Environmental DNA Detects Remaining Populations of Endangered Stream Salmon (Sichuan Taimen: Hucho bleekeri Kimura Salmonidae) in the Qinling Mountains

Abstract

Sichuan taimen (Hucho bleekeri Salmonidae) populations are declining, and the status of remnant populations in the Qinling Mountains is poorly known. In this study, eDNA and creel netting were used to estimate the distributions of Sichuan taimen in two upper tributaries of the Hanjiang River. A qPCR assay was applied to detect the eDNA of H. bleekeri, and the presence of Sichuan taimen in the Taibai River, but not in the Xushui River, was confirmed. Sampling during summer achieved a relatively higher detection rate. The utility of eDNA techniques for detecting low-density Sichuan taimen was demonstrated in intricate stream ecosystems. The eDNA method can increase the efficiency of the management of endangered freshwater species, such as Sichuan taimen, by providing reliable distribution data.

1. Introduction

An accurate knowledge of species distribution is crucial for effectively managing and implementing conservation measures for imperiled species. In China, fish monitoring programs in large rivers and streams primarily rely on conventional methods such as electrofishing, gill netting, and fishing. While these traditional methods are efficient sampling tools, they possess inherent limitations, including difficulties in sampling, labor-intensiveness, and dependence on professional expertise. Species identification in flowing streams using electrofishing or gillnetting is more challenging compared to still water systems due to the complexities of remote and inaccessible habitats, rugged streambeds, fast flow, and low individual density. Furthermore, the use of traditional monitoring techniques may cause severe injuries to both target and non-target fish, thereby increasing the risk of extinction, particularly for rare species [1,2]. These constraints associated with traditional methods impede expeditions in accurately surveying the distribution of streams inhabited by rare fish.

The environmental DNA (eDNA) method, based on detecting traces of DNA shed by organisms into the environment, has emerged as a powerful tool for the monitoring of aquatic and semi-aquatic species [3,4,5]. In the past decade, the applications of eDNA in biomonitoring have expanded, encompassing activities such as tracking endangered species and surveying biodiversity [6,7,8]. eDNA exhibits higher detection sensitivity compared to traditional sampling methods, making it particularly valuable for identifying rare or elusive species [9,10,11]. Research has demonstrated the superiority of the eDNA method over conventional monitoring techniques, as it is non-invasive, labor-efficient, and cost-effective [12,13]. Environmental DNA surveys are increasingly employed worldwide, providing opportunities for the conservation of imperiled species. However, the potential feasibility and practical application of eDNA methods for biomonitoring high mountain stream systems in China have yet to be explored.

All species in the genus Hucho are now listed as threatened or data-deficient on the IUCN Red List of Threatened Species [14]. A depressing case is the Sichuan taimen (Hucho bleekeri Salmonidae), a native salmonid that was once abundant in two regions of China [15,16]. One region is in northwest plateau of Sichuan, spanning Qinghai and Sichuan Provinces (including the upper and middle reaches of the Minjiang, Dadu, and Qingyi Rivers). The other region is in the Qinling Mountains in Shaanxi Province, encompassing the Taibai and Xushui Rivers in the upper reaches of the Hanjiang River. This species has experienced a significant range reduction and population decline in the latter part of the last century, largely due to hydropower development, habitat degradation, and overfishing [17]. According to expert judgment, the range loss has been estimated to be 98% since the 1960s and has been continuously declining [18]. Based on catch data from fishing and field surveys, the Sichuan taimen population within the northwestern plateau of Sichuan has nearly been extirpated. While Sichuan taimen sightings were still recorded in the last decade in the Makehe River, the Jiaomuhe River (downstream of the Makehe River, at the boundary of the Qinghai and Sichuan Provinces), and the Taibai River (in the Shaanxi Province), considering the ongoing deteriorating living status, including a severely fragmented range, continuous population loss, and degraded habitat, Sichuan taimen was listed as Critically Endangered (CR) on the IUCN Red List [19] and as a first-class protected wild animal by the Chinese government in 2021.

