Taylor UniversityPillars at Taylor University

Master of Environmental Science Thesis Collection

2006

Analysis of Response Calls to Diverse GroundPredators from Three Geographically SeparateFlorida Scrub-Jay (Aphelocoma coerulescens)SubpopulationsSarah A. (Bales) Douglass

Follow this and additional works at: https://pillars.taylor.edu/mes

Part of the Environmental Sciences Commons

This Thesis is brought to you for free and open access by Pillars at Taylor University. It has been accepted for inclusion in Master of EnvironmentalScience Thesis Collection by an authorized administrator of Pillars at Taylor University. For more information, please contact pillars@taylor.edu.

Recommended Citation(Bales) Douglass, Sarah A., "Analysis of Response Calls to Diverse Ground Predators from Three Geographically Separate FloridaScrub-Jay (Aphelocoma coerulescens) Subpopulations" (2006). Master of Environmental Science Thesis Collection. 3.https://pillars.taylor.edu/mes/3

ANALYSIS OF RESPONSE CALLS TO DIVERSE GROUND PREDATORS FROM

THREE GEOGRAPHICALLY SEPARATE FLORIDA SCRUB-JAY (APHELOCOMA

COERULESCENS) SUBPOPULATIONS

SARAH A. BALES1*, JAN M. REBER2, M. REBECCA BOLT3, and ROBERT T. REBER4

1Department of Earth and Environmental Science, 236 West Reade Ave, Taylor

University, Upland, Indiana 46989

2Department of Biology, 236 West Reade Ave, Taylor University, Upland, Indiana 46989

3The Dynamac Corporation, DYN-1, Kennedy Space Center, Florida 32899; and

4Department of Earth and Environmental Science, 236 West Reade Ave, Taylor

University, Upland, Indiana 46989

*Email: sarah.bales@gmail.com

2

ABSTRACT

Dialectic variation occurs in many bird species. Different factors have been

investigated regarding dialectic variation, including cultural and genetic transmission of

songs or calls, and geographic separation. In this study, the predator alarm call of the

Florida Scrub-Jay to ground predators was recorded in three geographically separate

subpopulations across Florida—Canaveral National Seashore/Merritt Island National

Wildlife Refuge, Lyonia Preserve and Leisure Lakes/Lake June in Winter Scrub State

Park—to examine any dialectic occurrence between these three subpopulations.

Additionally, many bird species are recognized as having highly evolved predator

recognition systems, often with different calls for specific predators. The Florida Scrub-

Jay has exhibited such varied call responses to predators. Three different ground predator

treatments, an artificial snake, a live snake, and a live cat, were used to elicit alarm calls.

Responses to each predator were then compared within a subpopulation. This study

revealed that significant differences existed between subpopulations responding to the cat

and artificial snake treatment, but no significant differences existed between

subpopulations for the live snake treatment. These results indicate dialectic divergence

between subpopulations in response to some species of ground predators, but not to the

snake, with which the birds have evolved and are most familiar. Florida Scrub-Jay

responses to the three ground predators were significantly different within two

subpopulations, at Leisure Lakes/Lake June in Winter Scrub State Park and Lyonia

Preserve, but not within the Canaveral National Seashore/Merritt Island National Wildlife

Refuge subpopulation. Results from this study could be valuable to scientists, wildlife

managers, and communities striving to improve their conservation efforts.

3

Key words: predator alarm call, ground predators, Florida Scrub-Jay, dialectic

variation, geographic separation, habitat fragmentation, conservation

4

INTRODUCTION

The Florida Scrub-Jay, Aphelocoma coerulescens, was officially designated in

1987 as Threatened due to a major decline in its population, an approximately 25-50

percent decrease over the previous decade (USFWS 1999). This decline in the Florida

Scrub-Jay (FLSJ) population was attributed to significant habitat loss, fragmentation, and

degradation that resulted primarily from an increase in urbanization, agriculture, and lack

of scrub management (USFWS 1999).

Florida Scrub-Jay families consist of a monogamous pair with additional

“helpers” that are typically previous offspring, and often number between two to eight in

total family size (McGowan and Woolfenden 1989). The risk of potential predation

decreases when birds live in groups because of the increased protection sentinels add.

This sentinel system involves an individual bird that is not foraging, but rather stands

guard, exchanging this task with other birds of the same family group (McGowan and

Woolfenden 1989). Particular behaviors, such as mobbing and alarm calls, are elicited

when a potential threat is spotted by a sentinel who then initiates an alarm call.

Florida Scrub-Jays exhibit mobbing behavior through loud calling, scolding, and

visual cues, described by Francis et al. (1989), to draw attention to possible predators.

They often mob predators on the ground, but hide from raptors spotted in the air

(McGowan and Woolfenden 1989). By drawing attention to a predator their survivability

increases. This is especially important because predation is the largest contributing factor

to FLSJ mortality, particularly during nesting (Woolfenden and Fitzpatrick 1984,

Fitzpatrick et al. 1991; Bowman and Woolfenden 2001; Thorington and Bowman 2003).

Snakes are considered to be a major nest predator on both nestlings and fledglings

5

(Schaub et al. 1992; Woolfenden and Fitzpatrick 1996). Free-ranging (including feral)

cats are increasingly becoming a threat to all birds, including the FLSJ. In a report by the

Florida Fish and Wildlife Conservation Commission (2001), it is estimated that free-

ranging cats in Florida may be killing 68 million birds per year.

The decline in FLSJ populations is not only attributed to high predation risks but

also to habitat fragmentation. Habitat fragmentation often disrupts and isolates smaller

portions of a population resulting in metapopulations and eventually subpopulations. As

these subpopulations become more isolated from one another, due to urbanization or fire

suppression, the risk of extinction will increase (Woolfenden and Fitzpatrick 1991; Stith

et al. 1996; Breininger et al. 1999). Once isolated, genetic variation may occur, which

can also result in dialect formation among spatially separated species. Song dialects have

been extensively studied and numerous hypotheses posed about their adaptive

significance (MacDougall-Shackleton and MacDougall-Shackleton 2001).

Many bird species are known to have different song and/or call dialects, including

the White-crowned Sparrow (Baker and Cunningham 1985; Cunningham et al. 1987),

Orange-fronted Parakeet (Bradbury et al. 2001), Song Sparrow (Harris and Lemon 1971),

Northern Cardinal (Lemon 1966), and Bewick’s Wren (Kroodsma 1974). Conclusions

from these studies indicate that song dialects may be influenced by factors that include

cultural learning, genetic transmission, and geographic separation. Relatively few studies

have examined the acoustical and genetic differences between populations of the FLSJ.

