Contents lists available at ScienceDirect

Applied Soil Ecology

journal homepage: www.elsevier.com/locate/apsoil

Review

Study of soil–vegetation relationships on the Butte Montceau inFontainebleau, France: Pedagogical exercise and training report☆

Alexandre Terrigeola, Marc-Frédéric Indorfa, Thierry Jouhaniquea, Florian Tanguya,Diane Hoareaua, Vida Rahimiana, Kenny Agésilas-Lequeuxa, Rebecca Bauesa,Marion Noualhgueta, Anice Cheraieta, Aleksander Miedziejewskia, Stéphane Bazotb,Jean-Christophe Latac, Jérôme Mathieuc, Christophe Hannotb, Jean-Michel Dreuillauxb,Augusto Zanellad,⁎

a Paris-Saclay University or Pierre et Marie Curie University, Paris, Franceb Université Paris Saclay, Paris, Francec Université Pierre et Marie Curie, Paris, Franced Università di Padova, Padova, Italy

A R T I C L E I N F O

Keywords:HumusHumusicaSoil ecologySoil biodiversitySoil training reportSoil and vegetationErasmus activityButte MontceauFontainebleau forest

A B S T R A C T

This article illustrates a short training course for students at the Master's level, which explores relationshipsbetween plants and soil. It takes place in the Forest of Fontainebleau (France), where a reception centre with alarge room (30 m2) is located containing a Berlese equipment, and many microscopes. On the first day, studentsare accompanied by their professors in the field and visit the four sites which are located along an ecologicaltransect. Vegetation and soils at each site are presented to them by specialists. In the laboratory, indications arewritten on a black board, explaining how to use relatively simple tools for biological investigations (microscopes,GPS, Berlese funnel, photometer, flora and fauna guides…). Students are then divided into 4 groups of 4–6students. Each group is assigned a site which is then described and analysed. At the end, each group produces awritten report. Are the measured parameters interrelated within each site? Are there functioning principles thatmay distinguish the four stations? Three professors are always present and accompany the students in the field orin the laboratory so as to provide help when necessary. After a brief moment of uncertainty, students are able toquickly organise themselves and after three days identify the essential elements of these ecosystems. They alsolearn how a real biologist observes a forest ecosystem. They discover that plants, soil and animals are inter-connected and form a natural functioning system. Students thus learn that nonetheless, it is difficult to haveclearly identified boundaries between the investigated forest stations, because the gradient between them isgradual and ecologically indefinite.

1. Foreword

This study is a pedagogical exercise, part of the academic courseSOLT, proposed yearly by Paris-Saclay and Pierre et Marie CurieUniversities (France) in the Biodiversity, Ecology, and EvolutionMaster. The training course takes place at the Station d’EcologieForestière de Fontainebleau-Avon (Université de Paris 7, Diderot).

The aim of the course is to give students an opportunity, through

field experience, to observe and measure the main ecological variablesin a forest ecosystem. Faced with an ecological question, students de-velop ways to answer it while dealing with field constraints: what canbe measured, with what means, and for what purpose?

SOLT is a five-day training course divided into three phases: 1) fieldwork, after a brief explanation recalling the content of previous courses,collection of vegetation and soil data (about 12 h); 2) data analysis in atraining room at the Ecological Station of Fontainebleau-Avon (12 h);

http://dx.doi.org/10.1016/j.apsoil.2017.07.021Received 22 March 2017; Received in revised form 3 July 2017; Accepted 12 July 2017

☆ Background music while reading?: Jean-Charles Gandrille: Toccata minimaliste (Coralie Amedjkane) d’Aivars KALĒJS: https://www.youtube.com/watch?v=Haxs7J_fg5U.⁎ Corresponding author.E-mail addresses: alex.terrigeol@gmail.com (A. Terrigeol), markindo@mac.com (M.-F. Indorf), thierry.jouhanique@laposte.net (T. Jouhanique), flo-917@hotmail.fr (F. Tanguy),

diane.hoareau@u-psud.fr (D. Hoareau), gaby493@orange.fr (V. Rahimian), kenny.agesilas-lequeux@u-psud.fr (K. Agésilas-Lequeux), rebecca.baues@u-psud.fr (R. Baues),marion.noualhguet@u-psud.fr (M. Noualhguet), anice.cheraiet@u-psud.fr (A. Cheraiet), aleksander.miedziejewski@etu.upmc.fr (A. Miedziejewski), stephane.bazot@u-psud.fr (S. Bazot),jean-christophe.lata@upmc.fr (J.-C. Lata), jerome.mathieu@upmc.fr (J. Mathieu), christophe.hannot@u-psud.fr (C. Hannot), jeanmichel.dreuillaux@u-psud.fr (J.-M. Dreuillaux),augusto.zanella@unipd.it (A. Zanella).

Applied Soil Ecology xxx (xxxx) xxx–xxx

0929-1393/ © 2017 Elsevier B.V. All rights reserved.

Please cite this article as: Terrigeol, A., Applied Soil Ecology (2017), http://dx.doi.org/10.1016/j.apsoil.2017.07.021

3) statistical analysis of data and report writing (16 h).Even if the report has the structure of a scientific paper, its content

has the value of only a few days of training experience, not one of ascientific publication. It lacks precision in the measurements (use offield tests for chemical data, lack of time for necessary scientific fieldand laboratory investigations, first time experience for many studentsto analyse a soil profile or make a phytosociological sampling list…).The report is also a training experience that places them in front of thedifficulties of data analysis interpretation. Students learn how datacollection, data analysis, and results are interconnected and inter-dependent. Another goal of the course is to show that the quality of anecological scientific publication strongly depends on feedback thathelps the researcher adopt and improve his or her preliminary in-vestigation plan and also develops her or his capacity to fruitfullycollaborate with other colleagues.

The teaching staff accompanied the students along a transect pas-sing through four different types of forest stations. The transect wasrelatively small (length< 1 km; width: 200 m) and could be con-sidered climatically homogeneous. The group stopped at each foreststation, making observations and measurements of the vegetation andsoil with the goal to be able to answer the following fundamentalquestions:

• Are there main ecological factors that can explain the observedsubdivisions in forest stations?

• What are the differences at the level of soil and vegetation biodi-versity and how is it possible to investigate and present these dif-ferences?

• Are soil and vegetation interdependent?

• Is it possible to predict the evolution of these systems in response toprocesses of climatic global warming?

The students were divided into 4 groups, each of them collectingdata from a single station. The data was then shared at the end of thesecond day with the other groups for statistical analysis. Each groupprepared their own statistical analysis and report. This article is basedon the report presented by Tanguy F., Jouhanique T. and Terrigeol A.,translated in English by Indorf M.-F. The other students made a similarreport and proofread this one which had a clearer structure. Theteaching staff accompanied the process while giving students liberty todevelop and carry out their own actions and initiatives. The corre-sponding author coordinated the redaction of this paper and realisedthe photographic report.

2. Introduction

Understanding of the soil functioning is crucial, especially in con-sidering agricultural needs and climate regulation. Exchanges are madebetween air, water, and living organisms but we still don’t understandall the processes involved. These compartments are not isolated asnumerous interactions occur between the soil, the flora, and the fauna.

Soil micro-organisms like bacteria et archaea are important actors inthis process. They have the ability to digest all the organic compoundsin the soil, even the humic or phenolic ones such as tannins (Lavelleet al., 1995a; Barot et al., 2007; Clause et al., 2014; Dickson and Broyer,1972; Lavelle, 2009). However, their activity remains limited withoutthe macro-organisms which modify the environment in time and space(Lavelle et al., 1995b).

Plants, for their part can influence soil properties by producingchemicals and organic compounds affecting litter, humus, and soil (Vander Putten et al., 2013; Ponge et al., 2014, 2011, 1997).

