Condor 85:220-242 0 The Cooper Ornithological Society 1983

STRUCTURE OF DISPLAY FLIGHTS IN THE LEAST

EDWARD H. MILLER

ABSTRACT. -Display Flights (DFs) of the Least Sandpiper (Calidrisminutilla) are described. DFs were prolonged flights (averaging 1.2 min in Manitoba and 3.4 min in Nova Scotia), about lo-30 m in altitude, by unmated males over their future nesting areas. Displaying males alternated brief glides and bouts of rapid shallow wingbeats (Flutters) throughout DFs; the two phasesaveraged 0.84 and 0.72 s, respectively. Flutters averaged 26 (Yukon) to 37 per min (Nova Scotia). DFs ended in several kinds of descents,from slow parachuting to rapid stooping. Three main call types were associatedwith DFs. The main Display Flight Vo- calization (DFV) was simple and stereotyped, and was emitted rhythmically throughout DFs. DFVs of 22 males averaged 390 ms long with intervals between them of 106 ms. DFVs thus occurred at a rate of about two per second, so were not synchronized with Flutters. Some features of DFs, suchas their monotonous, repeated calls, are sharedwith many other open-country specieswith aerial displays(e.g., plovers, pipits). Never- theless,DFVs are probably plesiomorphouswithin the Calidridinae. The adaptive significanceand ancestral/derived status of these and other features of DFs are discussed.

The Calidridinae (Scolopacidae)consists of 24 but adults and chicks generally feed in unde- closely related speciesof sandpipersthat nest fended areas distant from the nestinggrounds. in open habitat in the arctic and subarctic (Pi- Males assumean increasingrole in incubation telka et al. 1974). Their ecological similarities as it proceeds, and are mainly responsiblefor and close affinities make them ideal for study raising the brood (Miller 1977, 1979a). The of adaptive radiation and homologies of dis- aerial displaysof unpaired males in this plays in a non-passerine group. A good can- are very restricted in occurrenceand function: didate for such study is aerial display, which a male may engagein display flights for only is well developed in open-country speciesin a few days each spring, and these apparently general,and occursin numerous charadriiform serve solely to attract a female to a suitable species, including Calidridinae (Armstrong nesting area, while simultaneously repelling 1963, Pitelka et al. 1974). In this paper I pro- competing unpaired males. vide a detailed description of aerial displays in a representative calidridine, the Least Sand- STUDY AREAS AND METHODS piper ( minutilla). This will establish I studied Least at three locations: a baseline for future comparative studies, Sable Island, Nova Scotia (43”57’N, 59”55’W) thereby permitting assessmentof the relation- from 17 May to 23 July 1975, and 14 May to ship of the display’s structural features to the 3 August 1976; near Churchill, Manitoba different species’ spacingand mating systems, (58”24’N, 94”24’W) from 24 May to 6 July and allowing the species’ affinities to be con- 1978; and in the Blackstone River Valley, sidered. Ogilvie Mountains, Yukon Territory (64”39’N, C. minutilla is a subarctic species whose 138”27’W) from 30 May to 2 June 1979. Many nesting‘range extend$ from the Alaska Pen- were individually color-banded in the insula to Newfoundland, with some birds Nova Scotia study (see Miller 1977, 1979a). breeding as far south as Massachusettsand Display flights were timed with a stopwatch Nova Scotia(Anderson 1980, Johnsgard198 1). and were filmed with a super g-mm movie Males of this monogamous speciesoccupy the camera at 24 frames per second.Data on wind nesting grounds each spring in advance of fe- velocity were obtained from Environment males and establishephemeral territories over Canada, Atmospheric Environment. which they give Display Flights (DF; see Ap- Vocalizations on Sable Island were recorded pendix for a list of abbreviations). Males stop at 19 cm/s using Scotch tape 176 and a Uher displaying when they acquire a mate, although 4200 Report Stereo IC tape recorder, with a they nest in the area over which they displayed. Uher M5 17 dynamic microphone mounted in Both sexes incubate and care for the brood, a Dan Gibson parabolic reflector. All other

12201 LEAST SANDPIPER DISPLAY FLIGHTS 221 recordings were made on Scotch tape 208 matched to a Nagra IS tape recorder, and Senn- heiser MKH8 16 “shotgun” microphone cov- ered with a wind sock. Sound spectrogramswere prepared on a Kay Elemetrics Sona-Graph 7029A. Oscillograms were prepared on a Honeywell Visicorder sys- tem 1858. Time scalesare based on those dis- played by the machine. Power spectra(Fig. 10) were traced from plots obtained from a Prince- ton Applied Research Model 4513 FFT, and are based on 8 to 10 consecutive calls from each of the three males shown. RESULTS FIGURE 1. Frequency histogram of the durations of Display Flights (DFs) by male Least Sandpipers. CONTEXTS OF OCCURRENCE OF DISPLAY FLIGHTS; DURATIONS; GENERAL DESCRIPTION Display Flights (DFs) were performed only by observations in Nova Scotia and Labrador, males, especially unpaired males, and were seemsexaggerated. Birds in DFs appeared dis- longestand most frequent in the morning (Mil- tinctive because they hung in the air while ler 1977, 1979a, b). They began without evi- Fluttering the wingsand holding the wingsmo- dent social stimulus, or in responseto DFs by tionless,in alternating rhythm. I call this flying other males, often followed aerial chases of mode Punctuated Fluttering (PF). PF began females or intruding males, and occurred after near the end of ascent, and continued through fights, or during pausesin lengthy fights. the DF virtually without pause. Birds that be- DFs in Nova Scotia averaged 3.36 min in gan DFs spontaneously usually started emit- duration (median = 2.63 min), and had a fre- ting monotonous, rhythmically repeated Dis- quency distribution that was strongly right- play Flight Vocalizations(DFVs) shortly after skewed (ts = 19.1, df = GO,P < 0.001; Fig. 1). takeoff, but sometimes leapt into the air with One DF lasted 27 min and another only about Song or Chattering before starting DFVs, in 5 s. The few very brief DFs were atypical, be- the samecontext. DFVs were uttered in almost cause the displaying males never reached full unbroken rhythm throughout a DF, though altitude and hovered there, but just rose 5 or Song and Chattering were usual when descent 10 m then descended. DFs in Manitoba and was imminent. DFVs and PF were not in phase the Yukon were briefer, with median durations with one another (contra Townsend 1927:202). of 1.20 min and 2.15 min, respectively.Dwight (1895) reported that DFs of this species on DETAILED DESCRIPTIONS OF Sable Island lasted about 10 min, and Saun- MOTOR PATTERNS ders (1902:28) noted that the monotonous Qualitative description.A male’s ascentseemed calling in DFs there lasted “perhaps three or to be simple fast rising flight, though he was four minutes.” A similar estimate of DF du- conspicuous because of the rapidly repeated ration was given by Moore (19 12) for a exposure of the bright white undersidesof his disturbed at its nest on the Magdalen Islands, wings as he rose. This “twinkling” appearance Quebec. was particularly conspicuousin early morning A bird usually started a Display Flight by or evening, when the sun was low, and by itself flying up from the ground at a steep angle, occasionally evoked Song, Chattering, or Dis- reducing the speedand angle of ascentas max- play Flights from other males. A male’s flight imal altitude was approached, then levelling pattern changed near the end of ascent, when off when about 20 m high. He remained at he began Punctuated Fluttering. Males some- about that height throughout the DF while times continued to rise in altitude slightly after hovering over a small area (“treading”), then PF began, but typically stopped rising when letting the wind carry him over another area, about 10 to 30 m high, then engaged in PF and so on; a single DF crisscrossedan area up throughout most of the display. The onset of to 200 m acrossin this manner, though most each Flutter beganwith a very slight quivering covered less. In the Yukon, where the species of the wing tips, as determined by movie anal- nests at lower density, DFs were higher (per- ysis (Fig. 2a, frame 3). This was less obvious haps 30-40 m) and wider ranging;Townsend ’s in birds viewed from below, but in them the (1927:202) observation that they may attain beginningof each Flutter was signalledby slight heights of “50 or more yards,” based on his flexing of the wings, then backward movement 222 EDWARD H. MILLER

FIGURE 2. Postures of male Least Sandpipers during Punctuated Fluttering (PF) in Display Flights; numerals represent frame numbers on the movie film from which the sketcheswere made. a-PF as viewed from behind and slightly below. b, c-PF by two different males as viewed from below.

of the wings as part of the first Fluttering ac- c). Wings moved between positions of exten- tions (Fig. 2b, frames 1 to 3). sion and flexion, and from front to back The arc of the beating wings was predomi- throughout PF (Fig. 2b, frames 5 to 9; Fig. 2c, nantly below the horizontal early in a Flutter frames 3 and 5). At the end of a Flutter the (Fig. 2a, frame 5) then they beat rapidly in an wings were held flexed briefly, then they were arc from just below to about 40” above the extended and held motionless for the duration horizontal (frame 9). This was followed by the of the period between Flutters (the Inter-Flut- penultimate phase in which the arc was max- ter Interval or IFI). During Fluttering and IFIs imal, from nearly 45” above to 45-50” below many slight adjustments of one or both wings the horizontal (frame 12). Wingbeats showed could occur, especially in high winds, so this decreasing amplitude below the horizontal description should be considered normative thereafter (frames 20 and 2 l), and came to rest only. (frame 1). It was hard to identify correspond- Males in Display Flights held the head ing phasesin DFs seen from below (Fig. 2b, slightly erect (Fig. 3a, b), distended the throat LEAST SANDPIPER DISPLAY FLIGHTS 223

5 e a 13 &-.

