Using acoustic recorders and accounting for imperfect detection to understand spatial and temporal breeding patterns of a cryptic burrowing amphibian

doi:10.21203/rs.3.rs-5315869/v1

Download PDF

Research Article

Using acoustic recorders and accounting for imperfect detection to understand spatial and temporal breeding patterns of a cryptic burrowing amphibian

https://doi.org/10.21203/rs.3.rs-5315869/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Understanding the determinants of detection and occupancy at breeding habitats is crucial in the management of threatened amphibian species. This study investigates the spatial and temporal breeding patterns of the giant burrowing frog, Heleioporus australiacus australiacus, a cryptic and threatened amphibian species, through acoustic monitoring. We aimed to quantify the detectability of this species and identify environmental factors that influences its breeding behaviour. Passive acoustic recorders were deployed between May 2018 and November 2021 across eastern NSW. From 5,030 survey nights, the giant burrowing frog was only detected calling on 44 occasions. The estimated probability of detecting a giant burrowing frog during a single survey was p̂ = 0.01, confirming the cryptic nature of its breeding behaviour. Our analysis revealed that calling behaviour in the giant burrowing frog is correlated with climatic variables, with increased detection probabilities associated to higher nightly temperatures and impending rainfall. Conversely, higher daily rainfall and solar radiation levels were associated with a decrease in detection probability. Temperature, precipitation, seasonality and habitat aspect were identified as key factors influencing breeding site occupancy.

Overall, our study highlights the importance of accounting for imperfect detection in monitoring programs and provides valuable insight into the breeding behaviour of a difficult to detect species. These findings have implications for the management and conservation of giant burrowing frog populations, particularly in the face of changing environmental conditions. Continued monitoring efforts, focusing on tadpole surveys and passive acoustic monitoring, are recommended to improve our understanding of their ecology.

breeding patterns

frog

Heleioporus

occupancy modelling

passive acoustic monitoring

threatened species

Ecological monitoring provides critical information regarding species, community, and ecosystem trends, including those associated with human disturbance (Haughland, et al., 2010). However, false conclusions can be drawn from such monitoring if sampling is not carefully designed and conducted. This can have implications if results do not represent the real-world situation and may lead to inadequate and inappropriate conservation actions (Kéry, 2002). This is further complicated for species that are infrequently detected (e.g., cryptic < 30% detection; Spetch et al., 2017) or are rare in nature (< 30% occupancy probability; Spetch et al., 2017) since detection rates are typically low over large spatial scales and therefore characterized by a high frequency of zero observations during monitoring despite the species being truly present (reported as false negatives) (MacKenzie et al, 2006). Ecological monitoring should ideally report detection probabilities and confidence levels. Only by doing this will it be possible to quantitatively understand imperfect detection for some rare or cryptic species. By inclusion of these metrics, improvements in sampling design should be considered to ensure monitoring efforts maximize the potential for positive conservation outcomes.

Imperfect detection occurs when individuals, populations or species are not found during surveys despite their presence (Martin et al., 2005). This may lead to estimates of abundance, distribution and species richness being negatively biased (Kéry & Schmidt, 2008), and spatial or temporal patterns in species abundance, distribution and biodiversity being obscured (Link & Sauer, 1998; Pollock, et al., 2002). It is imperative that monitoring programs quantify imperfect detection by considering detection probabilities, thus allowing the removal of bias from estimates (Tanadini & Schmidt, 2011). Occupancy modelling is one such tool that can estimate occurrence while accounting for imperfect detection. An advantage of estimating detection probability for ecological monitoring programs is that conservation practitioners can optimize survey timing to ensure high detection rates (McGrath et al., 2015).

Improvements in technology may allow for more efficient and effective monitoring strategies to increase detectability, particularly for declining and/or cryptic species (Taylor et al., 2014; Diggins et al., 2020). Passive acoustic recorders (PARs) are one such technology that enable capture of animal vocalisations (Ross et al., 2023). By programming PARs to record their acoustic environment remotely for extended periods and at any time of the day, rich datasets can be gathered that are otherwise not feasible with in-field survey techniques (Celis-murillo, Deppe & Ward 2012). This can be coupled with in-field data loggers that measure abiotic variables to gather additional temporal and spatial data to identify covariates that might influence activity (Willacy et al. 2015; Sugai, et al., 2018). Additionally, presence/absence data obtained from PARs can be paired with traditional count surveys to improve estimates of abundance (Doser et al. 2021). These tools can also enable the construction of phenological models, enhancing ecological monitoring (Hoffmann & Mitchell, 2022).

Some amphibians are cryptic and rare, and their populations can have considerable natural fluctuations (Green, 2003; Beranek et al., 2022), which makes long-term monitoring and the detection of threat-related declines difficult (Barata, Griffiths & Ridout, 2017). Accurate monitoring of target species can be impacted by a range of ecological, environmental, economic, and logistical challenges (Bakker et al., 2010), including imperfect detection. These constraints have led to amphibians being under-represented in distribution and abundance studies, resulting in scarce historic data for many species (Recher & Lim 1990).

Effective conservation of threatened species relies on adequate information regarding their life history traits, such as those associated with reproduction, as it is the interaction between these traits that influence fitness (Roznik et al., 2015). For example, identifying breeding strategies offers an insight into why populations might vary in size and structure (Petranka & Thomas, 1995; Gould et al., 2022a), how frequent and for how long breeding events occur (Díaz-Paniagua, 1992; Paracuellos et al., 2022), as well as the timing of these events (Dixon & Heyer, 1968). The Wells (1977) explosive versus prolonged breeding paradigm presents a continuum of breeding strategies that varies among frogs and populations. Determining breeding strategies selected by species and populations can provide critical information for determining distribution and abundance and thus conservation assessment. While our understanding of the life history strategies of numerous anuran amphibians remains limited, deploying PARs at breeding habitats is useful because most anuran amphibians engage in vocalisation during their reproductive processes (Gerhardt and Huber, 2002; Köhler et al., 2017).

We combined acoustic monitoring with occupancy modelling to increase our understanding of spatial and temporal breeding patterns of a threatened and cryptic burrowing species, the giant burrowing frog, Heleioporus australiacus australiacus (HAA). We present this as a case study to demonstrate a process where rich data generated through acoustic surveys performed by PARs can be used to estimate detection probability with high precision to allow sufficient statistical power to model the probability of occupancy of cryptic species that vocalise.

We made several predictions based on the current understanding of spatial and temporal breeding patterns for this species (see Appendix Section A1 for a summary of HAA). We predicted that this species will have a low probability of detection (< 30%; Spetch et al., 2017) and a low probability of occupancy (< 30%; Spetch et al., 2017) driven by environmental and habitat specialisation for breeding (Stauber, 2006; Penman et al., 2005a; Penman, Lemckert, & Mahony, 2006a). Additionally, we hypothesized that breeding detectability would be positively correlated with temperature and rainfall as HAA is considered a late summer breeding species and has been shown to move to breeding habitats following rainfall (Penman et al., 2005b; Penman, Lemckert & Mahony, 2006a; Penman, Lemckert & Mahony, 2008a). Lastly, we predicted that breeding site occupancy would occur more often in ephemeral water bodies compared to permanent sites (Stauber, 2006) and positively correlated with the percentage of sandy soil profiles within the immediate landscape of the breeding site, as previous surveys indicate that HAA breeding habitats occur on sandy soils and, except for one population in Ulladulla, are not found on clay-based soils (Daly, 2019).

