Nematode community structure in relation to soil and vegetation characteristics

Applied Soil Ecology 1 ( 1994) 29-44

Nematode community structure in relation to soil and vegetation characteristics Ron G.M. de Goede*, Tom Bongers Department of Nematology, Wagenrngen Agricultural Unwerslty, PB 8123, NG6700 ES Wageningen, Netherlands (Accepted 14 October 1993)

Abstract To test the hypothesis that nematodes can contribute to an ecological soil classification, the nematode fauna of several Dutch terrestrial habitats was studied. A total of 209 samples from 44 nature reserves or slightly managed sites (n = 94) differing in vegetation (forest, shrubs, heathland, grassland) and soil type (clay, loam, sand) were studied. A selection of sites was studied over four seasons, and at one site variation in nematode fauna composition as a result of the selected sampling technique was studied. Nematodes extracted from bulk soil samples taken from the 0- 10 cm depth mineral soil, and were identified to genus. Multivariate analysis techniques were used to classify the nematode samples into seven sample groups (SG) as described by soil characteristics in combination with the vegetation as follows: ( 1) SG A grasslands, dwarf-shrub vegetation and forest gaps on sandy soils; (2) SG B grasslands and forests on clayey soils; ( 3) SG C-D deciduous forests on sandy-loam soils; (4) SG E-F deciduous forests on sandy soils; (5) SG G coniferous forests on sandy soils. The nematode fauna of SG D-G were very similar, and were dominated by ten taxa: Acrobeloides, Aphelenchoides, Cephalenchus. Filenchus A, Filenchus B, Plectus A, Prismatolaimus, Rhabditidae, Tylolaimophorus and Wilsonema. Variation due to seasonal fluctuations and sampling technique, was small compared with differences in nematode fauna structure between different sites. The actual vegetation of some sites was not in agreement with the natural vegetation expected on ‘site characteristics’. Analyses of the nematode fauna supported the observed inconsistencies between actual and natural vegetation. It was shown that for a range of terrestrial habitats nematode communities could be defined, and that these communities could be related to soil characteristics and vegetation. Key words: Classification; Community structure; Forest; Grassland, Nematodes; Soil type; Seasonal fluctuations

1. Intruduction

Awareness of the nature and extent of soil pollution has resulted in an increasing effort to research the consequences of soil pollution on soil organisms and the corresponding effects on soil *Corresponding author at: Biological Station of the Wageningen Agricultural University, Kampsweg 27, NL9418 PD Wijster, Netherlands. Tel: 31 5936 2441, Fax: 31 5936 2786. 0929-1393/94/$07.00 0 1994 SSDZ0929-1393(93)00001-A

processes. A result has been the practical examination of the possibilities of biological soil assessment systems to indicate the ecological condition of a soil. Nematodes possess many qualities which makes them suitable as bioindicator organisms for soil quality (Samoiloff, 1987; Freckman, 1988; Bongers 1990). Moreover, as free-living soil nematodes are represented at most trophic levels of the food web, they are thought to be

Elsevler Science B.V. All rights reserved

30

R GM

de Goede, T Bongers /Applied Sorl Ecology I (I 994) 29-44

closely connected to, and may reflect, fundamental ecological processes (e.g. decomposition, mineralisation, nutrient cycling) in soils. In arable soils, changes in the composition of the nematode fauna, were to be related to changes in activity and/or biomass of the microflora, and to microfloral mediated processes such as nutrient dynamics and nitrogen mineralisation (e.g. Sohlenius et al., 1987; Brussaard et al., 1990; Ettema and Bangers, 1993). Effects of disturbance on ecological processes in soils are expressed by functional or structural changes in the nematode community. At the community level changes in the composition of the nematode fauna have been found to indicate effects of pollutants (Cantelmo and Rao, 1978; Cantelmo et al., 1979; Sturhan, 1989; Bangers et al., 1991; Weiss and Larink 1991), liming (Hyviinen and Huhta 1989; Ratajczak et al., 1989; Hyvonen and Persson 1990; De Goede and Dekker 1993), acidification (Hyvonen and Persson 1990; Ruess and Funke 1992; Ruess et al., 1993 ), agricultural practices (Freckman and Ettema, 1993 ) and recovery after disturbance (Yeates et al., 1991; Ettema and Bongers 1993). Analysis of the composition of the nematode fauna could serve as a basis for ecological assessment of soils. Depending on the specific goals of the study, either direct interpretation of the taxa composition or evaluation of derived indices may be used. In monitoring studies, results can be compared with earlier investigations and observed changes can be interpreted. As with most indices, the results of nematological indices and their interpretation, must be related to the type of habitat studied. A reference system, which describes the ecological characteristics of soils (an ecological soil classification), could serve as a basis to assess the condition of a previously unsampled soil. If the aim of an ecological soil classification reference system is to provide a reference for a system to assess the biological quality of the soil, then the assessment and the reference system should be based on the same organisms (Bongers, 1990). Such an ecological soil classification reference system should, therefore, include habitat factors (i.e. abiotic and biotic habitat characteristics which determine the soil fauna com-

munity (Sinnige et al., 1992) ) which apply to the assessment organisms. The aim of the present investigation was to study relationships between the composition of the nematode fauna of several Dutch terrestrial habitats, their biotic and abiotic environment, and to indicate the possibility of using the nematode fauna to classify habitats into groups with similar soil characteristics. Ninety-four sites representing a variety of vegetation and soil types were sampled. Multivariate analyses were used to group samples with related nematode faunae. Vegetation and the physical and chemical soil properties of these sample groups (SG) are described, and the effects of seasonal fluctuations in nematode fauna composition on the classification are studied.

2. Materials and methods 2.1. Sites, sampling and uolation

A total of 209 samples from 94 different sites located at 44 nature reserves and rural estates were studied (Fig. 1) . In September 1985, 46 sites representing a variety of vegetation types (grassland, heather, shrubs, pine and deciduous forest and outer marshes) and soil types (clay, loam and sand) were sampled. Eight of these sites were also sampled at 3-monthly intervals and one of those was sampled ten-fold on 22 May 1987. From April to May 1988 another 121 samples were collected at 48 forest sites. These sites were selected on the basis of the potential distribution of the forest types within the Netherlands, to represent examples of ‘native’ forest types on sandy and sandy-loam soils (Van der Werf, 199 1) . Detailed descriptions of the topographic position and physical chemical characteristics of the sites are given in Bongers et al. (1989) and Van der Berg et al. ( 1990 ) . At each site a representative area (one in 1985 and 1987; two to three in 1988) of 100 m2 was selected, and 50 cores ( 17 mm diameter) were taken from the O-10 cm depth mineral soil following a zigzag pattern. Because of the imprac-

R.G.M. de Goede, T. Bongers /Applred Sod Ecology 1(1994) 29-44

31

Fig. 1. Geographical distribution of the 44 nature reserves and rural estates ( l ) that were studled.

