Biomechanics and ecological speciation

As ecological speciation progresses, the adaptive divergence of phenotypic traits can presumably strengthen any number of isolating barriers, both pre- and postmating. For the purposes of this chapter, we focus our attention on the role of biomechanics in premating isolation, and specifically in relation to mating displays. Mating displays include ritualized movements such as visual or vocal signaling, or the presentation of static traits such as exaggerated morphological characters [36]. Many dynamic displays, particularly those under intense sexual selection, appear to be costly or to require high levels of biomechanical proficiency in their production. Indeed, a number of studies have shown that dynamic mating displays require large energy investments [37-40] or are underpinned by adaptations for rapid neuromus-cular output [41-43]. High costs or high levels of required proficiency help to ensure that dynamic displays are "honest," because such displays tend to provide a reliable indication of a male's genetic or phenotypic quality [44-46].

The crux of the argument we develop here is that adaptive divergence in phe-notypic traits (morphology, physiology, or behavior) may influence, as a secondary consequence, the nature or strength of biomechanical constraints on mating displays. Numerous examples are discussed. Insofar as displays are costly or challenging to produce, even minor divergence in biomechanical systems could influence an animal's ability to produce these displays. This divergence could potentially influence reproductive isolation. Four lines of evidence would ideally be gathered to demonstrate that adaptive divergence influences mating displays and therefore speciation. First, a trait related to biomechanical performance should be shown to diverge adaptively among populations and species. Second, the corresponding variation in performance should be shown to influence mating displays. Third, variation in these displays should be shown to influence mate choice. Fourth, the resulting mate choice should be shown to influence speciation. No study has yet systematically examined these criteria for a single taxon, although the conceptual relationships between divergence, signal variation, and speciation have been considered previously at length [e.g., 11,47,48]. We now review three broad classes of biological adaptations — in body size, locomotion, and feeding — that may affect, as a secondary consequence, the biomechanical bases and expression of mating displays.

14.3.1 Mating Displays and Body Size

Body size evolves in response to a wide array of environmental factors. Cold temperatures, for example, tend to favor larger body sizes, as illustrated by Berg-mann's rule [larger body sizes at higher latitudes; 49-51]. In homeothermic animals, this trend might arise because of a positive relationship between body size and the ability to retain metabolic heat [52]. Large animals may also be favored in highly seasonal or unpredictable environments because they are comparatively resistant to starvation [53]. Moreover, body size tends to evolve in response to varying selection on life-history traits such as fecundity, reproductive rate, and dispersal. For example, selection for increased fecundity favors comparatively large body sizes, whereas selection for rapid offspring production often favors small body sizes [54,55].

Body size influences myriad aspects of organismal physiology and biomechanics [52,56]. Traits involved in communication are no exception. The maximum size of ornaments used in visual signaling is necessarily limited by body size [36,57]. Peacock tails or deer antlers, for example, are constrained to sizes and masses that can be effectively carried and displayed. Body size also shapes acoustic signals because of positive scaling between body size and the mass of acoustic source tissues [36,58]. Darwin's finches of the Galápagos Islands, for example, show a positive, nearly isometric relationship between body mass and syrinx (sound source) volume [59], and larger-bodied finches tend to sing at correspondingly lower vocal frequencies [60]. Similar relationships between vocal frequency and body size have been demonstrated in numerous taxa, especially anurans and birds [e.g., 61-66].

Body size influences how animals are able to execute visual and acoustic displays because of tradeoffs between body size and agility. Size-agility tradeoffs have been documented within some birds and butterflies. In these groups, the frequency, duration, and even the effectiveness of male aerial displays tend to be highest in species or individuals with the smallest body sizes [67-71]. This pattern is consistent with demonstrated negative impacts of body size on flight agility [67]. Negative impacts of body size on display production may help to explain selection for small body size ("reversed" sexual dimorphism) in species of birds and insects where males use aerial displays [68,69,72]. Body size also influences electric organ discharges (EODs) in electric fishes. Larger-bodied fishes can support larger populations of electrocytes, thus augmenting EOD intensity, and also express greater charge separation distances in their electrogenic organs, thus enhancing EOD range [36].

