Toxic Effects of Heavy Metals and Organic Polycyclic Aromatic Hydrocarbons in Sediment Porewater on the Amphipod Hyalella azteca and Zebraﬁsh Brachydanio rerio Embryos from Different Rivers in Taiwan

: The amphipod ( Hyalella azteca ) and zebraﬁsh ( Brachydanio rerio ) embryos were used for toxicological sediment porewater testing. Porewaters from 35 sampling stations of eight streams in southern Taiwan were screened for toxic effects and their relationship with 6 metal and 16 polycyclic aromatic hydrocarbons (PAHs). Concentration analysis results showed that the following PAHs, naphththalene, benzo(b)ﬂuoranthene, dibenz(a,h)anthracene, acenaphthalene, and the heavy metal cadmium were not detected in 35 sampling stations. The highest detection rate of 94.1% was caused by the PAHs ﬂuoranthene and pyrene. The highest detection rate of the metal zinc was 88.6% of 35 analyzed samples. The majority of samples (88%) were classiﬁed as level tier 1 according to USEPA national sediment inventory. This indicates the probability of adverse effects on aquatic life or human health. The results of a zebraﬁsh embryo test showed that heart rate and survival were signiﬁcantly reduced with all porewater samples. Therefore, ﬁsh exposed to contaminated river conditions may be affected in their cardiovascular functions. Looking at correlations between toxic effects of metals and PAHs, we found that phenanthrene, anthracene, pyrene, benzo(a)anthracene, chrysene, benzo(b)ﬂuoranthene, and benzo(a)pyrene were low, while ﬂuorene was highly correlated with toxic effects of metals.


Introduction
Sediments are deposits on the bottom of a water body and are naturally composed of sand, clay, soil, organic matter, and other minerals [1,2]. Sediments are also a sink, reservoir, and source of pollutants that harm natural water bodies and aquatic organisms [3][4][5]. Various chemical substances accumulate in sediments, which proves toxic effects on aquatic habitats and ecosystems [6]. Sediment porewater is defined as the water occupying the space between sediment particles which constantly remains in contact with sediments. Therefore, pollutants may be exchanged between sediment and porewater through dynamic equilibrium distribution [7]. Sediment particles are also where the benthic burrowing each sampling site, centrifuged after collection to obtain porewater, and stored in a refrigerator at 4 °C for subsequent biological and chemical experiments.

Sediment Porewater Toxicity Testing with the Amphipod Hyalella azteca
Hyalella azteca is an epibenthic and phytal amphipod which plays a vital role in ecosystem resilience due to its detritus-feeding and potential high abundance in nature. It is widely distributed in the freshwaters of North America, South America, and the Caribbean. Algae, epiphytes, and sediment particles are the main food sources of this amphipod. It is about 3-8 mm long (males are larger than females) and considered to be sensitive to aquatic pollutants [38]. It is known to tolerate slight salinity changes, but not to survive pH conditions of less than 6.0.
Centrifuged porewater (50 mL) was obtained and kept at 4 °C until analysis. Water parameters such as pH, conductivity, temperature, and dissolved oxygen were measured before and after this process at room temperature. Specimens of H. azteca, aged between 7 and 10 days, were selected to ensure that the test organisms were similar. H. azteca in reconstituted water (96 mg/L NaHCO3, 30 mg/L MgSO4, 4 mg/L KCl, 50 mg/L CaSO4, and 50 mg/L CaCl2 were dissolved in deionized water and aerated vigorously for at least 24 h) was used as controls. Operating conditions included a photoperiod of 16 L:8 D, temperature 23 ± 1 °C, and 20 mL reconstituted water/sample volume. Samples were tested in duplicates of 10 test organisms per beaker and the controls were reconstituted waters following a standard method. Ten individuals were placed in a beaker for 48 h exposure, and their survival rates were compared. Differential survival with samples was then compared to the control (80%) to estimate the level of toxicological risk.

