the influence of ecological infrastructures adjacent to ...open access the influence of ecological...

Submitted 31 October 2018Accepted 24 October 2019Published 10 January 2020

Corresponding authorEmilie Pecheurepecheurdoctuliegebe

Academic editorRodolfo Jaffeacute

Additional Information andDeclarations can be found onpage 15

DOI 107717peerj8094

Copyright2020 Pecheur et al

Distributed underCreative Commons CC-BY 40

OPEN ACCESS

The influence of ecological infrastructuresadjacent to crops on their carabidassemblages in intensive agroecosystemsEmilie Pecheur12 Julien Piqueray3 Arnaud Monty12 Marc Dufrecircne12 andGreacutegory Mahy12

1Gembloux Agro-Bio Tech Biodiversity and Landscape University of Liegravege Gembloux Belgium2Gembloux Agro-Bio Tech TERRA-AgricultureIsLife University of Liegravege Gembloux Belgium3Natagriwal asbl Gembloux Belgium

ABSTRACTBackground Conserving biodiversity and enhancing ecosystem services of interestin intensive agroecosystems is a major challenge Perennial ecological infrastructures(EIs) such as hedges and grassy strips and annual EI under Agri-Environment Schemesappear to be good candidates to promote both Our study focused on carabids anindicator group responding both at the species and functional trait level to disturbancesand supporting pest control and weed seed consumption servicesMethods We compared carabid assemblages at the species and functional traits levelssampled via pitfall trapping in three types of EIs (hedges grassy strips and annualflower strips) and crops We also tested via GLMs the effect of (1) the type of EIat the cropsrsquo border and (2) the distance from the cropsrsquo border (two meters or 30meters) on carabid assemblages of crops Tested variables comprised activity-densityspecies richness functional dispersion metrics (FDis) and proportions of carabids byfunctional categories (Diet generalist predatorsspecialist predatorsseed-eaters Sizesmallmediumlargevery large Breeding period springautumn)Results and Discussion Carabid assemblages on the Principal Coordinate Analysissplit in two groups crops and EIs Assemblages from all sampled EIs were dominated bymobile generalist predator species from open-land reproducing in spring Assemblagesof hedges were poor in activity-density and species richness contrarily to grassy andannual flower strips Differences in carabid assemblages in crops were mainly driven bythe presence of hedges The presence of hedges diminished the Community WeightedMean size of carabids in crops due to an increased proportion of small (lt5 mm)individuals while distance from cropsrsquo border favoured large (between 10ndash15 mm)carabids Moreover even if they were attracted by EIs granivorous carabid specieswere rare in crops Our results underlie the importance of local heterogeneity whenadapting cropsrsquo borders to enhance carabid diversity and question the relevance ofhedge implantation in intensive agrolandscapes disconnected from any coherentecological network Moreover this study emphasizes the difficulty tomodify functionalassemblages of crops for purposes of ecosystem services development especially forweed seed consumption as well as the role of distance from the cropsrsquo border in theshaping of crop carabid assemblages

Subjects Agricultural Science Biodiversity Conservation Biology Coupled Natural and HumanSystemsKeywords Intensive agroecosystems Ecosystem services Agri-environment schemes CarabidsFunctional diversity Biodiversity

How to cite this article Pecheur E Piqueray J Monty A Dufrecircne M Mahy Geacute 2020 The influence of ecological infrastructures adjacentto crops on their carabid assemblages in intensive agroecosystems PeerJ 8e8094 httpdoiorg107717peerj8094

INTRODUCTIONIt is well acknowledged that intensive agriculture the dominating production modelin Europe put farmland biodiversity in jeopardy (Donald Green amp Heath 2001 Stoateet al 2001) This raises concerns since biodiversity has been linked to the efficient anddurable delivery of ecosystem services in agroecosystems One of the most studied taxa inagroecosystems both for its contribution to ecosystem services and as an indicator group(Holland amp Luff 2000) are carabids Carabids respond to habitat changes and disturbances(Brooks et al 2012) causing shifts in community assemblages at the species and functionaltraits levels (Hanson et al 2016 Kotze amp OrsquoHara 2003) From an agronomic point of viewthey are potential providers of two ecosystem services of interest in agroecosystems pestcontrol (Kromp 1999) and weed seeds removal (Bohan et al 2011)

Preserving carabid diversity in agrolandscapes goes along with the implementation of agreen network of ecological infrastructures (EIs) Indeed carabid communities appear tobenefit from spatial landscape heterogeneity (Neumann et al 2016) Increased habitat typediversity as well as the presence of linear features varying in length and composition havepositive effects on carabid community compositions (Duflot et al 2016 Duflot et al 2017Fahrig et al 2015Woodcock et al 2005) Conservation of source habitats must be plannedsuch as forested patches and permanent grasslands (Purtauf et al 2005) and completed bya network of less disturbed perennial linear elements adjacent to fields such as hedges andgrassy strips (Fournier amp Loreau 2001) Potentially acting as connection elements allowingspecies to spread into the fields (Hof amp Bright 2010 Labruyere et al 2016 Labruyere Petitamp Ricci 2017) perennial linear elements are also of great importance to preserve locallybiodiversity in the fields notably by providing refuge areas or overwintering sites (Fournieramp Loreau 2001MacLeod et al 2004)

In the European Union perennial elements as well as a variety of annual flower stripscan be established under the framework of Agri-Environment Schemes (AES) Differentseed mixes aiming at various environmental purposes are sown each year in a ploughedstrip along crop borders Less disturbed than crops since chemical inputs are prohibitedthese strips could attract a variety of border-specific carabids species (Saska et al 2007)When sown along a perennial EI annual flower strips could also enhance carabid diversitythanks to the additive effect of both habitats (Boetzl Schneider amp Krauss 2016)

The successful establishment of communities of arthropods in EIs is largely constrainedby local and landscape scales (Gabriel et al 2010) The potential of AES regardingbiodiversity support in impoverished agrolandscapes may thus be limited (Kleijn et al2006) a fact underlying the theory that AES will perform best in moderately complexagrolandscapes (Concepcioacuten et al 2012) Still intensive agrolandscapes either havingundergone radical landscape simplification or being historically intensive (ie run underan open-field regime) represent a large share of the potential areas where biodiversitycould benefit from the implementationconservation of EIs Several studies have exploredcarabid assemblages in agrolandscapes dominated by a bocage pattern undergoing a lossof connectivity and a consequent fragmentation (Baudry Bunce amp Burel 2000 Fournier ampLoreau 2001 Neumann et al 2016) Regarding the influence of crops on carabids (Eyre