In this study, we utilized eDNA to determine the distribution of Sichuan taimens in the Qinling Mountains. First, the detection probabilities of eDNA for Sichuan taimen were assessed in different seasons along the Taibai River, where a recently discovered wild Sichuan taimen population exists. Second, an extensive and intensive field survey was conducted to identify the presence of Sichuan taimen in the Taibai and Xushui River basins. Our objectives were to (1) enhance the detection probability of Sichuan taimen eDNA by assessing the seasonal effects and (2) identify the remaining population of this endangered species in the Qinling Mountains. The findings from this study will enhance our understanding of the current distribution range of this stream fish and provide conservation recommendations.

2. Materials and Methods

2.1. Study Species

The Sichuan taimen displays a distinct landlocked life history separate from the other four anadromous species within the Hucho genus. During the Pleistocene glaciations, the Sichuan taimen migrated southward to the Qinling Mountains, and it is recognized as an epibiotic species [17]. Additionally, it holds a significant status as a flagship species in stream ecosystem conservation efforts. Biologically, Sichuan taimen prefers to dwell 700–3300 m above sea level in mountain streams with sandy and gravel substrates, narrow river beds, swift currents, high dissolved oxygen (>5 mg/L), and low water temperature (<15 °C) in the summer, and choose deep pools in winter [20]. In late winter, adult Sichuan taimen undertakes upstream migration to reach higher headwater tributaries where natural spawning beds are located. Spawning occurs during spring when water the temperature reaches 10 °C, and it takes place on stable streambeds [21]. Unfortunately, since the 1960s, the Sichuan taimen population in the Qinling Mountains has experienced a severe decline. The historic range has contracted to upstream tributaries of the Hanjiang River, the Taibai range of the Baohe River [22], and the upper reaches of the Xushui River basin [23].

2.2. Study Area

Currently, H. bleekeri is distributed in the Taibai and Xushui Rivers within the Qinling Mountains. These rivers originate from the southern slope of the Qinling Mountains and flow parallelly across southern Shaanxi, ultimately joining the Hanjiang River, one of the largest tributaries of the Yangtze River [24]. The Taibai River stretches over a length of 63 km and has a drainage area of 1349 km². With an average annual flow of 17 m³/s and a natural drop of 1588 m, the Taibai River provides a well-preserved riparian vegetation and riverbank, creating an ideal stream habitat for the remaining populations of Sichuan taimen. In 2012, during floods, 19 Sichuan taimens were captured near Taibaihe town as they were swept downstream [22]. In 2018, researchers using underwater visual observations and gill netting discovered the presence of wild young-of-the-year taimens in the river. The Xushui River is located on the northern latitude 33°11′–33°57′ and eastern longitude 107°02′–107°47′, with a watershed area of 2340 km² and a mainstream length of 167.40 km. The upper tributaries of the Xushui River primarily include the Hetaoping Stream, and the Dajian, Xiaojian, Maoer, Niuwei, and Heixiazi Streams. Several adult Sichuan taimens were caught in a deep pool in the Dajian Stream in 1999 [23].

2.3. Field Surveys

2.3.1. First Phase: Detection Rate of eDNA in Different Seasons

In the first phase of this study, we followed a protocol for detecting targeted DNA of Sichuan taimen in water collected from culture ponds [25]. A 236 bp D-loop region of mtDNA and common PCR were employed to compare the eDNA detection rates across different seasons to determine the best time for the distribution survey of Sichuan taimen.

We investigated seasonal effects on eDNA detection rates in the Taibai River, where wild populations of Sichuan taimen were known to be recently present. Between December 2015 and September 2016, we collected water samples every three months at six sites approximately 100 m apart along the Taibai River, covering a 600 m river section (sites 1–6, Figure 1). One sampling site was shifted downstream by approximately 50 m in December 2015 because the surface of the river was completely frozen (site 8, Figure 1).