Sediment Porewater Toxicity Testing with the Amphipod Hyalella azteca
Hyalella azteca is an epibenthic and phytal amphipod which plays a vital role in ecosystem resilience due to its detritus-feeding and potential high abundance in nature. It is widely distributed in the freshwaters of North America, South America, and the Caribbean. Algae, epiphytes, and sediment particles are the main food sources of this amphipod. It is about 3-8 mm long (males are larger than females) and considered to be sensitive to aquatic pollutants [38]. It is known to tolerate slight salinity changes, but not to survive pH conditions of less than 6.0.
Centrifuged porewater (50 mL) was obtained and kept at 4 • C until analysis. Water parameters such as pH, conductivity, temperature, and dissolved oxygen were measured before and after this process at room temperature. Specimens of H. azteca, aged between 7 and 10 days, were selected to ensure that the test organisms were similar. H. azteca in reconstituted water (96 mg/L NaHCO 3 , 30 mg/L MgSO 4 , 4 mg/L KCl, 50 mg/L CaSO 4 , and 50 mg/L CaCl 2 were dissolved in deionized water and aerated vigorously for at least 24 h) was used as controls. Operating conditions included a photoperiod of 16 L:8 D, temperature 23 ± 1 • C, and 20 mL reconstituted water/sample volume. Samples were tested in duplicates of 10 test organisms per beaker and the controls were reconstituted waters following a standard method. Ten individuals were placed in a beaker for 48 h exposure, and their survival rates were compared. Differential survival with samples was then compared to the control (80%) to estimate the level of toxicological risk.
In this study, after adjusting the results to the control group, samples with a relative survival rate of <75% were defined as being toxic to organisms following US EPA suggestions [39,40]. The relative survival rate was obtained from the ratio of the observed It was suggested that if the adjusted survival rate of organisms was less than 75%, this would cause adverse effects on aquatic organisms and human health. If the survival rate was between 75% and 90% for aquatic organisms, human health could be harmed, and if aquatic organisms showed a survival rate >90%, human health was certainly affected [40].

Sediment Porewater Toxicity Testing with Zebrafish Embryos
Mature wild zebrafish (AB strain wild type) were obtained from the Taiwan Zebrafish Core Facility at Academia Sinica (TZCAS). The experimental protocol was approved (approval no. NPUST-102-041) by the Institutional Animal Care and Use Committee (IACUC) of NPUST. Briefly, the zebrafish were raised at 28 • C in a circulating water tank. The photoperiod was set at 14/10 h light/dark cycle and mature wild zebrafish were fed with commercial ornamental fish food twice a day. The fertilized eggs obtained from the free spawning zebrafish were used for the assessment of developmental toxicity of the 35 porewater samples.
The one-cell stage of zebrafish fertilized eggs was evaluated for the toxicity of porewater and the experiment was carried out in 12-well plates. Thirty fertilized eggs of zebrafish were placed in each well containing 2 mL of aerated porewater samples. The porewater was refilled daily to record the survival rate, hatching rate, and deformity rate of zebrafish and the deformity pattern was observed using a dissecting microscope (Leica Z16-APO, Leica Microsystems Inc., Wetzlar, Germany). Fertilized eggs in aerated water (5.03 mM NaCl, 0.33 mM MgSO 4 ·7H 2 O, 0.17 mM KCl, and 0.33 mM CaCl 2 ·2H 2 O) were dissolved in deionized water, aerated vigorously for at least 8 h, and used as controls. Controls and all treatment groups were made in triplicate. The heartbeat rate of 6 hatched zebrafish per replicated treatment was evaluated on the third day as recorded by microscopic observation every minute. The experimental results were analyzed statistically by t-test and p < 0.05 was regarded as significantly different.

Analysis of Polycyclic Aromatic Hydrocarbons (PAHs)
The porewater samples were filtered through a 0.22 µm filter membrane (syringe filter, diameter 13 mm, PVDF) which was followed by solid phase extraction (LiChrolut RP18 ® 500 mg, 6 mL Merck, Darmstadt, Germany). Related to the concentration of the solvent, the PAHs concentration was measured by Ultra performance liquid chromatography (UPLC) (Waters, Milford, MA, USA) coupled with a photodiode array detector (PDA) and a fluorescence detector for analysis and quantification. The separation was carried out on a UPLC ® BEH Shield RP18 (2.1 × 150 mm, 1.7 µm) column, with ultra-pure water and acetonitrile as the mobile phase for gradient elution. For acenaphthylene, the UV detector wavelength was set at 228 nm. The elution conditions to separate the target compounds by fluorescence detection, and method detection limits (MDL) are shown as supplementary material (Tables S1-S3). The linear correlation coefficient (R) for all compounds was >0.995.

Analysis of Trace Metals
Sample pretreatment and analysis were followed by NIEA standard method (NIEA M353.01C). Porewater samples were filtered through cellulose acetate filters (pore diameter, 0.45 µm). The filtrate was acidified with 1% trace-metal grade nitric acid as a pre-treatment. The extracted portion was detected using an inductively coupled plasma optical emission spectrometer (ICP-OES, Optima 2100 Perkin-Elmer, Waltham, WA, USA) for Cr, Ni, Cu, Zn, Pb, and Cd analysis.

