The biology of gliding in flying lizards (genus Draco) and their fossil and extant analogs

The flying lizards of the genus Draco are among the most remarkable and successful clades of gliding vertebrates. Here, we evaluate the evolution of gliding in Draco and other lizards, describe the suite of morphological innovations that characterize Draco, discuss the ecological context of gliding in this genus, describe functions of their patagial membranes that are not related to gliding, and summarize the interspecific allometry of the Draco gliding apparatus, as well as the corresponding consequences for their now empirically quantified gliding performance. Several fossil reptilian lineages had morphologies similar to that of modern Draco, with patagial membranes supported by elongated ribs or rib-like dermal structures. Using Draco's snout-vent length/mass relationships, we provide improved estimates of wing loading for three of these fossil gliders (Icarosaurus seifkeri, Kuehneosaurus sp., Coelurosauravus elivensis) and then estimate absolute gliding performance for each taxon by extrapolating from Draco's wing loading/glide performance relationship. We find that I. seifkeri likely represented the best nonflapping terrestrial vertebrate glider yet described, whereas the larger Kuehneosaurus and Coelurosauravus probably required high descent velocities to achieve sufficient lift for gliding, with commensurately greater height loss with each glide.     

Wing design and scaling of flying fish with regard to flight performance

Fin and body dimensions of six genera of flying fish (Exocoetidae) were examined to study variation in morphological parameters in relation to aerodynamics performance. The fins are modified as wings for gliding flight. Fin area and fin span increase with increasing body mass, whereas the percentage of wing area contributed by the pectoral fins and the percentage of the caudal fin area contributed by the hypocaudal lobe remain constant. The aerodynamic design of flying fish approximates the monoplane-biplane classification proposed by Breder (1930). Scaling relationships for wing loading and aspect ratio indicate that wing morphology in the Exocoetidae is more similar to birds and bats than to other gliders. The flight performance of flying fish is a high-speed glide with a relatively flat trajectory. The wing, as indicated by the aspect ratio, is designed for high lift with low drag characteristics.     

Gliding flight in Chrysopelea: turning a snake into a wing

Although many cylindrical animals swim through water, flying snakes of the genus Chrysopelea are the only limbless animals that glide through air. Despite a lack of limbs, these snakes can actively launch by jumping, maintain a stable glide path without obvious control surfaces, maneuver, and safely land without injury. Jumping takeoffs employ vertically looped kinematics that seem to be different than any other behavior in limbless vertebrates, and their presence in a closely related genus suggests that gap-crossing may have been a behavioral precursor to the evolution of gliding in snakes. Change in shape of the body by dorsoventral flattening and high-amplitude aerial undulation comprise two key features of snakes' gliding behavior. As the snake becomes airborne, the body flattens sequentially from head to vent, forming a cross-sectional shape that is roughly triangular, with a flat surface and lateral "lips" that protrude ventrally on each side of the body; these may diminish toward the vent. This shape likely provides the snake with lift coefficients that peak at high angles of attack and gentle stall characteristics. A glide trajectory is initiated with the snake falling at a steep angle. As the snake rotates in the pitch axis, it forms a wide "S" shape and begins undulating in a complex three-dimensional pattern, with the body angled upward relative to the glide path. The head moves side-to-side, sending traveling waves posteriorly toward the tail, while the body (most prominently, the posterior end) oscillates in the vertical axis. These active movements while gliding are substantially different and more dynamic than those used by any other animal glider. As the snake gains forward speed, the glide path becomes less steep, reaching minimally recorded glide angles of 13°. In general, smaller snakes appear to be more proficient gliders. Chrysopelea paradisi can also maneuver and land either on the ground or on vegetation, but these locomotor behaviors have not been studied in detail. Future work aims to understand the mechanisms of production and control of force in takeoff, gliding, and landing, and to identify the musculoskeletal adaptations that enable this unique form of locomotion.     