Pecheur et al (2020) PeerJ DOI 107717peerj8094 220

Luff amp Leifert 2013) historically intensive agroecosystems are likely to have differentcarabid beetles assemblages

Addressing ecological functionality of carabid assemblages in intensive agroecosystemsrequires to consider the diversity of functional traits within the communities The interestof integrating functional traits to the usual diversity indices is twofold First species richnessalone is neither the best explanation nor the insurance of ecosystem functioning (CadotteCarscadden amp Mirotchnick 2011) Second landscape characteristics shape communities bya filtering operating on their functional traits (De Palma et al 2015 Duflot et al 2014b)Eventually this filtering applies a selection pressure on defined functional groups andmay lead to a homogenisation of communities (Barbaro amp Van Halder 2009) Thereforefunctional traits facilitate the inclusion of landscape history as a factor shaping existingspecies assemblages easing the generalisation of results (Gagic et al 2015)

Local planning however must not be neglected in the mitigation of biodiversity erosionin agroecosystems and themaintenance of ecosystem services Therefore this study focusingon historically intensive agrolandscapes aims at (1) comparing carabid assemblages andfunctional groups in three types of EIs grassy strips (perennial) hedges (perennial) andannual flower strips to evaluate the potential of these EIs in biodiversity conservation (2)testing how the type of EI can modify crop assemblages of carabids at the field border andat 30 m for purposes of promoting pest and weed seed control

MATERIALS amp METHODSThree study sites were selected in southern Belgium (Figs 1A 1B) Sampled fields wereall flat shared the same biogeographical context and were situated at the same altitudeIn ArcMap 102 we calculated for each crop the cover percentage of several land usecategories (see Annex 1) within radii of 50 and 500 m distances known to influencecarabid assemblages (Aviron et al 2005) Data for land cover was provided by the shapefileTOP10V ( ccopyIGN 2018) Located on loamy soils sites were all intensively managedsurrounded by at least 75 of cropped area at 50 and 500 m (Annex 1) Percentageof agricultural area of ecological potential ie set of (semi-) natural elements presenton arable lands was comprised between 15 and 5 (Piqueray et al 2013) Eventuallyan ANOSIM confirms that carabid communities from the crops were similar betweenthe three study sites (p-value = 0086 R= 0063) (Clarke amp Warwick 2001) This iscorroborated with results from homogeneity of groups dispersion tested with lsquobetadisperrsquofrom the lsquoveganrsquo package (Oksanen et al 2008) (ANOVA F(221) = 0942 p-value =040) The above-mentioned description attests that the three study sites had similar landuse occupancy as well as were intensively managed We therefore consider that landscapecontext shaping carabid assemblages was similar between our sites and will not take it anyfurther into account for the next analyses

In each study site four configurations of EIs bordering cereal crops were chosen(Fig 1C) Two configurations displayed a single perennial EI hedge or grassy strip whilethe two others showed an association of each type of perennial EI to an annual flowerstrip Annual flower strips were managed under regional AES Installed for a minimum


Figure 1 Geographical location of study sites and experimental design (A) Geographical emplacementin Europe of the Walloon region (in light blue) in Belgium (dark blue lines) (B) Localisation of the threestudy sites in Wallonia (C) Experimental design included for each study site four configurations of eco-logical infrastructures (EI) at the border of crops single hedge (H) single grassy strip (G) association ofa hedge to an annual flower strip (H + A) and association of a grassy strip to an annual flower strip (G +A) Carabids were captured in pitfall traps (black dots) positionned on transect lines in all EI as well as incrops at two and 30 m from the edge ccopyEuroGeographics Original product is freely available at eurogeo-graphicsorg Terms of the licence available at httpseurogeographicsorgproducts-and-servicesopen-datatopographic-data

Full-size DOI 107717peerj8094fig-1

of five years these strips adjacent to cropped fields are ploughed every year and sownwith a mix of cabbage and winter or spring cereals to which are usually added radish flaxbuckwheat and sunflowers The purpose of annual flower strips is to provide food resourcesto farmland birds during autumn and winter as well as a refuge for small farmland faunaGrassy strips coupled to annual flower strips were also under AES management Sownalong the annual flower strip they were installed for a minimum of five years normally leftuncut and made of a mix of grass species (ie mix of reference 45 Dactylis glomerata25 Phleum pratense and 30 Festuca rubra or 75 Dactylis glomerata and 25 Festucarubra)

Cereal crops (predominantly wheat but barley oat and spelt were present as well) alongwith three categories of EIs (hedges grassy strips annual flower strips) were thus sampledBoth in crops and EIs a transect line was drawn lengthwise the bordering EI (see Fig 1C)on which were placed pitfall traps to collect carabids In crops two transects were definedthe first at two meters from the border and the second at 30 m Traps were polypropylenecups of 85 cm diameter half-filled with a mix of polyethylene glycol diluted to 30 anda few drops of neutral liquid soap Their openings were flush with the soil surface and ametal grid (heightened of 15 cm with a mesh size of one cm) was used above every trap


to prevent catching rodents and other small vertebrates Three traps were equally spacedalong every transect Traps were simultaneously open for seven days during the secondweek of May June July and at the end of September 2016 Weather conditions were similarbetween sites during catching sessions and in case of rain no flooding occurred

Carabids were identified to species level using the identification keys of Coulon Pupier ampQueinnec (2011a) Coulon Pupier amp Queinnec (2011b) and Roger Jambon amp Bouger (2013)A list of all species and their associated traits is available in Table S2

Statistical analysis and plots were computed with R (version 351 R Core Team 2013)Exploring composition of carabid communities in the different sampled configurations ofEIs was performed using Principal Coordinate Analysis (PCoA)We plotted activity-densityper transect Data was standardised to take into account missing traps by averaging allindividuals by traps over the whole trapping period Then the number of individuals fromthe three trapswas summed to obtain an activity-density value by transect A distancematrixof Bray-Curtis was computed from the activity-density data set and root-transformed beforethe PCoA calculation in order to conserve its Euclidean property

Functional traits were used to analyse the structure of carabid assemblages Selected traits(Table 1) were either known to be response traits of carabids to landscape change (iesizeand wing type see Pakeman amp Stockan 2014) or inform about the potential provisioningof ecosystem service (ie diet) or species biology (ie breeding period) They were takenfrom the Ribera et al (1999) and Dufrecircne (1992) databases Functional trait diversity wasestimated using functional dispersion metrics (FDis index) from Laliberteacute amp Legendre(2010) and the associated R package lsquoFDrsquo (Laliberteacute Legendre amp Shipley 2014) Cailliezcorrection for non-Euclidean distance was applied to take into account the categoricaltraits Table 1 displays the trait occurrences for each EI type and crop completed by thecommunity weighted mean (CWM) size

