The Structure and Functions of the Contour Feathers of Water Birds Revisited

Abstract

The structural details of the contour feathers of water birds are known to serve a variety of functions ranging from intraspecific signaling to such physical qualities as thermal insulation, water repellency, resistance to impact and viscous drag reduction. All of them closely match the requirements of habitat and feeding habits. It comes as no surprise, therefore, that contour feathers are composed of an array of elements that confer these qualities to the optimal benefit of their avian bearer. In this chapter, we review the physical bases for these functions expressed in both structural and evolutionary terms. Some functions excel at the expense of others and many species have evolved an optimal balance between feather micro-structure and behavioral patterns that suit their specific environment. Several relationships between feather function and the structural properties of water bird feathers can be identified as specific evolutionary adaptations.

1. Introduction

Water birds belong to a large group of families that have open water in their habitat. They make their home in many diverse environments including oceans, shores, estuaries, rivers, lakes and wetlands. Many of them feed in salt water, taking their prey from the surface or catching it under water by swimming, plunging and deep diving. Some exploit the skies above pursuing their prey in an unobstructed three-dimensional space without ever alighting. Others inhabit only fresh water habitats and forage in a wide variety of ways such as skimming the surface or stalking under water Among them are families that have colonized the remotest parts of our earth and have adapted to the most extremes of climatic conditions. Indeed, water birds can be found foraging and often breeding at all latitudes. They can truly be said to have conquered the entire aquatic world.

Such wide-spread occurrence has exposed water birds to a multitude of evolutionary forces that have shaped their anatomy and behavioral patterns to optimally suit their specific environment. In this chapter, we will show how the feathers of water birds, in particular the contour feathers, vary among families and exhibit a range of properties that function, among other things, to regulate body temperature, repel water, prevent water from penetrating to the skin and resist the impact forces of diving, plunging and alighting. Some of these functions excel in extreme environmental conditions or in relation to specific feeding techniques, frequently at the expense of other functions. Others represent a balance between two or more opposing functions. In consequence, many of these functions are expressed in an array of feather characters that confer these qualities to the optimal benefit of the avian bearer.

To study the relationship between structure and function in some detail, a closer look at the composition of feathers is in order. The morphology of feathers has been well described in the ornithological literature [1, 2] and is reproduced here only to the extent necessary for the purpose of this chapter. All feathers, whether flight, tail or contour feathers, consist of essentially the same elements, only their relative prominence is different. At the base of the spine (or rachis) occur the downy or plumulaceous feathers, only a tuft in flight and tail feathers, but extensively present alongside the proximal two-third of the rachis of contour feathers. These are thought to primarily function as a means to regulate body temperature by entrapping air [3, 4]. More distally, a highly structured pattern is present with rami extending from the rachis in the plane of the feather, each sprouting barbules of which the distal ones have hooks that catch upon the curled proximal barbules of the barb next more distal. This continuous-looking, hook-and-flange arrangement provides this pennaceous part of the feather with the rigidity so critical for its mechanical and other properties.

In flight and tail feathers, the pennaceous part is by far the most dominant part – reflecting their importance in flight - but in contour feathers it occupies only the distal one-third of the length of the feather. The proximal two-third is made up of downy elements that work, among other things, as structural reinforcements limiting the bending of the downy barbules.

Efforts to correlate structure and function require an exact knowledge of all respective feather elements and a full understanding of their specific functions. However, much as we know about feather anatomy, not all details can be explained in terms of functional performance. For instance, it has been well established that suites of microscopic characters seen in downy barbules can be used to identify certain groups of birds [5] and, more recently, to correlate habitat and behavioral patterns with variations observed in these microscopic characters [6]. Because downy feathers are thought to primarily function as a means to regulate body temperature by entrapping air, the presence in the down of any species- or group-specific character suggests that environmental forces other than thermoregulation can and do exact evolutionary changes in down. We do not know what these forces are.

We also do not know all functions of feathers or their components. However, we will address here the functions that we do know and try to correlate these with the constituent feather elements, quantitatively or semi-quantitatively where possible.

2. Signaling

Birds signal to each other by behavior and the coloration of their feathers. Color can serve, among other things, to hide from predators, to attract mates or fight off rivals and is brought about by pigment cells within the feather. Melanin is an important pigment produced by melanocytes arrayed within the barbs and barbules. Another source of color is iridescence, a nano-structurally colored sheen on the feathers of several lineages of birds [7]. As far as we know, neither pigment nor iridescence interferes directly with the functionality of other feather elements.

