Litter additions reduce the side effects of biocides on soil nematode communities in Illicium verum forest

The experiment took place in Nanning, Guangxi province, China (107°45’-108°51’E, 22°13’-23°32’ N). The annual mean temperature and precipitation of this region are 21.6 ˚C and 1304 mm, respectively. The soil is classified as red soil (Udic Ferrosols) according to the Chinese Soil Taxonomy Classification (Gong, 1994). This experiment was arranged as a split-plot with different biocides treatments as main plots, and non-litter added and litter added treatments as subplots. Each plot has three replications. Within each plot, 4 treatments including control, insecticide treatment, fungicide treatment and insecticide × fungicide treatment are randomly distributed. Each treatment was divided into two subplots with or without litter addition. Each subplot (2 m × 2 m) was placed 2 m apart in order to minimize the risk of contamination.

Carbosulfan and benomyl-hymexazol are broad-spectrum biocides, which are often applied to forest pest control. For the insecticide treatment, carbosulfan (granular formulation and 5 % active ingredient) was applied at rate of 250 g/m² in July, it was mixed with soil (non-litter-added subplots) or litter (litter-added subplots) and spread under the crown geometry. For the fungicide treatment, benomyl-hymexazol (6 % of benomyl and 24 % of hymexazol) was applied at rate of 90 g/m² (3 times as recommended dose and diluted 300 times with water when in use). Because the fungicide is liquid, it could not be intercepted by litter layer (about 2 – 3 cm in depth), the fungicide was applied twice in each subplot on the litter layer (litter-added subplots) or bare soil (non-litter-added subplots), in July and August. Control plots received an equivalent volume of water. Insecticide × fungicide treatment applied both insecticide and fungicide as described above. I. verum forests can produce a large amount of litter every year. Shao et al. (2015) found litter provided basal resources for soil microorganisms and connects with soil organism closely in I. verum forest, so it was chosen as a remediator for soil ecosystem disturbance. Litter of I. verum was collected from the previous growth season. Upon collection, litter was oven dried to constant weight at 60 °C and then mechanically chopped to smaller particles (length=2 cm). Litter was completely mixed before adding to each treatment with pH 5.50, total C 483.0 g/kg and total N 15.9 g/kg. In litter-added subplot, 2000 g/m² dry litter was covered uniformly on the soil surface under the crown geometry, a layer of fine soil was sprinkled in order to prevent the litter from being blown away.

Soil samples were collected in October. Four random cores (10 cm × 10 cm × 10 cm) were combined to form one composite sample for each treatment. Each composite sample was placed in an individual plastic bag and sealed. All samples were kept at 4 °C for biological and chemical analysis.

Soil organic carbon (SOC) was measured by the dichromate oxidation, and soil total nitrogen (TN) was determined with an ultraviolet spectrophotometer after Kjeldahl digestion (Bao 2000). Total carbon and total nitrogen of plant litter were measured using a Vario MACRO cube Elementary Analyzer (Elementar Analysensysteme Vario MACRO cube, Germany).Microbial biomass carbon (MBC) microbial biomass nitrogen (MBN) were determined using the fumigation extraction method (Wu et al., 2006).

Nematodes were extracted from 100 g soil sample (fresh weight) by a modified cotton-wool filter method (Liang et al., 2009). The 100 individuals (or the total number in samples containing less than 100 individuals) of each sample were identified to genus level. The nematodes were assigned to the following trophic groups characterized by feeding habits (1) bacterivores; (2) fungivores; (3) omnivore-predators and (4) herbivores (Yeates et al., 1993).

The following ecological indices were calculated:

(1) Dominance index (λ), λ<!-- λ --> = ∑<!-- ∑ --> p i 2 , $\lambda=\sum p_{i}^{2},$where pi is the proportion of individuals in the i-th taxon (Simpson 1949);

(2) Shannon index (H^ʹ), H^ʹ = -Σp_i(lnp_i), where p_i is the proportion of individuals in the i-th taxon (Shannon & Weaver 1949);

(3) Generic richness GR = (S - 1)/ln(N), where S is the number of taxa and N is the number of total nematodes (Yeates & Bongers 1999);

(4) Nematode channel ratio (NCR), NCR= B/(B+F), where B is the number of bacterivores and F is the number of fungivores (Yeates 2003);

(5) Structure index S I = 100 ×<!-- × --> ∑<!-- ∑ --> k s n s / ∑<!-- ∑ --> k s n s + ∑<!-- ∑ --> k b n b ; $S I=100 \times\left(\sum k_{s} n_{s} /\left(\sum k_{s} n_{s}+\sum k_{b} n_{b}\right)\right);$