The first two criteria in support of the by-product speciation model are thus clearly met — body size has been shown to undergo adaptive divergence through natural selection, and body size variation can influence the expression of mating displays. Many lines of available evidence also support a role for body size in mate choice. In some taxa, females have been shown to express preferences for males with body sizes similar to their own [73,74]. In other taxa, females express general preferences for larger males, although intrasexual competition can limit female access to larger males and thus result in patterns of size-assortative mating [75]. Because body size is often highly correlated with aspects of behavior and courtship, which in turn provide proximate cues in mate choice, divergent selection on body size would seem to be relatively effective in promoting assortative mating [e.g., 76].

14.3.2 Mating Displays and Locomotion

Complex and highly specialized adaptations for locomotion are prevalent throughout the animal kingdom, and often entail substantial modification of broad suites of traits [77]. In terrestrial vertebrates, rapid sprint speeds are enabled by adaptations in limb length, aerobic capacity, and efficiency of pulmonary gas exchange [78]. In fishes that use their caudal fin for routine propulsion, sustained swimming is typically associated with fusiform bodies and high aspect ratio lunate tails, whereas burst swimming is typically associated with deep bodies and large fins, particularly in the caudal area [79]. In aerial vertebrates, powered (flapping) flight requires numerous adaptations including reduced body weight, aerodynamic body shape, broad lift surfaces, and efficient flight muscles [80]. In humans, selection for endurance running may have favored a broad suite of traits including springlike leg tendons, skeletal stabilization, plantar arches, forearm shortening, and expanded venous circulation for thermoregulation [81].

The ecological bases of locomotory adaptation are perhaps best studied through comparison of closely related species or populations. Many studies could be cited to this effect [82]; here we provide two representative examples. The first concerns Anolis sagrei, a lizard found throughout the Caribbean. In the late 1970s and early 1980s, this species was introduced to islands that contained no lizards. Ten to fourteen years later, the introduced populations were sampled and found to have undergone substantial divergence in hind- and forelimb length [83]. Moreover, these changes correlated positively with the mean diameter of available perches on the experimental islands, consistent with functional studies of limb length and locomotor efficiency [83,84]. The second example concerns Gambusia affinis, the western mosquitofish. Langerhans et al. [85] found that mosquitofish populations under high risk of predation have evolved comparatively large caudal regions, small heads, and elongate bodies, all of which are thought to improve escape ability. Interestingly, these "fast-start" adaptations may impair prolonged swimming ability, which could explain the retention of the opposing suite of traits in low-risk populations [see also 86].

Adaptive divergence in locomotory traits might, in turn, influence mating displays, given that displays often include, and sometimes even amplify, motor patterns used during normal locomotion. Courting displays in waterfowl, for example, include wing flapping, swimming, and changes in head posture similar to those that occur before flight [87,88]. Some other displays, such as courtship flights of hummingbirds and "strut" displays of grouse, are dominated by locomotion [89,90]. Indeed, a major preoccupation in early ethology was to explain ritualization, the process wherein common locomotory patterns become incorporated into stereotyped display sequences [91,92]. Beyond providing raw material for display patterns, selection for locomotory traits may also fine-tune animals' abilities to perform mating displays. The evolution of complex hummingbird flight displays, for instance, was presumably facilitated by selection for agile flight capabilities in other contexts, such as for food and territory defense. Another possible example concerns crickets and other ortho-pterans that produce acoustic signals through stridulation of the wings. The divergence of flight anatomy and biomechanics (e.g., wing size, flight muscle properties) presumably could influence the kinds of acoustic signals these animals produce and evolve.