3. Thermoregulation

As mentioned earlier, downy feathers are thought to function by entrapping air. The downy feather characters, including the nodes observed in many water bird and other families, are believed to work as structural reinforcements that limit the bending of the downy barbules. They may also function to catch other barbules and keep them from becoming entangled, thereby allowing the entrapment of more air and serve as a better thermal insulator [8, 9].

If the presence of nodes and other downy characteristics contribute to the conservation of heat by air convection, one may expect additional adaptations in the down of contour feathers for the conservation of body heat radiation. The nodes with their high content of dark granules, most likely consisting of melanin, may trap heat through infrared absorption and function by storing heat.

The surface of a bird’s skin loses energy in the form of heat through two mechanisms: one is convection heat, warmed air that is trapped by down and overlying pennaceous feathers as stagnant air, the other is heat by radiation, emitted from the skin of all warm-blooded animals as detectable by infrared cameras. Part of the radiation is absorbed by the feather keratin and, in turn, converted into convection heat and, partially re-emitted from the keratin or lost to the surrounding environment. By investigating the infrared absorption spectra of contour feathers it is possible to identify the contribution to radiation heat conservation by their respective downy and pennaceous parts. This has been done for the King Penguin (Aptenodytes patagonicus), the Wood Stork (Mycteria americanus) and the American Crow (Corvus brachyrhynchos) as representatives of divers, waders and land birds, respectively [10]. The results show essentially identical spectra for the three species over the range of 500 to 4500 cm⁻¹ wavenumbers suggesting there are no adaptative effects of foraging niche and habitat on the radiation heat conservation mechanisms of these and possibly all species.

There is, however, a prominent difference between the spectra of downy feathers and the pennaceous parts at 1700 cm⁻¹. Here, down shows a sharp transmittance peak that is strikingly absent in the pennaceous. The occurrence of this peak can be best explained by the Christiansen effect [11, 12] observed in binary systems, such a liquid and a powder, where the transmittance peak occurs at a wavelength for which both components have the same refractive index. In the case of down and air, this occurs at 1700 cm⁻¹. At all other wavelengths, scattering occurs due to the difference in the refractive index between down and air.

The values of the refractive index of materials, and the way these depend on wavelength, are determined by several factors. Apart from chemical composition, crystallinity with its various axial orientations, contributes significantly to these values. In long-chain molecules, such as keratin, the extent of crystallinity and the preferential orientation of crystalline regions provide a level of ordered structure that contributes as well. Non-crystalline alignment of polypeptide chains due to interchain bonding such as electrostatic bonds and hydrogen bonds, can also be expected to contribute to the value of the refractive index [10].

As the infrared spectra have shown, the chemical composition of all parts of the contour feathers is essentially the same and is dominated by the bulk of the beta-keratin component. However, the refractive indices for the various wavelengths for down and pennaceous are not the same due to their differences in crystallinity and structural orientation. As a result, the radiation peak transmitted by down at 1700 cm⁻¹ is not observed for the pennaceous part of the feather. Its absence is most likely explained by the limited range of wave numbers employed to coincide with the range of body heat radiation. Had the range of wave numbers been extended beyond the infrared, it is likely that a transmittance peak for the pennaceous could have been located, provided, of course, that a wavelength for which pennaceous and air have the same refractive index actually exist.

Similar considerations are expected to apply to radiation heat that originates from the environment. Here, infrared from solar radiation with wave numbers comparable to those of body heat radiation will be scattered or partly absorbed by the pennaceous feather parts and not directly reach the skin of the bird. For wave numbers extending beyond the infrared range, it is, again, possible that a transmittance peak could be located for the pennaceous portion of the feather, but the down is not likely to be permeable at that particular wave number. The difference in feather ultrastructure between the down and the pennaceous parts, with its concomitant difference in infrared transmittance, is the reason why radiation heat, whether from body heat or an environmental source, does not directly permeate the feather coat.

The Christiansen effect with its transmittance peak at 1700 cm⁻¹, as observed in the infrared spectra of down but absent in the pennaceous parts, is likely to have important consequences for the conservation of radiation heat by the bird. Heat radiated from the body will be scattered and re-scattered in all directions by the downy barbules, reflecting radiation back to the skin where it is absorbed. However, radiation at 1700 cm ⁻¹ will be fully transmitted by the down but not by the pennaceous, which will scatter it again, thereby conserving heat within the feather coat.

The nodes in the down, and in particular those that contain melanin, may aid in the process of radiation heat conservation by virtue of their heat capacitance but efforts to document infrared spectra on isolated nodes were unsuccessful [10]. Therefore, it has not been possible to determine whether nodes play any direct role in the conservation of body heat radiation.