(6) Enrichment index EI = 100 ×<!-- × --> ∑<!-- ∑ --> k e n e / ∑<!-- ∑ --> k e n e + ∑<!-- ∑ --> k b n b , $\text { EI }=100 \times\left(\sum k_{e} n_{e} /\left(\sum k_{e} n_{e}+\sum k_{b} n_{b}\right)\right),$

where k_b is the weight assigned to guilds BF₂ and FF₂ and n_b is the abundance of nematodes in guilds BF₂ and FF₂, which indicate basal characteristics of the food web; k_s is the weight assigned to guilds BF₃-BF₅, FF₃-FF₅, and OP₂-OP₅, n_s is the abundance of nematodes in these guilds, which represent the structure condition of the food web; k_e the weight assigned to guilds BF₁ and FF₂, and n_e is the abundance of nematodes in these guilds, which represent an enriched condition of the food web (Ferris et al., 2001). BF_x, FF_x, OP_x, (where x = 1-5) represent the functional guilds of nematodes that are bacterivores, fungivores, predators and omnivores where the guilds have the characters indicated by x on the colonizer-persister (cp) scale (1-5) following Bongers and Bongers (1998).

Plotting EI and SI provides a graphic representation of nematode faunal profile that indicate whether the soil community is basal, enriched, or structured and stable (Ferris et al., 2001). The faunal profile can be divided into four quadrants, quadrant A (SI<50, EI>50) shows that food web is affected by high disturbances and food web condition is disturbed; quadrant B (SI>50, EI>50) means that the disturbance level is low or moderate and food web condition is maturing; quadrant C (SI>50, EI<50) represents an undisturbed and structured food web; quadrant D (SI<50, EI<50) indicates a stressed and degraded food web (Ferris et al., 2001).

Data of nematode abundance was ln (x + 1) transformed prior to statistical analysis for normality of data. Three-way ANOVA was applied to assess the effects of litter addition, insecticide treatment, fungicide treatment and their interactions on soil nematodes communities. P-values < 0.05 were considered significant. The relationship between nematode genera and environmental factors was examined based on redundancy analysis (RDA) using the CANOCO software, version 5.0. A Monte Carlo permutation test (499 permutations) was used to test the significance of first and all canonical axes.

In non-litter-added plots, the application of fungicide decreased the abundance of omnivores-predators in comparison with control (CK); while an opposite trend was observed in the treatments with litter additions (Fig.1). Similarly, the decrease in the abundance of herbivores in the insecticide treatments was only observed in the non-litter-added plots. In addition, more bacterivores observed in the treatments with litter addition. The abundance of fungivores only showed response to the fungicide treatments, with higher abundance of fungivores observed in the treatments with litter addition.

Fig.1

The interactive effects between insecticide and litter addition significantly influenced nematode ecological indices (P < 0.05). In non-litter-added plots, the application of insecticide decreased the indices of H’, GR and NCR and increased λ. However, in litter-added plots, insecticide did not cause obvious variations in the values of λ and H’ (Table 1). The interactive effects between fungicide and litter addition were only found in GR. In comparison with the control, the application of fungicide increased GR in litter-added plots, but decreased GR in non-litter-added plots.

The RDA showed a correlation between soil nematodes and the abiotic factors (Fig. 2). The ordination of nematodes was strongly affected by litter addition and soil abiotic factors. Acrobeloides, Ditylenchus, Plectus, Rhabditidae, Aphelenchus and Panagrellus, associated with litter, SOC and TN positively, Basiria, Heterocephalobus and Rotylenchus associated with insecticide and CK without litter addition positively.

Fig.2

Nematode faunal profiles showed that nematodes from CK and insecticide × fungicide treated soil in litter-added plots were located in quadrant B. Nematodes form single insecticide and fungicide treated soil in litter-added plots and from CK and single fungicide treated soil in non-litter-added plots were located in quadrant C. While nematodes impacted by insecticide and insecticide × fungicide without litter addition were located in quadrant D (Fig. 3).

Fig.3

Our results showed that non-target soil nematodes were affected adversely by insecticide without litter addition. Neher et al.(2014) also reported soils treated with conventional insecticide contained less complex and successionally mature nematode communities. Insecticide declined nematode diversity via increasing dominant genera abundance (λ) and decreasing rare genera abundance (H’) in non-litter-added plots (Table 1). Both herbivores and bacterivores were impaired by insecticide significantly, and omnivore-predators disappeared in insecticide × fungicide treatment without litter addition (Fig.1). The loss of this high trophic level of the soil micro-food web may cause loss of ecosystem functions (Van der Wurff et al., 2007). Thus, insecticide may decrease the risk of trees disease by killing harmful insect and herbivores, but simultaneously disturb the balance of soil ecosystem by impacting biodiversity.

