The presence of amphibian larvae is restricted by both biotic and abiotic variables of the environment. Some of these variables are still undetermined in the septentrional eastern Palearctic where Rana amurensis, Strauchbufo raddei and Dryophytes japonicus are found in large numbers. In this study, we sampled 92 sites across Mongolia, Russia and the Democratic People’s Republic of Korea and measured biotic and abiotic water variables, as well as the height of flooded terrestrial and emergent aquatic vegetation at the breeding site. We determined that the presence of anuran larvae is generally, but not always, linked to pH and temperature. Rana amurensis was not significantly affected by any of the variables measured, while S. raddei was impacted by water conductivity and D. japonicus by pH, temperature and vegetation. Our results highlight a potential risk for these species due to the changes in aquatic variables in response to desertification.

anuran larvae, conductivity, Palearctic landscape, pH, salinity, septentrional Asia, species occurrence, vegetation, water biotic and abiotic properties

All amphibians above 40°N in East Asia rely on water bodies for the development of tadpoles. The characteristics of the aquatic environment is an important factor to the recruitment of populations and a limiting factor to the development of larvae (Altig and McDiarmid 1999; Loman and Lardner 2006). While biotic variables, such as aquatic vegetation, are important for predator avoidance (Kopp et al. 2006; Tavares-Junior et al. 2020), abiotic variables relate to water chemistry and specific physiological adaptations and metabolic processes (Hillman 2008). For instance, ion-exchange and respiratory circulation take place through a permeable skin allowing exchanges between internal and external fluids (Quaranta et al. 2009). Similarly to amphibians worldwide, species in Northeast Asia are also limited by water biotic and abiotic properties (Borzée et al. 2018; Kwon et al. 2021), including the three focal species of this study (Wang et al. 2009; Ko 2012; Huang et al. 2019).

Amphibian species, however, rely on different mechanisms to cope with aquatic environments and show varying degrees of tolerance (Hopkins and Brodie Jr. 2015), with variables, such as water chemistry (Viertel 1999), dissolved oxygen (Yu and Guo 2013), total dissolved solids (Browne et al. 2009) and water temperature (Indermaur et al. 2010), having different impacts on species. For instance, some species have a greater osmotic tolerance to brackish habitats (Wu et al. 2014), with larvae of species, such as Bufotes viridis, found in brackish water bodies with salinity up to 25 ppt (Katz 1973). Similarly, Fejervarya cancrivora larvae can be found in environments with salinity up to 35 ppt and can acclimatise up to 39 ppt (Hopkins and Brodie Jr. 2015). Vegetation is also important for some species as it provides shelter and food (Laurila 1998; Rieger et al. 2004) and, while most biotic variables are also critical, their impact is generally species specific (Altig and McDiarmid 1999).

This study focuses on three anuran species from Northeast Asia. Rana amurensis, S. raddei and D. japonicus are locally abundant species on the eastern Palearctic landscapes of Mongolia, Russia and the Democratic People’s Republic of Korea (DPR Korea hereafter). The water biotic and abiotic properties of the area are generally changing due to global change (Kelderman and Batima 2006) and aridification has become a threat (Shnitnikov 1975; Zolotokrylin et al. 2016). In addition, urbanisation, industrialisation and mining activities have contributed to the gradual pollution of waters (Hofmann et al. 2011). In this study, we correlate the presence of anuran tadpoles and the aquatic environment, so that a baseline is available to determine the ecological requirements of the focal species and it can be used in the future to examine the impact of regional and global changes.

Out of the 92 water bodies surveyed, we found R. amurensis tadpoles at 25 sites, S. raddei tadpoles at 34 sites and D. japonicus tadpoles at five sites (Fig. 1). The Generalised Linear Model used to determine the impact of the variables on the presence of tadpoles of all species together was significant (χ² = 159.53, df = 4, p < 0.001), but the results of the model showed that only pH (Wald-χ² = 62.24, B = 0.15, SE = 0.02, df = 1, p < 0.001) and temperature (Wald-χ² = 4.79, B = 0.01, SE = 0.01, df = 1, p = 0.029) had a significant impact on the presence of the anuran tadpoles, acknowledging that we did not test for interactions between variables. Conductivity (Wald-χ² = 0.69, B < 0.01, SE < 0.01, df = 1, p = 0.403) and vegetation (Wald-χ² = 0.45, B < 0.01, SE < 0.01, df = 1, p = 0.501) were not related to their presence. The presence of tadpoles was related to a marginally lower pH 8.71 ± 0.09 (mean ± SD) than their absence (8.82 ± 0.11), but their presence was related to a higher temperature (23.08 ± 3.39 °C) than their absence (21.17 ± 5.04 °C).

The three independent multinomial logistic regressions showed that for R. amurensis (n = 92), the overall model was significant (χ² = 21.81, df = 4, p < 0.001) and explained 43% of the variation (Nagelkerke R² = 0.43), although none of the variables was significantly explaining the presence of the species in relation to the four environmental variables (Table 1). In the case of S. raddei (n = 92), the model was significant (χ² = 16.55, df = 4, p = 0.002) and explained 22% of the variation (Nagelkerke R² = 0.22), and only conductivity significantly explained the presence of the species (Table 1). A higher conductivity was associated with the presence of the species (Table 2). Finally, for D. japonicus (n = 92) the model was significant (χ² = 102.01, df = 4, p < 0.001) and explained 89% of the variation (Nagelkerke R² = 0.89). The variables pH, temperature and vegetation significantly explained the presence of the species, albeit marginally for vegetation (Table 1). A higher pH, but lower temperature and higher vegetation were associated with the presence of the species (Table 2).