The containment of body heat radiation is likely to be of secondary importance in the overall conservation of body warmth in birds. No data on total heat conductivity have been experimentally determined for the radiation part in feathers, but estimates for diabase wool used in the construction sector are near 25 percent [13]. However, because most of the radiation in birds is scattered by down, converted to convection heat and partly re-emitted at a different, higher wavelength, it is evident that the transmittance leak at 1700 cm⁻¹ constitutes only a small loss of thermal energy by the down. The actual loss is probably minimal as this transmittance is scattered back to the skin by the pennaceous parts of the contour feather.

If the Christiansen effect is insignificant in the loss of radiation heat, then one can speculate whether the transmittance peak at 1700 cm⁻¹ and its deflection by the overlying pennaceous parts constitute an incidental physical phenomenon or represent an adaptation to the overall thermal balance. The available data do not provide an obvious answer. However, the absence of any feather components with specific infrared properties that stop the transmittance leak seems to suggest an incidental phenomenon rather than an evolutionary adaptation. No specific infrared absorption bands in either the keratin or the pigmented nodes have been observed that could account for such a stoppage.

The amounts of radiation heat that birds generate or lose are unknown and do not appear to be amenable to direct experimental determination, at least not by current technologies. Infrared spectra of feather components have demonstrated the mechanisms by which radiation heat is conserved and lost, but have not provided for a quantitative analysis.

The distal, pennaceous part of the contour feathers is not only exposed to environmental radiation but also to air and water. It is at this interface that the physical interaction with the external world occurs and where most adaptations to environmental factors can be found.

4. Water repellency and resistance to water penetration

One of the major functions of feathers is to prevent water from reaching the skin or weighing down the remiges and tail feathers in flight. With very few exceptions, all birds benefit from a plumage that optimally repels and resists the penetration of water. However, the manner in which this optimum is realized for each water bird family is closely associated with its behavior and interaction with its habitat and, as a result, the feather characters responsible for water repellency and resistance vary accordingly. To understand the way a water repellent/resistant structure functions, certain aspects of surface physics should be made clear.

The water repellency of feathers and other biological porous structures, such as the stomatal apparatus of leaves and the spiracles of insects, is governed by the fundamental principles of surface physics that apply to all porous surfaces whether natural or manmade. It is determined by the relative areas of solid – water and air – water interface and their respective interfacial energies regardless of the actual architecture of the repellent structure itself [14]. If the surface of the solid is coated with another material, such as paint or preening oil, it will assume the properties of the coating material. For feathers coated with uropygial gland oil, the feather – water interface is, in fact, an interface between gland oil and water.

A drop of water placed on a smooth feather surface such as the rachis, will pearl up and roll off easily. This surface is then said to be water repellent, the actual extent of which is determined by the contact angle θ, defined as the angle between the tangent to the curved water surface at the point of contact with the feather surface and the plane of the surface on which the drop is resting, measured through the water. When the drop is placed on the porous vane of the feather, it will entrap air in the hollows and interstices, forming additional air – water interfaces, which will cause considerable increase in the contact angle.

The manner in which this increase occurs has been described in previous publications [14, 15, 16] and need not be repeated here. Suffice to say that the contact angle and thus the water repellency of the distal one-third of the contour feather is quantitatively determined by (r + d)/r where 2r represents the diameter of the barbs measured in the plane of the long axes of the barbs separated by distance 2d. Note that the increase in contact angle is ascertained by the parameter (r + d)/r only and not by the separate values of r and d. Thus, θ for values of this parameter ranging between 2.4 (penguins, Spheniscidae) and 10 (land birds) would vary between about 126^o and 154^o, roughly correct by experimental verification [15]. These values are significantly higher than those attained for the most repellent of smooth surfaces which equal about 114^o [16].

These results have been derived solely from basic physico-chemical principles without reference to any specific dimensions of the porous surface. They are determined only by the areas of feather – water and air – water interface per unit of macroscopic surface areas without dictating the shape, curvature or configuration of these interfaces. Therefore, the relationship between the dimensions of a porous surface and its ensuing contact angle is a rigorous one, not an empirical one, and is of general validity. These premises have been tested experimentally and were found to be correct by Cassie and Baxter and Rijke using paraffinated stainless steel wire cages and grids [14, 17]. Many other studies including recent ones, have reported contact angle measurements on porous substrates including feathers and consistently confirmed the correctness of the above premises [16, 17, 18, 19].

In order to measure contact angles on smooth or porous surfaces correctly, certain experimental conditions have to be met, such as: the drop has to be small enough so as not to be perturbed by gravitational forces, but large enough to cover a representative area of the porous surface. The drop should be prevented from evaporation which would turn the advancing contact angle into a receding one. Feather specimens should be covered with fresh preening oil, not rinsed with an ethanol wash [18]. When these conditions are met, the correct contact angle is usually found to be within one-degree error as observed by multiple authors [14, 17, 20, 21, 22, 23]. These results have shown conclusively that contact angles and therefore water repellencies can be reliably calculated from and represented by the dimensions of the porous surface alone.

