Bardin 2005

GROUP CHARACTERISTICS AND SOCIAL AFFILIATION PATTERNS OF BOTTLENOSE DOLPHINS (TURSIOPS TRUNACTUS) INHABITING THE INDIAN RIVER LAGOON, FLORIDA

_______________

A Thesis Presented to the Faculty of San Diego State University _______________

In Partial Fulfillment of the Requirements for the Degree Master of Science in Interdisciplinary Studies: Animal Behavior _______________

by Erin Elizabeth Bardin Summer 2005

SAN DIEGO STATE UNIVERSITY

The Undersigned Faculty Committee Approves the Thesis of Erin Elizabeth Bardin:

Group Characteristics and Social Affiliation Patterns of Bottlenose Dolphins (Tursiops truncatus) Inhabiting the Indian River Lagoon, Florida

_____________________________________________ Richard H. Defran, Chair Department of Psychology

_____________________________________________ Scott C. Roesch Department of Psychology

_____________________________________________ David W. Weller Department of Biology and Department of Psychology

______________________________ Approval Date

iii

Copyright © 2005 by Erin Elizabeth Bardin All Rights Reserved

iv

DEDICATION I dedicate this thesis to my parents, Robert and Carole Bardin, whose unconditional love, faith, and support have made my passage through graduate school possible. I cannot thank you both for giving me the courage and confidence to weather this and many other storms.

v

ABSTRACT OF THE THESIS Group Characteristics and Social Affiliation Patterns of Bottlenose Dolphins (Tursiops truncatus) Inhabiting the Indian River Lagoon, Florida by Erin Elizabeth Bardin Master of Science in Interdisciplinary Studies: Animal Behavior San Diego State University, 2005 A long-term investigation of the group characteristics and social affiliation patterns of Atlantic coast bottlenose dolphins was carried out in the Indian River Lagoon (IRL), Florida. Dolphin group characteristics and photo-identification data were collected during 477 boatbased surveys conducted in the IRL from August, 1996 to June, 2004. Survey time totaled over 1874 h, during which 615 h were spent in direct observation of dolphins. A total of 2,132 groups were encountered and photographed over the study period. Group sizes ranged from 1 to 40, with an average overall group size of 4.1 (SD = 3.43, median = 3). Group size was higher a) between 1996 – 1998 (mean = 5.9) than between 1999 – 2004 (mean = 4.0), and b) for sightings in the afternoon (mean = 4.4) than in the morning (mean = 3.7). Unlike some other inshore populations, which have larger groups in the fall and winter months, the size of IRL dolphin groups did not vary by season. Group size overall (mean = 4.1) was small, even compared to other inshore bottlenose dolphin populations. Together, these group effects may indicate that prey characteristics vary diurnally as well as annually but are relatively uniform within years. Further, the size of groups containing at least one calf (mean = 5.5) was larger than groups without calves (mean = 2.4) suggesting that mothers often rely on the assistance of conspecifics in calf-rearing. Dolphins with nine or more sightings were examined for evidence of site fidelity. Although the majority of these dolphins exhibited a high degree of site fidelity to the IRL system as a whole, dolphins with high numbers of sightings who were encountered in the Mosquito Lagoon were rarely encountered elsewhere. Overall, half-weight Coefficients of Association (COA) indices for these frequently sighted dolphins ranged from 0.09 to 0.83 (maximum values), showed that they preferentially associate and that associations were higher within sex class than between sex class. The larger range in maximum COA values that males (0.10 – 0.83) exhibit over females (0.00 – 0.33) may indicate that some males, as they do in other areas, form longer-lasting bonds while females have several casual associates. Mothers and identified calves exhibited the strongest associations, with some associations lasting over three years. In general, these results reflect similar patterns of social structure to those observed in other residential inshore populations, including individuals who show a frequent change in group membership but maintain some long-term stable companionships.

vi

TABLE OF CONTENTS PAGE ABSTRACT...............................................................................................................................v LIST OF TABLES................................................................................................................... ix LIST OF FIGURES ...................................................................................................................x ACKNOWLEDGEMENTS..................................................................................................... xi INTRODUCTION .....................................................................................................................1 Mammalian Social Systems............................................................................................. 5 Cetacean Social Systems.................................................................................................. 8 Bottlenose Dolphin Social Systems ............................................................................... 10 Investigation of Social Systems ..................................................................................... 12 Social Affiliation Analysis............................................................................................. 13 Half-Weight Index ................................................................................................ 14 Twice-Weight Index ............................................................................................. 14 Simple Ratio Index ............................................................................................... 15 Methodological Issues ................................................................................................... 16 Importance of Studying Social Systems ........................................................................ 17 Habitat Structure ............................................................................................................ 18 Southeast Atlantic Bottlenose Dolphin Research .......................................................... 19 Population Structure of the IRL Dolphins ..................................................................... 20 Harbor Branch Oceanographic Institution Photo-Identification Program .................... 21 METHODS ..............................................................................................................................23 Survey Coverage............................................................................................................ 23 1996 to 2001 ......................................................................................................... 23 2002 to 2004 ......................................................................................................... 25 Photographic Survey Procedure..................................................................................... 25 Sorting, Matching and Cataloging ................................................................................. 26 Data Analysis ................................................................................................................. 27

vii Encounter Rates, Sighting Frequencies, and Group Characteristics..................... 27 Site Fidelity........................................................................................................... 29 Social Affiliation................................................................................................... 30 RESULTS ................................................................................................................................33 Group Size ..................................................................................................................... 33 Annual Variation................................................................................................... 35 Seasonal Variation ................................................................................................ 36 Regional Variation ................................................................................................ 39 Diurnal Variation .................................................................................................. 39 Shark Bite Scars.................................................................................................... 41 Group Composition........................................................................................................ 41 Seasonal Variation ................................................................................................ 41 Regional Variation ................................................................................................ 42 Site Fidelity and Residency ........................................................................................... 42 Social Affiliation............................................................................................................ 44 Female and Male Associations ............................................................................. 46 Mothers and Calves............................................................................................... 48 DISCUSSION ..........................................................................................................................50 Group Size ..................................................................................................................... 50 Habitat Structure ................................................................................................... 51 Resource Distribution ........................................................................................... 53 Predation Pressure................................................................................................. 56 Annual Effects ...................................................................................................... 57 Seasonal Effects .................................................................................................... 58 Diurnal Effects ...................................................................................................... 59 Site Fidelity and Residency ........................................................................................... 62 Social Affiliation............................................................................................................ 63 Male-Male Associations ....................................................................................... 65 Female-Female Associations ................................................................................ 66 Mother-Calf Associations ..................................................................................... 66 Environmental Considerations....................................................................................... 67 REFERENCES ........................................................................................................................69

viii APPENDIX SITE FIDELITY OF DOLPHINS SIGHTED NINE OR MORE TIMES ........................78

ix

LIST OF TABLES PAGE Table 1. Summary Information on Annual and Seasonal Survey Effort, Encounter Rate, Sighting Data, and Group Size. ..........................................................................34 Table 2. Descriptive Statistics for Dolphin Group Size by Study Year...................................35 Table 3. Mean Group Size by Sea Surface Temperature Range..............................................37 Table 4. Summary Information on Group Size by Day Period................................................40 Table 5. Summary Statistics for Group Composition by Season. Numbers in Parentheses Represent SD............................................................................................41 Table 6. Mean, SD, and CV of Real and Random Association Indices and P-values for Overall Study Period Using Daily, Monthly, and Yearly Sampling Periods.........45 Table 7. CV of Real Association Indices, CV of Random Association Indices, P-value of Permutation Test Using a Daily Sampling Period...................................................46 Table 8. Mantel Test t Statistics, P-Values, and Matrix Correlation Coefficients (r) for Between and Within Sex Class Associations...............................................................47 Table 9. Mean Group Size and Group Size Definitions from 14 Studies on Bottlenose Dolphins.......................................................................................................................52

x

LIST OF FIGURES PAGE Figure 1. Indian River Lagoon, Florida. ..................................................................................24 Figure 2. Study area regions. ...................................................................................................29 Figure 3. Frequency distribution of group size........................................................................34 Figure 4. Mean group size by study year................................................................................35 Figure 5. Frequency distribution of sea surface water temperatures. ......................................37 Figure 6. Recorded sea surface temperatures and dolphin group size for 2003. .....................38 Figure 7. Scatter plot of latitude on group size........................................................................39 Figure 8. Mean group size by day period. ...............................................................................40 Figure 9. Group size by study region.......................................................................................43 Figure 10. Discovery curve for dolphins identified in the IRL during on-surveys..................43 Figure 11. Frequency distribution of maximum association levels for dolphins identifited ≥ 9 times in the IRL....................................................................................45 Figure 12. Sociogram of significant sexed dyads from 2003. .................................................47 Figure 13. Maximum half-weight COA levels for dolphins identified at least nine times. Dashes in x-axis denote line breaks. .................................................................48 Figure 14. Coefficients of association between three mother-calf pairs by study year. Calves are listed first....................................................................................................49

xi

ACKNOWLEDGEMENTS Writing this thesis has been an adventure for which I am indebted to many people. First and foremost, I owe a tremendous amount of thanks to my advisor extraordinaire, R.H. Defran, for teaching me how to think like a scientist and write a good story. His mentorship has profoundly influenced my career path. I would also like to thank my other two committee members: Dave Weller, for his insightful reviews and warm words, and Scott Roesch both for taking a leap on a dolphin project and for making multivariate statistics palatable.This research would not have been possible without the generous donation of data from Harbor Branch Oceanographic Institution in Fort Pierce, Florida. I am deeply grateful to Marilyn Mazzoil and her thousands of hours following dolphins. No one knows those IRL dolphins like she does. I also thank Steve McCulloch for his immense contribution to the research and for securing me financial support. I give thanks and admiration to Elizabeth Murdoch for her sharp skills with a camera lens and Photoshop. A final thank you goes to Elisabeth Howells, Sarah Bechtel, and the many HBOI interns whose dedication to this work has made it an even more significant achievement. To CBL members, past and present, thank you for laughing your way through this with me! We’ve had quite a journey in that dark little lab. I give special thanks to Brittany Hancock and Jennifer Alongi for navigating the Socprog labyrinth with me. Another thank you should go to CBL alumnae for giving me research guidance: Karen Baker, Marthajane Caldwell, Kim Dudzik, and Aimee Lang. I also thank Dr. Hal Whitehead and Cindy Rogers for their invaluable advice on Socprog. I am additionally indebted to Charles Hurley and Linda Kilroy for helping me format and review my thesis. I thank all my friends who have put up with me these past years! Your questions about my thesis have been very thoughtful. I especially thank Helen Hong and Katie Nesmith for helping me review and edit my thesis. Most of all, I would like to thank my family for their support, and to Mike for his tough but never-ending love. Thank you for never letting me give up!

1

INTRODUCTION The purpose of this research project was to examine the social dynamics of the population of common bottlenose dolphins (Tursiops truncatus) that inhabit the Indian River Lagoon (IRL), Florida. Information on group characteristics, such as size and composition, as well as association patterns was investigated. Photo-identification methods were used to identify IRL dolphins, and long-term sighting histories of known individuals were used to examine association patterns. Comparisons of social affiliation patterns were made across several dimensions, including habitat type, season, and day period. Association patterns were compared to bottlenose dolphin populations from differing ecosystems. The study of mammalian social systems elevates the biological scale of investigation from the level of individual to the level of group. Interactions among group members, rather than individuals, become the focus of study. When attempting to answer questions about the social factors that affect a mammalian species, biologists often analyze association patterns. Social mechanisms that influence mammalian life histories vary across species and populations and are based on evolutionary adaptations to changes in the environment. Selective advantages for forming groups can include protection from predation, feeding efficiency, information sharing, mating facilitation, and energy maximization (Grier 1984). How much influence any one or combination of these factors exerts on a species is relative to the sum of the influence of all the factors, and these may contribute to either the individual’s fitness or the inclusive fitness of the group. Ecological factors that shape social systems in mammals include availability and distribution of resources, habitat characteristics, and predation pressure (Crook 1976). These in turn interrelate with species parameters, such as group size and composition and social affiliation (Caldwell 2001). Both terrestrial and aquatic mammals exhibit wide differences in the way their behavioral characteristics interact with environmental factors. These same variables also express themselves in substantial inter-population differences among the same species.

2 Among cetaceans, species tend to mirror certain qualities of the habitats they live in. Thus, species-specific social organization often depends on phylogenetic history and the environmental context of a given population. For example, many baleen whales tend to form dispersed groups, and they are often found alone. Although the apparent dispersion or isolation may be a result of a preference for independence, an alternative explanation is that due to their large size and capability for long-distance sound transmission, some groups may indeed be considered tightly packed despite significant spatial separation (Norris and Dohl 1980). Mysticete species range from very isolated, as seen in gray whales (Eschrichtius robustus), to very gregarious, as seen in humpback whales (Megaptera novaeangliae), although these generalizations to do not always hold true. Odontocete social dynamics and behavior have been widely studied, with special attention paid to killer whales (Orcinus orca) and bottlenose dolphins due to their sophisticated grouping strategies and relatively easy viewing access. Resident killer whales organize their pods along matrilineal lines, maintain very structured and stable group membership, and often remain in their natal pod for life (Baird 2000). Some researchers have found that bottlenose dolphins tend to form fission-fusion societies with present but low average associations among members (Bearzi 1987, Marsh 2000, Wells et al. 1987). The methods used to study odontocete social organization include social affiliation analysis, measures of group size and composition, genetic analysis, and acoustic analysis. In combination, these methodologies sometimes elucidate the existence of distinct communities with overlapping distributions. This study will focus on the first two methods, but attention will be paid to related studies that utilized genetics and acoustics. Social affiliation is a measure of how many times two or more animals are sighted together at the same time within a defined proximity (Ginsberg 1992). Coefficients of association are generated for each member of the group in question, and association trend analysis is often performed to examine how groups organize themselves. Environmental factors are frequently studied to determine if there are ecological correlates with association patterns. Group composition analysis, when paired with social affiliation analysis, can reveal how groups are delineated, by age, sex, or reproductive status, if such biological information about the population studied is available.

3 Several studies on dolphin social structure have incorporated genetic sampling to verify group member identity and uncover any genotypic differences among groups (Wells 1986; Caldwell 2001). Significant differences in the genetic structure of mtDNA have been found among communities that had no apparent geographic boundaries (Wells 1986; Caldwell 2001). Genetic variation within a population allows adaptation to changing environmental conditions. Many cetacean species produce sounds believed to facilitate group communication. Sound transmits rapidly underwater and over far distances. Thus, vocal communication is often the most efficient way of transmitting information over a large area. Vocal communication is believed to facilitate social organization, sometimes in a very complex way, for a number of cetacean species (Tyack 2000). Each methodology is adjusted to fit the life history of the population studied. Social affiliation research is often constrained by definitions of group and association chosen by researchers, with inevitable differences among studies (Norris and Dohl 1980). Data collection always has limitations, but systematic effort minimizes error in parameter estimates. Social affiliation analysis can provide information on the manner, frequency, and duration of aggregation among individuals. Defining these trends helps researchers to further depict the social structure of a population, as well as develop hypotheses about the selective advantages of grouping strategies. As top-level predators, bottlenose dolphins serve as important macro-indicators of the health and status of the systems they inhabit. Environmental features of a habitat may influence the way a population adapts to that habitat. For dolphins, the waters that a particular population inhabits may be geographically bounded in ways that would influence tides, prey distribution, water temperature and salinity. These environmental parameters, in turn, may influence population characteristics, such as home range, occurrence, distribution, and social structure. Differences in habitat features need to be considered when comparing differences in species characteristics between any two populations. The degree of a population’s openness may be described in geographical or demographical terms. A population that inhabits an enclosed or mostly enclosed waterway is considered geographically closed, such as is the case for the IRL dolphins. Conversely, a

4 population found in unbounded waters, like dolphins occurring along the California coast, is termed geographically open. Structurally complex habitats, which are more often inshore and tend to be more closed, may also provide more areas of refuge, protection from predation, and prey habitats. Dolphins found within these habitats tend to form small groups with strong associations, exhibit strong site fidelity, and use small home ranges (Caldwell 2001, Shane et al. 1986). Group size has been found to increase as habitats increase in their degree of openness across bottlenose dolphin populations (Gygax 2002). Regardless of geographical constraints, the degree that a population is demographically closed or open is a product of the genetic group’s stability and how much genetic transfer is present. Distinct populations with distinguishable mtDNA are sometimes found in areas where no physical boundaries exist that would otherwise separate the populations (Wells 1986, Caldwell 2001). These distinct genetic lines are probably preserved through inbreeding within communities. Delphinids often form fission-fusion societies: individuals aggregate but group membership changes frequently (Norris and Dohl 1980). Bräger et. al (1999) studied Hector’s dolphins (Cephalorhynchus hectori) off South Island, New Zealand and found that individuals did not associate significantly more than expected by random chance. Social structure in delphinids ranges from extreme stability, as seen in resident killer whales (Baird 2000), to extreme fluidity, evident in spinner dolphins (Stenella longirostris) (Norris et al. 1994). Bottlenose dolphins are found in temperate and tropical oceans throughout the world. Studies on the social systems of bottlenose dolphins are numerous and span many of the coastal and inshore waters of the world (Kino Bay, Mexico: Ballance 1990; Northern Adriatic Sea: Bearzi et al. 1997; Jacksonville, Florida: Caldwell 2001; Gulf of Guayaquil, Ecuador: Félix 1997; Bahamas: Rogers et al. 2004; Southern California Bight: Weller 1991; Sarasota Bay: Wells 1986). The evolution of various ecotypes in differing habitats suggests considerable phenotypic plasticity in response to changing environmental conditions. Wells (1986) found distinctive “groups” that interacted more internally than across groups in dolphins living in Sarasota Bay, Florida, and his research argued for the identification of “communities” within a broader population. Differences in associations among groups were greater than within groups, although the communities’ home ranges

5 overlapped and members from different groups associated on occasion. Thus, the argument for a separation of communities within the population is seemingly supported. The Indian River Lagoon, unlike the central west coast of Florida, is a narrow waterway open to the Atlantic Ocean at only five widely spread inlets, ranging in width between 0.93 km to 9.3 km (Figure 1). The majority of movement within the IRL is thus restricted in north and south directions. Therefore, for subgroups moving in opposite directions in many parts of the IRL, interactions between individuals are often inevitable. The physical dimensions of the IRL make it an almost entirely geographically closed region. If segregation among groups is present, it may occur along latitudinal segments of the IRL system. Whether or not the community-based population structure described by Wells (1986) will apply to the Indian River Lagoon population is unknown. Current research proposed on the IRL bottlenose dolphins has provided social and ecological information to help guide in the conservation of the species and the ecosystem as a whole. Use of the IRL for a variety of commercial and recreational functions has increased potential threats to the viability and sustainability of several species (Mazzoil et al. 2002). Identifying population patterns of bottlenose dolphins found in the lagoon will allow more informed management decisions to be made.