Calculations of Porewater Toxicity by Interstitial Water Benchmark Units (IWBUs)/Interstitial Water Toxic Units (IWTU)
The pollutant distribution between the sediment and the porewater is expected to cause different toxic effects. It is not proper to restrict toxicity evaluation to either result. Therefore, interstitial water benchmark units (IWBUs) can be used to predict the toxic effects of various metals on aquatic organisms. These IWBU values are calculated based on the porewater concentration and the final chronic value (FCV) as proposed by Di Toro et al. in 2005 (Equation (2), Table 1). For PAHs, the dissolved phase concentrations (C d ) of each nonionic organic contaminant are divided by its corresponding FCV to derive interstitial water toxic units (IWTUs) (Equation (3)).  In sediments in which IWBUs or IWTU is >1.0, benthic organisms are not protected and adverse effects may occur. Conversely, if the IWBUs/IWTU is ≤1.0, sediment toxicity due to mixing of contaminants is unlikely.

Results and Discussion
Four water parameters were routinely monitored during the toxicity testing procedures, including temperature, pH, dissolved oxygen, and conductivity. Since the test was commonly performed under controlled conditions, at constant temperature, water temperature was commonly the least varying factor. The pH, dissolved oxygen, and conductivity are shown in Table 3, before and after exposure. Houjin river-HJ, Donggang river-DG, Yanyan river-YS, Sanye river-SY, Dianbao river-DB, Agongdian river-AGD, Wuluo river-WR, and Niuchou river-LK.

Water Quality Monitoring of Porewater Samples
The test results showed that the pH values of all porewater samples were between 7.62 and 8.68, which approximated a pH ranging from 6.5 to 8.5. The dissolved oxygen values of 35 sampling sites ranged from 1.04 to 5.44 mg/L. Vaquer-Sunyer and Duarte [42] stated that most crustaceans begin to show elevated mortality when the dissolved oxygen reaches below 2.45 ± 0.14 mg/L, whereas dissolved oxygen levels in the rivers of HJ and DG in Appl. Sci. 2021, 11, 8021 7 of 17 this study were both lower than 2.45 mg/L and thus were worthy of further discussion of the physiological effects of dissolved oxygen. Conductivity increased due to an increase in salinity. Sampling sites HJ7 and DG1 were closer to the ocean, and the conductivity was relatively higher here than at the inland sites. In addition, site YS crossed the industrial area, and sites such as DB and SY were adjacent to the factory area. Therefore, discharged sewage led to relatively higher electrical conductivity.

Survival Rate of Hyalella azteca Exposed to Whole Sediment Porewater Samples
In this study, 88% of the 35 sample stations had adverse effects on aquatic organisms and with this on human health. The survival of the remaining 12% sites was between 75% and 90%, showing that the sediment pollutants of eight streams in the southern region had a great variation. The results of 35 survival rates of porewater exposure are shown in Figure 2, demonstrating different potential ecological risks.
Appl. Sci. 2021, 11, x FOR PEER REVIEW 7 of 19 DG in this study were both lower than 2.45 mg/L and thus were worthy of further discussion of the physiological effects of dissolved oxygen. Conductivity increased due to an increase in salinity. Sampling sites HJ7 and DG1 were closer to the ocean, and the conductivity was relatively higher here than at the inland sites. In addition, site YS crossed the industrial area, and sites such as DB and SY were adjacent to the factory area. Therefore, discharged sewage led to relatively higher electrical conductivity.

Survival Rate of Hyalella azteca Exposed to Whole Sediment Porewater Samples
In this study, 88% of the 35 sample stations had adverse effects on aquatic organisms and with this on human health. The survival of the remaining 12% sites was between 75% and 90%, showing that the sediment pollutants of eight streams in the southern region had a great variation. The results of 35 survival rates of porewater exposure are shown in Figure 2, demonstrating different potential ecological risks.   [39,43]. Higher amphipod survival rates (>80%) were found at the following sites: DB3, DB5, AGD1, HJ1, HJ4, DG1, DG2, and DG4. All amphipods died in porewaters from sites DB1, DB2, SY1, SY2, SY3, HJ7, LK1, LK3, and WR. This indicates that the Sanye River (sites designated as SY) and Niuchou River (designated as LK) showed acute toxic effects on benthic invertebrates.
It was suggested that if the adjusted survival rate of organisms was less than 75% this would cause adverse effects on aquatic organisms and human health. If the survival rate was between 75% and 90% for aquatic organisms, human health could be harmed and if aquatic organisms showed a survival rate > 90%, human health was certainly affected [40].