An expression for the pressure (P), required to force water between the barbs and barbules, can be derived from similar premises and shows P to be inversely proportional to r and (r + d)/r. As a result, the requirement of relatively large values for (r + d)/r to provide sufficient water repellency is opposed by the need for small values for this parameter to attain good resistance to water penetration. Thus, the structural characteristics compatible with optimal water repellency are, at least in part, in conflict with the requirements of resistance to water penetration. This conflict has important implications for water birds, which must realize a balance between these two opposing functions to cope with their respective habitats and behavioral patterns as indeed they do [15].

Experimental data on water repellency and resistance to water penetration for Double-crested Cormorants (Phalacrocorax auritus) and Anhingas (Anhingidae) have shown that results can be satisfactorily interpreted in terms of barb diameter and spacing only without recourse to barbules. Their (r + d)/r values for barbules are in the approximate range of 4.5 to 5.5 as found for almost all bird families regardless of their feeding habits or interaction with open water [15]. This suggests that the contribution of barbules to water resistance is real, but not based on the same mechanism as applies to barbs. Barbules provide an interlocking mechanism by preventing the barbs from separating under the increasing water pressure while increasing their own separation by their hooks sliding in the flanges, a process that can be verified under a low-powered light microscope. Similarly, water drops being repelled by the barbs and not involving the barbules can be observed with a magnifying glass.

The contact angle θ of water drops on smooth feather surfaces, such as the rachis or on a microscopic slide covered with preening oil, measures about 90^o as established by various authors [14, 16, 23]. The same value was found for water drops on polyethylene foil [20] and this is no coincidence: polyethylene almost exclusively consists of methylene groups (-CH₂-) which are the predominant chemical component of preening oil [24, 25].

That the specific architecture of a porous surface does not enter into the calculations allows the investigator to determine the contact angle from the value of (r + d)/r alone. For instance, Cassie and Baxter [14] found (r + d)/r for their duck feathers to be 5.9, which corresponds to a θ of 147^o in good agreement with their experimental value of 150^o. These results, corroborated by other workers [26], have shown that for feathers coated with fresh preening oil, both the water repellency in terms of the contact angle θ and the balance between water repellency and resistance expressed by the value of (r + d)/r, can be correctly predicted from the micro-structure of the feather alone. Furthermore, the value of 5.9 for duck feathers, when compared with 4.8 for the White-breasted Cormorant (Phalacrocorax carbo) [27] and 7.1 for the European Starling (Sturnus vulgaris) [28], suggests that the duck, and probably all dabblers, are more water repellent than cormorants, but less so than starlings. On the other hand, cormorants show a superior resistance to water penetration, particularly when compared with starlings.

Measurements on more than 160 species of about 45 bird families [26, 27, 29, 30, 31] have shown that (r + d)/r values vary from about 2.3 for penguins to about 6.5 for gulls (Laridae) and up to 10 for most terrestrial birds. This range in values for this parameter suggests that each water bird family has evolved a balance between water repellency and resistance to water penetration that suits its particular habitat and behavioral pattern.

The data on barb diameter, spacing, and (r + d)/r values published in the peer-reviewed literature are far from a complete inventory of bird plumage, but on the basis of what is available, the following observations can be made and tentative conclusions reached. First, the distal one-third of breast, abdominal, and back feathers shows the patterned structure that confers the water repellency and resistance to penetration. The proximal and medial parts show no such structure. The tail feathers and remiges, on the other hand, show a pennaceous structure over essentially the entire length of the feather and have values of (r + d)/r that are generally small, which prevent these feathers from becoming waterlogged. Among water bird families, contour feathers vary more in values of (r + d)/r, which range from 2 to 10, than rectrices and remiges both of which vary little and range from 2 to 4 [14, 30, 32]. Second, within most families, the contour feathers that protect the skin from coming in contact with water have, on the whole, very similar values for (r + d)/r, exceptions seen only when a species within a family behaves differently from its relatives. A typical example is the Brown Pelican (Pelecanus occidentalis), which, unlike its congeners, dives for its prey from the air. Third, data sets on feather structure suggest a relationship between barb diameter 2r and (r + d)/r values. Families such as penguins and other diving water birds, have wide barb diameters and small values for (r + d)/r, whereas the opposite holds true for terrestrial families such as the starlings and nightjars (Caprimulgidae). Birds that come into occasional contact with open water such as herons (Ardeidae) and gulls have intermediate values. As a result, penguins have excellent resistance to water penetration but poor water repellency as shown by their familiar ‘wet’ appearance when they exit the water. The breast feathers of terrestrial birds, on the other hand, are very water repellent but promise little in the way of resistance to water penetration. Those of herons and gulls fall somewhere in between.

