1. INTRODUCTION
Like many other developing countries, Sri Lanka has an abundance of natural resources, including vast tracts of remnant natural areas; floristic zones ranging from tropical marine to cool montane; a diversity and richness of wildlife; and varied landscapes, seascapes and geological features (Gunatilleke et al. 2008;Marasinghe et al. 2020a, In Review;Perera et al. 2015;Senevirathna and Perera 2013). In line with global trends, Sri Lanka has leveraged the demand for ecotourism experiences that is being driven by the growing environmental awareness and increased desire to reconnect with nature (Parker and Simpson 2018a,, 2020;Senevirathna and Perera 2013;Simpson and Newsome 2017). Before the COVID-19 pandemic, the dynamic and competitive ecotourism market segment strengthened and grew the internal economy of developing regions and countries (Sumanapala et al. 2015a;Marasinghe et al. 2020a). Ecotourism can provide a source of foreign exchange earnings, generate tax revenues, and increase employment (Hapsari 2018;Perera et al. 2012;Soldić Frleta 2014). Warnings have, however, began to emerge about the negative impacts of natural area mass tourism in terms of threats to local cultures, high environmental and social costs, marginal economic benefits, and leakage of money away from local communities (Kilipiris 2005;Newsome 2013;Rasoolimanesh et al. 2017). Despite these concerns, an increasing demand generated by growing numbers of environmentally conscious travellers with diverse needs and expectations was and is again likely to generate a demand for authentic ecotourism experiences in the future, post the COVID-19 pandemic (Newsome 2020;Patroni et al. 2019;Perera et al. 2012;United Nations World Tourism Organisation (UNWTO) 2017), 2018). Ecolodges are one response to meet the accommodation demand of this tourism segment and the desire of entrepreneurial operators to provide a delineated product in the increasingly competitive ecotourism market (Chan and Baum 2007a;Russell et al. 1995;Sumanapala et al. 2017).
Concurrent with that growth in the ecotourism market segment over the past three decades, ecolodges emerged as a popular option among environmentally aware tourists seeking nature-orientated accommodation that complements the nature-focused experiences that motivated their travel (Chan and Baum 2007b;Sumanapala et al. 2015a). Widely cited,Russell et al. (1995, 147) defined an 'ecolodge' as a 'nature-dependent tourist lodge that meets the philosophy and principles of ecotourism'. To that end, The International Ecolodge Guidelines (Mehta et al. 2002) specify that the three main characteristics of ecolodge accommodation should be conservation of neighbouring lands, benefits to local communities, and interpretation to both local populations and guests. However,Lai and Shafer (2005) andNewsome (2013) report that ecolodge operators often overlook the educational component.
There is a growing need for ecotourism operators to create demand by marketing tourism products that are more environmentally sustainable and socially responsible (Handriana and Ambara 2016;Patroni et al. 2019;Sotiriadis 2017;Yousaf et al. 2018). Understanding how to influence visitor satisfaction further allows ecotourism operators to develop and position their product(s) to boost return visits and word of mouth recommendation (El-Said and Aziz 2019;Handriana and Ambara 2016;Simpson et al. 2019;Smolčić Jurdana and Soldić Frleta 2011). There is a wealth of tourism literature that reports on tourists motives to travel, visitor satisfaction, and revisit intentions (e.g.Dutta et al. 2017;Lee 2009;Patroni et al. 2018b;Perera et al. 2012;Pérez Campdesuñer et al. 2017;Yousaf et al. 2018). Until recently, however, the ecolodge literature has predominantly focused on definitions, the physical environment, best practice management, and sustainability evaluations (Handriana and Ambara 2016;Bulatović 2017). Research is needed to provide a broader understanding of the behaviours of ecotourists and the factors that influence their destination/accommodation choices. Such research will help operators, managers, and governments to better cater to this specialized market segment to optimise visitor experience and revenue generation, as well as educating clients about the environment (Handriana and Ambara 2016;Mafi et al. 2019;Newsome 2013;Patroni et al. 2019;Sumanapala et al. 2017).
