Introduction

Mercury is a toxic heavy metal element that exerts biological effects. It can be accumulated in the body, impair organ functionality, and even threaten the survival. Migration of mercury in the food chain has been paid more attention for a long time (Zhang and Wang et al. 2006; Ullrich et al. 2007). In recent years, the transfer of mercury through the terrestrial food chain and the ecological risks it poses have raised concerns. Assimilation of methylmercury from rice is a very severe exposure route in humans (Horvat et al. 2003; Feng et al. 2008). High contents of mercury and methylmercury have been found in several predatory terrestrial species (Wang et al. 2005; Zheng et al. 2010), but studies of mercury accumulation in the terrestrial food chain are yet few, compared with those of aquatic ecosystems.

The nitrogen stable isotope technique is a practical method for analyzing nutrient levels in an ecosystem. Stable isotopes can accurately reflect the food source and nutrient levels (Thimdee et al. 2004). Over the past 20 years, stable isotope techniques have been applied to investigations into the food web, especially in oceanic, estuary, and coastal tidal areas, and used to analyze nutritional relationships between primary producers and the higher trophic levels (Hobson et al. 1992; Doi et al. 2005; Al-Maslamani et al. 2012; Clayden et al. 2014; Bai et al. 2017). Multiple studies have reported that nitrogen isotopes are enriched between adjacent trophic levels. Stable nitrogen isotope ratios can be used to explain nutritional relationships among the animals, describe trophic levels, and build the nutritional levels of the food web (Peterson and Fry 1987; Yoshioka et al. 1994; Peterson 1999).

Using 15N to assess nutritional level relationships is very effective (Wang et al. 2004). The relationships between food web structure and nutritional levels had been studied using 13C and 15N in ecosystems where multiple species coexistence (Zhang et al. 2010). The previous study measured the proportion of carbon and nitrogen in different crops and their seasonal variations and used the information to reveal the habitat features of lady beetles using 13C and 15N delta changes in the crop-aphids-lady beetle food chain (Ostrom et al. 1997). At present, although stable isotope analysis is widely used in ecosystem, it has not been reported in environmentally sensitive zones, such as nonferrous areas. Therefore, the objective of this study was to determine the trophic levels of the food chain in a zinc-smelting area using 15N in multiple species. Further, we analyzed the accumulation of mercury in the different trophic levels and the transferred mechanism in the food chain.

Materials and methods

Study area

Huludao City is an important nonferrous smelting and chemical industry area located in the Liaoning Province of northeast China. The Cishan River is the main body of water in Huludao City. Over the past decades, wastewater containing heavy metals was discharged into this river from zinc-smelting area, the state-operated Huludao zinc smeltery, which is the largest in Asia. The smeltery produced 3.3 × 105 tons of zinc per unit year; the soil and rivers in this area had been contaminated by waste materials and effluents from the metallurgical industry for many years, resulting in severe environmental pollution (Xue 2009). As an environmentally sensitive area, long-term smelting activities lead to the plant type is considerably single (Zheng et al. 2010, 2012).

Samples collection

Based on distances from smelters, we selected Laohe Tai (S1), the Southern Bridge of Cishan River (S2), Nanshan Park (S3), a zinc factory (S4), a railway bridge near the zinc-smelting area (S5), and the zinc-smelting area (S6) as the sampling sites (Fig. 1). We collected samples of the following plants: Echinochloa crus-galli (found at all six sites), Phragmites australis (S1 and S4), Setaria viridis (S2, S3, and S4), Deyeuxia Clarion (S5), and Digitaria sanguinalis (L.) (S3 and S6); Zea mays (S4 and S5), Atriplex patens (S1), Xanthium sibiricum Patr. (S1), and Artemisia argyi (S1). The usual arthropods were collected from the different communities in the smelting area, and included Locusta migratoria manilensis, Acrida chinensis, Paraten odera sinensis, Argiope bruennichi, Gryllidae, Monomorium chinense Santschi and Formica. We also found Monomorium chinense Santschis in S6 and Formica in S3 and S5.

Fig. 1
figure 1

Sampling sites

Full size image

Animals were manually collected and immediately placed in polyethylene bottles with a few drops of ether to euthanize the insects. The samples were identified in the laboratory, and then thoroughly rinsed with deionized water to remove pollutants from their surfaces. Moisture was removed using filter paper. The insects were then sealed in polyethylene bags and frozen at − 4 °C.

Sample analysis

The entire bodies of arthropods were measured in the laboratory. Grasshoppers, locusts, crickets, and spiders from the different habitats numbered more than eight each. There were four mantes’ samples and several ants’ samples. The stems, leaves, and roots of the plants collected were measured. Arthropod samples were dried in a lyophilizer, and plant samples were dried naturally.