26 /

diik FIGURE 3. Posturesof male Least Sandpipersin Display Flights (DFs). a, b-Male in DF as viewed obliquely from below and slightly behind (a), and from the side (b). c-Three sketchesof Parachutingduring descentfrom DFs. d- Early phasesof descentfrom a DF, the third sketch from the left suggestsa Butterfly posture. e-Rapid descent from a DF; numerals representframe numbers on the movie film from which the sketcheswere made. The film was exposed at 24 frames per second, so the sequenceshown lastedjust over a second.

(Fig. 3b), and opened and closed the bill spontaneously,but more often occurred in re- rhythmically as they emitted DFVs. The tail sponse to the sound or sight of another dis- was used much in aerodynamic adjustment, playing male. When neighboring males were and was sometimes fanned completely or only simultaneously in DF, they sometimes en- on one side, briefly or for lengthy periods. Seen gaged in unstereotyped brief chasesback and from below, the fanned tail was strikingly white forth at high altitude, during which Chattering, with a dark central stripe (due to the dark cen- Song and the Butterfly Song, or BS display tral rectricesand the legs).The pale “window” occurred (named after a display of the Baird’s on the wings was similarly conspicuouswhen Sandpiper, Calidris bairdii: see Fig. 6 of Drury backlit and seen from below, but was less so 196 1; seealso his Fig. 5 for “song to intruder” when viewed from above (I observed some by White-rumped Sandpipers).In BS the wings DFs from atop a 20-m tower and from the top were held stiffly and slightly flexed about 45” of some tall dunes). above the horizontal, as the displaying male Punctuated Fluttering continued in a DF un- sangcontinuously while gliding at a gentle an- til it was nearly over, but included short pe- gle downward (Fig. 3d, last two sketches).The riods when the male started to sing, fanned display was most pronounced when two neigh- and depressedthe tail slightly while the body boring males glided downward slowly in mu- assumed a more vertical position, with the tual BS along the common boundary between wings in a Parachuting posture (named after their territories (more accurately, this was gen- a similar display of male White-rumped Sand- erally an area of some overlap between the pipers, Calidris fuscicollis; see Fig. 5 of Drury areas over which they gave DFs), until they 196 1). He slowly descendedthus for one or a were near the ground. They remained side by few meters until the Song ended (Fig. 3~). This side throughout such formal descent. High al- was generally followed by a brief period of titude BS and mutual, descendingBS between silence, then DFVs, and PF resumed and al- tenured males apparently functioned to affirm titude was regained.Such periods of embedded boundaries between their adjacent DF areas. Parachuting Song (PS) sometimes occurred Males also used BS when they were close to 224 EDWARD H. MILLER females whom they were chasing; they then descentswas generally silent. The male pulled glided beneath and often slightly ahead of the out of his stoop at low altitude, decelerated female while singing repeatedly. This dis- just above ground level then landed gently. He played the upper wing surface and the pale sometimes showed a suggestionof Butterfly or wing stripes to the female. It was my impres- Parachuting very briefly just before touching sion that establishedmales (i.e., thosethat held ,down, or glided low over the ground for some DF areas) attempted to give BS display be- distance. Landing was sometimes accom- neath intruding males while chasingthem out, panied by a stereotyped posture in which one at least sometimes (see next paragraph). I did or both wings were held extended above the not discern this in mutual BSs between neigh- back, and this was usual when a male landed boring establishedmales, but it sometimes oc- beside his mate. Ground courtship could fol- curred during breaks in fights involving an es- low. However, most males that gave DFs were tablished male and an untenured male who not mated, and their landings were performed was trying to establish himself. without flourish. A male sometimes fluffed or Males that encroached upon an established started feeding, or ran into concealing vege- male’s Display Flight area were generally tation immediately after landing. chasedout in fast flight with Song, Chattering, Males remained approximately stationary or other calls, and such chases could take a over one or several small areas during DFs. defending male well out of his DF area. When PF broke down when males moved between the chasingmale was close to the chasedmale, areas;they sometimes increasedthe lengths of Butterfly Songcould occur below him, as men- their Flutters or simply glided, assistedby the tioned. Males also chasedfemales well beyond wind. Thus DFs were characterizedby PF while the limits of their DF areas. Such chasesoften treading over one small area, rapid irregular attracted several other males and could take movement with a breakdown in PF until over place at very high altitudes, when territorial another area, a resumption of PF and treading limits seemedto be ignored. Chasesof females there, and so on. Many minor adjustmentsoc- by unmated males occurred regardless of curred, of course. Males commonly responded whether they were paired and being attended to an intruding person, dog, or vehicle by shift- by their mates. ing the location of the DF so that they were The end of a DF typically started with a above the intruder, often at lower altitude than slight loss of altitude, cessationof Fluttering, normal. and a change from Display Flight Vocaliza- tions to silence, or an acceleration in the de- QUANTITATIVE DESCRIPTION; EFFECT livery of DFVs (which also became briefer), OF WIND VELOCITY; INDIVIDUALITY then Song. Descent took various forms. It Punctuated Fluttering by a male that was fairly could start with a slow, stepwise loss in alti- stationary in the air seemed very rhythmic. tude, coincident with changesin flight pattern Combined data on durations of Flutters and and callsjust noted. There was a gradual tran- Inter-Flutter Intervals are summarized in Fig- sition to a Parachuting posture, a transition ure 4. Means of Flutter and IF1 durations that could include Butterfly-like postures(Fig. summed to about 1.6 s. Distributions of both 3d), with or without Song, then the male par- were strongly right-skewed (for Flutters, t, = achuted slowly to the ground singing lustily 74.2, df = GO,and for IFIs, t, = 56.9, df = co; (Fig. 3~). Males could stall in the Parachuting for each, P < 0.001). Frequency histograms posture for several seconds,and rock the body based upon all raw data (equivalent to using back and forth while maneuvering the fanned weighted mean figures for different males; tail and wings, seeminglyto make their descent shaded histogram), and upon the unweighted as slow as possible. Most often, however, a means for males are distinguishedin Figure 4. male followed his incipient lossofaltitude with These differed little, although the unweighted a steep, very rapid stoop (Fig. 3e, frame 15). means provided slightly greater estimates of Most rapid descentsbegan with assumption of the mean and median of Flutter duration. The a posture with the wings partly flexed and held descriptive statisticsrefer to the raw data. above the back, suggestiveof a harrier (Fig. Different samples of Flutter and IF1 dura- 3e). From there the wings were flicked rapidly tions varied considerably; some extreme sam- in some manner (it was unclear how, even in ples are summarized in the lower half of Figure movies), and the bird descendedswiftly. Wing 4. Such differences may have arisen from dif- movement during such descentswas difficult ferencesamong males, or from effects of wind to detect by direct observation, but was ap- velocity, or both. I performed median testson parent in movies. Song overlapped the initial the lumped data for 12 males whose DFs were drop in altitude and the beginning of rapid sampled in high and low wind (low wind = 9 descents,but the terminal (fastest)part of such to 23 km per h; high wind = 25 to 37 km per LEAST SANDPIPER DISPLAY FLIGHTS 225

N=1587 IFIs’

Y=.84 set Per cent of N median =.65 see

:+ .I ..a ,.“ ’

Cumulative 60 - N per cent - 98 of N . . 76

I I I 1 2 3 4 5 6 7 1 2 3 4 5 6 7 8

Flutter duration tsec) IFI duration (set)

FIGURE 4. Frequency histograms(upper) and cumulative plots (lower) of the durations of Flutters (a, c) and Inter- Flutter Intervals (IFIs; b, d) in Display Flights by male Least Sandpipers.Frequency histogramsfor weighted data are stippled to distinguishthem from those for raw data. Chosen extreme distributions and some descriptive statisticsare shown in c and d. The central line in each of the cumulative plots representsall data summed. h). There were 873 estimates of Flutter dura- marized in Figure 5. The top record is of a tions in low wind and 657 in high wind (by male filmed in DF in a high wind. He showed median test, G = 0.49, df = 1, P < 0.5). There remarkable constancy in durations of Flutter- were 899 estimates of IF1 durations in low ing and IF1 phases over the period depicted wind and 699 in high wind (by median test, (about 140 each of Flutters and IFIs). Samples G = 0.34, df = 1, P < 0.5). The combined data of records for three other males are shown in therefore suggestedno systematic relationship the lower parts of Figure 5. IFIs of male 75- of durations of Flutters and IFIs to wind ve- 07 varied markedly regardlessof wind veloc- locity. If males were individually distinctive ity, but the most striking irregularities in Flut- in the lengths of their Flutter and IFI phases, ter durations occurred at the lower wind ve- however, lumping their records could obscure locity (sample (ii) at 10.6 km per h). systematic responses of particular males to Nevertheless, sustained periods of rhythmic changesin wind velocity. Analyses presented Fluttering occurred in both samples:compare elsewhere suggesta weak trend toward brief sample (i) at 10.6 km per h with the sample Flutters in high winds, for individual males; for 19.4 km per h. In general, durations of IF1 durations showed no systematic relation- Fluttering by male 75-07 varied lessand were ship to wind velocity (Miller 1977; see Fig. 5). briefer than were IF1 durations. The opposite My subjective impression was that males in was true of male 76-18. The overall patterns DFs in high wind were frequently buffeted by of IF1 durations for this male differed little gusts,and had to adjust their posture and di- between samples under different wind veloc- rection by flexing or extending one or both ities. He had highly irregular Fluttering dura- wings, fanning and rotating the tail, and other tions at the lower wind velocity (26.4 km per movements. h) but IF1 durations were much greater than Some unbroken PF sequencesby different those at 37.0 km per h only once. Fluttering males under various wind conditions are sum- durations of male 76-X were intermediate in 226 EDWARD H. MILLER

