Lovers in strange places: amphibian calling and amplexus detected in tidal mangrove creeks after rainfall

Site 1

The first observation occurred at night on 27 October 2022 within Dead Mangrove Creek, after 20.6 mm of rainfall that spanned 7 days before and 2 days after a new moon during a spring tide event (NSW Tides 2022–2023). We heard 10–15 adult males calling as part of a chorus in the middle portion of the creek over an area spanning 0.67 ha (Appendix S1). The section of creek where L. aurea were detected was sparsely populated by mangrove trees compared to adjacent sections (Fig. 2a, b). We visually detected five adult individuals within the system; an adult male 20–30 cm from the bank among mangrove pneumatophores (Fig. 3a), a calling adult male at the water’s surface floating on an algal matt 4 m from the bank (Fig. 3b), an amplecting male/female pair partially covered by an algal matt at the surface of the water 5 m from the bank (Fig. 3c), and an adult male on a mangrove tree branch 30 cm above the water and 10 m from the bank (Fig. 3d).

Fig. 2.

Mangrove creek lines where Litoria aurea breeding activity was detected on Kooragang Island, NSW, Australia during the 2022–2023 breeding season. Images depict the first (a, b) and second observation (c) sites.

Fig. 3.

Litoria aurea detected in mangrove creek lines on Kooragang Island, NSW, Australia during the 2022–2023 breeding season. Images depict (a) adult male among mangrove pneumatophores near the bank, (b) adult male calling while perched on an algal matt at the water’s surface, (c) amplecting male/female pair submerged in water, and (d) adult male perched on a mangrove branch.

The water was approximately 40–50 cm deep, with large green algae matts covering more than 50% of the water’s surface and G. holbrooki present. Water quality was measured 5 m from the bank at the water’s surface (temperature = 20.6°C, conductivity = 2.79 mS, pH = 7.76, salinity = 1.4 ppt, TDS = 1.98 ppt) and near the bottom of the water column (temperature = 21.0°C, conductivity = 3.01 mS, pH = 7.37, salinity = 1.5 ppt, TDS = 2.14 ppt). Salinity was re-checked at the site on 28 November and had risen to 9.72 ppt, and no tadpoles were observed.

We also detected five calling males in a neighbouring ephemeral wetland approximately 80 m from the site of mangrove calling which also contained G. holbrooki and had comparatively higher salinity (temperature = 22.3°C, conductivity = 7.96 mS, pH = 7.65, salinity = 3.98 ppt, TDS = 5.66 ppt). Additionally, calling of two males was detected at a permanent wetland approximately 280 m from the site of mangrove calling that is a known breeding site in the area and free of G. holbrooki (temperature = 23.1°C, conductivity = 1.47 mS, pH = 7.13, salinity = 0.73 ppt, TDS = 1.04 ppt). The calling aggregation at Dead Mangrove Creek was the largest for the 2022–2023 breeding season on Kooragang Island; there were 107 instances of pond calling detected, including 32 occasions where calling aggregations consisted of five or more males.

Site 2

The second observation occurred at night on 25 November 2022 in a separate mangrove creek, 11 days after a rainfall event (27 mm), and 1 day after a new moon during a spring tide event (NSW Tides 2022–2023). The creek ended in a large opening of water dotted with mangrove trees (Fig. 2c) and was more than 400 m away from any freshwater wetlands where L. aurea have historically been recorded on the island. We heard four males calling over an area of approximately 0.4 ha and detected one calling male perched on a mangrove tree branch 20–30 cm over the water’s surface 5 m from the bank. We also heard one calling male in flooded samphire, Sarcocornia quinqueflora, saltmarsh adjacent to the creek. The water was approximately 20 cm deep and G. holbrooki was present. Water quality measurements were taken at the surface near to where calling was detected in the mangroves (temperature = 24.3°C, conductivity = 11.01 mS, pH = 6.47, salinity = 5.49 ppt, TDS = 7.8 ppt). No calling was detected in any other ponds nearby. Subsequent water quality measurements and surveys for tadpole presence were not made.

Salinity patterns at observation sites

As our first observation occurred immediately after 7 days of rainfall, we hypothesise that fresh water had flooded into the creek, causing lower than usual salinity well within the tolerance for L. aurea tadpoles (Pyke and White 2022; Callen et al. 2023). This is supported by the increased salinity a month later that was at levels typical for mangrove systems on Kooragang Island (10–35 ppt) (Streever and Genders 1997). Yet, this was also a period only a few days after a spring tide, which would have led to greater than usual influx of salt water. Although we are unsure of the precise hydrological regime of the creek, it is possible that tidal flow in combination with rainwater influx led to lower salinity. The spring tide resulted in a higher than usual tide prior to our observation, which may have reduced salinity if it flushed hypersaline water from the creek; hypersaline conditions can result from less regular inundation which is influenced by tidal range, barriers to tidal flow, and evaporation (Benfer et al. 2007; Herbert 2007). The influx of rainwater may have (1) diluted tidal water that was delivered at this flood tide and remained after the ebb tide or (2) fed a buoyant, freshwater lens at the surface that was preserved during the flood tide and came to dominate the creek following the ebb tide, with the area subsequently experiencing limited further inundation of tidal water after the spring tide event. What remains undetermined is the effect of freshwater input on the river channels surrounding the island, which have been reported to drop to 1.2 ppt at the surface after significant rainfall (Brereton and Taylor-Wood 2010), and how this influences the salinity of water entering tidal creeks.

