Monday, September 30, 2013

METHODS: Mechanistic Approaches To The Study Of Animal Social Behavior & Social Organization...I, II

For some time now, and for several reasons, i have engaged my own thinking and the thinking of others about how the study of "real-time" social [here meaning inter-individual interactions among conspecifics] behavior, social interactions, and group "behavior" might be facilitated by using remote sensing. Figure and Table below summarize my information and thinking up to February 2014. 

I am interested in advances that are not summarized in Table and that may have been made since Feb. '14 in remote sensing or other mechanistic approaches that would facilitate "mapping" of individual & interindividual responses, within & between groups, onto critical resources. Ideally, identification of individuals & groups could be "mapped" onto resources &, beyond ideal!, remote sensing could tell us something about the dispersion [distribution & abundance], quality, and type of resources, particularly, plants. Other questions related to, say, access to mates, might also be envisioned. I understand that such remote sensing as i describe is a future project; however, i'm simply wondering whether advances have been made in that direction [beyond the approaches described in Table below]?

I acknowledge the input and expertise of Dr. Daniel Mennill [Univ of Windsor], who informed me of Encounternet, has answered numerous questions from me, and whose work, combined with that of others, has stimulated much of my thinking on these topics.


 
I. MAP: A map, including this “nested-vision” 3D schema, constitutes a systematic effort and tool to model (conceptualize, understand) an actual or potentially real problem, event, condition, situation, response, or, other, phenomenon, in the physical or perceived universe.  The depicted model represents a generic “nested-vision” 3D map, amenable to rotation and a variety of other alterations.  This structure might, for example, characterize a group (black square) from which individuals leave (lines) to forage in two habitats or patches (grey squares) of a home range (A, B), subsequently, returning to their group (lines).  Circles in each “patch” might represent different diameters at breast height (DBH) of trees, colored differentially for species recognition.  The grey squares might, alternatively, represent 2 sub-groups of a group (black square), with circles identifying individuals by age or dominance rank, or other features (e.g., degree of relatedness to a matriline or, in a polygynous group, a resident male).  Or, the black square might represent a source of water in an arid zone, with lines representing proportional (circle size) frequency of movement of different bands of different classes of taxa (within- or between-taxa), A and B.  Continuing to visualize the black box as a source of water, lines might represent individuals of different groups, A and B, with circles representing, for instance, age-size-class membership, frequency of transit, sub-group membership, or related variables.

Following Ware et al. (1997; Parker et al. 1998, Ware and Mitchell 2008), 3D graphs and maps make three assumptions: (1) that 3D is preferable to 2D visualization for large information structures; (2) that “nested” graphs are required to represent “complex” databases; and, (3) that maximum utility of these approaches includes “manual and automatic layout of structures”.  Other types of visualization utilities can be accessed at Colin Ware’s (University of New Hampshire) website: http://ccom.unh.edu/vislab/projects/networks, including, “node-link” diagrams with capacities for thousands of nodes and links, as well as, “interactive motion” structures whereby relevant information can be highlighted with a cursor.  In addition, conventional methods of graphing or mapping information can be expanded, such as the modified physical map presented by Jones (1995, Fig. 1, p 4).  Visualized applications are models constructed systematically to convert a researcher’s conceptualizations of hypothesized and “real-world” systems into graphic and mapped displays of imagined or actual information.  However, graphs, maps, and, related, utilities, do not substitute for mathematical modeling. ©Clara B. Jones
II. TABLE: This table identifies “mechanistic approaches” that may be employed to study groups of social mammals based on a review of the literature, and communication with researchers.  These technologies capture events of a species at one or more scale of analysis, from individuals to groups, to populations as well as abiotic (e.g., soil gradients) and other biotic features (plants, conspecific groups, animal species composition).  Appariti are systematically employed to convert a partial or complete array of real-world problems (migration, group foraging, contest competition, mate choice, cooperation, and the like) into analyzable data.  Challenges are encountered since, in social groups, one animal’s behaviors are a function of interactions with conspecifics, usually, other group members.  In most cases, these apparati will be used in association with traditional data-collection techniques (“focal” observations of animals, hand-held instrumentation, fruit-fall traps, DBH measurements: see, for example, Reich et al. 2004).  As Moorcroft (2012) pointed out, in addition to post-study (and real-time) advances in analyses, including model-fitting, the major contributions of newer technologies at present are increasing capacities to capture concurrent, fine-scale data within- and between-populations.  These utilities, also, permit assessments of environmental “grains” at different levels of analysis (Moorcroft 2012), an important capability for social biologists because events at one scale generally cannot be employed to predict events at lower or higher scales, and, because current technologies and near-generation mechanistic approaches permit a researcher to estimate static and dynamic population parameters (e.g., generation time, population growth).  I thank D.J. Mennill, W.J. Foley, S. Kawano, B. Nicolai, and N. Pettorelli for providing information via "personal communication" and remain grateful to Ted Fleming for assistance with the literature search. ©Clara B. Jones

