Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Ecomorphology of Rufous Hummingbirds : an investigation of maneuverability and agility in four age-sex… Moore, Janet L. 1997

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1997-0574.pdf [ 5.79MB ]
Metadata
JSON: 831-1.0087984.json
JSON-LD: 831-1.0087984-ld.json
RDF/XML (Pretty): 831-1.0087984-rdf.xml
RDF/JSON: 831-1.0087984-rdf.json
Turtle: 831-1.0087984-turtle.txt
N-Triples: 831-1.0087984-rdf-ntriples.txt
Original Record: 831-1.0087984-source.json
Full Text
831-1.0087984-fulltext.txt
Citation
831-1.0087984.ris

Full Text

ECOMORPHOLOGY OF RUFOUS HUMMINGBIRDS: AN INVESTIGATION OF MANEUVERABILITY AND AGILITY IN FOUR AGE-SEX CLASSES by JANET L. MOORE BSc. McGill University, Montreal, 1993. A THESIS IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES DEPARTMENT OF ZOOLOGY We accept this thesis as conforming to the required standard THE UNIVERSITY OF BRITISH COLUMBIA September, 1997 © Janet L. Moore, 1997 In presenting this thesis in partial fulfilment of the requirements tor an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of ^ ^ O o l O c j V-) The University of British Columbia Vancouver, Canada Date £7/(^7 DE-6 (2 /88) ABSTRACT Low and high velocity turning was investigated in four age-sex classes of rufous hummingbirds (Selasphorus rufus ). In the field, adult males outcompete all other age-sex classes for territories and it is assumed in the literature that adult males are more maneuverable (Feinsinger and Chaplin 1975, Kodric Brown and Brown 1978, Carpenter et al 1993 a,b). Using two experiments and a review of the flight literature, this thesis investigated high and low velocity turning in rufous hummingbirds. Rufous hummingbirds are sexually dimorphic in the length and shape of their wings. Adult males have short, narrow wings as compared to the females' long and broad wings. Juveniles of both sexes have longer wings than the adults. Wing disc loading is the ratio of mass to wing length and varies among classes as wing length varies. It has been assumed in the literature that birds with higher wing disc loading (adult males) are the most maneuverable. This assumption has led to numerous statements in the literature concerning the superior flight performance of adult male rufous hummingbirds. In a low velocity turning experiment, birds flew back and forth between two feeders 0.5 m apart. A computer recorded the elapsed time between feeders and a video camera recorded all flights from above. The average elapsed time recorded by the computer determined that juvenile females were the fastest of all four age-sex classes. The short-winged adult males were the slowest group and the adult female and juvenile males, which have similar wing lengths, had similar average elapsed times. The results were confirmed with the video data: adult males had the lowest maximum velocities and juvenile females the highest. These results clearly contradict those predicted in the current hummingbird literature. In a high velocity turning experiment, birds flew through a 4.8 m tunnel with 3 barriers that required the birds to alter their flight path en route to the i i feeder. The computer recorded the elapsed time between the perch and feeder and feeder and perch. Adult males flew faster flying from the feeder to the perch than all other age-sex classes. However, the classes did not differ significantly in elapsed time from perch to feeder. Adult males were the top performers in both angular and translational velocity during flights through the barriers towards the feeder. These results are predicted by current aerodynamic theory for bats and birds with non-hovering flight. An early chapter of this thesis examines the misuse of the terms maneuverability and agility throughout the hummingbird literature. Aerodynamic predictions concerning turning performance in hummingbirds have not been tested prior to this thesis. As these are the first measurements of maneuverability and agility in hummingbirds I feel that it is necessary for these terms to be operationilized for future use in hummingbird research. i i i TABLE OF CONTENTS Abstract ii Table of Contents iv List of Tables v List of Figures vi Acknowledgments viii Chapter One: General Introduction 1 Chapter Two: Defining, Predicting and Measuring Maneuverability 4 and Agility for Hummingbird Flight Chapter Three: Low Velocity Turning in Four Age-sex Classes of 15 Rufous Hummingbirds Methods 18 Results 26 Discussion 35 Chapter Four: High Velocity Turning in Four Age-sex Classes of 69 Rufous Hummingbirds Methods 72 Results 79 Discussion 83 Chapter Five: Conclusion and General Discussion 98 Literature Cited 101 iv LIST OF TABLES Table 1: General Linear Model of morphology as predicted by age and 65 sex class. Table 2: Average wing length, mass and WDL measurements for all four 65 age-sex classes. Table 3: Linear regressions of three morphological measures are 66 compared to four measures of translational performance. Table 4: Multiple regression analysis of wing length and mass for four 66 translational performance measures. Table 5: Linear regressions of three morphological measures are 67 compared to four measures of angular performance. Table 6: Multiple regression analysis of wing length and mass for five 67 translational performance measures with juveniles males removed. Table 7: General Linear Models of translational and angular 68 performance measures. Table 8: Linear regressions of four performance measures are compared 68 to elapsed time between feeders. v LIST OF FIGURES Figure 1: The experimental setup for low velocity turning. 46 Figure 2: Procedure for recording position and orientation of bird. 47 Figure 3: Comparison of elapsed time of video analysis and computer 48 elapsed time. Figure 4: Mean elapsed time compared to fastest elapsed time. 49 Figure 5: Frequency distribution of elapsed time from JM 36. 50 Figure 6: Average wing lengths of four age-sex classes. 51 Figure 7: Mass of 18 individuals is plotted against wing length. 52 Figure 8: Mean elapsed time is plotted against wing length. 53 Figure 9: Velocity and acceleration curves for translational and 54 angular motion for juvenile and adult males. Figure 10: Four measures of translational performance are plotted 55 against wing length. Figure 11: Four measures of translational performance are plotted 56 against wing disc loading. Figure 12: Four measures of translational performance are plotted 57 against mass. Figure 13: Four measures of angular performance are plotted against 58 morphology with and without juvenile males. Figure 14: Four performance measures in translational and angular 59 motion are plotted for each age-sex class. Figure 15: Mean number of degrees turned in the first 5 frames of a 60 flight. Figure 16: Schematic partial plan view of the experiment to illustrate 61 positions of maximal performance. v i Figure 17: Position of maximal performance for four age-sex classes. 62 Figure 18: Comparisons of translational performance measures; 63 including Vmax, Amax, tVmax and tAmax. Figure 19: Comparisons of angular performance measures; including 64 AVmax, AAmax, tAVmax and tAAmax. Figure 20: The experimental tunnel. 87 Figure 21: Procedure for recording position and orientation of bird. 88 Figure 22: Relationship between mean elapsed time and fastest elapsed 89 time through the tunnel. Figure 23: Average wing length and WDL for four age-sex classes. 90 Figure 24: Mean elapsed time between feeder and perch for all four 91 age-sex classes. Figure 25: Effect of WDL on four measures of translational velocity 92 through the barriers. Figure 26: Velocity at point of entry and exit of barriers for all four 93 age-sex classes. Figure 27: Maximum acceleration through the barrier section of the 94 tunnel as related to wing length and WDL. Figure 28: Maximum acceleration through the barriers in both 95 directions for all four age-sex classes. Figure 29: Maximum angular velocity through the barrier section of the 96 tunnel as related to wing length and WDL. Figure 30: Maximum angular velocity is plotted for all four age-sex 97 classes. vii A C K N O W L E D G M E N T S : I would like to thank a wide variety of people for their help and support of this thesis. To my advisor, Lee Gass who helped me develop as a scientist, writer and person and for inspiring me as a teacher. To Lara Chatters and Christianne Wilhelmson for helping catch birds in the field and feeding them in the lab, and for understanding the joys of the computer controlled hummingbird laboratory. To the computer masters; Lance Bailey and Dan Fandrich for their help in writing C programs, Mohammed Darwish for making the server work, and Micheal Weiss for his Macintosh wizardy. To my committee members John Gosline and Jamie Smith for their advice on early drafts and to Ron Ydenberg for peaking my interest in behavioural ecology. And finally to Mom and Dad, Jeff, Brian, Melissa, Viren and my friends for their incredible support system over the past three years. Thanks. This research was funded by a NSERC operating grant to Lee Gass. Personal funding was supplied by teaching assistantships from the University of British Columbia. viii \ CHAPTER 1 GENERAL INTRODUCTION How a hummingbird's morphology influences its ecology encompasses the fields of evolution, physics, aerodynamic theory, behavioural ecology, biomechanics and ecomorphology. In this thesis I discuss: 1) how age and sex influence flight performance of hummingbirds in high and low velocity turning, 2) how wing shape influences angular and translational motion, and 3) how changes in wing size and shape of hummingbirds are different from bats and other birds. Other topics include problems with the current aerodynamic terminology used when attempting to classify rufous hummingbird performance, the evolution of sexual dimorphism in rufous hummingbirds, and how wing morphology may directly and indirectly influence the ecology of hummingbirds. Interspecific comparisons of hummingbird morphology and their relation to ecology are prominent in the literature (Feinsinger and Colwell 1978; Feinsinger et al. 1979; Brown and Bowers 1985; Collins and Paton 1989). Feinsinger and Chaplin (1975) suggested that variations in mass and wing length (wing disc loading) of hummingbirds could be correlated with foraging strategy, where species with the highest wing disc loadings were territorialists and those with low wing disc loadings were trapliners. They predicted that in any community, hummingbirds with higher wing disc loadings would be better able to defend profitable territories. Since this declaration, many authors (Kodric Brown and Brown 1978; Feinsinger et al. 1979; Ewald and Rowher 1980; Collins and Paton 1989; Carpenter et al. 1993 a, b) have used this correlation to explain inter- and intraspecific differences in foraging strategies of hummingbirds. 1 Does the morphology of wings affect how individual hummingbirds behave in similar environments? How do sexual differences in wing shape influence foraging strategies of individual hummingbirds? Possible mechanisms for the evolution of sexual dimorphism include sexual selection, reproductive role division, sex specific foraging and ecological causation (Hedrick and Temeles 1989). It seems likely that the ecological significance of morphological divergence within a species is in adapting the sexes to different subniches (Selander 1966; Brown and Bowers 1985; Peters and Grubb 1983; Desrochers 1989). How does the variation in wing length found in rufous hummingbirds (Selasphorus rufus) influence the ecology of individuals within the species? What variation in foraging and territorial economics is maintained by these dimorphisms, mediated by variation in flight performance? Does the wing structure of any age-sex class afford time and/or energy advantages over others in performing particular movements in flight? I will examine how differences in morphology influence flight performance, specifically high and low velocity turning. In discussion, the term "maneuverability" is often used to describe the defense capabilities of aggressive, territorial hummingbirds (Feinsinger and Chaplin 1975; Kodric Brown and Brown 1978; Feinsinger et al. 1979; Collins and Paton 1989; Carpenter et al. 1993 a, b). These suggestions in the literature assumed that birds defending territories must be highly maneuverable despite empirical tests never having been done. This thesis examines the relationship between age, sex, wing morphology and performance, before drawing conclusions on foraging strategies resulting from variations in wing shape. I also examine the interrelations between the concepts of maneuverability and agility. Which age-sex class is the most "maneuverable" or "agile"? In chapter two I address the problems in the current literature that concern aerodynamic terminology. The terms maneuverability and agility are used improperly and 2 interchangeably throughout the hummingbird but not the bat literature. I suggest that the terminology be reassessed and operationalized specifically for hummingbirds. The maneuverability of 18 rufous hummingbirds was investigated in two experiments that involved both high and low velocity turning. In Chapter 3 I report an experiment in which hummingbirds flew back and forth between two adjacent feeders 0.5 m apart, mimicking intra-patch visitation to flowers in the field. Several performance measures of both translational and angular velocity and acceleration were used to rank individuals on these short flights. The four age-sex classes (adult males, adult females, juvenile males and juvenile females) were used to determine how differences in morphology, specifically wing length, mass and wing disc loading, influence flight time between feeders. Data from both the elapsed time and computerized video analysis are discussed. In Chapter 4, I investigated high velocity turning of hummingbirds in a cluttered environment. Birds flew from a perch to a feeder, through a 4.8 m tunnel with three barriers that prevented straight-line flight. This experiment tested the speed at which birds could maneuver through the tunnel while avoiding the barriers. A video camera recorded flights and video tape analysis determined maximum translational and angular velocity and acceleration in both directions through the barriers. Four age-sex classes were used in this experiment to determine if differences in wing length and mass influence the ability of a hummingbird to fly through a cluttered environment. Data were obtained from a computer that recorded elapsed time between perch and feeder. 3 CHAPTER 2 DEFINING, PREDICTING AND MEASURING MANEUVERABILITY AND AGILITY FOR HUMMINGBIRD FLIGHT The operationalization of terminology is an inherent and important component of science (Peters 1991). Both within and among disciplines, terminology is often misused, or used inconsistently by many authors. The misuse of terminology can and has led to many misunderstandings in the literature, when terms are cited and reused but their true definitions are lost. The terms maneuverability and agility are used interchangeably in the hummingbird literature to describe flight performance, and nowhere have they been distinguished, defined or related to the same terms used in other disciplines. In contrast, the bat and other bird literature use relatively consistent definitions for the same terms. Confusing and misleading use of these terms is localized to the subsect of the hummingbird literature that is the content of my thesis. The objective of this chapter is to operationalize definitions for both maneuverability and agility for hummingbirds. Defining Maneuverability: Norberg and Rayner (1987) clearly defined maneuverability in bats as the space required by a bat to alter its flight path while flying at a fixed speed. They also suggested that this is equivalent to the minimum turning radius that an organism can attain without losing speed or momentum. But in order to make tight turns (i.e. minimum turning radius) a flying organism must slow down. However, the tight turn referred to in Norberg and Rayner (1987) was defined without respect to speed. Consequently, maneuverability has been defined as the ability to turn in an enclosed space or fly amongst clutter (Aldridge 1986). How 4 is bat maneuverability related to bat morphology? Maneuverability is inversely proportional to minimum turning radius and those species with the lowest wing loading have the greatest maneuverability. In general, bat species with higher maneuverability have lower wing loading, because of either low body mass or large wing area (Aldridge 1986; Norberg 1987, 1990). Defining Agility As "maneuverability" does not consider the speed of a turn, another term "agility" is frequently used when describing turning flight. Agility is defined as the maximum roll acceleration during the initiation of a turn and measures the speed at which the flight path can be altered (Norberg and Rayner 1987). The relationship between agility and morphology is complex, because high agility can be achieved at both high and low speeds (Norberg and Rayner 1987). The highest angular roll acceleration is found in fast flying species with short wings and high wing loading. In general, angular roll acceleration increases with decreasing body mass; light bats turn faster than heavy bats. How are maneuverability and agility measured? Differences in maneuverability and agility over a broad spectrum of bat species allow for tight correlations between morphology and performance in both fast and slow flying species. Maximum agility is attained by broad rounded wings in slow fliers and small narrow wings in fast fliers (Norberg 1990). Morphological differences in wing shape and mass lead to variation in maneuverability and agility in both fast and slow flying bats. How can the maneuverability or agility of a flying organism be assessed? Is it possible to have similar tests for different types of organisms? Many ecomorphological studies of bats and birds have examined relationships between maneuverability, agility and morphology. Aldridge (1986), flew bats through a 5 tunnel with weighted strings as obstacles. Different inter-string distances were used to challenge each species of bat. The most cluttered arrangement that each bat could maneuver through was recorded. The most maneuverable bat was the one that could negotiate the most complex arrangement of strings (smallest inter-string distance). Not suprisingly, Aldridge (1986) found a significant correlation between the ability of a bat to maneuver through an obstacle course and its ability to turn tightly. Bats with the highest wing loading were considered the most agile, and those with the lowest wing loading the most maneuverable. Aldridge and Rautenbach (1987) measured bat maneuverability by flying them through an obstacle course of weighted strings in a tunnel. Again, maneuverability was scored as minimum inter-string distance that an individual could negotiate (minimum negotiable distance). They concluded that measurements of morphology are good predictors of ecological structures in communities and that differences in wing morphology could lead to spatial resource partitioning in bats. In another test of maneuverability, Moller (1991) flew swallows through a 4 m tunnel with 4 barriers on each side. The swallows' wing tips were dipped in black ink and the barriers were covered with white cloth. Maneuverability was measured by the number of black marks on the barriers; the fewer marks, the more maneuverable the bird was considered to be. Most studies of the flight performance of birds consider the association of increasing mass and the consequent risks of predation. Does heavy loading impede a bird's take-off speed and therefore increase risk of predation? In a recent study of the ability of small birds to take flight with changing masses, Metcalfe and Ure (1995) used the terms maneuverability and agility interchangeably. They asked how diurnal variation in mass influenced the ability to maneuver through a cage with 3 barriers. The positions at which the 6 birds flew through the barriers were recorded by a video camera. In this case the fastest bird was considered the most maneuverable. If we assume that the definition of