Linking forest dynamics to richness and assemblage of soil zoological groups and to soil mineralization processes

Linking forest dynamics to richness and assemblage of soil zoological groupsand to soil mineralization processes

Sandrine Salmon a,*, Lorenzo Frizzera b, Sylvaine Camaret c

a Museum National d’Histoire Naturelle, Departement Ecologie et Gestion de la Biodiversite, USM 301, UMR CNRS 7179, 4 Avenue du Petit-Chateau, 91800 Brunoy, Franceb Fondazione Edmund Mach, Centro di Ecologia Alpina, 38040 Viote del Monte Bondone, Italyc Universite de Savoie – Chambery - Laboratoire d’Ecologie Alpine (LECA) Campus scientifique, 73376 Le Bourget-du-Lac Cedex, France

Forest Ecology and Management 256 (2008) 1612–1623

A R T I C L E I N F O

Article history:

Received 8 January 2008

Received in revised form 27 June 2008

Accepted 9 July 2008

Keywords:

C mineralization

Diversity of zoological groups

Forest dynamics

Humus forms

Invertebrate communities

Spruce cycle phase

A B S T R A C T

We conducted a study in a spruce forest, grown on a sub-acidic bedrock, in the Italian Alps in order to

assess whether (1) forest dynamics influences animal communities, and in particular whether the

richness of zoological groups peaks during the regeneration phase, (2) the diversity of zoological groups is

correlated to C mineralization, considered as a measurement of soil functioning, (3) aspect influences

the above relationships. We compared soil animal communities, soil physico-chemical features and

nutrient mineralization in three developmental phases of spruce, with increasing tree cover (clearing,

regeneration and mature trees) and two sun exposures (North, South). Animal communities changed

with spruce dynamics. Mature spruce stands were characterized by higher densities of Acari, while

regeneration stands and clearings were mainly characterized by higher densities of Collembola and most

groups of macrofauna. As hypothesized, the richness of zoological groups was highest in regeneration

stands, especially in the south facing site, probably because of the simultaneous occurrence of a dense

herbaceous cover and spruce litter, leading to higher local soil heterogeneity. However, zoological group

diversity (Shannon index), which was lowest in mature stands, was better explained by the herbaceous

cover, i.e., by the quality of food resources, in both south and north facing sites. Variations of soil

characteristics with the developmental phase of trees, reflecting a higher litter input and slower litter

decomposition rate beneath mature trees, are in line with the distribution of zoological groups. As

expected, the diversity of zoological groups was positively correlated to C mineralization. Changes in

animal communities with phases of the forest cycle were much more pronounced in the south compared

to the north facing site. In light of previously published results, we discuss how the diversity and

composition of soil animal communities are plant driven.

� 2008 Elsevier B.V. All rights reserved.

Contents lists available at ScienceDirect

Forest Ecology and Management

journal homepage: www.elsev ier .com/ locate / foreco

1. Introduction

Identifying patterns and determinants of biological richness is amajor theme of community ecology and has a fundamentalimportance to ecosystem management and preservation ofbiological diversity (Schwartz et al., 2000). A related scientificallyand politically important question concerns the ecological con-sequences of biological diversity; that is, is there a relationshipbetween biological diversity and ecosystem functioning? Tradi-tionally, there are three factors that are believed to determinediversity in terrestrial and aquatic communities: resource supply,environmental heterogeneity (or habitat variety) and disturbance(Rosenzweig, 1995; Gaston, 2000; Bardgett, 2002; Bardgett et al.,

* Corresponding author. Tel.: +33 1 60 47 92 11; fax: +33 1 60 46 57 19.

E-mail address: [email protected] (S. Salmon).

0378-1127/$ – see front matter � 2008 Elsevier B.V. All rights reserved.

doi:10.1016/j.foreco.2008.07.009

2005; Dufour et al., 2006). The heterogeneous nature of the soilenvironment offers diversified resources and habitats, leading tothe coexistence of a high number of species (Jordana et al., 2000;Bardgett et al., 2005) provided that quantity and quality ofresources be favorable (Wardle, 2005). Recently, Bardgett et al.(2005) argued that the pattern of soil biodiversity is primarilyrelated to the heterogeneous nature of the soil environment atdifferent spatial and temporal scales. We may assume that soilheterogeneity could vary with phases of the forest cycle (e.g.,clearing, regeneration and mature trees) in lightly managed forestecosystems, since humus forms change with the age of trees(Bernier and Ponge, 1994; Salmon et al., 2006). This environmentalheterogeneity would be maximum in the intermediate succes-sional phase, which is the more diversified phase as a result ofhighest plant diversity (Connell, 1978). A diverse plant communitymay not only result in a heterogeneous habitat but also provide adiverse resource supply for soil invertebrates. If we consider the

mailto:[email protected]

http://www.sciencedirect.com/science/journal/03781127

http://dx.doi.org/10.1016/j.foreco.2008.07.009

S. Salmon et al. / Forest Ecology and Management 256 (2008) 1612–1623 1613

three main phases of natural dynamics of spruce forests (clearing,regeneration and mature trees), we may assume that theregeneration phase will constitute an intermediate level in whichthe richness of zoological groups should be at its highest level, dueto the simultaneous presence of herb layer and spruce litter thatincreases local soil heterogeneity and diversifies resource supply.

Several studies assessed changes in abundance and diversity ofinvertebrate species, as well as changes in species composition ofcommunities with phases of the forest cycle (Hagvar, 1982;Baguette and Gerard, 1993; Bernier and Ponge, 1994; Migge et al.,1998; Zaitsev et al., 2002; Chauvat et al., 2003; Trofymow et al.,2003; Grgic and Kos, 2005). However, each of these studies focusedon the species diversity of only one or a few animal taxa, so thatonly five taxa were studied separately in tree stands of various age:Acari, Collembola, Carabidae, Lumbricidae and Chilopoda. Severalimportant soil-dwelling invertebrates such as Araneida, Diptera,Diplopoda, Symphyla and Enchytraeidae were not studied. Paquinand Coderre (1997) studied multiple taxa to the exception ofCollembola and Acari, which account for a large part of forestarthropod communities. We may assume that functional differ-ences between species, resulting from ecological requirements, lifehistory-traits and size of organisms, are greater when speciesbelong to different zoological groups, than when they belong to thesame group, and that variations in assemblage and diversity ofzoological groups may impact more greatly soil functioning, e.g.,mineralization rates than variations at the species level of onegroup. Moreover, many microcosm experiments suggested thatthe functional composition of soil invertebrate communities ismore closely related with process rates than species richness(Vedder et al., 1996; Schwartz et al., 2000; Loreau et al., 2001;Cortet et al., 2003). Although many field and experimental studiesindicated that a large number of species is required to sustainecosystem functioning, at least aboveground (Naeem et al., 1996;Tilman et al., 1996; Loreau et al., 2001), several researchersemphasized that the relationship between species richness andfunction may be variable, negative or non-existent (Schwartz et al.,2000; Wardle et al., 1997). In particular, Setala et al. (2005) statedthat some animal species in soil communities are functionallyredundant because of the plasticity of their trophic regime, whichcould explain why soil functioning is not systematically correlatedto animal species richness. Consequently, considering the speciesdiversity of just one or a few number of taxa is not the betterapproach to the study of functional diversity. The diversity of allzoological groups may be a better and simpler mean to correlateinvertebrate diversity to soil processes such as mineralizationrates.

We may assume that changes in richness, diversity orassemblage of soil zoological groups with forest dynamics couldimpact soil functioning. Only four studies have related function,i.e., the decomposition of organic matter and nutrient cycling, toforest dynamics (Bernier and Ponge, 1994; Bauhus et al., 1998;Salmon et al., 2006, 2008). Bernier and Ponge (1994) observedchanges in humus forms, namely the result of interactions betweenvegetation and animal/microbial activity in the soil (Rusek, 1975;Kubiena, 1955), along a spruce chrono-sequence. More preciseanalyses of mineralizing activity were undertaken by Bauhus et al.(1998), who concluded that microbial C and N declined with standage as a result of a decreasing quality of soil organic matter inmature tree stands. Salmon et al. (2006, 2008) observed higher C/Nratio and C and N contents in mature than in regeneration stands ofspruce forests growing on acidic and calcareous bedrocks. Thesetwo studies (Salmon et al., 2006, 2008) also showed that theassemblages of soil zoological groups changed with developmentalphases of spruce on both acidic and calcareous bedrocks. Acomparison with the pattern observed here, on a sub-acidic

bedrock, will allow to ascertain whether the observed relation-ships can be generalized for different geological and climateconditions. As for humus forms, soil function and the compositionof invertebrate communities are affected by microclimate condi-tions (Toutain, 1987; Ponge, 1993). The effects of aspect on therelationships between forest dynamics, mineralization rate andanimal diversity have thus to be explored.