To test the influence of the type of EI and the distance to the crop edge on crop carabidassemblages we used General Linear Models (GLM) with the lsquoglmrsquo call

This test was performed on a set of variables carabid activity-density species richnessfunctional diversity index (FDis) proportions of carabid functional traits and communityweighted mean (CWM) size (see Table 1 for the traits categories) For the diet we groupedomnivorous and granivorous species as both can exert as weed seed suppressor We did nottake into account the wing morphology since brachypterous species were rare Distributionmodels for the GLMs were visually selected via the plotting of quantiles Most variablesfollowed a Gaussian distribution except activity-density granivorous carabids large (gt10mm) and very large carabids (gt15 mm) (log-transformed) and species richness (Poissondistribution) GLMs were run for each variable testing five models 1EI type Distance2EI type + Distance 3 Distance 4EI type and 5 the null model (sim1)For each variablewe selected the best fitted model using the Akaike Information Criterion (AIC (Burnhamamp Anderson 1998) with a correction for small sample (AICc (Hurvich amp Tsai 1989)Difference of AIC value from the best model (1 AICc) and AICc weight (wAICc) wereused to rank the models Eventually p-values were computed for models with a1 AICclt2We used a Likelihood Ratio Test (Type III Wald chisquare tests) through the commandlsquoAnovarsquo from the package lsquocarrsquo (Fox amp Weisberg 2011) If significance was detected a


Table 1 Proportions of selected functional traits () and CommunityWeightedMean (CWM) size in carabid assemblages from each ecological infrastructures (EI)and from crops Proportions of individuals displaying the traits of interest are given by transect (averaged values for three traps)

Size class () Wing () Diet () Breedingperiod ()

Typeof EI

1 lt5mm

2 5 mmltXle 10mm

3 10 mmltXle 15mm

4 gt15mm

CWMsize(mm)

Brachy-pterous

Dimor-phic

Macro-pterous

Generalistpredatorsa

Specialistpredatorsa

Omnivorous Granivorous Spring Autumn

H 1605 4395 2628 1072 882 1072 4433 4494 7795 426 1597 183 5300 4700G 2376 4726 2722 176 806 000 5504 4496 8741 481 142 636 6696 3304A 2823 2674 3647 858 886 786 4019 5195 7385 375 1380 860 8092 1908C2 2153 3995 3496 356 876 152 5285 4563 8686 718 536 060 5884 4116C30 1859 2722 5184 234 1004 012 5446 4542 9121 622 213 044 4322 5678

NotesEIs and crops are abbreviated as followH Hedges G Grassy strips A Annual flowers trips C2 Crops (2 m from the edge) C30 Crops (30 m from the edge)

alsquoSpecialist predatorsrsquo refer to species feeding mainly on Collembola while Generalist predators have a wider panel of preys

Pecheuretal(2020)PeerJDO

I107717peerj8094620

Table 2 Diversity indices of carabid assemblages from ecological infrastructures (EI) and cropsMeanvalues of carabid activity-density species richness and FDis index per transect (sum of three traps) foreach type of EI and crops Error is given by the standard deviation

Type of EI Activity-density Species diversity FDis

H 7306plusmn 7848 1133plusmn 497 025plusmn 012G 8206plusmn 6521 1433plusmn 513 029plusmn 004A 14211plusmn 3088 1867plusmn 513 032plusmn 009C2 24656plusmn 16081 1608plusmn 258 027plusmn 004C30 26906plusmn 20866 1617plusmn 221 027plusmn 004


Tukey contrast as post-hoc test was applied to the model with the lsquoglhtrsquo command fromlsquomultcomprsquo package (Hothorn Bretz amp Westfall 2008) Residuals from significant modelswere checked with spline-correlograms from the package lsquoncfrsquo (Bjornstad 2019) to detectpatterns of spatial autocorrelation We eventually checked for the need of a correlationstructure by usingGLSmodels comparing theAICofmodels with andwithout a correlationstructure

As our data did not allow to test the site effect we provided a descriptive table of thevalues tested per variables (Table S3) This makes it possible to visualize whether thereported effects were consistent between the three study sites

RESULTSA total of 7405 carabids belonging to 41 species (see Annex 2) were captured during thetrapping season Of all species 12 were represented by fewer than 10 individuals

The first two axes of the PCoA (Fig 2A) explained 2270 of the variability observed inthe dataset A differentiation was observed along the first axis between crops and borderingEIs Ecological infrastructures diverged along the second axis Herbaceous structures(grassy and annual flower strips) and crop at 2 m) shared similarities while communitiesat 30 m in the crops and in hedges were more distinct Activity-density of carabid speciescharacterising the crops was higher than those caught in the EIs (Fig 2B)

Activity-density was clearly higher in ploughed structures ie crops and annual flowerstrips especially compared to hedges (Table 2) Species richness was fairly low in theperennial EIs compared to the annual EI and the crops Overall the FDis index was quitesimilar between the different EIs and crops ranging from 025 to 032 The low valueobserved in hedges (025 plusmn 012) was driven by the value from the site of Thuin

Table 1 presents the percentage of occurrence of selected traits categories and CWM sizefor each EI and for crops

Small carabids (ltfive mm) represented between 16 and 23 of the carabid assemblagesin EIs and crops peaking to 28 in annual flower strips Hedges and grassy strips hadsimilar proportions of medium and large carabids (respectively around 45 and 27)while in hedgesmore than 10of caught carabids exceeded 15mmCWMsize was identicalin all EIs and at 2m from the crop border (around eight mm) but increased to 10 mm in the


Figure 2 Carabid assemblages similarities between ecological infrastructures and crops by PrincipalCoordinate Anaysis (PCoA) This PCoA displays the distribution of (A) transects (ie sum of the threetraps during all catching period) from all three study sites (grey dots) Transects are grouped under labelcentroids representing each type of ecological infrastructures and crops (C2= crops (2 m from the edge)C30= crops (30 m from the edge) A = annual flower strips G= grassy strips H= hedges) (B) carabidspecies and their activity-density proportional to the circles size The complete list of carabid species isavailable in Annex 2



Table 3 Results of the Generalized Linear Models applied on a selection of variables (per transect)testing both the type of ecological infrastructures and the distance from the cropsrsquo border on carabidassemblages in cropsModels included the type of ecological infrastructure along the crop (lsquolsquoTypersquorsquo) andthe distance from the crop border (lsquolsquoDistancersquorsquo) Akaike Information Criterion corrrected for small sam-ples (AICc) 1AICc and AICc weight (wAICc) are given for every model Selected models (1 AICc lt 2)are written in bold Non-Gaussian distribution models are specified in brackets (italic) for the concernedvariables

