DO WOODPECKER DRUMS ENCODE INFORMATION FOR SPECIES RECOGNITION?1

Page created by Scott Francis
 
CONTINUE READING
The Condor 103:143–150
q The Cooper Ornithological Society 2001

        DO WOODPECKER DRUMS ENCODE INFORMATION FOR
                   SPECIES RECOGNITION?1

                DANIELLE J. DODENHOFF2,3, ROBERT D. STARK2                AND   ERIC V. JOHNSON
     Biological Sciences Department, California Polytechnic State University, San Luis Obispo, CA 93407

              Abstract. We investigated whether woodpecker drums are exchanged interspecifically
            and attempted to ascertain the drum variables responsible for species recognition. Playback
            experiments were conducted on four woodpecker species (Nuttall’s Picoides nuttallii, White-
            headed P. albolarvatus, Hairy P. villosus, and Downy P. pubescens Woodpeckers) com-
            paring behavioral responses to sympatric hetero- and conspecific drums. There was a sig-
            nificant difference in behavioral responses to sympatric hetero- versus conspecific drums;
            individuals responded less intensely to drums of heterospecifics when drum cadences were
            dissimilar. Allopatric species drums with similar cadences were used to examine whether
            cadence was a variable that encodes species identity. There was no significant difference
            between responses to allopatric hetero- versus conspecific drums with a similar cadence,
            indicating that the drum’s cadence encodes information for species recognition. Further
            playbacks eliminated drum duration and spectral properties as important variables for species
            recognition. Results suggest that if an individual drums at a similar cadence to a sympatric
            heterospecific, then their signal may elicit a conspecific response.
              Key words: drumming, interspecific communication, nonvocal communication, species
            recognition, woodpeckers.

INTRODUCTION                                                of individuals (Short 1982, Wilkins and Ritchi-
Acoustic communication in birds normally oc-                son 1999). Given that many of these functions
curs by the use of songs and calls. However,                correspond to bird song, some researchers have
woodpeckers lack the ability to generate songs              concluded that drumming should have similar
similar to those used by passerines (Bracken-               characteristics and have even postulated drum-
bury 1982). Along with a vocal repertoire of                ming to be the evolutionary counterpart to pas-
calls, woodpeckers employ a long-distance non-              serine song in woodpeckers (Pynnönen 1939,
vocal acoustical signal aptly referred to as drum-          Lawrence 1967), although others disagree
ming. A woodpecker drum is a rapid, repetitive              (Winkler and Short 1978).
series of strikes with the bill on a substrate, not            Although there are several accounts of inter-
associated with foraging or cavity excavation               specific reaction to drumming, references have
(Bent 1939, Pynnönen 1939, Short 1974).                    tended to consolidate the variety of responses as
Drumming is unusual in that a separate instru-              equivalent to responses given to conspecific
ment in addition to the bird’s bill is required to          drums (Lawrence 1967, Winkler and Short
produce the signal (Skutch 1985).                           1978). The problem with comparing woodpeck-
   Limited information is available concerning              er responses to drums stems from the lack of
nonvocal acoustical signals in avian species                information on the function of the multiple sig-
(Prum 1998). As an instrumental signal, drum-               nals woodpeckers use to communicate. There
ming has been noted to be a form of long-dis-               are three classes of calls (call notes, intimate
tance communication that may or may not elicit              calls, and rattle calls) which may be given in
hetero- or conspecific responses (Crockett 1975,            response to a drum, and it is unknown if these
Duncan 1990). Functions attributed to drum-                 calls are equivalent responses to drumming
ming include territorial establishment, mate at-            alone.
traction, pair bond maintenance, and localization              There is no consensus as to whether drums
                                                            encode species-specific information detectable
  1 Received 2 March 2000. Accepted 27 September            by receivers. Four hypotheses have been sug-
2000.                                                       gested. First, drums are species-specific and not
  2 Present address: Department of Evolution, Ecology

and Organismal Biology, The Ohio State University,
                                                            normally exchanged interspecifically (Perrins
Columbus, OH 43210.                                         and Middleton 1985, Welty and Baptista 1988).
  3 E-mail: dodenhoff.1@postbox.acs.ohio-state.edu          Second, drums are not distinctive because other
                                                       [143]
144      DANIELLE J. DODENHOFF        ET AL.

