The influence of habitat type on the distribution of carabid beetles in traditionally managed “dehesa” ecosystems in NW Spain

2006·0.48 MB·

by  Angela Taboada

©EntomologicaFennica.11October2006
The influence of habitat type on the distribution
of carabid beetles in traditionally managed “dehesa”
ecosystems in NW Spain
AngelaTaboada,D.JohanKotze,JoséM.Salgado&ReyesTárrega
Taboada,A.,Kotze,D.J.,Salgado,J.M.&Tárrega,R.2006:Theinfluenceof
habitat type on the distribution of carabid beetles in traditionally managed
“dehesa”ecosystemsinNWSpain.—Entomol.Fennica17:284–295.
Weinvestigatedtheresponsesofcarabidbeetlestohabitattypeatthelocalscale
intraditionallymanaged“dehesa”ecosystemsinNWSpain.Thebeetleswere
sampled from two habitat types (central grassland and surrounding forested
zone)offive“dehesa”ecosystems,fromMaytoOctober2004.Atotalof4,374
carabidindividualsand55specieswerecapturedusingpitfalltrapping.Mostof
thespecieswerecollectedfrombothhabitattypes,probablyduetoopenspaces
createdbetweenold-growthtreesintheforestedzone.Differencesincarabidas-
semblagecompositionwerefoundmainlyduetoHarpalusspeciescharacteristic
fromthecentralgrasslandandafewforestspecialistspeciesexclusivelyfoundin
theforestedzone.Themaintenanceofthe“dehesa”structuremaybeofimpor-
tanceforcarabidspeciesadaptedtothevarietyofhabitatscreated,andrequires
thecontinuationoftraditionalmanagementpractices.
A. Taboada and J. M. Salgado, Department of Animal Biology, University of
León, Campus de Vegazana s/n, E-24071 León, Spain; E-mail: dbaatp
@unileon.es
D.J.Kotze,DepartmentofBiologicalandEnvironmentalSciences,P.O.Box65
(Biocenter3,Viikinkaari1),FI-00014UniversityofHelsinki,Finland
R.Tárrega,DepartmentofEcology,GeneticsandMicrobiology,Universityof
León,CampusdeVegazanas/n,E-24071León,Spain
Received12December2005,accepted1March2006
1.Introduction diversity(Gómez-Limón&DeLucíoFernández
1999).A“dehesa”isanagro-forestrysystemwith
Historicallanduseandmanagementpractices,in- extensivegrazing(sheep,goats,pigsandcattle);
cludingforestburningandcuttingtocreateagri- thesemultipleland-usesproducealandscapeof
cultural land and pastures for livestock grazing, high habitat heterogeneity (Blanco et al. 1997).
havealteredanddiversifiedthelandscapeinthe Theresultinganthropogenicmosaic(pasture,ar-
Iberian Peninsula (Blanco et al. 1997, Luis- able,woodland)producesarangeofenvironmen-
Calabuigetal.2000).Insomecases,thesetradi- talconditionsandnichesthatcanenhancediver-
tional practices have transformed the forested sity(e.g.,Pulido&Díaz1997,Martín&López
landtocreate“dehesa”ecosystems,wheretheuse 2002).
ofresourcescoexistswithareasofhighbiological The origin and evolution of the “dehesa”
ENTOMOL.FENNICAVol.17 (cid:127) CarabidbeetlesindehesaecosystemsinNWSpain 285
Figure1.Compositionofthedehesaecosystemanditsrelationtothesurroundingoakforest.a)Aerialviewand
b)detailedview.Habitattypes:G=Centralgrassland,F=Forestedzone,O=Surroundingoakforest.
landscape have been intrinsically related to branches(toprovideshade,browsingandacorn
transhumance pastoral systems for centuries production), and c) the enhancement of tree re-
(Blanco et al. 1997). Well-developed “dehesa” generation by the planting of new seedlings
systemswereestablishedasapermanentfeature (Blancoetal.1997).
oftheSpanishlandscapefromtheearly Middle Hereweinvestigatetheresponsesofcarabid
Ages,whenclearing,thinningandploughingof beetlestohabitattype(definedbythevariability
continuous oak forests were frequently per- inthecompositionofthevegetationlayer)atthe
formed. These management practices were de- local scale in traditionally managed “dehesa”
veloped in order to obtain regularly-spaced re- ecosystemsinNWSpain.WeselectedaQuercus
tainedtrees(between30and60treesperha)and pyrenaica“dehesa”systemthattypicallyconsists
to enhance grassland growth among the trees oftwomainhabitattypes:1)acentralgrassland
(Blancoetal.1997,Joffreetal.1999,Pulidoet areaand2)asurroundingforestedzonecharac-
al.2001).Mostcontemporary “dehesa”ecosys- terised by scattered old-growth trees, which are
temswerecreatedduringthesecondhalfofthe maintainedforlivestockshelterandfeedingpur-
19thandthebeginningofthe20thcenturiesasare- poses(Fig.1).
sult of increased human densities (Pulido et al. We aim to determine if the carabid assem-
2001). blages inhabiting the “dehesa” ecosystemdiffer
The degree of habitat diversity (i.e., differ- betweenthecentralgrasslandandthesurround-
encesinvegetationstructureandcomposition)in ingforestedzonedespitetheirspatialproximity.
