ARTICLE IN PRESS. Ecological Indicators xxx (2011) xxx xxx. Contents lists available at ScienceDirect. Ecological Indicators

Transcription

1 ECOIND-918; No. of Pges 1 Ecologicl Indictors xxx (211) xxx xxx Contents lists ville t ScienceDirect Ecologicl Indictors jo ur nl homep ge: Opertionl performnce indictors for litter mngement t festivls in semi-nturl lndscpes Arne Cierjcks,, Friederike Behr, Ingo Kowrik Deprtment of Ecology, Technische Universität Berlin, Rothenurgstr. 12, Berlin, Germny Ecocontrolling GR, Geisenheimer Str. 3, Berlin, Germny r t i c l e i n f o Article history: Received 1 Mrch 211 Received in revised form 1 June 211 Accepted 28 June 211 Keywords: Event mngement Environmentl mngement Festivl Littering Opertionl performnce indictor Solid wste s t r c t Huge open-ir festivls with more thn 5, visitors hve ecome common ut scrcely studied lnduse prctice in mny countries of the world. The environmentl consequences of crowds such s litter ccumultion nd dmge to vegettion re incresingly eing tken into ccount y event mngers. At present, mny festivls use overll solid wste mount s n indictor of their wste mngement performnce. However, these opertionl performnce indictors (OPIs) do not reflect either the success of individul litter reduction mesures or the dmge to vlule hitts. We developed specific OPIs for solid wste littering nd compred solute littering nd littering per person t three open-ir festivls in Germny cross different festivl zones nd phses. For one festivl, we estlished GIS to relte sensitive hitts to wste littering. Differences in litter indictors cross event zones nd phses s well s the impct of litter-reducing mesures were tested using generlized liner mixed models (GLMM). Wste littering reched vlues up to 35 g h 1 m 2, minly in the form of glss ottles. Stge, pth, food, nd snittion zones showed significntly higher littering per hour nd per re thn entrnce zones, nd littering ws significntly lower during the rrivl phse thn during the event nd deprture phses. Fencing in comintion with entrnce controls, in lloction, nd prticiptory litter collection systems decresed overll littering significntly. In ddition, requiring deposits for cups nd providing visitor informtion on nners significntly impcted littering of residul (non-glss) wste. Festivls led to cler short-term decrese in vegettion cover. Mpping wste littering in reltion to vlule hitts highlights the conflict etween festivl use nd nture conservtion. We conclude tht fencing, in lloction, nd prticiptory litter collection systems re the most efficient mesures to reduce littering t festivls nd should hve high priority in vlule hitts. In prticulr, event plnners should develop specific mesures for wste-intensive stge, pth, food, nd snittion zones. Overll, the study provides evidence tht specific OPIs re more effective t ssessing litter mngement t huge festivls thn generl OPIs. 211 Elsevier Ltd. All rights reserved. 1. Introduction Tourism-relted lnd use is closely linked to incresed solid wste genertion (e.g. Kuniyl et l., 23; Petrosillo et l., 26; Silv-Cvlcnti et l., 29). This lso pplies to the huge festivls tht hve ecome common tourist destintion in mny countries of the world (e.g. Jones, 29; Dickson nd Arcodi, 21). Wste ccumultion in semi-nturl res hs detrimentl effects on soil, wildlife, nd lndscpe ppernce (Sntos et l., 25; Moriwki et l., 29; Silv-Cvlcnti et l., 29). Therefore, mny event mngers mke sustntil effort to counterct wste production, efforts tht re lso ttriuted to high collection nd Corresponding uthor. Tel.: ; fx: E-mil ddress: rne.cierjcks@tu-erlin.de (A. Cierjcks). disposl costs. Consequently, severl guidelines nd numer of possile mesures for wste reduction nd seprtion hve een generted in recent yers (e.g. Bundesministerium für Umwelt, Nturschutz und Rektorsicherheit nd Deutscher Olympischer Sportund, 27; Bundesministerium für Umwelt, Nturschutz und Rektorsicherheit, 28). However, the success of wste mesures hs only scrcely een ssessed systemticlly using, for exmple, environmentl mngement methodologies such s the Eco-Mngement nd Audit Scheme (EMAS) of the EU or the interntionl stndrd for environmentl mngement systems ISO/EN ISO 141. These mngement schemes require detiled environmentl impct ssessment sed on environmentl performnce indictors (EPIs, in the sense of DIN EN ISO 1431, 1999). Such indictors help to monitor the implementtion nd the efficiency of wste mngement systems nd the resulting environmentl X/$ see front mtter 211 Elsevier Ltd. All rights reserved. doi:1.116/j.ecolind Plese cite this rticle in press s: Cierjcks, A., et l., Opertionl performnce indictors for litter mngement t festivls in seminturl lndscpes. Ecol. Indict. (211), doi:1.116/j.ecolind

2 ECOIND-918; No. of Pges 1 2 A. Cierjcks et l. / Ecologicl Indictors xxx (211) xxx xxx performnce of festivls rrngements. As wste genertion depends on the environmentl performnce of n orgniztion s opertions, wste indictors elong to the ctegory of opertionl performnce indictors (OPIs, DIN EN ISO 1431, 1999), which cn e clssified s environmentl policy indictors, ccording to the indictor concept of Heink nd Kowrik (21). Some festivls hve pulished generl wste indictors presenting the overll wste mount or relting it to the numers of visitors nd festivl dys (Tchonoglous et l., 26; Sthl et l., 26; Cierjcks nd Diefencher, 26; Cierjcks et l., 28). Although these my llow for the evlution of generl wste mngement, their suitility for plnning specific wste mesures hs een little tested in prctice, s hs een shown to e the cse for other indictor systems (Heink nd Kowrik, 21). In prticulr, they re unle to demonstrte the success of specific wste reduction mesures nd they do not offer festivl plnners informtion for predicting wste distriution in the field. Littering is mtter of incresing concern not only t festivls ut lso generlly in mny densely populted res of the world (Al- Khti et l., 29; Moriwki et l., 29; Dngi et l., 211) nd in nturl res sujected to tourism (Kuniyl et l., 23; Özdilek et l., 26). Littering is influenced y visitor density ut lso y site chrcteristics