Study area and site selection

Known HAA breeding sites (n = 22) were selected based on tadpole and/or calling male occurrences uploaded to the Atlas of Living Australia (ALA) and FrogID (Rowley et al., 2019). Additional sites in the Greater Blue Mountains Area were included after a citizen science campaign provided local community knowledge on undocumented HAA breeding habitats (n = 5). Sites where HAA had not yet been observed but were environmentally and climactically similar to known breeding sites were also included (n = 15, Fig. 1).

All sites were located within the Sydney Basin bioregion (NSW NPWS, 2003) which has a temperate climate and is characterised by warm summers with no dry season. Rainfall can occur throughout the year but varies across altitudes and proximity to the coast. The Sydney Basin bioregion consists of a geological basin composed predominantly of sandstone and shale. The considerable variety of rock types, topography and climate in the region has resulted in a diverse array of soil types and plant communities (NSW NPWS, 2003). A breakdown of the subregions in this study can be found in Appendix: Table A1.

Acoustic surveys

We collected breeding site occupancy data by deploying a single PAR at each survey site (Song Meter Mini, Wildlife Acoustics, and AudioMoth, Open Acoustic Devices). At sites that had historic or contemporary HAA records, we positioned the PAR as close as possible to these records. For all sites we positioned the PAR with microphones directly facing pooled and/or slow-moving water, as these areas were generally less noisy and are where HAA usually call from. All PARs were installed approximately one meter above the maximum water line and no more than two metres from the bank. Acoustic devices were set to sample at a rate of 48,000 Hz and recorded for five minutes every hour starting at 17:00 with the last survey beginning at 02:00. Every three months, deployed PARs had either their 32 GB micro-SD memory card replaced or were entirely replaced if they required a battery change. There were 5,030 survey nights over 42 sites conducted between May 2018 and November 2021. Total survey length per site ranged between 10 days and 559 days (mean = 119.76, SD = 94.33), however consecutive surveys ranged between 10 days and 480 days.

Sound files were manually analysed for the presence of HAA calls by an expert ecologist. This was performed as HAA calls are not loud compared to that of sympatric frogs, which is likely to make automated recognizers difficult to create for this species. Due to the large quantity of data collected (50,300 surveys total: 10 surveys per night for 5,030 survey nights) only acoustic recordings from 22:00 hours were included in the analysis, however when calling was detected the expert listened for periods of time prior to first detection and around the date of detection to confirm the bout of reproductive activity (Daly 1996; Lemckert & Mahony 2008). Additionally, we defined their total calling period by counting the number of consecutive nights of calling at a site, followed by the number of consecutive nights they didn’t call, repeating this process until their last recorded call at a site.

Variable of detection

We collated numerous environmental and climatic variables from various data sources to inform the analysis of detection probability. Temperature (˚C; temp) at the hour of each survey (22:00) and total daily rainfall (mm; rain1) were collected using the R package ChillR (v0.72.7, Luedeling & Fernandez, 2022). Julian date (Jdate) was associated with each acoustic recording, where the first day of March = 1 and the last day of February = 365, except in year 2020 when the last day of February = 366. Moon phase per night (moon) was sourced via the Astronomical Appplications Department (U.S. Naval Observatory, 2022). We extracted data on barometric pressure (hPA; baro), vapour pressure (hPa; vape), evaporation (mm; evap), solar radiation (MJ/m²; solar), minimum and maximum temperature (˚C; MinT and MaxT), and relative humidity at the min. and max. temperatures (%; RHmin and RHmax) on the day of each survey. These covariates were extracted through the SILO portal (Jeffrey et al., 2001). Additional calculations were made to determine the cumulative rainfall total for the previous five days before a survey night (mm; rain5), the number of days until daily rainfall total of 10 mm or greater (from here on referred to as a downpour; DTD); the temperature shift from the daily maximum to the time of survey (˚C; TShift), and the percentage of this temperature shift (%; PTShift). A summary of these variables is detailed in Table 1, with further information presented in Appendix: Table A2.

Table 1

Summary of breeding habitat covariates used in detection modelling for *Heleioporus australiacus australiacus* across southeastern NSW.
Covariate	Range	Mean (± SE)
temp (^°C)	1.71–30.43°C	14.54^oC (± 0.07°C)
rain1 (mm)	0–182.9 mm	3.92 mm (± 0.16 mm)
Jdate	1–366	158.81 (± 1.72)
baro (hPa)	994.5–1038.7 hPa	1017.89 hPa (± 0.10 hPa)
vape (hPa)	3.6–26.7 hPa	14.00 hPa (± 0.07 hPa)
evap (mm)	0–14.3 mm	3.54 mm (± 0.03 mm)
solar (MJ/m²)	2.5–32.4 MJ/m²	14.78 MJ/m² (± 0.09 MJ/m²)
MinT (^°C)	-4–25.1°C	10.97 ^oC (± 0.07°C)
MaxT (^°C)	4.2–42.7°C	21.53 ^oC (± 0.07°C)
RHmin (%)	12.8–100%	53.88% (± 0.22%)
RHmax (%)	21.5–100%	94.33% (± 0.15%)
rain5 (mm)	0–516.5 mm	19.79 mm (± 0.59 mm)
DTD	-108–0 days	-16.31 days (± 1.52 days)
TShift (^°C)	-21.27 – -1.15°C	-8.17 ^oC (± 0.04°C)
PTShift (%)	-81.25 – -6.01%	-36.45% (± 0.18%)

Variables of occupancy

Each breeding site was categorised as containing a permanent or ephemeral water body (hydro). The aspect relative to north (degrees; asp) of each breeding site was considered. Root-zone soil moisture (m³ m^− 3; RSM) and the normalized difference vegetation index (NDVI) were collated for each breeding site via the Central Resource of Sharing and Enabling Environmental Data in NSW (SEED) portal (NSW Government, 2023). Soil profile characteristics (%; clay; sand; silt) were extracted from Soil and Landscape Grid of Australia via the R package slga (v.1.2.0, O’Brien, 2021). Data was extracted at depth intervals of 0–5 cm, 5–10 cm, and 15–30 cm as HAA have been found buried at depths between 1–30 cm (Penman, Lemckert & Mahony, 2006b). As each depth profile varied in total depth measured, we took a weighted average of the extracted values to find the soil type percentages from 0–30 cm deep using the equation:

Equation (1)

$$\:\left(\left(Soilx\:\%\:of\:0-5cm\right)*0.1666667\right)+\left(\left(Soilx\:\%\:of\:5-15cm\right)*0.333333\right)+\left(\right(Soilx\:\%\:of\:15-30cm\left)\:*0.5\right)$$

Additional bioclimatic variables, including annual mean temperate (˚C; AMTemp), annual precipitation (mm; APrecip), temperature seasonality (TSeason), and the mean temperature during the warmest quarter of the year, from here on referred to as the warmest period (˚C; TWarm) were sourced from BIOCLIM via the R package dismo (v1.3-5, Hijmans, et al., 2021). Forest height (decimeters; f_height) was sourced via the global canopy height map (see, Potapov et al. 2021). A summary of these variables is detailed in Table 2 and further information and relevance can be found in Appendix: Table A2.