Despite the wealth of international literature, guidelines, and certification systems related to ecolodge management, the publication of empirical research about ecolodges remains limited (Mafi et al. 2019). Further, the research ofBandara (2009) andFernando and Kaluarachchi (2016) reports the importance of showing Sri Lankan ecolodge operators the relevance of that information in the local context. This study addresses those gaps in the literature by comparing ecolodge attributes and the motives, satisfaction, and revisit intentions of ecolodge visitors in Sri Lanka to similar research conducted in other countries through the application of structural equation modelling. As such, this research can enhance the ecological sustainability of ecolodge management in Sri Lanka and for similar accommodation at other forest and marine destinations in the region thatMarasinghe et al. (2020a) describe as Tropical Asia.
2. SPECIFICATION OF CONCEPTUAL MODEL
As previously mentioned, there now exists a substantial body of research reporting the relationships between tourist motives to travel, accommodation attributes, visitor satisfaction and tourist intention to revisit/recommend the experience. These attributes (indicators) of ecolodges and tourist motives to travel reported in the ecolodge literature guided the development of the questionnaire utilised in this study (seeTable 1).
Not surprisingly, high satisfaction increases the likelihood of repeat visitation and word-of-mouth (WOM) recommendation (Bigné et al. 2001;Chen and Chen 2010;Lee et al. 2011;Perera and Vlosky 2013;Yoon and Uysal 2005). Previous research also provides support for the hypothesis that visitor satisfaction is a mediating factor between tourist motives and ecolodge attributes and the revisit and recommendation intentions of tourists (Bigné et al. 2001;Chen and Chen 2010;El-Said and Aziz 2019;Handriana and Ambara 2016;Lee et al. 2011;Padlee et al. 2019;Yoon and Uysal 2005).
In addition to tourist motives and ecolodge attributes indirectly influencing revisit intentions via visitor satisfaction, the studies ofKozak and Rimmington (2002),Lai and Vinh (2013), andSom et al. (2012) suggest that tourist motives also directly influence revisit intention. Similarly, many studies provide evidence that accommodation/destination attributes (ecolodge attributes in this study) directly influence the intentions of tourists (e.g.El-Said and Aziz 2019;McDowall 2010;Padlee et al. 2019;Patroni et al. 2018a;Petrick 2004).
The conceptual model shown inFigure1 provides a visual summation of the studies highlighted above. Indicators (observed factors) for the latent factors of the conceptual model appear inTable 1
Source: Developed by authors to reflect relationships reported in the literature regarding ecolodges (as referenced in Section 3).
Moore et al. (2015) have, however, identified the need for additional measures and more in-depth research into the relationships between the latent factors revisit intention, tourist attitudes/motives, and satisfaction with the facilities that support nature tourism experiences. Consistent with that recommendation, this is the first study to explore and report on how the attributes Sri Lankan ecolodges influence travel motives, visitor satisfaction, and revisit intention. As previously reported, this study also explores the relevance of the existing ecolodge literature for the industry in Sri Lanka. Establishing these relationships is important because Sri Lankan ecolodge operators want to know that recommendations from the literature have relevance and will work in the local context (Bandara 2009;Fernando and Kaluarachchi 2016).
Based on the aims of this study and the literature presented above, the following six hypotheses define the conceptual model (Figure 1) of ecolodge attributes (EA), tourist motives (TM), visitor satisfaction (VS), and revisit intention (RI) for Sri Lankan ecolodges:
H1: EA positively influence TM to travel.
H2: EA positively influence VS.
H3: TM positively influence VS.
H4: VS positively influences RI.
H5: TM positively influence RI.
H6: EA positively influence RI.
3. Data Collection and Pre-treatment
A self-administered pen and paper semi-structured questionnaire captured the responses of ecolodge visitors for the indicators (observed factors) of the four latent factors described in the conceptual model (Figure1). The observed factors were used to model the relationship of the latent factors as reported in the ecolodge literature. Feedback from experienced local researchers and ecolodge operators adapted those factors to the Sri Lankan context.