The water samples were collected from rearing ponds to serve as positive controls (site 7, Figure 1). The fishery was situated near the bank of the Taibai River, approximately 350 m downstream from sampling site 1 (Figure 1). All the culturing ponds in the fishery were designed to allow for a mixture of groundwater and Taibai River water to flow into and through the ponds, eventually discharging into the Taibai River through an outflow located just below the fishery. Figure 1 presents the location map of the sampling sites, including the positive control (fishery), where the geographic information of all sampling sites was concealed to protect the species.

We conducted eDNA sampling at six sites and rearing pond from December 2015 to September 2016. At each site, three 2 L water samples were collected from the surface of the stream using a flow-through filter with a peristaltic pump and 0.45 mm cellulose nitrate filter paper. A blank control of distilled water was also filtered near one of the sampling sites to check for contamination. We removed each filter from the funnel with disposable gloves, folded it in half, and then rolled and preserved it in 95% ethanol in a 5 mL tube. Filters were stored on ice in the field, returned to the lab, and stored in a −20 °C freezer until DNA extraction. Meanwhile, 1 L distilled water was filtered as the negative control in subsequent experiments using the same sampling method [10]. A total of 80 water samples were filtered and transported back to the laboratory.

Water eDNA was extracted and amplified using a species-specific primer employing the same lab procedure described by Deng [25]. PCR products were run on 1% agarose gels to determine success, then purified using a Qiagen Min Elute PCR purification kit (Qiagen GmbH, Hilden, Germany), and were commercially sequenced in both forward and reverse directions on ABI3730 (Applied Biosystems, Foster City, CA USA). All water samples were run in triplicate to ensure the detection of degraded or low-quantity DNA. A sample was preliminarily judged as positive when a distinct band was detected in one of the nine PCR duplicates. Finally, we calculated the detection rate of the target DNA at each sampling time, and then considered the season with the highest detection rate as the optimal time for the following eDNA survey.

2.3.2. Second Phase: Determine Distribution Range of the Sichuan Taimen

In seasonal effect analysis, although the common PCR amplification method can obtain bright and single-target bands for the detection of culture water samples, occasionally there were several non-specific bands in the electrophoretic image when eDNA was amplified from the stream water samples. Therefore, a more sensitive TaqMan qPCR amplification procedure described by Deng [25] was applied to detect the targeted eDNA of Sichuan taimen along the Taibai River and the Xushui River.

In the second phase of this study, combining an eDNA survey and a baited creel-netting survey, eight streams of various sizes were chosen to assess the current distribution range of Sichuan taimen in the Qinling Mountains, primarily focusing on the Taibai River and the upper reaches of the Xushui River. Based on publications, news reports, and local interviews, 13 sampling sites distributed across six tributaries (including the Hetaoping Book, and the Dajian, Xiaojian, Maoer, Niuwei, and Heixiazi Streams) of the Xushui River, as well as 6 sites in the upper tributary of the Taibai River (Sujia Stream) were selected. The number of sites chosen in each stream varied based on historical records of Sichuan taimen presence. Detailed geographic information for all sampling sites can be found in

Table 1. Geographic information of the sampling sites in distribution survey.

Sample ID	Stream Name	Longitude (0)	Latitude (0)	Altitude (m)
Xushui River basin
S21	Hetaoping Stream	107.67271	33.81472	1535
S20	Hetaoping Stream	107.67643	33.81398	1540
S19	Hetaoping Stream	107.67777	33.81178	1542
S13	Dajian Stream	107.59071	33.76663	1714
S14	Dajian Stream	107.58762	33.76532	1659
S15	Dajian Stream	107.56453	33.76874	1532
S3	Xiaojian Stream	107.49081	33.74385	1274
S1	Xiaojian Stream	107.49025	33.74397	1276
S2	Xiaojian Stream	107.49981	33.74382	1254
S4	Xiaojian Stream	107.48938	33.743417	1310
S12	Maoer Stream	107.45415	33.73392	1069
S11	Niuwei Stream	107.35985	33.69751	1059
S22	Heixiazi Stream	107.34712	33.69592	951
Taibai River basin
S6	Sujia Stream	107.36523	33.91537	1613
S7	Sujia Stream	107.32536	33.91416	1628
S8	Sujia Stream	107.3298	33.9019	1632
S5	Sujia Stream	107.34709	33.90674	1692
S10	Sujia Stream	107.36503	33.90343	1733
S9	Sujia Stream	107.36512	33.90425	1756
S16	Taibai River			1235
S17	Taibai River			1227
S18	Taibai River			1224
Positive control	Fish pond
Negative control	Xianyi River	106.59852	34.82904
Blank control	distilled water