The goal of this study is to demonstrate how pedogenesis soil for-mation under the same climate can vary within short distances due todifferent geological subsoil structures, and how this drives the entireecosystem, from soil organisms to plant cover.

3. Materials and methods

3.1. The Butte Montceau

The Butte Montceau (N48°24′34”, E2°44′37”) is a 125 m high hilllocated east of Fontainebleau along the Seine River. The Butte is madeof several geological strata dating mainly from the Oligocene. TheSannoisian limestone form the first marine deposit, followed by the BrieLimestone (10–15 m thick). The 40 m thick stratum of sand forms themost imposing part of the Butte and dates back to the Stampian Stage.The Butte is mainly covered by forest vegetation. The different studystations are characterised by their soil, exposition, and floristic com-position (Fig. 1).

The Butte Montceau, along with other similar rises just south of thetown of Avon, is quite remarkable as it stands apart from the initialsedimentary stratum (the Beauce limestone) that can be found furthersouth, stretching from the town of Étampes to the Cher Valley. In thisregion, the initial sedimentary stratum is usually 40 m thick.

3.2. Sampling design

We studied four contrasted stations along a 500 m transect. Eachstation hosted a different subsoil structure and exposure. As mentionedabove, they are all located on the Butte Montceau and identified inFig. 1 by their number. Station 1 corresponds to summit, and station 4is next to the Seine River.

Flora and fauna: Five square patches of 100 m2 were laid out ateach station and these were used as replicates for floristic inventoriesand soil and litter samples. The patches were aligned along a transectthat was perpendicular to the slope, so that all replicates were at thesame altitude. A distance of 10 m was respected between each patchcorresponding to the width of a patch. The five patches provided aprospection zone large enough for the floristic inventory.

A quadrat of 50 cm x 50 cm was placed in the centre of each patch(Fig. 2a, b, c). All litter (approximately OL + OF horizons), inside thequadrats was then collected into a sack, weighed and identified for eachtree species. White rot was evaluated for each species on a basis of a 10-leaf sample for each species.

Soil properties: a soil chunk (soil without “litter”, i.e. soil withoutOL and OF horizons) of about 1000 cm3 (10 × 10 × 10) was thencarefully extracted from each quadrat, in order to preserve at best, thedifferent horizons for future chemical analyses. This chunk corre-sponded mainly to the A horizon in stations 1, 3 and 4, and to (OH

Fig. 1. Geological profile of the Butte Montceau at Fontainebleau going from the peak ofthe Butte to the bank of the Seine River. The following geological strata can be observed:the ridge is capped by a thin layer (3–4 m) of Beauce limestone covered with a thin windyloam (Limon des Plateaux, < 50 cm (1); the limestone layer is followed by an importantstratum of Fontainebleau sands (2), a soft slope of Brie limestone (3), a layer of Greenmarls (4) and finally the Champigny limestone which also forms the Seine bed (5). Inblack, the 4 sampling points corresponding to the different forest stations. (For inter-pretation of the references to colour in this figure legend, the reader is referred to the webversion of this article).

A. Terrigeol et al. Applied Soil Ecology xxx (xxxx) xxx–xxx

2

+ A+ BPhs horizons) in station 2. Each chunk was also stored in anidentified sack.

Soil profile: One pedological profile was done for each station. A60 cm deep pit was dug in order to observe the different soil horizons(Fig. 3a, b, c).

3.3. Methods

3.3.1. Humus and soil profileDue to varying soil depths at each station, a soil auger was some-

times used to analyse subsoils beyond 60 cm deep, and at other times, apit of 30 cm was sufficient. The different horizons were laid out on a flatsurface in order to facilitate observations (Figs. 4–7).

Names and codes of diagnostic horizons and soil correspond toRéférentiel Pédologique (AFES, 2009) and IUSS Working Group WRB

(2015) references; for the ones of humus systems and forms we fol-lowed Humusica 1 (Horizons: article 4, and classification: article 5).

Each horizon was then characterised by its thickness, colour, tex-ture, pH, and the presence of carbonate compounds. Various tools wereused such as a tape measure, field pH metre, Munsell soil colour chartand a solution of hydrochloric acid (HCl) to determine the presence ofcarbonates.

The leaves of the litter horizon (OL) were separated by speciesorigin and then weighed so as to know the percentages for each speciesin the litter make-up.

3.3.2. Soil physical and chemical measuresThe following physical and chemical measures were effectuated in

the lab on the soil chunks taken from each station. The presence ofcarbonates was determined by adding a small quantity of dilute HCldirectly onto the horizon. The presence or absence of effervescenceshowed the presence or absence of carbonates.

3.3.2.1. pH measurements. For field pH, several grams of soil werecollected from each horizon, to which a solution was added. Theresulting colour was compared to the pH scale to identify anapproximate pH of that horizon.

The method used was a pH 1:5 soil: water suspension. Five mixedgrams of soil were taken from each soil chunk and placed in a glass jar.A five-fold dilution was created by adding 25 mL of distilled water tothe 5 g of soil. The mixture was then agitated on a shaker table for30 min before being left to rest for 1 h. A calibrated pH meter was usedto measure the pH of the supernatant.

3.3.2.2. Residual water content measurements. The residual watercontent (RH) refers to the percentage of water contained in the soil.This factor makes it possible to convert concentrations for fresh soil intoconcentrations for dry soil.

Five grams of soil were collected in an aluminium weighing tin andkept overnight in an oven set at 105 °C. Upon removal, the tin was leftto cool in a desiccator, before being weighed.

The residual water content was calculated as follows:

RH = (mass of fresh soil − mass of dry soil)/mass of fresh soil

3.3.2.3. Nitrogen concentration measurements. Knowing the mineralnitrogen composition of a soil allows for a better understanding ofthe local ecosystem functioning, while also taking into considerationthe results from the floristic inventory, the litter, and the soil faunainventory. The first step was the extraction of all ions present in the soil,including those aggregated with fine particles. A 1 M KCl solutionserved as an extraction buffer, because of its high ionic strength andstability during the different preparatory stages. To avoid saturation ofthe extraction by the KCl, the 1:5 ratio was respected between soil mass:volume. Saturation would cause the KCl to become a limiting reagent,and thus interfere with the extraction of all ions.

Ten grams of soil were taken from the soil chunk, and 50 mL of KCladded. The mixture was homogenised before being set on a shaker tablefor 45 min. The prepared sample was then filtered (medium wide poresize) and the filtrate collected in a glass jar. The 3 forms of mineralnitrogen were then determined individually.

The method of determination of Nitrates was done in two steps. Thenitrates were first reduced into nitrites by cadmium ions (Cd2 + ).These nitrites could then be determined by using a variant of the Griessmethod. When sulphanilic acid is added, nitrite ions form a diazoniumsalt. This salt then reacts with an amine, α-Naphtylamine, and forms anazo dye that absorbs at 525 nm. Zeroing of the spectrophotometer wasdone with 6 mL of pure filtrate. A small amount of nitrate reagent (HI93728-0) was added to the tube then shaken vigorously for 10 s, fol-lowed by 50 s of less intense mixing. After this, the tube was left alone

Fig. 2. Students in the field at work: a) guided plant classification; b) at the students’ feet,the quadrat placed in the centre of each patch for collecting litter and soil samples; c)there was even a good footballer amongst them. (Color figures in the web version of thisarticle).

A. Terrigeol et al. Applied Soil Ecology xxx (xxxx) xxx–xxx

3

for 4′30 in order that the particles might be able to sediment beforemeasuring absorbance (Fig. 8). The measure given is the concentrationof nitrogen in the nitrate compound [N-NO3-]. In order to have theconcentration of the nitrates, a simple conversion must be done basedon the relation between the different molecular masses [M(NO3-)/M(N-NO3-)].