maneuverability is the ability to turn tightly, the elapsed time of a bird flying through obstacles is irrelevant, and instead a measure of agility (the ability to turn quickly). They predicted that increasing mass (increasing wing loading) would increase the bird's turning radius and therefore decrease its maneuverability. However, throughout the paper they refer to agility as the ability to out maneuver predators, despite only testing for maneuverability. In many instances researchers not only define maneuverability incorrectly, but use fundamentally invalid tests to measure maneuverability as I will describe later in this chapter. Another recent study investigated the influence of mass on escape ability of European starlings (Witter et al. 1994). Birds flew through a tunnel with poles in arrangements similar to the experimental design used to test bat maneuverability. The wings of the starlings were dipped in ink and the number of contacts were used to measure maneuverability. The elapsed time of the flight through the obstacles was recorded. As the mass of the bird increased (added weights on the legs of the bird), the number of contacts with poles increased, and as the mass of a bird decreased (through food deprivation) the number of contacts decreased. However, mass did not affect elapsed time through the obstacles; the birds did not compensate for their reduced "maneuverability" by slowing down. Therefore, mass did not affect the speed at which the birds maneuvered. If agility has been defined as the speed at which a maneuver is performed, then elapsed time through obstacles may be a good predictor of agility and not maneuverability. Which experimental designs most effectively test maneuverability and agility in bats and birds? Can these designs test turning performance in hummingbirds? I think that because of the exceptional flying capabilities of 7 hummingbirds, future measurements of hummingbird flight performance will require a unique design. Because hummingbirds, unlike other birds, rarely collide with obstacles, the design of past experiments would not apply for hummingbirds. The time taken to fly through an obstacle course is a measure of maneuverability that can easily be recorded. However, the velocity at which the hummingbirds enter the obstacles (dependent on the length of tunnel before obstacles) will greatly influence the results (Chapter 4) and must be considered. Ideally this would be incorporated in to studies as a critical experimental variable. What is an "agile maneuver"? This paper is not the first to address the habitual confusion of these two terms. Andersson and Norberg (1981) initially used "maneuverability" to encompass both maneuverability and agility. Aldridge (1986) and Norberg and Rayner (1987) later distinguished between the terms and suggested that both be used when describing maneuverability. Both acceleration in the roll plane (agility) and minimum turning radius (maneuverability) are important for agile maneuvers. Because low wing loading is predicted to increase maneuverability, and high wing loading is predicted to increase agility, how can the terms be used interchangeably? The problem lies in the fact that "a fast turn with small turning radius is rarely possible, and a tight turn must necessarily be slow" (Norberg and Rayner 1987). The terms maneuverability and agility should be used independently to describe turning in flapping flight but many authors continue to use them interchangeably. Norberg and Rayner (1987) clearly stated the importance of distinguishing between turning in a small space, turning without loss of speed, and initiating a turn rapidly. Maneuverability was carefully defined as the space 8 required by a bat to alter its flight path while flying at a fixed speed, and agility as the rate at which a turn is initiated. Discrepancies in the Literature Male rufous hummingbirds (Selasphorus rufus), having equal mass but shorter wings than females, have higher wing disc loading. According to the above definitions of maneuverability, females (low WDL) should be more maneuverable than males; better able to make tight turns. In contrast, males (high WDL) should be more agile than females i.e., better able to make fast turns. The hummingbird literature has repeatedly used these terms interchangeably to describe how wing length influences foraging strategies in the field. Chapter 3 and 4 test the turning ability of rufous hummingbirds at high and low velocity. The first paper to discuss the possible dependence of hummingbird maneuverability and wing disc loading was Feinsinger and Chaplin (1975), who stated that successful defense of a territory requires fast forward flight and high maneuverability. As adult male rufous hummingbirds defend the most profitable territories, this suggests that adult males are the most maneuverable. A hummingbird defends a territory by chasing intruders and in many cases engage in aerial encounters of aggression with these intruders. "Birds controlling rich and conspicuous flower clumps spent a large proportion of time in territorial defense, which often involved high speed aerial chases and acrobatic maneuvering" (Feinsinger and Chaplin 1975). Feinsinger et al. (1979) later suggested that the outcome of these chases depends on the pursuer's "acceleration and maneuverability". Predictions of foraging strategy in relation to WDL are made both between species and within species. Traplining hummingbirds which do not defend territories have lower wing disc loadings and lower energy requirements 9 as a result of straight forward flight (Feinsinger and Chaplin 1975). "Fast forward flight is unnecessary in nonagressive species and would be detrimental to careful searching behaviour. Selection should favor a low wing disc loading, provided by a relatively long wing which also increases drag and facilitates slow forward flight" (Feinsinger and Chaplin 1975). Feinsinger and Chaplin's (1975) correlation and interpretation have been carried through the literature by many authors. Carpenter et al. (1993a) stated that "high WDL is thought to improve agility. Because territorial defense in many hummingbird species requires quick acceleration, high WDL should be associated with superior territorial ability". Kodric Brown and Brown (1978) applied the same reasoning to explain their observation that short winged male rufous hummingbirds defended territories with higher densities of flowers than did females. The history of interpretation of Feinsinger and Chaplin's (1975) paper is revealing as it exemplifies how easily terms can become misused and confused in the literature. Feinsinger et al. (1979) clearly stated that " shorter wings probably increase turning speed, acceleration and other components of maneuverability necessary for successful interference". However, Kodric-Brown and Brown (1978) state that "Feinsinger and Chaplin (1975) suggest that selection for speed and maneuverability in aerial encounters has favored the evolution of relatively short wings in territorial hummingbirds." Later in this same paper the statement follows: "males apparently use the aerial agility conferred by their short wings to aggressively defend flowers that are sufficiently dense to pay their high foraging and defense costs." They appear to suggest that males are both highly agile and highly maneuverable, and that agility and maneuverability can be used interchangeably. A critical distinction between these terms must be made and maintained. 10 Counter Arguments In many studies of multi-species hummingbird assemblages, asymmetries among species have often been attributed to morphological characters thought to underlie flight performance and energetics (Feinsinger and Chaplin 1975; Feinsinger and Colwell 1978; Feinsinger et al. 1979; Snow and Snow 1980). This is not the first paper to suggest that Feinsinger and Chaplin (1975) generalized too broadly from a single regression. Snow and Snow (1980) and Gill (1985) disputed Feinsinger and Chaplin's (1975) correlation by demonstrating that many species of hummingbirds with high WDL are trapliners instead of being territorial as predicted by the generalization. Gill (1985) suggested that low WDL confers efficient hovering, but long wings (low WDL) hinder rapid acceleration and increase drag. Nevertheless, these authors concurred with Feinsinger and Chaplin (1975) that birds with high wing disc loading are more maneuverable, have greater acceleration and fly faster between flowers. In a recent study, Stiles (1995) found that few of the hummingbirds he studied at La Selva, Costa Rica fit the pattern of wing disc loading and foraging strategy suggested by Feinsinger and Chaplin (1975). For example, two species of hummingbirds with similarly broad wings (low wing loadings) are among the most and least territorial of the birds at LaSelva. Stiles (1995) found no general relationship between wing disc loading and foraging strategies for the group of hummingbirds that he studied. It is obvious from these studies that further investigations into the role of wing morphology and foraging strategy are necessary. Wing Loading Because hummingbirds hover while sweeping the air in the shape of a horizontal disc, wing disc loading is used as a morphometric measure instead of 1 1 wing loading. Wing loading is the mass of the bird divided by total wing area (Pennycuick 1975). Epting and Casey (1973) defined wing disc loading as the ratio of body weight to wing disc area (units of g/cm2). Wing area is measured in bats as the area of both wings, the tail and the area of the body between the wings (Norberg 1981). Initially, wing disc area in hummingbirds was calculated using the formula: wing disc area (wda) = n (b/2)2 where b=2.5 * wing length (wrist to tip of longest unflattened primary; Hainsworth and Wolf 1972). Wing span is the distance between the outstretched wing tips and includes the width of the body. Wing area of hummingbirds is predicted by wing span and was calculated using wing length instead of direct measurements of wing area. Feinsinger et al. (1979) suggested that Greenewalt's (1975) equation is a more precise measure of wing disc area and used the following formula to calculate wing disc loading: WDL= mass 71 (1 +0.404 1 0 6 ) 2 Recently, Stiles (1995) pointed out that this method of calculating WDL (Greenewalt 1975; Feinsinger et al. 1979) assumes that the proportion of wing span represented by wing chord is constant across species. This assumption has led to a bias in past comparisons of WDL among species; i.e. that wing shape is invariant. Stiles (1995) suggested using a measure of wing span that is 2 times the length instead of the current 2.5 times wing length. Again, this equation assumes that all hummingbirds have the same ratio of body width to wing length. This distinction is most important when comparing species, and may also lead to problems in intraspecific comparisons. There are obvious differences in the shape of male and female Rufous hummingbirds and there may be differences in the shape of the juveniles as well (in progress). Adult males have short narrow wings, compared to adult females' longer, broader wings. Juvenile males' wings are similar to adult females', and juvenile females 12 have the longest wings of all four age-sex classes. These differences in wing width and shape are not considered in calculating WDL, and do not modify aerodynamic predictions concerning maneuverability and agility. However, the incorrect assumptions underlying standardization of the measurements introduces errors. For example, assuming that male and female adult rufous hummingbirds wings are the same shape overestimates males' wing area and underestimates their wing disc loading. A Comparison of Flapping and Hovering Flight As a consequence of their small size and relatively large wing area (relative to lift-requiring body weight) hummingbirds and some insects can hover by pushing air downwards. Instead of flapping flight, hummingbirds use the most energetically expensive form of flying; hovering (Weisfogh 1972, 1973; Ellington 1975; Epting 1980). Hummingbirds are unique in their morphology and flight; in fact their flight patterns are more similar to insect flight than to that of any other group of birds. Hummingbirds, unlike any other birds, can turn while hovering at zero velocity. How can we discuss variation in the ability to turn tightly when all hummingbirds can turn 360 degrees without flying forward at all? There are no data yet in the literature on the relative turning ability of different species of hummingbirds, nor any measurements at all of hummingbird turning. Future Considerations How should the terms maneuverability and agility be defined and measured for an explicit description of hummingbird performance? Two obvious recommendations follow from this review. First, we need empirical studies of the relationship between turning tightly and quickly. Second, as all existing predictions about hummingbird maneuverability are based on 13 speculation, a detailed analysis of performance in a wide range of morphologies (and species) is necessary for all hummingbirds. Because all hummingbirds are highly maneuverable, hummingbird maneuverability cannot be defined as the ability to turn tightly. Instead I will use the term agility to define a hummingbirds ability to make fast, tight turns. How much do hummingbirds slow down when they turn quickly? How quickly can a bird turn in a hovering position? How are these measures related to wing shape, wing length and wing disc loading? The experiments in Chapters 3 and 4 are a first attempt to answer these types of ecomorphological questions for hummingbirds. This thesis investigates intraspecific differences in flight ability of one dimorphic species. Morphological variation among the four age-sex classes of rufous hummingbirds has led to many questions about the advantage of specific morphologies within these classes. The misuse of the terms maneuverability and agility is widespread in the hummingbird literature. This has contributed to a proliferation of unclear and confusing interpretations of resource partitioning and territoriality. Perhaps the problem lies in the fact that hummingbirds are unique flying machines compared to bats and other birds. The bat literature is filled with studies on maneuverability of different bat species, and the comparisons of morphology and maneuverability are generally clear and consistent. For the remainder of this thesis I will avoid the usage of terminology that is not clearly defined at this time in the hummingbird literature. 14 CHAPTER 3 LOW VELOCITY TURNING IN FOUR AGE-SEX CLASSES OF RUFOUS HUMMINGBIRDS An organism's morphology reflects its ecology. Morphological differences between groups of individuals lead to species occupying different niches (MacArthur 1958; Hutchinson 1959; Lack 1971 and Schoener 1974). In general, we expect ecological factors to explain the evolution of many intra-population variations in morphology, such as sexual dimorphism. Selander (1966) suggested that ecological polymorphism within populations is expressed by sexual dimorphism in the shape and size of feeding structures and related differential niches. Presently, there are two major hypotheses for the evolution of sexual dimorphism; the theory of ecological causation and Darwin's (1871) theory of sexual selection. Sexual selection theory states that certain characters give an advantage in competition for mates, and will be selected for in one sex (Darwin 1871). The ecological causation hypothesis states that sexual differences in morphologies have evolved due to ecological causes (Selander 1966, 1972; Slatkin 1984; Shine 1989). These intraspecific differences in morphology would adapt the sexes to different ecological niches. In many cases these two theories can act in conjunction and are difficult to dissect. Relationships between morphology and ecology can be complex. Morphological variation within species may be expressed in the rate or magnitude with which natural functions are performed. Subsequently, this leads to variation in behavioural choices of foraging microhabitats. Ecomorphology is the study of functional relationships between morphology and ecology as mediated by the performance or behaviour of an organism (Wainwright 1988, 1991; Miles and Ricklefs 1994). Within species, variation in morphology and 15 the variation in performance that expresses it, may lead to age-sex classes adopting different foraging strategies or different foraging microhabitats. Rufous hummingbirds (Selasphorus rufus) are sexually dimorphic in the length of their bills, the length and shape of their wings, and their colouration. Adult males have a rusty brown body with a bright red gorget and adult females are olive green with a small red patch of irridescent feathers on a white gorget. Juvenile males and juvenile females can be distinguished by the juvenile males' light bronze streaked gorget compared to the juvenile females' immaculate white gorget. Both sexes of juveniles have a green feathered body (Stiles 1972; personal observations). Adult males of this species have the shortest and narrowest wings of all four age-sex classes, and therefore the least wing area, and juvenile females have the longest wings (Kodric Brown and Brown 1978; Johnsgard 1983; Carpenter et al. 1993a,b). Adult females' wings are broad and similar in length to juvenile males' wings. Mass is independent of sex and age. Finally, adult females, having longer bills and longer tongues than adult males, can utilize deeper flowers and obtain nectar faster than adult males (Temeles and Roberts 1993). How might these polymorphisms in wing shape influence flight abilities and consequently resource use and social dominance among age-sex classes? While hovering, hummingbird wings move in a figure eight, sweeping a disc of air, and wing disc loading is the ratio of body mass to this area (Epting and Casey 1973; Feinsinger and Chaplin 1975). Wing length dimorphism generates dimorphism in wing disc loading (WDL); adult male rufous hummingbirds with the same mass but smaller wings than females have higher WDL than adult females (Feinsinger and Chaplin 1975; Kodric Brown and Brown 1978; Feinsinger et al. 1979; Carpenter et al. 1993). Juvenile males' WDL is similar to adult females', and juvenile females, having the longest wings, have the lowest WDL of all four age-sex classes. 