New performance indicators for water management in tourism

Tourism Management 46 (2015) 233e244

Contents lists available at ScienceDirect

Tourism Management journal homepage: www.elsevier.com/locate/tourman

New performance indicators for water management in tourism € ssling a, b, *, 1 Stefan Go a b

School of Business and Economics, Linnaeus University, 391 82 Kalmar, Sweden Department of Service Management and Service Studies, Box 882, 25108 Helsingborg, Sweden

h i g h l i g h t s
Develops and updates data on direct and indirect water use in tourism.
Provides first assessment of water embodied in foodstuffs consumed in hotels.
Identifies water use ‘lock-in’ as a significant problem for water management.
Calls for the development of a new set of indicators for water management.

a r t i c l e i n f o

a b s t r a c t

Article history: Received 25 November 2013 Accepted 26 June 2014 Available online

Tourism is increasingly recognized as a significant water-consuming sector on local, regional and global scales. As a consequence, the efficient use of water resources is now considered a key sustainability challenge for the tourism industry. To date, most research has focused on direct (on site) water consumption, with tourism water management based almost exclusively on direct water use benchmarks. This paper argues that such an approach overlooks the complexity of ‘local’ and ‘global’ water use, with local water use affecting sustainable water use in the destination and global water use representing the sustainability of water embodied in goods produced elsewhere, including fuels and food. Focussing on tourism accommodation as the locus of tourism water consumption, conventional water indicators are reviewed and discussed, and knowledge gaps identified. New data accounting for food consumption are then presented for a case study of resort hotels in Rhodes, Greece. The results are used to develop a novel set of performance indicators suitable for resort hotels and other accommodation, considering water -vis indirect water consumpavailability, planning and operation, as well as complexities of direct vis-a tion. The findings suggest a significant potential for water and related cost savings, indicating that holistic water management should be an operational imperative. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Embodied water Food Indicators Virtual water Water footprints Water lock-in Water management

1. Introduction Fresh water is an essential resource for tourism. It is consumed directly by tourists for hygienic purposes such as showering or flushing toilets; it is used for the irrigation of gardens and to fill up swimming pools; to provide opportunities for a wide range of leisure activities, such as golf; and it is needed for cleaning rooms and for washing bed and table linen. Water is often also part of the €rko €nen, 2006). landscapes that are attractive to tourists (Hall & Ha Indirectly, tourists consume water embodied in infrastructure (accommodation, roads, airports, etc.), food, fuel, consumption goods,

* School of Business and Economics, Linnaeus University, 391 82 Kalmar, Sweden. Tel.: þ46 70 4922634. E-mail address: [email protected]. 1 www.lnu.se, www.ism.lu.se. http://dx.doi.org/10.1016/j.tourman.2014.06.018 0261-5177/© 2014 Elsevier Ltd. All rights reserved.

and other services (Chapagain & Hoekstra, 2008; Cazcarro, nchez Cho  liz, 2014; Go €ssling, 2002; Pigram, 1995; Hoekstra, & Sa Worldwatch Institute, 2004). Recent research suggests that the water footprint (WF)2 of indirect (embodied, or global) water consumption may be far more significant than direct (local) water €ssling et al., 2012; see consumption alone (Cazcarro et al., 2014; Go also Sun & Pratt, 2014). Though people also consume water at home, there is strong evidence that tourism increases overall water consumption € ssling et al., 2012). In most countries, water use by tourism is (Go less than 5% of domestic water use, but there are certain countries where tourism is not only the main economic activity but also the main factor in water use and where the sector has great relevance

2 ‘Water footprint’ (WF) is defined for the purpose of this paper as the total volume of water used to produce a unit of a good or service consumed by a tourist.

234

€ssling / Tourism Management 46 (2015) 233e244 S. Go

for water security and competition for scarce resources. Such water and tourism ‘hot spots’ include a number of small islands in the Caribbean and the Mediterranean. Even more important is the role of tourism in regions where abstraction levels are high due to concentrated tourism development, and where the natural recharge of aquifers is limited. This is the case in for instance Malta, € ssling Cyprus, Mallorca and other islands (Clarke & King, 2004; Go et al., 2012; Hadjikakou, 2014). As freshwater availability is increasingly under pressure (WWAP, 2012), water consumption in tourism has received growing attention by organizations such as the World Tourism Organization (UNWTO, 2013), UNEP (2011), and OECD (2013), with calls made by these organizations to reduce water consumption. Water use in tourism has so far been studied from three different perspectives: i) Direct and indirect use, measured in L or m3 (e.g. Bohdanowicz & Martinac, 2007; Essex, Kent, & Newnham, €ssling et al., 2012). 2004; Go ii) Sustainability implications of water use, including water scarcity, competition for scarce resources between tourism and other economic sectors or local populations, as well as the transfer of water use between countries and continents as a result of global tourism flows (e.g. Cazcarro et al., 2014; €ssling, 2001a; Go € ssling et al., 2012; Cole, 2012, 2013; Go Hadjikakou, Chenoweth, & Miller, 2013; Page, Essex, & Causevic, 2014). iii) Water management, including all actions that can help to € ssling et al., 2012; OECD, 2013; reduce water demand (Go UNEP, 2011). As water is considered an increasingly scarce resource, various indicators to assess its availability and use intensities have been developed, generally with a view to reduce water consumption. Current use of indicators is largely focused on direct water use, i.e. the volume of local water consumed per tourist per day, which is usually restricted to accommodation. This excludes other areas of water consumption e such as activities, shopping or services -, as well as indirect (or imported/embodied) water needed for the production of infrastructure, fuels and foodstuffs. The magnitude of the omission of focussing solely on direct water use is evident. For € ssling (2002) estimated that the world's direct water instance, Go footprint of tourism amounted to 1 km3 of fresh water. In comparison, Cazcarro et al. (2014) presented an assessment of the net water footprint of tourism in Spain, i.e. the water embodied in goods and products consumed by tourism in this country. Including national water resources and adding imported water, but subtracting water exports, Cazcarro et al. concluded that the Spanish tourism system requires 6.9 km3 of fresh water annually, which is €ssling's (2002) global estimate. This implies almost seven times Go that, on a global scale, tourism, as a sector reliant on inputs of goods from other sectors (Briassoulis, 1991), has a considerable indirect water footprint. The insight that tourism is a far more relevant water-consuming sector than previously assumed requires a reconsideration of current approaches to water management. The locus of most water consumption in tourism is accommodation; it is here that tourists consume water directly during their stay, sign up to different activities, and eat a share or all of their food. Although most research on water consumption in tourism has, in fact, already focused on hotels and other accommodation, considerable knowledge gaps remain. Given recent changes in the understanding of the importance of different direct and indirect water use sub-sectors and resulting uncertainties, the following sections aim to provide an updated review of water footprints, specifically with regard to the

role of food and energy consumption; to better differentiate various end-use sectors of water consumption and their relevance, and to assess the role of food in water consumption, one of the main remaining research gaps. In order to overcome the paucity of data in some areas, additional results from a case study in Rhodes, Greece, are included in the analysis. 2. Water use in tourism While there is now a fair amount of information on direct water consumption, indirect water use values for food, constructions, and € ssling et al., 2012), and research fuels are still poorly understood (Go on water footprints is generally not as far developed as, for instance, studies of direct and indirect greenhouse gas emissions from tourism (e.g. Filimonau, Dickinson, Robbins, & Reddy, 2013). Recent studies indicate that the consideration of imported and exported water, as well as the amount of water abstracted ‘locally’, compared € ssling et al., to ‘global’ water, is important (Cazcarro et al., 2014; Go 2012; Hadjikakou, 2014; Hadjikakou et al., 2013). As the focus of this paper is the development of new indicators for water use occurring in accommodation, Table 1 suggests two ‘direct’ and six ‘indirect’ water use categories. It is thus conceptually similar to the ‘total water footprint’ concept developed by Hadjikakou et al. (2013), which is based on a bottom-up component-based approach, i.e. including an accommodation and activity footprint (direct) and a diet and fuel footprint (indirect). Even though the WF assessment in Table 1 has been extended to also include infrastructure, it is incomplete in that it excludes marketing & sales as well as shopping and other tourism-related services, for which limited data exists. It is also fundamentally different from the topdown inputeoutput approach provided by Cazcarro et al. (2014) for 76 sub-sectors of the Spanish economy. Both bottom-up and topdown water footprint approaches have merits and weaknesses (see e.g. Feng et al. 2011). For the purposes of the present study, where the focus is water use in accommodation, a componentbased bottom-up approach is used to generate new data and insight. Direct water use in accommodation ranges between 84 and 2425 L per tourist per day, including water use in rooms, for gardens and pools irrigation, with activities adding 10e875 L/guest  Tortella and Tirado 2011; Go €ssling et al., 2012; night (Deya Hadjikakou et al., 2013). The higher value for activities relates to golf, which appears to be the most water-intensive activity in tourism: According to Dey a Tortella and Tirado (2011) e who conducted a comprehensive study of 196 hotels in Mallorca -, accounting for golf courses as part of the hotel direct water

Table 1 Direct and indirect water use in tourism. Water use category e direct

Min-max in L/guest night

Estimated average L/guest night

Accommodation Activities

84e2425 10e875

350 20

Water use category e indirect Infrastructure Fossil fuels for transport Energy use at hotel Biofuels Food Other consumption

L/guest night 0.2 5e2500 0.3e200 2500 4500e8000 n.a.

