Behavioral Ecology Vol. 13 No. 6: 844-850
© 2002 International Society for Behavioral Ecology
Habitat structure and alarm call dialects in Gunnison's prairie dog (Cynomys gunnisoni)
a Grand Canyon Wildlands Council, PO Box 1594, Flagstaff, AZ 86002, USA b Northern Arizona University, Department of Biological Sciences, Flagstaff, AZ 86011-5650, USA
Address correspondence to C.N. Slobodchikoff. E-mail: con.slobodchikoff{at}nau.edu.
Received 4 September 2000; revised 10 March 2002; accepted 14 April 2002.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
We examined the relationship between habitat structure and alarm call characteristics in six colonies of Gunnison's prairie dogs (Cynomys gunnisoni) near Flagstaff, Arizona, before and after a mid-summer vegetation change. We found significant differences in alarm call characteristics between colonies, confirming the existence of alarm call dialects. Differences in frequency components but not temporal components of calls were associated with differences in habitat structure. Playback experiments revealed that differences in alarm call structure affected acoustic transmission of calls through the local habitat. Thus, we identify habitat structure as one factor that may contribute to alarm call differences between colonies of Gunnison's prairie dogs. Relationships between call characteristics and habitat structure changed over seasons. Playback experiments suggested that this changing relationship could reflect a change in the purpose of the alarm call between early and late summer. Some components of alarm calls seem tailored for attenuation over short distances in the early summer but for long-distance transmission at summer's end. These differences might arise because pups stay close to their natal burrows in the early summer and disperse throughout a colony in late summer. Alternatively, these differences in alarm call transmission between seasons could be caused by the increase in vegetation in the mid-summer. At the end of the summer prairie dogs could be more dependent on long-distance antipredator calls to offset the loss of visibility caused by the increase in vegetation in the late summer.
Key words: alarm calls, communication, C. gunnisoni, dialects, habitat structure, prairie dogs.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Dialects are variations in vocal signals between different populations of the same species. Dialects are well known among a wide variety of organisms from white-crowned sparrows (Marler and Tamura, 1962
) to
elephant seals (LeBouf and Peterson,
1969). Although there is evidence that dialects exist in species,
there is less certainty about what factors contribute to the formation and
persistence of dialects (Date and Lemon,
1993; Ryan and Brenowitz,
1985).
A combination of genetic, cultural, and environmental factors may
contribute to dialect formation in most species. If sufficient genetic
isolation exists between populations or if acoustic signals involve much
learning from one individual to the next, genetic and cultural factors can
have a strong impact on the formation of dialects
(Baker, 1975;
Handford and Lougheed, 1991;
Somers, 1973). If different
populations of the same species live in dissimilar habitats, then habitat
structure can have a strong influence on dialect formation
(Date and Lemon, 1993;
Morton, 1986;
Rothstein and Fleischer, 1987;
Tubaro and Segura, 1995).
Morton (1975), Hansen
(1979), and Marten and Marler
(1977) first showed that
habitat structure affects sound transmission, which in turn influences the
evolution of acoustical signals in animals. The local adaptation hypothesis
(Morton, 1975), the ranging
hypothesis (Morton, 1986), and
the closely related acoustical adaptation hypothesis
(Hansen, 1979;
Rothstein and Fleischer, 1987)
were developed from these observations. The local adaptation hypothesis
proposes the evolution of acoustic signals that optimize sound transmission
over maximum distances (Morton,
1975). The ranging hypothesis and the acoustical adaptation
hypotheses propose that some calls are selected for long-distance
transmission, while other calls may be structured to predictably degrade for
the purpose of establishing territory boundaries or to reduce the chance of
attracting predators or competitors (Brown
and Handford, 1996; Marler,
1955; Morton,
1986). All of these hypotheses stem from the common idea that
local acoustic habitats are highly variable due to the interplay of
atmospheric, vegetative, and ground effects. This variability causes
differences in sound transmission and ultimately drives selection for unique
acoustic dialects.
The acoustic environment affects vocal signals through attenuation and
distortion. Animals respond to the distortion and attenuation present in the
environment by changing temporal and frequency components of acoustic signals.
For example, distortion of sound is related to the length of the sound. So, in
less vegetated habitats, acoustic signalers often use short pulses with higher
rates of repetition to avoid distortion by atmospheric effects, whereas long
whistles are used in highly vegetated environments to avoid reverberation off
vegetation (Handford and Lougheed,
1991).