nd visitor ehvior (Sntos et l., 25; Silv- Cvlcnti et l., 29). Consequently, specific OPIs re needed tht llow expected litter mounts nd dmge to the environment to e modelled t given site. Bsed on such indictors, the efficiency of wste reduction mesures my e tested. In this study, we mesured littering t three open-ir festivls in northern Germny (Rheinkultur Festivl, Wcken Open Air Festivl, Fusion-Festivl) nd clculted specific OPIs for litter distriution in the festivl re. As littering depends on visitor density, we determined solute nd density-relted indictors nd tested the impct of litter reduction mesures. For the Rheinkultur Festivl, GIS ws developed tht comined the occurrence of vlule hitts nd the mount of litter to highlight possile conflicts etween festivl use nd nture conservtion. In prticulr, we elorted on the following hypotheses: (1) The proposed OPIs indicte differences in littering mong festivl zones nd phses. (2) The proposed OPIs llow for the evlution of litter reduction mesures. (3) Dmge to vlule hitts my e detected y pplying the OPI results in GIS. Bsed on the results, generl recommendtions for litter mngement plnning t open-ir festivls re given. The conclusions Fig. 1. Loction of studied festivls in Germny. my lso offer vlule insights for visitor mngement in seminturl res such s municipl prks. 2. Mterils nd methods 2.1. Study festivls Dt on solid wste littering were collected for three Germn festivls in 29 (Fig. 1). Rheinkultur Festivl represents the lrgest entrnce-free rock festivl in Germny with up to 2, visitors. This 1-dy festivl hs een orgnized yerly in the Rheinuen in Bonn, North Rhine-Westphli (5 7 N, 7 1 E) since The orgnizers re highly experienced in environmentl mngement nd hve received frequent wrds in the pst (e.g. Sounds for Nture Festivl since 25, Green N Clen Awrd in 28 nd 29). The Rheinuen re is used for vrious festivls throughout the yer, nd lthough mnged s n urn prk, it is chrcterized y high proportion of semi-nturl vegettion with typicl speciesrich riprin vegettion, including hrdwood nd softwood forests. In 29, 17, people visited the Rheinkultur Festivl nd produced totl of 51 t of wste (Tle 1). The entire festivl re is fenced, nd visitors re not llowed to ring food nd drink (except 1 l of wter) from outside. The festivl orgnizers provide 24-l ins nd tringulr fenced structures for wste collection nd orgnize ottle collection t the entrnces. In 29, ins for different solid wste components (glss, plstics nd other municipl wste; ctegories ccording to Sttistisches Bundesmt, 21) were locted t the mjority of the entrnces, nd visitors were informed of the Tle 1 Chrcteristics nd opertionl performnce indictors (OPIs) of the studied festivls. Festivl Fusion-Festivl Wcken Open Air Festivl Rheinkultur Festivl Festivl chrcteristics Festivl re (h) Totl wste (t) Visitors in 29 6, 75, 17, Durtion (d) Generl OPI Wste per visitor nd dy (kg pers. 1 d 1 ) Specific OPIs Numer of study plots Men (SE) littering (g h 1 m 2 ) 2.13 (.97) 2.83 (.74) 1.23 (6.7) Men (SE) littering inside the festivl re (g h 1 m 2 ) 2.38 (1.12) 2.66 (.7) 1.98 (.41) Men (SE) littering per person (g pers. 1 h 1 m 2 ).39 (.19).82 (.41) 3.1 (3.8) Men (SE) littering per person inside the festivl re (g pers. 1 h 1 m 2 ).42 (.21).83 (.42).18 (.6) Estimtion. Plese cite this rticle in press s: Cierjcks, A., et l., Opertionl performnce indictors for litter mngement t festivls in seminturl lndscpes. Ecol. Indict. (211), doi:1.116/j.ecolind

3 ECOIND-918; No. of Pges 1 A. Cierjcks et l. / Ecologicl Indictors xxx (211) xxx xxx 3 environmentl commitment of the Rheinkultur Festivl through nners nd n informtion desk. The Wcken Open Air Festivl is one of the most fmous hrd rock festivls in the world. Up to 75, fns visit the grrin lndscpe of Schleswig-Holstein every yer nd cmp for out 4 dys in August on the medows surrounding the smll villge of Wcken (54 N, 9 4 E). In ddition to the intensively mnged grss lnds, ecologiclly vlule hitts such s the typicl northern Germn hedgerows (Knicks) nd wetlnd hitts re found within the festivl re; most re thoroughly fenced off during the festivl. Also, the stge zone is fenced off, nd for security purposes the visitors re not llowed to ring outside food or drinks into this zone. The festivl orgnizers hve recently strted implementing severl dditionl wste collection fcilities to reduce the high mounts of litter (Tle 1). In 29, 2 h clen zone ws estlished in the cmping zone, where ins with wste seprtion fcilities nd visitor informtion nners were locted nd littered wste ws removed every dy. The ltter ws supported y prticiptory litter collection system in which promotionl fn mterils were given to volunteers who collected litter in the festivl re. The Fusion-Festivl is multimedi festivl locted on former Russin militry ir se in Lärz ner Müritz Ntionl Prk (Mecklenurg-West Pomerni, 53 3 N, 12 7 E). Founded in 1997, it tkes plce every June for 4 dys. Due to the proximity to the ntionl prk, the festivl re is chrcterized y high species richness nd different protected hitts. In 29, the Fusion- Festivl ttrcted out 6, visitors. The orgnizers consider the environmentl spects of festivl mngement nd hve estlished tight prticiptory litter collection system with wste deposit every visitor is oliged to py t the entrnce. The deposit is only returned if the visitor rings full g of litter to the collection points t the end of the festivl. In 29, dditionl glss ottle collection during the festivl ws crried out to reduce glss littering in the festivl re. Unfortuntely, exct numers on overll wste mount were not ville for 29 nd hd to e estimted y the orgnizers (Tle 1) Study design nd dt collection For the purposes of nlysis, ech festivl re ws divided into the following eight zones: (1) stge, (2) food (including eer grdens), (3) pths, (4) snittion, (5) cmping, (6) rest zone (Rheinkultur Festivl only), (7) entrnce (within the festivl re), nd (8) entrnce (outside the festivl oundries). At the 1-dy Rheinkultur Festivl, cmping zones did not exist, ut it ws the only festivl tht offered delineted rest zones where visitors