Page 1

A

Sd

RCTISPa

b

c

d

e

f

g

h

i

j

a

ARRAA

KLMSA

h0

ARTICLE IN PRESSG ModelPSOIL-2039; No. of Pages 9

Applied Soil Ecology xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Applied Soil Ecology

jo ur nal home page: www.elsev ier .com/ locate /apsoi l

oil macroinvertebrate communities and ecosystem services ineforested landscapes of Amazonia

aphael Marichala,b,e,∗, Michel Grimaldib, Alexander Feijoo M.c, Johan Oszwaldd,atarina Praxedese, Dario Hernan Ruiz Coboc, Maria del Pilar Hurtadof,hierry Desjardinsb, Mario Lopes da Silva Juniorg, Luiz Gonzaga da Silva Costag,zildinha Souza Mirandag, Mariana Nascimento Delgado Oliveirag, George G. Brownh,téphanie Tsélouikoa, Marlucia Bonifacio Martinse, Thibaud Decaëns i, Elena Velasquez j,atrick Lavellea,b,f

Université Pierre et Marie Curie (UPMC-Univ Paris 6), UMR BIOEMCO 211, Centre IRD Ile de France, 32 Av. Henri Varagnat, 93143 Bondy Cedex, FranceIRD, UMR BIOEMCO 211, Centre IRD Ile de France, 32 Av. Henri Varagnat, 93143 Bondy Cedex, FranceUniversidad Tecnológica de Pereira, Apartado Aéreo 97, Pereira, ColombiaUniversité de Rennes 2, UMR CNRS LETG 6554, Laboratory of Geography and Remote Sensing COSTEL, FranceMuseu Paraense Emilio Goeldi (MPEG), Coordenac ão de Zoologia, Av. Perimetral, n◦ 1901, CEP 66077-530 Terra Firme, Belém, Pará, BrazilCentro Internacional de Agricultura Tropical (CIAT), TSBF LAC, ap aereo, 6713 Cali, ColombiaUniversidade Federal Rural da Amazonia (UFRA), 2501 Av. Presidente Tancredo Neves, 66077-530 Bairro Montese, Belém, Pará, BrazilEmbrapa Florestas, Estrada da Ribeira, Km. 111, C.P. 319, Colombo PR 83411-000, BrazilUniversité de Rouen, ECODIV, Faculté des Sciences & des Techniques, Bâtiment IRESE A, Place Emile Blondel, F-76821 Mont Saint Aignan Cedex, FranceUniversidad Nacional de Colombia, Carrera 32 No 12-00 Chapinero, Vía Candelaria, Palmira, Valle del Cauca, Colombia

r t i c l e i n f o

rticle history:eceived 7 December 2012eceived in revised form 2 May 2014ccepted 16 May 2014vailable online xxx

eywords:andscapeacro-invertebrates

oil servicesmazonia

a b s t r a c t

Land use changes in the Amazon region strongly impact soil macroinvertebrate communities, whichare recognized as major drivers of soil functions (Lavelle et al., 2006). To explore these relations, wetested the hypotheses that (i) soil macrofauna communities respond to landscape changes and (ii) soilmacrofauna and ecosystem services are linked. We conducted a survey of macrofauna communities andindicators of ecosystem services at 270 sites in southern Colombia (department of Caqueta) and north-ern Brazil (state of Pará), two areas of the Amazon where family agriculture dominates. Sites representeda variety of land use types: forests, fallows, annual or perennial crops, and pastures. At each site weassessed soil macroinvertebrate density (18 taxonomic units) and the following ecosystem service indi-cators: soil and aboveground biomass carbon stock; water infiltration rate; aeration, drainage and waterstorage capacities based on pore-size distribution; soil chemical fertility; and soil aggregation. Signifi-cant covariation was observed between macrofauna communities and landscape metric data (co-inertiaanalysis: RV = 0.30, p < 0.01, Monte Carlo test) and between macrofauna communities and ecosystemservice indicators (co-inertia analysis: RV = 0.35, p < 0.01, Monte Carlo test). Points located in pastureswithin 100 m of forest had greater macrofauna density and diversity than those located in pastures withno forest within 100 m (Wilcoxon rank sum test, p < 0.01). Total macroinvertebrate density was signifi-cantly correlated with macroporosity (r2 = 0.42, p < 0.01), as was the density of specific taxonomic groups:Chilopoda (r2 = 0.43, p < 0.01), Isoptera (r2 = 0.30, p < 0.01), Diplopoda (r2 = 0.31, p < 0.01), and Formicidae

2

(r = 0.13, p < 0.01). Total macroinvertebrate density was also significantly correlated with available soilwater (r2 = 0.38, p < 0.01) as well as other soil-service indicators (but with r2 < 0.10). Results demonstratethat landscape dynamics and composition affect soil macrofauna communities, and that soil macro- fauna density is significantly correlated with soil services in deforested Amazonia, indicating that soil

Please cite this article in press as: Marichal, R., et al., Soil macroinlandscapes of Amazonia. Appl. Soil Ecol. (2014), http://dx.doi.org/10.1

macrofauna have an engineeri

∗ Corresponding author at: CIRAD, UR 34 Perennial Crops, Av. de l’Agropolis, TA B-34/0E-mail address: [email protected] (R. Marichal).

ttp://dx.doi.org/10.1016/j.apsoil.2014.05.006929-1393/© 2014 Elsevier B.V. All rights reserved.

vertebrate communities and ecosystem services in deforested016/j.apsoil.2014.05.006

ng and/or indicator function.© 2014 Elsevier B.V. All rights reserved.

2, 34398 Montpellier Cedex 5, France. Tel.: +33 467616524.

Page 2

ING ModelA

2 Soil Ec

1

DcsmeseRs(bsdomsi

2bAecaefv(ieesfi

(

tgi

2

2

IrBctl(wai

ARTICLEPSOIL-2039; No. of Pages 9

R. Marichal et al. / Applied

. Introduction

Deforestation is still intense in Amazonia (INPE-PRODES, 2010).eforestation has diverse, though largely related, origins: roadonstruction, wood exploitation, cattle ranching, and smallholderettlements (Le Tourneau, 2004). As forest is lost, landscape frag-entation increases (Ferraz et al., 2005), with significant negative

ffects on biodiversity (Laurance et al., 2001). Many studies havehown the importance of land use (Barros et al., 2002; Decaenst al., 2004; Lavelle and Pashanasi, 1989; Mathieu et al., 2004, 2005;ossi et al., 2010) and the influence of spatial heterogeneity atmall scales, from grass tufts to land use effects on soil macrofaunaMathieu et al., 2009). The role of landscape properties has rarelyeen addressed (Decaëns, 2010). Carvalho and Vasconcelos (1999)howed that species richness and density of litter-dwelling antsecreases with forest fragmentation, while Louzada et al. (2010)bserved that landscape configuration influenced dung beetle com-unities in Amazonian savannas. Beyond local effects at the plot

cale, effects of landscape changes on soil macrofauna communitiesn deforested areas of Amazonia remain largely ignored.

The loss in diversity observed at small scales (Mathieu et al.,005) likely affects ecosystem services, which are defined as theenefits people obtain from ecosystems (Millennium Ecosystemssessment, 2005). Soil macrofauna has an acknowledged influ-nce on soil formation, soil hydraulic properties, flood and erosionontrol, nutrient cycling, and primary production through directnd indirect plant stimulation and carbon dynamics (Brussaardt al., 2007; Lavelle, 2002; Lavelle et al., 1997, 2006). Earthworms,or example, are expected to greatly affect water-related ser-ices through their intense bioturbation and burrowing activitiesLavelle et al., 1997). Although soil macrofauna is broadly used as anndicator of soil quality (Rousseau et al., 2012, 2010; Ruiz-Camachot al., 2009; Turbe et al., 2010; Vasconcellos et al., 2013; Velasquezt al., 2007a), few studies have directly assessed the link betweenoil ecosystem services and soil macrofauna communities in theeld (van Eekeren et al., 2010).

To fill this gap, we tested the following two hypotheses:

(i) Soil macrofauna communities respond to landscape composi-tion and dynamics. Abundance and diversity of soil macrofaunais expected to decrease with landscape degradation.

ii) Soil macrofauna and the delivery of ecosystem services arecorrelated, mainly through the densities of soil engineers(earthworms, termites, ants) and soil processes.