Table 2

Summary of breeding habitat covariates used in occupancy modelling for *Heleioporus australiacus australiacus* across southeastern NSW.
Covariate	Range	Mean (± SE)
asp (^°)	29.4–343.8 ^°	186.37 ^° (± 13.42 ^°)
RSM (m³/m³)	0.35–0.69 m³/m³	0.44 m³/m³ (± 0.01 m³/m³)
clay (%)	14.5–32.3%	19.24% (± 0.66%)
sand (%)	46.7–60.9%	53.70% (± 0.59%)
silt (%)	13.7–19.2%	16.14% (± 0.23%)
AMTemp (^°C)	11.2–17.3°C	14.59 ^oC (± 0.31°C)
APrecip (mm)	726–1791 mm	1149.62 mm (± 27.42 mm)
TSeason	1.22–1.64	0.14 (± 0.02)
TWarm (^°C)	17.38–22.87°C	1.52 ˚C (± 0.23°C)
f_height (dm)	5–32 dm	6.50 dm (± 1.00 dm)

Statistical Analysis

All analysis was conducted using R Statistical Software (v4.3.1; R Core Team, 2023). A Pearson’s correlation matrix was calculated between all variables and those with correlations <-0.5 or > 0.5 were identified as collinear and not paired in models. Colinear covariates are identified in the Appendix: Table A3. Prior to modelling, covariates were standardised to Z-scores to allow for the direct comparison of model coefficients.

We utilized a single-season occupancy-detection model (MacKenzie et al., 2006) to assess the detectability of HAA. This model assumed that the outcome of each survey was influenced by two concurrent binomial processes: 1) the probability of HAA being present at a site (ψ) over long time periods; and 2) the probability HAA were present within the site and observed in any given survey (ρ). The state process was considered a Bernoulli random variable, such that:

Equation (2)

$$\:{z}_{i}∼Bernoulli\left({\psi\:}_{i}\right)$$

Here, ψ represents the probability of a species occupancy at site I, and z_i denotes the true state of occurrence. We further posited that the observation of a species at site I results from another Bernoulli random variable, described as:

Equation (3)

$$\:{y}_{ij}|{z}_{i}∼Bernoulli({z}_{i}{p}_{ij})$$

In this context, y_ij represents the observed ‘presence-absence’ data, and ρ_ij is the probability of detection. We incorporate the impact of covariates by employing the logit link function, expressed as:

Equation (4)

$$\:logit\left({p}_{ij}\right)=\alpha\:+\beta\:\text{c}\text{o}\text{v}\text{a}\text{r}\text{i}\text{a}\text{t}\text{e}\:\text{A}$$

In this scenario, the likelihood of detecting HAA at a site during a PAR survey (ρ_ij) is determined by a function involving an overarching intercept α and coefficients that characterize the influence of covariates.

We used a two-stage model selection process to determine the top candidate models. First, we kept the occupancy parameter (ψ) constant and assessed all combinations of the linear and quadratic covariates on the detectability component of the model with a dredge. Once the top detection model was identified it was retained and tested along with the occupancy covariates in a second dredge. We computed occupancy-detectability models through the unmarked package in R (Fiske & Chandler, 2011) and utilized Akaike’s information criterion (AIC) to evaluate the fit of each model; AIC values assess model fit by comparing how closely fitted values are to real values, while simultaneously accounting for the number of model parameters included (Agresti 2002). If competing occupancy models were within 2 AIC values, model averaged estimates were used to present predictions. We also plotted any covariates that were hypothesized to have a significant impact on occupancy that fell outside of this 2 AIC threshold. Dredges were performed with the MuMin R package (Bartoń, 2012). Model fit was scrutinized using two different tests involving 10,000 bootstraps: sum of squared errors and Freeman-Tukey Chi-squared. An alpha value of < 0.05 determined the presence of lack of fit.

Detectability

Heleioporus a. australiacus were detected calling on 44 out of 5,030 survey nights (p̂_naïve = 0.01). These detections occurred at ten of the 42 sites (ψ_naive = 0.24), between October 2020 and April 2021, with calling events ranging from one to eleven nights (mean = 4.40, SD = 3.75). Predominantly, detections took place towards the end of summer in February 2021 (n = 7) and early autumn in March (n = 15) and April 2021 (n = 11). Calling periods, i.e. the number of consecutive nights with calling, ranged from one night to seven nights (mean = 1.38, SD = 1.10). The number of nights between calling periods varied from two nights up to 49 nights (mean = 11.46, SD = 12.86). The longest period between the first and last calls at a site was 67 days (11 calling nights; mean = 29.6, SD = 23.34). A breakdown of the calling periods can be found in Appendix: Table A4.

The null model estimated that the probability of HAA occupying a site was ψ = 0.31 (0.40 SE), and that the probability of detecting the species on a single night through acoustic monitoring was p̂ = 0.01 (0.35 SE). For a reduced data set of only historic and contemporary HAA breeding sites (27 sites, 3,144 repeated surveys) the naïve occupancy was ψ_naive = 0.37, and the probability of detection was p̂_naïve = 0.01. Model fit testing provided no evidence for lack of fit (see Appendix: Table A5).

The model that best supported detectability of HAA at breeding habitats incorporated the temperature at the time of the PAR survey, the number of days until a downpour, and the amount of daily rainfall and solar radiation. According to the model’s predictions, the probability of detecting HAA would increase (from ~ 5% to ~ 40%) as nightly temperatures exceeded 20°C and approached 30°C (Fig. 2a). Detectability was also predicted to increase as the survey approached a day with a downpour of 10 mm or more. Surveys conducted 30 days before a downpour had less than 0.5% probability of detecting HAA, increasing to over 1% 14 days prior and over 2.5% the day before a downpour (Fig. 2b). Negative relationships were predicted between detectability and the daily amount of rainfall and solar radiation at breeding habitats. The probability of detecting HAA calling at a breeding habitat decreased with increased daily rainfall and was estimated at ~ 0% when there is 50 + mm of rainfall in a single day (Fig. 2c). The probability of detecting HAA calling at a breeding habitat also decreased as the amount of daily solar radiation increases (Fig. 2d). These four covariates appear in all ten of the top models and a list of these top models can be found in Appendix: Table A6.

Occupancy

The average temperature during the warmest period emerged as the most strongly correlated covariate for HAA breeding occupancy, appearing in eight of the top ten models. Temperature seasonality appeared in seven of the top ten models, while annual precipitation appeared in three models. These were the most recurring covariates and the combination of these three covariates made up the top-ranking model in our study (Table 3). However, no single model was identified that provided substantial support (ΔAIC ≤ 2) to best explain the occupancy probability of HAA at breeding habitats. As such we conducted model averaging on the five models within the ≤ 2 AIC threshold.

Table 3

Occupancy model selection statistics for the top ten models fitted to the occupancy data for *Heleioporus australiacus australiacus*. All models include the top detection model (D1), and Akaike Information Criterion (AIC) values are provided for each model.
Model	nPar	AIC	ΔAIC	Cumulative Weight
ρ _(D1) ψ _{(APrecip + TSeason + TWarm)}	9	383.75	0.00	0.23
ρ _(D1) ψ _{(TSeason + TWarm)}	8	384.75	1.00	0.37
ρ _(D1) ψ _{(asp + f_height + TWarm)}	9	385.43	1.67	0.46
ρ _(D1) ψ _{(hydro + TSeason + TWarm)}	9	385.45	1.70	0.56
ρ _(D1) ψ _(TWarm)	7	385.70	1.94	0.65
ρ _(D1) ψ _{(APrecip + silt + TSeason)}	9	385.90	2.14	0.73
ρ _(D1) ψ _{(clay + TSeason + TWarm)}	9	386.04	2.29	0.80
ρ _(D1) ψ _{(APrecip + asp + TWarm)}	9	386.10	2.34	0.87
ρ _(D1) ψ _{(asp + TSeason + TWarm)}	9	386.21	2.45	0.94
ρ _(D1) ψ _{(hydro + silt + TSeason)}	9	386.29	2.53	1.00

Covariate TWarm appeared in all top five models and demonstrated a positive relationship with HAA breeding site occupancy. The model estimated occupancy to be below 0.25 when warm period mean temperatures were below 21°C and over 0.75 when warm period mean temperatures exceeded 22.5°C (Fig. 3a). Covariate TSeason was the next most frequent amongst the top five models and demonstrated a negative relationship with HAA breeding site occupancy. Occupancy was predicted to decrease from ~ 0.35 to < 0.10 as temperature seasonality increased (Fig. 3b). Greater APrecip at breeding sites was predicted to increase the probability of HAA occupancy from ~ 0.10 at 1050 mm to over 0.35 as precipitation totals approached 1800 mm (Fig. 3c). The estimated effects of these variables in the top model are provided in Table 4.