The method used for the determination of Nitrites is adapted from amethod used by the Environmental Protection Agency of the UnitedStates (EPA). The nitrite ions created with this technique produce apurple-coloured compound that absorbs at 525 nm. Zeroing was onceagain done, but with 10 mL of the pure filtrate. A small amount of

reagent, Nitrite LR reagent (HI 93707-0) was added, and the solutionwas mixed for 15 s. This time, the spectrophotometer measured directlythe concentration of nitrites.

For the Ammonium test, a variant of a method developed by theAmerican Society for Testing and Materials (ATSM) was used. Thistechnique is based on the Nessler method. The Nessler reagent de-composes when in the presence of ammonium ions, and produces di-mercury diiodide. This compound can be used in colorimetry to de-termine the concentration of ammonium ions. The Nessler reagent isobtained by adding two solutions to the filtrate. First, lye or sodiumhydroxide (NaOH) is added, otherwise called solution A, followed by asecond solution B containing anhydrous potassium iodide (KI) and

Fig. 3. The soil profile was presented to the students during the first training day: a) Students take notes and are able to select the right soil horizons during their own investigation duringthe following days; b) Soil horizons laid out on a shovel in order to present them to the students. Horizons’ pH is also compared and discussed in the field c) a larger view of the sameprofile (the question was: can we detect in the soil profile 80 years after, the first signs of introducing conifers in a forest that was naturally without conifers?). (Color figures in the webversion of this article).

Fig. 4. Humus profile at station 1—Calcisol with a Mesomull humus. (Color figures in theweb version of this article).

Fig. 5. Humus profile at station 2—Entic Podzol with a Dysmoder.

A. Terrigeol et al. Applied Soil Ecology xxx (xxxx) xxx–xxx

4

mercury iodide (HGI2). The absorbance can be measured at 420 nm.Zeroing was done with 10 mL of filtrate in a tube. 200 μL of solution Awas added and mixed, then 200 μL of solution B was added and mixed.The spectrophotometric reading is given after a few seconds. Thismeasurement corresponds to the concentration of nitrogen in the am-monium compound [N-NH4 + ].

3.3.3. Litter and soil fauna inventoryThe fauna in the litter was manually collected in each sample taken

from all stations. A classification based on main functional groups wasalso produced. However, in two cases, only a portion of the sample wasused. The results were then extrapolated to the total litter mass.

Two techniques were used for collecting the soil fauna: manualsorting and Berlese-Tullgren funnel (Fig. 9a, b, c, d, e, f, g). Each time,500 g was taken from the soil chunk, and all the litter content (0.25 m2)was observed.

3.3.4. Floristic inventoryThe floristic diversity was evaluated based on the phytosociological

sampling method using the minimal area and the Braun-Blanquetabundance-coverage indices (Fig. 10a). The plot size is defined as “theminimal surface which has as a rule to be occupied by a sample of aplant community if the normal specific assemblage will be able to de-velop." (Westhoff and Van Der Maarel, 1978). Coverage quantifies thesurface that a plant occupies inside a given plot.

The analysis of the plot must be carried out on a plot larger or equalto the minimal area. This is done so that the analysed surface best re-presents the combination of characteristic species. For example, a plotsize of 400 m2 is usually considered representative of a woodland, witha woodland having certain species that characterise its structure.

• herbaceous stratum: plants less than 1 m;

• shrub stratum: plants between 1 m and 7 m;

• tree stratum: plants more than 7 m.

For each stratum, each species was classified (Fig. 10b) and assignedan abundance-coverage coefficient (Braun-Blanquet et al., 1952) whichcan be converted into a mean coverage percentage (Table 1). This isdone to facilitate statistical analyses.

Table 1 Correspondences between Braun-Blanquet coefficients andcoverage percentage.

4. Statistical analyses

ANOVA tests were used to compare stations’ fauna, flora and che-mical composition. However, if the variances were unbalanced, tests onmeans were also conducted. The threshold for the significance level wasset at 0.05.

Species diversity was calculated according to the Shannon Index,with S being the total number of species, ni the number of individuals ofspecies i, and N the total number of individuals.

All statistical analyses were performed using the R software. 3.2.2(R Core Team, 2015).

5. Results

5.1. Forest stations

The observed soil at station 1 is a Calcisol (AFES, 2009) and CambicCalcisol (IUSS Working Group WRB, 2015); with a Mull system and aMesomull humus form (Humusica 1, article 5) (Fig. 4).

The organic horizon is mainly composed of a discontinuous OLnhorizon. The OF and OH horizons are absent. This type of humus cor-responds to a Mesomull humus form. The organo-mineral horizon is acalcic A type (‘Aci’ AFES or IUSS = “Ah” FAO (2006)= biomacros-tructured maA, Humusica 1, article 4) with no reaction to HCl and aneutral pH and few earthworms. It has mainly a clay-loam texture. Thestructural horizon is a calcic Sci (AFES, 2009) equivalent to a Bk (FAO,2006).

There are no coarse calcareous fragments in the A, nor the S hor-izons. The adsorbent complex is almost or completely saturated (S/CEC> 80%) with Ca2+. The C horizon is a mixture of silt andFontainebleau sand with a sandy-loam texture. This is indicative of

Fig. 6. Soil profile at station 3—Luvisol with an Oligomull.

Fig. 7. Soil profile at station 4 —Calcisol with a Mesomull.

Fig. 8. Nitrogen measurements: Photometers multi-parametrical HANNA HI 8320 in-struments and kits for estimating Nitrates, Nitrites and Ammonium.

A. Terrigeol et al. Applied Soil Ecology xxx (xxxx) xxx–xxx

5

Fig. 9. Berlese-Tullgren funnels and samples of collected soil animals: a) Soil is placed in each funnel and lamps are lighted. Animals escape downwards until arriving in a little jar placedunder the funnel; b) small animals are placed in Petri dishes and observed under the microscope: white enchytraeids, spiders, insect larva in the centre and at the low edge, blackStaphylinidae on the left, two large centipedes yellow and stained, a small gamasidae in the centre; c) centipdes at the top, a group of white collembolan, an earthworm egg and very smallOribatid acarian aside a pseudo-scorpion; d) Gamasida acarians; e) Oribatida acarians; f) a springtail on the moon; g) students in laboratory during the third day of training. (Forinterpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).

A. Terrigeol et al. Applied Soil Ecology xxx (xxxx) xxx–xxx

6

leaching from the upper horizons (CaCO3 deposits) as the silt is foundfarther down in the soil after being carried downward by water. Thishorizon has a basic pH and shows effervescence after applying hydro-chloric acid. The numerous rock fragments present in the soil profile areformed by Beauce limestone.

The soil at station 2 is a Podzosol ocrique (AFES, 2009), corre-sponding to Entic Podzols (IUSS Working Group WRB, 2015) with aModer humus system and a Dysmoder humus form (Humusica, art. 5),(Fig. 5).

The substratum is constituted of Fontainebleau sands mixed with“Limon des Plateaux (windy loam). The organic horizon is subdividedinto the following horizons: OL, OH, and OH. The OH is more than 1 cmthick, and the transition from the OH to the A horizon is gradual. The Ais not clearly granular nor lumpy (Ah, AFES and IUSS;miA = biomicrostructured, sometimes msA =massive A, Humusica 1,article 4). This kind of humus is classified as Moder system (gradualtransition between OH and A horizons, Fig. 5) and Dysmoder form(thick OH horizon), in which the litter is slowly decomposed. The or-gano-mineral A horizon has an acidic pH (pHwater< 5). The slowdecomposition contributes to the acidification of this horizon. The

chocolate-coloured podzolic horizon B (BP, or BPhs, AFES = Bhs, IUSS)is where aluminium and iron accumulate. This horizon can be dividedinto a darker humic BPh = Bh (high carbon concentration) and aclearer sesquioxidic BPs = Bhs (Al and Fe are more present). Thecontrast between the horizons A and B is not very distinct, and thetransition is gradual. The podzolisation process is in its beginning stage,which justifies calling this soil a Podzosol ‘ochrique’ (AFES, 2009) or“Entic” Podzols (IUSS Working Group WRB, 2015). The upper horizonsare acidic (pH 4–5) and the loamy sand substratum contains little clay.In this situation, organic acids cause the leaching of the iron and alu-minium contained in clays. As they are carried downwards, the organicacids polymerise into dark-brown humic acids, which are larger andless soluble. The iron and aluminium hydroxides are then freed andwashed further downward where they are caught in the B podzolichorizon.

