Vertical distribution of soil free-living nematode in a playa habitat in the North-Western Negev desert, Israel

Study site (30º53’14’’N, 34º25’15’’E) is located at the Nizzana sand dune site within the Hallamish dune field eastern extension of the northern Sinai dune field – western Negev Desert, Israel (Yair, 1990). The dune field comprises longitudinal dunes aligned west-east, are 15 to 20 m high and are separated by a 50 m to 200 m wide interdune. The Nizzana experimental site is located about 45 km inland from the Mediterranean Sea. Mean annual precipitation is 95 mm and occurs during the winter months (November – April) and the average monthly temperatures are 9 ^oC in January and 25^oC in August. Potential annual evaporation is in the order of 2200 mm (Kidron, 2001; Kidron & Yair, 2001; Keck et al., 2016).

The playa area is a geomorphological inter-sand dune desert site unit comprised of thin, dense layers (1 – 2 mm thick) covered with dark brown clay skins, followed by 3 cm of angular blocky, deeper down, single grained sandy layers, with very few tubules and holes, alternating with loamy units, having a massive blocky structure (Kidron et al., 2009). All layers are nearly horizontal with textural boundaries, indicating a sedimentary origin and playa area soil riches in clay and silt and have low or no vegetation cover (Yair, 1990). The rusty mottles show that anaerobic conditions with stagnant water occur from time to time (Pen-Mouratov et al., 2011). The clayey layers are dense with pore sizes mainly in the range of 10 – 50 μm, thus retarding water infiltration and movement. Carbonate content at such places may exceed 30 %, with an alkaline pH. These geomorphological formations are surrounded by a sandy loessial plain area. This plain area soils have a surface crust, which is classified as a platy surface structure, with shallow cracks 1 – 2 mm deep filled with windblown sand. The lower part of the crust has a vesicular structure. The distribution of vegetation is mainly the perennial Chenopodiaceous species, Anabasis articulate in the inter-sand dune valley (Breckle et al., 2008).

Soil samples were collected at the study site at the end of the summer season, representing the driest period and in the middle of the winter season, representing the wet season, using a 7-cm diameter soil auger. Soil samples from the 0 to 50 cm at 10 cm intervals (corresponding to 0 – 10, 10 – 20, 20 – 30, 30 – 40, and 40 – 50 cm depths, four replicates) were randomly collected from both playa and loessial plain areas (as control). Each of the soil samples was deposited in an individual plastic bag which was placed in an insulated container and transported to the laboratory. These soil samples were kept in cold storage at 4 ^oC until processing. All soil samples were sieved (2 mm mesh size) before biological and chemical analysis in order to remove root particles and other organic debris.

Soil moisture – was determined gravimetrically by drying the samples at 105 ^oC for 48 h. The values were expressed as percentage of dry weight.

Organic matter concentration – was determined by oxidization dry soil sample with dichromate in the presence of H₂SO₄, without application of external heat (Rowell, 1994).

Electrical conductivity – was determined on a 5 g subsample that was agitated (half-hour at room temperature) with 50 ml distilled water. The filtrated solution was analyzed using a TH-2400 conductivity meter (ms cm^-1).

Soil free-living nematodes – were extracted from 100 g fresh soil samples using the Baermann funnel procedure (Cairns, 1960) and nematode populations were expressed per 100 g dry mass soil. The recovered organisms were counted and preserved in formalin (Steinberger & Sarig, 1993). Total one hundred individuals from each samples were randomly picked and used for identification to genus level, according to esophageal and morphology characteristics using an inverted compound microscope (Mal & Lyon, 1975; Ying, 1998).

