We conducted a quantitative analysis of the eDNA of ayu and F. psychrophilum in the rivers in Gifu Prefecture, Japan. In Gifu Prefecture, BCWD caused by F. psychrophilum was first identified in in 1996 (Hara et al. 2007) and has been a threat to fishery industries as well as to river ecosystems, and understanding its dynamics in rivers is expected to lead to the control of disease outbreaks. The eDNA distribution of both ayu and F. psychrophilum matched the known dynamics of both organisms: up- and downmigration of ayu in spring and autumn, respectively, and that of F. psychrophilum increased from November to May and decreased from June to September. Ayu and F. psychrophilum coexisted for a limited period in the river, during the up- and downmigration season of ayu, and therefore, the infection could be occurring during these seasons.
Distribution of ayu eDNA in the Nagara and Ibi rivers
The overall results of the eDNA survey for ayu were consistent with the known life history of ayu. In particular, the decrease in DNA concentration in the upstream region (NA9, IB6, and IB7) and the increase in DNA concentration in the middle reaches of the river (NA6 and IB5) on 30 September 2017, accurately reflected the spawning migration of ayu in these rivers (Komada 2016). In addition, NA5 and NA6 are known spawning grounds where many ayu spawn every year and the result matched the local ecology. This suggests that the eDNA survey is efficient for surveying the seasonal distribution of ayu in rivers over 100 km in length in addition to the shorter rivers surveyed previously (Saba River, Doi et al. 2017; Yodo River, Yamanaka and Minamoto 2016). The maximum eDNA concentration of ayu obtained in this study (8.3 × 106 copies/L) was larger than that recorded in other rivers (2.2 × 104 copies/L in the Gonokawa River [Inui et al. 2017] and 1.5 × 107 copies/L in the Tama River [Naito et al. 2018]; both rivers known to be rich ayu habitats), suggesting the presence of large ayu resources in the Nagara and Ibi rivers. However, since DNA concentrations increase during the spawning season, we need to be careful when comparing fish abundance based on eDNA levels. As a new finding, the eDNA of ayu was detected in some spots in January and February, after the spawning migration of most individuals, suggesting the presence of overwintering ayu that stay in a wide area of the river through winter. Overwintering ayu (those which live in the river in winter when other populations of ayu go down the river and live along the coast) have been reported in many rivers in Japan (Suzuki 1939; Sakae et al. 1996; Miyazaki 2008; Suzuki 2016). Although there are no official records of overwintering ayu in the Nagara and Ibi rivers, there is a possibility they may be present. In a previous study (Miyazaki 2008), F. psychrophilum was detected in the overwintering ayu in the Sho River, and they were suspected to be hosts of F. psychrophilum in winter and cause continuous infection by F. psychrophilum in the river.
Seasonal distribution of F. psychrophilum in the Nagara and Ibi rivers
Detection rate of F. psychrophilum in the rivers was very low from May to August. The eDNA concentration of F. psychrophilum in the downstream regions of the river increased in November, and maintained a higher concentration throughout the river during winter and early spring. The eDNA concentration of F. psychrophilum increased and decreased in the rivers without necessarily depending on the distribution of ayu, which suggested that the period in which ayu and F. psychrophilum coexisted in the river was limited. Generally, BCWD of ayu prevails before and after the rainy season, from May to July, which led to the hypothesis that F. psychrophilum would be detected mainly from May to July. However, our results suggest that ayu might be infected with F. psychrophilum earlier, during the up- and downmigration stages, before the apparent epidemic of BCWD.
The results of our surveys showed that F. psychrophilum was detected during the period of low water temperature and not during the high-temperature period. The GLMM results showed a negative effect of water temperature on the detection of F. psychrophilum eDNA, and the temperature threshold for the presence or absence of F. psychrophilum was 19.8 °C. In a previous laboratory-based heating experiment on F. psychrophilum, the bacterial population decreased at 23 °C and was inactivated at 28 °C. The decrease in detection rate of F. psychrophilum at higher temperatures is in line with previous findings, and shows the usefulness of eDNA analysis for the survey of F. psychrophilum.