0.2 130 75 e 6000 n.a.

Total per tourist/guest night

4600e12,000

6575

€ssling et al., 2012, updated based on Energies Nouvelles (2011), Rosello Source: Go , Cladera, and Martinez (2010), US Department of Energy (2012). Batie, Mola

€ssling / Tourism Management 46 (2015) 233e244 S. Go

consumption increases total water consumption by 87%. In terms of accommodation water use, systemic reviews of summer destina Tortella & Tirado 2011) and upscale tions such as Mallorca (Deya chains such as Hilton (Bohdanowicz & Martinac, 2007) or destination water footprints (Hadjikakou et al., 2013) have consistently reported direct mean water use values exceeding 300 L/guest night (see also WWF, 2004). Based on the aforementioned studies, average global direct water consumption values may be in the order of 350 L/day for accommodation, and 20 L/day for activities, i.e. values about 15% higher than averages reported in earlier studies €ssling et al., 2012). (Go Considerably greater uncertainty exists with regard to indirect water use, specifically for food, fossil fuels, energy use at the hotel, biofuels, or the construction of tourism-related infrastructure. With  -Batie et al. (2010) analyzed three hotels regard to the latter, Rosello in the Ballearic islands, finding water requirements for construction in the order of 85e97 L/m2. Depending on the size of the hotel's built area, these values can be used to determine water consumption per guest night, i.e. total water use embodied in constructions divided by the number of guest nights in a given year, and divided by 50 to account for the assumed lifetime (50 years) of the infrastructure. As an example, in a hotel with an average size of 20 m2 per bed e including public and administrative areas e, an assumed 200 guest nights per bed per year, and an average of 90 L/m2 of water used during construction, this would add about 0.2 L/guest night over the hotel's 50 year lifecycle. Construction-related water use may thus be seen as a negligible factor in water consumption, unless the use of other infrastructure for tourism e airports, ports, roads, constructions for activities (e.g. ski lifts), events, museums, €ssling, 2002) e is included as well. restaurants, etc. (cf. Go Fuel production, and hence power generation involving fossil fuels, is more water-intensive, though probably lower than the 18 L of water required to produce one litre of gasoline reported by €ssling et al. Worldwatch Institute (2004), and previously used in Go (2012) and Hadjikakou et al. (2013). Energies Nouvelles (2011) report, for instance, that 3e5 L of water are needed to produce 1 L of oil. Where oil production is mature, the ratio can however be as high as 14:1, due to an increasing share of water in the produced fluid (known as the water-oil ratio). The US Department of Energy (2012) reports that consumptive water use for US crude oil production ranges between 2.1 and 5.4 L of water per L of crude oil, while Saudi Arabian oil wells consume 1.4e4.6 L of water per L of crude oil, and Canadian oil sands 2.6e6.2 L. These figures refer to net water use, which is defined as the ”sum of water input less water that is recycled or reused” (US Department of Energy (2012): 3). Notably, water quality for crude oil production can vary, involving fresh water, saline water, as well as desalinated water. Depending on source, oil production may thus require between 1.4 L (crude oil minimum) and 6.2 L (oil sands maximum) of consumptive water, or 3e18 L of water per L of fuel, if measured as general water input, corresponding to reported minemax values (Energies Nouvelles 2011; Worldwatch Institute, 2004). Given the wide range in water use values, an estimated average value of 10 L of general water input per L of fuel is considered in this paper as an approximation of water use related to fossil fuel production. Given an average global return travel distance of 1898 km in 2010 and energy use of 1.123 MJ/pkm (domestic and international tourism; Peeters, 2013), fossil fuel use translates into approximately 130 L of water per guest night on global average. Minimum and maximum values will however vary considerably: the difference between a longer-stay, train-based trip and a short long-haul trip involving a cruise (e.g. Eijgelaar, Thaper, & Peeters, 2010) would indicate fossil fuel water use values of between 5 and 2500 L per guest night. Water use embodied in the use of energy required for water production (lifecycle of provisions, i.e., pumping, transport,

235

treatment, desalination) is not included in these estimates, nor is the lifecycle energy use for tourism infrastructure. As for desalination, energy requirements of up to 12 kWh/m3 of water have been reported (Gude, Nirmalakhandan, & Deng, 2010). Filimonau et al. (2013) also suggest that lifecycle assessments add 15e30% to direct energy use, significantly increasing WFs. Energy consumption at the hotel can be measured in MJ, kWh or L of fuel. Depending on the source of the energy consumed, i.e. electricity generated from fossil fuels, coal, gas, nuclear or renewables, as well as the amount of fossil fuels burnt in boilers, cars, etc. at hotels or for accommodation-related purposes, water embodied in energy use in accommodation will vary. Values in the literature suggest direct energy use of between 3.5 and 1536 MJ, or the € ssling, 2010). This equivalent of 0.1e42.2 L diesel per guest night (Go translates into water use values of 1e420 L/guest night, with an estimated average of 65 L/guest night, based on values as provided by Bohdanowicz and Martinac (2007) for the Scandic and Hilton chains, and the average water use value of 10 L for every litre of fuel consumed established above. Note that the energy used for construction of the hotel should be added to this, as it appears to be -Batie et al. (2010) suggest that integrated over a significant: Rosello 50-year lifetime of the infrastructure, energy use embodied in building materials is about 20% of annual operational energy use. Adding this, the estimated average of 65 L/guest night would increase to 75 L/guest night in water use for energy consumption in accommodation. Notably, while the direct construction WF is thus negligible, the WF related to energy use in constructions is significant. Biofuels are increasingly advocated as a sustainable fuel for tourist transport (air, surface-bound). However, the use of biofuels would increase the water footprint of tourism considerably. UNESCO (2009:11) reports that 44 km3 or 2% of all irrigation water are already allocated to biofuel production, noting that realization of all current national biofuel policies and plans would require an additional 180 km3 of irrigation water per year. UNESCO further reports that the production of 1 L of liquid biofuels currently requires a global average of 2500 L of water. Data provided by the US Department of Energy (2012) suggests consumptive water use of 14e336 L per L of corn ethanol. For cellulosic ethanol, values of 4.5e4.6 L of water per L of fuel have been reported. No average is provided in Table 1, as biofuel use in tourism appears still very limited, but the linkage to water consumption is relevant. It is important, however, to realize that “sustainable” options such as biofuels or desalination represent water trade-offs, which must be quantified before considering such measures (Hadjikakou et al., 2013). Finally, food production requires considerable amounts of water. Depending on the local climate, crop or livestock varieties and agricultural practices, it can take between 400 and 2000 L of water to produce 1 kg of wheat, or 1000 to 20,000 L of water to produce 1 kg of meat (UNESCO 2009; Mekonnen & Hoekstra, 2010a, 2010b). On global average, the production of one kg of food may require between 74 L (beer) to 17,196 L (chocolate) of water (Fig. 1). An important distinction needs to be made between the green, blue and grey water components of the food-related water footprint. Green water stands for the precipitation on land that becomes soil moisture in the unsaturated soil zone that plants then use to grow €m, 1993). Blue water refers to sur(sensu Falkenmark and Rockstro face and groundwater, i.e. water in its conventional sense found in rivers, lakes and aquifers and used for irrigation. Finally, grey water is defined as the amount of water that is required to dilute pollutants from food production (such as fertilisers and pesticides) to such an extent that water quality remains above given quality standards (cf. Water Footprint Network, 2013). Depending on the foodstuff and its origin and composition the amount of green/blue/grey water

236

€ssling / Tourism Management 46 (2015) 233e244 S. Go

Fig. 1. Total (sum of green, blue and grey water) in relation to blue water embodied in one kg of produce, approximate values. Source: Mekonnen & Hoekstra, 2010a, 2010b

contained in a kg of foodstuff can vary widely, both in absolute (L of water needed) and relative terms (percentage of green/blue/grey water) (see also Chenoweth, Hadjikakou, & Zoumides, 2014). Fig. 1 also shows that the share of blue water contained in different foodstuffs is small, varying between 1% (chocolate) and 30% (tomato) on global average. In comparison, green water accounts for 56% (lettuce) to 98% (chocolate), and grey water for 1% (chocolate) to 32% (lettuce) of total water requirements. These values represent global averages and can differ markedly between countries (Chenoweth et al., 2014). It has so far been estimated that daily water requirements to support human diets range from 2000 to 5000 L of water per € ssling et al., 2012), based on the UNESCO (2009) person per day (Go estimate of 1 L of water for 1 kcal of food. As demonstrated by Hadjikakou et al. (2013), this is likely to be a considerable underestimate for tourism, where a greater share of higher-order, protein-rich foods are consumed compared to the global average €ssling, Garrod, Aall, Hille, & Peeters, 2011). In Hadjikakou et al., (Go calculations for five hypothetical holiday food menus revealed water footprints of 4696e7876 L per day per tourist, and water use per calorie values of 1.38e2.34 L/kcal. Adding up all the water footprint components considered here, tourism water footprints have to be revised upwards, with a range of 4600e12,000 L, and an estimated average of 6575 L of water per tourist per day (Table 1). As results are conservative, and water embodied in shopping, services, or the commuting of employees not considered, results as presented in Fig. 2 would indicate that: a) tourism is considerably more water intensive than previously suggested; b) most of its water footprint is associated with indirect water consumption, c) food is by far the most important water use factor, accounting for an estimated 87% of total water consumption, and d) there are vast differences in minimum and maximum water consumption, indicating considerable scope to reduce water use (Table 1, Fig. 2). Yet, as discussed in the following section, in order to realize this potential, appropriate new indicators need to be developed.

