Accommodation, seasonality and domestic tourism to national parks: implications for environmental policy

sus ainabili y A icle Accommoda ion, Seasonali y and Domes ic Tou ism o Na ional Pa ks: Implica ions o En i onmen al Policy Flo a Ma ia Díaz-Pé ez 1, Ca los Gus a o Ga cía-González 1and Alan Fyall 2,*   Ci a ion: Díaz-Pé ez, F.M.; Ga cía-González, C.G.; Fyall, A. Accommoda ion, Seasonali y and Domes ic Tou ism o Na ional Pa ks: Implica ions o En i onmen al Policy. Sus ainabili y 2021,13, 5072. h ps://doi.o g/10.3390/su13095072 Academic Edi o : Jose And es Coca-S e aniak Recei ed: 25 Ma ch 2021 Accep ed: 27 Ap il 2021 Published: 30 Ap il 2021 Publishe ’s No e: MDPI s ays neu al wi h ega d o ju isdic ional claims in published maps and ins i u ional a il- ia ions. Copy igh : © 2021 by he au ho s. Licensee MDPI, Basel, Swi ze land. This a icle is an open access a icle dis ibu ed unde he e ms and condi ions o he C ea i e Commons A ibu ion (CC BY) license (h ps:// c ea i ecommons.o g/licenses/by/ 4.0/). 1Depa men o Applied Economics and Quan i a i e Me hods, Uni e si y o La Laguna, 38071 La Laguna, Spain; [email p o ec ed] (F.M.D.-P.); [email p o ec ed] (C.G.G.-G.) 2Rosen College o Hospi ali y Managemen , Uni e si y o Cen al Flo ida, O lando, FL 32819, USA *Co espondence: [email p o ec ed] Abs ac : The analysis o seasonali y and domes ic ou ism om he pe spec i e o he accommo- da ion sec o has, o da e, been unsa is ac o ily s udied in he domain o isi o s o na ional pa ks. In ligh o he scale o accommoda ion de elopmen and i s c ucial ole in ou ism, mos no ably wi h ega d o en i onmen al impac , in-dep h knowledge abou accommoda ion ma ke segmen s and hei speci ic cha ac e is ics and pa e ns o beha io a e in eg al o he de elopmen o ou ism policy. In he con ex o domes ic ou ism, unde pinned by an unde s anding o he heo y o planned beha io and push-and-pull mo i a ions, his s udy examines seasonali y and accommoda ion ype used by isi o s o na ional pa ks in he small islands o he Cana ies, Spain wi h he objec i e being o design he mos app op ia e en i onmen al ou ism policy. By adop ing a Mul inomial logis ic eg ession model wi h 1671 su eys ca ied ou , he s udy concludes ha domes ic isi o s demons a e a highe p opensi y o selec en i onmen al- iendly accommoda ion ypes du ing he high season. A he same ime, esiden s who isi he pa ks in high season assess he lowes alue o he p ese a ion le el o he pa ks’ na u al esou ces. Keywo ds: season; ype o accommoda ion; domes ic isi o s; na ional pa ks; mul inomial logis ic eg ession model; p ese a ion le el 1. In oduc ion The s udy o ou ism seasonali y om he speci ic pe spec i e o he accommoda ion sec o has, o da e, been unsa is ac o ily s udied in he academic li e a u e om an empi ical pe spec i e. Ne e heless, since he cons uc ion o any o m o new accommoda ion has a signi ican en i onmen al impac gene ally, and especially on small islands e i o ies, in-dep h knowledge abou hose ma ke segmen s wi h he highes p opensi y o selec en i onmen al- iendly o ms o accommoda ions is c i ical. Fo he impac o seasonali y on des ina ions, clea ly he economic impac is o pa - icula impo ance. An economic a ionale gene ally go e ns he de elopmen o ou is egions [ 1 ]; wi h emphasis in mos des ina ions on he high o “peak” season. Howe e , due o i s con ibu ion o o e coming nega i e en i onmen al e ec s, s udying low-season cha ac e is ics may be a use ul complemen , since a educed numbe s o ou is s could help p ese e he des ina ion’s na u al esou ces mo e e ec i ely. Al hough seasonali y in he Cana y Islands is low, i is possible o dis inguish wo seasons: om No embe o Eas e (high season), and om he end o Eas e o Oc obe (low season). Due o i s a o able clima ic condi ions, win e is included in he high season o he islands ha comp ise he Cana ies. Ne e heless, domes ic ou is s ha e a ema kably di e se beha io wi h sociological and/o ins i u ional easons (such as school aca ion pa e ns) explaining why summe con inues o ep esen he “peak” season. Small island e i o ies gene ally a e mo e agile and su e g ea e de e io a ion [ 2 , 3 ] han mainland des ina ions. Thus, since any kind o new accommoda ion building has a huge impac on small island e i o ies, he ques ion o which accommoda ion ype is Sus ainabili y 2021,13, 5072. h ps://doi.o g/10.3390/su13095072 h ps://www.mdpi.com/jou nal/sus ainabili y Sus ainabili y 2021,13, 5072 2 o 26 chosen by isi o s may be o a pa icula in e es o hose de e mining en i onmen al policy. In ac , as he na u al esou ces a e di e se so a e he e ec s o ou ism on hem: highly dis- o ganized land occupa ion (cons uc ions o la ge blocks o ho els, apa men s, es au an s, e c.) and was e gene a ion, o example. In such ins ances, excessi e de elopmen o he ou ism “p oduc ” can unde mine he o e i sel , making he des ina ions less a ac i e o ou is s. Mo eo e , ou is ac i i y is in ensi e in he consump ion o space. Tou is s and employees demand space o housing, es au an s, leisu e enues, and a he same ime consume sca ce esou ces such as wa e in addi ion o using he land as a was e holde . Hence, wi h he mass-de elopmen and mass-consump ion o ou ism, he e comes an excess o space consump ion, wi h pa icula challenges o small island s a es in gene al. In e es ingly, as ou ism is concen a ed in a eas o educed size, he e is usually a apid and uno ganized g ow h o p i a e cons uc ion. In ac , by lea ing de elopmen o he p i a e sec o , wi hou app op ia e egula ions, a “ ee ide ” p oblem can occu : he in es o s should coope a e wi h each o he o p e en de e io a ion o he en i onmen , bu in eali y, one in es o will maximize i s bene i s i he o he s main ain he en i onmen and o e use i . In si ua ions such as his, ma ke balance is achie ed wi h a widesp ead o e u iliza ion o he common good, in his case he en i onmen . Na ional pa ks a e na u al a ac ions globally ha a e o in e es o a g owing numbe o ou is s [ 4 , 5 ]. Ha ing he cha ac e is ics o p o ec ed a eas, hey p o ide a se ies o asse s ha con ibu e o he main enance o biodi e si y and en i onmen al quali y as well as elemen s ha can be conside ed ec ea ional, he apeu ic, spi i ual, cul u al, a is ic o educa ional [ 6 ]. In addi ion o hese in angible se ices p o ided by na ional pa ks, ano he se o angible se ices such as guides, ba and es au an se ices, mobili y, secu i y, e c., can be conside ed as well. These angible aspec s compose he pa k’s po olio o se ices [ 7 ]. Ano he key componen o hei a ac i eness is a clea sky, p o ec ed in he speci ic con ex o Spain by a law de eloped o pu poses un ela ed o he na ional pa ks bu a he o gua an ee he quali y o he as onomical obse a ion a Izaña (Isla de Tene i e) and Roque de los Muchachos (Isla de la Palma). This ea u e can be an added a ac ion, and no a less impo an one, o i s landscape c ea ed by an imp essi e geology, lo a and endemic auna. The case ocus o his s udy, he Cana y Islands, ecei es annually six imes he numbe o ou is s as i s o al popula ion [ 8 ]. Due o he la ge numbe o ou is isi s o he Cana y Islands Na ional Pa ks (CNP) ( wo ou o h ee ou is s) he de elopmen o con ol measu es wi hin he amewo k o he island’s ou ism policy a e necessa y gene ally, and pa icula ly wi h ega d o he conse a ion o essen ial na u al esou ces. The e o e, he de elopmen o a s udy whe e all he a iables abo e (domes ic/ o eign isi o s, accommoda ion ype and seasonali y) a e included is a equi emen o imp o e he decision-making p ocess in he case o na ional pa ks in he speci ic con ex o he Cana y Islands. 1.1. Tou ism and I s Impac on he En i onmen The ela ionship be ween sus ainable ou ism and he en i onmen can be desc ibed as symbio ic. In ac , ou ism p oduc ion is nou ished by he na u al he i age on which i depends. In his sense, ou ism has he capaci y o gene a e ex e nal bene i s om he en i onmen , con ibu ing o en i onmen al p o ec ion and conse a ion. Indeed, we can obse e how ou ism makes he local popula ion and ou is s become mo e awa e o en i onmen al alues. I can also se e as an ins umen o inance he p o ec ion o na u al a eas [ 9 ]. As s a ed by Ruiz [ 10 ] (p. 348). “Sus ainabili y implies a a ional ela ionship be ween man and na u e”. Sus ainable ou ism can be achie ed by c ea ing “bo om up” social sys ems in e es ed in sha ing expe iences, ideas and esou ces [ 11 ]. In his con ex , he egional scale has been ecognized [ 12 ] as being app op ia e o he implemen a ion o local pa icipa i e ac ions aimed a he sus ainable de elopmen o des ina ions. Howe e , a he same ime, he nega i e impac s o ou ism should no be igno ed, since hey can Sus ainabili y 2021,13, 5072 3 o 26 g adually des oy he en i onmen al esou ces on which ou ism depends. In his sense, i is a gued ha ou ism has become sensi i e o and, simul aneously, dependen on a high-quali y en i onmen [13]. Many ou is ac i i ies ake place in agile ecosys ems, such as p o ec ed na u al a eas o small islands. An ecosys em is a geog aphical a ea ha includes all li ing o ganisms (people, animals, plan s and mic oo ganisms), hei physical en i onmen (land, wa e and ai ) and he na u al cycles ha sus ain hem [ 9 ]. Howe e , he cons uc ion o ho els and ec ea ional in as uc u es gene a es sewage pollu ion, was ewa e ha pollu es i e s, lakes, o hinde s ma ine biological cycles. In addi ion, many na ional pa ks globally a e a ec ed by a ic conges ion and excessi e numbe s o isi o s [ 14 ]. Mo eo e , small islands a e conside ed he mos ulne able ecosys ems in he wo ld [ 15 ]. The physical impac s o ou ism on hese islands could lead o he deg ada ion o hei ecosys ems ha a e aluable ou ism asse s wi h an abundance o hei species. 