$76-Y at 31.7 km-hr- ’

, I 3- ., 1. ,, 7 :::I ,:I‘ 676-18 :; I ’ :: 6 75-07 2-i I\ :) I, ,j! :; 1 26.4 km-hr- ’ : I 10.6 km- hr- ’

19.4 5- km-hr- ’

4- Flutter or IFI (-1 (_.___/ duration tsec)

Or 76-18

6.5 ;- < :: :,, 676-X 8.8 km- hr- ’ ;: ,1 24.6 km-hr- ’ 3- ,:::s:,: 3- j j! j /: : : 2- 2- , / : !

M 10 flutters or IFIs’

Sequence __t FIGURE 5. Plots of the durations of Flutters and Inter-Flutter Intervals (IFIs) over time, in chosen sequencesof PunctuatedFluttering in Display Flights of male Least Sandpipers.The records are for the males referred to in Figure 5, plus one other. The wind velocity for each record is indicated.

certain respects.His Fluttering durations were The mean Flutter rate based on direct ob- only slightly longer in low than in high wind servations was 37.2 per min (Fig. 6) which is (this can be appreciated easily by scanning close to the figure of 38.5 obtained from data across both records at about the level of 0.5 on mean durations of Flutters and IFIs from s). His IF1 durations were longer and more movie analyses (using data shown in Fig. 4, variable at 8.8 than at 26.4 km per h and were mean durations of Flutters plus IFIs = 1.56 s; only slightly longer than Fluttering durations. 6011.56 = 38.5). Least Sandpipers near LEAST SANDPIPER DISPLAY FLIGHTS 221

N=122 samples for 15 c@

v= 37.2 y.84

3- 3 .5- 113. 3f 3 4f 56 FIGURE 7. Sound spectrograms of Display Flight Flutters-min- ’ Vocalizations (DFVs) of male Least Sandpipers. A se- FIGURE 6. Frequencyhistogram of Flutter rate in Dis- quenceofeight DFVs from a long seriesis shownin narrow play Flights of male Least Sandpipers, on summed raw (a) and wide (b) band representations.The first four of data. these are also shown in wide band representation on a different time scale (c). The time marker in the bottom right corner of each panel represents 500 ms. The fre- quency scale is in kHz. Analyzing filter bandwidths, 45 Hz (narrow) and 300 Hz (wide). Churchill and in the Yukon had a much lower rate of Fluttering, around 32.0 and 26.3 per min, respectively. Data on rates of Fluttering by different males them to occur at about the rate of 130 per min revealed individual differencesin Flutter rates, on Sable Island. The fundamental frequency and these remained fairly constant acrossthe of DFVs was between 1 and 2 kHz, and was range of wind velocities sampled(Miller 1977). always much weaker than its first harmonic I often guessedcorrectly the identity of some overtone (Figs. 7, 10). There was little detect- displaying males before seeingtheir leg bands, able energy in higher harmonics on sound just by observing their Flutter rates. spectrograms.A typical DFV beganwith a brief high amplitude element that ascendedand de- VOCALIZATIONS ASSOCIATED WITH scended in frequency suddenly. This was fol- DISPLAY FLIGHTS lowed by several briefer, lower amplitude and Display Flight Vocalizations: description, in- lower frequency elements, whose number and dividuality, variants, and sequences.Males in relative amplitude varied among males (Figs. DFs emitted a rhythmically repeated monot- 7, 8, 9). The last of these sometimes merged onous call, the DFV (Figs. 7, 8, 9). Simulta- with the long final element. This last element neous sequencesof DFVs and PF formed the rose slowly in frequency and amplitude; its longestand most conspicuousportions of DFs, maximal frequency and amplitude were near and they sometimes continued for minutes its end, and were about the same as for the without pause. Sequencesof DFVs were oc- lead element. Rhythmic amplitude modula- casionally given by males perched on the tions (AM) sometimesoccurred in the first part ground or up to several meters high on a prom- of the last element (parts of Fig. 9) and inence. Such sequencesgenerally led to air- rhythmic frequency modulation (FM) some- borne DFVs in a standard DF. With this ex- times occurred throughout it (parts of Fig. 8). ception, DFVs were unique to DFs, and hence A bivariate plot of durations of DFVs and were given only by males. ICIs revealed a clusteringof most males,though The unweightedmean of the mean durations a few males were extreme in one or both char- of DFVs for 22 males (totalling 866 DFVs) acteristics (Fig. 11). Nevertheless, using only was 390 ms, with individual means ranging these two temporal descriptors, many of the from 331 to 504 ms. Means of 880 interval 22 males could be recognized: 13 males were durationsbetween consecutive calls of 22 males distinguishablefrom all others, two were dis- (Inter-call Intervals, or ICIs) had an unweight- tinguishablefrom 20, three from 19, three from ed mean of 106 ms, and ranged from 79 to 18 and one from 16 (Fig. 11). Homogeneous 13 1 ms. Thus calls occurred at a rate of about subsetsfor ICI duration were larger and showed two per second; Dwight (1895: 18) estimated more overlap than did those for DFV duration 228 EDWARD H. MILLER

e 4 c

250 msec FIGURE 8. Examplesof Display Flight Vocalizations (DFVs) and partial DFV sequencesfor 10 male Least Sandpipers (a-j). These are ink tracings of the first harmonic of the fundamental frequency, based on sound spectrograms.The frequency scaleis in kHz. Analyzing filter bandwidth, 300 Hz.

(Fig. 11); this suggeststhat males exhibited If a male Parachuted all the way to the ground greater individuality in duration of DFVs than he sometimes sanguntil touchdown, or started of ICIS. Chattering as he neared the ground. Chattering Song and Chattering: contextsof use. Least generally became softer near the ground and Sandpipers had a complex, rich vocalization then continued more loudly for a brief period that I term Song (Figs. 12-l 6). This was given on the ground, sometimes after a brief period occasionallyduring the courseof DFs, but was of silence. Further Song was also sometimes most common around the time of descentfrom given after landing. DFs, and during aggressiveand sexual chases. Most Chattering and Song was not sponta- SpontaneousSong or Chattering was a nor- neous but was clearly elicited by the activities mal part of pre-descent and descentphases of or presence of other males or females. This DFs, and occurred in periods of Parachuting was true of Song in mutual BS and in most Song embedded in DFs also. As will be de- periods of PS embedded in DFs. Males on the scribed, the quality of DFVs changed during ground gave Song, Chattering or DFVs (or all those phasesand Song followed, generally dur- three) in responseto calling by other males, or ing breakdown of PF, or during the Parachut- even in responseto the sight of a neighboring ing phase, or both (but before stoop descent). male silently ascending into DF. Males in- LEAST SANDPIPER DISPLAY FLIGHTS 229

b

1 dw+4b+4+d 500 500

f D

125

FIGURE 9. Oscillogramsof Display Flight Vocalizations (DFVs) of male Least Sandpipers.Partial DFV sequences of six males are illustrated (a-f). g-Four DFVs on a different time scale.These correspondto the DFVs marked with the same letters (A-D) for males c, d, e, and f, respectively. Time markers are in ms.

variably gave Song(less frequently Chattering) and in ground courtship; these are described as they leapt from the ground to chaseintrud- and discussed elsewhere (Miller, in press). ers or to begin DF prompted by another male’s Chattering seemed to be unique to males. For DF. Males often gave Song or Chattering as present purposes,it is sufficient to emphasize they flew toward human intruders throughout that Song and Chattering were prominent and the nestingcycle, even after the chickshatched, integral parts of DF activity. although this was lesscommon (“alarm” calls Description of Song. Songswere in the fre- were more usual then). This is a common re- quency band l-4.5 kHz (Figs. 12, 13). Tran- sponse of nesting shorebirds (Dixon 19 18, sitions from DFVs to Song were variable, but Brown 1962, Portenko 1972). the first recognizableelement of Songwas usu- Forms of Song were given by both sexesin ally one with rhythmic FM of a carrier fre- certain contexts outside of DFs, e.g., in brood- quency which rose from about 2.5 to 3.5 kHz. ing chicks, in gathering chicks to be brooded, This element type had no detectable harmon-

1 2 3 4

Frequency-kHz

FIGURE 10. Power spectraof DFVs of three male Least Sandpipersfrom northern Manitoba. 230 EDWARD H. MILLER