ticability of homogenizing organic and mineral horizons and because of large differences in methodology for optimal nematode extraction from such horizons (Schouten et al., 1990; Schouten and Arp, 199 1 ), inclusion of both the organic horizons and the mineral soil in one bulk sample was avoided. The 0- 10 cm depth mineral soil was selected, because, in contrast to the organic horizon, it is present under most vegetation types found in the Netherlands. The cores were bulked and packed in polythene bags and stored at 4” C. Within 1 week after sampling, each bulk sample was homogenized and sub-samples were taken for nematode extraction and physical and chemical analysis. Nematodes were isolated using an Oostenbrink elutriator (Oostenbrink, 1960 ), fixed,

transferred to glycerin and identified from mass slides (Bongers et al., 1989) according to Bongers ( 1988). Nematode taxa lists for samples collected before 1988 are presented in Bongers et al. (1989), and those from 1988 are given in De Goede (1993). The physical and chemical analyses of the soil followed methods as described in Lagas et al. ( 1986) and Houba et al. ( 1989). Fresh soil was homogemzed and a sub-sample was extracted with 1M NaCl, whereafter concentrations of NO3 and NH4 (mg kg- ’ ) were determined using an autoanalyser. Water content (expressed as percentage fresh mass) was determined after drying at 105 “C. The remaining soil was air-dried at 40 oC and sieved through a 2 mm mesh sieve. The pH of KC1 and HZ0 were measured in soil : water

32

R G M de Goede, T Bangers / Apphed Sod Ecology I(1 994) 29-44

( 1: 5 v/v ) suspensions of distilled water and 1 M KCI, respectively. Total organic carbon (TOC, %) was measured spectrophotometrically as the proportion COZ produced after dry combustion at 900°C. The percentage of CaCO, was measured as the percent weight loss after addition of 4N HCl. To determine the cation exchange capacity (CEC, meq 100 g- ’ ), soil was leached with a 1M ammonia acetate solution (pH 7)) whereafter the absorbed ammonium was exchanged with MgCl, and measured with an autoanalyzer. Concentrations (meq 100 g- ’ ) of exchangeable Na, K, Ca and Mg were determined only in soils without CaC03. They were measured in the leachate of the CEC determination by atomic absorption spectrophotometry. Base saturation (%) was calculated as 100 x the ratio between total concentration of cations and CEC. Granular composition (fractions < 2 and 2-38pm) was determined by sedimentation velocity, after removal of organic matter, Fe- and Al-oxides, CaCO, and soluble salts. The > 38 pm fraction was calculated using these two measured granular fractions. Total nitrogen (mg kg-’ ) was measured calorimetrically by the indophenol blue method (Novozamsky et al., 1974), after destruction with sulphuric acid, selenium and hydrogen peroxide. All data are expressed on the basis of dry mass, unless indicated otherwise. The soil analyses were carried out at the National Institute of Public Health and Environmental Protection (RIVM ) Bilthoven, and all samples from 198 5 and a selection of the samples from 1988 (i.e. at least one sample per site) were analyzed. Vegetation releves were made only of the sites sampled in 1988; the plant nomenclature is according to Van der Meijden et al. ( 1990 ) . 2.2. Numerical analysis Multivariate analyses were based on families, genera and on the following lower taxa: ( 1) Filenchus A: small with filiform tail cf. Filenchus helenae; Filenchus B: small with short rounded tail cf. Filenchus ditissimus; Filenchus C: large with filiform tail cf. Filenchus vulgaris. (2) Hemicycliophora A: stylet < 110 pm; Hemicy-

cliophora B: stylet > 110 ,um; ( 3 ) Paratylenchus A: stylet < 17 pm; Paratylenchus B: stylet 22-33 pm; Paratylenchus C: stylet 44-66 ,um; Paratylenchus D: stylet > 70 pm; (4) Plectus A: c’> 8; Plectus B: c’< 8; Plectus C: cf. Plectus puslllus (long stoma) ; (5 ) Trzpyla A: cephalic setae > 0.25 cephalic diameter; Tripyla B: cephalic setae co.25 cephalic diameter; (6) Paramphidelus A: c= 3-4; Paramphidelus B: c= 22-27; (7) Criconematidae (excluding Hemicnconemoides) species: Crlconematidae A: stylet < 50 pm; Criconematidae B: stylet 50-84 pm; Criconematidae C: stylet > 84 pm. The data set which included the nematode taxa with their relative abundance, was analyzed using the clustering programme FLEXCLUS (Van Tongeren, 1986; Jongman et al., 1987) followed by detrended correspondence analysis (DCA ) using the computer program CANOCO (Ter Braak, 1987; Jongman et al., 1987). In the analyses relative abundance was transformed as follows: 2”% becomes (n+ 1 )%, with n=O-7. In both programmes the option ‘downweighting of rare species’ was used. These procedures served to normalize the distribution of relative abundance and reduced effects of relatively rare species. In DCA, detrending by second-order polynomials was used. Successive detrended correspondence analyses were necessary to interpret the structure of the data set. Each sequential analysis was primarily based on the same data set, but from which distinctive groups of samples detected in the graph of the first and second axes of earlier analyses were excluded (Peet, 1980; Verdonschot, 1990). SG were composed of one or more FLEXCLUS clusters from which the site scores for the DCA axes 1 and 2 did not differ significantly. Analysis of variance tested for significant differences. Similarities between nematode samples were calculated with the similarity ratio (SR) (e.g. Jongman et al., 1987) SR,, =cYI,Yk,/

(&Y$

+&Y&

-&Y,Y,)

with Yin and Y, being the proportion of taxon Y in sample i and j, respectively. Prior to calculation, data were transformed as described above.

33

R.G M. de Goede, T. Bongers /Applied Sod Ecology I (1994) 29-44

Differences in chemical and physical characteristics of soils representing the SG (based on nematode faunae) were tested using the MannWhitney U-test with PC 0.05.

SG A can be characterized as grass and dwarfshrub vegetation and forest gaps on sandy soils. SG B comprises grasslands and forests on mainly clayey soils. The SG C-D and E-G comprise forest habitats on sandy-loam soils and sandy soils, respectively, and silt fraction decreases in the sequence C-F. SG C-F mainly comprise deciduous forests, whereas most of the sites of SG G are coniferous forest. Table 1 gives the frequency of occurrence of the common taxa for each SG. Detailed soil chemical and physical characteristics of the seven SG and their vegetation analysis, are given in Tables 2-4.

3. Results 3. I. Multivariate analyses

We ended up with a data matrix of 209 samples and 130 nematode taxa. Using FLEXCLUS, 22 different clusters could be distinguished. By two subsequent DCA, these 22 clusters were combined into seven SG (Figs. 2 and 3 ) . SG A (containing 23 samples) and SG B ( n = 16 ) split off on axis 1 and 2 of the first DCA (Fig. 2). The eigenvalues of these axes were 0.303 and 0.157, respectively. The third axis had an eigenvalue of 0.089. Based on the first two axis of the second DCA, the remaining samples ( n = 170) were subdivided into another live SG (Fig. 3). The eigenvalue of the first three axes of the second DCA was 0.159,O. 106 and 0.080, respectively.

3.2. Nematode taxa In Table 1 the occurrence of 43 nematode taxa with a constancy 30.50 in at least one SG is given. SG A, B and C differed greatly from other SG as could be seen in the DCA, where they occupied the most extreme positions (Figs. 2 and 3). The nematode faunae of SG D-G, however, were similar. Ten taxa (Acrobeloides, Aphelen-

axis 2 120

100

so 00 40 20 0

-a 40 a0 a0 -100 -120

80

-20

20

60

100

140

180

220

200

axis 1 Fig. 2. Detrended correspondence analysis diagram for the 209 nematode samples. Samples are represented by letters corresponding to the SG to which they belong. The contour lines describe the total vanatlon of each SG. For clarity SG C-G are represented only by their contour lines. See text for charactenstlcs of SG.