TABLE 1. Descriptive statistics of the drum variables (mean 6 SD) for the playback tapes.

                                  Cadence             Duration                             Interstrike
        Playback signal        (strikes sec21)         (sec.)         No. strikes        interval (sec.)
Nuttall’s                       20.8 6 0.4        1.03 6    0.1       21.5 6   0.7      0.050 6 0.001
‘‘Slow’’ Nuttall’sa             17.1 6 0.5        0.78 6    0.2       13.1 6   3.2      0.064 6 0.001
Downy                           15.8 6 0.4        0.79 6    0.4       12.5 6   0.7      0.066 6 0.002
Hairy                           26.6 6 0.5        0.83 6    0.3       22.0 6   7.1      0.039 6 0.001
Northern Flicker                19.3 6 1.2        0.99 6    0.4       19.3 6   7.8      0.055 6 0.005
White-headed                    20.1 6 0.5        1.64 6    0.2       33.0 6   2.0      0.048 6 0.001
Three-toed                      15.3 6 0.6        0.78 6    0.1       12.0 6   1.0      0.070 6 0.007
Black-backed                    15.6 6 0.3        1.83 6    0.1       28.5 6   0.7      0.066 6 0.002
Ladder-backed                   31.1 6 1.2        1.13 6    0.1       35.0 6   1.2      0.033 6 0.001
Pulse tone 400 Hz                  15.9               1.3                 20                0.066
Nuttall’s altered to Hairy         26.3               1.0                 26                0.041
Hairy altered to Nuttall’s         18.8               1.0                 19                0.053
 a   Drums . 2 SD below species average.

signals, perhaps rattle calls, are used for species     other species were used as playback stimuli:
identification (Short 1974, 1982, Winkler and           Black-backed (P. arcticus), Three-toed (P. tri-
Short 1978). Third, drums are fairly species-spe-       dactylus), and Ladder-backed Woodpeckers (P.
cific even to the human ear, but woodpeckers            scalaris), and the Northern Flicker (Colaptes au-
may react indiscriminately (Winkler et al. 1995).       ratus). Determinations of sympatry were based
Fourth, drums are sympatrically, but not allo-          on each species’ breeding range within sampled
patrically, species-specific, with the drum ca-         areas. White-headed Woodpeckers could be con-
dence (in strikes sec21) as the predominant var-        sidered marginally sympatric with Nuttall’s and
iable for species identification (Stark et al.          Downy Woodpeckers, but only during nonbreed-
1998).                                                  ing months (Beedy and Granholm 1985, Gaines
   Two unresolved questions remain in the afore-        1988).
mentioned hypotheses. First, whether responses             The following procedure was used for all
to hetero- versus conspecific drums are different,      playback trials: Subject birds were observed for
and whether the variable(s) used by woodpeck-           a minimum of 5 min before each trial. Only in-
ers for species recognition are contained within        dividuals that were foraging, preening, or cavity-
the drum. We designed three experiments to re-          excavating, and within acoustical range were
solve these questions. Experiment 1 established         used. Playback experiments were conducted in
a method of scoring responses to drums, and ex-         a random, balanced, pairwise design (conspecif-
periment 2 compared behavioral responses to             ic:heterospecific), with a minimum of 30 min
playbacks of sympatric hetero- versus conspe-           between stimuli (based on log survivorship
cific drums. Experiment 3 compared behavioral           curves calculated for the maximum natural du-
responses to playbacks of allopatric heterospe-         ration a bird remained responsive during drum-
cific and computer-generated drums versus con-          ming sequences, Dodenhoff 1996). Birds unre-
specific drums, to test possible variables used for     sponsive to both stimuli were excluded from the
species recognition.                                    analysis.
                                                           Drum variables for playback tapes were based
METHODS                                                 on those reported by Stark et al. (1998), and in-
GENERAL PLAYBACK DESIGN                                 cluded duration (sec), number of strikes per
Populations in four regions of California, in-          drum, cadence, and interstrike interval in one
cluding the Los Padres, Sequoia, Sierra, and            drum (interstrike interval, in seconds; Table 1).
Inyo National Forests, were sampled during the          Tapes were generated by editing together field
1993–1996 breeding seasons. Four species were           recordings and commercially available drums
studied using playbacks: Nuttall’s (Picoides nut-       (Cornell Laboratory of Ornithology 1992a,
tallii), Hairy (P. villosus), Downy (P. pubes-          1992b), computer generated tones, and altered
cens), and White-headed Woodpeckers (P. al-             natural drums reset to specific cadences. Play-
bolarvatus). In addition, the drums from four           back tapes were standardized for decibel level.
DO WOODPECKER DRUMS ENCODE INFORMATION FOR SPECIES RECOGNITION?                            145