5. Water repellency, water resistance and behavioral patterns

The first effort to correlate the value of the parameter (r + d)/r - that is, the balance between water repellency and resistance to water penetration - with behavioral patterns was made more than 50 years ago [33]. In that paper, the well-known habit of cormorants of spreading their wings to the sun or breeze after a period in the water, a feature commonly referred to as “wing-drying,” was proposed to follow from the poor water repellency of their breast feathers, as evidenced by their low value for (r + d)/r in comparison to that for the Mallard (Anas platyrhynchos) and presumably other dabblers that do not spread their wings. Since then, further studies on the wing-drying of cormorants have overwhelmingly supported the notion that its function is the drying of contour feathers and not thermoregulation, balancing, intraspecific signaling or an aid to swallowing fish [31]. Yet the relation between the cormorant’s feather structure, specifically its parameter (r + d)/r, and this behavior has remained elusive. It is possible, though, to draw a number of conclusions from more recent data recorded with modern imaging software.

Water birds that regularly spread their wings include several species of cormorants, such as the Reed Cormorant (P. africanus), Bank Cormorant (P. neglectus), Cape Cormorant (P. capensis), White-breasted Cormorant and the Double-crested Cormorant, most of which have (r + d)/r values for their contour feathers between 4.3 and 4.9. Families and species with parameters under about 4.2, such as the divers (4.0), gannets (Sulidae) (3.8), auks (Alcidae) (3.4), penguins (2.3) and the Antarctic Blue-eyed Shag (P. atriceps) (3.8) never show wing-spreading behavior. Pelicans (Pelecanidae) (4.9 to 5.4), including the Brown Pelican (5.9), do so only very occasionally, but all other water birds do not with the notable exception of the darters (10 to 11). Darters have contour feathers that promote water to penetrate to the skin in order to reduce their buoyancy [34, 35] so their very large (r + d)/r value in unsurprising. It is reasonable to assume that, with the exception of the darters, all water birds benefit from a plumage with good water repellency and equally good resistance to water penetration. However, as we have seen, the structural requirements for these two qualities are partly opposed, so it is to be expected that each family or species will have made trade-offs and struck a balance that suits its specific demands of habitat and behavior. Spread-wing postures can then be explained as being part of a behavioral pattern in those birds that dive frequently and therefore require good resistance to water penetration, but this resistance comes at the expense of a measure of water repellency, which is compensated for by wing-drying.

The question as to whether it is the wings or the body plumage that is being dried by wing spreading was raised by Sellers [31] and can be addressed by the considering the difference in (r + d)/r values between flight and contour feathers. Values for flight feathers, in particular the outer coverts, measure 2 to 4 for both water and terrestrial birds, and these are therefore well protected from becoming waterlogged. Those for contour feathers, on the other hand, show much difference between these two groups of birds, with those of water birds that spend much time in the water and dive frequently ranging from about 2 to 4 and those of terrestrial birds ranging from about 7 to 10. Other water birds, including cormorants, have values that fall somewhere in between. Contour feathers with (r + d)/r values higher than about 4 are at risk of becoming waterlogged, which suggests that it is the exposed contour feathers rather than the flight feathers that need drying in cormorants and in darters.

Apart from (r + d)/r values, weather may also influence wing-spreading behavior. Cormorants reduce the extent to which they spread their wings with increasing wind speed, and at speeds of 4 on the Beaufort scale Sellers never saw birds to extend their wings by more than about 50%. Wind speeds may also be the reason why spread-wing postures are unknown in the Antarctic populations of the Blue-eyed Shag [29], but common in birds of this species breeding in Chile [36]. The persistent strong winds at high latitudes may well be the cause for the absence of wing-spreading behavior in the Antarctic populations.