1.2. Cons uc ion and I s Impac on he En i onmen The cons uc ion sec o is conside ed o be one wi h a high capaci y o d i e he econ- omy. This sec o is capable o s imula ing economic g ow h due o i s induced backwa d and o wa d e ec s on o he sec o s [ 16 – 19 ]. Howe e , cons uc ion ac i i ies also ha e a da k side [ 20 ] o he ex en ha he apid expansion o cons uc ion can cause a a ie y o di icul - o-sol e challenges mainly o a spa ial (land occupa ion) and en i onmen al na u e [ 19 – 23 ]. In his sense, a cons uc ion boom can lead o nega i e e ec s, such as he o e p oduc ion o building s ock [ 23 ] o he in ensi ica ion o en i onmen al de e io a ion h ough u ban sp awl pollu ion and des uc ion o ege a ion. In ecen decades, he emphasis on he nega i e e ec s o cons uc ion has ocused on en i onmen al s ess and sus ainabili y issues. Speci ically, he mos conside ed e ec s ha e been he en i onmen al deg ada ion gene a ed by pollu ion, he damage caused o na u al esou ces and u ban sp awl. Speci ically, i is wo h men ioning he des uc ion o lo a and auna, he use o ma e ials ha m ul o human heal h, and ca bon emissions caused by buildings and he pa icula me hods o p oduc ion and design [ 19 ]. These nega i e e ec s ha e made he concep o sus ainable cons uc ion lou ish, which has become widesp ead and applied, especially in mo e de eloped coun ies [24]. The concep o sus ainable cons uc ion en ails, among o he aspec s, he goal o educing ca bon emissions om buildings, he use o ecycled ma e ials and he use o cons uc ion echniques ha ha e ewe ha m ul e ec s on he en i onmen [ 25 – 27 ]. These goals ha e been achie ed by he go e nmen s o se e al de eloped coun ies h ough he implemen a ion o plans based on “G een building” echniques and he implemen a ion o en i onmen al quali y ce i ica ions [ 28 ]. In addi ion o hese speci ic measu es, he e a e many communi ies ha ask hei go e nmen s o u gen ly adop poli ical ac ions such as ins i u ional changes, o land use con ol measu es o help educe he nega i e e ec s o cons uc ion ac i i ies [16,20,22,27]. 1.3. Seasonali y Conside able e o s ha e been made by ou ism policymake s o educe seasonali y in many des ina ions. This is because i has been iewed as he eason o p oblems de i ed om peaking and o e use o acili ies and esou ces. In ac , seasonali y has been de ined by Bu le as a ‘ empo al imbalance in he phenomenon o ou ism, which may be exp essed in e ms o dimensions o such elemen s as isi o numbe s, isi o expendi u e, highway a ic and o he o ms o anspo a ion, employmen and admissions o a ac ions’ [ 29 ] (p. 5). The e o e, looking o a be e balance in he numbe o ou is s coming along each yea should be a ou ism policy a ge . In his espec , and in o de o measu e he seasonali y o ou is a i als o a pa icula des ina ion, a a ie y o quali a i e e ms can be used such as seasonal index o ac o s (using 12-mon h mo ing a e ages), Skewness, Ku osis o Gini coe icien [ 30 ]. In any case, he choice o measu e will depend on he esea ch ques ion and he deg ee o de ail equi ed. Sus ainabili y 2021,13, 5072 4 o 26 Rega ding he easons o seasonali y, hese can be ound no only in na u al ac o s, bu also in ins i u ional o social ac o s’ [ 31 , 32 ]. Re e ing o na u al ac o s, one o he mos s udied by he li e a u e is he clima e [ 1 , 33 – 35 ], bu he e a e o he ac o s ela ing o social issues, spo s seasons (ski, su ), and also o ine ia and adi ion [ 29 ]. Des ina ions based on clima e condi ions as he main ou is a ac ion, o example ski [ 36 ] o sun and sand ou ism [ 37 , 38 ] su e a s uc u al seasonali y p oblem. On he o he hand, social no ms and p ac ices ( aca ion pe iods, eligious pilg images, cong esses and con en ions) can play a de e minan ole in o he des ina ions. In his espec , he cha ac e is ics o demand, in e ms o he ypes o ou is s and hei coun ies o o igin, o example, can gene a e clea seasonali y p ac ices [39–42]. Di e en s udies ha e shown how ou is s’ coun ies o o igin explain he way hey beha e in e ms o seasonali y [ 43 , 44 ]. A he same ime, only a ew pape s ha e ocused he s udy o ou ism seasonali y om he pe spec i e o he accommoda ion sec o [ 45 – 47 ]. Those s udies ha e been o ien ed o he analysis o p oduc i i y, o he de e minan s o mul i-des ina ion ou ism ips, being he accommoda ion an explana o y a iable in he analysis, bu no an ou come. 1.4. Tou ism in Small Islands Tou ism o islands is second only o his o ical ci ies when i comes o isi a ion le - els [ 48 ] and his esul is no abou pe capi a isi a ion. Howe e , small islands cons i u e agile sys ems wi h high physical and socio-cul u al complexi y. The educed size o he e i o y and he physical condi ions o insula i y inc ease hei ulne abili y and exposu e, o a g ea e ex en han a con inen al e i o y, o hei social and na u al sys ems o he impac s o na u al agen s o o local o isi ing popula ions. In gene al, hese islands ha e high biodi e si y and unique landscapes. They also depend g ea ly on impo ed p oduc s and a e subjec ed o g owing p essu e o he de elopmen o ou ism. The abo e ci cums ances can de e employmen in hese small island e i o ies as well as discou aging he use o simila cons uc ion and design me hods o na u al ou ism ha can be seen in con inen al e i o ies [ 15 , 49 , 50 ]. The e o e, due o he agile sys em small islands cons i u es, he concep o ca ying capaci y should be applied o he s udy o ou ism [51–55]. The a ailabili y o quali y na u al esou ces (landscape, pu e a mosphe e, clean sea wa e , e c.) and unique cul u al asse s a e wha de e mine he a ac i eness o small islands in he con ex o he wo ld ou ism ma ke . The de elopmen o he ou ism sec o , howe e , has an ambi alen impac . On he one hand, i gene a es economic esou ces o he local popula ion, ac ing as an economic engine o small a eas, such as islands. On he o he hand, when ou ism demand eaches le els o o e c owding, i con ibu es o a de e io a ion o na u al esou ces, which, when excessi e, gene a es ad e se impac s in hese e i o ies. The p oblem is ha as he ou is en i onmen deg ades his in u n a ec s he supply o ou is goods. This esul s in a displacemen o ou ism demand owa ds segmen s wi h ewe ecological impac s and less o e c owding, which subsequen ly leads o he de e io a ion in hei en i onmen . 1.5. Objec i es and Resea ch Ques ions The p ima y objec i e o his esea ch is o examine isi o s’ accommoda ion choices when isi ing na ional pa ks in he Cana y Islands acco ding o hei domes ic (Cana y Island) esiden o non- esiden s a us and he impac o hese choices on he p ese a ion o he na u al en i onmen . Ul ima ely, he aim is o de elop di e en ia ed s a egies o each g oup depending on hei g ea e o lesse impac on he en i onmen whe e hey s ay. In pa icula , he s udy aims o explo e he mos ele an en i onmen al ou ism policy ecommenda ions o na ional pa ks in he Cana y Islands and hei insula en i onmen . Indeed, he s udy o ou ism seasonali y om he pe spec i e o he accommoda ion sec o o esiden isi o s has been unsa is ac o ily s udied in he academic li e a u e. In ac , no ele an pape has included he accommoda ion chosen by he ou is as a dependen Sus ainabili y 2021,13, 5072 5 o 26 a iable when adding he season and place o esidence as independen a iables. Ne e - heless, since he cons uc ion o any kind o new accommoda ion has signi ican impac on small island e i o ies, he in o ma ion abou he ma ke segmen s wi h he highes signi icance in luence on ce ain ypes o accommoda ion o his kind o islands may be c ucial. Wi h he gene al objec i e o achie ing a be e unde s anding o he