- = /

llO- ICI duration I ’ (msec) loo- I go- I ’

60- +

I I I I I I I I I I 330 350 370 390 410 430 450 470 490 510

DFV duration (msec) FIGURE 11. Bivariate plot of mean durations of Inter-Call Intervals (ICIs) and Display Flight Vocalizations (DF’Vs), for 22 male Least Sandpipers.The 95% confidencelimits on the means are also shown. Homogeneoussubsets on each of the two variables are shown above and to the right (by Student-Newman-Keul multiple range test, 01= 0.05). Those males that form homogeneoussubsets with one another on both variables are enclosedby dashed lines. its, and lasted about 0.33-0.5 s (Fig. 12c, ele- tained more energy during the initial dip in ment A; Fig. 13, lead element in a, b, c, e, f frequency (Fig. 12c, elements B, C; Fig. 12f, and h). Usually only one of these elements elements A, B). Amplitude profiles of this ele- occurred. Rhythmic AM occurred throughout ment type varied greatly. Those for two con- it and was linked to FM (compare element A secutive elements of one male were different in Figs. 12c and 14a, and the same element (Fig. 14a, elements B, C), although the first of type in Figs. 13a and 15a [A and B]). Otherwise these was similar to one of another male (Fig. the amplitude profile was marked only by a 15a, element C). These differenceswere in the terminal amplitude peak (Fig. 14a, element A; presence or absence of rhythmic AM. How- Fig. 15a, elements A and B). ever, these elements plus those of a third male The element describedwas usually followed (elements A and B of Figs. 12f and 16) shared by a second type with a much lower funda- the feature of amplitude-frequency coupling mental frequency (around 1.5-2 kHz) and for both rhythmic and non-rhythmic modu- strong harmonic structure (Fig. 12c, elements lations (e.g.,all examplesshowed an initial drop B, C; Fig. 12f, elements A, B; Fig. 13a, third in frequency and amplitude, and all showed a element; Fig. 13b, secondelement; etc. in Fig. terminal amplitude rise coincident with the 13c-g). This element startedaround 1.5-2 kHz, sudden frequency shift then). dropped in frequency just after its beginning, The next element type began in a very sim- then increasedto around its startingfrequency, ilar manner to the element just described,with increased slowly or remained nearly constant a drop, then an increase in frequency at about in frequency, and ended with a sudden rise to the starting level; it was also very similar in a brief, non-harmonic ending around 3-3.5 frequency band, harmonic structure, and the kHz. Rhythmic FM often occurred between common reversal of harmonic strength in the the early drop in frequency and the end (Fig. initial frequency dip (Fig. 12c, element D; Fig. 12c, element B, Fig. 13a, c, e). The frequency 12f, elements C-E; Fig. 13a, fourth through spectrum of this element changed over its twelfth elements; Fig. 13b, third through sixth length; most energy was in the fundamental elements;etc. in Fig. 13c-h). This element type frequency, but its first harmonic overtone con- undoubtedly had a common origin with the LEAST SANDPIPER DISPLAY FLIGHTS 231

4 a b

l- A B CD E F G -H 4- d e

FIGURE 12. Sound spectrogramsof Songsof male Least Sandpipers.a, b, c-One male’s Song shown in narrow (a) and wide (b, c) band representations.Song elements marked “A” to “H” in panel c correspond to the oscillograms marked the same in Figure 14. An ink tracing of this Song is shown in Figure 13~. d, e, f-A second male’s Song depicted in narrow (d) and wide (e, fj band representations.Song elements marked “A” to “G” in panel f correspond to the oscillogramsmarked the same in Figure 16. The Chattering which precedeselement A and follows element G is also shown as oscillogramsin Figure 16d, e and f. The time marker in the bottom right corner of eachpanel represents 500 ms. The frequency scale is in kHz. Analyzing filter bandwidths, 45 Hz (narrow) and 300 Hz (wide). secondelement type, but is considereddistinct spectrogramsprepared on the range SO-S,000 here becauseit occurred in a graded series of Hz (Fig. 12c, I), becauseof rapid rhythmic AM varying length and did not end with a sudden (Fig. 14, elements E-H, Fig. 15, elements M- upward frequency shift (except for the last one Q; Fig. 16, elements F, G). Amplitude in- in each series;see below). Grading was appar- creasedstrongly at the end. ent mainly as shortening on sound spectro- Songs of most males had all four types of grams, and also as declining amplitude in os- elements, in the order described. Two males cillograms (Fig. 15, elements D-L; Fig. 16, each had one element type missing in their elements C-E). Amplitude profiles began as Songs (Fig. 13d, h). Songs of males varied did the second element type, with a drop then mainly in the number of elements of the last recovery in amplitude. They thus had a dumb- type. Males tended to emit Songs in rapid bell shapeoverall, becauseof their brevity. At successionand to add elements of that type, least one of these elements was in every Song especially,when fighting or when chasingother examined. The seriesended with a sudden up- birds in flight. ward frequency shift to the last element type The integration of Song into DFs varied (for comments on grading of this element type, greatly. Some examples are shown in Figures see Miller 1979~). 17 and 18. DFVs were briefer and delivered The fourth type of Song element was about more rapidly as they blended into Song with- 3-3.5 kHz in frequency, and rose gently over out pause (Fig. 17d) with normal pause (Fig. its frequency range (Fig. 12c, elements E-G; 17f, g-3) or with a long silent period. DFVs Fig. 12f, elements F, G; Fig. 13a, last five ele- preceding Song sounded louder and more ments; Fig. 13b, last six elements; etc. in Fig. highly pitched than normal DFVs. DFVs just 13 c-h). It had no detectable harmonic struc- before Song were sometimes aberrant in form ture, and had a pulsed appearance on sound (Fig. 17c, g-2) or were followed by unusual 232 EDWARD H. MILLER

FIGURE 13. Examples of Song associatedwith Display Flights of eight male Least Sandpipers (a-h). Record g is incomplete; its initial elements are unclear on the tape. Record a startsimmediately from a Display Flight Vocalization, only the terminal portion ofwhich is depicted. These are ink tracingsof the fundamental frequencyand its first harmonic for all elements, based on sound spectrograms.The Song of male a is representedas oscillogramsin Figure 15. The Songof male c is representedas sound spectrogramsand oscillogramsin Figures 12 a-c and 14, respectively.The dash at the beginning of each record represents3 kHz. Analyzing filter bandwidth, 300 Hz.

elements that preceded Song (Fig. 17e, g- 1, introductory elementsto further Song,and that g-4). Chattering sometimes preceded Song. those elements commonly led into Chattering Transitions from Chattering to Song included (Fig. 18c, e-l, e-3). By far the commonest se- some minor variants in form (Fig. 17a, b), but quel to Song was Chattering, which was often were generally less variable than were transi- introduced by various aborted elements of oth- tions from DFVs to Song. er derivations (as described) or of Chattering The manner in which Song ended depended (Fig. 18e-2, f-l, f-2, f-3?, f-6?, f-7?, g-l, h-2, largely on its context. Songwas sometimes fol- h-3). Even if silence followed Song, there was lowed by resumption of DFVs that were aber- often a suggestionof one or a few Chattering rant in shape,and were brief and given in rapid elements immediately after the last Song ele- but declining cadence(Fig. 18g-1). Some were ment (Fig. 18f-4, f-5, h-4). followed by an aberrant DFV or two alone Description of Chattering. Chattering was a (Fig. 18d) or in transition to Chattering (Fig. compound-repetitive call associatedintimate- 18b, f-6?, g-2?, h- 1). Since Songwas sometimes ly with Song. It varied greatly among males, repeated several times without pause, it is not but usually consistedof triplets or quartets of surprisingthat some Songsended with aborted elementsrepeated in loud rapid succession(Fig.

FIGURE 15. Oscillogramsof a Song of a male Least Sandpiper. Elements marked “A” to “Q” correspond to the same calls in panels a to e of this figure. The element preceding“A” is a Display Flight Vocalization (DFV), and “A” startswith an incomplete DFV (see Fig. 13a, where this Songis shown as an ink tracing of a sound spectrogram).Time markers are in ms, except in panel a. a C D E

b

FIGURE 14. Oscillograms of a Song of a male Least Sandpiper. Elements marked “A” to “H” correspond to the sound spectrogramsmarked the same in Figure 12c, and also to the same calls in panels a to e of this figure. This song is also shown as an ink tracing of a sound spectrogramin Figure 13~. a-Entire Song. b-e-Song elements shown on a different time scale. Time markers are in ms.

a

c D------LM N

250 I 234 EDWARD H. MILLER

a

f

FIGURE 16. Oscillograms of Song and Chattering of a male Least Sandpiper. Elements marked “A” to “G” correspond1to the sound spectrograms marked the same in Figure 12f, and also to the same calls in panels a to c of this figure. The entire Song is shown in panel a and its elements and subsequentChattering are shown on different time scalesin panels b, c and d. The terminal part of a long Chattering seriesby a different male is shown on two time scalesin panels e and f. The Chattering sequencesmarked as “f’ and “H” in panel e are shown as sound spectrograms in Figure 19b and c; the segment “I” also correspondsto panel f in this figure. Panel e is indicated as “e” in Figure 19b. Time markers are in ms.