Many authors report that high levels of customer/visitor satisfaction are essential to ensure the success and sustainability of a tourism operation (e.g. Dutta et al. 2017;Patroni et al. 2019;Soldić Frleta 2017). For tourism experiences, visitor satisfaction is a measure of the cognitive difference between expectations, measured as tourist motives in this study, and the actual service delivery, measured as ecolodge attributes (Handriana and Ambara 2016;Parker and Simpson 2018b;Pinkus et al. 2016;Simpson et al. 2019;Soldić Frleta 2018). Destination image ‘pull factors’ such as natural landscapes, opportunities to view wildlife, local culture and lifestyle, and ecolodge attributes motivate tourists to visit or stay at particular locations (Chan and Baum 2007a(Chan and Baum 2007b;Hung et al. 2012;Madden et al. 2016). Motives to travel can influence the level of visitor satisfaction that tourists express regarding their ecotourism accommodation and experiences (Bigné et al. 2001;Dutta et al. 2017,Handriana and Ambara 2016;Lee 2009;Mlozi et al. 2013). Numerous studies also report that ecolodge attributes related to the facilities, location, and service level have a direct effect on visitor satisfaction (Bigné et al. 2001;Chan and Baum 2007a;Kozak and Rimmington 2002;Mandić et al. 2018).
Questions for factor-related questions used closed statements that ecolodge visitors rated using 7-point Likert scales. For the EA and VS factors, the Likert scales ranged from 1 = Very Dissatisfied to 7 = Very Satisfied. A Likert Scale of 1 = Not at all Important to 7 = Extremely Important was used to rank TM. Tourists ranked their RI using a scale of 1 = Definitely Not to 7 = Definitely.
The literature referenced above guided the development of the questionnaire regarding the indicators of EA, TM, VS, and RI. A panel of researchers and operators familiar with ecotourism surveys and the ecolodge industry in Sri Lanka provided feedback and the draft questionnaire was adapted to suit local conditions, which provided face validation of the survey instrument. The study experienced time constraints arising from the seasonal nature of the monsoon-influenced Sri Lankan tourism industry and difficulty engaging short-stay ecolodge visitors in the survey (discussed later). Therefore, colleagues and employees of the participating ecolodges provided the trial group for the pilot questionnaire. The small sample size for the pilot (10-15 people), meant that it was not possible to quantitatively check the construct validity and the reliability of the survey. The sample size of the trial was not a concern, because the preliminary analysis of the structural equation modelling (SEM) process (e.g. Cronbach-alpha/Internal Consistency check, homogeneity check, and exploratory factor analysis) confirms those characteristics for the full data set and therefore provides post-survey validation of the questionnaire (Bolarinwa 2015;Golob 2003;Sarantakos 2013;Schreiber et al. 2006;Weston and Gore 2006).
The target population for the survey was individuals aged eighteen years or older who were staying at least one night during the period between January 2014 and January 2015 in participating ecolodges in the Sri Lankan districts of Dambulla, Hambantota, Kandy, Matale Ratnapura, and Puualam. The four criteria specified in the earlier study ofKwan et al. (2010) guided the selection of sixteen ecolodges that had a focus on conservation, were designed and operated to have a minimal negative impact on the environment, provided educational programs for visitors, and contribution to the local community.
Front counter staff at the participating ecolodges opportunistically distributed questionnaires to survey visitors using convenience sampling technique (Hapsari 2018;Sarantakos 2013). Staff distributed the questionnaires one-per-room to visitors who were travelling together, travelling with their family, or were travelling as part of a group. Visitors returned their completed questionnaire to a drop-box at the front desk when checking out.