MAMMALIAN SOCIAL SYSTEMS Mammalian social systems have long been studied in both terrestrial and aquatic environments. Environments vary widely resulting in different behavioral responses at the group level. Species and even populations exhibit different social strategies related to the circumstances under which they live. Biologically intrinsic drives shape survival strategies in species. Risk of predation and resource availability vary widely from habitat to habitat, contributing to the development of socialization mechanisms. Flocking or herding usually occurs for populations with inconsistent resource distribution and without opportunities for concealment, although the mannerisms are often complex and vary (Gaskin 1982). Forces which simultaneously attract and repel group members can be described mathematically and involve the influence of the group itself as well as each participating individual (Breder 1954). Group size, grouping frequency, group composition, as well as the temporal and spatial environmental correlates to grouping are all

6 dimensions that have been studied in social mammals. Crook et al. (1976) created a system of classifying mammalian social systems according to rearing, mating, grouping, and dispersion strategies. Along each dimension, environmental variables, species parameters, and social system variables can interact. Within social systems, male investment, malefemale bonds, and grouping characteristics such as size, stability, resource utilization, and range exclusivity contribute to the determination of social structure. Social species exhibit a variety of aggregating strategies to help locate resources. Aggregating can give individual group members a number of survival advantages. A group of carnivores can sometimes be more efficient at foraging and often at actual feeding than an individual by itself. Prey locating when paired with communication allows predators to increase their foraging range, which is especially useful when prey is patchily distributed (Weller 1991). Familiarization with individuals having overlapping home ranges also contributes to foraging efficiency (Crook et al. 1976). Certain species of primates form fission-fusion societies with dynamic socializing similar to dolphins. Group size and composition affect associations in pygmy chimpanzees (Pan paniscus). Only in large parties do male-female associations become more significant than associations among females (White 1992). Chimpanzees (Pan troglodytes) are also found either alone or in groups which frequently change size and composition. Forest chimpanzees in Kibale, Uganda, exhibit sex differences in home range, with males covering an area 1.5 to 2 times greater. Results suggest this sex difference could be explained by female preference of community core areas or an avoidance of other communities (Chapman and Wrangham 1993). Chimpanzees exhibit wide variation in social structure and association as a result of group composition and size, which may be dependent on habitat ecology. Squirrel monkeys (Saimiri spp.) also exhibit social behavior differences among populations that appear to be resource dependent. When food is scarce, monkeys will tolerate close individual distances and body contact if it enables more efficient foraging, even in groups that did not engage in play (Baldwin and Baldwin 1973). African elephants (Loxodonta africana) demonstrate long migratory capability and adaptability to a variety of climates and habitat types, from deserts to dense forests. Like some delphinid species, such as killer whales, elephants live in matriarchal groups and exhibit strong maternal philopatry. Nyakaana and colleagues (2002) found significant levels

7 of genetic subdivision among regionally differentiated groups. At the head of each group was a matriarch member, usually the oldest female and mother to many of the group members. Evidence of the superior ability of matriarchs to distinguish between familiar and unfamiliar conspecific vocal calls suggests that the matriarch integrates a large amount of external information for each herd and is therefore a keystone member (McComb et al. 2001). It is hypothesized that killer whale matrilineal pods only dissolve when matriarchs die (Baird 2000). Carnivores respond behaviorally to both prey distribution and habitat complexity. Lions (Panthera leo) exhibit drastic sexual differences in the way they forage and also in the way they socialize. Male lions tend to associate with females more in open habitats than in structurally complex habitats (Funston et al. 1998). Females come into oestrus together, give birth synchronously, and raise and nurse their young communally (Van Orsdol et al. 1985). Maturing males leave their natal pride, and like dolphins, many form coalitions, which can aid in hunting success. Predation pressure shapes social systems in many mammals. Dolphin social structure has often been compared to that of ungulates for several reasons, primarily the similarities in environment and grouping behavior (Gaskin 1982). Flower (1883) was the first to suggest that cetaceans are closely related to the order Artiodactyla (the even-toed ungulates) based on gross-anatomy. Further studies on comparative blood chemistry and hormones have supported this claim (Reynolds et al. 2003). The three-dimensional realm that dolphins inhabit can be viewed as resembling the open plains that many ungulates subsist on, and neither habitat offers much opportunity for concealment from predators. The mannerisms with which ungulates aggregate to increase group vigilance also parallel the way several odontocete species form groups. For example, Hunter and Skinner (1998) found conclusive evidence that both wildebeest (Conochaetes taurinus) and impala (Aepyceros melampus) increase rate of looking and proportion of time spent looking, two indicators of vigilance, when predation pressure increased. For populations where food resources are not readily available and predation presents risk, grouping mechanisms must balance the costs of obtaining energy with the costs of avoiding predation. This paradox contributes to complex social patterns which are often context-specific. Associations may be beneficial in certain circumstances, such as foraging

8 for a particular type of resource or in a particular type of way. Pitcher et al. (1982) showed experimentally that food locating times decreased significantly with larger shoal size in goldfish (Carassius auratus) and minnows (Phoxinus phoxinus). Cooperative foraging has been shown in a variety of cetaceans (killer whales, Baird 2000; pilot whales, bottlenose dolphins and dusky dolphins, Würsig 1986). Alternatively, avoiding association with particular individuals, or altogether, may also have its advantages, either with respect to foraging success or escaping predators. Sperm whales (Physeter macrocephalus Linnaeus) off the coast of the Galápagos Islands forage in spaced-out rank formation, perpendicular to their travel direction. One benefit of such spacing may be decreased interference of foraging by neighboring conspecifics (Whitehead 1989).

CETACEAN SOCIAL SYSTEMS For the past forty or so years, cetacean social structure has been studied in great detail. Researchers have employed a number of different data collection and analysis methods, although boat-based observations continue to be the most prevalent. While the social structure of odontocetes has been given more attention than that of mysticete, several studies on baleen whales have contributed greatly to our understanding of their behavioral ecology. Gray whales and humpback whales migrate the longest distances of any mammal between feeding and breeding sites, which may be a result of great variation in food distribution. Eastern-Pacific grays feed primarily on amphipods and other invertebrates found in the mud in the Bering and Chucki Seas, and the colder, nutrient-dense arctic waters provide an ample food source. Giving birth in lagoons along Baja California, Mexico in the winter allows gray whale calves time to nurse and develop the thick blubber layer they will need to insulate themselves from the cold in the north. On feeding grounds during the summer, humpback mother-calf pairs are often sighted alone, and in general groups are small and unstable. Humpbacks found in aggregation tend to be foraging for sand lances or breeding (Clapham 2000). During the winter, humpbacks travel to breeding grounds in lower latitudes. Behaviors associated with socializing, like aerial behaviors, are much more common on breeding grounds (Whitehead 1985). Associations among members are brief and fluid, and lone females are frequently sighted in

9 escort by single males, as well as with females who are still lactating. Humpback whale songs, sung only by males, have the potential to travel hundreds of kilometers, and may function in mate-finding. Since humpbacks often associate with unrelated conspecifics, social groups may be based more on reciprocal altruism than kinship selection, with the exception of the mother-calf bond (Valsecchi 2002). Sperm whales exhibit the most extreme case of sexual dimorphism among cetaceans, with males being up to 1.5 times the length and three times the weight of females (Connor et al. 1998). Females are very social and exhibit wide variation in levels of organization: small units of about ten animals, stable groups of about 20 animals, temporary aggregations, and geographic concentrations (Whitehead and Weilgart 2000). Females care for each other’s calves by babysitting while others dive; if left alone, calves would possibly be at greater risk for predation. The depths at which sperm whales forage require long foraging bouts; staggering dives may offer a selective solution. After dispersing from their natal group, males form loose bonds with other males in bachelor groups. Although an organized social structure has not been readily defined, breeding males seem to avoid one another more than expected by chance (Whitehead 1993). Within the family delphinidae, variation in social structure is wide. Longitudinal studies of resident killer whales of the eastern North Pacific Ocean have found strongly defined matrilineal pods within two populations. Both males and females exhibit natal philopatry, which has not been documented in any other mammalian species (Baird 2000) except pilot whales (Globicephala melas) (Amos et al. 1993). Average pod size of both populations combined was found to be 12 (range 3-59) (Bigg et al. 1990). Transients are found in consistently smaller groups, and a group size of three seems to be optimal (Connor et al. 1998). Among killer whales, residents (which feed on fish) and transients (which prey primarily on marine mammals), exhibit cooperative feeding, which is an expected benefit of forming social groups. Prey location and capture may be more efficient in groups. Although there are no findings of predation on killer whales, solitary whales may be vulnerable to attacks by other killer whale groups. The reproductive costs of staying within the natal group do not appear to be high, since members seem to always mate outside their natal pod. Thus, selection seems to favor the formation of strong matrilineal bonds in killer whales.

10

BOTTLENOSE DOLPHIN SOCIAL SYSTEMS Bottlenose dolphins inhabit temperate and tropical systems worldwide, and are found in pelagic, coastal, and estuarine systems. They are the most widely recognized and most studied cetacean, given their history of associations with humans and their accessibility (Connor et al. 2000b). Populations of bottlenose dolphins demonstrate great morphological and behavioral variability, especially prevalent in prey selection and foraging techniques. A commonality of many well-studied populations is complex social grouping, in which associations are dynamic and fleeting. Dubbed fission-fusion societies, this type of socializing is also prevalent in higher order mammals like elephants (McComb 2001), chimpanzees (Chapman and Wrangham 1993), and lions (Van Orsdol 1985). Group living can provide numerous selective advantages for individual dolphins, in terms of increased foraging efficiency, protection from predation, transmission of information (Grier 1984) and facilitation of mating and rearing (Crook 1976). Dolphin aggregations reflect the available prey distributions as well as the existence of potential cooperative foraging techniques. The presence of conspecifics provides a form of concealment from potential predators, reduces the probability of individual capture, and increases collective vigilance (Crook 1976). Information transfer, through auditory, visual, tactile, and chemical signals, may be propagated more efficiently through group members who frequently change associates. In addition to receiving information about their environment from their own echolocation, dolphins are also thought to receive feedback signals from conspecific echolocation (Tayler and Saayman 1972), which further reinforces the advantages of group membership. Grouping according to age, sex, and reproductive status may also facilitate finding mates and calfrearing partners. Determining what bottlenose dolphins actually eat is difficult, and most findings come from stomach content analysis from stranded animals or those caught incidentally in fishing nets. Prey type information derived from stranded animals may not be the most accurate or representative of the general population (Barros and Odell 1990), although more accounts exist for stranded than by-caught animals. A wide variety of fish, cephalopods, and shrimp has been found in bottlenose dolphin stomachs. Barros and Odell (1990) identified and quantified stomach contents of 234 dolphins that stranded along the southeastern US,

11 reported from the Marine Mammal Stranding Network. Seventy-six stomachs were available for full analysis. Almost all of the individuals had consumed fish, and in over one-third, cephalopods had also been consumed. Most stomach contents revealed that the animals did not have sufficient food (3-6% of body weight) to maintain homeostasis. Although it is difficult to generalize the diet of stranded animals, which may have had compromised health, to the general population they represent, it is easy to argue that dolphins in the southeast Atlantic consume a wide variety of prey items and probably prefer fish to other prey types. Dolphins also exhibit different foraging methods, including cooperative strategies. There are several documented instances of dolphins schooling prey into a circle and taking turns feeding (Leatherwood 1975). Less conspicuous or more indirect examples of cooperative foraging probably exist and these mechanisms need to be further explored. Populations of bottlenose dolphins worldwide vary in their exposure to predation risk and resource availability, and these two ecological factors are important determinants of the grouping strategy dolphins employ. In certain areas, little or no predation has been documented, such as Turneff Atoll Lagoon, Belize (Campbell et al. 2002), and Moray Firth, Scotland (Wilson et al. 1997). The absence of crescent-shaped scars on individuals, which would indicate shark attacks, suggests minimal predation on these populations. Other potential threats to these populations, like interspecific or conspecific competition, disease and parasitism, as well as anthropogenic dangers, probably influence grouping strategies in several complex ways. Several populations have well-documented exposure to shark predation, most notably Sarasota, Florida (Wells et al. 1987) and Shark Bay, Australia (Connor and Heithaus 1996). In Sarasota, roughly one-fifth of the adult dolphins captured from 1975 to 1985 bore shark bite scars. None of the calves exhibited shark scarring, which is probably attributable to low survival rates for shark-bitten calves (Wells et al. 1987). Nursery groups, which consist of mother-calf pairs and other mature females and are seen in all dolphin populations, provide added vigilance against possible predators. Dolphins group according to sex and age. Numerous studies of free-ranging cetaceans have documented mother-calf groups that consist of several mother-calf pairs and juveniles (Wells 1987; Mann and Smuts 1999), and others have documented male-male coalitions (Smolker et al. 1992; Maze-Foley and Würsig 2002; Parson et al. 2003). Long-term studies

12 of well-known dolphins at Monkey Mia, Australia, have documented shared calf-rearing and protection among mothers and associated females. Tayler & Saayman (1972) found the formation of nursery groups in captive colonies in which females would take over swimming in echelon position with the calves of other females while mothers either rested or were fed. Calves stay with their mothers and continue to nurse for a variable length of time, often from three to seven years, depending on the population and offspring health (Connor et al. 2000b). The nutritional dependence of dolphin calves on their mothers probably explains the persistence of this bond (Crook 1976). Female and calf distribution is thought to directly reflect resource availability. The determination of male distributions, however, involves mate-finding as well as prey locating, and therefore is likely to be directly related to female distribution (Christal 1998). This divergence in drives as a function of sex may reflect differences in calf vs. noncalf groups, in distribution, size, or association patterns. Several studies have documented the existence of male alliances, composed of dyads or trios, that have significantly higher levels of association than would be expected by chance (Parsons et al. 2003; Connor et al. 2002). Parsons et al. (2003) investigated genetic relatedness between males in alliance on Little Bahama Bank and found significant kin selection among the six alliances tested. In a parallel study, Möller et al. (2001) revealed a lack of kinship among identified Indo-Pacific bottlenose dolphins (Tursiops aduncus) male alliances in southeastern Austrailia. The discrepancy in kinship findings suggests that selection of alliance partners is probably based on several mechanisms that vary among populations.

INVESTIGATION OF SOCIAL SYSTEMS Studies of social structure in mammals have included various quantitative approaches. Group size, group composition, social affiliation, and home range data have all provided important information about the social ecology of different study populations. This paper will focus primarily on group size and composition, as well as association patterns, which is measured as the number of times pairs of dolphins are seen together. The findings will be integrated with known information about distribution and home range. Association patterns will be of particular interest because the persistence or lack of persistence in associations among individuals over time and space reflect ecological adaptations. Careful

13 statistical analysis of sighting histories may also reveal complex association trends undetectable by other methodologies. The most widely known bottlenose dolphin study for its duration and depth is Wells (1986) on the dolphins inhabiting Sarasota Bay. Using photo-identification methods, capture and release, behavioral focal follows, and genetic testing, Wells and his colleagues have studied bottlenose dolphins inhabiting the Gulf of Mexico off the coast of central western Florida since 1970 and have presented evidence for the existence of three distinct communities. Further analysis of the Sarasota community has indicated the specific characteristics of the way groups were socially structured. Sarasota dolphin groups are delineated by sex, age, and familial lines, and are also thought to maintain a promiscuous mating system.

SOCIAL AFFILIATION ANALYSIS Association with a group is often included in social analyses as a descriptive dimension of social structure. Association indices used to measure the probability that an animal within a population will be sighted with other animals have been generated in several different ways. These indices yield quantitative measures of relationship stability, based on the indicated operational definition of association, which Cairns and Schwager (1987) define as the frequency with which two individuals are present in the same social group at the same time. The index itself represents how frequently two animals occur simultaneously in the same area. Three binary measures commonly used are the half-weight index, the twiceweight index, and the simple index ratio. Care should be undertaken when applying indices to measure association. The association analyses presented here base themselves on presence/absence data. In the subsequent index descriptions, the following notation from Cairns and Schwager (1987), modified by Gowans (1999) will be used: X = number of observation periods both individual a and individual b are sighted in the same group; Yab = number of observation periods in which both a and b are identified but not grouped; Ya = number of observation periods in which only a is identified; and

14 Yb = number of observation periods in which only b is identified.

Half-Weight Index The half-weight index, which was presented by Dice (1945) and modified by Cairns and Schwager (1987), reflects the proportion of occurrences in which two individuals, a and b, spend together over the average number of times either individual or both are sighted. X / [X + Yab + ½(Ya + Yb )] Ginsberg and Young (1992) suggested that this index is more likely to overestimate levels of association. The denominator averages the total number of times either individual is seen, which leads to occurrences being double-counted. This index is most appropriate for pairs that are more likely to be sighted when separate than when together (Cairns and Schwager 1987). The half-weight index has been used in several bottlenose dolphin studies (Sarasota, Wells 1986; Cedar Keys, Quintana-Rizzo and Wells 2001). Wide home ranges and inconsistency of good sighting conditions can make sampling dolphins evenly difficult. These limiting factors can lead to low sighting rates, which makes the half-weight index appropriate in these circumstances.

Twice-Weight Index Another index commonly used in association studies is the twice-weight index, which double counts the number of times individuals are seen separately. X / [X + 2Yab + Ya + Yb] This index will yield a larger denominator and therefore smaller value than the half-weight. In situations where sampling favors sighting both individuals together, the twice-weight index is the most appropriate to use (Cairns and Schwager 1987). Circumstances in which it is hard to delineate boundaries of association versus separation also favor the selection of this index.

15

Simple Ratio Index The simple-ratio index neither deflates nor inflates the denominator and provides the most unbiased estimate of association. X / [X + Yab +Ya + Yb] Ginsberg and Young (1992) add that calculating the number of observations in which both individuals are seen but not grouped can be tedious, but modifications to the formula can make it less complicated. The simple ratio index is often used in studies because it is the least biased (Caldwell 2001; Rossbach and Herzing 1999). In association analysis, valid results are predicated on representative sampling. In bottlenose dolphin photo-identification studies, consistently capturing (photographically) all members of a group, not just a particular subset, helps to minimize bias in estimates of social structure, residency, and abundance. Association indices provide statistical averages of the stability of populations. In this sense, they offer insight into dynamics of social structure. The degree of how stable or fluid a population is may be correlated with different species-specific or environmental factors. Although populations of short-beaked common dolphins (Delphinis delphis) in the eastern Ionian Sea, for example, are often characterized as highly fluid, they also exhibit some degree of non-random association, which may indicate patchy prey distributions and low predation pressure (Bruno et al. 2004). Another methodology often used to study social ecology is behavioral observation. Researchers use defined ethograms, focal follows and various other sampling strategies to examine behavioral patterns and draw conclusions about populations. Behavioral studies on dolphins and other cetacean species have tended to study surface activities due to the difficulty in observing them underwater. A common pitfall of behavioral studies is the tendency to focus on conspicuous behaviors and assign function to them without evidence (Gaskin 1982). Aerial behavior, vocalizations, water-slapping, and body contact are often accepted as behavioral events that indicate the behavioral state of socializing because all of these activities are attention-getting, whether by visual, tactile, or acoustic means. These behaviors may be transferring some kind of information. There may be alternative

16 explanations for the purposes of these behaviors, however, besides communication with other conspecifics. For example, an isolated event of tail-slapping or leaping may function as a way to shed parasites or fend off a possible predation attack. Agreement among researchers as to the function of behavioral events is lacking, most likely because each event must be evaluated with respect to the context in which it occurred. Other forms of communication between individual dolphins that are not readily obvious to the researcher also may exist. Finally, socialization is not necessarily dependent on the existence of these behavioral events. In squirrel monkeys, Baldwin and Baldwin (1973) showed that social integration was not necessarily dependent on the occurrence of social play, although the opportunity to play provided a foundation for forming more complex social interactions. Unraveling the patterns of social life in dolphins is dependent on good behavioral analysis that is sensitive to these methodological issues.

METHODOLOGICAL ISSUES When attempting to draw conclusions about the ethology of a population, several assumptions must be made. Operating definitions must be given for each research unit in question, and these definitions are constrained by their accuracy and precision. Researchers must carefully consider the parameters they are estimating. In ethology, the relationship between the predictor and response variables may be confounded by uncontrolled factors. Natural experiments, however, have higher external validity over laboratory experiments since they operate on existing dynamics. Studying one species in a variety of environmental contexts resembles the standard experimental design of observing behavioral variances after exposing a population to differing environmental conditions (Baldwin and Baldwin 1973). Finding comparable differences between ecotypes of the same dolphin species is the general goal of the present research. In social affiliation analysis, the research unit of interest is the group (or school). How bottlenose dolphin groups are defined varies widely depending on the population and the researcher. Connor et al. (2000) reviewed 17 studies on bottlenose dolphins, all of which have different definitions for “group.” These variations arise from situational differences and many times, for the convenience of the researchers. One result of the wide spread in definitions, however, is the reduced ability to compare studies across time and space.

17 In addition to the importance of defining research units, research questions must also be asked with care. Studies addressing mammalian sociality often fail to identify and define the overarching phenomena they are attempting to measure, for example “social structure” or “social organization” (Christal 1998). Both are related and interdependent, but they are not equivalent concepts. Christal chose to include only “social structure,” using the framework as described by Hinde (1976), which partitions social structure into the aggregation of patterns in relationships which are based on repeated interactions. Sampling cetacean interactions is limited and therefore biased to surface viewing. For this reason, with respect to social affiliation analysis, interactions are substituted by associations based on defined proximity. Evidence of repeated associations under the criteria for group membership reinforces the validity of making the jump from associations to relationship to social structure.