In the present study, our objectives were: (1) to determinewhether and how the diversity and the composition of soilinvertebrate communities are correlated with spruce forestdynamics and, more specifically, to examine the relationshipsbetween the richness of zoological groups and the diversity ofresources and habitats, assumed to peak in the regeneration phaseas a result of the simultaneous presence of spruce litter and denseherb layer; (2) to assess whether the mineralization rate, used as ameasurement of soil functioning, and humus forms parallelchanges in invertebrate communities, and may be explained bythe diversity of zoological groups; (3) to determine howenvironmental factors, such as aspect, influence the aboverelationships.

2. Material and methods

2.1. Study sites and sampling

Soil samples were taken at the beginning of October 2003 in amanaged, productive spruce forest (95% Picea abies, with a lowpercentage of Larix decidua and Pinus cembra) in the SoutheasternItalian Alps, located near the Lake Paneveggio in the Fiemme Valley(Province of Trento, Italy). Two sites were selected, 2.3 km apart(4681805400N, 1184401900E and 4681704500N, 1184500800E, respec-tively), at an altitude of 1750 m, on a sub-acidic bedrock and facingnorth and south, respectively. The two sites were located on slopesof the same valley, one directly opposite the other.

The management of these sites was conducted following‘naturalistic’ silviculture, which means that wide clearcuts werereplaced by small clearings about 200–1000 m2 area. Naturalregeneration of Picea occurs in these small areas, sometimes evenunder mature trees.

The substrate of the north facing site is derived from themoraine of the glacier of Pale di S. Martino Mountains, Permianvolcanic rocks, and Verrucano lombardo (a red, coarse-grained,crimbling, fluvial deposit). The mean annual temperature is 3.9 8C,and the mean annual rainfall 782 mm. The substrate of the southfacing site is moraine with acid post-Hercynian molasses, coveredwith mixed moraine. The most common component was volcanicrock, except in half of the clearing and regeneration areas wherecalcareous marls/calcareous schists also occurred. The meanannual temperature in this site is 4.4 8C, and the annual rainfall1103 mm.

In each site, sampling was completed in three contiguous areascorresponding to three developmental phases of spruce: regen-eration, mature trees and clearing. Characteristics of the six areasare given in Tables 1 and 2. Eight sampling points, (i.e., eightpseudoreplicates) were selected randomly in each area, except inthe north facing clearing area where four pseudoreplicates weredone. At each sampling point, distant from 3 to 9 m from eachother, two soil cores were extracted using a polystyrene crystalrectangular box 4.2 cm � 8 cm � 11.3 cm Lxlxh. Each sample corecontained organic and upper mineral layers and was used for: (1)collection of arthropods, respiration measurements, leachateanalysis, determination of soil carbon and nitrogen contents andsoil pH; (2) extraction of enchytraeids. Samples were transportedin their polystyrene crystal box to the laboratory and kept at 4 8Cfor 24–72 h until respiration measurement and leachate collection,

Table 1Tree age, and size and slope of the six sampling areas

Area Size (m2) Slope (8) Age of spruce (years)

Mean Minimum Maximum

South facing clearing 150 16 – – –

South facing regeneration 171 16 22 15 30

South facing mature trees 250 15 163 115 211

North facing clearing 205 23 5 4 6

North facing regeneration 205 23 20 15 30

North facing mature trees 190 20 130 110 145

Table 2Dominant plant species of the herbaceous layer with their mean relative cover and

mean herbaceous cover in the six sampling areas

Area Dominant plant species

(mean relative cover)

Mean

herbaceous

cover (%)

South facing clearing Calamagrostis villosa 50% 95

Melica nutans 41%

Senecio cacaliaster 16%

Cirsium erisithales 14%

Aposeris foetida 13%

Petasites albus 10%

South facing regeneration Melica nutans 40% 90

Calamagrostis villosa 35%

Aposeris foetida 15%

Cirsium erisithales 14%

Carex fer. australpina 12%

Aconitum lamarckii 11%

Petasites albus 10%

South facing mature trees Calamagrostis villosa 58% 26

Vaccinium myrtillus 31%

Deschampsia flexuosa 15%

Melampyrum sylvaticum 5%

Solidago vir. virgaurea 5%

North facing clearing Calamagrostis villosa 75% 90

Deschampsia cespitosa 28%

Senecio cacaliaster 19%


Thelypteris phegopteris 13%

Luzula sieberi 13%

Picea abies 13%


North facing regeneration Calamagrostis villosa 66% 91

Deschampsia cespitosa 30%



Picea abies 14%

Thelypteris phegopteris 13%

Luzula sieberi 13%

Carex frigida 11%

North facing adult trees Vaccinium myrtillus 75% 59

Calamagrostis villosa 26%


Oxalis acetosella 8%

Homogyne alpina 5%

Picea abies 5%

Vaccinium vit. vitis-idaea 5%

We estimated the cover percentage of species following Gallandat et al. (1995) and

Gillet (2000).

S. Salmon et al. / Forest Ecology and Management 256 (2008) 1612–16231614

which were performed at 15 8C. The soil volume treated wasadequate to rely precisely soil respiration measurements to thecomposition of invertebrate communities but was smaller thanthose commonly sampled for macrofauna.

Crystal polystyrene boxes that contained one set of soil cores (1)had a wide aperture on the top (covered with a nylon gauze of35 mm mesh to prevent invertebrates from escape), and a 5 mmhole at the bottom (also covered with gauze) to collect leachates.Respiration measurements and leachate release (by irrigating soilcores with deionized water) were performed, respectively, 4 and 8days after sampling on the soil cores contained in polystyrenecrystal boxes. These cores were then taken out of the boxes andplaced at the top of a Berlese funnel for the extraction ofarthropods. The dried soil remaining at the end of arthropodextraction was analyzed for carbon and nitrogen contents and forsoil pH (see below).

The other set of soil cores (2) was used for the extraction ofenchytraeids immediately after soil sampling.

2.2. Soil respiration

The estimate of carbon mineralization through respiration rateis one of the most commonly used methods for the assessment ofsoil functioning (Coleman et al., 2004). Carbon mineralization inthe soil cores was assessed by measuring CO2 release with aninfrared gas analyzer (IRGA method, using Polytron IR CO2

DIMARSOL, Drager1) modified by the Institut de Recherche pourle Developpement, Bondy, France. Each crystal polystyrene boxcontaining a soil core was put in an air-tight enclosure into whichfresh air was injected to obtain approximately the same basal CO2

concentration (around 350 ppm). An initial measurement of CO2

concentration was recorded at the beginning of the incubation,followed by a second measurement after 4 h incubation. Theamount of mineralized C released by each soil core (CO2–C) wascalculated from the measurement of CO2 concentration. Wecalculated the CO2–C to soil organic carbon ratio (CO2–C/C) thatwe called thereafter ‘‘C mineralization’’.

2.3. Soluble organic carbon and mineral nitrogen in leachates

Soil cores were irrigated with deionized water to obtain thesame volume of leachate (40 ml). Soil solutions were collected 12 hafter the irrigation, by means of a tube connected to the 5 mm holein the crystal polystyrene box. After their volume was measured,leachates were sent to the Laboratory of Soil Analysis of theNational Institute of Agricultural Research (INRA, Arras, France) tomeasure organic carbon, nitrate and ammonium contents. Allanalyses were performed within 24–48 h after collecting the soilsolution. Nitrate and ammonium were analyzed by a spectro-photometric method (Krom, 1980; Searle, 1984) using a Contin-uous Flow System. The concentration of soluble organic carbonwas calculated with an infrared C analyzer after burning 40 mlsubsamples at 950 8C.