Attenuation of sound is strongly dependent on frequency
(Morton, 1975). Generally,
frequencies between 1 and 4 kHz travel farthest and with most consistency in
any environment (Marten and Marler,
1977). However, acoustic signalers use a much larger range of
frequencies than 1-4 kHz. Ultimately, frequencies that are optimum for
different habitats depend on the purpose of the signal, the balance between
vegetative cover and atmospheric turbulence, and the height above the ground
at which a signal is transmitted (Linskens
et al., 1976; Marten and
Marler, 1977; Shy,
1983; Tubaro and Segura,
1995; Waas, 1988; Wasserman,
1979; Wiley and Richards,
1978).
In this study we explored the effects of habitat structure on acoustic
dialect formation in Gunnison's prairie dogs (Cynomys gunnisoni).
Gunnison's prairie dogs are colonial, ground-dwelling rodents found throughout
the grasslands of the Colorado Plateau in western North America
(Hall and Kelson, 1959).
Prairie dogs use alarm calls to detect and avoid predators, and the presence
of nepotism in alarm calling has been documented in both black-tailed prairie
dogs, C. ludovicianus, and C. gunnisoni (Hoogland,
1995,
1996). Alarm call dialects of
Gunnison's prairie dogs exist on both regional and local scales
(Slobodchikoff and Coast,
1980; Slobodchikoff et al.,
1998).
Although dialects are present between colonies of Gunnison's prairie dogs,
it is unclear what contributes to this intraspecific variation. On a regional
scale, colonies that are farther apart contain greater differences in call
characteristics, suggesting that genetic isolation contributes to differences
in alarm call structure (Slobodchikoff et
al., 1998; Travis et al.,
1997). However, on a local scale there is no relationship between
distance between colonies and differences in alarm calls, suggesting that
other factors are influencing local dialect formation
(Slobodchikoff and Coast,
1980; Slobodchikoff et al.,
1998).
Habitat structure can be extremely variable between prairie dog towns in
the vicinity of Flagstaff, Arizona, USA, which may cause differences in alarm
call dialects between colonies
(Slobodchikoff et al., 1988).
Furthermore, because of the existence of late summer monsoons, vegetation on
the same colony can increase drastically from early to late summer, which may
cause calls within the same colony to change between seasons. We hypothesized
that habitat structure influences variations in alarm call structure in
Gunnison's prairie dogs on a local scale. Specifically, we predicted that
differences in vegetation cover influence both temporal and frequency
characteristics of alarm calls in Gunnison's prairie dogs.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
We studied six colonies of Gunnison's prairie dogs (C. gunnisoni) located near the city of Flagstaff, Arizona. All colonies were located at the same elevation (2100 m) within alpine meadows surrounded by a zone of ponderosa pine (Pinus ponderosa). To determine whether differences in vegetative cover explained differences in alarm call dialects between colonies of Gunnison's prairie dogs, we conducted habitat analysis, recording, and playback sessions on all six colonies between May and September 1999.
Northern Arizona has a heavy monsoon season in late summer, which results in a dramatic increase in vegetation on most colonies. To establish whether prairie dogs change call characteristics in response to seasonal changes in vegetative cover, we conducted two sets of habitat analysis, recording, and playback sessions (premonsoon: 15 May-30 June 1999 and postmonsoon: 1 August-15 September 1999).
Variation in alarm calls
To determine if there were dialect differences between colonies, we
recorded and analyzed alarm calls of 10-12 different prairie dogs from each
colony in premonsoon season (n = 65) and alarm calls from 10-12
different prairie dogs from each colony in postmonsoon season (n =
60) for a total of 125 alarm calls from different adult prairie dogs in both
seasons. Recording sessions always took place between 0630 and 0930 h to
minimize variation in atmospheric conditions. Calls were recorded on
high-fidelity audiotape using a Sennheiser ME-88 directional microphone
(frequency response: 5-15 kHz sensitivity: 5mV/Pa) connected to a Sony
TC-D5PRO II cassette recorder (frequency response: 40-15 kHz).
To minimize the possibility of recording the same prairie dog twice, we marked the burrow where the prairie dog was calling from with a plastic flag, and no further alarm calls were taken from that burrow in premonsoon or postmonsoon seasons. Although marking all prairie dogs in all six towns would have been the optimum method, we were constrained by time. We had to finish the first recording session before the rains of the monsoon season started, and the sizes of the study colonies (>100 individuals each) prohibited us from trapping and marking all prairie dogs in time. The large size of these colonies likely served to decrease the probability of sampling the same prairie dog twice.