Visitors returned 385 questionnaires of the total of 450 questionnaires distributed. This raw response rate of 85.6% is significantly above the 70% level considered to be an excellent response rate for such surveys (Frankfort-Nachmias and Nachmias 1992;Denscombe 2014). An assessment of the validity of the returned surveys deemed 18 questionnaires unanalysable with a further five questionnaires removed based on being consistent outliers (Weston and Gore 2006). The remaining 362 questionnaires provided a very acceptable 80.4% response rate. Of those included questionnaires, six were missing data elements, corrected by substituting the individual case-mean for that factor (Byrne 2010;Schreiber et al. 2006). The sample size of 362 exceeds the acceptable minimum sample size of 200 for SEM analyses, the absence of feedback loops in the full structural model (Figure2) and the data checks and analyses reported below further validate the sample size of this study (Golob 2003;Weston and Gore 2006). The demographic profiles of the survey participants are published in two peer-reviewed articles bySumanapala et al. (2015a)Sumanapala et al. (2017), which are both available as full-text open-access/online articles.
4. FACTOR ANALYSIS
As recommended for SEM research, the factor analysis in this study utilised several methods of analyses (Golob 2003;Hair et al. 2005;Schreiber et al. 2006;Weston and Gore 2006). Data from the survey was checked and analysed with a variety of techniques using Version 20.0 of the Statistical Package for Social Scientists (IBM Corp. 2011). The multifaceted approach to factor analysis that is integral to the SEM technique overcomes the need for quantitative validation of the questionnaire and justifies applying findings reported in the ecolodge literature to inform the management of ecolodges in Sri Lanka.
Checks on the normality of the observed factors aligned to each latent factor (Field 2000;Gravetter and Wallnau 2014;Trochim and Donnelly 2006) showed strong approximations to the normal distribution with homogeneity of variances, and acceptable levels of skew (-1.433 to-0.342) and kurtosis (-0.702 to 1.594).
Cronbach’s alpha (Cα) is a measure of internal consistency applied to the validation of survey questionnaires as a measure of scale reliability that assesses how closely a set of items in a group are related (Dlačić et al. 2019;UCLA Statistical Consulting 2020). Cronbach’s alpha values validate both the internal consistency/reliability of the latent factors (acceptable values greater than 0.7) and the Cronbach-alpha if item deleted analyses (for which acceptable values are greater than 0.6 and less than the internal consistency Cronbach-alpha of the relevant latent factor) of the each observed factors (Lin and Huang 2018;Mohamad et al. 2015;Nunnally 1979).
Source: Developed by authors to explore relationships reported in the literature regarding ecolodges.
Checking of the multicollinearity between independent and dependant latent factors confirmed that tolerances were less than 0.1 and that variable inflation factors (VIF) were less than 10 (Mandić et al. 2018;Soldić Frleta and Smolčić Jurdana 2018a;Ziegler and Hagemann 2015).
Checking of the unidimensionality of the latent factors was based on Kaiser-Meyer-Olkin (KMO) measures (acceptable values are greater than 0.5 and preferably close to 1.0) and significant outcomes (p ≤ 0.05) for Bartlett’s test of sphericity (Malhotra and Morris 2009;Soldić Frleta and Smolčić Jurdana 2018a;Ziegler and Hagemann 2015;Subaskaran and Balasuriya 2016).
Cronbach-alpha scores of greater than 0.7 (Table 2) for the latent factors EA, TM, VS, and RI (Figure2) demonstrate that the observed factors (indicators) reliably describe each of the four latent factors (Nunnally 1979;Vo and Chovancová 2019). Corrected item-total correlation (ITC) scores greater than 0.3 demonstrate a similar level of variance in all questions related to each latent factor for all survey participants (Pedhazur and Schmelkin 1991;Yoon et al. 2001). That suggests all observed factors should remain in the model. Except for RI1 and RI3, deleting any other observed factors from the analysis would have reduced the Cronbach-alpha scores for the internal consistency of the latent factors.