, and the corresponding location map is provided in Figure 2. However, to ensure species protection, the geographic information for an additional three sites along the Taibai River, which were sampled due to the success of eDNA detection during the seasonal variation test, has been omitted.

In 2017–2018, we conducted eDNA sampling and baited creel netting at 22 sites from June 20th to July 10th in both years. At each site, three replicates were collected. Two 3 m nets with a 0.4 cm mesh were also set up at each site, with the mouth opening facing the two flanks. The nets were left submerged in the water for at least 24 h. All fish caught using the creel net were identified, and their total length and weight were measured.

A total of 66 water samples of H. Bleekeri were analyzed using the optimized protocol. Each qPCR run included two replicates per water sample and incorporated positive, negative, and blank controls. All extractions and PCR procedures were conducted in a dedicated room specifically used for handling low-quantity DNA sources. No fish DNA was previously processed in this room. The negative filter, extraction, and PCR controls were implemented at each step to monitor contamination.

Furthermore, to determine the positive sites, the water samples filtered from rearing ponds were utilized as positive controls. Additionally, a water sample from a traditional Brachymystax lenok tsinlingensis habitat, specifically the Xianyi River (a secondary tributary of the Yellow River), where H. Bleekeri was been recorded or captured, was designated as a negative control to eliminate interference from the eDNA of B. lenok tsinlingensis (a co-occurring salmon in study area). A positive qPCR replicate was defined as a sample that had at least one well with a cycle threshold (Ct) value lower than the Ct value of the negative control.

3. Results

3.1. Detection Rate of Sichuan Taimen eDNA in Different Seasons

We found a significant impact of the sampling season on the detection probability of Sichuan taimen in the Taibai River. Overall, the samples collected in summer yielded the highest detection rate (88%; 16 positive results out of 18 samples). The samples collected in winter yielded the lowest detection rate (1 positive result out of 18 samples). The detection rate in the autumn or spring samples was moderate, with a 67 percent probability for the former and 39 percent probability for the latter (Figure 3 and

Table 2. Detection results of the eDNA of H. bleekeri in the Taibai River in four seasons. A check mark indicates positive detection and an x-mark indicates negative detection.

Stream Name	Site	Sample	Winter	Spring	Summer	Autumn
Taibai River	S1	S1-1	×	×	√	×
		S1-2	×	×	√	×
		S1-3	×	×	√	×
Taibai River	S2	S2-1	×	√	√	√
		S2-2	×	√	×	√
		S2-3	×	√	√	√
Taibai River	S3	S3-1	×	√	√	√
		S3-2	×	√	√	√
		S3-3	×	√	√	√
Taibai River	S4	S4-1	×	×	√	×
		S4-2	×	×	√	√
		S4-3	×	×	√	√
Taibai River	S5	S5-1	×	×	√	×
		S5-2	×	×	×	√
		S5-3	×	×	√	×
Taibai River	S6	S6-1	√	×	√	×
		S6-2	×	×	√	√
		S6-3	×	√	√	×
Fish pond	Positive control	C+	√	√	√	√
Xianyi River	Negative control	C-	×	×	×	×
distilled water	Blank control	BC	×	×	×	×

). DNA sequences from a single PCR for each site confirmed that the positive samples were Sichuan taimen, and all sequences were identical.