Operationally it can be very difficult to study biomechanical impacts of locomotion on dynamic displays, simply because it is difficult to quantify the kinetics and dynamics of display movements in an animal that itself is moving through space [e.g., 89]. It is thus no surprise that most studies of display biomechanics have focused on animals that signal while stationary. An alternative approach is to study the biomechanical bases of multimodal signals, i.e., signals that involve multiple sensory channels. In a recent study of brown-headed cowbirds, for instance, Cooper and Goller [93] studied mating displays that feature simultaneous vocal output and wing movements. Analysis of dynamic changes in air sac pressure, wing movements, and vocal features provide evidence for a biomechanical interaction between wing movements and vocal displays. Specifically, wing position appears to constrain the timing of vocal output via biomechanical influences on the respiratory system [93]. Similar interactions between wing movements and breathing could presumably influence the evolution of vocal signals produced during flight [see also 94,95].

As in the previous section, the first two criteria for biomechanically driven ecological speciation are well supported. Morphological and physiological parameters certainly diverge adaptively through natural selection, and variation in loco-motory performance certainly affects the expression of mating displays. There are few data, however, that directly support a link in any given system between loco-motory adaptations, resulting divergence in mating displays, and mate choice. A promising model system on this front is the threespine stickleback, for which differences among sympatric morphs in body size and behavior are quite pronounced. While some attention has been given to causes and mating consequences of variation in body size in sticklebacks [73], less is known about intermorph differences in swimming performance, or about how such differences might affect mating displays and patterns. The intricacy and complexity of mating displays in this species, which has captivated behavioral biologists since Tinbergen [92], increases the likelihood that intermorph differences in display performance would be influenced by divergent selection on swimming performance.

14.3.3 Mating Displays and Feeding

Animals have evolved a wide range of morphological and behavioral adaptations for feeding [e.g., 96-99]. Fishes, for example, employ an impressive diversity of feeding modes including suction feeding, ram feeding, and prey capture through jaw protrusion [100-103]. Studies of variation within and among closely related species illustrate how ecological conditions may promote adaptive divergence in these traits and behaviors. Variation within a species in preferred prey (i.e., "resource polymorphisms"), for example, is sometimes mirrored by genetically based variation in feeding morphology, which in turn may provide the raw material for incipient speciation [3,104; but see 105]. The link between adaptation to alternative food resources and speciation is indeed evident in many classic adaptive radiations, including fishes in postglacial temperate regions [3], African cichlids [106,107], Galápagos finches [6,108], and Hawaiian honeycreepers [109].

In a majority of cases, feeding adaptations likely have little proximate impact on the biomechanics of display production. This is because the two functions often show little if any overlap in their mechanical and anatomical bases. This is certainly true for many familiar displays, such as plumage or color pattern in birds and fishes. In some taxa, however, feeding and mating adaptations make use of the same morphological structures. When they do, feeding and display functions may interact on both organismal and evolutionary scales. In an intriguing example, male giraffes use their long necks not only for foraging on high branches but also as weapons during intrasexual competition for females [110]. Feeding and display functions may sometimes oppose each other in biomechanical function. Male fiddler crabs, for example, use their claws during feeding and during displays to females; the small and agile claws are best for feeding and the large and conspicuous claws most useful for display. In response to this tradeoff, fiddler crabs have "assigned" each function to a different claw [111]. In other cases, however, feeding and display functions do not involve redundant structures, and morphological or biomechanical tradeoffs cannot be circumvented. One such case, on which we now focus, concerns overlap between feeding adaptations and mechanisms of vocal production in songbirds.