The Cronbach-alpha if deleted scores (Cα) provide evidence for removing RI1 and RI3 from the model. However,Hair et al. (2005) cautioned that the early removal of factors due to statistical issues is not advisable as the primary purpose of factor analysis is to explore the factor structure (Child 1990). Further,Golob (2003, 7) cites the ‘”three measure rule” [that] asserts a measurement model will be identified if every latent variable [factor] is associated with at least three observed variables’. For those reasons, testing of the measurement models included RI1 and RI3.
Multicollinearity checks between independent and dependant latent factors (Table 3) provided acceptable values of Tolerance (0.509 to 0.989) and VIF (1.0111 to 1.963). Unidimensionality checks for the latent factors (Table 3) were also acceptable with KMO values of 0.802 to 0.994 and significant responses (p<0.001) for Bartlett’s tests of sphericity.
5. STRUCTURAL EQUATION MODELLING
Having confirmed the validity and reliability of the observed factors as indicators of the latent factors/variables of the full model (Figure2), Version 18 of the AMOS (Analysis of Moment Structures) software package(Arbuckle 2007) was used to explore the hypotheses for the relationships between the latent variables.
Structural equation modelling is a confirmatory analysis technique (comparing theoretical models with empirical data) used to explore relationships between observed factors/variables/indicators and latent (unobserved) factors/variables/constructs that cannot be or are difficult to measure directly (Pérez Campdesuñer et al. 2017;Golob, 2003;Schreiber et al. 2006). The multifaceted SEM technique is ideally suited to exploring tourist attitudes, motives, satisfaction, and intentions as a complex system of independent and dependant factors that interact by direct and indirect influence (Dutta et al. 2017;Hung et al. 2012;Weston and Gore 2006).
5.1. Measurement Model Validation
For analyses based on SEM, testing of the measurement models are the equivalent of performing confirmatory factor analysis (Golob 2003;Schreiber et al. 2006;Weston and Gore 2006), and that was the approach adopted for this study. Moreover, 'the components of a non-recursive model can be broken into blocks, and if each block satisfies identification conditions, then the entire model is also identified'Golob (2003, 7). That was the approach used to validate the fit of each block of the measurement model for this study.
In line with the recommendations of Schreiber et al. (2006, 327) relating to the ‘one time analysis’ approach adopted by this study, the Comparative Fit Index (CFI), the root mean square error of approximation (RMSEA) and the Tucker-Lewis Index (TLI) were used to assess the models goodness-of-fit. However, there is variability regarding the values of indices considered to indicate a good fit of the model to the observed data (Table 4). Those values are impacted further by the observed factors being categorical or continuous, and by the sample size for the model (Weston and Gore 2006). For those reasons, the values for the goodness-of-fit indices and the acceptance values adopted by this study appear inTable 5
The model fit indices and compliance with the acceptance criteria for each of the latent factor blocks in the measurement model and the full model appear inTable 5. These results highlight the value of the guidance fromGolob (2003),Schreiber et al. (2006), andWeston and Gore (2006) that understanding and interpreting the significance of relationships suggested by an SEM requires the careful evaluation of multiple fit indices for the model. Based on these results, the observed factors in the four blocks of the measurement model provide an acceptable fit for each of the latent factors. This analysis confirms the factors included in the conceptual model and supports the retention of observed factors RI1 and RI3.
5.2. Testing the Full Structural Model
The aggregation of the analyses presented in the Factor Analysis section above and the model fit indices for the full structural model reported inTable 5 provide strong evidence that the proposed full model (Figure2) is suitable for testing relationships between the latent factors for ecolodges in Sri Lanka.
Testing the full model confirmed the first five hypotheses (Table 6). However, there was no evidence (p >0.05) to support the hypothesis that EA directly influence the RI of visitors to ecolodges in Sri Lanka. Instead, the attributes of Sri Lankan ecolodges indirectly influenced the RI of visitors through the strong effect that EA had on both TM and VS (Figure 3).