Stable isotopes and trophic levels

Samples were dried at 60 °C for 24 h, packed in tinfoil, and vacuum sealed. The samples were shipped to Third Institute of Oceanography, State Oceanic Administration to determine 13C and 15N. The analytical instrument was a stable isotope mass spectrometer (TC/EA IRMS). In this paper, we used the Pee Dee Belemnite (PDB) international standard for 13C and the standard for nitrogen in air for 15N. The analytical precision was δ 13C < 0.2 per thousand for 13C and δ 15N < 0.3 per thousand for 15N.

Trophic level relationships are the most important links among different organisms and form the foundation of community survival and the core of energy flows. The variation in δ15N between consumers and food is the nutritional fractionation and increases with trophic level, a process known as isotope enrichment (Vander Zanden et al. 1999; Post 2002). Therefore, we can divide the nutritional levels by the δ15N in animal tissues in adjacent trophic levels. We used δ15N to classify the trophic levels in the sampling sites. Formulas are shown as follows.

$$TL = \left( {\delta^{15}_{\text{consumer}} - \delta^{15}_{\text{rate}} } \right)/\Delta \delta^{15} N + \lambda$$
(1)

where TL is the trophic level of consumers, δ15 is the nitrogen stable isotope ratio of primary producers or primary consumers in the food web (λ = 1, the ratio of δ15 to δ15N for primary producers, when λ = 2, the ratio of δ15 to δ15N for primary consumers, when consumers trophic level is greater than 2, trophic level is not an integer). Generally, consumer spent not only the same trophic level of food (Vander Zanden et al. 1997).

Mercury measurements

Soil, plant, and insect samples were all digested using H2SO4–HNO3–V2O5 method (Liu et al. 2004). A Tekran-2600 CVAFS automated sampling analysis system with a detection limit of 0.005 ng L−1 was used to determine mercury contents in the samples. Blanks were also analyzed, and 20% of the samples were randomly selected for parallel testing. Hg levels were estimated in different organs using the concentrations by dry weight. All vessels were dipped in 3 mol L−1 HNO3 for 24 h and then rinsed with deionized water. The reagents were excellent grades.

The migration of mercury from plants to phytophagous insects could be analyzed. The biological enrichment coefficients (BCF) were calculated as follows.

$$BCF = I_{\text{hc}} /P_{\text{hc}}$$
(2)

where I hc is the mercury content in insects (mg kg−1), and P hc is the mercury content in plants (mg kg−1).

Quality control

Precision and accuracy were evaluated by comparing the measured mercury concentrations in with certified reference materials. The expected and measured concentrations in a hair reference (GBW-07601) were 0.36 ± 0.05 mg kg−1 and 0.40 ± 0.01 mg kg−1, respectively. The expected and measured concentrations in a soil reference (GBW-07405) were 0.290 ± 0.003 mg kg−1 and 0.290 ± 0.0025 mg kg−1, respectively.

SPSS and Excel were used for data analysis, and Arcgis9.0 was used to draw the sampling site map.

Results

Food web assessment

Herbivore food source analysis

Higher plants can be divided into C3, C4, and crassulacean acid metabolism (CAM) categories, depending on their carbon fixation pathways. The value of δ13C is approximately – 23 to − 30‰ in terrigenous C3 plants, and – 10 to − 14‰ in terrigenous C4 plants (Cai et al. 2001). We collected C4 plants from the study area (Table 1).

Table 1 Stable carbon and nitrogen isotope compositions of the different plants and animals
Full size table

The δ13C of consumers can reflect the δ13C of food from digestion and absorption and ranges from 0 to 1‰. Thus, δ13C is often used to analyze the sources of consumer food in the ecosystem (Vander Zanden and Rasmussen 1999). The Digitaria community in S6 are C4 plants. The values of δ13C and δ15N in animals are relatively higher. Plant communities in the S1–5 sampling sites consisted of C3 plants with the lower carbon levels, so the δ13C values in animals from these sites are all much lower. It further suggested that carbon and nitrogen contents in animals may be directly related to the surrounding environment. Owing to different food sources, isotope values from the same animals can vary. However, different animals assimilate the same food, their isotope values are very similar (Table 1).