R.G.M de Goede, T Bongers IApplIed Sod Ecology I (1994) 29-44

34

axis 2 90 So 70 60

50 D 40 30 20 10 0 -10

-20 -30 40 -50 -60 -70

80

Fig. 3. &trended correspondence analysis diagram for the 170 samples of the SG C-G. The samples of the SG A and B were not included in the analysis. See Fig. 2 for details. Note that scales differ from those m Fig 2.

choides, Filenchus A, Filenchus B, Tylolaimophorus, Cephalenchus, Wilsonema, Prismatolaimus, Rhabditidae and Plectus A) were present in

at least 70% of the samples of SG D-G. Moreover, 79%, 98%, 95% and 96% of the samples of respectively SG D, E, F and G had nematode fauna consisting of more than 50% of these ten taxa. Characteristic taxa for SG B were Neopsilenthus, Boleodorus, Paratylenchus B, Pro/Mesadorylaimus, Anatonchus, Pungentus, Mylonchulus, Tripyla A, Anaplectus, Aphelenchus and Coslenchus. Their occurrence in the other SG was

always at constancy < 0.30 and mainly restricted to SG A and C. Basiria and Cephalobus were also characteristic for SG B, but they occurred in other SG with constancies ~0.30. The nematode faunae of SG B (grasslands and forests on clayey soils ) were most similar to SG A and C (Fig. 2 ) . Genera with maximum occurrence in SG B and C were Cephalobus, Heterocephalobus and Rotylenchus (Table 1) and they were present in forests as well as in grasslands. However, Qudsianematidae, Wilsonema, Steinernema and

Cervidellus, which were common in all other SG, were almost lacking in SG B and C. Pratylenchus and Panagrolaimus had their highest constancy in SG A and B (Table 1) . Pratylenchus was present in both forests and grasslands, whereas Panagrolaimus occurred only in grasslands and one Scats Pine forest (with an unbroken herb layer of Deschampsiajlexuosa) . Plectus B was the only taxon characteristic for SG A, B and C together. The nematode fauna of the forested sites reflected gradual changes going from SG C via D to SG E or to SG F and G (Fig. 3 ) . Rotylenchus, Alaimus, Diphtherophora, Plectus C and Trichodoridae were characteristic taxa of SG C (Table 1). The characteristic taxa of SG B and C were absent or showed a decreasing constancy in SG D-G, whereas other taxa (e.g. Cephalenchus, Tylolaimophorus, Acrobeloides, Wilsonema, Steinernema and Cervidellus) showed an oppo-

site trend. SG E was an exception to the trend shown in Fig. 3. Some taxa (Drilocephalobus, Monhystera, Prismatolaimus, Qudsianematidae and Plectus A) which were important in the D-F-G se-

R GM. de Goede, T. Bongers /Applred Sod Ecology 1(1994) 29-44 Table 1 Frequency of occurrence

(constancy,

%)’ of the common* nematode

taxa for the seven SC Taxon

Paratyienchus A Panagrola~mus Aglenchus Paratylenchus Plectus B Paratylenchus B Boleodorus Pungentus Neopsdenchus Coslenchw Bawrra Rotylenchus Pkwur c Pro/Mesodoryiarmus Cephalobw Alarmus Dzphtherophora Heterocephalobus Rhabdttulae Malenchus Tnchodondae Tylencholalmus Hehcotylenchus Fdenchus C Cnconematudae C Paratylenchus C Cephalenchus Dnlocephalobus Monhystera Tylolalmophorus Fdenchus A Fdenchus B Acrobeloldes Aphelencholdes Dttylenchus Teratocephalus Prismaiolaimus Metateratocephahu Qudsranematudae Wrlsonema Steinernema Cenvdellus Plectus A

SC A

B

C

D

E

F

G

(23)

(16)

(14)

(34)

(50)

(44)

(28)

70 57 52 57 96 22 22 26 0 4 13 0 0 22 39 39 22 48 74 17 26

13 38 13 63 63 75 69 50 69 50 81 63 0 94 1100 25 19 75 94 31 6 6 19 81 19 13 6 6 0 0 6 69 63 75 75 19 50 13 44 13 13 0 6

0 7 29 14 71 0 0 0 7 7 36 100 50 29 64 57 57 71 100 71 71 57 36 64 79 71 64 14 21 64 100 79 50 100 79 57 86 79 29 29 14 21 57

3 3 3 9 24 6 0 0 0 0 3 38 6 3 18 18 3 44 79 53 26 9 38 65 76 65 91 41 50 68 91 97 100 94 56 41 85 44 74 56 26 24 74

2 8 4 4 12 6 2 0 2 0 2 6 0 4 12 10 6 10 62 56 18 10 10 18 58 20 90 22 14 78 88 96 100 98 90 56 40 54 54 64 68 52 40

4 0 2 11 11 0 0 0 0 0 0 22 2 2 2 7 0 27 78 47 18 27 20 40 62 22 98 51 84 93 100 100 100 98 58 93 93 96 73 96 71 53

0 4 4 11 25 0 0 0 4 4 0 0 11 0 7 4 4 18 75 61 29 32 25 14 7 0 7 7 43 75 64 86 100 96 57 68 93 82 79 93 71 75

91

100

48 52 4 9 0 9 0 4 35 70 43 100 87 65 52 70 57 65 65 61 57 43

‘Number of occurrences of a taxon in a SC/number of sues m that SC ‘Taxa Hrlth constancy t 50%, the total number of sues are gtven between brackets

35

quence, had exceptionally low constancies in SG E (Table 1). However, Ditylenchus reached maximum constancy (0.90) in SG E, whereas its constancy in SG D, F and G was less than 0.58. Plectus A, Wilsonema and Prismatolaimus reached maximum constancies in the forested sites on sandy soils (SG F and G) . Despite great similarity in composition of the nematode faunae of these SG, some clear differences existed. For example Tylolaimophorus, Teratocephalus, Metateratocephalus, Monhystera and Drilocephalobus had maximum constancy in SG F. Drilocephalobus, Criconematidae C and Cephalenchus all occurred in more than 50% of the sites of SG F, but had constancies of only 0.07 in SG G (Table 1). Many of the taxa with high constancies in the forests on sandy soils (SG D-G), also occurred in SG A (dwarf shrubs, grasslands and gaps in forests). The three taxa which were rare in SG G but common in SG F, were absent or rare in SG A. Taxa characteristic for SG A were Paratylenthus A, Aglenchus, Panagrolaimus, Criconematidae A, Aporcelaimellus and cf. Tripius (Table 1). 3.3. Seasonal sampling The seasonal replicates of the eight sites (S l8) which were sampled at 3-monthly intervals, as well as the ten replicates from the site which was sampled at one occasion (T 1), fell into single clusters and were thus classified within the same SG. Three such sites were assigned each to SG A and E, two to SG B, and one to SG C (Table5).Eachofthereplicatesamples (Sl-S8,Tl) had highest similarity with a replicate of the same site. The average similarity between seasonal replicates of the same site (intra-site similarity) ranged from 0.609 to 0.793, whereas the average similarity between the Tl replicates was 0.853. Compared with these intra-site similarities, average similarities between the replicate samples and samples collected at other sites that were classified within the same cluster or same SG (inter-site similarity), tended to be lower, and ranged from 0.414 to 0.645 and 0.224 to 0.508, respectively. From five of the eight sites sampled

