An Ethogram of dolphin behavior with emphasis on underwater activity This behavioral catalog was originally designed for use during a study of the communication of the Atlantic Spotted Dolphin (Stenella frontalis). Unpublished ethograms and valuable discussions were provided by J. Jahoda, J. Östman, D. Reiss, B. McCowan, C. Gubbins, K. Waples, and Annie Weaver in the development of this particular ethogram.
In the strictest sense, an ethogram is a catalog of the behavioral events (i.e., specific action patterns) that an animal or group of animals may use during different behavioral states or contexts. While a few researchers (e.g., Defran and Pryor 1980; Reiss 1989; Frohoff 1993) have developed behavioral catalogs for captive dolphins, this ethogram of underwater behavioral events represents one of the first such catalogs for free-ranging dolphins. Defran and Pryor (1980) present a list of behavioral units for five given contexts (e.g., social/contact behaviors, aggression) in Table 7.3. However, the behavioral events (e.g., threat posturing, leaping in unison) that Defran and Pryor (1980) describe appear to be mutually exclusive between contexts. This is not always the case for free-ranging animals (e.g., Packard and Ribic 1982; Weaver 1987); many behavioral units can be observed within more than one behavioral state. In behavioral research, a common problem is the choice of behavioral units and states to answer specific questions (Altmann 1974). The units should be discrete to increase observer reliability, as well as allow for rigorous statistical analysis. Most behavioral sequences are continuous but video recordings of behavior permit researchers to form discrete data sets of behavioral data. Video is invaluable for behavioral studies - many observers can independently view the same sequence, an archive of behaviors can be established, and confirmation of real-time observations can be facilitated. Most previous work on cetacean behavior (e.g., Tyack 1981; Chabot 1986; Silber 1986) are mainly descriptive discussions of general surface activities of the species in question. The behavioral units used in activity classification studies have usually been activity states such as feeding, resting, socializing, traveling (Würsig and Würsig 1980; Sjare and Smith 1986; Weilgard and Whitehead 1990).
The Behavioral Catalog
The catalog has been separated into sections based on individual dolphin actions, interactions between two or more dolphins, and vocalizations. The first section provides a list of definitions for anatomical terms commonly used in the study of delphinids.4 Dolphins The behavioral catalog is included on this site. Also included are a few photographic examples of some of the specific action patterns and appropriate descriptive definitions. The ethogram uses a three letter code, a name, and description of each definition, position and behavior. Each description has a unique three letter code. These codes are used for ease of data entry and analysis, and are alphabetized within each sub-section of the catalog. Definition codes are presented first to orient the observer with terminology used for anatomical classification. Swimming speeds and age & spot classes for spotted dolphins are included within this section, as they applied to my research in the Bahamas. "Group" is included in the definition section. The position codes are subdivided into individual posturing and the positioning of more than one dolphin. The infant codes are used for detailed analysis of adult/infant interactions, as well as the interactions between infants and more than one other adult. It is not always completely accurate to assign a class 1 or 2 spotter to a particular female as its mother: the young unmarked dolphins can not be reliably re-identified between (and at times within) years, and the young associate with individuals other than their mothers. The behavior codes have been divided into actions of individuals and interactions between two or more dolphins. This will help observers in separating context differences between single dolphins and the more detailed and involved situations with more than one individual. Sexual codes are presented separately because much of the posturing related to the genital area may be related to socio-sexual behavior. We do not mean to imply that these behaviors are present during successful intromission or that every context labeled as sexual implies a sexual function for the observed behaviors. This section is used as a way to classify the behaviors that may be related to sexual contexts but does not imply their function. Vocalization codes may be used concurrently with the other behavioral codes but are separated into a distinct category for ease of use from the catalog. A discussion of spectrograms, as well as sample sounds, are available elsewhere.
Behavioral states are analyzed to indicate the context for given behavioral events (such as pectoral fin rubbing). Contexts include aggressive, resting, foraging, play with object, and socio-sexual. The latter context is subdivided into two categories: large-group social activity and small-group social activity. One would expect less ambiguous behavioral events with more animals in an encounter, whereas more subtle (at least to the human observer) events might characterize the smaller, with possibly more closely related individuals, groups. We use matrices to decipher the data from videos and establish the occurrence of group size for each group type (e.g., young dolphins, male-female groups, female-calf groups) as our independent variable. Our dependent variable will be the action patterns and vocalizations for each of the group types.
The Behavioral Catalog can be downloaded in PDF format by clicking here. (32.5K)