Table 4

Coefficient estimates from the model averaging of site occupancy for *Heleioporus australiacus australiacus.*
Variable	Covariate of	Mean	Standard error
Intercept	Detection	-4.61	± 0.10
DTD	Detection	1.09	± 0.13
temp	Detection	1.20	± 0.08
rain1	Detection	-1.14	± 0.19
solar	Detection	-0.51	± 0.06
Intercept	Occupancy	-2.67	± 1.03
APrecip	Occupancy	0.91	± 0.16
TSeason	Occupancy	-1.05	± 0.19
TWarm	Occupancy	3.47	± 1.90
asp	Occupancy	-13.12	± 2.29
f_height	Occupancy	-3.87	± 0.71
hydro	Occupancy	-1.09	± 0.28

Covariates asp and f_height were featured together in the 3rd ranked model. Breeding habitats situated between 29 degrees and 100 degrees due north were predicted to have occupancy greater than 0.25. Predicted occupancy declined beyond 100 degrees before plateauing at sites situated ~ 140 degrees for a predicted occupancy of 0.13 (Fig. 3d). There was a slight negative relationship in the predicted occupancy across different forest heights (Fig. 3e). Confidence intervals for these covariates are relatively large. The hydro period of breeding habitats in the 4th ranked model showed similar probabilities of occupancy at sites with ephemeral and permanent water availability (Fig. 3f).

Soil profile

Given the inclusion of soil covariates in our hypotheses, and the placement of the percentage of silt and clay in the 6th and 7th ranked models, respectively, we present these covariates as they appeared in their best performing model. The 6th ranked model incorporated the percentage of silt in breeding habitat soil, revealing that as silt percentage increased, the probability of occupancy decreased. Occupancy was estimated to be highest (ψ ≈ 0.5) with silt percentage below 14%, but steadily declined to below ψ = 0.1 as silt percentage exceeded 18% (Fig. 4a). The 7th ranked model included the percentage of clay in breeding habitat soil. Like the silt percentage, a negative relationship was predicted with the highest probability of occupancy (ψ ≈ 0.15) estimated when the clay percentage was below 15%, decreasing to ψ ≈ 0.06 when clay percentage exceeded 30% (Fig. 4b).

In this study we set out to increase our understanding of the spatial and temporal breeding patterns of a threatened and cryptic amphibian using acoustic monitoring. Our model predicted that the probability of detecting HAA during any given survey was p̂ = 0.01, therefore, as hypothesized, we confirm that this species breeding behaviour is cryptic (detected in < 0.3 of repeated surveys) as per Spetch et al. (2017). Breeding site occupancy of HAA across all our study sites was estimated at ψ = 0.31, which is slightly greater than the ψ < 0.3 level required to be designated as rare (Spetch et al., 2017). However, we can confirm that this species’ breeding site occupancy is not common (ψ > 0.3 but < 0.5) under the same designations.

Our case study demonstrates that even with low detection rates, it is possible to correlate covariate that influences detection probability using a rich acoustic survey dataset, providing sufficient statistical power required to model the probability of occupancy for cryptic species. The detectability of target species with a specified method is an important consideration for conservation management and monitoring decisions for threatened species (Chadès et al., 2008). While our evaluation of detectability for HAA through acoustic surveys indicates low detectability, other studies have demonstrated the importance of accounting for imperfect detection by similarly using a large quantity of data. For example, a study conducted on agamids produced detection estimates of less than 0.001 (detected 1 every 1,000 surveys) and still achieved precise estimates of occupancy (McGrath et al., 2015), demonstrating the utility of controlling for detection probability.

Influences on detection and calling behaviour

Our analysis indicates that HAA calling behaviour is correlated with climatic variables. In particular, the probability of detection increased when nightly temperatures exceeded 20°C and a downpour was imminent. However, the probability of detection was likely to decrease with greater values of both daily rainfall and solar radiation.

The model found that the probability of detecting HAA via calling was greater as nightly temperatures surpassed 20°C and that calling was completely limited when temperatures fell below 10°C. This reflects previous findings for the species that have shown calling to be more prevalent at air temperatures between 12–22°C (Littlejohn & Martin 1967; Daly 1996). This estimated minimum calling temperature differs from its Western Australian (WA) congeners who have been identified calling in temperatures below 5°C (Lee, 1967). However, Lee (1967) points out that at low ambient temperatures these frogs tended to call from deeper within their burrows. It may be then that HAA do call when nightly temperatures fall below 20°C but from deeper within their burrows where PARs may not pick up their vocalisations. Across HAA’s range, suitable nightly temperatures occur between September through to May, and it is likely that breeding behaviours decline between June and August. Comparatively, it has a longer potential calling period than its WA congeners whose peak calling period begins April and ends early June (Lee, 1967).

The model also indicated that the probability of detecting HAA via calling was greater the closer a survey was to an upcoming day that would have at least 10 mm of rainfall. This suggests that HAA may have physiological mechanisms to sense changes in weather signalling rainfall (Lowe, Castley, & Hero, 2016), like many other amphibian species that respond to weather cues to commence reproductive activity (Henzi et al., 1995; Steelman & Dorcas, 2010). However, our results do not support this for HAA. Instead, we speculate that this response reflects breeding behaviours that centre around burrow construction. Prior to breeding Heleioporus construct burrows in dry soil near drainage lines for calling and attracting gravid females. It is at the bottom of these burrows that amplexus and oviposition occur (Lee 1967). Successfully attracting a mate prior to a downpour may be a strategy employed by HAA to increase the survivability of egg masses. The impending rainfall would saturate the soil, creating a moist chamber suitable for egg masses to develop. Alternatively, a downpour could flood their burrows washing away deposited egg masses into deeper drainage pools. Heleioporus are frequently found breeding in ephemeral habitats (Lee 1967). This scenario may therefore increase the survivability of developing eggs and tadpoles in the long run as they are relocated into pools which will likely remain wetter for longer than constructed burrows.

There is also the possibility that HAA are more likely to call prior to a downpour as a last-ditch effort to breed before breeding habitats are over saturated. The WA congeners of HAA are known to cease all calling following heavy rainfall because breeding sites and burrows are flooded. Heleioporus a. australiacus are active outside burrows following heavy rainfall and are frequently identified along roads during and after heavy rainfall (Lemckert & Brassil, 2003; Penman et al., 2005b; Penman, Lemckert & Mahony, 2006a; Penman, Lemckert & Mahony, 2008a). Our results suggest that activity following heavy rainfall is unlikely to include calling for a mate, supporting previous anecdotes that rainfall is not necessary to initiate calling in HAA as it is in WA congeners who await the onset of winter rainfall to begin reproductive activities (Lee, 1967; Lemckert, Brassil & McCray, 1998). Heleioporus in WA typically await this rainfall event as it leads to increased soil moisture which facilitates burrow construction. The difference in reliance on rainfall between HAA and its WA congeners may stem from disparities in annual precipitation and soil moisture levels between south-western and south-eastern Australia where Heleioporus reside.