All alarm calls were elicited by the same human adult female, dressed in
the same clothes. Humans have been regularly used in alarm call studies of
Gunnison's prairie dogs because approaches can be standardized more
effectively than if wild or captive nonhuman predators were used
(Slobodchikoff et al., 1991).
Humans have hunted prairie dogs in the Flagstaff area for hundreds of years,
and prairie dogs regularly alarm call in response to all humans as they would
to any other predator (Slobodchikoff et
al., 1991). Hopi and Navajo tribes near Flagstaff have culinary
recipes for prairie dogs (Gorman,
1981), and white settlers have treated prairie dogs as pests and
intensively eradicated them by hunting and poisoning for more than 150 years
(McNulty, 1971). Sportsmen in
the area regularly shoot prairie dogs to this day. The mean generation time
for Gunnison's prairie dogs is 1.5 years, which means that at least 100
generations of prairie dogs (not including those hunted by early native
Americans) have had contact with humans that kill them
(Rayor, 1985). Thus, we
believe that a human subject is sufficient to elicit real alarm calls from
prairie dogs.
We analyzed recorded alarm calls using the RTS real-time spectogram computer package (version 3.0; Engineering Design, Belmont, Massachusetts). Frequency resolution for each spectograph was set at 48.8 Hz, and sample rate was set at 25 kHz. We analyzed six different call variables for each alarm call. Three of these six call variables were temporal variables (syllables per bout, syllable length, and interval length). A bout was defined as a series of one or more calls followed by at least 3 s of silence by the individual animal (Waring, 1970). Three frequency variables were also measured (maximum, dominant and fundamental frequencies). Maximum, dominant, and fundamental frequencies were defined as the highest, mid-range, and lowest frequencies of the call, respectively, that had higher sound intensities than the surrounding harmonic frequency bands.
We analyzed 1-s time intervals of each call sequence using RTS. Time and frequency coordinates were digitized from points on each call and were used to calculate the above six variables (Figure 1). Fundamental, dominant, and maximum frequencies all have higher amplitudes than the harmonic frequencies existing in between the main frequencies. On the RTS color display the fundamental, dominant, and maximum frequencies were easily identified by their dark red or orange color, as opposed to the lighter yellow color of the other harmonics. In cases where the difference in intensity was difficult to differentiate from the color spectograph, we converted the call to a display graph that showed amplitude versus frequency and identified the main frequencies as those with the highest amplitudes on the graph. We used ANOVA to determine if calls differed between colonies in the pre- and postmonsoon seasons as well as between the same colony in different seasons.
|
Relating habitat structure and alarm call differences
To quantify differences in vegetative cover, we recorded percent vegetative
cover using 40 plots at each colony. Each plot was 1 m2. The plots
were further divided into 100 10 x 10 cm squares to aid in accurately
estimating percent cover. To test for differences in vegetative cover between
colonies in the same season, we used ANOVA. To determine differences in
vegetation within the same colony between seasons, we performed a
repeated-measures ANOVA.
To test whether alarm call differences were related to differences in
vegetative cover, we ran a regression between alarm call characteristics and
percent vegetation cover (Bonferroni adjustments made: premonsoon: 
/4 =
0.013, postmonsoon: /3 = 0.017). We used alarm call characteristics
that differed between colonies in the regression analysis. We used mean values
for call characteristics and habitat cover for each colony.
To determine if differences in call characteristics affected the transmission of a call through different environments, we conducted playback experiments. Because frequency components of calls were the only call components related to differences in vegetative cover, we only tested frequency components in the playback experiments. We did this by measuring the attenuation of frequency components of calls through different amounts of vegetative cover.
The playback tape we used consisted of a 4-kHz pure tone and also a representative call from each of the six study colonies. One playback tape was made for each season and played back on all six colonies. Calls belonging to the colony where they were played back were labeled reference calls. Calls not belonging to the colony at which they were played back were labeled foreign calls. If calls played on their home territory outperformed foreign calls and if calls played on their home territory transmitted better than when they were played on foreign territories, this would provide preliminary evidence that calls may be adapted for a specific environment and purpose.