   The results from the analysis of behavioral        servable changes from one behavior to another,
observations were used to construct a scoring         were noted for each individual during a drum-
method for comparing responses to playback            ming sequence (adapted from Halloran and Be-
signals for experiments 2 and 3. Although there       koff 1995). Behaviors that significantly preceded
are differences between calls and displays of         or followed drumming were determined by se-
woodpeckers, stereotypical behaviors are ob-          quence analysis using a first-order Markov
served within the family Picidae (Winkler et al.      chain.
1995). Therefore, generalizations from obser-            Independence between consecutive behaviors
vations of Nuttall’s Woodpecker correlated with       was tested using categorical modeling proce-
behaviors observed for other subject species. We      dures (SAS Institute 1990), which used maxi-
used these results to rank the behavioral data        mum likelihood values to estimate the G-test sta-
appropriately concerning responsiveness to            tistic and its associated x2 statistic (Sokal and
drum playbacks.                                       Rohlf 1981). Data collection excluded the pos-
   Behaviors observed more often during drum-
                                                      sibility of behaviors following themselves.
ming sequences from experiment 1 were used to
                                                      Therefore, the ‘‘expected’’ values where the pre-
create a categorical scoring method similar to
                                                      ceding and following behavior were the same
Emlen (1972) for experiments 2 and 3. Behav-
                                                      were considered ‘‘logical zeros,’’ and the ex-
iors included alarm calls, approach, drumming,
flights over the speaker, display, attack/supplant,   pected values were calculated using a log-linear
and frozen antipredation posture. These behav-        model (Fagen and Young 1978). Significant dif-
iors composed the intensity scale. The intensity      ferences were detected when the transformed
scale ranked the behavioral level achieved dur-       frequency zYz . 1.08.
ing each playback and ranged from 0–7 (low–              To investigate which behaviors occurred more
high). Given the subjective nature of categorical     often during drumming sequences versus other
scoring, we also employed a hybrid scale, which       activities, a comparison of the frequency of each
subdivided the intensity scale into differing lev-    behavior was conducted using a x2 analysis for
els of response (using frequency data), weighing      contingency tables. Twelve different Nuttall’s
each of the seven categories equally. Hybrid          Woodpeckers were observed for 30–60 min
scores ranged from 0–14 (low–high). Behavioral        each during the months of March and April,
responses were compared using two-tailed Wil-         1993, during non-drumming sequences. Behav-
coxon signed-ranks tests. For the test to be able     iors observed were classified categorically and
to detect significance at the a 5 0.05 level, n       compared to the behaviors observed during the
must be at least 6. If n 5 6, then differences had    drumming sequences.
to be of like sign (Sokal and Rohlf 1981).
                                                      EXPERIMENT 2: SYMPATRIC SPECIES
EXPERIMENT 1: BEHAVIORAL SEQUENCE                     PLAYBACKS
ANALYSIS
                                                      Behavioral responses of the four subject species
Since there were no previous measurements of
                                                      to sympatric hetero- and conspecific drums were
drum responses for these species, we conducted
                                                      compared. Trials with Nuttall’s Woodpeckers in-
a behavioral sequence analysis to categorize be-
                                                      cluded drums from three sympatric species
havioral responses to drumming. Observations
of Nuttall’s Woodpeckers during natural and in-       (Downy and Hairy Woodpeckers, and Northern
duced drumming sequences were made in the             Flickers), and trials with Downy Woodpeckers
Los Padres National Forest, San Luis Obispo           included drums from three sympatric species
County, California during the 1993–1995 breed-        (Nuttall’s and Hairy Woodpeckers, and Northern
ing seasons. Behaviors were identified according      Flickers). Northern Flicker drums used for play-
to previous descriptions (Short 1971, Winkler         back were recorded from individuals sympatric
and Short 1978) and categorized for analysis.         with Nuttall’s Woodpeckers. Hairy Woodpecker
Behaviors for this analysis included foraging,        trials were conducted using the drums from two
flying or changing locations, intimate calls,         sympatric species (Nuttall’s and Downy Wood-
preening, rattle calls, look around (with bill        peckers), whereas White-headed Woodpecker
wave), call notes, agonistic attacks, and copu-       trials consisted of drums from a single sympatric
lations. Behavioral transitions, defined as ob-       species (Hairy Woodpecker).
146    DANIELLE J. DODENHOFF        ET AL.