1 6 Hovering flight is energetically more expensive than any other type of flight (Weisfogh 1972, 1973; Ellington 1975; Epting 1980; Wells 1993 a,b). Hummingbirds hover at flowers to feed and spend a significant portion of their foraging time on short flights between flowers. It has been suggested that power output is positively correlated with WDL, such that the higher the WDL the more energy is required to hover and to fly (Pennycuick 1968; Weis-fogh 1972; Epting and Casey 1973; Epting 1980; Feinsinger and Chaplin 1975). What is the ecological significance of WDL? What advantages or disadvantages in flight performance are maintained by these dimorphisms? Do certain age-sex classes have a performance advantage over others? Numerous attempts have been made to understand the relationship between hummingbird morphology and foraging strategy (Feinsinger and Chaplin 1975; Kodric Brown and Brown 1978; Feinsinger and Colwell 1978; Feinsinger et al. 1979; Ewald and Rowher 1980; Brown and Bowers 1985; Collins and Paton 1989). It has been assumed that the short wings and high WDL of males make them more maneuverable and better able to defend territories against intruders. Through a correlation analysis of 12 species of hummingbirds, Feinsinger and Chaplin (1975) found that territorial hummingbird species have higher WDL than non territorial species. Referring directly to the sexual dimorphism in WDL of rufous hummingbirds, they concluded that adult males defend territories better than all other classes as a result of their higher WDL. The assumption is that greater maneuverability improves the ability to defend territories (see Chapter 2). This study is an experimental comparison of the flight performance of four age-sex classes of rufous hummingbirds. My specific objectives are to ask how 1) the specific morphology of wing structure affects how individual hummingbirds fly within similar environments 2) these differences in performance relate to morphology 3) turning ability reflects sex, age and wing length. Studies investigating 17 differences between adult males and adult females cannot remove the confound of sex and wing length. I included juveniles as well as adults of both sexes, which allowed differences in performance to be correlated with age and sex, thus removing this confound. Rufous hummingbirds forage on territories whose size depends on floral density and nectar production rate within the area (Gass et al. 1976; Kodric-Brown and Brown 1978). Foraging bouts include short flights between flowers and inflorescences and longer flights between patches of flowers. Because most (90-95%) short flights are less than 2 meters in length (Gass and Montgomerie 1981), most of a hummingbird's foraging time is spent hovering or flying between flowers on inflorescences, or among inflorescences on plants. Given the yet undescribed demands of performing these maneuvers, what wing shape is most profitable for short flights between flowers, or for territory defense? What consequences for resource partitioning do particular wing shapes have for each age-sex class? As a first step at answering these questions, I compared the performance of four age-sex classes of hummingbirds in a low velocity turning experiment. I estimated performance from elapsed flight time between two feeders arranged to require both a turn and a short forward flight. I also estimated angular and translational velocities and accelerations of all phases of this complex maneuver using video analysis, then correlated eight performance measures with morphology for age-sex classes and individuals within classes. Methods: The Subjects Five adult male, five adult female, four juvenile male and four juvenile female hummingbirds (Selasphorus rufus ) were selected for this experiment from a laboratory population of 33 birds. Birds were captured at Sumas 18 Mountain and Port Mellon, British Columbia in May 1995 and Rosewall Creek (Vancouver Island), British Columbia in June 1995. The birds were housed individually in wire mesh cages (60 x 60 x 60 cm), in a room with a controlled photoperiod (15h: 9h) at the UBC Animal Care Facility. The birds were fed Nektar plus (Nekton USA Inc.), a commercial hummingbird nectar with supplemented soya protein, during the weekdays, and are maintained on a 20% sucrose solution on the weekends. The experiments ran over a period of 4 months (August - November 1995). Experimental Environment A triangular experimental room, 2.1 m x 1.25 m x 2.54 m high, was constructed with lights and feeders on the two adjacent sides of the isosceles triangular room (Fig. 1). The angle of the walls between the feeders was 34.6 degrees. Two green metal panels which housed feeders and lights were located 1.25 m above the floor on the two adjacent walls. The holes for a feeder were 3 mm in diameter and were surrounded by an orange Avery label 19 mm in diameter. The LEDs were 4 mm in diameter, and were located 2 cm above each feeder. The orange labels and the light cued the food source and food availability, respectively. The perch was 0.75 m from the midpoint between the feeders, and was 1.25 m tall. The video camera was attached to the ceiling 0.9 m above the flight path between feeders. Since each frame included both feeders, the video record included the position and orientation of the bird each 1/30 s. A computer controlled red LED lights, dispensed food, recorded the time and duration of all visits to feeders and perches by monitoring photocells, and turned a video camera on and off (Tooze and Gass 1985; Tamm and Gass 1985; Brown and Gass 1993; Thompson 1994; Chatters 1996). Hummingbirds learn quickly to visit feeders that are associated with stationary light cues (Brown 19 1992; Thompson 1994). I used this fact to direct birds back and forth between feeders. When a bird arrived at a feeder, its bill triggered a photocell built into the feeder. In turn this triggered a solenoid pump (General Valve Corporation Series 3) to immediately dispense 2 Ltl of sucrose solution (20% sucrose w/w) into a reservoir. Birds fed while hovering by probing their bills through holes into a reservoir (disposable syringe needle fittings). The computer also controlled a video camera (Sony Handy cam CCD-V101) to record flight paths between feeders. A video camera attached to the ceiling, directly above the feeders, recorded all flights (Fig. 1). Training All birds received training prior to the experiment. For several days prior to experiments the ad libitum feeders in the home cages were equipped with a green facing and an orange Avery label to help birds learn the experimental arrangement. Then, I placed a bird in the triangular experimental enclosure with one ad libitum feeder and a telescoping perch raised close to the ceiling. The perch was gradually lowered, until the bird acclimated to the experimental room and the perch was at the same level as the feeders. This took from 2 to 48 hours depending on the individual. Then I removed the ad libitum feeder and exposed the two adjacent lights and feeders. The acclimation procedure began when a computer program turned the left light on. When the bird visited this feeder, the computer dispensed 5 (ll of 20% sucrose solution. The right light and feeder were exposed during the training program, however the light was off and the feeder would not supply food even when probed. The training program thus alternated left and right feeders and lights, supplying food only when the light was on. While a bird was in training, the waiting time between feeding bouts was increased from 10 seconds to 1.5 minutes as the bird learned to feed from the appropriate feeders. 20 I considered a bird ready for the experiment if it no longer paused in flying between feeders and it alternated between the appropriate feeders for 60 minutes, with a waiting time between feeding bouts of 1.5 minutes. During the experiment, trials began every 1.5 minutes, always with the left feeder and alternating between feeders 9 times (10 probes). If the trial began and the bird did not leave its perch for 20 seconds, the trial ended, and it had to wait for another 1.5 minutes until the next trial began. Then both lights were turned off, no feeders would deliver food if probed, and the bird returned to the perch for 1.5 minutes until the next trial. The series of events occurred as follows; the left light turned on, if the bird probed the left feeder the computer delivered 2 Lil to the left feeder and turned on the right light. If the bird probed the right hand light, it received 2 Lil of food, and the left light went on, etc. Trials were aborted if the bird probed the wrong feeder, perched, or took more than 20 seconds to probe the correct feeder. Experimental Protocol Each experiment lasted 2 hours and 30.5 ±1.8 trials on average. Birds probed feeders 393 times and consumed 786 ju.1 of 20% sucrose solution on average. Trials ended if birds probed an incorrect feeder, made 10 correct probes, or perched before making 10 correct probes. There was a waiting time of 1.5 min. between trials when no food was delivered to either feeder. The video camera turned on 5 s before trials and turned off at the end of the trial. Thus it recorded only flights during trials, although the complete computer record showed that birds sometimes probed feeders between trials. The running clock on the video camera was set according to the computer time at the start of each experiment, to ensure that computer data could easily be matched with the video tape. I used computer time to identify trials on the videotape, and timed events within trials by counting video frames. 2 1 After the experiment ended, I removed the bird from the enclosure, weighed it on an electronic balance (+0.0lg), measured the length of each wing with hand held calipers, and returned it to its home cage. Hummingbird mass may vary depending on the time of last feeding, the time of day, the season, and whether the bird has recently defecated. All birds were weighed immediately following the experiment to minimize this variation. Analysis of Video Records Using elapsed times between feeders in the computer record, I selected 6 flights by each bird for video analysis. I analyzed the fastest flight between feeders and the 5 flights nearest the individual's mode. These 6 flights allowed comparison of variation within and among individuals and classes. I considered the fastest flight to represent maximal performance of this task under the conditions and the modal flights to represent the most frequent performance. Two video tapes were destroyed during analysis, leaving data for 16 birds (5 adult males, 5 adult females, 4 juvenile males and 4 juvenile females) with 6 flights each. The videotapes were transferred to a Macintosh computer using the program Aldus Moviemaker. Each frame was processed as a digital image and placed into a stack of images which represented all frames in each flight. These data were saved on a CD-ROM to be analyzed with the program NIH Image. Using this program, I logged the position in the horizontal plane (x, y) and orientation (angle) of the bird in each frame of each flight. Figure 2 indicates both x and y directions within the horizontal plane. A flight started when the bird removed its bill from the first feeder and ended when the bird's bill probed the second feeder. Flight durations estimated in this way from the video record were highly correlated with more accurate estimates from the computer record (r2=0.83, p<0.003, 108 flights, Fig. 3). This supports the video analysis of the 2 2 birds, as the fastest birds according to the computer elapsed times were also fastest as calculated by video analysis. Video analysis of each flight captured all frames between the two feeders. For each frame (1/30 s) an image of the bird in flight appeared on the computer screen. Using a mouse, I placed a straight line from the base of the bill to the tip of the tail. The computer recorded the midpoint (x, y coordinates), angle and length of this line (Fig. 2). I also recorded the positions of both feeders in this manner. Since the feeders were always 0.5m apart, this calibrated the other measurements, all of which I recorded in meters. To record the orientation of the bird in each frame, I placed a computer drawn line from the back of the head of the bird to its tail (i.e. along its spine). To test the accuracy of this measure, I repeatedly measured one bird, removing the endpoints of the line and replacing them repeatedly on the bird. After measuring the angle 25 times my error was within 1 ± 1.05 degree, and I feel confident that I made no errors larger than two degrees. A typical source of error in this type of experiment is parallax from the lens of the video camera (Rayner and Aldridge 1985; Chatters 1996). As the distance between the feeders was only 0.5 m and the camera was 0.9 m away from the bird, parallax error should be negligible. I confirmed this empirically by recording a grid of known size under the camera in the plane in which the birds flew. Calculating Angular Motion Motion in the horizontal plane can be described in translational and rotational terms. Angular velocity and acceleration were calculated from orientation in successive frames of the video. Angular velocity is defined in this experiment as the rate of change in orientation and angular acceleration is its rate of change. Because I did not calculate the inertia of individuals during 2 3 flights, I have assumed that inertia is directly proportional to mass for all measures of angular motion. Angular acceleration should not be confused with angular roll acceleration, which is a component of the turning rate of bats (Rayner and Aldridge 1985; Norberg and Rayner 1987). Angular roll acceleration involves the differential twisting and flexing of the wings and the resulting roll of the organism (Norberg 1990). Smoothing the Data Smoothing data reduces the amplitude of variation. After smoothing, maxima are never exaggerated and in most cases are less than the unsmoothed maxima. I smoothed the data because slight errors in initial measurements result in large changes in velocities and larger changes in the accelerations. I applied a digital filtering equation (Winter 1990) with a cutoff frequency of 10 Hz to smooth the raw translational acceleration, angular velocity and angular acceleration data. I chose a cutoff frequency after a residual analysis of all possible frequencies (Winter 1990). The filter was applied in both directions in order to remove any phase shifts that may have occurred due to a single pass of the filter. It was necessary to pad the data, as a reflection of the actual data, at the beginning and end of the series due to the short length of the time series. Ten points were added to both the beginning and the end of the data for the filter to function properly. These points were necessary for the proper functioning of the filter, but did not significantly influence the values of the filtered data. As a consequence of smoothing, potentially spurious data points (the first, second and last data point) were not included in the analysis of the translational acceleration data. I did not remove the first two smoothed data points in the analysis of angular motion as the turn occurred in the first 5-7 frames. Interpretation of the turn would be significantly altered if these frames had been removed. 2 4 Video Analysis of Flights In ecomorphological studies, performance is interesting from both biomechanical and ecological perspectives. Maximal performance interests the biomechanic because it expresses limits imposed by structure (muscles, wings, etc.). Mean or modal performance satisfies the ecologist because mode is more likely to represent the "typical" flight in nature. Both measures are important in hummingbird ecology because birds may perform maximally under special conditions, and fly at mean performance levels for the remainder of the time. The two measures were correlated (r2=0.44, p=0.003, 18 birds, Fig. 4). Birds who flew fastest also had the shortest mean durations. As expected, birds did not perform the task appropriately in every flight. Frequency distributions were generally skewed towards longer elapsed times due to these unacceptable flights. After comparing videotape records and elapsed times for each bird, I concluded that on flights longer than 0.8 s the birds were pausing and therefore not accurately performing the task. I therefore excluded these from analysis. This removed the slowest 20% of flights by all age-sex classes. For example, the mean for a juvenile male (JM 36) before truncation was 0.74 s and after truncation 0.67 s, and the mode remained unchanged at 0.63 s (Fig. 5). The frame by frame analysis of the videotape allowed for a detailed analysis of a bird's position every 1/30 s. I calculated the Pythagorean distance traveled from frame to frame in the horizontal plane along the bird's trajectory (path). From this I calculated: maximum translational velocity (VMax), maximum translational acceleration (AMax), maximum angular velocity (AVMax) and maximum angular acceleration (AAMax). I also calculated the time (frame #) and position (along the path) at which these maxima occurred. Single factor analysis of variance (ANOVA) allowed for comparison of performance measures among the four age-sex classes (n=4). General Linear 2 5 Models (GLM) allowed detection of the effect of age and sex on all performance measures, including interactions between age and sex. Another G L M determined relationships between mass, wing length, sex and age. Linear and multiple regressions determined how wing length and mass affected the eight performance measures. Results: Morphological Comparisons The first step of this analysis was to determine how wing length, mass and wing disc loading are distributed among age and sex classes. Both sex (GLM p<0.001, Table 1) and age (GLM p<0.001, Table 1) are good predictors of wing length. Adult males have significantly shorter wings than juvenile females (Fig. 6). Adult females' and juvenile males' wings are intermediate and similar. Juveniles of both sexes have longer wings than adults, and sexes have significantly different wing lengths. It is often assumed that larger birds have longer wings. However the mass of my birds was independent of age and sex class (Table 2), and wing length was unrelated to mass (Fig. 7). Neither sex nor age are good predictors of mass (Table 1). I calculated wing disc loading as: WDL(g/cm2) = m (b/2)2 n , where b=wing span = 2.5 * wing length (cm) and m =mass (g) (Feinsinger et al. 1979). As expected, adult male WDL was highest, followed by juvenile males and adult females, and juvenile females were lowest. Sex is a good predictor of WDL (GLM, p=0.02), but age does not predict WDL (GLM, p=0.07, Table 1). 2 6 Performance and Morphology Birds of each age-sex class flew back and forth between feeders an average of 196.8 ±6.9 times in the 30.5 ± 1.8 trials of the experiment (120048 video frames in all). Elapsed time was significantly correlated with wing length such that birds with longer wings flew faster between feeders than birds with shorter wings (r2=0.42, p<0.001, n=18, Fig. 8). Mean elapsed time between feeders was the primary measure of performance. Juvenile females flew faster (0.59 ± 0.01 s) than adult males (0.68 s ± 0.02s, A N O V A p=0.01, Fig. 8), and adult males were significantly slower than all other age-sex classes combined (ANOVA, p=0.01, Fig. 8). Elapsed times of juvenile males and adult females were intermediate and not significantly different from one another. Given four classes of hummingbirds with variable morphology, how do these differences correlate with performance? The pattern of elapsed times suggests that long wings offer advantages for flying short distances and low speed turning. However, overall elapsed time encompasses several component actions: backing out of the first feeder, turning (angular velocity and acceleration), flying forward (translational velocity and acceleration), slowing down (deceleration) and probing the next feeder. I evaluated eight performance measures: maximum translational velocity (VMax), maximum translational acceleration (AMax), maximum angular velocity (AVMax), maximum angular acceleration (AAMax), and the time to reach each of these maxima (tVMax, tAMax, tAVMax, t A AMax). The pattern of translational and angular velocities and accelerations for adult and juvenile males is shown in Figure 9. This figure, representing all individuals in each class, exemplifies the difference between the two extreme age-sex classes. This figure also demonstrates how different flight components compose a short flight of 0.5m, as each class may exhibit maximal performance in different components of the same overall task. 27 I investigated all possible correlations between three morphological measures (wing length, wing disc loading and mass) and the eight performance measures. Does wing length directly influence performance? How does mass influence performance? Wing disc loading includes both wing length and mass and has received a vast amount of attention in the literature as to how it relates to performance (refer to Chapter 2). Does higher WDL afford higher maximal performance as claimed in the hummingbird literature (Feinsinger and Chaplin 1975; Kodric Brown and Brown 1978; Feinsinger et al. 1979)? Juvenile females (longest wings) were faster, overall, than all other age-sex classes; significantly faster than adult males (shortest wings). Elapsed time between feeders, as measured by the computer, was significantly correlated with wing length (r2=0.42, p<0.001, Fig. 8), and WDL (r2=0.28, p=0.02). However, mean elapsed time between feeders was unrelated to mass. Translational Motion To compare translational motion in the four age-sex classes, I calculated maximum translational velocity (VMax) and maximum acceleration (AMax) and the time to reach these maxima for individual flights. In general, individuals with the longest wings and lowest wing disc loading attained the highest maximum translational velocities and did so sooner than birds with short wings. Wing