3.2. Creel Survey Results

Conventional monitoring using creel netting caught a young-of-the-year Sichuan taimen once at site 16 (detection rate = 4.5%) in the Taibai River, but did not detect Sichuan taimen in the Xushui River basin in 2017–2018. Another co-occurring salmon, B. lenok tsinlingensis, was caught everywhere in those streams, except in the Taibai River. Overall, 11 different kinds of indigenous fish belonging to 11 genera in 3 families were obtained during the two years of monitoring using creel fishing. Rhynchocypris lagowskii, Pseudogobio vaillanti, and B. lenok tsinlingensis were among the most common fish caught in this region, accounting for approximately 96.48 percent of the total fish weight. The fish from the netting and abundance data from 2017 to 2018 are listed in

Table 3. Composition of the catch in 2017–2018 using creel netting. The numbers refer to the number of species individuals caught.

Species	Number	Weight (g)		Length (cm)		Quantity Percentage (%)	Weight Percentage (%)
		Observed Value	Average Value	Observed Value	Average Value
Taibai River
Hucho bleekeri	1	1.94	1.94	5	5	0.76	0.15
Rhynchocypris lagowskii	125	3.76–43.67	17.86 ± 3.87	4.5–15.5	10.31 ± 3.87	94.70	98.35
Botia superciliaris	3	2.3–6.43	4.74 ± 2.17	7–10	8.83 ± 1.61	2.27	1.08
Sarcocheilichthys nigripinnis	3	1.28–2.62	1.86 ± 0.76	5–6.5	5.83 ± 0.76	2.27	0.42
Sujia Stream
Brachymystax lenok tsinlingensis	3	15.09–163	74.95 ± 77.88	11.5–26	17.88 ± 7.42	3.26	24.94
Rhynchocypris lagowskii	86	2.8–31	14.26 ± 7.12	4–11.5	9.06 ± 2.79	93.48	71.63
Cobitis sinensis	3	7.1–13.22	10.32 ± 3.07	9–11.5	10.50 ± 1.32	3.26	3.43
Hetaoping Stream
Brachymystax lenok tsinlingensis	3	38–43	40.83 ± 2.57	11–15	13 ± 2	50	79.28
Rhynchocypris lagowskii	3	9–13	10.67 ± 2.08	4.5–6.5	5.17 ± 0.76	50	20.71
Dajian Stream
Brachymystax lenok tsinlingensis	17	22–305	63.12 ± 69.13	5–29	13.06 ± 6.95	94.44	99.28
Rhynchocypris lagowskii	1	6.5		5		5.56	0.72
Xiaojian Stream
Brachymystax lenok tsinlingensis	7	33.96–129	63.98 ± 31.6	15.2–21	18.03 ± 2.03	2.88	26.81
Rhynchocypris lagowskii	235	4–26.4	17.86 ± 2.87	4.5–15.2	10.11 ± 3.27	96.71	72.33
Paracobitis variegatus	1	14.44	14.44	17	17	0.41	0.86
Maoer Stream
Brachymystax lenok tsinlingensis	4	40.24–204	84.8 ± 79.55	16–26	18.88 ± 4.77	9.30	51.99
Rhynchocypris lagowskii	39	3.4–33.5	7.48 ± 2.35	2–15	6.45 ± 5.62	90.70	48.01
Niuwei Stream
Zacco platypus	7	8.37–21.3	14.15 ± 5.13	9–20	12.43 ± 3.54	10.61	10.50
Gnathopogon imberbis	5	4.26–16.35	9.71 ± 5.7	8.5–11.5	9.7 ± 1.44	7.58	5.14
Pseudogobio vaillanti	50	13.67–51.63	28.13 ± 11.12	12–19	14.86 ± 1.85	75.76	74.54
Trilophysa bleekeri	1	3.37	3.37	4	4	1.51	0.35
Brachymystax lenok tsinlingensis	1	57.25	57.25	18	18	1.51	6.06
Rhynchocypris lagowskii	1	26.55	26.55	15	15	1.51	2.81
Varicorhinus macrolepis	1	5.5	5.5	8	8	1.51	0.58
Heixiazi Stream (non catch were obtained)

.

