (30 kJ m⁻²), shown in figure 10a,d, for both horns there is a shallow dimple on the top (impact) surface with an area identical to the size of the hemispherical impactor tip (3.2 mm). On the bottom of the samples, the domestic sheep have a minor crack (length ∼ 3 mm), while the bighorn sheep horn exhibits a longer crack (length ∼ 10 mm), which indicates that domestic sheep can sustain more stress and exhibit a lighter damage when E_n is small. At high E_n (60 kJ m⁻²), for both horns a number of samples fail, and the typical failure modes are exhibited in figure 10c,f. On the bottom surface of the domestic sheep, circular and thin chips fractured away, while for the bighorn sheep thin chips were only punched out but not fractured away. Also, there is delamination combined with the dimple on the impact surface of the domestic sheep, which means that at high E_n level, bighorn sheep show a higher capability to withstand the impact load. The number of each damage mode under the same E_n is shown in figure 11. Bighorn sheep horn begins to fail when E_n = 40 kJ m⁻², whereas there is no failure damage for domestic sheep at this energy level, which indicates a higher impact resistance of domestic sheep under low E_n. According to Lee et al. [50], failure impact strength is defined when 50% of the samples fail (with failure defined when the bottom surface of the sample shows puncture damage). The failure impact strength of domestic sheep is 55 kJ m⁻² and for the bighorn sheep is 75 kJ m⁻²; 36% higher than the domestic sheep. The tubular structures are considered critical for absorbing impact energy in various structural materials [59]. The reason why bighorn sheep horns can withstand greater impact energy than domestic sheep horns could be due to the differences between their tubular structures. Higher tubular density in bighorn sheep horns can absorb more energy by plastic deformation. The failure impact strength reveals that the bighorn sheep horn has better impact resistance when the impact speed is high. Domestic and bighorn sheep have similar fighting styles, but with different fighting speeds. When fighting, domestic sheep stand closer to each other than bighorn sheep, thus generating a lower speed for clashing. The impact test results suggest that horn properties are tailored to fit fighting behaviour.

Figure 10.

External damage modes of domestic sheep (a–c) and bighorn sheep (d–f) with associated normalized impact energy (E_n). Radial direction is the impact direction, which is perpendicular to the imaged surface. Black lines are sample marks. (Online version in colour.)

Figure 11.

Damage types (bottom surface) with different normalized impact energy from the (a) domestic sheep and (b) bighorn sheep horn. (Online version in colour.)

4. Conclusion

The microstructures and mechanical properties of four different species: bighorn sheep, domestic sheep, mountain goat and pronghorn were characterized (optical and SEM, compression and tensile tests, impact tests) and compared. The main conclusions are as follows:

• — Keratin cells that form lamellae surrounding elliptically shaped hollow tubules were identified in the bighorn sheep, domestic sheep and pronghorn, while no tubules were found in the mountain goat. The bighorn sheep horn has the highest porosity (from the presence of the tubules) of 11.3%. The porosity of domestic sheep horn is 7.7% and the tubules have a deformed oval shape. The pronghorn horn has a porosity of 5.8%. The shapes, dimensions and tubule densities were different across the species.

• — Laminated keratin cells align parallel to longitudinal direction and show a high tensile strength in the pronghorn and mountain goat. The bighorn sheep horn lamellae arrange approximately 30° to tubule direction and the domestic sheep horn has a more disordered alignment, resulting in lower tensile strengths for the sheep compared to the pronghorn and mountain goat.

• — The pronghorn horn has the largest water content with a maximum of 52% that may be caused by the additional water stored in the nanopores inside the keratin cells. The water absorption in bighorn sheep horn is larger than that in domestic sheep, and mountain goat has the lowest water content. The amount of water absorption is positively related to the tubule density. In terms of chemical composition, horns from bighorn sheep and domestic sheep have similar protein compositions, while mountain goat horn has the highest cystine composition.

• — All of the horns have similar anisotropic mechanical behaviours related to loading direction: longitudinal (direction of tubules), radial (direction of impact) and transverse (perpendicular to both). In the dry condition, the longitudinal direction has the highest Young's modulus and yield strength. The radial direction has the largest toughness.

• — The mountain goat horn has highest yield strength and energy absorption in all directions, hydration conditions and loading states, and this likely benefits a stabbing combat style. Bighorn sheep horns are characterized by high energy absorption capability under compression in radial direction, which makes them suitable for absorbing energy generated by high-speed head butting. The pronghorn horn shows relatively high tensile strength that enables clashing combat and interlocking horn fighting behaviour.

• — Under low energy impact tests, the domestic sheep have higher impact resistance than the bighorn sheep. At high normalized impact energy (greater than 55 kJ m⁻²), the bighorn sheep have better impact resistance. The failure impact strength of domestic sheep is 55 kJ m⁻² and the impact strength of bighorn sheep is 75 kJ m⁻². The result indicates that bighorn sheep have developed horns that can survive large impact forces.

Different structures in horns are found to serve for certain specific mechanical functions. Cell lamellae in mountain goat and pronghorn are designed for high tension requirements, while tubular structures in bighorn sheep and domestic sheep enhance energy absorption during impact. Apart from the hierarchical structures, the chemical compositions in different horns could also affect the mechanical properties, which need further studies based on quantitative analysis of amino acids, peptide and proteomics. These findings based on the diversity of forms and functions that have evolved in Nature give further inspiration for the design of synthetic materials with a wide range of mechanical applications.

Data accessibility

Supporting data are available upon request from jmckittrick@eng.ucsd.edu.

Authors' contributions

J.M., C.H. and J.G. initiated the project. Y.Z. and W.H. conducted the experiments; C.H. and J.G. performed the evolutionary analysis; J.M. directed the project; Y.Z., W.H. and J.M. drafted the manuscript; W.H., C.H., J.G. and J.M. revised the manuscript. All authors gave the final approval for publication.

Competing interests

We have no competing interests.

Funding

This work is supported by a Multi-University Research Initiative (MURI) through the Air Force Office of Scientific Research of the United States (AFOSR-FA9550-15-1-0009) and a National Science Foundation Biomaterials Program grant no. 1507978. This work was performed, in part, at the San Diego Nanotechnology Infrastructure (SDNI) of UCSD, a member of the National Nanotechnology Coordinated Infrastructure, which is supported by the National Science Foundation (grant no. ECCS-1542148).