6. DISCUSSION
The primary focus of this study was to develop and test a conceptual model to determine if the indicators of tourist motivation, satisfaction, and revisit intention reported by the international ecolodge literature apply for ecolodges in Sri Lanka. The structural modelling reported in this study demonstrates that, in the main, the attitudes, behaviours, and motives of tourists visiting ecolodges reported in the international literature also apply for ecolodges in Sri Lanka of the style described byBandara (2009),Fernando and Kaluarachchi (2016), andKwan et al. (2010). As a result, most of the findings and management recommendations reported in the existing literature can inform the planning and management of ecolodges in Sri Lanka.
Source: Developed by authors to reflect relationships determined by this study. ** β significant at α = 0.01
The critical difference between the relationships reported by studies from other destinations and the modelling of this study (Figure 3) and (Table 6)) is that this study found no evidence that EA directly influenced the RI of tourists (p-value for H6 greater than 0.05). This finding is at odds with the research outcomes reported byMcDowall (2010),Patroni et al. (2018a) andPetrick (2004).
While not directly influencing the RI of tourists, the amenities, activities, and service provided by Sri Lankan ecolodges (i.e. EAs listed in (Table 1)) provide strong motivation for tourists to visit Sri Lanka and how satisfied visitors are with their ecolodge experience (Figure 3). For every unit increase (or decrease) in visitor perception of EA almost 84% transfers to TM (β = 0.842) and approximately 70% of that change is transmitted to VS (β = 0.709). This finding that the attributes of Sri Lankan ecolodges strongly influences the travel motives and satisfaction of visitors is consistent with ecolodge research from other destinations. (e.g.Hagberg 2011;Hays and Ozretic-Došen 2015;Mic and Eagles 2018;Osland and Mackoy 2004).
Motives for tourists to travel had a strong direct influence on the intentions of tourists to revisit ecolodges in Sri Lanka (Figure 3) with 80% of any change in motivation (TM) transferring to RI (β = 0.802). Consistent with other studies (Kozak and Rimmington 2002;Lai and Vinh 2013;Som et al. 2012), this finding demonstrates the importance of maintaining Sri Lanka’s destination image. That includes providing quality ecotourism experiences by correcting and avoiding the ongoing problems first identified byBuultjens et al. (2005) and then byNewsome (2013) and most recently byPrakash et al. (2019). Recurring themes in the articles of those authors are overcrowding at nature tourism destinations, the operation of motor vehicles and crowding of wildlife that occurs on so-called safari tours, and environmental degradation of natural landscapes and protected areas targeted for nature-based tourism.
While tourist motives to travel to Sri Lankan ecolodges significantly influenced visitor satisfaction (Figure 3), surprisingly only about 20% of any change in tourist motives manifests as a change in visitor satisfaction levels (β = 0.181). That may be evidence that visitor satisfaction with their experiences of ecolodges in Sri Lanka is so high that changes in pre-travel motives have little effect or are moderated (Antón et al. 2017). Alternatively, it may be that there is a disconnect between visitor expectations (measured as TM) and their level of satisfaction (Antón et al. 2017;Cohen et al. 2017). The review article of Cohen et al. (2017, 887) reports that “Several researchers have moved away from examining perceptions about the product and focus instead on the relationship between tourists and places as a determinant of satisfaction … [however] … considerably more consumer research is needed on these influences on satisfaction.” The Sri Lankan ecolodge industry and government agencies could benefit from additional research that further explores the relationships between visitor expectations and satisfaction with their nature-based tourism experience(s). The techniques of Importance-Performance Analysis could provide the basis for such research (e.g. Marasinghe et al. In Review;McGuiness et al. 2017;Simpson et al. 2019;Soldić Frleta et al. 2018Soldić Frleta et al. 2018aSoldić Frleta et al. 2018b;Taplin 2012).
Also consistent with the findings of several other studies (Bigné et al. 2001; El-Said and Aziz 2019;Lee et al. 2011;Padlee et al. 2019;Perera and Vlosky 2013;Yoon and Uysal 2005), visitor satisfaction with their Sri Lankan ecolodge experience had a strong direct influence on tourist intentions to revisit ecolodges and Sri Lanka more broadly (Figure 3) with 87% of any change in satisfaction transferring to revisit intention (β = 0.870). As noted in the previous paragraphs, visitor satisfaction with ecolodges in Sri Lanka is primarily driven by the amenities, activities, and service levels that tourists experience at an ecolodge.