a “dehesa” ecosystem is mainly dependent on We ask whether differences in environmental
managementpracticesanddifferentproductsob- conditionsduetotreedevelopmentintheforested
tained (e.g., forage, firewood, charcoal, acorn, areainfluencethepresenceofcarabidspeciesac-
cork,tannin).Basically,thevegetationstructure cording to their habitat preferences [see Reyes-
ofa“dehesa”iscomposedoftwolayers(Joffreet Lópezetal.(2003)forants].Shadeprovidedby
al. 1999): a savannah-like open tree layer with thescatteredtreesintheforestedzoneincreases
scattered matureoaks (mainly Quercus ilex and microhabitat humidity, decreases temperature
Q. suber) and a grass dominated layer (usually andcontributestoleaflitterdevelopmentandor-
native annual herbs for grazing). The mainte- ganicmattercontent(Joffreetal.1999),leading
nanceofthistwo-layeredvegetationstructurere- alsotomodificationsinthevegetationcomposi-
quires continuous management such as: a) con- tion.Thesechangesinenvironmentalconditions
trolofshrubinvasionbyuprooting,clearingand may make the forested zone suitable for forest
ploughing,b)preservationoftheold-growthtree specialist species, even though these zones are
layer by regular pruning to favour horizontal notconsideredtrueforestecosystems.Openhab-
286 Taboadaetal. (cid:127) ENTOMOL.FENNICAVol.17
itat species are expected to thrive in the central caryophyllea, Cynosurus cristatus, Medicago
grasslandandgeneralistspeciesareexpectedto lupulina, Ornithopus compressus, Trifolium ar-
belessinfluencedbythedifferencesbetweenthe vense,T.campestre,T.glomeratum,T.pratense,
twohabitattypes. T. strictum, and Vulpia bromoides (Díez et al.
1994, 1995). The surrounding forested zone is
characterised by a more complex vegetation
2.Materialandmethods structure consisting of several layers: oak trees,
oak saplings, shrubs and annual and perennial
2.1.Studyareaandsamplingmethod herbs(Table1).
Vegetation sampling in these forested zones
ThestudywasconductedinNWSpain(42º36’– were carried out in June–July 2004: (a) under-
42º39’N,4º58’–5º05’W),intheMediterranean story vegetation cover (shrubs, herbs and sap-
Region 980–1,030 m a.s.l. We selected five lings)wasvisuallyestimatedin20systematically
Quercus pyrenaica “dehesa” ecosystems [Fes- placed1m×1mquadratspersite;(b)oaktree
tuco heterophyllae–Querceto pyrenaicae sig- characteristics(perimeter,height,canopydiame-
metum;Penasetal.(1995)]atleast1kmapart, ter and distance between trees) were measured
surroundedbycontinuousoakforestsandwitha from40treespersite.
similargrazingintensity(sheepandgoats).Each Weusedplasticpitfalltraps(depth86mm,di-
“dehesa” ecosystem included the two habitat ameter60mm)coveredby10cm×10cmroofs
types described above, and we sampled carabid and partly filled with 25% propylene glycol to
beetlesinfivecentralgrasslandsites(sizerange collectthebeetles.Pitfallcatchesreflectcarabid
2.09–8.23ha)andfivesurroundingforestedsites activity-density(e.g.,Thomas1998)thatwillbe
(sizerange11.50–24.64ha). referredtohereasabundanceornumberofindi-
Grazedandopencentralgrasslandsaredomi- viduals.Fivesamplingpointswereplacedineach
nated by a variety of herb species such as Aira ofthetensites.Eachsamplingpointconsistedof
Table1.Structuralcharacteristicsofthevegetationatthefiveforestedzones.Dominantherb(annualandperennialherbs)and
scarce(<5%cover)shrubspeciesateachzonearegiven.Oaksaplings(<1mheight)areindicatedwhencoverwasgreater
than2%.Mean±StandardDeviationdistancebetweenoaktrees,treeheight,perimeterandcanopydiameterarealsogiven.