Cultural learning has been described as non-genetic song-learning, from

generation to generation (Liaolo et al. 2001). Songs or calls can be learned over time

(cultural-evolution) and the errors that occur in this learning process may contribute to

6

song variability (Lynch and Baker 1994). Genetic mutations that arise, as well as with

population size and migration, could largely affect call diversity within a population

(Lynch and Baker 1993). The influence geographic separation has on dialectic variation

also may involve cultural and genetic transmission factors. With the lack of interaction

between neighboring populations, these cultural and genetic factors could escalate the

rate of dialectic divergence. For example, if a population became more isolated from

another population, they are less likely to share songs and genes. If one of these

populations were to learn or establish songs that became distinctly different through time,

they would no longer share similar songs and, therefore, potentially not reproduce. This

could result in further genetic isolation. In addition, different environmental pressures

may influence song transmission within populations in geographically separate areas

(Lynch and Baker 1993). Possible environmental pressures may include different

predators or habitat types, factors pertinent to this study.

Dispersal distances also may contribute to population and dialectic variation. The

Florida Scrub-Jay is a non-migratory species with intense natal philopatry, thereby

remaining near its natal territory throughout life (Stith et al. 1996; Fitzpatrick et al. 1999).

Fitzpatrick et al. (1999) found that FLSJ disperse, on average, about 1.6 km from their

natal territory. This dispersal rate is noted as being one of the most restricted among all

bird species (Stenzler and Fitzpatrick 2002). Fitzpatrick et al. (1999) also found that

FLSJ are more likely to disperse within the corridors formed by historic sand ridges

found across Florida. The xeric oak scrub ecosystem (optimal FLSJ habitat) is associated

with Florida’s ancient dune systems that parallel the coasts of Florida (Woolfenden and

Fitzpatrick 1996). These sand ridges include the Lake Wales Ridge area, and both

7

coastal (Gulf and Atlantic) areas. The xeric oak scrub ecosystem has been described by

Menges (1999) in detail along with specific habitat requirements of the FLSJ. Patches in

this ecosystem that provide optimum habitat contain: low oak-dominated species about 2

m in height or less, minimal pine canopy, and open sandy patches for caching acorns

(Fitzpatrick et al. 1991; Breininger et al. 1995; Menges 1999; Bowman and Woolfenden

2002).

The levels of threat presented to birds by predators are species specific, and their

response (behaviorally and acoustically) has evolved to recognize and address diverse

predators and their associated danger. The Florida Scrub-Jay particularly exemplifies a

social unit which responds differently to different predator-related threats. The FLSJ has

been extensively studied, largely because of its unique family structure that involves

helpers, a cooperative breeding strategy, non-migratory nature, and specific habitat

requirements (Woolfenden and Fitzpatrick 1990). Most studies have examined FLSJ life

history patterns while few have actually analyzed their vocalizations, especially predator-

elicited calls.

An early study by Barbour (1977) categorized the FLSJ vocal repertoire and

recognized differences between their response to ground and aerial predators. The FLSJ

response to a ground predator is referred to as the “predator screech scold,” while the

response to an aerial predator is noted as an “aerial predator alarm weep” (Barbour 1977).

The response to an aerial predator is not only acoustically different but behaviorally

different than to a ground predator (Barbour 1977). Other animal species, such as the

California ground squirrel and Gunnison’s prairie dog, also have different responses to

aerial and ground predators (Placer and Slobodchikoff 2000). One aspect of this present

8

study focuses on the FLSJ acoustic response to ground predators, and the term “predator

screech scold” or “predator alarm call” are used interchangeably herein.

In another study (Elowson and Hailman 1991), FLSJ predator-elicited calls to

both ground and aerial predators were classified through dichotomous sorting. In

Elowson and Hailman’s (1991) work, predator-elicited calls of the FLSJ showed complex

variation encompassing twelve different call types and subtypes in response to various

aerial and ground predators. In particular, the predator alarm calls to snakes clustered

into a ‘flat’ call type, with four subtypes. For further description of the different call

types and subtypes see Elowson and Hailman (1991).

Another study examined the “hiccup” call between three sub-regional

populations across Florida (Russel unpubl. data 2005). Hiccup calls are exclusive to

females and often made in territorial disputes, for further description see Barbour (1977).

Interestingly, for the first time, Russel (unpubl. data 2005) found dialectic variation in the

hiccup call between all three sub-regional populations.

A highly developed predator recognition system also has evolved in other bird

species such as Carolina Wrens (Morton and Shalter 1977), Black-capped Chickadees

(Apel and Ficken 1981; Baker and Becker 2002), Tufted Titmice (Waite and Grubb

1987), and Lapwings (Walters 1990). These studies revealed some bird species have

diverse responses to various predators. These responses include different call structures

related to the level of threat presented by a predator (Morton and Shalter 1977; Baker and

Becker 2002) or distinct call types to various predators (Walters 1990). Some other

examples in the animal kingdom that incorporate specific behavioral and/or acoustical

signals for predator recognition include vervet monkeys (Cheney and Seyfarth 1990;

9

Brown et al. 1992), baboons (Fischer et al. 2001), and prairie dogs (Ackers and

Slobodchikoff 1999). The potential risk of predation decreases as a social unit cooperates

to watch for and detect predators.

Due to its specific habitat requirements, limited dispersal range, and increasing

fragmentation of its optimal habitat, the FLSJ is at high risk for further population

isolation. This may lead to not only greater dialectic variation but genetic isolation

between populations and, therefore, potentially to speciation or extinction. The focus of

this study was two-fold: to investigate possible dialectic variation in the alarm call

responses between three geographically separate FLSJ subpopulations, and to determine

any significant differences in FLSJ responses to three different ground predators. This

study hopes to contribute to the understanding of call types, dialectic variances, adaptive

significance, and conservation management for the FLSJ.