To test these hypotheses, we surveyed macrofauna communi-ies and ecosystem services in the diversity of landscapes found in aradient of land-use intensification in deforested areas of Amazonian Colombia and Brazil.

. Materials and methods

.1. Study sites

Sampling was conducted in two regions of Brazil and Colombia.n each country, three groups of nine farms were chosen that cor-espond to landscape units with different histories of colonization.razilian sites, located in the center of Pará State, were recentlyolonized: Palmares II is an old “fazenda” which was invaded byhe “Movimento dos Trabalhadores rurais sem Terra – MST” (Land-ess Workers’ Movement). Farms in Pacajá are located on a trail

Please cite this article in press as: Marichal, R., et al., Soil macroinlandscapes of Amazonia. Appl. Soil Ecol. (2014), http://dx.doi.org/10.1

“Travessão Sul 338”) perpendicular to the Trans-Amazonian high-ay. The Mac aranduba region is occupied by a group of former

gro-extractivist farmers who increasingly rely on cattle ranch-ng. Deforestation started in 1990, 1994 and 1997 in the three

PRESSology xxx (2014) xxx–xxx

landscape units, respectively. The three Colombian landscape units,located in the Caquetá Department (southwestern Colombia),are representative of three dominant farming systems: conven-tional livestock breeding in long-established degraded pastures,agrosylvo-pastoral and agro-forestry systems in the Canelos, Bal-canes and Aguadulce regions, respectively. Deforestation startedbetween 1940 and 1950 at all three Colombian sites.

On each of the 54 farms chosen, five sampling points werelocated equally along a transect corresponding to the longest diag-onal of the farm, thus representing a total of 270 points (135 ineach country). The distance between points (ca. 200 m) was equalto 1/6 of the transect length and varied according to farm area.Macrofauna and soil were sampled from April to June 2008.

2.2. Macrofauna sampling

The TSBF method (Anderson and Ingram, 1993) was used tosample soil invertebrates. At each of the 270 points, a central soilmonolith (25 cm × 25 cm, 20 cm deep) was dug, and two additionalsoil monoliths (25 cm × 25 cm, 10 cm deep) were dug 5 m east andwest from the central monolith. Thus, one sampled unit was com-posed of 3 monoliths. Overall, 810 monoliths were extracted andhand-sorted.

Macrofauna (groups in which more than 90% of individuals arevisible to the naked eye) in the litter and soil was hand-sorted andpreserved in 4% formaldehyde. All individuals were then sorted,counted and classified into the following taxonomic units: Formi-cidae, Isoptera, Blattaria, Diptera, Isopoda, Dermaptera, Hemiptera,Homoptera, Coleoptera (adults and larvae), Orthoptera, Lepi-doptera (larvae), Diptera (larvae), Araneae, Opiliones, Chilopoda,Diplopoda, Gastropoda, and Oligochaeta.

2.3. Land use and landscape analysis

A remote sensing approach was used to characterize landscapedynamics from 1990 to 2007 for each site. Landsat TM and ETM+(30-m spatial resolution, spectral recording adapted to land coveridentification) were acquired during the dry season for each site(1990, 1994, 1998, 2002 and 2007). Field validation measurementswere taken in 2007 and 2008 to classify landscape elements. Eachgeolocated measurement was linked to the spectral signature ofeach landscape element. A confusion matrix determined eight opti-mal classes of landscape elements.

For each site, supervised classification was performed with the2007 Landsat image. The spectral signature of each landscape ele-ment allowed us to reconstitute previous images (1990, 1994, 1998,and 2002). Five classifications for each site from 1990 to 2007 wereproduced by supervised classifications.

Nine farms were analyzed on each site. Multivariate analysiswas used to explain temporal dynamics of patches on each farm.Three-dimensional matrices were built with x (farm), k (land cover)and t (date). Then, the ACT (STATIS) method (Lavit et al., 1994) wasused to identify the (in)stability of spatial patterns over time (eachacquisition is integrated into a date-table). This method is basedon a date-table correlation to identify a trade-off (inter-structuralstep). The second step (intra-structural step) identifies trade-offreproducibility within each date-table. Similar date-tables indicatesimilar landscape spatial structure. This method alone, however,cannot explain the complexity of spatial organization within theagricultural mosaic. Several landscape metrics were necessary toanalyze the spatial organization of the landscape (Lausch andHerzog, 2002). Three groups of landscape metrics were identified:

vertebrate communities and ecosystem services in deforested016/j.apsoil.2014.05.006

fragmentation metrics (“Total Area” (ha), “Edge Density” (m ha−1),and “Mean Patch Density” (m ha−1)), diversity metrics (“PatchRichness”, “Shannon’s Diversity”, “Shannon’s Eveness”, and “Dom-inance index”) and fractal metrics (“perimeter/area”, “Mean Shape

Page 3

ARTICLE ING ModelAPSOIL-2039; No. of Pages 9

R. Marichal et al. / Applied Soil Ec

Table 1Land uses and simplified classification of land uses. X corresponds to land use thatexists in the landscape, but was not sampled for soil macrofauna.

Land use code Land use Simplified classification

P1 Pasture PasturesP2 Pasture with babac u (Attalea

speciosa)Pastures

P3 Pasture with trees PasturesP4 Pasture with scattered trees PasturesC Crop CropsO Orchard Tree plantationsCP Cocoa plantation Tree plantationsAP ac aï (Euterpe oleracea)

plantationTree plantations

PP African palm-tree plantation Tree plantationsAP Agroforestry plantation Tree plantationsSP Fodder shrub plantation Tree plantationsYFP Young fallow after pasture Fallows after pasturesOFP Old fallow after pasture Fallows after pasturesYFC Young fallow after crop Fallows after cropsOFC Old fallow after crop Fallows after cropsB Babac us Tree plantationsF3 Burned forest ForestsF2 Exploited forest ForestsF1 Forest ForestsLF Lowland forest ForestsLP Lowland pasture PasturesLB Lowland bush X

Ipnlc

sWr(

eadaft

2

pssr

pressure-plate apparatus (−30 kPa to −1600 kPa). From these mea-

TI

BS Bare soil XR River or pond X

ndex”). Landscape metrics were calculated using the softwareackage FRAGSTATS (McGarigal et al., 2002). Class area, patch rich-ess and a summary index of landscape integrity, all representing

and-use dynamics (1990–2007), were calculated in a 100 m radiusentered on each of the five sampling points in every transect.

From these metrics, 3D matrices were built to monitor land-cape spatial organization and temporal dynamics at the farm scale.

ard’s hierarchical ascending method was applied to farm facto-ial coordinates using principal component analysis (PCA) and CoASTATIS).

Farms were classified based on a PCA that included landscapelements and landscape spatial organization in 2007. Two factorialnalyses using ACT (STATIS) were applied to a multidimensionalate-table for the years 1990–2002 (Oszwald et al., 2011). In somenalyses, the 24 land-use types were reduced to 6 broad classes:orests, tree plantations, crops, fallows after crop, fallows after pas-ure, and pastures (Table 1).

.4. Ecosystem service assessment

Carbon stocks in soil and plant biomass, soil hydrodynamicroperties and indicators of soil quality were measured at each

Please cite this article in press as: Marichal, R., et al., Soil macroinlandscapes of Amazonia. Appl. Soil Ecol. (2014), http://dx.doi.org/10.1

ampling point (Tables 2 and 3). Several soil properties were mea-ured down to a depth of 30 cm since earthworms and roots mayeach this depth and consequently affect these variables.

able 2ndicators of soil ecosystem services measured: description, units and ranges.

Variable Description

SCS030 Soil carbon stock 0–30 cm deep

Physical Indicator of physical soil quality

Chemical Indicator of chemical soil quality

Morphological Indicator of morphological soil quality

Organic Indicator of organic soil quality

AW010 Plant-available soil water 0–10 cm deep

Macro010 Soil macroporosity 0–10 cm deep

BCStree Tree and shrub carbon biomass

INFIL Water infiltration into the soil

PRESSology xxx (2014) xxx–xxx 3

2.5. Carbon stock

Total carbon and nitrogen contents of the soil were measuredwith a CHNS analyzer on composite samples taken in four trenchesalong a transect (50 m), in the 0–10, 10–20 and 20–30 cm layersbelow the soil surface. These values were adjusted by the bulk den-sity measured in the same trenches and depths to estimate soilcarbon stock.