Toxic Effects of Sediment Porewater on Zebrafish Embryos
In this research, the effects of 35 sediment porewater samples on the development of zebrafish in different regions of Taiwan were evaluated after four days of exposure. The results of mortality, hatchability, and deformity rate on zebrafish embryos were recorded daily to evaluate the toxic effects of the sediment porewater as shown in  Embryos of nearly all fertilized eggs hatched on the third day of the general culture and the hatchability reached more than 98% in cases. In some instances, the control group could also show abnormal development due to physical differences between the organisms. However, the deformity ratio of the control group was low at about 1%. The samples tested were based on the results of mortality and nine sediment porewater samples, HJ7, DB 1~5, and SY 1~3 were classified as being acutely toxic where all zebrafish embryos died after one day of immersion. The toxicity control YS river SY river AGD river DB river HJ river LK river WR river DG river Survival rate (%)   [39,43]. Higher amphipod survival rates (>80%) were found at the following sites: DB3, DB5, AGD1, HJ1, HJ4, DG1, DG2, and DG4. All amphipods died in porewaters from sites DB1, DB2, SY1, SY2, SY3, HJ7, LK1, LK3, and WR. This indicates that the Sanye River (sites designated as SY) and Niuchou River (designated as LK) showed acute toxic effects on benthic invertebrates.
It was suggested that if the adjusted survival rate of organisms was less than 75% this would cause adverse effects on aquatic organisms and human health. If the survival rate was between 75% and 90% for aquatic organisms, human health could be harmed and if aquatic organisms showed a survival rate > 90%, human health was certainly affected [40].

Toxic Effects of Sediment Porewater on Zebrafish Embryos
In this research, the effects of 35 sediment porewater samples on the development of zebrafish in different regions of Taiwan were evaluated after four days of exposure. The results of mortality, hatchability, and deformity rate on zebrafish embryos were recorded daily to evaluate the toxic effects of the sediment porewater as shown in  Embryos of nearly all fertilized eggs hatched on the third day of the general culture and the hatchability reached more than 98% in cases. In some instances, the control group could also show abnormal development due to physical differences between the organisms. However, the deformity ratio of the control group was low at about 1%. The samples tested were based on the results of mortality and nine sediment porewater samples, HJ7, DB 1~5, and SY 1~3 were classified as being acutely toxic where all zebrafish embryos died after one day of immersion. The toxicity of four samples from YS4, WR1, WR 3, and WR 6 showed the second highest toxicity, which could also be classified as the highest toxicity due to the fact that all zebrafish embryos were dead after a two-day exposure. When zebrafish embryos were exposed to porewater samples from HJ3, DG1, LK1, LK2, HJ2, LK3, YS1~3, WR2, WR 4, and WR 5, either all of them died or the survival rate was lower than 20% on the third day. We classified such porewater samples as moderately toxic. The other 10 porewater samples (HJ1, DG3, HJ5, DG4, YS5, HJ6, AGD1, AGD2, HJ4, and DG2) were classified as less toxic. The survival rate here was still roughly higher than 15% on the fourth day of exposure. The embryonic development of zebrafish lasted for about 2 days and the main organs of zebrafish were developing thereafter until larval fishes hatched from their eggs. The development of zebrafish was seriously affected by pollutants contained in the porewater samples, causing incomplete development of organs in the fish embryos as shown in  In terms of deformity, a high deformity ratio was discovered in the moderate toxicity samples HJ3 and YS3. At site HJ3, 78% of the samples showed deformation while the survival rate was still below 69% on the second day. Recent research found that there were up to 85% deformities in the surviving individuals of the porewater sample from YS3. The remaining less toxic samples showed that mortality and deformity rates increased at longer treatment periods. Moreover, the hatching rate was low and showed delays until hatching. Although less toxic samples were classified with low toxicity, they still showed a significant toxic effect compared to the control group. of four samples from YS4, WR1, WR 3, and WR 6 showed the second highest toxicity, which could also be classified as the highest toxicity due to the fact that all zebrafish embryos were dead after a two-day exposure. When zebrafish embryos were exposed to porewater samples from HJ3, DG1, LK1, LK2, HJ2, LK3, YS1~3, WR2, WR 4, and WR 5, either all of them died or the survival rate was lower than 20% on the third day. We classified such porewater samples as moderately toxic. The other 10 porewater samples (HJ1, DG3, HJ5, DG4, YS5, HJ6, AGD1, AGD2, HJ4, and DG2) were classified as less toxic. The survival rate here was still roughly higher than 15% on the fourth day of exposure. The embryonic development of zebrafish lasted for about 2 days and the main organs of zebrafish were developing thereafter until larval fishes hatched from their eggs. The development of zebrafish was seriously affected by pollutants contained in the porewater samples, causing incomplete development of organs in the fish embryos as shown in Figures 3-5. In terms of deformity, a high deformity ratio was discovered in the moderate toxicity samples HJ3 and YS3. At site HJ3, 78% of the samples showed deformation while the survival rate was still below 69% on the second day. Recent research found that there were up to 85% deformities in the surviving individuals of the porewater sample from YS3. The remaining less toxic samples showed that mortality and deformity rates increased at longer treatment periods. Moreover, the hatching rate was low and showed delays until hatching. Although less toxic samples were classified with low toxicity, they still showed a significant toxic effect compared to the control group.     of four samples from YS4, WR1, WR 3, and WR 6 showed the second highest toxicity, which could also be classified as the highest toxicity due to the fact that all zebrafish embryos were dead after a two-day exposure. When zebrafish embryos were exposed to porewater samples from HJ3, DG1, LK1, LK2, HJ2, LK3, YS1~3, WR2, WR 4, and WR 5, either all of them died or the survival rate was lower than 20% on the third day. We classified such porewater samples as moderately toxic. The other 10 porewater samples (HJ1, DG3, HJ5, DG4, YS5, HJ6, AGD1, AGD2, HJ4, and DG2) were classified as less toxic. The survival rate here was still roughly higher than 15% on the fourth day of exposure. The embryonic development of zebrafish lasted for about 2 days and the main organs of zebrafish were developing thereafter until larval fishes hatched from their eggs. The development of zebrafish was seriously affected by pollutants contained in the porewater samples, causing incomplete development of organs in the fish embryos as shown in Figures 3-5. In terms of deformity, a high deformity ratio was discovered in the moderate toxicity samples HJ3 and YS3. At site HJ3, 78% of the samples showed deformation while the survival rate was still below 69% on the second day. Recent research found that there were up to 85% deformities in the surviving individuals of the porewater sample from YS3. The remaining less toxic samples showed that mortality and deformity rates increased at longer treatment periods. Moreover, the hatching rate was low and showed delays until hatching. Although less toxic samples were classified with low toxicity, they still showed a significant toxic effect compared to the control group.