METHODS

Study sites and populations.—There are five major sub-regions of the Florida

Scrub-Jay found in geographically separate locations, three of which are considered “core

populations” (Fig. 1) (Fitzpatrick et al. 1994 cited in USFWS 1999). This study focused

on three subpopulations within the core populations from June to August 2005. All three

sub-regional areas—the Atlantic Coast Subregion, Ocala Subregion, and Lake Wales

Ridge Subregion—occur along major sand ridges (Fitzpatrick et al. 1994 cited in USFWS

1999). The first study site was at Canaveral National Seashore/Merritt Island National

Wildlife Refuge (CANA/MINWR) and Kennedy Space Center (KSC), referred to as CA

hereafter. This rural study site was located in the Atlantic Coast Subregion. The second

study site was located at Leisure Lakes (LL) near Lake June in Winter Scrub State Park,

10

Lake Placid, Florida. This semi-rural study site was located in the Lake Wales Ridge

Subregion. The third study site, in the Ocala Subregion, was a 360-acre preserve called

Lyonia Preserve (LY). This preserve was located behind the Deltona Regional Library in

the city of Deltona, Florida. All three sites consisted of xeric habitats that ranged from

high quality managed oak scrub, to partially managed scrubby flatwoods, and restored

scrub habitat from sand pine and oak stands, respectively.

At each study site, six to eight families were randomly chosen and recorded.

Typically, these families were easily identified because most or all members were banded

with a US Fish and Wildlife metal band and a color band pattern issued by previous

researchers. This allows for identifying which family particular birds belong to. The

Florida Scrub-Jays were habituated to human observers at all sites.

Experimental procedure and equipment.—One artificial and two live predators

were used to elicit predator alarm calls. An Eastern Corn Snake, Elaphe guttata, and a

domestic cat, Felis catus, were used for the two live predator stimuli. An artificial blow-

up snake was used as the third stimulus. The model snake mimicked an Eastern Indigo,

Drymarchon couperi, in size and coloration. It was chosen because it is a natural snake

predator of the FLSJ. All trials were conducted in the mornings from 0730 to 1200.

Under no circumstances were the live animals involved in this study harmed in any way,

and at no time did they pose any harm to the FLSJ.

Three of the six to eight designated families were randomly chosen to be recorded

the first of a two-day period, and another three were recorded on the next day. This

schedule was repeated after a 24 h wait period. If all three families were not recorded on

a given day, a recording of the missed family was attempted on the third day. In some

11

cases, families were recorded over a three day period with a 24 hour wait period followed

by either two or three more days of recording. This procedure was repeated for each

treatment at each site. The exception occurred at Leisure Lakes where it was necessary to

record two families each day over a consecutive four day period. The initial failure of the

randomly selected families to appear on the first day was most likely due to the onset of

the acorn season (August). These families were possibly actively foraging and caching

acorns (R. Bowman, pers. comm.), which are an important food source and critical to

FLSJ survival (Menges 1999).

The artificial snake was the first stimulus exhibited to FLSJ families, followed by

the live Eastern Corn Snake and, finally, a house cat. The artificial snake was placed on a

saw palmetto or oak scrub species with a string attached to produce movement. The trial

began when at least one adult FLSJ was present. The trial ended when the birds ceased

scolding and apparently lost interest or simply flew away. Trial duration resulted in an

average of 3 minutes of recording time.

The live Eastern Corn Snake was placed in a Plexiglas and wood-structured cage

9 ¾ x 10 x 32 inches (w x h x l) that was set on a small foldable table and placed within

the scrub but still conspicuous to the birds. The trial ended when the birds flew away

after scolding, or when the snake was removed for its own well-being (heat stress), but

usually after about 3 minutes of recording.

Two male, spayed and de-clawed, gray littermates were used interchangeably at

CA and LY. A third cat, a black male that was spayed and de-clawed, was used at

Leisure Lakes. A cat carrier was utilized to transport and hold each cat until the top was

opened for exposure to the birds. Cats were leashed and attached to the carrier at all

12

times in the field. A cat was placed in a noticeable area near scrub when one or more

adult FLSJ were present. The trial ended when either the birds flew away or if the cat

became unsettled and was removed for its own well-being. Again, this resulted in an

average of 3 minutes of recording time per session.

Analysis of calls.—The predator screech scold, or alarm call, ranges between 1-8

kHz (Francis et al. 1989; Elowson and Hailman 1991). Francis et al. (1989) further

described the alarm call as a harsh vocalization often given consecutively in a long series

but sometimes singly. For further description of the predator alarm call, see Barbour

(1977) and Elowson and Hailman (1991).

Calls were recorded using a shotgun microphone (Audio-technica U.S.,

line+gradient condenser microphone, frequency response 20-20,000Hz, Audio-technica

U.S., Inc., Stow, Ohio), which was mounted on a 54 cm parabolic dish. The microphone

was integrated with an audio interface (Tascam US-122 USB, 24-bit depth, 44 kHz rate)

to digitally convert the calls for analysis into a COMPAQ laptop running the sound

analysis software Raven 1.2.1 (Cornell Lab of Ornithology) (Fig. 2). The first step was

to run calls through a 1,000 Hz High Pass filter to reduce background noise.

Spectrograms were then created in Raven with the following parameters: 128 point frame

length (1398.80 Hz bandwidth), 256 point Fourier transform size, Hamming, 512 samples.

Four random adult FLSJ calls from five randomly selected family groups in each

subpopulation for each of the three predators, a total of 180 calls, were measured for two

acoustical variables using Raven. The maximum frequency “Max Frequency” function,

as demonstrated by Miller, Vanderwerf, and McPherson (2003), and the duration, or

lengths, of the calls were analyzed (Fig. 3).

13

Statistical analyses.—For this study, a one-way analysis of variance (ANOVA)

was used to test differences between subpopulations (i.e. cat treatment between the three

subpopulations) and treatments within a subpopulation (i.e. between the cat, artificial and

live snake at LY). When statistical significance was determined, a Tukey Test (mean

separation test) was conducted.

Two datasets—maximum frequency for the artificial snake recordings between

subpopulations and call duration for the cat recordings between subpopulations—did not

follow normal distributions; thus, non-parametric analyses were utilized. Therefore, a

Kruskal-Wallis test, which acts similarly to a one-way ANOVA and is not dependent on

normal distribution, was used. This was followed by a non-parametric mean separation

test, the Mann-Whitney test, to determine specifically which treatment means were

significant.

RESULTS

Twenty calls per stimulus for a total of sixty calls per subpopulation (180 calls in

all) were analyzed for the two acoustic variables—max frequency and call duration. Data

were analyzed by treatments and subpopulations.