The model demonstrated that as daily rainfall levels exceeded 25 mm the probability of detecting HAA calling declined to zero. This is likely due to the oversaturation or potential flooding of theirs burrows which could diminish soil oxygen availability and draw them to the surface. Heavy rainfall can also pose a risk for individual frogs making them vulnerable to being swept away and reducing the probability of survivability. Heavy rainfall can also suppress HAA calling as their calls are likely drowned out by the sound of rain. This parallels the behaviour of other species that actively avoid calling during rainfall due to call interference (Dorcas and Foltz 1991). Indeed, the period when this study was conducted was during an exceptional La Niña event where record breaking rainfall was recorded in some parts of eastern Australia (Li et al., 2022), and therefore possible that detection opportunities were missed.

Finally, the model predicted that the probability of detecting HAA via calling would reduce with increased day-time solar radiation levels. On a clear day, solar radiation follows a normal distribution while air temperature satisfies the sine function (Shrestha, Thapa, & Gautam, 2019). That is – as the sun rises, solar radiation increases which leads to an increase in air temperature. As the sun passes its mid-point and begins to set solar radiation decreases which leads to a decrease in air temperature – known as radiative cooling. Our model showed that HAA were more likely to call when daily solar radiation levels are low and nightly temperatures are warm (above 20°C). We believe this indicates the presence of thick cloud cover which is known to reduce solar radiation levels during the day and traps escaping heat thereby reducing radiative cooling (Russak, 2009) and allowing for warm nights.

Influences of occupancy

As ectotherms with permeable skin, it is not surprising that temperature and precipitation had the strongest correlation on the breeding site occupancy of HAA. Breeding sites occupancy was positively related with the mean temperature during the warmest period and annual precipitation levels and negatively related to temperature seasonality and habitat aspect.

Physiological and behavioural performance in animals is greatly influenced by their ability to thermoregulate. Unlike endotherms, which produce metabolic heat, amphibians largely depend on their external environment to regulate their body temperature (Tracy, 1975). Habitat condition is critical for their survival and habitat selection often reflects a trade-off between their hydration and physiological performance. Habitats with higher ambient temperature can maximise performance but increases the rate of water loss, while those with lower ambient temperature reduce the risk of desiccation but impede performance (Köhler et al., 2011). Heleioporus a. australiacus are large and territorial frogs that breed over short periods during the warmer months of the year (Mahony et al., 2021), and their burrowing abilities award them year-round access to food and water resources. We therefore speculate that HAA occupy warmer breeding habitats to promote their physiological performance and increase the successfulness of their reproductive activities (i.e. burrow construction and defence, calling, and amplexus).

Animals must select breeding habitats that optimise the survival and fitness of their offspring (Rosenzweig, 1991). This is especially important for developing tadpoles, which lack parental care, are easily preyed upon and are highly sensitive to their environments. While the influence of temperature seasonality on the occupancy of HAA at breeding sites in our study likely reflects the species inhabiting coastal temperate forests, we speculate that it is also beneficial for developing HAA larvae. Temperature is one of the most important factors influencing young amphibian growth (Hochachka & Somero, 1984), and anuran larvae are exposed daily to temperature fluctuations. As temperatures fluctuate so too does the physiological rates controlling their ontogeny (Niehaus, Wilson & Franklin, 2006). Breeding in areas with lower temperature seasonality reduces the exposure of HAA larvae to thermal variation, which is especially important given their extensive developmental period can take 11 months (Anstis, 2007).

Precipitation is widely considered as one of the most significant factors that influences animal body size, physiology, and distribution (Parmesan & Yohe, 2003). Precipitation levels are on average higher, and rainfall occurs more frequently in the Sydney Basin bioregion than in the Jarrah Forest bioregions where many of the WA congeners inhabit. This disparity is likely to have led to the differing lengths of peak breeding season between the two populations of Heleioporus and may also explain why the larval period of WA congers is generally no longer than six months (Anstis, 2007). We therefore speculate that the relationship between occupancy and higher precipitation levels indicates a threshold requirement to increases the survivability of tadpoles and metamorphic HAA throughout their extended developmental period.

The aspect of a breeding habitat showed only moderate effects on HAA breeding site occupancy. Occupancy probability was highest (ψ > 0.25) for habitats situated at 25–100 degrees (NNE-EbS) and halved for sites beyond 150 degrees (SSE). This somewhat reflects previous findings where a higher frequency of HAA were observed in habitats located 45–135 degrees (NE-SE; Stauber, 2006). It’s possible this reflects a relationship between the habitats position and the amount of sunlight it receives. Sunlight facilitates tadpole development with vitamin D through the growth of algae that is consumed by tadpoles (Jäpelt & Jakobsen, 2013), and directly through UVB exposure (Whatley, et al., 2020). With the sun orientated more towards the north in the southern hemisphere, breeding habitats situated 0-90-degrees (N-E) and 270-360-degrees (W-N) would receive more daily sunlight than sites outside those ranges. Not reflected in our results was an increased in occupancy probability for sites 270 degrees (W) and beyond, which we believe is due to the intensity of the afternoon sun that is likely to increase the rate of evaporation and the risk of desiccation, and therefore decrease the probability of occupancy.

Soil profiles

The composition of soil profiles is an important factor that has been demonstrated globally to influence the distribution of burrowing species, and our results concord with these findings (Booth, 2006). Previous studies on radio-tracked H. australiacus found that individual frogs spent approximately 97% of their time in non-breeding habitats 20–250 m away from breeding habitats (Penman, Lemckert & Mahony, 2008b). We therefore hypothesized that occupancy would be positively related to breeding habitats with a higher percentage of sand within 500 m, as denser soils (i.e. those with a higher percentage of silt or clay) are likely to prevent or decrease the rate of burrow construction (Booth, 2006). Although soil profiles were not identified as the most correlative covariates with breeding habitat occupancy compared to other covariates, we found a negative correlation with silt and clay percentage in the soil profile. This supports previous understandings of habitat preference in HAA, and other burrowing species globally.

Breeding strategies

Heleioporus a. australiacus are often observed in low densities at breeding sites, and males are known to aggressively defend suitable microhabitats used for breeding (Mahony et al., 2021), which are characteristics of prolonged breeders, yet they are generally not observed at breeding sites for a period longer than 12 days (Penman, Lemckert, & Mahony, 2008b), indicative of explosive breeding. Despite being a relatively large frog (snout-vent length up to 80 mm) of robust body shape, reproduction generally occurs in ephemeral/permanent pools that occur along drainage lines (not permanent streams or rivers) and only rarely in permanent isolated ponds, which would also be indicative of an explosive breeding strategy. Furthermore, they have relatively large tadpoles (total length = 80 mm) that have an extended developmental period to metamorphosis which indicates a prolonged breeding strategy. While the exact climatic or seasonal triggers for reproduction were not previously well understood (Stauber, 2006), five congeners in WA have a tightly defined breeding season that is associated with autumn and winter rainfall (Lee and Main, 1954), with the frogs depositing eggs prior to the seasonal rainfalls occurring. We present evidence that reproduction in HAA is negatively associated with rainfall levels, although this climatic pattern is not a distinctive climatic seasonal pattern as in the south-west of WA. Similar to its congeners in WA, breeding in HAA is triggered prior to heavy rainfall, however, in the summer/autumn period (Stauber, 2006; Penman et al., 2006a). Together these indicate a prolonged breeding strategy. Lastly HAA occurs in a high rainfall region (average annual rainfall in Sydney Basin = 522–2395 mm) spread relatively across all seasons, which is also associated with prolonged breeding strategies. As such, it is unclear which breeding strategy (explosive or prolonged) this species leans more towards, instead we put it that HAA have explosive breeding events over prolonged periods.