applica ion o a speci ic s a is ical echnique, he mul inomial logis ic eg ession model (MLRM), he app oaches se ou abo e lead o he ollowing speci ic objec i es o his s udy: (1) To achie e a be e unde s anding o he ela ionship among he ollowing a iables and hei ca ego ies: accommoda ion, coun y o o igin and seasonali y when s udy- ing he popula ion o isi o s o a na ional pa k. (2) To be e unde s and he in luence o seasonali y in he con igu a ion o he di e en segmen s o he ou ism ma ke s cu en ly isi ing he na ional pa ks, dis inguishing domes ic (Cana y Island) esiden om non- esiden isi o s. (3) To unde s and he pe cep ion o each segmen o isi o s on he p ese a ion o na u al esou ces (4) To inco po a e, o he i s ime, a MLRM (mul inomial logis ic eg ession model) as an e ec i e and e icien me hod when analyzing he men ioned h ee a iables— seasonali y, accommoda ion ype and esiden /non- esiden o igin o he ou is . (5) Finally, i is expec ed ha his s udy leads o he de elopmen o app op ia e plans aimed a achie ing en i onmen al objec i es, h ough he con ol o he ype o accommoda ions dis ibu ion h oughou he yea in he con ex o he Cana y Islands. Based on he abo e objec i es, his esea ch se s ou o es he ollowing hypo heses applicable o ou ism ma ke segmen a ion: Hypo hesis 1 (H1). I is possible o conduc a quan i a i e analysis o he seasonali y o he accommoda ion used by he ou is , and pa icula ly when di e ences in domes ic (Cana y Island) esiden s and non- esiden isi o s’ beha io is a main esea ch opic. Hypo hesis 2 (H2). Seasonali y has a s ong in luence on he accommoda ion ype o he pa icu- la case o domes ic (Cana y Island) esiden ou ism. Hypo hesis 3 (H3). Domes ic isi o s (Cana y Island esiden s) ha e a di e en pe cep ion o he na u al esou ces p ese a ion s a us in na ional pa ks. Hypo hesis 4 (H4). Fo hose empi ical s udies in which no con inuous a iables a e conside ed, choosing he accu a e dependen (accommoda ion o p ese a ion alue) and independen a iables (domes ic/non- esiden o seasonali y) will be e ec i e using a MLRM. O cou se, he e ec i eness and e iciency a e gua an eed as he esea che has enough da a o apply he analysis o he poin o achie ing he esea ch objec i es. In he pa icula case o his pape , a ep esen a i e sample o he o al popula ion o isi o s o Cana y Island na ional pa ks du ing he 2016–Janua y 2020 pe iod was used. 2. Ma e ials and Me hods 2.1. Theo e ical Unde pinning The heo y o planned beha io (TPB) is among he mos ci ed social science heo ies ha a e used o explain beha io [ 56 ] and one o he mos widely applied heo ies in ma ke ing esea ch [ 57 ]. TPB aims o de elop he pa h owa d beha io al in en ion by ocusing on h ee p ima y componen s; hese a e an indi idual’s a i ude owa d a be- ha io (posi i e o nega i e p edisposi ion o unde lying a i ude), hei subjec i e no ms (likelihood ha signi ican o he s app o e o disapp o e o a pa icula beha io ), and hei pe cei ed beha io al con ol (di icul y pe cei ed in pe o ming a beha io ) [ 58 ]. The heo y has demons a ed i s u ili y in p edic ing indi iduals’ in en ions o engage Sus ainabili y 2021,13, 5072 6 o 26 in ce ain beha io s o a ying deg ees h ough he li e a u e. Fo example, by using a me a-analysis, Han and S oel [ 59 ] iden i ied signi ican (p< 0.001) ela ionships among he TPB componen s o a i ude and subjec i e no ms wi h indi iduals’ pu chase in en ions. The heo y has also been used o p edic and e alua e socially esponsible beha io s, such as sus ainable beha io s [ 60 ]; g een ho el isi a ion [ 61 ], ene gy conse a ion [ 62 ], and sus ainable anspo a ion [ 63 ] ha e all been examined. Mo eo e , wi hin he ma ke ing con ex , TPB may aid p ac i ione s in u he ing hei unde s anding o hei a ge ma ke segmen s [64]. People ha e a ying ypes and deg ees o wan s and needs ha canno be ul illed h ough any one pa icula ma ke ing p oposal [ 65 ]. The e o e, ma ke e s mus de elop ways o a ge hei e o s owa d ce ain g oups, and segmen a ion is one me hod ha can be used o achie e hei goal. Ma ke segmen a ion in ol es dissol ing la ge, di e se consume g oups in o smalle , mo e uni o m clus e s [ 66 ]. De eloping ma ke segmen s a o ds ma ke e s a mo e comple e unde s anding o a pa icula g oup, which ul ima ely in o ms mo e di ec and p oduc i e ma ke ing ini ia i es [ 67 ]. Howe e , ma ke e s need o iden i y how hey would like o de ine and de elop hei a ge segmen s. Bases o seg- men a ion exis o di iding g oups wi hin he ma ke ing li e a u e. Fo example, g oups may be di ided based on socio-demog aphics (gene a ion, educa ion), beha io al cha ac- e is ics (in ol emen le el, usage a e), o psychog aphic ac o s (li es yle, i.e., ad en u e o elaxa ion) [ 68 ]. People’s a i udes and beha io s can se e as bases o ma ke ing seg- men a ion [ 69 ]; including beha io al and psychog aphic in o ma ion allows ma ke e s o de elop a deepe unde s anding o hei a ge ma ke beyond basic demog aphic da a [ 64 ]. The e o e, TPB may help iden i y a ge ma ke segmen s since i elies on indi iduals’ a i udes and beha io s o p edic ou comes. In addi ion, ou ism esea ch has long ocused on wha ac o s d i e ou is s’ a el decisions [ 70 ]. Mo i a ion has been iden i ied as a p ima y d i e o ou is s’ a el beha io s [ 71 ]. O e ime, schola s ha e shaped and ad anced he unde s anding o he ole mo i a ion plays in ou is s’ decision-making p ocesses (e.g., [ 72 ]). Al hough a ious heo e ical pe spec i es exis , he push-pull mo i a ion amewo k has emained one o he dominan lenses h ough which a el mo i a ion has been s udied [ 73 , 74 ]. The amewo k can be used o unco e why ou is s make ce ain des ina ion- ela ed decisions, such as isi ing one des ina ion o e ano he [ 75 ]. Push-pull mo i a ion e ol es a ound wo p ima y concep s; push (pe sonal desi es and in e nal mo i es) and pull (ex e nal des ina ion cha ac e is ics) [ 74 , 76 ]. Push ac o s include he need o ad en u e, i ness, heal h, knowledge, o elaxa ion [ 77 , 78 ]. Pull ac o s may e ol e a ound he acili y quali y, s a demeano , p ice [ 73 ], en e ainmen alue, o singula i y [ 78 ]. Mul iple push and pull a iables o en play signi ican oles in d i ing di e en ou is s’ mo i a ion-d i en beha io s [ 79 ], including he decisions o na u e-based ou is s [ 80 ], who demons a e a ange o mo i a ions ela i e o hei des ina ion en i onmen [81]. 2.2. Cha ac e is ics o Da a Collec ion This s udy is quan i a i e in na u e, hus, in o de o ha e a p ecise knowledge o isi o s o he Cana y Islands Na ional Pa ks (CNP), a s uc u ed ques ionnai e di ided in o wo sec ions was de eloped. A i s sec ion wi h ques ions aimed a ob aining in o ma ion conce ning he mos ou s anding cha ac e is ics o he popula ion unde s udy, among hem, ype o accommoda ion o place o esidence. The second sec ion includes all o he ques ions aimed a collec ing in o ma ion ela ed o he speci ic objec i es o his esea ch, such as season (high o low). An ad-hoc su ey was hen ca ied ou om 2016 o Janua y 2020, allowing a much mo e p ecise knowledge abou he isi o s o he pa ks. The isi o s we e in e iewed in a ace- o- ace p ocedu e o a ound 15 min. The ad-hoc su ey allowed us o ob ain a much mo e p ecise knowledge o isi o s o he NP, wi h he ollowing composi ion: (1) Visi o s om mo e han 20 coun ies, wi h he highes pe cen ages o non- esiden isi o s being Ge man and B i ish; (2) Domes ic ou ism ep esen ed a ound one hi d o he o al popula ion o isi o s. Sus ainabili y 2021,13, 5072 7 o 26 On he o he hand, and wi h espec o he eam o in e iewe s, hey we e selec ed acco ding o hei language skills (English, Spanish, Ge man, F ench and Russian), hus esponding o he languages mos equen ly used by isi o s o he Cana ian Na ional Pa ks as a whole. This allowed each esponden o be in e iewed in his o he own language. Once selec ed, he in e iewe s we e speci ically ained o ca y ou he su ey. Fo example, in he case o isi s by o ganized g oups o amily o iends, he in e iew was ca ied ou wi h a single membe o he g oup o amily uni and he in e iews we e dis ibu ed among he di e en isi poin s, acco ding o he app oxima e equency o isi s o each poin . The in e iewees we e selec ed andomly using he simple andom p obabilis ic sampling p ocedu e, and we e in e iewed in si u in he di e en a eas o he Cana y Islands NP. Fo example, in he case o Teide Na ional Pa k, in e iews we e conduc ed in he ollowing a eas: Co dada de Ga cía, in on o he Pa ado , Mi ado del Llano de Ucanca, Pa ado de Tu ismo and in he Po illo a ea. Fo Calde a de Tabu ien e Na ional Pa k, in e iews we e conduc ed a he Cumb eci a iewpoin , Roque de los Muchachos and he Ba anco exi . In he case o Timan aya Na ional Pa k, isi o s we e in e iewed a Echade o de los Camellos and Mon añas de Fuego. In he case o Ga ajonay, isi o s we e in e iewed a he main access poin s o he pa k. Su eys we e conduc ed in wo phases du ing he pe iod o 2016–2020. The i s phase was conduc ed in he high season in win e and included Eas e . The second was in la e sp ing and ea ly summe which is low season on he Cana y Islands. 