The effect of soil type and ecosystems on the soil nematode and microbial communities

We examined the physical and chemical properties, nematode communities, and microbial attributes in soil samples collected from a Stagnosol (SS), a Cambisol (CS), and a Chernozem (CM) in each of a forest (FOR), a meadow (MEA), and an agricultural field (AGR) ecosystems. The soil types, ecosystems, locations, and vegetation characteristics are shown in Table 1.

Soil samples from each soil type and ecosystem were collected from five randomly established 1 × 1 m quadrats in selected plots of 20 × 20 m in May (M), July (J), and September (S) 2016. Five randomised subsamples were collected from the quadrats, one from each corner and one from the centre of the plots, for analysing the soil nematode communities, microbial activities, and physicochemical properties. The subsamples were bulked to produce a representative sample for the plot (1 kg). Samples were collected from a depth of 10 cm, excluding the surface humus layer. A total of 135 representative samples were collected; 5 from each ecosystem (FOR, MEA, and AGR, 5x3=15), from three soil type (SS, CS, and CM; 15x3=45), in three sampling date (M, J, and S; 45x3=135). The samples were transferred to the laboratory in sealed plastic bags and stored at 5 °C until processing for the nematode analysis or at -20 °C for the microbial analysis.

Total soil C and nitrogen (N) contents, soil moisture (SM) contents, and pH were measured in all samples. The organic C and total N contents were determined using a Vario MACRO Elemental Analyzer (CNS Version; Elementar, Hanau, Germany). Organic C content was determined based on the difference between total C and C bound in carbonates. SM content was estimated gravimetrically by oven-drying fresh soil at 105 °C overnight, and pH was measured potentiometrically in 1M KCl suspension by a digital pH meter separately for each representative sample.

Each sample was homogenised by gentle hand mixing, and stones were manually removed. The nematodes were extracted from 100 g of fresh soil by a combination of Cobb sieving and decanting (Cobb 1918) and a modified Baermann technique (van Benzoijen, 2006). One hundred grams of soil from each representative sample were soaked in l L of tap water for 60 min to disrupt soil aggregates and promote nematode movement. The soaked sample was carefully passed through a 1-mm sieve (16 mesh) to remove plant parts and debris, and this suspension was passed through a 50-μm sieve (300 mesh) 2 min later to remove water and very fine soil particles. The nematodes were then extracted from the soil/water suspension by a set of two cotton-propylene filters in the Baermann funnels. Two filter trays were used per sample to limit material thickness to <0.5 cm. Suspensions containing the nematodes were collected after extraction for 24 h at room temperature. The nematodes were killed and fixed in a hot 99:1 solution of 4 % formaldehyde and pure glycerol (Seinhorst, 1962). The all nematodes were microscopically (100, 200, 400, 600, and 1000× magnification) identified to the species level (juveniles to the genus level) from temporary slides using an Eclipse 90i light microscope (Nikon Instruments Europe BV, Netherlands). Nematode abundance was expressed as the number of individuals per 100 g of dry soil.

The nematodes were assigned to fifteen functional guilds integrating nematode feeding strategies (trophic groups) and the nematode coloniser-persister (c-p) scale (Bongers & Bongers, 1998). The five nematode trophic groups were: bacterivores (Ba), fungivores (Fu), carnivores (Ca), omnivores (Om), and plant parasites (Pp) (Wasilewska, 1997). The Pp group included both obligatory plant parasites and facultative plant parasites that may attack plants or fungi. Colonisers-persisters characterising nematode life strategies are classified on a scale of 1 to 5 (Bongers, 1990). C-p1 represents “r-strategists” (colonisers) with short life cycles, small eggs, high fecundity, high colonisation ability, and high tolerance to disturbance, eutrophication, and anoxybiosis. Colonisers generally live in ephemeral habitats. At the other end of the scale, c-p5 nematodes represent “k-strategists” (persisters) with the longest generation times, largest bodies, lowest fecundities, and the highest sensitivity to disturbance. Persisters are never dominant in a sample and generally live in stable habitats where they become very abundant (Bongers, 1990). C-p scaling allows the calculation of the basal maturity index (MI) for non-parasitic nematodes, the plant parasitic index (PPI) for plant parasites only (Bongers, 1990), and the summ maturity index (ΣMI) (Yeates, 1994) for all nematode taxa. Functional guilds allow the calculation of the enrichment index (EI), the structure index (SI), and the channel index (CI) proposed by Ferris et al., (2001). The species-diversity index (H´) defined by Shannon and Weaver (1949), the nematode channel ratio (NCR) defined by Yeates (2003), and trophic diversity (TD) defined by Heip et al., (1998) were also calculated.