Toxic Effects on Embryonic Cardiac Function
The cardiovascular system is the first organ system to form during embryonic development. The circulatory system is also a key factor for embryonic survival. In acute embryonic toxicity testing, the heart rate of hatching zebra fry was determined after the third day. The variations of embryonic heartbeat of zebrafish are shown in Figure 6. The porewater samples DG1, LK2, YS1, and YS2 caused a significant decrease in heartbeat compared to the control group. The average heartbeat rate of zebrafish in the control group was about 185 times per minute (taken as 100%), while the heartbeat rates were dropping to 133 and 115 per minute in toxic samples from stations YS1 and YS2. There show the most significant effects with heartbeat dropping down to 71% and 62%. The heartbeat rate of moderately toxic samples (HJ3, LK1, HJ2, LK3, YS3, WR2, WR4, and WR5), cannot be measured due to insufficient fish larvae which were dead or deformed on the third day. The effect of low toxicity (survival ratio was still roughly higher than 15% on the fourth day of exposure) samples (HJ1, HJ4, DG2, DG3, DG4, HJ5, HJ6, YS6, AGD1, and AGD2) on heart rate was evaluated and the results are shown in Figure 6. Low toxicity samples also caused a significant decrease in heartbeat rate. Compared to the control group, the degree of decline was between 80% (AGD1) to 93% (DG4). Overall, the decreasing tendency of the heartbeat rate was less in moderately toxic samples YS1 and YS2. The results showed that both medium toxic and low toxic samples can significantly

Toxic Effects on Embryonic Cardiac Function
The cardiovascular system is the first organ system to form during embryonic development. The circulatory system is also a key factor for embryonic survival. In acute embryonic toxicity testing, the heart rate of hatching zebra fry was determined after the third day. The variations of embryonic heartbeat of zebrafish are shown in Figure 6. The porewater samples DG1, LK2, YS1, and YS2 caused a significant decrease in heartbeat compared to the control group. The average heartbeat rate of zebrafish in the control group was about 185 times per minute (taken as 100%), while the heartbeat rates were dropping to 133 and 115 per minute in toxic samples from stations YS1 and YS2. There show the most significant effects with heartbeat dropping down to 71% and 62%. The heartbeat rate of moderately toxic samples (HJ3, LK1, HJ2, LK3, YS3, WR2, WR4, and WR5), cannot be measured due to insufficient fish larvae which were dead or deformed on the third day. The effect of low toxicity (survival ratio was still roughly higher than 15% on the fourth day of exposure) samples (HJ1, HJ4, DG2, DG3, DG4, HJ5, HJ6, YS6, AGD1, and AGD2) on heart rate was evaluated and the results are shown in Figure 6. Low toxicity samples also caused a significant decrease in heartbeat rate. Compared to the control group, the degree of decline was between 80% (AGD1) to 93% (DG4). Overall, the decreasing tendency of the heartbeat rate was less in moderately toxic samples YS1 and YS2. The results showed that both medium toxic and low toxic samples can significantly cause a decrease in heartbeat rate, which indicates that the heart beat and circulatory function of fish generally deteriorated when exposed to sediment leachates of the two rivers. cause a decrease in heartbeat rate, which indicates that the heart beat and circulatory function of fish generally deteriorated when exposed to sediment leachates of the two rivers. Figure 6. Effect of pollutants from sediment porewater on the heartbeat rate of zebrafish embryos. Hatching rate of zebrafish embryos exposed for 72 h at sampling stations with slightly toxic and moderately toxic porewaters. Error bars represent the standard deviation of the six zebrafish embryos. * p < 0.05 indicates significant differences between the control and exposure groups.