length predicted maximum velocity (r2=0.35 p=0.02, Fig. 10) and time to reach maximum velocity (r2=0.48, p=0.004, Fig. 10). Wing disc loading predicted maximum velocity (r2=0.55, p=0.001, Fig. 11) and time to reach maximum velocity (r2=0.58, p=0.001, Fig. 11). Whereas mass did not predict overall flight duration, it did predict maximum velocity (r2=0.33 p=0.02, Fig. 12), but not time to reach maximum velocity (Fig. 12). It follows that birds with the longest wings and lowest wing disc loading also reached higher maximum accelerations than birds with short wings; wing 28 length predicted maximum acceleration (r2=0.43 p=0.008, Fig. 10) and time to reach maximum acceleration (r2=0.31, p=0.03, Fig. 10). Wing disc loading predicted maximum acceleration (r2=0.41, p=0.01, Fig. 11) and the time to reach maximum acceleration (r2=0.27, p=0.048, Fig. 11). Maximum acceleration and time to reach maximum acceleration were unrelated to mass (Fig. 12, Table 3). To ensure that mass was not a key determinant of performance, multiple regressions removed the effect (if any) of mass on wing length (Table 4). There was no apparent relationship between mass and wing length (Fig. 7). However, linear regressions were also used to correlate performance measures with wing length and wing disc loading because mass is not an accurate measure of morphology (Table 3). Because wing length remained constant throughout the experiment and mass changed on a daily basis, I decided to use wing length as well as wing disc loading as measures of morphology. Carpenter et al. (1983) found that rufous hummingbirds gained up to 50 % of their body mass over a period of 1-2 weeks in the field. Hummingbird mass changes noticeably with feeding, defecating and hovering. Mass normally increases during the day, and it commonly fluctuates within individuals over days, weeks, and months, and therefore cannot be controlled during experiments. How do these changes in mass influence performance measures? After removing the effect of mass, elapsed time remained significantly correlated with wing length (GLM, p=0.04, Table 4). There was no relationship between mass and elapsed time between feeders. After removing mass, wing length remained significantly related to elapsed time (GLM, p=0.04), VMax (GLM, p=0.04), tVMax (GLM, p=0.01), and AMax (GLM, p=0.02, Table 4). After removing the effect of wing length, mass was also nearly significantly related to maximum translational velocity (GLM, p=0.05, Table 4). Mass was not related to any other measure of translational velocity or acceleration after the effect of wing length was 29 removed from the analysis (Table 4). Angular Motion An important component of total flight duration is the turn away from the first feeder along the flight path to the next feeder. The turn occurs in the first third of the flight and in most cases the bird has completed its turn in less than 10 frames or 0.33 s. The following section on angular motion reveals an interesting twist in this story. Juvenile males, with intermediate wing lengths, reached the highest maximum angular velocities. As a result there was no significant relationship between any measure of angular velocity (AVMax, tAVMax) or acceleration (AMax and tAMax) and morphology. Since juvenile males had the highest angular velocity and intermediate length wings, I suggest that there is no simple or direct relationship between angular velocity (at a low velocity) and wing morphology. If juvenile males are excluded from the analysis, maximum angular velocity for all other classes was highly correlated with wing length (r2=0.71, p=0.001, Fig. 13) and WDL (r2=0.66, p=0.002, Fig. 13), but not mass (Table 5). Similarly, maximum angular acceleration for all other classes was highly correlated with WDL (r2=0.52, p=0.01, Fig. 13) and mass (r2=0.44, p=0.03, Fig. 13), but not wing length (Table 5). Juvenile males had an advantage over all other age-sex classes in reaching the highest angular velocities, but this advantage was not correlated with age, sex or morphology. With juvenile males included in the analysis, mass and wing length were not correlated with any of the angular motion performance measures. However, after removing the effect of mass, wing length accurately predicted maximum angular velocity (GLM, p=0.002, Table 6), but no other measure of angular motion. 3 0 Comparisons Between Age-Sex Classes Translational Motion After completing the analysis of how morphology relates to translational and angular performance, it was necessary to determine whether particular age-sex classes outcompeted others as a function of their class and not their morphology. Juvenile females reached the highest maximum translational velocity of all age-sex classes at 1.35 ± 0.11 m/s, and adult males lowest (1.1 ± 0.07 m/s; Fig. 14). Juvenile males and adult females were intermediate in maximum velocities reached. The juveniles reached maximum velocity faster than both adult males and females (ANOVA, p=0.01). At the maximum, adult males accelerated significantly slower (5.1 ± 0.09 m/s2) than all the other classes combined (8.4 ± 1.5 m/s2, ANOVA, p=0.03, Fig. 14). Although the difference was not significant, juvenile males accelerated faster than the other classes (9.3 ± 1.5 m/s2). Adult males reached maximum acceleration (AMax) significantly later than all other classes (ANOVA, p=0.003). The adult males reached AMax significantly slower (0.15s) than juvenile males (0.08s) (ANOVA, p<0.001). Angular Motion Morphology did not clearly predict angular motion. Age class was a clear predictor of performance in angular motion, but sex did not predict angular motion (Table 7). Juvenile males reached the highest maximum angular velocity of all classes (ANOVA, p=0.02, Fig. 14). Adult males reached significantly lower angular velocities (1005 ± 100 degrees/s) than juvenile males (1326 ± 108 degrees/s, ANOVA p=0.02, Fig. 14). Adult and juvenile females reached similar AVMax. Juvenile males reached AVMax sooner than all other age-sex classes (0.09 ± 0.01s; ANOVA, p=0.01), and adult males were the slowest to 3 1 reach maximum angular velocity (0.14 ± 0.02 s), significantly slower than juvenile males (ANOVA, p=0.005). Juvenile males have a performance advantage making fast turns at low velocity. Juvenile males turned 128.6 ± 6 . 1 degrees in the first 5 frames (0.17s); significantly farther than adult males (86.2 ± 9.0 degrees; ANOVA, p=0.008, Fig. 15). Adult males turned significantly slower on this temporal scale than all other classes combined (ANOVA, p=0.01, Fig. 15). Turning speed over 5 frames was significantly correlated with maximum angular velocity reached by individuals (r= 0.68, p<0.001, Fig. 15), and was correlated with translational velocity averaged over whole flights (r2= 0.31, p=0.04). By frames 6-10, juvenile males had nearly finished their turns and were turning slower than all other classes combined (ANOVA, p=0.048), allowing them to spend more flight time in forward flight than other classes. During this straight forward flight, juvenile females compensated for slower turning by reaching higher translational velocity and did so sooner than juvenile males. Due to the large variation of accelerations within as well as between individuals, there was no difference between classes in angular acceleration. The adult males appear to have lower angular accelerations than any other classes, but this result is not significant (ANOVA, p=0.06). Juvenile males reached maximum angular accelerations (1.3 frames or 0.04 ± 0.007 s), in half the time as adult males (2.6 frames or 0.085 ± 0.02 s, p=0.01, Fig. 14). Juvenile females reached maximum angular acceleration after 1.93 frames, similar to adult females at 2.11 frames. Position of Maxima Do individuals or classes follow the same path between the two feeders, 50 cm apart? To determine this I examined the position at which the birds reached maximum velocity and acceleration for both translational and angular 3 2 motion. All birds reached the maxima of the components of this flight in the same sequence, but classes varied in where they reached them (Fig. 16). Angular acceleration peaked in the first few cm of the flight, followed by maximum translational acceleration near the beginning of the turn away from the feeder, then maximum angular velocity nearing the end of the turn. Maximum translational velocity occurred approximately midway between the two feeders. Juvenile females reached maximum translational velocity significantly closer to the origin of the flight (23.6 ± 1.59 cm) than adult males (27.09 ± 0.95 cm; ANOVA, p=0.01, Fig. 17). The maximum translational acceleration of juvenile males occurred closer to the origin of the flight than for all other age-sex classes, and significantly closer than adult males (ANOVA, p=0.01, Fig. 17). The location of the point of maximum translational velocity and acceleration along the path was not different between groups in the y direction of the horizontal plane. Juvenile males turned closer to the center in the x direction at 2.46 ± 0.56 cm compared to the rest of the classes, but this difference was not quite significant (ANOVA, p=0.06, Fig. 17). Juveniles reached maximum angular acceleration closer to the first feeder than adults (ANOVA, p=0.004, Fig. 17). The location of the point of maximum angular velocity and acceleration along the path was not different between groups in the y direction of the horizontal plane Performance Measures and Elapsed Time Angular velocity and translational acceleration of hummingbirds would be difficult to measure in the field. If these component measures of performance were tightly correlated with elapsed time between feeders in my laboratory test, this would encourage more field studies of this nature. However, a multiple 3 3 regression of the eight performance measures against elapsed time demonstrated that time to reach maximum translational velocity was the only component measure that clearly represented overall elapsed time between feeders (GLM, p=0.002). Using linear regressions, elapsed time between feeders was highly correlated with all measures of translational performance: VMax, tVMax, AMax and tAMax (Table 8). No measures of angular performance were correlated to elapsed time between feeders (Table 8). Comparing Performance Measures Assuming that the power available for acceleration is constant, I anticipated that reaching higher peak translational velocity would take longer than to reach lower peak translational velocity. However, birds that reached the highest maximum velocities did so sooner than slower birds (r2=0.79, p<0.001, Fig. 18), because they accelerated fast enough to peak sooner (r2=0.58, p=0.001, Fig. 18). It follows that maximum translational velocity and maximum translational acceleration were highly correlated (r2=0.75, p<0.001, Fig. 18). Similarly, time to reach VMax was highly correlated with the time to reach AMax (r2=0.46, p=0.01, Fig. 18). The situation for angular motion is not as clear. Time to reach maximum angular velocity was independent of maximum angular velocity (Fig. 19). Similarly, time to reach maximum angular acceleration was unrelated to maximum angular acceleration (Fig. 19). However, peak angular velocities were significantly correlated with angular acceleration (r2=0.37, p=0.016, Fig. 19). Similarly, birds which reached angular velocity fastest, also reached angular acceleration fastest (r2=0.81, p<0.001, Fig. 19). This suggests that birds reaching maximum angular velocity sooner had higher angular accelerations than birds reaching AAMax later. 34 Discussion: Detailed analysis of translational and angular motion allowed me to compare low velocity flight of four age-sex classes of rufous hummingbirds. In general, juveniles performed better than adults of the same sex in more than half of the comparisons. Juvenile females, with the longest wings, reached the highest maximum translational velocities and had the lowest elapsed times between feeders of all age-sex classes. Juvenile males had higher angular velocities than all other age-sex classes, and turned exceptionally quickly. In general, juveniles outcompeted adults of the same sex in all categories tested. All measures of translational motion were highly correlated with wing length and wing disc loading, but no measures of angular motion were predicted by morphology (wing length, mass or WDL). After removing the effect of mass, wing length strongly predicted elapsed time, maximum translational velocity, and maximum translational acceleration. Mass was uncorrelated with any of the translational or angular measures with the exception of maximum translational velocity. Sex was not a good predictor of any measure of performance tested here. Juveniles had a clear performance advantage in all translational performance measures except maximum velocity. Neither sex nor age predicted performance in angular motion with the exception that age predicted maximum angular velocity. Contrary to the basic assumption of ecomorphology, morphology does not always predict performance, and maximal angular performance was not predicted at all by the morphometric measures that I used. Juvenile males have a clear advantage in angular motion and reached higher angular velocities and accelerations than all other age-sex classes. Despite having intermediate wing lengths, juvenile males outcompeted all other classes, refuting the simplistic assumption that wing length determines performance. Other morphological 3 5 measurements such as tail length and shape were not considered in this experiment and may have contributed to this and other differences in performance. Performance contingent on wing morphology may allow age-sex classes to exploit somewhat different resources or adopt alternative foraging strategies. Why do adult male rufous hummingbirds have short narrow wings and adult females longer, broad wings? Why do juveniles of both sexes have longer wings than the adults? I will not attempt to answer these expansive questions from my experiments. My results demonstrate that correlations are not as simple as I expected based on my reading of the literature (Feinsinger and Chaplin 1975; Kodric Brown and Brown 1978) and that turning flight in hummingbirds is complex. This leads me to suggest that the way in which we interpret correlations in the literature must be changed. Morphology and Performance Maximum translational velocity, maximum translational acceleration and elapsed time between feeders were negatively correlated with wing disc loading. Within this population of wild-caught hummingbirds, adult males had the highest WDL, followed by adult females with intermediate WDL similar to juvenile males and juvenile females with the lowest WDL. Contrary to the literature (Feinsinger and Chaplin 1975; Feinsinger and Colwell 1978; Kodric Brown and Brown 1978; Feinsinger et al. 1979; Ewald and Rowher 1980; Collins and Paton 1989; Carpenter et al. 1993 a,b), birds with low WDL flew faster between feeders than birds with high WDL. Juvenile females (lowest WDL) were the top performers in translational motion (fastest short distance flyers). Angular velocity and acceleration were negatively correlated with WDL if juvenile males were excluded from the analysis. Again contrary to 36 expectations, birds with low WDL had high maximum angular velocities and accelerations. Juvenile males have a distinct advantage in angular motion, but the mechanism for this advantage is unclear. They turned faster, nearly completing the turn in the first 5 frames of the flight (0.17 s) leaving more time for straight forward flight. However, they did not reach top speed or reach the second feeder sooner than other classes. Despite juvenile males' high turning rates, juvenile females had higher translational velocities, allowing them to reach the second feeder sooner. Why is maximal performance important? How important are angular and translational motion in the daily movement of hummingbirds? Based on my results it is safe to conclude that the daily flight patterns of these four age-sex classes are influenced by their morphology. Different morphological characters predict different measures of performance. It is likely, though, that tradeoffs occur between angular and translational motion and that long wings hinder angular motion yet are beneficial for straight forward flight. The percent of the day that a hummingbird spends foraging depends on many factors and can range from 8-70% of time (Gass 1978; Sutherland et al. 1982; Hixon et al. 1983; Gass and Sutherland 1985). Number, species, and density of flowers in a territory, amount of nectar produced, ambient temperature, and intrusion rates are among the numerous factors involved in determining this percentage (Gass and Montgomerie 1981; Gass and Roberts 1992). In one system (Gass and Sutherland 1985) a negligible proportion of foraging time was spent on long flights from perch to patch and back and between patches. In that system, short flights between flowers occupied from 40% of foraging time in the morning, when flowers were fuller and birds spent more time hovering at individual flowers, to 60% in afternoon when they 37 visited more flowers on each bout. Approximately 90-95% of these intra-floral flights were less than 2 m in length (Gass and Montgomerie 1981; Gass and Sutherland 1985). All flights between flowers necessarily involve both translational and angular motion and should be considered in a hummingbird's daily energy budget. Other short flights include movements within a territory from perch to patch, patch to patch and chasing intruders from the territory. The translational flight in this experiment was less than 0.5 m, and the turning occurred at a low velocity. Straight, short flights consist almost entirely of acceleration and deceleration (Chatters 1996), and this was true for the short turning flight in this experiment (Fig. 9). Chatters (1996) found that in straight forward flight (over 5 m) females reached higher maximum translational velocity and higher maximum accelerations than males, and differences in wing length accounted for most of this variation. Maximum acceleration was higher for juveniles (low WDL) which contradicts the suggestions of past studies that high WDL confers higher acceleration (Feinsinger and Chaplin 1975; Kodric Brown and Brown 1978; Feinsinger et al. 1979; Carpenter et al. 1993a). Will the results of this low velocity experiment be consistent for high velocity or long distance flights? I predict that juveniles should reach higher velocities than adults in straight forward flight. How does WDL change over time? The mass of a hummingbird may change dramatically on an hourly, daily or seasonal basis. For example, a 5% change in mass can occur during one feeding or defecating event (Carpenter et al. 1983; Collins and Paton 1989). During migration, birds weighing 3.0 - 3.5 g gain between 1.5 - 2.0 g before resuming migration, an increase in up to 50% of their body weight in 7-14 days (Carpenter et al. 1983). If we assume that wing length remains constant during this time, WDL can increase from 10-40 %, and will have a large impact on a 38 bird's ability to fly. According to my results, birds with higher WDL will be significantly slower after weight gain, and this extra load will also influence angular motion and turning ability. Carpenter et al. (1993) suggested that fat juvenile females should outcompete lean juvenile males due to their higher WDL, but my results suggest the opposite. Theoretical predictions of morphology and flight performance between species do not clearly predict what may occur within species. Larger (heavier) birds fly faster than smaller birds of different species, but increasing the mass of an individual does not necessarily increase its flight velocity. This distinction has led to much confusion in the literature. In all other types of birds, increasing load increases turning radius and thus decreases maneuverability (Hedenstrom 1992). However, it is important to note that hummingbirds (unlike any other birds) can turn while hovering at zero velocity and represent a distinct group of flying organisms. The Ecomorphology of Social Dominance (in the field) Hummingbirds are energy limited and must compete with other hummingbirds and insects to obtain nectar. Hummingbirds defend territories (when resources are moderately scarce) in order to have access to flowers with a limited nectar supply. Dominant birds hold and defend the most profitable territories. Generally, males have higher WDL than females in hummingbird species exhibiting male dominance (Feinsinger and Chaplin 1975; Feinsinger and Colwell 1978; Kodric Brown and Brown 1978). There are many suggestions in the literature about the evolution of short wings (and high WDL) in adult males. Kodric Brown and Brown (1978) suggested that rufous males (with higher WDL) have sacrificed efficient flight for aggressive ability. Differences in aggressive and foraging behaviours are evident between the sexes. Both sexes of S. rufus are territorial, but males defend smaller territories 3 9 that contain more flowers per unit area, compared to females that defend larger, less dense territories (Kodric Brown and Brown 1978). Based on my results, however, I conclude that the aerodynamic characteristics that influence hummingbirds' aggressive ability are unknown. Speculation by various authors has correlated colouration, age, experience and wing disc loading with the ability to defend a territory (Kodric Brown and Brown 1978; Ewald and Rowher 1980; Carpenter et al. 1993 a,b). When adult males are present in the wild, they dominate territories with the highest flower density (Kodric Brown and Brown, 1978). After males had migrated, juvenile males dominated adult females and juvenile females rarely obtained territories (Carpenter et al. 1993a,b). Juvenile males displace adult females from their territories and juvenile males acquire territories of higher flower densities faster than adult and juvenile females (Carpenter 1993 a,b). The Advantages of Youth Juveniles may have a dual energy advantage over adults. Juveniles fly faster between flowers and turn faster than adults of the same sex. If juveniles are fastest between flowers, they will gain more nectar (energy) in less time if they can reach more flowers, assuming that they extract nectar at equal rates to the adults. Not only can they gain more energy this way, they may have lower energy costs as a result of their lower WDL. The relationship for wing disc loading and power output has been investigated only for hovering (Pennycuick 1968; Weis-fogh 1972; Epting and Casey 1973; Epting 1980) and may hold for turning or accelerating flight (Chatters 1996). If we assume that this relationship applies to all types of flight, juveniles should spend less time and energy flying, gain energy faster and have a clear advantage in energy gain over adults. 