Introduction
Fisheries depletion has become a global problem and aroused attention for decades (Hutchings et al. 2000; Worm et al. 2016). Of 4,714 fisheries assessed in 2012 around the world, 68% have slipped below the biomass target that supports maximum sustainable yield (Worm et al. 2016), and some rivers and lakes even have experienced many local fishery collapses (Mcintyre et al. 2016). While invasive species and overfishing are usually proposed to be contributors, early life stage mortality of fish is also a factor, due to its effect on population maintenance. For example, mass hatchery stocking failed to re-establish self-sustaining lake trout populations in the Great Lakes, USA, partly due to the fact that 50% of offspring died at an early life-stage (Smith et al. 1994; Palace et al. 1998).
Several man-made chemicals, mainly organochlorines, have been associated with larval and adult mortality in wild fish (Smith et al. 1994; Palace et al. 1998; Hamilton et al. 2016; Sepúlveda et al. 2004). Notably, fish undergo a series of rapid cell divisions and differentiations during embryonic development, including large-scale de novotranscription, and therefore fish embryos are much more sensitive to environmental pollutants than larval and adult fish (Mohammed et al. 2013; Cherr et al. 2017). Thus, organochlorines could cause embryonic mortality in wild fish and, over time, compromise natural population recruitment and ultimately induce fishery collapse. However, there is sparse evidence that widespread chemicals are lethal to embryos of wild fish since collecting fish embryo and tracking subsequent development are big challenges in aquatic ecosystems. Crucian carp (Carassius carassius ) is ubiquitous in the lakes, rivers, and reservoirs, and therefore has been usually used as a bio-indicator of fish resources (Zheng et al. 2015). During the reproductive season, crucian carp spawns adhesive eggs over substrates such as shallow-shore aquatic macrophyte, where eggs remain until hatching (Shin-ichiro et al. 2011), facilitating large-sample-size egg collection and observation. Such reproduction feature of crucian carp give us a chance to observe whether a pollutant could cause embryo mortality.
Tris(1,3-dichloro-2-propyl) phosphate (TDCIPP), a chloro-organophosphorus flame retardant, is produced industrially in high volumes. TDCIPP has been reported to cause embryo mortality in a zebrafish embryo exposure ex vivo (Volz et al. 2016; Kupsco et al. 2017). Therefore, in the present study, we utilized reproductive characterization of crucian carp in natural population to address following questions. (i) Does mass mortality of crucian carp embryo occur in aqueous ecosystem? (ii) Is the embryonic mortality observed in field associated with embryonic TDCIPP concentration; (iii) Can TDCIPP at environmental levels cause the mass mortality of crucian carp embryos under controlled laboratory conditions? We performed mortality investigation and causal analyses by clarifying the temporal and spatial distributions of embryonic mortality and TDCIPP concentrations in crucian carp over three years in a lake ecosystem, and laboratory exposure experiments were also conducted to determine whether environmental concentrations of TDCIPP could cause embryonic mortality in crucian carp from natural population.
Materials and Methods
Embryo Collection in Taihu Lake. In 2016 and 2017, embryos of crucian carp were collected from water grass in SZ, GH, MlL and ZS (see Table S1 for location information) during the crucian carp reproduction season (April to May), with the permission of the fishery administration. Fingerling identification was based on the external characteristics and DNA barcode analysis, as described in Supporting Information (SI). After collection, embryos were transported to laboratory immediately with incubators at 18 ± 2℃, at which temperature crucian carp embryos developed slowly (Laurila et al. 1987), Then, the embryos were cultured until hatching in active carbon-treated drinking water (1 L in 1.5 L-glass dishes; each dish contained approximately 500 embryos) at 26 ± 1°C and under a 16 h: 8 h (light: dark) light cycle. The transport and culture conditions were identical for all sampling sites. In 2018, embryos were collected from GH and ZS by constructing artificial fish nests for locally tracking the developmental progress of crucian carp embryo. Artificial nest was made by fishing net by cutting into pieces and tied together to increase the area for crucian carp to spawn. Artificial nests were put into shallow water nearshore in GH and ZS according to a time schedule (Fig. 2 A and B), and five nests were used in each site.