2.3. Va iables The a iables conside ed in he analysis a e o se e al ypes: economic (high/low sea- son), a el cha ac e is ics ( ype o accommoda ion, Na ional pa k isi ed), psychological (p ese a ion pe cep ion) and geog aphical (coun y o esidency o he ou is ). In ela ion wi h he coun y o esidency o he ou is , wo di e en modali ies a e s udied, domes ic (Cana y Island) esiden s and non- esiden isi o s. The ou come ‘ ype o accommoda ion’ is classi ied in ou modali ies: (1) ho el, (2) apa men , u al o en ed houses, (3) own house, amily o iend’s house and (4) c uise, camping and o he s (see Table 1). Due o he low pe cen ages o he emaining ca ego ies, we ha e g ouped hem in he las one due o he equi emen s o he s a is ical model ega ding he a ailabili y o he da a. In e e ence o he ype o accommoda ion and conside ing ha any kind o new building will des oy he en i onmen deeply and ha ‘c uise, camping and o he s’ has an e ec on he pollu ion, as well, i is possible o s ablish a classi ica ion om he modali y wi h he highes nega i e e ec on he en i onmen o he one wi h he lowes e ec . Such an o de will be as ollows: (1) ho el, (2) apa men , u al co age o en ed houses, (3) c uise, camping and o he s and (4) own house, amily o iend’s house. In o de o es ablish he p e ious classi ica ion, we ha e conside ed ‘new cons uc ion’ o be ou limi ing c i e ion, which is jus i ied by he a gumen s collec ed by he au ho s in he li e a u e e iew. In o he wo ds, his s udy pa s om he idea ha es ablishmen s ha a e al eady buil and a e no new ou is ic cons uc ions gene ally ha e a lesse impac on he e i o y and i s na u al esou ces han hose ha a e newly buil wi h he sole pu pose o ou is ic de elopmen . Indeed, ho els and apa men buildings buil o ou ism ep esen a g ea e impac a he momen o hei cons uc ion in e ms o occupancy and soil con amina ion han, o example, o he s ha a e simply ehabili a ed o his pu pose and ha we e al eady pa o he local scene y. Following his logic, using p i a e houses o amily and iends as ou is ic accommoda ions would be he es ablishmen s wi h he leas amoun o impac on he en i onmen . Fo seasonali y, ollowing [ 30 ] and in o de o measu e i o he Cana y Islands, Figu e 1shows ha seasonali y is no exclusi ely concen a ed in he summe mon hs, bu is in ac dis ibu ed h oughou he yea , wi h Ma ch and Oc obe s anding ou abo e he es . Fu he mo e, we can e i y ha he skewness is nega i e al hough qui e low, app oaching 0 (symme y) o he las wo yea s. Hence, he peaks would be ound in Sus ainabili y 2021,13, 5072 8 o 26 he second hal o he yea . Rega ding ku osis, we ealize ha , o e he yea s, i s alue inc eases, going om being pla yku ic o lep oku ic. The e o e, he peaks would be accen ua ed o e ime, ma king seasonali y. Table 1. Va iables. Va iables Ca ego ies (%) Residen (Cana y Islands) (1) Non- esiden 80.4 (2) Residen 19.6 Season (1) High 64.9 (2) Low 35.1 Accommoda ion (1) Ho el 44.1 (2) Apa men , u al o en ed house 29.6 (3) Own house, amily/ iend’s house 21.6 (4) C uise, camping and o he s 4.7 Accommoda ion Inside (1) Sleeping ou side he pa k 85.7 (2) Sleeping wi hin he pa k 14.3 Na ional Pa k (1) Ga ajonay 5.0 (2) Calde a Tabu ien e 4.4 (3) Teide 84.4 (4) Timan aya 6.2 P ese a ion (1) Low (0–5) 4.4 (2) Medium (6–8) 36.5 (3) High (9–10) 59.1 Sou ce: P epa ed by he au ho s wi h da a om he ques ionnai es. Figu e 1. Seasonali y o he numbe o ou is s’ a i als o he Cana y Islands (2016–2019). Ano he a iable included in he analysis is he Cana y Islands na ional pa ks, namely: Ga ajonay, Calde a Tabu ien e, Teide and Timan aya (see Figu e 2). In Spain, he e a e di e se ypes o na ional pa ks (NPs), he Cana y Island Na ional Pa ks a e some o hese. Howe e , e en among he Cana y Island NPs, you can ind di e ences. Fo example, wi hin he Teide, Ga ajonay and Calde a de Tabu ien e NPs going o walks and hiking in he pa ks a e allowed. By con as , in he Timan aya NP i is o bidden o go hiking in o de o p ese e he la a landscape as a na u al esou ce ha is conside ed an impo an pa o he island’s na u al he i age by he local popula ion. In ela ion o accommoda ion, he e Sus ainabili y 2021,13, 5072 9 o 26 a e some p i a e buildings inside he pa ks, bu ela i ely ew and new cons uc ions ha e been banned. Cu en ly, some beds a e o e ed by he public au ho i ies in shel e s—no p i a ely owned—and in camping a eas inside he pa ks. Ne e heless, he numbe o beds o e ed wi hin he pa ks is limi ed due o he en i onmen al impac hey could ha e. Figu e 2. The Cana y Islands Na ional Pa ks (CNP). Finally, he a iable p ese a ion is in ela ion wi h he e alua ion o na u al esou ces p ese a ion by he isi o s (low, medium and high). 2.4. Sampling Technique Simple andom p obabilis ic sampling was used wi h 1671 su eys ca ied ou and ieldwo k o ganized in wo phases. The i s phase o in e iews was made in he high season on he Islands, including he Eas e pe iod. The second phase in low season co esponded o la e sp ing, summe and beginning o au umn. A eliabili y o ± 2 δ was achie ed, as a unc ion o he sample size, n= 1671. The maximum es ima ion e o o he p opo ions calcula ed on he o al o main a iables, unde he assump ion o dicho omy, appea s in Table 2. Table 2. The maximum es ima ion e o . Va iable Maximum Es ima ion E o Residen ±1.96% Season ±2.29% Accommoda ion ±2.38% Accommoda ion_Inside ±1.68% Na ional Pa k ±1.74% P ese a ion ±2.37% No e: The p opo ion o he mos o en ca ego y is used. Table 3shows he sociodemog aphic p o ile o he pa icipan s ha comple ed he su ey and how hey ela e o he ypical ou is s expec ed in he Cana y Islands. As domes ic esiden s a e ou main esea ch objec i e, isi o s om he Cana y Islands a e included in he hi d ca ego y o ‘place o esidence’ wi h a ele an p opo ion. Sus ainabili y 2021,13, 5072 16 o 26 Pe haps by s aying longe and ou ing he pa ks ha e a close idea o he conse a ion o na u al esou ces wi hin he pa k. Non- esiden s basically s ay in he pa k o only a ew hou s. 3.2.2. C ossing o Con ingency Tables o SEAS_RESID wi h Type o Accommoda ion In sea ching o he ole played by seasonali y in domes ic o non- esiden isi o s o he CNPs, he dependen a iables conside ed in he mul inomial logis ic eg ession model calcula ed ha e been accommoda ion. In e ms o independen a iables, we ha e chosen o c ea e a a iable ha combines seasonali y (high and low) wi h he esiden o non- esiden na u e o a ou is . Mo eo e , a e adding he a iable ‘season’ he independen a iable chosen has been enamed SEAS_RESID wi h he ollowing ca ego ies: non- esiden —high season, non- esiden —low season, esiden —high season and esiden —low season. I s equency dis ibu ion appea s in Table 10. Table 10. F equency dis ibu ion o SEAS_RESID. SEAS_RESID F equency Pe cen 1. Non- esiden —high season 869 52.0 2. Non- esiden —low season 475 28.4 3. Residen —high season 215 12.9 4. Residen —low season 112 6.7 To al 1671 100.0 Sou ce: Da a om ques ionnai es. We ha e me ged some ca ego ies o he a iable ype o accommoda ion o a oid null equencies in he cells o he c ossing o con ingency ables wi h SEAS_RESID. So, he inal s uc u e o he a iable includes he ollowing ou ca ego ies: (1) ho els (all ca ego ies); (2) apa men , u al co age, en al house; (3) own house, amily o iend’s house; (4) c uise, camping and o he s. The esul ing c ossing o con ingency ables o SEAS_RESID wi h ‘ ype o accommoda ion’ is shown in Figu e 7. Figu e 7. Ba cha g ouped o ype o accommoda ion. In Figu e 7we can obse e ha o esiden s ‘own house, amily o iend’s house’ is by a he mos equen ca ego y ela i e o he nex one, ‘apa men , u al, o en ed house’, and ho els a e a a a hi d place. Despi e he weigh o ‘own house, amily o iends’ house’ o his esiden segmen , i should be men ioned ha di e en beha io s Sus ainabili y 2021,13, 5072 17 o 26 can be obse ed in his con ingency cha be ween esiden s who isi he pa ks du ing he high and he low season: he la e use mo e o en ‘own house, amily o iend’s house’, whe eas o he o me , ho els a e he es ablishmen used mo e o en. Rega ding non- esiden s, he e a e no no able di e ences be ween high and low season isi o s, bu he lesse ele ance o he ‘c uise, camping, o he s’ ca ego y should be no ed o non- esiden s coming du ing he low