We conducted playbacks on a 100-m transect from the burrow of the
representative caller for that colony. The recording tape recorder was placed
at 1, 5, 20, 40, 60, 80, and 100 m along the transect line. The broadcasting
tape recorder (Sony TC-D5PRO II) was attached to a Realistic MPA-25, 20-watt
mobile amplifier and loudspeaker (frequency response: 275-14 kHz). This
broadcasting system was placed at the caller's burrow at a height of 25 cm
(the height of a calling prairie dog) and in the direction the caller had been
facing (Slobodchikoff and Coast,
1980). The recording microphone was set at a height of 10 cm (the
height of a listening prairie dog;
Slobodchikoff and Coast,
1980). For consistency, we used the same recording system in the
sound transmission experiments that was used to record the original calls.
We standardized the sound decibel level of the playbacks between colonies
by calibrating the system to 45 dB, 10 m away from the source. This
calibration was used to mimic the actual sound level of an alarm call
(Slobodchikoff and Coast,
1980). The 4-kHz pure tone was synthesized using an NCH tone
generator (NCH Audio Action Software, 1999 version) and was used as a control
to represent the average dominant frequency of prairie dog alarm calls.
We analyzed playback recordings using Signal, a computer sound analysis
program (version 3.0; Engineering Design). We measured attenuation of
fundamental frequencies, dominant frequencies, maximum frequencies, and the
pure tone on a power spectrum (dB-volts) and converted to absolute dB SPL.
Excess attenuation (EA) was then calculated for the 4-kHz pure tone and for
each frequency component for all reference calls and foreign calls by
subtracting attenuation due to spherical spread: 20 log(far distance/near
distance) from original attenuation figures
(Marten and Marler, 1977). EA
in decibels at 100 m was used on dominant frequency components of the alarm
call, but because other frequencies did not travel the entire 100 m, we
calculated EA over shorter distances for maximum frequencies and fundamental
frequencies (20 m for maximum and 60 for fundamental frequencies).
We used only one representative call from each colony for each season for
the playback tape. Although randomly picking a representative call is a common
practice (Hurlbert, 1984),
there may be issues with pseudoreplication using this method
(Kroodsma, 1989;
Searcy, 1989). We dealt with
this problem by using a blocked design for our ANOVA. This more conservative
ANOVA is designed to handle repeated sampling methods
(Sokal and Rohlf, 1995).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Variation in alarm calls
Alarm call dialects existed between colonies in both seasons. In the premonsoon season maximum frequency, dominant frequency, fundamental frequency, and syllable length were significantly different between colonies (ANOVA, df = 5, 69; maximum frequency: p < .0001, F = 45.68; dominant frequency: p < .0001, F = 13.836; fundamental frequency: p < .0001, F = 9.06, syllable length: p < .0001, F = 7.77: interval length and syllables per bout: p > .056 for both, > 2.29 for both). In the postmonsoon season maximum frequency, dominant frequency, and fundamental frequency differed between colonies, but no temporal components were significantly different (ANOVA df = 5, 64; maximum frequency: p < .0001, F = 5.80; dominant frequency: p < .0001, F = 5.22; fundamental frequency: p < .0001, F = 6.57; syllable length, interval length, syllables per bout: p > .09 for all, F < 2.02 for all).
There were significant differences in maximum and dominant frequencies within the same colony, between seasons (ANOVA, season effect: p < .0001, F = 52.67, df = 1,6: maximum frequency: F = 93.97, p < .0001, df = 1,6; dominant frequency: F = 18.33, p < .0001, df = 1,6). However, no other call components changed within the same colony between seasons (ANOVA, season effect: fundamental frequency: F = .892, p = .347, df = 1,6; syllable length, interval length, syllables per bout: F < .92, p > .34, df = 1,6 for all).
Relating habitat structure and alarm call differences
Habitat structure was different both within and between our study colonies.
There were significant differences in vegetative cover between colonies
throughout the study period (ANOVA, premonsoon: p < .0001,
F = 80.31, df = 5,239; postmonsoon: p < .0001, F
= 67.5, df = 5,239). In addition, percent cover changed significantly within
the same colony between premonsoon season and postmonsoon season
(repeated-measures ANOVA by season: p < .0001, F =
103.05, df = 5,234).
If the observed differences in habitat structure on our study colonies
affect dialect differences between colonies, we would expect that colonies
with more similar vegetation structures would have more similar alarm calls.