The δ13C concentrations in the sampled communities are shown in Table 1. δ13C concentrations ranged from − 12.61 (Monomorium chinense Santschi) to − 29.63 (locusts). There were clear variations in δ13C values among animals from the different community habitats, and the δ13C of the same animal also can be different. δ13C between adjacent trophic levels accounted for approximately 0.4 to 1.0‰, limiting the study of δ13C in animal nutrition Li and Su (2012). However, there are certain enrichment phenomena in the different animals from the different community habitats in this study. The food source of locusts, grasshoppers, and crickets is mainly herbaceous plants. There are predator and prey relationships among ants, grasshoppers, and locusts. And there is none of the enrichment of stable relationship among the other animals. The δ15N values ranged from 1.73 (Monomorium chinense Santschi) to 9.94‰ (Paraten odera sinensis). The δ15N values in mantes and spiders were higher than in the other animals we sampled, so they occupy a higher trophic level.

Nutritional level analysis

Generally, primary consumers are the baseline organisms in the ecosystem and are perennial; δ15N variations among primary consumers can affect organisms occupying the higher trophic levels of the food web. As the most commonly found arthropods in the grassland ecosystem we sampled, locusts and grasshoppers are considered primary consumers. We determined the enrichment factor of the plants and animals by their difference of mean stable nitrogen isotope ratios. Variations in mean values (3.30‰) among the spiders and locust, grasshoppers and crickets in the different habitats indicated the enrichment factor of stable nitrogen isotopes. The stability of nitrogen isotope ratios of the consumers ranged from − 0.5 to 9.2‰ (mean 3.4 ± 1.1‰) and increased by approximately 3‰ (De Niro and Epstein 1981; Minagawa and Wada 1984). Increases in stable nitrogen isotope ratios are known to vary among species and ecosystems (Vander Zanden and Rasmussen 2001).

From the nutrition levels calculated, the mantes, crickets, locusts, and grasshoppers have predator and prey relationships in S1–6 sampling sites. Spiders, locusts, crickets, and grasshoppers have predator and prey relationships in sites S1–5 (Table 2). Compared with the other animals, the stable nitrogen isotope content was lower in ants, because ants are omnivorous and exhibit different metabolic characteristics.

Table 2 Relative trophic level of the selective animals in different sites
Full size table

According to the results, changes in relations among the trophic levels are mostly the herbivorous food chain. Most primary producers in the sampled habitats were herbaceous plants. Locusts, grasshoppers, and crickets mainly feed on them which are the second trophic level, namely primary consumers. By analogy, spiders and mantes mainly occupy the third trophic level. As omnivorous animals, ants have both characteristics of sarcophagus and phytophagous. We also can judge their diet characteristics from the trophic level positions. From the trophic level relationships among the different habitats, there are fewer species of arthropods in the nonferrous metallurgy area, but it constitutes a complex food web among the trophic levels, such as producers and consumers, primary consumers and secondary consumers.

Mercury concentrations in soil and plants

Mercury concentrations in the soil samples ranged from 0.129 to 28.182 mg kg−1 (mean 4.355 mg kg−1); this was approximately 117 times the background value (0.037 mg kg−1), indicating that mercury pollution in Huludao City was severe (Zheng et al. 2007). Mercury concentrations in plant leaves ranged from 0.029 to 0.191 mg kg−1 (mean 0.097 mg kg−1) and from 0.060 to 2.263 mg kg−1 in plant roots (mean 0.559 mg kg−1). Mercury concentrations were highest in sampling site S6 (Fig. 2).

Fig. 2
figure 2

Hg concentrations in herbaceous plants in the different sites

Full size image

Mercury concentrations in animals

Mercury concentrations in locusts range from 0.033 to 0.081 mg kg−1. The average is 0.064 mg kg−1. Mercury concentrations in Acrida chinensis range from 0.017 to 0.134 mg kg−1. The average is 0.077 mg kg−1. Mercury concentrations in Gryllidae range from 0.012 to 0.145 mg kg−1. The average is 0.088 mg kg−1. Mercury concentrations in ants range from 0.057 to 0.641 mg kg−1. The average is 0.372 mg kg−1. Mercury concentrations in Argiope bruennichi range from 0.024 to 1.502 mg kg−1. The average is 0.505 mg kg−1. Mercury concentrations in mantes range from 0.098 to 1.322 mg kg−1. The average is 0.397 mg kg−1. Mercury concentrations markedly differed in sampling sites (Fig. 3). Concentrations in sampling site S1 were notably lower than in the other five sites, and concentrations in the carnivorous spiders and mantes were significantly higher than concentrations in omnivorous animals and herbivores.

Fig. 3
figure 3

Mercury concentrations in animals from the different sites

Full size image

Mercury accumulation in arthropods from different community habitats

Mercury migration in plants—phytophagous insects

Locusta migratoria manilensis, Acrida chinensis, and Gryllidae are the common arthropods in Chinese grassland ecosystems during summer. Locusta migratoria manilensis and Acrida chinensis mainly feed on plant leaves, while Gryllidae feed on plant roots, stems, and leaves. Through the mercury migration in plants—Locusta migratoria manilensis, plants-Acrida chinensis, and plants-Gryllidae, the biological enrichment coefficients (BCF) can be calculated.