2.4. Collection and identification of soil fauna

Soil arthropods were extracted by the dry funnel method(Edwards and Fletcher, 1971) and stored in 90% ethyl alcohol.The total arthropod sample issuing from each soil sample corewas placed in a glass dissecting plate, the bottom of which wasdivided into 200 compartments (see Salmon et al., 2006, forfurther details). The total number of macroinvertebrates(Araneida, insects, Lumbricidae, Gastropoda, Diplopoda, Chilo-poda, Symphyla) was counted, while the number of microar-thropods was estimated by counting their numbers in 50randomly chosen compartments. Classification and countingwere achieved under a binocular microscope (40� magnifica-tion). Animals were identified at the level of the order, super-family or family and classified into morphotypes on the base ofmorphological features (Dindal, 1990; Dunger and Fiedler,1997). These features were body shape, presence of ptero-morphs, distance between anal and genital plates and presenceof leg nodules for mites, the presence and shape of eye spots,


body shape and pigmentation for Collembola, shape and size ofbody, head, eyes, antenna, and legs for other organisms.

Enchytraeids were extracted by the modified wet-funnelmethod (O’Connor, 1957), then immediately counted under amagnifying glass (Salmon et al., 2006).

2.5. Physical and chemical soil characteristics

Humus forms were identified in the field by the observation ofsoil profiles and thickness of humus layers (Brethes et al., 1995),and scaled according to the Humus index (Ponge et al., 2002). Thisindex increases from one to seven with the number, nature andthickness of organic layers and allows humus forms to be treated asnumerical data.

Organic carbon and total nitrogen contents were measuredin 50 mg dried and defaunated soil. Soil samples were groundto 100 mm and homogenized before measurements with a CHNatomic analyzer (PerkinElmer CHNS/O Analyzer 2400 Series II).

Table 3Mean abundance (�S.E.) in 35.28 cm2 of zoological groups, macrofauna, predators and total

Animal group Code MS

Collembola COLL 154.1 (88.2)

Entomobryomorpha C ENT 73.7 (34.5)

Poduromorpha C POD 80.4 (57.5)

Symphypleona C SIM 0.00 (0.00)

Neelipleona C NEL 0.00 (0.00)

Acari ACAR 892.2 (152.3)

Gamasina GAMA 24.6 (6.7)

Uropodina UROP 10.6 (3.5)

Oribatida ORIB 645.0 (137.5)

Other Acari O ACA 133.7 (48.0)

Acari larvae L ACA 78.2 (18.7)

Chilopoda CHIL 0.00 –

Geophilomorpha GEO 0.00 –

Lithobiomorpha LITH 0.00 –

Diplopoda DIPL 0.00 –

Iulidae IULI 0.00 –

Glomeridae GLOM 0.00 –

Symphyla SYMP 0.00 –

Diptera DIPT 4.00 (1.43)

Wingless Diptera D APT 0.63 (0.50)

Diptera larvae L DIP 3.38 (1.02)

Coleoptera COLE 1.50 (0.50)

Adult Coleoptera A COL 0.38 (0.26)

Coleoptera larvae L COL 1.13 (0.48)

Aphid (Homoptera) HOMO 1.50 (0.60)

Gastropoda GAST 0.00 –

Lumbricidae LUMB 0.00 –

Protura PROT 0.00 –

Araneida ARA 0.00 –

Enchytraeidae ENCH 11.1 (3.4)

Trichoptera TRIC 0.00 –

Formicidae FOR 1.13 (0.48)

Pauropoda PAUR 0.00 –

Blattidae BLAT 0.13 (0.13)

Total Fauna ABUN 1066a (186)

Predators PRED 24.6ab (6.9)

Macrofauna MACR 8.12b (1.89)

Animal group MN

Collembola 137.9 (70.6)

Entomobryomorpha 91.0 (57.6)

Poduromorpha 44.4 (20.1)

Symphypleona 0.50 (0.50)

Neelipleona 2.00 (1.51)

Acari 576.0 (91.8)

Gamasina 16.1 (4.2)

Uropodina 29.1 (11.9)

Oribatida 336.4 (57.1)

Before measures of organic carbon, any calcium carbonatepresent in the sample was removed by treatment with HCl.These analyses were executed in the Analytical Laboratory ofthe Centro di Ecologia Alpina, that had been provided withthe certification ISO/IEC 17025. Soil pH-H2O was measured onsoil mixed with deionized water (soil:water 1:5 v/v) for 5 min,and after 4 h incubation at room temperature (Anonymous,1999).

2.6. Statistical analysis

Prior to analysis, variables and indices were calculated tocharacterize animal communities: total abundance, number ofpredators (sum of Gamasida + Araneida + Chilopoda + Staphylini-dae and Carabidae), the richness of zoological groups (number ofmain zoological groups, i.e., names in bold type in Table 3),morphotype richness (number of morphotypes), Shannon indexbased on zoological groups (diversity of zoological groups).

fauna in each developmental phase of south and north facing sites and codes used in CA

RS CS

340.5 (53.6) 92.4 (31.4)

280.0 (43.0) 68.7 (27.0)

39.0 (12.2) 14.2 (5.3)

4.50 (1.92) 7.38 (2.43)

17.0 (9.8) 2.00 (1.38)

818.2 (143.2) 292.4 (77.2)

40.7 (9.3) 17.7 (6.1)

49.0 (20.6) 20.9 (9.8)

579.5 (109.0) 202.4 (58.5)

119.5 (21.4) 48.0 (16.1)

29.5 (11.6) 3.38 (2.04)

0.38 (0.26) 0.25 (0.16)

0.25 (0.16) 0.13 (0.13)

0.13 (0.13) 0.13 (0.13)

2.25 (0.41) 1.00 (0.42)

1.88 (0.40) 0.88 (0.40)

0.38 (0.26) 0.13 (0.13)

0.50 (0.38) 0.13 (0.13)

6.38 (0.98) 3.50 (0.89)

0.00 (0.00) 0.00 –

6.38 (0.98) 3.50 (0.89)

1.38 (0.50) 1.75 (0.53)

0.75 (0.31) 0.38 (0.26)

0.63 (0.26) 1.38 (0.38)

5.25 (4.97) 3.63 (2.70)

0.13 (0.13) 0.00 –

0.63 (0.18) 0.00 –

0.63 (0.50) 0.50 (0.50)

1.25 (0.49) 0.00 –

22.0 (4.8) 35.4 (7.8)

0.25 (0.16) 0.00 –

0.50 (0.38) 0.00 –

0.00 – 0.13 (0.13)

0.13 (0.13) 0.00 –

1200a (17) 431.0b (93.9)

42.7a (9.3) 18.0b (6.2)

18.4a (4.5) 10.12ab (2.26)

RN CN

101.5 (34.8) 111.5 (39.4)

79.7 (28.1) 88.7 (32.4)

18.2 (7. 8) 7.00 (1.19)

0.00 (0.00) 14.7 (7.1)

3.50 (2.06) 1.00 (0.71)

628.5 (231.7) 299.0 (82.0)

13.5 (3.9) 16.5 (9.3)

43.6 (12.9) 20.2 (4.5)

468.7 (189.8) 147.5 (51.9)

Table 3 (Continued )

Animal group MN RN CN

Other Acari 150.1 (54.0) 77.4 (31.4) 95.5 (9.3)

Acari larvae 44.2 (16.9) 25.2 (17.3) 19.2 (10.6)

Chilopoda 0.38 (0.38) 0.75 (0.41) 0.00 –

Geophilomorpha 0.00 (0.00) 0.00 (0.00) 0.00 –

Lithobiomorpha 0.38 (0.38) 0.75 (0.41) 0.00 –

Diplopoda 0.13 (0.13) 2.38 (1.00) 0.25 (0.18)

Iulidae 0.13 (0.13) 2.38 (1.00) 0.25 (0.18)

Glomeridae 0.00 (0.00) 0.00 (0.00) 0.00 (0.00)

Symphyla 0.13 (0.13) 0.25 (0.25) 0.25 (0.18)

Diptera 4.50 (1.31) 5.25 (1.24) 7.25 (1.33)

Wingless Diptera 0.63 (0.63) 0.13 (0.13) 0.00 –

Diptera larvae 3.88 (1.06) 5.13 (1.16) 7.25 (1.33)

Coleoptera 0.88 (0.48) 1.13 (0.48) 2.00 (0.29)

Adult Coleoptera 0.00 (0.00) 0.50 (0.27) 0.25 (0.18)

Coleoptera larvae 0.88 (0.48) 0.63 (0.26) 1.75 (0.18)

Aphid (Homoptera) 3.00 (1.64) 0.63 (0.42) 0.00 –

Gastropoda 0.00 – 0.25 (0.25) 0.00 –

Lumbricidae 0.00 – 0.25 (0.16) 0.00 –

Protura 0.00 – 0.13 (0.13) 0.00 –

Araneida 0.00 – 0.63 (0.63) 0.00 –

Enchytraeidae 33.1 (7.0) 32.7 (5.03) 57.5 (5.45)

Trichoptera 0.00 – 0.00 – 0.25 (0.18)

Formicidae 0.00 – 0.00 – 0.25 (0.18)

Pauropoda 0.00 – 0.00 – 0.00 –

Blattidae 0.00 – 0.00 – 0.00 –

Abundance 756.0 (163.7) 774.4 (263.0) 478.2 (117.2)

Predators 16.5 (4.3) 15.4 (4.6) 16.7 (9.3)

Macrofauna 8.87 (3.11) 11.2 (1.9) 10.0 (1.7)

Different letters inside a column indicate significant differences (p < 0.05) between developmental phases of trees in each site for macrofauna, predators and total fauna; MS:

south facing mature trees; RS: south facing regeneration; CS: south facing clearing; MN: north facing mature trees; RN: north facing regeneration; CN: north facing clearing.