Variables AICc 1 AIC wAIC

Activity-density (Gaussian distribution afterlog transformation)sim Type Distance 642 2025 0000sim Type + Distance 503 635 0027sim Distance 465 257 0177sim Type 468 281 0157sim1 440 000 0640Species richness (Poisson distribution)sim Type Distance 1398 1806 0000sim Type + Distance 1290 730 0018sim Distance 1241 239 0208sim Type 1258 407 0089sim1 1217 000 0685FDissim Type Distance minus598 2448 0000sim Type + Distance minus735 1077 0003sim Distance minus818 245 0222sim Type minus769 736 0019sim1 minus843 000 0755CWM sizesim Type Distance 1111 1281 0001sim Type + Distance 1010 272 0164sim Distance 1034 509 0050sim Type 983 000 0641sim1 1013 300 0143Spring breederssim Type Distance 74 1234 0001sim Type + Distance minus28 215 0128simDistance minus31 182 0152sim Type minus50 000 0376sim1 minus48 018 0343Autumn breederssim Type Distance 74 1234 0001sim Type + Distance minus28 215 0128simDistance minus31 182 0152sim Type minus50 000 0376sim1 minus48 018 0343

(continued on next page)


Table 3 (continued)


Generalist predatorssim Type Distance minus52 1823 0000sim Type + Distance minus166 685 0020simDistance minus220 144 0301sim Type minus188 466 0060sim1 minus235 000 0618Specialist predatorsim Type Distance minus528 2423 0000sim Type + Distance minus659 1113 0003sim Distance minus744 262 0208sim Type minus695 754 0018sim1 minus770 000 0771Omnivorous (Gaussian distribution afterlog transformation)sim Type Distance minus164 1800 0000sim Type + Distance minus286 577 0033simDistance minus330 140 0290sim Type minus307 368 0093sim1 minus344 000 0585Size class 1 (lt5 mm)sim Type Distance minus191 1053 0004sim Type + Distance minus261 360 0131sim Distance minus216 803 0014sim Type minus297 000 0797sim1 minus243 541 0053Size class 2 (5 mm ltX le10 mm)sim Type Distance minus120 1179 0001sim Type + Distance minus200 378 0067simDistance minus235 020 0402sim Type minus205 327 0086sim1 minus237 000 0443Size class 3 (10 mm ltX le 15mm) (Gaussian distribution afterlog transformation)sim Type Distance 87 1610 0000sim Type + Distance minus45 285 0098simDistance minus74 000 0406sim Type minus46 276 0102sim1 minus73 006 0395Size class 4 (gt15 mm) (Gaussian distribution afterlog transformation)sim Type Distance minus150 1865 0000sim Type + Distance minus256 806 0012simDistance minus319 172 0281sim Type minus282 549 0043sim1 minus337 000 0665


Table 4 Likelihood Ratio Test calculated on the selected GLMs explaining best the variables describingcarabid assemblages in crops Significant models (plt 005) are written in bold

Variables Selected model df LR χ2 P value

Activity-density sim1 ndash ndash ndashSpecies richness sim1 ndash ndash ndashFDis sim1 ndash ndash ndashCWM size sim Type 3 1264 0005SpringAutumn breeders sim Type 3 903 0029

sim Distance 1 093 0335Generalist predators sim Distance 1 112 0290Specialist predators sim1 ndash ndash ndashOmnivorous sim Distance 1 115 0283Class 1 sim Type 3 1609 0001Class 2 sim Distance 1 235 0126Class 3 sim Distance 1 260 0107Class 4 sim Distance 1 085 0358

crop at 30 m from theborder This is attributable to the percentage of large carabids (10ndash15mm) accounting for more than 50 of all individuals caught further in the crops Verylarge carabids (gt15 mm) tended to be scarce in all sampled elements Winged-phenotypesdominated in all of the EIs and in crops Brachypterous species were rare (only two specieswere identified Pterostichus madidus and Stomis pumicatus) Regarding diet all of theEIs and crops contained more than 73 of generalist predators If specialist species werepresent in similar proportions the percentage of omnivorous was higher in hedges andannual flower strips Granivorous carabids were more abundant in annual flower strips andgrassy strips Spring-breeding carabid species were the most abundant in EIs and cropsexcept at 30m from the crop edge

Results of the GLMs are fully presented in Table 3 Tested variables (EIs and distance)did not explain the patterns of species richness FDis index and proportion of specialistpredators For several variables both the null model and another model had a 1 AICc lt2in that case the model explained by the tested factor was never significant (Table 4) Thetype of EI modified the size and proportion of spring breeders in the crop assemblages(Table 4)The presence of a single hedge led to carabid assemblages of smaller size (Figs 3Aand 3B) with an increased proportion of spring breeders compared to an association ofgrassy stripannual flower strip (Fig 3C) Distance tended to be retained as an explanatoryvariable for other class sizes and diets but models were not significant Spline-correlograms(Fig S1) did not display spatial auto-correlation in residuals at short distances (ltfive km)confirming the spatial independence of transects However signals of potential negativeauto-correlation in model residuals were present at 30 km and 70 km Still the addition ofa correlation structure to the GLS models showed no improvement for all three variablesGLSmodels with the lower AIC values are those without the correlation structure (Table 5)We thus conclude that models without correlation structure are valid


Figure 3 Boxplots of functional traits of interest from crop carabid assemblages (by transects) Pre-sented plots display variables that responded significantly to the type of ecological infrastructure Sig-nificant Linear Models (LMs) concerned (A) Community weighted mean (CWM) size (B) Carabids ofsmall size (lt5 mm) and (C) Proportion of spring breeders Ecological infrastructures are abbreviated asfollow H hedges G grassy strips A annual flower strips) Distance from the crop border is given by theshade of the boxes (light grey= 2 m from the border dark grey= 30 m from the border) Letters abovethe plots indicate the results of the post-hoc Tukey test performed on the GLMs


Table 5 AIC values for GLSmodels with and without correlation structure All three variables (Com-munity Weighted Mean size Size class 1 and Spring breeders) were tested as a function of the type of eco-logical infrastructures

AIC values

Correlationstructure

CommunityWeightedMean(CMW) Size

Size class1 (lt5mm)

Springbreeders

No correlation structure 9162312 minus1501437 5573410corExp 9562312 minus1101437 7784194corLin 9562312 minus1101437 8829442corGaus 9562312 minus1101437 7225125corSpher 9562312 minus1101437 8829442corRatio 9562312 minus1101437 7695223

