Effects of environmental factors on vocalization pattern of dugongs revealed by generalized linear model

To quantitatively examine the complex environmental factors affecting acoustical behavior on marine mammals like dugongs, a combination of passive acoustic monitoring (PAM) and statistical modeling should be effective approach. We deployed automatic underwater sound monitoring systems (AUSOMS) on the sea floor of different two sites and recorded the dugongs' vocalization. The effects of the time period and tidal change to the vocalization rate, which is dugong calls per hour, were estimated with using generalized linear model assuming negative binomial distribution. Time periods is similarly affected to dugongs vocalization in both site; however, tidal effect different between the sites. We could provide the first insight for the effectiveness of the combination of PAM and statistical modeling to dugongs' acoustical behavior itself.