IMPORTANCE OF STUDYING SOCIAL SYSTEMS Assumptions about underlying causation are often made when drawing conclusions about social affiliation. The primary purpose for gathering sighting histories and generating association indices is to further understand a population’s social structure. The fact that there are regional differences in the amount of relationship stability among individuals indicates that there may be significant differences in adaptation to environmental differences among populations. The approach that social affiliation analyses take when making inferences about populations is indirect and requires a number of assumptions, especially when social structure is compared across regions. Social affiliation indices, even if low, indicating fluid or nonexistent associations, give meaningful information. For example, grouping or pairing with other individuals may hold fitness benefits for some animals at certain times. Conversely, avoiding individuals or the chance of forming long term bonds with other animals may also prove advantageous. Cooperative feeding has been demonstrated in several cetacean species indicating that grouping occurs in the presence of prey (Würsig 1986). Alternatively, some studies have shown increased feeding with more dispersal. Allen et al. (2001) found a negative correlation between group size and feeding frequency in bottlenose dolphins, as well as increased feeding with larger nearest neighbor difference. The link between social stability and foraging strategies has been suggested but has never been well explained in delphinids.

18

HABITAT STRUCTURE Interspecific comparisons of cetacean social structure by habitat have led to general conclusions of this relationship. The structural complexity of different habitats can be roughly divided into two types: unobstructed (open coastline and pelagic waters) and complex (estuaries and enclosed systems). This distinction does not capture the individual variations of specific habitats, but patterns reflected in the way dolphins group between the two types seem to reinforce this dichotomy. Large groups with wide home ranges tend to dominate open habitats (Defran et al. 1999), while smaller groups exhibiting greater regional site fidelity are more common in closed systems (Wells 1987). Caldwell (2001) attributes group size difference to balancing the costs associated with grouping strategies and with ranging patterns. Differences in habitat structure may affect food, shelter, and mate finding, and these effects may favor the development of different grouping and ranging patterns. Structurally complex aquatic environments may provide more opportunities for microhabitats to flourish than unobstructed habitats, and therefore more prey species for higher level predators. Shallow-water systems may offer more protection from predators than deep water, and enclosed systems may yield more opportunities for running into and thus reuniting with associates than open water. The IRL is a semi-closed system with few outlets to the open ocean and considerable variety in bottom topography, qualifying it as structurally complex. Mazzoil et al. (2002) has found average dolphin group sizes of 4.2, which are comparable to Galveston Bay, Texas (mean = 4.4: Fertl 1994), a system similar in topography but small compared to the open, unobstructed shoreline of southern California (mean = 19.8: Defran and Weller 1999). Certain aspects of association, however, seem to remain consistent across structurally varying habitats. Universally, the most stable bonds are between mother-calf pairs, followed often by sub-adult male dyads or triads (Connor et al. 2000b). The pattern of same-sex association appears to be somewhat uniform across systems with differing habitat structure, but the number of social and solitary individuals appears to be related to habitat characteristics (Quintana-Rizzo and Wells 2001).

19

SOUTHEAST ATLANTIC BOTTLENOSE DOLPHIN RESEARCH Recently, more systematic, longitudinal studies have been undertaken to better understand the population dynamics of bottlenose dolphins found off the Atlantic seaboard. Although Scott et al. (1998) originally hypothesized that a coastal migratory stock of dolphins ranged seasonally from the coast of Long Island to central Florida, several subsequent studies (Caldwell 2001, Zolman 2002) have questioned this account. Researchers have discovered the existence of separate populations, genetically distinct and with different but overlapping home ranges. The National Oceanic and Atmospheric Administration (NOAA) assessments of the western north Atlantic coastal stock of bottlenose dolphin have indicated the existence of at least four populations distributed from south of Long Island, New York around the peninsula of Florida on the basis of significant overall genetic variance (Hohn 1997). Zolman (2002) documented the existence of the northernmost site of a resident bottlenose dolphin community on the east coast of the United States in the Stono River Estuary, Charleston. Out of a catalog of 112 identified individuals, 19% (n = 21), were judged to be residents, based on having a sighting history of at least one in each of the four calendar seasons, regardless of year. It is possible that due to study limitations, actual residents comprise a greater percentage of the population. Preliminary mark-recapture studies of dolphins inhabiting the inshore waters of North Carolina yielded a preliminary abundance estimate of 1,033 (Read et al. 2003). Eventual comparisons of both the Stono River Estuary dolphins and those inhabiting the bays and estuaries of North Carolina to the IRL population could provide insight on how specific changes in latitude or habitat structure affect social structure. Using social affiliation analysis to determine community membership, Rossbach and Herzing (1999) distinguished two dolphin communities of 28 individuals inshore and 15 dolphins found over 27 km offshore near Grand Bahama Island, Bahamas. The researchers applied cluster analysis and multidimensional scaling techniques to group interdependent data which revealed relative association distances among groups. Mapping the simple ratio indices of association between the individuals sighted yielded clusters of association pools

20 which led the researchers to conclude that communities could be distinguished by association. Several studies have been carried out on the association patterns of dolphins living off the coast of Florida. Association analysis of dolphins sighted ten or more times within a twoyear period near Cedar Keys, Florida revealed that dolphins associated overall at low levels but more with individuals that exhibited similar sighting histories (Quintana-Rizzo and Wells 2001). Caldwell (2001) examined the social and genetic structure of dolphins in Jacksonville, Florida. Similar to the findings of Wells (1986), the existence of separate communities within the population were found. She looked at four regions: the coast, the St. John’s River, and the inshore waters to the north and south of the river. Coastal dolphins exhibited significantly different density and weaker association patterns than the other study areas. Genetic structure analysis of mtDNA haplotypes and nuclear microsatellite loci reinforced delineation of the Jacksonville population into three distinct communities, Northern, Southern, and Coastal, despite the absence of geographic barriers. Photographic and genetic evidence of separation among the communities allows investigators to study the variations in habitat that may explain the differences in social structure. Finding that coastal dolphins exhibited weaker association patterns was consistent with the habitat structural complexity hypothesis.

POPULATION STRUCTURE OF THE IRL DOLPHINS Earlier studies on bottlenose dolphins in the Indian River Lagoon have been informative but limited in duration. Most have been observational, lasting approximately one year or less in length, and have utilized boat and aerial surveys ( Leatherwood 1975; Rudin 1991; Spellman 1991; Fick 1995). Rudin (1991) examined the distribution and group composition of bottlenose dolphins within a specific 38.9 km sector of the IRL, from Melbourne Causeway, in the Indian River, north to Merritt Island Causeway, in the Banana River. Six aerial surveys and 73 boat survey days from January to June, 1991, yielded a total of 336 groups with an average group size of 3 ± 2.5 individuals. Most dolphins were found in water 1.8 m or shallower. Although 1006 dolphins were sighted, there is no report in the study of any individuals being identified.

21 Spellman (1991) concurrently photo-identified individual dolphins and recorded their behavior in a 30 km stretch of the IRL, south of Rudin’s (1991) study area: from Melbourne Causeway in the north to Sebastian Inlet in the south. Five aerial surveys and 59 boat surveys were conducted, and 75 individuals were identified. Thirty-two were sighted only once and ten were sighted in the northern section of the study area. Group size average was three (no SD reported). In another similar study, Fick (1995) conducted surveys of bottlenose dolphins in the IRL for 12 months to assess abundance, group size, and behavior. The study area extended between Eau Gallie Causeway, Melbourne in the south, to the Florida East Coast railroad spur in Titusville in the north. Survey coverage did not include all of the study area but was confined to four subsections. Each subsection had an average area of 20.42 km2 (SD = 0.46). Thus, certain areas between these study sections were not surveyed. The researchers flew sixteen aerial surveys to assess whether season or time of day had an effect on dolphin abundance throughout the study area. No significant effect of season, nor time of day, nor their interaction was found. One hundred and twenty-nine boat based surveys yielded 363 group sightings and 98 individually identified dolphins. Only seven dolphins were resighted during the study. Odell and Asper (1990) freeze-branded 134 dolphins in the lagoon between 1977 and 1979, and they followed with boat-based tracking and identification studies to determine the status of the freeze brands. Eighty-one of the original 134 freeze-branded dolphins were later encountered, 60 were resighted exclusively in the Indian River, and three had moved into the adjacent Mosquito Lagoon. At least one was sighted as far south as the Sebastian Inlet. Fick (1995) identified four of the freeze-branded dolphins, approximately 15 years later, and 14 dolphins were sighted approximately 20 years later in the IRL (Mazzoil et al. 2002).

HARBOR BRANCH OCEANOGRAPHIC INSTITUTION PHOTO-IDENTIFICATION PROGRAM Photographic surveys of bottlenose dolphins found in the Indian River Lagoon system were begun in August 1996 by the SEARCH Foundation of Harbor Branch (HBOI), which in 1999 became the Dolphin Research and Conservation Program (DCRP) under the Division of Marine Mammal Research at HBOI. Survey coverage of the IRL between August 1996 and

22 September 2001 was variable geographically as well as across seasons and years, but most surveys were concentrated between Sebastian Inlet and Ft. Pierce Inlet between 1999 and 2001. Beginning in July 2002 and continuing to August 2003, survey effort was extended to include the entire IRL and one complete survey of the entire IRL was attempted each month. The study period for this research spans the interval between August 1996 and August 2003 and details of photo-identification survey coverage for the entire study period of August 1996 to August 2003 are given in the methods section. Preliminary analysis of the dorsal fin photographic data collected has documented the occurrence of approximately 400 distinctively marked individuals (Mazzoil et al. 2000). Ongoing surveys will hopefully lead to further identification of individuals and uncover trends in social structure, which is the goal of the present research.

23

METHODS The Indian River Lagoon extends 256 km from Ponce de Leon Inlet in the north (29°08’N, 80°92”W) to Jupiter Inlet in the south (26°94’N, 80° 07”W) (Figure 1). The Indian River Lagoon includes three major bodies of water: the Mosquito Lagoon, the Indian River, and the Banana River (Figure 1). The Intracoastal Waterway (ICW), a dredged channel that runs from the northeastern US to Texas along and within the coastline, bisects the system from north to south. The IRL is an estuarine system which topographically consists of sea grass beds, sandy bottom expanses, and spoil islands. Five major inlets, Ponce de Leon, Sebastian, Ft. Pierce, St. Lucie, and Jupiter, as well as one locks system, connect the IRL with the Atlantic Ocean. Over 4000 species of fish, invertebrates, and plants are found inhabiting the lagoon system (Dybas 2002). Vessel traffic, mostly recreational boaters and fishermen, is moderate to heavy in fair weather conditions.

SURVEY COVERAGE 1996 to 2001 One hundred and ninety photo-identification surveys, conducted along the ICW in the IRL were carried out between August 1996 and September 2001, with the greatest number of surveys (n = 158, 83%) and most extensive coverage occurring between 1998 and 2001. The ICW survey track allowed coverage of both shallow (≤ 1m) and deeper (> 1m) sections of the IRL along the entirety of the study area. Most surveys were concentrated within the Indian River (IR) (n = 180) from its northern extreme to the St. Lucie Inlet in the south. A limited number of surveys extended into the Banana River (BR) (n = 6), the Mosquito Lagoon (ML) (n = 4) and the Barge Canal (BC) (n = 5) at the north end of the IRL. Three surveys were also conducted within the St. Lucie River (SLR) including the North and South Forks. The section of the IRL covered during individual surveys varied considerably in location and length as well as in frequency across this time period. Variability in the dimensions of survey coverage was due primarily to the developmental nature of the initial

24

Figure 1. Indian River Lagoon, Florida. phase of this work, and was also due to concurrent stranding, recovery and rehabilitation efforts by the HBOI Division of Marine Mammal Research. Survey coverage within the IRL during this portion of the study period was evaluated by partitioning the study area into 1 km latitude zones. The highest sampling frequency occurred between the Sebastian and Ft. Pierce Inlets and this range was designated as the primary study area (PSA) (Figure 1). The

25 mean sampling frequency for PSA zones was 64.6 (SD = 17.72). Surveys in the Indian River north of the PSA sampled zones 31 to 135. Within this northern section of the Indian River zones 31 – 114 were sampled an average of 5.0 times (SD = 2.85), while zones 115 - 135 averaged 20.2 times (SD = 2.34). Surveys in the Indian River south of the PSA sampled zones 189 to 237. Within this southern section of the Indian River, zones 180 - 212 were sampled an average of 23.2 times (SD = 3.91), while all zones between 213 to 237 were sampled five times.

2002 to 2004 Beginning in July 2002 and extending to June 2004 photo-identification survey coverage was extended to include the entire length of the IRL from Ponce de Leon Inlet in the north to Jupiter Inlet in the south and included the Indian River in its entirety, the Banana River from the southern tip of Merritt Island to the Barge Canal, the Mosquito Lagoon, and the St. Lucie River including the North and South Forks (Fig 1).

PHOTOGRAPHIC SURVEY PROCEDURE A variety of survey vessels ranging from 6 to 8 m in length and powered by outboard motors were deployed to conduct photo-identification surveys. During surveys the vessel motored along the middle of the ICW at about 17 km/h while searching for dolphins. Survey speed was lowered as needed, however, to accommodate to water, weather and visibility conditions as well as when entering manatee zones. Surveys were conducted once per day, and sometimes two vessels simultaneously surveyed different sections of the IRL. When dolphins were sighted, the survey vessel slowed and then stopped when close enough to make initial estimates of group size as well as to collect other information on the sighting time, GPS location, environmental conditions, and behavior. Once initial sighting information had been recorded, the vessel maneuvered closer to the group and individual dorsal fin notch patterns were photographed with 35-mm motor driven cameras, and a 400 mm telephoto lens. Between 1996 and 1999, Kodak 64 or 200 ASA Kodachrome slide film was used. Subsequently, dorsal fin photography was accomplished with a Canon digital camera system (EOS D2000 and EOS 1D, 100-400 mm telephoto zoom lens). In all cases, attempts were made to photograph every dolphin within a

26 group. Initial estimates of group size were revised as necessary and contact with the group was maintained until the photographic effort was completed. Identical procedures were repeated as the research vessel resumed travel on the ICW survey route and as additional dolphins were encountered. Field and photo-identification protocols and definitions closely followed those outlined in the Sarasota Dolphin Research Program Field Techniques and Photoidentification Handbook (Urian and Wells 1996). Following the definition of Wells (1987), a group was defined as all individuals within a 100 m radius, generally engaged in the same behavior and movement pattern. Calf and probable mother definitions used both appearance and behavior as criteria. Calf criteria included: constant or close affiliation with a large dolphin over the observation period, small size relative to the adjacent dolphin, fetal folds and colorations, and awkward swimming, submergence, and surfacing behavior (Weller 1991). In addition, dolphins regularly accompanying calves were considered probable mothers. Calf status was ultimately judged by observers in the field, and included most but not necessary all of the established criteria. For example, not all calves had fetal folds. In order to follow certain mother-calf pairs over time, calves of distinctive mothers were identified as long as the calves themselves had distinctive dorsal fin notches and/or other distinctive scarring patterns on their back, peduncle, or dorsal fin. Sighting and environmental data were entered into an Access database as soon as possible after each survey was completed, and all data were backed up daily on a server. Sighting and environmental data included sighting date, time, location, group size and calf estimates.

SORTING, MATCHING AND CATALOGING The laboratory analysis of all dorsal fin images closely followed the procedures described by Mazzoil et al. (2004b) and is briefly summarized. Clear photographs of distinctively marked dorsal fins were sorted by recognizable notch patterns, and the best photograph of each dolphin was selected as the “type specimen” to which all other photographs were compared (matching and cataloging). Subsequently, only unambiguous matches with the “type specimen” were accepted as re-identifications of known individuals.

27

DATA ANALYSIS Several analyses on the IRL data were conducted and compared to previous research findings in the study area and to other well-studied populations. Analyses included the following parameters: encounter rate, sighting frequency, group size frequency, mean group size by season, percentage and size of groups with and without calves, level and mean level of association, and mean number of affiliates. Data analysis methods closely followed Marsh (2000) and Rogers et al. (2004).

Encounter Rates, Sighting Frequencies, and Group Characteristics Encounter rates measured the frequency with which dolphins were sighted on surveys, including single and multiple sightings within a sampling period. High encounter rates indicate more usage of the study area, and comparisons of specific sub areas were made to examine potential habitat preference. Sighting frequencies represented the number of times individual dolphins were sighted. Patterns of group size frequency, derived from field estimates and confirmed by photo-identification, were plotted in order to examine their variability. Patterns in the shape of this distribution were studied for optimal group size or sizes for IRL dolphins. The analysis of sighting frequencies assisted in setting the minimum criteria for inclusion in social affiliation analyses which are discussed below. The primary dependent variables under investigation were group size and group composition. Group size distributions were positively skewed, the square-root of group size was used for analysis of variance calculations. A square-root transformation was used because it helped the data best approximate a normal distribution compared to other transformations (e.g., log and square). Group composition was partitioned by the presence or absence of calves, and was analyzed by analysis of variance ANOVA. The mean, range, and frequency parameters of group composition, and seasonal correlates of calf and non-calf groups were analyzed. Interest in the variations that may occur in the parameters of group composition is related, in part, to the possible function of calf groups as nurseries (Wells 1986).

28 Analysis of variance was used to test for possible seasonal effects on group size and composition. Seasons were defined in three ways: by calendar, by hurricane season, and by season according to study area water temperatures following Caldwell (2001). Calendar seasons were defined accordingly: spring (March through May), summer (June through August), fall (September through November), and winter (December through February). Temperatures were recorded for most surveys, and two seasons were identified. Winter surveys included those where water surface temperature dropped below 16°C, while summer surveys included those where water temperature was greater than or equal to 16°C. The distribution of group size and composition with respect to sighting location was investigated. The Harbor Branch photo-identification program partitioned the IRL study area into 1 km latitude zones, beginning at the extreme northern end of the Mosquito Lagoon (ML001), and ending in the south at the highest zone number (IRL237) at Jupiter Inlet. Evidence exists for a degree of site fidelity among particular IRL regions: Mosquito Lagoon (ML); the area from Barge Canal south to the end of Merritt Island, known as the Unusual Mortality Event (UME) area; and the St. Lucie River (SLR) (Mazzoil et al. 2004) (Figure 2). Group size data were partitioned into bins according to the IRL zone numbers in order to investigate grouping pattern differences among region. The effects of diurnal cycles on group sighting data were analyzed. Morning (AM) and afternoon (PM) sightings were first compared. Day period was next partitioned into four 3-hr periods, as defined by Hanson and Defran (1993): (1) early morning (dawn-0859), (2) late morning (0900-1159), (3) early afternoon (1200-1459), and (4) late afternoon (1500dusk). All four day periods were compared for group characteristic differences. In order to investigate the possibility of predation of sharks on IRL dolphins, shark bite scars were noted. Evidence of previous bites, identified as crescent or partial-crescent scars on the body of a photographed dolphin, were observed opportunistically and recorded for dolphins identified in the catalog. Since it is difficult to ascertain whether the estimate of shark bite scars was too high or too low due to measurement bias, no analysis on predation was conducted.

29

Mosquito Lagoon Area

ZONE 39

UME Area

ZONE 207

STL Area

Figure 2. Study area regions.

Site Fidelity Minimum sighting criteria for residency has varied widely from study to study, ranging from two to ten and often dependent on the duration of the study and specific association or residency patterns under investigation (Rogers et al. 2004). Placing too high of

30 a restriction on minimum sighting criteria limits the numbers of individuals and potential associates, and placing too low of a restriction may lead to including transient individuals that are not representative of the source population (Bejder et al. 1998). Given that the duration of the study lasted nine years, only dolphins sighted nine or more times over the course of the study were included for analysis of site fidelity. After selecting only individuals that were sighted nine or more times, the individual dolphins’ sighting histories were examined. An average site fidelity ratio (SFR) was calculated for each dolphin by dividing the number of years in which they were sighted over the number of years possible in which they could have been sighted (Mazzoil et al., in press). If approximate dates of death were available, the denominator of the SFR was reduced to reflect their life spans.