Allometric equations were used to estimate the carbon stockof aboveground biomass in trees (diameter at breast height, dbh≥10 cm) and bushes (dbh <10 cm, height >2 m) after measuringthe diameter and estimating the height of each tree and bush on50 m × 10 m and 50 m × 5 m plots, respectively, centered on eachpoint (Silva Costa et al., 2012).

2.6. Indices of soil quality

Four indices of soil quality were calculated (Velasquez et al.,2007a,b) which combine several soil morphological (aggregation),physical, chemical, and organic matter characteristics measuredat all points. Soil aggregates with distinct morphology and origin(physical, biogenic and root aggregates) and other soil com-ponents (plant and charcoal debris, stones) from a soil core10 cm × 10 cm × 5 cm deep were separated and weighed after dry-ing according to the method of Velasquez et al. (2007a) andVelasquez et al. (2007b). Sheer strength resistance and verticalresistance of the surface horizon were measured in the field witha hand torcometer and a penetrometer, respectively, and wererepeated at 4 locations. Particle-size distribution and chemicalproperties of the 0–10 cm soil horizon were measured on a com-posite of two pits. The chemical properties measured were pH H2O;CEC (cation exchange capacity); exchangeable K+, Ca2+, Mg2+, Al3+

and NH4+; and extractable phosphorus (Mehlich extractable), using

standard methodologies (Pansu and Gautheyrou, 2006). Each indi-cator was quantified over a common range of 0.1–1.0 (Table 3).

2.7. Indicators of soil water services: infiltration, available waterand macroporosity

Soil water infiltration rate (INFIL) was measured with theBeerkan test (Lassabatere et al., 2006) (250 cm3 of water poured ina simple ring 20 cm in diameter inserted into the soil to a depthof about 1 cm). This test was repeated four times near the soilpits. Measurements of water retention capacities at different waterpotentials followed a specific sampling protocol: cores with undis-turbed structure were taken from one of the five points on eachfarm (4 replicates). Water matrix potentials were −0.3 kPa, −1 kPa,−3 kPa, −10 kPa, −30 kPa, −100 kPa, −300 kPa, −1600 kPa and weremeasured with the sandbox method (0 to −10 kPa) and the Richards

vertebrate communities and ecosystem services in deforested016/j.apsoil.2014.05.006

surements, multiple linear regression models were generated toestimate water retention capacities at different water potentialsfrom the simplest soil variables measured at all points (all soil

Unit Range

Mg.ha−1 25.1–86.6None 0.1–1.0 (arbitrary)None 0.1–1.0 (arbitrary)None 0.1–1.0 (arbitrary)None 0.1–1.0 (arbitrary)Cm 0.2–1.5Cm 0.7–2.9Mg.ha−1 0–1163mm.h−1 16.18–10330.2

Page 4

ARTICLE IN PRESSG ModelAPSOIL-2039; No. of Pages 9

4 R. Marichal et al. / Applied Soil Ecology xxx (2014) xxx–xxx

Table 3Indicators of soil ecosystem services (mean ± SE) among land uses.

Variable Forests Fallows after crops Fallows after pasture Tree plantations Pastures Crops

Chemical 0.40 ± 0.03 0.47 ± 0.05 0.48 ± 0.02 0.52 ± 0.01 0.49 ± 0.01 0.59 ± 0.05Physical 0.49 ± 0.40 0.38 ± 0.05 0.48 ± 0.04 0.49 ± 0.03 0.48 ± 0.02 0.46 ± 0.05Organic 0.54 ± 0.02 0.49 ± 0.02 0.62 ± 0.04 0.66 ± 0.03 0.61 ± 0.01 0.55 ± 0.03Morphological 0.33 ± 0.01 0.34 ± 0.03 0.36 ± 0.02 0.40 ± 0.02 0.43 ± 0.01 0.36 ± 0.03SCS030 (Mg.ha−1) 45.06 ± 1.55 41.65 ± 2.28 51.84 ± 1.94 53.19 ± 1.60 50.43 ± 0.92 43.88 ± 2.43BCStree (Mg.ha−1) 156.59 ± 26.23 48.78 ± 14.76 48.01 ± 13.50 48.95 ± 12.46 6.48 ± 1.43 1.06 ± 0.47

−1 9 ± 48 ± 08 ± 0

poc−4

2

mPSbeauiKaarairldrlu(tAGa2

3

aMep

3“

isb((

INFIL (mm.h ) 2471.02 ± 290.48 1738.94 ± 384.52 2408.6AW010 (cm) 0.67 ± 0.03 0.62 ± 0.06 0.8Macro010 (cm) 1.74 ± 0.06 2.02 ± 0.13 1.6

hysical characteristics, CEC and carbon content). Soil macrop-rosity and plant-available soil water to a depth of 10 cm werealculated as the water volume drained between saturation and10 kPa (pF2) and between −30 kPa (pF 2.5) and −1600 kPa (pF.2), respectively.

.8. Statistical analyses

Soil fauna abundance was converted into density per squareeter (ind. m−2) for each site. Since the pantropical earthworm

ontoscolex corethrurus (Glossoscolecidae), native to the Guianahield (Righi, 1984), has a different response to ecosystem distur-ance than native earthworms (Marichal et al., 2010), we dividedarthworms into two categories for the analysis: P. corethrurusnd native earthworms. This resulted in a total of 18 taxonomicnits. Since the Shapiro–Wilk test (Siegel and Castellan, 1988)

ndicated non-normality of the data, we used the non-parametricruskal–Wallis rank sum test (Hollander and Wollfe, 1973; Kruskalnd Wallis, 1952) to compare macrofauna density and diversitymong landscape units. We used the non-parametric Wilcoxonank sum test to compare soil macrofauna diversity and densityt points in pastures within 100 m of a forest to those at pointsn pastures further than 100 m from a forest. All tables had 270ows, containing data measured at each point. The soil macrofauna,andscape metrics, ecosystem services and soil ecosystem servicesirectly influenced by soil macrofauna had 20, 26, 9 and 8 rows,espectively. We performed PCA of macrofauna community data,n(x + 1) transformed to reduce the effect of dominant taxonomicnits, and tested the effect of land use types with a Monte Carlo testManly, 1991). Co-inertia analyses were performed to test covaria-ions among datasets (Doledec and Chessel, 1994; Dray et al., 2003).ll statistical analyses were performed with R software (Ihaka andentleman, 1996; R Development Core Team, 2009) using the pack-ges ade4 (Chessel et al., 2004; Dray and Dufour, 2007; Dray et al.,007) and vegan (Oksanen et al., 2008) for multivariate analysis.

. Results

In total, we collected 26,375 invertebrates (6001 in Colombiand 20,374 in Brazil). Mean density was 520.9 ± 38.4 (S.E.) ind. m−2.aximum density was 5301.0 ind. m−2 in a primary Brazilian for-

st, with a minimum of 21.3 ind. m−2 in a degraded Colombianasture.

.1. Soil macroinvertebrate density and diversity amonglandscape units”

The mean number of taxonomic units per sample var-ed significantly among landscape units (Kruskal–Wallis rank

Please cite this article in press as: Marichal, R., et al., Soil macroinlandscapes of Amazonia. Appl. Soil Ecol. (2014), http://dx.doi.org/10.1

um test, p < 0.01), from 5.0 ± 0.3 taxa per sample (Colom-ian “Conventional”) to 10.0 ± 0.3 taxa per sample (Pacajá)Fig. 1a). Mean density varied significantly among “landscape units”Kruskal–Wallis rank sum test, p < 0.01), from 208.6 ± 23.4 ind. m−2

61.58 3114.89 ± 605.84 786.02 ± 132.1 2189.36 ± 517.25.04 0.99 ± .04 0.87 ± 0.02 0.76 ± 0.05.07 1.64 ± 0.05 1.64 ± 0.03 1.8 ± 0.1

(agrosylvo-pastoral) to 1038.7 ± 141.5 ind. m−2 (Mac aranduba)(Fig. 1b). Formicidae, the most abundant taxa, had densi-ties varying from 54.9 ± 5.9 ind. m−2 (Colombian “Conventional”)to 403.8 ± 103.8 ind. m−2 (Mac aranduba), while Isoptera densityranged from 9.2 ± 5.5 ind. m−2 (Colombian “Conventional”) to382.3 ± 83.0 ind. m−2 (Mac aranduba) (Table 4).