could sit nd et. In most festivl zones, we estlished 5 m 5 m study plots for the ssessment of litter, her cover, nd crowd density, for totl of 11 study plots (Tle 1). All zone types were present more thn once t ech festivl, which llowed for the estlishment of independently replicted study plots within ech festivl. We lso considered three festivl phses: (1) the rrivl phse efore the first concert, (2) the event phse, nd (3) the deprture phse fter the lst concert. For ech study plot, we recorded possile litter reducing fcilities such s the presence of ins within 5 m of the study plot, deposits on cups in eer grdens, prticiptory litter collection system nd in situ visitor informtion using nners. Bsed on their position in reltion to perimeter fences, study plots were grouped into three fencing ctegories: (1) no fencing (free ccess to study plot), (2) outside the fence efore visitor control, nd (3) inside the fence fter visitor control. All festivl zones, study plots, wste Tle 2 Clcultion of opertionl performnce indictors (OPIs). OPI (unit) Vriles considered Clcultion Totl littering (g h 1 m 2 ) Weight of litter collected + weight of new ottles per phse (g) Time since eginning of the festivl or since lst litter Are of study plot (m 2 ) Glss wste littering (g h 1 m 2 ) Weight of glss litter collected + weight of new glss ottles per phse (g) Time since eginning of the festivl or since lst litter Are of study plot (m 2 ) Residul wste littering (g h 1 m 2 ) Weight of residul litter collected + weight of new plstic ottles per phse (g) Time since eginning of the festivl or since lst litter Are of study plot (m 2 ) Totl littering per person (g pers. 1 h 1 m 2 ) Weight of totl litter collected + weight of new ottles per phse (g) Numer of persons within study plot during 3 s Time since eginning of the festivl or since lst litter Are of study plot (m 2 ) Glss wste littering per person (g pers. 1 h 1 m 2 ) Weight of glss litter collected + weight of new glss ottles per phse (g) Numer of persons within study plot during 3 s Time since eginning of the festivl or since lst litter Are of study plot (m 2 ) Residul wste littering per person (g pers. 1 h 1 m 2 ) Weight of residul litter collected + weight of new plstic ottles per phse (g) Numer of persons within study plot during 3 s Time since eginning of the festivl or since lst litter Are of study plot (m 2 ) Weight/(time * re) Weight/(time * re) Weight/(time * re) Weight/(numer of persons * time * re) Weight/(numer of persons * time * re) Weight/(numer of persons * time * re) Plese cite this rticle in press s: Cierjcks, A., et l., Opertionl performnce indictors for litter mngement t festivls in seminturl lndscpes. Ecol. Indict. (211), doi:1.116/j.ecolind

4 ECOIND-918; No. of Pges 1 4 A. Cierjcks et l. / Ecologicl Indictors xxx (211) xxx xxx ins, litter reduction mesures, nd fences t the Rheinkultur Festivl were mpped using ArcMp 9.2. The exct position of ins nd study plots ws determined using sic GPS hndheld unit (Grmin etrex H). Ech study plot ws visited during the festivl t lest once dy nd t lest once during ech festivl phse. During ech visit, litter except ottles ws collected, nd to document short-term impcts on vegettion, the cover of herceous vegettion ws estimted visully. Littered glss nd plstic ottles were counted seprtely to counterct ny influence of wste collection on littering s oviously clener study plots might led to decrese in littering. Glss nd plstic ottles were only counted t their first ppernce y clculting the numer oserved t given smpling minus the numer t lst smpling, ssuming tht no ottles disppered etween smplings. A smple of 15 ottles ws weighed to clculte the men weight of littered ottles resulting in 35 g per glss ottle nd 59 g per plstic ottle. In ddition, ll persons who were stnding inside the plot or pssed through it over period of 3 s were counted. To clculte the time spn since the lst smpling, we recorded the exct time. In cmping zones, we refrined from removing wste to reduce the disturnce to visitors nd their littering ehvior. In ddition, sptil heterogeneity ws higher in cmping res due to the presence of tents nd crs, nd ulky wste ccounted for high percentge of litter. Therefore, we estlished lrger plots of 15 m 15 m nd estimted the overll weight of the residul wste visully. Agin, glss nd plstic ottles were counted seprtely. 3. Theory/clcultion 3.1. Dt nlysis nd clcultion of opertionl performnce indictors (OPIs) Generl, nonspecific OPIs hve little prcticl relevnce, nd, s with other indictor systems (Heink nd Kowrik, 21), their direct correltion to individul litter reducing fcilities instlled y the festivl orgnizers hs scrcely een tested. Therefore, we developed specific OPIs sed on littering mounts for wste mngement t festivls nd wste ssessment of tourism-relted lnd use in semi-nturl res (Tle 2). Collected litter ws weighed, nd the percentges of glss (except entire glss ottles, which were counted seprtely) nd residul wste (except plstic ottles) were estimted. Overll littering ws clculted y dding in ottle weight. Then totl littering within ech phse ws summed if more thn one smpling per phse ws performed, nd the time spn etween given smpling time nd the festivl strt or the lst smpling ws determined for ech phse. Bsed on these dt, we clculted littering per hour nd squre meter (Tle 2). As glss nd residul wste re often treted differently for litter reduction mesures (e.g. specific glss ins, deposits for cups), we lso determined littering of glss wste per hour nd squre meter s well s littering of residul wste per hour nd squre meter. Differences mong these indictors my e relted to different visitor numers in study plots ut lso to distinct visitor ehvior (Sntos et l., 25). To disentngle these two possiilities, we dditionlly clculted OPIs stndrdized for visitor numer in ech plot nd determined litter per person, hour, nd squre meter oth for totl wste nd for glss nd residul wste seprtely (Tle 2) Sttisticl nlyses Due to the nested nd, in ddition, unlnced study design we used generlized liner mixed models (GLMM) for nlysis (Zuur et l., 