The soil at station 2 is also polycyclic. Below, the Podzosol AFES(= Podzols, IUSS) is an older Luvisol (same name AFES and IUSS) withboth E (II E) and BT (= Bt IUSS) horizons. The eluvial E horizon isdepleted in clay, while the BT (= Bt) horizon is enriched in clay. Foundat a depth of 120 cm, the BT horizon in not visible in Fig. 5).

The soil at station 3 is classified as a Luvisol (same name AFES andIUSS) with a Mull humus system and an Oligomull humus form(Humusica 1, article 5), (Fig. 6).

The substratum’s first part is composed of Fontainebleau sands. Theorganic horizon contains an OLn, pockets of OLv, and a scarce quantityof OF. There is no OH horizon, and the humus form is an Oligomull. Theorgano-mineral A horizon (Ah, AFES, 2009; FAO, 2006; correspondingto a biomesostructured meA, Humusica 1, article 4) is rather acidic witha pHwater of 4–5. This acidity is unusual in Mull systems and mostlikely a punctual event due to very recent climatic conditions and theincrease of acidophilous species. The leached horizon E is beige-co-loured and has a low clay content due to eluviation or leaching. The pHis slightly acidic (5–6). The argic horizon BT is ochre-coloured and has apH identical to that of E (5–6). The soil is active as plant litter is rapidlytransformed due to microorganisms and earthworms. Nonetheless,there are not enough earthworms to combine clay and organic matter ina clay-humus complex. The biomesostructured A horizon is made oflarge aggregates (> 4 mm) characterised by a very weak structure(they easily break under low pressure between fingers). As a result,leaching is prevalent. Under the BT is a C/R horizon composed of blocksof limestone and clays. The pH is neutral. The bedrock is mainlycomposed of Brie limestone.

The soil at station 4 is a Calcosol (AFES, 2009) = Haplic Calcisol(IUSS, 2015) with a humus Mull and a Mesomull form (Humusica 1, art.5), (Fig. 7).

The organic horizon is made up of a continuous OLn, while the OFand OH horizons are absent. This corresponds to a Mesomull (OLcontinuous) or almost an Eumull (OL discontinuous or absent). Theorgano-mineral horizon is an effervescent to HCl, calcaric (Aca, AFEES,2009 or Ahk, FAO, 2006). It has a lumpy structure because of theearthworms. Its pH is between 7 and 8. The structural horizon is cal-caric (Sca = Bk), with a gradual transition into the A horizon. Thestructure is polyhedric. There are no visible carbonate aggregates fromsedimentary rock formations in the A horizon, but some are present inthe S horizon. In this kind of soil, the absorbing complex is saturated (S/CEC> 95%), mainly by Ca2+. The Aca (Ahk) and Sca (Bk) horizonsshow effervescence with hydrochloric acid. The underlying C horizon isnot visible, but is composed of altered limestone bedrock. This bedrockis called Champigny Limestone.

5.2. Physical and chemical measures of horizon a

The water content, pH, and nitrate, nitrite and ammonia levels foreach station are shown in Fig. 11 (as well as in Table 2 of the Appen-dices). Stations 2 and 3 present an acidic pH ( < 5), distinctly differentfrom the 2 others stations which present a neutral pH (7). The nitrate

Fig. 10. Field (a) and laboratory (b) observation allow to classify animals (morpho-functional groups) and plants (species). (Color figures in the web version of this article).

Table 1Correspondences between Braun-Blanquet coefficients and coverage percentage

Abundance-dominance coefficient Coverage% Mean coverage%

5 75–100 87,54 50–75 62,53 25–50 37,52 5–25 151 1–5 2,5+ <1 0,5

A. Terrigeol et al. Applied Soil Ecology xxx (xxxx) xxx–xxx

7

and ammonium levels are not significantly different across the 4 sta-tions (ANOVA, p-values respectively being 0.63 and 0.16).

On the other hand, station 4 stands out by showing a higher con-centration of nitrites (ANOVA, p-value = 0.0018 and mean test forstations 3 and 4, p-value = 0.039). Moreover, stations 3 and 4 present amean residual humidity rate twice as high as in stations 1 and 2(Fig. 12Figs. 12–14 ), and therefore significantly different (ANOVA, p-value = 0.041).

The higher soil water content in station 4 may promote an anoxicenvironment that weakens the nitratation process. This would result inan increased nitrite content and a decreased nitrate content.

5.3. OL horizon characteristics

As mentioned in § 3.1, the amount of litter is more important on theacidic soils of stations 2 and 3 (Dysmoder and Oligomull) than on thesoils of stations 1 and 4 (Mesomull) (Fig. 12Fig. 13). There is a parti-cularly significantly smaller amount of litter at station 4 (ANOVA, p-

value = 0.044). White rot is more prevalent on leaves from stations 2and 3 than from station 1. The differences between means are con-sidered significant (test on mean between stations 1 and 2, and 1 and 3,p-value = 0.00045 and 0.003 respectively). White rot is caused byfungi and tends to be more present on older litter. Therefore, it is logicalto find more white rot in the Dysmoder and Oligomull of the stations 2and 3 than in the Mesomull of the station 1.

The large range of results at station 4 suggests a problem in theinterpretation due to the small quantity of litter present at this station.It could as well be caused by natural soil heterogeneity. Consideringonly the results obtained from the stations 1 to 3, the white rot per-centage is negatively correlated with soil pH. The regression slope issignificant (p-value = 0.0003).

The leaves at station 1 are mainly from pubescent oak trees. Atstation 2, they come from sessile oak and beech. Station 3 has sessileoak, beech, and hornbeam. Lastly, station 4 has hornbeam, beech,linden and chestnut. The Shannon diversity index (H) calculated onlitter leaves is shown in Fig. 12Fig. 14. The mass is used in place of thenumber of individuals. The H values are significantly different betweenstations (ANOVA, p-value = 0.0003). Station 1 has the lowest leaf litterdiversity, while station 4 has the highest.

5.4. Litter and soil fauna inventory

Fifty-one taxonomical groups were identified, taking into accountdistinctions between adult and larval stages. Taxonomical groups aregenerally identified to the Family, but in certain cases only to the Orderor Super-Family (Appendices, Table 3). Almost 2000 individuals wereidentified from all stations, using various resources (Moreau, 1990;Leraut et al., 2003; Chinery, 2005). A Shannon diversity index wascalculated for each replicate. Adults and larvae of the same taxonomicalgroup were combined. The calculated indices are shown in the upperpart of Fig. 15. Mean Shannon diversity indices are not significantlydifferent in each station for litter nor for soil (ANOVA, p-value = 0.98and 0.26 respectively). This is due to the important variability betweenreplicates. However, the stations 2 and 3 exhibit higher average valuescompared to the stations 1 and 4. More replicates would then be neededin order to strengthen the statistical test.

The number of individuals collected in soil samples is about 10times lower than that of the litter (lower part of Fig. 15). Similarly,diversity indices for soil samples are about half of those calculated forthe litter samples.

The diversity index calculated for the litter is slightly correlatedwith pH (Figure16 Fig. 16 Figs. 16 and 17 a). Two groups are observed,the first one, on the left correspond to the station 2 and 3, while station1 and 4 are on the right. Differences between these two groups aresignificant (p-value < 0.05) with a stronger faunal diversity at lowerpH.