3.2. Effect of land use on soil macroinvertebrate communities

The first two axes of the PCA for macroinvertebrate commu-nities accounted for 33.3% of the explained inertia (25.3% and8.0%, respectively, Fig. 2). Axis 1 clearly contrasted the pantropi-cal earthworm P. corethrurus, on the positive side of the axis, withall other taxa, especially Diplopoda, Chilopoda, Gastropoda, Formi-cidae, Coleoptera and Isoptera, on the negative side. The projectionof sites on the factorial plane suggested that the first axis representsa land-use gradient, from primary forest (F), fallows after crops (Fc),cultures (C) to fallows after pastures (Fp), tree plantation (Tp) andpastures (P). Differences in communities with different types ofland use were significant (20% of variance explained, p < 0.01, MonteCarlo test) despite the lack of a visible effect along Axis 2, whichranked sites mainly according to Araneae and Blattaria density.

3.3. Landscape and soil macroinvertebrate communities

Significant covariation (co-inertia analysis: RV = 0.30, p < 0.01,Monte Carlo test) was observed among macrofauna communi-ties and landscape metrics. The first axis of the co-inertia analysis(81.4% variance explained, Fig. 3) associated high densities oflitter-dwelling invertebrates (Diplopoda, Chilopoda, Gastropoda,Collembola) with forest (F1: primary, F2: exploited, LF: low lyingand F3: burned forest) and percentage of fallow areas and patchrichness. The second axis of the co-inertia analysis (4.9% of vari-ance explained) associated P. corethrurus density with palm-treeplantations, agroforestry plantations, and percentage of pastureswith scattered trees. Formicidae, native earthworms and Isopteradensities inversely covaried with pasture, fodder shrub plantationsand orchards.

Pasture points located within 100 m of a forest had greatermacrofauna density and diversity than pasture points more than100 m from a forest (Wilcoxon rank sum test, p < 0.01). Thesepoints may, however, have other types of land use such as fal-lows or crops, which stresses the importance of forest. The densityof Homoptera, Coleoptera, Formicidae, Isoptera, Diptera, Isopoda,Chilopoda, Diplopoda and Gastropoda followed a similar pattern(Table 5). Only the density of Isopoda and P. corethrurus werehigher at pasture points more than 100 m from forest than at pointswithin 100 m (Wilcoxon rank test, p < 0.05), whereas native earth-

vertebrate communities and ecosystem services in deforested016/j.apsoil.2014.05.006

worms, Dermaptera, Hemiptera, Orthoptera, Lepidoptera, Blattariaand Araneae and Opiliones showed no significant differences.Points in forests with no pasture within 100 m had higher diversity(Wilcoxon rank test, p = 0.058) and Diplopoda density (Wilcoxon

Page 5

ARTICLE IN PRESSG ModelAPSOIL-2039; No. of Pages 9

R. Marichal et al. / Applied Soil Ecology xxx (2014) xxx–xxx 5

F ndscapt outlieo

rs

3

wceaiiostalci

TM

ig. 1. Taxa per sample (a) and density (b) of soil macrofauna (all groups) under lahe upper quartile, with whiskers extending to the most extreme data point unless

pen circles. Ind: individuals. (p < 0.01, Kruskal–Wallis rank sum test).

ank sum test, p < 0.05) than those with pasture within 100 m. Noignificant differences were observed for the other groups (Table 6).

.4. Soil macrofauna and ecosystem services

Significant covariation (RV = 0.35, p < 0.01, Monte Carlo test)as also observed among macrofauna communities and the indi-

ators of ecosystem services (Fig. 4). The first axis of co-inertiaxplained 91.6% of the variance and associated Isoptera, Formicidaend native earthworm densities with macroporosity and infiltrabil-ty. P. corethrurus density was associated with the morphologicalndicator. The second axis of the co-inertia analysis explainednly 2.8% of the variance. Total macroinvertebrate density wasignificantly correlated with macroporosity (r2 = 0.42, p < 0.01), par-icularly Chilopoda (r2 = 0.43, p < 0.01), Isoptera (r2 = 0.30, p < 0.01)

Please cite this article in press as: Marichal, R., et al., Soil macroinlandscapes of Amazonia. Appl. Soil Ecol. (2014), http://dx.doi.org/10.1

nd Diplopoda (r2 = 0.31, p < 0.01) densities. P. corethrurus density,ocated on the opposite side of axis 1, was significantly but barelyorrelated with soil aggregation as measured by the morphologicalndex (r2 = 0.04, p < 0.05). Macrofauna communities also covaried,

able 4ean densities (ind. m−2) of macrofauna taxonomic groups among landscape units. Stand

Country Brazil

Landscape unit Mac aranduba Pacaja Palmar

Dermaptera 2.4 (0.7) 0.4 (0.4) 0.2 (Hemiptera 7.8 (2.3) 3.1 (0.7) 13.4 (Homoptera 1.2 (0.5) 2 (0.6) 0.9 (Coleoptera 62.9 (5.8) 60.6 (7.9) 64.4 (Formicidae 403.8 (103.8) 314.6 (83.7) 130 (19Isoptera 382.3 (83.0) 312 (72.5) 179 (52Orthoptera 0.8 (0.3) 1.1 (0.4) 2.3 (Lepidoptera (larvae) 3 (1.4) 0.7 (0.3) 2.1 (Diptera 10.1 (3.9) 9.7 (1.5) 5.1 (Dictyoptera 2.8 (0.6) 2.7 (0.6) 3.6 (Araneae 15.5 (2.7) 9.7 (1.4) 19.1 (Opiliones 0.2 (0.2) 0.6 (0.3) 0.1 (Isopoda 5.1 (2.7) 2.3 (0.6) 10 (2.5Native earthworms 43.3 (6.3) 26.9 (4.4) 46.7 (P. corethrurus 35 (7.7) 13.9 (2.6) 2.4 (Chilopoda 18.8 (2.7) 36.6 (5.2) 33.7 (Diplopoda 36.5 (7.9) 23 (3.7) 19.4 (Gastropoda 7.2 (1.9) 11.3 (3.9) 12.4 (

e units (logarithmic scale). The boxplots show the lower quartile, the median andrs (more than 1.5 times the interquartile range) are present, which are indicated as

although to a lesser extent (RV = 0.34, p < 0.01, Monte Carlo test),with the set of ecosystem services assumed to be influenced by fau-nal activities: carbon stocks; water infiltration; plant-available soilwater; macroporosity created by biological activities; soil organicmatter; and soil physical quality, aggregation and fertility mea-sured by organic, physical, morphological and chemical indicators.

4. Discussion

4.1. Macroinvertebrates, land use and landscape composition

Density and diversity of soil macrofauna were higher in theless deforested landscape units (Brazilian units of Pacajá andMac aranduba, with 40–70% of forest remaining) than in almostcompletely deforested areas (Colombian landscapes, with <10%

vertebrate communities and ecosystem services in deforested016/j.apsoil.2014.05.006

of forest remaining). They were significantly influenced by landuse at the sampling points. Density of all taxa except P. corethru-rus decreased along a gradient of deforestation and land-useintensification, from landscapes dominated by forests and/or

ard error in parentheses.