Litter resources play an important role in shaping nematode community structure and function (Zhao et al., 2021). In our study, the increased soil SOC and TN content correlated with most nematode species positively with litter addition (Fig. 2), which indicated that litter could influence nematode communities by altering soil nutrient content. Litter with high SOC content influences insecticide behavior differently than soil (Puglisi et al., 2012). Singh and Srivastava (2009) suggested clay and organic matter could accelerate adsorption of Carbofuran. Moreover, microarthropods that feed directly on litter and microbes can fragment litter and disperse microbes in the process of feeding (Neher et al., 2012). The increase in microbial carbon and microbial nitrogen confirmed that nematodes could access to more food resource which is conducive to maintaining community stability with litter addition (Supplementary Table 1). Although litter addition could keep nematode away from adverse effects of biocides, it may limit the effectiveness of insecticide via increasing the abundance of herbivores. Litter addition may result in higher soil moisture retention and nutrient content, and then increase abundance of herbivores. Consequently, further studies were needed to evaluate the proper quantity of litter applied in order to match the needs for both keeping insecticide effectively and maintaining soil ecosystem functions.

Unexpectedly, fungivorous seem not to be impaired by insecticide (Fig. 1), one possible reason is that relatively higher microbial carbon content in insecticide treatment can provide more food resources for fungivorous (Supplementary Table 1). Another reason could be that most fungivorous correlated with soil SOC and TN closely (Fig. 2), and those nutrients were not impacted by insecticide (Supplementary Table 1). Additionally, fungivorous are dependent partly on root-derived resources (Kudrin et al., 2021), which help them resist environment stress. Chen et al. (2016) also found some specific microorganisms adapted or improved their tolerance to hexaconazole.

The effects of fungicide on the nematode appeared not to be relevant in non-litter added plots, our results are in agreement with findings of Neher et al.(2019), in which they found that biocides had no effects on nematode trophic diversity, maturity indices and channel index, and assumed that crop rotation and tillage are main contributors to changes of nematode communities rather than biocides. Wang (2014) reported that soil nematodes showed a clear recovery trend 120 days after fungicide application. Fungicide was applied twice in July and August, and soil samples were collected in October. Therefore, we propose that spaced fungicide application gave nematodes time to recover and adapt to environment. Compensatory mechanisms such as the increase of tolerant nematode were also probably responsible for this phenomenon. For instance, Aphelenchoides which was most abundant in fungicide treatment with litter additions also showed tolerance to benomyl (Schmidt et al., 2000) (Supplementary Table 2).

Liu et al. (2019) reported the accumulation of litter addition may benefit the nematode community. In our study, fungicide increased the abundance of nematode markedly in litter-added plots (Fig.3). Our results were partly in accordance with the findings of Ekelund (1999) who found that fenpropimorph stimulated fungivorous flagellates and ciliates with glucose addition and suggested that the fungicide effect on soil protozoa was not mediated via effects on fungal populations. Hence, the effect of fungicide on fungivores, obviously, may also have no relation to the reduction of their fungal food source in this study.

Nematode faunal profile derived from EI and SI can provide information about the status of the soil food web (Ferris et al., 2001). In this study, the samples from insecticide and its interaction with fungicide treatments were all located in quadrant D and separated obviously from single fungicide and the control treatments without litter addition (Fig. 3). Therefore, we propose that soil food web length and connections are significantly reduced in insecticide and its interaction with fungicide treatments without litter addition. Our result was in line with Bai et al. (2017) who revealed that soil food web with high level of pesticides addition was stressed and degraded. Neher et al. (2014) also found SI value in insecticide treatment was lower than in non-insecticide treatment, and suggested insecticide shifted ecological succession back to earlier stages. The contributor for this phenomenon is the reduced abundance of high nutrient level nematodes which have ability to regulate soil food web (Fig.1). The normal operation of soil functions depends on the overall cooperation of soil microorganisms (Álvarez-Martín et al., 2016). SI and EI placed nematode communities into quadrant C and quadrant B with litter addition. It can be assumed that litter addition may help to reduce the adverse effects of biocides and maintained the structure and stability of soil food web. Furthermore, lowest SI value was found in insecticide × fungicide treatment (Fig. 3), suggesting that insecticide and fungicide may accumulate their effects and create worse influence on soil food web than single insecticide or fungicide treatment.

Liu et al. (2019) reported that litter was a limiting food source for the soil food web. Xu et al. (2013) found litter input is a crucial pathway for carbon and nutrient fluxes to the soil. In this study, litter addition changed soil biota community composition, which in turn may affect soil nutrient decomposition and mineralization. In non-litter-added plots, insecticide changed decomposition pathway from bacteria dominant (NCR > 0.5) to fungi dominant (NCR < 0.5) (Table 1). Because fungi have higher C assimilation efficiencies (Rousk & Bååth 2011), fungi dominant pathway could slow down soil nutrient cycling and decrease nutrient availability (Waring et al.,2013). Alternatively, fungal dominant channel maybe one of the defense mechanisms of soil food web that could promote resistance to environment change. Conversely, in litter-added plots, applications of insecticide resulted in bacterial based channels (NCR > 0.5). The present results are in line with the findings of Sauvadet et al. (2016), in which they show that the bacterial channel developed faster with leaf litter addition. Because bacterial based channels can accelerate decomposition rate, nutrient availability in litter-added plots was higher than in non-litter-added plots.