Comparative Results of PAH Concentrations
For chemical analysis, naphthalene, benzo(b)fluoranthene, dibenz(a,h)anthracene, and acenaphthalene were not detected in any of the porewater samples. Among the detected PAHs, fluoranthene was the most abundant quantified PAH in the porewater at all stations, followed by pyrene. High concentrations of the B[a]A, CHRY, and FLTH at site HJ were probably due to the five major industrial pollution sources (refineries, export processing zones, two industrial zones, and plastic manufacturing) found nearby, which bear a large amount of industrial wastewater and household sewage. Site DG1 was surrounded by livestock industries in the region which caused livestock wastewater pollution. The pollution sources in River YS could be related to domestic wastewater and additional wastewater discharge from the Linhai Industrial Zone and surrounding steel plant. Moreover, the mainstream of SY is crossing through two major industrial areas, coupled with discharged wastewater from the livestock industry. Pollution at river DB is due to the metal industry along the river, causing serious metallic and domestic wastewater pollution. These are caused by the major metal surface treatment industry and animal husbandry in the AGD upstream. The main pollution source impact of livestock wastewater on the WR and LK may be related to high concentrations of PAHs. The results of chemical analysis are shown in Table 4.
The ΣPAHs in the porewater of the 35 stations ranged from 0.085 to 378 μg/L, and the average ΣPAHs of the eight streams ranged from 4.807 ± 5.614 (AGD river) to 69.846 ± 151.487 (WR river).The sampling station with the highest ΣPAHs was WR1, followed by HJ1. Studies showed that the concentration of ΣPAHs in the porewater ranged between 48.2 and 205.7 μg/L from the Lanzhou Reach of the Yellow River (China). The concentration of ΣPAHs detected in the porewater of the Mersey Estuary (UK) ranged from 0.095 to 0.742 μg/L [44]. The concentration range within porewater of Xiamen Harbor (China) Figure 6. Effect of pollutants from sediment porewater on the heartbeat rate of zebrafish embryos. Hatching rate of zebrafish embryos exposed for 72 h at sampling stations with slightly toxic and moderately toxic porewaters. Error bars represent the standard deviation of the six zebrafish embryos. * p < 0.05 indicates significant differences between the control and exposure groups.

Comparative Results of PAH Concentrations
For chemical analysis, naphthalene, benzo(b)fluoranthene, dibenz(a,h)anthracene, and acenaphthalene were not detected in any of the porewater samples. Among the detected PAHs, fluoranthene was the most abundant quantified PAH in the porewater at all stations, followed by pyrene. High concentrations of the B[a]A, CHRY, and FLTH at site HJ were probably due to the five major industrial pollution sources (refineries, export processing zones, two industrial zones, and plastic manufacturing) found nearby, which bear a large amount of industrial wastewater and household sewage. Site DG1 was surrounded by livestock industries in the region which caused livestock wastewater pollution. The pollution sources in River YS could be related to domestic wastewater and additional wastewater discharge from the Linhai Industrial Zone and surrounding steel plant. Moreover, the mainstream of SY is crossing through two major industrial areas, coupled with discharged wastewater from the livestock industry. Pollution at river DB is due to the metal industry along the river, causing serious metallic and domestic wastewater pollution. These are caused by the major metal surface treatment industry and animal husbandry in the AGD upstream. The main pollution source impact of livestock wastewater on the WR and LK may be related to high concentrations of PAHs. The results of chemical analysis are shown in Table 4.
The ΣPAHs in the porewater of the 35 stations ranged from 0.085 to 378 µg/L, and the average ΣPAHs of the eight streams ranged from 4.807 ± 5.614 (AGD river) to 69.846 ± 151.487 (WR river).The sampling station with the highest ΣPAHs was WR1, followed by HJ1. Studies showed that the concentration of ΣPAHs in the porewater ranged between 48.2 and 205.7 µg/L from the Lanzhou Reach of the Yellow River (China). The concentration of ΣPAHs detected in the porewater of the Mersey Estuary (UK) ranged from 0.095 to 0.742 µg/L [44]. The concentration range within porewater of Xiamen Harbor (China) was <1 to 3548 ng/L [45]. The total PAHs detected in the Jiulong River Estuary and Western Xiamen Sea, China ranged from 158 to 949 µg/L [46]. Compared to other studies, the concentrations of the 16 PAHs in the porewaters of this study are still slightly higher than in other aquatic environments, indicating that the toxicity of the porewater is worthy of a follow-up discussion.