Taxonomical groups have been attached to one or more of the sevenecological functions proposed by Lavelle (1996) (Appendices, Table 3).

For the litter, half of the functions are predator functions, andpredation occurs at all different levels (macrofauna, mesofauna, andmicrofauna). The function of the litter transformer represents about athird of the total for the stations 1 to 3 and one fourth for the station 4(Mesomull). The difference is, however, not significant (ANOVA, p-value = 0.56).

On average, the ecosystem engineer function is better represented atthe stations 1 and 4 than at 2 and 3. But once again, the difference is notsignificant (ANOVA, p-value = 0.6). In the soil, the ecosystem en-gineers are more prevalent in station 4 than in stations 1 and 2. Theobserved difference between averages is significant (ANOVA, p-value = 0.015). Litter transformers are just as well present in the soilsampling. However, we found many litter transformers (compared tothe observed numbers) in the soil too. The differences on the observedaverages for the litter transformers are not significant (ANOVA, p-value = 0.09).

Fig. 11. pH levels and concentration for nitrates, nitrites, and ammonium (mg/kg drymass) for each of the 4 stations, 5 replicates per station.

Table 2Water content (Ralative humidity), pH, and nitrate, nitrite and ammonia levels for eachstation (also in Fig. 11).

Station Replicate RH pH Ammonium Nitrate Nitrite

1 1 13.9 7.3 31.0 0.0 0.231 2 11.0 6.8 21.9 12.4 0.171 3 13.9 7.2 30.7 102.9 0.171 4 21.7 7.3 24.1 59.4 0.261 5 18.0 7.3 23.9 29.7 0.242 1 16.2 3.8 28.9 52.9 0.482 2 35.2 4.3 21.9 51.3 0.312 3 12.0 4.1 20.9 42.8 0.232 4 38.7 5.1 23.2 3.6 0.332 5 15.4 4.5 22.7 44.5 0.183 1 28.4 4.4 27.1 0.0 0.703 2 29.8 5.7 29.4 63.1 0.433 3 16.7 4.6 22.7 71.8 0.243 4 49.5 4.1 29.2 0.0 0.303 5 14.0 4.3 22.0 41.2 0.124 1 27.0 7.7 27.8 0.0 0.414 2 53.0 7.2 61.5 0.0 3.094 3 29.0 7.3 28.5 102.9 2.894 4 18.5 7.3 22.4 38.1 1.414 5 25.0 7.1 28.4 14.8 1.47

A. Terrigeol et al. Applied Soil Ecology xxx (xxxx) xxx–xxx

8

5.5. Floristic inventory

At all 4 stations, 65 plant species were inventoried. The species listis presented in Table 4 of the Appendices. 13 species are listed in thetree stratum, 12 in the shrub stratum, and 40 in the herbaceous stratum.

From a floristic point of view, the 4 stations show some very dif-fering characteristics.

Station 1 is an Oak forest, made up of pubescent oak trees. Thesouthern exposure promotes the growth of sub-Mediterranean speciesin the northernmost limit of their range (Quercus pubescens, Carex hu-milis). Other species found in the tree and shrub strata are Fagus syl-vatica, Sorbus latifolia, Crataegus monogyna, and Ligustrum vulgare. In theherbaceous stratum the dominant species are Brachypodium pinnatum,Euphorbia amygdaloides, Euphorbia cyparissias, and Teucrium chamaedris.

Station 2 is an oak-beech forest (Quercus petraea, Fagus sylvatica)that has associated acidophilous species such as Pteridium aquilinum andCarex pilulifera. The herbaceous stratum is underdeveloped.

Station 3 is an Oak-hornbeam-beech forest (Quercus petraea, Quercusrobur, Carpinus betulus, Fagus sylvatica). Anemone nemorosa is largelyrepresented in the herbaceous stratum as well as Lonicera periclymenum.A few acidophilus species are present (Pteridium aquilinum, Deschampsiaflexuosa, Teucrium scorodonia).

Station 4 is also an Oak forest, but this time containing linden,maple, and European hornbeam (Quercus robur, Tilia platyphyllos, Acerpseudoplatanus, Carpinus betulus). The species-rich herbaceous stratumcontains Mercurialis perennis, Lamium galeobdolon, Ornithogalum pyr-enaicum, Ribes rubrum, Veronica officinalis, etc.).

A Shannon index has been calculated on all replicates, using theBraun-Blanquet abundance-dominance coefficients. Species labelledwith a ‘ + ’ or an ‘r’ are given a coefficient of 0.2 or 0.1 respectively.The calculated indices are given in Figure16 Fig. 16 Fig. 17. The speciesdiversity in the tree stratum is not significantly different between sta-tions (ANOVA test on the mean values of stations 1 and 2, p-value of0.25 and 0.64 respectively).

However, for the shrub stratum, station 2 differs from the stations 1and 4 with a lower level of diversity (test on mean, p-value of 0.016 and0.022 respectively). This difference is also observed for the herbaceousstratum between the stations 1 and 2, but not between the stations 2and 4 (test on mean, p-value of 0.016 and 0.11 respectively). For thissame stratum, station 3 shows a lower mean diversity than the stations1 and 4, but the difference is not significant. More replicates are neededto be more conclusive (test on mean, p-value of 0.07 and 0.26 respec-tively).

The diversity indices, calculated for the shrub and herbaceousstrata, show a slight correlation with the soil pH (Figure16 Fig. 16 b, c).First, we observed a higher diversity for the herbaceous stratum com-pared to the shrub stratum. Secondly, the diversity seems to increasewith the pH. Thirdly, no correlation was found for the tree stratum; thediversity was similar across all stations. Two groups were observed, oneclose to pH 4, which corresponds to stations 2 and 3 and the other one

Figs. 12–14. (12) Water content (relative humidity%) for each of the 4 stations, 5 replicates per station. (13) Masses of leaves per 0.25 m2 litter in the OL horizon (left) and percentages ofwhite rot (right) for each of the 4 stations, 5 replicates per station. (14) Shannon diversity indices calculated for leaf masses in the litter from each of the 4 stations, 5 replicates per station.

Table 3Fifty-one taxonomical groups were identified, taking into account distinctions betweenadult and larval stages. Taxonomical groups are generally identified to the Family, but incertain cases only to the Order or Super-Family. Taxonomical groups have been attachedto litter (O) or soil (A) samples and to one or more of the seven ecological functionsproposed by Lavelle (1996).