DISCUSSIONThe different sampled EIs sheltered distinct assemblages of carabids compared to croppedfields Regarding composition in functional traits communities from EIs and crops weredominated by predator carabids from open land displaying a good ability to disperseand mostly reproducing in spring The two last traits are known to mirror the filteringinitiated by landscape configuration on carabid assemblages (Duflot et al 2014b) selectinghighly mobile species able to cope with crop disturbances Even in the absence of a controlsituation without any EI we suspect that the landscape context largely shaped the poolof species and traits present in the agroecosystems and that local management could onlyslightly modulate these (Woodcock et al 2010 Trichard et al 2013)


Overall sampled carabid assemblages showed a functional homogeneity tempered bythe availability of EIs The similarity of FDis in crops bordered by different EIs suggested thateven though these EIs contributed to local diversity of carabid species their contributionto traits diversity may have been insufficient regarding the filtering exerted by the intensiveagrolandscapes When unusual traits for carabid assemblages in crops were detected iegranivory carabids were too few to raise functional diversity index (FDis) as FDis weightsspecies by their abundance

Still among the tested functional traits in crops the size of carabids was affected bythe presence of single hedges Community weighted mean (CWM) size of carabids in thecrops decreased in response to the decline of large (gt10 mm) carabids and the increase ofsmall carabids (ltfive mm) The distance to the crop border conversely tended to favour theproportion of large carabids As the size of predator species plays a key role in prey-predatorrelationships in terms of consumed species and predation rate (Rusch et al 2015) pestcontrol may be affected The dominance of large carabid species is suspected to favourintra-guild predation towards other carabid species of smaller size or other potentialpredators may occur possibly interfering with the desired pest control (Prasad amp Snyder2006) To the contrary according to allometric laws (Brose 2010) the larger proportion ofsmall carabid species in the presence of hedges could benefit to the consumption of smallpests such as aphids

The presence of single hedges also increased the proportion of spring breeders in thecrops This suggests that single hedges are a refuge habitat of lower quality at least for someautumn breeding species Poor habitat quality of hedges under anthropogenic managementis common (Magura Loumlvei amp Toacutethmeacutereacutesz 2017) and could partly explain the questioningresults of hedges characterized by a low diversity and activity-density of carabids Thiscontrasts with other studies where hedges as a part of a coherent network such as bocagelandscapes were commonly identified as distinct habitats hosting a mix of forest andopen-land species (Fournier amp Loreau 2001 Duflot et al 2014a) This was however notthe case on our study sites as hedges there were planted in open-field landscapes

Granivorous carabid species responded positively to the presence of grassy and annualflower strips adjacent to crops As granivorous species are highly mobile in agrolandscapesto exploit food resources (Labruyere et al 2016) they benefit from the establishment ofweeds in EIs (Fried et al 2009) fromwhich they may feed on seeds (Saska amp Jarosik 2001)Still they hardly dispersed within the crops and showed no response to the type of EIas it has been observed by Boetzl et al (2018) Literature shows that bordering habitatshave a limit of influence on crop species assemblages (Roume et al 2011) Acknowledgingthat some species show clear preferences for either boundaries or field-edges (Saska et al2007) which may retain them from dispersing within the neighbouring field (Boetzl etal 2018) the role of EIs in the enhancement of granivorous carabids in crops appearedlimited in intensive agroecosystems Moreover if distance from the edge tends to exert aselection pressure on size this could limit the range of granivorous carabid size diversityAs a consequence the variety of consumed seedmay be restrained as carabid size constrainsseed preferences (Honek et al 2007)


Though our results display clear trends the lack of repetition of EIs per site in ourstudy enables us to generalize them to intensive agrolandscapes Still our study addressestwo main issues (1) the planting of single hedges for carabids in intensive agriculturallandscapes outside any forested network and (2) the role of granivorous carabid speciesin the control of weed seed in crops As single hedges turned out to be of weak supportto carabids from agroecosystems potentially impeding their movement in the landscape(Mauremooto et al 1995) we encourage a deeper investigation of their role in intensiveagrolandscapes Regarding granivorous carabid species even if the activity-density ofcarabids does not ensure the delivery of an ecosystem service (Saska et al 2008) wesuggest as an experimental perspective to investigate the presence of granivorous carabidsin crops during ploughing period As ploughing exposes weed seeds from the seed bank itis likely an adequate time of action for their consumption by carabids if they can exert anycontrol Concomitantly as distance from the border appears as a potential limiting factorto the dispersion of carabids within crops we enjoin to investigate the spatial arrangementof EI in agroecosystems as a research perspective to promote efficient and stable pest andweed control in crops

CONCLUSIONOur results support that carabids can be very diverse in intensive agroecosystems (Vanbergenet al 2005) though displaying a homogenisation of functional traits Carabids fanned outin the different habitats at their disposal quite quickly as exampled by the colonization ofannual flower strips by granivorous species Thus adapting field edges with variate stripsprovide habitats and food resources to carabid species infrequent in crops or offer anadequate refuge to crop species Differences in carabid assemblages in crops were mainlydriven by the presence of hedges Changes concerned the CWM size which decreased in thepresence of a hedge while the proportion of large carabids increased with distance from thecrop border Considering carabid assemblages and functional guilds our results highlightthat the role of ecological infrastructures in the enhancement of ecosystem services may belimited by the supported pool of species and in view of these species habitat preferencesnotably for weed seed removal To conclude if the diversification of EIs is to be pursuedin intensive agrolandscapes their installation would benefit from a landscape-inspiredplanification based on a networking of farmersrsquo engagements to allow the development ofan efficient matrix of agri-environment schemes

ACKNOWLEDGEMENTSThe authors thank Benjamin Daigneux who identified the sampled carabids


ADDITIONAL INFORMATION AND DECLARATIONS

FundingEmilie Pecheur holds a PhD fellowship from the FRS- FNRS whose funding coveredthe realization of this study The funders had no role in study design data collection andanalysis decision to publish or preparation of the manuscript

Grant DisclosuresThe following grant information was disclosed by the authorsFRS- FNRS

Competing InterestsJulien Piqueray is employed by Natagriwal asbl

Author Contributionsbull Emilie Pecheur conceived and designed the experiments performed the experimentsanalyzed the data contributed reagentsmaterialsanalysis tools prepared figures andortables authored or reviewed drafts of the paper approved the final draftbull Julien Piqueray andArnaudMonty conceived and designed the experiments contributedreagentsmaterialsanalysis tools authored or reviewed drafts of the paper approved thefinal draftbull Marc Dufrecircne and Greacutegory Mahy conceived and designed the experiments authored orreviewed drafts of the paper approved the final draft