Colombia

es Agroforesty Agrosylvo-pastoral Conventional

0.2) 0(0) 0(0) 0(0)3.1) 3.2 (0.9) 3.1 (0.6) 2.5 (0.6)0.4) 0(0) 0(0) 0(0)5.4) 6.5 (1.3) 10.9 (2) 14.5 (2.5).7) 59.6 (7.3) 51.7 (4.6) 54.9 (5.9).7) 62.8 (13.0) 10.1 (5.0) 9.2 (5.5)0.6) 0.4 (0.2) 1.7 (0.6) 1.4 (0.4)0.7) 0.6 (0.3) 1.2 (0.5) 1.4 (0.7)0.9) 0.1 (0.1) 0.1 (0.1) 0.5 (0.5)0.8) 3.1 (0.7) 5 (1.1) 3.1 (0.9)3.2) 6.6 (1.2) 10.9 (3.1) 10.5 (2)0.1) 1.7 (0.6) 0.6 (0.3) 0(0)) 3 (1.3) 2.7 (1) 2.6 (2)

12) 41.2 (6.1) 24.8 (6.1) 17.8 (3.7)1.8) 83.7 (13.9) 82.6 (20.5) 109.3 (20.8)4.8) 1.3 (0.5) 0.4 (0.2) 0.4 (0.2)3.2) 0.5 (0.3) 3 (1.2) 0.4 (0.2)2.8) 0(0) 0(0) 0(0)

Page 6

ARTICLE IN PRESSG ModelAPSOIL-2039; No. of Pages 9

6 R. Marichal et al. / Applied Soil Ecology xxx (2014) xxx–xxx

Fig. 2. Ordination of the sampled sites in the factorial plane of a principal component analysis of community structure. (a) Correlation circle. Homo: Homoptera, Col:Coleoptera, Isopt: Isoptera, Ara: Araneae, Orth: Orthoptera, Lepi: Lepidoptera, N. earth: native earthworms, Derm: Dermaptera, Diplo: Diplopoda, Dip: Diptera, Blat: Blattaria,C ra, Opa e: F: foP

saGaleS

TMw

n

hilo: Chilopoda, Gast: Gastropoda, For: Formicidae, Isopo: Isopoda, Hemi: Hemiptexes. Letters correspond to the barycenters of sites sampled in each type of land us: pastures (Monte Carlo test on land uses significant, p < 0.01, Observation = 0.13).

econdary and planted tree covers to landscapes dominated bynnual crops and pastures. Densities of Diplopoda, Chilopoda andastropoda, mostly associated with litter abundance in forested

Please cite this article in press as: Marichal, R., et al., Soil macroinlandscapes of Amazonia. Appl. Soil Ecol. (2014), http://dx.doi.org/10.1

reas, decreased with deforestation. Formicidae and Isoptera fol-owed a similar trend. An opposite trend was observed for thearthworm P. corethrurus, an invasive species from the Guyanesehield (Righi, 1984) associated with human-created land uses,

able 5ean macrofauna densities (ind. m−2) ± SE at pasture points with and without forestithin a 100 m radius.

Pastures: forest within 100 m radius Wilcoxontest

No Yes

Total macrofauna 279.40 ± 26.09 650.00 ± 102.99 ***

Dermaptera 0.16 ± 0.11 1.18 ± 0.87 nsHemiptera 3.10 ± 0.52 4.15 ± 1.71 nsHomoptera 0.00 ± 0.00 0.99 ± 0.40 ***

Coleoptera 18.32 ± 2.61 70.67 ± 11.84 ***

Formicidae 69.50 ± 12.92 166 ± 36.09 **

Isoptera 34.15 ± 10.43 276.60 ± 81.62 ***

Orthoptera 1.19 ± 0.29 1.38 ± 0.73 nsLepidoptera (larvae) 1.50 ± 0.40 3.16 ± 2.22 nsDiptera 0.67 ± 0.27 9.87 ± 6.03 ***

Blattaria 3.47 ± 0.61 1.97 ± 0.71 nsAraneae 10.04 ± 1.67 8.09 ± 1.96 nsOpiliones 0.10 ± 0.07 0.00 ± 0.00 nsIsopoda 2.64 ± 1.03 2.17 ± 0.59 *

Native earthworms 25.62 ± 3.66 28.62 ± 7.37 nsP. corethrurus 102.30 ± 13.13 32.57 ± 10.63 *

Chilopoda 0.93 ± 0.35 20.33 ± 5.15 ***

Diplopoda 4.61 ± 1.96 16.58 ± 4.97 ***

Gastropoda 0.98 ± 0.50 4.54 ± 1.01 ***

s: p > 0.05.* p < 0.05.

** p < 0.01.*** p < 0.001.

i: Opiliones. (b) Ordination of the sampled sites in the plane defined by the first tworests, Fc: fallows after crop, Fp: fallow after pasture, C: crops, Tp: tree plantations,

especially pastures (Barros et al., 2003; Lavelle et al., 1987; Marichalet al., 2010; Rossi et al., 2010; Sanchez-de Leon et al., 2004). Fallowsderived from pastures or crops were distinctly different, the latterbeing projected closer to forests on PCA axes 1 and 2 than the for-mer. Soil compaction during the pasture stage (Barros et al., 2003;

vertebrate communities and ecosystem services in deforested016/j.apsoil.2014.05.006

Desjardins et al., 2000) probably had an impact on soil fauna at thefallow stage. Furthermore, since pastures are generally establishedafter the crop phase, more time may have elapsed since deforesta-tion. This is probably another reason why fallows after crops, which

Table 6Mean macrofauna densities (ind. m−2) ± SE at forest points with and without pasturewithin a 100 m radius.

Forest: pasture within 100 m radius Wilcoxontest

No Yes

Total macrofauna 966.10 ± 211.30 808.20 ± 195.43 nsDermaptera 2.17 ± 0.91 0.97 ± 0.57 nsHemiptera 7.50 ± 3.17 3.15 ± 0.83 nsHomoptera 1.58 ± 0.56 0.73 ± 0.53 nsColeoptera 53.30 ± 6.88 52.57 ± 6.18 nsFormicidae 400.30 ± 161.11 395.60 ± 167.44 nsIsoptera 309.30 ± 111.15 209.60 ± 70.73 nsOrthoptera 1.58 ± 0.48 0.97 ± 0.44 nsLepidoptera (larvae) 1.97 ± 0.65 0.73 ± 0.40 nsDiptera (larvae) 9.67 ± 2.51 10.42 ± 2.33 nsBlattaria 3.75 ± 0.84 4.36 ± 1.43 nsAraneae 21.71 ± 5.34 21.08 ± 3.57 nsOpiliones 0.59 ± 0.33 0.97 ± 0.45 nsIsopoda 10.07 ± 4.97 6.06 ± 2.44 nsNative earthworms 40.67 ± 2.44 27.13 ± 4.97 nsP. corethrurus 11.84 ± 3.24 7.75 ± 2.92 nsChilopoda 35.34 ± 6.87 37.31 ± 6.50 nsDiplopoda 41.65 ± 10.89 20.35 ± 4.88 *

Gastropoda 12.44 ± 3.44 7.99 ± 4.78 ns

ns: p > 0.05.* p < 0.05.

Page 7

ARTICLE IN PRESSG ModelAPSOIL-2039; No. of Pages 9

R. Marichal et al. / Applied Soil Ecology xxx (2014) xxx–xxx 7

Fig. 3. Results of co-inertia analysis between soil macrofauna and landscape metrics at the point level (radius of 50 m). (a) Contribution of soil macrofauna taxa densities(identified by their positions on the first two co-inertia axes) to the correlation with landscape metrics. Homo: Homoptera, Col: Coleoptera, Isopt: Isoptera, Ara: Araneae, Orth:Orthoptera, Lepi: Lepidoptera, N.earth: native earthworms, Derm: Dermaptera, Diplo: Diplopoda, Blat: Blattaria, Chilo: Chilopoda, Gast: Gastropoda, For: Formicidae, Isopo:Isopoda, Dip: Diptera, Hemi: Hemiptera, Opi: Opiliones. (b) Contribution of landscape metrics (identified by their positions on the first two co-inertia axes) to the correlationwith macrofauna taxa densities. PR: patch richness, Ltypo: landscape dynamic typology, landscape composition (% of land use): P1: pasture, P2: pasture with babac u, P3:p ntatioS pasturf wland

gfe

ata

sds1nfttbofcwopf“

asture with trees, P4: pasture with scattered trees, C: crops, O: orchard, CP: cocoa plaP: fodder shrub plantation, YFP: young fallow after pasture, OFP: old fallow after

orest, F2: exploited forest, F1: forest, LF: lowland forest, LP: lowland pasture, LB: lo

enerally start 2–3 years after deforestation, have greater macro-auna density and diversity than fallows after pastures (Mathieut al., 2005).