Nematode species were characterised as dominant at D >5 % (the species represents more than 5 % of the total nematode abundance in the ecosystem or soil type) and subdominant at D >2 % (the species represents more than 2 % of the total nematode abundance in the ecosystem or soil type) (Losos et al., 1984).

Microbial biomass C (Cmic) content was determined following the procedure described by Islam and Weil (1998). Ten grams of oven-dried equivalent (ODE) of field moist soil adjusted to 80 % water-filled porosity was irradiated twice by microwaves (MW) at 400 J g^-1 ODE soil to kill the microorganisms. The cooled samples were extracted with 0.5 M K₂SO₄, and the C content of the extract was quantified by oxidation with K₂Cr₂O₇/H₂SO₄. The same procedure was performed with a non-irradiated sample. Cmic content was determined as (Cirradiated content - Cnon-irradiated content)/KME, where KME represents the extraction efficiency (0.213) recommended by Islam and Weil (1998).

The functional diversity of the soil microbiota was determined using the methods described by Insam (1997). Each well in a BIOLOG EcoPlate received 150 μl of an extract prepared by resuspending of fresh soil in 0.85 % NaCl and diluted 1:10 000. The plates with the extracts were then incubated at 27 °C for 6 d, and absorbance at 590 nm was recorded every 24 h using a Sunrise Microplate reader (Tecan, Salzburg, Austria). The data were corrected against the initial readings at time zero and were expressed as optical densities of individual wells. The richness of the soil microbial community (Richn) was determined as the number of substrates used by the microbial community, i.e. the number of wells with a positive response after background correction. Hill’s diversity index (Diver) (Hill, 1973) based on Eq. 1 was calculated for estimating the diversities of the microbial functional groups:

in which p_i is the ratio of the activity on a substrate to the sum of activities on all substrates.

Data were log-transformed before analysis to improve normality. Soil and ecosystem types were included as fixed factors. The effects of soil type (SS, CS, and CM), type of ecosystem (FOR, MEA, and AGR), and sampling date (M, J, and S) on nematode trophic-web descriptors and functional guilds, soil properties, and microbial biomass, diversity, and richness were analysed by factorial analyses of variance (ANOVAs). Nonparametric Spearman’s correlation coefficient (rs) was calculated to test the relationships between nematode functional guilds, microbial parameters, and soil parameters for each sample using STATISTICA v9.0. Correlations obtained at P<0.05 were considered significant.

We then used multivariate analyses to evaluate the effects of soil and ecosystem types on nematode-community composition and the microbial characteristics. The composition of the nematode functional guilds and the microbial parameters were thus used as response variables, and the soil and ecosystem types were used as explanatory variables in a multivariate framework of a redundancy analysis (RDA). The soil physicochemical parameters were used as supplementary variables. Canoco 5 for Windows was used for the multivariate analyses (vers. 5.04; Ter Braak & Šmilauer, 2012).

The factorial ANOVA found that ecosystem type (FOR, MEA, and AGR) and soil type (SS, CS, and CM) significantly affected all soil properties (except SM content vs. soil type) and that sampling date (M, J, and S) affected only SM content (P<0.01, Table 2). The bi-factorial interaction ecosystem × soil significantly affected all soil properties, ecosystem × date affected half of the properties, and soil × date and the interaction of all three factors (ecosystem × soil × date) had minor or no effects on the soil properties. The values of the soil properties were generally higher in the FOR soils (except pH) than the MEA and AGR soils and higher in CM (including pH) than CS and SS. pH and the C/N ratio were correlated negatively in FOR but positively in AGR and MEA (Fig. 1).

Fig. 1

The three ecosystems and soil types contained 133 nematode species (32 bacterivores, 26 fungivores, 9 carnivores, 24 omnivores, and 42 plant parasites) (Table S1). Heterocephalobus eurystoma, Stegetellina leopolitensis, Ditylenchus parvus, Ditylenchus tenuides, Paraphelenchus obscurus, Boleodorus volutus, Cephalenchus intermedius, and Ecphyadophora tenusissima were new to the list of Slovak nematode fauna, increasing the total number of soil nematode species to 732. The number of species (99) and diversity were highest in the FOR soils, followed by the MEA (90) and AGR (53) soils. Nematode species number and diversity were higher in CM than CS and SS (102, 81, and 60, respectively) (Tables S1, 4). The most abundant nematode species by trophic group were Acrobeloides nanus and Chiloplacus propinquus (bacterivores), Aphelenchus avenae and Filenchus vulgaris (fungivores), Clarkus papillatus and Mylonchulus brachyuris (carnivores), Eudorylaimus carteri (omnivores), and Aglenchus agricola, Boleodorus thylactus, and Bitylenchus dubius (facultative and obligate plant parasites) (Table S1).