4 0 Alternative Strategies Suggestions in the literature abound as to why one class of hummingbirds dominates another (e.g. Kodric Brown and Brown 1978; Ewald and Rowher 1980; Carpenter et al. 1993a,b). It is most likely that a combination of colouration, WDL and experience accounts for differences in territory quality. If an individual does not gain an adequate territory, it may sneak nectar from flowers on other birds' territories, or obtain nectar from flowers in areas outside of a territory. If adult males' territories contain the highest density of flowers, they must defend them vigorously. Defense of a territory involves expensive chases with intruders that often escalate into aerial duels. What if holding the territory with the highest density of flowers were not the best strategy for all age-sex classes? Alternative strategies of holding less dense territories may be beneficial to some classes if 1) they fly quickly and 2) they have lower flight costs as a result of their lower WDL. For rufous hummingbirds, the birds that conform to these two criteria are the juvenile males and females. Rufous adult males are noisy as their feathers produce a buzzing noise that is specific to this class. Extra wing slots in the male's wing produce noise during flight that may assist them during aggressive behaviours and sexual displays. The suggestion (Miller and Inouye 1983; Craig 1984) that these wing slots influence speed and maneuverability has not yet been tested. Not only are adult males brightly coloured; they are the slowest of all classes in short flights between flowers. In contrast if juveniles are fast, inconspicuous, and silent, this must be advantageous for sneaking onto territories, winning chases and, stealing nectar without being detected. Within a similar system of dimorphic hummingbirds (Calypte anna), Ewald and Rowher (1980) suggested that duller coloration leads to better intrusion success for the females. They suggested that dominance occurs 4 1 because the contested resource (territory) is more valuable to adult males than to adult females and juveniles. Juvenile C. anna are better at stealing food because they are duller in colouration than the adults. They do not need to defend territories as stealing nectar allows them to gain nectar at an equal rate to that of the adult females. As intruders, juveniles spend more time in flight which is energetically less expensive than adult males. With this combination of dull coloration and cheap flight, juvenile C. anna may not need to defend dense territories of flowers. This argument may be true for rufous hummingbirds as the juveniles have similar colouration and wing length dimorphisms. The exploitation hypothesis suggests that different age-sex classes use alternative foraging strategies because their morphologies differ, allowing them to use alternate resources (Selander 1966; Desrochers 1989; Carpenter et al. 1993b). Ultimately each class prefers a resource most suited to its morphology, foraging capability, and ability to defend a territory (Carpenter et al. 1993a). Under the interference competition hypothesis, different age-sex classes have different morphologies which allow for behavioural dominance (Morse 1980; Peters and Grubb 1983; Morton 1990; Carpenter et al. 1993a). Carpenter et al. (1993) suggested that subordinate juveniles and adult females compensate with lower flight costs (low WDL), increased nectar robbing (juveniles' dull coloration) and even experience (adult females). One age-sex class is dominant and the others exploit resources or adopt behaviours to utilize other resources. These two hypotheses for rufous hummingbirds foraging strategies are not mutually exclusive (Carpenter et al. 1993 a, b). However, the results of this chapter lend support to the exploitation hypothesis because juveniles' long wings allow for an advantage in flight performance. By using different resources, each classes' foraging strategy is most suited to their morphology. An interesting observation in the field is that all age-sex classes gain mass at the same rate regardless of their foraging strategy (Carpenter et al. 1993b). 4 2 Despite juvenile males occupying territories with the highest flower density, juvenile and adult females gain weight at the same rate, simultaneously, but using different parts of a local habitat. Under the assumptions of Feinsinger and Chaplin (1975), Carpenter et al. (1993b) dismissed the hypothesis that females gain mass at an equal rate because they fly faster between flowers. Similarly, the high WDL of juvenile males should make them faster (Hedenstrom 1992; Webb et al. 1992; Norberg 1995). Granted that birds spend a large portion of their day flying between flowers, I predict that the birds that are fastest between flowers will reach more flowers and therefore gain mass at a faster rate. As a result, juveniles should gain mass as quickly as adults despite defending less dense territories. Juveniles may also gain an advantage with faster flights between flowers and due to lower flight costs than adults. If Temeles and Roberts' (1993) finding that females take nectar faster than males from flowers with long corollas applies to juvenile females as well (similar bill length), these factors may combine to explain the equality of mass gain between adult females and juvenile males. The Advantages of Short Wings If long winged individuals (juveniles) have energetic and speed advantages, why have adult males evolved short narrow wings that make them slower and noisier as well as making their flight more costly? Short wings must be advantageous in some way if they are costly in terms of power output. All of the age-sex classes engage in territorial disputes involving aerial combat. Short wings may confer advantages in these types of aerial maneuvers. This experiment does not investigate an individual's ability to maneuver in a high speed aerial fight. Ideally, the maneuvers involved in aerial combat should be analyzed, although they may be difficult to measure in captive birds. Chapter Four investigates high velocity turning through a tunnel with barriers in an 4 3 attempt to dissect differences between classes in both high and low velocity turning. The evolution of short wings in adult males may have been a result of selection pressure in performing elaborate courtship dives. This display involves repeatedly flying high in the air and diving towards the females at high velocities (Johnsgard 1983). On the other hand, adult females may have developed longer wings to help carry the weight of the eggs which can increase their body mass up to 17% (Wells 1993). Adult males spend no time foraging for the young in the nest, leaving the females to forage for insects and nectar while the young are still in the nest. Further investigation of the details of flight performance and morphology may help to answer these expansive questions concerning the evolution of wing size. Implications The results of this experiment clearly contradict the current literature on hummingbird performance. Juvenile females have the longest wings, lowest wing disc loading and fly faster between feeders than birds with shorter wings. This result has never been published. However, it was expected that birds with shorter wings and higher WDL would be more maneuverable and more agile. I suggested in Chapter 2 that the terms maneuverability and agility need to be operationalized to include definitions for hummingbirds before statements are made to defend or abandon these ideas. Aerodynamic theory makes no predictions about whether long or short winged hummingbirds should turn faster, or fly faster than one another at any speed. The more I read the literature, the more confused I become about the performance advantages and disadvantages high and low WDL should or should not afford. Rigorous testing in the laboratory of how variation in morphology influences performance between individuals and classes is necessary to clearly 4 4 answer these ecomorphological questions. Future studies in the field should continue to investigate the proportion of time each age-sex class spends in defense, forward flight, combat and especially turning. Ecomorphological studies in the laboratory of high speed translational and angular motion will aid in understanding how morphology of the birds influences their ecology. This study is not the first to question Feinsinger and Chaplin's (1975) conclusion that WDL reflects foraging strategies for all hummingbirds species. Other recent papers have suggested that the conclusions drawn from this correlation are invalid and that there are problems with the prediction that high WDL is correlated with the ability to remain on a profitable territory (Snow and Snow 1980; Gill 1985; Collins and Paton 1989; Stiles 1995). The functional relationship between morphology and ecology is complex and more studies of how morphology can influence the performance capability of individuals and classes may help in understanding dominance and territoriality more completely. A better understanding of ecomorphological relationships of hummingbirds both within and between species will lead to a greater understanding of maneuverability and agility in hovering organisms. 4 5 video camera light feeder perch F i g u r e 1 : The exper imental setup. Two feeders are 0.5 m apart. Birds fly back and forth between feeders 9 t imes within a trial as the l ights indicate food is avai lable. Bird returns to perch between trials. All f l ights within trials are recorded f rom above by the camera , and a computer records t imes of arrival and departure to a feeder and perch. Feeders are mounted 0.5 m apart on the long wal ls of an isosceles t r iangular room. 4 6 F i g u r e 2 : P r o c e d u r e f o r r e c o r d i n g p o s i t i o n a n d o r i e n t a t i o n o f b i r d ( A d u l t f e m a l e 1 6 ) ; f r a m e s 2 a n d 5 a r e s h o w n . L e f t i n s e t : e n l a r g e m e n t o f f r a m e 2 s h o w i n g l i n e a l o n g a x i s o f b i r d a n d i t s m i d p o i n t . R i g h t i n s e t : d i a g r a m m a t i c r e p r e s e n t a t i o n s h o w i n g p o s i t i o n a n d a n g l e . 47 F i g u r e 3: E l a p s e d t i m e f r o m v i d e o ana l ys i s is c o m p a r e d to e l a p s e d t i m e d a t a f r o m c o m p u t e r in s e c o n d s . 48 0 . 6 5 0 . 3 5 H 1 1 1 0.5 0 .6 0 .7 0 .8 Mean Elapsed T ime (s) F i g u r e 4 : M e a n e l a p s e d t i m e c o m p a r e d to t h e fas tes t e l a p s e d t i m e fo r e a c h b i rd . 49 30 2 5 + F i g u r e 5: F r e q u e n c y d is t r ibu t ion of e l a p s e d t i m e s fo r a r e p r e s e n t a t i v e j u v e n i l e m a l e b i rd ( J M 3 6 ) . T h e ver t i ca l l ine r e p r e s e n t s t h e cut -o f f po in t f o r d a t a u s e d in t h e a n a l y s i s . T h i s l ine e x c l u d e s e l a p s e d t i m e s 0.8 s a n d larger . 50 Age-Sex C lass F igure 6: Average wing lengths of the four age-sex classes used in experiments. Measurements were taken prior to each experiment. N=5 in each case. Bars represent standard errors. A M = adult males, A F = adult females, JM= juvenile males, JF= juvenile females. 51 4 7 4 5 E E 4 3 sz •*—' D) c 41 0) •J 3 9 • • • • • • • 3 7 3 5 3 .5 4 4 .5 Mass (g) 5.5 F i g u r e 7: M a s s of 18 ind iv idua ls is p lo t ted a g a i n s t w i n g l e n g t h . 52 0.8 Figure 8: Mean elapsed t ime is plotted against wing length for 18 individual birds. Closed circles represent the mean of each age-sex class, with bars representing standard error. A M = adult males, AF= adult females, JM= juvenile males, and JF= juvenile females. 53 1.4 1400 10 20 Time (frames) Time (frames) 20000 -20000 Time (frames) Time (frames) F i g u r e 9: V e l o c i t y a n d a c c e l e r a t i o n in t rans la t i ona l a n d a n g u l a r m o t i o n fo r j u v e n i l e m a l e s ( d i a m o n d s ) a n d adu l t m a l e s ( s q u a r e s ) is p lo t ted o v e r t i m e t o d e m o n s t r a t e d i f f e r e n c e s b e t w e e n c l a s s e s . J u v e n i l e f e m a l e s a n d a d u l t f e m a l e s a r e i n t e r m e d i a t e in b o t h c a s e s . A n g u l a r d a t a w a s cu t off a f ter 15 f r a m e s . 54 A B Figure 10: Four measures of translational performance are plotted against wing length. A) max imum velocity B) maximum acceleration C) t ime to reach VMax and D) t ime to reach AVMax. 55 A B 0.02 0.04 0.06 0.08 0.02 0.04 0.06 0.08 W D L ( g / c m A 2 ) W D L ( g / c m A 2 ) 0.04 0.06 W D L ( g / c m A 2 ) 0.08 0.02 0.04 0.06 W D L ( g / c m A 2 ) 0.08 Figure 1 1 : Four measures of translational performance are plotted against wing disc loading. A) max imum velocity B) maximum acceleration C) t ime to reach VMax and D) t ime to reach AMax. Each value is an average of 5 flights per bird, for a total of 15 individuals. 56 B 0.5 Figure 12: Four measures of translational performance are plotted against mass A) maximum velocity B) maximum acceleration C) T ime to VMax and D) Time to AMax. 57 A B 1800 1600 g 1400 <D 2. 1200 X re > < 1000 800 + 600 + + 35 40 45 W i n g L e n g t h ( m m ) 50 600 0.03 0.04 0.05 0.06 0.07 W D L ( g / c m A 2 ) 25000 <, 20000 (A (U U) .§ 15000 X re 5 < 10000 < 5000 <* 20000 (0 "w a> a> o « 15000 X re < 10000 5000 0.03 0.04 0.05 0.06 W D L ( g / c m A 2 ) 0.07 Figure 13: Four measures of angular performance are plotted against morphology with juveniles (closed squares) and without juvenile males (open diamonds). AVmax is plotted against A)wing length and B)WDL. AAmax is plotted against C) mass and D) WDL. Dashed line is linear regression with all data, and solid line is linear regression after juvenile males have been removed. 58 Translational B Class Class Angular 21000 3000 Class Class Figure 14: Four performance measures in translational and angular motion are plotted for each age-sex class. A) VMax B) AMax C) AVMax and D) AAMax Bars represent standard error. 59 B F igure 15: M e a n n u m b e r of d e g r e e s t u r n e d in t h e f i rst 5 f r a m e s ( 0 . 1 7 s ) . A ) fo r e a c h a g e - s e x c l a s s B) c o m p a r e d to A V m a x 60 Figure 16: Schematic partial plan view of the experiment to illustrate the positions of VMax, AVMax, AMax, and AAMax. Rectangles represent the approximate range of the parameters for all age-sex classes. 61 B VMax AMax X (cm) X (cm) A V M a x A A M a x 3.5 - i -3 --2.5 --? 2 1.5 -> 1 0.5 0 -0.5 X (cm) X (cm) F igure 17: Position of A) VMax B) AMax C) AVMax and D) AMax for four age-sex classes. Bars represent standard errors. A M = adult male, JM = juvenile male, AF= adult female, JF=juvenile female. 62 A B Figure 18: A) Maximum translational velocity is plotted against t ime to reach max imum translational velocity. B) Maximum translational acceleration is plotted against t ime to reach Amax C) Maximum translational velocity is plotted against maximum translational acceleration. D) t ime to reach VMax is plotted against t ime to reach Amax 63 1 8 0 0 6 0 0 0 0 .05 0.1 0 .15 0.2 T i m e t o reach A V M a x (s) B 2 5 0 0 0 ^ 2 0 0 0 0 0 0 .05 0.1 0 . 1 5 0.2 T i m e t o reach A A M a x (s) Figure 19: A) Maximum angular velocity is plotted against max imum angular acceleration. B) Max imum angular acceleration is plotted against t ime to reach AAmax C) Max imum angular velocity is plotted against maximum angular acceleration D) Time to reach AAmax is plotted against t ime to reach AVmax. 64 T a b l e 1 : G e n e r a l L i n e a r M o d e l o f m o r p h o l o g y a s p r e d i c t e d b y a g e a n d s e x c l a s s . T h e t h r e e m o r p h o l o g i c a l m e a s u r e s a r e a b b r e v i a t e d a s f o l l o w s : w i n g l e n g t h ( W L ) , m a s s , a n d w i n g d i s c l o a d i n g ( W D L ) . F R a t i o S e x p v a l u e F r a t i o A g e p v a l u e W L 6 5 . 6 9 < 0 . 0 0 1 * * 4 5 . 3 7 < 0 . 0 0 1 * * M a s s 0 . 6 4 0 0 . 4 4 1 0 . 1 3 5 0 . 7 2 0 W D L 7 . 6 7 0 . 0 1 8 * 4 . 1 7 0 0 . 0 6 6 ** p < 0 . 0 1 * p < 0 . 0 5 T a b l e 2 : A v e r a g e w i n g l e n g t h , m a s s a n d W D L m e a s u r e m e n t s f o r a l l f o u r a g e - s e x c l a s s e s . E a c h b i r d s ' m e a s u r e m e n t s w e r e t a k e n i m m e d i a t e l y a f t e r a n e x p e r i m e n t . M e a s u r e m e n t s a r e a n a v e r a g e o f t h e s a m p l e ± s t a n d a r d e r r o r . A M = a d u l t m a l e , A F = a d u l t f e m a l e , J M = j u v e n i l e m a l e , a n d J F = j u v e n i l e f e m a l e . A g e - S e x C l a s s (n ) W i n g L e n g t h ( m m ) M a s s (g ) W D L ( g / c m 2 ) A M (5) 3 9 . 4 8 ± 0 . 4 1 4 . 4 3 + 0 . 1 3 0 . 0 5 8 ± 0 . 0 0 3 A F (5) 4 2 . 4 7 ± 0 . 3 7 4 . 2 1 ± 0 . 3 6 0 . 0 4 8 ± 0 . 0 0 4 J M (4) 4 1 . 9 7 ± 0 . 1 3 4 . 2 8 ± 0 . 3 0 0 . 0 4 9 ± 0 . 0 0 4 J F (4) 4 5 . 3 2 ± 0 . 2 9 4 . 1 3 ± 0 . 2 0 0 . 0 4 1 ± 0 . 0 0 2 6 5 T a b l e 3: L i n e a r r e g r e s s i o n s o f t h r e e m o r p h o l o g i c a l m e a s u r e s a r e c o m p a r e d t o f o u r m e a s u r e s o f t r a n s l a t i o n a l p e r f o r m a n c e . R s q u a r e d v a l u e s a r e a b o v e p v a l u e s f o r e a c h c a t e g o r y ( n = 1 8 ) . A b b r e v i a t i o n s a r e a s f o l l o w s : e l a p s e d t i m e ( E T ) , m a x i m u m v e l o c i t y ( V M a x ) , t i m e t o r e a c h m a x i m u m v e l o c i t y ( t V M a x ) , m a x i m u m a c c e l e r a t i o n ( A M a x ) , a n d t i m e t o r e a c h m a x i m u m a c c e l e r a t i o n ( t A M a x ) . W i n g L e n g t h W i n g D i s c L o a d i n g M a s s E T r 2 = 0 . 4 1 9 0 . 2 8 4 0 . 0 8 2 p < 0 . 0 0 1 * * * 0 . 0 2 * 0 . 3 0 V M a x 0 . 3 4 5 0 . 3 3 4 0 . 3 3 4 0 . 0 2 * 0 . 0 2 * 0 . 0 2 * t V M a x 0 . 