TECHNOLOGY
CURRENT UTILITIES
REFERENCES
FUTURE UTILITIES/NOTES
Animal Patterns
Radio-tracking
Land-based telemetry system for tracking spatial ecology of individuals and groups
bats: Almenar et al. (2013); primates: Joly & Zimmerman (2011); birds and bats concurrently: Taylor et al. (2011); ungulates: Mueller et al. (2011)
2 or more animals can be studied concurrently (e.g., members of sub-groups); operates on relatively small spatial scales; short-term temporal data; difficulties associated with tracking in closed forest habitats when tracking on foot; recent advances enhance power and applications (Moorcroft 2012); relatively inexpensive compared with other tracking methods
Radio-tracking via airplane
Allows descriptions of landscapes relative to animal use when
Bats: Eby (1991)
Problems associated with length of battery life
Resource-selection analysis (RSA)
Used in combination with radio-telemetry, allowing descriptions of landscapes
Moorcroft (2012)
Capable of identifying spatial scales permitting “multi-layer” analyses; applicable to tests of socio-ecological hypotheses recording within-population dispersion of individuals and groups relative to resource dispersion; adaptable to studies of leadership and rank relations via differential use of space; “mechanistic home-range analysis” (Moorcroft 2012) applicable to social biology
Global-positioning system (GPS)
Satellite-based tracking system using solar- or battery-powered transmitters
Moorcroft (2012);  Holland and Wikelski (2009), Richter and Cumming (2008), Epstein et al. (2009); Tsoar et al. (2011), Markham and Altmann 2008; Tomkiewicz et al. (2010), Cagnacci et al. (2010)
Widespread scientific use relatively recent; permits deployment on animals smaller than large terrestrial and marine mammals; can be used to monitor physiological states; can track animal movements and use of space over large geographical ranges; databases can be created and managed for behavioral, ecological, and comparative studies (Tomkiewicz et al. 2010)
©Encounternet
Light-weight tags “enable automated mapping of social networks” (including, position and duration of interactions and signals
Rutz et al. (2012), Mennill et al. (2012a, b), Taylor et al. (2011)
Data received by a “grid of fixed receiver stations” yielding large, high-quality, high-resolution datasets; so far tested using birds; can be used for terrestrial and arboreal taxa (D.J. Mennill, personal communication)
“Proximity data-loggers”
Similar to and may be used in association with ©Encounternet technology for studying interactions of group-living animals
Ryder et al. (2012), Mennill et al. (2012a, b), Maynard et al. (2012)
Data transferred to receiver “grids” capturing frequency of contacts permitting construction of “weighted networks” characterizing “complex social dynamics and calculation of statistics; captures changes in individual and inter-individual responses, group structure, population processes, and resource dispersion, including, phonologies; useful for tests of sociobiological hypotheses (e.g., cooperation, sexual selection: Mennill et al. 2012)
Camera traps
Remote instruments that take photos or video when a sensor is triggered (mongabay.com)
Diaz et al. 2005, Harmsen et al. 2009, Norris et al. 2020
Use of robo-mammals lags behind studies of robo-mollusks or robo-amphibians; can adapt technology for estimates of animal interactions, such as, local predator-prey abundance and temporal distributions; can utilize for preliminary estimates of species distributions, including, relative occurrences of social and non-social taxa
Robotics
“Automated machines” capable of simulating biological events
fish: Ioannou et al. (2012: coordinated group movement); Handegard et al.(2012: group hunting and schooling prey); Kopman et al. (2013)
Use of robo-mammals lags behind studies of robo-mollusks, robo-amphibians, or robo-fish (“etho-robotics”); however, a  wide range of sociobiological questions is amenable to tests with robotic techniques, including, patterns of group dispersion relative to robots manipulated in various positions or configurations or simulated predators or prey (see references for fish) or manipulations of pelage or skin color and pattern relative to, for example, reproductive condition; in certain ways, these techniques can be employed in association with quantitative modeling (e.g., “agent-based” models)
Resource Patterns
Normalized Difference Vegetation Index (NDVI); Enhanced Vegetation Index (EVI); Moderate Resolution Imaging Spectroradiometer (MODIS); Light Detection and Ranging (LiDAR); Satellite-based remote sensing; Near Infra-red Spectroscopy (NIS); Imaging spectroscopy
Methods employed to assess plant food dispersion, type, and quality
Asner and Levick (2012), Bradbury et al. (2005), Youngentob et al. (2011)////, Saranwong et al. (2004), Saranwong et al. (2003), Nicolai et al. (2007), Pettorelli et al. (2005), Duffy and Pettorelli (2012)////, Xiao et al. (2006)
Research and development needed to for applications to covariation of events between plants (e.g., phenology, fruit type and ripeness) and mammal groups
Molecular genetics
Collection of tissue samples from animals at different locations, using data from mitochondrial and/or nuclear genes (e.g., microsatellites) to determine degree of genetic similarity between populations
Fleming (2010)
Measures of genetic similarities between seasonally-occupied habitats indicates connectivity of migratory movements; utility for studying migrants relative to particular resources in early stages of development; social biologists can use these techniques alone or in combination with other mechanistic approaches to assess genetic patterns within and between sub-groups (e.g., “fission-fusion” units) of the same species
Visualization
See  Map, above
Various approaches employed to visualize data/information, including, software structures; these utilities may incorporate motion, 3D, “fish-bowl”, “node-link”, and other information architectures
See Map, above
These mechanistic approaches require research and development for specific applications to questions, models, results, configurations (e.g., networks), and conceptualizations pertinent to Social Biology