Morpho-functional groups O A F1 F2 F3 F4 F5 F6 F7

Araneae 149 3 0 0 0 0 0 0 1Blattoptera 10 0 0 1 0 0 0 0 0Chilopoda_adults 38 5 0 0 0 0 0 0 1Coleoptera_larvae 16 3 0 0.3 0 0 0.3 0.3 0Coleoptera_adults 4 1 0 0.5 0 0 0 0.5 0Coleoptera cantharidae 3 1 0 0 0 0 0 1 0Coleoptera carabidae_adults 2 0 0 0 0 0 0 1 0Coleoptera carabidae_ larvae 5 3 0 0 0 0 0 1 0Coleoptera curculionoidea_adults 23 0 0 1 0 0 0 0 0Coleoptera curculionoidea_larvae 1 0 0 0.5 0 0 0.5 0 0Coleoptera dasytidae_ larvae 0 1 0 0.3 0 0 0.3 0.3 0Coleoptera elateridae_ larvae 1 4 0 0.3 0 0 0.3 0.3 0Coleoptera mycetophagidae 1 0 0 0.5 0 0 0 0.5 0Coleoptera geotrupidae 2 0 0 1 0 0 0 0 0Coleoptera staphylinidae_adults 29 0 0 0 0 0 0 1 0Coleopterasylphidae 1 0 0 0.5 0 0 0 0.5 0Collembola arthropleones 256 6 0.5 0.5 0 0 0 0 0Collembola symphypleones 11 1 0.5 0.5 0 0 0 0 0Diplopoda 14 0 0 1 0 0 0 0 0Diplopoda glomerida_glomeridae 4 0 0 1 0 0 0 0 0Diplopoda polydesmides 19 8 0 1 0 0 0 0 0Diptera_adults 9 0 0 1 0 0 0 0 0Diptera_ larvae 7 13 0 0.3 0 0 0.3 0.3 0Diptera_nymphs 2 0 0 1 0 0 0 0 0Embioptera 1 0 0 1 0 0 0 0 0Enchytraeida 123 43 0 0.5 0.5 0 0 0 0Gastropoda 67 4 0 0 0 0 1 0 0Hemiptera pentatomidae 1 0 0 1 0 0 0 0 0Hemiptera nabidae 1 0 0 1 0 0 0 0 0Thysanoptera 5 0 0 1 0 0 0 0 0Homoptera ciccadelloidea 2 0 0 0 0 0 1 0 0Homoptera coccoidea 16 1 0 0 0 0 1 0 0Hymenoptera apidae 1 0 0 0 0 0 1 0 0Hymenoptera apocrita 7 1 0 0 0 0 0 1 0Hymenoptera cynipoidea 2 0 0 0 0 0 0 1 0Hymenoptera Formicidae 187 1 0 0 0 1 0 0 0Isopoda oniscidea 45 13 0 1 0 0 0 0 0Lepidoptera_adultes 1 0 0 0 0 0 1 0 0Lepidoptera_larves 4 0 0 0.3 0 0 0.3 0.3 0Lumbriculida_anecic earthworms 5 0 0 0 0 1 0 0 0Lumbriculida_ endogeic earthworms 9 39 0 0 0 1 0 0 0Lumbriculida_ epigeic earthworms 23 2 0 0 0 1 0 0 0Mesostigma gamasida 229 13 0 0 0 0 0 1 0Nematoda 26 25 0.5 0.5 0 0 0 0 0Opiliones 1 0 0 0 0 0 0 0 1Sarcoptiformes oribatida 187 3 0 1 0 0 0 0 0Prostigmata trombidiidae 1 1 1 0 0 0 0 0 0Orthoptera gryllidae 5 0 0 0 0 0 1 0 0Psocoptera ectopsocidae 2 0 0 0.5 0 0 0.5 0 0Pseudoscorpionida 57 0 0 0 0 0 0 1 0Trichoptera_larvae 34 0 0 1 0 0 0 0 0

A. Terrigeol et al. Applied Soil Ecology xxx (xxxx) xxx–xxx

9

close to pH 7 corresponding to stations 1 and 4. Differences betweenthose two groups are not significant, neither for the herbaceous stratumnor the shrub stratum.

6. Discussion

Each of the four stations that were analysed in this study differ intheir soil profile, the physical and chemical properties and the char-acteristics of the litter (OL + OF horizons), the first 10 cm of “soilwithout litter” (OH, A, B horizons), the soil fauna and the vegetation.Some differences were higher between some stations, depending on soiland humus systems. Two different groups were observed according totheir pH values. However, stations 1 and 4, which have similar Shannonindexes and pH of the basic topsoil A horizons, were separated from thestations 2 and 3, which show different topsoil horizons, respectivelyacid A and (OH + A + B) horizons. In other words, pH seems to playan important role on the faunal and floral diversity. There is a re-lationship between pH and humus/soil systems as shown in the litera-ture (Toutain, 1981; AFES, 2009; Zanella et al., 2011a, 2011b). Fur-thermore, the humus/soil system influences the vegetation (Ponge,1999, 2003; Bigelow and Canham, 2002), and the optimum soil pH foreach plant formation corresponds to what was found during our study.Along the Fontainebleau transect the change in pH in the first 10 cm ofsoil (under OL and OF horizons, Figs. 4–7), is particularly important, asvery few plant individuals were found at pH levels between 5.0 and 6.5.The pH gradient along the transect (station 1, A horizon: pH about 7,station 2 (OH + A+ B horizons) and 3 (A horizon): pH near 5, station4, A horizon: pH 7) leads to several hypotheses, which will need furtherinvestigation for validation. It is possible that a) the high filtering ca-pacity of the Fontainebleau sands could easily lose its content in basesthrough leaching; b) the light slope of the Fontainebleau sand layer

does not allow for the accumulation of large calcareous colluviumscovering this acid sand layer and the abrupt physical passage fromcalcareous to acid stations is the only one present in the area; c) overtime the forest management of the beech forest growing on Podzol andLuvisol promotes an even-aged stand, which may have facilitated lix-iviation (Luvisol genesis) and podzolisation (Podzol genesis) of the fil-tering substratum, especially under the mature adult stands (as ob-served), where the shrub stratum is systematically cleared away bymanagerial interventions.

Another interesting fact is the higher faunal diversity for low soil pH(around 4) while the floral diversity seems to be higher with a higherpH value (around 7). The pH may not be the only factor explaining thischange in diversity, however it must play an important role. We did notmeasure the average weight of soil and litter fauna per gram, howeverthe diversity may already be a good indicator. It has already beenshown that the humus systems (Mull or Moder) could have some con-sequences on faunal diversity, with a higher diversity in the Mull soil(Schaefer and Schauermann, 1990; Zanella et al., 2011a; Humusica 1,articles 4 and 8 and Humusica 8, 2017). However, because of the lownumber of species observed, it is impossible to draw any firm conclu-sions. Indeed, soil and litter fauna, with their degradation process, canchange the soil reference. Faunal faeces, for example, can increase thepH, stimulate biological activity and increase the humification process(Bachelier, 1978; Toutain, 1981; Zanella et al., 2011b).

Furthermore, it has been shown that plant diversity can have im-portant consequences on faunal diversity (Ponge, 1999, 2003, 2013;Meier and Bowman, 2008; Wardle, 2006). Thus, faunal diversity canalso change soil properties and influence floral diversity and vice-versa.

Station 1: the soil is calcareous with a neutral pH. It is relatively dry,and as a consequence, there are few earthworms. Nonetheless, the litteris classified as Mesomull as it is efficiently degraded. Few fungus species

Fig. 15. Shannon diversity indices calculated for thefauna present in the litter and soil (above) and thetotal number of individuals collected (below), 5 re-plicates per station.

A. Terrigeol et al. Applied Soil Ecology xxx (xxxx) xxx–xxx

10

participate in this transformation as is noted by a lack of white rot.Arthropods are thus largely implicated in the transformation processes.

The large species diversity of the herbaceous and shrub strata islargely favoured by the southern exposure. The tree stratum is muchless developed in comparison to the other stations.

Station 2: the soil, which is very acid, is situated on a sandy subsoil,with a very low humidity and few earthworms. The clay-humus com-plex is fairly absent and thus the cations are not retained. The resultingsoil is not very rich. The topsoil, classified as Dysmoder, is transformedmainly by arthropods, but also by a few fungi. The slow transformationleaves behind large quantities of decaying and partially decayed or-ganic matter, which in turn allows for a large diversity of arthropodsinvolved in each step of the transformation process (the significanceremains to be determined statistically).

The vegetation is limited by the edaphic conditions which favouracidophil plants. Thereby the herbaceous and brush strata are not verydiverse.

Station 3: the acid soil is also positioned on a sandy subsoil, but ismore humid than the previous stations, and thus has a more activehumus. However, there are not enough earthworms to create a clay-humus complex. The litter at the top of an Oligomull, is efficientlydegraded by arthropods, but also by some earthworms and some fungi.The pedofauna diversity is rather important, but this needs to be con-firmed statistically.

The edaphic conditions are less restricting than for the station 2,which allows for a few more plant species to survive. The Shannondiversity index rises slightly for the shrub and tree strata, but onceagain this has yet to be confirmed statistically.