Annualherbs Perennialherbs Shrubs Mean±SD Mean±SD Mean±SD Mean±SD
treedist. treeht. treeperim. canopy
(m) (m) (cm) diam.(m)
1 Airacaryophyllea Agrostiscapillaris Crataegusmonogyna 7.70±3.76 8.46±4.42 73.33±71.47 6.49±4.66
Cynorosusechinatus Festucarubra
Tuberariaguttata Plantagolanceolata
2 Tuberariaguttata Agrostiscapillaris Ericacinerea 7.65±5.32 11.03±4.14 123.40±112.32 7.45±4.97
Carexmuricata Halimiumumbellatum
Festucarubra Oaksaplings
3 Airacaryophyllea Agrostiscapillaris Halimiumumbellatum 9.66±5.17 10.20±4.70 98.68±102.69 6.83±4.50
Cynosurusechinatus Oaksaplings
Tuberariaguttata
Vulpiabromoides
4 Airacaryophyllea Agrostiscapillaris Thymusserpyllum 7.18±4.34 10.18±4.96 87.80±91.43 6.08±4.27
Tuberariaguttata Hieraciumcastellanum
Vulpiabromoides Hieraciumgr.pilosella
Plantagolanceolata
5 Airacaryophyllea Agrostiscapillaries Callunavulgaris 5.53±3.13 13.88±2.10 86.20±48.26 6.64±3.03
Trifoliumcampestre Carexmuricata Ericaumbellate
Festucarubra Lithodoradiffusa
Hieraciumgr.pilosella Thymusserpyllum
Oaksaplings
ENTOMOL.FENNICAVol.17 (cid:127) CarabidbeetlesindehesaecosystemsinNWSpain 287
Table2.Generalisedlinearmodelresults.df=degreesoffreedom.ResidualDevianceistheamountofvariationnotexplained
bythemodels;pvaluesinparentheses(boldfaceifp<0.05;¤2distribution).TheCentralgrassland(Grassl.)coefficientsarethe
basicinterceptsofthemodels;Forest=forestedzone.Theta(withSE)istheaggregationvalue(fornegativebinomialmodels
only)withlowervaluesindicatingahigherdegreeofaggregation.
Resid.Dev.(p) Coeff.
NullDev. Size Ants Habitattype Size Ants Grassl. Forest Theta(SE)
Residualdf 9 8 7 6
Abundance 272.79 264.40(0.004) 263.56(0.361)16.91(<0.001) –0.051 9.49x10–5 –1.248 0.870
RarefiedRichness 4.26 4.25(0.902) 3.79(0.497) 3.79(0.976) 0.009 1.85x10–4 2.675 –0.009
Calathusfuscipes 14.56 14.17(0.536) 14.17(0.946) 10.80(0.067) –0.039 3.94x10–4 –3.419 1.409 2.00(0.86)
Carabusamplipennis 44.85 11.52(<0.001) 9.35(0.140) 7.44(0.167) 0.093 –1.75x10–3 –7.708 1.896 3.00(2.29)
Harpalusdistinguendus 58.67 16.61(<0.001) 14.25(0.124) 12.08(0.141) –0.207 –9.00x10–4 –0.791 –1.283 2.29(1.34)
Harpalusebeninus 25.40 24.85(0.459) 18.54(0.012) 7.56(<0.001) –0.119 5.91x10–3 –13.350 5.977 0.99(0.66)
Harpalusrufipalpis 29.81 15.01(<0.001) 11.39(0.057) 8.56(0.093) –0.317 1.92x10–3 –5.068 2.165 1.30(0.70)
Harpalusserripes 17.70 11.15(0.011) 8.64(0.113) 8.44(0.652) 0.070 5.17x10–3 –10.640 –1.408 0.44(0.26)
Nebriasalina 18.54 16.61(0.164) 12.86(0.053) 10.09(0.096) –0.114 –9.17x10–4 –4.407 2.107 1.69(0.88)
Poeciluskugelanni 35.23 28.33(0.009) 20.21(0.004) 12.76(0.006) 0.057 1.87x10–3 –7.920 2.063 2.60(1.69)
Steropusgallega 25.26 20.22(0.025) 14.99(0.022) 12.68(0.128) –0.095 –2.53x10–3 –3.567 1.505 1.54(1.24)
Steropusglobosus 22.44 22.17(0.603) 12.53(0.002) 10.57(0.161) 0.060 1.95x10–3 –5.782 0.771 3.71(1.76)
Syntomusfoveatus 25.31 13.30(<0.001) 13.14(0.690) 11.21(0.165) 0.005 4.69x10–4 –4.412 –2.550 1.11(0.58)
Trechusobtusus 16.13 12.91(0.073) 12.91(0.939) 11.96(0.329) –0.018 –2.22x10–4 –5.489 1.621 0.92(0.50)
Trechusquadristriatus 23.49 15.60(0.005) 15.10(0.476) 12.04(0.081) –0.034 –2.46x10–4 –6.256 2.911 0.97(0.61)
Zabrussilphoides 26.11 12.64(<0.001) 11.04(0.206) 10.54(0.478) –0.173 1.10x10–3 –4.973 0.786 1.66(0.99)
Generalistgroup 27.65 22.31(0.021) 16.89(0.020) 10.63(0.012) 0.193 2.00x10–3 –7.412 –3.274 2.43(1.46)
Openhabitatgroup 13.18 11.23(0.162) 10.37(0.354) 10.37(0.972) –0.006 7.55x10–4 –4.635 –0.033 2.91(1.39)
threetraps50–200cmapart(i.e.,15trapspersite, Generalisedlinearmodels(GLM)wereused
150trapsintotal).Samplingpointswereatleast to test for differences in carabid abundance and
30mapartandasfaraspossiblefromsiteedges. species richness between the two habitat types.