Cat treatment between subpopulations.—A one-way ANOVA for the average

max frequency, with normal distribution, was significant (P < 0.001). LL’s average max

frequency was significantly less (Tukey test, P < 0.001) than CA and LY. However, LY

and CA were not significantly different from each other. A Kruskal-Wallis test for the

average call duration, with non-normal distribution, revealed significance between

subpopulations (P < 0.001). All subpopulations were found to be significantly different

(Mann-Whitney test, P < 0.001) from each other (Fig 4).

14

Artificial snake treatment between subpopulations.—A Kruskal-Wallis test for the

average max frequency, with non-normal distribution, revealed significance between

subpopulations (P = 0.01). CA was significantly less than LL and LY (Mann-Whitney

test, P < 0.01), but LL and LY were not significantly different from each other. A one-

way ANOVA for call durations, with normal distribution, between subpopulations was

significant (P = 0.005). The average call duration of LY was significantly less (Tukey

test, P = 0.005) than CA and LL, but there was no significant difference between CA and

LL (Fig 5).

Live snake treatment between subpopulations.—A one-way ANOVA was not

significant for FLSJ responses to the live snake treatment at any of the three

subpopulations for the average max frequency (P = 0.236) or call duration (P = 0.572)

(Fig 6).

All treatments within CANA/MINWR.—A one-way ANOVA found no significant

differences for the average max frequency (P = 0.109) and average call duration (P =

0.229) between any treatment (Fig 7).

All treatments within Leisure Lakes.—A one-way ANOVA for the average max

frequency was significant (P < 0.001). The cat, artificial snake, and live snake treatments

were significantly different from each other (Tukey Test, P < 0.001). A one-way

ANOVA for the average call duration was significant (P < 0.001). The average call

duration for the cat treatment was significantly less (Tukey Test, P<0.001) than both

snake treatments, but the snake treatments were not significantly different from each

other (Fig 8).

15

All treatments within Lyonia Preserve.—In a one-way ANOVA analysis, there

were no significant differences found for the average max frequency between any

treatment (P = 0.766). A one-way ANOVA for the average call duration was significant

(P < 0.001). Similar to LL, the average call duration for the cat treatment was

significantly less (Tukey Test, P < 0.001) than both snake treatments, but the snake

treatments were not significantly different from each other (Fig 9).

DISCUSSION

Although the FLSJ has been studied extensively given its unique natural history,

its acoustic repertoire has typically been quantified and classified with minimal scientific

inquiry until recently. This study addressed dialectic variation between FLSJ

subpopulations with one particular acoustic entity, the predator alarm call. Dialectic

variation between FLSJ subpopulations has only been recently examined; using the

hiccup call, Russel (unpubl. data 2005) determined that there is indeed dialectic variation.

The current study confirms this conclusion, affirming dialectic variation within this bird

species.

Dialectic variation.—Dialect formation is widespread in many bird species (Harris and

Lemon 1971) and, therefore, it is not surprising to find dialectic variation between all

three FLSJ subpopulations. Results from this study for both maximum frequency and

call duration revealed that significant differences do exist between subpopulations in

response to the cat treatment and artificial snake treatment, but not between

subpopulations in regards to the live snake treatment (Figs. 4-6). Hence, dialectic

variation occurs between all three subpopulations when responding to two different

predator stimuli, the cat and artificial snake, but not to the live snake. Genetic isolation

16

and song or call learning between generations are potential causes for this dialectic

difference.

In general, dialectic variation has been extensively studied given its adaptive

significance for cultural-learning of songs and potential genetic isolation of populations

(Lachlan and Servedio 2004). Lemon (1966) suggested that dialects are influenced by

three factors: isolation of populations, generation to generation transmission of a dialect,

and a source of variation for the dialectic divergence. Since the FLSJ is highly sedentary

and disperses minimally away from its natal territory, isolation of the larger sub-regions

and even smaller subpopulations is becoming an influencing factor in this dialectic

divergence. Isolation, given time, often leads to genetic differentiation between

populations.

Song-learning.—Transmission of a dialect would be readily possible given the FLSJ

family structure, in which helpers often remain with the breeder pair for a year or more

(Woolfenden and Fitzpatrick 1996), thereby learning that particular dialect from the

family unit. This cultural-learning phenomenon is evident in many bird species, such as

Saddlebacks (Jenkins 1977), Indigo Buntings (Payne et al. 1981), and Chaffinches

(Lynch and Baker 1993), for example. In cultural transmission, or song learning, a song

or call is passed from one individual to another (Slater 1986). This type of song and call-

learning has not been extensively studied in the FLSJ. Questions Baptista (1975) posed

regarding song dialects among another non-migratory species, like the Nuttall’s White-

crowned Sparrow, may be appropriate to apply to the FLSJ given their similar natural

histories. Baptista addressed the following: What impacts do songs have on the gene

flow between neighboring populations? What is the state of song adaptation or

17

morphology on the borders of those populations with differing dialects? And, what

environmental “barriers” exist between populations?

The two initial questions Baptista (1975) posed would be interesting to apply to

FLSJ populations within a sub-regional area, for example, the Atlantic coast. A

comparison of a FLSJ call, studied or not previously studied acoustically, of families

within this region (CANA/MINWR to Kennedy Space Center) could yield interesting

comparisons. Baptista’s third question about potential environmental barriers is

addressed here.

Habitat fragmentation.—Another source for this dialectic variation may be due to the

highly fragmented nature of FLSJ habitat and disjointed landscape. Urbanization and

other anthropogenic sources (i.e. roads) have contributed greatly to the habitat loss that

has occurred across Florida (USFWS 1999). A disjointed landscape was evident in the

most urban study site (LY), where the 360 acre preserve of restored scrub sits within a

residential community (Fig. 10). The semi-rural site (LL) was broken up by roads and

interspersed with residential houses but it has an adjacent managed scrub habitat (Fig. 10).

The most rural site (CA) has had minimal human impact and has progressive

management practices for the present scrub habitat (Fig. 10).

With persistent pressures from habitat fragmentation and urbanization, FLSJ

survivability continues to significantly decrease, thereby increasing their risk for

extinction (Woolfenden and Fiztpatrick 1991; Breininger 1999, Breininger and Carter

2003). Several studies have examined the effects of habitat fragmentation and FLSJ

demographics within populations (see Breininger et al. 1996; Breininger 1999;

Breininger et al. 1999, Bowman and Woolfenden 2001). Breininger (1999) reported that

18

almost 30% of existing FLSJ populations inhabit fragments of habitats surrounded by

residential development. Breininger (1999) stated dispersal distances were longer in

urban areas than in rural or more natural areas. Also, demographic success (i.e.

population characteristics such as mortality, nesting, or fledgling success) in the FLSJ

was reduced in fragmented urban populations.

