3.1 Occurrence and Congener Group Patterns
Once ingested from diet, hydrophobic organic contaminants (HOCs) were taken up in the gastrointestinal tract, firstly passed through the liver and then redistributed to other organs such as muscle before being finally stored in adipose tissue (Du et al., 2020). The concentrations of PBDE congener groups in liver, muscle, and adipose tissues of two snake species are summarized in Table 1. Hexa-, octa- and nona-BDEs were detected in all tissue samples of two snakes. The total PBDE concentrations of liver, muscle, and adipose tissues ranged from 29.1-252.2, 10.5-48.9 and 4.36-106.4 ng/g lw, respectively. The average ΣPBDEs concentration in muscle samples of short-tailed mamushi (23.1 ng/g lw) was close to those of red-backed rat snake analyzed in the present study (33.2 ng/g lw) and our previous report (28.3 ng/g lw) (Zhou et al., 2016b), but was 1-2 orders of magnitude lower than of water snakes sampled from e-waste sites (Liu et al., 2018; Wu et al., 2013; Wu et al., 2020). Additionally, a significant positive correlation was observed between body weight of snakes and muscle concentrations of PBDEs in all samples (r2 = 0.56, p < 0.01; data not shown). This result suggested that the PBDE exposure of snakes might increase with their growth stages. The research focusing on PBDE levels in other tissue of reptiles was relatively limited. Wu et al. (2014) reported that the PBDE exposure of liver sample of Chinese alligators was 920 ng/g lw, significantly higher than the snakes observed in the present study.
Table 1
Tissue concentrations (ng/g lw) of tri- to deca-BDEs of two snake species.
species
|
red-backed rat snake
|
|
short-tailed mamushi
|
tissue
|
liver
|
muscle
|
adipose
|
|
liver
|
muscle
|
adipose
|
tri-BDE
|
0.09 ± 0.28
|
0.39 ± 0.8
|
0.002 ± 0.005
|
|
0.56 ± 0.77
|
0.43 ± 0.43
|
0.001 ± 0.003
|
(nd-0.83)
|
(nd-2.2)
|
(nd-0.01)
|
|
(nd-2.22)
|
(nd-1.13)
|
(nd-0.01)
|
tetra-BDE
|
0.95 ± 2.76
|
0.35 ± 0.62
|
0.660 ± 1.78
|
|
0.86 ± 0.7
|
0.11 ± 0.1
|
0.100 ± 0.1
|
(nd-8.32)
|
(0.02-1.97)
|
(0.01-5.41)
|
|
(nd-1.97)
|
(0.01-0.22)
|
(nd-0.26)
|
penta-BDE
|
5.43 ± 5.92
|
0.40 ± 0.41
|
1.310 ± 1.89
|
|
5.90 ± 9.2
|
0.25 ± 0.26
|
0.910 ± 0.52
|
(nd-15.35)
|
(0.02-1.2)
|
(0.23-6.24)
|
|
(nd-23.22)
|
(0.01-0.72)
|
(0.35-1.78)
|
hexa-BDE
|
33.70 ± 31.7
|
1.12 ± 0.63
|
7.980 ± 12.8
|
|
35.60 ± 22.1
|
3.42 ± 1.50
|
5.250 ± 3.82
|
(3.1-89.6)
|
(0.29-2.42)
|
(1.38-41.4)
|
|
(5.91-63.4)
|
(1.63-6.50)
|
(1.91-23.2)
|
hepta-BDE
|
8.05 ± 7.36
|
0.89 ± 0.61
|
2.360 ± 2.87
|
|
12.01 ± 7.67
|
2.19 ± 1.72
|
5.300 ± 8.97
|
(nd-24.8)
|
(nd-2.31)
|
(0.71-9.03)
|
|
(2.75-24.9)
|
(nd-5.53)
|
(1.20-35.6)
|
octa-BDE
|
24.80 ± 26.7
|
1.98 ± 2.42
|
3.180 ± 5.12
|
|
18.10 ± 16.1
|
2.89 ± 2.00
|
4.320 ± 6.47
|
(3.49-88.7)
|
(0.39-7.15)
|
(0.57-16.5)
|
|
(2.37-51.6)
|
(1.15-6.55)
|
(1.1-28.9)
|
nona-BDE
|
28.60 ± 15.3
|
8.68 ± 7.45
|
7.770 ± 8.61
|
|
40.50 ± 25.8
|
4.62 ± 4.17
|
6.480 ± 3.47
|
(12.3-58.3)
|
(1.62-19.6)
|
(1.29-26.6)
|
|
(7.20-71.6)
|
(1.26-10.7)
|
(3.31-12.9)
|
deca-BDE
|
31.00 ± 26.1
|
9.34 ± 5.4
|
0.350 ± 0.41
|
|
46.09 ± 28.9
|
6.25 ± 2.14
|
0.310 ± 0.22
|
(7.95-92.2)
|
(4.67-19.3)
|
(nd-1.29)
|
|
(7.62-76.75)
|
(3.91-10.17)
|
(0.14-0.79)
|
ΣPBDEs
|
172.60 ± 58.1
|
33.20 ± 10.4
|
28.600 ± 32.7
|
|
159.60 ± 84.1
|
23.10 ± 6.75
|
19.200 ± 15.8
|
(69.7-230.7)
|
(10.5-48.9)
|
(4.36-106.4)
|
|
(29.1-252.2)
|
(11.2-40.2)
|
(9.83-68.0)
|
The variations of PBDE congener group patterns in different tissues of two snake species, are presented in Figure 1. In liver samples of two snake species, the levels decreased in the order deca-BDE > hexa-BDE > nona-BDE > octa-BDE > hepta-BDE > penta-BDE > tetra-BDE > tri-BDE, contributing to 26.0%, 23.9%, 23.4%, 15.1%, 6.8%, 3.9%, 0.6% and 0.2% of the total PBDE levels, respectively. Deca-BDEs (36.6%) were dominant in snake muscle samples followed by nona-BDEs (31.6%). For adipose samples, nona- and hexa-BDEs were the two most abundant congener groups collectively contributing to 60.3% of total PBDEs. PBDE patterns predominated by deca-BDEs have been observed in various biotic and abiotic matrices (Li et al., 2018; Li et al., 2019; Wu et al., 2020), which may be due to the fact that deca-BDE was the