3.3. eDNA Survey Results

We detected Sichuan taimen eDNA at 5 (sites 6, 7 and 16–18; see Figure 2) of 22 sites, and the detection rate (22.73%) was obviously higher than that of creel netting. The Ct values were (19.97 ± 1.27) for the positive controls (H. bleekeri mtDNA in rearing pond) and (34.98 ± 0.50) for the negative controls (B. lenok tsinlingensis mtDNA of Xianyi River). The Ct values for positive detection ranged from 32.41 to 34.96 (mean of 33.68) (Figure S1, Table S1). The Ct values for the remaining samples ranged from 38.01 to 39.89, and we recorded those sites as negative detections, considering the fact that B. lenok tsinlingensis was caught everywhere at those sites, and the Ct values of those sites may have been caused by the amplification of B. lenok tsinlingensis mtDNA.

3.4. Distribution Range of Sichuan Taimen in Qinling Moutains

Among 12 sites (sites 5–10 and 13–18; see Figure 2) where they were expected to be present based on historical data since 1999, Sichuan taimen eDNA was detected at 5 sites. In the Taibai River and its upper tributary, Sujia Stream, nine sampling sites (sites 5–10 and 16–18; see Figure 2) were analyzed. Although Sichuan taimen eDNA was not detected at all sites, we considered the four negative sites (sites 5 and 8–10; see Figure 2) to have a very low density. The possibility remains that Sichuan taimen exists at these sites, but at very low levels that may be beyond the detection limits of our chosen methods. However, in the Xushui River basin, neither eDNA detection nor the creel survey supported the presence of H. bleekeri in any of the six streams (sites 1–4, 11–15 and 19–22; see Figure 2) in this study. It is highly likely that Sichuan taimen is absent from those sites in this region.

4. Discussion

4.1. Optimal Season for Monitoring Sichuan Taimen

The probability of detecting H. bleekeri eDNA varied throughout the year. The highest detection rates in June suggests that the target eDNA concentration may have increased in June. It is likely that some extra eDNA was discharged into the river during this period, which might be attributed to the increased fish biomass after the spawning season. For the Sichuan taimen in our study region, spawning typically occurs from late May to mid-June, followed by three weeks of hatching [21]. Interestingly, this period closely aligns with the peak eDNA concentration observed in June. Similar findings have been reported in other studies [26], indicating a positive relationship between eDNA concentration and breeding events [27,28,29]. These results suggest that directing sampling efforts toward the breeding season of target organisms may enhance the detection probabilities.

4.2. High Detection Capacity of eDNA

In the field survey, we performed a comparative analysis between the conventional tool and eDNA methods in terms of detection capacity and labor consumption. Among the 22 sites, one young-of-the-year taimen was caught at a sampling site in the Taibai River, and no more taimens were caught by the creel net during 2017–2018, not even at the other two sites in the Taibai River where we expected the occurrence of Sichuan taimen based on visual observation. Creel-netting surveys are likely to be less effective owing to the lower attractiveness of fish bait, observer error, and the low abundance of Sichuan taimen in these habitats. In the eDNA survey, Sichuan taimen were detected at the site where the creel net caught taimen and at four sites where no taimens were caught, indicating that eDNA methods are likely to be sensitive and effective tools for detecting Sichuan taimen with very low abundance. In addition, we found that the eDNA method is less labor-intensive, particularly for remote mountain regions where sampling sites are often inaccessible due to bumpy mountain tracks and slippery rocks at the stream bed. Our findings further support the notion that eDNA is superior to traditional sampling methods in terms of greater detection sensitivity, virtually unlimited sampling time, lower labor consumption, and routine sampling schemes in intricate aqueous systems [9,10].

To mitigate false detection, in the distribution survey, water samples were collected in triplicate at each of the 22 sites, among which all replicates producing positive amplification only occurred at one site, and one or two of the three were positive at four sites. This suggests that in this application, at least three samples should be collected at a site to reduce false negatives. Our results are consistent with those of previous studies [30,31].

