Mass marking of juvenile Schizothorax wangchiachii (Fang) with alizarin red S and evaluation of stock enhancement in the Jinping area of the Yalong River

Experimental fish and fluorochrome

Juveniles of S. wangchiachii used for marking and stocking in this study were produced at Jinping Hatchery using an artificial propagation technique. The breeding stocks were native spawners caught from the wild in the Jinping area of the Yalong River in 2011 and 2012. Juveniles were reared in numerous cylindrical tanks in the juvenile rearing room. These tanks are made of fiberglass, have a diameter of 2 m and a height of 1 m, and each has a temperature-controlled water supply (water temperature, 15.5–17.5 °C; dissolved oxygen concentration, 7.0–7.4 mg L⁻¹; pH, 7.1–7.4) from the recirculating aquaculture system.

The fluorochrome ARS (C₁₄H₇NaO₇S) used for marking was analytically purified powder. During immersion marking, ARS was dissolved directly in the rearing water according to the experimental design. To optimize marking quality and minimize juvenile mortality, several preliminary experiments were conducted in 2013, and results showed that immersing juvenile S. wangchiachii in water containing ARS doses ≤100 mg L⁻¹ for 24 h resulted in the lowest death rate while producing a mark that could be seen clearly in the otolith.

Small-scale marking pilot study

A small-scale marking pilot study was performed in the rearing tank. Approximately 7,000 40-day-old juvenile S. wangchiachii of the 2014 cohort (total length 23.52 ± 1.50 mm, mean ± S.D., n = 20) reared in a tank were selected for immersion marking on 5 May 2014. These juveniles were starved for 24 h prior to the treatment. The inner wall of the tank was carefully cleaned, and the rearing water was completely replaced with about 500 L of clean water. Thirty-five grams of ARS were pre-dissolved in about 10 L of water, which was immediately added to the tank. The juveniles were immersed in the ARS solution (70 mg L⁻¹) for 24 h. During immersion, the solution was aerated continuously and the fish were not fed. After immersion was completed, the ARS solution was discharged into a sewage pool, and at the same time clean water was pumped into the tank to thoroughly rinse out the remnant dye. One day after immersion, dead individuals were counted and removed from the tank. To check the visibility and persistence of the marks, 10 marked fish from the tank were haphazardly sampled and sacrificed with an overdose of MS-222 (100 mg L⁻¹) on 29 May 2014, 28 December 2014, 4 May 2015 and 24 January 2016. Sampled fish were kept in 100% ethanol until otolith examination.

Large-scale marking application

From late April to early May 2015, five batches of juvenile S. wangchiachii (total length 20.85 ± 1.41 mm, mean ± S.D., n = 140) of the 2015 cohort were marked at Jinping Hatchery using the ARS immersion protocol described above. In total, an estimated 600,000 fish were marked. To ensure the safety of the very small juveniles during immersion marking, the concentrations of ARS solution were controlled within a range of 30–50 mg L⁻¹. One day after immersion was completed, dead fish in each tank were counted and recorded. The mortalities of three batches of unmarked fish in the rearing room also were recorded. Five days after marking, both marked and unmarked fish were transferred to four outdoor fishponds in the hatchery. The fish were fed to satiation three times a day with a commercial artificial compound diet.

To assess effects of the marking process on growth, 100 marked and unmarked S. wangchiachii at similar daily age were haphazardly taken from two fishponds on 24 July 2015. These fish were starved for 24 h prior to further treatment. Afterwards, they were anaesthetized with MS-222 at a concentration of 100 mg L⁻¹. Their total length and standard length were measured to the nearest 0.01 mm with a digital caliper, and their wet mass was weighed to the nearest 0.0001 g with a precision electronic balance. To assess the mark effectiveness, 400 marked fish also were selected haphazardly from those in the four fishponds. They were anaesthetized with MS-222 at a concentration of 100 mg L⁻¹ and then stored in 100% ethanol until otolith examination.

