3.1. Soil properties and metal concentrations
The results of the soil properties (including the pH, TK, AK, TP, AP, TN, SOC, and CEC) and concentrations of Cd, Pb, Cu, and Zn in soil determined at the sampling sites are shown in Tables S2 and S3. In the study sites (Table S2), the pH values, TK, AK, TP, AP, TN, SOC, and CEC contents ranged from 7.9 to 8.3, from 11.18 to 25.18 g kg-1, from 82.24 to 342.1 mg kg-1, from 1.04 to 2.04 g kg-1, from 15.01 to 87.16 mg kg-1, from 1.08 to 2.54 g kg-1, from 6.62 to 11.88 g kg-1, and from 6.92 to 11.51 cmol kg-1, respectively. In the study sites, Cd, Pb, Cu, and Zn concentrations ranged from 0.17 to 5.94, from 24.23 to 283.2, from 16.39 to 462.3, and from 44.45 to 263.2 mg kg-1, respectively (Table S3). The descriptive data showed that Cd, Cu, and Pb concentrations exceeded the threshold legal values at sites S19 and S20 (as stipulated in the Chinese Environmental Quality Standard for Soils, prescribed by the Ministry of Environmental Protection of China (GB 15618-2018); Table S3].
3.2. Earthworm community
Earthworm species collected at the study sites are listed in Table 1. In total, 535 individuals were collected from all study sites (earthworm species were absent in sites S19 and S20), and comprised three families, six genera, and ten species across the sites. Five species belonged to the family Megascolecidae, with one species from the genus Amynthas (Amynthas hupeiensis) and other four from the genus Metaphire (including Metaphire guillemi, Metaphire asiatica, Metaphire kiangsuensis, and Metaphire tschiliensis). Two species belonging to the family Moniligastridae were Drawida gisti and Drawida japonica. The rare species Eisenia foetida, Aporrectodea trapezoids, and Bimastus parvus belong to the family Lumbricidae.
In the study sites, the ecological category with endogeic species was the most diverse and abundant group; five species were endogeic, four were epigeic, and only one was an anecic species (Table 1). The most abundant species was A. hupeiensis (Michaelsen, 1895) (accounting for 39.4% of the total number of individuals), followed by D. gisti (Michaelsen, 1931) (37.8%), and D. japonica (16.3%) (Michaelsen, 1892). The rare species in the study sites were M. guillemi (1.87%) (Michaelsen, 1895), M. tschiliensis (Michaelsen, 1928) (0.17%), M. asiatica (0.56%) (Michaelsen, 1900), M. kiangsuensis (Chen, 1930) (1.5%), and A. trapezoids (Duges, 1828) (0.73%). Furthermore, Shannon (H), Simpson (D), and Pielou (P) indexes were calculated to determine the alterations of earthworm diversity and species evenness, as shown in Table 2. The results showed that the values of H, D, and P indexes ranged from 0.45 to 1.34, from 0.28 to 0.68, and from 0.55 to 1.0, respectively.
3.3 Relationships between soil variables and earthworms
Relationship between earthworm community (including the total, adult, and juvenile individuals, species, diversity and evenness indexes), and soil properties (including the pH, TN, SOC, TP, TK, AK, AP, and CEC) and concentration of metals (Cd, Pb, Cu, and Zn) were analyzed using RDA (Figs. 2 and 3) to determine the potential impact of the soil variables on earthworm composition. The correlations between different soil properties, metals in soil, and earthworm parameters from the RDA are shown in Tables 3 and 4.
In Fig. 2, the RDA results showed the ranking of soil variables and earthworm parameters, with the first principal component axes explaining 96.4% of the total variance, suggesting that sites S5 and S4 differed from the other study sites. Pearson’s correlations between soil and earthworms showed that soil properties including TK, AK, and SOC contents significantly correlated with earthworm density (total, adult, and juvenile worm individuals) (p<0.05, Table 3), respectively, in study sites (except for the sites of S19 and S20 with no earthworms). The Cd and Pb concentrations influenced the earthworm density (total and adult worms) at the study sites (p<0.05). Juvenile worms were negatively correlated with pH values (p<0.05). In this study, the earthworm diversity and evenness variations by Shannon, Simpson, and Pielou indexes did not correlate with the soil variables (p>0.05, Table 3), except for the noted decreased and increased values of Shannon and Peilou index, respectively, with increasing soil TN content (p<0.05).
In Fig. 3, the RDA results showed the ranking of the soil variables and different earthworm species at each site; the first principal component axes explained 91.6% of the total variance. Pearson’s correlation coefficients between earthworm species and soil variables are shown in Table 4. RDA ranking results showed that soil variables, including the AK, SOC, and Cd and Pb concentrations affected the density of the dominant species A. hupeiensis and D. gisti (p<0.05, Table 4). In addition, the AP and TN contents were positively correlated with D. japonica (p<0.05), respectively. In contrast, AP levels in the soil were negatively correlated with the density of M. asiatica, M. kiangsuensi, and D. gisti (p<0.05), respectively.