Introduction

Nematodes are some of the most diverse soil animals (^{van den Hoogen et al., 2020}) and are capable to respond quickly to environmental changes; therefore, they have been used as indicators of soil health in different crop systems and native areas (^{Neher, 2001}; ^{van den Hoogen et al., 2020}). Bacterial-feeding nematodes can ingest a great number of bacterial cells per day and are attracted by high CO₂ concentrations in their soil. Their preying capacity releases NH₄⁺ readily available to plants and other animals in the soil solution (^{Lavelle, 1997}).

Abundance and diversity of nematodes in fields under different agricultural systems are used to infer disturbances to the soil fauna (^{van den Hoogen et al., 2019}), as well as enrichment and disturbance of soil communities (^{Bongers, 1990}; ^{Ferris et al., 2001}; ^{Berkelmans et al., 2003}). The intensification of agricultural practices generally leads to biodiversity loss (^{Matson et al., 1997}; ^{Tscharntke et al., 2005}), reducing the taxonomic richness of soil nematode communities (^{Yeates and Bongers, 1999}; ^{Yeates and Stirling, 2008}).

Reduced tillage practices are used to minimize soil disturbance, as they benefit many soil characteristics, such as soil structure and fauna (Sánches-Moreno et al., 2006). However, most studies focus on the effects of tillage and cropping patterns on plant-parasitic nematodes and, to a lesser extent, on the free-living nematode communities (Sánches-Moreno et al., 2006).

This study investigated the effects of different management systems on nematode communities in preserved areas of the Atlantic Forest in Ponta Grossa, Paraná State, Brazil, and in neighboring areas, based on plant-parasitic and free-living nematodes. The results are based on studies conducted under field conditions for over 30 years to (a) contribute to the knowledge of nematode diversity (plant-parasitic and free-living) and (b) study the influence of agricultural activities on nematode communities.

Materials and Methods

Results

Soil chemical attributes effects on nematode community

The PCA and clustering analyses showed variability between the management systems for nematode communities and soil chemical attributes (Figures 1A and 1B). Based on nematode community composition (Figure 1A), CT was grouped by the clustering analysis with MT, whereas NT formed a separate group near NF, in accordance with to the Bray and Curtis index classification. The soil chemical analysis (Figure 1B) showed that NF was grouped with NT, while MT was grouped with CT.

The canonical correspondence analysis (Figure 2) performed with data of nematode community and the soil chemical analysis showed that NF was positively correlated to higher concentrations of Al and C, whereas MT and NT were correlated to P concentrations. Other soil characteristics were not significant in the analysis and were excluded from the graphic representation. Regarding nematode communities (Figure 2), MT and NT were correlated to P content and omnivores, plant-parasitic, and to predators, to a lesser extent. NF was correlated to fungal-feeders, plant-parasitic nematodes, and high Al content, while CT was not correlated to the chemical analysis; nevertheless, it was correlated to bacterial-feeders and predators.

Figure 2 – Canonical correspondence analysis bi-plot of association of soil properties (P = phosphorus; C = carbon; Al = aluminum) and nematode community in different crop systems (NF = native forest; MT = minimal tillage; CT = conventional tillage; NT = no-tillage). Data represent all samples combined (n = 40). Eigenvalues (lambda) are 0.5113, 0.1273, and 0.1001 for the first (horizontal), second (vertical) and third axes, respectively. The first two axes explained 64.8 % of the variation. 

Discussion

Studies conducted in Uruguay and Brazil showed that discrimination of different cropping systems using only abundance values was not applicable, when native areas were compared to cropped areas (^{Korenko and Schmidt, 2007}; ^{Tomazini et al., 2008}). However, in systems under more intensive cropping, such as CT and MT, values of total density were higher. Several authors reported the predominance of higher density values in areas submitted to more intensive cultural traits when compared to native areas (^{Andrade et al., 2004}; ^{Goulart and Ferraz, 2003}; ^{Mattos et al., 2006}), in Brazilian ecosystems. This can be explained as some crops host plant-parasitic species in the field, increasing the number of nematodes in the samples, as observed for Helicotylenchus and Aphelenchus in our study. ^{Yeates and King (1997)} and ^{Valocká et al. (2001)} compared nematode communities in native pastures with cropped areas and observed that disturbances caused by agricultural practices increased the total abundance and richness of nematodes. This is possibly due to the increase in plant-parasitic nematode population, which could be so small in the soil and not be detected without the presence of the host plant.

Dorylaimida, such as Labronema and Discolaimoides, is cited as abundant in Cerrado soils and little detected in cropped soils (^{Goulart and Ferraz, 2003}); however, in this study, both species occurred in cropped areas, demonstrating low sensibility to agricultural land use. In addition, agricultural crops could increase abundance of bacterial-feeding nematodes, especially due to the presence of organic matter (^{Shaw et al., 2019}) and fungal-feeders, which sporadically parasitize on aerial parts of plants and could be favored or introduced by seeds of the agricultural crops (^{Yeates, 1991}).