Comparison of Results from PAH Analysis and Biological Survival Rates
The survival rate of test organisms and the concentration of PAHs were determined, after statistical analysis. The degree of correlation was classified as low when the correlation was ±0.3 (between 0.3 and −0.3) while moderate and high correlations were ±0.3 to 0.6 ±0.6 to 0.9 (between 0.3 and 0.6, −0.3 to −0.6). Since naphthalene, acenaphthene, benzo(b)fluoranthene, dibenz(a,h)anthracene, and acenaphthalene were not detected, the correlation could not be calculated. The biological survival rate showed a low correlation with phenanthrene, anthracene, pyrene, benzo(a) anthracene, chrysene, benzo(b) fluoranthene, and benzo(a) pyrene; medium correlation with fluoranthene and benzo(g,h,i)perylene; and a high correlation with fluorene.  Table 5 shows heavy metal concentrations in the porewater of sediments of the 35 sites. Test results showed that the concentrations of the trace metals nickel, copper, and lead were mostly below the detection limits. Chromium and zinc were the two most detected metals in sediment porewater. The highest metal concentration for total chromium was detected at sample site LK2; zinc showed the highest concentration at site WR4. ND The concentration of heavy metals measured in the porewater of the HJ3 and HJ7 sediment was low; however, we observed a low survival rate of H. azteca. A possible reason is that the death of H. azteca may be due to other classes of pollutants, or due to high content of TOC combined with free metal ions, which may decrease their bioavailability and, consequently, reduce their toxicity [47][48][49]. The literature mentioned that the concentration of heavy metals in porewater is high, not all of which comes from the isolation of heavy metals in sediments, but may be caused by the release of contaminated groundwater [50]. Studies have also pointed out that the copper concentration in the sediment porewater of lakes is positively correlated with organic matter content, but the mortality of chironomid larvae could not be explained by the copper concentration in the porewater [51]. Therefore, it is impossible to determine the cause of the mortality of H. azteca based on the concentration of trace metals in the porewater alone.

Interstitial Water Benchmark Units (IWBUs) and Interstitial Water Toxic Units (IWTUs) Used to Determine Toxicity in Sediment Porewater
In the present study, the survival rates of H. azteca and IWBUs were calculated as shown in Figure 7. The results of IWBUs (>1 means toxic) showed that 22 sample sites were toxic and 13 sample sites were not. The results of ∑IWTU showed that 77% of the sample sites exceeded 1.0 (Figure 8), benthic organisms were not protected, and adverse effects may occur. According to the survival rate of Hyalella azteca, 28 sample sites were toxic and 7 sample sites were not. In this study, IWBUs and ∑IWTUs were used to predict the biological toxicity of the porewater of the sediment of eight streams in southern Taiwan, with an accuracy of 79% and 85% (number of IWBUs (∑IWTUs)) > 1 site, and corresponding to the percentage of sites with Hyalella azteca survival rate < 80%, showing that both porewater toxicity calculations for metal and organic compounds can be used as a tool for initial toxicity predictions. When the sample station coincided with IWBUs > 1 and IWTUs > 1, H. azteca exposure to the sample also showed toxic effects (survival rate < 80%), showing that the combination of IWBUs and IWTUs indicators can more effectively predict the potential toxicity of porewater.
In the present study, the survival rates of H. azteca and IWBUs were calculated as shown in Figure 7. The results of IWBUs (>1 means toxic) showed that 22 sample sites were toxic and 13 sample sites were not. The results of ∑IWTU showed that 77% of the sample sites exceeded 1.0 (Figure 8), benthic organisms were not protected, and adverse effects may occur. According to the survival rate of Hyalella azteca, 28 sample sites were toxic and 7 sample sites were not. In this study, IWBUs and ∑IWTUs were used to predict the biological toxicity of the porewater of the sediment of eight streams in southern Taiwan, with an accuracy of 79% and 85% (number of IWBUs (∑IWTUs)) > 1 site, and corresponding to the percentage of sites with Hyalella azteca survival rate < 80%, showing that both porewater toxicity calculations for metal and organic compounds can be used as a tool for initial toxicity predictions. When the sample station coincided with IWBUs > 1 and IWTUs > 1, H. azteca exposure to the sample also showed toxic effects (survival rate < 80%), showing that the combination of IWBUs and IWTUs indicators can more effectively predict the potential toxicity of porewater.
The results of calculating the IWBUs and survival rate of zebrafish embryos in this study showed that when IWBUs > 1 and ∑IWTUs > 1 (indicating toxic action), it can accurately predict the toxic effects on zebrafish embryos (Figures 9 and 10). However, when IWBUs < 1 (indicating non-toxic action), the zebrafish embryos still exhibit toxic effects. This can be explained by the notion that IWBU is mainly for benthic organisms, and embryo exposure belongs to the early life cycle stage of organisms, so the sensitivity is higher than that of later stages [52]. Therefore, the toxicity thresholds for a test with embryos are lower than the set values of the original IWBU.   The results of calculating the IWBUs and survival rate of zebrafish embryos in this study showed that when IWBUs > 1 and ∑IWTUs > 1 (indicating toxic action), it can accurately predict the toxic effects on zebrafish embryos (Figures 9 and 10). However, when IWBUs < 1 (indicating non-toxic action), the zebrafish embryos still exhibit toxic effects. This can be explained by the notion that IWBU is mainly for benthic organisms, and embryo exposure belongs to the early life cycle stage of organisms, so the sensitivity is higher than that of later stages [52]. Therefore, the toxicity thresholds for a test with embryos are lower than the set values of the original IWBU.