season. Fo he o he ca ego ies, hei use is nea ly he same as ha o non- esiden s. 3.2.3. Resul s om he Mul inomial Logis ic Reg ession Model Applica ion A mul inomial logis ic eg ession model was pe o med o Type o accommoda ion (dependen a iable) on SEAS_RESID (independen a iable coded using dummies). The es included in Table 11 shows ha he independen a iable, SEAS_RESID, is s a is ically signi ican (wi h a alue o Chi-squa e o 1357.587 and a alue p= 0.000 ≤ 0.05). The e o e, he in luence o SEAS_RESID on Type o accommoda ion has been con i med. In e e ence o he quali y o he MLRM, we also ge high alues (0.557 and 0.594) o he pseudo R-Squa ed, so he model seems o be a good i . In his case, he pe cen age o co ec classi ica ions o he Type o accommoda ion o each ca ego y o SEAS_RESID achie es a easonable alue o 57.8%. Table 11. Model Fi ing In o ma ion o Type o accommoda ion. Model Fi ing C i e ia Likelihood Ra io Tes Quali y o MLRM E ec -2Log Likelihood Chi Squa e D Sig. Cox-Snell Nagelke ke % Co ec classi ica ion SEAS_RESID 1415.623 1357.587 12 0.000 0.557 0.594 57.80% To in e p e he eg ession pa ame e s mo e easily, we conside he MLRM wi hou in e cep (see Table 12). All he odds a ios and he es ima ed p obabili ies o each ca ego y o he dependen a iable ‘ ype o accommoda ion’ appea s in Table 13 and Figu e 8. Table 12. Pa ame e es ima es o he MLRM on ype o accommoda ion. Type o Accommoda ion SEAS_R ESID B S d. E o Wald d Sig. Exp B 95% Con idence In e al (Exp B) Ho el 1 1.908 0.132 209.330 1 0.000 6.742 5.206 8.732 2 3.622 0.383 89.464 1 0.000 37.429 17.669 79.286 3 2.140 0.748 8.196 1 0.004 8.500 1.964 36.790 4 1.386 0.645 4.612 1 0.032 4.000 1.129 14.175 Apa men , Ru al, Ren ed House 1 1.473 0.136 116.552 1 0.000 4.364 3.340 5.702 2 3.190 0.386 68.411 1 0.000 24.286 11.404 51.717 3 2.015 0.753 7.164 1 0.007 7.500 1.715 32.796 4 1.946 0.617 9.940 1 0.002 7.000 2.088 23.468 Own House, Family o F iend’s House 1 0.015 0.173 0.008 1 0.931 1.015 0.723 1.426 2 1.638 0.413 15.716 1 0.000 5.143 2.289 11.556 3 4.505 0.711 40.153 1 0.000 90.500 22.462 364.632 4 3.232 0.589 30.150 1 0.000 25.333 7.992 80.304 No e: The e e ence ca ego y o ype o accommoda ion is ‘c uise, camping, o he s. F om he analysis o he p e ious esul s, we can conclude ha : Fo non- esiden isi o s, he e is no seasonali y e ec on he ype o accommoda ion because he p obabili ies (and also he odds a io) a e e y simila o he high and low seasons. Fo hese isi o s, he mos p obable ype o accommoda ion is a ho el (nea ly 50%). This p obabili y is app oxima ely wo imes g ea e han he p obabili y o ‘apa men , u al co age o en ed house’ (see he odds a io o p1/p2 in Table 13). Sus ainabili y 2021,13, 5072 18 o 26 Table 13. Odds a io o he ca ego ies o ype o accommoda ion. SEAS_RESID p1/p4 p2/p4 p3/p4 p1/p2 p1/p3 p2/p3 1. Non- esiden —high season 6.742 4.364 1.015 1.545 6.642 4.299 2. Non- esiden —low season 37.429 24.286 5.143 1.541 7.278 4.722 3. Residen —high season 8.500 7.500 90.500 1.133 0.094 (10.638) 0.083 (12.048) 4. Residen —low season 4.000 7.000 25.333 0.571 (1.751) 0.158 (6.329) 0.276 (3.623) Figu e 8. Es ima ed p obabili ies o ype o accommoda ion. The beha io o esiden isi o s seems o be e y di e en because he mos p ob- able ype o accommoda ion is ‘own house, amily o iend’s house’. In his case, he seasonali y e ec appea s when we compa e, o example, he p obabili y o his ca ego y, p3, wi h, o example, p1 (ho el) and p2 (apa men , u al co age and en ed house). In he low season, he p obabili y o ‘ho el’ and ‘apa men , u al co age and en ed house’ is 29.5% (see he odds a ios o p1/p3 and p2/p3, espec i ely). Howe e , in he high season, he p obabili y o ‘ho el’, and ‘apa men , u al co age and en ed house’ is only 14.9%. The e o e, we can conclude ha he domes ic isi o conside ably inc eases he p obabili y o selec ing ano he ype o accommoda ion such as ‘ho el’ o ‘apa men , u al co age and en ed house’ when changing om he high o low season, ins ead o ‘own o iend’s/ amily’s house’. A mul inomial logis ic eg ession model was also pe o med o p ese a ion as dependen a iable on SEAS_RESID (independen a iable coded using dummies). The es included in Table 14 shows ha he independen a iable, SEAS_RESID, is s a is ically signi ican (wi h a alue o Chi-squa e o 999.980 and a alue p= 0.000 ≤ 0.05). The e o e, he in luence o SEAS_RESID on P ese a ion has been con i med. In e e ence o he quali y o he MLRM, we also ge high alues (0.454 and 0.510) o he pseudo R-Squa ed, hus he model seems o be qui e a good i , as well. In his case, he pe cen age o co ec P ese a ion classi ica ions o each ca ego y o SEAS_RESID eaches a alue o 59.7%. Conside ing he MLRM wi hou in e cep ion (see Table 15) esul s odds a ios and es ima ed p obabili ies o each ca ego y o he dependen a iable “p ese a ion”, as i is shown in Table 16 and Figu e 9. Sus ainabili y 2021,13, 5072 19 o 26 Table 14. Model Fi ing In o ma ion o P ese a ion. Model Fi ing C i e ia Likelihood Ra io Tes Quali y o MLRM E ec -2Log Likelihood Chi-Squa e D Sig. Cox-Snell Nagelke ke % Co ec classi ica ion SEAS_RESID 1042.748 999.980 8 0.000 0.454 0.510 59.7% Table 15. Pa ame e Es ima es o he MLRM on P ese a ion. Type o Accommoda ion SEAS_R ESID B S d. E o Wald d Sig. Exp B 95% Con idence In e al (Exp B) MEDIUM 1 2.383 0.209 130.003 1 0.000 10.840 7.196 16.329 2 2.645 0.287 84.886 1 0.000 14.007 8.020 24.708 3 1.366 0.224 37.172 1 0.000 3.920 2.527 6.081 4 1.629 0.346 22.193 1 0.000 5.100 2.589 10.045 HIGH 1 3.118 0.204 232.742 1 0.000 22.600 15.141 33.735 2 3.045 0.284 115.021 1 0.000 21.000 12.039 36.632 3 1.281 0.226 32.102 1 0.000 3.600 2.311 5.607 4 1.609 0.346 21.586 1 0.000 5.000 2.536 9.859 No e: The e e ence ca ego y o P ese a ion is LOW. Table 16. Odds a io o he ca ego ies o P ese a ion. SEAS_RESID p2/p1 p3/p1 p3/p2 1. Non- esiden —high season 10.840 22.600 2.084 2. Non- esiden —low season 14.007 21.000 1.499 3. Residen —high season 3.920 3.600 0.918 4. Residen —low season 5.100 5.000 0.980 Figu e 9. Es ima ed p obabili ies o p ese a ion. The column p3 (high)/p2 (medium) e lec s ha : Fo esiden s, he e is no e ec o seasonali y on he highe p ese a ion sco es, as he p obabili y ha hey will alue p ese a ion wi h a medium and high sco e in he high and low seasons is qui e simila (odd- a io close o 1). In he case o non- esiden s, di e ences do exis . To begin wi h, he p obabili y ha hey alue conse a ion wi h a high a ing is 50% supe io o he medium a ing in he low season and 100% in he high season. The e o e, he season does ha e an impo an in luence. Sus ainabili y 2021,13, 5072 20 o 26 The e o e, non- esiden s who isi he des ina ion in high season a e hose who gi e he bes a ing o he s a e o p ese a ion o he pa ks’ na u al esou ces. Residen s who isi he pa ks in high season, on he o he hand, a e he ones who gi e he lowes alue o he s a e o conse a ion o he pa ks. This shows a e y di e en pe cep ion o he s a e o conse a ion o na u al esou ces acco ding o he geog aphical o igin o he isi o s. 4. Discussion and Conclusions The esul s achie ed d aw a en ion o he exis ence o a po en ial con lic o objec i es. The a ailabili y o a ange o quali y na u al esou ces de e mines he a ac ion o an a ea in he con ex o ou is ma ke s. Howe e , he de elopmen o he ou ism sec o and he cons uc ion o buildings ha i en ails, ha e an ambi alen impac . On he one hand, ou ism c ea es employmen and gene a es economic income, imp o ing he quali y o li e o he local popula ion. On he o he hand, he de elopmen o his sec o , when i eaches le els o mass ou ism de elopmen , esul s in a de e io a ion o na u al esou ces leading ma ke s o an ad e se selec ion. The in luence o seasonali y o domes ic o non- esiden isi o s on a iables such as “accommoda ion” and “p ese a ion” has sca cely been ackled om an empi ical pe spec i e. The e o e, he ocus on he accommoda ion o p ese a ion as he dependen a iable wi h he wo o he s (season and place o esidence) as explica i e ones is an impo an con ibu ion o his analysis. In his sense, looking o he bes me hodological app oach o o e come he abo e di icul ies, he mul inomial logis ic eg ession model (MLRM) may o e an e ec i e solu ion. The main con ibu ion o his pape is de e mining di e ences in beha io be ween domes ic (Cana y Island) isi o s and non- esiden isi o s. Fo example, he esul s con i m ha esiden s who isi he pa ks in high season end o gi e he lowes alues o he s a e o he pa ks’ conse a ion. By con as , non- esiden s who isi he pa ks in high season gi e he highes alues o hei conse a ion. This shows he e y di e en pe cep ions o he s a e o conse a ion o na u al esou