Abundances of animal taxa were analyzed with correspondenceanalysis (CA), a multivariate method that gives a graphicalrepresentation of the data (in multidimensional space) bytransforming the statistical similarity between elements (chi-square distance) into an Euclidian distance (Greenacre, 1984).Active variables were defined by zoological taxa, the occurrence ofwhich was greater than five. Morphotype, developmental phases(regeneration, mature tree, clearing), aspect, animal communityindices (predator numbers, total abundance, Shannon index,zoological and morphotype richness), zoological taxa, the occur-rence of which was lower than six (Geophilomorpha, Glomeridae,wingless Diptera, Gastropoda, Protura, Trichoptera, Pauropauda,Blattidae), and soil characteristics (pH, organic carbon, nitrogen, C/N ratio, Humus index, CO2–C, CO2–C/C, mineral nitrogen andsoluble organic carbon in leachates) were used as passive variables.Passive variables were projected on factorial axes as if they hadbeen involved in the analysis, without contributing to the factorialaxes, thus allowing us to explain the distribution of zoological taxawithout constraining their ordination. All variables were standar-dized prior to analysis, following Ponge and Delhaye (1995). Eachvariable was then associated with a conjugate, varying in anopposite sense. Thus, higher and lower densities of each groupwere represented by two opposites, revealing any gradient offaunal abundance. In order to ensure a better legibility of thefigures, animal taxa and soil characteristics were representedseparately. Correlations between variables and factorial axes werecalculated using Spearman coefficient of rank correlation (Sokaland Rohlf, 1995).

Variation in soil parameters (C, N, CO2–C/C, pH, . . .) amongsampling points and consequently among developmental phaseswas analyzed with a Principal Components Analysis (PCA)(Legendre and Legendre, 1998). Calculations were based on a

matrix of Pearson’s correlation coefficients. Variables werestandardized prior to PCA.

Differences in total abundance, richness and diversity (Shannonindex) of zoological groups and morphotype richness between treedevelopmental phases were first tested using Kruskal–Wallis testprocedure in the two sites separately. When a difference wassignificant, the three means were compared two by two and thesignificance level of this multiple comparison was correctedaccording to Bonferroni (Dunn, 1964). In order to assess patterns oftotal abundance, richness and diversity of zoological groups, andmorphotype richness among developmental phases of spruce indifferent sites, south and north facing sites were considered asreplicates to test differences between developmental phases usinga one-way ANOVA with two replicates.

Correlations of Shannon index and richness of zoological groupswith C mineralization, i.e., CO2–C/C, were tested with Kendall test,for north and south facing sites pooled together. Relationshipsbetween richness of zoological groups and herbaceous cover wereanalyzed using linear regression for north and south facing sitespooled together. StaboxPro 5 and XLSTAT 2006.3 softwares(Grimmersoft1 and Adinsoft1) were used for all statisticalanalyses.

3. Results

3.1. Forest dynamics and animal communities

The abundance of zoological groups in each sampling area isgiven in Table 3. Values related to macrofauna must be consideredcautiously since the sampling size was smaller than thatcommonly sampled for macrofauna. The first two factorial axesof the CA extracted 20.7% and 11.3% of the total variance of

Table 4Shannon index, zoological and morphotype richness (mean for one sample of 35.28 cm2 � S.E.) in each developmental phase and exposure

Diversity of zoological

groups (Shannon index)

Richness of zoological

groups

Morphotype richness

South facing site MS 0.53b (0.10) 5.38b (0.50) 37.38ab (1.93)

RS 1.10a (0.06) 8.63a (0.65) 44.88a (2.71)

CS 1.29a (0.15) 6.38b (0.26) 29.00b (2.40)

North facing site MN 0.83b (0.09) 5.38 (0.32) 35.88 (1.81)

RN 1.14ab (0.15) 6.63 (0.46) 33.50 (3.65)

CN 1.44a (0.07) 4.64 (0.87) 27.26 (4.89)

Different letters inside a column indicate significant differences (p < 0.05) between the three developmental phases of trees, in south and north facing sites separately. For

abbreviations, see Table 3.


zoological taxa, respectively (Figs. 1 and 3). Axis 1 displays agradient of faunal abundance, morphotype richness and zoologicalgroup richness. This axis shows differences between develop-mental phases of spruce, especially between clearing andregeneration, particularly in the south facing site (Fig. 1). Thedensity of the great majority of zoological groups was higher in

Fig. 1. Projection of zoological taxa (active variables, coded with uppercase letters) along a

Table 3), indices related to soil fauna, and taxa, the occurrence of which was lower than fiv

of taxa names, see Table 3. Bold and italic types indicate highest and lowest values of qua

correlated to axes 1 and 2, respectively. EXPN: North exposure, EXPS: South exposure.

regeneration stands than in clearings, particularly in the southfacing site. In particular, axis 1 shows that the abundance of Acari(including Oribatida, Uropodina, Gamasida and undeterminedlarvae), Collembola (including Entomobryomorpha, Poduromor-pha and Neelipleona), Diplopoda (particularly Iulidae), Lumbrici-dae, Araneida, Diptera and Chilopoda (including Lithobiomorpha)

xes 1 and 2 issued from CA. Tree stands, sampling areas (MS, RS, CS, MN, RN, CN; see

e (coded with lowercase letters), were added as passive variables. For abbreviations

ntitative variables, respectively. Bordered and underlined variables are significantly

Fig. 2. Projection of soil parameters (active variables) along axes 1 and 2 issued from

PCA. C: soil organic carbon; N: soil nitrogen; min N and Sol org C: amount of total

mineral nitrogen and soluble organic carbon in leachates; CO2–C/C: CO2–C to soil

carbon content. For abbreviations of sampling points see Table 3.


was higher in the regeneration phase, especially in the south facingsite (Fig. 1, Table 3). These taxa were significantly correlated withaxis 1. Abundances of macrofauna and predators were also higherin regeneration stands than in clearings.

Axis 2 shows a clear positive gradient of richness and diversity ofzoological groups (Shannon index) and a negative gradient ofabundance (Fig. 1). This axis is explained by developmental phases ofspruce and shows opposite patterns for mature and regenerationstands, particularly in the south facing site, these two growth phasesbeing significantly correlated with the factorial axis. The density offive animal groups, Diplopoda (Iulidae and Glomeridae), Lumbri-cidae, Aranea, Symphyla and Trichoptera, was higher in regenerationthan in mature tree stands (Fig. 1, Table 3), especially in the southfacing site. These invertebrates have a relatively large body sizecompared to zoological groups that characterize mature stands,namely mites, particularly Oribatida, and Poduromorpha Collem-bola, although some Diptera were also found in mature stands.

Protura, that are strictly euedaphic invertebrates, were notinfluenced by developmental phases of spruce (Fig. 1. Table 3). Inthe north facing site, the three developmental phases were closerto each other and close to the origin, which indicated that theircommunities were very similar at the level of zoological groups(Fig. 1).

With regards to morphotypes, CA indicated that 19 Oribatidamorphotypes out of 50 were more abundant in adult than inregeneration stands. Regeneration areas were characterized byhigher densities of five Collembola out of 16, one Oribatida, oneGamasina, one Iulidae Diplopoda, and two Araneida morphotypes(data not shown).