The data were analyzed using the following approaches: (1) total nematode number in 100 g^-1 dry soil (TN); (2) trophic groups, including bacterivores (BF), fungivores (FF), plant parasites (PP), omnivores–predators (OP) (Yeates, et al., 1993; Liang et al., 1999; Pen-Mouratov et al., 2004); (3) F/B, fungivores/bacterivores ratio (Twinn, 1974); (4) NCR, nematode channel ratio, where NCR=BF/ (BF+FF) (Yeates, 2003); (5) T, trophic diversity, T = 1/Σpi2^, in which pi is the proportion of the i-th trophic group in the nematode community (Heip et al., 1988); (6) H’, Shannon Index, H’ = -Σpi(lnpi), where pi is the proportion of individuals in the i-th taxon (Shannon & Weaver, 1949); (7) Simpson index for dominance (λ), where λ=pi² (Simpson, 1949); (8) SR, species richness, SR = (S1)/ln(N), where S is the number of taxa and N is the number of individuals identified (Yeates & King, 1997); (9) MI, maturity index, MI = Σvipi, where vi, is the c-p value for free-living nematodes assigned by Bongers (1990) to the i-th nematode genus and pi is the proportion of the genus in the nematode community; (10) MMI, modified maturity index, including plant-feeding nematodes, MMI = Σvipi, where vi is the c-p value for free-living and plant parasitic nematodes assigned by Bongers (1990) to the i-th nematode genus and pi is the proportion of the genus in the nematode community (Yeates, 1994); (11) EI, Enrichment Index is calculated as EI = 100(e/(e+b)), where e is the abundance of individuals in guilds in the enrichment component weighted by their respective k_e values and b is the abundance of individuals in the basal component weighted by their k_b values (Ferris et al., 2001; 2004; Ferris & Matute, 2003); (12) SI, Structure Index is calculated as SI = 100(s/ (s+b)), where s is the abundance of individuals in the structure component weighted by their k_s values.

All obtained data were subjected to statistical analysis of variance (ANOVA) using the SAS model (GLM, Duncan’s multiple range test and Pearson correlation coefficient) and were used to evaluate differences between separate means. CANOCO redundancy analysis (ter Braak & Prentice 1996; ter Braak & Smilauer 2002) were used to evaluate the relationship among nematode trophic groups, soil depths and environmental variables. Differences obtained at levels of P < 0.05 were considered significant.

Soil moisture reached a maximum value of 7.2 % in the 0 – 10 cm soil layer at the playa site in the wet season, whereas a minimum value of 0.1 % was found in the dry season in the 0 to 20 cm soil layer in the loessial plain area. Soil moisture in the playa area was significantly (P < 0.05) higher than that in the loessial plain area during the dry and wet seasons (Table 1). Moreover, the univariate analysis elucidated a significant (P < 0.001) effect of the different factors, such as: sampling location, depth, season and interaction between them, except the combination of three of the factors (Table 2).

Soil organic matter content was found to reach a maximum value in the 0 – 10 cm layer in the playa site under dry the season and a minimum in the 20 – 40 cm soil layer in the loessial plain area during the wet season. Moreover, general high values were in the playa soil samples (Table 1). Significant difference in soil organic matter content were found by location (P < 0.0001) and depth (P < 0.01), rather than by seasonality (Table 2).

Soil electrical conductivity was found to follow the soil organic matter trend, with significantly (P < 0.0001) (Table 1) higher values in the playa area during the dry season than that in the loessial plain area during both seasons. Soil electrical conductivity was also significantly influenced by location (P < 0.0001) and depth (P < 0.01) (Table 2).

Total number of nematode in the playa area ranged from 1 to 15 individuals per 100 g dry soil, while total number of nematode in the loessial plain area was many fold higher, ranging from 60 to 631 individuals per 100 g dry soil and resulting in significant (P < 0.0001) differences between the sampling sites (Fig. 1). Moreover, total nematode abundance of every soil layer in the loessial plain area was significnatly (P < 0.0001) higher than that in the playa area during the dry and wet seasons. The upper soil layers were represented by a higher nematode abundance and the number of nematode had gradually decreasing trend with soil depths increasing at both sampling sites and seasons. In general, total number of soil free living nematode was significantly affected by sampling locations (P < 0.0001), depths (P < 0.001) and seasons (P < 0.01) on (Table 2). Total soil free living nematode was found to be significantly negatively correlated with soil moisture, soil organic matter content and electrical conductivity, respectively (Table 3).