29). Festivl nd study plots were treted s rndom fctors with the fctor study plots nested within the fctor festivl nd festivl phse nested within the fctor study plots. Festivl zone, festivl phse, nd litter reduction mesures such s in lloction, fencing, nners, nd cup deposits were treted s fixed fctors. For ech of the six OPIs (littering per hour nd squre meter nd littering per person, hour, nd squre meter for totl wste, glss wste, nd residul wste, respectively), we clculted seprte models. In ddition, we rn further clcultion for the Rheinkultur Festivl to divide festivl zones into different littering clsses. Here, we treted study plots s rndom fctors with festivl phse nested within the fctor study plots nd included festivl zone nd festivl phse s fixed fctors. This pproch led to the following significntly different groups of zone clsses (the entrnce zone just inside the festivl re, the stge zone, nd the entrnce zone outside the festivl re) nd one intermedite zone clss with ll other zones which were used for GIS nlysis (low littering, intermedite littering, high littering nd very high littering). All clcultions wre crried out using the glmmpql function in MASS pckge of R version GIS nlysis At the Rheinkultur festivl, ll hitts present within the festivl re were mpped in dvnce of the festivl in 29 ccording to the methodology of Lndesnstlt für Ökologie, Lndschftsentwicklung und Forstplnung Nordrhein-Westflen (1982) using ArcGis 9.2. We ssessed the qulity of ech hitt y pplying three-clss evlution scheme (low, intermedite, high qulity) dpted from Kule (1991). Criteri for evlution were legl conservtion sttus ccording to the regultions of North Rhine-Westphli ( 62 LG NRW) nd the Flor Fun Hitt Directive (92/43/EEC), representtiveness in terms of species composition nd vegettion structure, nd humn disturnce. In ddition, the men littering ws mpped cross the festivl zones sed on the four clsses resulting from GLMM for the Rheinkultur Festivl: low littering (. 1.4 g h 1 m 2 ), intermedite littering ( g h 1 m 2 ), high littering ( g h 1 m 2 ), nd very high littering ( 4.5 g h 1 m 2 ). The conservtion nd littering mps were comined to show possile conflicts etween nture conservtion nd festivl use of the re ccording to the following conflict clsses, which re sed on the rules given in Tle 3: low conflict potentil, intermedite conflict potentil, nd high conflict potentil. 4. Results 4.1. Differences in littering mong festivls, festivl zones, nd phses Wste littering rnged from. to 35. g h 1 m 2. Men littering showed cler differences mong the studied festivls (Tle 1). Both OPIs on littering nd littering per person re reflected in the overll mount of reported festivl wste per visitor nd dy if the festivl s interior re is considered lone. We found significnt differences mong festivl zones with the highest vlues in the stge, food, pth, nd snittion zones; intermedite littering in cmping nd rest zones; nd low littering in entrnce zones (Fig. 2). The high men for exterior entrnce zones is relted to the fencing impct. In the GLMM, there ws no significnt difference mong exterior nd interior entrnce zones. Littering of glss wste ws significntly lower in entrnce zones inside the visitor control res thn in ll other zones. Moreover, residul litter ws significntly lower in the entrnce nd rest zones Plese cite this rticle in press s: Cierjcks, A., et l., Opertionl performnce indictors for litter mngement t festivls in seminturl lndscpes. Ecol. Indict. (211), doi:1.116/j.ecolind

5 ECOIND-918; No. of Pges 1 A. Cierjcks et l. / Ecologicl Indictors xxx (211) xxx xxx 5 Tle 3 Rules for determining conflict clsses in GIS sed on littering clsses nd hitt vlue clsses t the Rheinkultur Festivl. Littering clss Hitt vlue Low Intermedite High Low littering (. 1.4 g h 1 m 2 ) Low Low Intermedite Intermedite littering ( g h 1 m 2 ) Low Intermedite High High littering ( g h 1 m 2 ) Intermedite High High Very high littering ( 4.5 g h 1 m 2 ) Intermedite High High thn in stge, food, nd cmping zones; pth nd snittion zones showed intermedite vlues nd did not show cler differences. Considering visitor density, food zones exhiited significntly higher littering per person thn entrnce zones, nd densityrelted littering in stge nd snittion zones ws significntly higher thn in interior entrnce zones (Fig. 2). Also glss littering per person ws significntly less pronounced in interior entrnce zones thn in other zones, nd cmping zones showed clerly lower glss littering per person thn stge, pth, food, rest, nd snittion zones. A similr pttern ws detected for littering of residul wste per person with interior entrnce zones exhiiting significntly lower vlues thn food, snittion, nd cmping zones. In ddition, our dt reveled pronounced temporl dynmics of littering (Fig. 3). Both littering nd littering per person were significntly higher during the event nd the deprture phses thn during the rrivl phse. We did not find significnt differences in glss littering specificlly, ut the pttern ws reflected in residul wste indictors with littering eing lower during the rrivl phse nd littering per person showing significnt differences cross ll phses Success of implementing litter reducing fcilities Litter reducing infrstructurl fcilities clerly influenced the mount of littered wste. Fencing in comintion with visitor control showed cler impct on littering (Fig. 4). The sence of fences resulted in intermedite increse in levels of totl littering nd littering per person, wheres littering in the res just eyond visitor control ws significntly lower. However, littering incresed significntly efore visitors were checked for ottles nd food. This pttern ws lso consistent in solute littering of glss wste, wheres residul wste ws overll significntly lower fter visitor control thn in other fencing situtions. In terms of littering per person, we did not detect ny consistent pttern for the individul wste frctions. Bin lloction significntly reduced totl littering nd littering of glss nd residul wste frctions (Tle 4). In ddition, we found significnt reduction in littering per person in res ner ins, ut this finding ws not consistent for glss nd residul wste littering per person. Also, prticiptory systems which rely on litter collection through the festivl s udience proved to significntly 7 6 Residul wste Glss Littering (g/h*m²) Littering per person (g/pers.