Fallows after crops had similar macrofauna diversity and densitys forests, which confirms that they may contribute significantlyo the conservation of soil macrofauna (Mathieu et al., 2005) andboveground biodiversity (Barlow et al., 2007).

Percentages of pasture, old fallows after pasture and fodderhrub plantations were associated with low soil fauna density andiversity. Our results indicated that land-use composition (mea-ured by patch richness and percentages of land use types within a00 m radius around the sampling point) is an important determi-ant of macrofauna communities. For example, pasture points with

orest within 100 m had higher macrofauna densities and diversityhan pasture points with no forest within 100 m. Most macroinver-ebrate taxa followed this pattern, showing that nearby forest cane a “source” of forest fauna in a pasture (Dias, 1996; Pulliam, 1988)r can increase the suitability of environmental conditions for soilauna by affecting a pasture’s microclimate. An exception was P.orethrurus, which had higher densities in pastures with no forestithin 100 m than in pastures with forest within 100 m. Diversity

Please cite this article in press as: Marichal, R., et al., Soil macroinlandscapes of Amazonia. Appl. Soil Ecol. (2014), http://dx.doi.org/10.1

f soil macrofauna and Diplopoda densities at forest points withasture within 100 m were lower than at forest points with onlyorest within 100 m. This pattern could be a consequence of theedge effect” associated with fragmentation (Gascon et al., 2000;

n, AP: ac aï plantation, PP: African palm-tree plantation, AP: agroforestry plantation,e, YFC: young fallow after crop, OFC: old fallow after crop, B: babac us, F3: burned

bush, BS: bare soil, R: river or pond. RV = 0.30, p < 0.01.

Laurance et al., 2001), since wind and light penetrating deeply intoa forest can change its animal and plant communities (Didham et al.,1998).

Time since deforestation of the original forest ecosystem andrate of deforestation (expressed in the landscape integrity param-eter) was the most important factor determining diversity anddensities of soil macrofauna communities, followed by the relativepercentage of forest cover. This result emphasizes the importanceof time since deforestation, regardless of land use (Mathieu et al.,2005).

4.2. Macroinvertebrates and soil ecosystem services

Soil invertebrates are both actors and indicators of soil services.Soil invertebrates, mainly earthworms, are known to consider-ably influence types and rates of soil services in diverse ways(Blouin et al., 2013; Lavelle et al., 2006). An increase in theirdensities is thus expected to increase the provision of services(an effect mechanism). This type of relation may also result fromcommunity responses to increases in services caused by other fac-tors (a response mechanism), such as organic matter storage or

vertebrate communities and ecosystem services in deforested016/j.apsoil.2014.05.006

nutrient accumulation in clayey soils (Lavelle and Spain, 2001).Invasion by P. corethrurus results in a significant increase in soilmacroaggregation due to the high production of solid casts in thesoil (Lavelle et al., 1994). Significant covariation observed between

Page 8

ARTICLE IN PRESSG ModelAPSOIL-2039; No. of Pages 9

8 R. Marichal et al. / Applied Soil Ecology xxx (2014) xxx–xxx

Fig. 4. Results of co-inertia analysis between soil macrofauna and indicators of ecosystem services. (a) Contribution of soil macrofauna taxa densities (identified by theirpositions on the first two co-inertia axes) to the correlation with ecosystem service indicators. Homo: Homoptera, Col: Coleoptera, Isopt: Isoptera, Ara: Araneae, Orth:Orthoptera, Lepi: Lepidoptera, N. earth: native earthworms, Derm: Dermaptera, Diplo: Diplopoda, Dip: Diptera, Chilo: Chilopoda, Gast: Gastropoda, For: Formicidae, Isopo:Isopoda, Blat: Blattaria, Hemi: Hemiptera, Opi: Opiliones. (b) Contribution of soil ecosystem services (identified by their positions on the first two co-inertia axes) to thec s; Phyc p), AWs soil.

nkomoaSbean(1ttpsb(1ai

5

dtpvhsGlp

orrelation with macrofauna taxa densities. BCStree: tree and shrub carbon biomashemical and organic quality, respectively; SCS030: soil carbon stock (0–30 cm deeoil Macro010: soil macroporosity (0–10 cm deep), INFIL: water infiltration into the

ative earthworm density and macroporosity is consistent with thenown effects of earthworm burrowing and bioturbation activitiesn these indicators of soil services (Blouin et al., 2013). However,esocosm experiments have shown that effects of earthworms

n soil functioning depend on land use (Hedde et al., 2013); thisspect requires further investigation in the Amazonian context.oil macrofauna influences several soil properties, but the distri-ution of species can be influenced by soil properties (Marichalt al., 2012; Vasconcellos et al., 2013). Covariation with carbon inboveground plant biomass, in contrast, simply reflects the fact thatative earthworm density is higher in forests than in other land usesFragoso et al., 1997; Marichal et al., 2010; Rombke and Verhaagh,992). Significant covariations of Formicidae and Isoptera densi-ies with plant-available soil water and macroporosity also suggesthat these taxa influence the ecosystem services indicated by theseroperties (Cowan et al., 1985; Lockaby and Adams, 1985). Theseoil ecosystem engineers can dig networks of galleries and cham-ers, with known effects on macroporosity, aeration and infiltrationFolgarait, 1998; Lobry de Bruyn and Conacher, 1990; Mando et al.,996). This is not the case for Chilopoda or Diplopoda, which takedvantage of existing forest conditions; these taxa are thus morendicators of than actors on soil services.

. Conclusion

Soil macrofauna communities respond to land use, landscapeynamics and composition, and their links to soil services makeshem good indicators of soil ecosystem services. Therefore, theotential of using soil invertebrates to assess soil ecosystem ser-ices is confirmed by our study. Further studies are needed,owever, to directly quantify effects of soil macrofauna on soil

Please cite this article in press as: Marichal, R., et al., Soil macroinlandscapes of Amazonia. Appl. Soil Ecol. (2014), http://dx.doi.org/10.1

ervices and link them to biological “effect” traits (Lavorel andarnier, 2002) involved in supplying services. This next step would

ead to the development of soil-service indicators calculated fromercentages or frequencies of biological traits.

sical, morphological, chemical, organic: indicators of soil physical, morphological,010: plant-available soil water (0–10 cm deep), INFIL: water infiltration into the

RV = 0.35, p < 0.01.

Acknowledgements

This work was a part of the AMAZ project (ANR-06-PADD-001-011 and ANR 06 BIODIV 009-01, coordinator: Patrick Lavelle)supported by the ANR (Agence Nationale de la Recherche, France)and CNPq (Conselho Nacional de Desenvolvimento Científico e Tec-nológico) and jointly implemented by IRD (Institut de Recherche pourle Développement), UFPA (Universidade Federal do Pará), UFRA (Uni-versidade Federal Rural da Amazônia), MPEG (Museu Paraense EmílioGoeldi), UTP (Universidad Tecnológica de Pereira–Colombia), andCIAT (Colombia). We thank Andres F. Carvajal for helping to sampleand separate macrofauna, Edward Guevara for landscape metrics,Max Sarrazin for soil analysis, and all those who participated infield and laboratory activities (students, technicians, farmers, fieldworkers). We also thank ECOS Colombia (AMAZ AGREG project,ECOS6 Nord/COLCIENCIAS//ICETEX).

References

Anderson, J.L., Ingram, J.S.I., 1993. Tropical Soil Biology and Fertility. A Handbook ofMethods, 2nd ed, Wallingford, UK.

Barlow, J., Gardner, T.A., Araujo, I.S., Avila-Pires, T.C., Bonaldo, A.B., Costa, J.E., Espos-ito, M.C., Ferreira, L.V., Hawes, J., Hernandez, M.I.M., Hoogmoed, M.S., Leite, R.N.,Lo-Man-Hung, N.F., Malcolm, J.R., Martins, M.B., Mestre, L.A.M., Miranda-Santos,R., Nunes-Gutjahr, A.L., Overal, W.L., Parry, L., Peters, S.L., Ribeiro-Junior, M.A.,da Silva, M.N.F., da Silva Motta, C., Peres, C.A., 2007. Quantifying the biodi-versity value of tropical primary, secondary, and plantation forests. PNAS 104,18555–18560.