environmental aspects (e.g. Miller, 2001). All indicators have in common that they measure performance, often with the goal to inform policy. As outlined by UNWTO (2004), it is important for companies to assess how their actions affect the assets and values of a destination, and to use indicators to measure what are regarded by UNWTO as ‘risks’ to the ecological, socio-cultural and economic functions of regions. According to Hunter and Shaw (2007: 48), the environmental impacts of tourism have received comparably scarce attention, and “the implications of supplying energy, food and water to destinations areas are often excluded from studies of the sustainability of tourism products and destinations”. Hence, in this article, focus is in on environmental indicators and, more specifically, conditions of water availability and pressure on water resources, in an attempt to translate theoretical insights into management recommendations. Where water consumption and its management have been discussed, assessments have usually focused on environmental indicators (see also Fang, Heijungs, & de Snoo, 2014). One of the first sets of indicators was presented by UNWTO (2004). This destination-based and less operational manual for sustainable

3. Water indicators in tourism There exist a wide range of indicators in tourism. These can be qualitative or quantitative, and may focus on social, economic, or

Fig. 2. Globally averaged water footprint, L per guest night.

€ssling / Tourism Management 46 (2015) 233e244 S. Go

tourism included indicators of water availability defined as ‘the % of annual supply in use’, ‘the number of days shortage per year’, and the ‘cost of new water’. In the first scientific review of water consumption in tourism, Bohdanowicz and Martinac (2007) identified water consumption in hotels, summarizing the results of reports and publications from the period 1990e2002, all of which expressed water use as consumption in L per room, guest night or m2. Bohdanowicz and Martinac (2007) also presented a unique data set for 73 Hilton and 111 Scandic hotels. As in previous studies, water use in these chains was expressed as ‘m3 per room and year’ and ‘L per guest night’. For the Hilton/Scandic sample, ‘laundry in kg per guest night’ was added as a new indicator, though no data for water use per kg of laundry was presented. In most other studies presented since then, ‘water use per guest night’, or, synonymously, ‘water use per tourist per day’, has remained the key indicator €ssling et al., 2012, Table 2). (Go lez, Guerrero, Exceptions include Blancas, Lozano-Oyola, Gonza and Caballero (2011), who suggest assessing the environmental dimension of sustainable tourism with regard to fresh water use based on i) water use, calculated as total volume consumed per day, and ii) ‘water saving’, calculated as the volume of reused water attributable to tourism (Table 2). The European Commission (EC, 2013), in its European Tourism Indicator System TOOLKIT for Sustainable Destinations, presents four indicators for water management: ‘Fresh water consumption per guest night compared to Table 2 Comparison of water use indicators.a Indicator

Reference/Organization


Water use per guest night

Antakyali, Krampe, and Steinmetz (2008); Bohdanowicz and Martinac € ssling (2007); Eurostat (2009); Go (2001a); Lamei, von Münch, van der Zaag, & Imam (2009b); Lamei, Tilmant, van der Zaag, and Imam (2009a); Langumier and Ricou (1995); Rico-Amoros, Olcina-Cantos, and Sauri (2009); WWF (2004) Alexander (2002); Deng and Burnett (2002) Cooley, Hutchins-Cabibi, Cohen, Gleick, & Heberger (2007); O'Neill & Siegelbaum and The RICE Group (2002) Bohdanowicz and Martinac (2007)


Water use per room


















m3 per room and year Laundry in kg per guest night Total volume consumed per day Volume of reused water The % of annual supply in use The number of days shortage per year Cost of new water Fresh water consumption per guest night compared to general population water consumption per person night; Percentage of tourism enterprises with low-flow shower heads and taps and/or dual flush toilets/ waterless urinals; Percentage of tourism enterprises using recycled water; Percentage of water use derived from recycled water in the destination

Blancas et al. (2011) UNWTO (2004)

European Commission (EC 2013)

a Other indicators on water use have been presented by e.g. Accor, Scandic Hotels, Rezidor SAS, Marriott, Six Senses, International Tourism Partnership and Green Globe 21, Green Key, American Hotel & Lodging Association, Considerate Hoteliers, Carbon Disclosure Project Reporting, UN Global Compact or the Global Reporting Initiative LEED, TraveLife, UK Green Tourism Business Scheme, TripAdvisor greenleaders.

237

general population water consumption per person night’, ‘Percentage of tourism enterprises with low-flow shower heads and taps and/or dual flush toilets/waterless urinals’; ‘Percentage of tourism enterprises using recycled water’, and ‘Percentage of water use derived from recycled water in the destination’. Finally, water management programmes as presented by the Accor group and Kuoni reveal different approaches. Accor (2011) defines water consumption to include all water used for irrigation systems (agriculture) and livestock consumption, as well as water metered in hotels, i.e. excluding well and rainwater. Kuoni (2013) presents a comprehensive water management manual for direct water use, including per capita consumption values and comparison with best practice, hot and cold water cost, laundry cost, flow measurements from fittings, towel re-use, wastewater treatment and cost-benefit analyses for laundry reduction and plumbing fixtures. The manual proposes a ‘water champion’ scheme, which uses an economic argument to generate interest in water reduction. Findings as presented above raise a number of issues regarding the current use of indicators for water management in tourism: 1. The most widely used indicator, water use per guest night, only considers direct water use, ignoring the importance of embodied water (food, fuels). In some assessments, such as by Accor, well water is not included, even though groundwater abstractions may be specifically relevant for sustainable water € ssling, 2001a). management (Go 2. ‘L per guest night’ is an indicator of relative use levels that can also be used for benchmarking purposes, but it does not indicate whether abstraction levels are sustainable in terms of aggregated (absolute) consumption, i.e. in comparison to available renewable water resources. As an example, Iceland has vast renewable and geothermally warmed amounts of water, and per tourist water use is largely irrelevant. In other areas, water consumption impacts are highly dependent on the time of the year when water is abstracted. In various Mediterranean islands, water has to be desalinated or imported by ship (Clarke & King, 2004), which may be considered unsustainable because of additional energy use and associated emissions of greenhouse gases. Tourism-related water use impacts are also relevant in terms of their opportunity cost (alternative uses), as well as equity aspects (the situation of local populations), which are not captured in WFs. 3. Water use as measured in ‘L per guest night’ may not be a sufficient criterion on its own for the purposes of informing water management. Only by combining it with an audit of sub-sectors can particularly water intensive end-uses be identified and measures for water conservation implemented. 4. A considerable share of water use is characterized by ‘lock-in’. For instance, as fixed installations, pools need to be filled and require constant water replenishment. Indicators thus need to distinguish between the planning of tourism infrastructure and its operation, considering an embedded and an operational WF. 5. There is evidence that water consumption in tourism is growing due to various trends, such as growing interest in energyeor water-intensive activities; higher hotel standards with larger pools and gardens; higher quality standards including in-room Jacuzzis; all-inclusive arrangements with large buffets; or the planned use of biofuels for transports. Such trends need to be addressed.