*h*m²) Stge Stge Food Food Pth Pth Snittion Cmping Rest zone Entrnce (interior) Snittion Cmping Rest zone Entrnce (interior) Residul wste c c c c c c Glss Entrnce (exterior) Entrnce (exterior) Fig. 2. Differences in men wste littering per hour nd squre meter () nd men wste littering per person, hour nd squre meter () mong festivl zones. Error rs: stndrd error. Lower cse letters refer to significnt differences in GLMM (p <.5). The high vlue for exterior entrnce zone is consequence of fencing nd not relted to zone type. Plese cite this rticle in press s: Cierjcks, A., et l., Opertionl performnce indictors for litter mngement t festivls in seminturl lndscpes. Ecol. Indict. (211), doi:1.116/j.ecolind

6 ECOIND-918; No. of Pges 1 6 A. Cierjcks et l. / Ecologicl Indictors xxx (211) xxx xxx 14 7 Residul wste Glss Littering (g/h*m²) Littering per person (g/pers.*h*m²) Arrivl Event Deprture Arrivl Event Deprture Fig. 3. Differences in men wste littering per hour nd squre meter () nd men wste littering per person, hour nd squre meter () mong festivl phses. Error rs: stndrd error. Lower cse letters refer to significnt differences in GLMM (p <.5). Littering (g/h*m²) No fencing c Fence (efore Fence (fter visitor control) visitor control) Littering per person (g/pers.*h*m²) No fencing Residul wste c Glss Fence (efore Fence (fter visitor control) visitor control) Fig. 4. Impct of fencing on men wste littering per hour nd squre meter () nd men wste littering per person, hour nd squre meter (). Error rs: stndrd error. Lower cse letters refer to significnt differences in GLMM (p <.5). reduce solute nd per-person littering. A deposit on cups, which ws exclusively relted to food zones, did not led to significnt littering reduction overll, ut decresed littering nd littering per person of the residul wste frction ws seen. Moreover, totl littering per person ws significntly lower in plots with cup deposit. A similr pttern ws found for the presence of visitor informtion in the vicinity of the study plots regrding wste disposl fcilities nd environmentl spects using nners. Here, we detected Tle 4 Differences in littering nd littering per person etween plots with nd without litter reduction mesures. p-vlues refer to significnt differences in GLMM. Litter reduction mesure Men littering (g h 1 m 2 ) Men littering per person (g pers. 1 h 1 m 2 ) Totl Residul wste Glss Totl Residul wste Glss Bin lloction No 9.75 (3.1) 1.34 (.24) 8.41 (2.96) 2.86 (1.88).33 (.13) 2.53 (1.86) Yes 1.96 (.27) 1.3 (.19).66 (.2).48 (.14).36 (.13).12 (.4) p Prticiptory litter reduction system No 6.63 (1.79) 1.49 (.17) 5.14 (1.76) 1.92 (1.11).39 (.11) 1.53 (1.1) Yes 1.46 (.56).26 (.6) 1.2 (.54).27 (.9).9 (.3).18 (.8) p.35 < <.1.38 Cup deposit No 6.38 (1.78) 1.38 (.17) 5. (1.74) 1.91 (1.1).39 (.11) 1.52 (1.9) Yes 2.85 (.71).93 (.25) 1.91 (.69).19 (.7).4 (.1).15 (.7) p Bnner No 6.18 (1.63) 1.36 (.16) 4.82 (1.6) 1.77 (1.1).35 (.1) 1.41 (.99) Yes.61 (.21).61 (.21). (.).17 (.6).17 (.6). (.) p Plese cite this rticle in press s: Cierjcks, A., et l., Opertionl performnce indictors for litter mngement t festivls in seminturl lndscpes. Ecol. Indict. (211), doi:1.116/j.ecolind

7 ECOIND-918; No. of Pges 1 A. Cierjcks et l. / Ecologicl Indictors xxx (211) xxx xxx 7 significnt reduction of residul litter oth in solute terms nd per person. hrdwood remnnts s well s woodlnd ptches with ntive species, which hve high protection sttus in conservtion lw Conflicts with nture conservtion Men cover of herl vegettion decresed during the festivls from 6.6% t visitors rrivl to 53.3% during the event nd to 47.8% t deprture, which represents men loss of 21.1% of the originl vegettion cover during the festivls. At Rheinkultur, GIS nlysis reveled severl res where conflicts etween nture conservtion nd festivl use were present (Fig. 5). Most conflicts were detected in forest ptches locted within the festivl re. These forests were riprin softwood nd 5. Discussion 5.1. Differences in littering mong festivls, festivl zones, nd phses Indictors of overll wste genertion t the Rheinkultur Festivl nd the Fusion-Festivl re in line with dt from other festivls nd events rnging from.2 to.4 kg pers. 1 d 1 ( kg pers. 1 d 1 for the Whole Erth Festivl, festivl of eduction, music, nd rt in Dvis, USA; Tchonoglous et l., 26;.44 kg pers. 1 d 1 for the sports event 26 FIFA World Fig. 5. Assessment of hitts (), men littering (), nd conflict potentil etween nture conservtion nd festivl use (c) t Rheinkultur Festivl, Bonn. Plese cite this rticle in press s: Cierjcks, A., et l., Opertionl performnce indictors for litter mngement t festivls in seminturl lndscpes. Ecol. Indict. (211), doi:1.116/j.ecolind

8 ECOIND-918; No. of Pges 1 8 A. Cierjcks et l. / Ecologicl Indictors xxx (211) xxx xxx Cup in Germny; Sthl et l., 26; kg pers. 1 d 1 for the religious nd culturl event Germn Protestnt Church Convention; Deutscher Evngelischer Kirchentg, 29). Compred to indictors on municipl wste (e.g. 1.2 kg pers. 1 d 1 in Germny; Sttistisches Bundesmt, 21;.5 1. kg pers. 1 d 1 in Chin; Zhng et l., 21), these results point to low overll wste genertion t festivls which is supported y the fct tht very different event types in different countries show similr generl OPIs. Interestingly, ll mentioned festivls nd events hve crried out wste reduction nd recycling mesures such s cup deposits nd visitor controls which my hve led to low generl OPIs. In contrst, the Wcken Open Air Festivl showed much higher vlues, which my indicte lower efficiency of wste mngement. However, the socio-economic ckground of the festivl udience, such s ge structure nd income, lso influences wste disposl nd littering ehvior nd my explin the differences mong festivls (Al-Khti et l., 29; De Feo nd De Gisi, 21). In ddition, further studies on other festivls in different countries will e necessry to