Beetles were collected continuously from 28 Species representing at least 50 individuals and
Mayto4Octoberin2004,andtrapswereemptied collectedfromatleastfiveofthetensiteswere
every20–25days.Beetleswereidentifiedusing analysed individually. The remaining species
standard keys (Jeannel 1941–1942, Lindroth werepooledintogroupsaccordingtotheabove
1974, Trautner & Geigenmüller 1987), and fol- mentioned habitat classification in order to in-
low the nomenclature in Serrano (2003). We clude them into the analysis. Calathus gra-
counted the number of ants (Formicidae) col- natensisandtheforestspecialistgroupwerenot
lected in the traps due to high abundances ob- tested, as they were clearly associated with the
tained. forestedzone.Theresponsevariablesinthemod-
els were number of individuals (overall and for
individualspecies)andoverallnumberofspecies
2.2.Dataanalysis (i.e., species richness). Abundance data were
modelledfollowinganegativebinomialerrordis-
Wepooledthecarabidcatchesofeachsiteandfor tribution(White&Bennetts1996)unlessspeci-
thewholetrappingperiod,andclassifiedthespe- fiedotherwise(seeTable2),andspeciesrichness
ciesasforestspecialists,generalistsoropenhabi- dataweremodelledfollowingaPoissonerrordis-
tat species, according to the literature (Jeannel tribution.Thepredictorvariablewashabitattype
1941–1942,Lindroth1974,Vázquez1990,Cam- (central grassland and forested zone). Site size
pos2003,Ortuño&Marcos2003&Peláez2004; and mean ant abundance per site were included
seeAppendix). into the models as exploratory variables. Ants
288 Taboadaetal. (cid:127) ENTOMOL.FENNICAVol.17
were included into the models because in pre-
vious studies potential competitor groups ex-
plainedalargeamountofthevariabilityfoundin
the carabid assemblages and individual species
(Taboadaetal.unpubl.).Asaresultofmanage-
mentpracticesandthe“dehesa”habitatconfigu-
ration,centralgrasslandsareunavoidablysmaller
than surrounding forested zones and, conse-
quently, size and habitat type were statistically
confoundedinthemodels(Pearson’scorrelation:
r=0.75,df=8,p=0.001).Therefore,thepre-
dicted values obtained (see Figs. 2 and 3) were
based on: a) the mean size of each habitat type
(i.e.,4.22haforcentralgrasslandsand17.65ha
forforestedzones)inordertogetamorerealistic
predictionforeachhabitattypeindependentlyof
size, and on b) the overall mean size (i.e., the
meansizeofthetensites,10.94ha,whichislo-
catedoutsidethetruesizerangesofbothhabitat
types)inordertoobtainavalid(althoughlessre-
alistic)comparisonbetweenhabitattypesbytak-
ingsizeintoaccount.
We performed a Detrended Correspondence
Analysis(DCA)tocorrelatecarabidspeciesand
sites(Jongmanetal.1995),assumingaunimodal
(Gaussian) response of carabid beetle abun-
dancestotheenvironment.Specieswithatleast
15individuals(26species)wereincludedinthis
analysis,avoidingtheinclusionofscarcespecies
accordingtoourdataset.
Unequaltrappingeffort(19%traploss)was
corrected by including the number of days the
trapspersitewereactiveasanoffsetterminthe
abundancemodels,andbyrarefyingspeciesrich-
nessdatato200individuals.Abundancedatafor Figure2.Mean(±95%ConfidenceInterval)predicted
theDCAanalysiswerestandardisedto100trap- numberofcarabidindividualsandspeciesatthetwo
ping days. Data analyses were carried out with habitattypes.Predictionswereobtainedwith:the
the R statistical package (R Development Core meansizeofeachhabitattype((cid:1))andtheoverall
Team2005). meansize((cid:1));seetext.Predictedvaluesforabun-
dancewerestandardisedtothemeannumberofac-
tivetrappingdayspersite(i.e.,1,543days).Predicted
valuesforspeciesrichnesswereobtainedafterrar-
3.Results
efactionto200individuals.
We collected 4,374 carabid individuals and 55
species (Appendix). Overall carabid abundance habitat (57.8% and 48.6% in central grasslands
was statistically significantly higher in the for- and forested zones, respectively) and generalist
ested zone (Table 2, Fig. 2a). Rarefied species species(42.2%and37.9%,respectively).Nofor-
richnessdidnotdiffersignificantly betweenthe estspecialistspecieswerecapturedfromthecen-
twohabitattypes(Table2,Fig.2b). tralgrasslands,but13.5%ofthespeciescaptured
Both habitat types supported mainly open exclusively from the forested zones were forest
ENTOMOL.FENNICAVol.17 (cid:127) CarabidbeetlesindehesaecosystemsinNWSpain 289
Figure3.Mean(±95%
CI)predictednumberof
individualsofthemost
representativespecies
atthetwohabitattypes.