Barros, E., Grimaldi, M., Sarrazin, M., Chauvel, A., Mitja, D., Desjardins, T., Lavelle,P., 2003. Soil physical degradation and changes in macrofaunal communities inCentral Amazon. Appl. Soil Ecol. 26, 157–168.

Barros, E., Pashanasi, B., Constantino, R., Lavelle, P., 2002. Effects of land-use sys-tem on the soil macrofauna in western Brazilian Amazonia. Biol. Fertil. Soils 35,338–347.

Blouin, M., Hodson, M.E., Delgado, E.A., Baker, G., Brussaard, L., Butt, K.R., Dai, J.,

vertebrate communities and ecosystem services in deforested016/j.apsoil.2014.05.006

Dendooven, L., Peres, G., Tondoh, J.E., Cluzeau, D., Brun, J.J., 2013. A review ofearthworm impact on soil function and ecosystem services. Eur. J. Soil Sci. 64,161–182.

Brussaard, L., de Ruiter, P.C., Brown, G.G., 2007. Soil biodiversity for agriculturalsustainability. Agric. Ecosyst. Environ. 121, 233–244.

Page 9

ING ModelA

Soil Ec

C

C

C

D

D

D

D

D

D

D

D

D

F

F

F

G

H

H

I

I

K

L

L

L

LL

L

L

L

L

LL

L

L

L

ARTICLEPSOIL-2039; No. of Pages 9

R. Marichal et al. / Applied

arvalho, K.S., Vasconcelos, H.L., 1999. Forest fragmentation in central Amazoniaand its effects on litter-dwelling ants. Biol. Conserv. 91, 151–157.

hessel, D., Dufour, A.B., Thioulouse, J., 2004. The ade4 package-I-One-table meth-ods. R News 4, 5–10.

owan, J.A., Humphreys, G.S., Mitchell, P.B., Murphy, C.L., 1985. An assessment ofpedoturbation by two species of mound-building ants, Camponotus intrepidus(Kirby) and Iridomyrmex purpureus (F. Smith). Aust. J. Soil Res. 22, 95–107.

ecaëns, T., 2010. Macroecological patterns in soil communities. Global. Ecol. Bio-geogr. 19, 287–302.

ecaens, T., Jimenez, J.J., Barros, E., Chauvel, A., Blanchart, E., Fragoso, C., Lavelle,P., 2004. Soil macrofaunal communities in permanent pastures derived fromtropical forest or savanna. Agric. Ecosyst. Environ. 103, 301–312.

esjardins, T., Lavelle, P., Barros, E., Brossard, M., Chapuis-Lardy, L., Chauvel, A.,Grimaldi, M., Guimaraes, F., Martins, P., Mitja, D., Muller, M., Sarrazin, M., TavaresFiho, J., Topall, O., 2000. Dégradation des pâturâges amazoniens. Description d’unsyndrome et de ses déterminants. Etude et gestion des sols 7, 353–378.

ias, C.D., 1996. Sources and sinks in population biology. Trends Ecol. Evol. 11,326–330.

idham, R.K., Hammond, P.M., Lawton, J.H., Eggleton, P., Stork, N.E., 1998. Beetlespecies respond to tropical forest fragmentation. Ecol. Monogr. 68, 295–323.

oledec, S., Chessel, D., 1994. Co-inertia analysis: an alternative method for studyingspecies-environment relationships. Freshw. Biol. 31, 277–294.

ray, S., Chessel, D., Thioulouse, J., 2003. Co-inertia analysis and the linking of eco-logical tables. Ecology 84, 3078–3089.

ray, S., Dufour, A.B., 2007. The ade4 package: implementing the duality diagramfor ecologists. J. Stat. Softw. 22, 1–20.

ray, S., Dufour, A.B., Chessel, D., 2007. The ade4 package-II: two-table and K-tablemethods. R News 7, 47–52.

erraz, S.F.D., Vettorazzi, C.A., Theobald, D.M., Ballester, M.V.R., 2005. Landscapedynamics of Amazonian deforestation between 1984 and 2002 in central Ron-donia, Brazil: assessment and future scenarios. For. Ecol. Manag. 204, 67–83.

olgarait, P.J., 1998. Ant biodiversity and its relationship to ecosystem functioning:a review. Biodivers. Conserv. 7, 1221–1244.

ragoso, C., Brown, G.G., Patron, J.C., Blanchart, E., Lavelle, P., Pashanasi, B., Senapati,B., Kumar, T., 1997. Agricultural intensification, soil biodiversity and agroecosys-tem function in the tropics: the role of earthworms. Appl. Soil Ecol. 6, 17–35.

ascon, C., Williamson, G.B., da Fonseca, G.A.B., 2000. Receding forest edges andvanishing reserves. Science 288, 1356–1358.

edde, M., Bureau, F., Delporte, P., Cecillon, L., Decaens, T., 2013. The effects of earth-worm species on soil behaviour depend on land use. Soil Biol. Biochem. 65,264–273.

ollander, M., Wollfe, D.A., 1973. Non-parametric statistical methods. John Wiley &Sons, New York.

haka, R., Gentleman, R., 1996. R: a language for data analysis and graphics. Comput.Graph. Stat. 5, 299–314.

NPE-PRODES, 2010. Monitoramento da floresta amazônica brasileira por satélite.www.obt.inpe.br/prodes

ruskal, W.H., Wallis, W.A., 1952. Use of ranks in one-criterion variance analysis. J.Am. Stat. Assoc. 47, 583–621.

assabatere, L., Angulo-Jaramillo, R., Ugalde, J.M.S., Cuenca, R., Braud, I., Haverkamp,R., 2006. Beerkan estimation of soil transfer parameters through infiltrationexperiments – BEST. Soil Sci. Soc. Am. J. 70, 521–532.

aurance, W.F., Lovejoy, T.E., Vasconcelos, H.L., Bruna, E.M., Didham, R.K., Stouffer,P.C., Gascon, C., Bierregaard, R.O., Laurance, S.G., Sampaio, E., 2001. Ecosystemdecay of Amazonian forest fragments: a 22-year investigation. Conserv. Biol. 16,605–618.

ausch, A., Herzog, F., 2002. Applicability of landscape metrics for the monitoring oflandscape change: issues of scale, resolution and interpretability. Ecol. Indic. 2,3–15.

avelle, P., 2002. Functional domains in soils. Ecol. Res. 17, 441–450.avelle, P., Barois, I., Cruz, I., Fragoso, C., Hernandez, A., Pineda, A., Rangel, P., 1987.

Adaptative strategies of Pontoscolex corethrurus (Glossoscolecidae, Oligochaete),a peregrine geophageous earthworm of the humid tropics. Biol. Fertil. Soils 5,188–194.

avelle, P., Bignell, D., Lepage, M., Wolters, V., Roger, P., Ineson, P., Heal, O.W., Dhillion,S., 1997. Soil function in a changing world: the role of invertebrate ecosystemengineers. Eur. J. Soil Biol. 33, 159–193.

avelle, P., Dangerfield, M., Fragoso, C., Eschenbrenner, V., Lopez-Hernandez, D.,Pashanasi, B., Brussaard, L., 1994. The relationship between soil macrofauna andtropical soil fertility. In: Swift, M.J., Woomer, P. (Eds.), Tropical Soil Biology andFertility. John Wiley-Sayce, New York, pp. 137–169.

avelle, P., Decaen, T., Aubert, M., Barot, S., Blouin, M., Bureau, F., Margerie, P., Mora,P., Rossi, J.-P., 2006. Soil invertebrates and ecosystem services. Eur. J. Soil Biol.42, S3–S15.

avelle, P., Pashanasi, B., 1989. Soil macrofauna and land management in PeruvianAmazonia (Yurimaguas, Loreto). Pedobiologia 33, 283–291.

avelle, P., Spain, A.V., 2001. Soil ecology. Kluwer Scientific Publications, Amsterdam.avorel, S., Garnier, E., 2002. Predicting changes in community composition and

ecosystem functioning from plant traits: revisiting the Holy Grail. Funct. Ecol.16, 545–556.

avit, C., Escoufier, Y., Sabatier, R., Traissac, P., 1994. The act (statis method). Comput.