ces acco ding o he geog aphical o igin o he isi o s. This esul would no be possible in an analysis wi hou segmen ing he ou ism ma ke by he domes ic/non- esiden a iable. Likewise, he success o he analysis o segmen using he domes ic/non- esiden a iable in a des ina ion wi h a low seasonali y is ano he impo an con ibu ion o his pape . In ac , no p e ious ele an pape s ha e been ound wi h hese esul s. The esul ing i ed model is in e p e ed wi hin he con ex o a speci ic applied p oblem, analyzing he seasonali y o domes ic o non- esiden isi o s, conside ing he ole played by “accommoda ion” chosen by isi o s o he CNP. Indeed, he impo an ole ound o accommoda ion by season in he esul s has been use ul o de e mining he cha ac e is ics ha seasonali y adds o his speci ic a iable. This is o a high impo - ance due o he di icul y o inding signi ican da a o low season in he majo i y o ou ism des ina ions. Based on he esul s ob ained, he main conclusion is as ollows: The i s hypo hesis is ul illed. The e o e, i is possible o pe o m a quan i a i e analysis o he seasonali y o he accommoda ion used by ou is s, pa icula ly when di - e ences in domes ic and non- esiden isi o s’ beha io is he main esea ch opic. The second hypo hesis is con i med as well, as seasonali y has a s ong in luence on accommo- da ion ype o he case o domes ic ou ism. Mo eo e , domes ic isi o s ha e di e en pe cep ions o he na u al esou ce conse a ion s a us in na ional pa ks. The in luence o he a iables season/ esiden –non- esiden on conse a ion has been con i med and esiden s who isi he pa ks in high season a e he ones who gi e he lowes alue o he pa ks’ conse a ion s a e. The e o e, he hi d hypo hesis is also ul illed. Finally, he ou h hypo hesis is es ed o hose empi ical s udies in which no con inuous a iables a e conside ed, choosing he accu a e dependen (accommoda ion o p ese a ion alue) and independen a iables (domes ic/non- esiden o season) will be e ec i e using a MLRM. Sus ainabili y 2021,13, 5072 21 o 26 This analysis can lead o he de elopmen o app op ia e ac ions aimed a achie ing en i onmen al objec i es, h ough he con ol o ype o accommoda ion dis ibu ion h oughou he yea in a des ina ion. In ac , he esul s highligh he impo an in luence o season o domes ic isi o s o he Cana y Island na ional pa ks when choosing ype o accommoda ion. In sum, and o his pa icula segmen , ‘own house, amily o iend’s house’ a e used in he high season and ‘ho el’ o ‘apa men , u al co age and en ed house’ in he low one. Howe e , and by con as , o he non- esiden segmen , no signi ican seasonali y e ec on choosing an accommoda ion was obse ed, o hese isi o s, he mos p obable ype o accommoda ion is a ho el du ing all he yea . As a ele an esul , oo, i is impo an o conside he di e en alue esiden s assess o he p ese a ion le el o na u al esou ces, lowe in high season. O special signi icance is he case o isi o s ha s ay inside he pa ks, hose ha e e y deep in o ma ion on he le el o p ese a ion o he na u al esou ces, and i s alua ion is lowe han isi o s sleeping ou side he pa ks. Policy Recommenda ions In ligh o hese esul s, i should be eminded ha he main objec i e o his esea ch is o unde s and he na ional pa k isi o s’ use o ou is ic accommoda ion acco ding o hei esiden o non- esiden s a us and i s alue o he na u al esou ces p ese a ion. Ul ima ely, he aim is o de elop di e en ia ed s a egies o each g oup depending on hei g ea e o lesse impac on he en i onmen whe e hey s ay and hei pe cep ions. In he analysis we ha e in oduced he ole o seasonali y based on he idea ha he di e en segmen s’ beha io would di e acco ding o each season, and he esul s ha e con i med his heo y. The applica ion o he logi model has clea ly shown ha non- esiden s ollow a simila pa e n du ing he high and low season, s aying mos ly in ho els. In con as , esiden s beha e di e en ly du ing each season. Fu he mo e, seasonali y can indeed be seen as a condi ioning ac o o domes ic isi o s o he island when choosing an accommoda ion es ablishmen . Thus, when de el- oping a se o en i onmen al policies o ou ism whe e he isi o s o his ou is ic na u al esou ce a e conside ed ( e e ing o na ional pa ks), policymake s will ha e o elabo a e a single s a egy o be applied h oughou he yea o non- esiden s on one hand, and on he o he , di e en measu es o domes ic isi o s depending on he season. S a ing om a gene al p inciple o non-applica ion o small islands, such as simila use o na u al ou ism policies as in con inen al e i o ies, he analysis allows us o e lec on he mos ele an en i onmen al ou ism policy ecommenda ions o na ional pa ks loca ed in small islands and hei insula en i onmen in gene al. Such ecommenda ions could include: Domes ic isi o s who isi he na ional pa ks in high season use hei own/ amily/ iends’ houses, which a e conside ed non-pollu ing accommoda ion. So, p omo ional plans o domes ic ou ism o na ional pa ks and he islands in high season should p omo e s aying mos ly in es ablishmen s ha a e no newly buil o in ones wi h he leas impac on wa e , ai and land pollu ion le els. This would be he mos ecommended segmen om an en i onmen al s andpoin . Howe e , domes ic isi o s ha come o he na ional pa ks in low season should be included in p omo ional plans oo, as hey con ibu e o balancing he numbe o isi s h oughou he yea . They s ay mo e equen ly in apa men s, u al o en ed houses. Some o hese buildings may ha e en i onmen al impac s as hey ha e in ol ed new cons uc ion while some o he s ha e no . The la e is he case o u al co ages, which a e usually eno a ed buildings. The e o e, o hese ypes o isi o s, he p omo ion s a egy should be o encou age hem o choose less o non-pollu ing accommoda ion. Addi ionally, no only p omo ional ac ions, bu o he kinds o measu es could be implemen ed. Fo example, ax incen i es and subsidies o en ep eneu s ha decide o eno a e u al houses and o e hem as ou is accommoda ion could be p o ided. Fo bo h domes ic and non- esiden isi o segmen s, egula ions ha o bid he cons uc ion Sus ainabili y 2021,13, 5072 22 o 26 o new ou is accommoda ions wi h a high land consump ion could be implemen ed. This is pa icula ly impo an in small islands whe e he land is a limi ed esou ce. Mo eo e , he e could be ini ia i es o he use o camping and shel e s in low season. In high season, gi en he abundan numbe o isi o s s aying wi hin he pa k, his measu e would no be possible due o i s po en ial en i onmen al impac . Thus, hese ini ia i es could be managed in he low season by o e ing new places in camping o shel e s in he a eas su ounding he na ional pa ks o by gi ing incen i es o special se ices. The incen i es could be ela ed o ee o discoun ed bed o e s, o example. The new se ices could include, among o he s, ee secu i y se ices in he camps o shel e s. Fo bo h domes ic and non- esiden segmen s and speci ically o high season isi o s, a ine could be imposed o beha io s ha gene a e nega i e en i onmen al impac s. Examples o hese kinds o undesi able beha io s could be s epping on he la a in he Timan aya o El Teide NPs o camping wi hou he equi ed au ho iza ion in La Calde a Na ional Pa k. The esul s show ha domes ic isi o s s aying wi hin he pa ks gi e he lowes alue o he conse a ion s a us o na u al esou ces. Pe haps by s aying longe and ou ing he pa ks, isi o s ha e a be e idea o he conse a ion o he na u al esou ces wi hin he pa ks. This is ue o he ex en ha hose who s ay in he pa k will ha e a be e knowledge o he s a e o conse a ion o he na u al esou ces since hey ou he pa ks. The e o e, i would be impo an o ca y ou pe iodic su eys o his second g oup on hei assessmen o he pa k. The su ey would also allow us o know hei demands o se ices om he pa ks’ managemen bodies. Mo eo e , i would be wo hwhile implemen ing some gene al ecommenda ions based on he idea o sus ainable buildings, as es ablished in he li e a u e e iew. De el- opmen o plans aimed a ehabili a ion a he han he new cons uc ion o ho el and non-ho el es ablishmen s in he Islands. Encou agemen o he e olu ion o he cons uc ion sec o on he Islands owa ds ‘g een building’ models, based on he use o non-pollu ing ma e ials and he use o ene gy e icien p oduc ion p ocesses. P omo ion o en i onmen al awa eness, il ing he demand o ou ism accommoda ion owa ds hose buildings ha ha e been buil wi h clean echnologies. Indeed, ou ism is no as ce ain beha io s seem o suppose, an economic ac i i y wi h- ou main enance cos s. In addi ion o he in as uc u es, main enance and condi ioning se ices a e needed ha also p e en de e io a ion o he na u al en i onmen . Fo he u u e, deepe and mo e sus ained