In the premonsoon season, dominant frequencies, fundamental frequencies,
maximum frequencies, and syllable lengths differed between colonies. However,
only maximum frequency was influenced by differences in habitat structure
(regression, Bonferroni corrected = .013:
R2adj = .737, p = .012, F =
19.07, df = 5; Figure 2a).
Dominant frequency, fundamental frequency, and syllable length were not
significantly related to habitat differences (regression, Bonferroni corrected
= .013, df = 5 for all: dominant frequency:
R2adj = .746, p = .017, F =
15.71, Figure 2b; fundamental
frequency: R2adj = .006, p = .367,
F = 1.03; syllable length: R2adj =
-.230, p = .811, F = .065). Thus, observed differences in
vegetation cover explain some but not all of the dialect differences that we
observed in the premonsoon season.
|
It is interesting that the relationship between vegetation cover and alarm
call characteristics changes with season. In the postmonsoon season, dominant
frequency and fundamental frequency of alarm calls were significantly
associated with vegetative cover (regression, Bonferroni corrected =
.017: dominant frequency: R2adj = .762,
p = .015, F = 17.00, df = 5; fundamental frequency:
R2adj = .831, p = .007, F =
25.58, df = 5; Figure 2c,d).
However, variation in maximum frequency was not explained by habitat in
postmonsoon season (regression, Bonferroni corrected = .017:
R2adj = .569, p = .051, F =
7.61, df = 5).
Playback experiments revealed that statistically significant relationships
between call characteristics and habitat cover translated into real
differences in transmission performance of calls through the environment. In
the premonsoon season, there was no difference in the excess attenuation of
reference calls and foreign calls for fundamental and dominant frequencies,
but the EA of maximum frequency was significantly greater in reference calls
compared to foreign calls (ANOVA randomized complete block design, Bonferroni
corrected = .017: maximum frequency: F = 6.27, p =
.016, df = 5,24; fundamental frequency, dominant frequency, p >
.05; Figure 3a). Unexpectedly,
this means that maximum frequencies did not travel as far on home colonies as
they did on foreign colonies in the premonsoon season.
|
Similar to premonsoon season, those call characteristics that were related
to habitat variation in postmonsoon season also showed significant differences
in actual transmission performance through the environment. However,
postmonsoon frequencies of reference calls attenuated less (traveled farther)
than foreign calls. Both fundamental and dominant frequencies in reference
calls experienced significantly less EA than foreign calls (ANOVA randomized
complete block design, Bonferroni corrected = .017: dominant
frequency: F = 7.14, p = .008, df = 5,24; fundamental
frequency: F = 7.59, p = .011, df = 5,24; maximum frequency:
F = .097, p = .758, df = 5,24;
Figure 3b).
In summary, there were differences in alarm calls both between colonies and within the same colony over seasons. Both seasons revealed relationships between habitat and frequency aspects of calls; however, the relationship changed with season. Finally, the statistically significant relationship between habitat structure and call frequencies was supported by differences in actual transmission performance of those frequencies through different environments.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The variation we uncovered in alarm calls between colonies supports the hypothesis that alarm call dialects exist in Gunnison's prairie dogs (Slobodchikoff and Coast, 1980
; Slobodchikoff et al.,
1998). To be classified as dialect differences, alarm calls within
a colony would need to be consistently more similar to each other than to
alarm calls from different colonies. The alarm calls of Gunnison's prairie
dogs did change within a colony through seasons. However, despite these
internal changes, alarm calls within the same colony were consistently more
similar to each other than they were to alarm calls of prairie dogs in other
colonies.
Our study suggests that dialects in Gunnison's prairie dogs are at least
partially influenced by differences in vegetative cover. This is consistent
with other studies that have found relationships between habitat cover and
structure of acoustic signals (Date and
Lemon, 1993; Handford and
Lougheed, 1991; Slobodchikoff
and Coast, 1980; Waas, 1988;
Wasserman, 1979). However,
although habitat structure does appear to influence dialects, it is clearly
not the only factor constraining dialect formation and persistence.
Differences in temporal aspects of alarm calls were not associated with
habitat differences. Other forces that regularly drive dialect selection, such
as genetic isolation (Baker,
1975) or cultural influences
(Handford and Lougheed, 1991),
could influence temporal components of alarm calls. For example, both
frequency and temporal components of dialects in pikas are highly influenced
by genetic isolation (Somers,
1973). In addition, other measures of habitat structure that we
did not evaluate in this study could affect call components. For example, stem
diameter of vegetation, wind patterns, and humidity may heavily influence
components of acoustic signals in open grassland environments
(Richards and Wiley, 1980;
Wiley and Richards, 1978).