Please cite this article in press as: Marichal, R., et al., Soil macroinlandscapes of Amazonia. Appl. Soil Ecol. (2014), http://dx.doi.org/10.1

Stat. Data Anal. 18, 97–119.e Tourneau, J.-F., 2004. Jusqu’au bout de la forêt? Causes et mécanismes de la

déforestation en Amazonie brésilienne, M@ppemonde 75.obry de Bruyn, L.A., Conacher, A.J., 1990. The role of termites and ants in soil mod-

ification: a review. Aust. J. Soil Res. 28, 88–93.

PRESSology xxx (2014) xxx–xxx 9

Lockaby, B.G., Adams, J.C., 1985. Pedoturbation of a forest soil by fire ants. Soil Sci.Soc. Am. J. 49, 220–223.

Louzada, J., Lima, A.P., Matavelli, R., Zambaldi, L., Barlow, J., 2010. Community struc-ture of dung beetles in Amazonian savannas: role of fire disturbance, vegetationand landscape structure. Landsc. Ecol. 25, 631–641.

Mando, A., Stroosnijder, L., Brussaard, L., 1996. Effects of termites on infiltration intocrusted soil. Geoderma 74, 107–113.

Manly, B.F.J., 1991. Randomization and Monte Carlo Methods in Biology. Chapmanand Hall, London.

Marichal, R., Feijoo, M.A., Praxedes, C., Ruiz, D., Carjaval, A.F., Oszwald, J., del PilarHurtado, M., Brown, G.G., Grimaldi, M., Desjardins, T., Sarrazin, M., Decaens, T.,Velasquez, E., Lavelle, P., 2010. Invasion of Pontoscolex corethrurus (Glossoscole-cidae, Oligochaeta) in landscapes of the Amazonian deforestation arc. Appl. SoilEcol. 46, 443–449.

Marichal, R., Grimaldi, M., Mathieu, J., Brown, G.G., Desjardins, T., da Silva, M.L.,Praxedes, C., Martins, M.B., Velasquez, E., Lavelle, P., 2012. Is invasion of defor-ested Amazonia by the earthworm Pontoscolex corethrurus driven by soil textureand chemical properties? Pedobiologia 55, 233–240.

Mathieu, J., Grimaldi, M., Jouquet, P., Rouland, C., Lavelle, P., Desjardins, T., Rossi,J.P., 2009. Spatial patterns of grasses influence soil macrofauna biodiversity inAmazonian pastures. Soil Biol. Biochem. 41, 586–593.

Mathieu, J., Rossi, J.-P., Grimaldi, M., Mora, P., Lavelle, P., Rouland, C., 2004. A multi-scale study of soil macrofauna biodiversity in Amazonian pastures. Biol. Fertil.Soils 40, 300–305.

Mathieu, J., Rossi, J.-P., Mora, P., Lavelle, P., Martins, P.F.D.S., Rouland, C., Grimaldi, M.,2005. Recovery of soil macrofauna communities after forest clearance in EasternAmazonia, Brazil. Conserv. Biol. 19, 1598–1605.

McGarigal, K., Cushman, S.A., Neel, M.C., Ene, E., 2002. FRAGSTATS: Spatial PatternAnalysis Program for Categorical Maps. Computer Software Program Producedby the Authors at the University of Massachusetts, Amherst, USA, Available athttp://www.umass.edu/landeco/research/fragstats/fragstats.html

Millennium Ecosystem Assessment, www.millenniumassessment.orgOksanen, J., Kindt, R., Legendre, P., O’Hara, B., Simpson, G.L., Stevens, M.H.A., Wagner,

H., 2008. vegan: Community Ecology Package.Oszwald, J., Gond, V., Doledec, S., Lavelle, P., 2011. Identification d’indicateurs de

changement d’occupation du sol pour le suivi des mosaïques paysagères. Boiset Forêts des Tropiques 307, 7–21.

Pansu, M., Gautheyrou, J., 2006. Handbook of Soil Analysis: Mineralogical, Organicand Inorganic Methods. Springer.

Pulliam, H.R., 1988. Sources, sinks, and population regulation. Am. Nat. 132,652–661.

R-Development-Core-Team, 2009. R: A Language and Environment for StatisticalComputing. R Foundation for Statistical Computing, Vienna, Austria, ISBN 3-900051-07-0, Available at http://www.R-project.org

Righi, G., 1984. Pontoscolex (Oligochaeta, Glossoscolecidae), a new evaluation. Stud.Neotrop. Fauna Environ. 19, 159–177.

Rombke, J., Verhaagh, M., 1992. About earthworm communities in a rain forest andan adjacent pasture in Peru. Amazoniana XII, 29–49.

Rossi, J.-P., Celini, L., Mora, P., Mathieu, J., Lapied, E., Nahmani, J., Ponge, J.F., Lavelle,P., 2010. Decreasing fallow duration in tropical slash-and-burn agriculture alterssoil macroinvertebrate diversity: a case study in southern French Guiana. Agric.Ecosyst. Environ. 135, 148–154.

Rousseau, G.X., Deheuvels, O., Arias, I.R., Somarriba, E., 2012. Indicating soil qualityin cacao-based agroforestry systems and old-growth forests: the potential ofsoil macrofauna assemblage. Ecol. Indic. 23, 535–543.

Rousseau, G.X., Silva, P.R.d.S., de Carvalho, C.J.R., 2010. Earthworms, antsand other arthropods as soil health indicators in traditional and no-fireagro-ecosystems from Eastern Brazilian Amazonia. Acta Zool. Mex. 26,117–134.

Ruiz-Camacho, N., Velasquez, E., Pando, A., Decaens, T., Dubs, F., Lavelle, P.,2009. Indicateurs synthétiques de la qualité du sol. Etude et gestion des sols16.

Sanchez-de Leon, Y., Zou, X., Borges, S., Ruan, H., 2004. Recovery of native earth-worms in abandoned tropical pastures. Conserv. Biol. 17, 999–1006.

Siegel, S., Castellan, N.J., 1988. Non Parametric Statistics for the Behavioural Sciences.MacGraw Hill Int., New York.

Silva Costa, L.G., Miranda, I.S., Grimaldi, M., Silva, M.L., Mitja, D., Lima, T.T.S., 2012.Biomass in different types of land use in the Brazil’s ‘arc of deforestation’. For.Ecol. Manage. 278, 101–109.

Turbe, A., De Toni, A., Benito, P., Lavelle, P., Lavelle, P., Ruiz, N., Van der Putten, W.H.,Labouze, E., Mugdal, S., 2010. Soil biodiversity: functions, threats and tools forpolicy makers. Bio Intelligence Services, IRD and NIOO. Report for EuropeanCommission (DG Environment)., pp. 250.

van Eekeren, N., de Boer, H., Hanegraaf, M., Bokhorst, J., Nierop, D., Bloem, J.,Schouten, T., de Goede, R., Brussaard, L., 2010. Ecosystem services in grass-land associated with biotic and abiotic soil parameters. Soil Biol. Biochem. 42,1491–1504.

Vasconcellos, R.L.F., Segat, J.C., Bonfim, J.A., Baretta, D., Cardoso, E.J.B.N., 2013. Soilmacrofauna as an indicator of soil quality in an undisturbed riparian forest andrecovering sites of different ages. Eur. J. Soil Biol. 58, 105–112.

Velasquez, E., Lavelle, P., Andrade, M., 2007a. GISQ, a multifunctional indicator of

vertebrate communities and ecosystem services in deforested016/j.apsoil.2014.05.006

soil quality. Soil Biol. Biochem. 39, 3066–3080.Velasquez, E., Pelosi, C., Brunet, D., Grimaldi, M., Martins, M., Rendeiro, A.C., Barrios,

E., Lavelle, P., 2007b. This ped is my ped: visual separation and near infraredspectra allow determination of the origins of soil macroaggregates. Pedobiologia51, 75–87.