major congener in in-use PBDE commercial products (Abbasi et al., 2015). Furthermore, deca-BDE would be debrominated down to penta- and octa-BDE congeners in vitro (Chabot-Giguere et al., 2013). Therefore, the higher concentrations of deca-BDE in liver and muscle compared to those in adipose tissue may be indicative of the metabolic fate of deca-BDE in snakes. The congener group patters of PBDEs showed the significant difference (p < 0.01) between the two snake species (Figure 1). The most fluctuating congeners were octa-BDE, nona-BDE and hepta-BDE for liver, muscle, and in adipose tissues. The interspecies difference for PBDE exposure may be attributed to different feeding habits and dermal exposure (Du et al., 2018), since the short-tailed mamushi generally lives in the terrestrial environment while the red-backed rat snake prefers the semi-aquatic condition (Du et al., 2020).
3.2 Tissue Distribution and Tissue-Specific Burden
Similar tissue distributions of total PBDE levels were observed in two snakes (Figure 1). In most snake samples, the tissue levels followed the order liver > muscle > lipid. This differed from the body distribution of chlorinated paraffins (CPs) in snakes,(Du et al., 2020) which coincided with the lipid contents in different tissues (Table 1). Result from the present study suggested that other factors (e.g. bioaccumulation pathway and metabolism) besides the lipid content might influence the tissue-specific accumulation of PBDEs in both red-backed rat snake and short-tailed mamushi. Sun et al. (2017b) investigated the tissue distribution of HOCs in freshwater fish community, and found that the significant accumulation of high lipophilic pollutants in liver tissue compared to other ones. Previous field study of frogs also detected the highest concentrations of HOCs in liver followed by muscle (Du et al., 2019), which might be explained by the fact that liver functioning as storage of glycogen and fat to apparent extent may cause the enhanced POP loads.
The snake liver and muscle samples shared close PBDE congener group patterns with enrichment of deca-BDE, which was very limited in snake lipid samples (Figure 1). The living habit may lead to the different congener group pattern in adipose tissues: reptiles would stop foraging and consume their body fat during brumation in winter; once adipose tissue was used, the highly lipophilic contaminants such as deca-BDE that were previously stored in lipid may redistribute among whole body. Although consumption of fat may reduce the HOC storage in snake adipose tissue, the redistribution process probably increased HOC in other tissues and eventually elevated the risk of exposure toxicity, which should be considered in biomonitoring (Du et al., 2020). Recent studies emphasized that the concentrations of PBDEs in different tissues may not entirely represent the tissue exposure, mainly due to the varying contaminant concentrations in wildlife tissues through time (Du et al., 2020). For example, HOC concentrations in adipose may be diluted by an increase of body fat (Yordy et al., 2010). In contrast, contaminant burdens are much steady and can reflect the whole exposure to tissues as well as individuals (Du et al., 2019).
In the present study, tissue-specific PBDE burdens of two snake species were calculated as the product of the tissue mass and tissue PBDE concentrations and presented in Figure 2. The body PBDE burdens (sum of three tissue burdens) in red-backed rat snake (400.2 ng) was two folds higher than those in short-tailed mamushi (188.6 ng). The red-backed rat snake individuals in the present study were generally thick in body condition since they had heavy body weight and high proportion in tissues (Table 1), in comparison with the short-tailed mamushi individuals which were considered to be thin. Yordy et al. (2010) also observed the higher contaminant burdens in fatter bottlenose dolphins. Animals with rich fat content may readily store lipophilic compounds in tissues. Simultaneously, there may be differences in metabolic activity between the two snake species, in particularly in snake liver and muscle, and these may be part of the explanation for the differences seen between the two species.