Other than wind speed, the relative temperatures of water and air may be a factor in wing-spreading. A case in point is the Flightless Cormorant (P. harrisi) of the Galapagos, which is known to spread its stubby wings after a dive in the cool waters of the archipelago and, in this respect, behaves no different from other cormorants. However, whereas most other cormorants have contour feather with barb diameters between 48 and 54 μm and (r + d)/r values between 4.3 and 4.9, those for the Flightless Cormorant are 31–41 μm and 7.1–7.4, respectively [21]. These numbers suggest that the Flightless Cormorant suffers a measure of water penetration through the barbs of its contour feathers, a feature that is more reminiscent of darters than of cormorants. As with darters, increased water penetration is thought to assist the underwater bottom-feeding habits of P. harrisi for which too much buoyancy would prove to be a disadvantage. Simple calculations appear to support this notion: the pressure that a surface-swimming Flightless Cormorant exerts on the water ranges between 630 and 780 N m⁻², whereas only 550 to 590 N m⁻² pressure is required to force water between the barbs [37, 38]. For other cormorants, the maximum weight for no water penetration between the barbs lies well above the bird’s weight range [26]. So, unlike those of other cormorants, the Flightless Cormorant’s contour feathers become waterlogged after a dive in cold water, but the bird can then proceed to dry its plumage in the warm tropical breezes on the lava rocks, an advantage denied to cormorants inhabiting high latitudes.

6. Other behavioral patterns

In the previous section, we attributed the occurrence or absence of spread-wing postures to the need for a trade-off between water repellency and resistance as reflected in the value of the parameter (r + d)/r. It is therefore to be expected that other behavioral patterns, directly or indirectly, relate to this parameter in a similar manner. As an example of the relationship between 2r, (r + d)/r, and behavioral pattern, gannets, cormorants, and shearwaters (Procellaridae) all have about the same barb diameter (50–51 μm), but gannets have a value for (r + d)/r of 3.8, which lies at the low end of the range (3.8–4.9), indicating a greater resistance to water penetration. This may well be an adaptation to the gannet’s habit of diving from the air and then pursuing prey under water. Brown Pelicans also dive from the air, but unlike gannets do not pursue their prey under water. Their breast feathers have smaller barb diameters and higher (r + d)/r values than those of gannets, producing an increased water repellency. American White Pelicans (Pelecanus erythrorhynchos), on the other hand, find their prey while swimming on the surface and have smaller values for (r + d)/r. Apparently, plungers, divers and birds that swim underwater benefit mostly from an increased resistance to water penetration, whereas surface feeders, such as the Brown Pelican, gulls and storm petrels (Hydrobatidae), profit from an increased water repellency. Similar findings were recorded for the five species of Dippers (Cinclidae), which among them show a slightly different water repellency and resistance in their contour feathers as an adaptation to their different feeding habits and river habitats [28]. Certain species of cranes (Gruidae) and rails (Rallidae) can also be regarded as having attained structural characters in their plumage that relate to their specific interaction with their watery feeding grounds [38, 39, 40].

It is likely that many more examples of contour feather structure correlating with specific behavior/habitat will be found once more data have been gathered. However, the above examples suffice to suggest that each feather substructure represents an evolutionary adaptation to a specific set of behavioral patterns and habitat conditions. It should be borne in mind that feather structure relates in the first place to behavior and habitat and secondarily to family identity and then only to the extent that family members behave in essentially the same way and inhabit similar habitats. As we have shown, congeners with different behavior/habitat patterns show a correspondingly different value for the structural parameter. That this behavior difference occurs in conjunction with a structural difference supports the existence of a correlation between feather structure and the habitat and behavior of its avian bearer.

7. Adaptations to the impact forces of diving, plunging and alighting

Unlike terrestrial birds, water birds physically interact with water at the interface between feather coat and water. Water is about 800 times denser than air and, as a result, the impact forces of diving, plunging and alighting are so much more severe than when operating in air. It is no surprise therefore, that the feathers of water bird are composed of stiffer elements to cope with these conditions. Since each family interacts with water in its own specific way variations in feather stiffness among families are to be expected for this reason alone.

All feathers are built of beta-keratin the elastic modulus of which is an inherent property of the keratin material itself. However, the actual stiffness of the various feather elements, rachis, barbs and barbules, is determined by the respective shapes and sizes of these elements.