EXPERIMENT 3: ALLOPATRIC AND                         quences to its occurrence during non-drumming
COMPUTER GENERATED PLAYBACKS                         sequences. The behaviors ‘‘preening’’ and ‘‘rat-
Behavioral responses of three subject species to     tle call’’ were observed more often during drum-
allopatric hetero- and conspecific drums were        ming sequences (x29 5 183, P , 0.01). Rattle
compared. Trials with Nuttall’s Woodpecker in-       calls have been noted to elicit drums, and are
cluded four allopatric species drums, one with a     considered similar in function to drumming
similar cadence (White-headed Woodpecker)            (Short 1971, Winkler and Short 1978). Although
and three with different cadences (Ladder-           ‘‘look around with bill wave’’ did not always
backed, Black-backed, and Three-toed Wood-           precede or follow drumming, it was observed
peckers). Trials with White-headed Woodpeck-         more often during drumming sequences. Forag-
ers included playbacks with Nuttall’s Wood-          ing was observed during both sequences more
peckers, whereas trials with Downy Woodpeck-         than expected. Previously, call notes have been
ers included three allopatric species: two with      considered a conspecific response to drumming
similar cadences (Black-backed and Three-toed        (Short 1982, Kilham 1983). However, our anal-
Woodpecker) and one with a dissimilar cadence        ysis indicated call notes were observed frequent-
(White-headed Woodpecker). Although Black-           ly in non-drumming rather than drumming se-
backed and Three-toed Woodpeckers’ drums are         quences. Furthermore, these call notes preceded
similar in cadence to Downy Woodpeckers              foraging and were more often observed during
(Short 1982), they differ in the spacing pattern     non-drumming activities. Thus, call notes with-
of individual strikes; Black-backed and Three-
                                                     out supporting behaviors were not considered a
toed Woodpeckers increase cadence towards the
                                                     response to drumming.
end of their drum, whereas Downy Woodpeckers
decrease (Stark et al. 1998). Also, the drum of      EXPERIMENTS 2 AND 3: RESPONSE TO
Black-backed Woodpeckers is significantly lon-       PLAYBACKS
ger than Downy Woodpecker drums (Stark et al.
1998).                                               Individuals demonstrated a significant difference
   Because behavioral responsiveness to allopat-     in the level of response between sympatric het-
ric drums may be observed, computer-generated        ero- versus conspecific drums, except between
drums were used to test whether drum cadence         Nuttall’s Woodpecker and the Northern Flicker
was the primary variable responsible for species     (Table 2). Furthermore, both Downy and Nut-
recognition (as predicted by Stark et al. 1998).     tall’s Woodpeckers were similar in their respons-
Two types of trials were conducted using com-        es to the Nuttall’s playback tape generated from
puter-generated drums to test spectral properties.   an individual greater than 2 standard deviations
First, a 400-Hz tone was repeated to simulate a      below the mean cadence for Nuttall’s Wood-
drumming Downy Woodpecker. Second, a Nut-            peckers (‘‘slow’’ Nuttall’s, Stark 1996).
tall’s Woodpecker strike was set at a cadence           Responses to conspecific drums were not sig-
similar to a Hairy Woodpecker drum, and recip-       nificantly different from responses to allopatric
rocally, a Hairy Woodpecker strike was set at a      species, provided the drum cadences of the two
cadence similar to a Nuttall’s Woodpecker drum.      species were comparable. Conversely, responses
                                                     to allopatric species with dissimilar cadences
RESULTS                                              were of significantly lower intensity. The results
EXPERIMENT 1: BEHAVIORAL SEQUENCE                    for the hybrid scale were similar to those for the
ANALYSIS                                             intensity scale (Table 2), except for a marginal
Nuttall’s Woodpeckers (n 5 68 birds, with 574        difference in the Hairy versus Ladder-backed
transitions) were observed for the behavioral se-    Woodpecker comparison; there was a significant
quence analysis. The sample size was large           difference for the intensity scale but a nonsig-
enough for the x2 distribution to approximate the    nificant difference for the hybrid scale (Table 2).
G distribution. Results of the categorical model     Hairy Woodpeckers responded to Ladder-backed
implied a strong dependency between behaviors        Woodpecker drums, but less intensely than to a
(x229 5 190, P , 0.01). Twelve Nuttall’s Wood-       conspecific signal (although the hybrid scale
peckers were used to analyze behavioral re-          score approached significance). The results from
sponses to drums by comparing each behavior’s        playbacks with computer-generated signals in-
frequency of occurrence during drumming se-          dicated no significant difference between the re-
DO WOODPECKER DRUMS ENCODE INFORMATION FOR SPECIES RECOGNITION?                                     147