Fig. 1

The general pattern obtained for the changes in total soil free living nematode were found to be reflected in the trophic group abundance. Bacterivores were found to be the dominant trophic group in the present study, and they were significantly greater (P < 0.05) in the loessial plain area than that in the playa area in each layer during the dry and wet seasons. In the playa and loessial plain area, bacterivores in the upper soil layer (0 – 10cm) was greater than in the other layers during the dry and wet seasons (Fig. 2). In general, bacterivores were significantly affected by sampling locations (P < 0.0001), depths (P < 0.01) and seasons (P < 0.05) (Table 2). Bacterivores exhibited a significant negative correlation with soil moisture and electrical conductivity (Table 3).

Fig. 2

Fungivores were found in lower numbers, whereas the pattern of their behavior was found to replicate that of the bacterivores, with a relatively higher population size in the loessial plain area than in the playa area in all sampling depths and season (Fig. 2). In the upper 0 – 10 cm soil layer in the loessial plain area, the fungivores had a higher abundance during the dry and wet season. Furthermore, the overall data revealed that the sampling locations (P < 0.0001), depths (P < 0.01) and seasons (P < 0.05) significantly effected on their number (Table 2). Fungivores manifested a significant negative correlation with soil moisture and electrical conductivity (Table 3).

Plant-parasitic nematodes were the least trophic group (Fig. 2). Plant-parasitic nematodes were significantly affected by sampling locations (P < 0.0001) and depths (P < 0.05) (Table 2) and significantly negatively correlated with soil moisture, soil organic matter content and electrical conductivity (Table 3).

Omnivores-predators were found to follow bacteria-feeding and fungi-feeding trophic groups, with a higher abundance in the loessial plain area and in the 0 – 20 cm soil layer (Fig. 2). Omnivores-predators were significantly influenced by sampling locations (P < 0.0001), depths (P < 0.01) and seasons (P < 0.05) (Table 2). Omnivores-predators were significantly negatively correlated with soil moisture, soil organic matter content and electrical conductivity (Table 3).

RDA analysis revealed that BF and OP were mainly distributed within the 0 – 10 cm depth and PP mainly in the 10 – 20 cm and 20 – 30 cm depths in the playa area during the dry season (Fig. 3). BF and PP were mainly distributed in the 0 – 10 cm depth in the playa area during the wet season. All trophic group of nematodes were mainly distributed within the 0 – 10 cm, 10 – 20 cm and 20 – 30 cm in the loessial plain area during the dry season, and correlated positively with soil organic matter content and negatively with soil moisture and electrical conductivity. However, the trophic group of nematodes was mainly distributed within the 0 – 10 cm and 10 – 20 cm in the loessial plain area during the wet season, and positively correlated with soil organic matter content, soil moisture and electrical conductivity.

Fig. 3

Total of 55 genera were observed, including 19 bacterivores, 10 fungivores, 12 plant-parasites, and 14 omnivore-predators (Table 4). Total of thirteen genera (including 6 bacterivores, 1 fungivores, 5 plant-parasites, and 1 omnivore-predators) were found in the playa area, while total of fifty-five genera (including 19 bacterivores, 10 fungivores, 12 plant-parasites, and 14 omnivore-predators) were found in the loessial plain area. Acrobeloides, Metateratocephalus and Acrobeles were the dominant genera in the playa area and were mainly present to the 0 – 20 cm depth, while Acrobeles, Metateratocephalus, Cephalobus, Aphelenchoides, Chiloplacus and Acrobeloides were the dominant genera in the loessial plain area and were also mainly present to the 0 – 20 cm depth.

F/B values at every soil layer in the loessial plain area were higher compared with the playa area during the dry and wet seasons (Table 5). Significant difference in the F/B values was found between locations (P < 0.01), but no significant differences were observed between depths and between seasons (Table 2). F/B values significantly negatively correlated with soil moisture (Table 3).

NCR values at every soil layer in the playa area were higher contrasted with the loessial plain area during the dry and wet seasons (Table 5). Significant difference in the NCR values was found between locations (P < 0.001), but no significant differences were observed between depths and between seasons (Table 2). NCR values significantly positively correlated with soil moisture and electrical conductivity (Table 3).