The mechanical forces involved in diving, plunging and alighting are not accessible to direct measurement in any reliable or representative way, at least not by current technologies. Any such data would not be meaningfully correlated to the resulting yield or flexure of barbs and vanes during forceful interaction with water. However, the bending and flexing of materials of different shapes and sizes have been well described in engineering physics and it is from these considerations that a number of conclusions can be drawn [41]. For one, the bending of the vane as a result of a force acting over its surface can be semi-quantitatively determined from the length, width and separation of the barbs, all of which can be easily measured. Specifically, the ratio of length to width of a barb can be shown to be the predominate factor in resisting bending of the vane, more so than the other contributor (r + d)/r. The deflection parameter (DF) that quantifies the bending per unit force applied to the vane is the product of this ratio and the wettability parameter (r + d)/r. For another, apart from feather stiffness, the resistance to impact forces is also determined by the extent of contour feather overlap and body feather density. The extent of overlapping can be approximated by the product of L_f, the length of the rachis, and the square root of the number of feathers per surface area. To estimate the latter, use was made of the data on number of feathers and body weights as reported by several authors [41, 42, 43, 44, 45, 46, 47]. By fitting a second-order polynomial to these data, an estimate of the number of contour feathers as a function of the mass of the bird could be obtained. For the relationship between body surface area and body mass, expressions proposed by Perez, Moye and Pritsos [48] and by Mitchell [49] were used to estimate surface area as a function of body weight. Combining the results of these two sets of calculations, contour feather densities expressed in number of feathers per surface area were found to be about 100,000 to 150,000 per m² for water birds weighing less than 1.2 kg for all families studied. This number increases with weight to 200,000/m² at about 7 kg. The extent of feather overlap, according to these calculations, yields about 10 to 15 feathers in a stack for families in the lower weight range with twice that number for heavier birds. Apparently, feather overlapping is the same for water birds in the lower weight range regardless of family identity and, as a result, the restriction that stacking provides to bending is also the same. Only for birds weighing more than 1.2 kg do we find an increase in feather density and overlap with weight – up to 250,000 per m² and stacks of 18 for a Pink-backed Pelican (P. rufescens) weighing 9.6 kg. This is in line with expectation as impact forces are directly proportional to mass [50].

The role of the barbules in resisting bending of the vane should be considered in the light of their primary function, i.e., keeping the barbs from separating under an applied force and doing so by their hooks sliding in the flanges of the barbule next more distal. Therefore, as well as for their small size, they are assumed to make only a minimal, if any, contribution to the over-all resistance to bending.

The above findings may be explained by any of two or both possibilities: 1) the feather density and number of feathers in a stack for the lower-weight families are sufficiently large to prevent feather bending regardless of behavioral pattern and 2) barb stiffness and resistance to water penetration of the contour feathers of each of these families are large enough to prevent water from reaching the skin on their own account and do not benefit from a further increase in feather density or stacking. Other than preventing water from reaching the skin, thermoregulatory adaptations can also be expected to affect feather density. Lowe [46] counted 48/cm² on a young Gentoo Penguin (Pygoscelis papua).

As mentioned earlier, the bending of the vane of a contour feather under the impact of forces associated with diving or landing on water surfaces consists of two factors: (1) the ratio of the length to the thickness of the barbs and (2) the wettability parameter (r + d)/r. The first factor indicates that short and thick barbs make the vane stiff resisting bending, whereas long and thin barbs favor flexibility that promotes bending. The appearance of the wettability parameter in the equation shows that feathers resistant to water penetration also help prevent their bending, whereas highly water-repellent feathers do not.

The deflection parameter values for 23 water bird species from 15 families have been assorted into six more or less distinct categories [50]. Deep divers, represented by four species of penguins and characterized by their habit of diving and pursuing prey under water, fall in the lowest category with a DF of 1.6 (10⁶) and therefore have the highest vane stiffness. The next category is made up of birds that swim and dive in pursuit of their prey and spend much time in and on the water. This category has a DF of 49 (10⁶) and includes the Common Diving Petrel (Pelecanoididae) and the cormorants. Category 3, into which fit the ‘true plungers’ such as petrels (Procellariidae), gannets and auks, show a DF of 194 (10⁶). Large surface feeders, such as pelicans, frigatebirds (Fregatidae) and skimmers (Rhynchopidae), form the next category with a DF of 387 (10⁶). Category 5, the shore birds, includes skuas (Stercorariidae), gulls and terns (Sternidae) that have the lowest vane stiffness with a DF of 839 (10⁶). These birds are not extended time swimmers, do not pursue their prey under water and spend much time in flight or on shore. Albatrosses, ranked as large birds of open oceans, are mostly airborne and alight only to take food from the surface or slightly below. In this respect, they behave much like category 5 families. Not listed are the Flightless Cormorant in category 2 and the Brown Pelican in category 4, because, as mentioned above, these species feed in a different way from their congeners, a feature expressed in the dimensions of their feather structure.

From these data can be concluded that the contour feathers of penguins, the most aquatic of families, are about thirty times stiffer than those of diving petrels and cormorants, and 120 times more so than those of plungers like gannets. Similarly, penguin feathers are 250 times more resistant to bending than those of surface feeders like pelicans, over 500 times more so than those of shorebirds such as skuas, gulls and terns and almost three orders of magnitude stiffer than albatross feathers. These large differences are directly related to feeding habits and interaction with water. Penguins find their prey exclusively under water and dive to great depths to catch it. Diving petrels and cormorants also dive, but spend more time on the surface and in the air. Plungers dive from the air with associated high pressure on impact, but catch their prey at lesser depths. Surface feeders do not dive and do not pursue their prey under water (brown pelicans dive from the air, but do not pursue under water). Shore birds feed from the water surface and are not extended time swimmers. Albatrosses, one of the most aerial of seabirds, alight only to feed from the surface and may occasionally dive at feeding frenzies.