Social Affiliation Social affiliation indices were calculated using SOCPROG 2.2, a series of MATLAB programs developed by Whitehead (2004). This software is arranged into a series of modules designed for different analytical tasks. First, Excel files containing sighting history information were inputted into Socprog and descriptive statistics about individuals and their associations were run. Next, the analysis of association module was run to provide inferential statistical tests of the association data. Sampling period and association type were set while running this module. The program permutes the data, giving several versions of randomized association and then outputs results for both random and real data sets, including mean of association indices, standard deviation of association indices, mean of non-zero association indices, and standard deviations of non-zero association indices. At each permutation step, the procedure selects two individuals and two groups, and then the four group-individual assignments are switched in order to keep the number of individuals in each group and the number of groups constant (Whitehead et al. 2005). This step is called a flip, and setting the number of flips is another factor of running permutation tests. Coefficients of association (COAs) were calculated in Socprog 2.2 using both the Simple Ratio Index (SRI) and the Half-Weight Index (HWI) (Whitehead 1999). Since the set of identified dolphins in the IRL is large and group sizes tend to be small, the half-weight was the index of primary interest because it gives weight to individuals that are more likely sighted apart than together (Cairns and Schwager 1987). An important starting point for the

31 analysis of social affiliation patterns is to set the criterion for the minimum number of sightings necessary for an individual to be included in association index analyses. The criterion must not be set too low in order to avoid having a spurious index value. The criterion must also not be set too high, however, in order to not seriously limit the sample size of individuals available for affiliation analyses. Dolphins with a minimum of nine total sightings were included in the analyses, which was appropriate to the individual sighting history dataset and in accordance with a review of similar studies. Mantel and permutation tests were also run using Socprog 2.2. Mantel tests examined whether gender patterns in affiliation, specifically whether same-sex associations, differed significantly from mixed-sex associations. Permutation tests compared distributions of association from the real dataset to distributions of association from random or permuted datasets. Permutation test statistics indicate a significant difference for a two-tailed test with α = 0.05 when P-values are either lower than 0.025 or higher than 0.975. Significant P-values greater than 0.975 are reported as one minus the value. A frequency distribution of maximum association levels was generated for all individuals meeting the minimum sighting criterion. Permutation tests on overall associations between sexed individuals were run using both the simple ratio index and the half-weight index. For each test, P-values stabilized at 20,000 permutations and 100 flips of the association matrices. A daily sampling period was initially chosen to examine association patterns. Further analyses using monthly and yearly sampling periods were compared to the daily analyses in order to examine temporal patterns of affiliation. To test whether or not dolphins are associating at a level significantly different from random, randomized permutations of the association dataset were run. Different types of permutations were run on the data in order to generate randomizations. ‘Permute groups within samples’ tests for the presence of preferred or avoided individuals, given the number of groups in which each animal was sighted during each sampling period. This test accounts for the likelihood of individuals sighted in many groups to group randomly. In this test, incidence matrices of groups by individuals are permuted, and a sampling period is chosen in which the data are to be altered. Using this option, both long-term (between sampling period) and short term (within sampling period) preferred companionships are tested. Long-term companionships are indicated by significantly higher standard deviation (SD) of real

32 association indices than SD of random association indices. Short-term companionships would be indicated by a significantly lower mean of the real association indices than the mean of the random association indices. SOCPROG v. 2.2 introduced a new affiliation measure, the Coefficient of Variation (CV), which is a ratio of mean to SD. The CV corrects for situations in which the within sampling period associations may drive down between sampling period associations, since the mean and SD are related. Circumstances preventing all individuals from being present in each sampling period due to birth, death, and emigration are also accounted for. The next test run was the ‘permute all groups’ test. The null hypothesis for this test is that there are no preferred or avoided companions given the total number of groups each animal was sighted in during the study. If the real CV is significantly higher than the random CV, the null hypothesis is rejected (Whitehead 2004). This test accounts for the tendency of dolphins sighted in many groups to associate randomly. This test, however, does not account for situations in which all dolphins were not present in every sampling interval because of birth, death, etc., which then may suggest instances of preferred companionship that are, in reality, coincidental occurrences. This test is most valid for brief studies in which all individuals are likely to be in the study area during each sampling interval; thus this test was used to examine annual affiliation patterns during the study period. Annual analyses only included individuals sighted at least nine times over the course of the study period. Further, only individuals sighted at least three times during any given year were included in each annual analysis. In summary, group size, composition, and social affiliation patterns of IRL dolphins were examined with respect to region, season, and day period. Comparisons were made within and across study years, as well as to other studied populations of dolphins.

33

RESULTS A total of 477 surveys were conducted in the Indian River Lagoon from August, 1996 to June, 2004. Sightings from these surveys were partitioned into two categories: on-surveys, which indicated dolphin encounters that occurred during the predetermined survey route (n = 368, 77%); and off-surveys, which indicated opportunistic dolphin encounters (n = 109, 23%). Only on-survey sightings were included in analyses of group size, composition, and affiliation. At least one dolphin was encountered on 97% (n = 355) of all on surveys. Survey time totaled over 1874 h, during which 615 h were spent in direct observation of dolphins.

GROUP SIZE A total of 2,132 groups were encountered and photographed over the study period (Table 1). Group sizes ranged from 1 to 40, with an average overall group size of 4.1 (SD = 3.43, median = 3) (Figure 3). Fifty-seven percent of all groups (n = 1207) had ≥ three individuals, and 76% of all groups (n = 1611) had ≥ five individuals. Summary statistics (i.e., means, medians, and standard deviations) were calculated and reported from raw data for ease of comprehension (Table 1). Since there were a low number of surveys in 1996 (n = 4), 1997 (n = 28), and 1998 (n = 19), and because group sizes did not appear to vary among these three years, they were combined into one time period for analysis. The rationale for this combination was further supported by a non significant ANOVA comparing mean group sizes for these three years (F(2,142) = 1.49, P = .228).

34

Table 1. Summary Information on Annual and Seasonal Survey Effort, Encounter Rate, Sighting Data, and Group Size. Annual

Seasonal

Overall

Study Period 1996 1997 1998 1999 2000 2001 2002 2003 2004

No. surveys 4 28 19 50 37 39 36 94 46

Total dolphins 120 464 265 677 844 1074 1179 2716 1359

Total groups 23 71 51 169 196 277 307 697 341

Group Size Mean SD 5.2 3.04 6.5 5.13 5.2 5.00 4.0 2.84 4.3 3.65 3.9 3.35 3.8 2.96 3.9 3.32 4.0 3.43

Range 1-12 1-28 1-28 1-20 1-30 1-23 1-18 1-40 1-22

Percent calf 14% 30% 35% 36% 29% 25% 17% 21% 22%

Spring Summer Fall Winter

96 88 82 88

1995 2076 2262 2365

505 536 494 597

4.1 3.9 3.9 4.4

3.28 3.23 2.95 4.03

1-28 1-23 1-28 1-40

25% 24% 22% 24%

354

8698

2132

4.1

3.43

1-40

24%

600 500

Frequency

400 300 200 100 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 28 30 40 Group Size Figure 3. Frequency distribution of group size.

35

Annual Variation Variations in group size were compared across years and showed that mean group sizes were higher in the combined 1996-1998 time period (mean = 5.9, SD = 4.66, median = 5.0) than in other study years (Table 2). Group sizes for 1999-2004 ranged from 1 to 40, with an overall mean equal to 4.0 (SD = 3.29, median = 3.0). An overall one-way ANOVA on annual group size averages was significant (F(6,2125) = 7.69, P < .01) (Figure 4). Post-hoc analyses indicated that the 1996-98 average was higher than all other annual values (P < .01), but that values between 1999 – 2004 did not differ from each other. Table 2. Descriptive Statistics for Dolphin Group Size by Study Year. Year Mean Median SD Variance Minimum N 1996 23 5.2 4 3.04 9.27 1 1997 71 6.5 5 5.13 26.34 1 1998 51 5.2 5 5.00 20.20 1 1999 169 4.0 3 2.84 8.04 1 2000 196 4.3 3 3.65 13.34 1 2001 277 3.9 3 3.35 11.22 1 2002 307 3.8 3 2.96 8.79 1 2003 697 3.9 3 3.32 11.01 1 2004 341 4.0 3 3.43 11.78 1

Maximum 12 28 28 20 30 23 18 40 22

6.5

Mean Group Size

6 5.5 5 4.5 4 3.5 3 1996-1998

1999

2000

Figure 4. Mean group size by study year.

2001 Year

2002

2003

2004

36

Seasonal Variation Group size did not appear to vary by calendar seasons (Table 1), and an ANOVA comparing average group size by calendar season was not significant (F (23,2100) = 0.99, P = .48). The effect of hurricane season on group size was also examined. Sighting data from hurricane season months (June through November, n = 1030) was tested against nonhurricane season months (December through May, n = 1102). Although group sizes in nonhurricane months (x-bar = 3.9, SD = 3.10) were slightly higher than in hurricane months (xbar = 4.3, SD = 3.71), this difference was not statistically significant (F (8,2115) = 1.49, P = .16). Sea surface temperature measurements were taken in approximately 89% of sightings (n = 1689) and ranged from 8 to 34 Co (Figure 5). Since temperatures measured in the current study exhibited daily fluctuations of up to 6 Co, and measurements were not taken during all sightings, SST measurements were categorized into equal-interval bins that ranged across six degrees (7 – 12 Co, 13 – 18 Co, 19 – 24 Co, 25 – 30 Co, and 31 -36 Co; n = 1817) (Table 3). Survey days in which no water temperature readings were taken were not included in the analysis, and no water temperature data were available for 1996. Although group size appears to fluctuate to some degree with respect to SST, the differences were not statistically significant (F(4,1812) = 1.87, P = 0.11). When a temperature shift occurs, conditions may foster recruitment and growth of prey species, but there may be a time lag before dolphins begin aggregating to take advantage of these prey conditions. Unfortunately, in the current study there were too few days in which temperature readings were taken in the early study period (1996: n = 0, 1997: n = 6, 1998: n = 5) in order to be able to compare to latter years in which there were more temperature data. In 2003, there were 89 days in which SST measurements were taken. A graph of SST from 2003 was compared with group sizes in order to investigate if there were lagged effects of SST on group size (Figure 6). From the beginning of the year, the temperature values increase into the warmer summer months and then fall towards the colder winter months resulting in an arc-shaped function, as predicted from the change of seasons. Meanwhile, group size values are highly dynamic within a short period, and this variability repeats itself uniformly across the year. This pattern does not support a view that SST has a time-lagged effect on group size.

37 180 160

Frequency

140 120 100 80 60 40 20 0 8

10

12

14

16

18

20

22

24

26

28

30

Temperature (ºC)

Figure 5. Frequency distribution of sea surface water temperatures.

Table 3. Mean Group Size by Sea Surface Temperature Range. Temperature Range Number of Sightings Mean Group Size 7-12 45 4.5 13-18 378 4.4 19-24 532 3.7 25-30 690 4.0 31-36 172 4.1

SD 4.13 4.23 2.78 3.13 3.38

32

34

1/ 14 /2 00 3 1/ 29 /2 00 3 2/ 13 /2 00 3 3/ 18 /2 00 3 4/ 17 /2 00 3 5/ 14 /2 00 3 6/ 11 /2 00 3 8/ 25 /2 00 3 9/ 22 /2 00 3 10 /2 1/ 20 03 11 /1 7/ 20 03 11 /2 6/ 20 03 12 /2 2/ 20 03

Sea Surface Temperature 30.0

25.0 30

20.0 25

15.0 20

10.0 15

5.0

0.0

SST

Group Size

35.0 45

40

35

10

5

0

Date

Group Size

Figure 6. Recorded sea surface temperatures and dolphin group size for 2003.

38

39

Regional Variation Group size was investigated among study regions. Group sizes within the ML (x-bar = 4.2, SD = 4.09), UME (x-bar = 4.1, SD = 3.24), and SLR areas (x-bar = 4.6, SD = 3.67), as well as the Primary Study Area (PSA) (x-bar = 4.0, SD = 3.23) were compared and did not appear to vary (F(3,1498) = 1.16, P = 0.32). Since the IRL extends in a north-south direction, the effect of latitude on group size was also examined using a simple regression model. Latitude did not appear to have a systematic effect on group size, and a significant linear relationship between the two variables was not found (F(1,2066) = 1.20, P = 0.15) (Figure 7). 45 40 35

Group Size

30 25 20 15 10 5 0 26.9

27.4

27.9

28.4

28.9

Latitude

Figure 7. Scatter plot of latitude on group size.

Diurnal Variation Diurnal changes in grouping patterns have been reported in other studies of bottlenose dolphins inhabiting the IRL (Fick 1995, Clement 1998), thus changes in group size by day period was examined. First, day period was separated into morning (AM, n = 970) and afternoon (PM, n = 1162) (Table 4). Groups sighted in the afternoon (x-bar = 4.4, SD = 3.78) were generally larger

40 Table 4. Summary Information on Group Size by Day Period. Day Period Total Groups Mean Early Morning 165 3.8 Late Morning 805 3.7 Morning Total 970 3.7 Early Afternoon 800 4.4 Late Afternoon 362 4.3 Afternoon Total 1162 4.4

SD 2.77 2.97 2.93 3.93 3.42 3.78

than those in the morning (x-bar = 3.7, SD = 2.93), and this difference was statistically significant (t (2125) = -3.9, P < 0.01). Group sightings were categorized by day period according to the sighting start time and were pooled across the study period (Table 4). Group size appeared to vary among day periods, with early afternoon groups being much larger than late morning groups (Figure 8). This apparent difference was confirmed by a one-way ANOVA (F(3,2128) = 5.14, P < 0.01). Posthoc test showed that the difference was due to variations between group sizes in the early afternoon and late morning (P < 0.01).

4.8 4.6

Group Size

4.4 4.2 4 3.8 3.6 3.4 3.2 3 Early Morning

Late Morning

Early Afternoon

Day Period

Figure 8. Mean group size by day period.

Late Afternoon

41

Shark Bite Scars Evidence of shark bite scars were examined in photographic histories of the identified dolphins. Photo-documented shark-bite scars exist for at least 19 (3%) of the identified IRL dolphins, ten of which were seen at least nine times. This count is probably an underestimate, since scars were noted opportunistically (Mazzoil, personal communication). In addition, some scars may have been hard to detect due to location or lack of severity.

GROUP COMPOSITION The composition of groups was evaluated by comparing calf and non-calf groups. Calf group status was based on the presence of at least one calf in the group. Of the 2132 groups encountered, 1162 (55%) contained at least one calf, and 24% of all individuals encountered were calves. Overall, calf groups were larger (x-bar = 5.5, SD = 3.80) than groups that did not contain calves (x-bar = 2.4, SD = 1.85), and this different was statistically significant (t(2130) = 27.82, P < 0.01).

Seasonal Variation The size of calf and non-calf groups was compared across calendar and hurricane seasons. Calf groups were consistently larger than non-calf groups for each calendar season and in both hurricane and non-hurricane groups (Table 5); therefore, calf and non-calf groups were examined separately for seasonal differences. Table 5. Summary Statistics for Group Composition by Season. Numbers in Parentheses Represent SD. Calf Groups Non-Calf Groups No. Groups Mean Group Size No. Groups Mean Group Size Winter 327 6.0 (4.54) 270 2.5 (2.09) Spring 288 5.4 (3.59) 217 2.3 (1.54) Summer 294 5.3 (3.49) 242 2.3 (1.79) Fall 253 5.1 (3.22) 241 2.5 (1.87) Hurricane 547 5.2 (3.37) 483 2.4 (1.83) Non-Hurricane 615 5.7 (4.13) 487 2.4 (1.87) Overall 1162 5.5 (3.80) 970 2.4 (1.85) Calf group sizes did not appear to vary by calendar season, and no statistical difference was found (F(3,1158) = 2.09, P = 0.10). Calf group size seemed to be slightly larger in non-

42 hurricane months (x-bar = 5.7, SD = 4.13) than in hurricane months (x-bar = 5.2, SD = 3.37), and this difference was marginally significant (F(1,968) = 3.54, P = 0.06). Non-calf group means did not display differences in size among calendar seasons (F(3,966) = 1.36, P = 0.25) or hurricane seasons (F(1,968) = 0.04, P = 0.85). Certain months had higher calf group encounter rates than non-calf group encounter rates. Calf groups were encountered more than non-calf groups in February (n = 126, 65% ), and non-calf groups were encountered more in April (n = 56, 35%) and September (n = 72, 42% ) (Figure 9). These encounter rate differences were statistically significant (P < 0.05).

Regional Variation Encounter rates between calf and non-calf groups were compared for the ML, UME, PSA, and SLR areas (Figure 5). The percentage of calf-group encounters were as follows: ML: 49% (n = 140); UME: 60% (n = 193); PSA 59% (n = 474); and SLR: 45% (n = 41). The overall calf-group encounter rates appeared highly variable, and these differences were statistically significant (χ2 = 14.2, P < 0.01). Further, post-hoc analyses revealed that the difference was due to the UME and PSA areas having more calf-group encounters than the ML or SLR areas (P < 0.01). Mean differences in group size for calf and non-calf groups between study regions were examined and compared (Figure 9). For calf groups, mean group size was larger in the SLR and ML areas than it was in the UME or PSA, which was confirmed by statistical comparison (P = 0.05). Non-calf group means did not appear to vary by region and were non-significant (P = 0.07).

SITE FIDELITY AND RESIDENCY Over the entire study period, 621 adult (non-calf) dolphins were identified, and 599 (96%) were from “on” surveys, excluding calves. The rate of discovery for dolphins in the IRL increased over the entire study period (Figure 10). Following the expansion of monthly coverage of the IRL in summer of 2002, the rate of discovery rises at a sharper rate, indicating an increase in the number of dolphins identified, which may in part be due to the increased frequency of sampling other areas besides the PSA.

43 8 6.9

Mean Group Size

7

6.2

6

5.3

5.2

5 4 3

2.3

2.7

2.4

Non-Calf Groups Calf Groups

2.1

2 1 0 Mosquito Lagoon

UME

Primary Study St. Lucie River Area

Study Area

Cumulative Number of Dolphins Identified

Figure 9. Group size by study region.

600 500 400 300 200 100 0 Aug- Apr- Dec- Aug- Apr- Jan- Sep- May- Jan- Sep- Jun96 97 97 98 99 00 00 01 02 02 03 Month

Figure 10. Discovery curve for dolphins identified in the IRL during on-surveys.

44 A total of 121 dolphins’ sighting histories were examined. Approximate dates of death were available for 11 individuals, and the denominator of the SFR was reduced to reflect their life spans. Possible SFR values range from 0-100% and describe a degree of residency over the study period for the sample. The average sight fidelity ratio of these dolphins was 76% (SD = 0.20, range 29%-100%). Further, removing the subset of dolphins with known dates of death left 110 individuals. At least 85.5% of these individuals were sighted in at least four or five (68.2%) of the seven study years (see Appendix A for SFR values). Occurrence within study regions of the most frequently sighted dolphins was examined. Eighty-one percent of their encounters were in the PSA. Of the eight individuals with sightings in the Mosquito Lagoon (ML), none had sightings anywhere else, including opportunistic sightings (off-survey). Almost half the dolphins sighted in the UME area (n = 50, 44%) were only sighted within ± 10 km of the UME area. Of the dolphins sighted at least once in the SLR (n = 50), only two had all of their sightings within the SLR, so no distinctive pattern in site fidelity emerged in the SLR.

SOCIAL AFFILIATION Individuals sighted at least nine times over the entire study period were included in social affiliation analyses. Only one sighting per day per individual was included in the analyses to avoid oversampling individuals in groups re-sighted along the survey route (Whitehead 1995). Maximum association levels ranged from 0.05 to 0.71 using the simple index, and 0.09 to 0.83 using the half-weight index (Figure 11). The entire study period showed a significantly higher CV for real associations compared to random associations, indicating non-random association between days. Results were similar for both the SRI real: CV = 2.73, random: CV = 2.10, P < 0.01) and the HWI (real: CV = 2.73, random: CV = 2.10, P < 0.01) (Table 6). Because of the similarity of the HWI and SRI, and for the sake of simplicity, only the HWI is reported in subsequent analyses. Association was tested over separate study years. As with prior analyses of group size, association data were pooled for 1996-1998 due to lower numbers of surveys in these years. In all seven study time periods, there was strong indication of non-random companionship (Table 7).

