Spread-Wing Postures and the Water Repellency of Feathers: A Test of Rijke's Hypothesis

Abstract

-Rijke (1967, 1968) proposed that the water repellency of feathers and the presence or absence of spread-wing postures in water birds could be explained by a structural mechanism first described for textiles. The textile model predicts that the tendency of water droplets to bead up on grid-like surfaces is a mathematical function in which the primary independent variable is an index, (r + d)Ir, where r is the radius of and d is one-half the distance between cylinders in the grid. Larger indices indicate more surfaces. Rijke found larger indices in the feathers of ducks than in the feathers of cormorants and anhingas; hence, he concluded that the latter birds must spread-wing to dry their wettable feathers. However, there were mathematical inconsistencies, undefined variables and concepts, and inadequate data in Rijke's papers. Despite these flaws, Rijke's hypothesis has been frequently cited in the ornithological literature. I report here my evaluation of the applicability of the textile model to feathers and my test of Rijke's prediction that species that assume spread-wing postures when wet have smaller (r + d)/r values for their ramus and barbule structure than species that do not. I used scanning electron microscopy to measure the feather structure of 14 species of water birds in 6 different categories (breast, back, and four regions of a remex). I found that the textile-feather analogy is not realistic, because feather structure is considerably more complex and variable than the geometric model that is fundamental to the textile equations. My (r + d)/r values show considerable overlap among three behaviorally distinct groups of water birds: those that predictably, occasionally, or never assume spread-wing postures. Statistically, the (r + d)/r values of the rami in some feather categories of the group of species that shows spread-wing behavior were smaller than those of the other two groups of birds (which did not differ). Index values of the barbule structure, which constitutes most of the feather surface, however, do not differ significantly among the three groups of birds. I also measured the shape of water droplets (by contact angles) on the breast and remex feathers of a Mallard (Anas platyrhynchos) and a Reed Cormorant (Phalacrocorax africanus) and compared these values between the two species as well as with those mathematically predicted by the textile model. In general, the observed water droplets have a shape more like that predicted by the (r + d)/r values of barbules than of rami. Droplets on the feathers of the Reed Cormorant were more bead-shaped than those on Mallard feathers, although the reverse should be true if the textile model holds for feathers. I conclude that Rijke's hypothesis is invalid for two reasons: the textile model cannot be applied reliably to feathers, and it does not account for the spread-wing behavioral differences among water birds. Received 12 April 1983, accepted 12 October 1983. RIJKE (1967, 1968) published two nearly identical papers in which he proposed that a mathematical model derived to explain the water repellency of textiles could also resolve three ornithological issues: (1) the water repellency of feathers; (2) the absence of spread-wing postures in most species of water birds; and (3) the function of this behavior in cormorants and anhingas. Rijke's model provided a plausible answer to the paradox that Townsend (in Bent 1922: 241) pointed out: if spread-wing postures are necessary to dry the wings, why is the behavior not shown by water birds other than cormorants and anhingas? Hailman (1969a, b), however, reviewed Rijke's papers (1967, 1968; hereinafter reference to Rijke refers to these years unless other dates are given) and pointed out several problems, most notably, several mathematical inconsistencies, an undefined variable, and an unsupported assumption. Despite these drawbacks, Rijke's conclusions have been generally accepted in the ornithological literature (see below). Hailman recommended that I look into Rijke's papers when I was preparing a note on my field observation of an Osprey (Pandion haliaetus) performing spread-wing behavior (ElowsonHaley 1982). In doing so, I found several more 371 The Auk 101: 371-383. April 1984 This content downloaded from 207.46.13.21 on Tue, 27 Sep 2016 05:12:58 UTC All use subject to http://about.jstor.org/terms 372 A. M. ELOWSON [Auk, Vol. 