TABLE 2. Wilcoxon signed-ranks test results for comparisons of responses to sympatric, allopatric, and com-
puter-generated drum signals versus the conspecific signal. Nonsignificant results indicate no significant differ-
ences in responses to conspecific and heterospecific signals. Listed are P-values (n birds). Int 5 Intensity scale;
Hyb 5 Hybrid scale; see Methods for a description of the two scales.

                                                                 Species
      Playback signal       Scale          Nuttall’s           Downy               Hairy          White-headed
Nuttall’s                    Int                              0.01   (8)         0.04 (9)           0.58 (10)
                             Hyb                              0.01   (8)         0.02 (9)           0.61 (10)
‘‘Slow’’ Nuttall’sa          Int           0.23   (9)         1.00   (9)
                             Hyb           0.17   (9)         0.83   (9)
Downy                        Int           0.01   (29)                           0.01 (11)
                             Hyb           0.01   (29)                           0.01 (11)
Hairy                        Int           0.01   (20)        0.01   (11)                           0.02 (7)
                             Hyb           0.01   (20)        0.01   (11)                           0.02 (7)
Northern Flicker             Int           0.16   (18)        0.04   (6)
                             Hyb           0.29   (18)        0.04   (6)
White-headed                 Int           0.40   (19)        0.02   (8)
                             Hyb           0.08   (19)        0.02   (8)
Three-toed                   Int           0.03   (11)        0.18   (13)
                             Hyb           0.01   (11)        0.23   (13)
Black-backed                 Int           0.03   (13)        1.00   (12)
                             Hyb           0.01   (13)        0.55   (12)
Ladder-backed                Int           0.04   (6)                            0.04 (13)
                             Hyb           0.04   (6)                            0.06 (13)
Pulse tone 400 Hz            Int                              0.06 (7)
                             Hyb                              0.06 (7)
Nuttall’s altered to         Int                                                 1.00 (7)
Hairy                        Hyb                                                 0.89 (7)
Hairy altered to Nut-        Int           0.28 (8)
tall’s                       Hyb           0.25 (8)

  a   Drums more than 2 SD below species average.

sponses of woodpeckers to these signals versus             These results differ from Duncan (1990), be-
conspecific drums.                                         cause Duncan’s comparisons did not include re-
                                                           ciprocal playbacks with species sharing similar
DISCUSSION                                                 cadences.
Although most researchers agree drums have                    Our results between species that share similar
species-typical characteristics (Short 1982, Kil-          cadences indicated interspecific responses may
ham 1983, Winkler et al. 1995), it was unknown             occur between sympatric species. The responses
whether observed interspecific responses to                by Nuttall’s Woodpeckers to the drums of North-
drums were predictable. Previous studies includ-           ern Flickers were similar to responses elicited
ed multiple responses as equivalent indications            by a conspecific drum. Thus, interspecific re-
of species recognition (Short 1982, Winkler et             sponses may occur between these sympatric spe-
al. 1995), which implied ambiguity within the              cies. Individual drums of both species often have
drum, but did not indicate the absence of dis-             a similar drum cadence, and were predicted to
crimination. Individuals may respond to a het-             have reciprocal heterospecific ambiguity in their
erospecific drum, but those responses can differ           drum (Stark et al. 1998). Despite this prediction,
from responses to conspecific drums, which                 we did not observe interspecific responses to
could indicate discrimination between stimuli.             natural (i.e., non-playback) drums between in-
Results from the playbacks of the ‘‘slow’’ Nut-            dividuals of these species over the course of this
tall’s stimulus to Nuttall’s and Downy Wood-               study. We attributed the lack of interspecific re-
peckers indicated that interspecific behavioral            sponse to observations of restricted drumming
responses to drums can occur naturally between             by Northern Flickers. Northern Flickers within
sympatric species, if the drum cadences overlap.           the study areas drummed for a few weeks during
148    DANIELLE J. DODENHOFF        ET AL.