Soil type and sampling date had significant effects on overall nematode abundance (P<0.01), but ecosystem type did not (Table 3). Ecosystem and soil types significantly influenced the abundances of all nematode functional guilds (except Om₅ and Pp₂, respectively), but the nematode-community compositions were similar.

The mean abundance of Ba₂ nematodes was significantly higher in FOR than MEA and AGR (P<0.01) and in CM than SS and CS (P<0.01). The amount of microbial biomass and microbial richness and diversity had tendencies similar to those of the Ba₂ nematodes; all were higher in FOR and CM (Table 3). Ba₄, Fu₃, and Pp_2,3 nematodes were most abundant in MEA, Ba₁ and Fu₂ were most abundant in AGR, and Ba₃, Ca_3,4, Fu₄, Om_4,5, and Pp₅ were most abundant in FOR. The majority of the nematode functional guilds were more abundant in CM than SS and CS. Only Ba₁ was significantly more abundant in CS (P<0.01). Sampling date only significantly affected the abundance of c-p2 nematode (Ba, Fu, and Pp) trophic groups, with higher values in M and S than J. Microbial richness was also affected by sampling date and was highest in J (P<0.01).

Nematode abundance, species number, and functional guilds and the microbial parameters were positively correlated with all soil properties. Only the Ba₂ nematode parameters were negatively correlated with SM content, and the Pp₂ nematode parameters were negatively correlated with the C/N ratio (Table 5). The RDA analysis, however, indicated that the abundance of most of the nematode guilds, total nematode abundance, nematode species number, and the microbial parameters tended to be higher in environments with higher pHs and that N and C contents tended to be higher in FOR and CM soil (Figure 1A), except for Fu₂ and Fu₃ nematodes. The presence and distribution of nematodes within functional guilds, number of species, nematode abundance, and the microbial parameters in MEA were more affected by soil type than soil properties.

The ANOVA found that ecosystem and soil types significantly affected all descriptors (except PP vs. ecosystem type). Sampling date had a significant effect on MI, ΣMI, PPI, and NCR (P<0.01, 0.05, Table 4). The interaction ecosystem × soil significantly affected all descriptors (P<0.01), ecosystem × date and soil × date significantly affected half of the descriptors, and the interaction of all three factors (ecosystem × soil × date) affected the majority of the descriptors. MI, ΣMI, PPI, H´, SI, and TD were generally higher in FOR than MEA and AGR soils and in CM than CS and SS. EI was highest in AGR and CS, CI was highest in MEA and SS, and NCR was highest in FOR and CS.

Ecosystem type was an important factor shaping soil nematode and microbial communities and affecting soil properties. The abiotic and biotic soil properties and interactions amongst them were best for the FOR ecosystem. FOR had the highest SM, C, and N contents and C/N ratio but the lowest pH. C and N contents were twice as high in FOR than AGR but were similar to those in MEA. The supposed benefits of management of agricultural land (e.g. tillage, fertilisation, and crop rotation) include increased soil C and N contents, fertility, water retention, and overall provision of ecosystem services (Garbach et al., 2017; Sánchez-Moreno et al., 2018). The low C and N contents in our AGR soils, however, suggested differences in the quantity and quality of inputs to the soil, nutrient inputs and losses, low plant diversity and stimulation of decomposition by soil disturbance compared to the semi-natural (MEA) and natural (FOR) ecosystems. These results are in agreement with many studies of differences in soil C and N contents and changes following conversion of forest to managed agricultural land, well summarised in a review by Murty et al., (2002). This review revealed that large amounts of C and N could be lost when forest is converted into cultivated land but that no changes in soil C and N contents were recorded when forests were converted to uncultivated pasture (similar to our meadow). In contrast, the abandonment and reforestation of agricultural land can substantially increase C and N storage (Compton and Boone 2000), due to increase in plant diversity (Lange et al., 2015). Additionally, cultivated soils usually have lower C/N ratios than forest soils (Murty et al., 2002), consistent with our and other results (Fernandes et al., 1997; Smil, 1999; Compton & Boone, 2000). Our C/N ratio was negatively correlated with pH in FOR, consistent with the results reported by Högberg et al., (2007).