As shown in Table 3, the highest enrichment of mercury in Acrida chinensis was found in S3, followed by S5. Both coefficients were greater than 1, the rest of the other sites are less than 1. However, the highest mercury concentrations in plant stems and leaves were found in S4 (Fig. 2). Enrichment coefficient of mercury by Acrida chinensis in plants is only 0.38 at S4. Enrichment coefficient of mercury by Locusta migratoria manilensis was also not too high in plant stems and leaves.

Table 3 Enrichment of Hg by locusts, grasshoppers, and crickets in plants in different sites
Full size table

Mercury migration in phytophagous insects—carnivorous insects

Mantes and spiders typically feed on Locusta migratoria manilensis, Acrida chinensis, and Gryllidae in grassland ecosystems. Mantes and spiders have different enrichment coefficient in the different sites shown in Tables 4 and 5. According to the food chain, such as locust—mantis, grasshopper—mantis, and cricket—mantis (Table 4), we found a significant enrichment of mercury. With the exception of S3, the enrichment coefficients of locusts and grasshoppers were greater than 2 at all sites. Enrichment coefficient in spiders—Locusta migratoria manilensis, spiders—Acrida chinensis, and spiders—Gryllidae was higher than the enrichment coefficient in mantes—Locusta migratoria manilensis, mantes—Acrida chinensis, and mantes—Gryllidae. It indicated that the capacity of mercury accumulation of spiders was higher than that of mantes (Tables 4 and 5).

Table 4 Enrichment of locust, grasshopper, and cricket to mantis
Full size table
Table 5 Enrichment of locust, grasshopper, and cricket to spider
Full size table

Discussions

This study used a stable carbon and nitrogen isotope method to assess the interspecies relationships of arthropods in a smelting area. This method can directly reflect the nutritional relationships among the various creatures. Stable nitrogen isotope ratios and nutritional relationship models can accurately determine trophic level. The primary producers are plants, and the primary consumers in the area we sampled were locusts, grasshoppers, and crickets at the second trophic level. Secondary consumers were spiders and mantes at the third trophic level.

The plant types we found in this area are relatively simple. It is the main characteristics of environmentally sensitive area (Zheng et al. 2012). Mercury concentrations in the soil and plants were higher than those in non-smelting areas. Through the food transfer, there are certain risks in the study area.

The δ15N stable isotope and enrichment factor in arthropods from the smelting area were higher than those reported by Li and Su (2012) in the karst cave delta. We found the highest δ15N values in mantes (9.94‰) and spider (7.95‰). These values are lower than those reported for swimming animals in the aquatic food web and birds, such as sea gulls (14.27‰), tadorna ferruginea (10.00‰), and the palm gull (13.25‰) (Yang et al. 2009). The enrichment factor throughout nonferrous metallurgy area was 3.30‰. It indicates that although community habitats were polluted, δ15N values were not affected by food source or biological metabolism.

The previous study reported that mercury concentrations obviously increased from primary producers to carnivorous animals through six nutritional levels (Beneditto et al. 2012). Enrichment coefficient of mercury by grasshoppers, locusts, or crickets in plant stems and leaves in this study was higher than that reported for earthworms in soil, which was from 0.040 to 0.539 (Zhang et al. 2009), while the mercury concentrations in earthworms and soil (Zhang et al. 2009) were higher than that in our study. Enrichment coefficient of mercury by grasshoppers, locusts or crickets in plant stems and leaves in this study was significantly lower than that by four species of acridid grasshoppers in plants, which was from 0.70 to 2.00 (Devkota and Schmidt 2000). Mercury concentrations measured in plants were much lower than those in our study. These findings suggested that higher mercury concentrations in plants and soil may lead to lower enrichment coefficient values. Primary consumers, such as locusts and grasshopper, have the lower mercury accumulations, while secondary consumers, namely carnivorous animals, such as mantes and spiders, have an obvious mercury accumulations, especially the spiders. Therefore, humans, located at the top of the food chain, are exposed to the certain risk of mercury.

Conclusions

This study focuses on the enrichment of mercury in the food chain. It indicates that the plant in the different community habitats is the first trophic level. The locusts and grasshoppers belong to the second trophic level. The third trophic level includes spiders and mantes. The enrichment of mercury in primary consumers is not obvious than in arthropods of the third trophic level. The mean enrichment coefficients in spiders and mantis is greater than 1. The δ15N values we recorded suggest the mercury concentrations accumulated from primary producers to top carnivores with trophic levels increased. Therefore, the biological amplification of mercury in food web is very significant in a zinc-smelting area.