Conclusions
The results of our chemical analysis from 35 porewater samples showed that the PAHs naphththalene, acenaphthalene, benzo(b)fluoranthene, dibenz(a,h)anthracene, and the metal cadmium were not detected at all ampling stations, and the detection frequency of the other 22 compounds, ranging from high to low, was fluoranthene (94.29%), pyrene (91.43%), and zinc (88.57%). The mortality rate of H. azteca exposed to 11 sampling stations (DB1, DB2, SY1, SY2, SY3, HJ7, LK1, LK3, WR1, WR3, and WR6) was 100% at the end of the porewater toxicity test, and the survival rate was not significantly correlated with the 6 trace metals and 16 PAHs. In addition, 100% mortality in the zebrafish embryo toxicity tests was not only observed at the previous 11 sites but also at the other 12 sampling sites (YS1, YS2, YS3, YS4, DB3, DB4, DB5, HJ2, LK2, WR2, WR4, and WR5). This indicates that the porewater of these 11 sites was toxic for both test organisms, and zebrafish embryos were more sensitive than the amphipod H. azteca. Based on the experimental results, the samples show a lack of acute mortality and can also significantly reduce the heartbeat rate

Conclusions
The results of our chemical analysis from 35 porewater samples showed that the PAHs naphththalene, acenaphthalene, benzo(b)fluoranthene, dibenz(a,h)anthracene, and the metal cadmium were not detected at all ampling stations, and the detection frequency of the other 22 compounds, ranging from high to low, was fluoranthene (94.29%), pyrene (91.43%), and zinc (88.57%). The mortality rate of H. azteca exposed to 11 sampling stations (DB1, DB2, SY1, SY2, SY3, HJ7, LK1, LK3, WR1, WR3, and WR6) was 100% at the end of the porewater toxicity test, and the survival rate was not significantly correlated with the 6 trace metals and 16 PAHs. In addition, 100% mortality in the zebrafish embryo toxicity tests was not only observed at the previous 11 sites but also at the other 12 sampling sites (YS1, YS2, YS3, YS4, DB3, DB4, DB5, HJ2, LK2, WR2, WR4, and WR5). This indicates that the porewater of these 11 sites was toxic for both test organisms, and zebrafish embryos were more sensitive than the amphipod H. azteca. Based on the experimental results, the samples show a lack of acute mortality and can also significantly reduce the heartbeat rate of zebrafish embryos. Therefore, it is concluded that the hatching rate of zebrafish embryos provided a more sensitive endpoint than mortality and deformity rate. These results also indicate that the cardiac functioning and circulatory system of fish embryos decreased in a moderate or even low toxic environment. Therefore, it is concluded that the index of heart rate is more sensitive than mortality, and deformity rates of early larvae towards polluted porewater areas of sediments can be affected to some extent. Pearson correlation analysis showed that the concentration of acenaphthene, fluorene, and anthracene had a significant positive correlation with zebrafish embryo survival rate (p < 0.05), the heart rate had a significant positive correlation only with fluorene and anthracene (p < 0.001), and the deformity rate had a significant positive correlation with anthracene (p < 0.05). The chemical analysis and toxicological approach provide a preliminary understanding of the possible risks of dissolved metals and PAHs in sediment porewaters collected from different regions of Taiwan.