Food, water, and temperature are the three primary factors that determine the habitats occupied by nematodes and microbes, the degree of species diversity, and the composition and structure of their communities. The availability of food, water, and temperature, however, are determined by ecosystem type, soil characteristics (e.g. structure, pH, and chemistry), plant composition, and microclimatic (Neher, 2010) or seasonal (Gaugler & Bilgrami, 2004) variations. More diverse nematode and microbial assemblages contribute to more resilient ecosystem services (Yeates, 2007; Fuhrman, 2009; Háněl, 2017). Forest soils, for example, contain more species than agricultural soils (Domsch et al., 1983; Neher et al., 2005), some with >400 nematode species (Yeates, 2007). This finding is consistent with our results; nematode species numbers and diversity (H´) were highest for FOR, even though FOR had the lowest overall nematode abundance, suggesting that established forests represent relatively stable environments providing suitable conditions for maintaining balanced and rich nematode trophic webs (Yeates, 2007). This was supported also by values of ecological and functional indices (Bongers, 1990; Ferris et al., 2001). All maturity indices (MI, ΣMI, PPI) as well as Structure index and Trophic diversity were generally higher in forests than in grasslands and/or cultivated soils. These results partially agree with those by Neher et al., (2005), who reported that MI, PPI and SI were higher in forests than in wetlands and agricultural soils. Ecosystem type has significant effect on values of CI, which was the highest in meadow soils in our study, indicates a higher proportion of fungal decomposition (fungal decomposition channels) and low abundance of c-p1 bacterial feeders (e.g. Rhabditidae and Panagrolaimidae) (Ferris et al., 2001). In contrast, Neher et al., (2005) revealed the highest CI value in forest soils.

We found several nematode species exclusively in one ecosystem e.g. Paraphelenchus obscurus in AGR, Paratylenchus microdorus in MEA, and Filenchus polyhypnus in FOR. Extreme disturbances, such as bulldozing, slash-and-burn management, windstorms, and wildfires in forests, however, can substantially reduce nematode diversity (Yeates, 2007; Čerevková et al., 2013). The species richness of the nematode fauna in FOR in our study was higher than in the soils of a protected forest in the Slovak Tatra National Park nine years after a windstorm and wildfire, likely due to the persistent influence of changes in the plant community and basal soil physicochemical properties (Renčo & Čerevková, 2015; Renčo et al., 2015).

The FOR soils also had the highest microbial biomass, richness, and diversity, what positively correlated with C and N contents, and was consistent with the observations of Yergeau et al., (2006)). Microbial biomass is involved in the control of the synthesis and decomposition of soil organic matter and acts as an accessible storage system for nutrients in ecosystems. Sites with high microbial biomass can therefore stock and recycle more nutrients for plant nutrition and thus improve the sustainability of an ecosystem (Kaschuk et al., 2010). In contrast, the number and diversity of nematode species and diversity of microbial functional groups in our study were lowest in AGR. Additionally, AGR had half the amount of microbial biomass than FOR and MEA, and microbial biomass was negatively correlated with C and N contents. These findings support our hypothesis that biological diversity would be lowest in the agricultural soils due to periodic perturbation, land management, and crop monoculturing, consistent with the results by Neher et al., (2005); even though AGR had the highest overall nematode abundance, likely due the periodic organic manure inputs (Hu et al., 2018).

Bacterivorous nematodes are often the most dominant feeding group in forest (Neher et al., 2005; Yeates, 2007; Renčo & Čerevková, 2017) and agricultural (Neher et al., 2005, Renčo et al., 2010) soils. The preponderance of Ba₂ bacterivores (A. nanus, C. persegnis, and C. propinquus) in all ecosystems in our study was likely due to the high microbial biomasses in FOR and MEA and to the management (tillage and fertilisation) in the corn monoculture in AGR. Microbial biomass was nevertheless significantly lower in AGR than FOR and MEA. Microbial diversity is often lower after a natural habitat has been cultivated (Buckley & Schmidt, 2001). These results are in agreement with Wasilewska (1997), who stated that a higher abundance of microflora would support larger numbers of bacterivorous nematodes. An increase in the abundance of this group is indicative of enhanced microbiological activity e.g. after the addition of cow and chicken manure or slurry (Wasilewska, 1997; Neher & Olson, 1999). Our study thus demonstrated the synchronisation between bacterivorous nematodes and their food resources, which has not been frequently reported (Wardle et al., 2001, Papatheodorou et al., 2004). Fungivorous nematodes (Fu₂) were the second most abundant trophic group in all ecosystems. AGR had the highest abundance of fungivores, mainly A. avenae, F. vulgaris, Ditylenchus intermedius, and Aphelenchoides parietinus, likely due to the high density of fungal hyphae and spores under Z. mays monoculture from the association of corn with arbuscular mycorrhizal fungi (Bai et al., 2008).