3.2. Forest dynamics and soil parameters

In the south facing site, humus forms were of the eumulltype with occasional inclusions of OH material in the clearing;they varied from eumull to amphimull and from amphimullto dysmoder in regeneration and mature stands, respectively.In north facing areas, humus forms varied from dysmull toamphimull in the clearing, from dysmull to dysmoder in theregeneration stand and were only comprised of dysmoder beneathmature trees.

Table 5Soil pH, organic carbon, total nitrogen, C/N ratio and Humus index (mean (S.E.)) in eac

pH H2O Organic carbon (g kg�1) Total

MS 4.34 (0.23) 337.13 (44.15) 14.44

RS 5.21 (0.28) 133.63 (41.21) 7.48

CS 5.78 (0.27) 93.63 (13.22) 4.96

MN 4.28 (0.21) 317.88 (50.78) 14.15

RN 4.31 (0.20) 331.63 (59.82) 16.83

CN 3.58 (0.71) 174.00 (35.17) 9.69

For abbreviations, see Table 3.

Table 6Amount of soluble organic carbon (Sol org C), ammonium–N, nitrate–N and total minera

through respiration, from soil cores in the different stands (mean of eight replicates an

NO3�–N

(mg soil core�1)

NH4+–N

(mg soil core�1)

min N (

MS 0 (0) 3.968 (0.492) 3.97

RS 1.729 (1.132) 4.634 (1.472) 6.36

CS 1.686 (1.686) 2.33 (0.615) 4.02

MN 0 (0) 2.297 (0.481) 2.30

RN 5.986 (0.160) 9.713 (2.505) 15.70

CN 7.073 (7.071) 8.870 (3.543) 15.94

For abbreviations, see Table 3.

The first two factorial axes of PCA extracted 44.4% and 20.8% ofthe total variance of soil parameters, respectively (Fig. 2). Axis 1displays a negative gradient of mineralization and decompositionof organic matter, and of soil pH, expressed by higher soil pH and Cmineralization (CO2–C/C) in south facing clearing and regenerationstand, and higher amounts of soil organic carbon and nitrogen, andhigher Humus index and C/N ratio in the south and north facingmature stands (Fig. 2, Tables 5 and 6). Soluble organic carbon inleachates and CO2–C released by soil cores were also higher inmature stands than in the two other areas where there was less

h developmental phase of spruce for each exposure

nitrogen (g kg�1) C/N Humus index

(2.03) 23.62 (0.61) 6.63 (0.26)

(2.20) 17.60 (0.54) 3.19 (0.68)

(0.54) 18.43 (0.73) 3.88 (0.45)

(1.79) 21.93 (1.49) 7.00 (0.00)

(2.88) 19.29 (0.86) 6.63 (0.37)

(1.9) 13.23 (2.61) 3.85 (0.75)

l nitrogen (min N) in leachates, and CO2–C to soil organic carbon (CO2–C/C) released

d standard error)

mg soil core�1) Sol org C

(mg soil core�1)

CO2–C/C (mg 24 h�1 org

C g�1)

(0.492) 16.99 (2.52) 0.1434 (0.0254)

(1.385) 6.68 (1.67) 0.2510 (0.0247)

(1.500) 3.63 (1.19) 0.2492 (0.0238)

(0.481) 5.46 (1.12) 0.0837 (0.0178)

(7.569) 5.71 (1.31) 0.0940 (0.0242)

(5.696) 4.69 (1.86) 0.1700 (0.0532)


organic matter in the superficial layers. Axis 2 of PCA indicates thathigh amounts of mineral nitrogen, in the form of both nitrate andammonium, occurred in leachates from the north facing clearingand regeneration stands.

Axis 2 of CA is strongly explained by soil parameters, added aspassive variables in order to explain the ordination of zoologicaltaxa (Fig. 3). Higher soil pH may explain higher diversity andrichness of zoological groups (Shannon index), and the distributionof the five zoological groups that were particularly abundant inregeneration stands. The abundance of oribatids and other Acariwere higher in sampling points with thicker organic layers andlower decomposition rate, i.e., with a higher amount of soil organicnitrogen, as well as a higher Humus index (Fig. 3).

3.3. Animal diversity/richness and developmental phases of trees

With regard to diversity of zoological groups, Shannon indexwas lower in the mature stand, while it was as higher in theregeneration stand as in the clearing of both south and north facingsites (Table 4). Herbaceous plant cover followed the same pattern(Kruskal–Wallis test, p = 0.001 and p < 0.0001, in north and southfacing areas, respectively). Zoological and morphotype richnesswere higher in the south facing regeneration stand, i.e., in theintermediate developmental phase. They increased from clearing

Fig. 3. Projection of tree stands, sampling areas (MS, RS, CS, MN, RN, CN; see Table 3 for

issued from CA. Variable codes in bold and italic type correspond to highest and lowest v

correlated to axes 1 and 2, respectively. EXPN: North exposure, EXPS: South exposure; C:

soluble organic carbon in leachates.

to regeneration then decreased from regeneration to maturestands (Table 4). They followed the same trend in the north facingsite but differences between developmental phases were notsignificant.

Considering the two sites as replicates, ANOVA indicated thatthe pattern of zoological group diversity was the same as in eachsite separately (p = 0.034), while differences in zoological andmorphotype richness between developmental phases were notsignificant (p > 0.05).

Relationships between herbaceous plant cover and richness ofzoological groups, and between herbaceous plant cover anddiversity of zoological groups (Shannon) were significantlyexplained by linear regression (R2 = 0.153, p = 0.01) and logarith-mic regression (R2 = 0.415, p < 0.001), respectively.

3.4. Linking forest dynamics, diversity of zoological groups and

C mineralization

C mineralization (CO2–C/C) was higher in clearings andregeneration stands than in mature tree stands (see above). Itwas also positively correlated with the herbaceous plant cover inboth south and north facing sites (Kendall’s correlation coefficient0.271, p = 0.01). The Shannon index based on zoological groups,which followed the same pattern, was significantly correlated with

abbreviations), aspect, and soil parameters (passive variables), along axes 1 and 2

alues of variables, respectively. Bordered and underlined variables are significantly

soil organic carbon, N soil nitrogen; min N and Sol org C: total mineral nitrogen and


CO2–C/C and mineral N in leachates (Kendall’s correlationcoefficient 0.209, p = 0.023 and 0.278, p = 0.004, respectively).Conversely, the richness of zoological groups was neithercorrelated with CO2–C/C nor with mineral N.

3.5. Influence of aspect

Axis 1 of CA indicates differences in community compositionbetween north and south facing sites, but these are too close to theorigin to be interpreted (Fig. 1). Nevertheless it highlights thatmost zoological taxa were more abundant in the south facingregeneration stand than in other stands, hence the greaterabundance of these taxa in south than in north facing site(Fig. 1, Table 3). In the same way, the richness of zoological groupsin the south facing site exceeded that in the north facing site. PCAhighlights contrasted values of Humus index, soil organic carbonand nitrogen between south and north facing areas (Fig. 2, Table 5),indicating that organic layers were thicker in the north facing thanin the south facing regeneration stand. C mineralization and soil pHshow the opposite trend since they were lower in south than innorth facing areas. The levels of mineral nitrogen in north facingclearing and regeneration stand exceeded that of south facingstands. Finally, the three developmental phases were more similarin the north than in the south facing site for soil parameterscorrelated to axis 1 of PCA (see above).

4. Discussion

4.1. Forest dynamics, soil parameters and mineralization processes

Our study indicated that humus forms in a spruce forestchanged from mull to dysmoder during succession from clearingsto mature tree stands. This evolution of humus forms, expressed byan increase of the Humus index, was confirmed by chemicalanalyses since soil from mature stands had a higher organic carbonand total nitrogen contents, and a higher C/N ratio, reflectingslower decomposition of litter than in regeneration stands andclearings. The pattern of soil pH, however, varied between the twosites: it was higher in clearing and regeneration than in maturestand in the south facing site, which is in agreement with previousstudies (Salmon et al., 2006, 2008) and with the pattern of C/Nratio. The lower value of soil pH in the north facing siteunexpectedly occurred in the clearing. Soil characteristics ofregeneration stands were more similar to clearing in the southfacing site, and to mature stands in the north facing site. Thisdifference with exposure may be explained by the fact that in anorth facing site, the amount of light that arrives at the soil surfaceis low and the height of trees during the regeneration phase isenough to decrease the amount of light to the same level as that inthe mature stand. In stands with a southern exposure, light wouldnot be intercepted by regeneration trees to the same extent as inthe mature stand, which would result in warmer air and soil andconsequently increased decomposition rate in the south facingregeneration stand (Imbeck and Ott, 1987) as indicated by higherCO2–C/C values. Changes in soil parameters during the growthphase of spruce are in agreement with the pattern observed in theItalian Alps on a more acidic parent rock (Salmon et al., 2006), inNorway (Hagvar, 1982), and in French Alps (Bernier and Ponge,1994). However, this evolution probably depends on altitude(Bernier, 1996) and forest management since the thickness oforganic layers and soil pH did not parallel those observed in spruceplantations in Germany (Zaitsev et al., 2002). The decrease in Cmineralization (CO2–C/C) with stand age confirms the decrease inC mineralization, microbial activity and bacterial number observedby Chauvat et al. (2003), while the fungal biomass increased. The

decline of microbial C and N concentrations with the age of treesrecorded by Bauhus et al. (1998) tallies with the decline of themineralization rate that we observed.