infer generl ptterns of overll wste genertion. Nevertheless, the trends visile in the generl wste indictors re lso reflected in our more specific littering indictors, which refer only to the interior of the festivl re. Thus, the proposed OPIs cn e used equivlently for ssessing the overll wste mngement. However, when the exterior entrnce zones, which re not considered in generl wste disposl dt, re included, the Rheinkultur Festivl shows the highest littering indictors due to prticulrly high glss littering in these zones resulting in high stndrd errors. Littering is known to e unevenly distriuted in the field (De Arujo nd d Cost, 27; Silv-Cvlcnti et l., 29). This ws lso true in our study festivls, where the comprison of OPIs reveled cler differences in littering mong the different festivl zones. Asolute littering ws higher in stge, food, pth, nd snittion zones compred to entrnce zones. These differences were less pronounced when compring density-relted OPIs, which implies tht most of these differences hve to e ttriuted to visitor density. However, our dt lso point to differences in visitor ehvior s littering per person ws significntly higher in food zones thn in entrnce zones nd significntly lower in interior entrnce zones compred to stge nd snittion zones. In food zones, visitors hve to del with the disposl of food nd pckge wste. Consequently, more wste is littered in this zone. In contrst, in interior entrnce zones, visitors lck mterils to consume nd therefore littering is expected to e lower. Concerning different wste frctions, our dt reveled differences in littering qulity in contrst to other littering studies (Silv-Cvlcnti et l., 29). Although glss littering ws more or less evenly distriuted cross the festivl res, except for the interior entrnce zones, glss littering per person ws lso lower in cmping zones compred to most other festivl zones. In contrst, littering of residul wste showed more complex pttern with greter levels in stge, food, snittion, nd cmping zones. Agin visitor ehvior seems to e relevnt, s littering of residul wste per person ws lso higher in food, snittion, nd cmping zones presumly due to higher mteril turnover in these zones. Aprt from sptil vrition, we lso detected pronounced temporl dynmic of littering. Both littering nd littering per person were lower during the rrivl phse, which is proly consequence of lower visitor densities ut lso of reduced littering disposition nd mteril consumption t the eginning of festivl. This pttern my e importnt for litter reduction mesures s it implies the need to djust litter mngement ccording to the festivl phse. Overll, the presented OPIs llow for the prediction of littering nd its distriution in the field during ll phses of festivl. This knowledge cn e used to optimize in numer nd plcement during festivl. As our results re sed on dt from different festivl udiences, these ptterns seem not to e strongly relted to the socio-economic ckground of the visitors. Therefore, these findings my e trnsferred to other festivls nd tourist events in semi-nturl lndscpes. In prticulr, high littering in food, stge, pth, nd snittion zones fter the rrivl phse requires explicit wste reduction mesures Success of implementing litter reducing fcilities All types of litter reducing fcilities nd instlltions proved to hve significnt impct. Fencing in comintion with visitor control is strightforwrd mesure to significntly reduce littering inside the festivl re. As oth solute nd reltive littering showed the sme pttern, this reduction is due to chnge in littering ehvior. Unfortuntely, the decrese in littering fter pssing through the visitor control re is countercted y greter levels efore visitor control, which my e ttriuted to incresed consumption efore the visitors enter the festivl re. The resulting litter in mny cses ws not mnged y the festivl itself if visitor control is crried out t the outer oundry of the festivl re, resulting in reduced disposl costs. Still, in terms of environmentl mngement, this spect should e considered when plnning litter reducing investments. Moreover, fencing nd visitor controls re cost-intensive nd require high mnpower. Therefore, this mesure is only prcticl if fencing lso mkes sense for security purposes. Bin lloction lso led to decrese in littering. This is n interesting result s other littering studies found tht wste littering nd in cpcity nd position were not necessrily linked (Hee et l., 25). We found significnt decrese in density-relted littering, which gin points to chnge in littering ehvior cused y the vilility of ins. Consequently, festivl mngers should try to optimize in distriution cross the festivl grounds. At the sme time, decresed willingness of visitors to move to the next in in zones with high visitor density (e.g. stge zones) should e considered s well s security spects. When wste frctions re nlyzed seprtely, littering per person of glss wste showed the sme trends (ut without significnce), which ws not consistent in residul wste. Whether this reflects rel pttern or rndom effects should e mtter for future studies. The estlishment of prticiptory litter collection system comined with regultions nd incentives cused significntly improved OPIs. Similr results hve een shown for municipl solid wste (Tylor, 2) nd office wste in Tiwn (Mrns nd Lee, 1993). These positive effects hve direct dvntges for event mngers, ut the ehviorl chnges my go further nd impct wste genertion ptterns when the visitors re ck t home. Consequently, events my ct s multipliers for environmentl wreness, which underscores the responsiility of festivl orgnizers (Brooks et l., 29). A similr pproch hs een recently used for reducing the cron footprint of visitors in everydy life y communicting spects of climte chnge during festivl (Deutscher Evngelischer Kirchentg, 29). The proposed OPIs (in prticulr the density-relted indictors) my e included into the informtion strtegy to enhnce the visitor s wreness of its own littering ehvior. Requiring deposits for cups to reduce litter t festivls is clssic method tht hs een proven to reduce overll mounts of residul wste (Tchonoglous et l., 26). Here we showed tht littering nd littering per person of residul wste were decresed y cup deposit; this trend ws visile ut not lwys significnt for ll OPIs. Finlly, we found tht in situ visitor informtion using nners lso chnged littering ehvior, which proved to e significnt Plese cite this rticle in press s: Cierjcks, A., et l., Opertionl performnce indictors for litter mngement t festivls in seminturl lndscpes. Ecol. Indict. (211), doi:1.116/j.ecolind