Release and recapture

On 24 July 2015, 840,000 young S. wangchiachii of the 2015 cohort, of which 400,000 individuals were had been marked by ARS, were released at sites 2 and 3 in the Jinping area of the Yalong River (Fig. 1). Site 2 (28°18′47.80″N, 101°38′51.19″E) is located in the wide and deep part of the river (>5 m maximum depth). Although it was not a suitable habitat for S. wangchiachii, site 2 was used because of stairs that provided access to the river. Site 3 (28°19′41.32″N, 101°38′52.18″E) was near a sand quarry, and the substrate was covered with gravel and small stones: this was an appropriate habitat for young fish. The distance between the two release sites was about 2 km. The fish were first captured from each fishpond with a nylon trawl, and 150 marked individuals were haphazardly selected for measuring and weighing. Afterwards, fish were transferred to the release sites by a pickup truck and released into the river using buckets. For comparison of growth, about 10,000 unmarked S. wangchiachii were raised as a control group in a fishpond in Jinping Hatchery. They were fed twice daily with commercial feed at a ratio of about 3% of body weight per day.

Before the recapture surveys were conducted, appropriate recapture sites along the Jinping area of the Yalong River were chosen. Criteria included ease of fishing, suitable habitats for young fish and the distances from the release sites. Seven sites, including the two release sites, along a 60 km stretch of the Yalong River in the Jinping area were selected for the recapture surveys (Fig. 1). Site 1 (28°17′46.38″N, 101°38′39.19″E) was a shallow riffle area (<0.5 m in depth) with a substrate of gravel and small stones located about 1.5 km upstream of site 2 and 7 km downstream of Jinping Dam II. Site 1 was an important nursery ground for Schizothorax fish at early life stages. Site 4 (28°20′50.68″N, 101°39′18.04″E), site 5 (28°24′10.28″N, 101°43′24.86″E), site 6 (28°27′41.38″N, 101°44′49.65″E) and site 7 (28°36′58.04″N, 101°55′56.04″E) were located about 3, 15, 20 and 50 km downstream of site 3, respectively, and had a similar substrate of small stones and occasional boulders.

Recapture surveys were carried out at three-month intervals (in October 2015, January 2016 and April 2016) using a similar method and sampling effort each time. At each site, three 9-m long fishing pots with 6-mm mesh were used for recapturing fish for four successive days, and catches in each pot were removed once a day. The fishing pot used in these surveys was a trap-type stationary fishing device that was especially suitable for catching small fish with total length <20 cm. Electrofishing permitted by the Sichuan Municipal Bureau of Aquatic Products was performed only one time at each site, and it involved using a 30-cm-diameter anode and a 6-mm mesh landing net to sample for 40 min along the river. Specimens of S. wangchiachii assumed to be from the 2015 cohort based on personal experience of age-total length were sacrificed with an overdose of MS-222 and measured and weighed, whereas other fish were released. Meanwhile, 50 individuals sampled from the hatchery control group also were measured. All sampled fish were stored in 100% ethanol for further processing.

Otolith removal and examination

Three pairs of otoliths were removed from all fish sampled, and the left three otoliths were mounted on glass slides using neutral balsam. To check the ARS mark and to read age, otoliths were observed under an Olympus BX40 fluorescence microscope fitted with a Q-Imaging MicroPublisher 5.0 RTV digital camera using the green laser and normal transmitted light (Yang et al., 2016).

Data analysis

In this study, instantaneous growth rate of mean total length (G_l) was calculated following Ricker (1975) as: ${\displaystyle G_l=\left(ln{l}_2-ln{l}_1\right)∕3}$ where l₁ is the mean total length in millimeters of S. wangchiachii of at a given time point, l₂ is the corresponding mean total length of the same batch three months later, and 3 is the sampling interval of three months.

Instantaneous growth rate of mean wet mass (G_w) was calculated as: ${\displaystyle G_w=\left(ln{w}_2-ln{w}_1\right)∕3}$ where w₁ is the mean wet mass in grams of S. wangchiachii at a given time, w₂ is the corresponding mean wet mass of the same batch three months later, and 3 is the sampling interval of three months.

The percent (P_i) of marked individuals out of all marked S. wangchiachii for each recapture survey was calculated as: ${\displaystyle P_i=n_i∕N_t}$ where n_i is the number of marked fish at site i of the given recapture date and N_t is the total number of marked fish of the given recapture.

The condition factor of S. wangchiachii was calculated following Fulton (1904) as: ${\displaystyle \text{condition factor}=10^6\times w∕l^3}$ where w is the wet mass in grams of S. wangchiachii at a given time point and l is the total length in millimeters of S. wangchiachii at the same time.

ANOVA tests for total length, wet mass and condition factor between marked and unmarked S. wangchiachii were performed with SPSS 19.0 software. Bonferroni Tests were used for post hoc tests when the variances were equal, and alternatively, Games-Howell Tests were used when the variances were unequal. The significance level was set as P < 0.05.