Mesocriconema, of the order Tylenchida, was observed in higher abundance in NF, demonstrating that this nematode is more sensitive to disturbances caused by the agricultural practices, such as tillage (^{Cares and Huang, 1991}; ^{Yeates, 1991}; ^{Gomes et al., 2003}). In the Cerrado region of Brazil, nematodes of Criconematidae family have played an important role in distinguishing native from cropped areas (^{Mattos et al., 2008}), since they are very sensitive to disturbances resulting from the replacement of native areas by annual crops (^{Goulart and Ferraz, 2003}). Other genera from Tylenchida were present in all sampled areas, but with great variation in the density values, such as Helicotylenchus, the dominant genus in Tylenchida in this study, possibly favored by soybean cropping in CT and MT (^{Machado et al., 2019}). The same can be concluded for Mononchida, found in all sampled areas and in great variation. These nematodes are predators and supposedly sensitive to environmental changes (^{Niles and Freckman, 1998}; ^{Wasilewska, 1997}); nevertheless, they were not observed in our study.

The Shannon-Weaver and the Bray and Curtis indexes showed high diversity in NF and MT samples and that these areas were more similar. Conversely, CT was more dissimilar to NT. These observations corroborated previous reports in which the number and richness of genera or species sometimes constitutes better tools to evaluate nematode diversity in different ecosystems (^{Mattos et al., 2006}); nevertheless, the efficiency of these parameters is controversy in literature. ^{Freckman and Ettema (1993)} considered richness and the Shannon-Weaver index little useful in distinguishing annual from perennial crops, while ^{Yeates and Bird (1994)} and ^{McSorley and Frederick (1996)} considered these approaches efficient.

In addition, cropped areas, except for MT, have a lower richness than NF. These findings agree with most similar studies in the literature, which report a higher number of taxa in native vegetation than in cropped areas (^{Freckman and Ettema, 1993}; ^{Háněl, 1995}; ^{Yeates and King, 1997}; ^{Mattos et al., 2006}; ^{Goulart and Ferraz, 2003}; ^{Gomes et al., 2003}; ^{Tomazini et al., 2008}).

Regarding the effect of soil chemical attributes on nematode communities, NF was positively correlated to higher Al and C concentrations, whereas MT was correlated to P concentration. Thus, we can infer that high Al concentration is characteristic of this native type of soil where the experiment was conducted, that is, Campos Gerais, characterized by poor and washed soils, deficient in P, Ca, N, and K, with low pH and Al high concentration (Bodziak Júnior and Maack, 2001). The low pH observed in these soils could lead to an increase in the Al toxicity, which might reduce microbial biomass, generally observed after N addition to the soil (^{Shaw et al., 2019}), influencing soil nematode communities, especially those dependent on microbial biomass.

The analysis of feeding groups provides information on changes in decomposition pathways between the sites studied. Abundance of bacterial-feeders could increase due to the microbial activity caused by fertilization or accelerated decomposition (^{Ettema and Bongers, 1993}) and respond to OM input depending on the quantity and quality of OM (^{DuPont et al., 2009}). Nevertheless, several bacteria in agricultural fields can solubilize insoluble forms of P to make them available for plant growth and have been considered as predominant source of P availability (^{Ilmer and Schinner, 1992}). In our samples, bacterial-feeders were also correlated with the P content, possibly due to high concentration of P-solubilizing bacteria.

On the other hand, enrichment opportunist fungal-feeding nematodes react rapidly to the addition of organic materials, because fungal-feeders have high respiration rates and low N needs, mineralizing N to plants (^{Chen and Ferris, 1999}; ^{Ferris et al., 1997}). This information if confirmed by our study where fungal-feeders were correlated with high C concentration. Besides, the higher concentration of C was also correlated with NF samples, reflecting the decomposition pathway of natural forests, predominantly by fungal-feeding nematodes. Besides, the black oat and wheat crops could explain high abundance of Aphelenchus in CT and MT in the winter, since these plant species can host this genus and favor the increase of its populations in the soil (^{Singh et al., 2013}).

Therefore, these results may serve as link between microorganism and energy flow in the soils analyzed, providing a valuable tool to access the nature of decomposition pathways. Population of fungal- and bacterial-feeders are useful for tracking disturbances within a crop system, although the responses of nematode communities to agricultural intensification are not easy to interpret and correlate to soil properties (^{Sánchez-Moreno et al., 2006}). Nematode communities can differentiate diverse habitats or cropping systems, especially in large temporal or land scales experiments (^{Levin, 1992}). Therefore, we observed that nematode communities were capable of grouping MT and CT in a separate cluster from NT and NF, indicating that aggressive cropping systems differ from conservative ones. ^{Popovici and Ciobanu (2000)} also found correlation between the composition of the nematode community and habitat characteristics, as different vegetation and crop systems.

In summary, the analyses of nematode communities successfully reported shifts in land use after 30 years in a field experiment in southern Brazil. Abundance of nematodes demonstrated that MT and NT areas were more similar and CT was more dissimilar to NT. Besides, cropped areas, except under MT, have a lower richness than NF. The PCA and clustering analysis also discriminated the impact of land use in the soil nematode communities, since NF was grouped with NT and MT with CT, based on the soil chemical analysis and nematode communities. These results indicated that nematode diversity is important for long-term stability of soil ecosystem and its use as soil health indicators merits further studies covering larger areas under soil conservation systems, such as NT and intercropping in Brazil.

Figure 1 – Dendrogram illustrating the dissimilarity between treatments (NF = native forest; NT = no-tillage; MT = minimal tillage; CT = conventional tillage) obtained by the UPGMA algorithm, based on the Euclidean distance, calculated from the average of nematode-feeding groups (A) and the soil chemical analysis (B).