the breeding season, usually when Nuttall’s           species identity. However, these results cannot
Woodpeckers were unresponsive to drums. This          exclude the possibility that either interstrike in-
may indicate a shift between species in the tim-      terval or duration encodes species identity for
ing of drumming during the breeding season;           Downy Woodpeckers that are sympatric with
species with similar drums could reduce ambi-         Black-backed or Three-toed Woodpeckers.
guity through temporal separation and maintain           Computer-generated drums tested whether
species distinctiveness (Stacier et al. 1996).        spectral properties were important in species
   Allopatric playbacks followed the same pat-        identification. Results from trials with Nuttall’s,
tern as observed in sympatric playbacks: drums        Hairy, and Downy Woodpeckers indicated that
of species with similar cadences elicited behav-      spectral properties of a drum were not charac-
ioral responses equivalent to conspecific signals,    teristics responsible for species recognition. In-
whereas those with dissimilar cadences did not.       stead, the spectral properties generated in drum-
This indicates that drum cadence is one variable      ming have been noted to depend on the drum-
that encodes species identity. Playback trials        ming substrate selected (Eberhardt 1997). Fur-
with Downy Woodpeckers tested whether the             thermore, differential signal attenuation of
variables ‘‘interstrike interval’’ and ‘‘duration’’   frequencies through an environment may make
were used to encode species identity. Downy           the spectral properties of a drum a poor variable
Woodpeckers responded similarly to conspecific,       for species-specificity and signaler identification
Black-backed and Three-toed Woodpecker                (Wiley and Richards 1982). However, these
drums, which indicated that drum duration and         spectral properties may be important in other as-
interstrike interval were not used by Downy           pects of communication, including ranging
Woodpeckers to differentiate between these spe-       (Morton 1986, Naugib 1995, 1998).
cies. This is similar to the finding that, for many      In summary, our results support the conclu-
passerines, signal duration is usually not a spe-     sion that the drum’s cadence encodes species
cies-typical variable (Becker 1982). However,
                                                      identity of the signaler, and that receivers per-
because heterospecific drums used as playbacks
                                                      ceive these differences within the acoustical en-
were no shorter than Downy Woodpecker
                                                      vironment. Individual woodpeckers did not react
drums, our test does not eliminate the possibility
                                                      indiscriminately to drums. Responses to heter-
of a minimum duration required for species rec-
                                                      ospecific drums with dissimilar cadences were
ognition. Shorter song durations may elicit less
                                                      less intense and did not include vocalizations or
intense responses, and minimum durations are
                                                      behaviors observed during playbacks with con-
required for species recognition for passerines
                                                      specific drums. Sympatric woodpecker drums
(Becker 1982).
   Our results also eliminated the interstrike in-    are not usually exchanged interspecifically, but
terval differences in drumming as a character-        interspecific responses may result if individuals
istic responsible for species recognition in          from different species have an overlapping drum
Downy Woodpeckers. Kaiser (1990) suggested            cadence. However, due to the amount of overlap
that the spacing pattern within unmodulated           between species, both sympatric and allopatric,
drums was related to the woodpecker’s foraging        we conclude drums are not species specific. The
method. In ‘‘wood-pecking’’ species the rhythm        lack of signal specificity does not prevent the
accelerates, whereas ‘‘gleaners’’ slow down to-       encoding of species identity. For example, dif-
ward the end or the rhythm and amplitude of           ferential timing of drumming between species
their drums remain uniform (Winkler et al.            within a breeding season could result in a lack
1995). Although Kaiser’s hypothesis was based         of selection pressure on drums to significantly
on small sample sizes, it suggested that the spac-    diverge. This may allow a few sympatric species
ing pattern within a drum may be controlled by        to utilize similar cadences to encode species
the musculature used for foraging. Thus, this         identity without resulting in ambiguity between
variable would have a greater likelihood of over-     species. Nonetheless, given the similarities be-
lapping with a sympatric species with a similar       tween allopatric species, it is unknown whether
foraging strategy. Given the high degree of over-     the trend of signal divergence observed in this
lap between the foraging strategies of many           study is applicable across broad geographic
sympatric woodpeckers, the interstrike interval       ranges or under differing acoustical environ-
of a drum may not be a reliable indicator of          ments. Clearly, considerable research still needs
DO WOODPECKER DRUMS ENCODE INFORMATION FOR SPECIES RECOGNITION?                                       149