Caveats

Due to the immenseness of the data available, analysis was conducted using only acoustic files recorded at a single hour of each night at each site. It is known that giant burrowing frogs also call outside this period and therefore possible that we missed sampling opportunities. Additionally, we speculated that HAA have explosive breeding events over prolonged periods,t is possible that we sampled in between these explosive bouts which may lead to underestimates of their calling period.

Our findings are framed in the context of our study period, which was characterised by widespread and intense bushfires, as well as a La Niña event. The 2019/2020 bushfires negatively impacted the occupancy and richness of threatened and common amphibian species (Beranek et al., 2023), and La Niña brought immense rainfall and flooding. Indeed, our low detection rates and the negative relationship observed with rainfall could be due to the absence of frogs having perished in the fire or avoiding flood conditions.

The giant burrowing frog is indeed a cryptic species, calling infrequently even in known breeding habitats. Breeding habitat detectability and occupancy was largely influenced by temperature and precipitation in our study and continued monitoring of HAA populations is vital as we enter a hotter and drier El Niño period that will likely alter this relationship.

The estimates of detectability and occupancy in our study provides essential information for those responsible for managing HAA and their environments. While passive acoustic monitoring has proven valuable to learning more about species phenology for known populations, we believe that it is not the best suited method for discovering new breeding populations due to the amount of monitoring required. We suggest instead frequent tadpole surveys be conducted as early as October through to May and PAR’s only be deployed once the species has been confirmed to breed in that location. We recommend that detectability be examined between PAR and visual encounter surveys for all anuran species as various life history and/or behavioural traits may influence the usefulness of these monitoring methods.

Author Contribution

All authors contributed to the study conception, design, data curation and investigation. Data analysis was performed, and the first draft manuscript was written by O Kelly. Authors CB, JG, SW, MM and AC commented on previous versions of the manuscript and all authors approved the final manuscript. Funding was obtained by MM, KK-T, and AC for fieldwork assessment and analysis through a grant from the bushfire recovery programme for wildlife and habitat, received from the Federal Department of Agriculture, Water, and the Environment (grant number: GA-2000241), and the 2019 Saving Our Species Contestable Grants Program (grant number: 2018 SSC 0048) from NSW NPWS Saving Our Species Program.

Acknowledgement

We acknowledge the Awabakal, Darkinjung, Dharawal, Dharug, Gundungurra, Wanaruah, Wiradjuri and Yuin People on whose land we conducted field work, and we pay our respects to elders past, present and emerging. We also acknowledge contributions from partner organisations including NSW NPWS and Forest Corporation NSW. We thank the Greater Blue Mountains wider community and especially Vic Giniunas for their knowledge on local giant burrowing frog habitats. We thank Shelby Ryan for GIS extractions and all the volunteers that have helped with field work and data collection.

Data Availability

The data that support the findings of this study are available by the Author upon reasonable request.

Conflict of Interest Statement

The authors declare no competing interests.