A primary feature in the radiation of songbirds is the exploitation of divergent feeding niches through divergence in the size, form, and function of beaks [108,109,112]. This divergence likely affects the evolution of vocal mating signals, i.e., songs, because of the recently identified contribution of beaks to vocal mechanics [113-115]. One prediction is that divergence in beak and vocal tract volume, and thus in vocal tract resonance properties, should affect the evolution of vocal frequencies. This is because larger-volume vocal tracts are best suited for low-frequency sounds, whereas small-volume vocal tracts are best suited for high frequency sounds [113,116]. In support of this prediction, fundamental frequency has been shown to vary negatively with beak length in neotropical woodcreepers [65]. Another prediction is that the evolution of increased force application, such as that required to crack larger and harder seeds, may detract from a bird's vocal performance capabilities [117]. Force-speed tradeoffs are a common feature of mechanical systems, and can be attributed to both biomechanical and muscular properties [118,119]. In the evolution of some kinds of "superfast" muscles, such as those used for sound production in the toadfish swimbladder, elevated rates of crossbridge detachment during contraction necessarily preclude strong force application [119]. The evolution of bite force in granivorous birds is expected to affect the rapidity with which they can adjust beak gape, with increases in bite force diminishing maximum rates of gape adjustment, and vice versa.

Of particular relevance to this latter prediction is the expression of song features that rely on changes in beak gape in their production. Gape changes are tightly correlated with changes in fundamental frequency [e.g., 120-124] and with the resonance function of the vocal tract filter [115,116]. Tradeoffs between beak gape speed and force, either at the level of jaw muscles or force transmission mechanics, could thus impede the evolution of high-performance songs, especially for strong biters. Recognition of this relationship suggests that two song features in particular, trill rate and frequency bandwidth, should be influenced by beak size evolution because the production of these features requires rapid beak gape cycling [125]. Some (but not all) available data support this prediction [126-130]. The nature of this relationship is illustrated in an ongoing study of a population of medium ground finches, Geospiza fortis, at El Garrapatero on Santa Cruz Island, Galápagos. This population shows a bimodal distribution of small and large beak sizes, with few intermediates [130] (A.P. Hendry et al., unpublished data). In this population, morphological variation is correlated closely with bite force capacities and with the frequency bandwidth of song, in directions predicted by biomechanical models of beak and vocal tract function (Figure 14.2) [130,131].

Large beak morph:

jmas* iMr

FIGURE 14.2 Sound spectrograms from 12 male medium ground finches (Geospizafortis) from El Garrapatero, Santa Cruz Island, Galápagos. Songs of large and small beak morphs differ significantly in frequency bandwidth, with the smaller beaked birds producing wider frequency bandwidth (and thus higher performance) songs. This is consistent with a hypothesis of vocal tract constraints on song production and evolution in these birds [130]. (Figure modified from Huber, S.K. and Podos, J., Biol. J. Linn. Soc., in press.)

FIGURE 14.2 Sound spectrograms from 12 male medium ground finches (Geospizafortis) from El Garrapatero, Santa Cruz Island, Galápagos. Songs of large and small beak morphs differ significantly in frequency bandwidth, with the smaller beaked birds producing wider frequency bandwidth (and thus higher performance) songs. This is consistent with a hypothesis of vocal tract constraints on song production and evolution in these birds [130]. (Figure modified from Huber, S.K. and Podos, J., Biol. J. Linn. Soc., in press.)

For Darwin's finches, three of the four criteria for demonstrating the role of biomechanics in ecological speciation are well met, which to our eye gives this system excellent potential in testing the model of speciation presented in this chapter. First, the adaptive divergence of feeding morphology has been linked to variation in local ecological conditions, with a degree of precision rarely achieved in natural systems [112,132-134]. Second, beak divergence has been shown to influence some fundamental song features, such as trill rate and frequency bandwidth [121,128,130]. Third, female choice is highly dependent on song as a mating signal [135-137]. It remains to be seen, however, whether those song features mechanically linked to variation in beak form and function also have a significant impact on mate and species recognition. Playback studies offer one method for addressing this question empirically. If such a link is present, it follows that beak adaptation could affect the evolution of reproduction isolation in these birds.

0 0

Post a comment