Since the FLSJ requires specific habitat conditions, habitat suitable for population

expansion is becoming increasingly harder to find, hence the longer dispersal distances in

urban settings. Thaxton and Hingtgen (1996) found that the FLSJ disperse from

fragmented habitat into larger land tracts but not from larger contiguous habitats to

smaller patchy habitats. These results, coupled with the limited average FLSJ dispersal

distance (1.6 km, Fitzpatrick 1999), may indicate a severe population interaction

limitation, even within the core populations found along the Atlantic and Gulf coasts, and

Lake Wales Ridge.

In addition to the factors mentioned previously, what else may be contributing to

this dialectic variation? This question may be answered by the second objective of this

study, which was to compare the response calls to three diverse ground predators within

each of the subpopulations.

When populations are separated from one another, thereby not communicating

with one another, one source of variation for dialectic divergence could be due to

different environmental stimuli (Nottebohm 1969). For instance, Thorington and

Bowman (2003) proposed that predator communities in urban and natural habitats follow

species-specific predation patterns. In suburban habitats, predators such as birds, rodents,

snakes, and terrestrial mammals may increase due to food availability. This seems most

19

likely, especially with cats. Different predator pressures may be influencing the dialectic

variation in the predator alarm calls within at least two study subpopulations.

Results from this part of the study for both max frequency and call duration

indicated significant differences in FLSJ responses to ground predators within LL and LY,

but not to ground predators within the CA subpopulation (Figs. 7-9). It is evident that

these subpopulations are exposed to different predator threats that, at this time, are

generating different acoustical responses.

Live snake.—There were no significant differences in response to the live snake

treatment between any of the subpopulations (Fig. 6). It is plausible that all of these

subpopulations have evolved with snakes and several snake species still exist in the

habitats of each subpopulation. Their response to the snake may have been influenced

mainly by the size of the snake stimulus (at least 58 in) but also perhaps due to its species

(Elaphe guttata). Elaphe guttata is a common snake found across Florida and is found

not only in rural areas but urban sites as well. Snakes account for the majority of FLSJ

deaths, especially young juveniles or nestlings (McGowan and Woolfenden 1989; Schaub

et al. 1992). Thorington and Bowman (2003) suggested that snake predation may

increase with human density because urban regions provide an area where nest predators

could flourish. However, there are few data that exist on snake communities and the

effects of urbanization on their populations and abundance or lack thereof in a suburban

setting (Bowman and Woolfenden 2001).

Artificial snake.—Some variation was evident in how the subpopulations responded to

the artificial snake. CA max frequency response to the artificial snake was significantly

different from LL and LY, whereas LL and LY were not significantly different from each

20

other (Fig. 5). Additionally, LY was significantly different from both LL and CA

subpopulations for the average call duration; however, there was not a difference in call

duration between CA and LL (Fig. 5).

The artificial snake mimicked an Eastern Indigo, which is a common snake

predator on the FLSJ (McGowan and Woolfenden 1989). However, due to the natural

history of this snake, it is more likely to frequent habitat-specific rural areas (USFWS

1999). This could explain why the max frequency of response to the artificial snake was

significantly different between CA and the other two sites. CA is the most natural, with

little human interference to significantly impact this snake’s habitat (Fig. 10). Duration

data indicated a significant difference between the two rural subpopulations (CA and LL)

with the more urban subpopulation (LY), but not between the rural subpopulations

themselves. Even low human density and development can impact snake populations

like the Eastern Indigo (Moler 1992; Bowman and Woolfenden 2001), thereby, again,

reinforcing perhaps D. couperi preference for the most rural areas. The two rural

subpopulations (CA and LL) have more exposure and awareness of the threat Eastern

Indigos pose compared to the urban subpopulation (LY) having little or no exposure to

these snakes. This could account for the different acoustic response to the artificial snake

treatment.

Domesticated cat.—An interesting aspect of this study observed cats as ground predators.

It is estimated that 60-100 million free-ranging cats exist within the United States (Mott

2004). Some experts estimate these cats kill millions of birds each year. Bobcats are a

natural predator of FLSJ, however, lower density and territoriality limit their impact on

the FLSJ. Domestic cats exist in high densities and are not territorial. This results in a

21

much more significant impact on local bird populations (Florida Fish and Wildlife

Conservation Commission 2001).

Results from the cat treatment were variable. For max frequency, LL was

significantly different from both CA and LY, while CA and LY were not different from

each other (Fig. 4). Interestingly, for call duration, all populations were significantly

different from each other. Lack of a clear pattern in the response of the FLSJ to domestic

cats may indicate that these birds could still be developing a response to this new

predation pressure. The more urban subpopulation (LY) has likely had the most

experience with cats as predators as opposed to the more rural sites (CA and LL) (Fig.

10). The FLSJ at CA responded to the cat similarly as to the other two predators,

therefore, there was no differentiation in threat levels presented by either the snakes or

cat. At LL, the more semi-rural subpopulation, data indicate perhaps a learning process

occurring within FLSJ families and their relationship with this emerging predator. A

dialectic divergence may become more distinct as the birds at LL are increasingly

exposed to cats. Their behavioral and acoustical response would continue to adapt to the

type of threat cats pose. Cats could easily be more apt to take fledglings and adults than

any other predator. In fact, cats could be one of the most burgeoning predator threats to

adult FLSJ. This could have significant ramifications for FLSJ subpopulations if they do

not learn and adapt their responses quickly to deal with this “new” predator.

Cats are often present in larger numbers near urban sites, even forming colonies,

because of the available food source provided by humans (Florida Fish and Wildlife

Conservation Commission 2001). If, by virtue of their relationship to urban areas, these

22

subpopulations continue to experience significantly different levels of predation by cats,

eventual dialectic variation or even extirpation of those subpopulations could occur.

Conclusions and Conservation Implications.—In the past, the FLSJ has evolved with and

adapted to predator threats with consistent responses as evidenced in this study with their

responses to the live snake. Historically, the FLSJ subpopulations were more connected,

but, today, the ever-growing urbanization has disrupted much of the optimal and

continual FLSJ habitat. This threat, along with increased exposure to cats, may occur in

random and isolated areas affecting the FLSJ sporadically. Therefore, subpopulations

would not have equal exposure to these new threats. This concomitant exposure of the

subpopulations to varying stimuli, including diverse ground predators, could encourage

greater dialectic variation.