4. Case study e Rhodes As the development of new indicators requires additional knowledge, for instance with regard to the use of food or pool

238

€ssling / Tourism Management 46 (2015) 233e244 S. Go

characteristics, a case study of three hotels was conducted in Rhodes, Greece. The Mediterranean island was chosen due to an opportunity to collaborate with tour operator Thomas Cook, as well as the fact that the Mediterranean is water scarce, and one of the most important tourism regions in the world (WWF, 2004; Clarke & King, 2004). While the share of water use for tourism in Rhodes is unknown, it reaches an estimated 4.8% of domestic water consumption €ssling et al., 2012), or 10% if including indirect in nearby Cyprus (Go water use (Hadjikakou, 2014). Rhodes received 1.42 million tourist arrivals in 2010, i.e. about 9.4% of all arrivals in Greece (Europtravel, 2013). With an average length-of-stay of 9.1 days (David Vanger, Europe Travel SA, personal communication, 14 November 2013), a total of about 12.9 million guest nights are annually spent in the island, where accommodation development has recently focused on upscale resort hotels. Arrivals peak in July and August when temperatures are at their highest and rainfall is practically non-existent (Hellenic National Meteorological Service 2013). For the purpose of this research, two five-star hotels (hereafter Hotel A and Hotel B) and one four-star hotel (hereafter hotel C) provided comprehensive data on water and energy use for 2012 and 2013. The case study is thus exploratory, and not representative of the wide range of different types of accommodation establishments in the world. Yet, the 4e5 star resort hotel is one of the most predominant types of hotel in many sun, sand, and sea destinations in the Mediterranean and around the world. All hotels studied are located directly adjacent to the sea, and have swimming pools and irrigated gardens of various sizes. Data on their water use was obtained by email in September/October 2013 through the respective general managers, in preparation for an on-site visit in October/November 2013. During the visit, water and energy systems of the hotels were inspected with technical staff, missing data collected, and a convenience sample of 101 mostly German tourists interviewed face-to-face on the basis of a questionnaire in Hotel A (local rating: five stars), including their water use habits (showers, taking baths, flushing the toilet; towel use) and food consumption patterns. Furthermore, water flow (L per minute) was measured in rooms and public areas for tabs and showers, and flush volumes for toilets were recorded. Discussions with technical and administrative staff helped to further clarify water consumption, especially with regard to room cleaning procedures. Hotel A also provided detailed data on food use in the form of a comprehensive purchasing list for 2013. As this hotel purchases foodstuffs together with another one, consumed food volumes were divided by the number of guest nights in both hotels, i.e. representing the average of food use in a 4-star and a 5-star establishment. The calculation also averages in 7800 guests visiting Hotel A during various events (conferences, weddings). For an estimated 50% of guests, food consumption covers three meals a day (full board, in all-inclusive packages), while for the other half, the food consumed only represents two meals per day (half-board). For the 7800 guests, this only represents participation in one meal. All meals are served as buffets, and have up to 100 components (breakfast & dinner) and half as many during lunch. No data was provided on food waste, though it was noted that employees were allowed to eat leftovers sent back to the kitchen after buffets had been removed.

Table 3 Overview key parameters hotels. Hotel A***** Year of construction Room/bed numbers Guest nights 2013 Area

2008 154/350 58,502 28,000 m2, out of this 11,000 m2 lawns and gardens Water sources and use (2012) 16,439 (83%) Well water m3 (%) Municipal water m3 (%) 3345 (17%) Total water use m3 19,784 L/guestnight 338 L (2012)

Hotel B*****

Hotel C****

2008 202/404 59,045 40,000 m2

1974 205/440 58,659 27,000 m2 gardens and beaches, plus buildings

39,789 (99%) 438 (1%) 40,227 675 L (2012)

e (0%) 13,700 (100%) 13,700 234 L (2013)

water consumption, in an attempt to account for the observed differences and, additionally, to suggest a distribution of water use by end-use sub-sector. Similarly, for indirect water use, new values for energy and food consumption are presented. 5.1. Direct water use There is a general lack of disaggregated end-use water data for gardens, pools, rooms, laundry and other purposes. Generally, water use per m2 of garden will vary greatly between different climates, depending on rainfall, evapotranspiration and plant species. Use values will generally be highest in dry, hot climates, and where plant species are not adapted to these conditions. In such climates, water use for irrigation can make up 50% of total water €ssling, 2001a). In Rhodes, water use values of use in hotels (Go irrigation were only provided by Hotel A, 100 m3 of water per day, i.e. one third of the well water pumped. Calculated per season, this amounts to 4380 m3, i.e. considering lower demand in the shoulder season and during periods of rain. Water use for irrigation is thus in the order of 75 L/guest night (Fig. 3). Regarding swimming pools, two factors affect water use after the initial filling, i.e. evaporation and filter backwash. Evaporation depends on various factors: pool water surface; whether the pool is undisturbed or occupied, with the number of people increasing water movement and evaporation surface, and causing water transfer out of the pool on bodies and in swimsuits; pool water temperature; as well as air temperature, humidity and vapour pressure of moisture in air (Shah, 2013). In Rhodes, initial fillings were calculated to amount to 40 L per guest night (2301 m3 divided by 58,502 guest nights; Hotel A), and 7 and 12 L/guest night (Hotel

5. Results Table 3 outlines differences in the three hotels studied, which are similar in terms of standard (4e5 stars), garden size (28,000e40,000 m2), bed numbers (350e440) and guest nights (58,502e59,045), though different in terms of direct water use (234e675 L/guest night), and the share of municipal versus well water use (0e100%). The following sections further discuss direct

Fig. 3. Direct water use in accommodation by end-use (L/guest night).

€ssling / Tourism Management 46 (2015) 233e244 S. Go

C and B). Spa indoor pool fillings would add 1 L/tourist per day to this (84 m3 divided by 58,502 guest nights; Hotel A). Limited data was found for evaporation. Rates of 6e8 L/m2/day for different Australian climates have been reported by a commercial pool cover provider (Daisy Pool Covers, 2013). Similar values have been measured by Hof and Schmitt (2011), who reported daily losses of 5 L/m2/day for a swimming pool in Mallorca, and annual water losses of 1.83 m3/m2. Notably, these values were recorded over the whole year. At estimated 6 L/m2/day in summer, this would translate into daily losses for outdoor pools of 9870 L/day, or 33 L per guest night in Hotel A, at an assumed 85% occupancy rate over the season. Again, the spa indoor pool would add about 1 L/tourist/ day to this. Filter backwash was not quantifiable in Hotel A, but amounted to 2.3 m3/day in Hotel B, corresponding to about 7 L/ guest night at 85% occupation over the season. As indicated in Table 4, pool area per bed appears to be an important explanatory variable with regard to the differences found between Hotel A and Hotel C in terms of water consumption for pools. With regard to room-related water use, the survey of tourists (n ¼ 101) indicated flush-frequencies of 3e15 times per tourist per day (toilets inside rooms, as well as public toilets in the hotel). Daily cleaning by staff adds 2 (hotel A, C) to 3 flushes (hotel B) per room (discussions with cleaning staff at hotels). On average, a value of 50 L per tourist per day was calculated based on average flushing times of tourists (5.7 times), staff (once, to account for double occupancy in rooms) and a water volume of 7 L/flush (a ‘small flush’ option does not exist). These values may apply when tourists stay in the hotel, but they are lower when tourists leave the hotel for excursions or other activities. A somewhat lower value of 45 L per tourist per day is thus assumed here. Given vast ranges in water use in toilets, with older models using more than 12 L/flush, water use for toilet flushing may reach more than 100 L per guest night in other hotels. Showers were on average used 2.6 times per tourist per day (n ¼ 100), with a range of 1e7 showers per tourist per day. This includes showers in rooms, at the pool, and in public areas. Tourists reported to shower on average 6 min, with values ranging between 1 and 15 min. For the purpose of calculations, lower values are used, i.e. 2 showers per tourist per day at 5 min, as showers at pools are unlikely to be long. It is possible that more showers are taken in July/August, when temperatures exceed 30  C. At measured water flows of 7e12 L/min, and an average flow of 7 L/min in rooms,

Table 4 Water use for pools and spas. Hotel A*****

Hotel B*****

Hotel C****

Pool area (m2) Pool volume (m3)

1645 m2 2301 m3

425 m2 420 m3

Evaporationa Filter number Filter backwashb (estimate)

9870 L/day 15 Backwash 1/day, exact volume unknown

2065 m2 No data available 12,390 L/day 12 2.3 m3/day

Spa indoor

56 m2/84 m3

Pool area per bed (excluding spa)

10.7 m2/bed

80m2/no data available 10.2 m2/bed

2550 L/day 6 Backwash 1/day, sometimes 2/day, exact volume unknown 48 m2/41 m3 2.1 m2/bed

a Assumed at 6 L/m2; dependent on temperature (water and air), humidity, wind, vapour pressure. b Filters are sized based upon pool volume (m3), pool turnover rate (the time in hours required to completely cycle the total volume of water through the filter, typically 2e4 h), and the particular flow rate for a given type of filter, such as sand media.