7. CONCLUSION
7.1. Key Findings of this Study
This study demonstrates how accommodation attributes and tourist motives to travel influence visitor satisfaction and tourist intentions to revisit ecolodges in Sri Lanka. Understanding these relationships is important for Sri Lanka as local ecolodge operators and managers require evidence that recommendations from the global literature have relevance for their lodge and will work in the local context. Ecolodge attributes strongly influence the motivation of tourists to travel to and within Sri Lanka and their level of satisfaction with that experience. Tourist motives weakly influence visitor satisfaction but have a strong direct influence on visitor intentions to revisit Sri Lanka or stay in an ecolodge. Visitor satisfaction is a moderator for the influence of both ecolodge attributes and tourist motives on revisit intentions. The alignment of these findings with research reported in the literature regarding the attitudes, behaviours, and satisfaction of ecolodge visitors in other countries demonstrates that the factors that attract visitors to ecolodges in those alternate destinations and ensure that they are satisfied with their visit apply equally for ecolodges in Sri Lanka.
7.2. Suggestions for Ecolodge Operators
The findings of this study show that operators and managers of Sri Lankan ecolodges can benefit by applying learnings from international research to maximise the satisfaction of their visitors and benefit from the personal and electronic word of mouth recommendations and repeat business that satisfied visitors provide. For those reasons, operators must maintain ecolodge standards (Table 1) and work to protect the cultural and natural resources that are crucial elements of quality ecolodge experiences to maintain the reputation of their lodge and the image of Sri Lanka as an attractive ecotourism destination.
Further, ecolodge operators and mangers in Sri Lanka should promote their ecolodge(s) by highlighting the uniqueness, history, and natural assets of the local area, and maintain an appropriate level of price as motivation for tourists to visit (Table 1 and (Table 3). The attributes of Sri Lankan ecolodges can be enhanced by having the lodge easily accessible, providing authentic cultural and ecotourism experiences in the local area, incorporating cultural and conservation education/interpretation activities, having a library of relevant local information, and ensuring visitor safety (Table 1 and (Table 3).
7.3. Limitations of Study and Additional Research
Data collection for this study relied on the support of ecolodge operators, managers, and staff to distribute questionnaires, and on visitors agreeing to participate. The managers of some ecolodges, including several the high-end best-practice lodges, declined to have the survey run at their establishment. Many visitors at the participating ecolodges were reluctant to complete a questionnaire. Visitors reported that was in part due to the short time that most stay in an ecolodge (1 to 3 days – Sumanapala et al. 2015a, 2015b, 2017) and due to being fully engaged in the cultural and nature-based activities associated with their stay at the lodge.
Increasing the number of lodges participating in the research could improve future ecolodge studies in Sri Lanka. That would provide a broader perspective regarding the offerings and operation of Sri Lankan ecolodges. The participating ecolodges could benefit from replicating this study to determine if there have been any changes in visitor responses in the five years since the data reported in this article was collected. Such a study could be even more beneficial given the COVID-19 pandemic. Further, a replicate study following the easing of international travel restrictions after the COVID-19 pandemic could establish a longitudinal program of ecolodge assessment in Sri Lanka.
ACKNOWLEDGEMENTS
The content of this article is based on research reported in the PhD thesis of Dr. Daminda P. Sumanapala. We thank the operators and management of the ecolodges who voluntarily distributed and collected the questionnaires that provided the data that underpins this study. We thank the anonymous ecolodge visitors who completed and returned questionnaires. We acknowledge the financial assistance provided by the National Research Council of Sri Lanka. The authors acknowledge the mentorship and guidance of Professor Sarath W. Kotagama in the development of the study that produced this article We also thank the THM Editors and the anonymous reviewers whose insightful comments and suggestions enhanced our article.