In Fontainebleau forest, the process of podzolisation has been wellinvestigated (Robin, 1968, 1979, 1990, 1993, 2005; Robin et al., 1983;review in Humusica 3, Delaporte et al.: Structural and functional dif-ferences in the belowground compartment of healthy and decliningbeech trees) and our Luvisols and Podzols perfectly fits the descriptionsof soil profiles studied in Robin’s articles. In Fontainebleau beech-oakforests, on sandy substrate (as in stations 1, 2 and 3), the soil fertility isstrongly related to the accessibility of Calcium ions. Stations 1, 2 and 3are placed on Fontainebleau sand between Beauce and Brie limestonelayers (Fig. 1):

Figs. 16 and 17. Relationship between the Shannon diversity indices calculated on a)litter fauna, b) herbaceous and c) shrub strata plants composition and the soil pH. Colourscorrespond to station 1 in black, station 2 in red, station 3 in green, and station 4 in blue.(17) Shannon index has been calculated on all replicates, using the Braun-Blanquetabundance-dominance coefficients. Species labelled with a ‘+’ or an ‘r’ are given acoefficient of 0.2 or 0.1 respectively. (For interpretation of the references to colour in thisfigure legend, the reader is referred to the web version of this article).

Table 4Floristic inventory. At all 4 stations, 65 plant species were inventoried. The species list ispresented in Table 4 of the Appendices. 13 species are listed in the tree stratum, 12 in theshrub stratum, and 40 in the herbaceous stratum.

Tree strutumAcer campestre Fagus sylvatica Quercus roburAcer platanoides Fraxinus excelsior Pinus sylvestrisAcer pseudoplatanus Prunus padus Tilia platyphyllosAesculus hippocastanum Quercus petraea Quercus pubescensCarpinus betulus

Shrub stratumCorylus avellana Ilex aquifolium Prunus spinosaCrataegus monogyna Lonicera periclymenum Sambucus nigraEuonymus europaeus Ligustrum vulgare Sorbus latifoliaHedera helix Prunus avium Viburnum sp

Herbaceous stratumAnemone nemorosa Galium palustre Ranunculus auricomusArum maculatum Geranium robertianum Ribes rubrumBrachypodium pinnatum Geum urbanum Ribes uva-crispaCarex flacca Glechoma hederacea Rubus sp.Carex sylvatica Lamium galeobdolon Ruscus aculeatusCarex humilis Melica uniflora Stacis sp.Carex pendula Pteridium aquilinum Tamus communisCarex pilulifera Hypericum perforatum Teucrium chamaedrisClematis vitalba Neoringia trinervia Teucrium scorodoniaDeschampsia flexuosa Mercurialis perennis Veronica officinalisDactilis glomerata Ornithogalum pyrenaicum Viola sp.Euphorbia amygdaloides Potentilla sterilis UrticadioicaEuphorbia cyparissias Ficaria verna Fragaria vesca

A. Terrigeol et al. Applied Soil Ecology xxx (xxxx) xxx–xxx

11

a) at the top, the Beauce limestone covering as a cap the MontceauButte, may feed the Calcisol of station 1. Applying hydrochloric acidalong the soil profile causes effervescence only at the bottom. A processof lixiviation removes the free CaCO3 dissolved in water, which afterlosing its water may precipitate near the bottom;

b) Beauce limestone transported by gravity water can also feed thestation 2 (on the same sand, in a lower position than station 1) on thecondition of having an impermeable substratum that would preventwater from leaving the system. In station 2, we found the top of a BT(horizon with clay accumulation) from an ancient Luvisol at a depth of120 cm. The pH corresponding to this deep BT horizon was about 6.This may explain the relative fertility of this Fago-Quercetum forest(adult beech and oak trees reaching a height of 25–30 m) even ifcharacterised by acidic humus system and soil;

c) at the bottom, the Brie limestone may furnish Calcium to theforest system of station 3, assuming that this spot be characterised by aLuvisol with a BT horizon that is not compact nor to thick. In someareas of the station 3, the limestone was near to the surface (< 60 cm),hindering deeper prospection in the soil and reducing the volume ofexploitable soil by roots. We believe that station 3 attained a nearlyneutral soil over time which can support a richer biodiversity as ourmeasurements proved.

The combination of the thickness and quality (more or less enrichedin silt or clay) of the Fontainebleau sands, the strength of the lixiviationprocesses, and the clay eluviation or podzolisation influencing nutrientavailability for sustained forest ecosystems, can explain the variabilityin conditions we found in stations 1, 2 and 3. Even though different,stations 2 and 3 were more similar when compared to the calcic station1, which tends to react similarly as the more calcareous station 4.

Station 4: the calcic soil has a pH of 7, corresponding to neutral. Thehumidity is important and the humus is active. The earthworms help informing the clay-humus complex by leaving behind a grumpy structure,which captures the nutriments necessary for plant development. Theorganic matter is efficiently transformed, creating a Mesomull. Theecosystem engineers − earthworms and ants − play a vital role in thisprocess.

The favourable soil conditions here allow for a species-rich vege-tation equal to that of the first station; this applies to each stratum. Thisis surprising with regards to its north-eastern exposure, which is lessfavourable than the first station's south-western exposure. However,this can be explained by the close proximity of water, which canespecially be beneficial during hot summers.

7. Conclusion

Various types of soils and the corresponding pedogenesis processeswere studied in this exercise. This study also analysed the relationshipbetween soils’ physical and chemical properties, the litter and soilfauna, and the corresponding vegetation. Interesting relations betweensoil properties and floral and faunal diversity have been shown. In onlya 500 m gradient, under the same climate conditions, it is very inter-esting to note such differences among soil profils.

The variability in the results between replicates sometimes skewedthe statistical analyses and occasionally limited the possibility to definetrends. To counteract this problem, it will be necessary to do morereplicates and have a better control of the experimental biases.

The litter and soil fauna densities per square metre cited in theliterature did not correspond to our results, but this is most likely due tothe methodology used. A lot more individuals were found in the sta-tions 1 and 4 than in the stations 3 and 4. It was also not possible for usto identify the micro-invertebrates. Molecular tools and techniques suchas DNA barcoding would greatly help in being more precise.

Regarding the recent climatic trend, litter decomposition could bemodified as well as plant communities (Verhoef and Brussaard, 1990).Changes in hygrometry and temperature could also have importantconsequences on the biodiversity of the Butte Montceau, especially if

management interventions are inadequate.

Teaching staff

Stephane Bazot, Université Paris Saclay, director of the Master «Biodiversité Ecologie Evolution », specialised in ecology of terrestrialenvironments

Jean-Christophe Lata, Université Pierre et Marie Curie, director ofthe Master « Sciences de l'Univers et de l'Environnement », specialisedin plant and soil relationships

Jean-Michel Dreuillaux, Université Paris Saclay, plant and phyto-sociology specialist

Jérôme Mathieu, Université Pierre et Marie Curie, specialised instatistical analysis applied to natural environments

Christophe Hannot, Université Paris Saclay, entomologist specia-lised in pedofauna, particularly in Coleoptera.

Augusto Zanella, University of Padua (Italy), Erasmus collaboration,professor of Dendrology and Soil ecology.

Acknowledgments

Teaching staff and students are very grateful to Odile Loison andCéline Hignard of Fontainebleau-Avon Station for warm reception andhelpful support during the duty period. Thank you to Michelle Hannirof the “Master SDUEE − UPMC” Office, and Nicola Benfatto of theInternational Mobility-Scholars Office of the University of Padua, fortheir always prompt and cordial collaborations.

References

AFES, 2009. Association Française pour l’Etude des sols. In: Quae (Ed.), Référentielpédologique 2008, (480 p.).