239

where most water is used, this would suggest average water use of 70 L per tourist per day. Bathtub use has to be added to these values. Even though only 28% of respondents reported to use bathtubs at all, with most of these trying the bath just once (68%), there was also a small share of ‘bathers’, using the tub every day (17%), though this share may be culturally dependent and could be significantly different in other samples. On average, bathtub use among ‘bathers’ was 2.2 times/week, and 0.6 times/week for the whole sample. At an estimated filling of 40 L per bathtub, this would amount to 3 L per guest per day. It should be noted that Hotel B had private Jacuzzis in each room, which, together with private pools, are likely to cause most of this hotel's particularly high water use. With regard to the last direct water use factor in rooms, the tap, no values were found, as it is difficult to identify the number of times the tap is used. In most cases, tourists will wash their hands after they have used the toilet (5.7 times), and when they brush their teeth. Assuming that tap water is also used during cleaning, overall use of tap water may be 10 times per tourist and day, and for an estimated 10 s each time. An exception may be tourists running taps constantly when brushing their teeth. At measured water flows of 4.5 L/min at Hotel A, this would correspond to about 8 L per guest night for tourists staying mostly in the hotel. Since no data on laundry weight could be determined, the values by Bohdanowicz and Martinac (2007) of between 0.7 and 3.1 kg/guest night for Scandic hotels (mean: 2 kg/guest night), and 0.7e16.2 kg per guest night for Hilton hotels (mean: 4.1 kg/guest night) may serve as an indication of laundry weight. At 11 L/kg of water use per kg of laundry (values Hotel A), this would correspond to 7e175 L per day. In Hotel A, at an estimated 2e3 kg per guest per day, water use for room towels (changed every second day), pool towels (changed upon request), bed sheets (changed daily) and bathrobes (changed once a week) would amount to an estimated 30 L/guest night. Overall, the results indicate direct water use of 317 L/guest night, without kitchen consumption. Compared to the measured consumption of 338 L/guest night for Hotel A, which is considered for comparison as data was collected here, the major missing variable, kitchen use, would account for the remainder of 25 L per guest night. Water use for cleaning, which is largely restricted to dry cleaning (marble floors), is negligible. This review of direct water use in Hotel A supports earlier findings according to which the presence of gardens automatically renders irrigation the main €ssling, 2001a), followed by showers and water use factor (cf. Go toilet flushing (Fig. 3). As for pools, both pool volume and size have been identified as factors in water use, as the volume affects the initial water demand for filling the pool as well as backwashing, while the surface area affects evaporation rates. The significance of water needed to fill pools in Hotel A, 40 L/guest night, indicates that ‘water lock-in’ demands careful design of hotels already in the planning stage. Finally, laundry volumes can be significant, specifically when high laundry volumes are washed in older machines. Note that wastewater treatment is not included in this calculation; a rough calculation indicates that treatment may require 0.5 kWh/ m3 of wastewater (DWA, 2011), and a corresponding indirect freshwater use (energy production, see below). 5.2. Indirect water Indirect water consumption is composed of energy (transport to/from destination), energy use at the hotel, and water embedded in food. While energy use for transport to/from the destination has been discussed earlier, the energy use of activities also needs to be considered, as these are often offered at or through the hotel, including rental cars for scenic drives or the visitation of attractions, as well as marine activities including boat-rides, diving or

240

€ssling / Tourism Management 46 (2015) 233e244 S. Go

snorkelling trips, water-skiing, banana-rides, paragliding, and others. For such activities, fuel use intensities can be in the order of 10e20 L per hour of operation (UNWTO-UNEP-WMO 2008). No additional data on activities was collected in Rhodes, however. As shown in Table 5, energy use per guest night varies in the hotels studied, with reported energy equivalent values of 17e42 kWh, i.e. at the lower end of the range for international €ssling (2010), and probably also below tourism reported in e.g. Go the average in Greek hotels, which however are difficult to compare due to the unit use of ‘kWh/m2’ (see Farrou, Kolokotroni, & Santamouris, 2012). At an estimated 0.8 L of water per kWh (Worldwatch Institute, 2004), this suggests energy-related WF of between 14 and 34 L/guest night. As shown in Table 5, there are currently small solar thermal areas installed per room, even though technical staff suggests that these are largely sufficient to provide warm water. None of the hotels studied produces photovoltaic electricity, even though Greece must be considered to have a good potential for the use of solar technology, and in all hotels, large empty spaces where found on rooftops. Food consumption patterns were investigated in greater detail in Hotel A. Averaged values were considered for a total of about 150 foodstuff items, some of them summarizing food groups, such as ‘spices’, as well as 60 categories of beverage items. While the average amount of each food/beverage item consumed per day often only includes a few grams, total weight of these can be considerable, which is relevant when environmentally more problematic foods are consumed, such as crustaceans including € ssling et al., 2011), or overfished lobster and giant shrimps (Go species such as swordfish, cod, or sharks (e.g. Fish Red List, Greenpeace, 2013). In total, an average weight of 3.122 kg of food is consumed per guest per day (Table 6), as well as 1.784 L of beverages, out of this 0.611 L of alcohol. Food group averages in Fig. 4 were calculated on the basis of weighted values for the globally averaged amount of water embodied in various foodstuffs, as provided by Mekonnen and Hoekstra (2010a, 2010b). Global averages were used because areas of origin for foodstuff purchases were impossible to determine. Together, these represent embodied water consumption of 4557 L for foodstuffs, and 940 L for beverages, i.e. 5497 L per guest night in total. Blue water consumption is an estimated 15e20% of this (Fig. 1), ranging from 4.4 L water per L of beer to 512 L of water per kg of olives. As indicated in Table 6, the total weight of food consumed per tourist per day amounts to 3.1 kg, plus 1.8 L of beverages. Note that this may included as much as one kg per tourist per day of food waste generated during preparation (peels, stones, bones, or food that has to be discarded for other reasons). In particular, the share of high protein foods is high, with meat consumption amounting to 0.385 kg per tourist per day, plus 0.139 kg of seafood & fish, and 0.294 kg of dairy products & eggs. These results confirm the highorder food intake of tourists found in earlier studies. One study in a 5-star resort hotel in Zanzibar, Tanzania, found meat consumption values of 0.201 kg per tourist per day, plus 0.429 kg of fish and

€ssling, 2001b). Results also appear to reflect the stateseafood (Go ment by echoed by 75% of the tourists interviewed (n ¼ 101) that they had “eaten more than at home”. Both greater food volumes and higher-order food consumption explain large water footprints. Notably, this excludes the water footprint of three foodstuff groups for which no data on embodied water was identified, i.e. Sweet foods (desserts), Other, Fish and Seafood (Table 6). Results are thus in line with the initial estimate of 6000 L of water per tourist per day (Fig. 2), and Hadjikakou et al.'s (2013) estimated range of water embodied in five hypothetical meal plans (4700e7900 L/tourist/ day). As indicated in Fig. 4, meat consumption contributes the largest share of indirect water consumption, 2.65 m3, confirming an inverse correlation between the weight of some food groups in comparison to their contribution to indirect water use. Both meats and dairy products have large water footprints in comparison to their weight e as well as their caloric content (cf. Hadjikakou et al., 2013) -, while Carbohydrates (bread, rice, etc.) are comparably less water-intensive. Fruit and vegetables, in comparison, have a more favourable weight-to-water ratio. Results thus confirm that buffets should offer a large share of vegetable-based dishes, as well as fresh fruits, while generally following rules for purchasing, preparation €ssling and presentation as outlined for energy management (Go et al., 2011). 6. Towards new indicators for water management Given the need to reduce fresh water consumption in tourism and concomitant evidence of the profitability of water saving measures (e.g. Kuoni, 2013), as well as expressions of commitment of large parts of the industry to engage in water management (cf. Table 2), the question arises as to how such savings can be realistically achieved. Clearly, water management requires action on various levels, including governance (policies and legislation), management, technological innovation and behavioural change (staff, tourists). Yet, such action needs to be guided, with indicators providing a vital foundation. Currently, only a few indicators are used and these may be of limited value for water management. Comprehensive water use indicators need to consider local water availability; direct and indirect water consumption, and planning/ operation. Moreover, as underlined by Bohdanowicz-Godfrey and Zientara (2014), only where detailed audits exist can opportunities for efficiency gains be identified and realized. While results as presented in this paper may primarily apply to resort-style hotels, they also have relevance for other forms of accommodation by highlighting the importance of indirect water consumption through fuels and food, in addition to the importance of gardens, showers, and pools in terms of the management of direct water use. Results also point at the significance of ‘water lock-in’, i.e. fixed water demand over the lifetime of the infrastructure, for instance for pools. These insights also need to be seen in light of evidence that water, energy and food consumption in

Table 5 Energy use at hotels. Energy use

Hotel A*****

Hotel B*****

Hotel C****

Energy consumption per guest night (2012)a Energy equivalent from above Energy water footprint (0.8 L water/kWh) Solar thermal installed Solar thermal per room

19 kWh þ 0.5 L oil 25 kWh/guest night 20 L/guest night 370 m2 0.8 m2/room

32 kWh þ 0.338 L diesel þ 0.597 m3 gas 42 kWh/guest night 34 L/guest night 248 m2 1.4 m2/room

16 kWh þ 0.05 L diesel 17 kWh/guest night 14 L/guest night 104 m2 0.5 m2/room

The Electric Company at Rhodes Island produces electricity with generators running on fuel oil and diesel. a Conversion factors from oil and diesel to kWh based on Carbon Trust (2013), values do not include transports, only generators/boilers; conversion factors from kWh to L of water based on Worldwatch Institute (2004), Energies Nouvelles (2011) and US Department of Energy (2012), with 1 kWh corresponding to an estimated 0.8 L of water.

€ssling / Tourism Management 46 (2015) 233e244 S. Go Table 6 Foodstuff use in kg per guest night.