Data AvailabilityThe following information was supplied regarding data availability

The raw data is available in the Supplementary Files

Supplemental InformationSupplemental information for this article can be found online at httpdxdoiorg107717peerj8094supplemental-information

REFERENCESAviron S Burel F Baudry J Schermann N 2005 Carabid assemblages in agricultural

landscapes impacts of habitat features landscape context at different spatialscales and farming intensity Agriculture Ecosystems amp Environment 108205ndash217DOI 101016jagee200502004

Barbaro L Van Halder I 2009 Linking bird carabid beetle and butterfly life-historytraits to habitat fragmentation in mosaic landscapes Ecography 32321ndash333DOI 101111j1600-0587200805546x

Baudry J Bunce RGH Burel F 2000Hedgerows an international perspective on theirorigin function and management Journal of Environmental Management 607ndash22DOI 101006jema20000358


Bjornstad ON 2019 ncf spatial covariance functions R package version 12-8 Availableat httpsCRANR-projectorgpackage=ncf

Boetzl FA Krimmer E Krauss J Steffan-Dewenter I 2018 Agri-environmental schemespromote ground-dwelling predators in adjacent oilseed rape fields diversityspecies traits and distance-decay functions Journal of Applied Ecology 56(1)10ndash20DOI 1011111365-266413162

Boetzl FA Schneider G Krauss J 2016 Asymmetric carabid beetle spillover betweencalcareous grasslands and coniferous forests Journal of Insect Conservation 2049ndash57DOI 101007s10841-015-9838-6

Bohan DA Boursault A Brooks DR Petit S 2011 National-scale regulation of theweed seedbank by carabid predators Journal of Applied Ecology 48888ndash898DOI 101111j1365-2664201102008x

Brooks DR Bater JE Clark SJ Monteith DT Andrews C Corbett SJ Beaumont DAChapman JW 2012 Large carabid beetle declines in a United Kingdom monitoringnetwork increases evidence for a widespread loss in insect biodiversity Journal ofApplied Ecology 491009ndash1019 DOI 101111j1365-2664201202194x

Brose U 2010 Body-mass constraints on foraging behaviour determine population andfood-web dynamics Functional Ecology 2428ndash34DOI 101111j1365-2435200901618x

BurnhamKP Anderson DR 1998Model selection and inference a practical information-theoretic approach New York Springer-Verlag DOI 101007978-1-4757-2917-7_3

Cadotte MW Carscadden K Mirotchnick N 2011 Beyond species functional diversityand the maintenance of ecological processes and services Journal of Applied Ecology481079ndash1087 DOI 101111j1365-2664201102048x

Clarke KRWarwick RM 2001 Change in marine communities an approach to statisticalanalysis and interpretation 2nd edition Plymouth PRIMER-E Ltd PlymouthMarine Laboratory

Concepcioacuten ED Diacuteaz M Kleijn D Baacuteldi A Bataacutery P Clough Y Gabriel D Her-zog F Holzschuh A Knop E Marshall EJP Tscharntke T Verhulst J 2012Interactive effects of landscape context constrain the effectiveness of localagri-environmental management Journal of Applied Ecology 49695ndash705DOI 101111j1365-2664201202131x

Coulon J Pupier R Queinnec E 2011a Faune n 94ndashColeacuteoptegraveres Carabidae de FranceParis Eacuteditions Faune de France 353

Coulon J Pupier R Queinnec E 2011b Faune n 94ndashColeacuteoptegraveres Carabidae de FranceParis Eacuteditions Faune de France 313

De Palma A KuhlmannM Roberts SPM Potts SG Boumlrger L Hudson LN LysenkoI Newbold T Purvis A 2015 Ecological traits affect the sensitivity of bees toland-use pressures in European agricultural landscapes Journal of Applied Ecology521567ndash1577 DOI 1011111365-266412524


Donald PF Green RE HeathMF 2001 Agricultural intensification and the collapse ofEuropersquos farmland bird populations Proceedings of the Royal Society B Biological Sci-ences 26825ndash29DOI 101098rspb20001325

Duflot R Aviron S Ernoult A Fahrig L Burel F 2014a Reconsidering the role of semi-natural habitat in agricultural landscape biodiversity a case study Ecological Research3075ndash83 DOI 101007s11284-014-1211-9

Duflot R Ernoult A Aviron S Fahrig L Burel F 2017 Relative effects of land-scape composition and configuration on multi-habitat gamma diversity inagricultural landscapes Agriculture Ecosystems amp Environment 24162ndash69DOI 101016jagee201702035

Duflot R Ernoult A Burel F Aviron S 2016 Landscape level processes drivingcarabid crop assemblage in dynamic farmlands Population Ecology 58265ndash275DOI 101007s10144-015-0534-x

Duflot R Georges R Ernoult A Aviron S Burel F 2014b Landscape heterogeneity as anecological filter of species traits Acta Oecologica 5619ndash26DOI 101016jactao201401004

DufrecircneM 1992 Biogeacuteographie et eacutecologie des communauteacutes de Carabidae en Wal-lonie Doctoral dissertation Universiteacute catholique de Louvain Louvain-la-NeuveBelgique

Eyre MD Luff ML Leifert C 2013 Crop field boundary productivity and disturbanceinfluences on ground beetles (Coleoptera Carabidae) in the agroecosystemAgriculture Ecosystems amp Environment 16560ndash67 DOI 101016jagee201212009

Fahrig L Girard J Duro D Pasher J Smith A Javorek S King D Lindsay KFMitchell S Tischendorf L 2015 Farmlands with smaller crop fields have higherwithin-field biodiversity Agriculture Ecosystems amp Environment 200219ndash234DOI 101016jagee201411018

Fournier E LoreauM 2001 Respective roles of recent hedges and forest patch remnantsin the maintenance of ground-beetle (coleoptera Carabidae) diversity in anagricultural landscape Landscape Ecology 1617ndash32 DOI 101023A1008115516551

Fox J Weisberg S 2011 An R companion to applied regression Second editionThousand Oaks Sage

Fried G Petit S Dessaint F Reboud X 2009 Arable weed decline in Northern Francecrop edges as refugia for weed conservation Biological Conservation 142238ndash243DOI 101016jbiocon200809029

Gabriel D Sait SM Hodgson JA Schmutz U KuninWE Benton TG 2010 Scalematters the impact of organic farming on biodiversity at different spatial scalesEcology Letters 13858ndash869 DOI 101111j1461-0248201001481x