Small-scale marking pilot study

One day after immersion was completed, no marked fish had died. Twenty-three days after marking, all sampled individuals showed a visible red-orange mark in their otoliths (Fig. 2A). Without polishing, visible marks also were easily identified in the otoliths of marked individuals sampled on 28 December 2014 (Fig. 2B), 4 May 2015 (Fig. 2C) and 24 January 2016 (Fig. 2D). More than one year after immersion, the ARS mark still remained visible, and there was no evidence that the mark was significantly decaying.

Large-scale marking application

In the large-scale immersion marking of juvenile S. wangchiachii, mortalities in each batch were very low (≤0.50%), and no significant difference in the mortality between marked and unmarked batches was detected (Table 1). Three months after being reared in outdoor fishponds, samples of marked and unmarked individuals showed no significant difference in total length, wet mass and condition factor (Table 1). A visible red-orange mark in the otoliths could be identified under green laser in all marked fish sampled.

Recapture and evaluation

Otolith checking confirmed that a total of 852 S. wangchiachii of the 2015 cohort were caught during the three recapture surveys. Of these fish, 262 individuals had a clear ARS mark in their otoliths, and 590 individuals had no mark (Fig. 3; Table 2). The percent of marked individuals in each recapture survey were 32.73% in October 2015, 26.36% in January 2016 and 30.53% in April 2016.

In the October 2015 recapture effort, total length, wet mass and condition factor of both marked and unmarked S. wangchiachii were significantly lower than those of the hatchery control group (Table 2; one-way-ANOVA with post hoc Games-Howell Tests, P < 0.001). The condition factor of marked individuals when they were recaptured was also significantly lower than when they were released (Table 2; independent t-test, P < 0.001). Between marked and unmarked individuals, there was no significant difference in wet mass (P = 0.134) and condition factor (P = 0.735), whereas a slight difference in total length was detected (P = 0.030). In addition, the total length, wet mass and condition factor differed significantly among samples caught from different recapture sites (Table 3; two-way ANOVA tests without the data for site 7, P < 0.001), but a significant difference was not observed between marked and unmarked individuals (P > 0.05). Recapture site and ARS mark had interaction effects on wet mass (P = 0.006).

In the January 2016 recapture effort, the total length and wet mass of both marked and unmarked S. wangchiachii were still significantly lower than those of the hatchery control group (Table 2; one-way-ANOVA with post hoc Games-Howell Tests, P < 0.001), but the difference in condition factor was not significant (P = 0.304). Between marked and unmarked individuals, no significant differences in total length (P = 0.828) and wet mass (P = 0.816) were detected. The total length, wet mass and condition factor of S. wangchiachii differed significantly among samples caught from different recapture sites (Table 3; two-way ANOVA tests without the data for sites 1 and 7, P < 0.001), but ARS mark and the interaction between it and recapture site did not have significant effects on the three indexes (P > 0.05).

In the April 2016 recapture effort, the hatchery control group was not sampled because of fishpond cleaning. There was no significant difference in total length, wet mass or condition factor between unmarked and marked S. wangchiachii (Table 2; independent t-test, P > 0.05). The total length, wet mass and condition factor of S. wangchiachii differed significantly among samples caught from different recapture sites (Table 3; two-way ANOVA tests, including only data for sites 3, 4 and 6, P < 0.001). A significant difference in condition factor (Table 3; P = 0.009) was observed between marked and unmarked individuals, but a significant difference was not observed for total length (P = 0.181) or wet mass (P = 0.07). The interaction effects of recapture site and ARS mark were not significant (P > 0.05).

The G_l and Gw of marked individuals tended to slowly increase after release. In the first trimester after release, the G_l and Gw values of marked individuals were 0.0566 and 0.1065, respectively, which were lower than those of the hatchery control group (0.1430 and 0.4964, respectively). In the second trimester, both G_l (0.0683) and Gw (0.2832) of marked individuals had increased slightly. At that time, the Gw of marked individuals was slightly higher than that of the hatchery control group (0.2569), whereas the G_l of marked individuals was still lower than that of the hatchery control group (0.1080). In the third trimester, the G_l of marked individuals had increased to 0.0878, whereas the Gw (0.2815) remained almost the same.