Frequency components of calls were clearly influenced by habitat structure,
and attenuation differed depending on frequency levels used. This leads us to
question how vegetative cover affects transmission of different frequencies.
Attenuation of sound is highly dependent on frequency
(Morton, 1975). Wiley and
Richards (1978) suggested that
wavelengths of vocal signals may be fine-tuned to the diameter of plants and
air pockets in the local environment, allowing for least attenuation (see also
Richards and Wiley, 1980).
Wavelength is proportional to frequency, with shorter wavelengths producing
higher frequency sounds. As vegetation cover changes, the diameter of plants
and air pockets may change, leading to more optimal use of slightly different
frequencies. The mechanics of sound propagation through vegetation are
extremely complex, but future studies that are designed to isolate the cause
of the relationships between frequency and habitat structure would provide an
appealing next step to this study.
Perhaps the most interesting and puzzling result of this study is the
seasonal change in the relationship between habitat structure and call
characteristics. The relationship between habitat and maximum frequencies of
calls in the premonsoon season actually produced calls that attenuated more
strongly on home colonies than on foreign colonies. In postmonsoon season, we
observed the opposite. Dominant and fundamental frequencies attenuated less on
home colonies than on foreign colonies. Why did we find this changing
relationship between habitat structure and alarm call structure? The
adaptation of an acoustic signal to its local acoustic environment depends on
the purpose of the signal as well as on the physical structure of the acoustic
environment. In some cases, acoustic signals will be selected for optimizing
sound transmission over maximum distances
(Morton, 1975). However, some
signals may be structured to predictably degrade for the purposes of
establishing territory boundaries or reducing the chance of attracting
predators or competitors (Brown and
Handford, 1996; Marler,
1955; Morton,
1986).
When looked at from this perspective, the different attenuation
performances of frequencies suggest that alarm calls could be designed for
different purposes that change seasonally. Specifically, certain call
components may be designed for maximum travel in postmonsoon season and
degradation over short distances in premonsoon season. Gunnison's prairie dog
alarm calls contain specific information on predators (i.e., type, danger
level, hunting style), and frequency components of calls carry this
information (Placer and Slobodchikoff,
2000,
2001;
Slobodchikoff et al., 1991).
Little is known about the actual coding of the information in the calls.
However, differential attenuation of certain frequencies in a call could
provide extra information about predators.
We cannot conclusively relate seasonal changes in the call—habitat
structure relationship to changes in call purpose without conducting studies
that relate changes in alarm calls to fitness levels. However, we do offer
some hypotheses. Premonsoon season corresponds to the time when pups emerge
from their natal burrows. Prairie dogs live in spatially and temporally fixed
territories within the colony, and pups reside in their natal territory until
dispersal at the end of the summer
(Robinson, 1989;
Travis and Slobodchikoff,
1993). Perhaps quickly degrading alarm calls in the premonsoon
season are directed toward the more vulnerable pups. Hoogland
(1996) suggests that alarm
calling in Gunnison's prairie dog females is primarily an expression of
parental care. The cost of calling long distances in the premonsoon seasons
could be too high, as it draws the attention of predators to the area of the
caller where the pups reside. Alarm calls that degraded quickly could lower
the cost of calling. Calling with lower amplitude might have the same effect
of lowering the cost of detection by predators. However, alarm calls in which
the maximum frequency drops out quickly would have a narrower bandwidth, which
would cause them to be less locatable even at a short distance. Alarm calls of
birds contain narrow bandwidths of frequencies that increase inconspicuousness
and decrease predator detection (Marler,
1955).
Thus, in the premonsoon season, calls with maximum frequencies that carried
shorter distances would have narrower bandwidths at a distance, yet contain
the full amount of information at close range. This may serve to warn pups
while minimizing the chance that a predator could hear and locate the call. In
the postmonsoon season juveniles disperse away from their natal territory
(Robinson, 1989). This could
alleviate the need for calls that degrade over short distances and promote
calls that carry long distances in postmonsoon season.
In addition to the decreased cost and increased benefit of calling over
long distances in postmonsoon season, alarm calls that carry long distances
could be tied to the decrease in visibility experienced with the increase in
vegetation after the monsoon. As a whole, vegetation cover increased between
preand postmonsoon season and was taller than the height of most prairie dogs.
