TY - JOUR
T1 - A Dynamic Individual-Based Model for High-Resolution Ant Interactions
AU - Wikle, Nathan B.
AU - Hanks, Ephraim M.
AU - Hughes, David P.
N1 - Funding Information:
Funding was provided by NSF EEID 1414296 and NIH GM 116927-01. We are grateful to Andreas Modlmeier and the many undergraduates in the Hughes Lab who tracked the ants. We thank Roland Langrock, Christen H. Fleming, and one anonymous reviewer for their helpful suggestions.
Publisher Copyright:
© 2019, International Biometric Society.
PY - 2019/12/1
Y1 - 2019/12/1
N2 - Ant feeding interactions (i.e., trophallaxis events) are thought to regulate the flow of nutrients and disease within a colony. Consequently, there is great interest in learning which environmental and behavioral factors drive ant trophallaxis. In this paper, we analyze ant trophallaxis behavior in a colony of 73 carpenter ants, observed at 1-s intervals over a period of 4 h. The data represent repeated observations from a dynamic contact network; however, traditional statistical analyses of network models are ill-suited for data observed at such high temporal resolution. We present a model for high-resolution longitudinal network data, where the network is assumed to be a time inhomogeneous, continuous-time Markov chain, with transition rates modeled as a function of time-varying individual and pairwise biological covariates. In particular, the high temporal resolution of the data leads to a tractable likelihood function, and likelihood-based inference procedures are utilized to explain which biological factors drive contact. Our results reveal how differences in ant social castes and individual behaviors, such as ant speed and activity levels, influence patterns of ant trophallaxis in the colony. Supplementary materials accompanying this paper appear online.
AB - Ant feeding interactions (i.e., trophallaxis events) are thought to regulate the flow of nutrients and disease within a colony. Consequently, there is great interest in learning which environmental and behavioral factors drive ant trophallaxis. In this paper, we analyze ant trophallaxis behavior in a colony of 73 carpenter ants, observed at 1-s intervals over a period of 4 h. The data represent repeated observations from a dynamic contact network; however, traditional statistical analyses of network models are ill-suited for data observed at such high temporal resolution. We present a model for high-resolution longitudinal network data, where the network is assumed to be a time inhomogeneous, continuous-time Markov chain, with transition rates modeled as a function of time-varying individual and pairwise biological covariates. In particular, the high temporal resolution of the data leads to a tractable likelihood function, and likelihood-based inference procedures are utilized to explain which biological factors drive contact. Our results reveal how differences in ant social castes and individual behaviors, such as ant speed and activity levels, influence patterns of ant trophallaxis in the colony. Supplementary materials accompanying this paper appear online.
UR - http://www.scopus.com/inward/record.url?scp=85073681759&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85073681759&partnerID=8YFLogxK
U2 - 10.1007/s13253-019-00363-5
DO - 10.1007/s13253-019-00363-5
M3 - Article
AN - SCOPUS:85073681759
SN - 1085-7117
VL - 24
SP - 589
EP - 609
JO - Journal of Agricultural, Biological, and Environmental Statistics
JF - Journal of Agricultural, Biological, and Environmental Statistics
IS - 4
ER -