We succeeded in estimating the distributional information of Japanese giant salamanders in Western Japan using eDNA analysis. The estimated distribution range of the Japanese giant salamander generally corresponded to the known natural distribution range of the species, as reported in previous studies (Biodiversity Center of Japan, 2001; Matsui et al., 2008). Although Fukumoto et al. (2015) examined the distribution of Japanese giant salamander using eDNA analysis in a single river system over two days, our study was able to survey its distribution much more broadly over 19 days, further indicating the usefulness of eDNA analysis for rapid and extensive surveillance of aquatic rare species distribution. Moreover, we showed a higher sensitivity of the nuDNA marker for detecting giant salamander eDNA and the negative effects of EC and alkaline conditions on eDNA detectability. Our study provides fundamental resources and methodology for surveying their entire distribution range using eDNA analysis and facilitating their conservation.
The results of a binomial GLMM showed that the nuDNA marker developed in this study had higher detectability of Japanese giant salamander eDNA than the mtDNA marker developed by Fukumoto et al. (2015). Given that the previous reports have shown higher fish eDNA detectability through multi-copy nuDNA markers (Minamoto et al., 2017; Dysthe et al., 2018), our results supported the higher eDNA detection sensitivity of multi-copy nuDNA markers and indicated its usefulness for efficient eDNA-based biomonitoring for amphibian species. Alternatively, the result may simply be due to the difference in PCR amplicon lengths between the markers (116 bp for nuDNA and 133 bp for mtDNA; Hänfling et al., 2016), although the eDNA concentration may not differ significantly between such small differences in PCR amplification lengths (Jo et al., 2020c). Nevertheless, the multi-copy nuDNA marker would allow for reducing the risk of false-negative eDNA detection and sensitively estimating the salamander’s distribution.
Our SODMs also revealed differences in the site-level eDNA occupancy probability between the two genetic markers in the studied rivers. Notably, when using nuDNA markers, surveys of at least five sites (eight sites if CIs were considered) were necessary to reliably infer the presence of salamander eDNA in the studied rivers. When using mtDNA markers, it was necessary to target twice as many sites (10 sites) to reliably estimate the presence of the target eDNA in the rivers. Assuming that the target eDNA and its source individual were adjacent, the results implied that nuclear DNA markers could infer the presence of the target species at fewer sites than the mtDNA marker, again suggesting the suitability of multi-copy nuDNA such as rRNA genes for sensitive and efficient surveillance of the distribution of aquatic rare species, including Japanese giant salamander, using eDNA analysis. Nevertheless, contrary to bacteria and fungi, sequences of nuclear genes available on the database (e.g., GenBank) are quite limited for vertebrate species (Handelsman, 2004; Toju et al., 2012; Minamoto et al., 2017). Database enrichment for vertebrate nuclear genes is highly important for eDNA applications with nuDNA markers in the future (Minamoto et al., 2017; Jo et al., 2022).
Binomial GLMM results also showed a significant negative effect of EC on the detectability of the Japanese giant salamander eDNA. High EC is an index of water quality decline, is related to the total dissolved solids and ion concentration in the water, and negatively impacts the occurrence, abundance, and reproductive success of aquatic species (Alavi & Cosson 2006; Bowles et al., 2006; Bodinof Jachowski et al., 2016). Some studies have previously documented negative relationships between EC and detectability of fish and amphibian eDNA (Pitt et al. 2017, Jo et al. 2020a; 2020b), which supports our findings. Although higher EC values can also relate to PCR inhibition (Schrader et al., 2012), it is unlikely that substantial PCR inhibitions occurred in our water samples (EC:0.00–0.25 mS/cm) given that Pitt et al. (2017) confirmed no evidence of PCR inhibition in river water samples with higher EC than ours (> 0.5 mS/cm). Moreover, in the 2014 survey, Japanese giant salamander eDNA tended to be detected less frequently in samples with higher pH. A higher pH (> 8.5) is also considered to reflect water quality decline due to wastewater, and thus, the result may have indicated a difficulty for the giant salamander to inhabit such areas. For example, the Sayo River tended to have the highest pH and EC of all the surveyed rivers, and eDNA detectability was generally low.
In contrast, the effect of pH on target eDNA detectability was not significant in the samples collected in 2015. This implies that although EC was a strong variable determining the giant salamander’s distribution, a higher pH may not always be observed at sites that the salamanders do not prefer to inhabit. Their distribution can also be determined by other environmental factors (e.g., annual mean temperature, precipitation, elevation, land use, and geological features) (Houlahan et al., 2000; Willson & Dorcas, 2003; Okada et al., 2008). Precipitation can be related to the flow status of the river, likely affecting the habitat of river-dwelling species, including the target species, and land use and elevation can be related to anthropogenic impacts on amphibian habitats (Johnson et al., 2011). Measurement of these parameters simultaneously with water sampling may have allowed for a more detailed assessment of the salamander's habitat suitability (Nicholson et al., 2020; Jo & Yamanaka, 2022).
Although the presence/absence of target eDNA in a water sample was the focus of this study, we may have been able to infer a more detailed ecology of the giant salamanders in the studied rivers by quantifying target eDNA concentrations. Environmental DNA concentrations can represent their relative abundance and activity in the environment (Pilliod et al., 2013; Spear et al., 2015; Iwai et al., 2019; Jo et al., 2020c). Continuous eDNA-based quantitative monitoring will inform time-series changes not only regarding their distribution but also their relative abundance in rivers. Moreover, as external fertilization (Kawamichi & Ueda, 1998) can increase the relative concentration of nuDNA to mtDNA in the water, the ratio of nuclear to mitochondrial eDNA concentrations could be used to estimate the timing and location of their spawning (Bylemans et al., 2017; Wu et al., 2022). These efforts will advance our understanding of the life history and reproductive ecology of giant salamanders and will make their conservation activities more efficient, although these points will be the subject of future studies.
Including our study, eDNA analysis has enabled the collection of broad-scale species distribution data on a considerably shorter timescale compared to conventional methods (Biggs et al., 2015; Deiner et al., 2017; Yao et al., 2022). Such an advantage of eDNA analysis can be quite useful not only for a rapid understanding of the species’ distribution and abundance, but also for revealing the relationship between their suitable habitats and environmental conditions. A few studies recently applied eDNA analysis to species distribution models (SDMs; a.k.a., ecological niche models) to link species occurrence records with environmental conditions and then estimated habitat suitability (the occurrence probability of a species at a site with a given environmental condition) (Riaz et al., 2018; Wilcox et al., 2018; Hashemzadeh Segherloo et al., 2022). Predicting species distribution has become increasingly important for conservation efforts owing to the impacts of recent climate change, including global warming caused by anthropogenic CO2 emissions. These effects are considered to cause habitat shifts and disturbances for various animals and plants, which may further accelerate in the future (Butchart et al., 2010; Cardinale et al., 2012; Ceballos et al., 2015). Combined with statistical approaches such as SDMs, eDNA analysis can promote biodiversity conservation and ecosystem management more efficiently and reasonably. Such studies would help to preserve the habitats of various rare species, including the giant salamander, and save them from extinction.