Animal-borne acoustic biologgers to broaden the applicability of passive acoustic monitoring to terrestrial mammals: a feasibility study for deer

Passive acoustic monitoring (PAM) has become a reliable option for assessing population metrics of terrestrial vocal mammals. Although the detectability of mammals using PAM can be affected by vocal activity rate within the effective detection space of monitoring devices, limited studies have identified situation-dependent factors that may determine such activity rates. To address this knowledge gap, we designed animal-borne acoustic biologgers. This study targeted sika deer (Cervus nippon) as a feasibility test and intended to identify which external factors determine the rutting call rate of bucks. We collected acoustic data from seven bucks with biologgers. Our Bayesian models for evaluating the diel rhythm of the two rutting call types (howls and moans) demonstrated that both call rates fluctuated little with time of day. We then constructed Bayesian generalised linear mixed models to identify the external factors that explain howling or moaning rates. Both models indicated that bucks rarely emitted rutting calls under noisy conditions caused by rainfall and wind. In addition, the current models demonstrated that, although situation-dependent factors in howling rate were detected, these factors could be handled by adjusting the PAM study design, making howl detection a useful tool for population-level estimates. In contrast, the moaning rate did not serve as a metric of population abundance but could function as an indicator of preferred breeding sites. Thus, our device serves as a key driver for expanding the applicability of PAM to various terrestrial mammals by bridging the knowledge gap between individual-level behaviour and population-level acoustic observations via PAM.