Extinction of FLSJ subpopulations is the end result of habitat fragmentation (i.e.

population isolation). If further fragmentation of FLSJ habitat occurs, speciation is also

likely, given time and rate of divergence; however, the FLSJ faces the greater threat of

going extinct before full speciation could occur.

Further Recommendations.—Because of the threats facing FLSJ subpopulations

today, suggestions for future study are mentioned. One recommendation is to conduct

research on FLSJ calls in conjunction with behavior. In addressing both behavior and

calls, for instance, re-analyzing the predator alarm call to diverse predators (especially to

a cat), could potentially provide insightful understanding of dialectic divergence.

A second recommendation is to repeat this study with other FLSJ subpopulations

across Florida or even within a sub-region (Atlantic coast). As mentioned previously,

two initial questions posed by Baptista (1975) addressed song adaptation and learning,

23

dialect formation, and gene flow among neighboring populations. This study could be

repeated with the focus of investigating these questions.

Thirdly, other FLSJ calls could be studied to determine dialectic occurrences

between and/or within subpopulations. Some call examples for future analyses include

the aerial predator alarm weep or the rapid multiple weep.

Lastly, conduct further studies of free-ranging cats as predators of FLSJ to

determine their impact on this threatened species. By determining the impact and level of

threat that free-ranging cats have upon the FLSJ, greater conservation initiatives could be

employed for the FLSJ. Ultimately, the research from this study and future studies

concerning dialectic variation could provide understanding on the rate of divergence,

prospect for speciation, and the potential of extinction of this threatened species.

24

ACKNOWLEDGEMENTS

In Florida, I would like to warmly thank John Stiner and Candace Carter of

CANA/MINWR, Bonnie Cary and Sarah Cleveland of Volusia Co. Environmental

Management, and Reed Bowman of Archbold Biological Station for guidance, help, and

permission of study site use. I would like to thank Becky Bolt, at The Dynamac

Corporation of Kennedy Space Center, and volunteers who lent time and/or animals—

Jenn Stiner, Marcia Rickey, and students involved in Park Youth Workers at

CANA/MINWR. In Indiana, a huge thanks to my advisor Jan Reber for all her time and

patience regarding this project, Ken Constantine and Rob Reber for their wonderful help

on statistics, Bob Davis for his expertise in acoustics, Dick Squiers and Leland Boren for

the opportunity to participate in a great environmental science program, and Kory Russel

for the pioneer work of acoustical analyses in FLSJ subpopulations across Florida.

25

LITERATURE CITED

Ackers, S. H. and C. N. Slobodchikoff 1999. Communication of stimulus size and shape

in alarm calls of Gunnison’s prairie dogs, Cynomys gunnisoni. Ethology 105(2):149-162.

Apel, K. M. and M. S. Ficken. 1981. Predator alarm calls of young Black-capped

Chickadees. The Auk 98: 624.

Baker, M. C. and A. M. Becker. 2002. Mobbing calls of Black-capped Chickadees:

effects on urgency on call production. Wilson Bulletin 114(4): 510-516.

Baker, M. and D. Logue. 2003. Population differentiation in a complex bird sound: A

comparison of three bioacoustical analysis procedures. Ethology 109: 223-242.

Baker, M. C. and M. A. Cunningham. 1985. The biology of bird-song dialects. The

Behavioral and Brain Sciences 8: 85-133.

Barbour, D. B. 1977. Vocal communication in the Florida scrub jay. Master’s thesis,

University of South Florida, Tampa.

Bowman R. and G. E. Woolfenden. 2001. Nest success and the timing of nest failure of

Florida scrub-jays in suburban and wildland habitats. In: Marzluff, J. M., Bowman, R.

and Donnelly, R. (eds), Avian Ecology and Conservation in an Urbanizing World.

Kluwer Academic Publishing, pp. 383-402.

26

Bowman, R. and G. E. Woolfenden. 2002. Nest site selection by Florida Scrub-Jays in

natural and human-modified habitats. Wilson Bulletin 114(1): 128-135.

Breininger, D. R., V. L. Larson, D. M. Oddy, R. B. Smith, and M. J. Barkaszi. 1996.

Florida Scrub-Jay demography in different landscapes. The Auk 113(3): 617-625.

Breininger, D. R., M. A. Burgman, and B. M. Stith. 1999. Influence of habitat quality,

catastrophes, and population size on extinction risk of the Florida scrub-jay. Wildlife

Society Bulletin 27(3): 810-822.

Breininger, D. R. and G. M. Carter. 2003. Territory quality transitions and source-sink

dynamics in a Florida Scrub-Jay population. Ecological Applications 13(2): 516-529.

Brown M. M., Kreiter N. A., Maple J. T., Sinnott J. M. 1992. Silhouettes elicit alarm

calls from captive vervet monkeys (Cercopithecus aethiops). Journal of Comparative

Psychology 106(4): 350-359.

Cheney, D. and R. Seyfarth. 1990. How monkeys see the world. Chicago: University of

Chicago Press.

Cunningham, M. A. and M. C. Baker, T. J. Boardman. 1987. Microgeographic song

variation in the Nuttall’s White-crowned Sparrow. The Condor 89: 261-275.

27

Elowson, A. M. and J. P. Hailman. 1991. Analysis of complex variation: dichotomous

sorting of predator-elicited calls of the Florida scrub jay. Bioacoustics: The International

Journal of Animal Sound and Its Recording 3: 295-320.

Fischer J., K. Hammerschmidt, D. L. Cheney, and R. M. Seyfarth. 2001. Acoustic

features of female chacma baboon barks. Ethology 107(1): 33-54.

Fitzpatrick, J. W., G. E. Woolfenden, and M. T. Kopeny. 1991. Ecology and

development-related habitat requirements of the Florida Scrub Jay (Aphelocoma

coerulescens coerulescens). Florida Game and Fresh Water Fish Commission, Nongame

Wildlife Program Technical Report NO. 8. pp. 49.

Fitzpatrick, J. W., R. Bowman, D. Breininger, M. O’Connell, B. Stith, J. Thaxton, B.