Perhaps decreased visibility makes prairie dogs more dependent on acoustical
communication for predator detection and avoidance in this season. Higher
selection pressure on acoustic signals in habitats that are less visibly open
has been shown in populations of Old World monkeys. Brown and co-workers
(1995) found a higher incidence
of acoustic signals modified for long-distance travel in species residing in
low-visibility areas such as forests. In contrast, savanna monkeys, residing
in high-visibility areas, did not have acoustic signals fine-tuned for
long-range communication.
Ultimately, in order to say that habitat structure influences the design of
alarm calls, the functionality of the alarm call structure in terms of fitness
needs to be described. The fact that alarm call frequencies differ in relation
to habitat (both between colonies and seasonally within the same colony) and
that this results in actual differences in attenuation through different
environments is a first step in discovering functionality of alarm calls. As a
first step, we can point to the fact that there is a relationship between
frequency components of alarm call dialects and vegetative cover. We can also
say that this relationship affects the transmission of the alarm call through
the environment, which implies that calls may be adapted to home environments.
Most studies examining the influence of habitat on acoustical signals
investigate territory advertisement, species recognition, and mating songs
(e.g., Baker, 1975;
Morton, 1986;
Wiley, 1991). Further research
on the relationship between habitat structure and prairie dog alarm calls has
the potential to incorporate new and different selection pressures such as kin
selection and predation into the study of the effects of habitat on acoustical
communication.
|
ACKNOWLEDGEMENTS |
|---|
We thank Chris Kennett, Jennifer Verdolin, Sissy Wong, Sarah Havins, and Dustin Stairs for their assistance in data collection; Drs. Drickamer, Scott, and Kearsley for their input on design and analysis of the experiments and editing, and Dr. Graydon Bell and Lia Mann from the Department of Mathematics and Statistics for assistance with statistical methods. We heartily thank the landowners that allowed us to conduct research on their property. We would not have had a project without them.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Baker C, 1975. Song dialects and genetic differences in white-crowned sparrows (Zonotrichia leucophyrus). Evolution 29: 226-241.
Brown CH, Gomez R, Wasser PM, 1995. Old world monkey vocalizations: adaptation to the local habitat? Anim Behav 50: 945-961.
Brown T, Handford P, 1996. Acoustic signal amplitude patterns: a computer simulation investigation of the acoustic adaptation hypothesis. Condor 98: 608-623.
Date E, Lemon R, 1993. Sound transmission: a basis for dialects in bird song? Behaviour 124: 291-312.
Gorman RC, 1981. Nudes and foodes: Gorman goes gourmet. Flagstaff, Arizona: Northland Press.
Hall E, Kelson K, 1959. The Mammals of North America, vol 1. New York: Ronald Press.
Handford P, Lougheed S, 1991. Variation in duration and frequency characters in the song of the rufous-collared sparrow, Zonotrichia capensis with respect to habitat, trill dialects and body size. Condor 93: 644-658.
Hansen P, 1979. Vocal learning: its role in adapting sound structures to long-distance propagation and a hypothesis on its evolution. Anim Behav 27: 1270-1271.
Hoogland J, 1995. The black-tailed prairie dog: social life of a burrowing mammal. Chicago: University of Chicago Press.
Hoogland J, 1996. Why do Gunnison's prairie dogs give anti-predator calls? Anim Behav 51: 871-880.
Hurlbert SH, 1984. Pseudoreplication and the design of ecological field experiments. Ecol Monogr 54: 187-211.
Kroodsma DE, 1989. Suggested experimental designs for song playbacks. Anim Behav 37: 600-609.
LeBouf B, Peterson R, 1969. Dialects in elephant
seals. Science 166:
1654-1656.
Linskens H, Martens M, Hendricksen H, Roestenberg-Sinniage A, Brouwers W, vanderStaak A, Strick-Jansen A, 1976. The acoustic climate of plant communities. Oecologia 23: 165-177.
Marler P, 1955. Characteristics of some animal calls. Nature 176: 6-8.
Marler P, Tamura M, 1962. Song dialects in three populations of white-crowned sparrows. Condor 64: 368-377.
Marten K, Marler P, 1977. Sound transmission and its significance for animal vocalization I. Temperate habitats. Behav Ecol Sociobiol 2: 271-290.