Relationship between body size and flying-related structures in Neotropical social wasps (Polistinae,Vespidae, Hymenoptera)

Body size influences wing shape and associated muscles in flying animals which is a conspicuous phenomenon in insects, given their wide range in body size. Despite the significance of this, to date, no detailed study has been conducted across a group of species with similar biology allowing a look at specific relationship between body size and flying structures. Neotropical social vespids are a model group to study this problem as they are strong predators that rely heavily on flight while exhibiting a wide range in body size. In this paper we describe the variation in both wing shape, as wing planform, and mesosoma muscle size along the body size gradient of the Neotropical social wasps and discuss the potential factors affecting these changes. Analyses of 56 species were conducted using geometric morphometrics for the wings and lineal morphometrics for the body; independent contrast method regressions were used to correct for the phylogenetic effect. Smaller vespid species exhibit rounded wings, veins that are more concentrated in the proximal region, larger stigmata and the mesosoma is proportionally larger than in larger species. Meanwhile, larger species have more elongated wings, more distally extended venation, smaller stigmata and a proportionally smaller mesosoma. The differences in wing shape and other traits could be related to differences in flight demands caused by smaller and larger body sizes. Species around the extremes of body size distribution may invest more in flight muscle mass than species of intermediate sizes.     

A new species of Rhabdias (Nematoda: Rhabdiasidae) from agamid lizards on Luzon Island, Philippines

Rhabdias odilebaini n. sp. is described on the basis of specimens found in the lungs of 2 species of agamid lizards: the Philippine flying lizard Draco spilopterus and the marbled bloodsucker Bronchocela marmorata . Specimens were collected in Aurora Province, Luzon Island, Philippines. The new species of Rhabdias is characterized by presence of 4 submedian lips, inconspicuous lateral lips, rounded cross-shaped oral opening, and tail end bent dorsally. This species is morphologically distinct from other Rhabdias spp. that parasitize reptilian and amphibian hosts, including 3 other species known to parasitize lizards of the Agamidae.     

Correlates of sprinting, jumping and parachuting performance in the butterfly lizard, Leiolepis belliani

Leiolepis belliani , a cursorial, beach-dwelling lizard, moves by running and jumping. The lizards' ability to flatten dorsoventrally, thereby increasing surface area and decreasing wing loading, may also confer parachuting ability. We measured locomotor performance of three ecologically relevant tasks: running, jumping and parachuting. In addition, we investigated whether, with the effect of size removed, locomotor performance capabilities are correlated, and whether they correlate with morphological features. Larger lizards fell and ran faster and jumped further. Lizards that were experimentally prevented from flattening fell faster than control lizards. When the effects of size were removed, limb length was uncorrelated with jumping and running performance; performance measures also were not correlated amongst themselves. The scant natural historical data available for this species suggests that lizards do not use their parachuting capability, and that dorsoventral flattening may have evolved for some other purpose. Leiolepis might serve as a useful model in understanding the evolution of gliding lizards (e.g. Draco ).     

The effect of size on the optimal shapes of gliding insects and seeds

Conventional aerodynamic arguments suggest that possession of high aspect ratio wings will always improve the flight performance of glides. Drag and power will be minimized at intermediate flight speeds. It is shown, however, that as the aspect ratio increases, these minimum drag speeds are reduced, and will fall below the stall speed of the glider. This will happen at lower aspect ratios in small gliders, which operate at higher profile drag coefficients. Increasing the aspect ratio further will improve performance less than this analysis suggests.
A detailed analysis is developed to calculate the optimum shape of small gliders. Profile drag increases with aspect ratio, owing to the fall in the Reynolds number, while induced drag falls with increasing aspect ratio. Minimum drag will be encountered and hence the glide angle will be minimized at intermediate values of aspect ratio. Best glide angles are achieved at low speeds (high lift coefficients) and the optimum aspect ratio increases with the mass of the glider.
Small natural gliders possess large, low aspect ratio wings. The aspect ratios are generally somewhat below those which would produce the best glide angle at stall speed, but should give a reasonable performance over a range of speeds.     