4.2. Forest dynamics and animal communities

Invertebrate communities in clearings were very similar incomposition to those of regeneration stands. Nevertheless,clearings were characterized by the low abundance of total faunaand of most zoological groups, which confirms results of otherstudies (Addison et al., 2003; Grgic and Kos, 2005). This is easilyexplained by the absence or the smaller amount of spruce litter, i.e.,the most favorable habitat for a number of microarthropods inconiferous forests, particularly at higher altitude. Mature sprucestands in this study were characterized by higher abundances ofAcari, particularly oribatids, which is in agreement with thevariations observed in a slightly more acidic and a calcareousbedrock, respectively (Salmon et al., 2006, 2008). This distributionof oribatid mites is inconsistent with that reported by Zaitsev et al.(2002) who observed a higher density of mites (and a lower pH) ina 25-year-old than in a 95-year-old spruce stand, but it supportsthe similarity of mite density between regeneration and maturestands in another spruce forest in Germany (Migge et al., 1998).

The abundance of Collembola was higher in regeneration standsin the south facing site, which supports the results of Salmon et al.(2006, 2008) on acidic and calcareous substrates. This pattern doesnot parallel the trend reported by other studies where higherdensities of Collembola occurred beneath old spruce stand (160years) (Hagvar, 1982), or did not vary among growth stages(Chauvat et al., 2003). Salmon et al. (2006) hypothesized that theunexpected lower density of Collembola in mature stands could beexplained by the competition between Collembola and oribatidsfor microbial food (Kaneko et al., 1995; Irmler, 2000). However,this explanation is not valid for the present study since theabundance of oribatids in mature stands did not exceedsignificantly that observed in the south facing regeneration stand.Instead, lower water and nutrient availability caused by thedensification of the root system of spruce (Babel, 1977), and theincreased rate of nutrient and water uptake by trees until canopyclosure (Miller, 1981) may be responsible for the lower abundanceof Collembola beneath mature trees, Collembola being lessresistant to desiccation than oribatids (Coleman et al., 2004).Lower levels of mineral N, especially nitrate–N, in mature standscorroborate the hypothesis of soil impoverishment under theinfluence of tree uptake. The higher percentage of herbaceouscover in regeneration compared to mature stands may also explainvariations in collembolan populations: herbaceous plants constitutea habitat for Collembola, particularly for epigeic and hemi-edaphicspecies, which mainly belong to one out of two super-families thatwere more abundant in regeneration stands, namely Entomobryo-morpha. Litter from herbaceous plants in regeneration stands alsocontributed to the diversification of food resources for Collembola,which in mature stands consisted almost entirely of spruce needles,particularly recalcitrant (Gallet, 1992) and less palatable thanherbaceous litter (Edwards, 1974; Ponge, 1991).

Most other groups, including saprophagous macrofauna(Diplopoda, Lumbricidae, Diptera larvae) and predators (Araneida),were also more abundant in regeneration than in mature stands,which supports previous observations indicating a higher abun-dance of these taxa in mull or soil at higher pH than in moder(including dysmoder), which is more acidic (Schaefer, 1991; Davidet al., 1993; Lavelle, 2001). The absence of most macrofauna inmature stands can be partially explained by the increase ofrecalcitrant spruce litter and the rarefaction of the herbaceouslayer during the second half of the intense growth phase of spruce


(i.e., from 40 to 80–90 years), since few organisms are able toconsume coniferous needles, oribatid mites being most efficient(Ponge, 1988, 1991). In addition, the accumulation of litterreinforces the decrease of soil pH, which affects the distributionof soil animals (Salmon and Ponge, 1999; Schaefer and Schauer-mann, 1990; Stevenson, 1994; Edwards, 1998), particularlymacrofauna. Our results support the assumption that during thesecond half of the build-up phase of succession, ecosystemdevelopment is plant driven and is ultimately determined bythe functional traits of the dominant species present, i.e., spruce(Wardle, 2002), through variations in the availability of water andnutrients, diversity and quality of food resources and habitat, andsoil pH.

4.3. Animal diversity and developmental phases of trees

As hypothesized, the richness of zoological groups peaked inregeneration stands, i.e., at the intermediate succession phase,which is the more diversified (Connell, 1978). In particular, thisintermediate state offered a thin layer of spruce litter (absent fromclearings), enriched by an increased number of herbaceous plants(scarce in mature stands), i.e., diversification and enhancement ofliving-space and food resources for soil fauna. Consequently, localsoil heterogeneity (i.e., heterogeneity at each sampling point) wasmaximized during regeneration phase, which may explain a higherrichness of zoological groups in this phase. This is in line with theobservation by Jordana et al. (2000) and the assumption byBardgett et al. (2005) that soil heterogeneity is a determinant ofsoil biodiversity. However, variations in richness of zoologicalgroups were significant only when considering south facing site,because the level of zoological group richness remained low innorth facing clearing and regeneration stands, which may beattributed to unfavorable micro-climate conditions.

The Shannon index, expressing the diversity of zoologicalgroups, did not follow exactly the same pattern as zoological andmorphotype richness. This index attributes less importance togroups with low abundance than the richness of zoological groups;it is consequently less sensitive to small variations in the faunacommunity, and better reflects functional diversity, giving a higherweight to best represented, and thus highly functional groups (seebelow). The pattern of Shannon index paralleled that of herbaceousplant cover; it was highest in clearings and regeneration standsand lowest in mature stands, this in the two sites. This pattern is inline with that observed in a spruce forest grown on calcareousbedrock (Salmon et al., 2008). After the regeneration phase,richness and diversity of zoological groups probably decreasedbecause of soil impoverishment caused by a decrease of the herblayer, an increase of recalcitrant and acidifying litter input, and treeuptake (see above), which results in the escape of mostmacrofauna, and the dominance of oribatid mites, the mostefficient group in the comminution of spruce needles. Thecorrelation of both diversity and richness of zoological groupswith herbaceous cover in contrasted sites, corroborates theassumption that the quality of food resources provided by plantcommunities (and depending on forest dynamics) drives, at leastpartly, the diversity and richness of zoological groups in soils(Wardle, 2005). The quality of food resources and habitat, seems toimpact the diversity of zoological groups more greatly than theheterogeneity of litter.

4.4. Linking diversity of zoological groups and soil functioning

The positive correlation between the diversity of zoologicalgroups (Shannon index on zoological groups) and C mineralizationis consistent with several other studies showing that a higher

biological diversity is required to sustain ecosystem functioning(Loreau et al., 2001). Our results indicate that the proportion(versus richness) of different zoological groups must be taken intoaccount when evaluating the relationships between diversity andmineralization processes since only the Shannon index (and notthe richness of zoological groups) was correlated to C mineraliza-tion. Nevertheless the greater amount of high quality resources inclearings and regeneration stands was probably the primarydeterminant of soil processes, as indicated by the correlationbetween C mineralization and herbaceous cover. We cannotexclude that an intensification of decomposition processes withincreasing faunal diversity could also result from a higher faunalbiomass, which was the outcome of a higher faunal diversity.