R G.M. de Goede, T. Bongers / Apphed Sod Ecology I (I 994) 29-44

36

Table 2 Mean physnzal and chemtcal charactenstrcs

of O-l 0 cm mmeral sod for each of the seven SG

SG -_ A

B

Texture (%) <2firn [2; 38> 238nm

2.0d 3.5d 94.5d

44.9a 38.la 17.0a

10.6b 23.8ab 65.6b

9 8b 22 2b 68 Ob

5 6c 12 8c 81 5c

1 8de 5.lde 88 6de

1 le 1.6e 97 3e

96 Water pF2 (%) pF4.2 (%)

23.7cd 24.3~ 7 oc

52.9a 53.la 23.4a

35.lb 43.3b 13.5b

33 7b 44.5ab 15 4ab

32.9bc 38.6b 10.7b

16.2d 8.5abc 4 2abc

150d 24.0~ 7.3c

TOC (%)

2 8b

5 6a

4.0a

5 5a

4 5a

2 5b

1.8b

pH(KC1) pH(HzG) %CaCO,

5.6b 6.0b 0.7b

6.7a 7.3a 5.9a

3.8~ 4.8~ O.Ob

3.2d 4.0d O.Ob

2 9e 3 7e 0 3b

3 2d 4 Id 0 Ob

3 3d 4.ld 0 Ob

CEC Base (%)’ K’ Ca’ Mg’

7.3c 35.0bc 0.26b 18.83d 18.15b

41.4a 84.0a 0.45a 26.16a 5.10a

14.lb 39.4b 0.27b 4.05d 0.91c

15 lb 9.9d 0 20b 1 15b 0 33d

12.6b 17 led 0 16b 6.05bd 0 31de

6.4~ lO.ld 0.08~ 0.26~ 0.19de

5.4c 13 5cd 0.05c 0.51bc 0.16e

NO, NO3 NH4 N(tot) CN

C

D

5 lab 2.9a 2988a 14.8a

F

E

6.0a 3.2a 2810a 20.4b

5 2a 1 3a 1901ab 22 2b

G

2.6b 0 8b 1237b 20 9b

0.7b 0 2b 339c 28.7b

‘Not determined for samples contammg CaCO,. %Water, so11water content at sampling, TOC. total orgamc carbon, K, Ca, Mg and CEC (catton exchange capacity) in meq 100 g- ‘; Base, base saturation; NO3 and N( tot ) III mg kg- ’ (dry weight ), stgmticant differences (PC 0.05, Mann-Whitney U-test) are mdtcated by dtfferent letters m the same row

over the season (S3-S7) all five replicates had highest inter-site similarity with samples classified within the same SG (Table 5). Furthermore, from three of these sites ah five replicates were most similar to sites within the same cluster (Table 5). From two sites (Sl and SS), most of the replicates (four and three, respectively) had highest inter-site similarity with samples classified within other SG ( SG G and SG F for S 1 and S8, respectively). Only one of the ten Tl replicates had highest inter-site similarity with a sample classified within another SG (SG F) . The relative abundance of most taxa varied between the sampling occasions, and taxa occurring in relatively low numbers were not detected in some seasonal samples of a site. The presence of each nematode taxon was related to its indi-

vidual maximum relative abundance. Eightyeight percent of the abundant taxa (i.e. any taxon reaching 10% in any season at a particular site ) were detected in all seasons at their respective localities; if less abundant taxa (5%) are considered, 66% were detected in all seasons. Thus, although variation in nematode faunal composition because of population dynamics and/or sampling error occurred, the similarity between the samples collected from the same site exceeded those from other, but related, sites. 3.4. Environmental characteristu of the SG SG A (dwarf shrubs, grasslands and gaps in forests) comprised soils with a sand content

37

R.G M de Goede, T. Bongers /Apphed Sol1Ecology l(l994) 29-44 Table 3 Frequency of vegetation types’ in SG A-G Vegetation type

SG B

A

Cladonro-hnetum Leucobryo-Anetum Empetro-Pmetum Empetro-Betuletum Convallario-Betuletum ConvallarwQuercetum dunese Betulo-Quercetum roborls Betulo-Quercetum molinitosum Fago-Quercetum molimtosum Degradated Fago-Quercetum Fago-Quercetum petraeae typrcum Mlro-Fagetum Luzulo-Fagetum Stellario-Carpinetum lonrceretosum Stellario-Carprnetum Melrco-Fagetum Pruno-Fraxmetum Fraxmo-Ulmetum typlcum Alrunon gluttosae Sabcetum Grassland Others (heathland,

forest gaps)

C

D

F

E

G

6 7

2 (6) 1 1

9 7 8 2 2

8 4

5 (9)

5

1

1

8

2 (16)

9

16 (20)

3

8 1 1 1

1

9

6 3 2 1

1

2 1 (5) 1

1

2 4 (8) 5 (9)

4 (8)

1 3

2

Total number of sites (::)

(1:)

(::)

(Z)

(Z)

(Z)

‘According to Van der Werf ( 199 1); the number of sites mcludmg seasonal replicates are gven between brackets.

> 89% and silt content < 4.1%. Soil chemical and physical characteristics were comparable with those for SG F and G, with only pH and nutrient concentrations differing significantly (Table 2 ) . According to the criteria defining SG, SG A could have been subdivided into three separate SG (grasslands; Scats pine forests with De&ramp siuflexuusu (L. ) Trin. dominating the herb layer; Cduna heather, Corynephoretum dune and gaps in forests where grasses dominated). However, as SG A comprised only 11 sites (23 sampling events), subdivision was not made. Only one grassland on sandy soil was not included in SG A, but in SG G. None of the nematode taxa characteristic of SG A (Table 1) were present at that site. The physical and chemical characteristics of that site were: texture classes <2~, 2-38~ and 238,~m, respec-

tively 5.7,4.4 and 90.0%; water content, 26.8%; pF2 and pF4 respectively 39% and 9%; TOC, 2.2%; pH (KCl) and pH (H,O), respectively 3.5 and 4.2; carbonate content, 0.0%; base saturation, 17%; CEC and concentration K, Ca and Mg respectively 7.4 meq 100 g-l, 0.06 meq 100 g-l, 0.84 meq 100 g-l and 0.13 meq 100 g-l. SG B consisted of clayey soils (including one loamy soil) with an average clay content of 44.9% (Table 2) and a sand content < 33%. Soil water content at sampling, moisture retention at pF2 and pF4.2, concentrations of K, Ca and Mg, CEC and base saturation were signifkantly higher than in other SG. Owing to the presence of carbonates in six of the nine sites, average soil pH was highest for SG B ( pH-KCl= 6.7 ) . Both grasslands and forests were present in SG B (Table 3), which were not separated in different clusters. The for-

38

R. GM. de Goede, T Bangers /Applied Sod Ecology 1(1994) 29-44

Table 4 Relative frequency of occurrence (O-10%) of a selection of higher plants and ferns in SG C-G. forests on sandy-loam to sandy sol1 Species

SG C

D

E

F

G

6

32

20

43

21

Anus sylvestrzs Betula spp Quercus spp. Fagus sylvatlca Acer pseudoplatanus Corylus avellana

0 33 83 50 17 33

5 34 81 31 19 34

0 15 15 95 35 15

14 37 84 44 23 14

95 33 24 5 0 0

Rhamnus frangula Sorbus aucuparra Crataegus spp. Ilex aquifohum Prunus spp. Sambucus mgra

0 17 17 17 17 33

16 50 13 9 38 19

0 40 0 30 40 20

14 60 21 26 33 16

10 19 5 0 5 0

Empetrum mgrum Deschampslaflexuosa Lonicera periclymenum Vaccimum myrtrllus Polygonatum multljlorum Pterrhum aqurbnum Rubus spp. Mlrum effusum Moluua caerulea Dryopterzsspp. Hedera hehx Oxalis acetosella Anemone nemorosa Galeobdolon luteum