A wing-assisted running robot and implications for avian flight evolution

DASH+Wings is a small hexapedal winged robot that uses flapping wings to increase its locomotion capabilities. To examine the effects of flapping wings, multiple experimental controls for the same locomotor platform are provided by wing removal, by the use of inertially similar lateral spars, and by passive rather than actively flapping wings. We used accelerometers and high-speed cameras to measure the performance of this hybrid robot in both horizontal running and while ascending inclines. To examine consequences of wing flapping for aerial performance, we measured lift and drag forces on the robot at constant airspeeds and body orientations in a wind tunnel; we also determined equilibrium glide performance in free flight. The addition of flapping wings increased the maximum horizontal running speed from 0.68 to 1.29 m s?1, and also increased the maximum incline angle of ascent from 5.6° to 16.9°. Free flight measurements show a decrease of 10.3° in equilibrium glide slope between the flapping and gliding robot. In air, flapping improved the mean lift:drag ratio of the robot compared to gliding at all measured body orientations and airspeeds. Low-amplitude wing flapping thus provides advantages in both cursorial and aerial locomotion. We note that current support for the diverse theories of avian flight origins derive from limited fossil evidence, the adult behavior of extant flying birds, and developmental stages of already volant taxa. By contrast, addition of wings to a cursorial robot allows direct evaluation of the consequences of wing flapping for locomotor performance in both running and flying.     

Shake rattle and roll: the bony labyrinth and aerial descent in squamates

Controlled aerial descent has evolved many times independently in vertebrates. Squamates (lizards and snakes) are unusual in that respect due to the large number of independent origins of the evolution of this behavior. Although some squamates such as flying geckos of the genus Ptychozoon and the flying dragons of the genus Draco show obvious adaptations including skin flaps or enlarged ribs allowing them to increase their surface area and slow down their descent, many others appear unspecialized. Yet, specializations can be expected at the level of the sensory and neural systems allowing animals to maintain stability during controlled aerial descent. The vestibular system is a likely candidate given that it is an acceleration detector and is well-suited to detect changes in pitch, roll and yaw. Here we use conventional and synchrotron μCT scans to quantify the morphology of the vestibular system in squamates able to perform controlled aerial descent compared to species characterized by a terrestrial or climbing life style. Our results show the presence of a strong phylogenetic signal in the data with the vestibular system in species from the same family being morphologically similar. However, both our shape analysis and an analysis of the dimensions of the vestibular system showed clear differences among animals with different life-styles. Species able to perform a controlled aerial descent differed in the position and shape of the inner ear, especially of the posterior ampulla. Given the limited stability of squamates against roll and the fact that the posterior ampulla is tuned to changes in roll this suggests an adaptive evolution of the vestibular system in squamates using controlled aerial descent. Future studies testing for similar differences in other groups of vertebrates known to use controlled aerial descent are needed to test the generality of this observation.     

Glide angle in the genus Petaurus and a review of gliding in mammals

The gliding angle of the Mahogany Glider Petaurus gracilis and the Sugar Glider Petaurus breviceps was determined from field studies by measuring the height of launch and landing of glides and the distance travelled. This showed no significant difference between these two species in glide ratio, which averaged 1.91 and 1.82 m distance per 1 m loss in altitude, respectively, nor in glide angle which averaged 28.26° and 29.69° for the Mahogany Glider and Sugar Glider, respectively. Significant differences were found between them for height of launch (19.75 and 11.96 m, respectively), height of landing (4.48 and 1.95 m, respectively), diameter at breast height of landing tree (44.12 and 23.22 cm, respectively), and glide distance (29.71 and 20.42 m, respectively). An examination of the ratio of interorbital width to maximum skull width of gliding and nongliding possums was measured from museum skulls to examine whether gliders have eyes wider apart, to allow triangulation of distance in preparation for gliding. Gliding possums showed a trend toward having a larger interorbital width than nongliding possums, although there appear to be several factors acting on the interorbital width. Museum study skins of all gliding marsupials were measured to determine the relationship between patagium surface area and body mass which showed a clear relationship (r² = 0.9688). A comparison of gliding behaviour, patagium, development of limbs, tail morphology and mass was also made between gliding marsupials and other gliding mammals.     

Maximal horizontal flight performance of hummingbirds: effects of body mass and molt

Hovering and fast forward flapping represent two strenuous types of flight that differ in aerodynamic power requirement. Maximal capabilities of ruby-throated hummingbirds (Archilochus colubris) in hovering and forward flight were compared under varying body mass and wing area. The capability to hover in low-density gas mixtures was adversely affected by body mass, whereas the capability to fly in a wind tunnel did not show any adverse mass effect. Molting birds that lost primary flight feathers and reduced wing area also displayed mass loss and loss of aerodynamic power and flight speed. This suggests that maximal flight speed is insensitive to short-term perturbations of body mass but that molting is stressful and reduces the birds' speed and capacity for chase and escape. Hummingbirds' flight behavior in confined space was also investigated. Birds reduced their speeds flying through a narrow tube to approximately one-fifth of that in the wind tunnel and did not display differences under varying body mass and wing area. Hence, performance in the flight tube was submaximal and did not correlate with performance variation in the wind tunnel. For ruby-throated hummingbirds, both maximal mass-specific aerodynamic power derived from hovering performance in low-density air media and maximal flight velocity measured in the wind tunnel were invariant with body mass.     