T values at every soil layer in the loessial plain area were higher compared with the playa area during the dry and wet seasons (Table 5). Significant sampling locations effect was found on the T values (P < 0.0001), but no significant differences were observed between depths and between seasons (Table 2). T values correlated significantly negatively with soil moisture, soil organic matter content and electrical conductivity (Table 3).

H’ values at every soil layer in the loessial plain area were higher than that in the playa area during the dry and wet seasons (Table 5). Significant sampling locations and seasons effects were found on the H’ values (P < 0.0001), but no significant difference in H’ values was observed between depths (Table 2). H’ values significantly negatively correlated with soil moisture, soil organic matter content and electrical conductivity (Table 3).

The λ values at every soil layer in the loessial plain area were lower compared with the playa area during the dry and wet seasons (Table 5). Significant sampling location effect was found on the λ values (P < 0.0001), but no significant differences in the λ values were observed between depths and between seasons. Significant interaction effect of locations and seasons (P < 0.0001) was observed in the λ values (Table 2). The λ values significantly positively correlated with soil moisture, soil organic matter content and electrical conductivity (Table 3).

SR values at every soil layer in the loessial plain area were higher contrasted with the playa area during the dry and wet seasons (Table 5). Significant sampling location effect was found on the SR values (P < 0.0001), but no significant differences in the SR values were observed between depths and between seasons. Significant interaction effect of locations and seasons (P < 0.01) was detected in the SR values (Table 2). SR values significantly negatively correlated with soil moisture, soil organic matter content and electrical conductivity (Table 3).

MI values at every soil layer in the loessial plain area were higher compared with the playa area during the wet season (Table 5). Significant sampling locations effect was found on the MI values (P < 0.0001), but no significant differences in the MI values were observed between depths and between seasons. Significant interaction effects of locations and depths, locations and seasons (P < 0.05) were detected in the MI values (Table 2). MI values significantly negatively correlated with soil moisture, soil organic matter content and electrical conductivity (Table 3).

MMI values at every soil layer in the loessial plain area were higher contrasted with the playa area, except the 30 – 40 cm layer during the wet season (Table 5). Significant sampling location effect was found on the MMI values (P < 0.01), but no significant differences in the MMI values were observed between depths and between seasons (Table 2). MMI values significantly negatively correlated with soil moisture, soil organic matter content and electrical conductivity (Table 3).

EI values in every soil layer in the loessial plain area were higher compared with the playa area (except the 0 – 10 cm layer during the wet season) during the dry and the wet seasons (Table 5). Significant differences in the EI values were found between locations (P < 0.01) and between seasons (P < 0.05), but no significant difference was observed between depths (Table 2). EI values significantly negatively correlated with electrical conductivity (Table 3).

SI values in every soil layer in the loessial plain area were higher contrasted with the playa area (except for the 0 – 10 cm and 40 – 50 cm layers during the dry season) during the dry and wet seasons (Table 5). Significant difference in the SI values was found between locations (P < 0.0001), but no significant differences were observed between depths and between seasons (Table 2). Significant interaction effects of locations and depths (P < 0.05), locations and seasons (P < 0.01) were detected in the SI values. SI values significantly negatively correlated with soil moisture, soil organic matter content and electrical conductivity (Table 3).

Integrating all the data using RDA (Fig. 3) for each of the sampling sites and seasons and soil depths indicated a strong effect between the dry (Fig. 3 A, B) and wet periods (Fig. 3 C, D) of each treatment. The relatively higher values of soil moisture (SM) during the wet period were found to explain the significant difference between the sampling sites. In the playa area (Fig. 3A, B) the SM was found to move to deeper layers, whereas the SM in loessial plain area remained close to the upper soil layer (Fig. 3 C, D). Different behavior of SM between the two sites explain the importance of soil pedological characteristics. The second variable known to play an important role in basic food source availability is the OM whose behavior was found to remain stable in the upper soil layers in both cases. These main variables were found to have the greatest significant effect on the soil free living nematode community abundance, distribution and trophic groups (Fig. 3), where in the playa area the OM was found to have a great effect on TN compared with the loessial plain area. Redundancy analysis reflected that the nematode community mainly distributed in the upper soil layers (0 – 20 cm) under the playa and loessial plain area.