esea ch in o he accommoda ion sec o and i s con ibu ions o ou ism mo e b oadly need o be in es iga ed, as does he ole accommoda ion does, o mo e likely can, con ibu e o he wide sus ainabili y o des- ina ions. No longe can accommoda ion be managed, o esea ched, in isola ion, wi h i ep esen ing a co e and in eg al componen o he wide ou ism sys em. By way o example: local lodging es ablishmen s such as Ai BnB ha e been banned by he Cana y Islands go e nmen du ing mos o he pe iod when we we e doing he esea ch. Howe e , u al houses a e legal o ms o he accommoda ion es ablishmen , and hey ha e had easonable occupa ion a es du ing he COVD19 c isis. The e o e, i would be bene icial o u u e esea ch o p o ide some de ails on he complemen a i y o local accommoda ion acili ies wi h local li e: he le el a which hey a e sha ed wi h he local popula ion, which is ob iously essen ial o hei en i onmen al impac and he sus ainabili y o ou ism as a whole. In his way, he e a e mo e han jus economic easons in ol ed in choosing be ween high o low season, he e a e also social ones. As an example, in high season, domes ic isi o s use hei own/ amily/ iends’ houses. This is mainly because in high season, such as win e and Eas e , social habi s among he local popula ion gi e ise o isi s o hei amily and iends o aca ions. Sus ainabili y 2021,13, 5072 23 o 26 Au ho Con ibu ions: Concep ualiza ion, F.M.D.-P. and A.F. ( om Janua y 2019); me hodology, F.M.D.-P. and C.G.G.-G.; so wa e, C.G.G.-G.; alida ion, C.G.G.-G.; o mal analysis, C.G.G.-G.; in es iga ion, F.M.D.-P.; esou ces, F.M.D.-P. and A.F.; w i ing—o iginal d a p epa a ion, F.M.D.-P.; w i ing— e iew and edi ing, A.F. and F.M.D.-P.; isualiza ion, C.G.G.-G.; p ojec adminis a ion, F.M.D.-P.; unding acquisi ion, F.M.D.-P. All au ho s ha e ead and ag eed o he published e sion o he manusc ip . Funding: This esea ch has been suppo ed by unding om he Go e nmen o he Cana y Islands, (Spain) and he Regional De elopmen Eu opean Fund (EU), Re . P oID201/010125. The E hics Commi ee ha app o ed his esea ch was he Agencia Cana ia de In es igación, Inno ación y Sociedad de la In o mación, ACIISI, Go e nmen o he Cana y Islands, Spain. Ins i u ional Re iew Boa d S a emen : No applicable. In o med Consen S a emen : No applicable. Da a A ailabili y S a emen : Supplemen a y con en ela ed o his a icle has been published online a h ps:// igsha e.com/s/0474ce9871103709dda0 (accessed on 29 Ap il 2021). Con lic s o In e es : The au ho s decla e no con lic o in e es . The unde s had no ole in he design o he s udy; in he collec ion, analyses, o in e p e a ion o da a; in he w i ing o he manusc ip , o in he decision o publish he esul s. Re e ences 1. Hadwen, W.L.; A hing on, A.H.; Boon, P.I.; Taylo , B.; Fellows, C.S. Do Clima ic o Ins i u ional Fac o s D i e Seasonal Pa e ns o Tou ism Visi a ion o P o ec ed A eas ac oss Di e se Clima e Zones in Eas e n Aus alia? Tou . Geog . 2011 ,13, 187–208. [C ossRe ] 2. Ramdas, M.; Mohamed, B. Visi o Pe cep ions on he Impac s o Tou ism Ac i i ies, De elopmen and In as uc u e on he En i onmen o Pe hen ian Islands. SHS Web Con . 2014,12, 01081. [C ossRe ] 3. Zubai , S.; Bowen, D.; Elwin, J. No qui e pa adise: Inadequacies o en i onmen al impac assessmen in he Maldi es. Tou . Manag. 2011,32, 225–234. [C ossRe ] 4. Deng, J.; King, B.; Baue , T. E alua ing na u al a ac ions o ou ism. Ann. Tou . Res. 2002,29, 422–438. [C ossRe ] 5. Reinius, S.W.; F edman, P. P o ec ed a eas as a ac ions. Ann. Tou . Res. 2007,34, 839–854. [C ossRe ] 6. Ha man, D. In angibles alues o p o ec ed a eas: Wha a e hey? Why do hey ma e ? Geo g. W igh Fo um 2004,21, 9–22. 7. K uge , M.; Van De Me we, P.; Saayman, M.; Slabbe , E. Unde s anding accommoda ion p e e ences o isi o s o he K uge Na ional Pa k. Tou . Hosp. Res. 2017,19, 170–185. [C ossRe ] 8. Ins i u o Nacional de Es adís ica (INE). A ailable online: www.ine.es (accessed on 11 Janua y 2019). 9. Rabbany, G.; A in, S.; Rahman, A.; Islam, F.; Hoque, F. En i onmen al E ec s o Tou ism. Am. J. En i on. Ene gy Powe Res. 2013 , 1, 117–130. 10. Ruiz, L. Ad e se En i onmen al E ec s o Tou ism on Communi ies in he Ecuado ian Highlands Region. Cu . U ban S ud. 2017 , 5, 348–355. [C ossRe ] 11. Dickinson, J.E.; Filimonau, V.; Hibbe , J.F.; Che e , T.; Da ies, N.A.J.; No ga e, S.; Speed, C.; Wins anley, C. Tou ism communi ies and social ies: The ole o online and o line ou is social ne wo ks in building social capi al and sus ainable p ac ice. J. Sus ain. Tou . 2016,25, 163–180. [C ossRe ] 12. Wang, Y.; Sun, M.; Wang, R.; Lou, F. P omo ing egional sus ainabili y by eco-p o ince cons uc ion in China: A c i ical assessmen . Ecol. Indic. 2015,51, 127–138. [C ossRe ] 13. Hassan, S.S. De e minan s o Ma ke Compe i i eness in an En i onmen ally Sus ainable Tou ism Indus y. J. T a el Res. 2000 ,38, 239–245. [C ossRe ] 14. P akash, S.L.; Pe e a, P.; Newsome, D.; Kusuminda, T.; Walke , O. Reasons o isi o dissa is ac ion wi h wildli e ou ism expe iences a highly isi ed na ional pa ks in S i Lanka. J. Ou doo Rec ea . Tou . 2019,25, 102–112. [C ossRe ] 15. Amoamo, M. The Mi iga ion o Vulne abili y: Mu iny, Resilience and Recons i u ion. A Case S udy o Pi cai n Island. Shima In . J. Res. Isl. Cul . 2011,5, 69–93. 16. Dang, T.H.G.; Low, S.P. Role o cons uc ion in economic de elopmen : Re iew o key concep s in he pas 40 yea s. Habi a In . 2011,35, 125–128. 17. Lewis, T.M. The cons uc ion indus y in he economy o T inidad & Tobago. Cons . Manag. Econ. 2004 ,22, 541–549. [C ossRe ] 18. Su, C.K.; Lin, C.Y.; Wang, M.T. Taiwanese cons uc ion sec o in a g owing ‘ma u i y’ economy, 1964–1999. Cons . Manag. Econ. 2003,21, 719–728. [C ossRe ] 19. Yo ucu, V.; Keles, R. The cons uc ion boom and en i onmen al p o ec ion in No he n Cyp us as a consequence o Annan plan. Cons . Manag. Econ. 2007,25, 77–86. [C ossRe ] 20. Abidin, N.Z. In es iga ing he awa eness and applica ion o sus ainable cons uc ion concep by Malaysian de elope s. Habi a In . 2010,34, 421–426. [C ossRe ] Sus ainabili y 2021,13, 5072 24 o 26 21. Balaban, O. Planlama e Mima lık Ekseninde TOKÐ Uygulamala ı. TMMOB Ken le¸sme e Ye el Yöne imle Sempozyumu Bildi ile Ki abı; TMMOB Yayınla ı: Anka a, Tu key, 2004. 22. A eline, N.; Li, L.H. Reconside ing land policies in Asia. In P ope y Ma ke s and Land Policies in No heas Asia, he Case o Fi e Ci ies: Tokyo, Seoul, Shanghai, Taipei and Hong Kong; A eline, N., Li, L.H., Eds.; Maison F anco-Japonaise & Cen e o Real Es a e and U ban Economics HKU: Hong Kong, China, 2004; pp. 298–314. 23. Feagin, J.R. The seconda y ci cui o capi al: O ice cons uc ion in Hous on, Texas. In . J. U ban Reg. Res. 1987 ,11, 172–192. [C ossRe ] 24. Ko kmaz, S.; E en, D.; Syal, M.; Po bha e, V. A e iew o g een building mo emen imelines in de eloped and de eloping coun ies o build an in e na ional adop ion amewo k. In P oceedings o he Fi h In e na ional Con e ence on Cons uc ion in he 21s Cen u y: Collabo a ion and In eg a ion in Enginee ing, Managemen and Technology, Is anbul, Tu key, 20–22 May 2009; pp. 1749–1757. 25. Robe s, S. E ec s o clima e change on he buil en i onmen . Ene gy Policy 2008,36, 4552–4557. [C ossRe ] 26. Tan, Y.; Shen, L.; Yao, H. Sus ainable cons uc ion p ac ice and con ac o s’ compe i i eness: A p elimina y s udy. Habi a In . 2011,35, 225–230. [C ossRe ] 27. Wong, J.M.W.; Thomas Ng, S.; Chan, A.P.C. S a egic planning o he sus ainable de elopmen o he cons uc ion indus y in Hong Kong. Habi a In . 2010,34, 256–263. [C ossRe ] 28. Balaban, O. The nega i e e ec s o cons uc ion boom on u ban planning and en i onmen in Tu key: Un a eling he ole o he public sec o . Habi a In . 2012,36, 26–35. [C ossRe ] 29. Bu le , R.W. Seasonali y in ou ism. Issues and Implica ions. In Seasonali y in Tou ism; Baum, T., Lund o p, S., Eds.; Taylo & F ancis: New Yo k, NY, USA, 2001; Chap e 2; pp. 5–22. 30. Koenig-Lewis, N.; Bischo , E.E. Seasonali y esea ch: The s a e o he a . In . J. Tou . Res. 2005,7, 201–219. [C ossRe ] 31. Capó, J.; Rie a, A.; Rosselló, J. Accommoda ion de e minan s o seasonal pa e ns. Ann. Tou . Res. 2007,34, 422–436. [C ossRe ] 32. Lim, C.; McAlee , M. Mon hly seasonal a ia ions: Asian ou ism o Aus alia. Ann. Tou . Res. 2001,28, 68–82. [C ossRe ] 33. Becken, S. Measu ing he E ec o Wea he on Tou ism: A Des ina ion- and Ac i i y Based Analysis. J. T a el Res. 2013 ,52, 156–167. [C ossRe ] 34. Ridde s aa , J.; Odube , M.; C oes, R.; Nijkamp, P. Impac s o seasonal pa e ns o clima e on ecu en luc ua ions in ou ism demand: E idence om A uba. Tou . Manag. 2014,41, 245–256. [C ossRe ] 35. Ru in, J. Coas al Tou ism: A Compa a i e S udy be ween C oa ia and Tunisia. Tou . Geog . 