Introduction:
Passive acoustic monitoring (PAM) is a well-established method used to
monitor acoustically active species and the habitat they reside in
(Merchant et al. , 2015). Over the years PAM technology has
greatly advanced the field of cetacean ecology, allowing a
cost-effective alternative to extensive visual surveys that is not
reliant on daylight hours, favourable weather conditions and
availability of observers. Passive acoustic monitoring of cetaceans has
resulted in high-resolution data providing us with insights of
population size and abundance (Marques et al. , 2013; Amundinet al. , 2022), habitat use (Fleming et al. , 2018; Palmeret al. , 2019), and behaviour (Pirotta et al. , 2014;
Malinka et al. , 2021; Todd et al., 2022) for many species. Such
technology is also fundamental for long-term monitoring, particularly
with the increase in coastal developments and potential disturbance from
construction, marine renewable devices, shipping, and fisheries (e.g.,
Todd et al., 2020, 2022; Omeyer et al. , 2020; Ramesh et
al. , 2021; Fernandez-Betelu, Graham and Thompson, 2022).
While there are many useful applications of PAM, fixed autonomous
acoustic recording devices can increase deployment times and sampling
frequencies (Sousa-Lima et al. , 2013). Data loggers or
echolocation click detectors, such as the C-POD (Cetacean POrpoise
Detector) (Chelonia Ltd., 2022) are a user-friendly, relatively
inexpensive device which can be deployed for continuous monitoring
periods of 3-6 months. C-PODs detect individual echolocation clicks
between 20-160 kHz and have been a popular tool used to study odontocete
ecology and behaviour worldwide (e.g. Carstensen, Henriksen and
Teilmann, 2006; Simon et al., 2010; Nykänen, 2016; Jaramillo-Legorreta
et al., 2017; Garagouni, 2019). Although no waveform data are stored by
the devices, summary data on each click are preserved allowing
post-deployment classification of detected sounds into sequences called
click trains. Further data analysis is then performed where click trains
are assigned to dolphin or porpoise origins based on frequency and
bandwidth. While it is often not possible to differentiate between
dolphin species (Robbins et al., 2015), based on in-field testing,
Roberts and Read (2015) reported that C-PODs perform well with a
relatively high accuracy in detecting cetacean echolocation. C-PODs have
been used for over a decade and now form the basis of valuable long-term
monitoring datasets. The F-PODs (Full waveform capture Pods) are the
successor of the C-PODs, and the manufacturer is recommending a
transition from C-PODs to F-PODs as availability and support for C-PODs
will be limited in the coming years. This may have important
implications for long-term monitoring programmes (and the associated
archival data from such) as C-PODs are replaced due to equipment loss
(often as a result of storms, theft etc.) or reach the end of their
operational lifetimes. The F-PODs have been designed to improve and
upgrade the data associated with C-PODs by recording more details of
selected clicks including position of loudest cycle, frequency range and
capture of full waveform (Chelonia Ltd., 2022). These new features
enhance train detection, providing increased sensitivity with lower
false positive rates compared to C-PODs (Chelonia Ltd., 2022).
One of the main advantages of PAM is its potential to be implemented in
long-term monitoring programmes to study the change in species
occurrence and behaviour over longer temporal scales. Many studies
currently using C-PODs need to ensure the longevity of their data for
monitoring purposes, particularly in the light of climate change and
habitat alterations through coastal developments. However, to date there
have been no studies reporting how the C-POD and its successor the F-POD
compare in detection capacity and ability to identify trends in
spatio-temporal drivers of detected cetaceans. In this study, we used
data from a co-deployed C-POD and F-POD to compare the performance of
the PODs in detecting habour porpoise (Phocoena phocoena) across
various commonly used detection metrics. Additionally, Generalised
additive models (GAMs) were used to explore how both PODs identifed
spatiotemporal variation in harbour porpoise occurrence and foraging
activity in relation to environmental variables.