4.5. Influence of aspect

Strong differences were observed between northern andsouthern exposures. Generally higher values of faunal abundanceand diversity were observed in south facing sites compared tonorth facing ones, especially in the regeneration phase. There wereno, or only small differences in the composition of animalcommunities, humus forms and C mineralization betweendevelopmental phases of spruce in the north facing site, whilethese factors varied greatly with tree age in the south facing site.Striking differences also concerned diversity and richness ofzoological groups (higher in south facing regeneration stand andclearing). Such north–south variations have been observed in aspruce forest grown on calcareous bedrock in the Alps (Salmonet al., 2008), which allows to attribute the observed pattern to theeffect of aspect although only one site was studied for eachexposure type, all the more since geology, topography and climatewere similar in the two sites. This contrast between south andnorth facing sites emphasizes the impact of interactions betweenspruce forest dynamics and micro-climate (temperature, moistureand light intensity) on microflora and fauna. Since the length of theperiod of direct sunlight is correlated with air and soil warming(Imbeck and Ott, 1987), contrasts of temperature and consequentlyvariations in biological activity between the three stands could beenhanced at southern exposure, especially between open (clear-ing) and closed (mature stands) habitats.

5. Conclusions

Our study showed that profound modifications in thecomposition of animal communities and soil functioning,expressed by C mineralization rate, occur between clearing,regeneration and mature stands. Variations among developmentalphases of spruce, observed here, parallel those of two other studiesconducted in three varied sites differing by the quality of thebedrock (Salmon et al., 2006, 2008), which allows to generalize theobserved pattern. Our results also support the hypothesis thatforest dynamics drives the richness of zoological groups, commu-nities composition and soil functioning, through quality (herbac-eous cover) and quantity (mineral nitrogenous nutrients) of foodresources and, to a lesser extent, through soil heterogeneity due tothe simultaneous presence of herb layers and spruce litter in theregeneration phase. Our study also emphasizes the relationshipbetween diversity of zoological groups and soil functioningexpressed by mineralization rate; it highlights particularly thatthe proportion of zoological groups must be taken into account. Ifdecrease of herbaceous cover, soil impoverishment, accumulationof spruce needles, and soil acidification, which occur during theforest cycle, are actually responsible for changes in soil functioningand in the composition and diversity of animal communities, thena decrease in the size of stand management areas (particularly in


lowland forests where the size of stands is very large), wouldreduce those processes and consequently improve the diversity ofzoological groups. A patchwork of stand management areas shouldincrease the heterogeneity of the soil environment on a largerspatial scale and, consequently, facilitate the coexistence of speciesfrom diverse groups on a macrogeographic scale (Loyola et al.,2006; Toljander et al., 2006).

Acknowledgements

This study was funded by the Fund for Research Projects of theAutonomous Province of Trento and the Centro di Ecologia Alpinaof Trento (Italy), ‘‘DINAMUS – Humus forms and forest dynamics’’and ‘‘INHUMUSnat2000 – Humus forms-indicators of functionality for

Nature 2000 sites’’ projects, decisions G.P. n. 437/2002 e G.P. n.1587/2004.

The authors are grateful to Gerard Bellier from IRD (Institut deRecherche pour le Developpement, Bondy, France) for hiscollaboration to soil respiration measurements. They thankJean-Francois Ponge and Ceryl Techer from the Museum Nationald’Histoire Naturelle, Paris for fruitful discussions and manuscriptcorrections, and the making of soil sampling boxes, respectively.The authors thank Lorenzo Tarasconi for classifying arthropods,Giorgia Mattiuzzo and Marco Clementel for enchytraeid extraction,Silvia Chersich and Roberto Zampedri from the Centro di EcologiaAlpina for geological, temperature and rainfall data, GianlucaSoncin for providing the history of our study sites, Mirco Tomasiand Matteo Girardi for carbon and nitrogen analyses, and AnnalisaLosa and Heidi C. Hauffe for linguistic revision.

References

Addison, J.A., Trofymow, J.A., Marshall, V.G., 2003. Abundance, species diversity, andcommunity structure of Collembola in successional coastal temperate forestson Vancouver Island, Canada. Applied Soil Ecology 24, 233–246.

Anonymous, 1999. Qualite des sols, vol 2. AFNOR, Paris, pp. 408.Babel, U., 1977. Influence of high densities of fine roots of Norway spruce on

processes in humus covers. Ecological Bulletins 25, 584–586.Baguette, M., Gerard, S., 1993. Effects of spruce plantations on carabid beetles in

southern Belgium. Pedobiologia 37, 129–140.Bardgett, R.D., 2002. Causes and consequences of biological diversity in soil. Zoology

105, 367–374.Bardgett, R.D., Yeates, G.W., Anderson, J.M., 2005. Patterns and determinants of soil

biological diversity. In: Bardgett, R.D., Hopkins, D.W., Usher, M.B. (Eds.), Bio-logical Diversity and Function in Soils. Cambridge University Press, Cambridge,pp. 100–118.

Bauhus, J., Pare, D., Cote, L., 1998. Effects of tree species, stand age and soil type onsoil microbial biomass and its activity in a southern boreal forest. Soil Biologyand Biochemistry 30, 1077–1089.

Bernier, N., 1996. Altitudinal changes in humus form dynamics in a spruce forest atthe montane level. Plant and Soil 178, 1–28.

Bernier, N., Ponge, J.-F., 1994. Humus form dynamics during the sylvogenetic cyclein a mountain spruce forest. Soil Biology and Biochemistry 26, 183–220.

Brethes, A., Brun, J.-J., Jabiol, B., Ponge, J.-F., Toutain, F., 1995. Classification of foresthumus forms: a French proposal. Annales des Sciences Forestieres 52, 535–546.

Chauvat, M., Zaitsev, A.S., Wolters, V., 2003. Successional changes of Collembola andsoil microbiota during forest rotation. Oecologia 137, 269–276.

Coleman, D.C., Crossley, D.A.J., Hendrix, P.F., 2004. Fundamentals of Soil Ecology.Elsevier Academic Press, Amsterdam, p. 386.

Connell, J.H., 1978. Diversity in tropical rain forests and coral reefs. Science 199,1302–1310.

Cortet, J., Joffre, R., Elmholt, S., Krogh, P.H., 2003. Increasing species and trophicdiversity of mesofauna affects fungal biomass, mesofauna community structureand organic matter decomposition process. Biology and Fertility of Soils 37,302–312.

David, J.F., Ponge, J.-F., Delecour, F., 1993. The saprophagous macrofauna of differenttypes of humus in beech forests of the Ardenne (Belgium). Pedobiologia 37, 49–56.

Dindal, D.L., 1990. Soil Biology Guide. Wiley, New York, p. 1349.Dufour, A., Gadallah, F., Wagner, H.H., Guisan, A., Buttler, A., 2006. Plant species

richness and environmental heterogeneity in a mountain landscape: effects ofvariability and spatial configuration. Ecography 29, 573–584.

Dunger, W., Fiedler, H.J., 1997. Methoden der Bodenbiologie. Gustav Fischer Verlag,Jena, p. 539.

Dunn, O.J., 1964. Multiple comparisons using rank sums. Technometrics 6, 241–252.

Edwards, C.A., 1974. Macroarthropods. In: Dickinson, C.H., Pugh, D.J.F. (Eds.),Biology of Plant Litter Decomposition, 2. Academic Press, New York, pp.533–554.

Edwards, C.A., 1998. Earthworm Ecology. St. Lucie Press, Boca Raton, p. 389.Edwards, C.A., Fletcher, K.E., 1971. A comparison of extraction methods for terres-

trial arthropods. In: Phillipson, J. (Ed.), Methods of Study in Quantitative SoilEcology: Population, Production and Energy Flow. Blackwell, Oxford, pp. 150–185.

Gallandat, J.D., Gillet, F., Havlicek, E., Perrenoud, A., 1995. Typologie et SystemiquePhyto-Ecologiques des Paturages Boises du Jura Suisse. Laboratoire d’EcologieVegetale et de Phytosociologie, Universite de Neuchatel, Neuchatel, p. 68.

Gallet, C., 1992. Apports de la biochimie a la connaissance du fonctionnement desecosystemes forestiers: role des composes phenoliques dans une pessiere amyrtille. Doctorate thesis, Universite Claude-Bernard, Lyon, p. 277.

Gaston, K.J., 2000. Global patterns in biodiversity. Nature 405, 220–227.Gillet, F., 2000. La Phytosociologie Synusiale Integree. Guide Methodologique.

Institut de Botanique, Universite de Neuchatel, Neuchatel, p. 371.Greenacre, M.J., 1984. Theory and Applications of Correspondence Analysis. Aca-

demic Press, London, p. 364.Grgic, T., Kos, I., 2005. Influence of forest development phase on centipede diversity

in managed beech forests in Slovenia. Biodiversity and Conservation 14, 1841–1862.

Hagvar, S., 1982. Collembola in Norwegian coniferous forest soils. I. Relations toplant communities and soil fertility. Pedobiologia 24, 255–296.