eDNA of F. psychrophilum was absent from river surface water during summer, and their whereabouts remain unknown. It is possible that F. psychrophilum lives inside the body of ayu, in other fish that do not develop the disease, in attached algae on the riverbed, or in the deeper layers of the river, and requires further investigation. Furthermore, detailed investigation is needed to determine the localization of F. psychrophilum in winter. One possibility is that F. psychrophilum is held by the overwintering ayu. As mentioned above, the wintering ayu behaves differently from those with the well-known life cycle, and a case has been reported (Miyazaki 2008), in which F. psychrophilum was isolated from a wintering ayu when it was caught in April. There is a possibility that the overwintering ayu transmits the pathogen to the next generation of ayu.
Infection status of ayu during the upmigration period
From the above results, we hypothesized that ayu is likely to be infected with F. psychrophilum during the up- and downmigration periods, and the infectious status was checked during the upmigration period. We found that ayu harbors F. psychrophilum in early April when it migrates from the ocean coast to the river. This result is consistent with previous reports that showed that ayu harbors F. psychrophilum just after freshwater acclimation in April and May before the epidemic (Amita et al. 2000; Fujii 2009).
Considering the overall results, we suggest a mechanism of F. psychrophilum infection: during the up- and downmigration periods, when the ayu concentrates in a particular area of the river, ayu is infected with F. psychrophilum, which causes its death in June or October. Although the mechanism of transgenerational transmission of BCWD is not clearly understood, it has been shown that intra-egg infection is caused by the invasion of F. psychrophilum when it attaches to the egg surface in salmonid fish (Kumagai and Nawata 2010). Because eDNA of F. psychrophilum was also detected at a high concentration in areas below the spawning area of ayu, it is considered that F. psychrophilum is concentrated in the river water near the spawning ground. Additionally, because it is possible that a high concentration of eDNA of F. psychrophilum is released from the carcasses of ayu that have finished spawning (Ohara et al. 2010), there is a possibility of horizontal infection of eggs at the spawning ground.
In our study, because more than 90% of individuals were infected with F. psychrophilum after June, we suggest that many ayu have the possibility of developing the disease. The factors involved in the development of BCWD are suggested to be: (1) the increase in the growth efficiency of F. psychrophilum in ayu owing to the decrease in water temperature during the rainy season, and (2) the decrease in the immunity of ayu due to various stresses, such as rapid changes in water temperature, territorial conflict, and spawning behavior (Iguchi and Matsubara 2002; Iguchi 2003). In the spawning and downmigration seasons, it is possible that ayu larvae are infected with F. psychrophilum.
Perspectives for future countermeasures against BCWD
Based on the results of this study, the optimization of the timing of the release of ayu is considered to be an effective measure against damage from BCWD. In Japan, ayu has been released when the water temperature is higher than 13 °C according to the guidelines of the Ayu BCWD Countermeasures Council, the Ministry of Agriculture (Sato 2018). While this approach is believed to be empirically correct, it is not based on the mechanism of infection. This study revealed the period when F. psychrophilum spreads in rivers. In the future, by examining the release time using the amount of F. psychrophilum in rivers as an index, the damage caused by BCWD can be further suppressed. In Gifu Prefecture, the improvement in ayu seedlings resistant to BCWD is underway (Kuwada et al. 2010). Constructing an effective release plan for ayu possessing this resistance and conventionally farmed ayu in the future and conducting eDNA surveys in rivers will help to promote epidemic prevention in rivers by using the habitat survey of F. psychrophilum as an index.
As described above, F. psychrophilum invades and settles in the river ecosystem and becomes a threat to ayu owing to past release projects. In addition to ayu, various biological release projects are currently being conducted globally, which can cause the spread of alien organisms and the spread of pathogenic bacteria such as Edwardsiella ictaluri (Nagai and Iida 2008). By utilizing the eDNA survey used in this research as a method for host and pathogen monitoring, early countermeasures against various problems caused by alien pathogens can be taken and damage to wild and planted organisms can be reduced or prevented. Through such surveys, it is possible to control the disease by identifying the infection time and onset time of the disease. Through such a control measure of BCWD, we can protect ayu not only as commercial resources but also as cultural heritage.