9 ECOIND-918; No. of Pges 1 A. Cierjcks et l. / Ecologicl Indictors xxx (211) xxx xxx 9 for residul wste. Environmentl informtion is known to contriute to wste reduction (Tchonoglous et l., 26). Our dt highlight this fct nd lend support to litter mngement pproches tht comine optimized wste logistics with trgeted visitor informtion. In summry, our dt provide evidence tht single wste reduction mesures cuse cler decrese in littering. The ctivities nlyzed should e considered prticulrly in food, stge, pth, nd snittion zones with high littering. Furthermore, other mesures such s pckge-poor food supply (Sthl et l., 26) wit detiled ssessment of their effect on litter reduction Conflicts with nture conservtion Music festivls nd other temporry events re known to directly or indirectly cuse severl negtive environmentl impcts (Heinzl nd Zimmermnn, 199; Vecchi et l., 28; Vn Hulle et l., 28; Brmn et l., 29). A severe direct consequence of festivl tourism is the dmge to hitts. Hitts my suffer from litter ccumultion ut lso directly from visitor trmpling. The ltter is clerly illustrted y the short-term decrese in vegettion cover during the festivls. At the Fusion-Festivl nd the Wcken Open Air Festivl, long-term consequences on vlule hitts re impeded through fencing. In ddition, species-poor grsslnd in the cmping res is most commonly restored y sowing fter the festivls. In contrst, the Rheinkultur Festivl re is frequently used for festivls throughout the yer without regulr restortion mesures. Consequently, long-term consequences my e expected which is supported y the fct tht the herl vegettion in woodlnds in the udience s rech showed low overll cover of 2 2% in comprison to fenced woodlnd in the re with 29 3% vegettion cover (ccording to 5 vegettion relevés crried out in 29, dt not shown). Woodlnds were the most vlule hitts within the Rheinkultur Festivl re. A comintion of visitor trmpling nd littering is in conflict with conservtion gols nd should therefore e impeded. The OPI for overll littering, which indictes visitor presence nd therefore, indirectly, trmpling, proved to e n effective indictor to detect conflicts using GIS pproch. To counterct dmge to high vlue hitts, fencing of the woodlnds ws proposed for the Rheinkultur Festivl in 21. Other negtive impcts such s nutrient nd orgnic wste input re ssumed to e less relevnt s most of the vegettion of the respective semi-nturl res (grrin lndscpe, urn prk) is dpted to high soil nitrogen. Only in xeric grsslnds in the Fusion- Festivl re, fencing seems dequte to reduce nutrient input. 6. Conclusions The proposed OPIs re highly efficient for predicting litter distriution in the festivl re nd for ssessing the efficiency of litter reduction mesures. Moreover, they reflect the overll wste mounts t the festivls. Our indictors thus proved to e suitle for prcticl ppliction through event mngers nd plnning uthorities, necessry feture ccording to Heink nd Kowrik (21). The indictors re reltively esy to mesure, nd dt collection my e dpted to the specific informtion needs of festivl plnners. Costs for dt collection nd nlysis re expected to e rther low s most of the work cn e tken over y volunteers. This effort my e esily compensted y reduced disposl nd restortion costs in the course of improved wste mngement. In ddition, the findings of our study comining the results of three festivls with different udiences in terms of socil ckground re redily trnsferle to other festivls. The festivl zones we studied re normlly present t ll huge events, nd differences mong zones nd phses re expected to e similr. The sme pplies to the efficiency of litter reducing instlltions nd fcilities, which might e generlized. However, ny environmentl mngement pproch hs to e dpted to the chrcteristics of the festivl oth in terms of the festivl orgnizers nd the udience (Cierjcks nd Diefencher, 26). Therefore, prticiptory litter collection systems nd even the detils of visitor informtion nd refunding of cup deposits hve to e dpted for ech festivl. The success of mny further wste reduction mesures performed y other events such pckge-poor food supply or specific wste recycling fcilities (Sthl et l., 26) hs s yet not een tested. Future studies should focus on filling this gp. The findings on OPIs my e used for mngement of litter in other res such s municipl prks. Bin lloction nd visitor informtion long pth zones nd in food nd stge zones seem promising mesures here. Also nturl res impcted y tourism my profit from trgeted in lloction nd visitor informtion. In ddition, prticiptory litter collection systems comined with incentives seem fesile. However, visitor ehvior my e inherently different in these settings, nd further reserch is needed to understnd the resulting ptterns of littering. Nevertheless, solute nd visitor-relted OPIs my help with detiled nlysis nd to distinguish the impcts of density nd ehvior. Acknowledgments We thnk Sine Funk from RhEINKULTUR GmH, Bonn/Germny, Mrtin Eulenhupt from Kulturkosmos Müritz e.v., Lärz/Germny, responsile for the Fusion-Festivl nd Holger Hüner from ICS GmH, Dörpstedt/Germny, who orgnized the Wcken Open Air Festivl. Oliver Kliemt, Alexndr Tolnd, Stefnie Ferchlnd, Andre Steinmnn, Cludi Fuchs, Birte Schwenkenecher, Corneli Senz, Justin Döring, Arvid Mrkert, Juli Pinter, Mnuel Kelm, Johnn Friee, Juli Schneider, Mrie Pieper, Melnie Fischer, nd Stephn Göel helped with field work nd dt processing. Wilfried Roloff prepred Fig. 1 nd Keline Vrgs kindly checked our English. We thnk two nonymous reviewers for vlule comments to n erlier version of the mnuscript. The dt collection nd nlysis for the Rheinkultur Festivl ws funded y the Deutsche Bundesstiftung Umwelt (DBU, AZ 26775). References Al-Khti, I.A., Arft, H.A., Doud, R., Shwhneh, H., 29. Enhnced solid wste mngement y understnding the effects of gender, income, mritl sttus, nd religious convictions on ttitudes nd prctices relted to street littering in Nlus Plestinin territory. Wste Mnge. 29, Brmn, S.C., Singh, R., Negi, M.P.S., Bhrgv, S.K., 29. Fine prticles (PM2.5) in mient ir of Lucknow city due to fireworks on Diwli festivl. J. Environ. Biol. 3, Brooks, S., Mgnin, A., O Hllorn, D., 29. Rock On!: ringing strtegic sustinle development to music festivls. Prog. Ind. Ecol. 6, Bundesministerium für Umwelt, Nturschutz und Rektorsicherheit, Deutscher Olympischer Sportund, 27. Green Chmpions für Sport und Umwelt Leitfden für umweltfreundliche Sportgroßvernstltungen. 