Agresti A (2002) Categorical Data Analysis. John Wiley and Sons Incorporated, Hoboken, New Jersey
Anstis M (2007) Tadpoles of south-eastern Australia: A guide with keys, 2nd edn. New Holland, Frenchs Forest, N.S.W.
Bakker VJ, Julia K, Baum, Jedediah F, Brodie AK, Salomon BG, Dickson HK, Gibbs OP, Jensen, Peter BMI (2010) The changing landscape of conservation science funding in the United States. Conserv Lett 3(6):435–444
Barata IM, Griffiths RA, Martin SR (2017) The power of monitoring: optimizing survey designs to detect occupancy changes in a rare amphibian population. Sci Rep 7:16491
Bartoń K (2012) MuMIn: Multi-model inference. R package version 1.7.2. http://CRAN. R-project. Org/package = MuMIn
Beranek CT, Sanders S, John Clulow, and, Mahony M (2022) Factors influencing persistence of a threatened amphibian in restored wetlands despite severe population decline during climate change driven weather extremes. Biodivers Conserv 31(4):1267–1287. https://doi.org/10.1007/s10531-022-02387-9
Beranek CT, Andrew J, Hamer SV, Mahony A, Stauber SA, Ryan J, Gould S, Wallace et al (2023) Severe wildlife promoted by climate change negatively impact forest amphibian metacommunities. Divers Distribution 29(6). https://doi.org/10.1111/ddi.13700
Booth DT (2006) Effect of soil type on burrowing behaviour and cocoon formation in the green-striped burrowing frog. Cyclorana alboguttata Can J Zool 84(6):832–838. https://doi.org/10.1139/z06-062
Celis-Murillo A, Deppe JL, Ward MP (2012) Effectiveness and Utility of Acoustic Recordings for Surveying Tropical Birds. J Field Ornithol 83(2):166–179. http://www.jstor.org/stable/23256481
Chadès I, McDonald-Madden E, McCarthy MA, Wintle B, Linkie M, Hugh PP (2008) When to stop managing or surveying cryptic threatened species. Proceedings of the National Academy of Science 105(37), 13936–13940. https://doi.org/10.1073/pnas.0805265105
Daly G (1996) Observations on the eastern owl frog Heleioporus australiacus (Anuran: Myobatrachidae) in Southern New South Wales. Herpetofauna 26:33–42
Daly G (2019) The distribution and status of the Giant Burrowing Frog Heleioporus australiacus in the Shoalhaven region of south-eastern New South Wales. Australian Zoologist 40(2):256–271. https://doi.org/10.7882/AZ.2018.011
Díaz-Paniagua C (1992) Variability in timing of larval season in an amphibian community in SW Spain. Ecography 15(3):267–272. https://doi.org/10.1111/j.1600-0587.1992.tb00034.x
Diggins CA, Michelle Gilley L, Kelly CA,W. Mark Ford (2020) Using Ultrasonic Acoustics to detect cryptic flying squirrels: Effects of season and habitat quality. Wildl Soc Bull 44(2):300–308. https://doi.org/10.1002/wsb.1083
Dixon JR, Heyer WR (1968) Anuran secession in a temporary pond in Colima, Mexico. Bull South Calif Acad Sci 67:129–137
Dorcas ME, Foltz KD (1991) Environmental effects of anuran advertisement calling. Am Zoologist 31(5):A111–A111
Doser JW, Andrew O, Finley AS, Weed, Elise FZ (2021) Integrating automated acoustic vocalization data and point count surveys for estimation of bird abundance. Methods Ecol Evol 12(6):1040–1049. https://doi.org/10.1111/2041-210X.13578
Fiske IJ, Chandler RB (2011) Unmarked: An R package for fitting hierarchical models of wildlife occurrence and abundance. J Stat Softw 43(10):1–23. https://doi.org/10.18637/jss.v043.i10
Gerhardt H, Carl, Franz Huber (2002) Acoustic communication in insects and anurans. University of Chicago Press
Gould J, Beranek C, Jose Valdez,Michael Mahony (2022) Quantity versus quality: A balance between egg and clutch size among Australian amphibians in relation to other life-history variables. Austral Ecol 47(3):685–697. https://doi.org/10.1111/aec.13154
Green DM (2003) The ecology of extinction: Population fluctuation and decline in amphibians. Biol Conserv 111(3):331–343. https://doi.org/10.1016/S0006-3207(02)00302-6
Haughland DL, Hero J-M, Schieck J, Castley G, Boutin SA, Solymos P, Lawson BE, Holloway G, Magnusson WE (2010) Planning forwards: Biodiversity research and monitoring systems for better management. Trends Ecol Evol 25(4):199–200. https://doi.org/10.1016/j.tree.2009.11.005
Henzi SP, Dyson ML, Piper SE, Passmore NE, Bishop P (1995) Chorus attendance by male and female painted reed frogs (Hyperolius marmoratus): Environmental Factors and Selection Pressures. Funct Ecol 9(3):485–491. https://doi.org/10.2307/2390013
Hijmans RJ, Phillips S, Leathwick J, Elith J (2021) Dismo: Species Distribution Modeling. R package version 1.3-5. https://CRAN.R-project.org/package=dismo
Hochachka PW, George NS (1984) Biochemical Adaptation. Princeton University Press, Princeton, NJ
Hoffmann EP, Nicola JM (2022) Breeding phenology of a terrestrial-breeding frog is associated with soil water potential: Implications for conservation in a changing climate. Austral Ecol 47(2):353–364. https://doi.org/10.1111/aec.13122
Jäpelt RB,Jette Jakobsen (2013) Vitamin D in plants: a review of occurrence, analysis, and biosynthesis. Front Plant Sci 4:136. https://doi.org/10.3389/fpls.2013.00136
Jeffrey SJ, Carter JO, Moodie KB, Alan RB (2001) Using spatial interpolation to construct a comprehensive archive of Australian climate data. Environ Model Softw 16(4):309–330. https://doi.org/10.1016/S1364-8152(01)00008-1
Kéry M (2002) Inferring the Absence of a Species—A Case Study of Snakes. J Wildl Manage 66(2):330–338. https://doi.org/10.2307/3803165
Kéry M, Benedikt RS (2008) Imperfect detection and its consequences for monitoring for conservation. Community Ecol 9(2):207–216. https://doi.org/10.1556/ComEc.9.2008.2.10
Köhler A, Sadowska J, Olszewska J, Trzeciak P, Berger-Tal O,Christopher R. Tracy (2011) Staying warm or moist? Operative temperature and thermal preference of common frogs (Rana temporaria) and effects on locomotion. Herpetological J 21:17–26
Köhler Jörn, Jansen M, Rodríguez A, Philippe JR, Kok LF, Toledo M, Emmrich F, Glaw, Celio FB, Haddad, Mark-Oliver, Rödel (2017) & Miguel Vences. The use of bioacoustics in anuran taxonomy: Theory, terminology, methods and recommendations for best practice. Zootaxa 4251(1), 1–124. https://doi.org/10.11646/zootaxa.4251.1.1
Lee AK, Albert RM (1954) Two new species of burrowing frogs of the genus Helioporus Gray from southwestern Australia. Western Australia Naturalist 4:156–158
Lee A (1967) Studies in Australian amphibia II. Taxonomy, ecology and evolution of the genus Heleioporus Gray (Anura: Leptodactylidae). Australian J Zool 15(2):367–439. https://doi.org/10.1071/ZO9670367
Lemckert F, Traecey Brassil, McCray K (1998) Recent records of the giant burrowing frog (Heleioporus australiacus) from the far south coast of NSW. Herpetofauna 28:32–39
Lemckert F,Traecey Brassil (2003) Movements and habitat use by the giant burrowing frog, Heleioporus australiacus. Amphibia-Reptilia 24(2):207–211. https://doi.org/10.1163/156853803322390453
Lemckert F,Michael Mahony (2008) Core calling periods of the frogs of temperate New South Wales, Australia. Herpetological Conserv Biology 3(1):71–76
Li, Xiaofan Z-Z, Hu Yu-heng, Tseng Y (2022) Liu, & Ping Liang. A historical perspective of the La Niña event in 2020/2021. Journal of Geophysical Research: Atmospheres 127(7), e2021JD035546. https://doi.org/10.1029/2021JD035546
Link WA, Sauer JR (1998) Estimating population change from count data: application to the North American Breeding Bird Survey. Ecol Appl 8:258–268. https://doi.org/10.1890/1051-0761(1998)008[0258:EPCFCD]2.0.CO;2
Littlejohn MJ, Angus AM (1969) Acoustic interaction between two species of leptodactylid frogs. Anim Behav 17(4):785–791. https://doi.org/10.1016/S0003-3472(69)80027-8
Lowe, Katrin J, Guy Castley, Jean-Marc H (2016) Calling phenology and detectability of a threatened amphibian (Litoria olongburensis) in ephemeral wetlands varies along a latitudinal cline: Implications for management. Austral Ecol 41(8):938–951. https://doi.org/10.1111/aec.12386
Luedeling E, Fernandez E (2022) chillR: Statistical Methods for Phenology Analysis in Temperate Fruit Trees. R package version 0.72.7. https://CRAN.R-project.org/package=chillR
MacKenzie DI, Nichols JD, Royle JA, Pollock KH, Bailey LL, Hines JE (2006) Occupancy estimation and modelling: inferring patterns and dynamics of species occurrence. Academic, Burlington, Vermont, USA
Mahony MJ, Penman T, Bertozzi T, Lemckert F, Bilney R, Stephen CD (2021) Taxonomic revision of south-eastern Australian giant burrowing frogs (Anura: Limnodynastidae: Heleioporus Gray). Zootaxa 5016(4):451–489. https://doi.org/10.11646/zootaxa.5016.4.1
Martin TG, Brendan A, Wintle JR, Rhodes, Petra M, Kuhnert, Scott A, Field SJ, Low-Choy AJ, Tyre, Hugh PP (2005) Zero tolerance ecology: improving ecological inferences by modelling the source of zero observations. Ecol Lett 8:1235–1246. https://doi.org/10.1111/j.1461-0248.2005.00826.x