4 8 1 0 . 5 8 0 . 2 7 4 0 . 0 0 4 * * 0 . 0 0 1 * * * 0 . 0 4 5 * A M a x 0 . 4 3 0 . 4 1 0 . 1 3 9 0 . 0 0 8 * * 0 . 0 1 * * 0 . 1 7 t A M a x 0 . 3 1 0 . 2 6 9 0 . 0 6 7 0 . 0 3 * 0 . 0 4 8 * 0 . 3 5 ** p < 0 . 0 1 * p < 0 . 0 5 T a b l e 4: M u l t i p l e r e g r e s s i o n a n a l y s i s o f w i n g l e n g t h a n d m a s s f o r f i v e t r a n s l a t i o n a l p e r f o r m a n c e m e a s u r e s ( n = 1 8 ) . T h e p e r f o r m a n c e m e a s u r e s a r e a b b r e v i a t e d a s f o l l o w s : e l a p s e d t i m e b e t w e e n f e e d e r s ( E T ) , m a x i m u m v e l o c i t y ( V M a x ) , t i m e t o r e a c h m a x i m u m v e l o c i t y ( t V M a x ) , m a x i m u m a c c e l e r a t i o n ( A M a x ) , t i m e t o r e a c h A m a x ( t A M a x ) . W i n g l e n g t h M a s s F r a t i o p v a l u e F r a t i o p v a l u e E T 5 . 5 7 7 0 . 0 3 6 * 0 . 2 6 8 0 . 6 1 3 V M a x 5 . 2 7 5 0 . 0 4 0 * 4 . 9 9 5 4 0 . 0 4 5 * t V M a x 1 0 . 1 2 0 . 0 0 8 * * 3 . 8 0 9 0 . 0 7 5 A M a x 7 . 6 8 0 0 . 0 1 7 * 0 . 9 7 2 0 . 3 4 4 t A M a x 4 . 5 4 9 2 0 . 0 5 4 0 . 2 4 5 0 . 6 2 9 ** p < 0 . 0 1 * p < 0 . 0 5 6 6 T a b l e 5 : L i n e a r r e g r e s s i o n s o f t h r e e m o r p h o l o g i c a l m e a s u r e s a r e c o m p a r e d t o f o u r m e a s u r e s o f a n g u l a r p e r f o r m a n c e ( n = 1 8 ) . L i n e a r r e g r e s s i o n s a r e r e p e a t e d a f t e r t h e e x c l u s i o n o f j u v e n i l e m a l e s ( n = 1 4 ) . R s q u a r e d v a l u e s a r e t h e t o p v a l u e s , a n d p v a l u e s a r e t h e b o t t o m f o r e a c h c a t e g o r y . A b b r e v i a t i o n s a r e a s f o l l o w s : m a x i m u m a n g u l a r v e l o c i t y ( A V M a x ) , t i m e t o r e a c h m a x i m u m a n g u l a r v e l o c i t y ( t A V M a x ) , m a x i m u m a n g u l a r a c c e l e r a t i o n ( A A M a x ) , a n d t i m e t o r e a c h m a x i m u m a n g u l a r a c c e l e r a t i o n ( t A A M a x ) . W i t h W L J u v e n i l e W D L M a l e s M a s s W i t h o u t W L J u v e n i l e W D L M a l e s M a s s A V M a x 0 . 2 3 4 0 . 0 6 7 0 . 2 6 0 . 0 5 2 0 . 0 9 7 0 . 2 5 8 0 . 7 1 1 0 . 0 0 1 * * 0 . 6 5 8 0 . 0 0 2 * 0 . 2 1 7 0 . 1 4 9 A A M a x 0 . 0 9 5 0 . 2 6 3 0 . 2 2 9 0 . 0 7 1 0 . 1 5 6 0 . 1 4 6 0 . 1 9 3 0 . 1 7 6 0 . 5 2 3 0 . 0 1 2 * 0 . 4 3 5 0 . 0 2 7 * ** p < 0 . 0 1 * p < 0 . 0 5 T a b l e 6: J u v e n i l e m a l e s h a v e b e e n r e m o v e d f o r t h i s a n a l y s i s ( n = 1 4 ) . M u l t i p l e r e g r e s s i o n a n a l y s i s o f w i n g l e n g t h a n d m a s s f o r f o u r a n g u l a r p e r f o r m a n c e m e a s u r e s . T h e p e r f o r m a n c e m e a s u r e s a r e a b b r e v i a t e d a s f o l l o w s : m a x i m u m a n g u l a r v e l o c i t y ( A V M a x ) , t i m e t o r e a c h m a x i m u m a n g u l a r v e l o c i t y ( t A V M a x ) , m a x i m u m a n g u l a r a c c e l e r a t i o n ( A A M a x ) , t i m e t o r e a c h A A m a x ( t A A M a x ) . W i n g l e n g t h F r a t i o p v a l u e M a s s F r a t i o p v a l u e A V M a x 1 9 . 1 0 0 . 0 0 2 * * 1 . 9 9 5 0 . 1 9 6 t A V M a x 1 . 2 8 4 0 . 2 9 0 0 . 6 6 8 0 . 4 3 1 A A M a x 1 . 1 4 8 0 . 3 1 5 5 . 0 5 5 0 . 0 5 6 t A A M a x 1 . 9 9 3 0 . 1 9 6 2 . 8 4 6 0 . 1 3 0 ** p < 0 . 0 1 * p < 0 . 0 5 67 Tab le 7: G e n e r a l L i n e a r M o d e l s o f t r a n s l a t i o n a l a n d a n g u l a r p e r f o r m a n c e m e a s u r e s . C a t e g o r i e s o f a g e a n d s e x w e r e a n a l y s e d f o r n i n e p e r f o r m a n c e m e a s u r e s . T h e p e r f o r m a n c e m e a s u r e s a r e a b b r e v i a t e d a s f o l l o w s : e l a p s e d t i m e b e t w e e n f e e d e r s ( E T ) , m a x i m u m v e l o c i t y ( V M a x ) , t i m e t o r e a c h m a x i m u m v e l o c i t y ( t V M a x ) , m a x i m u m a c c e l e r a t i o n ( A M a x ) , t i m e t o r e a c h A m a x ( t A M a x ) , m a x i m u m a n g u l a r v e l o c i t y ( A V M a x ) , t i m e t o r e a c h A V M a x ( t A V M a x ) , m a x i m u m a n g u l a r a c c e l e r a t i o n ( A A M a x ) , a n d t i m e t o r e a c h A a m a x ( t A A M a x ) . S e x F r a t i o p v a l u e A g e F r a t i o p v a l u e E T 0 . 8 9 5 0 . 3 6 5 5 . 7 4 8 0 . 0 3 5 * V M a x 1 . 0 7 5 0 . 3 2 2 3 . 4 8 5 0 . 0 8 9 t V M a x 1 . 7 2 3 0 . 2 1 6 1 1 . 0 3 0 . 0 0 7 * * A M a x 1 . 2 2 4 0 . 2 9 2 6 . 9 7 8 0 . 0 2 3 * t A M a x 0 . 1 7 5 0 . 6 8 4 1 7 . 1 2 0 . 0 0 2 * * A V M a x 0 . 1 0 2 0 . 7 5 6 1 1 . 1 8 0 . 0 0 7 * * t A V M a x 1 . 1 1 9 0 . 3 1 3 4 . 1 5 3 0 . 0 6 6 A A M a x 0 . 7 7 9 0 . 3 9 6 0 . 2 1 3 0 . 6 5 3 t A A M a x 0 . 1 7 0 0 . 6 8 8 2 . 8 1 0 . 1 2 2 ** p < 0 . 0 1 * p < 0 . 0 5 Tab le 8: L i n e a r r e g r e s s i o n s o f f o u r p e r f o r m a n c e m e a s u r e s a r e c o m p a r e d t o e l a p s e d t i m e b e t w e e n f e e d e r s ( n = 1 8 ) . R s q u a r e d v a l u e s a r e a b o v e p v a l u e s f o r e a c h c a t e g o r y . A b b r e v i a t i o n s a r e a s f o l l o w s : m a x i m u m v e l o c i t y ( V M a x ) , t i m e t o r e a c h m a x i m u m v e l o c i t y ( t V M a x ) , m a x i m u m a c c e l e r a t i o n ( A M a x ) , t i m e t o r e a c h m a x i m u m a c c e l e r a t i o n ( t A M a x ) m a x i m u m a n g u l a r v e l o c i t y ( A V M a x ) , t i m e t o r e a c h m a x i m u m a n g u l a r v e l o c i t y ( t A V M a x ) , m a x i m u m a n g u l a r a c c e l e r a t i o n ( A A M a x ) , a n d t i m e t o r e a c h m a x i m u m a n g u l a r a c c e l e r a t i o n ( t A A M a x ) . T r a n s l a t i o n a l M o t i o n E T A n g u l a r M o t i o n ET V M a x r * = 0 . 6 6 3 A V M a x r ^ O . 2 5 6 p < 0 . 0 0 1 * * p = 0 . 0 5 4 t V M a x 0 . 7 1 7 t A V M a x 0 . 1 3 9 < 0 . 0 0 1 * * 0 . 1 7 1 A M a x 0 . 6 5 3 A A M a x 0 . 1 3 9 <o.oor* 0 . 1 7 t A M a x 0 . 5 3 1 t A A M a x 0 . 1 7 9 0 . 0 0 2 * * 0 . 1 7 ** p < 0 . 0 1 * p < 0 . 0 5 68 CHAPTER 4 HIGH VELOCITY TURNING IN FOUR AGE-SEX CLASSES OF RUFOUS HUMMINGBIRDS Hummingbirds fly at high speeds through the clutter of underbrush during chases in the field and courtship displays. High speed flights often involve changing direction to avoid predators, chase intruders, display for potential mates, and avoid collisions with objects. The type and degree of maneuverability necessary may be highly variable, both among individuals of different morphology and among situations. Although general statements are often made in the literature, the specific wing and body morphologies that allow most hummingbirds to be highly maneuverable have not yet been investigated. Aerodynamic predictions of how morphology influences maneuverability are based on the current bat and bird literature, and do not necessarily apply to hummingbirds. Wing size and shape may influence the ability of hummingbirds to turn during high speed flight. Maneuverability is defined as the minimum space required to turn at a given speed, the ability to turn tightly and the ability to fly through clutter (Aldridge 1986; Norberg and Rayner 1987). Conversely, agility is defined as the ability to turn quickly or the rate at which turns can be initiated, and can be achieved during fast or slow turns (Norberg and Rayner 1987). In the bird literature, maneuverability is often used as a blanket term to encompass both maneuverability and agility, and here I will define only agility, as the ability to make tight, fast turns (see Chapter 2). Aerodynamic theory also predicts that individuals with low wing loading are the most maneuverable because it allows them to make small radius turns (AldridgeT 986; Norberg and Rayner 1987; 69 Norberg 1990). It follows that decreases in mass would increase maneuverability (Hedenstrom 1992), suggesting that light birds should turn more tightly (be more maneuverable) than heavier birds. In addition, aerodynamic theory predicts that individuals with high wing loading reach higher flight speeds than those with low wing loading (Pennycuick 1975; Norberg and Rayner 1987; Hughes and Rayner 1991). Either mass or wing length can be altered experimentally to manipulate wing loading. Because it is unusual to alter the wing length of animals, most studies investigate the relationship between wing loading and maneuverability by manipulating mass (Aldridge and Rautenbach 1987; Hughes and Rayner 1991; Chandler and Mulvihill 1992; Witter et al. 1994; Metcalfe and Ure 1995). All of these authors found that increasing an individual's mass (therefore its wing loading), decreased its maneuverability. This contradicted the theoretical prediction that increasing mass will increase flight speed. However, artificially loading individuals to test the effects of increased mass on maneuverability may not be a valid test because most forms of loading significantly alter posture during flight (Metcalfe and Ure 1995). Among species, speed is proportional to mass such that larger birds fly faster than smaller birds, and speed is proportional to the square root of wing loading (Pennycuick 1975; Norberg and Rayner 1987; Rayner 1988). Within species it is predicted that cruising velocity should be directly proportional to mass. Hughes and Rayner (1991) stated that there is a lack of intraspecific and intra-individual comparisons of theoretical predictions concerning morphology and maneuverability. It is important to consider that all aerodynamic predictions mentioned apply to flying organisms using only flapping flight, and do not include hovering flight used by hummingbirds. Hummingbirds are compared using wing disc loading instead of wing loading as it considers the area of air swept by the wing 70 disc. Feinsinger and Chaplin (1975) predicted that individual hummingbirds with the highest wing disc loading defend the most profitable territories, as a result of their increased maneuverability. However, they used maneuverability to describe kinds of flight thought to be important to hummingbirds, and they neither measured it nor attempted to estimate it quantitatively. Although hummingbird maneuverability has never been measured, the Feinsinger and Chaplin (1975) study has been used to support many inter- and intraspecific comparisons of morphology and foraging strategies. This chapter is a first attempt at measuring the turning ability of hummingbirds. In recent years, many authors (Aldridge 1986; Aldridge and Rautenbach 1987; Moller 1991; Witter et al. 1994; Metcalfe and Ure 1995) have assessed the maneuverability of bats and birds by flying them through tunnels equipped with hanging strings, poles, and partitions to represent cluttered environments. Individuals who contact the barriers least often are considered most maneuverable. This design is inappropriate for hummingbirds because they rarely bump into obstacles, unlike other birds (Cuthill and Guilford 1989). For example, in field situations while other species tangle themselves in mistnets, hummingbirds artfully dodge the nets as they come into view (pers. obs.) If birds that can fly through the most cluttered environment are the most maneuverable and hummingbirds do not often hit obstacles, then the time taken to maneuver through a given degree of clutter may be a more appropriate measure of turning ability. Instead of counting the contacts with barriers, I measured the translational and angular velocity and acceleration of individuals continuously as they flew through an obstacle course of barriers. I considered hummingbirds that flew fastest through the barriers to be the most "agile". Although the design of this experiment was typical of tests of maneuverability in bats and birds, I believe that the time taken to fly through clutter is a better 7 1 measure of agility; especially under the relatively high-speed conditions I used. Aerodynamic theory predicts that birds with long wings (low wing loading) are more "maneuverable" and those with short wings more "agile". In Chapter three I demonstrated that hummingbirds with the longest wings and lowest WDL have the lowest elapsed time between two feeders 0.5 m apart. This short flight included several component actions: backing out of the first feeder, turning (angular velocity and acceleration), flying forward (translational velocity and acceleration), and slowing down (deceleration) before probing the next feeder. Longer wings allowed hummingbirds to reach higher translational and angular velocities than short winged individuals. The design of the experiment in Chapter 3 has never been used for other types of birds, and this type of short flight involving a turn may be difficult to categorize as either maneuverability or agility. Morphological differences in wing length, mass and WDL should influence agility directly. Here I compare elapsed time of four age-sex classes of rufous hummingbirds through a tunnel with three barriers to investigate turning at high speed. Video analysis allowed for a detailed examination of how individuals negotiated turns as they flew around the barriers. Methods: Experimental Environment A rectangular tunnel, 4.8 m long x 0.56 m wide x 0.6 m high, was constructed with 3 barriers 0.3 m wide x 0.6 m high (Fig. 20). A green aluminum panel covered one end of the tunnel and housed a feeder and LED. The 3 mm diameter feeder was surrounded by an orange Avery label 19 mm in diameter. A 4 mm diameter LED was located 2 cm above the feeder, and another LED was attached to the first barrier. The orange label and the lights 72 cued the food source and food availability, respectively (Brown and Gass 1993). The perch was 47 cm high and was located 16 cm from the end of the tunnel. The straight line distance between the perch and feeder was 4.64 m, but flight paths were necessarily longer than this. A computer controlled LEDs, dispensed nectar solution, and recorded the elapsed time between the perch and feeder by monitoring photocells (Tooze and Gass 1985; Tamm and Gass 1985; Brown and Gass 1993; Thompson 1994; Chatters 1996). When a bird arrived at the feeder or perch, its bill or body triggered a photocell built into the feeder. In turn this triggered a solenoid pump (General Valve Corporation Series 3) to immediately dispense 2 |il of sucrose solution (20% sucrose w/w) into a reservoir. Birds fed while hovering, by probing their bills through a hole into a reservoir (disposable syringe needle fittings). The tunnel was equipped with barriers 45 cm apart and overlapped by 7 . , cm so that the bird could not view the feeder when sitting on the perch (Fig. 20). From the perch, the bird flew 2.91 m before encountering the first barrier. The videocamera (Sony Handycam CCD-V101) was attached to the ceiling above the three barriers, 1.4 m from the top of the tunnel and it recorded all flights within an experiment. Plastic mesh covering allowed for the camera to view the bird, without the bird escaping from the enclosure (the tunnel section was 0.9 m x 0.3 m wide). The computer recorded the time the bird left the perch, arrived at the feeder, left the feeder, and arrived back at the perch. From this information I calculated elapsed time between perch and feeder, and feeder and perch. The Subjects Five adult male, five adult female, four juvenile male and four juvenile female hummingbirds (Selasphorus rufus ) were selected from a laboratory 73 population of 33 birds. Birds were captured at Sumas Mountain and Port Mellon, British Columbia in May 1995 and Rosewall Creek (Vancouver Island), British Columbia in June 1995. The birds were housed individually in wire mesh cages (60 x 60 x 60 cm), in a room with a controlled photoperiod (15h: 9h) at the UBC Animal Care Facility. They were fed Nektar plus (Nekton USA Inc.), a commercial hummingbird nectar with supplemented soya protein, on weekdays, and 20% sucrose solution on weekends. The experiments ran for 4 months (August - November 1995). Training I trained all birds prior to the experiment. For several days the ad libitum feeders in the home cages were equipped with a green facing and an orange Avery label to help birds learn the experimental arrangement. First, I placed a bird in half the length of the tunnel with an ad libitum feeder at one end and a perch at the other end. After it perched and fed consistently from the ad libitum feeder, I opened the entire tunnel and moved the feeder to the end of the tunnel opposite the feeder. The bird was left for 4-6 hours in the 4.8 m tunnel with the perch and the ad libitum feeder. In order to feed, it had to fly from its perch at one end of the tunnel, to the feeder at the opposite end of the tunnel, a distance of 4.64 m and back. During this time, the green barriers were out of the way, flat against the sides of the tunnel. Next, I trained the bird to feed from the computer controlled wall feeder. LEDs at the perch end of the tunnel and above the feeder were turned on to indicate the food source and food availability. The bird could not see the LED above the feeder when the barriers were in place, so it was necessary to use a LED nearer the perch. When the bird was acclimated to the tunnel and feeding from the wall feeder, the three barriers that had been lying flat against the tunnel 74 walls were moved into place, in stages over a 2 hour period. Initially a path was visible from the perch to the feeder, but as the barriers were moved into place the bird could no longer see the feeder from its perch and had to swerve to avoid hitting them. With the barriers in place, the birds trained until they were flying consistently around the barriers without pausing. A video monitor enabled me to assess this from outside of the room with the tunnel. Experimental Protocol Trials began with the LEDs turning on. The bird left its perch, flew 2.91 m to the first barrier, around the 3 barriers, and decelerated to arrive at the feeder. The bird then probed the feeder to receive 2 Lil of 20 % (w/w) sucrose solution. It backed out, turned away from the feeder, flew 0.83 m to the first barrier, through the barriers and back to the perch. Only when it perched did the LED turn back on, indicating that more food was available. To ensure no asymmetries or preferences for left and right turns, each bird was trained with two barriers on the right or left side, tested for 45 minutes on that side, trained again on the opposite side until it was flying consistently without hesitation, and then tested for 45 minutes on the opposite side. Each experiment lasted 1.5 hours. During an experiment birds flew back and forth (round trip) through the tunnel an average of 273.3 ± 25.7 times and consumed 547 Lil of sucrose solution. The running clock on the video camera was set according to the computer time at the start of each experiment, to ensure that computer data could be matched easily with the video tape. I used computer time to identify trials on the videotape, and timed events within trials by counting video frames. After the experiment ended, I removed the bird from the enclosure, weighed it on an electronic balance (±0.0 lg), measured the length of both wings with hand held 75 calipers, and returned it to its home cage. Because hummingbird mass varies depending on the time of last feeding, the time of day, the season, and whether the bird has recently defecated, I weighed birds immediately following the experiment to minimize this variation. Wing disc loading was calculated as the mass divided by the area swept out by the wings using the formula: WDL(g/cm 2)= m (b/2)2 n , where b=wing span = 2.5 * wing length (cm) and m= mass (g) (Feinsinger et al. 1979). Analysis of Video Records Using elapsed times between perch and feeder in the computer record, I selected 10 flights by each of 10 birds for video analysis (5 flights towards the feeder and 5 flights towards the perch). For each bird, I retained and analyzed the 5 flights with elapsed times nearest the individual's mode in both directions of flight through the tunnel. Some video tapes were destroyed during analysis, leaving a total of 10 birds (2 adult males, 2 adult females, 2 juvenile males and 4 juvenile females) with 10 flights each to analyze. The videotapes were transferred to a Macintosh computer using the program Aldus Moviemaker. Each frame was processed as a digital image and placed into a stack of images which represented all frames in each flight. These data were saved on a CD-ROM for analysis with the program NIH Image. Using this program, I logged the position in the horizontal plane (x,y) and orientation (angle with respect to the tunnel) of the bird in each frame of each flight. Analysis started as the bird passed the first barrier and ended as it passed the third barrier. 