Flight speed of seabirds in relation to wind speed and direction

We studied flight speed among all major seabird taxa. Our objectives were to provide further insight into dynamics of seabird flight and to develop allometric equations relating ground speed to wind speed and direction for use in adjusting seabird density estimates (calculated from surveys at sea) for the effect of bird movement. We used triangulation at sea to estimate ground speeds of 1562 individuals of 98 species. Species sorted into 25 “groups” based on similarity in ground speeds and taxonomy. After they were controlled for differences inground speed, the 25 groups sorted into eight major “types” on the basis of response to wind speed and wind direction. Wind speed and direction explained 1664% of the variation in ground speed among seabird types. For analyses on air speed (ground speed minus apparent wind speed), we divided the 25 groups according to four flight styles: gliding, flap-gliding, glide-flapping and flapping. Tailwind speed had little effect on air speed of gliders (albatrosses and large gadfly petrels), but species that more often used flapping decreased air speed with increase in tailwinds. All species increased air speeds significantly with increased headwinds. Gliders showed the greatest increase relative to increase in headwind speed and flappers the least. With tailwind flight, air speeds were greatest among species with highest wing loading for each flight style except gliders, which showed no relationship. For headwind flight, species with higher wing loading had higher air speeds; however, the relation was weaker in flappers compared with species using some amount of gliding. In contrast, analyses for air speed ratio (i.e. difference between air speed in acrosswinds [with no apparent wind] and speed flown into headwinds, or with tailwinds, divided by speed acrosswind) revealed that among species using some flapping, and with lower wing loading (surface-feeding shearwaters, small gadfly petrels, storm petrels, phalaropes, gulls and terns), adjusted air speeds more than those with higher wing loading (alcids, “diving shearwaters”, “Manx-type shearwaters”, pelicans, boobies and cormorants). As a result, most flappers of low wing loading flew much faster than V_mr (the most energy efficient air speed per distance flown) when flying into headwinds. We suggest that better-than-predicted gliding performance with acrosswinds and tailwinds of large gadfly petrels, compared with albatrosses, resulted from a different type of “soaring” not previously described in seabirds.     

Behavioral and hormonal correlates of alternative reproductive strategies in a polygynous lizard: Tests of the relative plasticity and challenge hypotheses

Species with alternative reproductive tacts are good models to investigate the poorly understood question of whether individual variation within sexes results from the same physiological mechanisms that control variation between sexes. We have shown previously that adult male tree lizards, Urosaurus ornatus, of different throat color morphs express different levels of aggression in the laboratory. Further field results support the suggestion that the two morphs practice alternative reproductive tactics because the two morphs express different levels of aggressive behavior under field conditions and exhibit dramatic and opposite responses to aggressive challenges. However, despite these behavioral differences, the two morphs do not differ in levels of testosterone or corticosterone either in undisturbed situations or following aggressive challenge. These results are consistent with the relative plasticity hypothesis which proposes that organizational, rather than activational, actions of steroid hormones will be more important in morph differentiation when morphs are fixed in adult life, as they are in tree lizards. These results also support the hypothesis that steroid hormonal levels are insensitive to social modulation in males of species such as U. ornatus without paternal care.     

The gliding reptiles of the Upper Permian

The skeleton of a long-ribbed reptile from the Upper Permian Marl Slate of north-east England is described. The animal is assigned to the genus Weigeltisaurus , previously recorded from the Kupferschiefer of West Germany. Long-ribbed reptiles from the Upper Permian of Madagascar are also considered. Daedalosaurus Carroll, 1978 is a junior synonym of Coelurosauravus Piveteau, 1926. The European and Madagascan genera can be accommodated within a single family Coelurosauravidae, Infraclass Eosuchia. The skull is diapsid with an incomplete lower temporal arcade. Comparison with the modern Draco and the Upper Triassic kuehneosaurids supports the conclusion that the coelurosauravids were effective gliders.     