45

18 16 14

Frequency

12 10

Simple Ratio Index

8

Half-Weight Index

6 4 2 0 0.09 0.11 0.13 0.15 0.17 0.19 0.21 0.24 0.27

0.3

0.32 0.41 0.56 0.63 0.65 0.73 0.83

Maximum Association Level

Figure 11. Frequency distribution of maximum association levels for dolphins identifited ≥ 9 times in the IRL.

Table 6. Mean, SD, and CV of Real and Random Association Indices and P-values for Overall Study Period Using Daily, Monthly, and Yearly Sampling Periods. Sampling Real Random PReal Random PPeriod Mean Mean value SD SD value Daily NA NA NA 0.04 0.04 0.94 Monthly 0.23 0.23 0.36 0.15 0.15 0.76 Yearly 0.35 0.35 0.51 0.20 0.19 0.91 *Significantly higher real value compared to the random value.

Real CV 2.41 0.66 0.57

Random CV 2.02 0.65 0.56

Pvalue <0.01* 0.77 0.86

Association patterns were viewed over different time scales: day, month, and year. Results indicated that individuals preferentially associate between days (Table 7). The proportion of non-zero indices was also greater for the real association matrix than the random association matrix, when viewed over days, months, and years (P < 0.01) months, which indicates that some individuals in the population avoid each other over each of the time scales under investigation. There appears to be a temporal pattern to the duration of companionships among IRL dolphins.

46 Many IRL dolphins seem to associate for a number of days, segregate, and in some cases reunite with individuals after longer time periods, even years. Table 7. CV of Real Association Indices, CV of Random Association Indices, P-value of Permutation Test Using a Daily Sampling Period. Time 19961999 2000 2001 2002 2003 2004 Total Period 1998 Study Period CV real 2.54 1.82 3.14 3.00 2.61 3.16 3.51 2.41 CV 2.18 1.70 2.70 2.46 1.89 2.78 2.20 1.85 random P-value <0.01* 0.02* <0.01* <0.01* <0.01* <0.01* <0.01* <0.01* *Significantly higher CV of the real associations compared to the CV of the random associations.

Female and Male Associations Of the 121 dolphins sighted nine or more times, information on sex was available for 67 individuals. Dolphins were identified as females in the field if they were consistently sighted with smaller individuals, presumably calves (Weller 1991). Additionally, data on sex were available from prior capture and release collections. A total of 46 females and 21 males were identified, which represents 55% of individuals sighted at least nine times. A Mantel test for differences in associations between and within sex class was run to test whether or not females and males associate more with members of their own sex than with the opposite sex. The null hypothesis is that dolphins do not associate significantly more or less with members of their own sex than members of the opposite sex. The test yields a t statistic, and if the corresponding P-value is not between 0.025 and 0.975 and the matrix correlation coefficient is positive, the null hypothesis of differences in associations between classes can be rejected. Overall, females and males were found to associate significantly more within their own sex class than between sex classes across days (t = 2.38, P < 0.01) and years (t = 2.13, P < 0.07). When examined by year, only one of the seven time periods, 2003, showed significant differences in between and within sex class differences (Table 8). A sociogram exhibiting the strength relationships among sexed individuals for 2003 is given (Figure 12). Males exhibited a larger range in maximum COA values (0.10 – 0.83) than females (0.00 – 0.33) (Figure 13).

47 Table 8. Mantel Test t Statistics, P-Values, and Matrix Correlation Coefficients (r) for Between and Within Sex Class Associations. Time 1996- 1999 2000 2001 2002 2003 2004 Total Total Total Period 1998 by by by Day Month Year 0.85 1.32 1.68 -1.03 -0.31 2.37 1.22 2.38 -0.92 2.13 t 0.81 0.91 0.95 0.14 0.41 0.01* 0.84 <0.01* 0.18 0.02* P 0.06 0.08 0.08 -0.08 -0.02 0.08 0.15 0.06 -0.04 0.07 r *Same-sex associations are significantly larger than mixed-sex associations at the α = 0.05 level.

TOOT STUB (M) (F)

WSCR (F)

LISL (M)

HOLE (M)

REGR (F)

SMOO (F)

OMCF (F) ALLG (M)

TTIP (F)

GASH (F)

RING (F)

SHAM (F) NIBB (F)

PERI (F) REPE (F) HAND (F) SMAS (M) FBD6 (M)

HILO (M)

ELMU (F)

UUMA (F)

CROS (F)

HCBT (F)

REEL (F)

CONE (F)

0.67 0.43 0.20

Figure 12. Sociogram of significant sexed dyads from 2003.

48

8 7

Frequency

6 5 4 3 2 1 0 0

0.11

0.13

0.15

0.17

0.19

0.21

0.24

'0.28

0.33

0.83

COA Female

Male

Figure 13. Maximum half-weight COA levels for dolphins identified at least nine times. Dashes in x-axis denote line breaks.

Mothers and Calves A total of 9 females and their respective calves were identified following the calf and presumed mother identification criteria presented in the Methods section were entered into social affiliation analyses. For the entire study period, mothers and calves had a range of 0.11 to 1.00 maximum COA levels, with a mean maximum level of 0.64. A permutation test revealed that they associated non-randomly over the entire study period (P < 0.01). Mothers and calves continued to associate after the first year of sightings for the calf, although the associations changed during subsequent years. The high association levels between three mother and calf pairs sighted over three years indicates that companionship extends beyond the first year of life. Each of the three pairs showed an average COA of 0.84, but COAs continued to be high even after the fourth year of life (Figure 14).

49

1

Half-Weight COA

0.9

0.8

0.7

0.6

FRED-SCAR NICA-MICA UNEV-EVEN

0.5

0.4 1999

2000

2001

2002

2003

2004

Year

Figure 14. Coefficients of association between three mother-calf pairs by study year. Calves are listed first.

50

DISCUSSION This study presents an examination of grouping and association patterns of bottlenose dolphins within the IRL system. Studies on the group characteristics of IRL dolphins have been conducted in the past, but they were relatively short in duration and geographically limited to certain sections of the IRL. Fick’s (1995) year-long study covered four non-continuous sections of the Indian River between Titusville and Melbourne, adjacent to Merritt Island, which totaled 81.7 km in distance; Rudin (1991) and Spellman (1991) conducted six-month concurrent studies in adjacent areas: Rudin’s study covered Melbourne Causeway to the Merritt Island Causeway, and Spellman’s study covered Melbourne Causeway to the Sebastian Inlet. Together they totaled only 68.9 km in distance. Studies of such relatively short duration and small area are limited in their ability to reveal important aspects of dolphin social ecology. In order to examine long-term association patterns, dolphins must be studied over a long time period, relative to their life span, which usually requires years of data collection. The present study sought to develop a more comprehensive assessment of IRL dolphin social structure by analyzing grouping and association patterns over a longer temporal and greater spatial scale. During part of the present study period, surveys covered the entire IRL system which permitted group characteristic comparisons to be drawn among regions within the IRL. In addition, the expanded temporal frame of this study allowed a greater opportunity to draw conclusions about important variables which may affect social structure in IRL dolphins. Findings pertaining to group characteristics, site fidelity, and social affiliation are discussed in the following sections.

GROUP SIZE Dolphin group size ranged from 1-40 in the present study. Despite this variation, however, both measures of central tendency, the mean (4.1) and median (3.0), were small and suggest that smaller groups are more optimal in the IRL. The overall average group size was similar to mean group size findings from other IRL dolphin studies of shorter duration (Rudin 1991: M = 3.0, Spellman 1991: M = 3.0, Fick 1995: M = 2.8, Clement 1998: M = 3.0). Three variables which are thought to shape group size-- habitat structure, resource distribution, and

51 predation pressure-- were examined to determine their influence on dolphin grouping patterns and will be discussed in the following sections.

Habitat Structure Group sizes for coastal or inshore bottlenose dolphin populations tend to be smaller than for offshore or pelagic populations, which are partially assumed to reflect differences in habitat structure (Norris and Dohl 1980, Quintana-Rizzo and Wells 2001, Eisfeld 2003). Although it is important to avoid oversimplifying the complexities inherent in specific ecosystems, patterns in the way dolphins aggregate seem to relate to general characteristics of their respective underwater environments. Habitat differences may affect predation risk and response. Dolphins living in structurally complex habitats, like enclosed estuaries and bays, may be affected by the available topographic features, such as sandbars and shallow continental slopes. These geographic features could be used to help evade predators, as well as reduce the volume of water that dolphins must monitor (Shane et al. 1986). In contrast, open or pelagic areas offer few structures for concealment from predators (Caldwell 2001). In these areas, dolphins may form larger groups, in part, for better predator vigilance and defense. In Sarasota, larger groups have been found inhabiting the deeper waters of the Gulf of Mexico than they do in the shallower areas near barrier islands (Wells 1980). The IRL is a relatively shallow, narrow waterway compounded by a high level of human activity. Spoil islands, seagrass flats, and sandy bottom expanses make the confines of the IRL structurally complex. Group size findings in the present study are consistent with those from studies on other populations occurring in similar inshore habitats: Sarasota Bay, M = 4.8 - 7.0 (Wells et al.1987); Shark Bay, Australia, M = 4.8 (Smolker et al. 1992); and Jacksonville, M = 4.5 (Caldwell 2001). In contrast, IRL dolphins form much smaller groups than open-coast populations, for example San Diego, M = 19.8 (Defran and Weller 1999); Santa Barbara, M = 12.7 (Marsh 2000); and Moray Firth, Scotland, M = 13.0 (Eisfeld 2003). A summary of group size differences from 14 studies on bottlenose dolphins is given in Table 9. Resources in enclosed estuaries are generally known to be more uniformly distributed as single prey items (Barros 1993). This type of prey concentration may accommodate the foraging requirements of small groups or solitary dolphins rather than large groups. In a year-long IRL prey-utilization study, Clement (1998) found the predominate size of IRL dolphin groups

52

Table 9. Mean Group Size and Group Size Definitions from 14 Studies on Bottlenose Dolphins. Citation Shane 1990

Study Area Sanibel Island, FL

Habitat Inshore bay

Fick 1995

IRL, FL

Inshore estuary

Campbell et al. 2002 Brager 1994 Caldwell 2001

Turneffe Atoll, Belize Galveston Bay, TX Jacksonville, FL

Atoll/inshore lagoon Inshore bay Inshore estuary

Rogers et al. 2004 Smolker et al. 1992

Little Bahama Bank Shark Bay, Australia

Coastal Inshore bay

Wells et al. 1987

Sarasota Bay, FL

Inshore bay

Maze-Foley & Wursig 2002 Marsh 2000

San Luis Pass, TX

Inshore bay

Santa Barbara, CA

Coastal

Eisfeld 2003

Moray Firth, Scotland

Coastal

Ballance 1990

Kino Bay, Mexico

Inshore bay

Defran & Weller 1999 Saayman & Tayler 1973

San Diego, CA

Coastal

Southeast Cape, South Africa

Open Coast

Group Size Definition Dolphins in apparent association, moving in same direction, usually engaged in same behavior A single dolphin or assemblage of dolphins that interacted during a sighting Shane 1990

Mean Group Size 2.4-7.4 (depending on behavioral state)

2.8

3.8

Shane 1990 Individuals within a 100 m radius Shane 1990 Group of dolphins with relatively close-knit spatial cohesion Individuals within a 100 m radius Smolker et al. 1992

4.4 4.5

Dolphins in continuous association with each other and in visual range (Weller 1991) Aggregations of individuals within 500 km of each other, engaged in similar activities, and moving in the same direction Dolphins sighted in bay at one time Shane 1990

12.7

Dolphins generally dispersed over several square kilometers

4.6 4.8

7.0 10.6

13.0

15.0 19.8 140.3

53 engaged in feeding behavior to be three or less, across all seasons observed (spring, summer, and fall). It is likely that habitat parameters which shape fish distribution, in turn, affect dolphin concentration, driving optimal group size down. Alternatively, the larger groups formed by coastal bottlenose dolphins within the Southern California Bight reflect ephemeral but rich concentrations of prey species, which often include schooling fish and market squid (Weller 1991, Lang 2002). The presence or absence of certain geographic structures may shape dolphin response to predation pressure (Wells et al. 1987, Scott et al. 1990, Caldwell 2001). Sarasota Bay dolphin habitats are characterized mainly by channels and passes, seagrass flats, and the open-water expanses between barrier islands and the mainland. The Sarasota community tends to increase their use of shallow water expanses during the spring and summer seasons, when bull sharks are most abundant (Wells et al. 1987, Scott et al. 1990). Inhabiting a more confined amount of space may increase the chance of predator detection and thus decrease the need for predator surveillance (Wells et al. 1980). A small group of individuals may be ample enough to locate predators in a structurally complex habitat like the IRL. Predation risk must be balanced with foraging payoffs in each habitat, and both factors will usually shape grouping patterns (Jarman 1974, Terborgh and Janson 1986, Lima and Dill 1989). Fine-scale changes in habitat within the IRL may also affect group size. Clement (1998) found that group sizes were related to habitat type. In spoil-island habitats and tributaries, dolphins were most often found in groups of four or more, but in the other habitat types (sandbar and shoreline), dolphins were most commonly found in groups of three or less. Continued research on the grouping variations among these habitat types within the IRL may reveal whether these patterns persist over time and relate to substrate differences.

Resource Distribution Studies on dolphin feeding ecology have often been limited to extrapolating from the stomach contents of stranded dolphins to the living members of their populations. When feeding events are witnessed, it can be difficult to identify prey species, and dolphins often prey on a wide variety of species (Shane 1990, Connor et al. 2000). Thus, in many instances, the only prey items known with certainty are identified from stranded dolphins.

54 Barros (1993) identified prey species from stomach contents collected from stranded bottlenose dolphins along the central east coast of Florida. These dolphins were judged to be “resident” if they were found along the IRL and “oceanic” if they had stranded on the adjacent ocean beaches. These distinctions were not founded on previous sighting histories of the stranded individuals. Without prior knowledge of the home ranges of these individuals, there is not enough information to label them as one ecotype or the other. Whether or not dolphins occurring in the IRL spend any time in the open ocean is unknown; no photographic or tagging data exist to date. Therefore, it may be unfounded to assume that where an individual was found stranded can reliably indicate whether it spent its life within the IRL or in the adjacent coastal waters. Further, Barros noted an absence of pseudo-stalked barnacles (genus Xenobalanus) on the bodies of stranded dolphins found within the IRL. These ectoparasites were commonly found, however, along with whale-lice (family Cyamidae) on oceanic dolphins. Barros argued for the possible use of the presence or absence of Xenobalanus for stock differentiation of dolphins in the central east coast of Florida. His argument was justified by the presence of only one record of a dolphin stranded in the IRL with soft barnacles on its flukes in 17 years of stranding coverage. In contrast, the Harbor Branch photo catalog has evidence of over 24 dolphins with soft barnacles attached that were sighted within the IRL. Thus, there are dolphins with ectoparasites that inhabit the IRL for at least some portion of their lives. The distinction between “resident” and “oceanic” dolphins is somewhat undetermined; however, some characteristics of the prey contents analyzed in resident dolphins are consistent with IRL dolphin long-term habitat utilization. Thirty-eight of the resident dolphin stomach samples were analyzed and prey items were identified and partitioned into fish, cephalopods, and crustaceans. Samples analyzed from the resident population indicated that fish accounted for 99.2% of all prey items taken, and that the “fish-only” category occurred in 32 (84%) of samples. Of the fish prey items, 15 families were represented with Sciaenidae (croakers and drums) being the most dominant. Four species were present in more than 30% of the samples analyzed: spotted sea trout (Cynoscion nebulosus), silver perch (Bairdiella chrysoura), Atlantic croaker (Micropogonias undulates), and oyster toadfish (Opsanus tau). In terms of total prey biomass, all four of these prey species, plus striped mullet (Mugil cephalus), constituted the majority. All four of the most frequent prey species were present in samples taken from all four seasons, and are known to inhabit the IRL throughout the year, which may indicate that dolphins

55 utilize these prey species year-round. Unlike other estuaries where sciaenids move offshore to spawn, in the IRL some species, such as spotted sea trout and silver perch, remain in the estuary throughout their entire life cycles (Barros 1993). The fish most frequently identified in dolphin stomach samples are, for the most part, non-schooling fish associated with seagrass beds (Barros 1993). Many fish species in the IRL are patchily distributed with respect to vegetative cover and salinity (Mulligan and Snelson 1983). The patchiness with which prey items are distributed, along with their small school size, may influence dolphins to form smaller groups. The foraging hypothesis proposed by Rodman (1988) predicts that group size will adjust to resource abundance and dispersion. Effective prey detection and capture must be balanced with the capacity of a prey patch (Weller 1991). Increased group size provides more members for prey search, but in habitats with predictable but low concentrations of prey, perhaps fewer individuals are needed to locate a prey patch. Further, when a patch is located it can then accommodate all members of the smaller dolphin group, thereby decreasing conspecific competition (Weller 1991, Campbell et al. 2002). Therefore, group size may often be dependent on prey concentration (Terborgh and Janson 1986, Rodman 1988). Prey concentration or patchiness may have differential effects on group formations. Research conducted on certain terrestrial species has provided insights into the extent that foraging demands shape group size. Identification of prey species, prey biomass consumed, and foraging tactics may be more readily observed for terrestial, rather than aquatic animals. One study on lions found that foraging party group size was dependent on the size and abundance of prey (Caraco and Wolf 1979). In another study, data analyzed from ten habitats on lion social organization indicated no relationship between group size and food supply (Van Orsdol et al. 1985). Prey capacity was expected to influence foraging group sizes, although this was not found. The authors speculated that in addition to ecological factors, group size may be influenced by kinship and other proximate factors. Lion pride size, which consists of resident females, their offspring, and attendant males, does seem to vary with food abundance, especially during seasons of low prey availability (Van Orsdol et al. 1985, Packer et al. 1990). Thus, food supply factors may exert influence on social structure, but on a broader level than is apparent in a single sighting. The variability in prey abundance that lions experience may differ for IRL dolphins, which seem to experience fewer fluctuations in prey size and location. Barros (1993) did not find

56 seasonal differences in prey characteristics for dolphins found stranded within the IRL. Perhaps the ability to secure food resources year-round diminishes the need to form large, stable groups in dolphins, which could suffer from increased intra-group competition for prey. Favored prey species of IRL dolphins are all highly soniferous. Many of these species are demersal, solitary, cryptic or tend to hide in shelters or refuges which may make them harder to locate visually (Barros 1993). Barros (1993) suggested that dolphins may use passive listening in order to locate prey that may otherwise be difficult to find using visual cues. Dolphins may be selectively feeding on these sonorous fish and thus may have adapted foraging strategies appropriate to secure these resources. Maintaining small group sizes may allow more stealth and effective detection of these solitary, sound-producing fish.