The relative contributions from liver, muscle and adipose tissues to the total PBDE burdens (sum of three tissue burdens) were similar between red-backed rat snake and short-ailed mamushi. The total burdens of two snakes were dominant by adipose (85.1% and 80.0%, respectively) while liver (8.1% and 12.5%, respectively) and muscle (6.9% and 7.5%, respectively) tissues had low proportion of the total burdens. For congener groups, high proportion of tri-BDE burdens were found in muscle tissue of red-backed rat snake (69.1%) and short-tailed mamushi (76.4%). Up to 80% of tetra- to nona-BDE burdens were enriched in adipose tissue of two snakes. For deca-BDE, liver tissue was the main storage tissue contributing to 43.6% of the burdens. These results indicated that the adipose tissue in snakes could serve as a more effective storage and sink for PBDE congeners compared to liver or muscle tissues. In HOC biomonitoring researches, measuring muscle or liver tissue concentrations was a common approach for assessing the chemical exposure in wildlife (Glüge et al., 2018; van Mourik et al., 2016). However, the overlook of adipose data may underestimate the exposure of PBDEs, in particular for animals with rich fat contents.
3.3 Biomagnification Potential
Biomagnification factors (BMFs) of HOC can be calculated by dividing the predator concentration by the prey concentration. Biomagnification is suggested to take place when the BMF value is higher than one. It is of great importance to confirm the predator and prey species in the food web (Glüge et al., 2018). Although snakes are generally seemed as a high trophic species, the diet of these feeders would vary with the environment. Most of red-backed rat snake preferred to live in/near water and choose the aquatic organisms that is the relative abundant and readily available. According to a snake dissection study in YRD, the pond loach, rice field eel and black-spotted frog were supposed as potential major diet for red-backed rat snakes in the local environment (Huang et al., 1990). Further assurance of predator-prey relationship was carried out using stable isotope analysis, and the analyzed δ13C and δ15N values of red-backed rat snake and its prey candidates have been described in a previous study using the same samples (Du et al., 2020). δ13C values of red-backed rat snake and black-spotted frog were close (-23.76 ± 0.89‰ and -24.83 ± 0.41‰, respectively) whereas red-backed rat snake was enriched in δ15N (9.33 ± 0.62‰ SD) compared with their diet (7.36 ± 1.21‰ SD). For pond loach and rice field eel, the δ13C values were significantly lower than those observed in red-backed rat snakes (p < 0.05), which indicated that the pond loach and rice field eel were not the major carbon sources of red-backed rat snakes. These results confirmed the direct predator-prey relationship between the red-backed rat snakes and black-spotted frogs in the paddy field food web.
In the present study, biomagnification factors (BMFs) of PBDE congeners were calculated by dividing the lipid-normalized muscle concentration of red-backed rat snakes by the lipid-normalized muscle concentration of black-spotted frog. As shown in Figure 3, more than half of individual BDE congeners (13 out of 22) displayed BMF values above one. The BMF values of tri-BDE, tetra-BDE, penta-BDE, hexa-BDE, hepta-BDE, octa-BDE, nona-BDE and deca-BDE averaged at 0.75, 2.30, 1.62, 2.61, 3.10, 4.30, 0.85 and 0.74, respectively. In general, tetra- to octa-BDEs exhibited biomagnification potentials through the frog-snake food chain. The BMFs of PBDEs in the present study were comparable with those reported between predators and prey from an e-waste recycling site in South China (0.13-5,30) (Wu et al., 2020), but lower than those observed in the North Sea (0.6-15) (Weijs et al., 2009), the Atlantic coast (14-54) (Shaw et al., 2012), and the feeding experiment (7.2-20.6) (Luo et al., 2013). These results may suggested the trophic magnification potentials were affected by the behavior of predator and distinctive ecology, in which the food resource would vary markedly.
Parabolic relationship was observed between BMFs and log Kow values of PBDE congeners (Figure 3). The BMFs increased with increasing number of bromine atoms up to 8 and then rapidly declined as bromine atom number raised. It was noteworthy that the BMFs plots dropped dramatically after hepta-BDE indicating the limited biomagnification potential of nona- and deca-BDEs (Br ≥ 9) probably due to their very high molecular weight. Similar results have also been observed in bioaccumulation of PBDE (Wu et al., 2020), bioaccumulation of CPs (Sun et al., 2017a), trophic magnification of CPs (Zhou et al., 2018) and bioaccumulation study for other halogenated HOC (Zhu et al., 2015). PBDE congeners with higher bromine number tend to have higher log Kow values (Tittlemier et al., 2002; Yue and Li, 2013) resulting in higher biomagnification potential. However, the breakpoint from positive to negative correlation was approximately log Kow 8.6. The parabolic trend showed that the biomagnification potential of PBDE congeners appealed to be related to both hydrophobicity and molecular size of chemicals. The congener groups with low-moderate log Kow values (tri- to octa-BDEs) have smaller molecular size, and therefore, lipophilicity is the dominant factor. For congener groups with high log Kow values (nona- and deca-BDEs), larger molecular size may increase the difficulty of transmembrane and limit their bioavailability and biomagnification potential.