Bachelier, Georges, 1978. La faune des sols, son écologie et son action. ORSTOM, Paris.Barot, S., Blouin, M., Fontaine, S., Jouquet, P., Lata, J.C., Mathieu, J., 2007. A tale of four

stories: soil ecology, theory, evolution and the publication system. PLoS One. http://dx.doi.org/10.1371/journal.pone.0001248.

Bigelow, S.W., Canham, C.D., 2002. Community organization of tree species along soilgradients in a north-eastern USA forest. J. Ecol. 90 (1), 188–200.

Braun-Blanquet, J., Roussine, N., Nègre, R., 1952. Groupements végétaux de la Franceméditerranéenne.

Chinery, M., 2005. Association des amis du Laboratoire d'entomologie du Muséum(France). In: Legrand, J. (Ed.), Insectes de France et d'Europe occidentale.Flammarion.

Clause, J., Barot, S., Richard, B., Decaëns, T., Forey, E., 2014. The interactions betweensoil type and earthworm species determine the properties of earthworm casts. Appl.Soil Ecol. 83. http://dx.doi.org/10.1016/j.apsoil.2013.12.006.

Dickson, R.E., Broyer, T.C., 1972. Effects of aeration, water supply, and nitrogen sourceon growth and development of tupelo gum and bald cypress. Ecology 53, 626–634.http://dx.doi.org/10.2307/19347764.

FAO, 2006. Guidelines for Soil Description, 4th edition. FAO, Rome.Humusica 1, article 4: Terrestrial humus systems and forms – Specific terms and diag-

nostic horizons. Applied Soil Ecology Special Issue.Humusica 1, article 5: Terrestrial humus systems and forms – Keys of classification of

humus systems and forms. Applied Soil Ecology Special Issue.Humusica 1, article 8: Terrestrial humus systems and forms – Biological activity and soil

aggregates, space-time dynamics. Applied Soil Ecology Special Issue.IUSS Working WRB Group, 2015. World reference base for soil resources 2014, update

2015. International Soil Classification System for Naming Soils and Creating Legendsfor Soil Maps. World Soil Resources Reports No. 106. FAO, Rome.

Lavelle, P., Chauvel, A., Fragoso, C., 1995a. Faunal activity in acid soils. Plant-SoilInteractions at Low pH: Principles and Management. Springer Netherlands, pp.201–211.

Lavelle, P., Lattaud, C., Trigo, D., Barois, I., 1995b. Mutualism and biodiversity in soils.Plant Soil 170 (1), 23–33.

Lavelle, P., 1996. Diversity of Soil Fauna and Ecosystem Function Biologie International,vol. 33. pp. 3––16.

Lavelle, P., 2009. Ecology and the challenge of a multifunctional use of soil. Pesqui.Agropecuària Bras 44, 803–810.

Leraut, P., Blanchot, P., Hodebert, G., 2003. In: Delachaux, Lonay, Niestlé (Eds.), Le guideentomologique.

Meier, C.L., Bowman, W.D., 2008. Links between plant litter chemistry, species diversity,and below-ground ecosystem function. Proc. Natl. Acad. Sci. 105 (50), 19780–19785.

Moreau, A., 1990. In: Boubée, N. (Ed.), Atlas des larves d'insectes de France: Vers blancs,chenilles asticots.

Ponge, J.-F., Arpin, P., Sondag, F., Delecour, F., 1997. Soil fauna and site assessment inbeech stands of the Belgian Ardennes. Can. J. For. Res. 27, 2053–2064. http://dx.doi.

A. Terrigeol et al. Applied Soil Ecology xxx (xxxx) xxx–xxx

12

org/10.1139/x97-169.Ponge, J.F., Jabiol, B., Gégout, J.C., 2011. Geology and climate conditions affect more

humus forms than forest canopies at large scale in temperate forests. Geoderma 162,187–195. http://dx.doi.org/10.1016/j.geoderma.2011.02.003.

Ponge, J.F., Sartori, G., Garlato, A., Ungaro, F., Zanella, A., Jabiol, B., Obber, S., 2014.The impact of parent material, climate, soil type and vegetation on Venetian foresthumus forms: a direct gradient approach. Geoderma 226–227, 290–299. http://dx.doi.org/10.1016/j.geoderma.2014.02.022.

Ponge, J.F., 1999. Horizons and humus forms in beech forests of the Belgian Ardennes.Soil Sci. Soc. Am. J. 63, 1888–1901.

Ponge, J.F., 2003. Humus forms in terrestrial ecosystems: a framework to biodiversity.Soil Biol. Biochem. 35, 935–945.

Ponge, J.F., 2013. Plant-soil feedbacks by humus forms: a review. Soil Biol. Biochem. 57,1048–1060.

R Core Team, 2015. R: A Language and Environment for Statistical Computing. https://www.r-project.org/.

Robin, A., Guillet, B., Duchaufour, P., 1983. Écologie des podzols du bassin parisien:Exemples en forêts de Fontainebleau et Villers-Cotterêts. Revue Forestière Française35, 35–46.

Robin, A., 1968. Contribution à l’étude des processus de podzolisation sous forêt defeuillus These de Doctorat. Université Paris 6.

Robin, A., 1979. Genèse et évolution des sols podzolisés sur affleurements sableux duBassin Parisien Thèse d'état. Université Nancy 1.

Robin, A., 1990. Les sols sur sables soufflés de la forêt de Fontainebleau. Lettre de lasociété botanique française 137, 211–220.

Robin, A., 1993. Catalogue des principales stations forestières de la forêt deFontainebleau. Université Pierre et Marie Curie.

Robin, A., 2005. Épisodes majeurs de la podzolisation en forêt de Fontainebleau (France):Essai de synthèse à l'aide du radiocarbone naturel. Comptes Rendus Geoscience 337,599–608.

Schaefer, M., Schauermann, J., 1990. The soil fauna of beech forests: comparison betweena Mull and a Moder soil. Pedobiologia 34 (5), 299–314.

Toutain, F., 1981. Les humus forestiers: structures et modes de fonctionnement. Rev. For.Fr. 33, 449–477.

Van der Putten, W.H., Bardgett, R.D., Bever, J.D., Bezemer, T.M., Casper, B.B., Fukami, T.,... Suding, K.N., 2013. Plant–soil feedbacks: the past, the present and future chal-lenges. J. Ecol. 101 (2), 265–276.

Verhoef, H.A., Brussaard, L., 1990. Decomposition and nitrogen mineralization in naturaland agroecosystems: the contribution of soil animals. Biogeochemistry 11 (3),175–211.

Wardle, D.A., 2006. The influence of biotic interactions on soil biodiversity. Ecol. Lett. 9(7), 870–886.

Westhoff, V., Van Der Maarel, E., 1978. The Braun-Blanquet approach. Classification ofPlant Communities. Springer Netherlands, pp. 287–399.

Zanella, A., Jabiol, B., Ponge, J.F., Sartori, G., De Waal, R., Van Delft, B., Graefe, U.,Cools, N., Katzensteiner, K., Hager, H., Englisch, M., Brethes, A., Broll, G., Gobat,J.M., Brun, J.-J., Milbert, G., Kolb, E., Wolf, U., Frizzera, L., Galvan, P., Kollir, R.,Baritz, R., Kemmers, R., Vacca, A., Serra, G., Banas, D., Garlato, A., Chersich, S.,Klimo, E., Langohr, R., 2011a. European Humus Forms Reference Base. http://hal.archivesouvertes.fr/docs/00/56/17/95/PDF/Humus_Forms_ERB_31_01_2011.pdf(Accessed 31 May 2017).

Zanella, A., Jabiol, B., Ponge, J.F., Sartori, G., De Waal, R., Van Delft, B., Graefe, U.,Cools, N., Katzensteiner, K., Hager, H., Englisch, M., 2011b. A European morpho-functional classification of humus forms. Geoderma 164, 138–145.

A. Terrigeol et al. Applied Soil Ecology xxx (xxxx) xxx–xxx

13