Toland, and G. Woolfenden. 1994. Habitat conservation plans for the Florida Scrub

Jay: a biological framework. Unpublished draft report. On file at U.S. Fish and Wildlife

Service, South Florida Ecosystem Office; Vero Beach, Florida.

Fitzpatrick, J. W., G. E. Woolfenden, and R. Bowman. 1999. Dispersal distances and its

demographic consequences in the Florida Scrub-Jay. In; Proc. 22nd International

Ornithological Congress, Adams, N.J. and R. H. Slotow (Eds.). Birdlife South Africa. pp.

2465-2479.

Florida Fish and Wildlife Conservation Commission. 2001. Impacts of feral and free-

28

ranging domestic cats on wildlife in Florida. Florida Fish and Wildlife Conservation

Commission.

Francis, A. M., J. P. Hailman, and G. E. Woolfenden. 1989. Mobbing by Florida scrub

jays: behaviour, sexual asymmetry, role of helpers and ontogeny. Animal Behavior 38:

795-816.

Harris, M. A. and R. E. Lemon. 1971. Songs of song sparrows (Melospiza melodia):

individual variation and dialects. Canadian Journal of Zoology 50: 301-309.

Jenkins, P. 1977. Cultural transmission of song patterns and dialect development in a

free-living bird population. Animal Behavior 25: 50-78.

Kroodsma, D. E. 1974. Song learning, dialects, and dispersal in the Bewick’s wren. Z.

Tierpsychology 35: 352-380.

Lachlan, R. F. and M. R. Servedio. 2004. Song learning accelerates allopatric speciation.

Evolution 58(9): 2049-2063.

Lemon, R.E. 1966. Geographic variation in the song of Cardinals. Canadian Journal of

Zoology 44: 413-428.

Laiolo, P., A. Rolando, A. Delestrade, and A. De Sanctis. 2001. Geographic

29

diversification in the call repertoire of the genus Pyrrhocorax (Aves, Corvidae).

Canadian Journal of Zoology 79: 1568-1576.

Lynch, A. and A. J. Baker. 1993. A population memetics approach to cultural evolution

in Chaffinch song: meme diversity within populations. The American Naturalist 141(4):

597-620.

Lynch, A. and A. J. Baker. 1994. A population memetrics approach to cultural evolution

in Chaffinch song: differentiation among populations. Evolution 48: 351-359.

MacDougall-Shackleton E. A. and S. A. MacDougall-Shackleton. 2001. Cultural and

genetic evolution in Mountain White-crowned Sparrows: song dialects are associated

with population structure. Evolution 55(12): 2568-2575.

McGowan, K. J. and G. E. Woolfenden. 1989. A sentinel system in the Florida scrub jay.

Animal Behavior 37: 1000-1007.

Menges, E. S. 1999. Ecology and conservation of Florida scrub. In Savannas, Barrens,

and rock outcrop plant communities of North America (R.C. Andersons, J.S. Fralish, and

J.M. Baskin, Eds.) Cambridge University Press.

Miller, E. H., E.Vanderwerf, and L. McPherson. 2003. Vocalizations of the Tuamoto

Sandpiper Prosobonia cancellata. Wilson Bulletin 115(4): 455-463.

30

Moler, P. E. 1992. Eastern indigo snake, p. 181-186. In P. E. Moler (ed.), Rare and

Endangered Biota of Florida. Vol. 3, Amphibians and Reptiles. University Press of

Florida. Gainesville, Florida.

Morton, E. S. and M. D. Shalter. 1977. Vocal response to predators in pair-bonded

Carolina Wrens. The Condor 79: 222-227.

Mott, M. 2004. U.S. faces growing feral cat problem.

http://news.nationalgeographic.com/news/2004/09/0907_040907_feralcats.

Nottebohm, F. 1969. The song of the Chingolo, Zonotrichia capensis, in Argentina:

description and evaluation of a system of dialects. Condor 71: 299-315.

Payne, R. B., W. L. Thompson, K. L. Fiala and L. L. Sweany. 1981. Local song traditions

in Indigo Buntings: cultural transmission of behavior patterns across generations.

Behavior 77: 199-221.

Placer, J. and C. N. Slobodchikoff. 2000. A fuzzy-neural system for identification of

species-specific alarm calls of Gunnison’s prairie dogs. Behavioral Processes 52: 1-9.

Russel, K. 2005. Call variations between sub-regional Florida Scrub-Jay (Aphelocoma

coerulescens) populations. Thesis. Unpublished data.

31

Schaub, R., R. L Mumme, and G. E. Woolfenden. 1992. Predation on the eggs and

nestlings of Florida Scrub Jays. The Auk 109(3): 585-593.

Slater, P. J. B. 1986. The cultural transmission of bird song. TREE 1(4): 94-97.

Stenzler, L. M. and J. W. Fitzpatrick. 2002. Isolation of microsatellite loci in the Florida

Scrub-Jay Aphelocoma coerulescens. Molecular Ecology Notes 2: 547-550.

Stith, B. M., J.W. Fitzpatrick, G. E. Woolfenden, and B. Pranty. 1996. Classification and

conservation of metapopulations: a case study of the Florida scrub-jay. In: D.R.

McCullough (ed), Metapopulations and Wildlife Conservation Island. Washington D.C.,

pp. 187-215.

Thaxton, J. E. and T. M. Hingtgen. 1996. Effects of suburbanization and habitat

fragmentation on Florida Scrub-Jay dispersal. Florida Field Naturalist 24: 25-37.

Thorington, K. and R. Bowman. 2003. Predation rate on artificial nests increases with

human housing density in suburban habitats. Ecography 26: 188-196.

U. S. Fish and Wildlife Service. 1999. South Florida multi-species recovery plan. Atlanta,

Georgia. 2172 pp.

Waite T. A. and T. C. Grubb, Jr. 1987. Dominance, foraging and predation risk in the

32

Tufted Titmouse. The Condor 89: 936-940.

Walters, J. 1990. Anti-predatory behavior of Lapwings: field evidence of

discriminative abilities. Wilson Bulletin 102(1): 49-70.

Woolfenden, G. E. and J. W. Fitzpatrick. 1984. The Florida Scrub Jay: demography of a

cooperative breeding bird. Princeton University Press, Princeton, New Jersey.

Woolfenden, G. E. and J. W. Fitzpatrick. 1990. The Florida Scrub Jay: a synopsis after 18

years of study. In: Cooperative breeding in birds: long-term studies in ecology and

behavior. Ed. P.B. Stacey and W.D. Koenig. Cambridge University Press.