241

Table 6 (continued )

Foodstuff

kg per guest night

Meats Cold Cuts Pork Rabbits Beef Chicken Duck Turkey Lamb Subtotal

0.063 0.115 0.002 0.066 0.084 0.008 0.018 0.029 0.385

Vegetables & fruits Dry fruits Frozen vegetables Frozen potatoes Fresh fruits Fresh vegetables Lemon Olives Subtotal

0.016 0.148 0.082 0.571 0.773 0.010 0.013 1.613

Other Pulses Oils Nuts Vinegar Soup Snails Stock Subtotal

0.005 0.113 0.020 0.041 0.001 0.001 0.037 0.218

Dairy & eggs Yoghurt Cream Butter &Margarine Cheese Eggs Subtotal

0.080 0.040 0.010 0.086 0.078 0.294

Carbohydrates Flour Cereals Sugar Pasta Mashed potatoes Semolina Pastry Bread Croissants Cookies & biscuits Rice Subtotal

0.025 0.059 0.024 0.034 0.002 0.001 0.017 0.064 0.013 0.014 0.018 0.271

Sweet foods & spices Traditional sweets Ketchup* Jam Spices Salt Honey Ice cream Subtotal

0.003 0.022 0.010 0.069 0.007 0.010 0.081 0.202

Seafood Lobster Calamari Shrimps Shellfish Octopus Squid Caviar Other seafood Subtotal

0.001 0.009 0.011 0.009 0.007 0.004 0.000 0.003 0.044

Foodstuff

kg per guest night

Fish Smoked fish Pangasius Shark Cod Swordfish Perch Salmon Tuna Sardine Bream Other Subtotal

0.005 0.022 0.020 0.005 0.004 0.014 0.011 0.002 0.000 0.002 0.010 0.095

Beverages (L) Soft drinks Refreshments Juices Milk Tea Coffee Wine Beer Spirits Subtotal

0.037 0.038 0.072 0.300 0.612 0.114 0.288 0.280 0.043 1.784

Subtotal foodstuffs Subtotal beverages Total

3.122 1.784 4.906

Note: The averaged volume of some of the foodstuffs consumed is less than 1 g per guest night, and appears as 0.000 in the table. Yet, accumulated to 181,262 guest nights, absolute volumes can become significant. For instance, the amount of Caviar consumed in the hotels is almost half a ton per year (497 kg), and the weight of lobster and giant shrimps exceeds 2.2 t. The list excludes 13 kg of salted fish; * Item ’ketchup' includes ketchup, mustard & mayonnaise.

tourism are increasing. For instance, specific electricity consumption in accommodation may have grown by up to 30% over the past decade (Hawkins & Bohdanowicz, 2011). Higher standards in hotels, with spas and multiple (heated) swimming pools, additional private pools, in-room Jacuzzis, sports and health centres, more laundry and more luxurious ‘rain’ showerheads are factors that will also increase water consumption. Consequently, unless new indicators for water consumption are identified, adjusted to sustainable water abstraction levels both for direct and indirect consumption, and allowing to measure and monitor water use, it seems unlikely that global water consumption in tourism will decrease significantly in either relative (per guest night) or absolute terms (total volumes). Water management thus needs to be optimized on the basis of disaggregated indicators that address different water use categories (garden and pools, rooms, restaurants), as well as a hotel's planning and operational phase. A precondition for the development of such indicators is an understanding of renewable water availability in the area, as well as the area's future development under various human development and climate change scenarios, determining benchmark levels for sustainable regional water use. Notably, this will have to consider peak water use, which might coincide with the period of lowest aquifer recharge or water availability. Regarding indirect water use, general rules for food management may render more complex, specific assessments for various foodstuffs redundant, though such assessments may be vital to avoid food purchases from particularly water scarce areas. Based on these considerations, the present study proposes 8 interrelated indicators for sustainable water management in accommodation, addressing the water situation in the area in

242

€ssling / Tourism Management 46 (2015) 233e244 S. Go

Indicator 4: Area of solar thermal and PV installed per bed Indirect water use is influenced by energy consumption. Hotels and virtually all forms of accommodation can in most climates reduce their indirect water demand by producing warm water with solar thermal installations and electricity from photovoltaic (PV) cells. Alternatively, various forms of € ssling, 2010). At the geothermal cooling have been explored (Go same time, accommodation should also be built in a way that minimizes energy consumption.

6.3. Operating accommodation

Fig. 4. Weight of foodstuff groups consumed per guest night, and water use in L. For a list of foodstuffs contained in each group see Table 6. Note: the following values were used where water footprint data was missing: dry fruits as fruits; cream, ice cream and yoghurt as milk; jam as fruits/sugar; cold cuts as pork; sweets, croissants, biscuits and pastry as pasta. Not considered due to lack of data: stock, ketchup, mustard, mayonnaise, seafood and fish, spices, salt, honey, snails, soup. The water footprint of beverages assumes the same values for soft drinks and refreshments (240 L/L, based on average values provided by Ercin, Aldaya, and Hoekstra (2011)), and the value for wine is also used for spirits (872 L/L). Averages do not account for buffet ‘leftovers’ offered to staff, potentially replacing a share of meals otherwise eaten at home. Source: data provided by Hotel A, Mekonnen and Hoekstra (2010a, 2010b).

question (1 indicator), the infrastructure planning process (3 indicators), and operations (4 indicators). 6.1. Area situation Indicator 1: Renewable water resources per guest night in peak season This indicator assesses a watershed or region's water supply system with regard to renewable water availability, and considering the region's future development (additional beds planned) under different climate change scenarios. It focuses on renewable water resources, as neither the use of fossil or desalinated water can be considered sustainable. The peak season is generally the driest season and thus an indicator of pressure on water resources in the time of their most limited availability. The outcome of the calculation of this indicator determines benchmarks for the following indicators. 6.2. Planning accommodation Indicator 2: Area of irrigated land per bed Gardens have been identified as a major water-consuming factor in many destinations, even though it is acknowledged that many city hotels do not have gardens. Where gardens exist, their size and design, as well as the plant species chosen influence water demand. A proxy for these variables is ‘area of irrigated garden per bed’, the general rule being that the smaller the irrigated area, the lower water consumption for irrigation. Indicator 3: Area of pool per bed Pools, like gardens, are major water consuming factors. Both pool size and pool volume influence water consumption as a result of the initial filling, evaporation and backwashing. The design of pools consequently has great relevance for future water use and ‘water lock-in’, and can be assessed on the basis of a calculation of pool area per bed. Where no pools exist, this indicator is redundant. In some destinations/hotels, the use of sea water for pools may be accepted or even appreciated by guests.

Indicator 5: Amount of meats and dairy products per guest night The importance of food in water consumption has been clearly outlined, and meat and dairy products have been identified as key food groups in water-intensive diets. For instance, Hadjikakou et al. (2013) find in their analysis of different meal compositions that 75% of the food water footprint is related to meat and dairy products. Reducing the share of these foodstuffs is thus of great importance for reducing indirect water demand. Indicator 6: Energy use per guest night Energy used in accommodation can be produced by the hotel itself (indicator 4), or be purchased from renewable sources. Where this is the case, the indicator becomes less relevant, even though reducing energy use is always relevant to improve environmental performance. Where energy use per guest night is high and costly, the indicator can help to identify major energy consumption sub€ ssling, 2010). sectors, which are likely to include A/C systems (Go Indicator 7: Share of rooms fitted with low-flow options Room-specific water use depends primarily on the showerheads chosen as well as low-flow toilets with dual flush options, and lowflow taps. The installation of Jacuzzis should generally be avoided, as these are water and energy intensive. A wide range of low-flow plumbing fixtures are now available, which can be chosen based on guest comfort perceptions and regional water availability (Indicator 1). Indicator 8: kg of laundry used per guest night This indicator refers to both water and energy use for laundry. The general rule is that the less laundry produced, the lower will be the direct and indirect water consumption. Again, specific benchmarks have to be set against water availability (Indicator 1). It is evident that several of the indicators presented above require benchmarks of what constitutes sustainable use values, i.e. these are dependent on the respective region's specific water situation. Other indicators may be used generically in accommodation where resource savings are a priority irrespective of water security concerns. Currently, none of the environmental management systems studied (Table 2) appears to account for renewable water capacity (see also Hsiao, Chuang, Kuo, & Yu, 2014 for a review of ISO 14000 and green hotel assessment systems); yet, this should determine sustainable benchmarks and hence the level of water saving activities in the hotel. For other indicators, such as the share of meats and dairy products consumed, the preparation of more vegetarian alternatives can be an option for sustainable water management, until these can be refined to account for regional or product-specific differences in water consumption (cf. Chenoweth et al., 2014). Evidence suggests that water management will be cost-efficient (e.g. Kuoni, 2013). Other aspects, such as the amount of sustainable water available for tourism in a given area is also a topic best discussed at the destination level to create a common understanding for all businesses with an interest in water management and longer-term investment plans.