Predictionswereob-
tainedwith:themean
sizeofeachhabitat
type((cid:1))andtheover-
allmeansize((cid:1)).Pre-
dictedvalueswere
standardisedtothe
meannumberofactive
trappingdayspersite
(i.e.,1,543days).
specialists,includingCalathusrotundicollis,Ca- ciatedtotheforestedzone,independentlyofsize
losoma inquisitor, Carabus lusitanicus and No- (for example, Calathus fuscipes, Harpalus
tiophilus biguttatus (Appendix). Calathus gra- ebeninus, Poecilus kugelanni and Steropus gal-
natensis,ageneralistspeciesaccordingtothelit- lega;Table2,Fig.3);(b)twospeciesresponded
erature, was also exclusively collected fromthe asgeneralistsinthestatisticalsense(i.e.,nosta-
forestedzones. tistically significant difference between the two
The most abundant species were collected habitattypes)andasopenhabitatspeciesinthe
frombothcentralgrasslandsandforestedzones: ecological sense (Harpalus distinguendus and
Calathus fuscipes (1,071 individuals, 24.5% of Harpalusserripes;Table2,Fig.3);and(c)two
thetotalcatch),Steropusglobosus(769individu- species responded highly negatively to habitat
als,17.6%)andHarpalusdistinguendus(612in- size(HarpalusrufipalpisandZabrussilphoides;
dividuals,14%),allgeneralistspecies.Onlytwo Table2,Fig.3).
(Harpalus ebeninus and Poecilus kugelanni) of Inaddition,speciesintheopenhabitatgroup
the14speciestested,respondedstatisticallysig- (24 species) were not statistically significantly
nificantly to habitat type (i.e., responded posi- associatedwitheitherofthehabitattypes,while
tivelytotheforestedzone,Table2).Nineofthe thegeneralistgroup(11species)wasstatistically
14 species tested responded statistically signifi- significantly associated with the central grass-
cantlytothesizeofthestudiedsites,andthenum- lands(Table2).Theforestspecialistgroup(five
berofantswasasignificantpredictoroftheabun- species) responded to habitat type as expected
danceoffourbeetlestested(Table2). (not tested as it was clearly associated with the
Ingeneral,habitatsizehadanegligibleeffect forestedzone).
onthespeciesresponsestohabitattypeasmostof Results from the DCA analysis (Fig. 4)
thespeciestestedrespondedinthesamedirection showedthatthecarabidassemblagecomposition
whenwecomparedthepredictedvaluesobtained ofthetwohabitattypesclearly differed,mainly
fromtheoverallmeansizetothepredictedvalues due to the presence of Harpalus species (open
obtainedfromthemeansizeofeachhabitattype andgeneralistspecies)inthecentralgrasslands,
(Fig.3). and forest specialist species (e.g., Calathus
Weidentifiedthreetypesofresponsesofthe rotundicollisandCalosomainquisitor)andother
speciestohabitattype:(a)mostofthespeciesre- abundantlycollectedspeciesattheforestedzones
spondedeitherasgeneralistsoraremainlyasso- (e.g.,Poeciluskugelanni).
290 Taboadaetal. (cid:127) ENTOMOL.FENNICAVol.17
Figure4.Detrended
CorrespondenceAnaly-
sis(eigenvaluesand
axislengths:0.52and
2.86SDforaxis1,and
0.21and1.54SDfor
axis2,respectively)for
carabidspeciesand
sites.Habitattypes:G=
Centralgrassland,F=
Forestedzone.Carabid
namesareacombina-
tionof4(genusname)
by4(speciesname)
letters.Forexample,
calaambi=Calathus
ambiguus.
4.Discussion movementratesandconsequentlyahigheractiv-
ity-densitymeasurefrompitfalltraps).
Eventhoughthetwohabitattypessampledhere On the other hand, some species were more
werequitedifferentintermsofvegetationstruc- abundantlyfoundintheforestedzone,butcannot
ture,weshowedthattheywerequitesimilarre- be considered as true forest specialist species.
gardingtothecarabidbeetles–mostspeciesdid Thesespeciesmaybeassociatedtospecificenvi-
not respond to habitat type (i.e., were collected ronmentalconditionscreatedundertheoaktrees
frombothcentralgrasslandsandforestedzones). (see Reyes-López et al. 2003). For example,
Some of these species were collected from the Poecilus kugelanni was abundantly found in
forestedzonesbutweremainlycharacteristicof meadow areas characterised by non grazed tall
theopencentralgrasslands,suchasspeciesofthe grasses(Taboadaetal.unpubl.)thatarealsochar-
genusHarpalus,asshownintheDCAanalysis. acteristicoftheimmediateoaktreesurroundings
Thesespeciesmayprefertheenvironmentalcon- intheforestedzone.
ditionsthatcharacterisethecentralgrassland,like Differences between the two habitat types
highertemperatureandlowermoisture,butmay werealsofoundintermsofhigheroverallabun-
alsooccasionallymoveintothesurroundingfor- dance and carabid assemblage composition,
estedzone.Human-createdspacesbetweenold- mainlyduetohighnumbersofHarpalusspecies
growthtreesintheforestedzonemayallowthese collectedfromthecentralgrasslandsanddueto
speciestomovefromandtothecentralgrassland species collected exclusively from the forested
(Koivula&Niemelä2003).Besides,astempera- zones, as reflected in the DCA analysis. The
ture has a significant effect on carabid beetle maintenance of the forested zone characterised
movement(Raworth&Choi2001), itmay also bymatureoaktrees,withgreatcanopycover,per-
have influenced pitfall catches in the central imeterandheight,positivelyinfluencedthepres-
grasslands(i.e.,highertemperaturesimplyhigher enceofafewforestspecialistspecies.Thesespe-
ENTOMOL.FENNICAVol.17 (cid:127) CarabidbeetlesindehesaecosystemsinNWSpain 291
cies may move from the well-developed sur- simplifiedagriculturallandsincaseofintensifi-
roundingoakforest(seeTaboadaetal.2004)to cation,orintoclosedoakforestedlands(i.e.,due
theforestedzoneduetoshadeandhumiditypro- to invasion by young trees and shrubs fromthe
vided by tree development, and probably the surroundings)incaseofabandonment,havelead
morediversemicrohabitatscreatedbyleaflitter to the alteration of the vegetation composition
accumulation (Koivula et al. 1999). Some spe- and structure in these ecosystems (Díez et al.