Gliding locomotion of Siberian flying squirrels in low-canopy forests: the role of energy-inefficient short-distance glides

Short glides of less than 20 m seem energy inefficient for the Siberian flying squirrel Pteromys volans as with the northern flying squirrel Glaucomys sabrinus. However, Siberian flying squirrels in low-canopy forests frequently use short glides. Therefore, we sought to clarify the gliding patterns of Siberian flying squirrels for energy-efficient gliding transport in low-canopy forests (mean tree height, 15.3 m) in Hokkaido, Japan, based on records of 66 glides and 35 launch and landing trees. Mean launch height, landing height, and horizontal glide distance were 14.4, 2.7, and 21.4 m, respectively. For short distances, horizontal glide distance was strongly correlated with launch heights but not with launch tree height. For glides of more than 20 m, horizontal glide distance was significantly correlated with both launch height and launch tree height. The mean heights of launch and landing trees for short glides were 15.6 and 19.5 m, respectively. For long glides, these heights were 22.7 and 19.2 m. For short glides, mean launch tree height did not differ from overall mean tree height. However, for long glides, the mean launch tree height was greater than the overall mean tree height. Also, for short glides, the height of the landing tree was greater than that of the launch tree. Launch trees used for long glides were as high as the landing trees used in short glides. From these results, we conclude that Siberian flying squirrels in low-canopy forests save energy by gliding initially from a tree with sufficient height to permit a glide to a taller tree. This taller tree then permits long-distance glides that are energetically more efficient.     

The highest kingdom of Anolis: Thermal biology of the Andean lizard Anolis heterodermus (Squamata: Dactyloidae) over an elevational gradient in the Eastern Cordillera of Colombia

Vertebrate ectotherms may deal with changes of environmental temperatures by behavioral and/or physiological mechanisms. Reptiles inhabiting tropical highlands face extreme fluctuating daily temperatures, and extreme values and intervals of fluctuations vary with altitude. Anolis heterodermus occurs between 1800 m to 3750 m elevation in the tropical Andes, and is the Anolis species found at the highest altitude known. We evaluated which strategies populations from elevations of 2200 m, 2650 m and 3400 m use to cope with environmental temperatures. We measured body, preferred, critical maximum and minimum temperatures, and sprint speed at different body temperatures of individuals, as well as operative temperatures. Anolis heterodermus exhibits behavioral adjustments in response to changes in environmental temperatures across altitudes. Likewise, physiological traits exhibit intrapopulation variations, but they are similar among populations, tended to the “static” side of the evolution of thermal traits spectrum. The thermoregulatory behavioral strategy in this species is extremely plastic, and lizards adjust even to fluctuating environmental conditions from day to day. Unlike other Anolis species, at low thermal quality of the habitat, lizards are thermoconformers, particularly at the highest altitudes, where cloudy days can intensify this strategy even more. Our study reveals that the pattern of strategies for dealing with thermal ambient variations and their relation to extinction risks in the tropics that are caused by global warming is perhaps more complex for lizards than previously thought.     

Intraspecific differences in behaviour and physiology: effects of captive breeding on patterns of torpor in feathertail gliders

Studies on the physiology of mammals and birds are often conducted using captive-bred individuals and it is commonly assumed that the resulting data are representative of individuals living in the field. To investigate whether these assumptions are justified, we quantified morphological, behavioural, and physiological variables of the small marsupial feathertail glider (Acrobates pygmaeus). We compared three populations: (i) individuals from a cool-temperate, montane area, (ii) individuals form a subtropical, coastal area, and (iii) captive-bred individuals. Captive-bred gliders differed from the montane field gliders in morphology (longer tails and snouts), behaviour (longer activity periods) and physiology (less frequent torpor, shorter torpor, shallower torpor, higher metabolic rates during rest and torpor, and slower rates of rewarming). Most of these differences were also apparent between the captive-bred and the coastal field gliders. Unlike both field populations, captive-bred gliders often became hypothermic and were unable to rewarm. In contrast to the other physiological variables, the minimum body temperatures defended during torpor and the corresponding air temperatures differed between the montane and coastal field gliders, but were similar in coastal field and captive-bred gliders. Our study shows that morphology, behaviour and physiology can be strongly affected by breeding in or acclimation to captivity. The poor expression of torpor and thermal performance of the captive-bred gliders raises the question of whether they possess the physiological capability for survival in the wild. Even though captive breeding appears to have only minor effects on some physiological variables, data from captive-bred individuals should only be extrapolated to the field with caution.     