0 0 33 0 10 0 17 17 0 50 67 67 67 33

0 13 31 16 6 19 59 16 38 28 16 16 9 9

0 5 25 15 1.5 25 70 25 0 10 40 25 10 5

9 42 40 42 12 16 40 5 9 26 21 9 5 9

43 52 19 0 0 0 5 0 0 5 0 0 0 0

N

N IS the number of vegetation descnptlons.

ests were Salicetum, Alnetum and Fraxino-UImetum typicum (Table 3 ) . The observed overlap of SG C-G in the ordination diagram of the second DCA (Fig. 3 ) also occurred in the chemical, physical and floristic characteristics (Tables 2-4). SG C and D were composed of sites with sandy-loam soils with comparable soil texture, soil water content, pF2, pF4.2, total organic carbon content, CEC and concentrations of nitrogenous compounds and K. Compared with SG D, concentrations of Ca and Mg, base saturation and soil pH of SG C were

significantly higher, whereas the C: N-ratio is lower. Except for soil texture and pH, the soil chemical and physical characteristics of SG E were very similar to SG C and D (Table 2). Soil texture of SG E was intermediate to the sandy-loam soils of SG C and D and the sandy soils of SG F and G. Average bulk soil pH of SG E was the lowest found in this study with pH (KCl) = 2.9. SG F and G comprised sandy soils with an average sand fraction > 93% and only differed in total nitrogen concentration, being lower for SG G (Table 2 ) . Compared with the other forested SG (B-E), soil water content, pF, total organic carbon, CEC, NO3 : NH,-ratio and K, Mg, NO3 concentrations were lower in SG F and G. The observed gradient in soil chemical and physical characteristics for SG B-G, corresponded with changes in the composition of the vegetation of these groups (Tables 3 and 4). Pinus sylvestris L. was present in 95% of the sites of SG G, but was rare in the other SG. Within SG G, 75% of the sites were classified as Dicrane-pmton. These forests had a very speciespoor herb layer, with only Deschampsiajlexuosa and Empetrum nigrum L. present. ConvallarioQuercetum dunense, Betulo-Quercetum roboris and forested former heathlands (i.e. degraded form of Fago-Quercetum petraeae) were characteristic forest types of SG F. Sorbus aucuparia L., Lonicera periclimenum L. and Vaccinium myrtillus L. were found more frequently in this SG as compared with the other SG. SG E mainly consisted of Fago-Quercetum petraeae typicum and Milk-Fagetum. The dominant tree species Fagus sylvaticaL. was present in 95% of the sites, and Rubus occurred in 70% of the sites. SG D is a transitional stage between SG E-F and C, with moist sub-associations of Fago-Quercetum petraeae and Betulo-Quercetum roboris or relatively floristic rich forest types from the Fagetalia sylvaticae. Similar to SG E and F S. aucuparia, Prunus, Rubus and fewer also Ptertdium aquilinum (L.) Kuhn were found frequently in SG D. The Fago-Quercetum petraeae molinietosum and Betulo-Quercetum roboris molinietosum sites were typified by the presence of Molinea caerula (L. ) Moench. The Stellario-

39

R.G.M de Goede, T. Bongers /Applred Sod Ecology l(l994) 29-44 Table 5 Intra-site similarity for sites sampled over seasons (S-S8) and a site which was sampled ten-fold at one occasion (Tl site similarity between these sites and samples of other sites classtfied withm the same cluster or same SG Site code

SG

CN’ N2

Intra-site similarity

Mean

Sl

A

5

5

S2

A

6

8

S3

A

7

7

s4

B

1

11

S5

B

1

11

S6

C

9

10

S7

E

18

34

S8

E

20

5

Tl

E

18

34

0.785 (0.735-0.920)5 0.708 (0.596-0.874) 0.609 (0.490-0.701) 0.758 (0.706-0.856) 0.744 (0.681-0.806) 0.689 (0.556-0.795) 0.793 (0.643-0.924) 0.650 (0.452-0.807) 0.853 (0.758-0.953)

), and inter-

NSG’

NC?

0.0813

1

-

0.0876

4

4

0.1215

5

3

0.044 1

5

5

0.0720

5

5

0.0534

5

5

0.1545

5

4

0.1272

2

-

0.0990

9

8

Inter-site similarity with Samples of same cluster

Samples of same SG

Mean

Mean

s.d. s.d.

0.0520 0.0758 0.0747 0.0439 0.0468 0.0777 0.0940

0.515 (0.285-0.752) 0.414 (0.277-0.550) 0.443 (0.290-0.631) 0.446 (0.290-0.580) 0.525 (0.366-0.646) 0.645 (0.481-0.793)

0.1425 0.0922 0.0898 0.0734 0.0828 0.0706

0.1078 0.0389

0.576 (0.115-0.816)

0.1479

0.361 (0.226-0.575) 0.303 (0.11 l-0.483) 0.308 (0.102-0.575) 0.224 (0.163-0.291) 0.321 (0.206-0.469) 0.382 (0.288-0.475) 0.523 (0.172-0.794) 0.416 (0.083-0.754) 0.508 (0.267-0.754)

s.d.

‘Cluster number, ‘Number of samples within cluster (all replicate samples of a site fell into single clusters); 3number of replicate samples which had highest similarity to a sample of another site within the same SG (N is 5 and 10 for respectively Sl-S8 and Tl ) ; ‘number of replicate samples which had highest similarity to a sample of another site within the same cluster; ‘range; S 1: Leucobryo-Pinetum, S2: Calluna vulgans heather, S3: grassland, S4: Fraxmo-Ulmetum typtcum, S5: grasland, S6: StellarroCarpmetum, S7: Fago-Quercetum petraea typwm, S8 and S9: degraded Fago-Quercetum; see text for details.

Carpinetum and Pruno-Fraxinetum sites of SG C had a species rich shrub and herb layer with Anemone nemorosa L., Hedera helix L., Oxalis acetosella L. and Dryopteris spec. being characteristic plant species. The forest types present in SG B were unique to this group. Generally each SG contained several forest types, but most of the forest types were restricted to, or had their optimum of occurrence in, one of the SG (Table 3). However, the forest type Milio-Fagetum was dominant in three different SG. In conclusion, the results showed that for a range of terrestrial habitats nematode communities can be defined, and these communities can be related to soil characteristics and vegetation. Soil texture, pH and vegetation structure are

found to be important discriminating characteristics.

habitat

4. Discussion Studies such as the present work do not permit elucidation of causal relationships between nematodes and biotic or abiotic environmental factors. However, stepwise regression with the sample scores of the DCA as the dependent variable, showed soil texture, pH and water content are best correlated with nematode populations. They explained > 82% of the variation present. In several studies (Yeates, 1968, 1974; Johnson et al., 1973; Baujard et al., 1979; Scotto La Massese and Boulbria, 1980) soil water characteristics were found to be of overriding importance