cies,suchasCarabuslusitanicusandNotiophilus 1991, Gómez-Limón & De Lucío Fernández
biguttatus,werefoundinlownumbersinthesur- 1999, Plieninger et al. 2003, Peco et al. 2005).
rounding forested zone, but in case of the rare Thislossinhabitatdiversityinthe“dehesa”eco-
Calosoma inquisitor the forested zone of the systemmayaffectthefauna,andespeciallythose
“dehesa”probablyprovidesasuitablehabitat. specialistspeciesadaptedtohabitatheterogene-
Ourresultsfurthersuggestthatantsareasig- ity(Niemeläetal.1996,Stoateetal.2001).In-
nificantpredictorofthedistributionoffourcara- deed,incaseofcarabidbeetles,theassemblage
bid species tested. We found either positive or compositionisexpectedtovaryalonglandscape
negativeresponsesofcarabidstoants.Although transformations due to land use intensification
antsappeartoaffectcarabidabundanceandspe- (Fournier&Loreau1999,Vanbergenetal.2005)
cies richness negatively (Lövei & Sunderland andabandonment(Burel&Baudry1995,Purtauf
1996,Hawesetal.2002),individualspeciesre- etal.2004).
spondeddifferentlytothepresenceofwoodants, Preservingthecharacteristichabitatdiversity
dependingontheirdailyactivitypatterns[e.g.,in of“dehesas”inNWSpainmeansthatspecificand
spruceborealforests;Koivulaetal.(1999,2002)] traditional practices are needed, such as shrub
andtheirabilitiestoavoiddirectencountersand control(bycuttingorgrazingbygoats),treecare
predation by ants [in mixed pine-birch forests; (cutting branches, removing old trees, sanitary
Reznikova&Dorosheva(2004)]orintheirabil- care, planting of new trees) and livestock man-
ity toproducedefensivechemicals[e.g.,formic agement(i.e.,balancedfeedinginthewholearea)
acid;Willetal.(2000)]. (Blancoetal.1997,Plieningeretal.2003).Ulti-
In terms of carabid beetle assemblages, the mately, the preservation of “dehesas” requires
forestedzonesofthe“dehesas”arenotthatdiffer- thatconservationpoliciesintegrateparticipatory
entfromthecentralgrasslandsandcannotbecon- planning approaches that consider the motiva-
sidered as true forest ecosystems. Furthermore, tionsoflocalpeople(landownersandland-users)
maindifferencesincarabidbeetlesareexpected andtheappreciationof“dehesa”incomes(Plie-
betweentheopen“dehesa”ecosystem(i.e.,cen- ninger et al. 2004) as most “dehesas” are pri-
tralgrasslandandsurroundingforestedzoneto- vatelyownedorlocalrentingsystems.
gether) and the closed oak forest, a hypothesis
thatwillbeaddressedinthefuture.Itmaywellbe
that the “dehesa” carabid beetles constitute a
uniqueassemblagethatdependsonhabitatdiver-
sityofthe“dehesa”ecosystem,makingitworthy
ofconservationefforts.
The long-term maintenance of the “dehesa”
landscape strongly depends upon varied local Acknowledgements.WethankL.CalvoandH.Roblesfor
uses (Gómez-Limón & De Lucío Fernández collaboration,S.LehvävirtaforusefulcommentsandB.
O’Haraforstatisticaladvice.A.Vanbergenandananony-
1999).Overthelast60years,greatchangesinthe
mous reviewer improved the manuscript considerably.
Mediterranean rural landscape, together with
LandscapemeasurementswerecarriedoutwithSIGPAC
changes in the agriculture policies of the Euro- (http://sigpac.mapa.es/fega/visor).WethanktheServicio
peanUnion,resultedineithertheintensification TerritorialdeMedioAmbientedelaJuntadeCastillay
ofagriculturalandlivestockfarmingortheaban- Leónwhograntedaccesstotheforestsandseveralforest
guards for collaboration. The study was supported by
donmentoftraditionalandnon-competitiveland
Ministerio de Ciencia y Tecnología (project reference
uses, leading to the disappearance of “dehesa”
REN2003-05432/GLO). A. Taboada was financed by
ecosystems. Transformation of “dehesas” into JuntadeCastillayLeón,SpainandCIMO,Finland.
292 Taboadaetal. (cid:127) ENTOMOL.FENNICAVol.17
References Biodivers.Conserv.11:1269–1288.