The second observation occurred at a site with higher salinity (5–6 ppt), which has been shown to have no impact on L. aurea tadpole survival or development (Callen et al. 2023), but is below typical salinity for tidal creeks on Kooragang Island (Streever and Genders 1997). This observation also occurred soon after a spring tide but several weeks after a rain event and so it is possible that the freshwater lens had already begun to lose its integrity and experience mixing with tidal water to cause higher salinity. This may have provided a unique scenario where conditions were still optimal for offspring while providing a level of salt exposure for adults that has been shown to be beneficial against chytrid, reducing the risk of chytrid transmission for breeding adults that physically touch (Clulow et al. 2018).

Site selection

Our observations suggest that reproductively active males were able to detect the presence of fresh or low salinity water within mangrove creeks after rainfall. Although we are unsure of the initial trigger for movement to these sites, it could be a result of the species’ preference for recently refilled wetlands for breeding purposes (Beranek et al. 2021), with adults perceiving freshwater lenses as such. Although this sensory capacity requires further investigation, L. aurea is known to synchronise spawning at wetlands with rainfall, including those that have been previously dry (Pyke and White 2001; Beranek et al. 2022). Yet, only one previous record has been made of calling activity in mangrove creeks across the species’ range (Hamer 2002), suggesting conditions are rarely suitable to trigger movement into these systems.

Our observation at Dead Mangrove Creek in particular shows that reproductively active adults congregated within the creek line instead of neighbouring freshwater sites that were also available, some of which are used for breeding. We did detect low levels of calling activity in these ponds, suggesting that males were not forced to call from the creeks as sub-optimal sites due to density-dependent constraints on space.

Potential benefits of using freshwater lenses

There are potential benefits of selecting creeks as oviposition sites when they have converted to fresh water. As freshwater lenses are effectively newly created waterbodies, they have not yet been colonised by predators that would be found in more permanent freshwater systems (e.g. dragonfly larvae: Beranek et al. 2022). Although G. holbrooki was present at both sites it appeared to be at low densities, which may allow for successful breeding outcomes (van de Mortel and Goldingay 1998; Goldingay 2008), despite the fish being a voracious predator of L. aurea offspring (Pyke and White 2000). If timed correctly, it is also possible that a gradual increase in salinity at the very end of the offspring developmental period could provide early protection against chytrid, as metamorphs show an increase in keratin production which the fungus feeds on (Marantelli et al. 2004).

Brackish creeks as breeding traps

Exploiting tidal systems for reproduction is a risky strategy if freshwater lenses do not remain available long enough for offspring to complete development, becoming reproductive traps. Akin to the risk of using ephemeral wetlands that lose fresh water as they begin to dry out (Gould et al. 2022), tidal creeks will lose their freshwater lenses as the rain subsides and there is an influx of salt via tidal influences, which may result in mass mortality of tadpoles. Although L. aurea adults select breeding sites based on cues that allow an assessment of current breeding conditions (e.g. Pollard et al. 2017), this may not guarantee breeding success if those conditions are not retained throughout offspring development. The sustained presence of fresh water may not be quantifiable based on the conditions that parents experience during site selection and egg deposition (e.g. Gould et al. 2022), and total clutch failure due to freshwater ephemerality has been recorded for this species (Beranek et al. 2020). Species developmental rate will also influence the probability of offspring being able to successfully metamorphose prior to the loss of fresh water, which for L. aurea ranges from 5 to 12 weeks (Pyke and White 2001).

Habitat modifications potentially benefiting Litoria aurea

Anthropogenic disturbances to hydrology on Kooragang Island have benefited L. aurea, including the installation of bunds around ponds to prevent flooding from estuaries (Beranek et al. 2021). We suggest that L. aurea has unintentionally benefited from restrictions to tidal influences in mangrove creeks via the addition of infrastructure such as culverts and roads (Streever and Genders 1997), which may allow freshwater input from rainfall to remain present for extended periods. It remains to be determined how L. aurea will be affected by the removal of culverts from several creeks on the island to increase wetland productivity, as this has led to improvements in tidal penetration (Streever et al. 1996).

Conclusion

These observations provide evidence that L. aurea exploits generally high salt systems for breeding when fresh/brackish or low-salinity water becomes temporarily available after rainfall. The population dynamics of amphibians in environments with a salt water–fresh water interface is important to investigate as they may reveal novel ideas that can be applied to species conservation (Hettyey et al. 2019), or reveal ecoevolutionary pathways that amphibians may gain to withstand chytridiomycosis (Fisher et al. 2021).