Marten K, Quine D, Marler P, 1977. Sound transmission and its significance for animal vocalization II. Tropical forest habitats. Behav Ecol Sociobiol 2: 291-302.
McNulty F, 1971. Must they die? Garden City, New York: Doubleday.
Morton E, 1975. Ecological sources of selection on avian sounds. Am Nat 108: 17-34.
Morton E, 1986. Predictions from the ranging hypothesis for the evolution of long distance signals in birds. Behaviour 99: 65-86.
Placer J, Slobodchikoff CN, 2000. A fuzzy-neural system for identification of species-specific alarm calls of Gunnison's prairie dogs. Behav Process 52: 1-9.[ISI][Medline]
Placer J, Slobodchikoff CN, 2001. Developing new metrics for the investigation of animal vocalizations. Intelligent Automat Soft Computing 7: 1-11.
Rayor LS, 1985. Effects of habitat quality on growth, age of first reproduction, and dispersal in Gunnison's prairie dogs (Cynomys gunnisoni). Can J Zool 63: 2835-2840.
Richards D, Wiley R, 1980. Reverberations and amplitude fluctuations in the propagation of sound in a forest: implications for animal communication. Am Nat 115: 381-399.
Robinson A, 1989. Dispersal of the Gunnison's prairie dog (Cynomys gunnisoni) (Masters thesis). Flagstaff: Northern Arizona University.
Rothstein S, Fleischer R, 1987. Vocal dialects and their possible relation to honest status signaling in the brown-headed cowbird. Condor 89: 1-23.
Ryan M, Brenowitz E, 1985. The role of body size, phylogeny and ambient noise in the evolution of bird song. Am Nat 126: 87-100.
Searcy WA, 1989. Pseudoreplication, external validity and the design of playback experiments. Anim Behav 38: 717-720.
Shy E, 1983. The relation of geographical variation in song to habitat characteristics and body size in North American tanagers (Thraupine: Piranga). Behav Ecol Sociobiol 12: 71-76.
Slobodchikoff CN, Ackers S, Van Ert M, 1998. Geographic variation in the alarm calls of Gunnison's prairie dogs. J Mammal 79: 1265-1272.
Slobodchikoff CN, Coast R, 1980. Dialects in the alarm calls of prairie dogs. Behav Ecol Sociobiol 7: 49-53.
Slobodchikoff CN, Kiriazis J, Fisher C, Creef E, 1991. Semantic information distinguishing individual predators in the alarm calls of Gunnison's prairie dogs. Anim Behav 42: 713-719.
Slobodchikoff CN, Robinson A, Schaack C, 1988. Habitat use by Gunnison's prairie dogs. In: Habitat use and management of small mammals and amphibians. Washington, DC: US Forest Service.
Sokal R, Rohlf J, 1995. Biometry, 3rd ed. Oxford: W.H. Freeman.
Somers P, 1973. Dialects in southern Rocky Mountain pikas, Ochotona princeps. Anim Behav 21: 124-137.
Travis S, Slobodchikoff CN, 1993. Effects of food resources on the social system of Gunnison's prairie dogs. Can J Zool 71: 1186-1192.
Travis S, Slobodchikoff CN, Keim P, 1997. DNA fingerprinting reveals low genetic diversity in Gunnison's prairie dog (Cynomys gunnisoni). J Mammal 78: 725-732.
Tubaro P, Segura E, 1995. Geographic, ecological and subspecific variation in the song of the rufous-browed peppershrike (Cyclarhis gujanensis). Condor 97: 792-803.
Waring GH, 1950. Sound communication of black-tailed, white-tailed, and Gunnison's prairie dogs. Am Midl Nat 83: 167-185.
Wass J, 1988. Song pitch—habitat relationships in white-throated sparrows: cracks in acoustic windows? Can J Zool 66: 2578-2581.
Wasserman F, 1979. The relationship between habitat and song in the white-throated sparrow. Condor 81: 424-426.
Wiley R, 1991. Associations of song properties with habitats for territorial oscine birds of eastern North America. Am Nat 138: 973-993.
Wiley R, Richards D, 1978. Physical constraints on acoustic communication in the atmosphere: implications for the evolution of animal vocalizations. Behav Ecol Sociobiol 3: 69-94.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. Nettle Language and genes: A new perspective on the origins of human cultural diversity PNAS, June 26, 2007; 104(26): 10755 - 10756. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||