Woolfenden, G. E. and J. W. Fitzpatrick. 1991. Florida Scrub Jay ecology and

conservation. Pages 542-565 in Bird population studies (C.M. Perrins, J.D. Lebreton, and

G.J.M. Hirons, Eds.) Oxford University Press, New York.

Woolfenden, G. E. and J. W. Fitzpatrick. 1996. Florida scrub jay. The Birds of North

America No. 228: 1-26.

33

FIGURE LEGENDS

Figure 1. Five major sub-regions of the Florida Scrub-Jay including the three “core

populations” pertinent to this study. They include the Atlantic Coast, Ocala, and Lake

Wales Ridge sub-regions. The white stars indicate the study subpopulation locations

within each sub-region. Adapted by K. Russel (unpubl. data 2005) from Fitzpatrick et al.

1994 cited in USFWS 1999.

Figure 2. Diagram of the sound recording equipment and setup. Model from K. Russel

(unpubl. data 2005).

Figure 3. Duration of the call was measured from the waveform while max frequency

was taken from the spectrogram. a. Waveform and spectrogram figure from a family

group at CA responding to the cat treatment. b. Waveform and spectrogram figure from

a family at LL responding to the live snake treatment. c. Waveform and spectrogram

figure from a family at LY responding to the artificial snake treatment.

Figure 4. Mean max frequency and call duration to the cat treatment between

subpopulations. The different letters (a,b,c) indicate significance between subpopulations.

Figure 5. Mean max frequency and call duration to the artificial snake treatment between

subpopulations. The different letters (a,b,c) indicate significance between subpopulations.

34

Figure 6. Mean max frequency and call duration to the live snake treatment between

populations. There were no significant differences between subpopulations; hence the

letter “a” was applied for all three sites.

Figure 7. Mean max frequency and call duration between stimuli within CANA/MINWR

subpopulation. There were no significant differences between treatments; hence the letter

“a” was applied for all three treatments.

Figure 8. Mean max frequency and call duration between stimuli within Leisure Lakes

subpopulation. The different letters (a,b,c) indicate significance between treatments.

Figure 9. Mean max frequency and call duration between stimuli within Lyonia Preserve

subpopulation. The different letters (a,b,c) indicate significance between treatments.

Figure 10. Aerial photographs revealing landscape patterns in the three study sites with

the FLSJ families studied outlined in black. CANA/MINWR is the most rural, and study

families are located on land managed by the Fish and Wildlife Service, National Park

Service, and Kennedy Space Center. Leisure Lakes is a semi-developed residential

community adjacent to a state park with managed scrub habitat, and is therefore ‘semi-

rural.’ Lyonia Preserve, a 360 acre restored scrub ecosystem and managed scrub habitat,

is located within the City of Deltona and is surrounded by residential communities. The

scale for the zoomed in aerial photo on the right is 1 in to 200 m.

35

FIGURES

Figure 1.

36

Figure 2.

37

a.

b.

Figure 3.

38

c.

Figure 3.

39

CAT TREATMENT

a

b

a

2500

2700

2900

3100

3300

3500

3700

3900

CA LL LY

Subpopulation

Ma

xim

um

Fre

qu

en

cy

(k

Hz)

CAT TREATMENT

c

b

a

0.2

0.22

0.24

0.26

0.28

0.3

0.32

0.34

0.36

0.38

0.4

CA LL LY

Subpopulation

Ca

ll D

ura

tio

n (

se

co

nd

s)

Figure 4.

40

ARTIFICIAL SNAKE TREATMENT

bb

a

2500

2700

2900

3100

3300

3500

3700

3900

CA LL LY

Subpopulation

Ma

xim

um

Fre

qu

en

cy

(k

Hz)

ARTIFICIAL SNAKE TREATMENT

b

aa

0.2

0.22

0.24

0.26

0.28

0.3

0.32

0.34

0.36

0.38

0.4

CA LL LY

Subpopulation

Ca

ll D

ura

tio

n (

se

co

nd

s)

Figure 5.

41

LIVE SNAKE TREATMENT

a

aa

2500

2700

2900

3100

3300

3500

3700

3900

CA LL LY

Subpopulation

Ma

xim

um

Fre

qu

en

cy

(k

Hz)

LIVE SNAKE TREATMENT

aa

a

0.2

0.22

0.24

0.26

0.28

0.3

0.32

0.34

0.36

0.38

0.4

CA LL LY

Subpopulation

Ca

ll D

ura

tio

n (

se

co

nd

s)

Figure 6.

42

TREATMENTS WITHIN CANA/MINWR

aa

a

2500

2700

2900

3100

3300

3500

3700

3900

Cat Artificial snake Live snake

Treatment

Ma

xim

um

Fre

qu

en

cy (k

Hz)

TREATMENTS WITHIN CANA/MINWR

aa

a

0.2

0.22

0.24

0.26

0.28

0.3

0.32

0.34

0.36

0.38

0.4

Cat Artificial snake Live snake

Treatment

Ca

ll D

ura

tio

n (s

eco

nd

s)

Figure 7.

43

TREATMENTS WITHIN LEISURE LAKES

c

b

a

2500

2700

2900

3100

3300

3500

3700

3900

Cat Artificial snake Live snake

Treatment

Ma

xim

um

Fre

qu

en

cy (k

Hz)

TREATMENTS WITHIN LEISURE LAKES

b

b

a

0.2

0.22

0.24

0.26

0.28

0.3

0.32

0.34

0.36

0.38

0.4

Cat Artificial snake Live snake

Treatment

Ca

ll D

ura

tio

n (s

eco

nd

s)

Figure 8.

44

TREATMENTS WITHIN LYONIA PRESERVE

aaa

2500

2700

2900

3100

3300

3500

3700

3900

Cat Artificial snake Live snake

Treatment

Ma

xim

um

Fre

qu

en

cy (k

Hz)

TREATMENTS WITHIN LYONIA PRESERVE

bb

a

0.2

0.22

0.24

0.26

0.28

0.3

0.32

0.34

0.36

0.38

0.4

Cat Artificial snake Live snake

Treatment

Ca

ll D

ura

tio

n (s

eco

nd

s)

Figure 9.

45

CANA/MINWR

LEISURE LAKES

Figure 10.

46

Figure 10.

LYONIA PRESERVE