McGrath TP, Guillera-Arroita G, Lahoz-Monfort J, Osborne W, Hunter D,Stephen Sarre (2015) Accounting for detectability when surveying for rare or declining reptiles: turning rocks to find the grassland earless dragon in Australia. Biol Conserv 182:53–62. https://doi.org/10.1016/j.biocon.2014.11.028
Niehaus AC, Robbie S, Wilson, Craig E, Franklin (2006) Short- and long-term consequences of thermal variation in the larval environment of anurans. J Anim Ecol 75(3):686–692. https://doi.org/10.1111/j.1365-2656.2006.01089.x
NSW Government (2023) NSW Biodiversity Conservation Information System. NSW Department of Planning, Industry and Environment. https://www.seed.nsw.gov.au/ (accessed 17 April 2022)
NSW National Parks and Wildlife Service (2003) The Bioregions of New South Wales: their biodiversity, conservation and history. NSW National Parks and Wildlife Service, Hurstville
O’Brien L (2021) Slga: Data Access Tools for the Soil and Landscape Grid of Australia. R package version 1.2.0 https://github.com/obrl-soil/slga
Paracuellos M, Rodríguez-Caballero E, Villanueva E, Santa M, Alcalde F, Dionisio MA, Juan R, Fernández, Cardenete et al (2022) Citizen science reveals broad-scale variation of calling activity of the Mediterranean tree frog (Hyla meridionalis) in its westernmost range. Amphibia-Reptilia 43(3):251–261. https://doi.org/10.1163/15685381-bja10094
Parmesan C,Gary Yohe (2003) A globally coherent fingerprint of climate change impacts across natural systems. Nature 421(6918):37–42. https://doi.org/10.1038/nature01286
Penman T, Lemckert F, Chris Slade,Michael Mahony (2005a) Nonbreeding habitat requirements of the giant burrowing frog, Heleioporus australiacus (Anura: Myobatrachidae) in south-eastern Australia. Australian Zool 33(2):251–257. https://doi.org/10.7882/AZ.2005.022
Penman T, Mahony MJ, Towerton AL, Francis Lemckert (2005b) Bioclimatic analysis of disjunct populations of the giant burrowing frog, Heleioporus australiacus. J Biogeogr 32:397–405. https://doi.org/10.1111/j.1365-2699.2005.01205.x
Penman T, Francis Lemckert,Michael Mahony (2006a) Meteorological effects on the activity of the giant burrowing frog (Heleioporus australiacus) in south-eastern Australia. Wildl Res 33(1):35–40. https://doi.org/10.1071/WR04081
Penman T, Francis Lemckert,Michael Mahony (2006b) A preliminary investigation into the potential impacts of fire on a forest dependent burrowing frog species. Pac Conserv Biology 12(1):78–. https://doi.org/10.1071/PC060078
Penman T, Francis Lemckert,Michael Mahony (2008a) Applied conservation management of a threatened forest dependent frog, Heleioporus australiacus. Endanger Species Res 5(1):45–53. https://doi.org/10.3354/esr00111
Penman T, Francis Lemckert,Michael Mahony (2008b) Spatial ecology of the giant burrowing frog (Heleioporus australiacus): implications for conservation prescriptions. Australian J Zool 56(3):179–186. http://dx.doi.org/10.1071/ZO08077
Petranka JW, Daphne AG, Thomas (1995) Explosive breeding reduces egg and tadpole cannibalism in the wood frog, Rana sylvatica. Anim Behav 50(3):731–739. https://doi.org/10.1016/0003-3472(95)80133-2
Pollock KH, Nichols JD, Theodore R, Simons GL, Farnsworth LL, Bailey, John RS (2002) Large scale wildlife monitoring studies: statistical methods for design and analysis. Environmentrics 13:105–119. https://doi.org/10.1002/env.514
Potapov P, Li X, Hernandez-Serna A, Tyukavina A, Hansen MC, Kommareddy A, Pickens A et al (2021) Mapping global forest canopy height through integration of GEDI and Landsat data. Remote Sens Environ 253:112165. https://doi.org/10.1016/j.rse.2020.112165
Core Team R (2023) R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing. https://www.R-project.org/
Recher HF (1990) & Leong Lim. A review of current ideas of the extinction, conservation and management of Australia’s terrestrial vertebrate fauna. In Australian ecosystems: 200 years of utilization, degradation and reconstruction: proceedings of a symposium held in Geraldton, Western Australia, edited by D.A. Saunders, 287–301. Chipping Norton, N.S.W., Australia: Surrey Beatty & Sons for the Ecological Society of Australia
Rosenzweig ML (1991) Habitat selection and population interactions: the search for mechanisms. Am Nat 137:S5–28. http://www.jstor.org/stable/2462286
Ross SRP-J, Darren P, O’Connell JL, Deichmann C, Desjonquères A, Gasc JN, Phillips, Sarab S, Sethi CM, Wood, Zuzana Burivavola (2023) Passive acoustic monitoring provides a fresh perspective on fundamental ecological questions. Funct Ecol 37:959–975. https://doi.org/10.1111/1365-2435.14275
Rowley JJ, Corey T, Callaghan T, Cutajar C, Portway K, Potter S, Mahony DF, Trembath Paul Flemons, & Adam Woods. 2019. FrogID: Citizen scientists provide validated biodiversity data on frogs of Australia. Herpetological Conserv Biology 14(1), 155–170
Roznik EA, Sarah J, Sapsford DA, Pike L Schwarzkopf, Ross A, Alford (2015) Condition-dependent reproductive effort in frogs infected by a widespread pathogen. Proceedings of Biological Sciences 282(1810): 20150694. https://doi.org/10.1098%2Frspb.2015.0694
Russak V (2009) Changes in solar radiation and their influences on temperature trends in Estonia (1955–2007). J Phys Res 114. http://dx.doi.org/10.1029/2008JD010613
Shrestha AK, Thapa A,Hima Gautam (2019) Solar Radiation, Air Temperature, Relative Humidity, and Dew Point Study: Damak, Jhapa, Nepal. Int J Photoenergy Article ID 8369231. https://doi.org/10.1155/2019/8369231
Specht HM, Henry T, Reich F, Iannarilli MR, Edwards SP, Stapleton MD, Weegman MK, Johnson BJ, Yohannes, Todd W, Arnold (2017) Occupancy surveys with conditional replicates: An alternative sampling design for rare species. Methods Ecol Evol 8:1725–1734. https://doi.org/10.1111/2041-210X.12842
Stauber A (2006) Habitat requirement and habitat use of the Red-crowned Toadlet, Pseudophryne australis, and Giant Burrowing Frog, Heleioporus australiacus, in the Sydney Basin. PhD Thesis. University of Technology, Sydney
Steelman CK,Michael E. Dorcas (2010) Anuran Calling Survey Optimization: Developing and Testing Predictive Models of Anuran Calling Activity. J Herpetology 44(1):61–68. http://www.jstor.org/stable/40665007
Sugai LSM, Thiago SF, Silva, Jose W, Ribeiro Jr, Diego Llusia (2018) Terrestrial passive acoustic monitoring: Review and perspectives. Bioscience 69(1):15–25. https://doi.org/10.1093/biosci/biy147
Tanadini LG, Benedikt R, Schmidt (2011) Population Size Influences Amphibian Detection Probability: Implications for Biodiversity Monitoring Programs. PLoS ONE 6(12):e28244. https://doi.org/10.1371/journal.pone.0028244
Taylor BD, Ross L, Goldingay,John M. Lindsay (2014) Horizontal or vertical? Camera trap orientations and recording modes for detecting potorros, bandicoots, and pademelons. Australian Mammalogy 36(1):60–66. http://dx.doi.org/10.1071/AM13012
Tracy CR (1975) Water and energy relations of terrestrial amphibians: insights from mechanistic modelling. Perspectives of Biophysical Ecology. Springer, Berlin, Heidelberg, pp 325–246. https://doi.org/10.1007/978-3-642-87810-7_19
United States Naval Observatory (2022) Fraction of the Moon Illuminated. Astronomical Applications Department. https://aa.usno.navy.mil/data/MoonFraction (Accessed: 11 March 2022)
Wells KD (1977) The Social Behaviour of Anuran Amphibians. Anim Behav 25(3):666–693. https://doi.org/10.1016/0003-3472(77)90118-X
Whatley C, Tapley B, Chang Y-MR, Newton-Yowens J, Mckendry D, Christopher JM (2020) Impacts of UVB provision on serum vitamin D3, pigmentation, growth rates and total body mineral content in Mallorcan midwife toad larvae (Alytes muletensis). J zoo aquarium Res 8(1):37–44. https://doi.org/10.19227/jzar.v8i1.434
Willacy RJ, Mahony M, Newell DA (2015) If a frog calls in the forest: Bioacoustic monitoring reveals the breeding phenology of the endangered Richmond Range Mountain frog (Philoria richmondensis). Austral Ecol 40(6):625–633. https://doi.org/10.1111/aec.12228

No competing interests reported.

SupplementaryMaterials.docx

Download PDF

Editor assigned by journal
24 Oct, 2024
Submission checks completed at journal
23 Oct, 2024
First submitted to journal
23 Oct, 2024

You are reading this latest preprint version

Using acoustic recorders and accounting for imperfect detection to understand spatial and temporal breeding patterns of a cryptic burrowing amphibian

Status:

Version 1

Abstract

Figures

INTRODUCTION

METHODS

Study area and site selection

Acoustic surveys

Variable of detection

Variables of occupancy

Equation (1)

Statistical Analysis

Equation (2)

Equation (3)

Equation (4)

RESULTS

Detectability

Occupancy

Soil profile

DISCUSSION

Influences on detection and calling behaviour

Influences of occupancy

Soil profiles

Breeding strategies

Caveats

CONCLUSION

Declarations

Author Contribution

Acknowledgement

Data Availability

Conflict of Interest Statement

References

Additional Declarations

Supplementary Files

Status:

Version 1