Plants and their root systems serve as food for plant parasitic nematodes (Flis et al., 2018; Le et al., 2019) before they serve as a food source for microbivorous nematodes during decomposition. Root systems are more diverse in natural ecosystems with rich communities of plant species than for monocultured crops. Root growth is also more extensive and less ephemeral in perennial plants than annual crops and supports a soil community with many species of plant parasites, omnivores, and predators (Neher, 2010). Plant parasites are common in natural grasslands (Popovici & Ciobanu, 2000; Čerevková, 2006). The abundance of plant parasites, such as Boleodorus thylactus, Malenchus exiguus, and Paratylenchus microdorus (Pp₂) or Amplimerlinius macrurus, Geocenamus microdorus, and Bitylenchus dubius (Pp₃) was highest in MEA.

The importance and high population densities of plant parasitic nematodes in agriculture are mainly associated with specific crop pests, e.g. root-knot and cyst-forming endoparasites (e.g. Meloidogyne, Heterodera, and Globodera). The high overall abundance of Pp nematodes in AGR (Aglenchus agricola Pp₂, Paratylenchus bukowinensis Pp₂, Bitylenchus dubius Pp₃, and Helicotylenchus dihystera Pp₃ are all ectoparasites) suggests their close relationship with cultured crops. Omnivores and carnivores were significantly more abundant in MEA and FOR than AGR, consistent with previous findings by Neher et al., (2005) and Renčo et al., (2010).

Soil type was also an important factor affecting the nematode and microbial communities and soil properties. Soil properties were best in CM, with a neutral pH and the highest C and N contents and C/N ratio, followed by CS and SS. C and N contents were twice as high in CM than SS, in agreement with the general soil classification (www.vupop.sk).

Soil type was more important than ecosystem type for both the nematode and microbial communities. For example, nematode abundance, number of nematode species, and microbial biomass or diversity positively correlated in the CM soil type in two out of three ecosystems studied. Significant effects of soil type on the composition of nematode communities have been documented by Alphei (1998) and Lišková et al., (2008) in forests, by Popovici and Ciobanu (2000) in grasslands, and by Neher et al., (2005) in agricultural land. The populations of bacterivores (mainly A. nanus, Eucephalobus striatus, and C. propinquus) and fungivores (A. avenae and F. vulgaris) and microbial biomass in our study were highest in CM with aerobic conditions, a neutral pH, and a high humus content beneficial to microbial activities and associated nematodes (Wasilewska, 1997). In contrast, the abundances of bacterivores and fungivores were low in SS because of its oxygen deficiency and acidic conditions. These results partially agreed with those by Lišková et al., (2008), who reported that Cephalobidae bacterivores (Acrobeloides, Acrobeles and Cervidellus) were more abundant in a light sandy Regosol with a high pH, but disagreed with those by Wasilewska (1997) and Lišková et al., (2008), who reported that fungivores were more abundant in an acidic Cambisol.

The abundance of facultative plant parasites (Pp₂) did not differ amongst the soil types. A. agricola in CM, Malenchus bryophilus in CS, and B. thylactus in SS were nevertheless the most abundant, supporting the preference of various species of Pp₂ nematodes for different soil types, also reported by Lišková et al., (2008). In contrast, obligate plant parasites (Pp₃) were most abundance in CM, followed by CS and SS, probably due to the different levels and distributions of food sources between these soil types, as also suggested by Popovici and Ciobanu (2000) and Lišková et al., (2008). Natural ecosystems are characterised by high proportions of omnivores and predators (Wasilewska, 1997; Ferris et al., 2001). Omnivores and predators were most abundant in CM, but only in FOR and MEA.

In our study soil type was also as important factor affecting values of all ecological and functional indices, contradicting findings of Lišková at al., (2008), who reported that only fungal to bacteria (F/B) ratio and channel index (CI) was significantly different among Cambisol, Regosol, Fluvisol and Rendzina soil types. Ruess (2003) studied CI and F/B at various sites and stated that soil and climate affect CI more strongly than does ecosystem type. In our study CI was significantly affected by both, ecosystem and soil type as well as their interactions, and sampling date has no impact on CI values.

In general, season (sampling date) in our study had relatively minor effects on both the abiotic and biotic characteristics. Only SM content fluctuated with the season (lowest in summer) what significantly affecting microbial biomass, confirming results of Buchanan and King (1992). Similar overall nematode abundance influences sampling date, which can partly be explained by changes in SM, in agreement with observation of Sohlenius and Boström (2001) from Swedish Scot pine forest soils. Out of functional guilds, Ba₂, Fu₂, and Pp₂ nematodes were influenced by sampling date, however only Ba₂ were negatively correlated with SM content.