Gagic V Bartomeus I Jonsson T Taylor AWinqvist C Fischer C Slade EM Steffan-dewenter I EmmersonM Potts SG Tscharntke TWeisserW Bommarco R 2015Functional identity and diversity of animals predict ecosystem functioning betterthan species-based indices Proceedings of the Royal Society B Biological Sciences28220142620 DOI 101098rspb20142620


Hanson HI Palmu E Birkhofer K Smith HG Hedlund K 2016 Agricultural landuse determines the trait composition of ground beetle communities PLOS ONE111ndash13 DOI 101371journalpone0146329

Hof AR Bright PW 2010 The impact of grassy field margins on macro-invertebrateabundance in adjacent arable fields Agriculture Ecosystems amp Environment139280ndash283 DOI 101016jagee201008014

Holland JM Luff ML 2000 The effects of agricultural practices on Carabidae intemperate agroecosystems Integrated Pest Management Reviews 5109ndash129DOI 101023A1009619309424

Honek A Martinkova Z Saska P Pekar S 2007 Size and taxonomic constraintsdetermine the seed preferences of Carabidae (Coleoptera) Basic and Applied Ecology8343ndash353 DOI 101016jbaae200607002

Hothorn T Bretz F Westfall P 2008 Simultaneous inference in general parametricmodels Biometrical Journal 50346ndash363 DOI 101002bimj200810425

Hurvich CM Tsai CL 1989 Regression and time series model selection in small samplesBiometrika 76297ndash307 DOI 101093biomet762297

Kleijn D Baquero RA Clough Y Diacuteaz M De Esteban J Fernaacutendez F Gabriel DHerzog F HolzschuhaJ oumlhl R Knop E Kruess A Marshall EJP Steffan-DewenterI Tscharntke T Verhulst J West TM Yela JL 2006Mixed biodiversity benefitsof agri-environment schemes in five European countries Ecology Letters 9243ndash254DOI 101111j1461-0248200500869x

Kotze DJ OrsquoHara RB 2003 Species declinemdashbut why Explanations of carabidbeetle (Coleoptera Carabidae) declines in Europe Oecologia 135138ndash148DOI 101007s00442-002-1174-3

Kromp B 1999 Carabid beetles in sustainable agriculture a review on pest controlefficacy cultivation impacts and enhancement Agriculture Ecosystems amp Environment74187ndash228 DOI 101016S0167-8809(99)00037-7

Labruyere S Bohan DA Biju-Duval L Ricci B Petit S 2016 Local neighbor and land-scape effects on the abundance of weed seed-eating carabids in arable fields a nation-wide analysis Basic and Applied Ecology 17230ndash239 DOI 101016jbaae201510008

Labruyere S Petit S Ricci B 2017 Annual variation of oilseed rape habitat quality androle of grassy field margins for seed eating carabids in arable mosaics Agriculturaland Forest Entomology 20(2)234ndash245 DOI 101111afe12250

Laliberteacute E Legendre P 2010 A distance-based framework for measuring functionaldiversity from multiple traits Ecology 91299ndash305 DOI 10189008-22441

Laliberteacute E Legendre P Shipley B 2014 FD measuring functional diversity frommultiple traits and other tools for functional ecology R package version 10-12hAvailable at https cranr-projectorgwebpackagesFDFDpdf

MacLeod AWratten SD Sotherton NW ThomasMB 2004 Beetle banks asrefuges for beneficial arthropods in farmland long-term changes in preda-tor communities and habitat Agricultural and Forest Entomology 6147ndash154DOI 101111j1461-9563200400215x


Magura T Loumlvei GL Toacutethmeacutereacutesz B 2017 Edge responses are different in edges undernatural versus anthropogenic influence a meta-analysis using ground beetles Ecologyand Evolution 71009ndash1017 DOI 101002ece32722

Mauremooto JRWratten SDWorner SP Fry GLA 1995 Permeability of hedgerows tocarabid beetles Agriculture Ecosystems amp Environment 52141ndash148DOI 1010160167-8809(94)00548-S

Neumann JL Griffiths GH Hoodless A Holloway GJ 2016 The compositionaland configurational heterogeneity of matrix habitats shape woodland carabidcommunities in wooded-agricultural landscapes Landscape Ecology 31301ndash315DOI 101007s10980-015-0244-y

Oksanen J Kindt R Legendre P OrsquoHara B Simpson GL Solymos P Stevens MHHWagner H 2008 The vegan package Community ecology package Available athttp r-forge r-project orgprojects vegan

Pakeman RJ Stockan JA 2014 Drivers of carabid functional diversity abiotic environ-ment plant functional traits or plant functional diversity Ecology 951213ndash1224DOI 10189013-10591

Piqueray J Richard CWalot T Le Roi A De Tillesse M Hance T Mahy G 2013Proceedings of the conference nature conditionnaliteacute et verdissement de la PACun tournant wallon 9th 2011 Espace Senghor Gembloux Belgium BiotechnologieAgronomie Societe et Environnement 17344ndash351

Prasad RP SnyderWE 2006 Polyphagy complicates conservation biological con-trol that targets generalist predators Journal of Applied Ecology 43343ndash352DOI 101111j1365-2664200601129x

Purtauf T Roschewitz I Dauber J Thies C Tscharntke TWolters V 2005 Landscapecontext of organic and conventional farms influences on carabid beetle diversityAgriculture Ecosystems amp Environment 108165ndash174 DOI 101016jagee200501005

R Core Team 2013 R a language and environment for statistical computing Vienna RFoundation for Statistical Computing

Ribera I Foster GN Downie IS McCracken DI Abernethy VJ 1999 A comparativestudy of the morphology and life traits of Scottish ground beetles (ColeopteraCarabidae) Annales Zoologici Fennici 3621ndash37

Roger JL Jambon O Bouger G 2013 Cleacute de deacutetermination des Carabidae paysagesagricoles du Nord Ouest de la France Rennes SAD-Paysage Institut National deRecherche Agronomique

Roume A Deconchat M Raison L Balent G Ouin A 2011 Edge effects on groundbeetles at the woodlot-field interface are short-range and asymmetrical Agriculturaland Forest Entomology 13395ndash403 DOI 101111j1461-9563201100534x

Rusch A Birkhofer K Bommarco R Smith HG Ekbom B 2015 Predator body sizesand habitat preferences predict predation rates in an agroecosystem Basic andApplied Ecology 16250ndash259 DOI 101016jbaae201502003

Saska P Jarosik V 2001 Laboratory study of larval food requirements in ninespecies of Amara (Coleoptera Carabidae) Plant Protection Science 37103ndash110DOI 10172218372-PPS


Saska P Van DerWerfW De Vries E Westerman PR 2008 Spatial and tem-poral patterns of carabid activity-density in cereals do not explain levels ofpredation on weed seeds Bulletin of Entomological Research 98169ndash181DOI 101017S0007485307005512