to be done in the area of avian mechanical                     quences, p. 79–113. In P. Colgan [ED.], Quantita-
acoustical signals, including woodpecker drums.                tive ethology. John Wiley and Sons, New York.
                                                           GAINES, D. 1988. Birds of Yosemite and the east slope.
                                                               Artemisia Press, Lee Vining, CA.
ACKNOWLEDGMENTS                                            HALLORAN, M., AND M. BEKOFF. 1995. Cheek rubbing
We thank D. Frey, R. Gambs, W. Koenig, D. Nelson,              as grooming by Abert squirrels. Anim. Behav. 50:
and two anonymous reviewers for helpful comments               987–993.
which significantly improved this manuscript; J. Daly,     KAISER, M. 1990. Untersuchungen zur Biomechanik
R. Schmidt, and R. Peck for assistance with the statis-        des Balztrommelns der Specte (Picidae). Beitr.
tics; and the Biological Sciences Department and De-           Vogelkd. 36:129–159.
partment of Statistics, California Polytechnic State       KILHAM, L. 1983. Woodpeckers of eastern North
University, San Luis Obispo for use of equipment, fa-          America. Dover, New York.
cilities, and for financial support of this research. We   LAWRENCE, L. DE K. 1967. A comparative life-history
especially thank R. Bowker for technical assistance            study of four species of woodpeckers. Ornithol.
and sound-analysis software.                                   Monogr. 5.
                                                           MORTON, E. S. 1986. Predictions from the ranging hy-
                                                               pothesis for the evolution of long-range commu-
LITERATURE CITED
                                                               nication. Behavior 99:65–86.
BECKER, P. 1982. The coding of species-specific char-      NAUGIB, M. 1995. Auditory distance assessment of
    acteristics in bird sounds, p. 214–244. In D.              singing conspecifics in Carolina wrens: the role of
    Kroodsma and E. Miller [EDS.], Acoustic com-               reverberation and frequency-dependent attenua-
    munication in birds. Vol. 1. Academic Press, New           tion. Anim. Behav. 50:1297–1307.
    York.                                                  NAUGIB, M. 1998. Perception and degradation in
BEEDY, E., AND S. GRANHOLM. 1985. Discovering Si-              acoustical signals and its implications for ranging.
    erra birds: western slope. Yosemite Nat. Hist. As-         Behav. Ecol. Sociobiol. 42:139–142.
    soc. and Seqouia Nat. Hist. Assoc. Mariposa, CA.       PERRINS, C. M., AND A. L. A. MIDDLETON. 1985. The
BENT, A. C. 1939. Life histories of North American             encyclopedia of birds. Oxford Univ. Press, New
    woodpeckers. U.S. Natl. Mus. Bull. 174.                    York.
BISHOP, Y. M. M., S. E. FIENBURG, AND P. W. HOLLAND.       PRUM, R. O. 1998. Sexual selection and the evolution
    1975. Discrete multivariate analysis: theory and           of mechanical sound production in manakins
    practice. M.I.T. Press, Cambridge, MA.                     (Aves: Pipridae). Anim. Behav. 55:977–994.
BRACKENBURY, J. 1982. The structural basis of voice        PYNNÖNEN, A. 1939. Beitrage zur Kenntnis der Biol-
    production and its relationship to sound charac-           ogie finnischer Spechte. Ann. Zool. Soc. Zool.
    teristics, p. 53–73. In D. Kroodsma and E. Miller          Bot. Fenn. Vanamo 7:1–166.
    [EDS.], Acoustic communication in birds. Vol. 1.       SAS INSTITUTE INC. 1990. SAS user’s manual. Version
    Academic Press, New York.                                  6, 4th ed. SAS Institute Inc., Cary, NC.
CORNELL LABORATORY OF ORNITHOLOGY. 1992a. Inter-           SHORT, L. 1971. Systematics and behavior of some