Imbeck, H., Ott, E., 1987. Verjungungsokologische Untersuchungen in einem hoch-staudenreichen subalpinen Fichtenwald, mit spezieller Berucksichtigung derSchneeablagerung und der Lawinenbildung. Mitteilungen des EidgenossischesInstitute fur Schnee- und Lawinenforschung 42, 1–202.

Irmler, U., 2000. Changes in the fauna and its contribution to mass loss and N releaseduring leaf litter decomposition in two deciduous forests. Pedobiologia 44, 105–118.

Jordana, R., Arpin, P., Vinciguerra, M.T., Gonzalez, S., 2000. Biodiversity acrossecotones in desertificable Mediterranean areas. In: Balabanis, P., Peter, D.,Ghazi, A., Tsogas, M. (Eds.), Mediterranean Desertification Research: Resultsand Policy Implications. Official Publications of the European Communities,Luxembourg, pp. 497–505.

Kaneko, N., McLean, M.A., Parkinson, D., 1995. Grazing preference of Onychiurussubtenuis (Collembola) and Opiella nova (Oribatei) for fungal species inoculatedon pine needles. Pedobiologia 39, 538–546.

Krom, M.D., 1980. Spectrophotometric determination of ammonia: a study of amodified Berthelot reaction using salicylate and dichloroisocyanurate. Analyst105, 305–316.

Kubiena, W.L., 1955. Animal activity in soils as a decisive factor in establishment ofhumus forms. In: Mc Kevan, D.K. (Ed.), Soil Zoology. Butterworths, London, pp.73–82.

Lavelle, P., 2001. Soil Ecology. Kluwer, Dordrecht, p. 654.Legendre, P., Legendre, L., 1998. Numerical Ecology, 2nd English edition. Elsevier,

Amsterdam, p. 853.Loreau, M., Naeem, S., Inchausti, P., Bengtsson, J., Grime, J.P., Hector, A., Hooper, D.U.,

Huston, M.A., Raffaelli, D.G., Schmid, B., Tilman, D., Wardle, D.A., 2001. Biodi-versity and ecosystem functioning: current knowledge and future challenges.Science 294, 304–308.

Loyola, R.D., Brito, S.L., Fereira, R.L., 2006. Ecosystem disturbances and diversityincrease: implications for invertebrate conservation. Biodiversity and Conser-vation 15, 25–42.

Migge, S., Maraun, M., Scheu, S., Schaefer, M., 1998. The oribatid mite community(Acarina) of pure and mixed stands of beech (Fagus sylvatica) and spruce (Piceaabies) of different age. Applied Soil Ecology 9, 115–121.

Miller, H.G., 1981. Forest fertilization: some guiding concepts. Forestry 54, 136–157.Naeem, S., Hakansson, K., Lawton, J.H., Crawley, M.J.L.J.T., 1996. Biodiversity and

plant productivity in a model assemblage of plant species. Oikos 76, 259–264.O’Connor, F.B., 1957. An ecological study of the enchytraeid worm population of a

coniferous forest soil. Oikos 8, 161–169.Paquin, P., Coderre, D., 1997. Changes in soil macroarthropod communities in

relation to forest maturation through three successional stages in the Canadianboreal forest. Oecologia 112, 104–111.

Ponge, J.-F., 1988. Etude ecologique d’un humus forestier par l’observation d’unpetit volume. III. - La couche F1 d’un moder sous Pinus sylvestris. Pedobiologia31, 1–64.

Ponge, J.-F., 1991. Succession of fungi and fauna during decomposition of needles ina small area of Scots pine litter. Plant and Soil 138, 99–113.

Ponge, J.-F., 1993. Biocenoses of Collembola in Atlantic temperate grass-woodlandecosystems. Pedobiologia 37, 223–244.

Ponge, J.-F., Chevalier, R., Loussot, P., 2002. Humus Index: an integrated tool for theassessment of forest floor and topsoil properties. Soil Science Society of AmericaJournal 66, 1996–2001.

Ponge, J.-F., Delhaye, L., 1995. The heterogeneity of humus profiles and earthwormcommunities in a virgin beech forest. Biology and Fertility of Soils 20, 24–32.

Rosenzweig, M.L., 1995. Species Diversity in Space and Time. Cambridge UniversityPress, Cambridge, p. 436.

Rusek, J., 1975. Die bodenbildende Funktion von Collembolen and Acarina. Pedo-biologia 15, 299–308.

Salmon, S., Artuso, N., Frizzera, L., Zampedri, R., 2008. Response of soil faunacommunities and humus forms to forest dynamics and solar radiation. SoilBiology and Biochemistry 40, 1707–1715.


Salmon, S., Mantel, J., Frizzera, L., Zanella, A., 2006. Changes in humus forms and soilanimal communities in two developmental phases of Norway spruce on anacidic substrate. Forest Ecology and Management 237, 47–56.

Salmon, S., Ponge, J.-F., 1999. Distribution of Heteromurus nitidus (Hexapoda,Collembola) according to soil acidity: interactions with earthworms and pre-dator pressure. Soil Biology and Biochemistry 31, 1161–1170.

Schaefer, M., 1991. Fauna of the European temperate deciduous forest. In: Rohrig,E., Ulrich, B. (Eds.), Ecosystems of the World 7. Temperate Deciduous Forests.Elsevier, Amsterdam, pp. 503–525.

Schaefer, M., Schauermann, J., 1990. The soil fauna of beech forests: comparisonbetween a mull and a moder soil. Pedobiologia 34, 299–314.

Schwartz, M.W., Brigham, C.A., Hoeksema, J.D., Lyons, K.G., Mills, M.H., van Man-tgem, P.J., 2000. Linking biodiversity to ecosystem function: implications forconservation ecology. Oecologia 122, 297–305.

Searle, P.L., 1984. The Berthelot or indophenol reaction and its use in the analyticalchemistry of nitrogen. Analyst 109, 549–568.

Setala, H., Berg, M.P., Jones, T.H., 2005. Trophic structure and functional redundancyin soil communities. In: Bardgett, R.D., Hopkins, D.W., Usher, M.B. (Eds.), Bio-logical Diversity and Function in Soils. Cambridge University Press, Cambridge,pp. 236–249.

Sokal, R.R., Rohlf, F.J., 1995. Biometry. Freeman, New York, p. 887.Stevenson, F.J., 1994. Humus Chemistry. Genesis, Composition, Reactions, 2nd ed.

Wiley, New York, p. 496.

Tilman, D., Wedin, D., Knops, J., 1996. Productivity and sustainability influenced bybiodiversity in grassland ecosystems. Nature 379, 718–720.

Toljander, Y.K., Lindahl, B.D., Holmer, L., Hogberg, N.O.S., 2006. Environmentalfluctuations facilitate species co-existence and increase decomposition incommunities of wood decay fungi. Oecologia 148, 625–631.

Toutain, F., 1987. Activite biologique des sols, modalites et lithodependance.Biology and Fertility of Soils 3, 31–38.

Trofymow, J.A., Addison, J., Blackwell, B.A., He, F., Preston, C.A., Marshall, V.G., 2003.Attributes and indicators of old-growth and successional Douglas-fir forests onVancouver Island. Environmental Reviews 11, S187–S204.

Vedder, B., Kampichler, C., Bachmann, G., Bruckner, A., Kandeler, E., 1996. Impact offaunal complexity on microbial biomass and N turnover in field mesocosmsfrom a spruce forest soil. Biology and Fertility of Soils 22, 22–30.

Wardle, D.A., 2002. Communities and Ecosystems. Linking the Aboveground andBelowground Components. Princeton University Press, Princeton, p. 391.

Wardle, D.A., 2005. How plant communities influence decomposer communities.In: Bardgett, R.D., Hopkins, D.W., Usher, M.B. (Eds.), Biological Diversity andFunction in Soils. Cambridge University Press, Cambridge, pp. 119–138.

Wardle, D.A., Zackrisson, O., Hornberg, G., Gallet, C., 1997. The influence of islandarea on ecosystem properties. Science 1296–1299.

Zaitsev, A.S., Chauvat, M., Pflug, A., Wolters, V., 2002. Oribatid mite diversity andcommunity dynamics in a spruce chronosequence. Soil Biology and Biochem-istry 34, 1919–1927.

Documents

Linking forest dynamics to richness and assemblage of soil zoological groups and to soil mineralization processes