12 d-c Figs. 16-20). Most Chattering showed I have described transitions between Song lessenergy in the fundamental frequency than and Chattering. Chattering sometimes also in its first harmonic overtone (Fig. 19). It was graded directly into DFVs and vice versa (Fig. the most variable call emitted by males in DFs, 17h). DFVs that blended into Chattering or and showed strong changesin amplitude, fre- followed it were less aberrant than were those quency, and harmonic content even within that preceded or followed Song, but their du- single series (Figs. 16-20). However, such ration and cadence followed the same trends: changeswere gradual, and prolonged seriesof DFVs were briefer and of faster cadence im- loud Chattering usually varied little; they were mediately before and after Chattering,and were particularly constant in the intervals between increasingly normal as they fell away to either elements and in the relative amplitudes and side of it. frequencies of the elements. The latter two characteristicsplus the number of component DISCUSSION elementsper compound unit of repetition were The foregoing descriptions should facilitate characteristic of individual males, though the recognition and description of similar and dif- number of elementsper compound unit varied ferent display componentsin aerial displays of with behavioral context; e.g., the number of related species.Few adequate descriptions of elements was high just before and after Song relevant displays of calidridines and other (Fig. 12d-f; Fig. 16a, d; Fig. 17; Fig. 18). charadriiform species have been published LEAST SANDPIPER DISPLAY FLIGHTS 235

I I 250 msec FIGURE 17. Beginningsof Songby six male Least Sandpipers(a-g), and transition from Chattering to Display Flight Vocalizations (DFVs; h). These are ink tracingsof the fundamental frequency and its first harmonic, based on sound spectrograms.The fundamental is replacedby a dashedline where it was either too faint to portray or was not involved in transitions; it is omitted altogether from panel h, though it is present for that entire series. a-Early descentfrom Display Flight (DF); the rhythm of PunctuatedFluttering was broken, but the Song occurred before stoopingdescent. b-During descentfrom DF. c-Brief ParachutingSong embedded in DF. d-During slowing- down of DF (i.e., lengthening of Inter-Flutter Intervals), with slight loss of altitude. e-Preceding descent from DF, with Chattering. h-Transition from Chattering to DFV’s, about 10 s after Song during the same DF. Frequency markers are in kHz. Analyzing filter bandwidth, 300 Hz.

(acoustic displays are reviewed by Miller [in Sandpiper (C. pusilla)in northern Manitoba, press]).Most of the following discussiondeals while breeding density increasesin the same with convergent and adaptive features of aerial order (pers. observ.). Display height is low in displays of the Least Sandpiper, and evaluates the semi-colonial Western Sandpiper (C. mau- the homologous status of only some display ri; Holmes 1971, 1973) and in the densely characteristicsin the species. nesting Temminck’s (C. temminckiz),’ a specieswhich also relies heavily on songposts CONVERGENCE AND ADAPTATION IN (Hilden 1965, 1975, 1978, 1979). I know of DISPLAY FLIGHTS no data on the relationship of breeding density Birds that inhabit open country commonly to display height within species. perform aerial displays in long-distance sig- The long broadcastdistances for specieswith nalling, since the higher they are, the farther medium to low breeding densities must set soundswill carry (Armstrong 1963). Altitude some limits to the structure of aerial displays. also increasesa bird’s visual conspicuousness, Such displays must be obvious, stereotyped, and the distinctive flight pattern of male Least redundant, and physically adapted to with- Sandpipersin Display Flight is detectable over stand attenuation and environmental degra- long distances. Considering this relationship dation, for maximal transmissionand accurate of transmission distance to display height, it reception to occur. Features of the main part is unsurprising that display height varies in- of the DF of C. minutillaillustrate some con- versely with breeding density among species. formity with thesepoints. First, the distinctive For example, the Red Knot (Culidriscanutus) pattern of Punctuated Fluttering during DFs nests at low densities, and male knots have is visible over long distances, and is stereo- very high aerial displays(Manniche 19 10, Net- typed and redundant through its rhythmical tleship 1974). Height of aerial display is great- repetition. Furthermore, the approximately est in the Stilt Sandpiper (C. himantopus),low- stationary position of displaying birds aids re- er in the Dunlin (C. alpina),lower yet in Least ceivers in locating them. The same points ap- Sandpiper, and lowest in the Semipalmated ply to Display Flight Vocalizations. Redun- 236 EDWARD H. MILLER

____ _. 2 , \ ______--__- ______..------J

FIGURE 18. Endings of Song by eight male Least Sandpipers (a-b). These are ink tracings of the fundamental frequency and its first harmonic, based on sound spectrograms.The fundamental is replaced by a dashed line where it was too faint to portray. a- During joint Butterfly Song(BS) descentwith a neighbor. b-In descent;Chattering continues beyond the sequence illustrated. c-During BS about 1 m high, after a leap from the ground in responseto audible Display Flight (DF); four further triplets of Chattering conclude this record (not illustrated). d-Slowing-down in DF (i.e., lengtheningof Inter- Flutter Intervals), with a slight loss in altitude. e-l -During DF, in responseto a pair flying into the DF area of the calling male. e-2-During descent from DF, Chattering continues beyond the sequenceillustrated. e-3-Song in DF long before descentstarts. f-l -Spontaneous Songduring DF; silencefollows the record shown. f-3-Song during mutual BS descentwith a neighbor; Chattering continuesbeyond the sequenceillustrated. f-4, f-S-Silence follows the records shown. f-6-Song during high altitude aerial tussle with another (presumed) male; Chattering continues beyond the sequenceillustrated. f-7-Song during high altitude aerial tussle with another (presumed) male; Chattering declines then stops shortly after this record. g-l, h-l-Silence follows the records shown. h-2-Chattering continues beyond the sequenceillustrated. h-3-Song early in descent from DF. The rhythm of Punctuated Fluttering was broken, but the Song occurredbefore stooping descent. h-4-During DF; silence follows the record shown. Frequency markers are in kHz. Analyzing filter bandwidth, 300 Hz.

dancy in both cases is manifest not through lengthy aerial displays of many taxa, including the correlation of parts of a complex pattern, Furnariidae, Motacillidae, and Charadriidae but rather through rhythmical repetition of a (pers. observ.). Such widespread convergence simple one. This reflects the simple, unchang- suggestsadaptiveness, which may lie in the ing information content of DFs over most of opportunities of signal averaging afforded to their course. listeners, in an environment that severely de- Rhythmically repeated, simple calls occur in grades sounds because of wind and tempera- LEAST SANDPIPER DISPLAY FLIGHTS 237

. . . ..I..“..

b

FIGURE 19. Sound spectrogramsof Chattering by a male Least Sandpiper. A sequenceis shown in narrow (a) and wide (b, c) band representations.Segments marked “e” and “f” correspondto the oscillogramsin Figure 16e and f, respectively;the segmentmarked “H” correspondsto the segmentmarked the same in Figure 16e. These letters also correspondto the same segmentsin panels b and c of this figure. The time marker in the bottom right corner of each panel represents 500 ms. The frequency scale is in kHz. Analyzing filter bandwidths, 45 Hz (narrow) and 300 Hz (wide). ture (see Schleidt 1973, Wiley and Richards sult that spectral structure remains fairly con- 1982). One feature of DFVs themselves,which stant regardlessof transmissiondistance. Also, is probably adaptive, is their relatively narrow this representsefficient “packaging” of energy band structure. This characteristic minimizes into calls (see Miller and Baker 1980). Most differential frequency attenuation, with the re- features of the calls themselvesare lessreadily

250msec

FIGURE 20. Examples of Chattering from seven male Least Sandpipers (a-g). These are ink tracings of the first harmonic of the fundamental frequency,based on sound spectrograms.Frequency markers are in kHz. Analyzing filter bandwidth, 300 Hz. 238 EDWARD H. MILLER