2010,12, 264–277. [C ossRe ] 36. Pegg, S.; Pa e son, I.; Ga iddo, P.V. The impac o seasonali y on ou ism and hospi ali y ope a ions in he alpine egion o New Sou h Wales, Aus alia. In . J. Hosp. Manag. 2012,31, 659–666. [C ossRe ] 37. Ka yopouli, S.; Kou a, C. Cyp us as a Win e Des ina ion: An Explo a o y S udy. Tou . Anal. 2012,17, 495–508. [C ossRe ] 38. Kou a, C.; Ka yopouli, S. Cyp us’ image a sun and sea des ina ion as a de imen al ac o o seasonal luc ua ions. Explo a ion in o mo i a ional ac o s o holidaying in Cyp us. J. T a el Tou . Ma k. 2013,30, 700–714. [C ossRe ] 39. Hinch, T.; Hickey, G. Tou ism a ac ions and seasonali y: Spa ial ela ionships in Albe a. In P oceedings o he T a el and Tou ism Resea ch Associa ion; Uni e si y o Mani oba: Winnipeg, MB, Canada, 1997; pp. 69–76. 40. Spence , D.M.; Holecek, D.F. Basic cha ac e is ics o he all ou ism ma ke . Tou . Manag. 2007,28, 491–504. [C ossRe ] 41. Fe nández, A.; Mayo ga, M.C. Seasonal concen a ion o he ho el demand in Cos a del Sol: A decomposi ion by na ionali ies. Tou . Manag. 2008,29, 940–949. [C ossRe ] 42. Volo, S. Resea ch no e: Seasonali y in Sicilian ou ism demand—An explo a o y s udy. Tou . Econ. 2010 ,16, 1073–1080. [C ossRe ] 43. And az, J.M.; Rod igues, P.M. Moni o ing ou ism lows and des ina ion managemen : Empi ical e idence o Po ugal. Tou . Manag. 2016,56, 1–7. [C ossRe ] 44. Mason, P.; Cheyne, J. Residen ’s a i udes o p oposed ou ism de elopmen . Ann. Tou . Res. 2000,27, 391–411. [C ossRe ] 45. San os, G.E.D.O.; Ramos, V.; Rey-Maquiei a, J. De e minan s o Mul i-Des ina ion Tou ism T ips in B azil. Tou . Econ. 2012 ,18, 1331–1349. [C ossRe ] 46. San os, G.E.D.O.; Ramos, V.; Rey-Maquiei a, J. Leng h o S ay a Mul iple Des ina ions o Tou ism T ips in B azil. J. T a el Res. 2014,54, 788–800. [C ossRe ] 47. Sai o, H.; Romão, J. Seasonali y and egional p oduc i i y in he Spanish accommoda ion sec o . Tou . Manag. 2018,69, 180–188. [C ossRe ] 48. Robe s, S.; Lewis-Came on, A. Small Island De eloping S a es Issues and P ospec s. In Ma ke ing Island Des ina ions; Else ie : Ox o d, UK, 2010; pp. 1–10. 49. Mu ay, I. Indica o s o Socio-Ecological Sus ainabili y o he Balea ic Islands (2003–2008). G upd’ In es igacióSos enibili a i Te i o i, Uni e si a de les Illes Balea s. 2010. A ailable online: h p://www.uib.es/os /es udi/ (accessed on 21 Feb ua y 2019). 50. Gu ié ez, L.R. Impac Assessmen o Tou ism Cons uc ion in Cuba. J. Build. Cons . Plan. Res. 2015,3, 10–17. [C ossRe ] 51. Bu le , R.W. Ca ying capaci y in ou ism: Pa adox and hypoc isy. Tou . Res. A 2010,20, 53–64. 52. Coccossis, H. Island ou ism de elopmen and ca ying capaci y. In Island Tou ism and Sus ainable De elopmen : Ca ibbean, Paci ic and Medi e anean Expe iences; Apos olopoulos, Y., Gayle, D.J., Eds.; P aege : Wes po , CT, USA, 2002; Chap e 8; pp. 131–143. 53. Coccossis, H.; Mexa, A. The Challenge o Tou ism Ca ying Capaci y Assessmen : Theo y and P ac ice; Rou ledge: London, UK, 2017. 54. Dodds, R.; Bu le , R. The phenomena o o e ou ism: A e iew. In . J. Tou . Ci ies 2019,5, 519–528. [C ossRe ] 55. Wall, G. F om ca ying capaci y o o e ou ism: A pe spec i e a icle. Tou . Re . 2020,75, 212–215. [C ossRe ] Sus ainabili y 2021,13, 5072 25 o 26 56. Ga ay, L.; Fon , X.; Co ons, A. Sus ainabili y-O ien ed Inno a ion in Tou ism: An Analysis Based on he Decomposed Theo y o Planned Beha io . J. T a el Res. 2018,58, 622–636. [C ossRe ] 57. Wang, L.; Wong, P.P.W. Ma ke ing o en i onmen ally iendly ho els in China h ough eligious segmen a ion: A heo y o planned beha iou app oach. Tou . Re . 2020. [C ossRe ] 58. Ajzen, I. The heo y o planned beha io . O gan. Beha . Hum. Decis. P ocesses 1991,50, 179–211. [C ossRe ] 59. Han, T.-I.; S oel, L. Explaining Socially Responsible Consume Beha io : A Me a-Analy ic Re iew o Theo y o Planned Beha io . J. In . Consum. Ma k. 2016,29, 91–103. [C ossRe ] 60. Si, H.; Shi, J.-G.; Tang, D.; Wu, G.; Lan, J. Unde s anding in en ion and beha io owa d sus ainable usage o bike sha ing by ex ending he heo y o planned beha io . Resou . Conse . Recycl. 2020,152, 104513. [C ossRe ] 61. Ve ma, V.K.; Chand a, B. An applica ion o heo y o planned beha io o p edic young Indian consume s’ g een ho el isi in en ion. J. Clean. P od. 2018,172, 1152–1162. [C ossRe ] 62. Lopes, J.R.N.; de A aújo Kalid, R.; Rod íguez, J.L.M.; Á ila Filho, S. A new model o assessing indus ial wo ke beha io ega ding ene gy sa ing conside ing he heo y o planned beha io , no m ac i a ion model and human eliabili y. Resou . Conse . Recycl. 2019,145, 268–278. [C ossRe ] 63. Cai, S.; Long, X.; Li, L.; Liang, H.; Wang, Q.; Ding, X. De e minan s o in en ion and beha io o low ca bon commu ing h ough bicycle-sha ing in China. J. Clean. P od. 2019,212, 602–609. [C ossRe ] 64. Daellenbach, K.; Pa kinson, J.; K isjanous, J. Jus How P epa ed A e You? An Applica ion o Ma ke ing Segmen a ion and Theo y o Planned Beha io o Disas e P epa a ion. J. Nonp o i Public Sec . Ma k. 2017,30, 1–31. [C ossRe ] 65. Die ich, T.; Rundle-Thiele, S.; Kubacki, K. Segmen a ion in Social Ma ke ing; Sp inge : Singapo e, 2017; Volume 1. 66. Smi h, W.R. P oduc di e en ia ion and ma ke segmen a ion as al e na i e ma ke ing s a egies. J. Ma k. 1956 ,21, 3–8. [C ossRe ] 67. Rundle-Thiele, S.; Kubacki, K.; Tkaczynski, A.; Pa kinson, J. Using wo-s ep clus e analysis o iden i y homogeneous physical ac i i y g oups. Ma k. In ell. Plan. 2015,33, 522–537. [C ossRe ] 68. Dolnica , S. Ma ke segmen a ion in ou ism. In Tou ism Managemen , Analysis, Beha iou and S a egy; CABI: Camb idge, MA, USA, 2008; pp. 129–150. 69. Wedel, M.; Kamaku a, W. Ma ke segmen a ion: Concep ual and me hodological ounda ions. J. Classi . 2000,17, 143–145. 70. De esa, M.; Laguna, M.; Palacios, A. The ole o mo i a ion in isi o sa is ac ion: Empi ical e idence in u al ou ism. Tou . Manag. 2010,31, 547–552. [C ossRe ] 71. Suha an o, D.; B ien, A.; Suma jan, N.; Wibisono, N. Examining a ac ion loyal y o ma ion in c ea i e ou ism. In . J. Qual. Se . Sci. 2018,10, 163–175. [C ossRe ] 72. Mohammad, B.A.; Som, A.P. An analysis o push and pull a el mo i a ions o o eign ou is s o Jo dan. In . J. Bus. Manag. 2010,5, 41–50. 73. Dean, D.; Suha an o, D. The o ma ion o isi o beha io al in en ion o c ea i e ou ism: The ole o push-pull mo i a ion. Asia Pac. J. Tou . Res. 2019,24, 393–403. [C ossRe ] 74. Chan, C.S.; Yuen, S.K.; Duan, X.; Ma a a, L.M. An analysis o push-pull mo i a ions o isi o s o Coun y Pa ks in Hong Kong. Wo ld Leis. J. 2018,60, 191–208. [C ossRe ] 75. Said, J.; Ma yono, M. Mo i a ion and Pe cep ion o Tou is s as Push and Pull Fac o s o Visi Na ional Pa k. In P oceedings o he E3S Web o Con e ences; Online Mee ing. 2018; Volume 31, p. 08022. 76. Dann, G.M. Anomie, ego-enhancemen and ou ism. Ann. Tou . Res. 1977,4, 184–194. [C ossRe ] 77. P ebesen, N.K.; Woo, E.W.; Chen, J.S.; Uysal, M. Mo i a ion and in ol emen as an eceden s o he pe cei ed alue o he des ina ion expe ience. J. T a el Res. 2013,52, 253–264. [C ossRe ] 78. Yoon, Y.; Uysal, M. An examina ion o he e ec s o mo i a ion and sa is ac ion on des ina ion loyal y: A s uc u al model. Tou . Manag. 2005,26, 45–56. [C ossRe ] 79. Suni, J.; Pesonen, J. Hun e s as ou is s—An explo a o y s udy o push–pull mo i a ions. Scand. J. Hosp. Tou . 2019 ,19, 175–191. [C ossRe ] 80. Chen, Y.-F.; Mo, H.-E. A Su ey o Push and Pull Mo i a ions o G een E en Tou is s. In . J. In . Educ. Technol. 2014,4, 260–263. [C ossRe ] 81. Konijnendijk, C.C.; Anne s ed , M.; Ma u ha ee an, S.; Nielsen, A.B. Bene i s o U ban Pa ks: A Sys em-A ic Re iew (Repo o he In e na ional Fede a ion o Pa ks and Rec ea ion Adminis a ion). 2013. A ailable online: h p://www.i p a.o g/images/ pa k-bene i s.pd (accessed on 30 Sep embe 2018). 82. B ida, J.G.; Os i, L.; Ba que , A. Segmen ing esiden pe cep ions owa ds ou ism—A clus e analysis wi h a mul inomial logi model o a moun ain communi y. In . J. Tou . Res. 2010,12, 591–602. [C ossRe ] 83. Can, V. Es ima ion o a el mode choice o domes ic ou is s o Nha T ang using he mul inomial p obi model. T ansp. Res. Pa A Policy P ac . 2013,49, 149–159. [C ossRe ] 84. Eila , Y.; Eina , L. De e minan s o in e na ional ou ism: A h ee-dimensional panel da a analysis. Appl. Econ. 2004 ,36, 1315–1327. [C ossRe ] 85. Eymann, A.; Ronning, G. Mic oeconome ic models o ou is s’ des ina ion choice. Reg. Sci. U ban Econ. 1997 ,27, 735–761. [C ossRe ] 86. G igolon, A.B.; Bo ge s, A.W.; Kempe man, A.D.; Timme mans, H.J. Vaca ion leng h choice: A dynamic mixed mul inomial logi model. Tou . Manag. 2014,41, 158–167. [C ossRe ]