The following pattern of feather structure in relation to feeding habits/behavior emerges:

Barb width and spacing determine the relative water repellency and resistance to water penetration of feathers. Diving birds, and in particular deep diving birds, benefit from a mostly water resistant plumage with little in the way of water repellency. Less aquatic families, such as gannets and to a greater extent cormorants, show more repellency, but at the expense of some of their water resistance. Some cormorants compensate for this by their habit of wing spreading. Swimming and hovering birds that catch their prey from the surface, shore birds and those operating mostly in the skies show a predominantly water repellent plumage.

The length and diameter of the barbs of contour feathers vary widely among water birds. Barb stiffness varies with barb length and width and is largest for deep diving birds, less so plungers and very much less so for surface feeders ranging over three orders of magnitude. These structural differences in the feather plumage are thought to represent evolutionary adaptations to feeding habits and, in some cases, environmental conditions.

8. Viscous drag and feather geometry

Birds that swim and dive will also benefit from reduced drag for their locomotion in water, a consideration that applies less to waders and shore birds and not at all to land birds. Viscous drag in water is dependent on the surface microstructure of the distal one-third of the contour feather, but also on the shape of its surface in contact with water, an aspect of feathers that has so far received little or no attention. Drag in air, such as in flight, on the other hand, has been the topic of several studies.

That the shape of the surface area in contact with water varies among bird families has been noticed in the course of previous studies. It was seen to be nearly circular in land birds with a length-to-width ratio (L/W) of approximately 1.0, but oblong with an L/W of about 4 in penguins, the most aquatic of families. Birds less intimate with open water showed intermediate values for L/W.

Here, we consider the interface between the distal one-third and flowing water to calculate viscous drag for feather shape geometry. Assuming the flow to be parallel to the long axis of the feather, i. e. zero angle of attack, one can derive the total drag coefficient, composed of viscous pressure and frictional drag, from the computational and experimental results of studies on model ship hulls of varying length-to-diameter ratios using solutions to the Reynolds-averaged Navier-Stokes equations [51]. The drag coefficient then follows from the value of L/W for values of L/W less than 7 which is within the range of feather geometry. The relationship between drag coefficient and L/W predicts that oblong shapes of the tips of contour feathers reduce drag facilitating swimming and diving, whereas a more circular shape would cause an increase in frictional drag. A similar reasoning could be applied to the shape of the area that the body of a swimming bird has in contact with water. If this area is assumed to be elliptical, a drag coefficient for body surface area in contact with water can be determined using the same considerations.

Grouping bird species according to their interaction with open water can be achieved by assigning them to foraging niches as proposed by Pigot et al. [52], applying a standardized protocol for foraging niche delimitation. Using this procedure, a total of thirty niches has been identified for all of the approximately 10,000 bird species of the world. Of these six major foraging niches were categorized as Aquatic.

Aquatic birds can be subdivided into swimmers and waders. Their values for the drag coefficient show a viscous drag for swimmers significantly lower than that of waders and, predictably, land birds. These categories, when further subdivided into eight aquatic foraging niches and, for comparison, two terrestrial ones, show that divers have the lowest recorded drag coefficient increasing in order for plungers, surface feeders, aerials, herbivore surface feeders, ground feeders, perchers to herbivore ground feeders. Land birds experience an even higher drag with no significant difference between ground feeders and those that catch their prey by aerial or sally sorties.

Adaptations in feather microstructure and body surface area in contact with water that bring about a reduction in viscous and frictional drag while swimming increase according to the bird’s intimacy with open water. Swimming and diving birds, such as penguins and grebes, benefit the most from reduced viscous drag, more so than plungers such as gannets. Aerials such as terns even less so, but much more than herbivore surface feeders such as ducks. The body feathers of ducks, in turn, appear to be better adapted to their watery habitat than those of aquatic ground feeders such as herons or kingfishers. The drag coefficient of the Spotted Dikkop (Burhinus capensis), an herbivore aquatic ground feeder of open scrubby habitat with comparatively little interaction with open water, was found to be more in line with those of land birds in which adaptations to locomotion in water are not expected to have evolved.

Land birds do not only show drag coefficients higher than those of water birds, they also show no significant difference among the two foraging niches examined so far. This is in line with expectation as their lack of interaction with open water and their locomotion in air only suggest that forces that foster reduced drag in water have been absent in their evolutionary history.