Saska P VoddeM Heijerman TWesterman P Van derWerfW 2007 The sig-nificance of a grassy field boundary for the spatial distribution of carabidswithin two cereal fields Agriculture Ecosystems amp Environment 122427ndash434DOI 101016jagee200702013

Stoate C Boatman ND Borralho RJ Carvalho CR De Snoo GR Eden P 2001 Ecologi-cal impacts of arable intensification in Europe Journal of Environmental Management63337ndash365 DOI 101006jema20010473

Trichard A Alignier A Biju-Duval L Petit S 2013 The relative effects of local manage-ment and landscape context on weed seed predation and carabid functional groupsBasic and Applied Ecology 14235ndash245 DOI 101016jbaae201302002

Vanbergen AJ Woodcock BAWatt AD Niemelii J 2005 Effect heterogeneityscale landscape at the on carabid communities of land-use Ecography 283ndash16DOI 101111j0906-7590200503991x

Woodcock BA Redhead J Vanbergen AJ Hulmes L Hulmes S Peyton J NowakowskiM Pywell RF HeardMS 2010 Impact of habitat type and landscape structure onbiomass species richness and functional diversity of ground beetles AgricultureEcosystems amp Environment 139181ndash186 DOI 101016jagee201007018

Woodcock BAWestbury DB Potts SG Harris SJ Brown VK 2005 Establishing fieldmargins to promote beetle conservation in arable farms Agriculture Ecosystems ampEnvironment 107255ndash266 DOI 101016jagee200410029


INTRODUCTIONIt is well acknowledged that intensive agriculture the dominating production modelin Europe put farmland biodiversity in jeopardy (Donald Green amp Heath 2001 Stoateet al 2001) This raises concerns since biodiversity has been linked to the efficient anddurable delivery of ecosystem services in agroecosystems One of the most studied taxa inagroecosystems both for its contribution to ecosystem services and as an indicator group(Holland amp Luff 2000) are carabids Carabids respond to habitat changes and disturbances(Brooks et al 2012) causing shifts in community assemblages at the species and functionaltraits levels (Hanson et al 2016 Kotze amp OrsquoHara 2003) From an agronomic point of viewthey are potential providers of two ecosystem services of interest in agroecosystems pestcontrol (Kromp 1999) and weed seeds removal (Bohan et al 2011)

Preserving carabid diversity in agrolandscapes goes along with the implementation of agreen network of ecological infrastructures (EIs) Indeed carabid communities appear tobenefit from spatial landscape heterogeneity (Neumann et al 2016) Increased habitat typediversity as well as the presence of linear features varying in length and composition havepositive effects on carabid community compositions (Duflot et al 2016 Duflot et al 2017Fahrig et al 2015Woodcock et al 2005) Conservation of source habitats must be plannedsuch as forested patches and permanent grasslands (Purtauf et al 2005) and completed bya network of less disturbed perennial linear elements adjacent to fields such as hedges andgrassy strips (Fournier amp Loreau 2001) Potentially acting as connection elements allowingspecies to spread into the fields (Hof amp Bright 2010 Labruyere et al 2016 Labruyere Petitamp Ricci 2017) perennial linear elements are also of great importance to preserve locallybiodiversity in the fields notably by providing refuge areas or overwintering sites (Fournieramp Loreau 2001MacLeod et al 2004)

In the European Union perennial elements as well as a variety of annual flower stripscan be established under the framework of Agri-Environment Schemes (AES) Differentseed mixes aiming at various environmental purposes are sown each year in a ploughedstrip along crop borders Less disturbed than crops since chemical inputs are prohibitedthese strips could attract a variety of border-specific carabids species (Saska et al 2007)When sown along a perennial EI annual flower strips could also enhance carabid diversitythanks to the additive effect of both habitats (Boetzl Schneider amp Krauss 2016)

The successful establishment of communities of arthropods in EIs is largely constrainedby local and landscape scales (Gabriel et al 2010) The potential of AES regardingbiodiversity support in impoverished agrolandscapes may thus be limited (Kleijn et al2006) a fact underlying the theory that AES will perform best in moderately complexagrolandscapes (Concepcioacuten et al 2012) Still intensive agrolandscapes either havingundergone radical landscape simplification or being historically intensive (ie run underan open-field regime) represent a large share of the potential areas where biodiversitycould benefit from the implementationconservation of EIs Several studies have exploredcarabid assemblages in agrolandscapes dominated by a bocage pattern undergoing a lossof connectivity and a consequent fragmentation (Baudry Bunce amp Burel 2000 Fournier ampLoreau 2001 Neumann et al 2016) Regarding the influence of crops on carabids (Eyre






















Typeof EI

1 lt5mm

2 5 mmltXle 10mm

3 10 mmltXle 15mm

4 gt15mm

CWMsize(mm)

Brachy-pterous

Dimor-phic

Macro-pterous




H 1605 4395 2628 1072 882 1072 4433 4494 7795 426 1597 183 5300 4700G 2376 4726 2722 176 806 000 5504 4496 8741 481 142 636 6696 3304A 2823 2674 3647 858 886 786 4019 5195 7385 375 1380 860 8092 1908C2 2153 3995 3496 356 876 152 5285 4563 8686 718 536 060 5884 4116C30 1859 2722 5184 234 1004 012 5446 4542 9121 622 213 044 4322 5678




I107717peerj8094620





















Table 3 (continued)














AIC values



Size class1 (lt5mm)

Springbreeders
















































































































Typeof EI

1 lt5mm

2 5 mmltXle 10mm

3 10 mmltXle 15mm

4 gt15mm

CWMsize(mm)

Brachy-pterous

Dimor-phic

Macro-pterous




H 1605 4395 2628 1072 882 1072 4433 4494 7795 426 1597 183 5300 4700G 2376 4726 2722 176 806 000 5504 4496 8741 481 142 636 6696 3304A 2823 2674 3647 858 886 786 4019 5195 7385 375 1380 860 8092 1908C2 2153 3995 3496 356 876 152 5285 4563 8686 718 536 060 5884 4116C30 1859 2722 5184 234 1004 012 5446 4542 9121 622 213 044 4322 5678




I107717peerj8094620





















Table 3 (continued)














AIC values



Size class1 (lt5mm)

Springbreeders
















































































































Table 3 (continued)














AIC values



Size class1 (lt5mm)

Springbreeders






































































































AIC values



Size class1 (lt5mm)

Springbreeders




























































































the influence of ecological infrastructures adjacent to ...open access the influence of ecological...

Documents