4.3. Distribution of Sichuan Taimen in the Qinling Mountains

In the Qinling Mountain region of the Shaanxi Province, by combining the eDNA method with creel netting from 2017 to 2018, we found that Sichuan taimen was present in the Taibai River and its upper tributary (Sujiagou Stream); a historic distribution range also reported in recent studies [22]. On the other hand, the results for four sites in the upper reach of the Taibai River, where Sichuan taimen eDNA was not detected, were unexpected. These four sites were located in the Sujiagou Stream, which is the headwater of the Taibai River and is near the upstream extent of the range of the spawning habitats in the basin. Moreover, this tributary had rapid water flow and the lowest water temperature among the sampled tributaries during the surveillance period. It is possible that with a small amount of DNA and a fast-flowing stream flow, the density of DNA was too low to be detected by our technique. In another traditional habitat in the upper reaches and tributaries of the Xushui River basin, particularly in the Dajiangou Stream, neither eDNA nor creel surveys detected the presence of Sichuan taimen, indicating that Sichuan taimen might have become extinct in the Xushui River basin. Our result is consistent with the findings of fishery resource monitoring in the Xushui River basin carried out by our colleagues in the last two years (unpublished data).

Habitat retraction and population decline in Sichuan taimen not only occurred in the Qinling Mountains, but also in other traditional habitats in the Sichuan and Qinghai Provinces, where the situation is even worse.

Recent fishery resource monitoring programs have shown that no Sichuan taimens were caught in the Qingyi, Minjiang, and Dadu rivers using traditional tools in the last few years [32,33,34]. Considering the low detection rate of the traditional survey in this study, we suggest that the high-sensitivity eDNA method could be employed in those rivers to detect the remnant populations of Sichuan taimen in the Sichuan and Qinghai Provinces.

Moreover, it is worth noting that these trends of ongoing fragmentation and reduction in habitat for Sichuan taimen are likely to continue into the future; therefore, more vigorous strategies for the conservation of this precious fish, particularly with the implementation of reintroduction and habitat reconstruction, are urgently needed. Fortunately, the Chinese government and related agencies have already paid much attention to rare freshwater fish protection after the extirpation of the Yangtze River Dolphin (Lipotes vexillifer), and the artificial propagation and reintroduction of Sichuan taimen have been undertaken in the Qinghai and Sichuan Provinces.

4.4. Low eDNA Concentration in the Taibai River

qPCR analysis can be used to quantify the amount of DNA present in a sample and estimate the relative abundance of individuals [35,36,37]. In the distribution survey, Sichuan taimen eDNA was detected at five positive sites along the Taibai River and its upper tributary, Sujia Stream, but the Ct value of these positive detections was much higher than that of the positive control (i.e., rearing pond with fish density of 0.08 kg/m³). Although our limited data are insufficient to make an accurate estimate of the abundance of Sichuan taimen in the Taibai River, the gap between the two Ct values in this study showed some signs of an extremely low abundance of Sichuan taimen in the Taibai River.

5. Conclusions

In our study using a time-series design, we observed that the efficacy of eDNA surveys for stream salmon was influenced by the sampling season. This suggests that considering the sampling time is crucial to achieve a relatively high detection rate, which is desirable for monitoring rare species. We successfully detected eDNA at sites where the presence of Sichuan taimen was confirmed through catches using baited creel netting. Additionally, we detected eDNA at sites where Sichuan taimens were not detected using traditional tools. These findings indicate that the eDNA method proved to be useful and effective in detecting the remaining population of Sichuan taimen, despite their very low abundance. Our results suggest that Sichuan taimen can be reliably monitored using eDNA and that this approach may be useful for obtaining distribution data for other rare species. Furthermore, eDNA can be used for quick surveys before the application of traditional surveys with traps and electrofishing, thus making the best use of limited time when management decisions are urgently needed.