interpreted, however (see Wiley and Richards characteristics probably vary in what those 1982): distancesare. Experimental evidence is needed 1. The frequency spectra of DFVs of Least to evaluate further the adaptiveness of DFV Sandpipers and other calidridines (including structure. syntopic species) vary substantially (see The adaptive significance of other display Kroodsma and Miller 1982). This would not components of DF’s can not be evaluated in be expected if spectra were closely adapted to a similar way, because they are not used in long-distance transmission in such simple, simple long-distance communication, with similar habitats. likely adaptations to the physical environ- 2. Frequency modulation is a good way to ment. Song, Chattering, and associatedvisual encode information for long-distance signal- displaysby males presumably evolved through ling, and amplitude modulation is a poor way, sexualselection, as have DFVs and Punctuated because of its high susceptibility to environ- Fluttering. Unlike the latter two components, mental degradation. Both increase gradually they are often transmitted over short distances over the course of DFVs, and FM is also ex- to particular receivers of known sex, identity, pressedin the lead element of DFVs, with AM and pairing status,and they encode more com- in comparable or greater detail in the subse- plex messages,and can assumemore complex quent soft, brief elements. Thus, patterns of and variable form than can DFVs. All are AM and FM in DFVs do not differ in a way emitted during major changesin locomotion, predicted by signal detection theory. including ascent into DFs and descent from In an adaptational perspective, several al- them, as well as during diverse aerial approach ternative explanations for these observations behavior toward males or females. Descents can be proposed. First, calidridine sandpipers from DFs show particularly striking visual and differ in their breeding densities, so the fre- acoustic displays, a seemingly universal char- quency spectraof their long-distancecalls may acteristic in (see next section). In brief, reflect closeadaptation to species-typicalspac- DFVs and Punctuated Fluttering encode non- ing patterns and microhabitat. Wasserman behavioral information about a male’s pairing (1979) and Gish and Morton (198 1) have doc- status, individual identity, and location, and umented adaptive differences for maximal they encode general sorts of behavioral infor- transmission in song of White-throated Spar- mation about a male’s activity and attentive- rows (Zonotrichia albicollis) and Carolina ness.More complex display components differ Wrens (Thryothorusludovicianus) between primarily in also encoding behavioral infor- habitat types and from different parts of the mation about locomotory, approach, and in- species’ range, respectively. Second, different teractive behavior (see Smith 1977). It is these features of DFVs may encode different sorts messagesthat have molded the complexity of of information, some for nearby neighboring Song and Chattering and the graded nature of males, others for more distant, anonymous re- Chattering, under the influence of sexual se- ceivers. Call characteristicsthat are important lection. for individual identification among neighbors may be degradable or not. Hence, they may HOMOLOGIES OF VISUAL COMPONENTS apply to such features as temporal patterning and amplitude envelopes of the soft, brief in- For homologies to be useful in suggestingaf- termediate elements of DFVs, as well as to finities, they must represent shared derived grosstemporal patterning within DFV-calling states (synapomorphies), and not shared an- sequences.Call characteristicsmay also be de- cestral or uniquely derived ones (symplesio- signed to provide information to receivers morphies and autapomorphies, respectively; about their distance from callers (though such Eldredge and Cracraft 1980). information must be very approximate in an Visual displaysassociated with DFs in Least environment so physically variable becauseof Sandpipers show strong similarities to some turbulence, etc.), and such information can sketched by Drury (196 l), notably Stooping come from sound features that degrade pre- descent,Parachuting, and Punctuated Flutter- dictably with distance. Evidence for differen- ing. The first of these is widespread in aerial tial responsivenessof receivers to songswith display of birds, and has probably arisen a different levels of natural degradation has been number of times. Stooping is apparently ubiq- provided by D. G. Richards (review in Wiley uitous in waders, and provides little insight and Richards 1982). In summary, long-range into relationships. Parachuting and Butterfly acousticsignals like DFVs are probably adapt- occur sporadically in Calidridinae, the former ed for transmission over optimal, not maxi- in Least and White-rumped sandpipers, the mal, distancesto biologically relevant receiv- latter in the same two species plus Baird’s ers (Lemon et al. 1981) and particular sound Sandpipers (Drury 1961). The seemingly re- LEAST SANDPIPER DISPLAY FLIGHTS 239 stricted occurrenceof thesetwo visual displays and referencesgiven below; Little ; La- may just reflect inadequate description, so butin et al. 1982), patrol flights (; evaluation of their ancestral/derived status Bergman 1946, Nettleship 1973; R. E. Gill, must wait. pers. comm.), switchback flight paths (Green- Punctuated Fluttering occurs in various shank, nebularia; Nethersole-Thomp- species outside the Calidridinae, in- son and Nethersole-Thompson 1979) undu- cluding the American (Scolopax lating flight (LesserYellowlegs, Tringaflavipes, minor; Brewster 1894) the Willet (Catoptro- and Redshank, Tringa totanus; Grosskopf phorus semipalmatus; Vogt 1938, Sordahl 1963, Rowan 1929) and numerous other elab- 1979) the Upland Sandpiper (Bartramia lon- orate variants (e.g., Pitelka 1943, von Frisch gicauda; Ailes 1976, pers. observ.), and the 1956, Lind 196 1, Haverschmidt 1963, Glutz Little Curlew (Numenius minutus; Labutin et von Blotzheim et al. 1975, 1977, Dabelsteen al. 1982). It is also probably homologous to a 1978, Allen 1980). The evolutionary origins component of aerial displays in certain Trin- of many such variants may lie in the correla- ginae and (Numenius), in which flut- tion between Display Flight termination and tering occurs during rises and gliding occurs display complexity. For example, most cali- during falls (Rowan 1929, Grosskopf 1963, Bi- dridines have fairly simple flight paths in DFs cak 1977, Glutz von Blotzheim et al. 1977). (exceptionsare the Spoonbill Sandpiper [Eury- Thus, Punctuated Fluttering is plesiomor- norhynchuspygmeus] and Red Knot; Dixon phous within the Calidridinae. 1918 [but see Portenko 19571, Nettleship A distinct form of aerial display in Calidri- 1974). The most complex motor patterns oc- dines has been mentioned by several authors, cur during terminal dives, which are almost though I have never seen it in Least Sandpi- always accompanied by Song, the most com- pers. Male Western Sandpiperstake flight with plex vocal utterance associatedwith DFs. Ter- a “slow but deep wingbeat,” then, at an alti- minal dives and flourishes appear to be given tude of about 5 m, slowly patrol the territory by all or nearly all wader species.Ritualization (Holmes 1973: 109). Slow wing beats also oc- of these during Display Flights would seem to cur in Baird’s Sandpiper (Drury 196 1, Holmes be a straightforward process,and has probably and Pitelka 1964, Holmes 1973) Dunlin occurred in various groups, including lap- (Cramp et al. 1983) the Little Stint (Cafidris wings, , tringines, and snipes. In the minuta; Cramp et al. 1983) the Curlew Sand- latter, ritualization of terminal diving has pro- piper (Calidris ferruginea; Holmes and Pitelka moted the evolution of non-vocal acoustic 1964), and the Purple Sandpiper (Culidris displays, based on vibration of the outer rec- maritima; Swanberg 1945, Bengtson 1970, trices in the air current modified by the set Cramp et al. 1983). Portenko (1972: 359) de- wings (Tuck 1972, Rutschke 1976, Reddig scribedmales of the Rufous-necked Sandpiper 1978, Sutton 198 1; seealso Labutin et al. 1982). (C a ld‘i rls ruji co11 1s ’ ) as “slowly beating their It is probably safe to consider simple flight wings like a bat that had just taken flight,” paths, without embedded dives, to be plesio- while flying low above the ground. Compa- morphous within the Calidridinae. rable displays also occur in Common Snipe The tail is not an important visual display ( gallinago), Eurasian Woodcock feature in Display Flights, despite its obvious (Scolopax rusticola), Charadriidae, Haemato- functional correlation with them. This is pre- podidae, Laridae (jaegers, Stercorarius), and sumably becauseof its inconspicuousnessand Thinocoridae, at least (Drury 196 1, Andersson its involvement in stabilizing adjustments un- 1973, Maclean 1969, Shorten 1974, Glutz von der windy conditions. In the Least and Upland Blotzheim et al. 1975, Miller and Baker 1980, sandpipersthe tail is spread only as necessary Phillips 1980, pers. observ.). The display is to adjust position (Ailes 1976). Spreading oc- thus plesiomorphous within the Calidridinae, curs consistently during the descent phasesof and its absencein the Least Sandpiper is apo- undulating flight in the LesserYellowlegs (and morphous. probably in other species,whose displays are The kind and amount of movement during yet undescribed), but this also appears to be Display Flights varies greatly across species. simply for aerodynamic reasons, probably to Considering such variations, it is easy to vis- slow the descent(Rowan 1929). Future obser- ualize the evolutionary elaboration of flight vations may disclose display functions of the path for display purposes. In Charadrioidea, tail in DFs of some calidridines. If so, such DFs include simple ascent followed by gliding states must be considered apomorphous. descent (Thinocoridae; Maclean 1969, pers. To summarize, the Least Sandpiper exhibits observ.), Punctuated Fluttering and treading plesiomorphy in Display Flights in Punctuated (certain speciesdiscussed in this paper), reg- Fluttering, simple flight pattern with no ritu- ular intermittent dives (snipes; Kliebe 1974, alized embedded dives (occasional brief and 240 EDWARD H. MILLER variable Song or Chattering false descentsare mous and nest at moderate to low population embedded), and lack of display use of the tail. densities. It is apomorphous in lacking “bat display.” CONCLUDING COMMENTS HOMOLOGIES OF ACOUSTIC COMPONENTS Inference about adaptationsin the structure of The few published accounts of aerial displays Display Flights is simplest for those charac- in waders, particularly calidridines, preclude teristics which encode simple behavioral in- comprehensive assessmentof homologies of formation, and which are transmitted over long Song and Chattering and of the ancestral/de- distances:Punctuated Fluttering and approx- rived status of particular features of Display imately stationary position. The significance Flight Vocalizations at present. of the form of Songand Chattering lies in their Homologues to DFVs may occur even in proximate effectson receivers, not on effective distantly related groups. Thus, the distinctive transmissibility, and assessmentof their adap- long call of curlews is probably homologous to tiveness depends upon which paradigm of the DFV of calidridines, because of its sim- communication one is working within (Marler plicity, repetitiveness, physical features, and 1967, Beer 1977, Smith 1977, Dawkins and association with Display Flights (Forsythe Krebs 1978, Morton 1982). Different para- 1967, 1970, Skeel 1976, 1978, Glutz von digmspredict different evolutionary modes and Blotzheim et al. 1977). Male Willets emit a patterns, and these are of varied significance repeated “pill-will-willet” during DFs, a call for the origin and evaluation of homologies. which is also a likely homologue of the DFV For example, predictions of signal structure of calidridines (Vogt 1938, Sordahl 1979). Fi- basedon models of maximal vs. optimal trans- nally, godwits (Limosa) utter simple, repeated mission differ, as do those based on models of calls in the prolonged circling phase of their deceptive (manipulative) vs. non-deceptive aerial displays (Bent 1927, Lind 1961, Hagar communication. In this paper, I have adhered 1966, Nowicki 1973). DFVs are thus plesio- to the view that signalsare adaptive in struc- morphous within the Calidridinae, and their ture, largely for consistency. It seems best to absencein certain speciesis apomorphous. sidestepthese admittedly crucial issuesat pres- Convergence, though, is partly responsible ent, and await further information on adaptive for the loss of Display Flight Vocalizations. radiation of calidridine Display Flights from The lekking Buff-breasted Sandpiper (Tryn- future studies. gites subrujicollis) has a silent flight display which is simply ritualized flutter-fighting, and ACKNOWLEDGMENTS it utters only soft “tick” soundsduring mate- This researchbegan at the suggestionof I. A. McLaren, attraction displays on the ground (Parmelee et whom I thank for his boundless support,ideas, and en- al. 1967, Sutton 1967); males of the Ruff (Phi- couragement.Many people have helped me by reviewing lomachuspugnax) have a similar flight display parts or all of drafts, collecting data, and assistingwith and are silent on their lekking hills (Hogan- field work: K. Abraham, F. Cooke, D. Fairfield, B. Fenton, D. Finch, R. Frisch, R. Gill, C. Handel, S. Hills, R. Holmes, Warburg 1966, van Rhijn 1973); the related J. Jehl, Jr., D. Jenni, K. Kovacs, P. Myers, L. Oring, D. Sharp-tailed and Pectoral sandpipers(Culidris Paulson, and W. J. Smith. I am grateful to H. Lundgren acuminata and C. melanotos)utter entirely dif- and B. Culik for their careful help with translations; E. ferent kinds of calls during low, rapid display Longpresand H. Hosford for help with the production of flights over their mating territories (Pitelka diagrams;and the Natural Sciencesand EngineeringRe- searchCouncil of Canada for financial support (operating 1959, Flint and Kishchinskii 1973, Myers grants to I. A. McLaren and E. H. Miller). 1982) (are these derivatives of DFVs?); and White-rumped Sandpipers possessa well-de- LITERATURE CITED veloped hovering flight but alsoproduce unique callsduring it (Sutton 1932, 196 1, Drury 196 1, AILES,I. W. 1976. Ecology of the Upland Sandpiper in central Wisconsin. M.Sc. thesis, Univ. of Wisconsin, Holmes and Pitelka 1962; however, the “ter- Stevens Point. ritorial song” shown as a sound spectrogram ALLEN, J. N. 1980. The ecology and behavior of the in Fig. 9 1 of Glutz von Blotzheim et al. [ 19751 Long-billedCurlew in southeasternWashington. Wildl. is certainly suggestiveof DFVs). The absence Monogr. 73. of DFVs in these speciesis partly due to an ANDERSON,K. S. 1980. Least Sandpiper (Calidris min- utilla) breeding in Massachusetts.Am. Birds 34~867. increasedemphasis on visual signals,promot- ANDERSSON,M. 1973. Behaviour of the Pomarine Skua ed by the moderate- to high-density breeding Stercorariuspomarinus Temm. with comparative re- conditions which characterize these polygy- marks on Stercorariinae.Omis Stand. 4: 1-16. nous forms. At least some calidridine species ARMSTRONG,E. A. 1963. A study of bird song. Oxford Univ. Press, London. are polygynousand possessDFVs (e.g., Curlew BEER,C. G. 1977. What is a display?Am. Zool. 17:155- Sandpipers [Holmes and Pitelka 1964]), but 165. most species that have DFV’s are monoga- BENGTSON,S.-A. 1970. Breedingbehaviour ofthe Purple LEAST SANDPIPER DISPLAY FLIGHTS 241