40

R G M de Goede, T Bongers /Appired Sod Ecology 1 (I 994) B-44

for the occurrence of nematode species. Because of the strong co-correlations between most environmental variables, no single factor can be selected in the present work. Moreover, differences in soil characteristics coincided with differences in vegetation. Especially when soil characteristics were comparable, further classification of the nematode fauna seemed to depend on vegetation structure and related differences in e.g. nutrient dynamics and microclimate, as was shown for the sandy soils. Grass and dwarf-shrub habitats were separated from forested sites with comparable soil characteristics. Within the grass and dwarf-shrub habitats, the nematode fauna of the grasslands differed most from those of the forested sites, whereas the heathlands and patches under canopy gaps lay in between. The occurrence of some nematode species, however, probably can be related directly to soil morphological properties. Yeates ( 1980, 198 1) observed correlations between the occurrence of nematode species with relative large labial probolae or cephalic setae and soil macroporosity. Comparable results were found in the present study for Wilsonema, Cervidellus, Teratocephalus and Metateratocephalus, which had maximum occurrence in the coarse-textured soils. These correlations are thought to be related to locomotion and/or feeding requirements of the nematode species (Yeates, 198 1; De Ley, 199 1) . Except for root-feeding nematodes, causal relationships between occurrence and habitat are unclear for most other species. Based on the composition of the nematode fauna, soils could be classified into clusters (SG) which could also be described by combinations of soil physical, chemical and floristic properties. As in many classifications, transitions between clusters were gradual, and this may hamper classification of specific samples, especially for samples similar to SG D-G (forests on sandyloam to sandy soils ) . The apparent clear distinction between SG A, B and C-G, probably reflects the selection of only a few SG A- and Bsites ( 11 and eight sites, respectively), as opposed to 149 sites for the SG C-G. Extending the number of forested clayey and loamy soils and

grassland sites m the classification, will probably make the transitions between the SG also more gradual. Moreover, such an intensified sampling could also result in a further sub-division of SG A and B. Nevertheless, taking into account the relatively large differences in soil properties and vegetation type present in SG D-G, the high similarity of the nematode fauna of those SG is remarkable. This agrees with the findings of Johnson et al. ( 1972 ) who found, despite differences in pedology and forest type, high similarities ( > 60%) between 15 of the 18 forests studied. Wasilewska ( 1970) assumed that poor natural habitats on sandy soils in Northern and Central Europe are characterized by similar nematode species. Such high similarities between geographically isolated sites with similar vegetation and edaphic factors were found also by Norton and Hoffmann ( 1974) for plant parasitic nematodes in the USA. In the present study the high similarity between the nematode faunae of different vegetation types on sandy soils, could be related to the trend of decreasing pH in this series of soils. These forested sandy soils had very low pH, with SG E having an average pH (KCl) of less than 3.0 (Table 2 ) . In De Goede and Dekker ( 1993 ) artificial acidification of forest soils resulted in a decrease of nematode taxa that, based on the nematode Maturity Index scaling (Bongers, 1990 ), were sensitive to environmental disturbance. Compared with the other SG, taxa that were sensitive to environmental disturbance were scarce in many sites in SG E (De Goede et al., 1994). The nematode fauna of SG E was dominated by taxa thought to be relatively stress tolerant (De Goede et al., 1994). Moreover, in the samples of SG C-G there was a significant (P,< 0.0 1) positive correlation between pH and the proportion of taxa that were sensitive to environmentally disturbance, and in addition SG E was the only SG in which fungal feeding nematodes were more abundant than bacterial feeding nematodes. One of the applications of an ecological soil reference classification system may be the assessment of the condition of soils. Although only nature reserves or lightly managed soils were

R GM. de Goede, T. Bongers / Applred Sod Ecology I (1994) 29-44

studied, examples of the potential of the nematode fauna composition as an indicator was demonstrated. In the Scats pine forests (SG G), the nematode fauna may indicate that the vegetation of certain sites was not the natural vegetation. Jordana et al. ( 1987), also found signiflcant differences in nematode fauna structure between natural and exotic vegetation on comparable soil types. Two sites (Hernense Bos and Wageningen forest) were placed in SG A, between heathlands and grass-covered gaps in forests and grasslands. The actual vegetation of both forests (Scats pine forest with a herb layer dominated by Deschampsia jlexuosa) was comparable with several forests within SG G. However, the soil characteristics of both forests were inconsistent with those characteristic for a Dicraw-pinion (Van der Werf, 199 1). The natural vegetation of Wageningen forest is thought to be a degraded form of Fago-Quercetum petraeae (Van der Werf, personal communication, 1989). Thus, although the nematode fauna of both sites possessed characteristics of the Dicrano-Pinion (Cephalenchus and Criconematidae C absent, Plectus A present ), the presence of nematodes typical of grasslands (Paratylenchus A, Helicotylenchus) and dwarf-shrub vegetation (Aglenthus) indicate inconsistencies between actual and natural vegetation on these soils. At Hernense Bos many plant parasitic nematode taxa characteristic to SG A (Paratylenchus A, Aglenchus, Pratylenchus, Helicotylenchus and Hemicriconemoides) were found. This may be related to the low vitality of the stand. These examples showed that an ecological soil reference system which is based on habitat factors that apply for nematodes, can be used as a reference to assess soil quality. Based on the habitat factors found in this study, the Hernense Bos and Wageningen forest would have been classified into SG F-G. Subsequent comparison of the nematode fauna of Hemense Bos and Wageningen forest and these SG then would have indicated the deviant status of these sites. Data on the sensitivity of nematode community classifications to seasonal fluctuations are scarce. Jordana et al. ( 1987 ), who compared soil fauna populations of different vegetation on

41

comparable soil types, showed that seasonal fluctuations in the composition of the nematode fauna were minor to differences due to vegetation type. Similar findings were reported by Yeates ( 1984), who concluded that site was more important than month in determining population composition of grazed pastures. In our study differences in nematode fauna structure owing to seasonal fluctuations and sampling strategy, also appeared to be relatively small compared with differences in nematode fauna structure between different sites. All replicate samples (seasonal replicates and the replicates from a site sampled on one occasion) had highest similarity with the replicates of the same site, and for seven sites most of the replicates had highest inter-site similarities with samples classified within the same SG. However, the latter was not the case for two sites that were sampled over season (Sl Hernense Bos, S8 Oak forest Wageningen) . The high similarity between Hemense Bos and samples within SG G can be explained by the deviant status of this site (see discussion above). The three ‘deviant’ replicates of Oak forest Wageningen (degraded Fago-Quercetum ) had highest intersite similarities with sites of related vegetation types ( Fago-Quercetum petraea typicum, degraded Fago-Quercetum) and an Empetro-Betuletum, which further demonstrates the high similarity of the nematode fauna of those SG. Restricted sampling of only O-10 cm depth mineral soil resulted in the exclusion of the principal part of the nematode fauna of soils with a mor or moder humus (most soils of SG D-G). The nematode distribution in the soil profile largely depends on the distribution of organic material (Yeates and Coleman, 1982). In mor and moder humus, highest nematode abundances were found in the fermentation horizon of the organic layer (Popovici, 1980; De Goede et al., 1993), whereas in mull humus highest numbers occurred in the upper centimetres of the mineral soil (Popovici, 1980). The exclusion of the organic horizon will, as was shown in this study, probably not be inconvenient in a habitat classification to assess groups of soils with similar ecological requirements. Also Arpin and Ponge ( 1986) showed that the nematode fauna