3f Design, Drmstdt. Bundesministerium für Umwelt, Nturschutz und Rektorsicherheit, 28. Leitfden. Einführung eines Energie- und Umweltmngementsystems ei ntionlen und interntionlen Großvernstltungen. Bundesministerium für Umwelt, Nturschutz und Rektorsicherheit, Berlin. Cierjcks, A., Teichert, V., Diefencher, H., 28. Umweltmngement von Großvernstltungen. Digitl Business nd Printing GmH, Berlin. Cierjcks, A., Diefencher, H., 26. Vom Recyclingppier is zur Nullemission Umweltmngement eim Deutschen Evngelischen Kirchentg. In: Altner, G., Leitschuh-Fecht, H., Michelsen, G., Simonis, U.E., v. Weizsäcker, E.U. (Eds.), Jhruch Ökologie 27. Verlg C.H. Beck, München, pp Dngi, M.B., Pretz, C.R., Urynowicz, M.A., Gerow, K.G., Reddy, J.M., 211. Municipl solid wste genertion in Kthmndu, Nepl. J. Environ. Mnge. 92, De Arujo, M.C.B., d Cost, M.F., 27. Visul dignosis of solid wste contmintion of tourist ech: Pernmuco, Brzil. Wste Mnge. 27, De Feo, G., De Gisi, S., 21. Pulic opinion nd wreness towrds MSW nd seprte collection progrmmes: sociologicl procedure for selecting res nd citizens with low level of knowledge. Wste Mnge. 3, Plese cite this rticle in press s: Cierjcks, A., et l., Opertionl performnce indictors for litter mngement t festivls in seminturl lndscpes. Ecol. Indict. (211), doi:1.116/j.ecolind

10 ECOIND-918; No. of Pges 1 1 A. Cierjcks et l. / Ecologicl Indictors xxx (211) xxx xxx Deutscher Evngelischer Kirchentg, 29. Aktulisierte Umwelterklärung 29. Deutscher Evngelischer Kirchentg, Fuld. Dickson, C., Arcodi, C., 21. Promoting sustinle event prctice: the role of professionl ssocitions. Int. J. Hosp. Mnge. 29, DIN EN ISO 1431, Umweltmngement Umweltleistungsewertung Leitlinien. Beuth Verlg, Berlin. Hee, J., Aleidinger, M., Berger, T., Hoffelner, W., 25. Littering ein Schweizer Prolem? Eine Vergleichsstudie Schweiz-Europ. Bundesmt für Umwelt, Wld und Lndschft (BUWAL), Wien. Heink, U., Kowrik, I., 21. Wht re indictors? On the definition of indictors in ecology nd environmentl plnning. Ecol. Indic. 1, Heink, U., Kowrik, I., 21. Wht criteri should e used to select iodiversity indictors? Biodivers. Conserv. 19, Heinzl, R., Zimmermnn, M., 199. Hnduch Umweltschonende Großvernstltung. Leitfden für Plnung und Durchführung unterschiedlicher Vernstltungstypen, second ed. Erich Schmidt Verlg, Berlin. Jones, M., 29. Sustinle Event Mngement: A Prcticl Guide, first ed. London, Erthscn Ltd. Kule, G., Arten-und Biotopschutz, second ed. UTB, Stuttgrt. Kuniyl, J.C., Jin, A.P., Shnnigrhi, A.S., 23. Solid wste mngement in Indin Himlyn tourists treks: cse study in nd round the Vlley of Flowers nd Hemkund Shi. Wste Mnge. 23, Lndesnstlt für Ökologie, Lndschftsentwicklung und Forstplnung Nordrhein- Westflen, Biotopkrtierung Nordrhein-Westflen. Grundlgen des Biotop- und Artenschutzes Nr. 4. Lndesnstlt für Ökologie, Lndschftsentwicklung und Forstplnung Nordrhein-Westflen, Recklinghusen. Mrns, R.W., Lee, Y.J., Linking recycling ehvior to wste mngement plnning: cse study of office workers in Tiwn. Lndscpe Urn Pln. 26, Moriwki, H., Kitjim, S., Kthir, K., 29. Wste on the rodside, poi-sute wste: its distriution nd elution potentil of pollutnts into environment. Wste Mnge. 29, Özdilek, H.G., Ylç ın-özdilek, Ş., Ozner, F.S., Sönmez, B., 26. Impct of ccumulted ech litter on Cheloni myds L (green turtle) htchlings of the Smndg cost, Hty, Turkey. Fresenius Environ. Bull. 15, Petrosillo, I., Zurlini, G., Grto, E., Zccrelli, N., 26. Indicting frgility of socioecologicl tourism-sed systems. Ecol. Indic. 6, Sntos, I.R., Friedrich, A.C., Wllner-Kersnch, M., Fillmnn, G., 25. Influence of socio-economic chrcteristics of ech users on litter genertion. Ocen Cost. Mnge. 48, Silv-Cvlcnti, J.S., de Arujo, M.C.B., d Cost, M.F., 29. Plstic litter on n urn ech cse study in Brzil. Wste Mnge. Res. 27, Sttistisches Bundesmt, 21. Umwelt. Afll. 28. Fchserie 19 Reihe 1. Sttistisches Bundesmt, Wiesden. Sthl, H., Hochfeld, C., Schmied, M., 26. Green Gol TM Legcy Report. Orgnistionskomitee FIFA Fussll-Weltmeisterschft 26, Frnkfurt, Bundesministerium für Umwelt, Nturschutz und Rektorsicherheit (BMU), Berlin. Tylor, D.C., 2. Policy incentives to minimize genertion of municipl solid wste. Wste Mnge. Res. 18, Tchonoglous, G., Krginnidis, A., Leverenz, H., Cdji, M., Antonopoulos, I.S., 26. Sustinle wste mngement t specil events using reusle dishwre: the exmple of Whole Erth Festivl t the University of Cliforni, Dvis. Fresenius Environ. Bull. 15, Vn Hulle, S.W.H., Audenert, W., Decostere, B., Hogie, J., Dejns, P., 28. Sustinle wstewter tretment of temporry events: the Drnouter Music Festivl cse study. Wter Sci. Technol. 58, Vecchi, R., Bernrdoni, V., Cricchio, D., D Alessndro, A., Fermo, P., Lucrelli, F., Nv, S., Plzzlung, A., Vlli, G., 28. The impct of fireworks on irorne prticles. Atmos. Environ. 42, Zhng, D.Q., Tn, S.K., Gerserg, R.M., 21. Municipl solid wste mngement in Chin: sttus, prolems nd chllenges. J. Environ. Mnge. 91, Zuur, A.F., Ieno, E.N., Wlker, N.J., Sveliev, A.A., Smith, G.M., 29. Mixed Effects Models nd Extensions in Ecology with R, first ed. Springer, Berlin. Plese cite this rticle in press s: Cierjcks, A., et l., Opertionl performnce indictors for litter mngement t festivls in seminturl lndscpes. Ecol. Indict. (211), doi:1.116/j.ecolind