SandpiperCalidris maritima in West Spitsbergen.Cr- minck’s Stint Calidris temminckii. Omis Fenn. 56: nis Stand. 1:17-25. 56-74. BENT,A. C. 1927. Life historiesof North American shore HOGAN-WARBURG,A. J. 1966. Social behaviour of the birds. Part 1. Bull. U.S. Natl. Mus. 142. Ruff, Philomachuspugnax (L.). Ardea 54: 109-229. BERGMAN,G. 1946. Der Steinwalzer, Arenaria i. in- HOLMES,R. T. 1971. Density, habitat, and the mating terpres.(L.), in seiner Beziehung zur Umwelt. Acta system of the Western Sandpiper (Calidris maurt). Zool. Fenn. 47:1-136. decologia 7: 19l-208. BICAK, T. K. 1977. Some eco-ethological aspectsof a HOLMES.R. T. 1973. Socialbehaviour of breedina West- breedingpopulation of Long-billed Curlews (Numen- emSandpipers. Ibis 115:107-123. ius americanus),in Nebraska. MSc. thesis. Univ. of HOLMES,R. T., AND F. A. PITELKA. 1962. Behavior and Nebraska, Omaha. taxonomic position of the White-rumped Sandpiper. BREWSTER,W. 1894. Notes and song-flightof the Wood- Proc. Alaska Sci. Conf. 12:19-20. cock (Philohela minor). Auk 11:29l-298. HOLMES,R. T., AND F. A. PITELKA. 1964. Breeding be- BROWN,R. G. B. 1962. The aggressiveand distraction havior and taxonomic relationships of the Curlew behaviour ofthe Western SandpiperEreunetes mauri. Sandpiper. Auk 8 11362-379. Ibis 104:l-l 1. JOHNSGARD,P. A. 1981. The plovers, sandpipers, and CRAMP,S., ET AL. [EDS.]. 1983. Handbook of the birds snipesof the world. Univ. of NebraskaPress, Lincoln. of , the Middle East, and North Africa. The KLIEBE, K. 1974. Beobachtungen zur Flugbalz der birds of the Western Palearctic.Vol. 3. Oxford Univ. Zwergschnepfe (Lymnocryptesminimus). Vogelwelt Press, Oxford. 95:30-33. DABELSTEEN,T. 1978. An analysis of the song-flight of K~OODSMA,D. E., AND E. H. MILLER [EDS.]. 1982. In- the Lapwing (Vanellus vanellusL.) with respect to troduction, p. xxi-xxxi. In Acoustic communication causation, evolution and adaptations to signal func- in birds. Vol. 1. Academic Press, New York. tion. Behaviour 66: 136-178. LABUTIN,Y. V., V. V. LEONOVITCH,AND B. N. VEPRINTSEV. DAWKINS,R., AND J. R. KREBS [EDS.]. 1978. 1982. The Little Curlew Numenius minutus in Si- sianals: information or maninulation?. v. 282-309. beria. Ibis 124:302-3 19. In Behavioural ecology: an evolutionary approach. LEMON,R. E., J. STRUGER,M. J. LECHOWICZ,AND R. F. Sinauer Assoc., Sunderland, MA. NORMAN. 1981. Song features and singing heights DIXON, J. 1918. The nestinggrounds and nestinghabits of American warblers:maximization or optimization of the Spoon-billed Sandpiper. Auk 35387-404. of distance?J. Acoust. Sot. Am. 69:1169-l 176. DRURY,W. H., JR. 1961. The breedingbiology of shore- LIND, H. 1961. Studies on the behaviour of the Black- birds on Bylot Island, Northwest Territories, Canada. tailed (Limosa limosa(L.)). Munksgaard,Co- Auk 78:176-219. penhagen,Denmark. DWIGHT,J., JR. 1895. The Ipswich Sparrow Ammodra- MACLEAN,G. L. 1969. A study of Seedsnipein southern musprinceps(Maynard) and its summer home. Mem. . Living Bird 8:33-80. Nuttall Omithol. Club No. 2. MANNICHE.A. L. V. 1910. The terrestrial mammals and ELDREDGE,N., AND J. CRACRAFT. 1980. Phylogenetic birds of north-east Greenland. Medd. Gronl. 45:1- patterns and the evolutionary process:method and 200. theory in comparativebiology. Columbia Univ. Press, MARLER,P. 1967. Animal communication signals.Sci- New York. ence 157:769-774. FLINT, V. E., AND A. A. KISHCHINSIUI. 1973. Data on MILLER,E. H. 1977. Breedingbiology ofthe Least Sand- the biology ofthe Siberian PectoralSandpiper, p. 1OO- piper, Calidrisminutilla (Vieill.), on SableIsland, Nova 105. In Fauna and ecologyofwaders, Issue 1. Moscow Scotia, Ph.D. diss., Dalhousie Univ., Halifax, Nova Society of Naturalists, Moscow. Scotia. FORSYTHE,D. M. 1967. Vocalizations of the Long-billed MILLER, E. H. 1979a. Functions of display flights by Curlew. M. SC.thesis, Utah State Univ., Logan. males of the Least Sandpiper, Calidris minutilla FORSYTHE.D. M. 1970. Vocalizations of the Lonn-billed (Vieill.), on Sable Island, Nova Scotia. Can. J. Zool. Curlew. Condor 72:213-224. 57:876-893. GISH, S. L., AND E. S. MORTON. 1981. Structural adap- MILLER, E. H. 1979b. Flight displays of Least Sandpi- tations to local habitat acousticsin Carolina Wren pers. Bull. Wader Study Group 26:44-45. songs.Z. Tierpsychol. 56: 74-84. MILLER, E. H. 1979~. An approach to the analysis of GLUTZVON BLOTZHEIM, U. N., K. M. BAUER,AND E. BEVEL gradedvocalizations of birds. Behav. Neural Biol. 27: [EDS.]. 1975. Handbuch der Vogel Mitteleuropas, 25-38. Band 6, (1. Teil). Akademisch Ver- MILLER, E. H. In press. Communication in breeding lagsgesellschaft,Wiesbaden, West Germany. shorebirds.In J. Burger and H. Olla [eds.], Behavior GLUTZVON BLOTZHEIM. U. N., K. M. BAUER.AND E. BEZZEL of marine . Vol. 5. Shorebirds. Plenum Publ. [EDS.]. 1977. Handbudh der Vogel Mitteleuropas, Corp., New York. Band 7, Charadriiformes (2. Teil). Akademisch Ver- MILLER,E. H., AND A. J. BAKER. 1980. Displays of the lagsgesellschaft,Wiesbaden, West Germany. Magellanic Oystercatcher(Haematopus leucopodus). GROSSKOPF,G. 1963. Weitere Beitrage zur Biologie des Wilson Bull. 92: 149-168. Rotschenkels.Vogelwelt 84:65-84. MOORE,R. T. 1912. The Least Sandpiper during the HAGAR.J. A. 1966. Nesting of the Hudsonian Godwit nestingseason in the Magdalen Islands. Auk 29:2 1O- at Churchill, Manitoba. Living Bird 5:5-43. 223. HAVERSCHMIDT,F. 1963. The Black-tailed Godwit. E. J. MORTON,E. S. 1982. Grading, discreteness,redundancy, Brill, Leiden, Netherlands. and motivation-structural rules, p. 183-2 12. In D. E. HILDBN,0. 1965. Zur Brutbiologiedes Temminckstran- Kroodsma and E. H. Miller [eds.], Acoustic com- dlaufers, Calidris temminckii (Leisl.). Omis Fenn. 42: munication in birds. Vol. 1. Academic Press, New l-5. York. HILDBN,0. 1975. Breeding system of Temminck’s Stint MYERS,J. P. 1982. The promiscuousPectoral Sandpiper. Calidris temminckii. Omis Fenn. 52: 117-146. Am. Birds 36: 119-l 22. HILDBN,0. 1978. Population dynamics in Temminck’s NETHERSOLE-THOMPSON,D., AND M. NETHERSOLE- Stint Calidris temminckii. Oikos 30: 17-27. THOMPSON.1979. Greenshanks.T. & A. D. Poyser, HILDBN,0. 1979. Territoriality and site tenacity of Tem- Berkhamsted,England. 242 EDWARD H. MILLER