42

R GM. de Goede, T Bongers /Applied Sod Ecology l(l994) 29-44

of mineral soil horizons possessed discriminating abilities between habitats. They compared the composition of the nematode fauna of different soil horizons of a Scats pine (Pinus syhesfris), an oak (Quercus petraea (Mattuschka) Lieblein ) and a mixed pine-oak forest. Differences between these forests were best revealed by differences in nematode fauna structure of the deeper soil layers, corresponding to the 6- 10 cm depth mineral soil of the Scats pine forests. It should be noted however, that the similarity between corresponding horizons of the different forests studied by Arpin and Ponge ( 1986) appeared to be larger than the similarity among the different horizons within each forest. Similar results were obtained in De Goede et al. ( 1993), where nematological changes during primary succession of Scats pine forest were studied. Inclusion of the nematode fauna of these organic horizons may contribute to a more refined classification of SG D-G. The abundance distribution of nematode species in soil samples is often skewed, with a majority of species occurring in relatively low numbers (e.g. Wasilewska, 1976; Platt et al., 1984, personal observation). Several such relatively rare species (e.g. those in the genera Boleodorus, Pungentus, Coslenchus, Basiria and Diphtherophora) (Bongers et al., 1989), appeared important to the classification. However, as the detection level of a species depends on the relative abundance of the species, the position of relatively rare species in a classification needs special attention. Besides, many nematodes show high dispersion abilities (Orr and Newton, 197 1; Krnjaic and Krnjaic, 1972; Proctor, 1984), and can therefore be found also outside their optimum home range. Wilsonema, a genus restricted to sandy soils, was found in one of the grasslands with a clayey soil, and Criconematidae were reported from submerged sediments in an oil harbour (Tamis, 1986). The classification described in this study was based on all nematode taxa occurring in the samples. However, in recent studies (Zell, 1989; Ettema and Bongers, 1993 ) it was concluded that the abundance of species within the Rhabditidae in some situations depended on brief periods of

increased food availability, rather than habitat structure. Similar relationships can be expected also for species of Diplogasteridae and Panagrolaimidae which also are characterized as opportunists (Bongers, 1990 ). During such periods of increased food availability these taxa can make up > 50% of the total nematode fauna. Extreme dominance of these species in soil samples can also be the result of accidental occurrence (e.g. a rabbit dropping or a small animal corpse in a core). The insect parasitic genus Steinernema, the juveniles of which were found frequently in both grasslands and forests on sandy soils also shows pockets of abundance; in infected insects numbers in excess of 800 000 individuals of Steinernema have been recorded (Poinar, 1983). Although opportunistic nematodes have important potentials as indicators of environmental conditions (Bongers, 1990; Ettema and Bongers, 1993), their value to an ecological soil classification is debatable (Bongers and Korthals, unpublished data, 1993 ) .

5.

Acknowledgements

The authors like to thank H.H.B. van Megen for technical assistance, the members of the Department of Nematology for assistance in the field, S. van der Werf for being helpful with the selection of forest sites, Prof. Dr. A.F. van der Wal and Dr. G.W. Yeates for discussion and improving the English and the owners of the Dutch nature reserves for their permission to collect samples.

6. References Arpm, P. and Ponge, J -F , 1986. Influence d’une lmplantatlon rkcente de pin sylvestre sur le comportement de la nkmatofaune du sol, par comparalson avec un peuplement femllu pur et un peuplement mClangC. Pedobiolopa, 29: 39 l-404. BauJard, P., Comps, B. and Scotto la Massese, C., 1979. Introduction B I’ktude Bcologique de la ndmatofaune tellurlque du masslf landaus (France). Rev. Ecol. Biol Sol, 16: 61-78.

R.G.M. de Goede, T. Bongers /Applzed Sod Ecology l(l994) Bongers, T., 1988. De Nematoden van Nederland. KNNV Bibliotheekuitgave nr. 46, Pirola, Schoorl. Bongers, T., 1990. The Maturity Index: an ecological measure of environmental disturbance based on nematode species composition. Oecologia, 83: 14- 19. Bongers, T., De Goede, R.G.M., Kappers, F.I. and Manger, R., 1989. Ecologische typologie van de Nederlandsebodem op basis van de vnjlevende nematodenfauna. RIVM rapport 718602002, Bilthoven. Bongers, T., Alkemade, R. and Yeates, G.W., 1991. Interpretation of disturbance-mduced maturity decrease in marine nematode assemblages by means of the maturity mdex. Mar. Ecol. Prog. Ser., 76: 135-142. Brussaard, L., Bouwman, L.A., Gems, M., Hassink, J. and Zwart, K.B., 1990. Biomass, composition and temporal dynamics of soil organisms of a silt loam soil under conventional and mtegrated management. Netherlands J. Agric. Sci., 38: 283-302. Cantelmo, F.R. and Rao, K.R., 1978. Effects of pentachlorophenol (PCP) on metobenthrc communities established in an experimental system. Mar. Biol., 46: 17-22. Cantelmo, F.R., Tagatz, M.E. and Rao, K.R., 1979. Effect of barite on meiofauna in a flow-through expenmental system. Mar. Environ. Res., 6: 301-309. De Goede, R.G.M., 1993. Terrestrial nematodes m a changing environment. Ph.D. Thesis, Agric. Univ., Wageningen. De Goede, R.G.M. and Dekker, H.H., 1993. Effects of hming and fertilization on nematode commumties in coniferous forest soils. Pedobiologia, 37: 193-209. De Goede, R.G.M., Bongers, T. and Ettema, C.H., 1994. Graphical presentation and interpretation of nematode community structure: c-p triangles. Meded. Fat. Landbouwwetenschappen, University of Gent, m press De Goede, R.G.M., Georgieva, S.S., Verschoor, B.C. and Kamerman, J.W., 1993. Changes m nematode community structure m a primary succession of blown-out areas m a drift sand landscape. Fundamental Appl. Nematol., 16: 501-513. De Ley, P., 1991. The nematode community of a marginal soil at Camberene, Senegal, with special attention to functional morphology and niche partitioning in the family Cephalobidae. Meded. Kon. AC. Wet. Lett. Sch. Kunst. Belgie, 53: 108-153. Ettema, C.H. and Bongers, T., 1993. Characterization of nematode colonization and nematode succession in disturbed soils using the Maturity Index. Biol. Fertil. Soils, 16: 79-85. Freckman, D.W., 1988. Bacterivorous nematodes and organic-matter decomposition. Agric. Ecosystems Environ., 24: 195-217. Freckman, D.W. and Ettema, C.H., 1993. Assessing nematode communities in agroecosystems of various human intervention. Agric. Ecosystems Environ., 45: 239-26 1. Houba, V.J.G., van der Lee, J.J., Novozamsky, I. and Walmga, I., 1989. Soil and Plant Analysis - a senes of syllabi, part 5: Soil analysis procedures. Wagenmgen Agncultural University, Wageningen.