After being released, the hatchery-produced S. wangchiachii began to move away from the release area. In October 2015, marked individuals were recaptured at all recapture sites upstream and downstream of the release sites (Table 2). Site 2 was not suitable as nursery ground for S. wangchiachii, but P₂ (18.90%) nevertheless represented a high percent of total recaptures (Fig. 4). In the subsequent recaptures, catches at site 2 were very small because of unsuitable habitat (Table 2). In the three surveys, mean P₃ (28.69 ± 1.03%) and P₄ (31.62 ± 25.98%) were much higher than that of P₅ (11.49 ± 7.80%), P₆ (8.58 ± 4.38%) and P₇ (4.17 ± 3.23%). P_i significantly decreased with distances from the release sites. P_i for the distant sites 6 and 7 in the latter two recaptures increased slightly compared to that in the first recapture, but the values were still much lower than those of sites 3–5 (Fig. 4). This implies that stocked fish were mainly distributed over a 10–15 km long stretch around the release sites.

Feasibility of ARS mass marking

Previous studies of marking different fish species demonstrated that ARS treatment produces an excellent mark quality and has no significant harmful effects on the fish (Baer & Rösch, 2008; Bashey, 2004; Caraguel et al., 2015; Liu et al., 2009). However, faced with sustained pressure to produce enough fish seed to achieve the annual goals of release programs, managers of many hatcheries continue to worry that mass marking using the ARS method will cause high mortality, and this concern has a negative impact on the use of marking to evaluate fish stock enhancement. In this study, marking juvenile S. wangchiachii (mean total length 23.52 ± 1.50 mm) by immersion in 70 mg L⁻¹ ARS solution for 24 h did not cause death. In the following large-scale marking application, the mortality of five marked batches and three unmarked ones was negligible (≤0.50%), and no significant difference in mortality between marked and unmarked fish was detected. After rearing for three months in outdoor fishponds, no significant differences in total length, body mass, or condition factor between marked and unmarked groups were detected. Because juvenile S. wangchiachii experienced natural mortality, the extremely low mortalities that occurred during the immersion marking process might not have been due to ARS solution. In addition, immersion marking was carried out directly in the rearing tanks, which avoided manipulations of fish, reduced stress and costs.

The ARS mark that develops in the otolith remains highly readable for several years, whether fish are reared in the laboratory or in the field (Champigneulle & Cachera, 2003; Nagiec et al., 1995; Partridge et al., 2009; Poczyczyński et al., 2011). Because sunlight and turnover of skeletal calcium can cause fluorescent marks to fade, external fluorescent marks on scales and fin rays cannot be readily detected over time (Bashey, 2004; Elle, Koenig & Meyer, 2010). In contrast, otoliths are protected by the skull and previously deposited otolith materials are not resorbed (Campana & Neilson, 1985), which prolongs the lifetime of the mark. In the 2014 marking effort, S. wangchiachii marked with 70 mg L⁻¹ ARS retained highly visible marks on otoliths after rearing for about 20 months in an indoor tank, and they did not present clear signs of significant fading. In the mass marking of 2015, fish marked with 30–50 mg L⁻¹ ARS were transferred to outdoor fishponds and reared for about three months. Jinping Hatchery is located in the arid river valley region of the western Sichuan Province, where sunshine is very strong all year long (Yuan, Li & Lin, 2013). Nevertheless, ARS marks of fish sampled from each marking batch were highly visible. All of the ARS marks on the otoliths of recaptured fish were as clear as they had been at the time of release. However, because otoliths continuously grow and thicken, over time the mark can be covered by otolith materials, and marks can become faint and difficult to detect unless exposed by grinding and polishing the otoliths (Baer & Rösch, 2008; Sánchez-Lamadrid, 2001; Taylor, Fielder & Suthers, 2005). In this study, although it was not experimentally tested, the final retention time of the ARS mark in the otoliths of S. wangchiachii was long enough to monitor the released individuals to evaluate stocking effectiveness.