Finally, the three binary logistic regressions were significant for S. raddei (Wald = 5.11, df = 1, p = 0.024), but not for R. amurensis (Wald = 1.24, df = 1, p = 0.265) or D. japonicus (Wald = 0.61, df = 1, p > 0.433). Rana amurensis tadpoles were found in water with a maximum salinity of 2.87 ppt, S. raddei with a maximum salinity of 2.06 ppt and D. japonicus in a maximum salinity of 0.59 ppt (Fig. 2). A maximum of 4.14 ppt was recorded in the absence of any of the focal species.

Figure 2. 

Box plots displaying salinity tolerance in the three focal species of this study: Rana amurensis, Strauchbufo raddei and Dryophytes japonicus. Data were collected in Mongolia, Russia and DPR Korea during the breeding season in 2017 and 2018. Sample size: R. amurensis: 25, S. raddei: 34, D. japonicus: 5.

Our results highlighted the significant impact of pH and temperature on the presence of anuran larvae in the shallow water bodies of the septentrional eastern Palearctic. These results are in line with the general literature on amphibian tadpoles and the species of this region seem to follow the same broad patterns (Altig and McDiarmid 1999; Kuzmin et al. 2017). The relationship with pH is likely related to the acidity of the water or other chemicals for which pH is a proxy, resulting in a set of limiting factors potentially leading to the death of individuals in case of relatively weak variations in values (Clark and Lazerte 1985; Horne and Dunson 1995). Sensitivity to pH and other variations in water parameters, such as TDS (here correlated with all other variables) and salinity (here correlated with pH), may be linked to the mineral salts present in large concentration in the region and leaching into the environment, as well as pollutants of human origins (Linhoff et al. 2011; Afonina and Tashlykova 2018). In addition, greater changes are expected in the near future as global changes, locally resulting in an increase in the rate of desertification and thus linked environmental variables, are becoming increasingly prominent threats (Shnitnikov 1975; Zolotokrylin et al. 2016).

The sensitivity and tolerance to variations in water chemistry are, however, likely to be regionally mitigated by genetic variation and likely adaptation (Pierce and Wooten 1992). For instance, a difference in genetic expression between populations of Dryophytes cinereus makes coastal populations resistant to salinity levels that are deadly to the same species in other regions (Albecker et al. 2021). In addition, while the increase in specific variables is beneficial to the development of tadpoles, this relationship is generally limited to a specific level. For instance, while the significant correlation with an increase in temperature is likely related to the minimum requirement for growth at a relatively rapid pace due to the short duration of the warm season at high latitudes (Herreid and Kinney 1967), an increase in temperature above a certain threshold will have a negative impact on larvae when approaching the critical thermal temperature (Deutsch et al. 2008; Simon et al. 2015).

The fact that none of the variables tested was significant for R. amurensis may relate to the species ability to withstand hypoxia and the related increase in osmotic pressure (Shekhovtsov et al. 2020) and because the species may be able to partially detoxify internal fluids. The sensitivity of S. raddei to conductivity is likely linked to local variations in the landscape as minerals leaching chemicals into the environment are unlikely to be continuously present in the environment. In addition, local adaptations to the environment can also be mitigated by the synergetic action of environmental variables. For instance, we found adult individuals in water with salinity up to 4 ppt, but only at sites where we did not find tadpoles. The significant relationship of D. japonicus to pH, temperature and vegetation may be related to the points raised above for the two first variables. In addition, the occurrence of the species may be linked to the requirements of adults for vegetation cover when breeding in this landscape (Borzée et al. 2019). It is, however, important to note the risk of artefacts for D. japonicus due to the low sample size, especially in view of the species variable vegetation requirements in relation to humidity (Maslova 2001). Despite these restrictions due to environmental limitations, some populations are able to adapt to harsh environments by physiologically adjusting. For instance, R. amurensis and S. raddei were found in unusually high salinity, a pattern mirroring the tolerance of numerous species (Smith et al. 2007; Rios-López 2008; Hua and Pierce 2013; Hopkins and Brodie Jr. 2015).

In future studies, such patterns could be related to other variables and especially variation between populations, in relation to genetic and landscape connectivity. Intra-clade plasticity may be related to adaptation to changing environments and, thus, correlated with the potential to adapt to climate change. Predictions highlight the risk of salination of landscapes and clades with coastal populations able to cope with higher salinity may be able to adapt better in the coming decades.

We are grateful to Yoonjung Yi, Sungsik Kong, Minjee Choe, Kyungmin Kim, Desiree Andersen, Erdenetushig Purvee, Tumenkhuslen Munkhsaikhan and Solongo Gansukh for their help in the field. This project was supported by Rural Development Agency of Korea (PJ015071022021) to YJ. The research was partially supported by the budgetary project АААА-А21-121011390004-6 (Russia) to NS and the National Natural Science Foundation of China for Young Scientists (QN2021014013L) to AB.