Predation Pressure A number of shark species have been documented as bottlenose dolphin predators: white sharks (Carcharodon carcharias), tiger sharks (Galeocerdo cuvier), bull sharks (Carcharhinus leucas), sixgill sharks (Hexanchus griseus), and dusky sharks (Carcharhinus obscurus) (Heithaus 2001). Of these, bull sharks are known to inhabit surface, coastal, and brackish waters in temperate or warm temperate latitudes and are known to attack prey larger than themselves (Heithaus 2001). Bull sharks are known to inhabit the IRL, are distributed throughout inlets, seagrass flats, mangroves, sandy bottom, river mouths, and freshwater canals (Gilmore 1977). Dusky sharks and tiger sharks inhabit neritic zones and tiger sharks are often found in the offshore reef areas adjacent to the IRL (Gilmore 1977). To some extent, these shark species may prey on dolphins in the IRL. Several studies have documented shark predation on dolphins. Roughly 37% of dolphins surveyed in Moreton Bay, Queensland, bore scars from shark attacks (Corkeron et al. 1987). Of dolphins caught in gill nets off Natal, South Africa, 10.3% showed evidence of shark bites; however, only 1.2% of over 6000 sharks caught contained cetacean remains (Cockcroft et al. 1989). In Sarasota, about one-third of the non-calves examined during captures had shark bite scars (Wells et al. 1987). Tiger, dusky, and bull sharks, which are known bottlenose dolphin predators, occur throughout Sarasota Bay. Bull sharks, the most prevalent, occur both inshore and offshore, with abundance highest in the summer. During this season, dolphins were sighted more often in the shallower, more obstructed habitats, and in deeper waters, larger dolphin

57 groups were more common. Thus, bull shark abundance may influence dolphin range and grouping patterns (Wells et al. 1987). Predator occurrence in the IRL appears to be similar to that in Sarasota. Wells (1987) reported that of the 86 dolphins examined during a capture-release from 1975-1985, 21.9% noncalves bore obvious shark bite scars. The presence of shark bite scars on many of the photographed IRL dolphins indicates that shark predation exists to some degree and that some dolphins survive these attacks. Dolphin mortality due to shark predation has been estimated for a number of studies, although direct observations of predation are rare (Connor et al. 2000). Heithaus (2001) proposes that the stomach contents of various sharks be analyzed in order to identify major dolphin predators, after controlling for scavenging on dead, seriously ill or injured dolphins. To a degree, observing shark bite scars on cetaceans indicates the occurrence of predation attempts, but mortality rates can still be difficult to calculate. A large proportion of a population may bear scars, the results of unsuccessful predation attempts, but not experience high mortality. On the contrary, shark attacks may be frequently successful and leave few survivors. Nevertheless, dolphins probably employ strategies to reduce predation events, which may include forming groups to enhance predator detection and defense (Wells et al. 1980). Predation pressure may also influence mother-calf groups to form larger nursery schools, which may be more successful at fending off shark attacks (Wells et al. 1999). The trade-off between net energy requirements and predation risk probably shapes group size more than either factor by itself. In combination, both factors are the main determinants for individual fitness (Van Schaik and Van Hooff 1983). An optimal number of dolphins which balances both predator detection and resource needs lays the foundation for group size, and the evolution of social living further refines group sizes actually observed.

Annual Effects Group sizes were higher in the combined 1996-1998 time period than in all other study years considered individually. This difference in mean group size may reflect differences in resource distribution, movement, or recruitment related to annual variation (Connor et al. 2000). Other studies have noted annual variations in group size. Dolphins inhabiting Turneffe Atoll, Belize, exhibited larger group sizes in 1993 and 1994 than in 1992, 1995, or 1996, although these differences were not explained (Campbell et al. 2002). In the Aransas Pass area of

58 southeastern Texas, larger dolphin group sizes were found in colder than in warmer water years (Weller 1998). The decrease in group size may be related indirectly to decreases in primary productivity. In the winter season, as water temperature drops and daily photoperiod decrease, primary productivity decreases in estuaries (Lyford and Phinney 1968). Lower levels of net productivity may have a bottom-up effect on food availability for top-level predators such as bottlenose dolphins. During periods of food scarcity, dolphins may rely more on conspecifics to locate prey patches, which would account for bigger dolphin groups during colder periods. Sea surface temperature fluctuations due to year-scale climatic shifts such as ENSO (El Niño Southern Oscillation) warm water incursion events have been related to shifts in the home range of Pacific bottlenose dolphins along the California coast (Wells et al. 1990). Dolphins may be responding to shifts in prey movement patterns related to environmental factors. In addition, dolphins may have to alter their foraging strategies in the event that prey distributions change. One approach may involve adjusting group size to increase hunting efficiency. Although mean water temperatures were not found to differ among study years in the present study, other environmental differences, such as salinity, in the early years of the study may have contributed to shifts in resource distribution, which in turn drove dolphin optimal group size up. Salinity has been known to affect fish distributions in the Aransas Pass area to a certain degree, particularly at the larval stage (Weller 1998). Further integration of environmental variables and dolphin group data will need to be conducted in order to further investigate this possible link.

Seasonal Effects No seasonal variability was found in group size among IRL dolphins. Neither Julian calendar season nor hurricane season changes affected group size in the current study. Seasonal effects on group size in other studies suggested variations in the type, availability, and distribution of preferred prey items (Weller 1991, Weller 1998, Gubbins 2000, Caldwell 2001, Campbell et al. 2002). In each case, these authors proposed that seasonal shifts of prey characteristics may in turn affect dolphin foraging strategies. Dolphins in San Diego display lower group sizes in the spring season (Weller 1991). Upwelling during April, May, and June is thought to increase the nutrient level in the water, and many fish species spawn during this time of year. Both the increase in nutrients and change in reproductive cycles may make prey species

59 more uniformly distributed during the spring, which may lessen the need for large aggregations of prey-searching dolphins. Most of the fish species identified in the IRL occur year-round and have less dramatic fluctuations in concentration than in the Southern California Bight (Mulligan and Snelson 1983), which may explain the lack of seasonal variability in dolphin group size. Whether changes in environmental conditions over the course of a year impact IRL dolphin grouping strategies remains uncertain. Tucuxi dolphins (Sotalia fluviatilis) inhabiting Guanabarra Bay, Brazil, which is also an estuarine system, also do not display wide variation in group size among calendar seasons. This finding seems to be consistent among other tucuxi populations (Azevedo 2005). Rather than adjust group size, some Guanabarra dolphins appear to leave the area entirely when conditions fail to provide adequate food supply, decreasing total dolphin abundance. In contrast, prey conditions in the IRL seem to be more uniform throughout the year (Barros 1993), although further investigation into dolphin abundance and home range needs to be conducted. In the current study, water temperature changes did not seem related to group size. Failing to find a relationship between these two factors, however, must be viewed with caution. Group size changes may stem from a complicated relationship between dolphin grouping strategies and environmental conditions. Given the wide distribution of bottlenose dolphins throughout temperate and tropical oceans throughout the world (see Connor et al. 2000 for a review), dolphins appear to be very adaptable to changes in temperature (Reynolds et al. 2000, Barco et al. 1999). Although dolphins may not adjust group size in direct response to fluctuating sea surface temperatures, several fish species have been known to be sensitive to water temperature fluctuations (Gilmore 1977), and dolphins are thought to respond secondarily to changes in prey species movements (Barco et al. 1999, Weller 1991).

Diurnal Effects Diurnal changes in group size suggest that IRL dolphins respond to changes in environmental variables as the day progresses, forming larger groups in the afternoon than in the morning. Changes in group characteristics throughout the day period have been noted for other populations in addition to IRL dolphins. Increased feeding behaviors during crepuscular periods were noted for Pacific Coast bottlenose dolphins related to the higher activity levels of important prey species during nocturnal hours (Hanson and Defran 1993, Ward 1999). Many of these prey

60 species transition between resting and feeding modes during dusk and dawn, and it was suggested that prey caught in a transition state are more accessible and easy to capture (Hanson and Defran 1993). Other dolphin predators, such as sharks, may also be more active during crepuscular periods, (Heithaus 2001). Lower light levels throughout certain periods in the day may make dolphins more vulnerable to predator attacks (Lima and Dill 1990). An increase in potential predator activity during the late afternoon may also influence dolphins to be defensive and aggregate in anticipation of potential predator attacks. Cockroft et al. (1989) suggest that dolphins inhabiting coastal waters near Natal, South Africa consistently move offshore during the evening to avoid encounters with sharks, which move inshore to feed at night. Future research into diurnal changes in the behavioral states of IRL dolphins, as well as day period changes in both predator and prey activity levels, will further elucidate how dolphins respond to daily changes in their environment. Group Composition In the IRL, calf groups are typically larger than non-calf groups. Given the maternal energy demands of calf-rearing, specifically nursing and protection, the possibility of having assistance from other females may foster the aggregation of mother and calf pairs. Several studies from various parts of the world have found larger mean group sizes when calves are present (Sarasota Bay, Florida: Wells 1986; San Diego, California: Weller 1991; Gulf of Guayaquil, Ecuador: Félix 1997; Aransas Pass, Texas: Weller 1998; Santa Barbara, California: Marsh 2000; Turneffe Atoll, Belize: Campbell et al. 2002; Moray Firth, Scotland: Eisfeld 2003; Little Bahama Bank, Bahamas: Rogers et al. 2004; Charleston, South Carolina: Laska 2005). The primary variable leading to the formation of larger calf groups is probably calf protection. Predation pressure in many populations significantly contributes to larger calf-group sizes. Larger groups allow for more predator vigilance and defense (Crook 1976). In addition, associating with more conspecifics, particularly females, provides more potential care-givers for calves. Females have been known to associate more with other females in similar reproductive stages and rearing calves in a group may provide mutual benefits for females (Connor et al. 2000b, Mann et al. 2000)). The occurrence of larger calf groups in non-hurricane months may be related to seasonal cycles in breeding. In Fick’s (1995) one-year study, Indian River Lagoon dolphins engaged in more social behavior during dry (non-hurricane) months than in wet (hurricane) months. The

61 increase in social activity may indicate more mating during the dry spring season. Bottlenose dolphins gestate for approximately one year (Reynolds et al. 2003), and some studies have shown that dolphins typically give birth in spring or summer (Wells et al. 1987, Mann et al. 2000). Further, it is possible that conditions may be more favorable for rearing neonates in the summer, for the following reason. Dolphins rearing calves have often been associated with habitats high in seagrass coverage (Mann et al. 2000, Scott et al. 1990). Seagrass biomass in the IRL reaches a peak during summer (June and July) and a low in late winter (February-May) (Virnstein and Carbonara 1985). Mann et al. (2000) found peak birth months to be between October and December for dolphins inhabiting Shark Bay, Australia. Reproductive success was defined as the number of calves surviving at least three years (typical weaning age) within the ten-year study period. Water depth alone predicted female reproductive success; females encountered more often in shallower waters had more surviving offspring. Similar to the IRL, the shallower habitats have high seagrass bed coverage which provide habitat to many juvenile fish species (Barros 1993, Mann et al. 2000). In addition, dolphins were found to be more likely to conceive in the same birth season if they had lost a calf, regardless of season, which emphasizes the drive for high reproductive success. Calf mortality was estimated at 29% for first year of life, which suggests a high degree of predation pressure and other threats to the viability of young dolphins in this population. In Shark Bay, it is apparent that reproductive success strongly influences adult female behavior. Further research in the IRL that targets individual mother-calf pairs and their associates will need to be conducted in order to investigate female reproductive strategies. Differences in regional calf encounter rates indicate that there may be preferred areas for rearing calves in the IRL. The PSA and UME areas had higher calf-group encounter rates than both the SLR and ML areas. Habitat parameters may also differ among IRL regions. Greater numbers of calves have been sighted in certain regions of Sarasota Bay, which have been dubbed nursery areas. One such area, Palma Sola Bay, is enclosed and shallow with large expanses of productive seagrass meadows which probably provide enough food to meet energy requirements of lactating mothers. Palma Sola Bay’s shallow waters may also provide protection from predation for calves (Scott et al. 1990). Although all four IRL areas under current study have high seagrass coverage, the UME and PSA areas have actually experienced some of the greater declines in seagrass coverage (IRLNEP 2004). Current studies estimate that the area south of

62 Merritt Island through Palm Bay has experienced aout 70 percent of seagrass coverage loss since the 1940’s (IRLNEP 2004). Future research on fine-scale differences in habitat may determine whether conditions in some parts of the IRL are more favorable for rearing calves.

SITE FIDELITY AND RESIDENCY Many IRL dolphins exhibit a high degree of overall site fidelity. The time period of the present study spans nine years, but evidence for even longer-term site fidelity comes from resightings of dolphins freeze-branded by Odell and Asper from 1979-1981 (1990). Fourteen of the original 34 dolphins were sighted during the current study period, over 20 years after the initial freeze-brands were applied (Mazzoil et al., in press). At least some dolphins have inhabited the IRL for a long part of their lifetime, although how continuously is uncertain. The more frequently a dolphin is encountered may indicate site fidelity. The residency patterns of dolphins encountered near Cedar Keys, Florida were analyzed in a one-year, high effort study. Residence was quantifed as an index which represented the proportion of sightings of an identified individual to the number of complete surveys in a month. The more times an individual was sighted, the higher its average residence index (ARI)(Quintana-Rizzo and Wells 2001). The authors suggested that the positive relationship between sightings and ARI values confirms using the actual number of sightings of an individual as useful indicator of residency patterns. The current study used Site Fidelity Ratios (SFR) as an index of residency. Values indicate the proportion of possible sighting years in which individuals were sighted. Over longer study periods or larger study areas, it may be more prudent to use an SFR rather than an ARI to quantify long-term site fidelity, since sightings over years are pooled. Further, the size of the IRL necessitates that surveys of the total study area be conducted over several days, which decreases the probability that any one given individual will be sampled in a single survey day. In other words, although the individual may have been present in the IRL on a given day, it could have been ranging in a region outside the survey route. In addition to SFRs, encounter rates also indicate which regions dolphins utilize. Encounter rates were not very high for the PSA, UME, and SLR areas. The PSA had the highest percentage of encounters (81%), which is probably due to the greater survey effort which occurred in this area; however, individuals seem to move considerably among the UME, PSA,

63 and SLR areas. In contrast, dolphins sighted within the ML had high encounter rates in the ML region. Eight of the most frequently sighted individuals with sightings in the ML had no sightings anywhere else, including opportunistic sightings (off-survey). Evidence for site fidelity to the ML is further supported by an examination of an additional 33 individuals who were sighted at least five times. Only four of these individuals had sightings outside the ML, and all four had at least half of their sightings in the ML. The preceding analysis was consistent with recent distributional analyses of IRL dolphins (Mazzoil et al., in press). Most of the dolphins sighted in the ML were only sighted in the ML (n = 69, 76%). The regional site fidelity to the Mosquito Lagoon could indicate the existence of a separate ML community. Future research, which will include the more consistent, monthly sampling of the ML that began in 2002, will yield more information about whether there is longterm regional site fidelity. There is an emerging argument for managing IRL bottlenose dolphins as a distinct stock (Mazzoil et al., in press), given the frequency of year-round sightings and the limited access to open water. In the future, survey findings of the adjacent coastal and offshore areas could be compared with individual sighting histories in the IRL to examine any movement out of the IRL. Evidence for movement out of estuarine and into coastal waters exists for dolphins studied in the coastal waters near Charleston, South Carolina (Laska 2005, Zolman 2002). Whether dolphins occurring within the IRL travel out of the estuary system to the open ocean is still unknown. An amendment to the Marine Mammal Protection Act specifies a maximum number of marine mammals that can be removed due to human causes (Potential Biological Removal) based on the abundance, maximum rate of increase, and recovery factors for a given stock. Therefore, populations must be defined in order to establish these parameters and accurately manage them (Hohn 1997). Future research should address the home ranges and range limits of dolphins occurring within the IRL.

SOCIAL AFFILIATION Association analyses of bottlenose dolphins in the IRL reveal a complex social system in which individuals preferentially associate for short periods of time, forming some stable bonds over years. The range in level of association is wide for IRL dolphins; individuals may employ different strategies with respect to the formation, or lack of, stable bonds.

64 Dolphins in the Cedar Keys display a wide range in coefficients of associations, and the individuals with the highest level of association also had the highest residency values (QuintanaRizzo and Wells). Rogers et al. (2004) modified COA value categories to describe strength of association from five (Well et al. 1987) to three: low ≤ 0.39, moderate 0.40 – 0.79, and high ≥ 0.80. Although there were few IRL individuals who had high enough COAs to be compared with SFR values, the individuals whose maximum COA values were in the moderate or high category (n = 18) also had high SFR values. Eleven individuals had SFR values 0.86 or higher, and all but one had SFR values of 0.57 or higher. Evidence exists that some dolphin populations are actually assemblages of overlapping, sympatric communities (Wells 1986, Caldwell 2001). In Jacksonville, the Northern and Southern Communities were not only behaviorally distinct, but also displayed affiliation differences. The mean association index was higher in the Southern than in the Northern regions. Since the average group size was higher in the Southern region, individuals would be exposed to more potential affiliates than in the Northern region. Although both communities were not separated by geographic barriers, they appear to be socially distinct. Whether distinct regional communities exist in the IRL remains unanswered, although the high level of site fidelity to Mosquito Lagoon exhibited by many dolphins may suggest the existence of a separate social community or communties. A concentrated effort to reveal long-term association patterns of ML dolphins will be necessary to test the hypothesis of any distinct dolphin communities. Dolphin abundance or density may contribute to social affiliation patterns. If associating with any given individual yields similar fitness advantages, then it can follow that mean affiliation levels would be lower if dolphin density increased. Under these circumstances, dolphins would be exposed to more potential associates, and there would be no advantage to spending more time with any particular individual. Alternatively, if separate individuals differ in their abilities to offer protection, foraging skills, and mate procurement, an increase in density may not necessarily encourage more sociality (Connor 2000). Instead, dolphins may have more choice in whom they form bonds with, and once they do, those bonds may become more lasting. Overall male-male and female-female dyads have higher companionship than mixed sex pairs, which may indicate sex-differential survival strategies. The two within-sex-class association categories were examined, as well as the bonds between mothers and calves.

65

Male-Male Associations Males exhibited a wide range of maximum COA values: 0.10 – 0.83. Of particular interest were two males, LISL and HOLE, who had a very high HWI: 0.83. The pair was sighted together 23 times over eight years (1997-2004). Each was observed an additional four times apart from one another. LISL, who was included in the 2003 HERA capture-release project was aged as 15 in 2003 (Mazzoil, personal communication). The formation of stable male pairs has been reported for several bottlenose dolphin populations (Parsons et al. 2003, Smolker et al. 1992, Wells et al. 1987), and pair bonds seem to emerge around the time of sexual maturity. Another dyad with a high HWI (0.67) was FBC6 and FBC7, who were sighted once together in 1996, once in 1997, nine times in 1999, and four times in 2000. FBC6 was treated under the care of HBOI from 8/31/00 to 3/5/01, and was found dead on 6/12/01. From August, 1996 until August, 2000, FBC7 was seen without FBC6 only three times. Not all males appeared to form significant coalitions with other males, and year to year variability in male affiliations was present. This variation indicates diversity in survival and reproductive strategies which may relate more to the conditions which vary in “fitness payoffs” (Connor et al. 2000). A male like FBC7, who is more accustomed to maintaining strong affiliations, and who has lost his most significant companion, might be expected to form a new strong companionship with another male. Future observations on male affiliation patterns and how they change over life cycles will hopefully provide more insight into the complexities among male social relationships. Male associate selection is thought, in part, to reflect intra-sexual competition and mate choice (Connor et al. 2000). Competition will probably influence male relationships more when males are more likely to encounter other male rivals (Connor et al. 2000). Having an alliance partner may better chances of defense in an aggressive encounter or improve access to females. Many dolphins appear to range over most areas of the IRL, which may increase encounter probabilities among males. Pair-bonding between the adult males of the Sarasota community is considered the norm, and males sighted without partners are considered to be in a transition stage (Owen et al. 2002). In Shark Bay, Australia, males form pairs and trios in order to maintain consortships with individual females over varying lengths of time (Connor et al. 1992). As mentioned above, some IRL male pairs have high association levels, which may foster successful mating. Further identifications of males and analysis of association patterns could determine the

66 extent to which males form consortships. In addition, future research into genetic relatedness of male-male dyads may indicate whether nepotism plays a role into alliance formation.

Female-Female Associations Female associations did not range as high as male associations (range of maximum COA: 0.00-0.33) in the IRL. Females in other populations tend to affiliate to some degree based on reproductive status (Rogers et al. 2004), which may facilitate shared calf-rearing, more efficient mating consortships, and defense against male harassment (Connor et al. 2000). In the IRL, females with calves tended to have more affiliations than females who did not have calves, indicating that the formation of calf groups provides females with associates that can add in predator vigilance and baby-sitting. Maximum COA values tend to vary less across populations for females than males, and it often appears that females have larger numbers of casual associates (Smolker et al. 1992, Quintana-Rizzo and Wells 2001). In Shark’s Bay, although females were found to form fewer stable associations than males, they tended to form larger social networks (Smolker et al. 1992). The wide fluctuation in energy demands that females experience, particularly in calf-rearing, suggests that females benefit from a larger range of associates. Thus, while rearing calves, it may be beneficial for a female to affiliate with other currently nursing mothers. Limits to longer term companionships may stem from individual variation in reproductive cycles.