Effectiveness of stocking enhancement

After release, trimonthly recapture surveys confirmed that some of the stocked S. wangchiachii had survived. Assuming that the percent of marked (47.62%) and unmarked (52.38%) S. wangchiachii remained unchanged in the stocked cohort, the percent of catches that originated from stock enhancement were estimated to be 68.73% in October 2015, 55.35% in January 2016 and 64.11% in April 2016, respectively. This demonstrated that stocked S. wangchiachii constituted an important part of the young fish with a mean level of 62.73% recruitment of the 2015 cohort. In previous successful stock enhancements, such as those for vendace Coregonus albula (Poczyczyński et al., 2011), brown trout Salmo trutta (Caudron & Champigneulle, 2009), Japanese Spanish mackerel Scomberomorus niphonius (Obata et al., 2008), and red sea bream Pagrus major and Japanese flounder Paralichthys olivaceus (Kitada & Kishino, 2006), hatchery-produced fish contributed considerably to population recruitment, and stocking successes were often attributable to appropriate release sizes and environmental conditions at release sites. In this study, the relatively high percent of released S. wangchiachii might be explained by the fact that young fish for release were fully covered with scales, which would protect the skin against mechanical injury and bacteria and parasites (Yan et al., 2014). In addition, the negligible fishing pressure on young S. wangchiachii under age two and few predators, such as Percocypris pingi and Silurus asotus, might have had positive effects.

However, the comparative analysis of recapture data showed that total length, wet mass, and condition factor of recaptured S. wangchiachii differed significantly among different recapture sites (Table 3). The body sizes of fish recaptured at release sites were often smaller than those at other sites for both marked and unmarked fish (Table 2). This finding suggests that fish at release sites did not grow as well as fish at other sites. Pebbly nursery grounds in shallow waters are essential for Schizothorax fish at early life stages. In the Jinping area of the Yalong River, water flow sharply decreases (at a maximum percent of about 95%) due to the upstream dam of Jinping Dam II, which leads to marked physical habitat degradation (Wang et al., 2007). Excavation of sand in the river, which take place frequently at five sites along the 60 km long survey area, further destroyed the habitats. This reduction of essential habitats could have a significant negative effect on the river’s carrying capacity for Schizothorax fishes. It is likely that the released fish moved very slowly so that nine months after release most of them still were distributed over a 10–15 km long stretch near the release sites, although a few marked fish were caught about 50 km downstream three months after release. The relatively slow migration speed would maintain a high density of fish in the release area, which would result in both released and wild fish having to compete intensely with each other for resources.

Wild fish can be replaced with hatchery-reared fish when the latter are released in numbers that exceed the carrying capacity, but it is difficult to verify the extent to which they replace the wild ones (Kitada & Kishino, 2006). The surveys conducted in this study showed that, there are some spawning grounds for Schizothorax fishes in the Jinping area of the Yalong River, where some naturally born juveniles were caught in April 2014. This study showed that natural recruitment still accounted for a sizeable percent (approximate 40%) of S. wangchiachii recruitment of the 2015 cohort. Additionally, the non-significant difference in total length, wet mass and condition factor between the marked and unmarked fish indicated that ARS marking did not have significant harmful effects. This finding suggests that the stocked S. wangchiachii should grow as the naturally born fish in the Jinping area of the Yalong River. Releases at the scale used in this study might have not exceeded the carrying capacity of release sites or replaced wild fish, but repeated large annual releases might do so. To maintain the fish population at sustainable levels, new release sites around site 7 should be added (Fig. 1). Moreover, sand excavation in the river should be stopped immediately, and nursery habitats must be restored to expand the carrying capacity of the river.

Wild S. wangchiachii live in rapid flowing river water and consume adherent alga (Bacillariophyta) using the sharp outer horny sheath on their lower jaw to scrape it off the substrate. In contrast, the hatchery-reared fish are reared in still water ponds and fed on commercial diets. It is well known that environmental differences between hatchery-reared and wild fish can influence their behavior, especially foraging behavior and avoidance of predators, which may subsequently affect post-release success (Hervas et al., 2010; Le Vay et al., 2007; Johnsson, Brockmark & Näslund, 2014). On first release into the wild, hatchery-reared fish must not only avoid predators but also adapt to a new food supply (Blaxter, 2000). In this study, condition factors of the stocked S. wangchiachii recaptured three months after release were significantly lower than those of the hatchery control group as well as their own cohort at the time of release (Table 2). This suggests that these hatchery-produced fish might need to be acclimatized to the wild habitat, as their growth was negatively affected by the change of environmental conditions. Perhaps, hatchery-reared fish suffered a high level of short-term post-release mortality during the first trimester after release. However, it was difficult to precisely estimate this mortality due to lack of historical data on fisheries and prior investigation in the area. In the subsequent recaptures, the condition factor of stocked fish had returned to a level as good as that of the hatchery control group. In addition, the G_l and Gw of marked fish displayed a slowly increasing trend. Therefore, survival of hatchery-produced S. wangchiachii suggests that they gradually adapted to the wild habitat, and they exhibited favorable growth six months after release.