7 6 For each frame (1/30 s) an image of the bird in flight appeared on the computer screen (Fig. 21). Using a mouse, I placed a straight line from the base of the bill to the tip of the tail (i.e. along its spine). The computer recorded the midpoint, angle, and length of this line. I also recorded the positions of the three barriers in this manner. Since the barriers were always 0.45m apart, I used this to calibrate the other measurements. To test the repeatability of this measure, I measured the angle of one bird 25 times, removing the endpoints of the line and replacing them repeatedly on the bird. This analysis determined that my measurement error was within 2 ± 0.34 degrees. A typical source of error in this type of experiment is parallax from the lens of the video camera (Rayner and Aldridge 1985; Chatters 1996). As the distance between the barriers was 0.9 m and the camera was 1.4 m away from the bird, parallax error was negligible in the plane across the barriers. I confirmed this empirically by recording and analyzing a grid of known size under the camera in the plane in which the birds flew. Another source of error in this experiment is the height at which birds were flying through the tunnel. I have assumed that all birds were flying less than 10 cm from the top of the tunnel, as the height of the perch was 13 cm from the top of the tunnel, and that birds did not fly 10 cm from the bottom of the tunnel. Since the flight trajectory would appear shorter if it were farther from the camera, i.e. lower in the tunnel, my assumption would introduce a 12.5 % - 24 % underestimate of distance travelled if it flew within this 40 cm range in the middle of the tunnel. This error would overestimate translational velocity and acceleration but would not influence measures of angular motion. 77 Smoothing the Data: I applied a digital filtering equation to smooth time series (Winter 1990), with a cutoff frequency of 10 Hz for translational velocity and acceleration and 8 Hz for angular velocity and angular acceleration. These cutoff frequencies were chosen after a residual analysis of all possible frequencies (Winter 1990). The filter was applied in both directions to remove phase shifts that occurred due to a single pass of the filter. It was necessary to pad the data, as a reflection of the actual data, at the beginning and end due to the short length of the time series. The ten points added to the beginning and the end were necessary for the proper functioning of the filter, but did not significantly influence the values of the filtered data. However, I did not include the first, second and last smoothed data point in the analysis. Video Analysis of Flights As expected, birds did not perform the task appropriately on every flight. Frequency distributions were generally skewed towards the right (slower flights) due to these unacceptable flights. After comparing videotape records and elapsed times for each bird, I concluded that on flights longer than 5.0 s the birds were pausing and therefore not correctly performing the task, and excluded them from analysis. The frame by frame analysis of the videotape allowed for a detailed analysis of a bird's position every 1/30 s. For each frame in a bird's flight, an x and y coordinate of the center of the bird was used to calculate distance traveled in the horizontal plane and distance traveled along the bird's trajectory (path). I investigated both maximal and mean performance in this experiment. If the mean accurately represents the fastest time of a bird, then the mean accurately 78 represents the individual's peak performance. Birds with the fastest flights also had fastest mean elapsed times (r2=0.77, p<0.001, 18 birds, Fig. 22). Results: As a result of damage to videotapes, sample size in this experiment was reduced to 10 individuals: 2 adult males, 2 adult females, 4 juvenile females, and 2 juvenile males (1 juvenile male had only flights returning to the perch) were analyzed using videotape, and comparisons between age and sex classes were limited. Morphological Comparisons As expected, adult males had the shortest wings of all four age-sex classes, and juvenile females the longest. The adult female and juvenile males had intermediate and similar wing lengths (Fig. 23). Males had higher wing disc loading (WDL; Feinsinger et al. 1979) than all other age-sex classes (ANOVA, n=4, p<0.001, Fig. 23). Juvenile females, on average, were heavier than during the low velocity turning experiment, but their wings were the same. The higher mass of juvenile females resulted in this class having similar WDL to juvenile male and adult females (Fig. 23). Performance and Morphology Each bird spent 45 minutes with 2 barriers on the left side of the tunnel (140.8 ± 13.3 flights), then switched to two barriers on the right side for 45 minutes (132.5 + 15.6). There was no relationship between length of either right or left wings and the elapsed time between perch and feeder or feeder and perch. Elapsed time did not vary with trial number or (after switching sides of the barriers) the "handedness" of the task. 79 Overall elapsed time between the perch and the feeder encompassed several component actions. On each feeding flight, the bird left the perch, accelerated for a portion of the longest part of the tunnel to fly 2.91 m until it reached the first barrier. Then it maneuvered past the three barriers, making one large turn around the middle barrier, and flew 0.83 m after the last barrier, decelerating to the feeder. It hovered while feeding, turned 180 degrees away from the feeder when finished, flew 0.83 m to the first barrier, and maneuvered through the three barriers before flying 2.91 m to the perch. The elapsed time from perch to feeder and feeder to perch (not including time spent feeding) did not differ with age-sex class and there was much variation within classes (Fig. 24). Interestingly, adult males flew at the same average speed to and from the feeder, but the other three classes (adult females, juvenile males and juvenile females) were all faster towards the feeder than when returning to the perch (ANOVA n=4, p=0.01, Fig. 24). This pattern led to an investigation of the difference between negotiating a cluttered environment at both high and low velocity. I considered the flights towards the feeder performance in clutter at a high velocity and flights towards the perch as performance in clutter at a lower velocity. Assuming birds flew the same distance through the tunnel, average velocity towards the feeder was 10 % faster than returning to the perch. Excluding adult males, all other classes had average velocities 12.2 % faster on flights towards the feeder than returning to the perch. The videotape analysis allowed for a detailed investigation of flight parameters as individuals flew through the barriers, but not for the remainder of the flight. Elapsed time through the barriers, as measured by frames of the videotape (1/30 s), was correlated with the computer elapsed time for adult females (r2=0.24, p=0.03), juvenile males (r2=0.36, p=0.05) and juvenile females (r2=0.35, pO.OOl), but not for adult males. 8 0 Using the logged positions of the birds in the tunnel for flights in both directions, I calculated translational velocity at the first (Vin), and last barriers (Vout). Maximum translational velocity (VMax), average velocity (VAvg), and maximum translational acceleration (AMax) through the barriers were also calculated for each flight. I calculated angular motion through the barriers using maximum angular velocity (AVMax) and maximum angular acceleration (AAMax). I compared three morphological measures (wing length, mass and wing disc loading) to measures of both translational and angular motion. Individuals with higher WDL reached the first barrier at a significantly higher velocity (Vin, r =0.54, p=0.02), reached higher maximum velocity through the barriers (Vmax, r =0.59, p=0.01), and reached higher average velocity through the barriers (Vavg, r2=0.41, p=0..04, Fig. 25) than those with lower WDL. However, velocity leaving the barriers was not correlated with wing disc loading (Vout, Fig. 25). Towards the feeder, maximum velocity through the barriers, but no other measure of velocity, was correlated with mass (r2=0.49, p=0.03). No measure of velocity through the barriers in the direction towards the feeder was correlated with wing length. For flights back to the perch, no measures of velocity through the tunnel were correlated with any morphological measures. After accelerating to reach the first barrier at a higher velocity than on the return trip, all birds accelerated significantly between the first and last barriers (ANOVA, n=4, p=0.02, Fig. 26). This was unexpected as turning around the barriers should decrease velocity, unless birds expended extra power to counteract this. However, on flights back to the perch, classes differed in this respect. All birds left the barriers faster than they entered except for adult males (ANOVA, n=4, p=0.002, Fig. 26), who entered and left at the same velocity. On flights towards the feeder, adult males entered the barriers (Vin), at a higher 8 1 velocity than all other classes (ANOVA, n=4, p=0.05, Fig. 26). For adult males, this implies higher average acceleration from the perch, and lower average acceleration from the feeder than other classes. The turn away from the feeder towards the perch may impede adult males' ability to reach peak accelerations within a short distance. Maximum acceleration through the barriers on flights towards the feeder was significantly correlated with wing length (r2=0.41, p=0.05) and wing disc loading (r2=0.67, p=0.003, Fig. 27), but not mass. Individuals with shorter wings and higher WDL accelerated faster through the barriers than those with long wings and low WDL. On return flights to the perch, maximum acceleration through the barriers was not correlated with any of the morphological measures. Adult males had nearly significantly higher maximum accelerations than all other classes towards the feeder (ANOVA, n=4, p=0.05, Fig. 28), but there was no difference between classes on returning flights. Angular motion through the barriers was measured by the change in angle over time or angular velocity. Maximum angular velocity was significantly correlated with wing length (r2=0.44, p=0.04), and wing disc loading (r2=0.64, p=0.01, Fig. 29), but not mass. Individuals with the shortest wings and highest WDL reached higher maximum angular velocities than birds with long wings and low WDL. Adult males had an advantage in angular velocity within the barriers towards the feeder (ANOVA, n=4, p<0.001, Fig. 30), but not towards the perch (Fig. 30). On return flights towards the perch, angular velocity was not correlated with any of the morphological measures (mass, wing length or WDL). Maximum angular acceleration between the barriers in both directions was not correlated with any morphological measures. 8 2 Discussion: The ability to negotiate turns amongst a cluttered environment has been used as a measurement of maneuverability for many flying organisms (Aldridge 1986; Aldridge and Rautenbach 1987; Moller 1991; Witter et al. 1994; Metcalfe and Ure 1995). If maneuverability is defined as the ability to turn tightly and agility the ability to turn quickly, then this experiment measures both, and I argued in the introduction that during antagonistic encounters hummingbirds turn fast and hard whether or not their environments are cluttered. In this experiment, all birds maneuvered through the barriers without collisions or pauses, and I considered all birds highly maneuverable. The computer elapsed time data demonstrated that adult males traveled to and from the feeder in the same time. For adult females, juvenile males and juvenile females, the time taken to fly to the feeder was significantly less than to return to the perch. On flights towards the feeder, birds with high wing disc loading reached the first barrier at higher maximum velocities, and reached higher maximum and average velocities within the barriers than birds with low wing disc loading. The heaviest birds had the highest maximum velocity through the barriers. My results are consistent with the predictions of aerodynamic theory, that increasing mass will increase flight speed (Webb et al. 1992) but most tests of this theory with other organisms do not concur. In fact, many studies (Hughes and Rayner 1991; Witter et al. 1994; Metcalfe and Ure 1995) have demonstrated that increasing mass decreases flight speed and reduces maneuverability. For example, Witter et al. (1994) found that as the mass of a starling increased (added weights on the legs of the bird), the number of contacts with poles in an obstacle course increased, and as the mass of the starling decreased (through 83 food deprivation) the number of contacts decreased. However, mass did not affect elapsed time through the obstacles. I expected that birds would slow down as they passed through the barriers, however in both directions birds were faster leaving the barriers than entering, with the exception of adult males who entered and left at the same velocity on flights towards the perch. I also expected that birds entering the barriers on the way to the feeder would do so at a higher velocity than entering them after the feeder, as the distance from the perch to the first barrier was almost 3 times the distance from the feeder to the first barrier when returning to the perch. Despite faster elapsed times of flights towards the feeder than towards the perch, there was no difference in velocity entering the first barrier in either direction. However, in flights towards the perch, adult males left the barriers at a lower velocity than all other age-sex classes. This is the only performance measure in which adult males were slower than all other age-sex classes. Adult females, juvenile males and juvenile females may have "caught up" in elapsed time during this portion of the whole flight through the tunnel. During flights towards the feeder, birds with the shortest wings and highest WDL had the highest maximum accelerations through the barriers. Adult males had higher maximum accelerations than juvenile females on the way to the feeder, but this measure was equal for these two classes on the way back to the perch. Adult males may not be able to reach peak acceleration on flights towards the perch due to the short distance before the barriers. This may explain why there was no difference in elapsed times for adult males in both directions through the tunnel. There was no relationship between morphology and maximum acceleration on flights returning from the feeder to the perch. The same pattern exists for angular velocity. Birds with the shortest wings and highest WDL had the highest maximum angular velocity through the 84 barriers towards the feeder, but not on flights returning to the perch. There was no significant difference between age-sex classes in maximum angular velocity through the barriers towards the perch. Maximum angular acceleration was not related to morphology, age-sex class, or direction of the flight. Adult males were the top performers in angular and translational velocity through the barriers on the way to the feeder. There was no significant difference between age-sex classes in elapsed time or velocity through the tunnel towards the perch. If flights towards the feeder are considered high velocity turning and flights away from the feeder are low velocity turning, my results suggest that adult males, with their high WDL and short wings, have an advantage in high speed agility (ability to make fast turns), and neither class nor morphology influence low speed agility. These results are predicted by current aerodynamic theory for bats and birds with non-hovering flight (Norberg and Rayner 1987). Birds with low wing loading are predicted to have high maneuverability and low agility, and birds with high wing loading are predicted to have low maneuverability and high agility. In my low velocity turning experiment (Chapter 3), birds with the longest wings had an elapsed time advantage flying between feeders 0.5 m apart. Longer wings and lower WDL (e.g. juvenile females) allowed individuals to reach higher translational and angular velocities. These results counter those of this high velocity experiment. I believe the reason for these conflicting results is the pattern of usage of these two types of flight in the daily life of rufous hummingbirds. The adult males are the only age-sex class who displays for mating purposes. This display involves high speed dives towards the females that occur on a regular basis during the breeding season. There are no published attempts at measuring the speed of these courtship dives, which are much faster than any flights measured in this thesis. The ability to turn quickly at the bottom 85 of a courtship dive is necessary to avoid crashing into the female or the underbrush, suggesting that adult males wings may be sexually selected for mating display purposes. It is important to note that the dives involved in courtship are fundamentally different than the turns in these experiments. Conversely, the low velocity turning experiment demonstrated that juvenile females flew short distances between two feeders faster and turned faster at close range. If they fly faster between feeders, and therefore visit more flowers than slower classes in a foraging bout, they may gain mass faster. This speed advantage in combination with juvenile females' dull coloration and silent flight may allow them to sneak nectar from highly profitable territories defended by the conspicuous adult males. If this is the case, selection may favour longer wings for faster flights between flowers in a patch. Due to the small sample of videotapes available for analysis in this experiment, I consider the results preliminary and believe the experiment should be repeated. The analysis demonstrates a significant trend, in which adult males have a performance advantage at high velocities. The length of the tunnel should be varied systematically to determine how velocity when negotiating turns influences the speed at which individuals can maneuver. Altering the distance between the barriers, and the complexity of the barriers or clutter would also aid in determining which classes have an advantage in maneuvering through clutter. Intraspecific comparisons of performance will help in understanding how morphology influences ecology. Aerodynamic models are necessary for birds with hovering flight so that both inter- and intraspecific comparisons of hummingbird performance can be made. 86 F e e d e r E n d 0 . 8 3 m 0 . 4 5 m 0 . 4 5 m 3 . 0 7 m 4.8 m P e r c h E n d F i g u r e 20: T h e e x p e r i m e n t a l t u n n e l . T h e v i d e o c a m e r a w a s a b o v e t h e b a r r i e r s , 0.9 m f r o m t h e t u n n e l . L E D ' s a r e s i t u a t e d a b o v e t h e f e e d e r a n d b e f o r e t h e f i r s t b a r r i e r . A c o m p u t e r r e c o r d s w h e n a b i r d a r r i v e s a n d l e a v e s t h e p e r c h a n d t h e f e e d e r . 87 Figure 2 1 : P r o c e d u r e f o r r e c o r d i n g p o s i t i o n a n d o r i e n t a t i o n o f b i r d ( A d u l t f e m a l e 1 6 ) . E v e r y s e c o n d f r a m e i s s h o w n . L e f t i n s e t : e n l a r g e m e n t o f f r a m e 1 2 s h o w i n g l i n e a l o n g a x i s o f b i r d a n d i t s m i d p o i n t . R i g h t i n s e t : d i a g r a m m a t i c r e p r e s e n t a t i o n s h o w i n g p o s i t i o n a n d a n g l e . 88 5 CO « o — LL 2 1.5 -1 2 3 4 5 Mean Elapsed time (s) F i g u r e 2 2 : R e l a t i o n s h i p b e t w e e n m e a n e l a p s e d t i m e a n d f a s t e s t e l a p s e d t i m e t h r o u g h t h e t u n n e l f o r 18 ind iv idua ls . 