    active audio. A field guide to bird songs, Eastern         North American woodpeckers, genus Picoides
    and Central edition. 2nd ed. Houghton Mifflin,             (Aves). Bull. Am. Mus. Nat. Hist. 145:1–118.
    Boston.                                                SHORT, L. 1974. Habits and interactions of North
CORNELL LABORATORY OF ORNITHOLOGY. 1992b. Inter-               American Three-toed Woodpeckers (Picoides arc-
    active audio. A field guide to bird songs, Western         ticus and Picoides tridactylus). Am. Mus. Novi-
    edition. 2nd ed. Houghton Mifflin, Boston.                 tates 2547:1–42.
CROCKETT, A. B. 1975. Ecology and behavior of the
                                                           SHORT, L. 1982. Woodpeckers of the world. Delaware
    Williamson’s Sapsucker in Colorado. Ph.D. diss.,
                                                               Mus. Nat. Hist., Greenville, DE.
    Univ. Colorado, Boulder, CO.
DODENHOFF, D. J. 1996. Interspecific and intraspecific     SKUTCH, A. F. 1985. Life of the woodpecker. Ibis Publ.,
    communication: a quantitative analysis of drum-            Santa Monica, CA.
    ming behavior using four species of California-        SOKAL, R. R., AND F. J. ROHLF. 1981. Biometry. W. H.
    occurring woodpeckers (Family Picidae). M.Sc.              Freeman, San Francisco.
    thesis, California Polytechnic St. Univ., San Luis     STACIER, C. A., D. A. SPECTOR, AND A. G. HORN. 1996.
    Obispo, CA.                                                The dawn chorus and other diel patterns in acous-
DUNCAN, S. 1990. Auditory communication in breed-              tical signaling, p. 339–355. In D. Kroodsma and
    ing Northern Flickers (Colaptes auratus). Ph.D.            E. Miller [EDS.], Ecology and evolution of acous-
    diss., Univ. Wisconsin, Milwaukee, WI.                     tic communication in birds. Cornell Univ. Press,
EBERHARDT, L. S. 1997. A test of an environmental              Ithaca, NY.
    advertisement hypothesis for the function of           STARK, R. D. 1996. Minimizing signal equivocation: a
    drumming in Yellow-bellied Sapsuckers. Condor              quantitative analysis of the drumming patterns of
    99:798–803.                                                eleven species of California occurring woodpeck-
EMLEN, S. T. 1972. An experimental analysis of the             ers: (Family Picidae). M.Sc. thesis, California
    parameters of bird song eliciting species recogni-         Polytechnic State Univ., San Luis Obispo, CA.
    tion. Behaviour 41:130–171.                            STARK, R. D., D. J. DODENHOFF, AND E. V. JOHNSON.
FAGEN, R., AND D. YOUNG. 1978. Temporal patterns of            1998. A quantitative analysis of woodpecker
    behavior: durations, intervals, latencies, and se-         drumming. Condor 100:350–356.
150     DANIELLE J. DODENHOFF          ET AL.

WELTY, J. C., AND L. BAPTISTA. 1988. The life of birds.      tion and possible causation. J. Field Ornithol. 70:
    Harcourt Brace Jovanovich, Orlando, FL.                  578–586.
WILEY, R. H., AND D. G. RICHARDS. 1982. Adaptation for    WINKLER, H., AND L. SHORT. 1978. A comparative
    acoustic communication in birds: sound transmission      analysis of acoustical signals in pied woodpeckers
    and signal detection, p. 132–183. In D. Kroodsma         (Aves, Picoides). Bull. Amer. Mus. Nat. Hist. 160:
    and E. H. Miller [EDS.], Acoustic communication in       1–109.
    birds. Vol. 1. Academic Press, New York.              WINKLER, H., D. CHRISTIE, AND D. NURNEY. 1995.
WILKINS, H. D., AND G. RITCHISON. 1999. Drumming             Woodpeckers: an identification guide to the wood-
    and tapping by Red-bellied Woodpeckers: descrip-         peckers of the world. Houghton Mifflin, Boston.
You can also read