Embryos from four sites in Taihu Lake were collected for TDCIPP analyzing prior to culture (approximately 50 mg of embryos per sample; n = 6). The dead embryos and contemporaneous live embryos during culture were also sampled to determine TDCIPP concentrations [approximately 50 mg of embryos per sample; n = 6 in 2017; n = 3 in 2018 (due to limited numbers of dead embryos)].
Exposure Experiment. One- to three-year old wild adult crucian carp including females and males, aged by measurement of dorsal spines and pectoral fin rays according to a previous study (Vilizzi et al. 2018), were captured from a natural population in October 2017, and then were exposed in a laboratory setting to TDCIPP for six months (October 2017–April 2018) at nominal concentrations of 8, 40, and 200 μg/L. Exposure was conducted in 420-L glass tanks (30 fishes per tank, with each group containing three replicates) under a natural temperature and light cycle (to imitate the natural maturation of ovaries). The fish were fed on fish premix twice daily until the end of exposure. A flow-through system (Fig. S3) was used to renew the water daily, and water samples were collected monthly (n = 6) for TDCIPP analysis during the exposure period. After six months of exposure, embryos were obtained through artificial fertilization; a portion of these were sampled (approximately 50 mg each sample, n = 6) for TDCIPP analysis and a portion were prepared for culture to enable study of embryonic development. Culture conditions of embryo and method for cell height quantification were according to a paper reported previously (Kupsco et al. 2017), and more details were provided in SI.
In April 2018, embryos were obtained from the adult crucian carp cultured in the control group in the exposure experiment as described above, and the larvae were exposed to TDCIPP from half-month-old (beginning eating) to sex maturation (June 2019) at nominal concentrations of 50, 500, 5000, and 50000 ng/L. Exposure was conducted in 420-L glass tanks (50 fishes per tank, each group containing three replicates) equipped with a flow-through system. During the exposure, water samples were collected quarterly (n = 5) for TDCIPP analysis. Fish were fed on live brine shrimp (Artemia nauplii ) twice daily until fish were one-month-old, and then fed on fish premix twice daily until the end of exposure. After a one-year exposure, embryos were obtained through artificial fertilization, and samples were collected and developmental processes were observed according to the same method described above. After sample preparation, the TDCIPP concentrations in water and embryos were determined by liquid chromatography-tandem mass spectrometry (LC-MS/MS) analysis.
Chemical Analysis. All of the samples were collected in glass vials, with all vials having been pretreated in a muffle furnace at 450°C for 6 h to prevent contamination. Samples were frozen immediately at -20°C after collection, and stored at this temperature until analysis.
Methods for sample preparation and TDCIPP quantification in water and embryo samples were according to our previous study with some modification (Zhao et al. 2018), and the details were provided in SI. TDCIPP-d15 was used as the isotopic internal standard to correct the loss of TDCIPP during the sample preparation and to compensate for variations in LC-MS/MS instrumental response from injection to injection. Recoveries of TDCIPP in the water and embryo samples were 73.4 ± 1.4% and 94.1 ± 1.7%, respectively. TDCIPP was detected in the procedural blanks at concentrations of 0.1 ± 0.03 ng/L in water samples and 0.9 ± 0.14 ng/g lw in embryo samples. The limits of quantitation (LOQs) were defined as 10 times the standard deviation of the procedural blanks, and thus the LOQ was 0.3 ng/L for water samples and 1.4 ng/g lw for embryo samples. The final concentrations reported in this study were blank-subtracted, and all of the chemical analysis experiments were performed in triplicate.
Statistical Analyses. Differences in mortality and developmentally delayed rates of embryos from different sites in Taihu Lake were analyzed by the χ2 test. Concentrations of TDCIPP, mortality of embryos from the exposure experiment, and relative embryo height were analyzed by one-way analysis of variance (ANOVA) with Dunnett’s test (95% confidence interval). When the concentrations were below the LOQ, the values of the LOQs divided by 2 were used for calculation and statistical analysis. Differences were analyzed with SPSS (v22.0, IBM, Armonk, NY) and considered statistically significant at p< 0.05.