89 B 0.08 0.07 + 0.06 CM < E o 3 0.05 Q 0.04 0.03 0.02 Age-Sex Class Age-Sex C lass F igure 23: A ) A v e r a g e w i n g l e n g t h s o f t h e f o u r a g e - s e x c l a s s e s u s e d i n t h e e x p e r i m e n t . B ) A v e r a g e W D L o f f o u r a g e - s e x c l a s s e s u s e d i n t h i s e x p e r i m e n t . M e a s u r e m e n t s w e r e t a k e n p r i o r t o t h e e x p e r i m e n t . B a r s r e p r e s e n t s t a n d a r d e r r o r . A M = a d u l t m a l e , J M = j u v e n i l e m a l e , A F = a d u l t f e m a l e , J F = j u v e n i l e f e m a l e . 90 A 4.5 4.3 2 4"1 | 3.9 •= 3.7 1 3.5 i Q. Qj 3-3 3.1 H 2.9 2.7 2.5 B A M (5) J F ( 4 ) Age-Sex Class Age- Sex Class F igure 24: M e a n e l a p s e d t i m e b e t w e e n f e e d e r s i s p l o t t e d f o r e a c h o f t h e f o u r a g e - s e x c l a s s e s . A ) T h e f l i g h t s t o w a r d s a n d r e t u r n i n g f r o m t h e f e e d e r a r e c o m b i n e d . B ) F l i g h t s a r e s e p a r a t e d . T h e d i a m o n d s r e p r e s e n t f l i g h t s f r o m t h e p e r c h t o t h e f e e d e r , a n d t h e s q u a r e s r e p r e s e n t f l i g h t s f r o m t h e f e e d e r t o t h e p e r c h . A M = a d u l t m a l e s , A F = a d u l t f e m a l e s , J M = j u v e n i l e m a l e s , J F = j u v e n i l e f e m a l e s . B a r s r e p r e s e n t s t a n d a r d e r r o r . 91 A B 0 0.02 0.04 0.06 0.08 0 0.02 0.04 0.06 0.08 WDL (g /cm A 2) WDL ( g / c m A 2 ) F igu re 25: E f f e c t o f W D L o n f o u r m e a s u r e s o f t r a n s l a t i o n a l v e l o c i t y t h r o u g h t h e b a r r i e r s t o w a r d s t h e f e e d e r . A ) v e l o c i t y a t p o i n t o f e n t r y o f b a r r i e r s B ) v e l o c i t y a t p o i n t o f e x i t C ) m a x i m u m v e l o c i t y t h r o u g h t h e b a r r i e r s a n d D ) a v e r a g e v e l o c i t y t h r o u g h t h e b a r r i e r s . 92 T o w a r d s Feeder T o w a r d s Perch Age-Sex Class Age-Sex Class Figure 26: V e l o c i t y a t p o i n t o f e n t r y ( d i a m o n d s ) o f b a r r i e r s a n d p o i n t o f e x i t ( s q u a r e s ) f o r f l i g h t s i n b o t h d i r e c t i o n s . A M = a d u l t m a l e s , A F = a d u l t f e m a l e s J M = j u v e n i l e m a l e s , J F = j u v e n i l e f e m a l e s . B a r s r e p r e s e n t s t a n d a r d e r r o r . 93 Figure 27: M a x i m u m a c c e l e r a t i o n t h r o u g h t h e b a r r i e r s e c t i o n o f t h e t u n n e l a s r e l a t e d t o w i n g l e n g t h a n d W D L . A l l f l i g h t s a r e t o w a r d s t h e f e e d e r . 94 14.1 12.1 10.1 + < (0 X (0 8.1 6.1 A M i A F J F 4.1 2.1 Age-Sex Class F igure 28 : M a x i m u m a c c e l e r a t i o n o f a l l f o u r a g e - s e x c l a s s e s t h r o u g h t h e b a r r i e r s i n b o t h d i r e c t i o n s . D i a m o n d s r e p r e s e n t f l i g h t s t o t h e f e e d e r , a n d s q u a r e s r e p r e s e n t f l i g h t s b a c k t o t h e p e r c h . B a r s r e p r e s e n t s t a n d a r d e r r o r . A M = a d u l t m a l e s , A F = a d u l t f e m a l e s , J M = j u v e n i l e m a l e s , J F = j u v e n i l e f e m a l e s . 95 B F i g u r e 2 9 : M a x i m u m a n g u l a r ve loc i t y t h r o u g h t h e ba r r ie r s e c t i o n of t h e t u n n e l a s re la ted to w i n g l eng th a n d W D L . Al l f l ights a re t o w a r d s t h e f e e d e r . 96 800 700 600 </> a> <D i . O) a> S 500 x CO > A M 400 + • A F I J M • J F 300 200 Age-Sex Class F igure 30: M a x i m u m a n g u l a r v e l o c i t y i s p l o t t e d f o r e a c h o f t h e f o u r a g e - s e x c l a s s e s . D i a m o n d s r e p r e s e n t f l i g h t s t h r o u g h t h e b a r r i e r s i n t h e d i r e c t i o n t o w a r d s t h e f e e d e r a n d s q u a r e s r e p r e s e n t f l i g h t s t h r o u g h t h e b a r r i e r s f r o m t h t h e f e e d e r t o t h e p e r c h . A M = a d u l t m a l e s , A F = a d u l t f e m a l e s , J M = j u v e n i l e m a l e s J F = j u v e n i l e f e m a l e s . B a r s r e p r e s e n t s t a n d a r d e r r o r . J M n = 1 . 97 CHAPTER 5 CONCLUSION AND G E N E R A L DISCUSSION This thesis investigated intraspecific differences in flight ability of one dimorphic species. Morphological variation among the four age-sex classes of rufous hummingbirds has led to questions about the advantage of specific morphologies within these classes. Morphological differences in wing length, mass and WDL are predicted to directly influence maneuverability and agility. The misuse of the terms maneuverability and agility is widespread in the hummingbird literature. This misuse has contributed to a proliferation of unclear and confusing interpretations of resource partitioning and territoriality. Perhaps the problem is that hummingbirds are unique flying machines compared to bats and other birds. In Chapter 3's low velocity turning experiment birds with the longest wings and lowest WDL (e.g. juvenile females) had faster elapsed times than birds with short wings and high WDL (e.g. adult males). Longer wings allowed hummingbirds to reach higher translational and angular velocities than short winged individuals. The design of the experiment in Chapter 3 has never been used for other types of animals, and this type of short flight involving a turn may be difficult to categorize as either maneuverability or agility. Juvenile males with intermediate wings outcompeted all other classes in angular motion. The results of this experiment clearly contradict the current literature on hummingbird performance. The low velocity turning experiment demonstrated that juvenile females flew short distances between two feeders faster and turned faster at close range. This low speed advantage in combination with juvenile females' dull coloration and silent flight may favour longer wings for faster flights between flowers in a 98 patch. Many studies (Pennycuick 1968; Weis-Fogh 1972; Epting and Casey 1973; Epting 1980; Feinsinger and Chaplin 1975) have suggested that long wings and low WDL allow reduced power output during flight, however this thesis is the first to suggest that juveniles' long wings are beneficial in terms of turning performance. In Chapter 4, I compared elapsed time of four age-sex classes of rufous hummingbirds to fly through a tunnel fitted with three barriers to investigate maneuverability at high speed. Adult males were the top performers in angular and translational velocity through the barriers on the way to the feeder. There were no significant differences between age-sex classes in elapsed time or velocity through the tunnel towards the perch. If flights towards the feeder are considered high velocity turning and the flights away from the feeder are low velocity turning, my results suggest that adult males, with their high WDL and short wings, have an advantage in high speed agility (ability to make fast turns), and neither class nor morphology influence low speed agility. These results are predicted by current aerodynamic theory for bats and birds with non-hovering flight (Norberg and Rayner 1987). I predict that the optimal foraging habitat for each of the four age-sex classes of rufous hummingbirds are not alike. These experiments demonstrated that different age-sex classes have performance advantages for distinct flight tasks. Choice experiments involving flowers or feeders arranged at various distances may dissect these differences. I predict that if hummingbirds select habitats that allow them to minimize foraging costs, each class would choose a foraging microhabitat best suited to their wing morphology. Rigorous testing in the laboratory of how variation in morphology influences performance between individuals and classes is necessary to clearly answer these ecomorphological questions. Field and laboratory studies of high speed translational and angular 9 9 motion will aid in understanding how morphology of the birds influences their ecology. The functional relationships between morphology and ecology are complex and more studies of how morphology can influence the performance capability of individuals and classes may help in understanding dominance, territoriality, and foraging strategies more completely. A better understanding of ecomorphological relationships of hummingbirds both within and between species will lead to a greater understanding of maneuverability and agility in hovering organisms. 1 0 0 Literature Cited: Aldridge, H. 1986. Manoeuverability and ecological segregation in the little brown (Myotis lucifugus) and Yuma (M. yumanensis) bats (Chiroptera: Vespertilionidae). Canadian Journal of Zoology. 64: 1878-1882. Aldridge, H. D. J. N. and I. L. Rautenbach. 1987. Morphology, echolocation and resource partitioning in insectivourous bats. Journal of Animal Ecology. 56: 763-778. Andersson, M and R. A. Norberg. 1981. Evolution of sexual size dimorphism and role partitioning among predatory birds, with a size scaling of flight performance. Biological Journal of Linnean Society. 15:105-130. Brown, G. S. 1992. Factors affecting spatial association learning by hummingbirds. Ph.D. Thesis. University of British Columbia. Brown, G. S. and C. L. Gass. 1993. Spatial association learning by hummingbirds. Animal Behaviour. 46: 487-497. Brown, J. H. and M. A. Bowers. 1985. Community organization in hummingbirds: relationships between morphology and ecology. Auk. 102: 251-269. Carpenter, F. L., Paton, D. C. and M. A. Hixon. 1983. Weight gain and adjustment of feeding territory size in migrant hummingbirds. Proceedings of the National Academy of Science. 80: 7259-7263. Carpenter, F. L., M. A. Hixon, R. W. Russell, D. C. Paton, E. J. Temeles. 1993. Interference asymmetries among age-sex classes of rufous hummingbirds during migratory stopovers. Behavioural Ecology and Sociobiology. 33: 297-304. Carpenter, F. L. , M. A. Hixon, E. J. Temeles, R. W. Russell, D. C. Paton. 1993. Exploitative compensation by subordinate age-sex classes of migrant rufous hummingbirds. Behavioural Ecology and Sociobiology. 33: 305-312. Chandler, C. R. and R. S. Mulvihill. 1992. Effects of age, sex and fat level on wing loading in dark-eyed juncos. Auk. 109: 235-241. Chatters, L. M. 1996. The dynamics of hummingbird flight. M.Sc. Thesis. University of British Columbia. 101 Collins, B. G. and D. C. Paton. 1989. Consequences of differences in body mass, wing length and leg morphology for nectar-feeding birds. Australian Journal of Ecology. 14: 269-289. Craig, J. L. 1984. Wing noises, wing slots, and aggression in New Zealand honeyeaters (Aves: Meliphagidae). New Zealand Journal of Zoology. 11: 195-200. Cuthill, I. and T. Guilford. 1990. Perceived risk and obstacle avoidance in flying birds. Animal Behaviour. 40: 188-190. Darwin, C. 1871. The descent of man and selection in relation to sex. London. Desrochers, A. 1989. Sex, dominance and microhabitat use in wintering black-capped chickadees: a field experiment. Ecology. 70 (3): 613-645. Ellington, C. P. 1975. The aerodynamics of normal hovering flight: three approaches. In: Comparative Physiology- water, ions and fluid mechanics. Schmidt-Neilsen. K., Boris, L. and Maddrell, S. H. P., eds. Cambridge University Press: Cambridge, pp. 327-345. Epting, R. J. and T. M. Casey. 1973. Power output and wing disc loading in hovering hummingbirds. American Naturalist 107: 761-765. Epting, R. J. 1980. Functional dependence of the power for hovering on wing disc loading in hummingbirds. Physiological Zoology. 53: 347-357. Ewald, P. W. and S. Rowher. 1980. Age, coloration and dominance in nonbreeding hummingbirds: a test of the asymmetry hypothesis. Behavioural Ecology and Sociobiology. 7: 273-279. Feinsinger, P. and S. B. Chaplin. 1975. On the relationship between wing disc loading and foraging strategy in hummingbirds. American Naturalist. 109: 217-224. Feinsinger, P. and J. T. Colwell. 1978. Community organization among neotropical nectar-feeding birds. American Zoologist. 18: 779-795. Feinsinger, P., R. K. Colwell, J. Terborgh, and S. B. Chaplin. 1979. Elevation and the morphology, flight energetics, and foraging ecology of tropical hummingbirds. American Naturalist. 113: 481-496. 1 0 2 Gass, C. L., G. Angeher, and J. Centa. 1976. Regulation of food supply by feeding territoriality in the rufous hummingbird. Canadian Journal of Zoology. 5 4 ( 1 2 ) : 2046-2054. Gass, C. L. 1978. Experimental studies of foraging in complex laboratory environments: interaction between food availability and feeding behaviour in hummingbirds. American Zoologist. 18 : 729-738. Gass, C. L. 1985. Behavioural foundations of adaptation. In: Perspectives in Ethology, Vol. 6. Bateson, P. P. G. and Klopfer, P. H. (eds.) Plenum Publishing Corp.: New York. pp. 63-107. Gass, C. L. and R. D. Montgomerie. 1981. Hummingbird foraging behaviour: decision-making and energy regulation. In: Foraging Behavior: Ecological, Ethological, and Psychological Approaches. Kamil, A. C. and Sargent, T. D., eds. Garland Publishing: New York. pp. 159-194. Gass, C. L. and G. D. Sutherland. 1985. Specialization by territorial hummingbirds on experimentally enriched patches of flowers: energetic profitability and learning. Canadian Journal of Zoology. 6 3 : 2125-2133. Gass, C. L. and W. M. Roberts. 1992. The problem of temporal scale in optimization: three contrasting views of hummingbird visits to flowers. American Naturalist. 1 4 0 : 829-853. Gill, F. B. 1985. Hummingbird flight speeds. Auk 1 0 2 : 97-101. Greenewalt, C. H. 1975. The flight of birds. Transactions of the American Philosophical Society. 6 5 : 1-67. Hainsworth, F. R. and L. L. Wolf. 1972. Power of hovering flight in relation to body size in hummingbirds. American Naturalist 1 0 6 : 589-596. Hedenstrom, A. 1992. Flight performance in relation to fuel load in birds. Journal of Theoretical Biology. 1 5 8 : 535-537. Hedrick, A. V. and E. J. Temeles. 1989. The evolution of sexual dimorphism in animals: hypotheses and tests. Trends in Ecology and Evolution. 4 : 136-138. Hixon, M. A., Carpenter, F. L. and D. C. Paton. 1983. Territory area, flower density, and time budgeting in hummingbirds: an experimental and theoretical analysis. American Naturalist. 1 2 2 : 366-391. 103 Hutchinson, G. E. 1959. Homage to Santa Rosalia, or Why are there so many kinds of animals? American Naturalist. 93: 145-159. Hughes, P. M and J. M. V. Rayner. 1991. Addition of artificial loads to long-eared bats Plecotus auritus: handicapping flight performance. Journal of Experimental Biology. 161: 285-298. Johnsgard, P. A. 1983. The hummingbirds of North America. Smithsonian Institution Press. Washington D. C. Kodric-Brown, A. and J. H. Brown. 1978. Influence of economics, interspecific competition, and sexual dimorphism on territoriality of migrant rufous hummingbirds. Ecology. 39(2): 285-296. Lack, D. 1971. Ecological isolation in birds. Harvard University Press, Cambridge, MA. MacArthur, R. H. 1958. Population ecology of some warblers of north-eastern coniferous forests. Ecology. 39: 599-619. Metcalfe, N. B and S. E. Ure. 1995. Diurnal variation in flight performance and hence potential predation risk in small birds. Proceedings of the Royal Society of London B. 261:395-400. Miller, S. J. and D. W. Inouye. 1983. Roles of the wing whistle in the territorial behaviour of male broad-tailed hummingbirds (Selasphorus platycercus). Animal Behaviour. 31: 689-700. Moller, A. P. 1991. Sexual ornament size and the cost of fluctuating asymmetry. Proceedings of the Royal Society of London B. 243: 59-62. Morton, E. S. 1990. Habitat segregation by sex in the hooded warbler: experiments on proximate causation and discussion of its evolution. American Naturalist. 135: 319-333. Norberg, U. M. 1990. Vertebrate flight: Mechanics, physiology, morphology, ecology and evolution. Springer-Veriag. Berlin. Norberg, U. M. 1995. How a long tail and changes in mass and wing shape affect the cost for flight in animals. Functional Ecology. 9: 48-54. 104 Norberg, U. M. and J. M. V. Rayner. 1987. Ecological morphology and flight in bats (Mammalia; Chiroptera): wing adaptations, flight performance, foraging strategy and echolocation. Philosophical Transactions of the Royal Society of London B. 316: 335-427. Pennycuick, C. J. 1968. Power requirements for horizontal flight in the pigeon Columba livia . Journal of Experimental Biology. 49: 527-555. Peters, R. H. 1991. A critique for ecology. Cambridge University Press. Cambridge. Peters, W. D. and T. C. Grubb Jr. 1983. An experimental analysis of sex-specific foraging in the downy woodpecker, Picoides Pubesceris. Ecology. 64(6): 1437-1443. Rayner, J. M. V. 1988. Form and function in avian flight. Current Ornithology. 5:1-66. Rayner, J. M. V. and H. D. J. N. Aldridge. 1985. Three-dimensional reconstruction of animal flight paths and the turning flight of microchiropteran bats. Journal of Experimental Biology. 118: 247-265. Ricklefs, R. E. and D. B. Miles. 1994. Ecological and evolutionary inferences from morphology: An ecological perspective. In: Ecological Morphology: Integrative organismal biology. Wainwright P. C. and S. M. Reilly (eds.) University of Chicago Press, Chicago and London. Ridgway, R. 1911. The birds of North and Middle America. Part V. U.S. National Museums Bulletin. 5 0 : 1-859. Snow, D. W. and B. K. Snow. 1980. Relationships between hummingbirds and flowers in the Andes of Colombia. Bulletin of the British Museum of Natural History (Zool.). 38:105-139. Schoener, T. W. 1974. Resource partitioning in ecological communities. Science. 19: 189-213. Selander, R. K. 1966. Sexual dimorphism and differential niche utilization in birds. Con dor. 68:113-151. Selander, R. K. 1972. Sexual selection and dimorphism in birds. In: B. G. Campbell (ed) Sexual selection and the descent of man, 1871-1971. Chicago. 1 0 5 Shine, R. 1989. Ecological causes for the evolution of sexual dimorphism: a review of the evidence. Quarterly Review of Biology. 64(4): 419-441. Slatkin, M. 1984. Ecological causes of sexual dimorphism. Evolution. 38(3): 622-630. Stiles, G. F. 1972. Age and Sex determination in rufous and alien hummingbirds. Condor. 74: 25-32. Stiles, G. F. 1995. Behavioural, ecological and morphological correlates of foraging for arthropods by the hummingbirds of a tropical wet forest. Condor. 97: 853-878. Sutherland, G. D., Gass, C. L. , Thompson, P. A. and K. P. Lertzman. 1982. Feeding territoriality in migrant rufous hummingbirds: defense of a yellow-bellied sapsucker (Sphyrapicus varius) feeding sites. Canadian Journal of Zoology. 63: 2249-2253. Tamm, S. and Gass, C. L. 1985. Energy intake rates and nectar concentration preference by hummingbirds. Oecologia. 70: 20-23. Temeles, E. J. and W. M. Roberts. 1993. Effect of sexual dimorphism in bill length on foraging behaviour: an experimental analysis of hummingbirds. Oecologia. 94: 87-94. Thompson, J. 1994. What spatially structured information do hummingbirds use in spatial association tasks? M.Sc. Thesis. University of British Columbia. Tooze, Z. J. and C. L. Gass. 1985. Responses of rufous hummingbirds to midday fasts. Canadian Journal of Zoology. 63: 2249-2253. Wainwright, P. C. 1988. Morphology and ecology: Functional basis of feeding constraints in Caribbean lavrid fishes. Ecology. 69: 635-645. Wainwright, P. C. 1991. Ecomorphology: Experimental functional anatomy for ecological problems. American Zoologist. 31: 680-693. Webb, P. I, J. R. Speakman and P. A. Racey. 1992. Inter-and intra-individual variation in wing loading and body mass in female pipestrelle bats: theoretical implications for flight performance. Journal of Zoology London. 228: 669-673. 106 Weis-Fogh, T. 1972. Energetics of hovering flight in hummingbirds and Drosophila. Journal of Experimental Biology. 56: 79-104. Weis-Fogh, T. 1973. Quick estimates of flight fitness in hovering animals, including novel mechanisms for lift production. Journal of Experimental Biology. 59: 169-230. Wells, D. J. 1993. Ecological correlates of hovering flight in hummingbirds. Journal of Experimental Biology. 1 7 8 : 59-70. Winter, D. A. 1990. Biomechanics and Motor Control of Human Movement, pp. 36-45. Witter, M. S. I. C. Cuthill, R. H. C. Bonser 1994. Experimental investigations of mass-dependent predation risk in the European starling, Sturnus vulgaris. Animal Behaviour. 4 8 : 201-222. 107 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0087984/manifest

Comment

Related Items