Mother-Calf Associations Mothers and calves had the strongest association levels of all dyads examined in the present study. Other analyses of mother-calf associations have indicated that companionship starts to decrease with time (Mann and Smuts 1999, Quintana-Rizzo and Wells 2001, Rogers et al. 2004). For example, a distinctly identified calf named c1PERI was sighted a total of 24 times (1999: 2, 2000: 5, 2001: 4, 2002: 3, 2003: 9, and 2004: 1), and his/her COA level rose to a peak in 2001 and then started to fall during the fourth and fifth year. The high COA level in 2004 may be a spurious value due to a low number of sightings. Some calves were never sighted without their mothers; c1REEL was sighted a total of ten times between 2000 and 2004, each time with his/her mother. Overall, the high mean level of association between mothers and calves (M = 0.67) and lasting strong companionship for the first three-four years of the calf’s life is similar to that

67 found in other populations, specifically Sarasota (Wells et al. 1987) and Shark Bay (Smolker et al. 1992). In both these areas and in the IRL, calf mortality due to shark predation poses real threats to developing calves. Prolonged associations between mothers and calves may benefit calf survivorship. Occasional sightings of calves without their mothers may indicate brief bouts of separations, perhaps to encourage self-reliance (Mann and Smuts 1999). Gibson and Mann (2003) found that the sociality of calves in Shark Bay often mirrored that of their mothers, in terms of number of associates and time spent associating with dolphins other than their mothers. The extent to which dolphins socialize may, to some degree, represent some vertical cultural transmission, with offspring either inheriting or learning alternative association strategies from their parents. Future longitudinal research on the social development of calves in the IRL may elucidate either consistent or variable patterns in sociality.

ENVIRONMENTAL CONSIDERATIONS The social structure of IRL dolphins is ultimately affected by ecological correlates such as prey distribution and predation pressure. This link suggests that the examination of grouping and affiliation patterns can provide some insight into how environmental conditions shape dolphin behavior. Further, it is important to address the health status of the IRL as a whole in order to ascertain how changes in the ecosystem will affect dolphin behavior. Past research on IRL dolphins has revealed that dolphins are associated with seagrass habitats (Barros 1993). Stomach contents analysis of stranded dolphins in the IRL indicates a preference for preying on fish rather than cephalopods, crustaceans, or other prey items (Barros and Odell 1990, Barros 1993). Stomach content analysis of dolphins inhabiting Sarasota Bay had similar findings: all 16 dolphins examined had fed exclusively on fish. Further, most of the fish species identified from both populations were associated with seagrasses. Seagrass beds are biologically very productive; they are known to have much higher densities of benthic invertebrates than sandy bottom substrates, even in small patches of seagrass (Virnstein et al. 1983). Seagrasses provide habitat for many species in the IRL and are regarded as a general indicator of biological integrity within the IRL (Barros 1993, Sigua et al. 2000). Research has suggested that declines in species diversity in seagrass communities within the IRL stem from water pollution, disruption in the natural pattern of water circulation, and changes in the freshwater inflows (Sigua et al. 2000).

68 Seagrass proliferation is largely controlled by available sunlight, which is in turn controlled by water transparency. Pollutant loading increases the turbidity of the water and, thereby, decreases the sunlight penetration through the water column. Limited outlet to the open ocean makes the IRL vulnerable to high pollution levels. The entire system is connected to coastal waters by only five inlets; this restricted connection can severely limit water exchange and lagoon flushing (Pitts 1996). The impact of increased pollution due to anthropogenic activities is serving to transform the lagoon from a seagrass or macrophyte-based system to an algal-based system (Sigua et al. 2000). This transformation may have an impact on the abundance and diversity of primary consumers, notably fish species, which in turn may have an effect on food source availability for fish predators such as bottlenose dolphins (Allen et al. 2001). As of 1996, seagrass acreage was considered to cover 62% of the total 113,073 potential coverage area (IRLNEP 2004). Some regions in the IRL, such as south of Merritt Island to Palm Bay area, which overlaps with the UME area, have lost up to 70% of seagrass coverage from the 1940’s (IRLNEP 2004). The present study provided nine years of baseline information on dolphin social ecology and spanned most parts of the IRL system. If grouping, residency, and association parameters are monitored along with environmental variables such as seagrass coverage, possible changes in the ways dolphins organize could reflect habitat transformation. For example, if seagrass beds provide habitat for dolphin calf-rearing, reduction in size or quality of seagrass areas may be detrimental to calf survivorship. Calf group encounter rates were higher in the PSA and UME areas, which have lower levels of seagrass coverage in general than the ML and SLR areas (IRLNEP 2004). If seagrass abundance continues to decline, dolphins may shift their distributions to areas with higher seagrass coverage or be otherwise affected. In order to understand the relationship IRL dolphins have with their habitat, more studies of fine-scale habitat utilization need to be conducted.

69

REFERENCES Allen, M. C., A. J. Read, J. Gaudet, and L. S. Sayigh. 2001. Fine scale habitat selection of foraging bottlenose dolphins Tursiops truncatus near Clearwater, Florida. Marine Ecology Progress Series 222:253-264. Amos, B., C. Schlotterer, D. Tautz. 1993. Social structure of pilot whales revealed by analytical DNA Profiling. Science 260:670-672. Azevedo, A. F., S. C. Viana, A. M. Oliveira, and M. Van Sluys. 2005. Group characteristics of marine tucuxis (Sotalia fluviatilis) (Cetacea: Delphinidae) in Guanabara Bay, southeastern Brazil. Journal of the Marine Biological Association of the United Kingdom 85:209-212. Baird, R. 2000. The killer whale: foraging specializations and group hunting. Pages 127-153 in J. Mann, R. C. Connor, P. L. Tyack, and H. Whitehead, eds. Cetacean Societies. University of Chicago Press, Chicago, IL. Baird, R. and H. Whitehead. 2000. Social organization of mammal-eating killer whales: group stability and dispersal patterns. Canadian Journal of Zoology 78: 2096-2105. Baldwin, J. D. & J. I. Baldwin. 1973. Role of play in social organization: comparative observations on squirrel monkeys (Saimiri). Primates 14:369-381. Ballance, L. T. 1990. Residence patterns, group organization, and surfacing associations of bottlenose dolphins in Kino Bay, Gulf of California, Mexico. Pages 267-283 in S. Leatherwod and R. R. Reeves, eds. The bottlenose dolphin, Academic Press, San Diego, CA. Ballance, L. T. 1992. Habitat use patterns and ranges of the bottlenose dolphin in the Gulf of California, Mexico. 1992. Marine Mammal Science 8:262-274. Barco, S. G., W. M. Swingle, W. A. McLellan, R. N. Harris, and D. A. Pabst. 1999. Local abundance and distribution of bottlenose dolphins (Tursiops truncatus) in the nearshore waters of Virginia Beach, Virginia. Marine Mammal Science 15:394-408. Bearzi, G., G. Notabartolo-Di-Sciara, E. Politi. 1997. Social Ecology of Bottlenose Dolphins in the Kvarnerić (Northern Adriatic Sea). Marine Mammal Science 13:650-668. Bejder, L., D. Fletcher, and S. Bräger. 1998 A method for testing association patterns of social animals. Animal Behaviour 56:719-725. Barros, N. B. 1993. Feeding ecology and foraging strategies of bottlenose dolphins on the central east coast of Florida. Ph.D. dissertation, University of Miami, Miami FL. 329 pp. Barros, N. B. and D. K. Odell. 1990. Food habits of bottlenose dolphins in the southeastern United States. Pages 309-328 in S. Leatherwod and R. R. Reeves, eds. The bottlenose dolphin, Academic Press, San Diego, CA.

70 Barros, N. B. and R. S. Wells. 1998. Prey and feeding patterns of resident bottlenose dolphins (Tursiops truncatus) in Sarasota Bay, Florida. 1998. Journal of Mammalogy 79:10451059. Bejder, L., D. Fletcher, and S. Bräger. 1998. A method for testing association patterns of social animals. Animal Behavior 56:719-725. Bigg, M. A., P. F. Olesiuk, G. M. Ellis, J. K. B. Ford, and K. C. Balcomb. 1990. Social organization and genealogy of resident killer whales (Orcinus orca) in the coastal waters of British Columbia and Washington State. Pages 383-405 in P.S. Hammond, S. A. Mizroch, and G. P. Donovan, eds. Individual recognition of cetaceans: Use of photoidentification and other techniques to estimate population parameters. Report of the International Whaling Commission (Special issue 12). Bräger, S., B. Würsig, A. Acevedo, and T. Hennigsen. 1994. Association patterns of Bottlenose dolphins (Tursiops truncatus) in Galveston Bay, Texas. Journal of Mammalogy 75:431437. Bräger, S. 1999. Association patterns in three populations of Hector’s dolphin, Cephalorhynchus hectori. Canadian Journal of Zoology 77:13-19. Breder, Jr., C. M. 1954. Equations descriptive of fish schools and other animal aggregations. Ecology 35:361-370. Bruno, S., E. Politi, and G. Bearzi. 2004. Social organisation of a common dolphin community in the eastern Ionian Sea: evidence of a fluid fission-fusion society. European Research on Cetaceans 15:49-51. Cairns, S. J. and S. J. Shwager. 1987. A comparison of association indices. Animal Behavior 35:1454-1469. Caldwell, M. 2001. Social and genetic structure of bottlenose dolphin (Tursiops truncatus) in Jacksonville, Florida. Ph.D. dissertation, University of Miami, Miami, FL. 143 pp. Campbell, G. S., B. A. Bilgre, and R. H. Defran. 2002. Bottlenose dolphins (Tursiops truncatus) in Turneffe Atoll, Belize: occurrence, site fidelity, group size, and abundance. Aquatic Mammals 28:170-180. Chapman, C. A. and R. W. Wrangham. 1993. Range use of the forest chimpanzees of Kibale: implications for the understanding of chimpanzee social organization. American Journal of Primatology 31:263-273. Chivers, B. L. and P. J. Corkeron. 2002. Association patterns of bottlenose dolphins (Tursiops aduncus) off Point Lookout, Queensland, Austrailia. Canadian Journal of Zoology 80: 973-979. Christal, J. 1998. An analysis of sperm whale social structures: patterns of association and genetic relatedness. Doctoral dissertation, Dalhousie University, Halifax, Nova Scotia, Canada. 192 pp. Christal, J. and H. Whitehead. 2001. Social Affiliations within Sperm Whale (Physeter macrocephalus) Groups. Ethology 107:323-340.

71 Clement, D. 1998. Assessment of Tursiops truncatus utilization and prey presence within proposed feeding habitats in the Indian River Lagoon, FL. M. Sc. thesis, Florida Institute of Technology, FL. 181 pp. Cockcroft, V.G., G. Cliff, and G.J.B. Ross. 1989. Shark predation on Indian Ocean bottlenose dolphins Tursiops truncatus off Natal, South Africa. South African Journal of Zoology 24:305-310. Connor, R. C. 2000. Group living in whales and dolphins. Pages 199-218 in J. Mann, R. C. Connor, P. L. Tyack, and H. Whitehead, eds. Cetacean Societies. University of Chicago Press, Chicago, IL. Connor, R. C. and M. R. Heithaus. 1996. Approach by a great white shark elicits flight response in bottlenose dolphins. Marine Mammal Science 12:602-606. Connor, R. C., A. J. Read, and R. Wrangham. 2000a. Male reproductive strategies and social bonds. Pages 247-269 in J. Mann, R. C. Connor, P. L. Tyack, and H. Whitehead, eds. Cetacean Societies. University of Chicago Press, Chicago, IL. Connor, R. C., R. S. Wells, J. Mann, A. J. Read. 2000b. The bottlenose dolphin: social relationships in a fission-fusion society. Pages 91-126 in J. Mann, R. C. Connor, P. L. Tyack, and H. Whitehead, eds. Cetacean Societies. University of Chicago Press, Chicago, IL. Connor, R. C., J. Mann, P. L. Tyack, and H. Whitehead. 1998. Social evolution in toothed whales. Trends in Ecology and Evolution 13:228-232. Connor, R. C., R. A. Smolker, and A. F. Richards. 1992. Two levels of alliance formation among bottlenose dolphins (Tursiops sp.). Proceedings of the National Academy of Science, USA 89:987-990. Corkeron, P. J., R. J. Morris, and M. M. Bryden. 1987. Interactions between bottlenose dolphins and sharks in Moreton Bay, Queensland. Aquatic Mammals 13:109-113. Crook, J. H., J. E. Ellis, and J. D. Goss-Custard. 1976. Mammalian social systems: Structure and Function. Animal Behavior 24:261-274. Defran, R. H. 2004. The distribution of Indian River Lagoon bottlenose dolphins. Technical report. Harbor Branch Oceanographic Institution. 11 pp. Defran, R. H. and D. W. Weller. 1999. The occurrence, distribution, and site fidelity of bottlenose dolphins (Tursiops truncatus) in San Diego, California. Marine Mammal Science 15:366-380. Eisfeld, S. J. 2003. The social affiliation and group composition of bottlenose dolphins (Tursiops truncatus) in the outer southern Moray Firth, NE Scotland. M.Sc. thesis. University of Wales, Bangor, Wales. 85 pp. Félix, F. 1997. Organization and social structure of the coastal bottlenose dolphin Tursiops truncatus in the Gulf of Guayaquil, Ecuador. Aquatic Mammals 23:1-16 Fertl, D. 1994. Occurrence patterns and behavior of bottlenose dolphins (Tursiops truncatus) in the Galveston Ship Channel, Texas. Texas Journal of Science 46:299-317.

72 Fick, K. 1995. An investigation of abundance, group size, and activities of Atlantic bottlenose dolphins in Indian River, Brevard County, Florida. M. Sc. thesis, Florida Institute of Technology, FL. 98 pp. Flower, W. H. 1883. On whales, present and past and their probable origin. Proceedings of the Zoological Society of London 1883:466-513. Funston, P. J., M. G. L. Mills, H. C. Biggs, and P. R. K. Richardson. 1998. Hunting by male lions: ecological influences and socioecological implications. Animal Behavior 56:13331345. Gibson, Q. A. and J. Mann. 2003. Individual variability in wild bottlenose dolphin calf social relationships in Shark Bay, Western Australia. Paper presented at the Fifteenth Biennial Conference on the Biology of Marine Mammals, Greensboro, NC, December. Gaskin, D.E. 1982. The Ecology of Whales and Dolphins. Heinemann Educational Books, Ltd., London, UK. Gilmore, R. G. 1977. Fishes of the Indian River Lagoon and adjacent waters, Florida. Bull. of the Florida State Museum, Biol. Sci. 22:101-147. Ginsberg, J. R., and T. P. Young. 1992. Measuring association between individuals or groups in behavioral studies. Animal Behavior 52:377-379. Grier, J. W. 1984. Biology of Animal Behavior. Times Mirror/Mobey, College Publishing, St. Louis, MO. Gubbins, C. 2000. Behavioral ecology and social structure of bottlenose dolphins in South Carolina. Ph.D. dissertation, University of Nevada Reno, Reno, NV. 162 pp. Gubbins, C. 2002. Use of home ranges by resident bottlenose dolphins (Tursiops truncatus) in a South Carolina estuary. Journal of Mammalogy 83:178-187. Gygax, L. 2002. Evolution of group size in the dolphins and porpoises: interspecific consistency of intraspecific patterns. Behavioral Ecology 13:583-590. Ha, R. R., P. Bentzen, J. Marsh, and J. C. Ha. 2003. Kinship and association in social foraging northwestern crows (Corvus caurinus). Bird Behavior 15:65-75. Hanson, M. T., and R. H. Defran. 1993. The behavior and feeding ecology of the Pacific Coast bottlenose dolphin, Tursiops truncatus. Aquatic Mammals 19.3:127-142. Heithaus, M. R. 2001. Predator-prey and competitive interactions between sharks (order Selachii) and dolphins (suborder odontoceti): a review. J. Zool., Lond. 253:53-68. Hinde, R. A. 1976. Interactions, relationships, and social structure. Man 11:1-17 Hohn, A. 1997. Design for a multiple-method approach to determine stock structure of bottlenose dolphins in the mid-Atlantic. NOAA Technical Memorandum NMFS-SEFSC401, 22 pp. Irvine, B., M. D. Scott, R. S. Wells, J. H. Kaufman. 1981. Movements and activities of the Atlantic bottlenose dolphin, Tursiops truncatus, near Sarasota, Florida. Fishery Bulletin 79:671-687.

73 IRLNEP. 2004. Indian River Lagoon: the health and future of this Estuary of National Significance. http://sjr.state.fl.us/programs/outreach/pubs/irl_insert1.html (accessed June 20, 2005). Jarman, P. J. 1974. The social organisation of antelope in relation to their ecology. Behaviour 48:215-267. Krebs, C. J. 1999. Ecological methodology. Addison-Wesley Educational Publishers, Inc., Menlo Park, CA. Laska, D. L. 2005. Occurrence and distribution of bottlenose dolphins, Turiops truncatus, in the coastal waters near Charleston, South Carolina M.Sc. thesis. College of Charleston, Charleston, SC. 118 pp. Leatherwood, S. 1975. Some observations of feeding behavior of bottle-nosed dolphins (Tursiops truncatus) in the northern Gulf of Mexico and (Tursiops cf. T. gilli) off southern California, Baja California, and Nayarit, Mexico. Mar. Fish. Rev. 37:10-16. Lettevall, E., C. Richter, N. Jaquet, E. Slooten, S. Dawson, H. Whitehead, J. Christal, and P. Mcall Howard. 2002. Social structure and residency in aggregations of male sperm whales. Canadian Journal of Zoology 80:1189-1196. Lima, S. L. and L. M. Dill. 1989. Behavioral decisions made under the risk of predation: a review and prospectus. Canadian Journal of Zoology 68:619-640. Lyford, J. H. and H. K. Phinney. 1968. Primary productivity and community structure of an estuarine impoundment. Ecology 49:854-866. Mann, J. 2000. Unraveling the dynamics of social life: long-term studies and observational methods. Pages 45-64 in J. Mann, R. C. Connor, P. L. Tyack, and H. Whitehead, eds. Cetacean Societies. University of Chicago Press, Chicago, IL. Mann, J., R. C. Connor, L. M. Barre, and M. R. Heithaus. 2000. Female reproductive success in bottlenose dolphins (Tursiops sp.): life history, habitat, provisioning, and group-size effects. Behavioral Ecology 11:210-219. Mann, J. and B. Smuts. 1999. Behavioral development in wild bottlenose dolphin newborns (Tursiops sp.). Behavior 136:529-566. Marsh, J.A. 2000. School characteristics and social affiliation patterns of California bottlenose dolphins (Tursiops truncatus) M. Sc. thesis, San Diego State University, San Diego, CA. 84 pp. Maze-Foley, K. and B. Würsig. 2002. Pattens of social affiliation and group composition for bottlenose dolphins (Tursiops truncatus) in San Luis Pass, Texas. Gulf of Mexico Science, 2002:122-134. Mazzoil, M., S. McCulloch, E. Murdoch, and R.H. Defran. 2004a. The distribution of bottlenose dolphins in the Indian River Lagoon, Florida: applications to epidemiology. Abstracts of the Southeast and Mid-Atlantic Marine Mammal Symposium. Ft. Pierce, FL. March 26 – 28.