Koivula,M.&Niemelä,J.2003:Gapfellingasaforest
harvestingmethodinborealforests:responsesofcara-
Blanco, E., Casado, M. A., Costa, M., Escribano, R.,
bid beetles (Coleoptera, Carabidae). — Ecography
García, M., Génova, M., Gómez, A., Gómez, F.,
26:179–187.
Moreno,J.C.,Morla,C.,Regato,P.&Sáinz,H.1997:
Koivula,M.,Punttila,P.,Haila,Y.&Niemelä,J.1999:
Losbosquesibéricos.Unainterpretacióngeobotánica.
Leaflitterandthesmall-scaledistributionofcarabid
— Editorial Planeta, Barcelona, Spain. 572 pp. [In
beetles(Coleoptera,Carabidae)intheborealforest.—
Spanish.]
Ecography22:424–435.
Burel,F.&Baudry,J.1995:Speciesbiodiversityinchang-
Lindroth,C.H.1974:HandbooksfortheIdentificationof
ingagriculturallandscapes:AcasestudyinthePays
BritishInsects.Coleoptera,Carabidae.—RoyalEnto-
d’Auge,France.—Agric.Ecosyst.Environ.55:193–
mologicalSociety,London.148pp.
200.
Lövei,G.L.&Sunderland,K.D.1996:Ecologyandbe-
Campos,A.M.2003:EstudiodelosCarabidae(Coleop-
haviourofgroundbeetles(Coleoptera:Carabidae).—
tera)deGalicia.—PhDthesis,Univ.ofSantiagode
Annu.Rev.Entomol.41:231–256.
Compostela,Spain.660pp.[InSpanish.]
Luis-Calabuig,E.,Tárrega,R.,Calvo,L.,Marcos,E.&
Díez,C.,Luis,E.,Tárrega,R.&Valbuena,L.1991:Deg-
Valbuena,L.2000:Historyoflandscapechangesin
radation process in traditional systems of sylvo-
NorthwestSpainaccordingtolanduseandmanage-
pastoralmanagementindehesasystemsdominatedby
ment.—In:Trabaud,L.(ed.),LifeandEnvironment
Quercuspyrenaica.—Proceedingsofthe“IVCon-
intheMediterranean:43–86.WitPress,Southampton,
grèsInternationaldesTerresdeParcours”,pp.107–
U.K.
109.
Martín,J.&López,P.2002:TheeffectofMediterranean
Díez,C.,Luis,E.,Tárrega,R.&Alonso,P.1994:Estudio
dehesamanagementonlizarddistributionandconser-
delasleguminosasendehesasderobledeQuercus
vation.—Biol.Conserv.108:213–219.
pyrenaicaysurelaciónconlascaracterísticasedáfi-
Niemelä,J,HailaY.&Punttila,P.1996:Theimportanceof
cas.—Proceedingsofthe“XXXIVReuniónCientí-
small-scaleheterogeneityinborealforests:variation
ficadelaSociedadEspañolaparaelEstudiodelos
indiversityinforest-floorinvertebratesacrossthesuc-
Pastos(S.E.E.P.)”,pp.71–77.[InSpanish.]
cessiongradient.—Ecography19:352–368.
Díez,C.,Luis,E.,Tárrega,R.&Alonso,P.1995:Estudio
Ortuño, V. M. & Marcos, J. M. 2003: Los Caraboidea
de las gramíneas en dehesas de roble de Quercus
(Insecta:Coleoptera)delacomunidadautónomadel
pyrenaicaysurelaciónconlascaracterísticasedáfi-
PaísVasco.TomoI.—ServicioCentraldePublica-
cas.—Proceedingsofthe“XXXVReuniónCientífica
cionesdelGobiernoVasco,Spain.573pp.[InSpan-
delaSociedadEspañolaparaelEstudiodelosPastos
ish.]
(S.E.E.P.)”,pp.43–47.[InSpanish.]
Peco,B.,DePablos,I.,Traba,J.&Levassor,C.2005:The
Fournier,E.&Loreau,M.1999:Effectsofnewlyplanted effectofgrazingabandonmentonspeciescomposition
hedgesonground-beetlediversity(Coleoptera,Cara- andfunctionaltraits:thecaseofdehesagrasslands.—
bidae)inanagriculturallandscape.—Ecography22: BasicAppl.Ecol.6:175–183.
87–97. Peláez, M. C. 2004: Estudio faunístico, ecológico y
Gómez-Limón, J. & De Lucío Fernández, J. V. 1999: biogeográfico de los Carabidae (Coleoptera) del
Changesinuseandlandscapepreferencesontheagri- macizodelSueve(Asturias,España).—PhDthesis,
cultural-livestock landscapes of the central Iberian Univ.ofLeón,Spain.893pp.[InSpanish.]
Peninsula (Madrid, Spain). — Landsc. Urban Plan. Penas,A.,García,M.E.&Herrero,L.1995:Seriesde
44:165–175. vegetación.Atlasdelmedionaturaldelaprovinciade
Hawes,C.,Stewart,A.J.A.&Evans,H.F.2002:Theim- León.—InstitutoTecnológicoGeominerodeEspaña
pactofwoodants(Formicarufa)onthedistribution yDiputacióndeLeón,León,Spain.104pp.[InSpan-
andabundanceofgroundbeetles(Coleoptera:Carabi- ish.]
dae)inaScotspineplantation.—Oecologia131:612– Plieninger,T.,Pulido,F.J.&Konold,W.2003:Effectsof
619. land-usehistoryonsizestructureofholmoakstandsin
Jeannel,R.1941–1942:ColéoptèresCarabiques.Faunede Spanish dehesas: implications for conservation and
France,vol.39and40.—Lechevalier,Paris.1173pp. restoration.—Environ.Conserv.30:61–70.