Flight speeds among bird species: allometric and phylogenetic effects

Flight speed is expected to increase with mass and wing loading among flying animals and aircraft for fundamental aerodynamic reasons. Assuming geometrical and dynamical similarity, cruising flight speed is predicted to vary as (body mass)^1/6 and (wing loading)^1/2 among bird species. To test these scaling rules and the general importance of mass and wing loading for bird flight speeds, we used tracking radar to measure flapping flight speeds of individuals or flocks of migrating birds visually identified to species as well as their altitude and winds at the altitudes where the birds were flying. Equivalent airspeeds (airspeeds corrected to sea level air density, U_e) of 138 species, ranging 0.01–10 kg in mass, were analysed in relation to biometry and phylogeny. Scaling exponents in relation to mass and wing loading were significantly smaller than predicted (about 0.12 and 0.32, respectively, with similar results for analyses based on species and independent phylogenetic contrasts). These low scaling exponents may be the result of evolutionary restrictions on bird flight-speed range, counteracting too slow flight speeds among species with low wing loading and too fast speeds among species with high wing loading. This compression of speed range is partly attained through geometric differences, with aspect ratio showing a positive relationship with body mass and wing loading, but additional factors are required to fully explain the small scaling exponent of U_e in relation to wing loading. Furthermore, mass and wing loading accounted for only a limited proportion of the variation in U_e. Phylogeny was a powerful factor, in combination with wing loading, to account for the variation in U_e. These results demonstrate that functional flight adaptations and constraints associated with different evolutionary lineages have an important influence on cruising flapping flight speed that goes beyond the general aerodynamic scaling effects of mass and wing loading.     

Caudal autotomy and regeneration in lizards

Caudal autotomy, or the voluntary self-amputation of the tail, is an anti-predation strategy in lizards that depends on a complex array of environmental, individual, and species-specific characteristics. These factors affect both when and how often caudal autotomy is employed, as well as its overall rate of success. The potential costs of autotomy must be weighed against the benefits of this strategy. Many species have evolved specialized behavioral and physiological adaptations to minimize or compensate for any negative consequences. One of the most important steps following a successful autotomous escape involves regeneration of the lost limb. In some species, regeneration occurs rapidly; such swift regeneration illustrates the importance of an intact, functional tail in everyday experience. In lizards and other vertebrates, regeneration is a highly ordered process utilizing initial developmental programs as well as regeneration-specific mechanisms to produce the correct types and pattern of cells required to sufficiently restore the structure and function of the sacrificed tail. In this review, we discuss the behavioral and physiological features of self-amputation, with particular reference to the costs and benefits of autotomy and the basic mechanisms of regeneration. In the process, we identify how these behaviors could be used to explore the neural regulation of complex behavioral responses within a functional context.     

Thermophysiology,microclimates, and species distributions of lizards in the mountains of the Brazilian Atlantic Forest

Thermophysiological traits, particularly thermal tolerances and sensitivity, are key to understanding how organisms are affected by environmental conditions. In the face of ongoing climate change, determining how physiological traits structure species’ ranges is especially important in tropical montane systems. In this study, we ask whether thermal sensitivity in physiological performance restricts montane lizards to high elevations and excludes them from the warmer environments reported at low elevations. For three montane lizard species in the Brazilian Atlantic Forest, we collect thermophysiological data from lizards in the highest elevation site of each species’ distribution, and ask how well the individuals exhibiting those traits would perform across the Atlantic Forest. We use microclimatic and organism‐specific models to directly relate environmental conditions to an organism's body temperature and physiological traits, and estimate measures of thermophysiological performance. Our findings demonstrate that thermophysiological constraints do not restrict montane lizards to high elevations in this system, and thus likely do not determine the warm boundaries of these montane species’ distributions. Results also suggest that competition may be important in limiting the warm boundaries of the species’ ranges for two of the focal species. These experimental results suggest that caution should be used when claiming that physiology drives patterns of diversity and endemism within montane environments. They also highlight the importance of interdisciplinary experimental studies that bridge the fields of evolution and ecology to improve predictions of biological responses to future environmental shifts.