€ssling / Tourism Management 46 (2015) 233e244 S. Go

243

7. Conclusions

Acknowledgement

With the continued growth in tourism and the trend towards higher-standard accommodation and more water-intensive activities, water pressure is bound to increase in many destinations, and in particular in regions already facing water security threats (e.g., Caribbean, China, southeast Asia, Mediterranean; € ro € smarty, Green, Salisbury, and Lammers, 2000). Regional Vo conflicts over water use have already been reported (ITP, 2013), and are projected to increase in the future due to increasing demand and declining supplies. Water pollution, population growth and climate change will create further pressure on freshwater resources (WWAP, 2012), to the extent that all water consumption has already become global in relevance (Hoekstra & Mekonnen, 2012). In order to efficiently manage water resources, it is imperative to understand direct (local) and indirect (global or embodied) water consumption, and to reduce risks associated with resource scarcity or declining water quality. Water consumption is closely linked to energy and food production, and best addressed in accommodation, the locus of most tourism consumption. This article confirms some general rules established in the literature, such as the fact that upscale accommodation has significantly higher water € ssling use values than more basic forms of accommodation (Go et al., 2012), and it underlines the importance of a distinction of direct and indirect water use. New insights have been provided with regard to end-use sectors of direct water consumption, as well as the role of fuel and food provisions in generating indirect water footprints. As these are currently not sufficiently addressed by industry (e.g. Kuoni, 2013), there is a need to implement more comprehensive water management indicators in an attempt to bridge the gap between current scientific opinion and industry practices. This paper focused on resort hotels in warm climates, where gardens and pools has are of specific relevance. Results suggest that large gardens, initial pool fillings, evaporation and backwashing all have significant influence on water consumption. This calls for the consideration of water ‘lock-in’ in built tourism infrastructure and landscaping, and the distinction of indicators for planned as opposed to already operational accommodation. With regard to food, results suggest that upscale all-inclusive tourism models encourage higher-order food consumption, and concomitant high levels of both energy and water use for their production. Energy consumption aspects, including fossil fuels, and also including the future use of biofuels, deserve greater attention as contributing factors to overall water use. To address this situation, 8 new indicators have been suggested to better capture the complexities of water consumption. These are relevant to different stakeholders, and their use may help to better understand water security risks, while lowering resource use costs. Evidence suggests that virtually all hotels have opportunities to reduce water consumption, probably in the order of at least 20% per € ssling guest night (Bohdanowicz-Godfrey and Zientara 2014; Go et al., 2012). As most measures also have short amortization times, water efficiency should be a focus of the global tourism industry. Future research needs to reconfirm results as presented in this paper for different tourist cultures and destinations, and to further explore qualitative aspects of water use. Currently, ‘water’ is not distinguished by its properties, i.e. whether it is abstracted in water-rich or -poor regions, whether its abstraction has social or environmental implications, whether it affects biodiversity, whether its use is consumptive, or whether it is treated after use. These as well as the quantification of renewable water capacities in destinations require continued research efforts.

I am indebted to Michalis Hadjikakou for a very thorough review of an earlier version of this paper. His advice is gratefully acknowledged, as is the support of the Thomas Cook group and in particular Swantje Lehners. In Rhodes, Nikos Portokallas facilitated detailed data collection. Without his help, it would have been impossible to carry out the project and to develop the new performance indicators presented in this paper. Nikitos Nikitaras facilitated contacts with hotel managers in Rhodes. Toby Meyer and Angelika Sotirakis conducted interviews. Ann-Christin Andersson helped preparing the data. Their help is gratefully acknowledged. References Accor. (2011). The Accor group's environmental footprint. First Multi-criteria Lifecycle Analysis for an International Hospitality Group. Available http://www.accor. com/fileadmin/user_upload/Contenus_Accor/Developpement_Durable/img/ earth_guest_research/2011_12_08_accor_empreinte_environnementale_dp_ bd_en.pdf Accessed 20.11.13. Alexander, S. (2002). Green hotels: Opportunities and resources for success. Portland, Oregon, USA: Zero Waste Alliance. Available http://www.zerowaste.org/ publications/GREEN_HO.PDF Accessed 20.11.13. Antakyali, D., Krampe, J., & Steinmetz, H. (2008). Practical application of wastewater reuse in tourist resorts. Water Science and Technology, 57, 2051e2057. Blancas, F. J., Lozano-Oyola, M., Gonz alez, M., Guerrero, F. M., & Caballero, R. (2011). How to use sustainability indicators for tourism planning: the case of rural tourism in Andalusia (Spain). Science of the Total Environment, 412-413, 28e45. Bohdanowicz, P., & Martinac, I. (2007). Determinants and benchmarking of resource consumption in hotels e case study of Hilton International and Scandic in Europe. Energy and Buildings, 39, 82e95. Bohdanowicz-Godfrey, P., & Zientara, P. (2014). Environmental management and online environmental performance assessment tools in the hotel industry: €ssling (Eds.), Handbook of theory and practice. In C. M. Hall, D. Scott, & S. Go Tourism and Sustainability. London: Routledge (in press). Briassoulis, H. (1991). Methodological issues: tourism input-output analysis. Annals of Tourism Research, 18(3), 485e495. Carbon Trust. (2013). Conversion factors. Available at http://www.carbontrust.com/ media/18223/ctl153_conversion_factors.pdf Accessed 28.10.13.  liz, J. (2014). The water footprint of Cazcarro, I., Hoekstra, A. Y., & S anchez Cho tourism in Spain. Tourism Management, 40, 90e101. Chapagain, A. K., & Hoekstra, A. Y. (2008). The global component of freshwater demand and supply: an assessment of virtual water flows between nations as a result of trade in agricultural and industrial products. Water International, 33(1), 19e32. Chenoweth, J., Hadjikakou, M., & Zoumides, C. (2014). Quantifying the human impact on water resources: a critical review of the water footprint concept. Hydrology and Earth System Sciences, 18, 1e18. Clarke, R., & King, J. (2004). The atlas of water. London: Earthscan. Cole, S. (2012). A political ecology of water equity and tourism: a case study from Bali. Annals of Tourism Research, 39(2), 1221e1241. Cole, S. (2013). Tourism and water: from stakeholders to rights holders, and what tourism businesses need to do. Journal of Sustainable Tourism, 22(1), 89e106. Cooley, H., Hutchins-Cabibi, T., Cohen, M., Gleick, P. H., & Heberger, M. (2007). Hidden oasis. Water conservation and efficiency in Las Vegas. Boulder, Colorado: Pacific Institute, Oakland, California, and Western Resource Advocates. Daisy Pool Covers. (2013). Fact sheet evaporation. Available http://www. daisypoolcovers.com.au/daisy-covers/evaporation/ Accessed 20.11.13. Deng, S.-M., & Burnett, J. (2002). Water use in hotels in Hong Kong. International Journal of Hospitality Management, 21, 57e66. Deutsche Vereinigung fuer Wasserwirtschaft, Abwasser und Abfall e.V. (DWA). €ranlagen 2011. Available at http://de. (2011). Leistungsvergleich kommunaler Kla dwa.de/tl_files/_media/content/PDFs/Abteilung_WAW/mj/Leistungsvergleich_ 2012_LOW.pdf Accessed 13.11.13.  Tortella, B., & Tirado, D. (2011). Hotel water consumption at a seasonal mass Deya tourist destination. The case of the island of Mallorca. Journal of Environmental Management, 92, 2568e2579. Eijgelaar, E., Thaper, C., & Peeters, P. (2010). Antarctic cruise tourism: the paradoxes of ambassadorship, last chance tourism' and greenhouse gas emissions. Journal of Sustainable Tourism, 18(3), 337e354. Energies Nouvelles. (2011). Water in fuel production. Oil Production and Refining. Available at http://www.ifpenergiesnouvelles.com/actualites/evenements/ nous-organisons/panorama-2011 Accessed 06.11.13. Ercin, A. E., Aldaya, M. M., & Hoekstra, A. Y. (2011). Corporate water footprint accounting and impact assessment: the case of the water footprint of a sugarcontaining carbonated beverage. Water Resources Management, 25(2), 721e741. Essex, S., Kent, M., & Newnham, R. (2004). Tourism development in Mallorca. Is water supply a constraint? Journal of Sustainable Tourism, 12(1), 4e28. European Commission (EC). (2013). European tourism indicator system TOOLKIT for sustainable destinations. DG Enterprise and Industry. February 2013. Available at