Joffre,R.,Rambal,S.&Ratte,J.P.1999:Thedehesasy- Plieninger,T.,ModolellyMainou,J.&Konold,W.2004:
stemofsouthernSpainandPortugalasanaturaleco- Landmanagerattitudestowardmanagement,regener-
systemmimic.—Agrofor.Syst.45:57–79. ationandconservationofSpanishholmoaksavannas
Jongman,R.H.G.,TerBraak,C.J.F.&VanTongeren,O. (dehesas).—Landsc.UrbanPlan.66:185–198.
F.R.1995:DataAnalysisinCommunityandLand- Pulido,F.J.&Díaz,M.1997:Linkingindividualforaging
scapeEcology.—CambridgeUniversityPress,Cam- behaviourandpopulationspatialdistributioninpatchy
bridge.299pp. environments: a field example with Mediterranean
Koivula,M.,Kukkonen,J.&Niemelä,J.2002:Borealca- bluetits.—Oecologia111:434–442.
rabid-beetle (Coleoptera, Carabidae) assemblages Pulido,F.J.,Díaz,M.&HidalgoDeTrucios,S.J.2001:
alongtheclear-cutoriginatedsuccessiongradient.— SizestructureandregenerationofSpanishholmoak
ENTOMOL.FENNICAVol.17 (cid:127) CarabidbeetlesindehesaecosystemsinNWSpain 293
Quercusilexforestsanddehesas:effectsofagroforest- ron.Manage.63:337–365.
ryuseontheirlong-termsustainability.—For.Ecol. Taboada,A.,Kotze,D.J.&Salgado,J.M.2004:Carabid
Manage.146:1–13. beetleoccurrenceattheedgesofoakandbeechforests
Purtauf,T.,Dauber,J.&Wolters,V.2004:Carabidcom- inNWSpain.—Eur.J.Entomol.101:555–563.
munitiesinthespatio-temporalmosaicofaruralland- Thomas,C.F.G.,Parkinson,L.&Marshall,E.J.P.1998:
scape.—Landsc.UrbanPlan.67:185–193. Isolatingthecomponentsofactivity-densityfortheca-
RDevelopmentCoreTeam2005:R:Alanguageandenvi- rabidbeetlePterostichusmelanariusinfarmland.—
ronmentforstatisticalcomputing.—RFoundationfor Oecologia116:103–112.
Statistical Computing, Vienna, Austria. ISBN 3- Trautner, J. & Geigenmüller, K. 1987: Tiger Beetles,
900051-07-0;URLhttp://www.R-project.org Ground Beetles: Illustrated Key to the Cicindelidae
Raworth,D.A.&Choi,M-Y.2001:Determiningnumbers andCarabidaeofEurope.—Margraf,Germany.488
ofactivecarabidbeetlesperunitareafrompitfall-trap pp.
data.—Entomol.Exp.Appl.98:95–108. Vanbergen,A.J.,Woodcock,B.A.,Watt,A.D.&Nie-
Reyes-López,J.,Ruiz,N.&Fernández-Haeger,J.2003: melä,J.2005:Effectofland-useheterogeneityonca-
Communitystructureofground-ants:theroleofsingle rabid communities at the landscape scale. — Eco-
treesinaMediterraneanpastureland.—ActaOecol. graphy28:3–16.
24:195–202. Vázquez,M.G.1990:Estudiofaunístico,biogeográficoy
Reznikova,Z.&Dorosheva,H.2004:Impactsofredwood ecológico de los Caraboidea (Coleoptera) entre las
antsFormicapolyctenaonthespatialdistributionand cuencasdelosríosBernesga,ToríoyPorma(León,
behaviouralpatternsofgroundbeetles(Carabidae).— España).—PhDthesis,Univ.ofLeón,Spain.455pp.
Pedobiologia48:15–21. [InSpanish.]
Serrano,J.2003:CatálogodelosCarabidae(Coleoptera) White, G. C. & Bennetts, R. E. 1996: Analysis of fre-
delaPenínsulaIbérica.—MonografíasdelaSociedad quencycountdatausingthenegativebinomialdistri-
EntomológicaAragonesavol.9,Zaragoza,Spain.130 bution.—Ecology77:2549–2557.
pp.[InSpanish.] Will,K.W.,Attygalle,A.B.&Herath,K.2000:Newde-
Stoate,C.,Boatman,N.D.,Borralho,R.J.,RioCarvalho, fensivechemicaldataforgroundbeetles(Coleoptera:
C.,DeSnoo,G.R.&Eden,P.2001:Ecologicalim- Carabidae):interpretations in a phylogenetic frame-
pactsofarableintensificationinEurope.—J.Envi- work.—Biol.J.LinneanSoc.71:459–481.