74 Mazzoil, M., S. D. McCulloch, R. H. Defran, and E. Murdoch. 2004b. The use of digital photography and analysis for dorsal fin photo-identification of bottlenose dolphins. Aquatic Mammals 30:209-219. Mazzoil, M., S. D. McCulloch, R. H. Defran. In Press. Observations on the site fidelity of bottlenose dolphins (Tursiops truncatus) in the Indian River Lagoon, Florida. In Press, Florida Scientist. Mazzoil, M., S. D. McCulloch, R. H. Defran. 2002. Occurrence and distribution of Atlantic bottlenose dolphins (Tursiops truncatus) in the Indian River Lagoon (1996-2001). Technical Report, Dolphin Research and Conservation Program, Harbor Branch Oceanographic Institution. McComb, K., C. Moss, and S. M. Durant. 2001. Matriarchs as repositories of social knowledge in African elephants. Science 292:5516-5526. Möller, L. M., L. B. Beheregaray, R. G. Harcourt, and M. Krützen. 2001. Alliance membership and kinship in wild male bottlenose dolphins (Tursiops aduncus) of southeastern Australia. Proceedings of the Royal Society of London, Series B 268: 1941-1947. Mulligan, T. J. and F. F. Snelson, Jr. 1983. Summer-season populations of epibenthic marine fishes in the Indian River Lagoon system, Florida. Florida Scientist 46:250-276. Norris, K. S. and T. P. Dohl. 1980. The structure and function of cetacean schools. Pages 211261in L. M. Herman, ed. Cetacean Behavior: Mechanisms and Functions. WileyInterscience, 211-261. Norris, K. S., B. Würsig, R. S. Wells, and M. Würsig. 1994. The Hawaiian spinner dolphin. University of California Press, Berkeley, CA. Nyakaana, S., P. Arctander, and H. R. Siegismund. 2002. Population structure of the African savannah elephant inferred from mitochondrial control region sequences and nuclear microsatellite loci. Heredity 89:90-98. Odell, D., E. Asper. 1990. Distribution and movements of freeze-branded bottlenose dolphins in the Indian and Banana Rivers, Florida. Pages 515-540 in S. Leatherwood and R. R. Reeves, eds. The bottlenose dolphin, Academic Press, San Diego, CA. Parsons, K., J. W. Durban, D. E. Claridge, K. E. Balcomb, L. R. Noble, and P. T. Thompson. 2003. Kinship as a basis for alliance formation between male bottlenose dolphins, Tursiops truncatus, in the Bahamas. Animal Behavior 66:185-194. Pitcher T. J., A. E. Magurran, and I. J. Winfield. 1982. Fish in larger shoals find food faster. Behavioral Ecology and Sociobiology 10:149-151. Pitts, P. A. 1996. An investigation of tidal and low-frequency current processes in Florida’s Indian River Lagoon. Florida Scientist 59:205-215. Quintana-Rizzo, E. and R.S. Wells. 2001. Resighting and association patterns of bottlenose dolphins (Tursiops truncatus) in the Cedar Keys, Florida: insights into social organization. Canadian Journal of Zoology 79:447-456. Read, A. J., K. W. Urian, B. Wilson, and D. M. Waples. 2003. Abundance of bottlenose dolphins in the bays, sounds, and estuaries of North Carolina. Marine Mammal Science 19:59-73.

75 Reynolds, J. E. III, R. S. Wells, and S. D. Eide. 2003. The bottlenose dolphin: biology and conservation. University Press of Florida, Gainsville, FL. Rodman, P. S. 1988. Resources and group sizes of primates. Pages 83-108 in C. N. Slobodchikoff, ed. The ecology of social behavior, Academic Press, San Diego, CA. Rogers, C. A., B. J. Brunnick, D. L. Herzing, and J. D. Baldwin. 2004. The social structure of bottlenose dolphins, Tursiops truncatus, in the Bahamas. Marine Mammal Science 20:688-708. Rossbach, K. A. and D.L. Herzing. 1999. Inshore and offshore bottlenose dolphin (Tursiops truncatus) communities distinguished by association patterns near Grand Bahamas Island, Bahamas. Canadian Journal of Zoology 77:581-592. Rudin, M. 1991. An assessment of distribution, group composition, and behavior of bottlenose dolphin, Tursiops truncatus, through photoidentification, in the Indian River Lagoon system, Florida. M. Sc. thesis, Florida Institute of Technology, FL. 39 pp. Scott, M. D., R. S. Wells, A. B. Irvine. 1990. A long-term study of bottlenose dolphins on the west coast of Florida. Pages 235-244 in S. Leatherwood and R. R. Reeves, eds. The bottlenose dolphin, Academic Press, San Diego, CA. Shane, S. H. 1990. Behavior and ecology of the bottlenose dolphins at Sanibel Island, Florida. Pages 245-65 in S. Leatherwood and R. R. Reeves, eds. The bottlenose dolphin, Academic Press, San Diego, CA. Shane, S. H., R. S. Wells, and B. Wursig. 1986. Ecology, behavior and social organization of the bottlenose dolphin: a review. Marine Mammal Science 2:34-63. Sigua, G. C., J. S. Steward, W. A. Tweedale. 2000. Water-quality monitoring and biological integrity assessment in the Indian River Lagoon, Florida: status, trends, and loadings (1988-1994). Environmental Management 25:199-209. Smolker, R. A., A. F. Richards, R. C. Connor, and J. W. Pepper. 1992. Sex differences in patterns of association among Indian Ocean bottlenose dolphins. Behaviour 123:38-68. Spellman, A. C. 1991. A study of the distribution, identity, and activities of the Atlantic bottlenose dolphin, Tursiops truncatus, in the Indian River Lagoon, Florida, based on photoidentification. M. Sc. thesis, Florida Institute of Technology, FL. 42 pp. Stolen, M. K., D. K. Odell, and Barros, N. B. 2002. Growth of Bottlenose dolphins (Tursiops truncates) from the Indian River Lagoon system, Florida, U.S.A. Marine Mammal Science 18:348-357. Tayler, C. K. and G. S. Saayman. 1972. The social organisation and behavior of dolphins (Tursiops aduncus) and baboons (Papio ursinus): some comparisons and assessments. Ann. Cape Prov. Mus. Nat. His. 9:11-49. Tayler, C. K. and G. S. Saayman. 1973. Imitative behaviour by Indian Ocean bottlenose dolphins (Tursiops truncatus) in captivity. Behaviour 44:286-298. Terborgh, J. and C. H. Janson. 1986. The socioecology of primate groups. Annual Review of Ecology and Systematics 17:111-136.

76 Tyack, P. L. 2000. Functional aspects of cetacean communication. Pages 270-307 in J. Mann, R. C. Connor, P. L. Tyack, and H. Whitehead, eds. Cetacean Societies. University of Chicago Press, Chicago, IL. Urian, K. W. and R. S. Wells. 1996. Bottlenose dolphin photo-identification workshop: March 21-22, 1996, Charleston, South Carolina. NOAA Tech. Mem. NMFS-SEFSC-393. Valsecchi, E., P. Hale, P. Corkeron, and W. Amos. 2002. Social structure in migrating humpback whales (Megaptera novaeangliae) Molecular Ecology 11:507-520. Van Orsdol, K. G., J. P. Hanby, and J. D. Bygott. 1985. Ecological correlates of lion social organization (Panthera leo). Journal of Zoology, London (A) 206:97-112. Van Schaik, C. P.and M. A. Van Noordwijk. 1983. On the ultimate causes of primate social systems. Behaviour 85:91-117. Ward, B. G. 1999. Movement patterns and feeding ecology of Pacific Coast bottlenose dolphins (Tursiops truncatus). MA thesis, San Diego State University, San Diego, CA. 98 pp. Weller, D. W. 1991. The social ecology of Pacific coast bottlenose dolphins. MA thesis, San Diego State University, San Diego, CA. 91 pp. Weller, D. W. 1998. Global and regional variation in the biology and behavior of bottlenose dolphins. Ph.D. dissertation, Texas A&M University, College Station, TX. 142 pp. Wells, R. S. 1986. Structural aspects of dolphin societies. University of California, Santa Cruz, CA. Ph.D. dissertation. 234 pp. Wells, R. S., D. J. Boness, and G. B. Rathbun. 1999. Behavior. Pages 324-422 in J. E. Reynolds and S. A. Rommel, eds. Biology of Marine Mammals. Smithsonian Institution Press, Washington, D.C. Wells, R. S., L. J. Hansen, A. Baldridge, T. P. Dohl, D. L. Kelly, and R. H. Defran. 1990. Northward extension of the range of bottlenose dolphins along the California coast. Pages 421-31 in S. Leatherwood and R. R. Reeves, eds. The bottlenose dolphin, Academic Press, San Diego, CA. Wells, R. S., M. D. Scott, and A. B. Irvine. 1987. The social structure of free-ranging bottlenose dolphins. Pages 247-305 in Current Mammalogy, vol. 1, H. Genoways, ed. Plenum Press, New York, NY. Wells, R. S., M. D. Scott, and A. B. Irvine. 1980. The social ecology of inshore odontocetes. Pages 263-217 in Cetacean Behavior: Mechanisms and Processes, L. M. Herman, ed. Wiley & Son, New York, NY. White, F. J. 1992. Pygmy chimpanzee social organization: variation with party size and between study sites. American Journal of Primatology 26:203-214. Whitehead, H. S. 1985. Humpback whale breaching. Invest. Cetacea 17:117-155. Whitehead, H. S. 1993. The behavior of mature male sperm whales on the Galapagos breeding grounds. Canadian Journal of Zoology 71:689-99. Whitehead, H. S. 1989. Formations of foraging sperm whales, Physeter macrocephalus, off the Galápagos Islands. Canadian Journal of Zoology 67:2131-2139.

77 Whitehead, H. S. 1995. Investigating structure and temporal scale in social organizations using identified individuals. Behavioral Ecology 6:199-208. Whitehead, H. S. 1999a. Testing association patterns of social animals. Animal Behaviour 57:F26-F29. Whitehead, H. S. 2004. Programs for analyzing social structure. SOCPROG 2.2; http://is.dal.ca/~whitelab/index.htm (accessed December, 2004). Whitehead, H. S. , L. Bejder, and C. A. Ottensmeyer. 2005. Testing association patterns: issues arising and extensions. Animal Behaviour 69:e1-e6. Whitehead, H. S., J. Christal, and P.L. Tyack. 2000. Studying cetacean social structure in space and time. Pages 65-87 in J. Mann, R. C. Connor, P. L. Tyack, and H. Whitehead, eds. Cetacean Societies. University of Chicago Press, Chicago, IL. Whitehead, H. S. and L. Weilgart. 2000. The sperm whale: social females and roving males. Pages 154-172 in J. Mann, R. C. Connor, P. L. Tyack, and H. Whitehead, eds. Cetacean Societies. University of Chicago Press, Chicago, IL. Wilson, B., P. M. Thompson, and P. S. Hammond. 1997. Habitat use by bottlenose dolphins: seasonal distribution and stratified movement patterns in the Moray Firth, Scotland. Journal of Applied Ecology 34:1365-1374. Wilmhurst, J. F., J. M. Fryxell, B. P. Farm, A.R.E. Sinclair, and C.P. Henschel. 1999. Spatial distribution of Serengeti wildebeest in relation to resources. Canadian Journal of Zoology 77:1223-1232. Würsig, B. 1986. Delphinid foraging strategies. Pages 347-359 in R. J. Schusterman, J. A. Thomas, and F. G. Wood, eds. Dolphin cognition and behavior: a comparative approach, Lawrence Erlbaum Associates, Hillsdale, NJ. Zolman, E. S. 2002. Residence patterns of bottlenose dolphins (Tursiops truncatus) in the Stono River Estuary, Charleston County, South Carolina, U.S.A. Marine Mammal Science 18:879-89.

78

APPENDIX SITE FIDELITY OF DOLPHINS SIGHTED NINE OR MORE TIMES

79 Sight

1996-

ID

Freq

98

1999

2000

2001

2002

2003

2004

SFR

BULB

10

0

0

0

0

0

7

3

0.29

COGS

10

0

0

0

0

0

6

4

0.29

CURL

9

0

4

5

0

0

0

0

0.29

ACID

13

7

3

3

0

0

0

0

0.43

CATA

10

0

0

0

0

3

4

3

0.43

DING

12

0

0

0

3

3

6

0

0.43

DUPS

9

0

0

0

0

3

5

1

0.43

EASY

11

4

1

6

0

0

0

0

0.43

FETA

10

0

0

0

0

3

6

1

0.43

FRED

11

0

0

0

1

0

9

1

0.43

PACU

10

0

0

0

0

3

6

1

0.43

SISS

9

0

0

0

0

2

4

3

0.43

UUMA

10

0

0

3

0

1

6

0

0.43

WHIT

9

0

1

5

3

0

0

0

0.43

WTIP

10

0

0

0

0

1

6

3

0.43

WWWW

9

0

0

0

0

1

5

1

0.43

ATIE

10

0

0

1

2

2

5

0

0.57

BIGS

10

4

4

1

1

0

0

0

0.57

CITY

9

0

0

0

2

1

3

3

0.57

CLIP

9

0

0

0

2

2

4

1

0.57

ELMU

9

0

0

1

3

0

4

1

0.57

JOBO

9

0

1

3

1

0

4

0

0.57

LDHS

16

3

6

5

2

0

0

0

0.57

LIBI

9

0

0

3

3

1

0

2

0.57

MOON

10

0

0

0

1

1

7

1

0.57

NIBB

13

0

0

2

1

2

8

0

0.57

ORTH

10

0

0

0

1

2

6

1

0.57

PARN

12

0

0

0

1

3

6

2

0.57

REEL

10

0

0

1

0

1

6

2

0.57

RIGA

10

0

0

0

2

1

4

3

0.57

SLAB

12

0

1

1

0

6

4

0

0.57

SMAS

10

2

0

0

2

1

5

0

0.57

STEP

9

0

0

0

1

2

5

1

0.57

TTIP

15

1

2

0

0

0

8

4

0.57

80 ZODI

9

0

0

0

2

1

5

1

0.57

CROS

11

0

0

1

2

1

5

2

0.71

DBLS

9

4

0

0

2

2

0

1

0.71

DENT

22

0

0

3

5

6

7

1

0.71

GOST

11

0

0

3

2

1

4

1

0.71

HILO

15

0

0

2

3

2

6

2

0.71

KIKO

14

3

4

2

2

0

3

0

0.71

NICA

11

0

0

1

5

1

3

1

0.71

PUNC

9

0

0

3

1

2

2

1

0.71

REGR

16

0

0

2

2

3

5

4

0.71

ROCK

10

0

0

1

1

2

5

1

0.71

RUAL

12

0

3

2

0

2

3

2

0.71

SIGM

10

2

1

3

2

0

2

0

0.71

SMIL

11

0

0

1

2

3

2

3

0.71

STIP

14

4

2

2

0

0

4

2

0.71

TSLW

13

3

1

5

1

0

3

0

0.71

YUNG

9

1

0

0

3

1

3

1

0.71

ZIGG

14

0

1

8

1

0

3

1

0.71

BANG

14

2

1

1

3

0

5

2

0.86

BFLG

18

3

4

3

2

0

5

1

0.86

CHIC

18

0

1

1

1

2

9

4

0.86

CHIP

12

2

3

1

0

1

2

3

0.86

COAT

10

1

1

2

0

2

2

2

0.86

CONE

14

1

0

1

2

4

3

3

0.86

COUT

11

4

2

1

1

2

1

0

0.86

DIGG

22

4

6

3

3

1

5

0

0.86

DSLA

12

1

2

1

1

1

6

0

0.86

EMPI

14

0

1

2

2

2

5

2

0.86

FANG

26

4

4

3

3

4

8

0

0.86

FLAP

20

6

3

1

3

1

6

0

0.86

GASH

20

2

0

5

3

3

6

1

0.86

HAND

14

1

1

2

1

0

8

1

0.86

INCH

10

0

1

1

2

2

2

2

0.86

JIMM

18

1

5

2

0

1

7

2

0.86

LIPN

17

2

3

3

0

1

6

2

0.86

LIPS

10

1

1

1

3

1

3

0

0.86

81 MFIN

13

0

1

1

1

2

7

1

0.86

MLAS

13

4

0

1

4

1

2

1

0.86

MSIG

15

0

1

2

3

2

6

1

0.86

OVER

15

0

1

1

3

1

8

1

0.86

PERW

10

2

1

1

0

1

4

1

0.86

QMOO

20

0

6

7

1

1

3

2

0.86

REPE

15

1

1

4

4

0

4

1

0.86

RING

24

0

4

6

2

1

9

2

0.86

SAIN

12

2

1

3

1

4

1

0

0.86

SAWB

23

0

4

4

4

5

4

2

0.86

SLTV

12

2

0

1

2

1

5

1

0.86

STUB

14

1

1

1

1

0

8

2

0.86

TOOT

14

1

0

2

1

4

4

2

0.86

TOPS

9

1

2

2

2

1

1

0

0.86

TSTR

9

2

0

1

1

1

3

1

0.86

TTTP

15

3

0

2

3

1

5

1

0.86

UPTN

11

2

1

4

2

1

1

0

0.86

ZIPP

26

0

4

6

5

2

6

3

0.86

ALLG

17

1

2

3

2

1

5

3

1.00

BELO

20

1

3

3

2

4

5

2

1.00

CHEK

20

2

1

3

2

3

8

1

1.00

DNLA

25

1

5

3

1

1

11

3

1.00

EVEN

21

4

5

3

2

2

4

1

1.00

FBC7

28

1

9

5

4

3

5

1

1.00

HCBT

27

5

8

3

2

1

7

1

1.00

HOLE

33

6

2

5

4

5

5

6

1.00

KEYP

19

1

2

5

1

1

5

4

1.00

LEDV

15

2

2

1

2

1

5

2

1.00

LISL

32

4

3

7

4

4

4

6

1.00

LOWR

34

3

4

6

5

3

12

1

1.00

MKEY

22

6

2

2

4

4

3

1

1.00

MOMW

12

1

3

2

1

1

3

1

1.00

PATC

17

1

2

2

1

2

7

2

1.00

PERI

26

3

6

4

5

2

5

1

1.00

POJI

17

2

1

3

1

3

4

3

1.00

PTNT

17

4

4

2

1

1

3

2

1.00

82 SFIN

28

5

5

7

3

2

4

1

1.00

SHAM

18

1

1

1

3

1

8

2

1.00

SKIN

18

1

2

2

2

4

5

2

1.00

SMOO

28

6

7

5

1

1

7

1

1.00

ABSTRACT OF THE THESIS Group Characteristics and Social Affiliation Patterns of Bottlenose Dolphins (Tursiops truncatus) Inhabiting the Indian River Lagoon, Florida by Erin Elizabeth Bardin Master of Science in Interdisciplinary Studies: Animal Behavior San Diego State University, 2005 A long-term investigation of the group characteristics and social affiliation patterns of Atlantic coast bottlenose dolphins was carried out in the Indian River Lagoon (IRL), Florida. Dolphin group characteristics and photo-identification data were collected during 477 boat-based surveys conducted in the IRL from August, 1996 to June, 2004. Survey time totaled over 1874 h, during which 615 h were spent in direct observation of dolphins. A total of 2,132 groups were encountered and photographed over the study period. Group sizes ranged from 1 to 40, with an average overall group size of 4.1 (SD = 3.43, median = 3). Group size was higher a) between 1996 – 1998 (mean = 5.9) than between 1999 – 2004 (mean = 4.0), and b) for sightings in the afternoon (mean = 4.4) than in the morning (mean = 3.7). Unlike some other inshore populations, which have larger groups in the fall and winter months, the size of IRL dolphin groups did not vary by season. Group size overall (mean = 4.1) was small, even compared to other inshore bottlenose dolphin populations. Together, these group effects may indicate that prey characteristics vary diurnally as well as annually but are relatively uniform within years. Further, the size of groups containing at least one calf (mean = 5.5) was larger than groups without calves (mean = 2.4) suggesting that mothers often rely on the assistance of conspecifics in calf-rearing. Dolphins with nine or more sightings were examined for evidence of site fidelity. Although the majority of these dolphins exhibited a high degree of site fidelity to the IRL system as a whole, dolphins with high numbers of sightings who were encountered in the Mosquito Lagoon were rarely encountered elsewhere. Overall, half-weight Coefficients of Association (COA) indices for these frequently sighted dolphins ranged from 0.09 to 0.83 (maximum values), showed that they preferentially associate and that associations were higher within sex class than between sex class. The larger range in maximum COA values that males (0.10 – 0.83) exhibit over females (0.00 – 0.33) may indicate that some males, as they do in other areas, form longer-lasting bonds while females have several casual associates. Mothers and identified calves exhibited the strongest associations, with some associations lasting over three years. In general, these results reflect similar patterns of social structure to those observed in other residential inshore populations, including individuals who show a frequent change in group membership but maintain some long-term stable companionships.