244

€ssling / Tourism Management 46 (2015) 233e244 S. Go

http://ec.europa.eu/enterprise/newsroom/cf/_getdocument.cfm?doc_id¼7826 Accessed 07.11.13. Europtravel. (2013). Various websites. Available at http://www.europetravel.gr Accessed 13.11.13. Eurostat. (2009). Medstat II: ‘Water and tourism’ pilot study. Eurostat, European Commission. Available at http://epp.eurostat.ec.europa.eu/cache/ITY_OFFPUB/ KS-78-09-699/EN/KS-78-09-699-EN.PDF. Falkenmark, M., & Rockstrom, J. (1993). Curbing rural exodus from tropical drylands. Ambio, 22(7), 427e437. Fang, K., Heijungs, R., & de Snoo, G. R. (2014). Theoretical exploration for the combination of the ecological, energy, carbon, and water footprints: overview of a footprint family. Ecological Indicators, 36, 508e518. Farrou, I., Kolokotroni, M., & Santamouris, M. (2012). A method for energy classification of hotels: a case-study of Greece. Energy and Buildings, 55, 553e562. Feng, K., Chapagain, A. K., Suh, S., Pfister, S., & Hubacek, K. (2011). Comparison of bottom-up and top-down approaches to calculating the water footprints of nations. Economic Systems Research, 23(4), 371e385. Filimonau, V., Dickinson, J., Robbins, D., & Reddy, M. V. (2013). The role of ‘indirect’ greenhouse gas emission in tourism: assessing the hidden carbon impacts from a holiday package tour. Transportation Research Part A, 78e91. Greenpeace. (2013). Red list fish. Available http://www.greenpeace.org/usa/en/ campaigns/oceans/seafood/red-fish/ Accessed 20.11.13. €ssling, S. (2001a). The consequences of tourism for sustainable water use on a Go tropical island: Zanzibar, Tanzania. Journal of Environmental Management, 61(2), 179e191. €ssling, S. (2001b). Tourism, environmental degradation and economic transition: Go interacting processes in a Tanzanian coastal community. Tourism Geographies, 3(4), 230e254. €ssling, S. (2002). Global environmental consequences of tourism. Global EnviGo ronmental Change, 12, 283e302. €ssling, S. (2010). Carbon management in tourism: Mitigating the impacts on climate Go change. London: Routledge. €ssling, S., Garrod, B., Aall, C., Hille, J., & Peeters, P. (2011). Food management in Go ̈ tourism. Reducing tourism's carbon ‘foodprint’. Tourism Management, 32(3), 534e543. €ssling, S., Peeters, P., Hall, C. M., Dubois, G., Ceron, J. P., Lehmann, L., et al. (2012). Go Tourism and water use: supply, demand, and security. An international review. Tourism Management, 33(1), 1e15. Gude, V. G., Nirmalakhandan, N., & Deng, S. (2010). Renewable and sustainable approaches for desalination. Renewable and Sustainable Energy Reviews, 14, 2641e2654. Hadjikakou, M. (2014). Measuring the impact of tourism on water resources: Alternative frameworks (PhD thesis). UK: University of Surrey. Hadjikakou, M., Chenoweth, J., & Miller, G. (2013). Estimating the direct and indirect water use of tourism in the eastern Mediterranean. Journal of Environmental Management, 114, 548e556. €nen, T. (2006). Lake tourism: An integrated approach to lacustrine Hall, C. M., & H€ arko tourism systems. Clevedon: Channel View Publications. Hawkins, R., & Bohdanowicz, P. (2011). Responsible hospitality: Theory and practice. Woodeaton, Oxford: Goodfellow Publishers Ltd,. Hellenic National Meteorological Service. (2013). Various websites. Hoekstra, A. Y., & Mekonnen, M. M. (2012). The water footprint of humanity. Proceedings of the National Academy of Sciences, 109(9), 3232e3237. Hof, A., & Schmitt, T. (2011). Urban and tourist land use patterns and water consumption: evidence from Mallorca, Balearic Islands. Land Use Policy, 28(4), 792e804. Hsiao, T.-Y., Chuang, C.-M., Kuo, N.-W., & Yu, S. M.-F. (2014). Establishing attributes of an environmental management system for green hotel evaluation. International Journal of Hospitality Management, 36, 197e208. Hunter, C., & Shaw, J. (2007). The ecological footprint as a key indicator of sustainable tourism. Tourism Management, 28, 46e57. ITP (International Tourism Partnership). (2013). Water risk assessment. Stockholm International Water Institute (SIWI). Kuoni. (2013). Kuoni water management manual for hotels. Available http://www. kuoni.com/docs/kuoni_wmp_manual_0.pdf Accessed 20.11.13. Lamei, A., von Münch, E., van der Zaag, P., & Imam, E. (2009b). Optimum contracted for potable water supply for hotels in arid coastal regions. Water Science and Technology, 59(8), 1541e1550. Lamei, A., Tilmant, A., van der Zaag, P., & Imam, E. (2009a). Dynamic programming of capacity expansion for reverse osmosis desalination plant: Sharm El Sheikh, Egypt. Water Science and Technology: Water Supply, 9, 233e246. Langumier, A., & Ricou, C. (1995). Le fonctionnement estival des services d'eau dans les communes du littoral normand. Analyse de l'impact des populations et activit es saisonni eres. Paris: Agence de l'eau Seine Normandie. Mekonnen, M. M., & Hoekstra, A. Y. (2010a). The green, blue and grey water footprint of crops and derived crop products. Value of Water Research Report Series No.47, UNESCO-IHE.

Mekonnen, M. M., & Hoekstra, A. Y. (2010b). The green, blue and grey water footprint of farm animals and animal products. Value of Water Research Report Series No.48, UNESCO-IHE. Miller, G. (2001). The development of indicators for sustainable tourism: results of a Delphi survey of tourism researchers. Tourism Management, 22, 351e362. O'Neill & Siegelbaum and The RICE Group. (2002). Hotel water conservation, A Seattle demonstration. Seattle: Seattle Public Utilities Resource Conservation Section. OECD. (2013). Effective policies for growth e In progress report. CFE/TOU. 10 (p. 24). Paris: OECD. September 2013. Page, S. J., Essex, S., & Causevic, S. (2014). Tourist attitudes towards water use in the developing world: a comparative analysis. Tourism Management Perspectives, 10, 57e67. Peeters, P. (2013). Developing a long-term global tourism transport model using a behavioural approach: implications for sustainable tourism policy making. Journal of Sustainable Tourism, 21(7), 1049e1069. Pigram, J. J. J. (1995). Resource constraints on tourism: water resources and sustainability. In R. W. Butler, & D. Pearce (Eds.), Change in tourism: People, places, processes (pp. 208e228). London: Routledge. Rico-Amoros, A. M., Olcina-Cantos, J., & Sauri, D. (2009). Tourist land use patterns and water demand: evidence from the western Mediterranean. Land Use Policy, 26, 493e501. -Batie, B., Mola , A., Cladera, A., & Martinez, V. (2010). Energy use, CO2 Rosello emissions and waste throughout the life cycle of a sample of hotels in the Balearic Islands. Energy and Buildings, 42, 547e558. Shah, M. M. (2013). New method for calculating evaporation from occupied swimming pools. HPAC engineering. Available at http://hpac.com/humidity-control/newmethod-calculating-evaporation-occupied-swimming-pools Accessed 12.11.13. Sun, Y. Y., & Pratt, S. (2014). The economic, carbon emission, and water impacts of Chinese visitors to Taiwan: eco-efficiency and impact evaluation. Journal of Travel Research. http://dx.doi.org/10.1177/0047287513517420. United Nations Educational, Scientific and Cultural Organization (UNESCO). (2009). United Nations World Water Development Report 3. Water in a changing world. Available at http://www.unesco.org/water/wwap/wwdr/wwdr3/pdf/WWDR3_ Water_in_a_Changing_World.pdf Accessed 04.09.13. United Nations Environment Programme (UNEP). (2011). Towards a green economy: Pathways to sustainable development and poverty eradication. Geneva & Nairobi: UNEP. UNWTO. (2004). Indicators of sustainable development for tourism destinations. A guidebook. Madrid: Spain. UNWTO. (2013). UNWTO tourism highlights, 2013 Edition. Madrid: UNWTO. United Nations World Tourism Organization (UNWTO), United Nations Environmental Programme (UNEP), Word Meteorological Organization (WMO). (2008). Climate change and tourism: Responding to global challenges. Madrid: United Nations World Tourism Organization; Paris: United Nations Environment Program; Geneva: World Meteorological Organization. U.S. Department of Energy, Argonne National Laboratory. (2012). Consumptive water use in the production of ethanol and petroleum gasoline e 2011 update. Available at http://greet.es.anl.gov/publication-consumptive-water Accessed 06.11.13. € ro € smarty, C. J., Green, P., Salisbury, J., & Lammers, R. B. (2000). Global water reVo sources: vulnerability from climate change and population growth. Science, 289, 284e288. Water Footprint Network. (2013). Glossary. Available http://www.waterfootprint. org/?page¼files/Glossary Accessed 20.11.13. Worldwatch Institute. (2004). Rising impacts of water use. Available at http://www. worldwatch.org/topics/consumption/sow/trendsfacts/2004/03/03/ Accessed 04.09.13. WWAP. (2012). The United Nations World Water Development Report 4: Managing water under uncertainty and risk. Paris, France: UNESCO. World Wide Fund for Nature (WWF). (2004). Freshwater and tourism in the Mediterranean. Available http://wwf.panda.org/?14217/Freshwater-and-Tourism-inthe-Mediterranean Accessed 20.11.13. € ssling is a professor at the School of Business Stefan Go and Economics at Linnaeus University, Kalmar, and the Department of Service Management, Lund University, both Sweden. He studied geography and biology, and holds a PhD in human ecology from Lund University. Stefan has worked with aspects of sustainable tourism since 1994, focussing primarily on transports, energy, and water.