NETTLESHIP,D. N. 1973. Breedingecology of Turnstones SUTTON,G. M. 1961. Bonaparte’s Sandpiper Calidris Arenaria interpresat Hazen Camp, Ellesmere Island, fuscicollis(Vieillot), Life history, p. 3221326. In D. N.W.T. Ibis 115:202-2 17. A. Bannerman led.1. The birds of the British Isles. NETTLESHIP,D. N. 1974. The breeding of the Knot Cal- Vol. 9. Scolopa&dae.Oliver & Boyd, London. idriscanutus at Hazen Camp, EllesmereIsland, N.W.T. SUTTON, G. M. 1967. Behaviour of the Buff-breasted Polarforschung44:8-26. Sandpiper at the nest. Arctic 20:3-7. NOWICKI,T. 1973. A behavioral study of the Marbled SUTTON,G. M. 1981. On aerial and ground displays of Godwit in North Dakota. M.Sc. thesis,Central Mich- the world’s snipes. Wilson Bull. 931457-477. igan Univ., Mount Pleasant. SWANBERG,P. 0. 1945. Om skarsnlppan, Calidris m. PARMELEE,D. F., H. A. STEPHENS,AND R. H. SCHMIDT. maritima (Briinnich), i de svenska ijgllen och dess 1967. The birds of southeasternVictoria Island and hickningsbiologi. Fauna Flora (Stockh.) 40: 108-l 33. adjacent small islands. Bull. Natl. Mus. Canada 222. TOWNSEND,C. W. 1927. Pisobiaminutilla (Vieillot), Least PHILLIPS,R. E. 1980. Behaviour and systematicsof New Sandpiper, p. 202-204. In A. C. Bent, Life histories Zealand plovers. Emu 80: 177-197. of North American shore birds. Part 1. Bull. U.S. PITELKA,F. A. 1943. Territoriality, display, and certain Natl. Mus. 142. ecologicalrelations of the American Woodcock. Wil- TUCK, L. M. 1972. The Snipes: a study of the son Bull. 55:88-l 14. Capella. Can. Wildl. Serv. Monogr. No. 5. PITELKA,F. A. 1959. Numbers, breeding schedule,and VAN RHIJN,J. G. 1973. Behaviouraldimorphism in male territoriality in Pectoral Sandpipersof northern Alas- RUES,Philomachus pugnax (L.). Behaviour 57: 153- ka. Condor 6 11233-264. 229. PITELKA,F. A., R. T. HOLMES,AND S. F. MACLEAN,JR. VOGT,W. 1938. Preliminary notes on the behavior and 1974. Ecology and evolution of social organization ecology of the eastern Willet. Proc. Linn. Sot. N. Y. in Arctic sandpipers.Am. Zool. 14:185-204. 49:8-42. PORTENKO,L. A. 1957. Studien an einigen seltenenLim- VON FRISCH,0. 1956. Zur Brutbiologie und Jugenden- icolen aus dem niirdlichen und iistlichen Sibirien. I. twicklung des Brachvogels(Numenius arquata L.). Z. Die Lijffelschnepfe-Eurvnorhvnchus.pwmaeus - (L.). Tierpsychol. 13:50-8 1. J. Ornithol. 98:454-466.- WASSERMAN,F. E. 1979. The relationship between hab- PORTENKO.L. A. 1972. Birds ofthe Chukotskii Peninsula itat and songin the White-throated Sparrow. Condor and irangel Island, Part I. Nauka, Leningrad. 8 1~424-426. REDDIG,E. 1978. Der Ausdruckflugder Bekassine(Ca- WILEY, R. H., AND D. G. RICHARDS. 1982. Adaptations Della aallinapo ,!zallinapo).J. Ornithol. 119:357-387. for acoustic communication in birds: sound trans- ROWAN, w. 1925. hates &Alberta Waders included on mission and signal detection, p. 13l-l 8 1. In D. E. the British list. Part VII. Tringa jlavipes, Yellows- Kroodsma and E. H. Miller [eds.], Acoustic com- hank. Br. Birds 23:2-17. munication in birds. Vol. 1. Academic Press, New RLJTSCHKE,E. 1976. Zur Struktur des Schallfedem von York. Bekassinen(Capella). Beitr. Vogelkd. 22: 12-25. SAUNDERS,W. E. 1902. Birds of Sable Island, N. S. Ot- VertebrateZoology Division, British Columbia Provincial tawa Nat. 16:15-3 1. Museum, Victoria, British Columbia VSV 1X4, Canada SCHLEIDT,W. M. 1973. Tonic communication:continual and BiologyDepartment, Universityof Victoria, Victoria, effectsof discretesigns in animal communication sys- British Columbia VSW 2Y2 Canada. Received 1 April tems. J. Theor. Biol. 42:359-386. 1982. Final acceptance26 October 1982. SHORTEN,M. 1974. The EuropeanWoodcock (Scolopax rusticola).A searchofthe literature since 1940. Game Conservancy,Fordingbridge, Hampshire, England. SKEEL,M. A. 1976. Nesting strategiesand other aspects of the breeding biology of the Whimbrel (Numenius APPENDIX. Abbreviations used in the text for features phaeopus)at Churchill, Manitoba. M.Sc. thesis,Univ. of the Display Flight and vocalizations of the Least Sand- of Toronto, Toronto, Ontario. piper. SKEEL,M. A. 1978. Vocalizations of the Whimbrel on its breeding grounds. Condor 80: 194-202. AM Amplitude Modulation SMITH,W. J. 1977. The behavior of communicating:an BS Butterfly Song ethological approach. Harvard Univ. Press, Cam- DF Display-Flight bridge. MA. DFV Display Flight Vocalization SORDAH~T. A. 1979. Vocalizations and behavior of the FM Frequency Modulation Willet. Wilson Bull. 9 1:55l-574. ICI Inter-call Interval SUTTON,G. M. 1932. The exploration of Southampton IF1 Inter-Flutter Interval Island, Hudson Bay. Part II, Zoology. Section 2. The PF PunctuatedFluttering birds of Southampton Island. Carnegie Institute, PS ParachutingSong Pittsburgh.