29-44

43

Hyviinen, R. and Huhta, V., 1989. Effects of lime, ash and nitrogen fertilizers on nematode populations in Scats pine forest soils. Pedobiologia, 33: 129-143. Hyviinen, R. and Persson, T., 1990. Effects of acidification and liming on feeding groups of nematodes in coniferous forest soils. Biol. Fertil. Soils, 9: 205-210. Johnson, S.R., Ferris, V.R. and Ferris, J.M., 1972. Nematode community structure of forest woodlots. I. Relationships based on similarity coefficients of nematode species. J. Nematol., 4: 175-183. Johnson, S.R., Feni, J.M. and Ferris, V.R., 1973. Nematode community structure of forest woodlots. II. Ordmation of nematode communities. J. Nematol., 5: 95-107. Jongman, R.H.G., ter Braak, C.J.F. andvan Tongeren, O.F.R., 1987. Data Analysis m Community and Landscape Ecology. Pudoc, Wageningen. Jordana, R., Arbea, J.I., Moraza, L., Montenegro, E., Mateo, M.D., Hemandez, M.A. and Herrera, L., 1987. Effect of reafforestation by conifers m natural biotopes of middle and South Navarra (Northern Spain). Revue Suisse Zool., 94: 491-502. KmJaic, D.J. and Km~ax, S., 1972. Dispersion of nematodes by wind. Boll. Lab. Entomol. Agrana Filippo Silvestri Portici, 30: 66-70. Lagas, P, van der Berg, S., van Dordrecht, P. and MestersBakhuys, H., 1986. Voorschriften voor bodemanalyse. RIVM Rapport 847221001, Bilthoven. Norton, D.C. and Hoffmann, J.K., 1974. Distribution of selected plant parasitic nematodes relative to vegetation and edaphic factors. J. Nematology, 6: 8 l-86. Novozamsky, I., van Eck, R., van Schouwenburg, J.C. and Wahnga, I., 1974. Total nitrogen determmation m plant material by means of the mdophenol blue method. Neth. J. Agnc. Sci., 22: 3-5. Oostenbnnk, M., 1960. Estimating nematode populations by some selected methods. In: J.N. Sasser and W.R. Jenkins (Editors), Nematology. The University of North Carolma Press, Chapel Hill, NC, pp. 85-102. Orr, C.C. and Newton, O.H., 1971. Distnbution of nematodes by wmd. Plant Dis. Reporter, 55: 61-63. Peet, R.K., 1980. Ordination as a tool for analyzing complex data sets. Vegetatio, 42: 171-174. Platt, H.M., Shaw, KM. and Lambshead, P.J.D., 1989. Nematode species abundance patterns and their use in the detection of environmental perturbations. Hydrobiologia, 118: 59-66. Popovici, I., 1980. Distnbution and dynamics of soil nematodes m mixed and spruce fir forest ecosystems. Rev. Roum. Biol. Biol. Anim., 25: 17 1- 179. Poinar, G.O., 1983. The National History of Nematodes. Prentice-Hall, Englewood Chffs, NJ. Proctor, D.L.C., 1984. Towards a biogeography of free-living nematodes. I Changing species nchness, diversity and densities with changing latitude. J Biogeog., 11: 103- 117. Ratajczak, L., Funke, W. and Zell, H., 1989. Die Nematodenfauna emes Fichtenforstes: Auswukungen anthropogener Einfltlsse. Verh. Ges. Gkol., 17: 391-396. Ruess, L. and Funke, W., 1992. Effects of experimental aci-

44

R G M de Goede, T Bongers /Apphed Sod Ecology I(1 994) 29-44

dificatton on nematode populations m soil cultures. Pedobtologta, 36: 231-239. Ruess, L.. Funke, W. and Breumg, A., 1993. Influence of expertmental acnhficatton on nematodes, bacteria and fungi: soil macrocosms and field experiments. Pedobiologra, 120: 189-199. Samotloff, M R., 1987. Nematodes as mdtcators of toxic environmental contaminants. In: J.A. Veech and D W Dickson (Editors), Vistas on Nematology. E.O. Painter. DeLeon Sprmgs, FL, pp. 433-439. Schouten, A.J. and Arp, 1991. A comparative study on the efftciency of extraction methods for nematodes from dtfferent forest litters. Pedobiologa, 35: 393-400. Schouten, A.J., van Breemen, E.M. and Slob, W., 1990. Esttmatmg nematode abundance m natural sandy forest sod: a comparison of two sot1 samplers. Pedobiologia, 34: 239246. Scotto la Massese, C and Boulbrta, A.. 1980. Essai d’mterpretatton ecologrque de la nematofaune de la for& landaise Ann. Sci. Forest, 37: 37-5 1. Smmge, C.A.M., Tamis, W.L.M. and KhJn, F., 1992 Indelmg van bodemfauna in ecologtsche soortengroepen. CML Report 80, Section Ecosystems and Environmental Qualtty, Letden Sohlemus, B., Bostrom, S. and Sandor, A , 1987. Long-term dynamics of nematode commumties m arable soil under four cropping systems. J. Appl. Ecol., 24: 13 1- 144. Sturhan, D., 1989. Nematodes as potential indicators of heavy metals m natural systems. In. A.F. Wal and R.G M. de Goede (Editors), Nematodes in Natural Systems: A Status Report. Meded. 199, Agricultural University, Wagenmgen, p. 41. Tamts, W.L.M., 1986. Nematoden m Vlaardmgs havenshb en m papterslume m het Apeldooms kanaal. Hydr. Adv. Bur. Klmk, Wagenmgen, Rapp. Meded., 22. Ter Braak, C.J.F., 1987. CANOCO - a FORTRAN program for canonical community ordmation by [partial] [detrended] [canonical ] correspondence analysis, prmctpal components analysrs and redundancy analysts (version 2.1) ITI-TNO, Wagemngen, 95 pp. Van Tongeren, O., 1986. FLEXCLUS, an interactive program for classificatton and tabulation of ecologtcal data Acta Bot. Neerlandica, 35: 137-142. Van den Berg, S., De Goede, R.G.M. and Kappers, FL, 1990

Fysisch-chemtsche analyses van bodemmonsters en topografische beschnJving van bemonsterde locattes t.b.v het proJect Bodemclassrficatte. RIVM rapport 7 188 1900 1, Btlthoven. Van der Meyden, R , Weeda. E J , Holwerda, W.J. and Hovenkamp, P.H., 1990. Heukels’ Flora van Nederland. Wolters-Noordhoff, Gronmgen. Van der Werf, S , 1991. Natuurbeheer m Nederland 5; Bosgemeenschappen Pudoc, Wageningen. Verdonschot, P EM., 1990. Ecological charactenzatton ot surface waters m the provmce of Over&e1 (The NetherPh.D. Thesis, lands). Agncultural University, Wageningen Wasilewska, L , 1970. Nematodes of the sand dunes m the Kampinos Forest. I. Species structure Ekologa Polska, 18: 429-443. Weiss, B. and Larmk, O., 1991 influence of sewage sludge and heavy metals on nematodes m an arable soil. Biol Fertil. Sods, 12: 5-9 Yeates, G.W., 1968. An analysts of annual vanation of the nematode fauna m dune sand, at Hrmatangi beach, New Zealand. Pedobiologta, 8. 173-207 Yeates, G W., 1974 Studies on a chmosequence of sods m Tussock grasslands. 2. Nematodes. N Z. J. Zool., 1. 17 l177. Yeates, G.W., 1980 Populations of nematode genera m sods under pasture. III. Vertical dtstrtbutton at eleven sites N.Z.J Agm. Res, 23. 117-128 Yeates, G.W , 198 1 Populations of nematode genera m sods under pasture IV Seasonal dynamics at live North Island sues. N.Z. J. Agnc. Res ,24. 107-I 21. Yeates, G.W., 1984. Vartatton m sot1 nematode diversity under pasture wtth soil and year. Sot1 Btol. Btochem , 16 95-102. Yeates, G W and Coleman, D C., 1982 Role of nematodes m decomposttton. In: D.W Freckman (Editor), Nematodes m Soil Ecosystems Universtty of Texas Press, Austin. TX, pp 55-80 Yeates, G W., Barnforth, S S , Ross, D.J , Tate, K.R. and Sparlmg, G.P., 199 1. Recolomzatton of methyl bromide stenhzed soils under four different field condrtions. Btol Fertil. Soils, 11: 18 1- 189. Zell, H , 1989. Lebensraum Buchenwaldboden 13. Dte Nematoden. Verh. Ges. Gkol 117 125-l 30.