101 inconsistencies (detailed below) that prompted me to reevaluate the hypothesis with respect to issues (1) and(2) above. I present the results of that study in this paper. I have not considered here the function of spread-wing postures, which is a complex question that has been investigated elsewhere (Bernstein and Maxson 1982, Hennemann 1982, Winkler 1983, and the references therein). Rijke based his hypothesis on a resemblance of feather structure to textile structure. The textile model (Cassie and Baxter 1944, Baxter and Cassie 1945) states that the water repellency of porous surfaces increases with the size of an index that is calculated from the dimensions of elements in the surface structure (i.e. the threads). Cassie and Baxter (1944) even suggested the applicability of their model to feathers, which may have motivated Rijke to test the idea. Using light microscopy, he measured the feather structure of seven species and found larger indices in ducks than in cormorants. Therefore, he concluded that cormorants, unlike ducks, have wettable feathers that require drying by spread-wing postures. Rijke's work has been noted in the literature in two somewhat different contexts: feather structure determines its water repellency and the feathers of cormorants and anhingas are wettable. Some authors have cited the original textile papers (Cassie and Baxter 1944, Baxter and Cassie 1945) as having established the first concept (Thompson 1953; Kennedy 1970a, b, 1972; Rutschke 1976). Rutschke (1960) took measurements from mallard feathers and concurred with the estimates Cassie and Baxter (1944) gave. Many more authors, however, have cited Rijke's conclusions with respect to feather structure and water repellency (Clark 1969; Kennedy 1970b; Stettenheim 1972, 1976; Rutschke 1976; Rhijn 1977; Schreiber 1977; Jones 1978; Winkler 1983). Bernstein and Maxson (1982) discussed Rijke's conclusions in both contexts. Although they measured the feather elements of the Antarctic Blue-eyed Shag (Phalacrocorax atriceps) to contrast with Rijke's data for other species of cormorants, they noted Hailman's (1969a, b) objections to the model. Without reference to the underlying process, McAtee and Stoddard (1945) and Owre (1967) proposed that cormorants and anhingas have a wettable plumage or feather coat (nomenclature from Humphrey and Parkes 1959; the feather coat is the aggregate of feathers worn a. 9 90g Fig. 1. Contact angles of water droplets resting on (a) wettable and (b) surfaces. by a bird at any given time). Despite the fact that Rijke did not evaluate the wettability of the feather coat as a whole, several authors have cited him as providing a mechanism that makes this suggestion plausible (Clark 1969, Kennedy 1971, Kahl 1971, George and Casler 1972, Siegfried et al. 1975, Mahoney 1981, Bernstein and Maxson 1982, Hennemann 1982). Finally, Kennedy (1969) has cited Rijke's work as having established a drying function for the spreadwing postures of cormorants and anhingas. Neither Rijke's hypothesis nor my test of it is comprehensible without some explanation of the textile model itself. In what follows, I present that model followed by Rijke's data and finally the predictions that should hold if the hypothesis were true. The textile model and feathers.-The distinction between water-repellent and surfaces needs clarification, as authors (Rowen and Gagliardi 1947, Crisp 1963) have pointed out multiple uses of the latter term. Truly waterproof materials, such as a yellow rain slicker, are almost impermeable to water. Water-repellent surfaces, by contrast, cause water to bead up and roll off under brief exposures in ordinary atmospheric conditions, but such surfaces will become wetted upon extended exposure or under increased pressure. Denim fabrics and feathers are surfaces, and Rijke's model deals with repellency per se. The textile model (Cassie and Baxter 1944, Baxter and Cassie 1945) is considered to be the state of the art for the water repellency of porous surfaces (Crisp 1963). The water repellency of a surface is determined by whether water droplets on it bead up (repellent) or flatten and spread out (wettable) (Fig. 1). Thus, surface wettability can be operationally expressed as the angle made by the surface and a tangent to a droplet's curvature at the point of contact, This content downloaded from 207.46.13.21 on Tue, 27 Sep 2016 05:12:58 UTC All use subject to http://about.jstor.org/terms April 1984] Water Repellency of Feathers 373